American Canyon City Council
March 21, 2023

Type of Services Geologic Fault Investigation
Project Name Giovannoni Property
Location Devlin Road
American Canyon, California
SECTION 1: INTRODUCTION
This geologic fault investigation report was prepared for the sole use of Orchard Partners, LLC (Orchard) for the lands between the west end of Devlin Road and north of Green Island Road in the northern end of American Canyon in Napa County, California. The location of the site is shown on the Vicinity Map, Figure 1. For our use, we were provided with the following documents pertaining specifically to the project site:
▪ Sheet 1. Confirmed Aquatic Resources Delineation Map, Giovannoni Project Site, provided by Monk and Associates, dated October 25, 2016. ▪ Un-named topographic map of the Giovannoni property, not dated.

1.1 PROJECT DESCRIPTION
A new gravity sanitary sewer main will be constructed starting at the Napa Logistics Park north of Devlin Road and will cross the property south of Devlin Road (Giovannoni property) and terminate at a pump station to be located just north of Green Island Road. The proposed sewer alignment will run in a north-south direction across the Giovannoni property within a mapped Alquist-Priolo Earthquake Fault Zone likely intersecting the West Napa Fault.

1.2 SCOPE OF SERVICES
Our scope of services was presented in our proposal dated March 22, 2017, and consisted of two tasks: 1) file review and field reconnaissance, and 2) a fault investigation to locate the potential surface trace of the West Napa Fault where it crosses the planned sewer line.

Following requests by the City of American Canyon and the Property Owner’s representatives to combine the easement for the new sanitary sewer main within the setback for the West Napa Fault the scope of services was expanded as defined in our Confirmation of Requested Services (CRS) #1 dated June 30, 2017. The additional tasks consisted of: 1) a second 2075

GIOVANNONI PROPERTY
582-4-1

trenching investigation at the northern end of the property; and 2) a ground water investigation at the northern and southern trenching investigation.

Following the lack of evidence of ground rupture within the southern trench the scope of was expanded as defined in our Confirmation of Requested Services (CRS) #2 dated August 11, 2017. The additional task consisted of extending the southern trench 150 to 200 feet to locate the potential trace of the West Napa Fault on the southern side of the site.

Brief descriptions of our exploration programs are presented below.

1.3 GROUND WATER EXPLORATION PROGRAM
Previous investigations to the north of the site by Myers (1983) and Westling and Hanson (2008), and our firm (Cornerstone Earth Group, 2016), identified a correlation between disruption of the ground water table and the West Napa Fault Zone (WNFZ). Noting this phenomenon, we endeavored to bracket the general location of the West Napa Fault zone by advancing 12 borings (MW-1 through MW-12) on July 26, 2017, with direct push drilling equipment. The borings were spaced approximately 50 feet apart and drilled to depths of 15 to 20 feet.

The borings were backfilled with cement grout in accordance with local requirements; exploration permits were obtained as required by local jurisdictions. The approximate locations of our exploratory borings are shown on the Site Plan, Figure 2. Detailed locations of the borings are presented in the trench logs, Figures 8A through 8C and 9A through 9B.

1.4 FAULT INVESTIGATION EXPLORATION PROGRAM
As previous workers conducting fault investigations along the WNFZ observed that the fault forms a barrier to groundwater movement resulting on a higher phreatic surface on the east (Myers, 1983; Wesling and Hanson, 2008), we prefaced our field investigation by conducting an array of borings across the northern portion of the site, and across the southern portion of the site (see section 3.3 for a description of that investigation).

The fault investigation consisted of three trenches excavated, cleaned, logged, and inspected under the supervision of Cornerstone’s Certified Engineering Geologist. Trenches were excavated to approximate depths of 7½ to 9½ feet with a backhoe equipped tractor. Exposed walls were supported with hydraulic shoring and cleaned with hand tools for logging of subsurface conditions. Trench 1 extended to 215 feet to depths of 7½ to 9½ feet and was located on the northern side of the property, south of the unnamed creek traversing the mapped trace of the West Napa Fault. Trenches 2A and 2B extended to a combined length of 394 feet to approximate depths of 7½ to 9½ feet and was located on the southern half of the site, traversing the mapped traces of the West Napa Fault. All trench segments were viewed in the field by the city-designated peer reviewing certified engineering geologist (Mr. James Joyce, CEG, of Joyce Associates).
2076

GIOVANNONI PROPERTY
582-4-1

1.5 ENVIRONMENTAL SERVICES
Environmental services were not requested for this project. If environmental concerns are determined to be present during future evaluations, the project environmental consultant should review our geotechnical recommendations for compatibility with the environmental concerns.

SECTION 2: REGIONAL SETTING

2.1 REGIONAL GEOLOGIC SETTING
The site is located in the southern portion of Napa Valley that is characterized as a relatively large north-west trending alluvial valley located within the Northern California Coast Range geomorphic province. The valley is at the southernmost end of the Mayacmas Mountains.

South of the City of Napa, the hills on the western side of the valley terminate at the marshes bordering the northern end of San Pablo Bay; whereas the hills on the northeast continue to near Sulphur Springs Mountain near the City of Vallejo. The bedrock ridges on each side of the Napa Valley trend northwest, parallel to the general north-northwest structural trend of the North Coast Ranges. Pre-Quaternary bedrock is generally restricted to the foothills, but locally there are low knolls or hills of Tertiary age bedrock in the central and western parts of the valley.

Quaternary alluvial fan deposits shed from the hills on the east, and fluvial deposits associated with the Napa River and its tributary valleys comprise the youngest deposits with in the Napa Valley (Sowers et al., 1995; Bezore et al., 2002, 2005; Clahan et al., 2004, 2005; Wagner et al., 2006). Late Pleistocene estuarine deposits formed during the last interglacial stage are postulated to underlie a broad geomorphic surface in the southern end of the valley where the subject site is situated.

Within the region, the San Andreas Fault system, which distributes shearing across a complex assemblage of primarily right lateral, strike-slip, parallel and sub-parallel faults that includes the Hayward and Calaveras Faults and others (see the “Faulting” section of this report). The mountainous topography west of Napa Valley resulted from latest Pliocene and Quaternary uplift associated with the younger structures. A published regional-scale geologic map of the Cuttings Wharf Quadrangle forms the base of the Vicinity Geologic Map, Figure 3.

2.2 LOCAL GEOLOGY
Several published geologic maps cover the area, including those of Sims et al., (1973), Fox et al., (1973), Sowers (1995), Knudsen et al. (2000), Blake et al., (1974). Bezore, et al., (2002), Graymer et al., (2007), Wagner and Guiterrez (2010). The map of Bezore represents the best mapping and at the best scale (1:24,000) and therefore was used as our Local Geologic Map, Figure 3. The site is in an area adjacent to the San Francisco Bay where Quaternary alluvial deposits dominates the geology of Cuttings Wharf 7.5’ Quadrangle. The Bezore, et al., (2002) map depicts the majority of the site and adjacent areas as underlain by Late Pleistocene to Holocene alluvial fan deposits (“Qf”), which are bound to the north and on the west by older geologic units including Late Pleistocene fan deposits (“Qpf”) and Early to middle Pleistocene alluvial fan or terrace deposits (“Qoa”). Specifically, the Qf unit is relatively younger than and was deposited over the Qpf and the Qoa units. The Qf unit is described as; “Gently sloping, 2077

GIOVANNONI PROPERTY
582-4-1

fan-shaped, relatively undissected alluvial surfaces where late Pleistocene or Holocene age was uncertain or where the deposits of different age interfinger such that they could not be delineated at the map scale. Sediments include sand, gravel, silt, and clay, that are moderately to poorly sorted, and moderately to poorly bedded. The Qpf unit is described as; “Late Pleistocene fan deposits. Gently sloping, fan-shaped alluvial surfaces where late Pleistocene age is indicated by slight dissection and/or the development of alfisols.” The Qoa unit is described as; Moderately to deeply dissected alluvial deposits capped by alfisols, ultisols, or soils containing a silica or calcic hardpan.

The Huichica Formation (Pliocene) is mapped on the western edge of the site. The formation is described as “Fluvial gravel, sand, silt, and clay”. Holocene alluvium is mapped within the meandering, channelized path of the unnamed creek that extends into the site (Figure 3). An abandoned channel of this creek extends into the site and was encountered within our Trench 1 (see discussion of faulting in the “Faulting” section of this report). The un-named creek continues south decreasing in width and depth until its upper reaches transitions into ephemeral wetlands south of Trench 2A. Detailed mapping of the geologic units found on-site is presented on the Fault Investigation Plan, Figure 7.

2.3 REGIONAL SEISMICITY
While seismologists cannot predict earthquake events, geologists from the U.S. Geological Survey have recently updated earlier estimates from their 2014 Uniform California Earthquake Rupture Forecast (Version 3) publication. The estimated probability of one or more magnitude

6.7 earthquakes (the size of the destructive 1994 Northridge earthquake) expected to occur somewhere in the San Francisco Bay Area has been revised (increased) to 72 percent for the period 2014 to 2043 (Aagaard et al., 2016). The faults in the region with the highest estimated probability of generating damaging earthquakes between 2014 and 2043 are the Hayward (33%), Rodgers Creek (33%), Calaveras (26%), and San Andreas Faults (22%). In this 30- year period, the probability of an earthquake of magnitude 6.7 or larger occurring is 22 percent along the San Andreas Fault and 33 percent for the Hayward or Rodgers Creek Faults.

During such an earthquake the danger of fault surface rupture at the site is slight, but very strong to severe ground shaking would occur.

Important among these younger structures in the map area are the Quaternary-active, including Holocene-active, faults of the San Andreas Fault system, such as the Maacama, Healdsburg, Rodgers Creek, and West Napa Faults, shown as magenta (Holocene-active; Hart and Bryant, 1999) and orange (Quaternary-active) on the map. A regional fault map is presented as Figure 4, illustrating the relative distances of the site to significant fault zones. The faults considered capable of generating significant earthquakes are generally associated with the welldefined areas of crustal movement, which trend northwesterly. The table below presents the State-considered active faults within 25 kilometers of the site. 2078

GIOVANNONI PROPERTY
582-4-1

Table 1: Approximate Fault Distances
Fault Name
Distance
(miles)
Distance
(kilometers)
West Napa Fault <0.1 0.1
Green Valley Fault 7.4 12.0
Rogers Creek Fault 8.1 13.0
Hayward (Rogers) 11.7 18.8
Great Valley (Gordon Valley) 13.1 21.1
Hayward (Total) 12.8 20.6
Great Valley (Pittsburg Kirby Hills) 18.0 29.1
*Distances are from estimated surface projection of each fault.

A regional fault map is presented as Figure 4, illustrating the relative distances of the site to other significant fault zones.

SECTION 3: SITE CONDITIONS

3.1 RECENT HISTORY
The site is located in an area dominated by topographically level, low-lying alluvial areas adjacent the intertidal marshland that borders San Pablo Bay.

Historic aerial stereo pair photographs at the United States Geological Survey (USGS) were reviewed covering a period from 1948 through 1973. Additionally, Google Earth® (2015) aerial images were reviewed that show the site and vicinity from 1991 through 2017. The photos from 1948 through 1991 show the subject parcel and adjacent areas were primarily large agricultural parcels used for livestock grazing. An unnamed, seasonal creek located in the northwest portion of the site trends in a northwest-west direction. Three east-west trending troughs along the northern property line and approximately 190 and 350 feet south of the northern property line, and a north-south trending trough approximately 1400 feet east of the western property line were presumably created as tributary drains to an unnamed creek. By 2004, livestock trails in various directions and a fence trending in an east-west direction across the central portion of the site are present. By 2010 the embankments for the railroad track overpass and the western end of Devlin Road are present along the northwest side of the site. By 2012 the overpass for the railroad to the northeast of the site is present. The aerial images clearly show faint linear tonal lineaments that define the West Napa Fault zone trending through the site as generally mapped by Bezore et al., (2002) and other geologists. SAR imagery collected following the August 24, 2017, South Napa Earthquake (SNE) displays two northwest trending linear features trending through the area south of the Napa Airport. The eastern lineation aligns between the previously 2079

GIOVANNONI PROPERTY
582-4-1

mapped fault traces of the West Napa Fault within the Site (CGS, 2017). Amongst the various published maps, there some minor differences in the interpretation of these surface features.

3.2 SURFACE DESCRIPTION/GEOMORPHOLOGY
Cornerstone personnel performed a reconnaissance of the site in June 2017. Due to the undeveloped nature of the site, its relatively flat topography and ground cover vegetation, there are few exposures of natural earth materials at the subject site except for local erosion scars on sloping ground within the central trough and stream banks of the unnamed creek near the northern edge of the site. At the time of the reconnaissance the site was undeveloped and being used for livestock pasture. Grasses and weeds exist across the ground surface. The central portion of the site is crossed in a north-south sense by a seasonal creek which has resulted in subtle depressions accented by wetland vegetation. A topographic rise on the southwest corner of the site corresponds to area where the published maps show the Huichita Formation forming an oblong outcrop at the ground surface.

The site was accessed via the property to the north. From the west end of Devlin Road a dirt road and path was traversed to the northwest corner of the site where a gate is located. Our site reconnaissance revealed geomorphic features suggesting the presence of fault surface traces including: subtle linear topographic depressions along the northerly trending creek adjacent to the north property line, a dextrally offset drainage, and linear edge along the west edge of wetlands (see the “Faulting” section of this report).

3.3 GROUND WATER INVESTIGATION
Ground water was encountered in our twelve borings (MW-1 through MW-12) on July 26, 2017, at depths ranging from 8.7 to 19.4 feet below current grades. The depths to ground water encountered in our explorations are summarized in the table below. All measurements were taken at the time of drilling and may not represent the stabilized levels that can be higher or lower than the initial levels encountered.

Borings MW-1 through MW-7 were drilled approximately 5 feet north of Trench 1 after the trench had been logged. The array of borings was spaced 50 feet apart and centered (MW-4) where tonal aerial photographic and geomorphic evidence suggested the fault trends through the property. First evidence of ground water was measured as the borings were advanced. Ground water appeared to stabilize within 15 to 20 minutes of being encountered. The phreatic surface differed by 8.3 feet (higher on the east) over an array length of 100 feet between borings MW-3 and MW-5.

Borings MW-8 through MW-12 were drilled 50 feet apart in an array on the south side of the property where tonal and subtle geomorphic evidence had suggested the fault trended through that area. Furthermore, the wetland to the south of the array has a northwest trending western edge that was suspected to be controlled by faulting. The shallow ground water conditions that result in this wetland were thought by our investigators to be the result of a ground water barrier along the eastern border of the WNFZ. The 200-foot long array of borings were centered on a north-south trending seasonal creek channel at the northwest end of a wetland. 2080

GIOVANNONI PROPERTY
582-4-1

Ground water was not encountered in borings MW-8 and MW-9. Ground water was encountered west of and at MW-10. Based on the known relationship between observed faultrelated features in Trench 1 and the change in the phreatic surface across MW-1 through MW7, Cornerstone concluded a southern trench centered on MW-10 would be a likely location to observe faulting on the southern side of the site.

An approximately 200-foot long trench, centered on MW-10 was excavated, cleaned and logged. No evidence of fault surface traces was observed. The trench was subsequently extended further to the west where faulting was observed (see the “Faulting” section of this report for further discussion).

Fluctuations in ground water levels occur due to many factors including seasonal fluctuation, underground drainage patterns, regional fluctuations, and other factors. Trench 2B encountered the fault in older (Pliocene) well consolidated sediments with no ground water observed in Trench 2B. Cornerstone did not bore to measure ground water across the zone of faulting in Trench 2B. The subsequent trenching revealed that the older and more consolidated Huichica Formation existed in the area of MW-8 and MW-10 whereas the younger and significantly less consolidated alluvial fan deposits (Qf) were present in the area of MW-10. Hence, whereas the groundwater conditions along the northerly array were controlled by faulting (with the fault extending through the alluvial fan deposits), ground water conditions along the southern array were controlled or impacted by stratigraphy and formational conditions (where the fault extends through the Huichica Formation).

Table 2: Depth to Ground Water
Boring Number
Corresponding
Trench
Depth to
Ground Water
(feet)
Ground Water
Elevation1
(feet)
Depth of
Boring
MW-1 1 19.4 4.6 20.0
MW-2 1 18.9 5.1 20.0
MW-3 1 19.1 4.9 20.0
MW-4 1 15.2 8.3 20.0
MW-5 1 11.4 11.6 20.0
MW-6 1 16.3 6.7 20.0
MW-7 1 8.7 14.3 15.0
MW-8 2A >20.0 <11.0 20.0
MW-9 2A >20.0 <10.8 20.0
MW-10 2A 16.6 13.9 20.0
MW-11 2A 11.2 19.8 15.0
MW -12 2A 14.3 17.5 20.0
1 Mean Sea Level, referenced from topographic map provided by client. 2081

GIOVANNONI PROPERTY
582-4-1

SECTION 4: GEOLOGIC HAZARDS

4.1 FAULTING
As previously noted, there are a number of Quaternary faults in the northern Bay Area region including the San Andreas Fault system, the Maacama, Healdsburg, Rodgers Creek, Cordelia, and West Napa Faults (Hart and Bryant, 1999).

More locally, the West Napa Fault (WNF) is shown on several published maps as trending through the general region of the site with a northwesterly trend of approximately N45°W. The West Napa Fault zone historically has been considered to be a 30- to 35-km-long fault that trends parallel to the western margin of the Napa Valley and extends to the southeast along the eastern margin of marshlands that border northeastern San Pablo Bay (Figure 4). Fault traces in the vicinity of the West Napa fault were first mapped by Weaver (1949), and subsequent mapping indicated that the fault zone consists of northwest-trending fault traces (Fox et al., 1973; Sims et al., 1973; Helley and Herd, 1977; Pampeyan, 1979; Wagner and Bortugno, 1982; Fox, 1983; Jennings, 1994; Bezore et al., 2002, 2005; Clahan et al., 2004, 2005; Wagner et al., 2004). The fault is the northernmost element of a series of relatively short north-to-northwesttrending en echelon faults that also includes the Franklin and Southampton faults. These faults traverse the East Bay Hills and lie between the Hayward – Rodgers Creek Fault zone on the west and the Calaveras – Concord – Green Valley Fault zone on the east. Based on the trenching results within the Napa region and existing regional mapping, Geomatrix (1998) hypothesized that the apparent recent activity on these and possibly other faults in the East Bay Hills accommodate the transfer of slip from the northern Calaveras Fault to the West Napa Fault. Unruh et al. (2002) and Kelson et al. (2005) also proposed slip transfer from the northern Calaveras Fault to the West Napa Fault through a series of left-stepping, en echelon dextral faults and lineaments that they collectively term the Contra Costa Shear Zone (CCSZ).

Evidence regarding recency of activity of the WNF was originally evaluated by the state as part of Fault Evaluation Report 129 (Bryant, 1982). The surface traces of the WNF within the Cuttings Wharf 7.5‘ Quadrangle were inferred from tonal lineaments in aerial photographs, and subtle scarps in Pleistocene-age alluvium (Bryant, 1982). More recently, and in response to the 2014 South Napa Earthquake, an updated Fault Evaluation Report (FER 256) was conducted based on evaluating ground rupture observed after earthquake and remote sensing techniques of lidar and synthetic aperture radar (SAR) (CGS, 2017; Baize et al., 2015; Catchings et al., 2016). The fault traverses several different types of terrain that influence its expression in the landscape including agricultural land on low-relief alluvial fans and terraces and low-relief salt marsh/floodplain. The area surrounding the subject site is an example of the latter. The landscape in the area of the site has remained relatively unmodified, and therefore has preserved the geomorphic features found along its WNF surface trace. Wesling and Hanson, in 2008 conducted a study of the West Napa Fault Zone and characterized the geomorphology around the area of the subject site with the following discussion; “The scarps along this reach align with right-deflected ephemeral drainages that flow northeast to southwest across the fault. A relatively prominent ephemeral drainage directly south of the Napa County Airport appears to be deflected about 245 meters in a 2082

GIOVANNONI PROPERTY
582-4-1

right lateral sense across the main trace of the fault and nearly 70 meters across a secondary fault trace. Similar small deflected drainages occur locally along this reach of the fault. Although the trace of the fault is poorly expressed in the active floodplain and salt marshes along the river, it is important to note that the Napa River makes a broad right swing of nearly 2,000 meters across the projection of the fault.” Bryant noted the significance of the following features in the area of the Napa County Airport; ''Geomorphic features such as linear vegetation contrasts, scarps in Holocene alluvium, deflected drainage, and possible closed depressions, strongly indicate Holocene faulting. The Napa Airport, built in 1941, is constructed on natural ground with minimal fill at the southwest end of the runways (W. Partain, airport manager, p.c., December 1981 in Myers, 1983). It is unlikely that the sharp tonal lineaments through the airport are artificial.”
Subsequent to the evaluation of Bryant, an Alquist-Priolo Earthquake Regulatory Zone was established for the southern portion of the WNF (CDMG, 1983). The WNF and associated Earthquake Regulatory Zone regulatory boundaries (“AP Zone”) trends through the central portion of the site as shown on the Alquist-Priolo Earthquake Regulatory Zone Map, Figure 5.

Traditionally the WNF has been considered a late Pleistocene-Holocene active fault (Helley and Herd, 1977; Pampeyan, 1979; Bryant, 1982; Wagner and Bortugno, 1982; Fox, 1983; Jennings, 1994). More recent work by Langenheim and others (2006) shows that the 2000 (M5.2) Yountville earthquake probably occurred on the West Napa Fault. Significantly however is the fact that the West Napa Fault experienced surface ruptured during the August 24, 2014, Magnitude 6.0 South Napa Earthquake (“SNE”). Studies of the event show that SNE with its epicenter located approximately 2.5 kilometers (km) west of the Napa County Airport resulted in as much as 20 km of surface rupture consisting of both coseismic and post seismic slip along on the near-vertical, strike-slip West Napa Fault. Up to 50 centimeters (cm) of dextral displacement was observed on the principal rupture, with displacements of less than 10 cm on several parallel surface traces east of the primary rupture, in a zone up to 2 km wide (CGS, 2017). The largest coseismic slip occurred in the northern half of the rupture area (Lienkaemper et al., 2014), with less slip in the southern half nearer to the subject site. The post-seismic slip shows a complementary pattern, with large post-seismic slip in the south nearly equaling the coseismic slip in the north (Hardebeck and Shelly, 2016; Hudnut, 2014; Hudnut et al., 2014).

The boundary between predominately post-seismic slip and predominately coseismic slip corresponds to a change in the character of the surface rupture, from a single north-northweststriking rupture in the south to a multistranded more-northerly striking rupture to the north (Brocher et al., 2015).

During and immediately following the South Napa Earthquake, minor right-lateral offset, likely less than a few centimeters, was observed trending through two taxiways at the Napa County Airport north of the site (EERI Special Earthquake Report, 2014; Ponti et al., 2017; CGS, 2017).

This surface rupture is located on the mapped trace of the Airport Section of the West Napa fault zone as designated by Bryant (2000). Surface displacements were detected initially through high resolution remote sensing techniques (X-band InSAR and UAVSAR imagery) but the surface rupture at the airport was verified by geologists in the field (EERI Special Earthquake Report, 2014). The southern extent of the surface rupture appears to have died out 2083

GIOVANNONI PROPERTY
582-4-1

in the southern part of the airport near where the fault takes a step onto a westerly trace (as mapped by Bezore, 2002) which continues through the subject site. The fact that the 2014 rupture didn’t jump the gap over to the westerly trace suggests the more eastern trace served as a “releasing step”. The concept of a “releasing step” in en-echelon fault surface patterns within a strike-slip fault zone is explained by Wesnousky (2006).

4.1.1 Proposed Sewer Line
An existing forced sewer line oriented north-south crosses the site approximately 750 feet east of the western property line. A new gravity sewer line is planned to replace the existing line.

Appropriate building methods will need to be employed where the line crosses the zone of faulting. The current land owner has also requested that the new sewer line fall within the fault setback area to reduce the number of easements affecting the property. Locating the fault in the north and south end of the property will allow for both goals to be met.

4.2 FAULT SURFACE RUPTURE
Surface fault rupture involves shearing, differential movement, and ground breakage along the trace of the fault during moderate to strong earthquakes. The resulting movement can severely damage structures and utilities that are located across the fault trace. Thus, studies are undertaken to identify the location of fault traces, to determine the activity of the fault and to provide building setbacks where Quaternary active faults are identified. Evaluation of surface fault rupture is based on the premise that future fault rupture will take place along previous ruptures. Consequently, accurate determination of the location and character of previous fault ruptures is required for surface fault hazard assessment. In terms of fault rupture hazard evaluations, faults are considered “active” if they display evidence of movement within Holocene time (the last 11,000 years), and “potentially active” if they display evidence of movement within Quaternary time (i.e., within the last 1.6 million years). As previously discussed, the western portion of the site is located within a State-designated Alquist-Priolo Earthquake Regulatory Zone. As the proposed sewer easement will cross the WNFZ at a location yet to be determined, we have provided estimates of surface displacement for the fault and recommendations to help minimize the potential impact of fault surface rupture for the utility (see the “Mitigating Fault Surface Ruptures” section of this report).

4.2.1 Previous Consultant’s Studies
As a result of the state-designated zonation of the fault, numerous subsurface investigations have been completed to locate and/or “clear”1 sites of fault surface traces as part of geologic and geotechnical studies for commercial and residential developments (e.g., ENGEO, 1977; EMRI, 1979; DMA, 1983; DHA, 1984; GEI, 1996; Giblin Associates, 2005; Raney Geotechnical, 2003a, b, 2007). Previous Local Investigations, Figure 6, indicates the locations of some of the more proximal consultant’s investigation sites. Several previous fault studies performed in the vicinity of the site (within 2¼ miles) were compiled and reviewed in an effort to better constrain the locations of faults on or near the site.

1 i.e., prove lack of evidence for fault surface traces trending through the investigation area. 2084

GIOVANNONI PROPERTY
582-4-1

Darwin Myers Associates in 1983 performed an investigation at the Napa County Airport property. The Myers study included three magnetometer survey lines and logging of two exploratory trenches located in the southern portion of the adjacent airport property. These explorations were located to intersect the mapped projection of the easterly surface trace, which is the sole trace mapped on the airport property. The geophysical survey generally produced results that were inconclusive in terms of identifying or suggesting diagnostic features suggesting of faulting. Myers stated: “No prominent slickensided shear planes were observed in the alluvium. However, the truncation of sandy clay unit at Station 0+71 in T-1 and 0+20 in T-2, shear planes in the topsoil, the break in slope, and the change in moisture content are regarded as compelling evidence of faulting. Because the fault cuts Holocene deposits, the trench data indicate that the fault is active. Moreover, by utilizing the geomorphic features, it is possible to present [sic] the main trace of the fault from the exploratory trenches as far north as the north boundary of the airport property and to the south as far as Green Island Road. Minor subparallel fault traces and branching faults may occur within 200 to 300 feet of the main trace” (Darwin Myers Associates, 1983).

In 2008 the USGS performed mapping along the West Napa Fault (WNF) which included trench logging near the southern edge of the Napa Airport property (Wesling and Hanson, 2008). Their trench (Trench GC-1) was located on the same mapped surface trace as that investigated by Myers in 1983 and about 300 feet south of the Myers Trench T-1. Wesling and Hanson indicated;
“The trench was located at a very sharp linear tonal contrast that is coincident with a 1- to 1.5-meter-high west-facing scarp that traverses a broad terrace (corresponding to Elevation 20 feet) on the south side of the Napa County Airport. The lineament trends N16°W. The lineament and scarp, are now obscured by up to 1 meter of fill that was placed by the County to smooth the surface across the scarp. Groundwater was encountered at a depth of 1.5 to 2 meters. The stratigraphic and structural relations exposed in Trench GC-1 clearly indicate that late Pleistocene and Holocene alluvial deposits are displaced by the West Napa Fault zone. The cumulative late Pleistocene vertical displacement (during approximately the past 120 to 125 thousand years [kyr] or 80 kyr) is more than 1.2 meters down on the west. Both the tectonic setting of the fault and geomorphic evidence suggest the fault is predominantly a rightlateral strike-slip fault; the total net slip could not be determined based on the trench exposure. The middle- to late-Holocene vertical displacement (i.e., the base of Unit 5) is less than 0.5 meter, which indicates there have been multiple surface-faulting events during the late Quaternary. The number and timing of individual surface faulting events has been obscured by shrinking and swelling of the clayey soil in the upper part of the trench.”
Both the Myers (1983) and the Westling and Hanson (2008) investigations confirmed that the fault trace at the airport site forms a groundwater barrier.

Several studies intended to locate active traces of the fault, or to “clear” building envelopes of 2085

GIOVANNONI PROPERTY
582-4-1

fault traces have been completed to the south of the Darwin Myers Associates (1983) study for commercial real-estate development between Green Island Road and extending on to the northern part of Oat Hill. Some noteworthy studies that located active fault traces include: EMRI & Associates (1979, 1983), Bailey Scientific (1989, 1990), Kleinfelder & Associates (KA) (1983, 1984a, 1984b, 1990a, 1990b, 1990c), Balbi & Chang Associates (1984), Herzog & Associates (1988a, 1988b, 1989a, 1989b), Bay Soils, Inc. (1979), Earthtec (1989), and Darwin Myers Associates (1985). The locations of some of the more proximal studies are shown on Previous Local Investigations, Figure 6. Some of these studies are discussed briefly below.

EMRI & Associates (1979, 1983) and Bailey Scientific (1989) performed a subsurface fault investigation approximately 0.75 miles to the south of the building site. The study area included the entire AP-Zone, and targeted the single mapped trace of the WNF. The studies incorporated trenching, subsurface borings, and magnetometer surveys. No evidence of active faulting was reported.

Bailey Scientific (1990) investigated an area east of and adjoining the study areas of EMRI & Associates (1979, 1983) and Bailey Scientific (1989). The area of this study includes the eastern margin of the AP-Zone, and is located about 0.33 mile south of the building site. No evidence of active faulting was reported.

Kleinfelder Associates (KA) (1984a, 1984b) performed a subsurface fault investigation approximately 1 mile south of the building site, roughly centered on Oat Hill. Their study area included the mapped trace of the WNF. They reported that no evidence of active faulting was found on the eastern side of Oat Hill; evidence for active faulting was found only on the western side of Oat Hill, southwest of the subject site, and west of the mapped trace.

Balbi & Chang Associates (1984) conducted a geotechnical Investigation for an access road partially shadowing the building site west of Oat Hill. The study included a continuous roadcut roughly perpendicular to the mapped fault trace, and borings. No evidence for active faulting was reported.

Kleinfelder Associates (1983) (KA) performed a subsurface fault investigation approximately 400 feet south of the building site near Oat Hill. KA inferred active faulting on the basis of disruption of soils at depths of 3 to 4 feet, but observed no disruption of surface soils. The azimuth of the fault trace is not certain, but appears to project west of the building site, along a topographic break at the foot of Oat Hill.

Additionally, more recent studies have identified active fault traces on the flanks of Oat Hill (i.e., Giblin Associates, 2005; Raney Geotechnical, 2003a, 2007). The recent studies demonstrate the presence of active traces along the strongly expressed lineament on the northeast flank of the hill as mapped in Bryant (1982).

A previous investigation by Cornerstone (2016) found no evidence of faulting in three trenches to the east and west of the unnamed creek to the north of the current investigation. The traces of the West Napa Fault are presumed to lie closer to the axis of the stream channel where it trends north-northwest.
2086

GIOVANNONI PROPERTY
582-4-1

In summary, some of the previous consultant’s investigations performed within the area confirm that strands of the West Napa Fault trend through the region and studies north of the project site on the adjacent airport indicates an active fault trace or traces trend towards the western portion of the subject site. A north-northwesterly trending strand is projected crossing the subject site.

This trace continues north into the adjacent property where it steps east to a trace that was encountered in the trenches of Darwin Myers Associates (1983) and by the USGS (2008).

These previous consultant investigations found conclusive evidence of Holocene-active faulting with multiple fault surface rupture episodes occurring in the Holocene.

4.2.2 Current Fault Investigation
The initial purpose of our investigation was to locate and bracket the West Napa Fault where it will cross the planned gravity sewer line. Following approval of the initial proposal the City of American Canyon (City) and the Property Owner’s representatives preferred to combine the easement for the new sanitary sewer main within the setback for the West Napa Fault. A second trench was added to the scope of the investigation to locate or bracket the West Napa Fault in the northern and southern portions of the site. Following the combined 400 feet of trenching investigation the West Napa Fault had been identified only on the north side of the site; Trench

1. With the agreement of the 3rd party reviewer for the City, Jim Joyce, C.E.G., and approval by the client, Trench 2 was extended approximately 200 feet to the west. Trench 2 was split into two sections to avoid exposing or damaging the existing force sewer line which was reported to lie at a depth of about 4 feet below the existing grades. Trench 2B and Trench 2A are offset 70 feet perpendicular to each other. As the trend of the sewer line is oriented approximately 80 degrees compared to the trenches, the east end of Trench 2B projects to the west end of Trench 2A along strike of the fault. We assume there is little to no loss of visibility between the trenches despite the presence of the sewer line.

Before and during the ensuing trenching investigation the City’s reviewing geologist (Mr. James Joyce, C.E.G.) was provided a fault trenching plan. The area of Trench 1 exists to the south upslope of an outside bend of the unnamed creek. The average elevations along the mix of nearly level stream terrace and fan deposits is approximately 23 to 24 feet. Trenches 2A and 2B trend from a topographic low lying seasonal wetland to the mid-slope of a gentle hill (approximate elevations ranging from 31 to 38 feet).

Three trenches were excavated, cleaned, logged and inspected under the supervision of Cornerstone’s Certified Engineering Geologist. Trenches were excavated to approximate depths of 7½ to 9½ feet with a backhoe equipped tractor. The depths of the trenches were dictated by the depth to the local groundwater table. Exposed walls were supported with hydraulic shoring, and cleaned with hand tools for logging of sub-surface conditions. We focused our logging on the southerly trench wall but followed certain features across to the opposing trench wall where they were exposed further by cleaning. The northerly wall of Trench 2B was cleaned completely along with about 50 percent of the southerly wall. Trench 2B logging (Figure 9C) is flipped for a uniform presentation of trench log views. The trenches were located within the “AP Zone” to identify the location of the West Napa Fault, which is mapped as a single trace in a north-northwest direction across the site (plotted on the Fault Investigation 2087

GIOVANNONI PROPERTY
582-4-1

Map, Figure 7). All segments of the trenches were reviewed in the field by a city-designated peer reviewing geologist (C.E.G.), James Joyce. Mr. Joyce clarified during his visits that he concurred with our field interpretations.

4.2.2.1 Trench 1
Trench 1 was located approximately 210 feet east of the western property line and approximately 85 feet south of a bend in the unnamed creek on the north side of the site. The 205-foot-long trench was excavated in a westerly direction (S88W) to depths of 7½ to 9½ feet.

Trench 1 was placed to intersect the trace of the West Napa Fault that was mapped in a northnorthwesterly trend parallel to the creek channel in the adjacent property to the north.

The trench is located in an area where Bezore’s mapping indicates the surficial deposits transition from (Holocene) Qha in the eastern portion of the trench, into (middle Pleistocene) Qoa in the western one-third of the trench. Generally, A- and B-horizon soils overlie alluvial sedimentary units. The faulting is expressed in a series of vertical cracks with a gray sandy texture in comparison to the finer and darker surrounding alluvial units. Evidence of faulting was observed in Trench 1 from 130 to 140 feet as measured from the eastern end of the trench (Figure 8B). Two prominent sets of cracks merge at depth and thicken to 1-inch wide. The gray sandy features cross the trench at attitudes of N10°W and N6°W, steeply dipping to the west.

Dark brown clay was observed on the edges of the gray sandy features. Little (less than 1 inch) to no vertical offset was observed on either side of the vertical features. The upper soil horizons are not offset and the vertical features do not extend to the surface. The upper 12 to 20 inches of the vertical features are thin (less than a 5 millimeters) and in some places, have a 1- to 2- millimeter gap where white and translucent sand grains are clinging to the vertical faces of the cracks. These gaps may be remnant from the recent earthquake as clay has not infilled the void created by recent faulting. A detail of the faulted area is presented in Figure 8C.

4.2.2.2 Trench 2A
Trench 2A was excavated in a westerly direction (S83W) extending for 294½ feet from a topographically low portion of land rising approximately three feet in elevation to the west.

Trench 2A was stopped on the western end as it would encounter the current sanitary sewer force main line. Trenching continued (Trench 2B) on the western side of the sewer line approximately 80 feet to the north. Trench 2B was excavated in a westerly direction (S88W) for approximately 85 feet. Trench 2A was excavated to depths of 7½ to 9½ feet. Trench 2a extended through an area where Bezore’s map shows four geologic units (which include from east to west; (Pleistocene) Qf, (Holocene) Qha, (Pleistocene) Qoa, and then (Pliocene) Th].

The majority of Trench 2A exposed residual soils overlying semi-consolidated alluvial sediments. Residual soils 1A and 1B on the eastern two thirds of the trench correspond to the mapped Holocene to Pleistocene unit Qha. The western soils (1Ca and 1Cb) and underlying alluvial sediments correspond to the mapped Pleistocene unit Qoa. A moderately consolidated unit (Huichica Formation, Th) was observed at depth along the western 54 feet of Trench 2A.

No evidence of faulting was observed in Trench 2A (Figures 9A through 9C). 2088

GIOVANNONI PROPERTY
582-4-1

4.2.2.3 Trench 2B
Trench 2B started west of the sewer line and was excavated in a westerly direction (S88W) for approximately 85 feet. Trench 2B was excavated to depths of 7½ to 9½ feet and consisted of residual soil (Qoa) overlying the Huichica Formation (Th) (Figure 9D). Faulting was observed in Trench 2B from 20 to 35 feet as measured from the eastern end of Trench 2B. The primary fault trace was expressed as a vertical wedge of yellowish fault gouge (4c) with carbonate precipitate marking the main faulting features. Within this fault trace, horizontal striations were observed on a slickenslided surface (N39W, 53SE, rake: 90) within a 1-inch-thick fault gouge zone indicating strike-slip style of deformation. The main faulting feature (fault gouge surrounded by slickenslides and carbonate) was measured across the trench with an attitude of N26W, 80NE. An arcuate subsidiary trace (as in a “flower structure”) with down-dip striations (N61W, 48NE, rake: 3NE) was observed just east of the primary feature. Bedding within the Huichica Formation is somewhat disrupted proximal to the zone of faulting as observed between Station 36 to Station 42 feet. A shear zone was observed well west of the fault zone between Station 80 to Station 87. No evidence of shearing extended into the residual (A and B horizon) soils above these older shear features. Within this zone a series of faulted surfaces juxtaposing the coarse and fine members of the Huichica Formation were completely replaced by 1 to 3 inches of carbonate accumulation. The Huichica Formation clay, sand, and gravel were substantially calcified and large nodules were associated with the carbonate. The extent of and degree of carbonate formation would have required a significant amount of time to develop and the cementation along the shear surfaces indicates a substantial haitus since the last fault movement along the fault surfaces. Based on the amount of carbonate precipitate and hardness of the sedimentary unit across the shear zone we interpret this area as an older strand of shearing that may predate the Quaternary period (i,e, confined to the Pliocene).

The approximate locations of the trenches are shown on the Fault Investigation Map, Figure 7.

Detailed trench logs are presented as Figures 8A through 8C, and 9A through 9D.

4.2.3 Encountered Geologic Units in Exploratory Trenches Surficial soil exposed in all trenches are predominantly dark brown to very dark grayish brown/very dark gray clay with some fine to medium sand of medium to high plasticity (1A, 1B, and 1C). The upper 6 to 8 inches of soil structure was commonly disturbed from the activities of cattle grazing. The underlying geologic units (2 and 3) are interpreted as alluvial in origin but not consolidated. The oldest unit (4 – Huichica Formation) is moderately lithified and generally redder in color.

Unit 1 – Holocene to Pleistocene brown to dark grayish brown, clayey well-developed A- and B-horizon soils. A-horizon soils are developed in place by the accumulation of organic material and degradation of minerals into to clay. Rain and surface water strip material from the Ahorizon where it is transported down. B-horizon soil is developed below A-horizon soil, in place by the accumulation of organic constituents, clay particles, and minerals (iron, hematite, calcium, etc.) derived from the A-horizon.
2089

GIOVANNONI PROPERTY
582-4-1

Qha 1A Clay [CL/CH] stiff to very stiff, moist, dark brown to brown [10YR 3-2/1] with fine sand and occasional fine to coarse gravels. Block and prismatic A- and B-horizon soils.

Qha 1B Sandy clay [CL] stiff, moist, brown to gray brown [2.5YR 4-3/2] with fine to medium sand and occasional fine sub-rounded gravels. Weak to moderate blocky Bhorizon soils.

Qoa 1C Sandy clay [CL/CH] medium stiff to very stiff, moist, with fine sands. 1Ca1 Dark gray [10YR 4/1] clay with fine sands. A-horizon soils with moderately granular to weak blocky structure.
1Ca2 Dark gray [5YR 4/1] clay with fine sands and occasional well-rounded gravels at base of horizon. A-horizon soils with massive to moderately granular structure.
1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels. B-horizon soils with strong blocky to prismatic structure. 1Cb2 Brown to gray brown [10YR 4/3 – 5/2] clay with fine sands. B-horizon soils with moderate blocky structure.

Unit 2 – Pleistocene light brown to gray clayey sand and sandy clay alluvial sediments.

Qoa 2 Clayey sand and sandy clay with gravel [SC] dense and stiff, moist to wet, light brown to gray [2.5Y 6/2-4/3] fine to medium sands and clays. 2a Sandy clay [SC] stiff, moist to wet, light brown [2.5Y 6/3] with fine to medium sands.
2b Clayed sands [CL] dense, moist to wet, light brown [2.5Y 5/2] with occasional sub-angular to sub-rounded gravels. Orange mottles [10YR 5-4/4] common above the saturated zone.

Unit 3 – Sandy channel deposits within Unit 2.

Qoa 3 Clayey sand with gravels [SC] dense, moist, light brown to brown [10YR 4/2-3/4] with sub-angular to sub-rounded shale and chert.

Unit 4 – Th - Huichica Formation (Pliocene). Moderately lithified, degraded where faulted.

Th 4a Sandy clay [CL] stiff to very stiff, moist, clay with fine sands, yellowish brown [10YR 5/6], brown [7.5YR 4-5/4-6] and reddish brown [2.5YR 6/3]. Bedding present between sandier and clayey packets of sediment.

Th 4b Clayey sands with gravels [SC] very dense, moist, yellowish red to reddish brown [5YR 4-5/4-6].
2090

GIOVANNONI PROPERTY
582-4-1

Th 4c Clay with silt [CL] medium stiff, moist, olive yellow [2.5Y 6/6], some fine to coarse angular to sub-angular gravel and fine carbonate nodules. Carbonate accumulations formed on the vertical margins where the slide planes of the fault were observed.

Disseminated carbonate (I and II) (forward diagonal marks on trench logs) is common in the middle of unit 2 and 5 and at the base of B-horizon soils. Hematite nodules (thin reverse diagonal marks on trench logs) are found within Trench 2A within B-Horizon soils and the upper portion of Unit 2. Horizontal contacts between soils horizons and precipitates are features that are formed over significant amounts of time and therefore can serve as “marker horizons”.

Nodules of carbonate (Stage I and II) were observed along vertical ped faces on the lower half of B-Horizon soils and sporadically throughout Unit 2. When broken apart these nodules contained a core of a fine chert gravel or coarse sand that act as a locus for precipitation.

Based on similar studies by Borchadt (2016) the soils developed beyond the modern stream channel and sedimentary units are interpreted to be up 21 thousand years (Holocene to Pleistocene).

4.2.4 Field Investigation Summary
The West Napa Fault was identified on the northern (Trench 1) and southern (Trench 2B) portions of the property. An approximate 12-foot-wide zone of faulting was observed between Station 130 and Station 142 within Trench 1. An approximately 15-foot-wide zone faulting was observed from Station 20 to Station 35 feet of Trench 2B. The difference in the fault zone width between the northerly exposure versus the southerly exposure may be due to the fact that at Trench 1 the fault is cutting a younger and less consolidated geologic unit whereas at the southerly location the fault is exposed deeper in the stratigraphic section and in a more consolidated material. In fact, the exposure of the fault within Trench 1 provides a glimpse of how the deformational style of the fault zone varies as it propagates up into a younger and softer material. Although the WNFZ is clearly a Holocene active fault, no fractures, truncations or variations in horizon thickness were observed in the A or B soil horizons overlying the identified fault zones. This does not preclude the absence of surface rupture extending up into the upper soils as these clayey soils are expansive and subject to shrink and swell processes and livestock grazing activities. These processes may have destroyed fracturing features in the soil horizons. Table 2 above presents the ground water levels and corresponding elevations encountered in the borings. The disruption of the phreatic surface across the northern array of borings (MW-1 through MW-7) indicate the presence of the faulting found in Trench 1.

SECTION 5: CONCLUSIONS

5.1 SUMMARY
The West Napa Fault was confirmed at our exploratory excavations to be trending through the site and plotted in relationship to the current sewer line on Figure 7. Our recommended building setback lines that are depicted in Figure 7 and extend along the eastern and western bounding limits of the fault zone and apply to any future habitable structures at the site. These setbacks are equal to 50 feet along the eastern and western edge of the West Napa Fault Zone as observed in Trench 1 and Trench 2B and defined in the trench logs (Figures 8B and 9D). 2091

GIOVANNONI PROPERTY
582-4-1

Stakes were placed in the field and noted in the logs to assist in defining the fault zone and setbacks. The City of American Canyon surveyed the northern (1835227, 6482837) and southern stake (1834229, 6483276) in the California State Plane coordinate system.

Task C, presented in our original proposal, offers observation during construction activities if further definition of the fault zone is needed. Cornerstone staff can observe the sewer line trench walls during construction. Provided the sewer line excavation is dewatered our geologist will observe the trench walls to potentially identify the trace of the West Napa Fault. If located, this operation could reduce the length of pipe needing mitigation.

5.2 MITIGATING FAULT SURFACE RUPTURE
The current development concept calls for the installation of a sewer pipe easement along the West Napa Fault Zone and it will cross the fault zone at a location yet to be determined.

Cornerstone has provided estimated of fault offset or displacement and then provide guidelines intended to help minimize the effects of fault surface rupture for the pipeline. For the magnitude of fault surface rupture (displacement) we considered the methods of Wells and Coppersmith (1994), Wesnousky (2006), and Stirling et al., (1998) all of whom have developed regression curves relating the parameters of fault surface length, and estimated moment magnitude (Mw).

Of these sources that of Stirling et al. (1998) uses a larger data set for dip slip faults collected worldwide and consequently gives a more conservative estimate of average displacement for dip slip fault movement. We used the Mw established in the literature for the “airport segment” of the WNFZ as defined by Kelson and others (1998) being an 8-kilometer-long segment.

Displacement along the WNFZ could be distributed across more than one slip surface, however for design purposes it is advisable to assume that any individual shear within each fault zone could rupture. However, as we and other investigations (Darwin Meyers, 1983; and Wesling and Hanson, 2008) have identified up to three fault surface traces within the fault zone, it is not possible to know if such a surface displacement event would be distributed throughout the fault zone.

The use of structural fills is not considered an appropriate method of mitigating (primary) fault surface rupture for habitable structures.

As already noted, the identified “habitable building exclusion zone shown on the map is comprised of a wide zone where up to four fault surface traces trend through the site. Following the method outlines by Wells and Coppersmith (1994) and adopting conservative assumptions regarding the West Napa Fault Zone, we calculated a potential surface displacement of up to

2.6 feet within the fault zone. Such an event would most probably be distributed amongst more than one fault trace. In order to accommodate for a surface displacement event distributed through the fault zone (the habitable building exclusion zone) we recommend that future design teams for the planned pipelines refer to the provided guidelines for the seismic design and assessment of pipelines (Appendix A).
2092

GIOVANNONI PROPERTY
582-4-1

SECTION 6: LIMITATIONS
This report, an instrument of professional service, has been prepared for the sole use of Orchard Partners, LLC (Orchard) specifically to support the identification of the West Napa Fault in relationship to the proposed sewer line project in American Canyon, California. The opinions, conclusions, and recommendations presented in this report have been formulated in accordance with accepted geotechnical engineering practices that exist in Northern California at the time this report was prepared. No warranty, expressed or implied, is made or should be inferred.

Recommendations in this report are based upon the soil and ground water conditions encountered during our subsurface exploration. If variations or unsuitable conditions are encountered during construction, Cornerstone must be contacted to provide supplemental recommendations, as needed.

Orchard Partners, LLC may have provided Cornerstone with plans, reports and other documents prepared by others. Orchard Partners, LLC understands that Cornerstone reviewed and relied on the information presented in these documents and cannot be responsible for their accuracy.

Cornerstone prepared this report with the understanding that it is the responsibility of the owner or his representatives to see that the recommendations contained in this report are presented to other members of the design team and incorporated into the project plans and specifications, and that appropriate actions are taken to implement the geotechnical recommendations during construction.

Conclusions and recommendations presented in this report are valid as of the present time for the development as currently planned. Changes in the condition of the property or adjacent properties may occur with the passage of time, whether by natural processes or the acts of other persons. In addition, changes in applicable or appropriate standards may occur through legislation or the broadening of knowledge. Therefore, the conclusions and recommendations presented in this report may be invalidated, wholly or in part, by changes beyond Cornerstone’s control. This report should be reviewed by Cornerstone after a period of three (3) years has elapsed from the date of this report. In addition, if the current project design is changed, then Cornerstone must review the proposed changes and provide supplemental recommendations, as needed.

An electronic transmission of this report may also have been issued. While Cornerstone has taken precautions to produce a complete and secure electronic transmission, please check the electronic transmission against the hard copy version for conformity. Recommendations provided in this report are based on the assumption that Cornerstone will be retained to provide observation services during construction to confirm that conditions are similar to that assumed for design, and to form an opinion as to whether the work has been performed in accordance with the project plans and specifications. If we are not retained for these services, Cornerstone cannot assume any responsibility for any potential claims that may arise during or after construction as a result of misuse or misinterpretation of Cornerstone’s report by others. 2093

GIOVANNONI PROPERTY
582-4-1

Furthermore, Cornerstone will cease to be the Geotechnical-Engineer-of-Record if we are not retained for these services.

SECTION 7: REFERENCES
Aagaard, B.T., Blair, J.L., Boatwright, J., Garcia, S.H., Harris, R.A., Michael, A.J., Schwartz, D.P., and DiLeo, J.S., 2016, Earthquake outlook for the San Francisco Bay region 2014–2043 (ver. 1.1, August 2016): U.S. Geological Survey Fact Sheet 2016–3020, 6 p., http://dx.doi.org/10.3133/fs20163020.

Bailey Scientific, 1989, Geologic hazards investigation, Green Island Industrial Park, Napa County, California: Unpublished consulting report prepared for Panattoni Development Company, September 22. (AP-2328)
Bailey Scientific, 1990, Geologic hazards investigation, Green Island Industrial Park No. 3, Napa County, California, Price Property: Unpublished consulting report prepared for Mr. Harry Price, August 28. (AP-2451)
Bailey Scientific, 1991, Geologic Hazards Investigation, Struble Property, Parcel “B”, Green Island Road, Napa County, California, their proj. no. 378, date August 9.

Baize, S. Scotti, Oona, 2015, The Napa Earthquake, California (M=6; 24/08/2014), PostSeismic Report with Specific Focus on Surface Faulting, Institut De Radioprotection, Et De Surete Nucleaire, Rapport PRP-DGE n° 2015-00018.

Balbi & Chang Associates (BCA), 1984, Alquist-Priolo requirements, geologic study, proposed Tribotech facilities, Oat Hill, Napa County, California: Unpublished consulting report prepared for Ward Maher and Sons, Inc., February 1. (AP-1668).

Bay Soils, Inc., 1978 (1979), Geological and soil reconnaissance for a site near the Eagle Crest Property in American Canyon referenced in Kleinfelder (1992) and Terrasearch (1994).

Bezore, S., Randolph, C.E., and Witter, R.C., 2002, Geologic Map of Cuttings Wharf 7.5-Minute Quadrangle, Napa and Solano Counties, California, Version 1.0.

Bezore, S., Clahan, K.B., Sowers, J.M., and Witter, R.C., 2005, Geologic map of the Yountville

7.5′ Quadrangle, Napa County, California: A digital database: California Geological Survey, scale 1:24,000.

Blake, M.C., Bartow, J.A., Frizzell, V.A., Schlocker, Julius, Sorg, D.H., Wentworth, C.M., and Wright, R.H., 1974, Preliminary geologic map of Marin and San Francisco Counties, and parts of Alameda, Contra Costa, and Sonoma Counties, California: U.S. Geological Survey, Miscellaneous Field Studies Map MF-574, scale 1:62,500. 2094

GIOVANNONI PROPERTY
582-4-1

Borchart, G., 2016, Pedochronological Report for the Brigantino Property, Hollister, California, Prepared for Quantum Geotechnical, Inc., dated November 29.

Bortugno, E.J., 1982, Map showing recency of faulting in the Santa Rosa quadrangle, California, 1:250,000: California Division of Mines and Geology, Regional Geologic Map 2A, scale 1:250,000.

Brocher et al., 2015, The Mw 6.0 24 August 2014 South Napa Earthquake, Seismological Research Letters, Volume 86, Number 2A, March/April.

Bryant, W.A., 1982, West Napa fault zone and Soda Creek (east Napa) fault: California Division of Mines and Geology Fault Evaluation Report FER-129, 9 p., scale 1:24,000.

Bryant, W.A., compiler, 2000, Fault number 36b, West Napa fault, Napa County Airport section, in Quaternary fault and fold database of the United States: U.S. Geological Survey website, http://earthquakes.usgs.gov/hazards/qfaults, accessed 9/31/2017.

California Division of Mines and Geology, 1983, Special Studies Zones, Cuttings Warf Quadrangle, scale 1:24,000.

California Department of Conservation Division of Mines and Geology, 1998, Maps of Known Active Fault Near-Source Zones in California and Adjacent Portions of Nevada, International Conference of Building Officials, February, 1998.

California Division of Mines and Geology (2008), “Guidelines for Evaluating and Mitigating Seismic Hazards in California, Special Publication 117A, September.

California Geological Survey, 1983, State of California Seismic Hazard Zones, Cuttings Wharf

7.5-Minute Quadrangle, California: Seismic Hazard Zone Report 129, dated April 13.

California Geological Survey, 2017, West Napa Fault in the Napa and Cuttings Warf 7.5-minute Quadrangles, Napa and Solano County, California: Fault Evaluation Report 256, dated July 10.

Catchings, R. D., Goldman, M. R., Li, Y. G., and Chan, J. H., 2016, Continuity of the West Napa-Franklin fault zone inferred from guided waves generated by earthquakes following the 24 August 2014 MW 6.0 South Napa Earthquake: Bulletin of the Seismological Society of America, v. 106, no. 6, p. 2721–2746.

Clahan, K.B., Wagner, D.L., Saucedo, G.J., Randolph-Loar, C.E., and Sowers, J.M., 2004, Geologic map of the Napa 7.5′ Quadrangle, Napa County, California: A digital database:

California Geological Survey, scale 1:24,000.

Clahan, K.B., Wagner, D.L., Bezore, S.P., Sowers, J.M., and Witter, R.C., 2005, Geologic map of the Rutherford 7.5′ Quadrangle, Sonoma and Napa counties, California: A digital database:

California Geological Survey, scale 1:24,000.
2095

GIOVANNONI PROPERTY
582-4-1

Cornerstone Earth Group, 2016, Geotechnical and Geologic Fault Investigation, Napa Logistics Park – Phase 2, Devlin Road, American Canyon, California, Job No. 582-2-2, dated July 29, 2016.

Darwin Myers Associates (DMA), 1983, Geologic and Seismic Hazards Reconnaissance, Napa Airport Study Area, Napa County: Unpublished consulting report, August 10, 1 plate, 1 appendix, 31 pp.

Darwin Meyers Associates (DMA), 1984 (1985), Technical review of Alquist-Priolo report and addendum prepared for Tentative Parcel Map #2359-PM (Medeiros); report and map prepared by Kleinfelder & Associates (report dated 11-18-83; addendum dated 9-6-84): Unpublished consulting report prepared for Napa County Planning and Development Department, September

12. (AP-1669)
Donald Herzog & Associates, Inc. (DHA), 1984, Soil investigation Murdock Property, 1133 Larkin Way, Napa, California: Unpublished consulting report, May 5, 6 plates, 26 pp. (C-569) Earthquake Engineering Research Institute, 2014, Special Earthquake Report, M60. South Napa Earthquake of August 24, 2014.

Earthtec ltd., 1989, Preliminary Geotechnical/Geological Study, Struble Industrial Park, Green Island Road, Napa County, California, their job no., 88-274, dated February 24.

Environmental Management & Research, Inc. (EMRI), 1979, Fault hazard investigation, preliminary soils data, Green Island Industrial Park, Napa County, California: Unpublished consulting report, August 1, 24 pp. (AP-2328)
Environmental Management & Research, Inc. (EMRI), 1979, Fault hazard investigation, preliminary soils data, Green Island Industrial Park, Napa County, California: Unpublished consulting report, August 1, 24 pp. (AP-2328) ENGEO, 1977, Geologic Investigation for Mead Property, Napa County, California: Unpublished consulting report prepared for Ward-Ryder Company, March 21, 30 pp. (C-352) Fox, K.F. Jr., Sims, J.D., Bartow, J.A., and Helley, E.J., 1973, Preliminary geologic map of eastern Sonoma County and Western Napa County, California: U.S. Geological Survey Miscellaneous Field Studies Map MF-483.

Fox, K.F., Jr., 1983, Tectonic setting of late Miocene, Pliocene, and Pleistocene rocks in part of the Coast Ranges north of San Francisco, California: U.S. Geological Survey, Professional Paper 1239.

Graymer, R.W., Jones, D.L., and Brabb, E.E., 2002, Geologic map and map database of northeastern San Francisco Bay region, California -- most of Solano County and parts of Napa, Marin, Contra Costa, San Joaquin, Sacramento, Yolo, and Sonoma Counties: U.S. Geological Survey, Miscellaneous Field Studies Map MF-2403, scale 1:100,000. 2096

GIOVANNONI PROPERTY
582-4-1

Geomatrix Consultants, Inc. (Geomatrix), 1998, Final Report, Walnut Creek Water Treatment Plant Expansion, Seismic Study—Phase II: Unpublished consulting report prepared for East Bay Municipal Utility District, October 30, 5 tables, 9 figures, 2 appendices, 63 pp.

Geotechnical Engineering Inc. (GEI), 1996, Soil Investigation Including Geologic Hazard Evaluation, New Assembly Hall, Kingdom Hall & Bridge, American Canyon Assembly of Jehovah’s Witnesses, Commerce & Hanna Drives, American Canyon, California:

Unpublished consulting report, 32 pp. (AP-2328).

Giblin Associates, 2005, Geologic Hazards and Preliminary Geotechnical Investigation, Oat Hill Subdivision, American Canyon, CA: Unpublished.

Google Earth, 2015, version 7.3.0, American Canyon and Napa, California, DigitalGlobe December 2015, https://www.google.com/earth/download/ge/ .

Graymer, R.W., Bryant, William, McCabe, C.A., Hecker, Suzanne, and Prentice, C.S., 2006, Map of Quaternary-active faults in the San Francisco Bay region: U.S. Geological Survey, Scientific Investigations Map SIM-2919, scale 1:275,000.

Hardebeck, J.L., Shelley, D.R., 2016, Aftershocks of the 2014 South Napa, California Earthquake: Complex Faulting on Secondary Faults, Seis Soc America Vol 106, No. 1100-1109, June 2016.

Harding Lawson Associates, 1987, Site reconnaissance south of Oat Hill in American Canyon. referenced in Kleinfelder (1992) and Terrasearch (1994).

Hart, E.W., and Bryant, W.A., 1999, Fault-rupture hazard zones in California: California Division of Mines and Geology Special Publication 42, 38 p. revised 1997, supplements 1 and 2 added 1999.

Helley, E.J. and Herd, D.G., 1977, Map showing faults with Quaternary displacement, northeastern San Francisco Bay region, California: U.S. Geological Survey, Miscellaneous Field Studies Map MF-881, scale 1:125,000.

Helley, E.J., Graymer, R.W., Phelps, G.A., Showalter, P.K., and Wentworth, C.M., 1994, Quaternary geology of Santa Clara Valley, Santa Clara, Alameda, and San Mateo Counties, California: a digital database: U.S. Geological Survey, Open-File Report OF-94-231, scale 1:50,000.

Hudnut, K., 2014, Fault Afterslip Forecast for the 24 August, 20-14, Report from U.S.G.S. to FEMA for Interagency Agreement HSFE09-15-X-0805, DR-4193. (Donald) Herzog & Associates, Inc. (DHA), 1988a, Report geologic hazard investigation, 37- Acre Residential Subdivision AP #85-040-21, south of Eucalyptus Drive, American Canyon, California: Unpublished consulting report prepared for Mr. David Strunk, April 13. (AP-2233) 2097

GIOVANNONI PROPERTY
582-4-1

(Donald) Herzog & Associates, Inc. (DHA), 1988b, Geologic investigation for a portion of Eagle Crest Property in American Canyon referenced in Kleinfelder (1992) and Terrasearch (1994). (Donald) Herzog & Associates, Inc. (DHA), 1989, Report addendum, 37-Acre Residential Subdivision AP #85-040-21, south of Eucalyptus Drive, American Canyon, California: Unpublished consulting report prepared for Mr. David Strunk, February 10. (AP-2233) Hudnut, K.W., Brocher, T.M., Prentice, C.S., Boatwright, J., Brooks, B.A., Aagaard, B.T., Blair, J.L., Fletcher, J.B., Erdem, J.E., Wicks, C.W., Murray, J.R., Pollitz, F.F., Langbein, J., Svarc, J., Schwartz, D.P., Ponti, D.J., Hecker, S., DeLong, S., Rosa, C., Jones, B., Lamb, R., Rosinski, A., McCrink, T.P., Dawson, T.E., Seitz, G., Rubin, R.S., Glennie, C., Hauser, D., Ericksen, T., Mardock, D., Hoirup, D.F., and Bray, J.D., 2014, Key recovery factors for the August 24, 2014, South Napa earthquake: U.S. Geological Survey Open-File Report 2014–1249, 51 p., https://dx.doi.org/10.3133/ofr20141249.

Joyce Associates, 2016, Peer Review of Fault Hazard Investigation of West Napa Logistics Napa, California.

Kelson, K.I., Unruh, J.R., and Baldwin, J.N., 2005, Late Quaternary deformation in the northeastern East Bay hills, San Francisco Bay region (abs.): Geological Society of America, Abstracts with Programs, v. 37, no. 4, p. 106.

Kleinfelder & Associates (J.H. Kleinfelder), 1984a, Addendum to fault location report, tentative parcel map #2359-PM, Mederios Subdivision: Unpublished consulting report, September 6. (AP1669) Kleinfelder & Associates (J.H. Kleinfelder), 1984b, Engineering geology evaluation report, Oat Hill Industrial Subdivision, TR Contract No. 84-2, Napa County, California:

Unpublished consulting report, August 2. (AP-1668)
Kleinfelder & Associates (J.H. Kleinfelder), 1984c, Response to comments concerning engineering geology evaluation report, Oat Hill Industrial Subdivision, TR Contract No.

Donald Herzog & Associates, Inc. (DHA), 1984, Soil investigation Murdock Property, 1133 Larkin Way, Napa, California: Unpublished consulting report, May 5, 6 plates, 26 pp. (C-569)
Kleinfelder and Associates (J.H. Kleinfelder), 1990a, Preliminary geological engineering evaluation, commercial development at the southwest corner of the intersection of Broadway Street and American Canyon Road, Napa County Assessor’s Parcel No. 059-110-30 & 33, Napa County, California: Unpublished consulting report prepared for Elliott Real Estate, June

11. (AP-2483)
2098

GIOVANNONI PROPERTY
582-4-1

Kleinfelder and Associates (J.H. Kleinfelder), 1990b, Addendum to preliminary geological engineering evaluation dated June 11, 1990, for the proposed commercial development to be located on Napa County Tax Assessor’s Parcel No. 059-110-30 and 33, Napa County, California: Unpublished consulting report prepared for Elliott Real Estate, September 18. (AP-2483)
Kleinfelder and Associates (J.H. Kleinfelder), 1990c, Second addendum to preliminary geological engineering evaluation dated June 11, 1990 and September 18, 1990, for the proposed commercial development to be located on Napa County Tax Assessor’s Parcel No. 059-110-30 and 33, Napa County, California: Unpublished consulting report prepared for Elliott Real Estate, November 5. (AP-2483)
Kleinfelder and Associates (J.H. Kleinfelder), 1992, Fault investigation report Eagle Crest Subdivision: Unpublished consulting report prepared for American Canyon Associates, May 28. (AP-2950)
Knudsen, K.L., Sowers, J.M., Witter, R.C., Wentworth, C.M., Helley, E.J., 2000, Preliminary geologic maps of the Quaternary deposits and liquefaction susceptibility, nine-county San Francisco Bay Region, California: A digital database: U.S. Geological Survey Open-File Report 00-44, version 1.0, scale 1:52,500.

Langenheim, V.E., Graymer, R.W., and Jachens, R.C., 2006a, Geophysical setting of the ML

5.2 Yountville, California, earthquake: Implications for seismic hazard in Napa Valley, California:

Bulletin of the Seismological Society of America, v. 96, no. 3, pp. 1192-1198.

Langenheim, V.E., Roberts, C.W., McCabe, C.A., McPhee, D.K., Tilden, J.E., and Jachens, R.C., 2006b, Preliminary isostatic gravity map of the Sonoma Volcanic field and vicinity, Sonoma and Napa counties, California: U.S. Geological Survey, Open-File Report 2006-1056, scale 1:100,000.

Pampeyan, E.H., 1979, Preliminary map showing recency of faulting in coastal north-central California: U.S. Geological Survey, Miscellaneous Field Studies Map MF-1070, scale 1:250,000.

Ponti, D.J., Rosa, C., and Blair, J.L., 2017, Observations of ground deformation from the August 24, 2014 South Napa Earthquake: U.S. Geological Survey Open-File Report (in press).

Raney Geotechnical, 2003a, Fault Rupture Hazard Investigation, Green Island Industrial Park, Approximately 50-Acre Site, Hanna Drive South of North Slough, American Canyon, California: Unpublished consulting report prepared for Industrial & Commmercial Contractors, L.P., January 24.

Raney Geotechnical, 2003b, Geotechnical Investigation, Commerce 197 Building, Commerce Boulevard, American Canyon, California: Unpublished consulting report. 2099

GIOVANNONI PROPERTY
582-4-1

Raney Geotechnical, 2007, Supplemental Fault Rupture Hazard Investigation, Green Island Industrial Park, Approximately 50-Acre Site, Hanna Drive South of North Slough, American Canyon, California: Unpublished consulting report.

Sims, J.D., Fox, K.F., Bartow, J.A., and Helley, E.J., 1973, Preliminary geologic map of Solano County and parts of Napa, Contra Costa, Marin, and Yolo Counties, California: U.S. Geological Survey, Miscellaneous Field Studies Map MF-484, scale 1:62,500.

Sowers, J.M., Noller, J.S., and Lettis, W.R., 1998, Quaternary geology and liquefaction susceptibility, Napa, California 1:100,000 quadrangle, a digital database: U.S. Geological Survey, Open-File Report OF-98-460, scale 1:100,000.

Stirling, M., Rhoades, D., and Berryman, K., 1998, Evaluation of Wells and Coppersmith (1994) Earthquake and Fault Relationships in the New Zeland Context, Institute of Geological & Nuclear Sciences Ltd., prepared for the Earthquake Commission Research Foundation.

Terrasearch, Inc., 1994, Fault hazard investigation on Eagle Crest Property, American Canyon, California: Unpublished consulting report, October 6. (AP-2950) Unruh, J.R., Kelson, K.I., and Barron, A., 2002, Critical evaluation of the northern termination of the Calaveras fault, eastern San Francisco Bay area, California: Final Technical Report, U.S.

Geological Survey, National Earthquake Hazards Reduction Program Award Number 00-HQGR-0082, July, 72 pp.

U.S. Geological Survey and California Geological Survey, 2006, Quaternary fault and fold database for the United States, accessed December 15, 2001, from USGS web site:
http://earthquake.usgs.gov/regional/qfaults/
U.S. Geological Survey Earthquake Hazards Program. 2010. Quaternary Fault and Fold Database for the United States. Available: <http://earthquake.usgs.gov/hazards/qfaults/>. Last updated March 31, 2010. Accessed April, 2016
Wagner, D.L., and Bortugno, E.J., 1982, Geologic map of California, Santa Rosa sheet:

California Division of Mines and Geology, scale 1:250,000.

Wagner, D.L., Clahan, K.B., Saucedo, G.J., Randolph-Loar, C.E., and Sowers, J.M., 2004, Geologic map of the Sonoma 7.5′ quadrangle, Sonoma and Napa counties, California: A digital database: California Geological Survey, scale 1:24,000.

Wagner, D.L., Gutierrez, C.I., and Clahan, K.B., 2006, Geologic map of the south half of the Napa 30x60-minute quadrangle, California.: California Geological Survey, Preliminary Geologic Maps , scale 1:100,000.

Wagner, D.L. and Gutierrez, C.I., 2010, Preliminary Geologic Map of the Napa 30’ X 60’ Quadrangle, California.
2100

GIOVANNONI PROPERTY
582-4-1

Weaver, C.E., 1949, Geology and mineral deposits of an area north of San Francisco Bay, California: California Division of Mines Bulletin 149, 135 pp.

Wesling, J.R., Hanson, K.L., 2008, Mapping of the West Napa Fault Zone for Input into the Northern California Quaternary Fault Database; USGS Final Technical Report, USGS External Award Number 05HQAG0002.

Wesnousky, S. G., 2006, Predicting the endpoints of earthquake ruptures, Nature, 444, 358- 360, doi:10.1038/nature0525
Working Group on California Earthquake Probabilities, 2015, The Third Uniform California Earthquake Rupture Forecast, Version 3 (UCERF), U.S. Geological Survey Open File Report 2013-1165 (CGS Special Report 228). KMZ files available at:
www.scec.org/ucerf/images/ucerf3_timedep_30yr_probs.kmz Aerial Photographs: Geomorphic features on the following aerial photographs were interpreted at the U.S. Geological Survey in Menlo Park as part of this investigation:

Date Flight Frames Scale Type
March 1, 1948 GS-EF-2 117, 119 1:28,400 black & white
October 2, 1958 CSI-4V 115, 116 1:20,000 black & white
June 19, 1973 3567-3 222, 223 1:12,000 black & white
September 27, 1973 Area 2 3-14, 15 1:20,000 natural color 2101
Project Number
Figure Number
Date Drawn By
582-4-1
Figure 1
RRN
Vicinity Map
N
October 2017
Southern Pacific Railroad
Giovannoni Property
American Canyon, CA
Green Island Rd
Green Island Rd
Commerce Blvd
Paoli Loop Rd Napa Vallejo Hwy
SITE
2102
N
0 400 800

APPROXIMATE SCALE (FEET)
Base by Google Earth, dated 5/20/2017
Overlay: Topographical Map, undated
Project Number
Figure Number
Date Drawn By
582-4-1
Figure 2
RRN
Site Plan
Giovannoni Property
American Canyon, CA
October 2017
Earthquake
Regulatory Zone
Earthquake
Regulatory Zone
Proposed
lift station
Existing
sewer line
Devlin Road
Southern Pacific Railroad
Green Island Road
Approximate
Site Boundary
Approximate location of trench
Legend
Trench 1
Trench 2B
Trench 2A
Approximate mapped location of the West Napa Fault
Bryant et al., 1982
Gryamer et al., 2002
MW-8 through MW-12
(numbered west to east)
MW-1 through MW-7
(numbered west to east)
Approximate location of temporary
monitoring well
2103
SITE
Project Number
Figure Number
Date Drawn By
Figure 3
Vicinity Geologic Map
N
0 4000

APPROXIMATE SCALE (FEET)
2000
Contact- dashed where approximate,
dotted where concealed
Fault - solid where accurately located,
dashed where approximately located;
short dash where inferred;
dotted where concealed
Geologic Units Explanation
Artificial fill
Holocene alluvium, undifferentiated
Late Pleistocene to Holocene fan deposits
Late Pleistocene fan deposits
Early to middle Pleistocene fan or terrace deposits
Huichica Formation (Pliocene)
Fluvial gravel, sand, silt, and clay
af
Qf
GEOLOGIC MAP OF THE CUTTINGS WHARF 7.5' QUADRANGLE
NAPA AND SOLANO COUNTIES, CALIFORNIA: A DIGITAL DATABASE by Stephen Bezore, Carolyn E. Randolph-Loar, and Robert C. Witter Giovannoni Property
American Canyon, CA
582-4-1
October 2017 RRN
Qpf
Qha
Qoa
Th
2104
Project Number
Figure Number
Date Drawn By
Figure 4
RRN
Regional Fault Map
Base by USGS, California Quaternary Faults, Cooperator California Geological Survey N
0 4 8

APPROXIMATE SCALE (MILES)
SITE
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
2105
Project Number
Figure Number
Date Drawn By
Figure 5
AP Earthquake Regulatory Zone Map
N
0 4000

APPROXIMATE SCALE (FEET)
2000
Giovannoni Property
American Canyon, CA
582-4-1
October 2017 RRN
Base: State of California, Special Studies
Zones, Cutting Wharf Quadrangle
SITE
2106
Giovannoni Property
American Canyon, CA
Project Number
Figure Number
Date Drawn By
582-4-1
Figure 6
RRN
Previous Local Investigations within 2¹⁄ Miles
October 2017
Base by Google Earth, dated 5/26/2014
N
0 2000 4000

APPROXIMATE SCALE (FEET)
1 2
3 4
5 6
8 7
9 10
11 12
13 Approximate Location of Previous Investigation
1 - Darwin Meyers Associates, 1983
2 - Wesling and Hanson, 2008
3 - Earthtec Limited, 1989
4 - EMRI, 1979
5 - EMRI, 1983
6 - Bailey Scientific, 1989
7 - Bailey Scientific, 1990
8 - Kleinfelder, 1984a, 1984b
Balbi and Chang, 1984
9 - Kleinfelder, 1983
10 - Herzog, 1988, 1989a, 1989b
11 - Bay Soils, 1979
12 - Darwin Meyers Associates, 1985
13 - Kleinfelder, 1990a, 1990b, 1990c
14 - Cornerstone, 2016
SITE
14 2107
Earthquake
Regulatory Zone
Proposed
lift station
Existing
sewer line
Devlin Road
Southern Pacific Railroad
Green Island Road
Approximate
Site Boundary
N
0 400 800

APPROXIMATE SCALE (FEET)
Approximate location of trench
Approximate Earthquake
Regulatory Zone
Approximate location of
seasonal unnamed creek
Legend
Base by Google Earth, dated 5/20/2017
Overlay: Topographical Map, undated
Project Number
Figure Number
Date Drawn By
Figure 7
RRN
Fault Investigation 582-4-1
Giovannoni Property
American Canyon, CA
October 2017
Trench 1
Trench 2B Trench 2A
West Napa
Fault
Qoa
Qha Qpf
Qf
Th
Qha
Geologic Units
Holocene alluvium, undifferentiated
Late Pleistocene to Holocene fan deposits
Late Pleistocene fan deposits
Early to middle Pleistocene fan or terrace deposits
Huichica Formation (Pliocene)
Fluvial gravel, sand, silt, and clay
Qf
Qpf
Qha
Qoa
Th
Approximate location of
West Napa Fault
(this investigation)
50 ft setback
2108
Project Number
Figure Number
Date Drawn By
Figure 8A
RRN
Trench 1
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South)
1”=5’ H:V
Trench 1
S 80 W
18 0 5 10 15 20 25 35 40
Distance (feet)
Elevation (feet)
MATCHLINE:

SEE BELOW
30 23
45 5
0 70 75 80 85 90 95 105 110
Distance (feet)
100
22 1
21 2
20 3
19 4
17 6
16 7
15 8
MATCHLINE:

SEE FIGURE 8B
55 60 65
50 55
Depth (feet)
MATCHLINE:

SEE ABOVE
5 0
1 2
3 4
6 7
8 Depth (feet)
14 9
9 10
13 10
18 Elevation (feet)
23 22
21 20
19 17
16 15
14 13
1B
2 1A
1B
2 1A
MW-7, 15 feet East
MW-6
MW-5
Gravels Carbonate clay
pad surfaces
1Ca1
1Cb1
Unit 1 – Holocene soils
1A Clay [CL/CH]:
stiff to very stiff, moist, dark brown to brown [10YR 3-2/1] with fine sand and occasional fine to coarse gravels. Block and prismatic A- and B-horizon soils. 1B Sandy clay [CL]:
stiff, moist, brown to gray brown [2.5YR 4-3/2] with fine to medium sand and occasional fine sub-rounded gravels. Weak to moderate blocky B-horizon soils. 1Ca1 Dark gray [10YR 4/1] clay with fine sands. A-horizon soils with moderately granular to weak blocky structure.
1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure.

Unit 2 – Holocene to Upper Pleistocene
2 Clayey sand and sandy clay with gravel [SC]:
dense and stiff, moist to wet, light brown to gray [2.5Y 6/2-4/3] fine to medium sands and clays.
sandy lense
Disseminated carbonate (I & II) Qha
Qoa
Qoa
Qoa
Qoa
Qha
Qha
2a
Qoa
2109
Project Number
Figure Number
Date Drawn By
Figure 8B
RRN
Trench 1 (continued)
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South)
1”=5’ H:V
Trench 1 (continued)
S 88 W
180 185 190 195 200 205 215
Distance (feet)
165 170 175 210
MATCHLINE:

SEE ABOVE
MW-2, 27 feet West
18 Elevation (feet)
23 22
21 20
19 17
16 15
14 13
24 18
Elevation (feet)
23 22
21 20
19 17
16 15
14 13
24 MW-3
2b
2a
MW-1, 83 feet West
110 115 120 125 130 135 145 150
Distance (feet)
MATCHLINE:

SEE BELOW
140 155
5 0
1 2
3 4
6 7
8 160 165
Depth (feet)
9 10
MATCHLINE:

SEE FIGURE 8A
11 18
Elevation (feet)
23 22
21 20
19 17
16 15
14 13
2b
1Ca1
1Cb1
MW-4 Stake in field
at 134 ft.
2a
3 3
Hand-auger, dry 2b
2a
West Napa Fault Zone
N 10 W, 89 W N 6 W, 90 W
Unit 2 – Holocene to Upper Pleistocene
2a Sandy clay [SC]:
stiff, moist to wet, light brown [2.5Y 6/3] with fine to medium sands. 2b Clayed sands [CL]:
dense, moist to wet, light brown [2.5Y 5/2] with occasional sub-angular to sub rounded gravels. Orange mottles [10YR 5-4/4] common above the saturated zone.

Unit 3 – sandy channel deposits
3 Clayey sand with gravels [SC]:
dense, moist, light brown to brown [10YR 4/2-3/4] with sub-angular to sub-rounded shale and chert.

Unit 1 – Holocene soils
1Ca1 Dark gray [10YR 4/1] clay with fine sands. A-horizon soils with moderately granular to weak blocky structure.
1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure. 1Cb1
1Ca1
Disseminated carbonate (I & II) Qoa Qoa
Qoa
Qoa
Qoa
2110
Project Number
Figure Number
Date Drawn By
Figure 8C
RRN
Detail of Trench 1 (130' to 142')
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South)
1”=5’ H:V
Detail of Trench 1 (130' to 142')
S 88 W
2a
130 132 134 136 138 140
Distance (feet)
Elevation (feet)
142
23 0
Depth (feet)
21 2
19 4
17 6
15 8
13 10
20 3
22 1
18 5
16 7
14 9
12 11
Stake in field at 134 ft
N 1 W, 88 W
across trench
N 10 W, 89 W
across trench
N 6 W, 90 W
across trench
2b
3 MW-4 Carbonate on
fracture surfaces Unit 1 – Holocene soils
1Ca1 Dark gray [10YR 4/1] clay with fine sands. A-horizon soils with moderately granular to weak blocky structure.
1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure.

Unit 2 – Holocene to Upper Pleistocene
2a Sandy clay [SC]:
stiff, moist to wet, light brown [2.5Y 6/3] with fine to medium sands. 2b Clayed sands [CL]:
dense, moist to wet, light brown [2.5Y 5/2] with occasional sub-angular to sub rounded gravels. Orange mottles [10YR 5-4/4] common above the saturated zone.

Unit 3 – sandy channel deposits
3 Clayey sand with gravels [SC]:
dense, moist, light brown to brown [10YR 4/2-3/4] with sub-angular to sub-rounded shale and chert.

Carbonate on
fracture surfaces
1Ca1
1Cb1
3 gravel lenses
Qoa
Qoa
Qoa
Qoa
Disseminated carbonate (I & II)
2111
Project Number
Figure Number
Date Drawn By
Figure 9A
RRN
Trench 2A
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South)
1”=5’ H:V
Trench 2A
N 83 E
26 05 10 15 20 25 35 40
Distance (feet)
Elevation (feet)
MATCHLINE:

SEE BELOW
30 31
45 50
301
292
283
274
256
247
238
50 55
Depth (feet)
22 9
21 10
60 1Cb1
2 1Ca1
1Cb2
MW-12
Gravels
MW-11
75 80 85 90 95 100 110 115
Distance (feet)
105
MATCHLINE:

SEE BELOW
60 65 70
MATCHLINE:

SEE ABOVE
501234678
Depth (feet)
9 10
120
26 Elevation (feet)
31 30
29 28
27 25
24 23
22 21
MW-10
2 1Cb1
1Ca1
1Cb2
Carbonate
nodules
1Ca2
Unit 1 – Holocene soils
1Ca1 Dark gray [10YR 4/1] clay with fine sands. A-horizon soils with moderately granular to weak blocky structure.
1Ca2 Dark gray [5YR 4/1] clay with fine sands and occasional well-rounded gravels at base of horizon. A-horizon soils with massive to moderately granular structure. 1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure. 1Cb2 Brown to gray brown [10YR 4/3 – 5/2] clay with fine sands. B-horizon soils with moderate blocky structure.

Unit 2 – Holocene to Upper Pleistocene
2 Clayey sand and sandy clay with gravel [SC]:
dense and stiff, moist to wet, light brown to gray [2.5Y 6/2-4/3] fine to medium sands and clays.

Qf
Qoa
Qoa
Qf
Qoa
Qoa
Hematite concretions
Disseminated carbonate and
carbonate concretions (I, II, & III)
2112
Project Number
Figure Number
Date Drawn By
Figure 9B
RRN
Trench 2A (continued)
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South)
1”=5’ H:V
Trench 2A (continued)
N 83 E
195 200 205 210 215 220 230 235
Distance (feet)
225
MATCHLINE:

SEE BELOW
180 185 190
MATCHLINE:

SEE FIGURE 9A
5 0
1 2
3 4
6 7
8 Depth (feet)
9 10
240
MW-8
26 Elevation (feet)
31 30
29 28
27 25
24 23
22 21
1Cb2
1Ca2
2 1Cb1
3 Carbonate
nodules
Unit 2 – Holocene to Upper Pleistocene
2 Clayey sand and sandy clay with gravel [SC]:
dense and stiff, moist to wet, light brown to gray [2.5Y 6/2-4/3] fine to medium sands and clays.

Unit 3 – sandy channel deposits
3 Clayey sand with gravels [SC]:
dense, moist, light brown to brown [10YR 4/2-3/4] with sub-angular to sub-rounded shale and chert.

Unit 4 - Huichica Formation (Pliocene)
4a Sandy clay [CL]:
stiff to very stiff, moist, clay with fine sands, yellowish brown [10YR 5/6], brown [7.5YR 4-5/4-6] and reddish brown [2.5YR 6/3]. Bedding present between sandier and clayey packets of sediment.

Unit 1 – Holocene soils
1Ca2 Dark gray [5YR 4/1] clay with fine sands and occasional well-rounded gravels at base of horizon. A-horizon soils with massive to moderately granular structure. 1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure. 1Cb2 Brown to gray brown [10YR 4/3 – 5/2] clay with fine sands. B-horizon soils with moderate blocky structure.
120 125 130 135 140 145 155 160
Distance (feet)
MATCHLINE:

SEE ABOVE
150 165
5 0
1 2
3 4
6 7
8 170 175
Depth (feet)
9 10
180
MATCHLINE:

SEE FIGURE 9B
26 Elevation (feet)
31 30
29 28
27 25
24 23
22 21
MW-9
2 1Cb2
1Ca2
Carbonate
nodules
Roots 1Cb1 Krotovina Root Roots
Qoa
Qoa
Qoa
Hematite concretions
Disseminated carbonate and
carbonate concretions (I, II, & III)
Disseminated carbonate (I & II)
Th 4a
Qoa Qoa
Qoa
Th
2113
Project Number
Figure Number
Date Drawn By
Figure 9C
RRN
Trench 2A (continued)
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South)
1”=5’ H:V
Trench 2A (continued)
N 83 E
255 260 265 270 275 280 290 294.5
Distance (feet)
240 245 250 285
MATCHLINE:

SEE ABOVE
26 Elevation (feet)
31 30
29 28
27 25
24 23
22 21
26 Elevation (feet)
31 30
29 28
27 25
24 23
22 21
32 33
Carbonate
nodules
Sandstone
gravels
Carbonate precipitate
in joints
1Cb2
1Ca2
2 3
1Cb1
4a
Unit 2 – Holocene to Upper Pleistocene
2 Clayey sand and sandy clay with gravel [SC]:
dense and stiff, moist to wet, light brown to gray [2.5Y 6/2-4/3] fine to medium sands and clays.

Unit 3 – sandy channel deposits
3 Clayey sand with gravels [SC]:
dense, moist, light brown to brown [10YR 4/2-3/4] with sub-angular to sub-rounded shale and chert.

Unit 4 - Huichica Formation (Pliocene)
4a Sandy clay [CL]:
stiff to very stiff, moist, clay with fine sands, yellowish brown [10YR 5/6], brown [7.5YR 4-5/4-6] and reddish brown [2.5YR 6/3]. Bedding present between sandier and clayey packets of sediment.

Unit 1 – Holocene soils
1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure. 1Cb2 Brown to gray brown [10YR 4/3 – 5/2] clay with fine sands. B-horizon soils with moderate blocky structure.

Qoa Qoa
Qoa
Th
Qoa
Th Qoa Qoa
Hematite concretions
Disseminated carbonate (I & II)
2114
Project Number
Figure Number
Date Drawn By
Figure 9D
RRN
Trench 2B
Giovannoni Property
American Canyon, CA
October 2017
582-4-1
(View Looking South, north wall
cleaned, log is reversed)
1”=5’ H:V
Trench 2B
N 88 E
29 0 5 10 15 20 25 35 40
Distance (feet)
Elevation (feet)
MATCHLINE:

SEE BELOW
30 34
45 33
32 31
30 28
27 26
50 55
25 24
60 29
Elevation (feet)
34 33
32 31
30 28
27 26
25 24
35 36
37 1Ca2
1Cb1
4a
4a
4b
Gravels
Shearing
N 61 W, 48 NE
Rake: 0° Dip Slip Fault
N 50 W, 60 SW
Dip Slip Fault
S 41 E, 22 SW
Rake: 2° SW
Sandy
lenses
Fault
S 39 E, 53 SE
Rake: 80° SEW
Stike Slip Fault
N 24 W, 85 NE
Bedding
N 34 W, 34 SE
Unit 1 – Holocene soils
1Ca2 Dark gray [5YR 4/1] clay with fine sands and occasional well-rounded gravels at base of horizon. A-horizon soils with massive to moderately granular structure. 1Cb1 Brown [10YR 4/3] clay with fine sands and occasional fine sub-angular gravels.

B-horizon soils with strong blocky to prismatic structure. 75 80 85 90 95 100
Distance (feet)
60 65 70
MATCHLINE:

SEE ABOVE
29 Elevation (feet)
34 33
32 31
30 28
27 35
36 37
29 Elevation (feet)
34 33
32 31
30 28
27 35
36 37
38 1Ca2 39
1Cb2
4b 4b
4a
4a
Gravelly
lense
Carbonate
accumulation
Sandy
lense
Jointing
N 55 E, 80 SE Shear zone
N 26 W, 80 NE
4a
4c
West Napa Fault Zone
Unit 4 - Huichica Formation (Pliocene)
4a Sandy clay [CL]:
stiff to very stiff, moist, clay with fine sands, yellowish brown [10YR 5/6], brown [7.5YR 4-5/4-6] and reddish brown [2.5YR 6/3]. Bedding present between sandier and clayey packets of sediment.
4b Clayey sands with gravels [SC]:
very dense, moist, yellowish red to reddish brown [5YR 4-5/4-6]. 4c Clay with silt [CL]:
medium stiff, moist, olive yellow [2.5Y 6/6], clay with silt and some fine to coarse angular to sub-angular gravel and fine carbonate nodules. Carbonate accumulations formed on the vertical margins where the slide planes of the fault were observed.

Fault gouge.

Stake in field at 34 ft
Qoa Th
Th
Th
Qoa
Th
Qoa
Th
Disseminated carbonate (I & II)
2115
582-4-1
APPENDIX A : SEISMIC DESIGN AND ASSESSMENT GUIDLINES
2116
Catalog No. L51927
Guidelines for the Seismic Design and Assessment of Natural Gas and Liquid Hydrocarbon Pipelines
Contract PR-268-9823
Prepared for the
Pipeline Design, Construction & Operations Technical Committee of
Pipeline Research Council International, Inc.

Prepared by the following Research Agencies:

D.G. Honegger Consulting
D.J. Nyman and Associates
Authors:

Douglas G. Honegger
Douglas J. Nyman
Publication Date:

October 1, 2004
2117
“This report is furnished to Pipeline Research Council International, Inc. (PRCI) under the terms of PRCI PR-268-9823, between PRCI and D.G. Honegger Consulting. The contents of this report are published as received from D.G. Honegger Consulting. The opinions, findings, and conclusions expressed in the report are those of the authors and not necessarily those of PRCI, its member companies, or their representatives.

Publication and dissemination of this report by PRCI should not be considered an endorsement by PRCI or D.G. Honegger Consulting, or the accuracy or validity of any opinions, findings, or conclusions expressed herein.

In publishing this report, PRCI makes no warranty or representation, expressed or implied, with respect to the accuracy, completeness, usefulness, or fitness for purpose of the information contained herein, or that the use of any information, method, process, or apparatus disclosed in this report may not infringe on privately owned rights. PRCI assumes no liability with respect to the use of, or for damages resulting from the use of, any information, method, process, or apparatus disclosed in this report.

The text of this publication, or any part thereof, may not be reproduced or transmitted in any form by any means, electronic or mechanical, including photocopying, recording, storage in an information retrieval system, or otherwise, without the prior, written approval of PRCI.”
Pipeline Research Council International Catalog No. L51927 Price:

Copyright, 2005
All Rights Reserved by Pipeline Research Council International, Inc.

PRCI Reports are Published by Technical Toolboxes, Inc. 3801 Kirby Drive, Suite 340
Houston, Texas 77098
Tel: 713-630-0505
Fax: 713-630-0560
Email: [email protected]
2118
PR-268-9823
Acknowledgements
This report was prepared for the PRCI Design, Construction, and Operations Technical Committee. Ad hoc steering group members participating on this project over its fiveyear duration include the following:

Lynn A. Christensen
Colorado Interstate Gas Company
Martin McLamb
BP
Richard W. Gailing (Chair)
Southern California Gas Company
Moness Rizkalla
Visitless Integrity Assessment Ltd.

Alan G. Glover
TransCanada PipeLines Ltd.

Joe Zhou
TransCanada PipeLines Ltd.

Stanley R. Kitt
Foothills Pipe Lines, Ltd.

This project would not have been possible without the financial contributions of Southern California Gas Company and the authors greatly appreciate their support. We also recognize M. McLamb and J. Zhou for their contributions to refine the material on strain acceptance criteria and M. Rizkalla for his valuable review comments on early versions of the guidelines. We are especially thankful to R. Gailing for his leadership in support of the project and his thorough review of the various drafts of these guidelines.

The authors are indebted to the individuals contracted to serve as expert reviewers during the development of these guidelines. Their comments and suggestions were valuable in refining the content and organization of these guidelines.

Jean M.E. Audibert (2000 review)
Fugro-McClelland Marine Geosciences,
Inc.

Michael J. Rosenfeld (1999 review)
Kiefner and Associates, Inc.

James D. Hart (2000 review)
SSD, Inc.

Robert W. Warke (1999 review)
Southwest Research Institute, Inc.
(Currently at Le Tourneau University)
William A. Lettis (2000 review)
William A. Lettis and Associates, Inc.

Thomas J.E. Zimmerman (1999 and
2000 review)
Center for Engineering Research, Ltd.

David W. Murray (1999 review)
D.W. Murray Engineering, Ltd.

Finally, we are grateful to E.L. Von Rosenberg, Consulting Engineer, Materials and Welding Technology, for his valuable assistance in improving the guidance related to specifications of pipe material and welding procedures provided in the commentary. 2119
This Page Intentionally Left Blank
2120
Introduction
Project PR-268-9823 was initiated in February 1998 by PRCI to provide current seismic guidelines for the design and assessment of natural gas transmission pipelines. These guidelines were refined using two rounds of review by outside technical experts in 1999 and 2000. The PRCI ad hoc steering group for the project also provided regular input regarding the scope and technical content for these guidelines. A decision was made in late 2000 to expand the scope to liquid hydrocarbon pipelines (crude oil and refined products) based upon the identical analytical treatment of seismic design and assessment of these types of pipelines.

Much of the current seismic practice can be traced to research conducted in the 1970s to develop design criteria and procedures for the Trans-Alaska oil pipeline. Many of the procedures in these guidelines have been implemented, in one form or another, by the major natural gas pipeline operators in California. The recognition of credible seismic hazards in other parts of the United States has been the primary driver for extending current practices in California to the rest of the United States. The experience in designing pipelines in seismically active regions of the United States has also provided a basis for specifying pipeline seismic design practices in other parts of the world.

The first comprehensive set of seismic guidelines for oil and gas pipeline systems was published in 1984 by the American Society of Civil Engineers1 . These guidelines were
the product of a collaborative effort among pipeline operators, consultants, and researchers. The 1984 ASCE guidelines have become accepted as a de-facto standard for seismic design within the oil and gas industry. Since their publication in 1984, considerable advancements have been made in seismic hazard definition and the understanding of pipeline response to earthquake hazards.

The present guidelines are intended to be an update of the 1984 ASCE guidelines relating to buried pipelines transporting natural gas and liquid hydrocarbons. To that end, preparation of these guidelines attempts to take full advantage of recent research findings with respect to soil loading on buried pipelines, acceptable strain-based pipeline limit states, and current analysis tools. To the extent possible, these guidelines have been stated in a concise manner with minimal discussion so as to keep the main text clear and concise. More in-depth discussions of the bases for recommendations in these guidelines are provided in a comprehensive Commentary at the end of this document. These guidelines are the product of considerable peer review by experts within the pipeline industry, key investigators at various research organizations, and consultants providing seismic design and assessment services to natural gas pipeline operators.

There continue to be advancements in understanding pipeline behavior and seismic hazard definition. It is anticipated that these guidelines will be updated as ongoing research is incorporated into actual practice.

1 Nyman, D.J. (ed.), 1984. Guidelines for the Seismic Design of Oil and Gas Pipeline Systems, American Society of Civil Engineers.

PR-268-9823
2121
This Page Intentionally Left Blank
2122

GUIDELINES FOR THE SEISMIC DESIGN

AND ASSESSMENT OF NATURAL GAS AND

LIQUID HYDROCARBON PIPELINES
PART I

GUIDELINES AND

RECOMMENDED PROCEDURES
PR-268-9823
2123
This Page Intentionally Left Blank
2124
Table of Contents

1.0 GENERAL PROVISIONS ..................................................................................1-1

1.1 Application ................................................................................................... 1-1

1.2 Process..........................................................................................................1-1

1.3 Alternate Procedures.....................................................................................1-3

1.4 Required Information....................................................................................1-3

1.4.1 Pipeline Information ...........................................................................1-3

1.4.2 Soil Information ..................................................................................1-4

1.4.3 Ground Deformation Hazard Definition .............................................1-4

1.5 Acronyms...................................................................................................... 1-4

1.6 Nomenclature................................................................................................1-5

1.7 References...................................................................................................1-10

1.8 Units............................................................................................................1-11

2.0 QUANTIFYING SEISMIC HAZARDS.............................................................2-1

2.1 Surface Faulting............................................................................................2-1

2.1.1 Empirical Methods for Estimating Fault Displacements....................2-2

2.1.2 Orthogonal Displacement Components..............................................2-3

2.1.3 Design Fault Displacements................................................................2-4

2.2 Peak Ground Acceleration............................................................................2-6

2.2.1 Near Source Earthquakes in the Western United States.....................2-7

2.2.2 Cascadia Earthquakes .........................................................................2-8

2.3 Liquefaction..................................................................................................2-8

2.3.1 Assessing Liquefaction Potential........................................................2-9

2.3.2 Displacements Associated with Relative Buoyancy.........................2-13

2.3.3 Displacements Related to Liquefaction-Induced Settlement ............2-14

2.4 Lateral Spread Movement........................................................................... 2-14

2.5 Landslides...................................................................................................2-16

2.6 Seismic Wave Propagation.........................................................................2-18

2.7 Transient Ground Deformation...................................................................2-19

3.0 PIPELINE PERFORMANCE CRITERIA .......................................................3-1

3.1 New Pipelines and Pipelines Compatible with Modern
Construction Practices.................................................................................. 3-2

3.1.1 Performance Goal: Maintain Pressure Integrity.................................3-2 PR-268-9823
2125

3.1.2 Performance Goal: Maintain Normal Operability..............................3-3

3.1.3 Load-Controlled Conditions................................................................3-4

3.2 Considerations for Offshore Pipelines.......................................................... 3-5

3.3 Pipelines Not Compatible with Current Construction Practices...................3-6

4.0 PIPELINE ANALYSIS PROCEDURES ...........................................................4-1

4.1 Pipe Element Definition ............................................................................... 4-1

4.2 Pipe Stress-Strain Definition ........................................................................4-1

4.3 Soil Spring Definition................................................................................... 4-2

4.3.1 Axial Soil Springs...............................................................................4-2

4.3.2 Lateral Soil Springs.............................................................................4-3

4.3.3 Vertical Uplift Soil Springs.................................................................4-4

4.3.4 Vertical Bearing Soil Springs..............................................................4-5

4.4 Extent of Pipeline Model.................................................................................. 4-5

4.5 Representation of Applied Ground Movement............................................. 4-6

4.6 Sensitivity Analyses......................................................................................4-6

5.0 MITIGATION OPTIONS ...................................................................................5-1

5.1 Modifying Pipeline Loading Conditions...................................................... 5-1

5.1.1 Reducing Soil Loads by Minimizing Soil Strength Properties...........5-1

5.1.2 Unanchored Length.............................................................................5-4

5.1.3 Isolating Pipelines from Ground Displacement..................................5-4

5.2 Modify Pipeline Configuration..................................................................... 5-5

5.3 Modify Emergency Response Procedures.................................................... 5-5 APPENDIX A: Considerations for Establishing Performance Goals APPENDIX B: Seismic Design Provisions in Existing Standards APPENDIX C: Probabilistic Seismic Hazard Analysis
APPENDIX D: Example Calculations
APPENDIX E: Guideline Reviewers
PR-268-9823
2126
SECTION 1: General 1-1

1.0 GENERAL PROVISIONS
These guidelines provide recommended procedures and methods for the assessment of new and existing natural gas and liquid hydrocarbon pipelines subjected to seismicrelated loading conditions. These guidelines are believed to be representative of, and consistent with, current practices for seismic hazard mitigation for these pipelines.

The specific procedures contained in these guidelines are related to determining pipeline response to predefined seismic hazards (e.g., liquefaction, fault displacement, slope instability). The process of defining these hazards is not part of the scope of these guidelines, although considerable guidance is given on typical practice and simplified approaches for estimating the potential severity of seismic hazards. The suitability of the methods described in these guidelines to define seismic hazards should be confirmed by specialists with appropriate knowledge and expertise.

1.1 Application
These guidelines are limited to onshore and offshore buried steel pipelines fabricated with full-penetration girth welds. Although the procedures of these guidelines may be extended to other applications (e.g., low-pressure distribution mains), the special considerations that apply to buried piping systems with conditions such as numerous tieins of distribution facilities and potential fatigue effects of large cyclic pressure fluctuations may occur during operation are not addressed.

1.2 Process
The general process for performing a design or an assessment of a buried pipeline is illustrated by the flow chart in Figure 1.1. As indicated in the flow chart, application of the procedures in these guidelines may require input from experts in specialized technical disciplines, particularly in the areas of seismology, geology, soil mechanics, and materials and welding technology. The content of these guidelines is limited to guidance regarding typical practices in these areas as they relate to the seismic assessment of buried pipelines. This information is also valuable to facilitate interaction among pipeline engineers and individuals having the requisite specialized expertise.

Most of the approaches for quantifying seismic hazards presented in these guidelines are based upon empirical relationships developed from past earthquake observations. As a result, each such approach is limited to the particular conditions characteristic of the earthquake data used to develop the underlying empirical relationship. Understanding the relative importance of these limits and rational approaches for implementing modifications for site-specific conditions is the primary reason for relying on individuals with special technical expertise.

The procedures in this document assume that the performance criteria and critical pipeline locations have been identified. The level of required seismic performance, as PR-268-9823
2127
SECTION 1: General 1-2
defined by the annual probability of occurrence of an unacceptable condition, is a decision that needs to be made on a case-by-case basis considering public safety, environmental impacts, operational requirements, economic consequences, regulatory requirements and corporate policies. Factors to be considered when establishing performance requirements are discussed in Appendix A.

The state of knowledge in the field of pipeline response does not allow the definition of well-defined post-yield strain limits associated with some levels of pipeline performance.

This is particularly true for cases where detailed knowledge of the pipeline material properties is lacking and the performance level is associated with maintaining pipeline pressure integrity. Considerable uncertainty also exists in defining seismic hazards. In these guidelines, the specification of partial safety factors associated with a rigorous reliability analysis has not been attempted. While such an approach may be desirable for specific pipelines, it is not considered practical for general application and, therefore, is not provided in these guidelines. The philosophy adopted in these guidelines is to consider conservative estimates of seismic hazards that are likely to occur. The likelihood of experiencing a seismic hazard is assumed to be equal to the acceptable probability of not achieving the performance goals for the pipeline. Pipeline response to the seismic hazard defined in this manner is assumed to utilize estimates of pipeline strength and strain capacity that are equal to or slightly more conservative than the median. The determination of what constitutes a “conservative” hazard definition and “median” pipeline strength and strain capacity is based largely upon judgment and procedures implemented in many pipeline projects and approved by various regulatory and oversight agencies over the past two decades.

These guidelines address the site-specific evaluation of a pipeline to a particular seismic ground displacement hazard. For a single pipeline, the number of site-specific conditions to be evaluated is typically based upon an assessment of conditions within 200 m to 300 m of the pipeline right-of-way. For transmission systems providing multiple feeds to large metropolitan regions, an alternate approach may be to first identify critical pipeline segments within the system. The critical segments are those portions of the system necessary to maintain life safety, avoid environmental damage, and provide reliable service. Critical segments of the pipeline may also include locations where options are not available for rapidly reconfiguring the system to isolate earthquake damage and reroute flow to maintain service. Identification of seismic hazards would then focus on the critical pipeline segments within the system. For large transmission systems in areas with numerous seismic hazards, this approach has the advantage of significantly reducing the level of effort for quantifying seismic hazards. A drawback of this approach is that damage that might occur to less critical portions of the pipeline system is not identified.

Of course, when examining a single pipeline for the level of reliable service, the entire pipeline becomes a critical segment as damage anywhere along the pipeline can lead to service interruption.

PR-268-9823
2128
SECTION 1: General 1-3

1.3 Alternate Procedures
These guidelines do not preclude the use of alternate procedures for determining pipeline performance when subjected to seismic-induced load conditions. Alternative methods may be equally acceptable if based upon sound engineering and rational analysis satisfying the intent of provisions given in these guidelines. In most cases, alternative procedures are justified and feasible if they accommodate the use of site- or projectspecific conditions such as those listed below:

1. Site-specific seismic hazard definition based upon more detailed knowledge of the tectonic and geologic setting in which the pipeline is operated;

2. Site-specific assessment of geotechnical conditions, e.g., potential liquefaction areas;

3. Improved characterization of pipe-soil interaction based upon field testing of in situ soil conditions or laboratory tests representative of in situ conditions; and

4. Determination of alternative pipeline limit-state criteria based upon laboratory tests representative of the full range of expected pipeline cross-sections, materials, applied load conditions and deformation states.

Alternative analysis methods must be capable of correctly capturing non-linear soil behavior, the influence of large pipeline deformations on computed pipeline strains, and post-yield or post-buckling pipeline strength characteristics.

1.4 Required Information
The analysis procedures contained in these guidelines require the following engineering information.

1.4.1 Pipeline Information

1. Outside diameter and wall thickness of the pipe, elbows or induction bends;

2. Stress-strain relationships representative of the pipe material;

3. Toughness properties of the pipe, girth weld, seam weld, and weld heat-affected zone;

4. External coating specifications (as it relates to the pipe-soil friction interface) and thickness and type of insulation or shielding material (if any);

5. Maximum allowable operating pressure;

6. Design temperature differential between installed and operating conditions;

7. Pipeline alignment details (plan, profile, and location of fittings); and

8. For existing pipelines, reduced strain limits based upon type of girth-weld, expected material and weld defects in the girth weld and seam weld, and corrosion condition over a length of pipeline experiencing strains resulting from permanent ground displacement.

PR-268-9823
2129
SECTION 1: General 1-4

1.4.2 Soil Information

1. Depth of soil cover over a length of pipeline experiencing strains from permanent ground displacement;

2. Backfill specifications;

3. Depth to water table;

4. Water depth for offshore pipelines in the area of potential permanent ground displacement;

5. Soil strength parameters over a length of pipeline experiencing strains from permanent ground displacement and to a depth equal to the bottom of the pipe trench.

The needed parameters include:
a. Total unit weight
b. Internal friction angle
c. Cohesion

1.4.3 Ground Deformation Hazard Definition

1. Expected amount and dominant direction of permanent ground deformation (usually defined at the ground surface);

2. Length of pipeline exposed to permanent ground deformation;

3. Variation of permanent ground deformation with depth, if any, along the length of the pipeline under consideration;

4. Variation in the direction of permanent ground deformation; and

5. Uncertainty in the location of the fault with respect to the pipeline crossing.

1.5 Acronyms
APE annual probability of exceedance
API American Petroleum Institute
ASCE American Society of Civil Engineers
ASME American Society of Mechanical Engineers
ASTM American Society of Testing and Materials
CPT cone penetrometer test
EPS expanded polystyrene
HAZ heat affected zone
NPS nominal pipe size (NPS 12 corresponds to 12-inch and 323 mm pipe) SMYS specified minimum yield strength
PR-268-9823
2130
SECTION 1: General 1-5
SPT standard penetration test
TGD transient ground displacement
XPS extruded polystyrene

1.6 Nomenclature
To the extent possible, the nomenclature in these guidelines matches the nomenclature in equations from various reference sources. In several cases, this leads to duplication in nomenclature. Since many terms are well established in various technical disciplines, the potential for confusion in the use of unfamiliar terms is considered a more significant drawback than duplication.

Ac critical acceleration to trigger seismic slope instability AD average surface fault displacement in a vertical plane parallel to fault strike c cohesion representative of the soil backfill
c apparent wave propagation velocity for estimating strains from wave propagation
c′ effective cohesive strength of slope material for landslide assessment CB borehole diameter correction (assume 1.0 if unknown) CE hammer energy correction (CE = Eeff/60; assume 1.0 if unknown) CN overburden correction ( Pa ve /. σ ≤ 20 )
CQ correction factor for assessing liquefaction using CPT test data CR rod length correction (CR = 0.75 + 0.25(Lrod - 4); 0.75 ≤ CR ≤ 1.0) CRR cyclic resistance ratio for assessing liquefaction
CS sample liner correction (1.2 if no liner; if unknown, assume 1.0) CSR cyclic stress ratio for assessing liquefaction
ds average depth to top of seismogenic rupture zone in attenuation relationship D outside pipe diameter
D depth to basement rock
D′ ovalization parameter
D5015 mean soil particle diameter for saturated, liquefiable soils with (N1)60 values less than 15
Dbore SPT bore hole diameter
DF total fault displacement
PR-268-9823
2131
SECTION 1: General 1-6
DFS component of fault displacement in a vertical plane parallel to the fault strike (quantity provided by the Wells and Coppersmith empirical fault displacement relationship)
Dmin minimum pipe diameter from assumed ovaling
Dr relative density, %
Dw depth to water table
E random error term used in ground shaking attenuation relationships E modulus of elasticity
Eeff percentage of hammer energy delivered to sampling rod in SPT test f pipe coating factor for estimating interface friction angle from internal friction angle
fs sleeve friction measured in a CPT conducted in accordance with ASTM D-3441
F attenuation modification factor (0 for normal or strike-slip faulting, 1 for reverse or thrust faulting, if type of faulting unknown, use 0.5) FFH free face height for lateral spread displacement equation, m FFL distance from toe of free face, m
F15 percentage of material passing through a #200 sieve from a sample of liquefiable soil with an (N1)60 value less than 15
FA factor of safety against liquefaction, CRR/CSR
FC fines content of soil as measured by percentage of a sample passing a #200 sieve
g acceleration due to gravity
h thickness of potential sliding soil mass
H depth from the ground surface to the centerline of a buried pipeline H rupture depth, km, in attenuation relationship
HB depth to bottom of fault, km
HF horizontal component of fault displacement parallel to fault strike HS depth to top of seismogenic zone, km
HT depth to top of fault, km
I0.5 index parameter in CPT liquefaction assessment procedure computed for n = 0.5
I1 index parameter in CPT liquefaction assessment procedure computed for n = 1.0
PR-268-9823
2132
SECTION 1: General 1-7
In index parameter in CPT liquefaction assessment procedure computed for a specific value of n
Kc fines content correction factor for liquefaction assessment using CPT data Ko coefficient of earth pressure at rest
Kσ overburden correction factor to account for overburden stresses greater than 100 kPa
Lanchor length of burial sufficient to develop yield in the pipe under relative axial soil displacement
LD estimated slope displacement for assessing whether or not large landslide movements are possible
Lrod length of rod connected to SPT sampler, m
LSD mean lateral spread displacement
M earthquake moment magnitude
MD maximum surface fault displacement in a vertical plane parallel fault strike MSF magnitude scaling factor for computing cyclic resistance factor n factor in CPT liquefaction assessment related to fines content of soil N SPT blow counts collected in accordance with ASTM D-1586 (N1)60 standard penetration resistance normalized to an overburden pressure of 1 tsf (96 kPa) and a hammer energy efficiency ratio of 60% (N1)60FC (N1)60 corrected for fined content of the soil Nc soil bearing capacity factor for vertically downward loading in clay Nch soil bearing capacity factor for horizontal loading in clay Ncv soil bearing capacity factor for vertically upward loading in clay Nq soil bearing capacity factor for vertically downward loading in sand Nqh soil bearing capacity factor for horizontal loading in sand Nqv soil bearing capacity factor for vertically upward loading in sand Nγ soil bearing capacity factor for vertically downward loading in sand p internal pipeline pressure
py internal pipeline pressure that produces a hoop stress equal to σy Pa reference pressure (100 kPa ≈ 14.5 psi ≈ 1 tsf)
Pcr predicted collapse pressure from external pressure
Pe elastic buckling pressure
Pext net external pressure on pipe (total external pressure - internal pressure) Pfr average annual probability of fault displacement
PR-268-9823
2133
SECTION 1: General 1-8
Pg average annual probability associated with exceeding specified performance goal for the pipeline (e.g., 0.001 annual probability for loss of pressure integrity)
Pu maximum lateral soil load caused by pipe lateral movement relative to the surrounding soil
Py axial load in a pipeline corresponding to a uniform tensile stress equal to the SMYS
PGA peak ground acceleration at surface
qc tip resistance measured in a CPT conducted in accordance with ASTM D-3441
qc1N qc normalized to an overburden pressure of Pa
(qc1N)FC qc1N corrected to account for fines content
Qd maximum lateral soil load caused by pipe vertically downward movement relative to the surrounding soil
Qu maximum lateral soil load caused by pipe vertically upward movement relative to the surrounding soil
r amount of rattle-space for culvert mitigation concept rd cyclic stress reduction coefficient for assessing liquefaction using SPT data R earthquake source distance parameter (variously defined in equations) Rfp fault displacement reduction factor accounting for probability of fault displacement
RS closest distance to seismogenic rupture surface, km
R* effective distance from site to earthquake epicenter, km, used in lateral spread displacement equation
RLD fault rupture length at depth
sr fault slip rate
S percentage of surface slope for estimating lateral spread displacement SHR attenuation relationship surficial soil correction parameter (0 for soft rock, alluvium and firm soil, and 1 for hard rock)
SSR attenuation relationship surficial soil correction parameter (0 for hard rock, alluvium and firm soil, and 1 for soft rock)
SRL surface rupture length, km
t pipe wall thickness
tu maximum soil force on pipeline from relative axial movement T15 thickness of saturated, liquefiable, soils with (N1)60 values less than 15 TF component of fault displacement perpendicular to fault strike PR-268-9823
2134
SECTION 1: General 1-9
VF component of vertical fault displacement
Vmax maximum ground velocity from ground shaking
W free face ratio, expressed as percent, for estimating lateral spread displacement
W expected down-dip fault width, km
z depth below ground of SPT or CPT test measurement
ZT factor in attenuation relationship, 0 for interface event, 1 for intra-slab event in attenuation relationship
α fault dip angle, degrees
α coefficient for determining fines correction factor in liquefaction assessment α adhesion factor for estimating axial pipeline soil load αε ground strain coefficient for computing axial strains from wave propagation ακ ground strain coefficient for computing bending strains from wave propagation
β angle of regional stress azimuth for estimating fault displacement, degrees β coefficient for determining fines correction factor in liquefaction assessment δ soil interface friction angle
δo initial ovalization of pipe cross-section
∆ girth weld offset used to compute compression strain limit ∆d vertical downward relative pipeline displacement caused by liquefaction ∆p relative displacement between pipe and soil in lateral direction necessary to develop Pu
∆qd relative displacement between pipe and soil in vertically downward direction necessary to develop Qd
∆qu relative displacement between pipe and soil in vertically upward direction necessary to develop Qu
∆t relative displacement between pipe and soil in axial direction necessary to develop Tu
∆u vertical upward relative pipeline displacement caused by liquefaction εcl longitudinal compression strain limit for load-controlled conditions εco longitudinal compression strain limit for continued operation εcp longitudinal compression strain limit for pressure integrity εcr critical longitudinal compression strain
εcr-p critical longitudinal compression strain for conditions with external pressure PR-268-9823
2135
SECTION 1: General 1-10
εtl longitudinal tension strain limit for load-controlled conditions εg ground strain related to seismic wave propagation
εto longitudinal tension strain limit for continued operation εtp longitudinal tension strain limit for pressure integrity εv volumetric strain experienced by soil as the result of liquefaction φ soil internal friction angle
φ′ effective angle of internal friction for slope stability assessment Φ standard normal probability function
γ total unit weight of soil
γ effective (or submerged) unit weight of soil
γmax maximum soil shear strain
γw unit weight of water
κg ground curvature from seismic wave propagation
λ ratio of pore pressure to overburden stress
σ standard deviation
σt axial stress in pipeline from thermal differential
σy pipe material yield stress
σvo total soil overburden pressure
σ ′
vo effective soil overburden pressure
σAD standard deviation of average fault displacement
σMD standard deviation of maximum fault displacement
θ slope angle

1.7 References
For clarity, references have been largely omitted from the Procedures section of these guidelines. References to the basis documents used in these guidelines are noted in the Commentary section along with a complete reference listing.

PR-268-9823
2136
SECTION 1: General 1-11

1.8 Units
Seismic design and assessment of buried pipelines invariably crosses multiple disciplines with an equal variability in the degree to which English and metric units are adopted. As many of the equations in these guidelines are adopted from other sources, the original unit system has been maintained resulting in a mix of English and metric units within these guidelines. The majority of equations in these guidelines can be used with any consistent set of units. Exceptions are noted and typically relate to empirical relationships or equations fitted to published curves to facilitate calculations using spreadsheets or other computer-based applications.

PR-268-9823
2137
SECTION 1: General 1-12
DEFINE

PERFORMANCE
GOALS

IDENTIFY CRITICAL

PIPELINE LOCATIONS

TO BE ASSESSED

COLLECT NECESSARY

EARTHQUAKE, SOIL,

AND PIPELINE DATA

QUANTIFY SEISMIC

HAZARDS

ANALYZE PIPELINE

RESPONSE

RESPONSE
ACCEPTABLE?

DEVELOP

MITIGATION

MEASURES
END
YES

MATERIALS AND NO

WELDING
TECHNOLOGY***
***SPECIAL TECHNICAL

EXPERTISE TYPICALLY

REQUIRED
GEOLOGY***
SEISMOLOGY***
SOIL MECHANICS***
SAFETY

OPERABILITY

ECONOMICS

ENVIRONMENTAL IMPACT

REGULATORY

CORPORATE POLICY

CONTRUCTION
TECHNOLOGIES***
Figure 1.1: Seismic Design and Assessment Process for Assessing Seismic Hazards for Buried Natural Gas and Liquid Hydrocarbon Pipelines PR-268-9823
2138
SECTION 2: Quantifying Seismic Hazards 2-1

2.0 QUANTIFYING SEISMIC HAZARDS
Credible seismic hazards for buried steel pressure pipelines are generally limited to those that produce permanent deformation of the ground along the pipeline alignment or large transient ground displacements that result in significant permanent deformation following an earthquake. These hazards primarily include surface faulting, liquefaction, liquefaction-induced lateral spread movement, landslides, and near-surface settlement.

Seismic wave propagation is also a consideration for buried pipelines, but it generally does not have a serious effect on buried welded steel pipelines. Transient ground deformation generally has not been a problem, except in a few special situations that are discussed in Section 2.7.

This section presents the methods and approaches commonly used to define seismic hazards for buried pipelines. Detailed coverage of the theory and development of procedures for collecting data and quantifying seismic hazards is beyond the scope of these guidelines. The information provided relies heavily on project experience and recent advances identifying and characterizing seismic hazards. It is recommended that the definition of seismic hazards be performed by individuals experienced with the methods described in this section and possessing the requisite background in seismology, geology, and soil mechanics.

The discussion of seismic hazards is organized by category and is presented in accordance with the development of parameters to define the hazard. The focus is on direct seismic hazards as opposed to potential indirect hazards from such causes as debris impact (e.g., rockfall impact) or collateral damage resulting from earthquake damage to adjacent pipelines in a common right-of-way or utility corridor.

The definition of potential seismic hazards is often developed during a separate geotechnical or geologic investigation. In lieu of such investigations, the procedures provided in this section can be followed to provide an estimate of the amount and likelihood of ground movement. These procedures are best used as a tool to determine if a pipeline is at significant risk from seismic hazards. The procedures also provide a means to identify the need for, and scope of, a more rigorous geologic or geotechnical investigation.

2.1 Surface Faulting
Surface fault rupture is an important consideration for buried pipelines, because pipelines crossing fault zones must deform longitudinally and in flexure to accommodate ground displacement. If a pipeline crosses an active fault, it is necessary to delineate its location, orientation, width of fault rupture zone, and the amount and direction of potential fault displacement.

The term “active fault” implies a judgment on the part of a geologist that the fault could experience surface rupture within the time frame established by the performance goals for PR-268-9823
2139
SECTION 2: Quantifying Seismic Hazards 2-2
the pipeline (e.g., an average annual probability of exceedance less than 1x10-3).

Generally, a fault is considered active if it can be demonstrated to have displaced the ground surface during the Holocene epoch (i.e., within the past 11,000 years).

For some faults, a characteristic earthquake and associated recurrence interval may have been established through field studies and historical seismicity. For such cases, the expected fault displacement may be estimated as the cumulative slip between characteristic earthquakes (i.e., the slip rate per year times the recurrence interval).

Otherwise, it is more meaningful to rely on empirical relationships such as those described in Section 2.1.1.

Some faults experience aseismic creep in addition to surface rupture produced by an earthquake. For pipeline design, the total fault movement, including creep, is important since both lead to deformation of the pipeline. The effects of creep need to be considered separately because the slow rate of displacement is consistent with drained soil loading conditions while fault displacement occurs rapidly under undrained conditions. This impacts the determination of soil loads on the pipeline as discussed in Section 4.

It is recommended that fault displacement estimates for the design or assessment of pipeline response be approached as a three-stage process:

1. Estimate the expected fault displacement of an active fault;

2. Adjust the estimate of expected fault displacement based upon the consequences of pipeline damage and pipeline performance objectives to obtain the design fault displacement; and

3. Convert the design fault displacement to orthogonal displacement components relative to the fault (i.e., vertical component, horizontal component parallel to the fault strike, and horizontal displacement perpendicular to the fault strike).

2.1.1 Empirical Methods for Estimating Fault Displacements The most common method for estimating surface fault displacements relies on observations of fault displacements during past earthquakes. Wells and Coppersmith (1994) developed a series of empirical relationships that relate moment magnitude, M, surface rupture length, subsurface rupture length, down-dip rupture width, rupture area, and maximum and average displacement per event. The most commonly used Wells and Coppersmith relationships that are used when other information is lacking provide maximum and average fault displacements (MD and AD) as function of fault surface rupture length or moment magnitude.

Estimating MD and AD as a function of surface rupture length is generally preferred because the determination of fault displacement is based directly upon the geologic attributes of the fault. Using earthquake moment magnitude to estimate MD and AD is more indirect because it is based upon a two-stage correlation of moment magnitude to fault parameters (typically either surface rupture length or seismogenic rupture area) and PR-268-9823
2140
SECTION 2: Quantifying Seismic Hazards 2-3
fault displacement to magnitude. However, relating fault displacement to earthquake magnitude is preferable for situations where earthquake magnitude has been previously established by other geologic investigations.

For estimating fault displacement (MD or AD) as a function of surface rupture length (SRL), the following regression formulae from Wells and Coppersmith (1994) apply:
log( ) 1.38 1.02 log( ) MD =− + ⋅ SRL [log (σMD) = 0.41] (2-1a) log( ) 1.43 0.88 log( ) AD =− + ⋅ SRL [log (σAD) = 0.31] (2-1b) where:

AD = Average surface fault displacement, m
MD = Maximum surface fault displacement, m
SRL = Surface rupture length, km. The surface rupture length is the maximum length of the fault segment crossing the pipeline that, by reasonable and qualified judgment, can be expected to rupture during an earthquake. σMD = Standard deviation of maximum displacement regression σAD = Standard deviation of average displacement regression The data sets for the two regression equations listed above include displacement data for all types of slip, i.e., strike-slip, normal and reverse faulting. The displacements estimated by the relationships of Wells and Coppersmith (1994) is the vector sum of the horizontal and vertical slip components along the fault strike. The implications of this definition on determining three-dimensional components of fault displacement are discussed further in Section 2.1.2.

Wells and Coppersmith (1994) also provide relationships for estimating maximum and average fault displacements as a function of moment magnitude (M). These relationships are provided below.
log( ) 5.46 0.82 MD =− + ⋅M [log (σMD) = 0.42] (2-2a)
log( ) 4.80 0.69 AD =− + ⋅ M [log (σAD) = 0.36] (2-2b)

2.1.2 Orthogonal Displacement Components
As a final step in defining fault displacement, the design fault displacement determined in Section 2.1.2 must be transformed into orthogonal components of fault displacement.

This is necessary only if fault displacement is estimated using equations (2-1) or (2-2), since these expressions only provide the net displacement in a vertical plane perpendicular to the strike of the fault. The transformation to orthogonal displacement components relative to the fault strike is illustrated in Figure 2.1. The following equations provide the transformation:

PR-268-9823
2141
SECTION 2: Quantifying Seismic Hazards 2-4
Total fault displacement, DF:
( )( ) 2 22 sin cos cos
FS
F
D D
α αβ
=
+
(2-3a)
Vertical displacement, VF:

VD FF = sinα (2-3b)
Horizontal displacement parallel to the fault strike, HF:
cos cos HD FF = α β (2-3c)
Transverse horizontal displacement perpendicular to the fault strike, TF:

TD FF = cos sin α β (2-3d)
Apparent dip angle, α:
tan tan sin α = δ β (2-3e)
where:

DFS = fault displacement in a vertical plane parallel to the fault strike δ = fault dip angle corresponding to “average” dip angle inherent in equations (2-1) and (2-2) (see Commentary). Lacking other information, the dip angle can be taken as 75° for strike slip faults, 60° for normal slip faults, and 45° for reverse or thrust faults.
β = horizontal angle between fault strike and regional stress azimuth. For faults that are predominantly strike slip, this angle is 0°, and for faults that are predominantly reverse or normal faults, this angle will be 90°. Other faults will have some combination of strike and reverse/normal slip, with an angle, β, between 0° and 90°.

2.1.3 Design Fault Displacements
Determination of a design fault displacement is based upon estimates of expected average and maximum fault displacements described in Section 2.1.1. Design fault displacements are determined based upon consideration of the performance requirements for the pipeline and the consequences for loss of pipeline pressure integrity. Two modifications are recommended for determining design fault displacements, one to account for the consequences of loss of pipeline pressure integrity and another to account for faults with a low likelihood of occurrence relative to the pipeline performance criteria.

PR-268-9823
2142
SECTION 2: Quantifying Seismic Hazards 2-5

2.1.3.1 Accounting for Consequences of Loss of Pressure Integrity It is preferable that design fault displacements be based upon the uncertainty in fault displacement and site-specific consideration of the potential risk (likelihood of a particular type of damage and consequences) associated with earthquake-related pipeline damage. In lieu of such considerations, the following design factors are recommended to account for the consequences of loss of pipeline pressure integrity and the expected recurrence interval for fault displacement:

1. Consider the mean maximum fault displacement for natural gas pipelines and flammable or explosive petroleum products pipelines in Location Class 4 areas (ASME B31.8).

2. Consider two-thirds of the mean maximum fault displacement for natural gas pipelines and flammable or explosive petroleum products pipelines in Location Class 3 areas (ASME B31.8).

3. Consider two-thirds of the mean maximum fault displacement for petroleum products pipelines located in environmentally sensitive areas.

4. Other natural gas or liquid hydrocarbon pipelines in other areas than above should be designed for the mean average fault displacement.

Prior to use, the above guidelines should be carefully reviewed and evaluated by the pipeline owner or its representatives for consistency with project objectives and regulatory requirements.

2.1.3.2 Accounting for Expected Occurrence of Fault Displacement The time frame for defining a fault to be active is typically an order of magnitude greater than the typical performance goals for a pipeline (e.g., 11,000 years for the Holocene epoch versus a pipeline performance goal of a 1,000-year average return period for experiencing loss of pressure integrity). To account for this, it is recommended that the expected fault displacements be reduced when the estimated likelihood of fault displacement is less than the acceptable probability of not achieving pipeline performance goals. A simple correction factor based upon the ratio of the probability of fault displacement and pipeline performance goal is recommended as defined by equation (2-4):
2 fr
fp
g
P
R
P = (2-4)
where:

Rfp = reduction factor accounting for probability of fault displacement Pfr = average annual probability of fault displacement
Pg = average annual probability associated specified performance goal (e.g., 1x10-3 mean annual probability for loss of pressure integrity) PR-268-9823
2143
SECTION 2: Quantifying Seismic Hazards 2-6
If fault displacements are determined by other means than equation (2-1) or (2-2), appropriate steps should be taken to determine orthogonal components in the vertical, horizontal and transverse directions.

2.2 Peak Ground Acceleration
Estimates of peak ground acceleration (PGA) are required for quantifying permanent ground displacement hazards related to liquefaction, lateral spread movement, and landslide movement. Ground shaking is estimated using empirical attenuation equations representing the decrease in ground motion with increasing distance from the fault rupture plane. There are many attenuation relationships available, and many are developed to represent specific types of earthquakes and specific geologic and tectonic environments.

Peak ground acceleration can be obtained from the results of probabilistic hazard analyses or deterministically based upon a postulated earthquake on a known fault.

Probabilistic hazard analyses typically form the basis for building code earthquake hazard maps. The details of how probabilistic hazard analyses are performed is a specialized field and is beyond the scope of these guidelines. However, a summary of the general procedures is provided in Appendix C to familiarize the reader with the general concepts.

Deterministic estimates of peak ground acceleration are relatively straightforward to apply provided an attenuation relationship is selected that is appropriate for the earthquake mechanism and tectonic and geologic characteristics of the area.

A deterministic approach to estimating peak ground acceleration is required for assessing other hazards related to liquefaction and some approaches for assessing slope instability.

Where probabilistic hazard analyses are available, it may be possible to infer a magnitude and acceleration value from the details of the calculation of the probabilistic hazard.

Selecting a suitable combination of magnitude and acceleration from a probabilistic hazard estimate should be performed by individuals familiar with the generation of the probabilistic hazard and the seismicity of the region.

The two relationships presented in Sections 2.2.1 and 2.2.2 are provided to illustrate some of the general characteristics and differences in attenuation relationships. These relationships were developed for near-source earthquakes (western United States) and subduction zone (Cascadia region) earthquakes. It is strongly recommended that users seek out specific attenuation relationships applicable for the region in which the pipeline is located.

Both attenuation relationships contain an error term, E, that represents a factor that can be used in conjunction with a standard normal probability table to estimate the probability of exceeding the mean of ln(PGA) represented by the equation. For example, for a mean estimate, the standard normal probability, Φ(x) is 0.5 (50%) and the value of x is 0. If the probability of exceedance is desired to be less than 16%, Φ(x) is 0.84, x is 1.0 and the value of ln(PGA) is increased by E which is equivalent to multiplying the PGA by e E
.

PR-268-9823
2144
SECTION 2: Quantifying Seismic Hazards 2-7
Other percentile levels may be established by selecting values of Φ(x) consistent with the percentile associated with the desired probability of non-exceedance.

The attenuation relationships also illustrate the potential variability in the amount and types of information needed for their application. This is particularly true for definitions of source-to-site distance. The near source earthquake relationship requires more descriptive information on the fault plane and the portions of the fault capable of generating earthquake ground motions (the seismogenic part of the fault) in order to compute this distance. There is also a difference in the number of parameters used to characterize the site.

2.2.1 Near Source Earthquakes in the Western United States [] [
[]
2 20.647
1 ln( ) 3.512 0.904 1.328ln 0.149

1.125 0.112ln( ) 0.0957 0.440 0.171ln( )

0.405 0.222ln( ) ( ) ( )
S
M
SS
S HR A
PGA M R e
R MF R S] SR
R S fD E x −
=− + − + ⎡⎤ ⎣⎦
+− − + −
+ − + +Φ
(2-5)
where:

PGA = peak ground acceleration, g
M = moment magnitude
F = 0 for normal and strike-slip faulting
= 1 for reverse or thrust faulting
= 0.5 if type of faulting unknown
RS = closest distance to seismogenic rupture, km
= 22 R S + d
R = horizontal projection of closest site to seismogenic rupture distance, km dS = average depth to the top of the seismogenic rupture zone, km = [] 1 sin( ) 2
H HW H BT ++ ≥ α S
HB = depth to bottom of fault, km
HS = depth to top of seismogenic part of the crust, km
HT depth to top of fault, km
W = expected down-dip fault width, km
=

1.01 0.32 10−+ M
α = fault dip angle
PR-268-9823
2145
SECTION 2: Quantifying Seismic Hazards 2-8
() Af D = 0 if D ≥ 1 km, otherwise
= {[0.405 0.22ln( ) 0.440 0.171ln( ) (1 )(1 ) − − − −− RSS ] [ RS D S ] SR} SR D = depth to basement rock, km
E = random error term = 0.889 − 0.691M
Φ = standard normal probability function
x = cumulative probability of actual PGA being less than or equal to PGA from the attenuation relationship
SSR = 0 for hard rock, alluvium, and firm soil
= 1 for soft rock
SHR = 1 for hard rock
= 0 for soft rock, alluvium, and firm soil
For generic rock sites, SSR = 1, SHR = 0, and D = 1 km.

For generic soil sites, SSR = 0, SHR = 0, and D = 5 km.

2.2.2 Cascadia Earthquakes
ln( ) . . . . ln .
. . ()
. PGA M R e
H ZE x
M
T
=+ − +
+++ −
0 2418 1414 2 2 552 17818
0 00607 0 3846
0 554
1 c h
Φ
for rock (2-6a)
ln( ) . . . ln .
. . ()
. PGA M R e
H ZE x
M
T
=− + − +
+ ++ −
0 6687 1438 2 329 1097
0 00648 0 3643
0 617
1 c h
Φ
for soil (2-6b)
where:

H = rupture depth, km
R = closest distance to fault rupture, km
ZT = 0 for interface event, 1 for intra-slab event
E = 1.45 0.1 − M but not greater than 0.65

2.3 Liquefaction
Liquefaction hazards to pipelines include flotation, sinking, and general ground settlement from dissipation of excess pore water pressure. Pipeline flotation or sinking requires the pipe to be located below the ground water table within a zone of liquefiable soil. Ground settlement occurs when a liquefiable soil layer beneath a layer of competent PR-268-9823
2146
SECTION 2: Quantifying Seismic Hazards 2-9
soil densifies. If the pipeline is located in the layer of competent soil near the surface, it will be subjected to displacement associated with subsidence of the ground.

Quantification of liquefaction hazards requires site-specific information on the subsurface soil properties and peak ground acceleration. Soils information can most commonly be obtained through subsurface investigations using the standard penetration test (SPT) and cone penetration test (CPT). These investigations should be performed by personnel experienced in the interpretation of soil boring logs and liquefaction assessment.

Parameters used in the assessment of liquefaction potential are summarized in Table 2.1.

Liquefaction also may cause lateral spreads or flow slides. These hazards are discussed in Sections 2.4 and 2.5.

2.3.1 Assessing Liquefaction Potential
Liquefaction is a phenomenon primarily limited to loose saturated sand deposits.

Normalized SPT blow counts (defined in Section 2.3.1.1) greater than 30 can be taken as an indication that the soil densities are sufficiently high to prevent liquefaction. In addition, work primarily from Chinese investigators, suggests that some saturated clayey soils may also be susceptible to significant loss of strength similar to liquefaction. The criteria for identifying potentially sensitive clayey soils are as follows:

1. Percent of grain size finer than 0.005 mm is less than 15%.

2. Liquid limit is less than 35.

3. Water content is greater than 90% of the liquid limit.

The information in Table 2.1 is required for performing a liquefaction assessment for a particular subsurface location where SPT or CPT data are collected. In addition, it is necessary to know if sample liners were used and the energy efficiency of the hammer used during collection of SPT data.

Both the SPT and CPT methods for assessing liquefaction potential compare the resistance to liquefaction, measured by the cyclic resistance ratio (CRR), to the cyclic stress ratio (CSR). Liquefaction is likely when CSR is greater than CRR. The cyclic stress ratio is primarily a function of the peak ground acceleration (PGA) and the in situ soil stress. The equation for computing CSR is the same for evaluating liquefaction potential using SPT or CPT data.

CSR PGA r vo
vo d =⋅ ⋅ ⋅ F
H
G I
K
J 065 . '
σ
σ
(2-7)
where:

PGA = peak ground acceleration, g
PR-268-9823
2147
SECTION 2: Quantifying Seismic Hazards 2-10
rd =

1.5 1.5 2
1 0.4113 0.04052 0.001753
1 0.4177 0.05729 0.006205 0.001210
zz z
zz z
− ++
− +− + z
z = depth of interest (z must be in meters for computing rd) σ vo = total overburden soil pressure at test location = γz σ vo
′ = effective soil overburden pressure = γz - γw(z - Dw) Dw = depth to water table
γ = unit weight of soil
γw = unit weight of water

2.3.1.1 Assessment of Liquefaction Using SPT Data
The CRR is computed as follows:
for x ≥ 3,
23 4 2
13 2 4 3
(4.8(10) 4.721(10) 6.136(10) 1.673(10) )
1 1.248(10) 9.578(10) 3.285(10) 3.714(10)
5 3
6 4
MSF K x x x CRR
x xx
σ
−− −
−− −
⋅ −+ − = −+ − + x
−
− (2-8a)
for x < 3,
CRR MSF K =⋅ ⋅ 0.05 σ (2-8b)
where:
x = (N1)60FC, the normalized SPT data corrected for fines content MSF = magnitude scaling factor for approximately a 32% chance of liquefaction
= 103.74/M4.33
Kσ = overburden pressure correction factor for σ ′
vo > Pa
2 1 for

0.897 0.411 0.514 for 10

0.6 for 10
vo a
a vo
vo vo
a a
vo a
P
PP
P P
P
σ
σ
σ σ
σ
a
=≤ ′
=+ − < < ′ ⎛⎞ ′ ⎛⎞ ′ ⎜⎟ ⎜⎟ ⎝⎠ ⎝⎠
=≥ ′
Pa = reference pressure (100 kPa ≈ 14.5 psi ≈ 1 tsf)
PR-268-9823
2148
SECTION 2: Quantifying Seismic Hazards 2-11
The expression for CRR is based largely upon past earthquake observations for earthquakes with magnitudes of approximately 7.5. This expression is depicted graphically in Figure 2.2 as a function of (N1)60FC for MSF and Kσ equal to 1.0.

Two steps are used to correct the measured SPT data. First, the measured SPT value, N, is corrected using equation (2-9) to account for specific test procedures and normalized to an effective overburden pressure equal to Pa. The second correction accounts for the greater liquefaction resistance of soils with larger quantities of fine-grained material. In this case, the fines content, FC, is determined as the percentage of material passing through a #200 sieve (0.074 mm opening). The normalized SPT value, (N1)60, is computed as follows:
() 1 N NC C C C 60 =⋅ ⋅ ⋅ ⋅ ⋅ NEBR CS (2-9)
where:
() 1 60 N = SPT value normalized to Pa and 60% hammer energy efficiency N = measured SPT
CN = overburden correction = Pa vo /. σ ′ ≤ 20
CE = hammer energy correction
= Eeff/60 (assume 1.0 if unknown)
Eeff = hammer energy efficiency, percentage
CB = borehole diameter correction (assume 1.0 if unknown) CR = rod length correction
= 0.75 + 0.04( 4) Lrod − (0.75 ≤ CR ≤ 1.0)
Lrod = length of rod in SPT test (Lrod must be in meters for computing CR) CS = sample liner correction (1.2 if no liner; assume 1.0 if unknown) The following expression corrects the (N1)60 value for the effect of fines content, FC:
1 60 1 60 () () NN FC =+ α β (2-10)
where α and β are defined as follows:

Fines Content, FC α β
5% or less 0 1.0
Greater than 5% to less than 35% 2
190 1.76
FC e − 1.5

0.99 1000
FC
+
35% or greater 5 1.2
PR-268-9823
2149
SECTION 2: Quantifying Seismic Hazards 2-12

2.3.1.2 Assessment of Liquefaction Using CPT Data
Data collected from cone penetrometer tests (CPT) need to be corrected for normalized test conditions and the fines content of the soil. Since no samples are collected in CPTs, the fines content correction relies on inferred conditions from the ratio of measured tip bearing, qc, and friction value, fs. The equation for computing CRR from CPT data is given below:

For (qc1N)FC < 50
CRR MSF qc N FC = ⋅ + L
N
M O
Q
P
0833
1000
0 05 1 .() . (2-11a)
For 50 ≤ (qc1N)FC < 160
CRR MSF qc N FC = F
H
G I
K
J +
L
N
M
M
O
Q
P
P
93 1000
0 08 1
3 () . (2-11b)
where:
(qc1N)FC = tip resistance normalized to Pa and corrected for fines content The measured tip resistance is normalized to an effective overburden pressure equal to Pa and corrected for fines content. These corrections are dependent upon the value of In, a parameter related to the granular nature of the soil.
2 (3.47 log ) (1.22 log ) nn IQ = − ++ 2 F (2-12)
( ) 100
100
n
c vo
n
vo
q Q σ
σ
− ⎛⎞ ⎜⎟ ′ ⎝⎠
= (2-13)
100% s
c vo
f
q
F
σ
⎡⎤
⎢⎥ ⎣⎦ − = (2-14)
where:
qc = CPT tip bearing pressure
fs = CPT side friction pressure
PR-268-9823
2150
SECTION 2: Quantifying Seismic Hazards 2-13
The value of n used in equations (2-13) and (2-14) is, in turn, dependent upon the resulting value of In. Determination of n requires computing In for values of n equal to

0.5 (I0.5) and 0.1 (I1). The value of n is then determined as follows:
n = 1.0 if I1 greater than 2.6 (clayey)
n = 0.5 if I1 and I0.5 less than 2.6 (granular)
n = 0.7 if I1 less than 2.6 and I0.5 greater than 2.6 (silty) The normalized CPT tip resistance, (qc1N)FC, is then computed as shown below:
1 () c
c N FC c Q
a
q q KC P
⎛⎞ = ⎜
⎝⎠⎟ (2-15)
where:

Kc = 1.0 for In ≤ 1.64
= −+ 1788 3375 2163 5581 0403 − + − 23 .. . . . III nnn 4 In for In > 1.64
CQ =
n
a
vo
P
σ
⎛⎞ ⎜⎟ ′ ⎝⎠

2.3.2 Displacements Associated with Relative Buoyancy
Pipeline flotation is a displacement-controlled phenomenon for onshore pipelines. The amount of displacement is limited by the distance between the bottom of the pipe and the water table or the depth of soil cover plus about 1/3 the pipe diameter for cases where the water table is at the ground surface. This is an upper limit assuming that the liquefied state is maintained for a long time. Otherwise, the displacement is further limited by the velocity at which the pipeline moves through the liquefied material and the “drag” as the pipe moves through the liquefied soil. Offshore pipelines or pipelines crossing bodies of water will tend to float if they are not at least neutrally buoyant with respect to the unit weight of the liquefied soil.

Pipelines that are negatively buoyant with respect to the unit weight of the liquefied soil are subject to sinking if they are located within a liquefiable soil deposit. The amount of sinking is limited by the distance between the bottom of the pipe and the depth to nonliquefiable soil. This is an upper limit assuming that the liquefied state is maintained for a long time. Otherwise, the displacement is further limited by the velocity at which the pipeline moves through the liquefied material similar to flotation.

The length of pipeline exposed to liquefaction is dependent upon the subsurface extent of liquefiable soil and the alignment of the pipeline through the zone of liquefaction.

Determining the length of exposed pipeline can be made through subsurface investigations at regular intervals along the pipeline. Alternatively, the extent of PR-268-9823
2151
SECTION 2: Quantifying Seismic Hazards 2-14
exposure may be estimated based upon obvious changes in surface geology (e.g., a transition from deltaic deposits to rock or glacial till).

2.3.3 Displacements Related to Liquefaction-Induced Settlement Liquefaction-induced settlements are credible hazards to pipelines located within liquefiable layers of soil. For the more typical onshore case where the pipeline is located above the water table, liquefaction-induced settlement is a credible hazard if the liquefiable layer is close enough to the surface to produce settlement in near-surface soil layers. One means to determine this possibility is to compare the thickness of the liquefied layer and the depth to the liquefied layer. The screening curves in Figure 2.3 are based upon observations of where surface settlements have been sufficient to cause structural damage in past earthquakes. Since buildings are generally more susceptible to damage from settlement, the use of Figure 2.3 can be used as a conservative screening guideline for buried pipelines.

The amount of liquefaction-induced settlement can be estimated for each sandy soil layer using Figure 2.4. For each soil layer, the amount of volumetric strain is taken from Figure 2.4 based upon the factor of safety against liquefaction and the relative density of the soil or the corresponding corrected blowcount, (N1)60FC. The total amount of settlement is estimated by multiplying the volumetric strain by the thickness of each soil layer below the invert of the pipeline. Utilization of Figure 2.4 is made by computing the factor of safety against liquefaction, FA , as given by the ratio of CRR to CSR.

2.4 Lateral Spread Movement
The most appropriate method to estimate lateral spread movement is to rely on computerbased approaches that account for the actual strength properties of the liquefied soil.

These programs typically are limited to two-dimensional representations of the soil.

They have the advantage of being able to quantify variation in the magnitude and direction of soil displacements. Examples of two commonly used programs are SOILSTRESS and DESRA, both developed at the University of British Columbia. In addition, there are many research-oriented programs and special purpose programs used within the geotechnical community.

In lieu of computer analyses, an estimate of lateral spread movement can be obtained using empirical relationships developed from lateral spread and soil data collected in past earthquakes in the United States and Japan (Youd et al., 2002), as shown below.

For sloping ground conditions (Figure 2.5):
*
15 15 15 0.1)
( ) 16.213 1.532 1.406

0.012 0.338 0.540

3.413 (100 ) 0.795 ( 50
Log LSD M LogR
R LogS LogT
Log F Log D +
=− + −
−+ +
+ −−
(2-16a)
PR-268-9823
2152
SECTION 2: Quantifying Seismic Hazards 2-15
For free face conditions (Figure 2.6):
*
15 15 15 0.1)
( ) 16.713 1.532 1.406

0.012 0.592 0.540

3.413 (100 ) 0.795 ( 50
Log LSD M LogR
R LogW LogT
Log F Log D +
=− + −
−+ +
+ −−
(2-16b)
where:

LSD = horizontal lateral spread movement (m)
M = earthquake moment magnitude [6.0 < M < 8.0]
R = epicentral distance, km
R* = R + Ro
Ro =

0.89 5.64 10 M −
S = ground slope (%) [0.1 < S < 6.0]
W = free face ratio (%) = 100(FFH/FFL) [1 < W < 20]
FFH = height of free face
FFL = distance from base of free face
T15 = thickness (m) of saturated cohesionless soils with (N1)60 < 15 [1 < T15 < 15]
(N1)60 = standard penetration blowcount normalized to Pa and 60% driving efficiency
F15 = average fines content (%) in T15 [0 < F15 < 50]
D5015 = average median particle size (mm) in T15 [0 < D5015 < 50] The limits associated with the above variables represent the range of values in the Bartlett and Youd data set. The simplified approach may result in significant error if site-specific conditions deviate significantly from these limits. Note that the range of (N1)60 is limited to less than 15. Available data indicate that lateral spread displacement does not occur in sediments with (N1)60 greater than 15.

Variation in ground displacement away from a free face can be estimated by computing the displacement using the free face equation (2-16b) with increasing values of FFL (i.e., decreasing free face ratio, W). This is relatively straightforward because there is only one term dependent upon the free face ratio.

The foregoing methods for estimating permanent ground movement relate to maximum displacements in a lateral spread. For the evaluation of pipeline behavior, it is necessary to estimate a transverse distribution of the pattern of ground deformation within the soil mass that moves across or along a buried pipeline. Considerations for defining the pattern of ground displacement are provided in Section 4.5.

PR-268-9823
2153
SECTION 2: Quantifying Seismic Hazards 2-16

2.5 Landslides
Strong ground shaking during earthquakes may trigger landslides in a variety of geologic and topographic settings as depicted in Figure 2.7. The potential threat to pipeline performance is a function of the following:

1. The amount of landslide displacement;

2. The depth of the landslide relative to the depth of the pipeline;

3. The type of ground displacement associated with the landslide movement (i.e., blocktype (coherent) movement or disrupted movement); and

4. The direction of landslide movement relative to the pipeline.

An approximate estimate of the magnitude of triggered landslides can be made using the approach sometimes adopted for regional identification of the extent of landslide hazards.

The basis of the recommended approach is the sliding block analysis approach developed by Newmark (1965). The approach is applicable to both disruptive and coherent slides.

Disruptive slides typically are shallow and are broken during movement into small blocks, rock fragments, or soil particles. Coherent slides typically are deeper seated movements of intact blocks of soil or rock. For both types of slides, a hazard to the pipeline exists only if the pipeline is above the interface of the potential slide zone.

The recommended approach for identifying the potential for triggered landslides requires an estimate of the critical acceleration, Ac, that will trigger down-slope displacement for the type of slide and material. The critical acceleration can be computed using the following equation developed for the assumption of a constant slope and material strength:

A c
h c = ′⋅ +− ′ − γ
( ) tan cos sin 1 λφθ θ (2-17)
where:
h = thickness of the potential landslide
θ = slope steepness angle
c′ = effective cohesion of the slope material under dynamic conditions φ′ = effective friction angle of the slope material under dynamic conditions λ = ratio of the pore pressure to overburden stress = γw(h - Dw)/γh ≤ 0.6 Dw = depth to water table
γ = unit weight of soil
γw = unit weight of water
PR-268-9823
2154
SECTION 2: Quantifying Seismic Hazards 2-17
Equation (2-17) can be used to plot values of Ac as a function of slope angle for various values of c′ , φ′ , h, γ, and Dw.

An example of a plot of Ac versus slope angle is illustrated in Figure 2.8. The curves in Figure 2.8 are based upon common slope material properties listed in Table 2.2 and assume h = 10 ft and γ = 100 pcf. The two groups of curves in Figure 2.8 correspond to completely saturated conditions (i.e., the water table is at the ground surface) or unsaturated conditions (i.e. the water table depth is greater than h).

The computation of slope displacement generally requires the double integration of earthquake ground accelerations larger than the critical acceleration, Ac, that cause the forces driving slope displacement to exceed the resisting forces. This approach implies that one or more ground motion time histories representative of the site under consideration are available. As a simpler alternative, empirical relationships that envelop the displacements computed for a range of time histories are available for estimating slope displacement as a function of peak ground acceleration, PGA.

The amount of landslide movement, LD, can be estimated using the following bounding relationships.

For very firm ground or rock conditions
2 1.7 (1 ) c
D
c

A PGA L PGA
PGA A
⎛⎞ = ⋅ ⋅− ⋅⎜
⎝⎠⎟ (2-17a)
For soil conditions
2 3.0 (1 ) c
D
c

A PGA L PGA
PGA A
⎛⎞ = ⋅ ⋅− ⋅⎜⎟ ⎝⎠
LD in inches for Group C (2-17b)
where:

Ac = critical acceleration (% of gravity) from Figure 2.8 Triggered landslides should be considered possible if LD is greater than 1 inch (25 mm) for disruptive landslides or 4 inches (100 mm) for coherent landslides. If the possibility of a triggered landslide exists, site-specific investigations should be performed to quantify the magnitude, extent, and likelihood of landslide movements.

The evaluation of pipeline behavior in landslide areas requires the definition of a displacement pattern for the displaced soil mass as shown in Section 4.5.

PR-268-9823
2155
SECTION 2: Quantifying Seismic Hazards 2-18

2.6 Seismic Wave Propagation
Body waves, including compression waves and shear waves, propagate radially from the source of earthquake energy release (hypocenter) into the surrounding rock and soil medium. Compression waves cause axial compression and tension strains in the ground in a radial direction away from the hypocenter. Shear waves cause shear strains in the ground perpendicular to these radial lines. When the compression waves and shear waves are reflected by interaction with the ground surface, surface waves (Love waves and Rayleigh waves) are generated. Except at very large distances from the epicenter, the magnitudes of surface waves are much less than body waves.

A pipeline buried in soil that is subject to the passage of these ground waves will incur longitudinal and bending strains as it conforms to the associated ground strains. In most cases, these strains are relatively small, and welded pipelines typically do not incur damage. Propagating seismic waves also give rise to hoop membrane strains and shearing strains in buried pipelines, but these strains are even smaller and may be neglected.

A relatively simple method based upon an approach developed by Newmark (1965) can be used to obtain an upper-bound estimate of ground strain due to a propagating seismic wave with a constant shape (sine wave). The maximum ground strain, εg, is given by:
max g
V
εc
ε
α = (2-19)
where:

Vmax = Maximum horizontal ground velocity in the direction of wave propagation c = Apparent propagation speed of seismic wave (i.e., the component of wave speed parallel to the pipeline under consideration)
αε = Ground strain coefficient corresponding to the most critical angle of incidence and type of seismic wave
= 1.0 for compression and Rayleigh waves (ASCE, 1984)
= 2.0 for shear waves (ASCE, 1984)
The apparent propagation velocity is the inverse of the “slowness,” a term used in seismology. The slowness is a function of the source-to-site geometry and the wave propagation velocity along the entire path of propagation. Tables of earthquake wave travel times will exist wherever detailed earthquake location work has been performed.

An apparent propagation velocity of 2 km/sec can generally be taken as a lower bound estimate to provide a maximum estimate of ground strain.

If there is no slippage of the pipeline relative to the surrounding soil, then the maximum axial strain in the pipeline is equal to maximum ground strain. The ground strain coefficient corrects this ground strain for the effects of pipe orientation relative to the PR-268-9823
2156
SECTION 2: Quantifying Seismic Hazards 2-19
particular wave motions. If slippage occurs, then pipeline longitudinal strains will be less than the ground strain; hence the assumption that pipeline strain is equal to ground strain is a conservative upper-bound.

The development of the maximum ground strain in the pipeline is dependent upon whether or not there is sufficient pipeline length and a sufficiently long wave length over which to establish equilibrium without longitudinal slipping of the pipeline within the soil. For very soft soils, the assumption of no slipping may lead to a significant overestimate of the strains that can be transferred from the ground to the pipeline.

The maximum ground curvature, κg, is the second derivative of the transverse displacement and is given by:
g 2
PGA
κc
κ
α = (2-20)
where:
ακ = Ground curvature coefficient corresponding to the most critical angle of incidence and type of seismic wave
= 1.6 for compression waves (ASCE, 1984)
= 1.0 for shear and Rayleigh waves (ASCE, 1984)
The bending strains calculated by equation (2-20) are of small magnitude. Therefore, as a rule, the effect of curvature associated with seismic wave propagation is neglected.

For buried piping or pipeline segments with significant bends, the strain associated with seismic wave propagation will be influenced by both the virtual anchorage caused by the bend and the local flexibility of the pipe bend. An analysis approach that accounts for these conditions for buried pipelines subjected to thermal strain is provided in Appendix VII of ASME/ANSI B31.1. The same analysis approach may be suitable for determining the effects of seismic wave propagation by equating the seismic strains to thermal strains.

2.7 Transient Ground Deformation
Large transient ground deformations, associated with shear failure of weak soils or liquefaction of subsurface soil layers, have been identified as a potential cause of pipeline damage in recent earthquakes. Two conditions for transient ground deformation (TGD) have been identified based upon investigations following the 1989 Loma Prieta and 1994 Northridge earthquakes in California.

One condition susceptible to TGD is a site close to the earthquake epicenter (typically within 20 km) that is characterized by a narrow valley filled with weak soil overlying much stronger soil deposits and bounded by steep side walls. Such valleys are susceptible to direct lateral loading as the surrounding side walls experience long period displacement pulses associated with the directivity of earthquake rupture (i.e., earthquake PR-268-9823
2157
SECTION 2: Quantifying Seismic Hazards 2-20
“fling”). During the earthquake, the motions imparted to the soft soil by the surrounding side walls of the valley are sufficient to generate shear failure in the weak soil. This shear failure decouples the soft soil from the motions of the surrounding side walls and can lead to large transient relative motions at the interface between the weak soil and the stronger soil deposits at depth.

The second condition susceptible to TGD is a site experiencing liquefaction in soil confined by non-liquefiable material. In this case, liquefaction acts to decouple the soil response from that of the surrounding non-liquefiable material leading to the potential for large transient relative motions at the boundaries between liquefied and competent soils.

There are no established procedures for identifying and quantifying TGD. Techniques that have been used to identify TGD are limited to studies to identify the cause of failures in older pipelines not typical of current construction practices (i.e., girth welds with severe defects, pipelines with large-angle mitre joints, or pipelines with corrosion defects). The few past studies performed have used detailed subsurface information and analyses incorporating actual earthquake time histories from nearby ground motion recording stations.

PR-268-9823
2158
SECTION 2: Quantifying Seismic Hazards 2-21
Table 2.1 Information Required for Performing Liquefaction Assessment Information Required Notation
SPT blow counts collected in accordance with
ASTM D-1586
N
SPT bore hole diameter Dbore
Length of rod connected to SPT sampler Lrod
Percentage of hammer energy delivered to
sampling rod in SPT
Eeff
SPT Assessment
Fines content of soil as measured by % of a
sample passing a #200 sieve
FC
CPT tip resistance in accordance with ASTM
D-3441
qc
CPT
Assessment
CPT sleeve friction in accordance with
ASTM D-3441
fs
Depth to water table at the time of the SPT or
CPT
Dw
Earthquake moment magnitude M
Peak ground acceleration at surface PGA
SPT and CPT
Unit weight of soil γ
Assessment
Depth below ground of test measurement z
Table 2.2 Slope Material Strengths for Three Common Slope Materials

LABLE IN
FIGURE 2.8

SLOPE MATERIAL / SOIL

CONDITION
Effective
Cohesion,
c′ (psf)
Effective
Friction
Angle φ′
A Strongly cemented rocks, crystalline
rock, and well-cemented sandstone
300 0
B Weakly cemented rocks, sandy soil,
and poorly cemented sandstone
0 35°
C Argillaceous rocks, clayey soil, and
shale
0 20°
PR-268-9823
2159
SECTION 2: Quantifying Seismic Hazards 2-22
Apparent Dip
Angle, α
Dip Angle, δ
Regional Stress Angle, β
Wells &
Coppersmith
Displacement,
AD or MD
Total Fault
Displacement, DF
Vertical Component of
Fault Displacement, V
Horizontal
Component of Fault
Displacement, H
Transverse Component of
Fault Displacement, H
Figure 2.1 Determination of Components of Actual Fault Displacement Based Upon Fault Dip Angle and Regional Stress Azimuth
PR-268-9823
2160
SECTION 2: Quantifying Seismic Hazards 2-23

LIQUEFACTION NO LIQUEFACTION
Figure 2.2 CRR Required to Prevent Liquefaction vs. (N1)60FC (from Youd and Idriss, 1997)
PR-268-9823
2161
SECTION 2: Quantifying Seismic Hazards 2-24
1g ≈ 1,000 gal

NO DAMAGE

NO DAMAGE DAMAGE
DAMAGE

NO DAMAGE
DAMAGE
Figure 2.3 Screening Chart for Determining whether Consolidation of Liquefied Soil is a Credible Seismic Hazard (from Ishihara, 1985)
PR-268-9823
2162
SECTION 2: Quantifying Seismic Hazards 2-25
γmaz = maximum soil shear strain
Figure 2.4 Chart to Estimate Earthquake-Related Volumetric Strain in Saturated Clean Sand Soil Deposits (from Ishihara, 1990)
PR-268-9823
2163
SECTION 2: Quantifying Seismic Hazards 2-26
Figure 2.5 Geometry Assumed for Estimating Lateral Spread Displacement for Ground Slope Conditions (from Bartlett and Youd, 1992)
PR-268-9823
2164
SECTION 2: Quantifying Seismic Hazards 2-27
Figure 2.6 Geometry Assumed for Estimating Lateral Spread Displacement for Free Face Conditions (from Bartlett and Youd, 1992)
PR-268-9823
2165
SECTION 2: Quantifying Seismic Hazards 2-28
Earth Block Slide (coherent)
Earth Flow (occurs very slow to
rapid and can be coherent to
disruptive in nature)
Debris Avalanche (disruptive and
occurs very rapid to extremely rapid)
Rock Topple (disruptive)
Earth Slump (coherent)
Rock Fall (disruptive)
Figure 2.7 Examples of Various Types of Landslides (Varnes, 1978) PR-268-9823
2166
SECTION 2: Quantifying Seismic Hazards 2-29

0.00 0.10

0.20 0.30

0.40 0.50
0 1020 3040 50
Slope Angle, degrees
A dry
A wet
B dry
B wet
C dry
C wet
Figure 2.8 Critical Acceleration for Constant Slope and Soil Strength PR-268-9823
2167
SECTION 3: Pipeline Acceptance Criteria 3-1

3.0 PIPELINE PERFORMANCE CRITERIA
The most common type of loads generated on a buried pipeline by seismic hazards is generated by ground movement relative to the pipeline. The pipeline sees no further soil load once it has deformed sufficiently to match the ground movement. This type of loading is commonly referred to as displacement-controlled. Guidance is provided for two performance goals associated with displacement-controlled conditions, maintaining strains to a level that normal operations can continue and maintaining the pressure integrity of the pipeline.

Pipelines buried with minimal soil cover or in relatively weak soils and subjected to high axial loads as a result of ground movement may experience upheaval buckling. While initial pipeline response is displacement controlled, upheaval of the pipeline can lead to some strain relaxation and a related concentration of compression strain in the vicinity of pipeline upheaval. The amount of strain concentration that can occur is limited by the axial soil restraint on the portion of the pipeline that remains buried. This type of behavior is referred to as displacement-limited. Development of significant pipeline compression sufficient to cause upheaval buckling should be avoided by proper orientation of the pipeline relative to the direction of ground movement. Where upheaval buckling can not be avoided, the potential for pipe strains to concentrate needs to be investigated. Such an investigations will typically consider the bending moment capacity of the pipeline, the length of pipeline that might release strain into the portion pipeline undergoing upheaval displacement, and changes in soil restraint as a function of pipeline upheaval displacement. Full treatment of upheaval buckling is beyond the scope of these guidelines.

In some cases, a seismic hazard results in a condition where loads act continuously on the pipeline. This type of loading is commonly referred to as load-controlled. Pipeline strains for load-controlled loading conditions must remain below the strain associated with reduced load carrying capacity of the pipeline. Guidance is also provided for strain limits associated with load-controlled conditions.

All of the strain limit equations in this section are based upon recommendations from previous testing programs. In nearly all cases, strains were measured over a length of pipe, often referred to as the gauge length, equal to 0.5 to 2.0 pipe diameters. For tests that produced wrinkles or buckling in the pipe wall, the gauge length encompassed the wrinkle and buckle. Therefore, the strain limits in this section are applicable to analyses that can resolve global pipe response over a distance of one pipe diameter or less. More refined analytical representation of the pipe (e.g., use of shell elements) will provide information on the strains at a location of local pipe wall distortion. Alternate strain criteria need to be developed to relate pipe performance to these localized strains.

Pipeline performance criteria are defined in terms of allowable longitudinal compression and tension strains for onshore pipelines. The strain criteria assume that an analysis of pipeline response will be performed using finite element techniques and bilinear representation of soil loads as described in Section 4.0. The criteria defined in this section are not applicable to analyses based upon approximations of pipeline response PR-268-9823
2168
SECTION 3: Pipeline Acceptance Criteria 3-2
carried out using hand calculations that cannot resolve pipeline strains within a length of approximately one diameter.

3.1 New Pipelines and Pipelines Compatible with Modern
Construction Practices
The strain limits provided in this section are applicable to pipelines that meet the following conditions:

1. Pipe material and fabrication equivalent to that specified in API 5L

2. Joint connections using full-penetration butt welds in portions of the pipeline expected to experience stresses near or above yield (e.g., no belled-end welded joints, slip joints, lap-welded joints, sleeves)

3. Weld quality consistent with current welding procedures (e.g., API 1104)

4. Weld strength matching or exceeding the expected actual pipe strength

5. Pipe, weld, and heat-affected zone toughness:
(a) Average Charpy energy greater than 40 J (30 ft-lb), or (b) Average CTOD toughness greater than 0.20 mm.

6. Ratio of yield strength to tensile strength (Y/T) less than 0.92 Strain limits for computed longitudinal compression and tension strains (hereafter referred to as simply compression and tension strains) are provided for three conditions:

1. Strains resulting from imposed ground displacements with the goal of maintaining pressure integrity of the pipeline with a high likelihood of repairs to the pipeline following the earthquake;

2. Strains resulting from imposed ground displacements with the goal of maintaining normal operation of the pipeline with a low likelihood of repairs to the pipeline following the earthquake; and

3. Strains resulting from load conditions in which the displacement of the pipeline is not limited by ground displacement.

3.1.1 Performance Goal: Maintain Pressure Integrity
The performance goal, “maintain pressure integrity,” accepts significant ovalization and possible initiation of wrinkle formation in the body of the pipe. With this performance goal, post-earthquake response should include provisions to reduce operating pressure and perform site investigations to identify indications that permanent ground deformation has occurred. Replacement of the damaged portion of the pipeline may be necessary to resume normal operations.

PR-268-9823
2169
SECTION 3: Pipeline Acceptance Criteria 3-3
Longitudinal Compression Strain Limit

1.76 cp
t
D
ε = ≤ the longitudinal tension strain limit (3-1)

Longitudinal Tension Strain Limit

0.02 to 0.04 tp ε = (see Commentary) (3-2)

3.1.2 Performance Goal: Maintain Normal Operability
The performance goal, “maintain normal operability,” provides a high level of confidence of no significant pipeline damage. The expression in equation (3-1) is valid for diameter to pipe wall thickness ratios of 100 or less and provides a term to account for an assumed level of pipe ovaling.

Longitudinal Compression Strain Limit
Recommended relationships for allowable longitudinal compression strains for continued operation are provided for two types of typical pipeline stress-strain curves: (1) those exhibiting typical rounded shapes at yield and (2) those that exhibit a distinct yield plateau.

Rounded Stress-Strain Curve with Plain Pipe Imperfection 1 0.70 1.72 0.086
co 0.437 1 0.892 1.09
yy
t pE
Dp t
ε
σ
−
⎛⎞ ⎡⎤ ⎛ ⎞⎛ ⎞ ⎛⎞ ⎛ ⎞ ∆ =− ⎢⎥ ⎜ ⎟⎜ ⎟ ⎜⎟ − ⎜⎟ ⎜ ⎟ ⎝⎠ ⎢⎥ ⎝ ⎠ ⎣⎦ ⎝ ⎠⎝ ⎠ ⎝⎠ (3-3a)
Rounded Stress-Strain Curve with Pipe End Offset at Girth Weld 1 0.85 1.59 0.15
co 0.056 1 0.868 1.27
yy
t pE im
Dp t
ε
σ
−
⎡⎤ ⎛ ⎞⎛ ⎞ ⎛⎞ ⎛⎞ ⎛ ⎞ =− ⎢⎥ ⎜ ⎟⎜ ⎟ ⎜ − ⎜⎟ ⎜ ⎟ ⎜ ⎝⎠ ⎢⎥ ⎝ ⎠ ⎣⎦ ⎝ ⎠⎝ ⎠ ⎝⎠ p ⎟
⎟ (3-3b)
Distinct Yield Plateau with Plain Pipe Imperfection
1 0.70 20
co 1.06 1 0.50 1.10
yy
t pE
Dp t
ε
σ
−
⎛⎞ ⎡⎤ ⎛ ⎞⎛ ⎞ ⎛⎞ ⎛ ⎞ ∆ =− ⎜⎟ ⎢⎥ ⎜ ⎟⎜ ⎟ ⎜ −⎜ ⎟ ⎜ ⎝⎠ ⎢⎥ ⎝ ⎠ ⎣⎦ ⎝ ⎠⎝ ⎠ ⎝⎠ .09
⎟
⎟ (3-3c)
PR-268-9823
2170
SECTION 3: Pipeline Acceptance Criteria 3-4
Rounded Stress-Strain Curve with Pipe End Offset at Girth Weld 1 0.80 20
co 0.404 1 0.906 1.12
yy
t pE im
Dp t
ε
σ
−
⎡⎤ ⎛ ⎞⎛ ⎞ ⎛⎞ ⎛⎞ ⎛ ⎞ =− ⎜⎟ ⎢⎥ ⎜ ⎟⎜ ⎟ ⎜ − ⎜ ⎟ ⎜ ⎝⎠ ⎢⎥ ⎝ ⎠ ⎣⎦ ⎝ ⎠⎝ ⎠ ⎝⎠ .15
p ⎟
⎟ (3-3d)
where:
t = pipe wall thickness
D = pipe outside diameter
p = internal pressure
py = internal pressure to produce hoop stress equal to yield stress σy = yield stress
E = modulus of elasticity of pipe material
imp = blister-type initial imperfection expressed as a percentage of wall thickness
∆ = offset between pipe joints at girth weld
Longitudinal Tension Strain Limit
ε to = 0.01 to 0.02 (see Commentary) (3-4)

3.1.3 Load-Controlled Conditions
The primary earthquake hazards to buried pipelines result in load conditions that are controlled by the ability of the pipeline to conform to the displacement of the ground (displacement-controlled). It is possible that situations can arise in which the loading on the pipeline is not related to permanent ground displacement. For these conditions, pipeline strength is of primary importance. The strain criteria for load-controlled conditions are based upon preventing formation of local wrinkling of the pipe wall and preventing large post-yield tension strains.

Longitudinal Compression Strain Limit:

1.59 0.75 2.42 cl co
t
D
εε
⎛⎞ =≥ ⎜⎟ ⎝⎠ (3-5)
Longitudinal Tension Strain Limit:
ε tl = 0.005 (3-6)
PR-268-9823
2171
SECTION 3: Pipeline Acceptance Criteria 3-5
Note that the maximum tension strain limit is equal to the nominal yield strain for pipe materials consistent with API 5L. Given this, the use of the pipe yield stress as a controlling acceptance criterion is also acceptable. If compression strain limits are determined to be less than 0.005, the allowable compression stress limit should be established based upon the stress-strain properties of the pipe material.

3.2 Considerations for Offshore Pipelines
Offshore pipelines that are simply laid on the sea floor with minimum embedment, so as not to be constrained by burial in the soil, can generally accommodate significant levels of deformation without damage. Sections of a pipeline that are trenched into the sea floor will exhibit seismic response characteristics similar to those of onshore pipelines. Both trenched and untrenched pipelines are vulnerable to hazards similar to onshore pipelines including submarine landslides and liquefaction-induced lateral spreads, flow failures, and debris flows. Like onshore debris flows, submarine debris flows can move very quickly and may present a significant load-controlled design condition for an offshore pipeline. In addition, offshore pipelines may be subjected to turbidity currents related to submarine ground failure.

Analysis of offshore pipelines proceeds in the same manner as for onshore pipelines.

Assessment of computed pipeline strains needs to address the effects of external pressure to reduce the available compression bending strain. For offshore pipelines, the compression longitudinal bending strain limit is given by equation (3-7):
ext
cr p cr
cr
P
g P
εε −
⎛ = − ⎜
⎝⎠
⎞
⎟ (3-7)
where:
εcr-p = longitudinal compression strain for condition with external pressure εcr = limiting compression strain from Section 3.1
Pext = net external pressure = total external pressure – internal pressure Pcr =
PP
PP
ey
ey
⋅
+22 = predicted collapse pressure
Pe = 22
3 . ⋅ F
H
G I
K E J
t
D = elastic buckling pressure with Poisson’s ratio = 0.3 Py =
2⋅⋅ σ y t
D = pressure to produce cross-section yielding
PR-268-9823
2172
SECTION 3: Pipeline Acceptance Criteria 3-6
g =
1 1
1 1
22 2
2 2
+
+−
F
H
G I
K
J
+ F
H
G I
K
J
≤
P
P
D
t
D
t
P
P
y
e
oo
y
δδ e
δo = initial ovalization

3.3 Pipelines Not Compatible with Current Construction
Practices
Older pipelines may not conform to the attributes listed in Section 3.1, particularly with respect to pipe toughness and absence of corrosion or weld defects. These conditions may be accounted for in the assessment by appropriate reduction of the allowable strain limits in Sections 3.1.1 through 3.1.3.

Some older pipelines were constructed with a variety of girth weld details with oxyacetylene or arc welding techniques that predated modern day controls on filler material and flux composition and power supply, and lacked good quality control. Belland-spigot or bell-bell-chill-ring details are examples of older girth weld details. In addition, older pipelines may have been fabricated in manner that would not be permitted under current codes with such details as mitre bends, wrinkle bends, hot and cold field bends to accommodate changes in alignment. The strain capacity of these types of joints and conditions is highly variable and generally low compared to modern construction.

Strain acceptance criteria for these types of pipelines need to be developed on a case-bycase basis. Acceptable approaches for establishing strain criteria include full-scale testing, finite element shell analyses, wide-plate testing, and fracture mechanics evaluations.

More rigorous approaches to establishing allowable strains based upon pipeline inspections and fracture mechanics principles are strongly encouraged. Selecting methods appropriate for specific pipe and load conditions is an area of active research that is not sufficiently developed for incorporation into general seismic guidelines.

PR-268-9823
2173
SECTION 4: Pipeline Analysis Procedures 4-1

4.0 PIPELINE ANALYSIS PROCEDURES
Finite element analysis of pipeline response to imposed ground movements is the preferred means to assess pipeline response using the provisions of these guidelines, particularly considering the advent of high-speed, desktop computers and the proliferation of relatively inexpensive, user-friendly finite element analysis software.

The recommended approach is to represent the pipeline with pipe or beam elements and represent the soil loading on the pipeline with discrete spring elements. This approach is illustrated in Figure 4.1.

There are no restrictions on the analysis software that can be used as long as it is capable of capturing the non-linear effects of non-linear soil springs, user-defined stress-strain curves for the pipe material, and large changes in pipeline geometry. The definition of soil springs assumes that the spring forces always act in the axial, horizontal, and vertical directions relative to the pipeline. To maintain this condition for situations where ground and pipeline movements are large, the analysis model should be capable of defining the soil spring displacements relative to the local pipeline coordinate system. The analysis should be capable of accounting for the effects of internal pressure because the hoop stress from internal pressure reduces the longitudinal stress at which the material will yield. It is also very beneficial if the analysis software has the capability to provide computed strains at multiple locations around the circumference of the elements used to represent the pipeline.

4.1 Pipe Element Definition
The pipe element length in regions where the pipe strain is expected to exceed the yield strain (typically at abrupt transitions in ground displacement or locations with abrupt changes in soil restraint such as elbows) generally should not exceed one pipe diameter.

Multiple straight or curved pipeline elements can be used to model elbows and bends in the pipeline alignment. The maximum length of pipeline elements used to model elbows should not represent more than a 15° angular change (e.g., at least 6 elements to model a 90° elbow). It is preferable to use multiple curved pipe elements at bends and elbows if they are available. In most cases, the differences in results between using curved pipe elements versus many straight pipe elements are not significant.

4.2 Pipe Stress-Strain Definition
Nonlinear material representation in analysis software is typically based upon a definition of a uniaxial engineering stress-strain curve that is converted to a true stress-strain curve within the software application. Users typically have several options to choose from regarding the modeling of plasticity. For the analysis methodology adopted in these guidelines, bilinear or multilinear isotropic hardening rules based upon a von Mises yield criterion are adequate. Alternate approaches to modeling plastic behavior can also be adopted provided they have a sound basis in engineering mechanics.

PR-268-9823
2174
SECTION 4: Pipeline Analysis Procedures 4-2

4.3 Soil Spring Definition
Soil loading on the pipeline is represented by discrete nonlinear springs as illustrated in Figure 4.1. The maximum soil spring forces and associated relative displacements necessary to develop these forces are computed using the equations given in the sections below.

Soil properties representative of the backfill should be used to compute axial soil spring forces. Other soil spring forces should generally be based upon the native soil properties.

Backfill soil properties are appropriate for computing horizontal and upward vertical soil spring forces only when it can be demonstrated that the extent of pipeline movement relative to the surrounding backfill soil is not influenced by the soils outside the pipe trench.

Although tests have indicated that the maximum soil force on the pipeline decreases at large relative displacements, these guidelines are based upon the assumption that the soil force is constant once it reaches the maximum value. The dimension for the maximum soil spring force is force per unit length of pipeline. The equations below are based upon buried pipelines in uniform soil conditions.

For deeply buried pipelines with variable soil properties between the ground surface and the pipeline depth, the equations below may not be representative of the true soil loading conditions. Guidance on how to proceed with variable soil conditions is provided in the Commentary.

Horizontal soil loads on offshore pipelines resting on the sea floor increase more gradually with displacement due to the formation of a mound of soil in front of the pipeline. Determination of the soil spring characteristics for this condition requires special treatment by experienced practitioners and is not covered in these guidelines.

The expressions for maximum soil spring force are based upon laboratory and field experimental investigations on pipeline response, as well as general geotechnical approaches for related structures such as piles, embedded anchor plates, and strip footings. Several of the equations have been derived to fit published curves to facilitate their use in spreadsheets or other computer-based applications.

4.3.1 Axial Soil Springs
T D c DH K u
o =+ F +
H
G I
K πα π γ δ J 1
2 tan( ) (4-1)
where:

D = pipe outside diameter
c = soil cohesion representative of the soil backfill
PR-268-9823
2175
SECTION 4: Pipeline Analysis Procedures 4-3
H = depth to pipe centerline
γ = effective unit weight of soil
Ko = coefficient of pressure at rest
α = adhesion factor, a curve fit to plots of recommended values is α = 23

0.274 0.695 0.608 0.123
11 c
cc
− −+
+ +
where c is in ksf or kPa/50
δ = interface angle of friction for pipe and soil = fφ
φ = internal friction angle of the soil
f = coating dependent factor relating the internal friction angle of the soil to the friction angle at the soil-pipe interface
Representative values of f for various types of external pipe coatings are provided below.

PIPE COATING f
Concrete 1.0
Coal Tar 0.9
Rough Steel 0.8
Smooth Steel 0.7
Fusion Bonded Epoxy 0.6
Polyethylene 0.6
∆t = displacement at Tu
= 0.1 inches (3 mm) for dense sand
= 0.2 inches (5 mm) for loose sand
= 0.3 inches (8 mm) for stiff clay
= 0.4 inches (10 mm) for soft clay

4.3.2 Lateral Soil Springs
P N cD N HD Q u ch qh d =+ ≤ γ (4-2)
where:

Nch = horizontal bearing capacity factor for clay (0 for c = 0) Nqh = horizontal bearing capacity factor (0 for φ = 0°) The expressions below for Nch and Nqh are closed-form fits to published empirical (plotted) results (see Commentary).

PR-268-9823
2176
SECTION 4: Pipeline Analysis Procedures 4-4
Nch = horizontal bearing capacity factor for clay (0 for c = 0) = 23 ( 1) ( 1)
cd a bx
xx
++ +
++
≤ 9
Nqh = horizontal bearing capacity factors for sand (0 for φ = 0°) = 23 a bx cx dx ex ++ + + () ( ) ( ) ( )4
Factor φ x a b c d e
Nch 0° H/D 6.752 0.065 -11.063 7.119 --
Nqh 20° H/D 2.399 0.439 -0.03 1.059(10)-3 -1.754(10)-5
Nqh 25° H/D 3.332 0.839 -0.090 5.606(10)-3 -1.319(10)-4 Nqh 30° H/D 4.565 1.234 -0.089 4.275(10)-3 -9.159(10)-5 Nqh 35° H/D 6.816 2.019 -0.146 7.651(10)-3 -1.683(10)-4 Nqh 40° H/D 10.959 1.783 0.045 -5.425(10)-3 -1.153(10)-4 Nqh 45° H/D 17.658 3.309 0.048 -6.443(10)-3 -1.299(10)-4 Nqh can be interpolated for intermediate values of φ between 20° and 45°. ∆p = displacement at Pu
= 0.04 0.10 to 0.15
2 D HD ⎛⎞ +≤ ⎜⎟ ⎝⎠ D

4.3.3 Vertical Uplift Soil Springs
The equations for determining upward vertical soil spring forces are based upon smallscale laboratory tests and theoretical models. For this reason, the applicability of the equations is limited to relatively shallow burial depths, as expressed as the ratio of the depth to pipe centerline to the pipe diameter (H/D). Conditions in which the H/D ratio is greater than the limit provided below require case-specific geotechnical guidance on the magnitude of soil spring force and the relative displacement necessary to develop this force.

Q N cD N HD u cv qv =+ γ (4-3)
where:

Ncv = vertical uplift factor for clay (0 for c = 0)
Nqv = vertical uplift factor for sand (0 for φ = 0°)
PR-268-9823
2177
SECTION 4: Pipeline Analysis Procedures 4-5
Ncv = 2 H
D
⎛⎞ ⎜⎟ ≤ ⎝⎠ 10 applicable for 10 H
D
⎛⎞ ⎜⎟ ≤ ⎝⎠
Nqv = tan( ) φ φ
44 F
H
G I
K
J
F
H
G I
K
J H
D
≤ Nq (see Section 4.3.4 for definition of Nq)
∆qu = displacement at Qu
= 0.01H to 0.02H for dense to loose sands ≤ 0.1D
= 0.1H to 0.2H for stiff to soft clays ≤ 0.2D

4.3.4 Vertical Bearing Soil Springs
Q N cD N HD N D
dc q =+ + γγγ
2 2 (4-4)
where:

Nc, Nq, Nγ = bearing capacity factors
Nc = cot( . ) tan . tan( . ) φ πφ φ ++ F +
H
G I
K
J − L
N
M O
Q
P + 0 001 45 0 001
2 1 0 001 2 e
Nq = eπφ tan( ) φ tan2 45
2 + F
H
G I
K
J
Nγ = (this is a curve fit to plotted values of N (0.18 2.5) e φ− γ) γ = total unit weight of soil
∆qd = displacement at Qd
= 0.1D for granular soils
= 0.2D for cohesive soils

4.4 Extent of Pipeline Model
The length of pipeline modeled outside of the zone of applied ground movement is dependent upon the specific pipeline alignment. The length of pipeline needs to be sufficient to assure that small elastic deformations outside the area of high pipeline strains do not significantly change the magnitude of the highest computed longitudinal strains. An estimate of the extent of the model for which the pipe can be considered anchored to the soil, Lanchor, can be computed using the following formula:

L
Dt
T anchor
y
u
= π σ(4-5)
PR-268-9823
2178
SECTION 4: Pipeline Analysis Procedures 4-6
The length Lanchor is the distance necessary for the axial soil force, Tu, to generate axial yield in the pipeline. Termination of pipelines within a zone of ground deformation should force the pipeline to move with the ground. For example, analysis of straight pipelines crossing faults will have one end of the pipeline model fixed at a distance Lanchor from the fault and the other end displaced an amount consistent with the fault movement.

For pipelines that do not have a straight alignment outside of the zone of ground movement (e.g., side bends, over bends, sag bends), a virtual anchor location may exist at a distance considerably less than that estimated by equation (4-6). This actual point at which the pipeline is anchored needs to be considered on a case-by-case basis. If the location of anchorage is not readily apparent, the modeling can include the actual pipe configuration for a pipe length equal to Lanchor. Any anchorage provided by the pipe configuration (e.g., bends, elbows, changes in soil cover) will then be captured in the analysis.

4.5 Representation of Applied Ground Movement
Ground movements representative of the displacement patterns and amplitudes for lateral spreads or landslides are applied to the base of the soil spring elements. The ground deformations should be specified based upon estimates of relative ground movement at the depth of the pipeline. Ideally, the expected pattern of ground displacement should be determined on the basis of geotechnical field investigations, but such may not be feasible in all cases considering the spatial extent of pipelines and the potential for numerous hazard areas.

It is always conservative to assume that the ground displacement occurs abruptly. Abrupt ground displacements have been observed at surface faults, as well as the head of landslides and lateral spreads. In some locations, it may be possible to infer future ground displacement patterns from past earthquake displacements. For locations with little earthquake experience to aid in the determination of ground deformation pattern, an abrupt offset cannot be ruled out without the assistance of a specialist in geology or geotechnical engineering.

4.6 Sensitivity Analyses
The selection of material properties, soil strengths, and ground movement patterns is an inherently uncertain process. Generally, additional analyses should be performed to provide information on the sensitivity of the computed strain levels to changes in input parameters when the level of conservatism associated with the input parameters is not well understood. This information can be used to better define the “best estimate” of pipe response and provide information that can be used to assess the level of confidence in the expected pipeline performance. The range of variation to be used in these sensitivity analyses generally can be estimated in conjunction with the selection of baseline parameters and any geotechnical investigations of soil properties and ground deformations. Unless other information is available to determine the amount of variation, PR-268-9823
2179
SECTION 4: Pipeline Analysis Procedures 4-7
the following are suggestions for examining the sensitivity of the results obtained by analysis:

1. Upper-bound estimates of pipe material strength (as opposed to specified minimum values);

2. Variation in soil strength to capture the reasonable range of upper-bound and lower-bound ranges;

3. Increased applied ground displacement (this can normally be done at the time of the analysis); and

4. Modifications in the ground displacement transitions at boundaries with zones of potential ground movement.

The above variations typically result in a range of acceptable ground displacements or a range of strains for a particular ground displacement value. If a balanced range of parameters is selected, the “best estimate” of pipe response can typically be estimated from the resulting range in pipeline response determined from the variation in input parameters.

PR-268-9823
2180
SECTION 4: Pipeline Analysis Procedures 4-8

B) IDEALIZED REPRESENTATION OF SOIL WITH DISCRETE SPRINGS

A) ACTUAL THREE-DIMENSIONAL SOIL RESTRAINT ON PIPELINE

C) BI-LINEAR SOIL SPRINGS USED TO REPRESENT SOIL FORCE ON PIPE ∆qu
Qd
Q
∆t Qu
Pu
Tu
∆p ∆t
∆qd ∆p
Tu
Pu
P T
Figure 4.1 Pipeline Modeling Approach
PR-268-9823
2181
SECTION 5: Mitigation Options 5-1

5.0 MITIGATION OPTIONS
There are several options to improve post-earthquake conditions if the pipeline response is found to exceed the acceptance criteria defined in 3.0. Selection of a particular approach is dependent upon considerations that vary with pipeline location, expected failure mode, potential for collateral damage, risk acceptance philosophy, and estimated mitigation costs. Mitigation options fall into one of four categories: (1) relocate the pipeline to avoid the hazard, (2) modify/design the pipe construction to reduce the soil loads on the pipeline, (3) modify the pipeline configuration to increase its ability to resist ground displacement, and (4) modify post-earthquake response procedures to address the consequences of pipeline damage. Relocation to avoid the hazard requires no further explanation. The following discussion focuses on pipeline installation changes and operational measures to improve pipeline performance.

5.1 Modifying Pipeline Loading Conditions
Soil loads on buried pipelines can be reduced in several ways. The most common approach is to minimize the strength of the soils surrounding the pipeline or the frictional characteristics of the pipeline. The alignment of the pipeline can also be modified to change the orientation and distribution of soil loads with respect to the pipeline. Finally, provisions can be made to isolate the pipeline from relative soil movement. Potential options for implementing changes to modify the soil loading on buried pipelines are summarized below. Many of the options have limited applicability because of restrictions related to right-of-way access, the need to avoid existing subsurface structures and utilities, or the compaction requirements associated with various types of land use.

Urban environments are particularly restrictive with respect to the feasibility of mitigation options to improve pipeline response.

5.1.1 Reducing Soil Loads by Minimizing Soil Strength Properties The capacity of a buried pipeline to withstand ground displacements can be improved by minimizing the longitudinal, lateral, and uplift soil resistance to pipe movements.

5.1.1.1 Loose Granular Backfill
A practical means for achieving minimum soil restraint is to bury the pipeline in a shallow trench filled with a loose granular backfill. As depicted in Figure 5.1, the trench walls should be sloped at an angle of about 30° to 45° for horizontal ground displacement components and about 60° for vertical ground displacement components. This trench geometry will allow the soil to fail within the backfill material rather than in the higher strength, undisturbed soil outside the trench.

For typical pipeline trench conditions, loose granular backfills (sand or gravel) will offer less resistance to pipe movement than compacted cohesive backfill materials (clay or silty clay). A granular material with an angle of internal friction of 35° or less is PR-268-9823
2182
SECTION 5: Mitigation Options 5-2
recommended. To satisfy an angle of internal friction of 35° or less, the backfill material should be a well-graded granular material with 100 percent of the aggregate less than one inch (25 mm) in diameter. The material should be obtained from a natural well-graded fluvial deposit; crushed rock is not acceptable. The backfill should be moderately compacted to a relative density, Dr, of 66 percent or less if achievable.

The assumption of any particular soil condition and the development of spring restraint properties must be consistent with field conditions. In particular, for horizontal relative ground displacement, the logarithmic spiral failure surface (depicted in Figure 5.2) must be enveloped by the limits of the excavated pipe trench that is backfilled with the selected material. Similarly, for vertical relative ground displacements, the upward breakout must occur within the designated backfill. If these trench excavation and backfill requirements are not satisfied, soil parameters applicable to the hybrid situation of in situ soils and trench backfill must be considered in the development of soil restraints for the pipeline, and these restraint properties are typically much higher than for loose granular backfill.

5.1.1.2 Locating Pipeline at or Above Grade
Lateral soil loads can be greatly reduced by placing the pipeline on the ground surface or on aboveground supports. Typcially this is done by attaching sliding shoes to the pipeline that bear on structural steel members tied to the ground or mounted in an aboveground configuration (e.g., aboveground segments of the Trans-Alaska pipeline).

Teflon, or other low-friction materials, can be incorporated into the construrction of the sliding shoes to improve the ability of the pipeline to accommodate ground displacement by sliding laterally. Several options for such a modification are shown in Figure 5.3.

Where soil cover is required to protect the pipeline from third-party damage, an earthen berm can be used in lieu of burial. Locating pipelines above ground is rarely a practical solution outside of controlled access areas or very remote regions. However, in some cases, it is feasible to construct a sliding support configuration at the bottom of an open trench. The top of the open trench is will generally be covered (e.g., steel grating) to protect the pipeline from vandalism and reduce the impacts on surface activities.

5.1.1.3 Low-Friction Coating or Protective Wrapping
Axial soil friction loads can be further reduced over what is achieved by loose backfill by the use of smooth, hard, low-friction coatings. Larger reductions in axial soil friction can be obtained by creating a preferred slip surface other than the pipe-soil interface. One means of achieving this is to use two separate layers of geosynthetic wrapping as shown in Figure 5.4a. This type of installation forces axial slip to occur at the interface between the two layers of geotextile fabric and can reduce the interface friction angle used to calculate maximum axial soil spring force to less than 10°. In practice, the reduction in axial soil friction force obtained from a double geotextile fabric wrapping should be verified by field tests under soil and backfill conditions that replicate the actual installation. The effects of aging on the low friction interface should also be evaluated.

PR-268-9823
2183
SECTION 5: Mitigation Options 5-3

5.1.1.4 Geosynthetic Lining of Sloped Trench Walls
Additional reduction in horizontal soil load over what can be achieved using loose granular backfill in conjunction with sloped trench walls (see 5.1.1.1) is possible by lining the walls of the trapezoidal trench with two layers of geosynthetic fabric. The two layers of geosynthetic fabric create a low-friction failure surface in lieu of the logarithmic spiral failure surface that would be developed in the backfill material. The load necessary to overcome the friction between the two layers of geosynthetic fabric is much less than that required to develop a shear failure in the backfill soil.

5.1.1.5 Replacing Soil with Geofoam
Geofoam materials offer a means to reduce axial, lateral, and upward vertical soil loads.

Geofoam is a rigid cellular plastic foam of either expanded polystyrene (XPS) or extruded polystyrene (EPS). Geofoam has been used extensively in northern Europe for subgrade insulation in regions susceptible to frost heave. Another usage of geofoam in Europe and the U.S. is as low-density fill for construction over weak soils. One common application is to use geofoam as fill for bridge approaches and abutments. Geofoam varies in weight from about 160 N/m3
(1 lb/ft3) to 470 N/m3 (3 lb/ft3). Compressive
strength of XPS is generally less than EPS although the compressive strength of both increases with density. The typical range of compressive strengths is 140 kPa to 240 kPa (20 psi to 35 psi) for EPS and 200 kPa to 500 kPa (30 psi to 75 psi) for XPS.

Replacing soil above the pipeline with geofoam reduces axial friction force by effectively reducing overburden stresses acting normal to the pipeline. Care must be taken to maintain a proper balance between limiting pipeline restraint for ground movement, yet providing sufficient restraint to prevent upheaval buckling of straight pipe and excessive bending stress at pipe bends due to operating load conditions.

Lateral loads/restraints on buried pipelines also can be reduced by using geofoam to replace much of the backfill soil as shown in Figure 5.4b. This load reduction is achieved due to two factors. First, the boundary between soil and geofoam forms a failure surface that is weaker than that corresponding to the logarithmic spiral failure surface depicted in Figure 5.2. Second, the weight of material displaced along the soil-geofoam boundary is much less. The shearing force can be reduced even further by placing loose sand between the geofoam and the native soil or through the use of dual layers of geosynthetic fabric at the interface.

The use of geofoam has some advantages in urban settings because the compressive strength of the geofoam is sufficient to handle light traffic loads. Potential significant drawback to geofoam applications include increased vulnerability to applied external loads, need to prevent contact with gasoline and other organic fluids or vapors, high flammability, and cost.

5.1.1.6 Use of Crushable Material to Limit Maximum Loads on the Pipeline Controlled strength material around the pipeline can be used to limit the lateral loads that can be exerted normal to the pipe wall. Geofoam and cellular concrete are two materials PR-268-9823
2184
SECTION 5: Mitigation Options 5-4
well-suited for this purpose since they can experience large compression strain under near-constant compression load. Cellular concrete is a mixture of sand, cement, and water to which a foaming agent or polystyrene beads are added to create small air pockets. The use of crushable material as a mitigation measure is typically only a practical consideration for fault crossings in rock or rock-like materials where very large lateral soil-spring forces can be generated and excavating a trapezoidal trench that can be filled with loose granular material is difficult or impractical. Controlled-strength materials improve pipeline response by allowing the pipe to bend in a more gradual manner to accommodate the imposed ground displacement.

The distance from the pipeline that is to be filled with controlled-strength material is largely governed by the amount of compressive irrecoverable strain that can be accommodated before the material begins to exhibit much higher compressive strength.

This strain level is commonly referred to as the lock-up strain. The lock-up strain for cellular concrete can vary from 15% to 35%. The lock-up strain for EPS or XPS geofoam can vary from 25% to more than 50%.

5.1.2 Unanchored Length
The capacity of a buried pipeline to withstand ground displacement components can be improved by maximizing the distance from the deformation zone (fault rupture, landslide, lateral spread, etc.) to points of virtual anchorage, typically side bends, overbends, and sagbends. Sharp bends, tees, branch fittings, valves, etc. also will have a tendency to anchor the pipeline against axial movement and should be avoided within or near a zone of ground displacement. Good design practice is to provide a straight segment of pipeline as long as practical through and beyond the ground displacement zone to maximize the length of pipeline available to distribute strain.

5.1.3 Isolating Pipelines from Ground Displacement
Soil loads on buried pipelines are the result of relative movement between the pipe and the surrounding soil. This relative movement can be minimized or eliminated by providing space around the pipeline that is greater than the relative movement associated with poor pipeline response. Isolating the pipeline from ground displacement is generally practical only when the ground displacements are relatively small, the length of pipeline requiring isolation is relatively short, or the pipeline diameter is relatively small.

Mitigation concepts using culverts allow the pipeline to respond in a manner similar to an on-grade condition. The term culvert refers to any buried structure built partially or completely around the pipeline to provide an unobstructed space for the pipeline to deform in a direction transverse to its axial alignment. Three conceptual culvert configurations are illustrated in Figure 5.5. Culvert concepts are essentially specialized casings and the same problems that can arise for cased pipelines generally apply to culverts. In addition, caution is needed to assure that axial loads from thermal changes or internal pressure do not lead to buckling of the pipe within the culvert. Axial buckling can typically be prevented by incorporating bends or expansion loops in the pipeline.

PR-268-9823
2185
SECTION 5: Mitigation Options 5-5

5.2 Modify Pipeline Configuration
In some cases, minor modifications to proposed or existing pipeline configurations can greatly improve performance. Such modifications include increasing the pipe wall thickness, increasing the strength and toughness of the pipe material, and replacing sharp bends and elbows with induction bends or gradual pipeline field bends.

Increasing the pipe wall thickness increases the allowable longitudinal compression strain and increases the bending and axial strength of the pipeline relative to the soil. Avoiding sharp bends and elbows can minimize the degree to which the pipeline must accommodate the ground displacement through local bending.

Isolation valves, automatic or remotely controlled, may be provided on each side of the zones of ground displacement to mitigate the consequences of possible pipeline ruptures.

It should be recognized that the use of automatic or remotely-controlled valves on liquid lines also involves consideration of the effects of hydraulic transients to avoid having a rapid shutdown result in risks from possible overpressurization more significant than the risks from the seismic hazards.

5.3 Modify Emergency Response Procedures
Knowing that a pipeline is susceptible to earthquake damage and planning for the damage in the event an earthquake occurs is often sufficient to assure that the performance criteria for the overall system are met. Planning for pipeline earthquake damage is not fundamentally different than planning for damage from any other source. However, normal emergency response procedures are typically inadequate for dealing with postearthquake recovery, because there may be multiple emergencies occurring simultaneously at a time when the general civil infrastructure is damaged (e.g., roads, bridges, electric power, communication) and access to the pipeline or coordination of inspection and repair efforts may be impeded.

Knowledge of the conditions under which emergency response procedures will be implemented is vital. A partial list of some conditions to consider is provided below:

1. Normal electric power may be lost for several hours to several days.

2. Water and sanitary services may be lost for several hours to weeks.

3. There likely will be reduced demand for product.

4. Telephone services, wire and wireless, may be interrupted or heavily congested.

5. Highway bridges and over passes may be damaged, resulting in road closure.

6. Traffic may be congested due to road damage and emergency activity.

7. Personnel may be unavailable due to property damage, injury, or family needs.

PR-268-9823
2186
SECTION 5: Mitigation Options 5-6

8. Maintenance and construction contractors may be unavailable due to increased demand for their services or damage to their facilities.

9. Repair crews may be hampered by food and fuel shortages and interruption of electric power, water, and sanitary services.

10. Damage to adjacent facilities may hamper access to pipeline repair areas.

11. Replacement pipe, fittings, and valves may be unavailable in the quantities required.

Planning for rapid earthquake response needs to be coordinated with local and regional government authorities, as well as key customers. Identifying potential post-earthquake response needs with local and regional governments can minimize delays in accessing damaged portions of the system and problems that may arise with securing a work site.

Also, governmental agencies can provide important information on the earthquake hazard and expected levels of earthquake damage to buildings and services such as electricity, water, sewer, and transportation. Making customers aware of the potential severity and duration of service interruption can allow them to incorporate the information into their own emergency response planning.

Finally, it is not sufficient to simply have an earthquake response plan. Because of the infrequent nature of earthquakes, regular earthquake simulation exercises are necessary to maintain personnel readiness and identify potential planning deficiencies. These exercises should be coordinated with local and regional planning exercises to identify coordination issues and take full advantage of current information on earthquake hazards and other earthquake damage that could potentially jeopardize a rapid response to pipeline damage.

PR-268-9823
2187
SECTION 5: Mitigation Options 5-7
Native Soil
A A
B
45°-φ/2 (typical)

0.3 m Cap, Trench Spoil or Equivalent
A and B selected to assure soil logarithmic spiral
failure surface lies within the trapezoidal trench
Native Soil

0.3 m Cap, Trench Spoil or Equivalent
~60°
Figure 5.1 Use of Loose Granular Backfill to Reduce Soil Loads PR-268-9823
2188
SECTION 5: Mitigation Options 5-8
Caving Active
Wedge
Central Wedge Heaving Passive
Wedge
Logarithmic Spiral
Failure Surface
-20
0 20
40 60
80 0 20 40 60 80 100 120 140 160 180 200 220 240
Horizontal from Pipe Centerline
Vertical from Bottom of Pipe
tan
o r reθ φ =
Y
X
θ
ro φ = 35°
Figure 5.2 Theoretical Logarithmic Spiral Failure Surface PR-268-9823
2189
SECTION 5: Mitigation Options 5-9

OPTIONAL BUMPER TO

RESTRICT MOVEMENT
A. PIPE ON GROUND SURFACE B. PIPE ON GROUND SUPPORT
D. PIPE ON ELEVATED SUPPORT
C. PIPE IN ABOVEGROUND BERM
F. PIPE HUNG FROM ABOVEGROUND SUPPORT
E. PIPE IN ABOVEGROUND CULVERT
Figure 5.3 Aboveground Mitigation Concepts
PR-268-9823
2190
SECTION 5: Mitigation Options 5-10
Native Soil
Trench Spoil or Equivalent
A is selected to provide
the desired angle for the
sliding surface formed by
the dual layer of
geosynthetic.
~0.2
Select Granular Backfill
Around Pipe Dual Layer of Geosynthetic
A A
A. USE OF GEOSYNTHETIC FAILURE SURFACE
If direction of ground displacement is not known,
common for typical precision in locating fault crossings, DIRECTION OF GROUND geofoam wedge required on both sides of the pipe.

DISPLACEMENT
Optional single or double
layer of geosynthetic to
form slip plane and
reduce sliding friction of
geofoam block.

Geofoam wedge to resist lateral
soil loads on pipe. (Also works
with top geofoam block to
reduce axial soil friction forces in
granular material.)
Top geofoam block above
pipe reduces axial load on
pipe. Block should extend
beyond projection of pipe
only if axial load reduction is
of interest (i.e., no geofoam
wedge present).

Soil Backfill
Regular or Select Backfill
Geofoam
B. USE OF GEOFOAM TO REDUCE LATERAL AND AXIAL SOIL LOADS Figure 5.4 Application of Geosynthetic Geofoam Materials to Reduce Soil Loads PR-268-9823
2191
SECTION 5: Mitigation Options 5-11
r/2 r/2
Cover to
Carry
Soil/Surface
Loads
Cover Support
Sacrificial
Culvert
Frangible Pipe
Carrier Support r/2 r/2
Internal Pipe
Carrier
r/2 r/2
Sacrificial
Culvert
Figure 5.5 Culvert Concepts to Isolate Pipelines from Relative Ground Displacement (Ground displacement reduced by “rattlespace”, r)
PR-268-9823
2192

GUIDELINES FOR THE SEISMIC DESIGN

AND ASSESSMENT OF NATURAL GAS AND

LIQUID HYDROCARBON PIPELINES

PART II

COMMENTARY
PR-268-9823
2193
This Page Intentionally Left Blank
2194
Table of Contents
C1.0 GENERAL PROVISIONS ............................................................................... C1-1 C1.1 Application ................................................................................................ C1-1 C1.2 Process ....................................................................................................... C1-1 C1.3 Alternate Procedures.................................................................................. C1-3 C1.4 Required Information................................................................................. C1-3 C2.0 QUANTIFYING SEISMIC HAZARDS.......................................................... C2-1 C2.1 Surface Faulting......................................................................................... C2-1 C2.1.1 Empirical Methods for Estimating Fault Displacements .............. C2-3 C2.1.2 Orthogonal Displacement Components......................................... C2-4 C2.1.3 Design Fault Displacements.......................................................... C2-5 C2.2 Peak Ground Acceleration......................................................................... C2-7 C2.3 Liquefaction............................................................................................... C2-8 C2.3.1 Assessing Liquefaction Potential .................................................. C2-9 C2.3.2 Displacements Associated with Relative Buoyancy ................... C2-10 C2.3.3 Displacements Related to Liquefaction-Induced Settlement ...... C2-11 C2.4 Lateral Spread Movement........................................................................ C2-12 C2.5 Landslides................................................................................................ C2-13 C2.6 Seismic Wave Propagation...................................................................... C2-16 C2.7 Transient Ground Deformation................................................................ C2-18 C3.0 PIPELINE PERFORMANCE CRITERIA .................................................... C3-1 C3.1 New Pipelines and Pipelines Compatible with
Current Construction Practice.................................................................... C3-1 C3.1.1 Performance Goal: Maintain Pressure Integrity........................... C3-5 C3.1.2 Performance Goal: Maintain Normal Operability........................ C3-8 C3.1.3 Load-Controlled Conditions.......................................................... C3-9 C3.2 Considerations for Offshore Pipelines..................................................... C3-10 C3.3 Pipelines Not Compatible with Current Construction Practices.............. C3-11 C4.0 PIPELINE ANALYSIS PROCEDURES ........................................................ C4-1 C4.1 Pipe Element Definition ............................................................................ C4-2 C4.2 Pipe Stress-Strain Definition ..................................................................... C4-3 PR-268-9823
2195
C4.3 Soil Spring Definition................................................................................ C4-3 C4.4 Extent of Pipeline Model........................................................................... C4-5 C4.5 Representation of Applied Ground Movement.......................................... C4-5 C4.6 Sensitivity Analyses................................................................................... C4-6 C5.0 MITIGATION OPTIONS ................................................................................ C5-1 C6.0 REFERENCES .................................................................................................. C6-1 PR-268-9823
2196
SECTION C1: General C1-1
C1.0 GENERAL PROVISIONS
The presentation of these guidelines utilizes a format in which a concise set of general procedures is provided separate from more detailed discussion. This Commentary is meant to provide the necessary background and discussion in order to properly implement the recommendations contained in Part I: Guidelines and Recommended Procedures. This format was selected to provide easy reference for users once they have familiarized themselves with the entire guidelines document.

C1.1 Application
These guidelines should not be interpreted as a set of minimum requirements for a particular application. Instead, the goal is to provide consistency among the definition of seismic hazard, analytical methods, and acceptance criteria. Modifications to the level of analysis or the definition of seismic hazard will necessarily lead to a change in the acceptance criteria.

The acceptance criteria in these procedures are based upon consideration of steel pipe with butt-welded pipe joints. This has been the typical configuration for natural gas and liquid hydrocarbon transmission pipelines for the past 40 to 50 years. These guidelines do not cover piping configurations that include appurtenances such as valves, meters, filters, branch line connections, flange connections and other configurations with deformation limits considerably different from butt-welded steel pipe.

C1.2 Process
The procedures in these guidelines are consistent with the philosophy that the earthquake hazards are defined in a conservative fashion, and the subsequent pipeline assessments are based upon more liberal strain acceptance criteria. With this philosophy in mind, the performance goal for a pipeline defines the expected occurrence rate of the earthquake hazard.

Defining the target performance goal for assessing pipeline response to seismic hazards is a fundamental and critical task. It is incumbent on the pipeline owner to determine an appropriate level of seismic risk considering public safety, service reliability to customers, regulatory agency approval, potential property and environmental damage, and protection of capital investment. The definition of performance goal should be in terms of an acceptable median annual probability for experiencing an unacceptable level of pipeline performance (e.g., continued function or pressure integrity).

Identification and quantification of seismic hazards requires expertise in the fields of geology and seismology. In addition, quantification of the soil properties necessary for the evaluation of pipeline response requires expertise in the field of geotechnical engineering. It is assumed that sufficient information on the seismic hazard and soil conditions is available for use with these procedures. In most cases, estimates of soil PR-268-9823
2197
SECTION C1: General C1-2
strength parameters will require subsurface site investigations and laboratory tests of soil samples.

While the scope of these procedures is directly applicable to earthquake-generated hazards, the procedures can be extended to other non-seismic hazards that produce permanent ground movements along a pipeline alignment such as slope movement and ground settlement provided such hazards are defined in a similar fashion (i.e., the hazards are directly related to the target performance goal). These procedures may also be extended to other pipeline systems (e.g., chemical, water) provided adequate consideration is given to the operational characteristics that differ significantly from natural gas and liquid hydrocarbon pipelines. Such characteristics may include alternate pipe material and alternate pipe construction methods.

The approach in these guidelines differs from typical practices embodied in building codes that rely on partial safety factors to achieve a target level of reliability. The simplified approach in these guidelines is justified by the current inability to reasonably quantify the uncertainty in estimating various seismic hazards, soil loading parameters, and pipeline performance under specific strain levels. For most building codes, conservatism in design methods and specification of safety factors are specifically intended to result in a probability of not meeting a performance objective (e.g., building collapse) that is less than the probability of experiencing the design seismic ground motions. As an example, U.S. building codes are intended to achieve a mean annual probability of building collapse from earthquakes no greater than 4x10-4. The building design is based upon earthquake ground motions with a mean annual probability of exceedance of approximately 1x10-3 to 2x10-3. Thus, use of these guidelines with seismic hazard definitions based upon design values specified in building codes will not generally result in the same level of performance intended for buildings. Instead, to achieve a level of reliability that approximates that for buildings, these guidelines require that the annual probability of exceeding seismic hazards be the same as the probability of not meeting the building performance objective. Using the above example for the U.S. leads to defining the mean annual probability of exceedance for the seismic hazard as 4x10-4.

In actual practice, it is more typical for performance goals to be defined by mean annual probabilities of 1x10-3 to 2x10-3 for experiencing loss of pressure integrity from a seismic event. Adopting lower performance goals compared to what are intended for new buildings is justified considering the relatively low potential for significant safety consequences from damage to natural gas and liquid hydrocarbon pipelines compared to large occupied buildings.

In some cases, pipeline operators may wish to achieve a very high level of reliability for especially critical pipelines. The approach in these guidelines is to introduce additional conservatism by considering less frequent and more severe earthquake hazards. Users are cautioned that this approach may not be appropriate when performance goals dictate that earthquake hazards with mean annual probabilities of exceedance are considerably less than 4x10-4. Caution is required because of the fact that the majority of the available methods for estimating seismic ground displacements are empirical and based upon a limited set of historical observations. Because of this, extrapolation to extremely rare PR-268-9823
2198
SECTION C1: General C1-3
seismic events can be highly imprecise. If it is necessary to achieve extraordinarily high levels of reliability, the necessary conservatism needs to be achieved through a combination of a low likelihood of experiencing a seismic hazard and a high likelihood (higher than considered in these guidelines) of achieving specific pipeline capacity without compromising pipe performance (e.g., maintaining operability or pressure integrity). Establishing appropriate levels of conservatism in hazard definition and pipeline capacity will typically require a more formal formulation incorporating state-ofthe-art methods that are beyond the scope of these guidelines.

C1.3 Alternate Procedures
The relationships for soil loads and pipeline limit states are based upon information that can be obtained readily through available construction records or standard geotechnical investigations. Field and laboratory tests of actual pipelines and soil conditions are generally much better representations of actual behavior. However, specific tests are not typically performed because of the relatively high costs and difficulty in extrapolating data from a few locations to represent conditions present over long sections of the pipeline alignment. For these reasons, the procedures have been formulated based upon what is believed to be readily available information.

Nonlinear, large-displacement finite element analysis, with the pipeline represented by beam-like elements with nonlinear stress-strain properties, is the recommended approach.

In some cases, bounding analyses using simplified calculation procedures may be sufficient to assess pipeline response. It is also possible to perform more complex finite element analyses with the pipeline represented by shell-elements. Such refined analyses are not prohibited by these guidelines and have been shown to provide good correlation with pipeline behavior in controlled tests. However, it should be recognized that the methods described in these guidelines for determining soil restraint boundary conditions do not apply to finite element models represented by shell elements. Furthermore, it is doubtful that shell element soil restraint properties could be developed without extensive, full-scale test programs. Consequently, the use of shell element models for the pipeline should be restricted to investigations of localized response of the pipeline where nonlinear deformations (e.g., wrinkling, buckle formation) are expected.

C1.4 Required Information
The information requirements listed are based upon data necessary to implement the procedures in these guidelines. The sources of this information may include mill tests, laboratory tests, geotechnical investigation, and seismic hazard assessment.

PR-268-9823
2199
SECTION C2: Seismic Hazard Quantification C2-1
C2.0 QUANTIFYING SEISMIC HAZARDS
Seismic hazards affecting buried pipelines are divided into two categories: permanent ground deformations (surface faulting, lateral spread movement, subsidence, landslide) and dynamic strains from wave propagation.

The response of buried pipelines to inertial shaking effects (i.e., the pipeline vibrating within the soil) are not considered in these guidelines, as past experience and prior studies show that these do not have a significant effect on pipeline stress or strain (Nelson and Baron, 1981; Parnes and Weidlinger, 1979).

C2.1 Surface Faulting
Faulting is the deformation associated with the relative displacement of adjacent parts of the earth's crust. Fault displacements can occur rapidly during an earthquake (duration on the order of one to several seconds). In addition, relatively minor displacements may accumulate gradually over many years as aseismic creep.

Faulting that results in surface rupture is an important consideration for buried onshore pipelines, because pipelines crossing fault zones must deform longitudinally and in flexure to accommodate ground surface offsets. For offshore pipelines that are simply laid on top of the seafloor, fault movements are generally of little, if any, consequence.

Due to the unrestrained pipeline configuration, strains induced by fault movement will be only a small fraction of those that would be developed in a buried pipeline with similar dimensional characteristics and properties. It is possible that offshore faulting could produce a vertical offset that would cause a segment of the pipeline to be elevated above the seafloor for a short distance, and, hence, cause concern for vortex-induced oscillations from water currents. It is not practical to attempt to design for such a remote possibility in advance, particularly considering that fatigue damage to the pipeline would not occur for some time, probably years. For offshore pipelines that are intentionally buried (typical for pipelines in less than 100 feet of water) or are expected to become buried as a result of sea bottom sediment transport processes, the treatment of offshore pipelines mirrors that for onshore pipelines.

Faults are classified on the basis of their direction of movement with respect to the ground surface. The various types or classifications of fault movement are illustrated schematically in Figure C2.1. A strike-slip fault is one in which the predominant component of ground movement is horizontal displacement. If the movement of one side of the fault when viewed from the other side is to the right, the fault is called a rightlateral strike-slip fault. When the movement is to the left, the fault is called a left-lateral strike-slip fault. Normal-slip and reverse-slip faults are those in which the overlying side moves downward and upward, respectively, in relation to the underlying side of the fault.

In many cases, faults exhibit a combination of strike-slip and normal or reverse-slip movements.

PR-268-9823
2200
SECTION C2: Seismic Hazard Quantification C2-2
The effect of right-lateral strike-slip movement on a buried pipeline is illustrated in Figure C2.2. For some faults, the maximum credible surface movements can be as large as 10 meters for strike-slip faults and 5 meters for normal-slip faults. In some cases, ground movements for a particular fault may occur over a number of closely-spaced parallel fault traces or splays within a fault zone, and consideration must be given to the total as well as the individual movements within the zone.

The evaluation or design of a pipeline crossing a fault generally requires the following information:

1. Location of the pipeline-fault intersection and error bounds on location (e.g., ±50 meters).

2. Expected allocation of fault slip among multiple fault splays, if applicable.

3. Angle of intersection of the fault with the pipeline alignment at the point of crossing.

4. Type of slip expected (e.g., strike-slip, normal-slip or reverse-slip).

5. Width of the zone of disturbed ground. For strike-slip faults, the zone of ground distortion will be on the order of 2 to 5 meters. For normal or reverse faults, the principal zone of surface disturbance is typically on the order of 5 to 15 meters. For reverse faults, secondary surface cracking and heaving may occur over a distance of as much as 200 meters on the hanging wall block of the fault.

6. Fault dip angle.

7. Design fault displacement (defined as orthogonal components of displacement in three-dimensional space).

Generally, a fault is considered active if it can be demonstrated to have displaced the surface of the ground during the Holocene Epoch, i.e., within the past 11,000 years. This, of course, assumes that faults with evidence of such historic movement will be the ones that move again. This has not always been the case. The 1999 Hector Mine earthquake in southern California occurred on a fault that was shown on fault maps to be “inactive,” albeit some geologists believe that the fault rupture was closely tied to an adjacent active fault system.

In some jurisdictions, fault activity may be defined by regulatory bodies. In California, the Alquist-Priolo Special Studies Zones Act of 1972 provides special regulations regarding land uses near faults that are “well defined” and “sufficiently active.” In this context, well-defined implies that the fault can be detected by a trained geologist as a physical feature, at or just below the ground surface. “Sufficiently active” is defined as having evidence of Holocene surface displacement along one or more segments.

Unfortunately, there are no reliable means for predicting the probability of future fault displacements. The basic mechanisms that govern such phenomena are measured in terms of thousands of years, i.e., orders of magnitude greater than the design life of the PR-268-9823
2201
SECTION C2: Seismic Hazard Quantification C2-3
pipeline. Consequently, for the evaluation or design of pipeline crossings of active surface faults, it is necessary to utilize judgment based upon experience and consideration of the best available data.

The preferred method for estimating fault displacement is to rely on historical data on fault displacement or the findings from site-specific studies that may include examination of topographic features, observations of geomorphic features, or trenching across the fault to identify historical evidence of past fault displacement. For most pipelines, this preferred approach is only practical where site-specific studies have previously been performed for the purposes of refining regional seismic hazard estimates. In lieu of sitespecific information, estimates of fault displacement often rely on empirical methods tempered by the judgments of individuals experienced in their development and limitations.

C2.1.1 Empirical Methods for Estimating Fault Displacements A number of investigators have compiled statistics regarding historic fault displacements as a function of earthquake magnitude, fault rupture length, and other related parameters.

The most recent compilation was done by Wells and Coppersmith (1994) using a worldwide data base of source parameters for 421 historical earthquakes. In their study, Wells and Coppersmith developed a series of empirical relationships among moment magnitude, surface rupture length, subsurface rupture length, down-dip rupture width, rupture area, and maximum and average displacement per event.

The data sets for the two regression equations given in these guidelines include displacement data for all types of slip, i.e., strike-slip, normal-slip and reverse-slip faulting. Regression equations that are unique for particular slip types are also available in Wells and Coppersmith, but the correlation is unsatisfactory for reverse faults, probably due to the difficulty in making reliable field observation of reverse-slip movements. Wells and Coppersmith (1994) recommended using the regressions for an aggregate of “all” types of slip.

The surface rupture length is not necessarily the entire length of the fault trace visible on the surface but, rather, it is the maximum length of the fault segment crossing the pipeline that, by reasonable and qualified judgment, can be expected to rupture during an earthquake. Such judgments should be made by geologists with knowledge of regional seismotectonics and with reference to site-specific information and studies. As an upperbound estimate of fault displacement, it may be appropriate to assume that the fault can rupture along its entire length. However, this could lead to unnecessarily conservative design displacements.

In regions where probabilistic or deterministic ground shaking hazard maps have been developed, information on fault length may be readily available (e.g., consider the fault map data provided for use with the 2000 International Building Code in the U.S.) Other information sources include the U.S. Geological Survey (U.S.), Canadian Geological Survey, Pacific Geological Survey (British Columbia, Canada), and the California PR-268-9823
2202
SECTION C2: Seismic Hazard Quantification C2-4
Geologic Survey. In the absence of published information, it may be necessary to conduct geologic field studies in order to estimate fault length. Surface expression of fault movement can be considered unlikely for earthquakes with moment magnitudes less than 6.0.

The potential variability in the empirical expressions for fault displacement can be illustrated by a sample calculation. Using equations (2-1a) and (2-1b) for a surface rupture length of 75 km yields mean maximum and mean average fault displacements of

3.4 m and 1.7 m, respectively. Using equations (2-2a) and (2-2b) with a moment magnitude of 7.5, yields mean maximum and mean average fault displacements of 4.9 and 2.4 m, respectively. For the case of a fault with a 75 km surface rupture, the maximum and average fault displacements within one standard deviation of the mean are

1.3 m to 8.7 m and 1.8 m to 7.6 m, respectively. The range is somewhat larger for fault displacements estimated based upon earthquake magnitude. Thus the empirical formulae produce a range of fault displacements that vary by a factor of about 4 to 7 in order to capture the actual fault displacement slightly better than half of the time. For this reason, projects that pose an extreme threat to safety or the environment (e.g., large LNG facilities, nuclear power plants) typically require site-specific investigations to better define fault displacement. Site-specific investigations may include trenching, soil borings, and paleoseismic studies to establish the approximate dates of prior fault displacement and the amount of displacement that has occurred in past earthquakes. For most pipeline projects, however, this level of effort is not justified by the potential consequences of earthquake damage.

C2.1.2 Orthogonal Displacement Components
For the design of a pipeline crossing a fault, the relative displacement of the soil block on one side of the fault must be known in a three-dimensional sense relative to the soil block on the other side. The user should be aware that it is not uncommon for geologists to define fault displacement in terms of the resultant two-dimensional displacement vector in a vertical plane parallel to the strike of the fault. In other words, the fault displacement is the resultant of the observed vertical offset of the ground surface and horizontal displacement parallel to the fault.

If the fault is known to be predominantly strike-slip, it usually is sufficient to assume that the resultant displacement will be parallel to the fault strike. However, in the case of reverse faults, the nature of the fault displacement may be highly variable due to an irregular (nonlinear) surface trace such that the displacement at the pipeline crossing may be a combination of horizontal, vertical and compressional-extensional displacement components relative to the strike of the fault.

Fault displacement estimated using the Wells and Coppersmith regression equations is the projection of a resultant fault displacement on a vertical plane that intersects the fault strike at the surface. The database for the regression equations does not include the vertical and horizontal components of displacement in the vertical plane that intersects PR-268-9823
2203
SECTION C2: Seismic Hazard Quantification C2-5
the strike, nor does it include the transverse component of reverse-slip or normal-slip components of displacement perpendicular to the fault strike.

Empirical approaches, other than those developed by Wells and Coppersmith, are available for estimating fault displacement. These guidelines do not prevent the use of these approaches but encourage the use of alternate relationships that may be more appropriate for specific regions or fault systems.

The recommended approach to determine three-dimensional fault displacement components from fault displacement defined in a manner similar to the Wells and Coppersmith regression equations requires some estimate of the fault geometry and regional seismic stress regime as defined by (1) the angle of intersection between regional stress azimuth and fault strike, β, and (2) the fault dip angle, δ.

The displacement of a hypothetical fault displacement is shown in Figure C2.3. The net displacement in a vertical plane parallel to the fault strike is the resultant of two of the three orthogonal components of fault displacement, i.e., the horizontal component of slip, HF, parallel to and the vertical component, VF, perpendicular to the strike of the fault.

The individual horizontal and vertical displacement components, HF and VF, as well as the transverse component of fault displacement perpendicular to the fault strike, TF, are unknown.

If it is assumed that the total fault displacement, DFS, is in the direction of the regional stress azimuth, the total fault displacement and its constituent orthogonal displacement components can be determined from the geometry of right triangles. The resulting relations for the three orthogonal components are given in equations (2-3b), (2-3c) and (2-3d) in these guidelines.

The values for the fault dip angle provided with equation (2-3) are approximate values that are believed to be generally representative of the database used in the development of the Wells and Coppersmith empirical relationships. It is possible that empirical relationships developed for specific seismic regions are available. In this case, the generic fault dip angles provided with equation (2-3) may not be appropriate. In general, the determination of fault movement components is best left to an experienced geologist having knowledge of local plate tectonics and the regional stress regime in the earth’s crust.

C2.1.3 Design Fault Displacements
Recommended design fault displacements are based upon the potential consequences associated with pipeline damage and defined in terms of the expected maximum or average fault displacement consistent with the probability of exceeding pipeline performance criteria established by the pipeline owner. The process for determining design fault displacement consists of three basic steps:

PR-268-9823
2204
SECTION C2: Seismic Hazard Quantification C2-6

1. Determine the performance requirements for the pipeline in terms of an average annual probability of exceeding strains associated with maintaining normal operability or pressure integrity;

2. Determine the average and maximum fault displacement consistent with the probability established in Step 1; and

3. Determine design fault displacement values by modifying the fault displacements in Step 2 based upon the consequences associated with the pipeline not meeting its performance objectives, the average probability of the occurrence of fault displacement, and the acceptable annual probability for the pipeline exceeding its performance objectives.

Pipeline performance requirements are established by the pipeline owner and will typically vary depending upon the factors discussed in Section C1.2.

The maximum fault displacement that occurs during a seismic event typically occurs along a short length of the fault rupture, with lesser amounts of movement along most of the fault rupture length. For faults with no information from historical rupture patterns or detailed fault studies, estimating the distribution of maximum displacements along the fault rupture length is extremely uncertain. Since that maximum is generally considerably greater than the average displacement (about double), it is reasonable to use a value smaller than the maximum credible displacement when developing design values for engineering design at active fault crossings.

Determination of fault displacements consistent with the performance goals of the pipeline requires considerable judgment, particularly for faults where the episodic nature of earthquake activity has not been investigated. Reduction of fault displacement estimates for design are warranted when the average probability of surface rupture is less than the performance goals of the pipeline. The approach adopted in these guidelines is to simply factor the fault displacement estimates by the ratio of the acceptable probability for the pipeline not meeting its performance goal by twice the average probability associated with the fault displacement estimate. The factor of 2 on the average probability of fault displacement can be reduced if there is good information on fault occurrence based upon historical records or fault trenching investigations.

The approach for estimating design fault displacement in these guidelines implicitly accepts the risk associated with uncertainties in understanding fault behavior. Some of these uncertainties include the following:

1. Actual displacement on a fault may be considerably greater than that produced by earthquakes of similar magnitude.

2. The likely fault displacement, based upon estimates of fault slip rate and expected earthquake magnitude (the most common approach), is not a reliable means to assess the difference in the probability of fault displacement in the near future (i.e., 50-year to 100-year time frame).

PR-268-9823
2205
SECTION C2: Seismic Hazard Quantification C2-7
C2.2 Peak Ground Acceleration
Estimates of peak ground acceleration (PGA) are necessary in order to estimate potential seismic hazards related to liquefaction, lateral spread movement, and slope stability.

Peak ground accelerations are estimated using empirical attenuation equations that relate the dissipation of earthquake energy with distance from the source of the earthquake.

Attenuation relationships utilize data recorded from strong motion accelerographs. The acceleration time history data are typically segregated according to the local site conditions on which the accelerograph is founded, such as hard rock, soft rock, firm soil, and soft soil. In addition, many recent attenuation relationships segregate the data by the type of faulting (e.g., strike-slip, normal-slip, reverse-slip). Attenuation relationships also have been developed to represent spectral accelerations at various frequencies. However, only PGA is of importance for use in these guidelines.

All attenuation relationships require as input the earthquake magnitude and a measure of distance from the earthquake source. The selection of a particular measure of magnitude (e.g., moment magnitude, surface wave magnitude, and local magnitude) and distance (e.g., epicentral, hypocentral, and shortest distance to fault plane) is largely a matter of personal preference on the part of the individual developing the relationship. Figure C2.4 provides a definition of earthquake source parameters often incorporated into attenuation relationships.

There is typically a great deal of scatter in the recorded ground motion data for a particular distance from the earthquake source. Directivity associated with the orientation of fault rupture is believed to be a significant factor in the observed scatter.

Another factor in the amount of observed scatter is local interaction of seismic waves as they propagate to a single site from multiple locations along the fault rupture surface.

This scatter introduces uncertainty in the attenuation relationships that is represented by the random-error term E. The random-error term is typically a function of earthquake magnitude.

Setting E to zero results in a best estimate of the fit of the attenuation equation to the data. The random-error term is used to estimate the probability that accelerations might be different than that estimated from the best-estimate equation. In application, the random-error term is multiplied by a factor representing the number of standard deviations from the mean producing the desired standard normal probability value. Thus, using 1(E) produces an estimate of PGA that envelops 84% of the data, 1.65(E) produces an estimate of PGA that envelopes 95% of the data, and 2(E) produces an estimate of PGA that envelops 98% of the data.

Some measure of the precision with which attenuation relationships estimate PGA can be gained by examining the factors representing ±25% variation from the mean. This corresponds to the factors on the mean that would provide a range of accelerations that have a 50% chance of occurring. A mean plus 25% estimate corresponds to an estimate that envelops 75% of the data or is 0.68 standard deviations from the mean. A mean minus 25% estimate corresponds to an estimate that envelops 25% of the data or is –0.68 PR-268-9823
2206
SECTION C2: Seismic Hazard Quantification C2-8
standard deviations from the mean. For the two attenuation relationships provided, the random-error terms for a magnitude 7 earthquake are 0.41 for the near-source earthquake equation and 0.75 for the Cascadia earthquake. The corresponding factors on the mean estimate to give a 50% chance of experiencing the estimated PGA are computed as follows:

Near-Source: Factor = e0.68(0.41) = 1.32
Cascadia: Factor = e0.68(0.75) = 1.67
Assuming the PGA estimate was 0.5 g, the range of PGA values having a 50% chance of occurrence is 0.38 g to 0.66 g for the near-source earthquake and 0.30 g to 0.84 g for the Cascadia earthquake. This example illustrates the wide variability typically encountered in estimating seismic hazards. It is also clear that stating estimates of PGA within increments of 0.05 g or greater is generally sufficient.

Since attenuation relationships are developed from recorded ground motion data, users should attempt to identify attenuation relationships developed for the specific region of interest. If such relationships cannot be identified, the selected attenuation relationship should be based upon records from regions that are tectonically and geologically similar.

Often, multiple attenuation relationships may be identified as appropriate for a particular location. Where possible, it is recommended that peak ground acceleration be estimated as the average of at least three representative attenuation relationships.

A good source of information for attenuation relationships applicable to the United States is Seismological Research Letters (SSA, 1997). This reference provides a collection of recently developed attenuation relationships and is a starting point for identifying attenuation relationships for locations other than the west coast of the United States. For other parts of the world, it is generally necessary to identify applicable attenuation relationships from past earthquake investigations, seismic hazard analyses, or accepted practice in past estimates of seismic ground motion hazards for the region.

C2.3 Liquefaction
Liquefaction is the transformation of a saturated cohesionless soil from a solid to a more liquid state as a result of increased pore-water pressure and concomitant loss of shear strength. Loss of shear strength gives rise to bearing failures and large deformations in surface structures founded on liquefied soil. Liquefaction often leads to the formation of sand boils, mud volcanoes, fissures and other channels through which water and sediments are ejected. These ejections cause volume loss resulting in differential settlement even though no significant lateral movement occurs.

Liquefaction is a phenomenon that is generally limited to saturated sandy soils that have been deposited in the Holocene (the past 11,000 years) and most commonly to fluvial sediments deposited in the past 1,500 years. These characteristics typically limit consideration of liquefaction to locations where the water table is less than 15 m below PR-268-9823
2207
SECTION C2: Seismic Hazard Quantification C2-9
the ground surface. Liquefaction is most likely to occur within a kilometer of bodies of water (e.g., shorelines, streams, drainage channels) and within deltaic deposits where the water table is typically shallow.

Existing maps and databases pertaining to surficial geology should be used to identify areas with soils of depositional characteristics and geologic age that are known to have liquefied in previous earthquakes. The procedures proposed by Youd and Perkins (1978) should be followed. Table C2.1 (Youd and Perkins, 1978) summarizes the liquefaction susceptibility of various deposits according to geologic age. For most pipeline projects, it is only necessary to consider sedimentary deposits with high or very high liquefaction susceptibility.

C2.3.1 Assessing Liquefaction Potential
Collecting information necessary to assess the potential hazards from liquefaction requires site-specific subsurface data. Unless adequate subsurface data are available from past projects at the site of interest, site investigations consisting of subsurface sampling, standard penetration tests, cone penetrometer tests, and laboratory tests are required.

Assessing liquefaction potential is based upon both peak ground acceleration and earthquake magnitude. Estimating peak ground acceleration at the site of interest can be performed using probabilistic or deterministic approaches. The deterministic approach is based upon selecting a magnitude for a particular fault and using attenuation relationships that relate peak ground acceleration to magnitude, distance, local soil conditions, and type of earthquake faulting. The probabilistic approach is similar to the deterministic approach only in the sense that attenuation relationships are used to relate peak ground acceleration to earthquake magnitude, distance, and type of earthquake faulting.

In a probabilistic approach, the effects of numerous earthquake sources, each with different rates of occurrence for earthquakes of various magnitudes, are included.

Additional information on probabilistic seismic hazard estimation is provided in Appendix C. The results of a probabilistic approach for estimating PGA are typically represented in maps of PGA contours for different annual probabilities of exceedance (Figure C2.5) or site-specific curves representing the variation of PGA with annual probability of exceedance (Figure C2.6). For liquefaction assessment, it is necessary to examine the contribution of earthquake magnitudes and distances to the ground shaking hazard. This is termed deaggregation of the hazard. The deaggregation of the hazard is often represented by plots of magnitude and distance contributions such as those shown in Figure C2.7. There are no established rules for selecting an appropriate magnitude and distance from a deaggregation of the probabilistic hazard analysis. Three common approaches are to select either the mean contribution, estimates enveloping mean plus one standard deviation (84%) of the contributions, or a magnitude and distance combination that has the largest contribution to the probabilistic hazard. These approaches are illustrated in Figure C2.7.

PR-268-9823
2208
SECTION C2: Seismic Hazard Quantification C2-10
Assessment of potential liquefaction hazards requires site-specific information on the density and grain-size distribution of subsurface soil layers. As a general consideration, site-specific investigations should be performed at the types of areas listed in Table C2.1 as having moderate to very high risk of liquefaction where the peak ground acceleration is expected to be greater than about 0.15 g.

The procedures presented for assessing liquefaction are based upon the findings of a 1997 workshop (Youd and Idriss, 1997). The proceedings of this workshop provide an excellent summary of the state-of-practice in assessing liquefaction hazards. Two changes in the manner in which liquefaction is assessed resulting from the workshop deserve mention. First, the workshop participants agreed to modify the hand drawn curve developed by Seed and Idriss (1982) to reflect a minimum CRR value of 0.05. The modified curve for which the equations in these guidelines are developed is shown in Figure C2.8. Second, the workshop participants agreed that the magnitude scaling factors originally developed by Seed (Seed and Idriss, 1982) were far too conservative for earthquakes with magnitudes less than 7. Although specific magnitude scaling factors were not recommended by the workshop participants, the recommended range of magnitude scaling factors represent an increase of 50% to 100% in the CRR.

C2.3.2 Displacements Associated with Relative Buoyancy
Vertical pipeline movement resulting from buoyancy forces has not been a significant hazard to buried onshore pipelines in past earthquakes. The combination of limited displacement potential and the length of pipeline typically exposed to liquefaction generally prevents the development of significant pipeline strains.

Onshore pipelines located above the water table are not subjected to relative buoyancy forces. Pipelines located beneath the water table in liquefaction-susceptible soils can experience vertically upward or downward displacements depending upon the weight of the pipeline with coating and contents, the weight of liquefied soil displaced by the pipeline, and the residual shear strength of the liquefied soil. The time over which the soils remain in a liquefied state is a key variable to determining the amount of vertical movement. No studies to quantify the duration of liquefaction and its effect on movements caused by differential buoyancy could be identified. As a first approximation, the duration of liquefaction can be assumed to be equal to the duration of ground motion sufficient to cause liquefaction, another poorly defined parameter. The procedure in these guidelines is limited to a means to identify the maximum potential movement of the pipeline.

Practically, relatively extreme conditions of buoyancy mismatch appear to be necessary to produce relative pipe movement. This is illustrated by the following example.

Consider an NPS 24 pipe with a wall thickness of 9.5 mm (0.375 inches). The weight of the pipe is 1.42 kN/m (97 lb/ft) and the effective density of the pipe is 4.9 kN/m3 (31
lb/ft3
). If the total unit weight of liquefied soil is 17.4 kN/m3 (110 lb/ft3), the net buoyancy force on the pipe is 3.62 kN/m (248 lb/ft) in the upward direction. It is PR-268-9823
2209
SECTION C2: Seismic Hazard Quantification C2-11
assumed that the depth to the centerline of the pipe is 2 m (6.56 ft), the water table is at the ground surface, and the residual shear strength of the liquefied soil is 5% of the effective stress (typical range is 1% to 10%). These assumptions lead to a residual shear strength of only 0.976 kPa (20.4 psf) and a soil resistance to vertical movement of 12.9 kN/m (884 lb/ft) based upon the approaches in Section 4.3.3. Since the soil resistance is about 3.6 times the upward buoyancy force, no movement would be expected.

Given this example, it is expected that vertical movements from pipeline buoyancy will only be a significant hazard for large diameter pipelines within soils expected to exhibit unusually low residual shear strength in a liquefied state. Calculations, similar to those in the above example, can be used to gauge the potential for significant pipeline vertical movement from buoyancy forces.

Evaluating displacements from relative buoyant forces for underwater pipelines can be treated in the same fashion as onshore pipelines. However, buoyancy is typically not an issue for offshore pipelines or inland water crossings as they are required to be negatively buoyant to remain stable on the sea floor. Achieving negative buoyancy is typically done by coating the pipeline with concrete or using discrete weights attached to the pipeline. If buoyancy is an issue, additional conservatism may be warranted to account for the fact that upward movement sufficient to uncover the pipeline will expose it to potential hazards related to flow-induced oscillation and impact with debris carried by water currents (e.g., exposed river crossings).

C2.3.3 Displacements Related to Liquefaction-Induced Settlement Although a methodology is presented for estimating the amount of settlement related to liquefaction, assessing pipeline performance also requires an estimate of the extent of liquefaction along the pipeline alignment. Estimating the extent of soil layers susceptible to liquefaction can best be determined from subsurface profiles produced from boring log data. Borings may be drilled at regular intervals along the pipeline alignment to identify the limits of liquefaction. In some locations, detailed maps may be available describing surficial soil deposits and their depositional environments. These may be used for preliminary screening to identify locations more prone to liquefaction.

It is possible that potentially liquefiable deposits extend for several hundred meters along the pipeline. As it is unlikely, based upon historical observations, that the pipeline will experience uniform liquefaction over large distances, some alternate approach is needed to define the extent of a pipeline subjected to relative settlement. This requires the judgment of individuals familiar with the local geology and performing liquefaction assessments.

The graphical methods presented are based upon both observational data (Figure 2.2 from Ishihara, 1985) and laboratory tests (Figure 2.3 from Ishihara, 1990). Both graphs are based upon sands with a fines content less than 5% (“clean” sands). For engineering purposes, this factor can often be ignored as the determination of what constitutes a liquefiable layer and the factor of safety against liquefaction incorporates the presence of PR-268-9823
2210
SECTION C2: Seismic Hazard Quantification C2-12
fines content. Applying the charts for other soil types generally overestimates settlement since the presence of fines would tend to fill some voids and reduce the amount of soil grain reorganization responsible for liquefaction-induced settlement.

C2.4 Lateral Spread Movement
Lateral spreads involve the horizontal movement of competent surficial soils due to liquefaction of an underlying deposit (see Figure C2.9). Lateral spreads can be especially destructive to buried pipelines, albeit the degree of damage depends on the magnitude and extent of ground movement and the configuration of the pipeline. Offshore, liquefaction requires special treatment as the amount and extent of ground movements associated with liquefaction can be much greater than for onshore installations.

Lateral spread movement is the most common and one of the most severe earthquake hazards for buried pipelines. Techniques for estimating the magnitude of lateral spread movement have only been developed in the past 15 years.

The simplified approach developed by Youd et al. (1999) based upon multiple linear regression of data from past earthquake observations represents the best available method for estimating lateral spread movement outside of explicit finite element techniques. This approach is a modification of a similar method developed by Bartlett and Youd (1992).

The 1999 modifications included data from the 1995 Kobe, Japan earthquake and corrected some of the data records in the 1992 investigation. The expected variability in the simplified approach can be assessed by comparing the predicted lateral spread displacements with the observed displacements. This is provided in Figure C2.10 where it can be seen that nearly all data points are within a factor of two of the predicted displacement.

One weakness of the Youd et al. (1999) approach is its inability to provide information on the extent of lateral spread movement and the variation of ground movement with depth and within the body of the lateral spread. Because of the importance of the spatial variation in ground displacement along the pipeline, the recommended approach for estimating lateral spread movements is to employ an analysis program based upon a finite-element, finite-displacement, or discrete-element representation of the soil. Some examples of programs that have been used in past pipeline projects include DESRA-2 (Lee and Finn, 1978), TENSI-M (Larkin and Marsh, 1982), and SOILSTRESS. Many geotechnical engineering firms also maintain proprietary in-house analysis programs. A comparison between the results of the Bartlett and Youd appoach and the results from a SOILSTRESS analysis are shown in Figure C2.11 for a hypothetical river crossing. In the analyses represented in Figure C2.11, the Bartlett and Youd approach was used to benchmark the maximum surface displacements estimated by the SOILSTRESS analysis.

With this benchmark established, the SOILSTRESS results are then used to determine the variation with respect to depth below the ground surface for relative ground movements normal to the river.

PR-268-9823
2211
SECTION C2: Seismic Hazard Quantification C2-13
C2.5 Landslides
Landslide activity most commonly triggered by earthquake shaking includes rock falls, disrupted soil slides, rock slides, soil slumps, soil block slides, and soil avalanches (Varnes, 1978). Several common types of landslides are illustrated in Figure C2.12.

Except for rock avalanches, locations susceptible to earthquake triggered landslides are most often associated with areas of past landslide activity. The most significant landslide hazards that can affect buried pipelines are slumps, shallow slides, and deep slides.

Slides can occur onshore on relatively steep slopes or offshore in soft sediments on relatively shallow slopes. Landslides may develop into debris flows and impart large dynamic loads on the pipeline. For offshore pipelines, turbidity currents created by the slide may also lead to significant loads on the pipeline.

Slumps and shallow slides are caused primarily by inertial forces, but are often assisted by densification of loose soil or liquefaction of underlying sediments. These movements occur mostly along the margins of embankments, cut-and-fill slopes, and slopes with relatively shallow cover in hilly or mountainous terrain.

Deep slides involving significant components of translation and rotation of a soil mass often develop catastrophically and affect large areas. A landslide frequently causes underthrusting in soils near the base of its slope, so that substantial compression and bending may be transferred to pipelines located there.

Areas of past landslide activity can be identified through reconnaissance of the pipeline alignment and stereo aerial photographs or sidescan sonar for offshore pipelines.

Earthquake-triggered landslides typically occur during the earthquake. If the earthquake forces are insufficient to completely develop a failure plane, earthquake-related landslides may occur several days following the earthquake due to reduced resistance to slide movement.

Reconnaissance of a pipeline alignment can be used to identify and estimate the relative activity of existing slides. The findings from the site reconnaissance can be used to modify new pipeline alignments to avoid many potential landslide hazards. Determining the amount of expected deformation requires a site-specific determination of the existing soil strength properties and an assessment of the factor of safety under seismic loading conditions.

In lieu of a detailed assessment of landslide potential, an approximate estimate of potential landslide movement can be made based upon the existing slope and a general description of the near-surface material. This approximate approach is implemented in these guidelines and is to be used as a screening tool to determine the potential for slide movement.

The approach to estimating the potential severity of landslides is based upon bounding relationships for sliding-block displacement developed by Newmark (1965) and estimated levels of acceleration associated with slope instability developed by Wilson and Keefer (1985). The approach to estimating landslide movement developed by PR-268-9823
2212
SECTION C2: Seismic Hazard Quantification C2-14
Wilson and Keefer utilizes two relationships based upon Arias intensity, one for the acceleration levels associated with 2 cm or 10 cm of landslide movement and the other for the variation of acceleration as a function of earthquake magnitude.

Wilson and Keefer adopted the use of Arias intensity as a means to avoid explicit calculation of movement using multiple time histories and the methodology developed by Newmark. The Wilson and Keefer approach is not considered generally applicable because of the lack of validation of the Arias intensity relationships for a large number of historical earthquakes.

To provide a general indication of the potential magnitude of landslide movements, a bounding empirical relationship developed by Newmark is used. The analytical model used by Newmark is quite simple. A block resting on a sloping surface is prevented from sliding by the friction force between the block and the surface. Motion of the block is produced when acceleration of the block in the downslope direction exceeds the frictional resistance reduced by the component of block weight in the downslope direction. The factor of safety against block sliding is equal to the resistance to sliding divided by the acceleration acting on the block. The sliding resistance in the downslope direction is the frictional resistance minus the component of block weight in the downslope direction.

The upslope movement of the block can occur if the acceleration in the upslope direction exceeds the sum of the friction resistance and the component of block weight in the downslope direction. For the simplified approach used in these guidelines, the Newmark model is identical to the model assumed by Wilson and Keefer. In fact, Wilson and Keefer used explicit analyses using Newmark’s approach to develop their approach.

The duration of motion is limited to the time period over which the acceleration exceeds the friction force. Newmark analyzed the movement of a hypothetical block using 5 earthquake time histories normalized to a peak ground acceleration of 0.5 g. These analyses did not consider vertical motions and assumed the horizontal earthquake acceleration acted in a direction parallel to the sliding surface. He then examined the fit of several possible analytical expressions to the data, as illustrated in Figure C2.13, and developed several bounding relationships based upon ground velocity, peak ground acceleration and the factor of safety. The equations for estimating landslide displacement in equation (2-18) is based upon an assumption of peak ground velocity equal to 36 and 48 inches per second per g (91 and 122 cm per second per g) for rock and soil sites, respectively. The bounding displacement equation used to derive equation (2-18) is given by equation (C2-1).
2 1
2 V NP D
gN PGA N
⎛⎞ =− ⎜⎟ ⎝⎠
GA (C2-1)
where:

D = computed displacement
V = peak ground velocity
PGA = peak ground acceleration
PR-268-9823
2213
SECTION C2: Seismic Hazard Quantification C2-15
N = resistance coefficient = Ac
g = acceleration of gravity
Rearranging the terms and substituting Ac for N and V as x(PGA) gives:
22 1
2 cc
x PGA PGA D
gA A
⎛⎞ = ⎜
⎝⎠
− ⎟ (C2-2)
Equation C2-2 depends on having a velocity ratio, x, the ratio of the peak ground velocity to PGA. Several investigators have proposed such relationships and representative values are provided in Table C2.2. Equations (2-15) and (2-16) result from substituting values of 36 inches per second per g (0.91 meters per second per g) for rock and 48 inches per second per g (1.22 meters per second per g) for the velocity ratio, x.

The displacements necessary to initiate coherent and disruptive slides (10 cm and 2 cm, respectively) are adopted from Wilson and Keefer. If the displacements computed using the relationships in these guidelines exceed these threshold values, the potential for landslides exist, and additional geotechnical investigation is warranted. Such investigations typically involve borings to identify different soil deposits, sample retrieval for ascertaining soil strength properties, and slope stability calculations to estimate the amount of landslide movement.

A pseudostatic analysis of slope stability is often used in the assessment of slope movements. The pseudostatic analysis should account for the strength properties of the slope material expected during the earthquake. Pseudostatic analyses represent the effects of an earthquake by applying static horizontal and/or vertical accelerations to a potentially unstable mass of soil. The inertial forces induced by these pseudostatic accelerations increase the driving forces and may decrease the resisting forces acting on the soil. Pseudostatic analyses are not appropriate for soils that build up large pore pressures or show more than about 15% degradation of strength due to earthquake shaking. Stability is expressed in terms of a pseudostatic factor of safety calculated by limit equilibrium procedures. Selection of an appropriate pseudostatic acceleration requires great care; values considerably smaller than the peak acceleration of the sliding mass are usually used.

If conditions exist where such strength loss may occur, it is generally necessary to perform a local site response analysis in order to obtain an estimate of the shear strains produced and the appropriate soil strength for slope stability analysis. Inertial instabilities are most commonly analyzed by pseudostatic, sliding block (Newmark, 1965), or stress-deformation analyses. The Makdisi-Seed approach, based upon the results of sliding-block analyses, is also used frequently.

The pseudostatic acceleration required to bring a slope to the point of incipient failure is known as the yield acceleration. If earthquake-induced accelerations in a slope momentarily exceed the yield acceleration, the unstable soil will momentarily accelerate relative to the material beneath it. Sliding-block analyses can be used to calculate the amount of displacement that occurs. The total displacement depends on the amount by PR-268-9823
2214
SECTION C2: Seismic Hazard Quantification C2-16
which the yield acceleration is exceeded (a function of the ground motion amplitude), the time over which the yield acceleration is exceeded (a function of the frequency content of the ground motion), and the number of times the yield acceleration is exceeded (a function of ground motion duration). Given the highly variable nature of ground motion characteristics, computed displacements can be quite variable. The Makdisi-Seed procedure is based upon sliding-block analyses of earth dams and embankments.

Knowing the fundamental period of vibration of the dam/embankment and the yield acceleration of the slope, simple charts can be used to estimate earthquake-induced permanent displacements.

It is important to note that the factor of safety based upon equivalent static slope stability assessment can be substantially less than 1.0 before significant sliding displacement occurs. Thus, concluding that large slope movements are highly probable because static slope analyses indicate that the slope cannot withstand the PGA often greatly overestimates the likelihood that a significant landslide hazard exists.

Several commercial software packages are available that greatly simplify the application of sliding-block analysis. Stress-deformation analyses have been used to estimate permanent deformations caused by inertial instabilities. Strain potential and stiffness reduction approaches allow estimation of permanent deformations from relatively simple analyses; their estimates are highly approximate. Although the computational effort is dramatically increased, permanent deformations can be analyzed more rigorously using nonlinear finite-element techniques. As the accuracy of constitutive models for soils improve, the use of nonlinear finite-element analyses is likely to increase. Quantification of slope stability under static and dynamic conditions with analytic techniques for estimating potential slope movements requires special expertise and should be carried out by individuals experienced with the local geologic conditions.

A slope stability analysis is only one part of a comprehensive evaluation of slope stability. Prior to the analysis, detailed information on geologic, hydrologic, topographic, geometric, and material characteristics must be obtained. The accuracy of the analysis will be only as good as the accuracy of this information.

C2.6 Seismic Wave Propagation
Wave propagation strains are related to the incoherency between seismic ground motions at two locations along a linear structure. Well-constructed buried natural gas and liquid hydrocarbon pipelines in good condition generally have not been affected by seismic wave propagation. This is borne out by the lack of a single reported case of failure of ductile, full penetration welded natural gas or liquid hydrocarbon pipeline attributable to wave propagation alone. Recent earthquake experience (Honegger, 1999a) has indicated that wave propagation is a credible earthquake hazard for pipelines only in cases of extremely poor quality girth welds or corrosion defects subjected to very high levels of seismic ground motion. In both instances, these guidelines recommend a case-specific assessment of the longitudinal strain to failure. Such an assessment may require removal of sufficient girth weld samples to develop a statistical description of strain capacity PR-268-9823
2215
SECTION C2: Seismic Hazard Quantification C2-17
based upon laboratory tests. Other approaches may involve detailed non-destructive examination, fracture mechanics models, and laboratory tests of a limited number of pipeline specimens. In either case, the details of such case-specific investigations are currently under development in the industry and are beyond the scope of these guidelines.

The only instance where seismic wave propagation has been linked to pipeline is a case study of the 1985 Michoacan (Mexico City) earthquake. The welded pipeline was constructed of 42-inch x 0.313-inch wall API 5L X42 pipe. The reported damage was compressive wrinkling at intervals of approximately 500 feet and occurred in soft clay deposits. The pipeline was influenced by unusual seismic and geologic conditions.

Because of the long distance separating Mexico City and the earthquake epicenter (nearly 250 miles), the predominant ground shaking was the result of Rayleigh waves with a relatively high period of incoming waves (2 to 3 seconds). Site amplification contributed to a high peak particle velocity, measured at 35 cm/sec, and the soft lake sediments were characterized by a wave propagation velocity as low as 40 to 100 m/sec. The combination of high particle and low wave propagation velocities would have promoted high ground strains, and the long period would have promoted a relatively large development length for mobilizing shear resistance between the pipe and surrounding soil. O’Rourke (1988) concluded that calculated pipe strains were sufficient to conclude that traveling ground waves were responsible for the damage.

For these reasons, the conclusions of O’Rourke (1988) can, at best, only be considered applicable to geologic settings similar to those from the case study, i.e., very soft deep sediment deposits prone to harmonic excitation by Rayleigh waves. The soil deposits near Mexico City and the tectonic regime surrounding Mexico are unique in this respect.

Furthermore, estimating wave propagation strains for any region of similar geology for which strong motion earthquake records do not exist requires an attenuation relationship to estimate ground motions associated with Rayleigh waves, which do not exist.

Although the potential for damage to welded steel pipelines due to seismic wave propagation is generally low, it is occasionally necessary to compute the stress or strain in a pipeline caused by seismic waves. A procedure for estimating apparent wave propagation velocity was first proposed by Newmark (1965). The basis for estimating wave propagation strains is illustrated in Figure C2.14. Earthquake body waves (compression and shear waves) emanate from the earthquake source region as a continuous wave front. Ground shaking at the surface results from the propagation of this wave front up through the near-surface soils. At two locations, A and B, the apparent propagation of body waves is related to the difference in arrival times of the wave front at these locations. The arrival time is a function of the source-to-site distance for A and B and the propagation velocity along the path between A and B. The straight propagation paths shown in Figure C2.14 are a simplification, as the actual propagation paths are likely to be curved as a result of passage through strata of varying density. On the surface, the seismic wave front appears to propagate along the surface at a velocity that is closely related to the propagation of the seismic waves at depth (e.g., 3 km to 30 km).

The approach proposed by Newmark (1965) leads to an estimate of maximum ground strain between any two locations on the surface that is proportional to the ground velocity generated by the response of surface soils to the input motion divided by the apparent PR-268-9823
2216
SECTION C2: Seismic Hazard Quantification C2-18
velocity of the seismic wave front relative to the surface. In this model, locations directly above the earthquake source zone experience virtually no strain as they experience nearsimultaneous arrival of the seismic wave front which translates to a near infinite apparent velocity.

There are some cases that are outside the scope of these guidelines that occasionally merit special consideration. These include transitions between very stiff and very soft soils, penetration into buried valve enclosures, branch connections, pipe fittings, and valves. In addition, pipelines with significant weld flaws, corrosion defects or not connected with full-penetration butt welds (e.g., compression couplings, bell and spigot joints) may also be vulnerable to damage from seismic wave propagation. In such cases, an assessment of the strain capacity of the pipe and comparison with the magnitude of wave propagation strains can be used to judge the expected performance of the pipeline.

C2.7 Transient Ground Deformation
Another potential seismic hazard not related to permanent ground deformation is transient ground deformation (TGD). Transient ground deformations are produced when local site conditions allow weaker surface soils to respond differently from adjacent nearsurface soils that are considerably stronger and can possibly damage buried pipelines.

This type of behavior has been observed repeatedly in past earthquakes where pipeline damage appears to be more prevalent near the boundaries between strong and weak soils.

Transient ground deformation has been identified as a possible mode of local soil response that can lead to pipeline damage. One of the first investigations to identify transient ground deformation as a cause of pipeline damage was undertaken by O'Rourke (O’Rourke, 1991, Pease and O’Rourke, 1997) in a study of water and sewer pipeline damage in the Marina District of San Francisco following the 1989 Loma Prieta earthquake. O’Rourke postulated that liquefaction in the Marina District caused the hydraulic fill soils to be decoupled from the response of the underlying stiff soils and respond more like a sliding mass.

Following the 1994 Northridge earthquake, investigations into the causes of pipeline damage in Potrero Canyon identified a similar decoupling mechanism as a probable cause of the ground cracking patterns and locations of pipeline damage (Honegger, 2000). The underlying mechanism in Potrero Canyon was postulated to be shear failure within the weak canyon soils that were subjected to a velocity pulse of more than 1 m/s and a period of 5 to 6 seconds. The magnitude and long period of the velocity pulse is attributed to the proximity of Potrero Canyon to the projection of the caU.S.tive fault plane. This type of near-field motion has been recorded in past earthquakes and is commonly referred to as earthquake “fling.”
There are no widely accepted procedures for identifying sites susceptible to TGD or estimating potential TGD displacements. A large part of the difficulty lies in the inability to characterize near-field ground motions and the characterization of the strength properties of soils undergoing variable states of liquefaction. The following guidance for PR-268-9823
2217
SECTION C2: Seismic Hazard Quantification C2-19
identifying sites potentially susceptible to TGD are provided based upon the sparse amount of field data:

1. Presence of a relatively uniform layer of soil with low shear strength (or high potential for liquefaction).

2. Presence of firm soil or rock adjacent to, or confining, deposits of weak or liquefiable soils.

3. Soft soil sites located in the near-field of a potential earthquake with a likelihood for experiencing ground motions characteristic of earthquake “fling”.

Techniques for modeling TGD include two-dimensional finite element analysis and simplified sliding-block analysis. Representative models are illustrated in Figure C2.15.

Finite element methods are complicated by the uncertainty associated with modeling slip along the soil failure plane and, in some cases, the need to account for separation at the boundary between stiff and soft soils. Sliding-block analogies face similar impediments with respect to defining a limiting value of shear strength for the soft soil and the sliding friction associated with the decoupled soil mass. Both methods require an explicit input ground motion time history that adds an additional layer of uncertainty in the modeling process.

At this time (2001), assessment of TGD remains a research topic. Unless there is compelling evidence to suspect that TGD displacements exceed a few centimeters, consideration of TGD is believed to be unnecessary for new pipelines or pipelines constructed in a manner consistent with modern construction. This position is supported by the fact that there is no evidence that TGD has been a factor in damage to such pipelines.

Older pipelines with potential weld or corrosion defects or not constructed with buttwelded pipe joints are at greater risk of damage from TGD. Given the lack of an accepted method for estimating TGD displacements, it does not seem prudent to devote large resources to estimating TGD for potentially susceptible pipelines. Instead, TGD damage occurring at unknown locations should be recognized as a possible consequence of the earthquake, and the need to potentially respond to TGD damage should be incorporated into earthquake response plans.

PR-268-9823
2218
SECTION C2: Seismic Hazard Quantification C2-20
Likelihood that Cohesionless Sediments (saturated)
would be Susceptible to Liquefaction (by age of deposit) Type of Deposit
General
Distribution of
Cohesionless
Sediments in
Deposits
Less than
500 years Holocene Pleistocene
PrePleistocene
Continental Deposits
River channel Locally variable Very high High Low Very low Flood plain Locally variable High Moderate Low Very low Alluvial fan and plain Widespread Moderate Low Low Very low Marine terraces and plains Widespread N/A Low Very low Very low Delta and fan-delta Widespread High Moderate Low Very low Lacustrine and playa Variable High Moderate Low Very low Colluvium Variable High Moderate Low Very low
Talus Widespread Low Low Very low Very low
Dunes Widespread High Moderate Low Very low
Loess Variable High High High Unknown
Glacial til Variable Low Low Very low Very low
Tuff Rare Low Low Very low Very low
Tephra Widespread High High Unknown Unknown
Residual soils Rare Low Low Very low Very low
Selka Locally variable High Moderate Low Very low
Coastal Zone
Delta Widespread Very high High Low Very low
Estuarine Locally variable High Moderate Low Very low
Beach: high wave energy Widespread Moderate Low Very low Very low Beach: low wave energy Widespread High Moderate Low Very low Lagoonal Locally variable High Moderate Low Very low
Fore-shore Locally variable High Moderate Low Very low
Artificial
Uncompacted fill Variable Very high N/A N/A N/A
Compacted fill Variable Low N/A N/A N/A
Table C2.1 Liquefaction Susceptibility Screening Table
(from Youd and Perkins, 1978)
PR-268-9823
2219
SECTION C2: Seismic Hazard Quantification C2-21
Material V/PGA
(inches/s-g)
V/PGA
(cm/s-g)
Rock 26 66
Stiff Soil 57 144
Cohesionless Soil 55 140
Alluvium 48 122
Table C2.2 Velocity Ratios for Various Types of Surface Soils (from O’Rourke and El Hamdi, 1988)
PR-268-9823
2220
SECTION C2: Seismic Hazard Quantification C2-22
Hanging Wall Side (Block) of Fault
c) Reverse-Slip Faulting
b) Normal-Slip Faulting
a) Strike-Slip Faulting
(Left Lateral)
Figure C2.1 Fault Movement Classification
PR-268-9823
2221
SECTION C2: Seismic Hazard Quantification C2-23
A
B
C
A B
C
Fault Crossing Angle
Initial Configuration
Displaced Configuration
Analytical Model
Relative Displacement
Between Pipe and Soil
NOTE: For clarity, only horizontal
soil springs are illustrated. For
analysis of pipeline response, each
node would also have axial and
vertical soil springs
Spring Base Displacement
Ground Displacement Profile
Pipe Displacement
Bilinear Soil Spring Representation
Figure C2.2 Effect of Right Lateral Strike-Slip Fault Movement on a Buried Pipeline PR-268-9823
2222
SECTION C2: Seismic Hazard Quantification C2-24
Apparent Dip
Angle, α
Dip Angle, δ
Regional Stress Angle, β
Wells &
Coppersmith
Displacement,
AD or MD
Total Fault
Displacement, DF
Vertical Component of
Fault Displacement, V
Horizontal
Component of Fault
Displacement, H
Transverse Component of
Fault Displacement, H
Figure C2.3 Orthogonal Components of Fault Displacement Based Upon Estimates Using Wells & Coppersmith (1994) Empirical Equations
PR-268-9823
2223
SECTION C2: Seismic Hazard Quantification C2-25
A A
B
B
Section A-A
Section B-B

SOURCE OF

ENERGY RELEASE
SITE
Rsf = closest distance to
projection of fault on surface
Re = epicentral distance
Rh = hypocentral distance
Rfp = closest distance
to fault plane

DEPTH TO
ENERGY

RELEASE

DEPTH TO
ENERGY

RELEASE
Figure C2.4 Various Definitions of Earthquake Parameters Adopted in Attenuation Relationships
PR-268-9823
2224
SECTION C2: Seismic Hazard Quantification C2-26
(from the Supplement to the National Building Code of Canada (1990) for PGA with a 10% probability of exceedance in 50 years)
Figure C2.5 Example of Mapped Representation of Probabilistic Ground Motion PR-268-9823
2225
SECTION C2: Seismic Hazard Quantification C2-27
0 10
20 30
40 50
60 70
80 90
100
0 500 1000 1500 2000 2500
Return Period, years
Los Angeles
Seattle
San Diego
Salt Lake City
St. Louis
New York
Raleigh
Denver
Figure C2.6 Example Hazard Curve Representation of Probabilistic Ground Motion PR-268-9823
2226
SECTION C2: Seismic Hazard Quantification C2-28
< 25 50 75 100 125 150
5 5.5
6 6.5
7 7.5
0 20
40 60
80 Contribution, %
Distance, km
Magnitude
60-80
40-60
20-40
0-20
5 6
7 < 25
50 75
100
125
150
0 20
40 60
80 100
Contribution
Magnitude
Distance
80-100
60-80
40-60
20-40
0-20
A) Largest Contributor to Seismic Hazard
50% contribution to hazard
from M = 6.75; 84%
contribution from M = 7.2,
both at a distance of 25 km or
less
B) Cumulative Contribution of Magnitude and Distance
Figure C2.7 Example of Extraction of Deaggregated Hazard Information from the 2,500-year Firm Rock PGA Hazard at Salt Lake City
PR-268-9823
2227
SECTION C2: Seismic Hazard Quantification C2-29
Figure C2.8 Liquefaction CSR Curve (from Youd and Idriss, 1997) PR-268-9823
2228
SECTION C2: Seismic Hazard Quantification C2-30
Liquefaction occurs in the cross-hatched zone. The surface layer moves laterally down the mild slope or toward a free breaking up into blocks which may jostle back and forth and settle unevenly during the spreading process.

Figure C2.9 Schematic Representation of a Lateral Spread (from Youd, 1984) PR-268-9823
2229
SECTION C2: Seismic Hazard Quantification C2-31
(arrows reflect changes from approach of Bartlett and Youd (1992) for selected data points) Figure C2.10 Comparison of Actual Lateral Spread Displacements with Estimates Using Youd et al., 1999 (from Youd et al., 1999)
PR-268-9823
2230
SECTION C2: Seismic Hazard Quantification C2-32
HORIZONTAL DISTANCE, meters
HORIZONTAL DISTANCE, meters
Figure C2.11 Sample Comparison of Bartlett and Youd Method with SOILSTRESS (Bartlett and Youd (1992) results denoted by “MLR”)
PR-268-9823
2231
SECTION C2: Seismic Hazard Quantification C2-33
Earth Block Slide (coherent)
Earth Flow (occurs very slow to
rapid and can be coherent to
disruptive in nature)
Debris Avalanche (disruptive and
occurs very rapid to extremely rapid)
Rock Topple (disruptive)
Earth Slump (coherent)
Rock Fall (disruptive)
Figure C2.12 Various Types of Landslides (after Varnes, 1978) PR-268-9823
2232
SECTION C2: Seismic Hazard Quantification C2-34
Displacement from normalized earthquakes for
a sliding block with unsymmetrical resistance
(inclined sliding plane)
Displacement from normalized earthquakes for
a sliding block with symmetrical resistance
(horizontal sliding plane)
Figure C2.13 Generic Sliding-block Relationships Using Actual Earthquake Records Scaled to a PGA of 0.5 g (from Newmark, 1965)
PR-268-9823
2233
SECTION C2: Seismic Hazard Quantification C2-35
α1 α2
L1
L2
t
L
c
t
L
c
c
cL L
LL apparent
1 1
2 2
21 21
1 2
21 = ⋅
= ⋅
= −
− L
N
M O
Q
P
cos( )
cos( )
cos( ) cos( )
α
α
αα
bg
A B
Figure C2.14 Wave Propagation Paths Leading to Incoherent Seismic Ground Strains PR-268-9823
2234
SECTION C2: Seismic Hazard Quantification C2-36
Support Motion M
Friction Boundary
A) Sliding Block on Frictional Surface
Boundary is modeled with interface
sliding/gap elements capable of
transmitting only compression and
shear across the boundary.

Earthquake displacements are
imparted through boundary elements.

B) 2-D Finite Element Representation of Valley
Figure C2.15 Modeling Techniques with Potential Application to Estimating TGD PR-268-9823
2235
SECTION C3: Pipeline Performance Criteria C3-1
C3.0 PIPELINE PERFORMANCE CRITERIA
Determining limiting pipeline strains associated with varying levels of performance is an ongoing area of investigation within the pipeline industry. While there have been several notable projects using full-scale tests and detailed finite element analyses, the differences in load conditions, measurement techniques, and project goals make direct comparison difficult. As a result, defining appropriate relationships for seismic assessment of pipeline response is largely based upon available test data reported in the open literature and the collective judgment of the authors and reviewers contributing to these guidelines.

C3.1 New Pipelines and Pipelines Compatible with Current Construction Practice
There is experience in the pipeline industry of strain based design. Most of this is generally related to offshore applications such as pipe reeling for installation, high temperature, high pressure pipelines that can buckle due to high axial loading, and deepwater pipelines. Much of this experience relates to “load-controlled” situations on small diameter, low D/t (diameter to wall thickness), thick-walled pipes. In these situations, there is a high likelihood that the pipe will experience the design strain (in the range of 2% nominal strain1
), thus validating the methodologies used. However, the strains in these applications may not be as large as those expected for seismic design.

There are situations where pipelines have survived much larger strains, as well as instances where pipelines have failed at much lower strains.

The selection and specification of both pipe and welding for a strain-based design pipeline system includes consideration of a number of factors that may not be adequately addressed in a conventional stress-based design, which is thoroughly addressed in various codes and standards such as ASME B31.4, B31.8 and API 1104. Considerable research efforts in the pipeline industry have been directed at understanding the behavior of pipe at high strains, with this effort increasing over the last few years with more focus on strainbased design. Although the industry now has a vastly improved understanding of high strain behavior, significant advancements continue to be made.

The overall strain capacity of a pipe is dependent upon a number of factors:

1. Compressive Strain:
a. The properties of the pipe material including the shape of the stress-strain curve, the ratio of yield strength to tensile strength (Y/T), and uniform elongation strain (εUL);

1 Nominal strain is the strain that the pipe “globally” experiences over a certain gauge length (anywhere between 0.5 to 2 diameters). However, locally the strains can be much higher due to small geometric imperfections, and once a pipe starts to wrinkle and locally buckle, the strains that can be experienced in the vicinity of the buckle can be many times that of the nominal strain.

PR-268-9823
2236
SECTION C3: Pipeline Performance Criteria C3-2
b. The D/t ratio, with a greater wall thickness having a higher strain capacity for the same pipe diameter (lower D/t ratio); and
c. Geometric anomalies associated with out of roundness (ovality) and wall thickness variations need to be considered for single pipe joints and at girth welds, where they can introduce wall thickness step changes when joints are butted together. The effect of geometric anomalies is to reduce the strain to initiate pipe wall buckling.

2. Tensile Strain: The tensile strain capacity depends upon the pipe material properties mentioned above, as well as the following factors related to the resistance of girth welds to the presence of welding defects:
a. Increased wall thickness reduce the effects of a loss of cross-section from weld defects and increases the amount of material available to resist crack growth;
b. Weld properties (yield strength, ultimate strength, and toughness) that meet or exceed the pipe metal reduce the potential for flaw growth; and c. Weld quality specifications and inspection requirements increase the likelihood that girth welds will achieve the required properties and defect tolerances.

Key aspects of the above issues are briefly discussed below:

Pipe Steel Properties: The pipe properties of interest for both compressive and tensile strain capacities are the longitudinal properties. Circumferential properties are typically only of interest for assuring that the material is not anisotropic.

1. The stress-strain curve should be monotonically increasing, and should contain no plateau regions in the post yield portion of the curve. A high yield strength for the grade is not necessarily a benefit in strain-based assessments. It is good practice to obtain several stress-strain curves from the mill from full thickness, longitudinal specimens for the pipe.

2. The Y/T and εUL influence strain capacity with increases in strain capacity generally obtained by lower Y/T and higher εUL. Both of these values vary with pipe grade and the pipe toughness with increases in the grade and toughness generally resulting in higher Y/T ratios and lower εUL and lower strain capacities (i.e., X65 will generally have a higher strain capacity than X80). The required value of Y/T and εUL depends upon each application and the required strain capacity. Values of Y/T and εUL quoted in pipe material specifications need to be considered carefully, as “standard” values used in conventional specifications may not be sufficient. Specifying unusually low values of Y/T and high values of εUL may significantly increase costs for manufacturing the PR-268-9823
2237
SECTION C3: Pipeline Performance Criteria C3-3
limited amount of pipe that might be needed for seismic design and result in additional costs for testing to show that the pipe meets specifications. Some pipeline projects have taken steps to segregate pipe by heat and, therefore, Y/T properties. They then use the high Y/T pipe at high strain areas.

3. Weld heat affected zone (HAZ) properties are an important consideration as there is a tendency of certain low alloy materials to soften in the HAZ. This softening reduces the original pipe yield strength near the weld, leaving stronger material on either side of the zone. Once yielding occurs in the HAZ, all of the plastic strain will accumulate in this narrow zone, rapidly exhausting the available ductility and leading to plastic instability. Careful specification of the carbon level and the presence of secondary hardening alloys in the pipe steel is necessary to minimize the potential for softening in the HAZ.

Pipe Geometric Properties:

1. Wall thickness considerations include both the absolute thickness and variability.

A thicker pipe wall increases both tension and compression strain capacities. This results in an additional benefit from using lower grade pipe as a lower grade pipe will require a thicker pipe wall. For example, if X65 is used for seismic design instead of X80, the strain capacity is increased because of both the lower grade (lower Y/T and higher εUL) and thicker wall (lower D/t). Careful consideration needs to be given to the allowable wall thickness tolerances in the pipe specifications as geometric misalignment decreases the compression strain for initiation of wrinkling (a concern for strain levels associated with continued operation). There is a trade-off between minimizing the geometric misalignment across the girth weld and the extra cost associated with having tighter tolerances.

2. Control of tolerances on pipe ovality, like control of pipe wall thickness, is important because ovality can reduce the compression strain for initiation of pipe wrinkling in a pipe joint. Furthermore, ovality of the ends of the pipe can result in geometric misalignment of the pipe ends at the girth weld that reduces the compression strain for initiation of pipe wrinkling in the same way as wall thickness variation. There is also a trade-off between minimizing pipe ovality and the extra cost associated with having tighter tolerances.

Welding Materials and Procedures: The presence of small imperfections or flaws in the weld deposit, and particularly those having a planar geometry and located near the root or cap of the weld, severely degrade the ability of the material to uniformly distribute plastic strain. Flaw extension in either a brittle or ductile mode from continuing strain at the extremities of these flaws leads rapidly to rupture. Because the imperfections cannot be entirely eliminated, a weld metal having slightly overmatching yield strength should be selected to minimize the opportunity for over-strain of the surrounding material. The degree to which the weld metal will overmatch the pipe metal must account for the possibility that the actual pipe strength may be significantly greater than the specified PR-268-9823
2238
SECTION C3: Pipeline Performance Criteria C3-4
minimum values. The need to achieve overmatching weld strength is another reason to consider the use of lower grade pipe steels in designing the pipeline to resist ground movement. A welding process should be chosen to minimize both the number and size of the imperfections. In the absence of highly skilled manual pipeline welders, the welding procedure should consider the use of mechanized GMAW or TIG processes.

Quality Control Over the Welding Process: The experienced-based workmanship requirements (flaw acceptance criteria) contained in the stress-based welding codes such as API 1104 or ISO 13847 may not be adequate for strain-based designs. New acceptance limits (generally smaller and fewer allowable flaw sizes) may need to be calculated based upon fracture mechanics technology. The fracture mechanics calculations used to determine weld acceptance are semi-empirical and require estimates of several parameters, particularly those related to flaw geometry, flaw location, material strength, and material toughness. Unlike approaches aimed at fitness-for-purpose assessments, the approach in these guidelines will not generally require the same high level of reliability that the pipeline will be capable of withstanding a particular tension strain for the reasons noted in Section C3.1.2. The level of required reliability needs to be considered when making decisions on the appropriate level of conservatism associated with the many assumptions necessary to establishing strain limits using fracturemechanics approaches.

The new and smaller allowable flaw size presupposes that there will be an inspection system in place capable of detecting and sizing any flaws remaining in the weld. It is well established that flaw height and the distance to the pipe surface are the critical parameters. Conventional radiographic techniques are not suitable for either of these measurements and the use of a mechanized ultrasonic procedure such as that commonly used for offshore lay barges is recommended. Accuracies of about 0.040 inches are achievable for flaw height measurement in pipe with wall thickness less than about one inch. The net effect of the special welding procedure and the extensive post-weld inspection may lead to a high weld rejection rate, thereby reducing productivity and increasing cost per weld. However, this should be economically practicable considering the relatively few segments of the pipeline where seismically induced permanent ground displacements may occur.

Achieving strain limits as high as 2% to 4% may not be possible under certain circumstances, particularly for the assessment of existing pipelines with flaws or defects or where the use of high-grade pipe steel reduces the likelihood that girth welds will have strength and toughness characteristics superior to the pipe steel such that welds develop the full strength and strain capacity of the pipe steel. Specific criteria for assessing the strain capacity of girth welds is beyond the scope of these guidelines but techniques for performing such assessments are well-established. Two of the more common approaches to qualify pipe material and welding specifications for a specific project application depend upon either curved wide plate testing of weld specimens that conform to project specifications or theoretical fracture mechanics models.

PR-268-9823
2239
SECTION C3: Pipeline Performance Criteria C3-5
It is noted that there may be additional proprietary data that would lead to an alternate basis for establishing pipeline strain acceptance criteria. Users of these guidelines may elect to review such additional data prior to establishing strain criteria for specific projects. As noted in Section C1.2, users that need to demonstrate levels of seismic reliability significantly higher than has been accepted in typical practice will likely need to adopt more conservative strain acceptance criteria than provided in these guidelines.

C3.1.1 Performance Goal: Maintain Pressure Integrity
Failure in a strain-based design is normally taken to mean loss of the pressure boundary integrity. The failure can result from axial tensile overload or excessive bending.

Bending involves both tension and compression strains with failure frequently occurring on the compression side of the bend as a result of a buckle in the pipe wall followed by a wrinkle. This leads to gross tension strain in local areas of the wrinkle. Most failures involve first the initiation of a partial thickness crack followed by brittle or ductile crack propagation through the wall. The latter may also involve out-of-plane ductile tearing.

Longitudinal Compression Strain Limit
In compression, local instabilities such as wrinkling can develop at strains much less than the allowable tension strain limits. Wrinkling of the pipeline wall does not, in itself, constitute a failure condition. While wrinkling may also lead to a significant reduction in pipeline moment capacity, this is not a concern for displacement-controlled conditions, as the pipeline deformation is not related to an applied load.

Under sustained loading conditions, further compressive shortening would be expected to concentrate at points of initial wrinkling. Typically, the initiation of compressive wrinkling occurs in the range of 0.3% to 0.6% strain for most large diameter pipes, or about one-tenth the strain level noted above for tension. Loss of pressure boundary integrity is associated with strains far greater than that associated with the initiation of compressive wrinkling for moderate D/t ranges.

In recommending longitudinal compression strain limits, the clear evidence of substantial compression strain capacity prior to rupture was tempered by the concerns regarding establishing recommendations on the maximum compression strain capacity that are much greater than past practice for seismic load applications within the pipeline industry (e.g., ASCE, 1984). The following discussion describes some of the test data used in assessing the suitability of equation (3-1).

Pipe tests performed by Mohareb et al. (1994) focused on conditions where the axial load was constant (five of the seven analysis cases utilized constant axial load). The axial force applied in the pipe tests by Mohareb et al. corresponded to a 45°C temperature differential, a tension force equivalent to the force necessary to counteract axial shortening from the Poisson effect of internal pressure, and a compressive force to counteract the tension produced by the closed-end conditions of the test specimens.

PR-268-9823
2240
SECTION C3: Pipeline Performance Criteria C3-6
Ghodsi et al. (1994) repeated the tests of Mohareb et al., but include a girth weld in the center of the test specimen to assess the impact of a girth weld on initiation of wrinkling.

The testing performed by Ghodsi et al. demonstrated that the presence of a girth weld significantly reduces the strain associated with a visual indication of pipe wrinkling.

However, the reduction in the strain that could be reached in the tests was not as significant as the reduction in strain at initiation of wrinkling.

Pipe tests at the Center for Engineering Research (C-FER) in Edmonton, Alberta, Canada, have also been performed to examine post-wrinkling behavior. The details of the tests are proprietary but some results, including a relationships for compression strain limits for X70 steel, have been published by Zimmerman et al. (1995). The Zimmerman et al. relationship is reportedly based upon limiting maximum computed tension strain in the crest of the buckle formed after wrinkling to 10%.

Another testing program at the University of Alberta was carried out to investigate strains associated with the initiation of wrinkling (Dorey et al., 2001). These tests were performed on X70 pipe with a D/t ratio of 92 and internal pressures producing hoop stresses ranging from 0% to 80% of the specified minimum yield stress. These tests are relevant for comparing with equation (3-1) since they were taken past the point of wrinkle formation and development of maximum pipe moment capacity. Dorey et al. recommend several relationships for estimating the longitudinal strain in the pipe at the point of maximum moment. These relationships account for imperfections in the pipe wall and at the girth weld and also consider two typical shapes for the stress-strain curve of the pipe material.

Data from the available papers and test reports for the above tests are plotted in Figure C3.1 with strains associated with the end of testing plotted in Figure C3.1(a) and strains associated with initial wrinkling or maximum moment plotted in Figure C3.1(b). Several data points are plotted for the same D/t ratios. These data points correspond to different levels of internal pressure and net axial load applied during the testing. Internal pressure in the tests varies from zero to an internal pressure that produces hoop tensile stresses equal to 72% (Mohareb et al., 1994 and Ghodsi et al., 1994) or 80% (Zimmerman et al., 1995 and Dorey et al., 2001) of the pipe yield stress. Higher strains for the same D/t ratios are from tests with higher internal pressure.

The potential conservatism in equation (3-1) is clearly evident in Figure C3.1(a) by the fact that the maximum local compression strains imposed by laboratory tests with moderate to high internal pressure are typically at least two times greater than those given by equation (3-1). The only exceptions to this are the C-FER data. The smaller strains for the C-FER data may be a result of the presentation of limited data in the publicdomain paper, the lack of interest in testing well beyond maximum moment, or the physical limitations of the testing configuration. It should be noted that all of the tests used to generate the data in Figure C3.1 were carried out without the pipe losing pressure integrity; hence, the actual safety margin against loss of pressure integrity is indeterminate.

PR-268-9823
2241
SECTION C3: Pipeline Performance Criteria C3-7
Longitudinal Tension Strain Limit
Determining an appropriate tension strain limit will vary with welding specifications and inspection criteria and needs to be done on a project-by-project basis. The likelihood of achieving a particular strain capacity need not be unnecessarily restrictive. It is the intent of these guidelines that any acceptable strain criteria used on a particular project should be based upon the concepts of achieving a balanced seismic design as discussed in Appendix A. Estimates of the probability of pipeline failure at the specified strain limits in the range of 10% to 30% are judged acceptable for achieving a balanced seismic design in these guidelines. This level of weld failure, which may seem high compared to other applications such as fitness for purpose assessments, is justified in these guidelines where the likelihood of seismic hazard is the same as the acceptable likelihood experiencing seismic damage. Other factors supporting a higher acceptable probability of a weld not meeting a particular strain level are related to the low probability of a severe undetected flaw at a location of high strain.

It is expected that the allowable longitudinal tension strains will range from 2% to 4% for pressure integrity when evaluating pipeline response to permanent ground deformation.

Longitudinal tensile strains of 3% to 5% have been adopted for assessing the ability of pipelines to maintain pressure integrity when subjected to earthquake-generated ground displacement for more than 20 years and are recommended in ASCE (1984).

Advancements have been made in the use of fracture-mechanics approaches to estimate failure strains and the comparatively low strain capacities that often resul from fracturemechanics approaches have raised questions regarding the suitability of historically adopted strain limits.

The magnitude of permissible computed tension or compression strain is limited to no greater than 4% based upon what is believed to be the practical limits of the ability of the analytical approach recommended in these guidelines to adequately represent pipeline behavior. Strain limits supported by additional test data, more refined analysis methods, or corroboration of the analysis approach with test or case history data are generally preferred.

Resolving differences between historic practice, test data, and fracture-mechanics approaches is complicated by several factors. Typical fracture-mechanics approaches were originally developed for applications that require a very low likelihood of failure.

This low failure probability is achieved, in part, by adopting conservative assumptions with respect to the type of flaw, the location of the flaw, material toughness, and degree of weld overmatching. Thus, the strain estimates obtained by existing fracture mechanics methods are likely to be associated with a probability of failure much lower than required by the approach adopted for defining seismic hazards in these guidelines. Revising fracture-mechanics approaches to adequately match test data for strain capacities much greater than 1% and provide a statistical estimate of strain capacity is an area of active research within the pipeline industry.

PR-268-9823
2242
SECTION C3: Pipeline Performance Criteria C3-8
The need to justify high strain capacities will vary among projects and the nature of the seismic hazards faced. In some cases, the required strain capacity necessary to withstand earthquake-generated ground movement is much less than 2% to 4% range recommended in these guidelines as generally achievable. There are also cases, particularly for highcost projects, when a substantial increase in pipeline wall thickness required by adopting conservative fracture-mechanics based estimates of tension strain capacity may not have an economically significant impact, considering the increased wall thickness is only required at locations where earthquake ground movement is expected. Finally, ground displacements that are much more likely to occur compared to earthquake-related ground movement (e.g., frost heave, thaw settlement, active landslides, mining subsidence), generally require more conservative estimates of strain capacity to meet desired performance goals. For such cases, guidance on longitudinal strain limits is provided in Tables C3.1 through C3.3 based upon a fracture mechanics study by EMC2 (Wang,
2003). The three tables provide three levels of longitudinal tension strain limits based upon material toughness and Y/T ratio. Comparison with test results described in the EMC2 report indicate the tension strain limits in Tables C3.1 through C3.3 may have a mean factor of safety as high as 1.8 on strain. Users should carefully review the EMC2 report to fully understand the basis for the recommendations in Tables C3.1 through C3.3 before adopting them for a particular project.

C3.1.2 Performance Goal: Maintain Normal Operability
The compression strain limits to maintain normal operability for displacement-controlled conditions are largely based upon judgment. The magnitude of longitudinal tension and compression strains are limited to 2%. The reasons for the 2% limit are related to concerns about the overall distortion of the pipe cross-section that might be associated with coating damage or impair the passage of internal pigs.

Longitudinal Compression Strain Limits
Longitudinal compression strains associated with buckling instability of the pipe wall are limited to the value estimated using the maximum moment expression developed by Dorey et al. (2001). Although there is evidence that strains greater than those corresponding to the maximum moment onset of wrinkling can be sustained without appreciable deformation of the pipe cross-section, a conservative approach is adopted to reduce the likelihood of conditions that might impair the long-term integrity of the pipeline. Equations (3-3a) and (3-3b) are the equations developed by Dorey et al. (2001) for offsets at the girth weld. The compression strain limit from equation (3-3a) is plotted in Figure C3.1(b) for the case of X70 pipe with a p/py ratio of 0.72 and a ∆/t ratio of 0.10.

Also plotted in Figure C3.1(b) are the pressurized test data points corresponding to initiation of wrinkling and maximum moment from tests performed at the University of Alberta. For comparison, data assembled by Stephens et al. (1991) from a wide variety of unpressurized pipe tests are also shown to illustrate similarity with the University of Alberta results. As shown by Figure C3.1(b), the 2% limit on strains for continued PR-268-9823
2243
SECTION C3: Pipeline Performance Criteria C3-9
operation governs for the conditions selected at D/t ratios less than about 57. The reasons for the 2% limit are related to concerns about the overall distortion of the pipe crosssection that might be associated with maintaining corrosion coating integrity or impair the passage of internal pigs.

Longitudinal Tension Strain Limits
Similar to the selection of tension strain limits for pressure integrity, the longitudinal tension strain limit associated with continued operation are associated with strains that can be accommodated without degrading corrosion coating or internal pig passage. The literature review for these guidelines did not identify information in the open literature relating the impact of longitudinal tension strain on the long-term integrity of the pipeline. Longitudinal tension strains of approximately 1% to 2% seem to be reasonable considering that these strains are generally less than what might be generated during field cold bending operations. Lacking additional information or project-specific test data, it is suggested that longitudinal tension strains for continued operation be taken as half of the tension strain limits for pressure integrity.

C3.1.3 Load-Controlled Conditions
The loading conditions considered in these guidelines are typically governed by imposed displacement of the surrounding soil (displacement-controlled). There are some situations where the pipeline loads may not be related to displacement (load-controlled).

In the ASME codes, load-controlled conditions are considered primary loads and displacement-controlled conditions are considered secondary loads. Seismic inertial loads are an example of primary loads for aboveground piping.

Examples of load-controlled conditions include flow-type ground failures and situations where ground failure removes soil restraint around a section of pipeline. Figure C3.2 illustrates pipeline configurations corresponding to these two ground displacement conditions. Other examples of load-controlled pipeline conditions not related to earthquake hazards can be found in the offshore industry where pipe laying operations can impose strains in the pipe that must be accommodated without degradation in pipeline strength. For load-controlled situations, local buckling of the pipeline can lead to progressive deformation and pipeline rupture.

In developing load-controlled strain criteria, consideration was given to establishing a uniform strain limit of 0.005, the strain at which the specified minimum yield strength is determined. For most pipeline steels historically used for gas transmission pipelines, stresses below the specified minimum yield stress (corresponding to a strain of 0.5%) do not pose a significant risk for progressive deformation. However, for compression strains, this limitation underestimates strain capacities for pipes with low D/t ratios and may underestimate strain capacity for pipes with very high D/t ratios. It is also desirable to account for the beneficial effects of internal pressure in reducing the loss of bending strength of the pipe once the maximum moment is reached. The approach adopted is to PR-268-9823
2244
SECTION C3: Pipeline Performance Criteria C3-10
base the compression strain limits on the maximum moment relationships developed by Dorey et al. (2001) that are recommended for continued operation. For load-controlled conditions, the relationships developed by Dorey et al. (2001) are factored by 0.75 to provide a reasonable margin against exceeding the maximum moment. The Stephens et al. (1991) relationship is used as a lower limit for compression strains under loadcontrolled conditions as this relationship represents a 95% lower bound fit to a large database of tests on unpressurized pipes assembled by Battelle.

C3.2 Considerations for Offshore Pipelines
The strain criteria for offshore pipelines undergoing bending and external pressure is adopted from the approach of Murphey and Langner (1985). This approach was adopted because of the relatively simple forms of the governing equations. Alternate approaches such as BS 8010 or API Bulletin 5C3 were also considered. Alternate approaches can be used but the degree of variation in the resulting allowable bending strains are well within the expected variation in hazard quantification.

Tension in offshore pipelines reduces the estimated collapse pressure by reducing the hoop stress at which yielding occurs. The approach adopted in these guidelines is consistent with past recommendations (AGA, 1990) as well as recent investigations (Foeken and Gresnigt, 1998). The approach limits the effective von Mises stress to the material yield stress. For a particular value of axial stress, the hoop stress corresponding to a von Mises yield condition is estimated as follows:
σσ
σ
σ
σ
σ yh red yh
a
yh
a
ya
− = −− L
N
M
M
O
Q
P
P
1 3
42 2
2 (C3-3)
where:
σyh-red = reduced hoop yield stress
σyh = hoop yield stress
σah = axial yield stress
σa = applied axial stress
With the simplifying assumption that the axial and hoop yield stresses are equal to the nominal yield stress, σy, equation (C3.3) limits the tension strain limit to less than the yield strain.

PR-268-9823
2245
SECTION C3: Pipeline Performance Criteria C3-11
C3.3 Pipelines Not Compatible with Current Construction Practices
Pipelines not covered by the strain limits in these guidelines include pipelines with one or more of the following characteristics:

1. Alternate girth weld details

2. Welds with detectable crack defects or not acceptable by current welding criteria

3. Corrosion defects

4. Undermatched welds
Pipelines with Alternate Connections Details
Older pipelines may have been constructed with welded slip joints, bell-bell-chill-rings, or other non-butt-welded pipe joints. In addition, pipelines may have been fabricated with mechanical connections such as Dresser® or Victaulic® couplings. The specific details of the configuration of these joints can significantly reduce their strain capacity.

For example, depending upon the shape of the bell for welded slip joints, local buckling under longitudinal compression may occur at strains well below the yield strain of the pipe (Tawfik and O’Rourke, 1985). The vulnerability of older pipelines with non-buttwelded pipe joints is often compounded by the fact that the quality of the welds is typically much worse than obtained by more recent construction practices.

It is not possible to define generic recommendations for establishing strain limits for nonbutt-welded pipe joints. Detailed finite element analyses utilizing shell elements capable of capturing the effects of material yielding and local wall buckling is perhaps the most cost-effective approach to assess strain capacity for non-standard girth welds. If the welds are suspected of having defects, the finite element analyses will need to be combined with fracture mechanics-based assessment of strain capacity for the defective weld.

Pipelines with Corrosion or Weld Defects
Evaluating the effects of corrosion or weld defects is a potential issue for the assessment of existing pipelines. Explicit consideration of the effect of defects is limited to cases where evidence from past pipeline repairs or findings from internal inspections permit quantification of defect size and distribution. Assessing pipelines for weld and corrosion defects requires expertise in fracture mechanics and experience in characterizing defect geometry and material toughness parameters. Test data on pipe or weld toughness properties are normally not available for existing pipelines. The lack of data generally leads to removing samples of material for testing, fabrication of simulated test specimens, or estimating parameters based upon correlations with similar pipe material, chemical composition, or Charpy test data.

PR-268-9823
2246
SECTION C3: Pipeline Performance Criteria C3-12
Procedures for estimating the strain capacity of pipelines with weld or corrosion defects is an area of ongoing research. Much of the work is related to estimating to what extent defects, in combination with relatively small amounts of induced bending, reduce the safe pipeline operating pressure. For seismic assessment purposes, the more common situation is longitudinal strains exceeding yield from axial or bending response with the pipeline operated well below its maximum permissible pressure.

There are many references that can be consulted regarding the assessment of weld and corrosion defects and non-standard joint details. Examples include ASME B31G (ASME, 1991), Roberts and Pick (1998), Rosenfeld (1996), Wang et al. (1996, 1998), Wang (2003), Warke and Amend (1997), and Warke et al. (1997).

For the majority of older pipelines, available information is often limited to pipe size, wall thickness, and material specification. In some cases, there may be coupon tensile test data but this is typically from a statistically insignificant number of pipe samples.

This lack of data can severely limit the analytic tools available to estimate strain capacity.

In many cases, a decision will need to be made as to whether or not it is necessary to take a pipeline out of service to remove test samples for determining mechanical properties.

Information gathered from the operational history of the pipeline, limited excavation to examine the pipeline, information from field personnel on site during construction of the pipeline, and historical experience with the performance of pipe suppliers and pipeline contractors can provide sufficient information to form a judgment on the strain capacity of a pipeline suspected of having weld or corrosion defects. Extraction of specimens for testing is typically limited to cases where potential pipeline damage from a seismic hazard is judged to be extremely serious.

It is not necessary to establish a very conservative acceptable strain limit for the purposes of the seismic assessment. The typical approach to determining an acceptable strain limit for pipelines with weld or corrosion defects is to assume that the defect is located on the circumference coincident with the occurrence of maximum pipe strain. This approach can be very conservative. The combination of maximum pipe strain with a defect is dependent upon the variation in location of of ground displacement, the direction of ground displacement with respect to the pipeline alignment, and the variation in the distribution of weld defects around the circumference or the distribution of corrosion defects around the circumference and along the pipeline. It should be remembered that the approach described in these guidelines is assumed to contain sufficient conservatism in the definition of the seismic hazard. If highly conservative approaches to assess the impact of defects are applied, it is reasonable to consider reducing the conservatism in defining the seismic hazard.

Pipelines with Undermatched Girth Welds
For the purposes of these guidelines, girth welds are considered to undermatch the pipeline material if the actual weld metal yield strength and toughness do not exceed the actual values of yield strength and toughness of the pipe material. There is considerable PR-268-9823
2247
SECTION C3: Pipeline Performance Criteria C3-13
evidence (e.g., Denys, 1992, 1993, 1994; Horsely et al., 1997) that undermatching girth welds prevent the development of gross-section yielding generally assumed to occur in the analysis of pipeline response to permanent ground displacement. If weld undermatching is suspected, tension strains for pipeline assessment should be based upon limiting pipe stresses to the yield stress of the weld unless data are available to support higher values.

Studies investigating the effect of weld undermatching on compression strains have not been identified. It seems reasonable to assume that the effect of undermatched girth welds is much less under compressive loading. Unless there is reason to believe that local yielding of the weld metal substantially reduces the occurrence of local buckling of the pipe wall, it is recommended that no adjustment is necessary to the longitudinal compression strain limits for undermatched girth welds.

It must be pointed out that undermatched welds are rare for most existing pipelines and the majority of new pipelines. Weld undermatching has only become a significant concern with the recent introduction of high-strength pipelines steels consistent with API 5L Grade X70 and above. However, without careful weld specifications, it is possible to have undermatched welds for lower strength pipe. The difficulty in achieving weld overmatching may limit the ability of higher grade pipe steels from being used when it is necessary to resist significant seismic ground movements.

Assessing the Risk of Not Meeting Performance Goals
The ability of the pipeline to achieve the desired level of performance, expressed as an annual probability estimate, should be taken as the product of the annual probability that the pipeline will experience ground displacements greater than the capacity assuming factored strain limits and the probability that the pipe defects more severe than those accounted for exist in portions of the pipeline experiencing high strain levels. This can be expressed analytically in the following manner:

PP P G GD D
n
= −− 11bg (C3-2)
where:

PG = computed performance expressed as an annual probability of exceedance PGD = annual probability of exceedance for the earthquake producing the ground displacement hazard representing the pipeline capacity with reduced strain limits
PD = probability that the reduced strain limit for any individual weld is less than expected accounting for the variability in defect distribution and methods for relating defect size to strain capacity
n = number of welds exposed to strains above the reduced strain limits PR-268-9823
2248
SECTION C3: Pipeline Performance Criteria C3-14
As an example, consider the following conditions:

Analysis of a portion of a pipeline assuming nominal properties indicates that a pipeline subjected to permanent ground movement is adequate to meet the desired performance goal, defined as an annual probability of loss of pressure integrity no greater than 0.2%. The ground displacement corresponding to a 0.2% annual probability of exceedance is 2 m (6.6 ft) and is governed by longitudinal tension strain. Inspections of another portion of the pipeline built at the same time and by the same contractor indicate a potential for girth weld or corrosion defects. A reduced longitudinal tension strain capacity is determined based upon the most severe flaws identified by the prior inspection and fracture mechanics analyses. It is estimated that there is a 5% chance that the actual reduced strain capacity will be less than the determined reduced strain limit for any one weld. The reduced longitudinal tension strain capacity is exceeded for ground displacements greater than 1.25 m (4.1 ft). The analysis of pipeline response indicates that the length of pipeline exceeding the reduced strain limits is approximately 120 m at the full ground displacement of 2 m. The annual probability of exceeding 1.25 m of ground displacement is estimated to be 0.5%. The pipeline was constructed with 12-m (40-ft) joints of pipe.

The annual probability of loss of pressure integrity accounting for the potential for defects is estimated as follows:

PGD = 0.005
PD = 0.05
n = 120 m ÷ 12 m/joint = 10 girth welds
PG = −− = 0 005 1 1 0 05 0 002 10.. bg .

The pipeline with defects can be considered as meeting the performance goal since the estimated probability of exceeding the reduced strain limit is equal to the

0.2% defined by the performance goal. While the annual probabilities of exceeding acceptable longitudinal tension strain limits are the same, the probabilities associated with a particular time interval can be different.

For a 50-year time frame, T, the probability of ground movement exceeding 2 m (i.e., not meeting the desired performance goal) is computed as follows:

PP eT e
T = − − = −− = 1 1 1 1 002 0 0952 50 () . . bg
PR-268-9823
2249
SECTION C3: Pipeline Performance Criteria C3-15
where:

PeT = probability of exceeding nominal strain limit in time T Pe = annual probability of earthquake more severe than defined by the performance goal
T = time frame for computing PeT
The above calculation may greatly overestimate the likelihood of failure in 50 years because it assumes that the pipeline has a 100% probability of failure at the strains produced by a 2-m ground displacement.

The probability that the reduced longitudinal tension strain will be exceeded in the same 50-year time frame is computed as follows:
( ) ()
( ) ()
[ ][ ]
50 10
11 11
1 1 .005 1 1 .05

0.2217 0.4013 0.0890
Tn PP P erT GD D = −− −− ⎡ ⎤⎡ ⎣ ⎦⎣
= −− −− ⎡⎤⎡ ⎣⎦⎣
==
⎤
⎦
⎤
⎦
where:

PerT = probability of exceeding reduced strain limit in time T T = time frame for computing PerT
The differences in the above example probability calculations are not significant considering the level of uncertainty in estimating earthquake hazards. This is generally the case when the pipe meets the desired performance goal considering the effects of potential defects in a probabilistic manner. When it is useful to express the probability in terms of a reference time frame, basing calculations on only the annual probability of exceeding acceptable longitudinal strain capacity is generally sufficient.

PR-268-9823
2250
SECTION C3: Pipeline Performance Criteria C3-16
Table C3.1 Level 1 Longitudinal Tension Strain Limits for Special Applications Grade Y /T
(ksi) (inch) (mm) Defect Length
= 12.5 mm
Defect Length =

25.0 mm

0.250 6.4 0.98 0.55

0.375 9.5 1.55 0.90

0.500 12.7 1.88 1.20

0.750 19.1 2.22 1.66

1.000 25.4 2.34 1.84

0.250 6.4 0.96 0.54

0.375 9.5 1.51 0.88

0.500 12.7 1.85 1.17

0.750 19.1 2.18 1.63

1.000 25.4 2.26 1.81

0.250 6.4 0.87 0.49

0.375 9.5 1.39 0.79

0.500 12.7 1.74 1.09

0.750 19.1 2.07 1.54

1.000 25.4 2.16 1.72

0.250 6.4 0.85 0.49

0.375 9.5 1.37 0.78

0.500 12.7 1.72 1.07

0.750 19.1 2.05 1.52

1.000 25.4 2.14 1.70

0.250 6.4 0.78 0.45

0.375 9.5 1.26 0.71

0.500 12.7 1.61 0.99

0.750 19.1 1.94 1.44

1.000 25.4 2.01 1.63

0.890

0.917
Strain Limit (%)
X80
X52
X60
X65
X70
Wall
Thickness

0.841

0.850

0.883
The strain limits may be applied if all of the conditions apply:

1. Actual Y/T ratio no greater than the list values,

2. Minimum Charpy energy no less than 20 J (15 ft-lb) and average Charpy energy no less than 27 J (20 ft-lb),

3. Weld strength overmatching,

4. High-low misalignment within workmanship criteria, and

5. Undercut within workmanship criteria.

PR-268-9823
2251
SECTION C3: Pipeline Performance Criteria C3-17
Table C3.2 Level 2 Longitudinal Tension Strain Limits for Special Applications Grade Y /T
(ksi) (inch) (mm) Defect Length
= 12.5 mm
Defect Length =

25.0 mm

0.250 6.4 1.44 0.78

0.375 9.5 2.27 1.31

0.500 12.7 2.78 1.78

0.750 19.1 3.27 2.46

1.000 25.4 3.41 2.71

0.250 6.4 1.39 0.76

0.375 9.5 2.20 1.27

0.500 12.7 2.70 1.73

0.750 19.1 3.19 2.41

1.000 25.4 3.33 2.66

0.250 6.4 1.25 0.67

0.375 9.5 2.03 1.14

0.500 12.7 2.56 1.60

0.750 19.1 3.05 2.28

1.000 25.4 3.19 2.54

0.250 6.4 1.22 0.66

0.375 9.5 1.99 1.11

0.500 12.7 2.52 1.57

0.750 19.1 3.01 2.24

1.000 25.4 3.15 2.51

0.250 6.4 1.09 0.59

0.375 9.5 1.83 1.00

0.500 12.7 2.36 1.44

0.750 19.1 2.85 2.11

1.000 25.4 2.99 2.38
Wall
Thickness Strain Limit (%)
X52 0.841
X60 0.850
X65 0.883
X70 0.890
X80 0.917
The strain limits may be applied if all of the conditions apply:

1. Actual Y/T ratio no greater than the list values,

2. Minimum Charpy energy no less than 30 J (22 ft-lb) and average Charpy energy no less than 40 J (30 ft-lb),

3. Minimum CTOD toughness no less than 0.1 mm and averaged CTOD toughness no less than 0.15 mm,

4. Weld strength overmatching,

5. High-low misalignment within workmanship criteria, and

6. Undercut within workmanship criteria.

PR-268-9823
2252
SECTION C3: Pipeline Performance Criteria C3-18
Table C3.3 Level 3 Longitudinal Tension Strain Limits for Special Applications Grade Y /T
(ksi) (inch) (mm) Defect Length
= 12.5 mm
Defect Length =

25.0 mm

0.250 6.4 1.88 1.01

0.375 9.5 2.98 1.72

0.500 12.7 3.65 2.35

0.750 19.1 4.05 3.23

1.000 25.4 4.25 3.56

0.250 6.4 1.81 0.97

0.375 9.5 2.88 1.65

0.500 12.7 3.55 2.28

0.750 19.1 3.95 3.16

1.000 25.4 4.15 3.48

0.250 6.4 1.62 0.85

0.375 9.5 2.66 1.48

0.500 12.7 3.34 2.11

0.750 19.1 3.74 2.99

1.000 25.4 3.94 3.33

0.250 6.4 1.58 0.83

0.375 9.5 2.60 1.44

0.500 12.7 3.29 2.06

0.750 19.1 3.69 2.94

1.000 25.4 3.89 3.28

0.250 6.4 1.40 0.73

0.375 9.5 2.37 1.29

0.500 12.7 3.08 1.88

0.750 19.1 3.48 2.76

1.000 25.4 3.68 3.10
Wall
Thickness Strain Limit (%)
X52 0.841
X60 0.850
X65 0.883
X70 0.890
X80 0.917
The strain limits may be applied if all of the conditions apply:

1. Actual Y/T ratio no greater than the list values,

2. Minimum Charpy energy no less than 40 J (30 ft-lb) and average Charpy energy no less than 55 J (41 ft-lb),

3. Minimum CTOD toughness no less than 0.13 mm and averaged CTOD toughness no less than 0.20 mm,

4. Weld strength overmatching,

5. High-low misalignment within workmanship criteria, and

6. Undercut within workmanship criteria.

PR-268-9823
2253
SECTION C3: Pipeline Performance Criteria C3-19

0.00 0.02

0.04 0.06

0.08 0.10

0.12 0.14

0.16 30 40 50 60 70 80 90 100
D/t
Longitudinal Compression Strain
(a) Comparison of Test Data on Maximum Tested Strain Over Gauge Length

0.00 0.01

0.02 0.03

0.04 0.05
10 20 30 40 50 60 70 80 90 100
D/t
Longitudinal Compression Strain
Grey circles from tests to onset of
wrinkling with no internal pressure
(Stephens et al., 1991)
(b) Comparison of Test Data on Wrinkling and Maximum Moment Equation (3-1)
Equation (3-3a) for X70, p/py = 0.72, ∆/t = 0.10
Mohareb et al. (1994)
Ghodsi et al. (1994)
Zimmerman et al. (1995)
Dorey et al. (2001)
Figure C3.1 Comparison of Recommended Compression Strain Limits with Test Data PR-268-9823
2254
SECTION C3: Pipeline Performance Criteria C3-20
(a) Pipe Span Created by Landslide or Lateral Spread
(b) Soil Load Created by Shallow Soil Slide
Figure C3.2 Examples of Load-Controlled Configurations for Buried Pipelines PR-268-9823
2255
SECTION C4: Pipeline Analysis Procedures C4-1
C4.0 PIPELINE ANALYSIS PROCEDURES
Experience in the oil and gas industry with respect to the analytical evaluation of buried pipeline response to large permanent ground deformation dates back to the mid-1970s.

This experience includes simple evaluation methodologies as well as more sophisticated finite element approaches. A brief summary of this history and the experience of pipelines undergoing large ground deformations is presented as a means to introduce the discussion of analytical methods and results in these guidelines.

Differences in the relative stiffness between the pipe and the soil lead to relative pipe-soil movements. Large relative movement is an indication that the pipeline has considerably more strength than the surrounding soil at that particular location. Therefore, highstrength pipeline material is generally more beneficial in resisting highly localized soil deformation.

The geometry of the pipeline may allow for concentration of loading and relative displacement, leading to the development of locally large strains in the pipeline. Because these load conditions are produced by an imposed displacement, relatively large strains can be accepted provided the pipeline is in good condition and the girth welds are capable of developing gross-section yielding of the pipe. To accommodate the large strains, it is beneficial to have a ductile material (such as the steels considered in this evaluation) that hardens gradually between its yield and ultimate strength capacity. This allows for a more uniform distribution of strain and lessens the likelihood of local tearing of the pipe wall.

The first approach for estimating the response of buried pipelines to large ground deformations is generally credited to Newmark and Hall (1975). Their approach provided a means to quantify the amount of strain that could be accumulated in a buried pipeline crossing a fault. The basic components of the analysis are the stress-strain properties of the pipeline steel, the ground deformation geometry, and an estimate of the variation of the axial loading placed on the pipeline by the soil. The analysis assumed a straight pipeline configuration across a fault-like discontinuity.

A modification of the Newmark and Hall approach, developed by Kennedy et al. (1977), allowed the incorporation of bending effects into the analysis. The Kennedy et al. method accounted for bending by relating the lateral soil forces on the pipeline to an assumed amount of imposed pipeline curvature. The modified approach of Kennedy et al. is best suited for instances of large ground movement for which the bending strength of the pipeline, while significant, is not a dominant feature of the pipeline response. For small amounts of deformation where bending effects are significant, Kennedy et al. outlined the process for including a local bending analysis into their methodology.

However, this process is cumbersome and is rarely considered practical compared to finite element methods.

Both the Newmark and Hall and Kennedy et al. approaches are practically limited to consideration of initially straight pipelines, single components of lateral offset (either vertical or horizontal), constant soil-pipeline interaction parameters and negligible pipe PR-268-9823
2256
SECTION C4: Pipeline Analysis Procedures C4-2
bending stiffness. By the late 1970s, use of nonlinear finite element techniques were accepted as the preferred means to analyze all but the most simple of problems. Finite element approaches provide a means to rapidly investigate the effects of changes in backfill characteristics, pipeline material, wall thickness, and pipeline alignment. In addition, the assumptions used to formulate closed-form approaches like those of Newmark and Hall and Kennedy et al. have not been substantiated in comparison with finite element analyses as recommended in these guidelines.

The mechanics of implementing a finite element analysis have changed little in the last 15 to 20 years. The primary advancement in performing analyses has been in the availability of powerful desktop personal computers and compatible nonlinear analysis software that accounts for material yielding and large deformations.

The recommended analysis approach models the pipeline with beam-like elements capable of capturing nonlinear material behavior and the effect of large deformations. It is preferred that the analysis software include a pipe element in its element library with the capability to model internal pressure and provide output at various circumferential locations. These characteristics limit the amount of effort necessary to define pipe properties and the effort related to interpretation of the output from the analysis.

The recommended analysis approach can not account for localized straining that occurs once pipe wall buckling initiates. The recommended approach will underestimate local and global post-buckling strains in the region of the buckle. However, because of the displacement-controlled loading, and the fact that research into pipeline behavior has not, to date, been able to develop monotonically increasing strains of sufficient magnitude to produce rupture in a buckle (i.e., no rupture except by burst testing), this is considered acceptable. Also, refined calculations into the post-buckling region, based upon the use of nonlinear shell elements, have not yet been established to the point where they can be considered to be design tools.

C4.1 Pipe Element Definition
The one-half diameter limitation on the length of pipe elements in areas of high strain is related to the gauge length typical of full-scale pipeline tests. Shorter pipe elements promote the localization of plastic strains that have been observed in the tests.

Stress intensification and flexibility factors are based upon elastic pipeline response and are not applicable to analysis of nonlinear pipe behavior. The use of multiple straight pipe elements to represent bends and elbows introduces localization of strains that may be overly conservative considering the greater flexibility typical of elbows. However, the beneficial effect of elbow flexibility decreases with increasing D/t and internal pressure.

Conservatism in the treatment of elbows is warranted because of the lack of sufficient tests to determine applicable strain limits for elbows. However, tests on Grade B elbows have demonstrated that they have the ability to withstand severe deformation while maintaining internal pressure integrity (Yoshizaki et al., 1998).

PR-268-9823
2257
SECTION C4: Pipeline Analysis Procedures C4-3
The use of sections of elbows cut to provide angle changes other than standard 90° or 45°. The wall thickness of elbow fittings is typically not uniform around the circumference will vary depending among different manufacturers. Because of these factors, sections of elbow fittings require special attention because of the potential for a significant mismatch in wall thickness between the pipe and the elbow section. This mismatch in wall thickness can lead to stress risers and reduce the strain capacity of the welded connection.

The length of the pipeline model should be sufficient to adequately capture the anchoring effects of the soil outside the zone of ground movement. Extending the pipeline model a considerable distance outside the zone of ground movement typically does not significantly lengthen the solution time for the analysis as the pipeline response is mostly elastic. In cases where a particular analysis program is impacted by an extended pipeline model, a shorter length can be used if analyses are performed to confirm that longer models do not appreciably change the maximum computed strains.

It is assumed that the pipeline analyses are performed using a multi-linear uniaxial stressstrain curve to represent the pipeline material. The procedure for defining the compressive stress-strain curve is an attempt to capture trends exhibited in full-scale pipe tests. Pipeline test results demonstrate an abrupt softening in the moment versus curvature behavior of unpressurized pipelines strained beyond the point of maximum moment capacity (Yoosef-Ghodsi et al., 1994, Mohareb et al. 1994, Zimmerman et al., 1995).

Not capturing the reduction in moment capacity that accompanies formation of a local buckle in the pipe wall is not a significant shortcoming. As discussed in Section C4.2, the representation of the soil loading does not account for the reduction in maximum soil load at large relative pipe-soil displacement. The use of a constant maximum soil load forces continued deformation of the pipeline once yielding is reached. Localization of bending strains is not a concern because the pipeline is limited to the displacement experienced by the surrounding soil.

C4.2 Pipe Stress-Strain Definition
(No additional commentary provided)
C4.3 Soil Spring Definition
Relative movement of the surrounding soil with respect to a buried structural element imparts loads on that element. The magnitude of these loads was initially investigated to understand the performance of footings, piles and soil anchors. In the early to late 1970s, several research programs developed soil loading relationships specifically for buried pipelines. A key characteristic of soil loading is that it increases only to the point at which gross failure of the soil occurs. For example, the maximum lateral load that can be imparted to a buried pipe is related to the load necessary to develop a failure plane in the PR-268-9823
2258
SECTION C4: Pipeline Analysis Procedures C4-4
soil. Once this load has been reached, further relative displacement serves primarily to move soil along the failure plane.

The limited nature of soil loading on the pipeline is commonly represented in analytical approaches by modeling the soil with bilinear springs. This approach is conservative because forces necessary to carry soil along the established failure planes are typically less than those needed to initially generate the failure plane in the soil. An approximate analogy in mechanics is the relationship between static and sliding friction.

It is possible to estimate the reduction in soil loads for large relative displacements based upon soil mechanics analyses or test data. The analysis approach using beam-like pipe elements in the analysis of pipe response is not compatible with such soil loading representations because of the inability to capture the reduction in moment capacity at large pipe strains. If a more refined estimate of soil loading is used, the modeling of the pipe should be modified to capture the localized straining that occurs in the vicinity of a buckle in the pipe wall.

There is considerable uncertainty regarding the relationships used to compute soil spring properties. Most of this uncertainty is related to estimates of the soil strength parameters.

The uncertainty in estimating soil strength parameters for pipeline analyses in cohesive soils is further complicated by the fact that pipelines are typically located above the water table and within the desiccation zone of the soil. The strength of partially saturated desiccated soils is not well defined in soil mechanics practice.

The equations provided for the parameters Nqh, Nch, Nqv Ncv, and Nγ are developed from empirical curves found in the literature. The equations are particularly useful when implementing calculations in a spread sheet or computer program. The curves from which the equations were developed are provided in Figures C4.1 to C4.4.

Most of the soil loading relationships in these guidelines are nearly the same as those recommended by the Gas and Liquid Fuel Lifelines Committee of the ASCE Technical Council on Lifeline Earthquake Engineering (ASCE, 1984). One exception is the relationship for estimating axial soil loads in cohesive soils. These guidelines utilize the adhesion curve developed by Honegger (1999b).

There is considerable recent research supporting the use of the approach in these guidelines for estimating soil loads. Examples include centrifuge modeling tests (Paulin et al., 1995, 1996), field tests (Cappelletto et al., 1998, Honegger, 1999b, Rizkalla et al., 1996), and finite element analysis (Altaee et al., 1996). There is some evidence from laboratory tests that the lateral soil loads determined by the equations in these guidelines may overestimate the maximum soil loads (Paulin et al., 1998). However, this evidence is from an ongoing proprietary research program, and there are a number of technical issues relating to modeling similitude and load rate that require resolution. Until the findings from this proprietary research become available for review or other data are published from tests at seismic loading rates, the expressions contained in these guidelines represent the best approaches available. As an alternative, it is always permissible to utilize test data for a particular soil condition.

PR-268-9823
2259
SECTION C4: Pipeline Analysis Procedures C4-5
C4.4 Extent of Pipeline Model
(No additional commentary provided)
C4.5 Representation of Applied Ground Movement
The use of abrupt transition in the applied ground movement at faults and boundaries of lateral spreads and landslides is required in these guidelines unless site-specific information is available to justify other patterns. This is not an unrealistic assumption based upon the observation of abrupt transition patterns in past earthquakes. The assumption of an abrupt transition is conservative in that the computed strains are likely to be substantially greater than those computed assuming a more gradual transition. This approach is consistent with the representation of soil loading and analysis methods used in these guidelines. That is, the definition of ground deformation hazard, as represented by the soil loads and deformation pattern, is conservative while the assessment of pipeline response, represented by the strain criteria and pipeline modeling approaches, is believed to be more median-centered.

An alternate approach for expressing ground displacement based upon observations from Japanese earthquakes is to assume a pattern defined by a cosine function raised to the power of n as given in equations (C4-1) and (C4-2).

For slide-type displacements:
( ) 1 cos for 0.5
1 cos 1 for 0.5
n
ss
n
ss
xx
yx WW
xx
WW
π δ
δπ
⎛⎞ =− ⎜⎟ ≤
⎝⎠
⎛⎞ ⎡⎤ ⎛⎞ =− ⋅ − ⎜⎟ ⎢⎥ ⎜⎟ >
⎝⎠ ⎣⎦ ⎝⎠
(C4-1)
For fault-type displacements:
( ) 1 cos 2
n
f
x
yx W
π δ
⎛⎞ =− ⎜⎟
⎝⎠
(C4-2)
where:
y(x) = displacement as a function of location, x, within the zone of deformation x = distance across the zone of deformation
H = rupture depth, km
δ = amount of displacement
Ws = width of slide zone
PR-268-9823
2260
SECTION C4: Pipeline Analysis Procedures C4-6
Wf = width of fault zone
The cosine functions given by equations (C4-1) and (C4-2) are graphed in Figure C4.5 for various powers of n.

A displacement pattern corresponding to an n of 10 is representative of slide and fault displacements. A more abrupt transition, approximated by an n value greater than 100 has also been observed in some earthquakes. Other displacement patterns established based upon historical observation or rational analysis may be used for characterizing ground movements.

Practically, current empirical or analytical methods for quantifying seismic ground movement hazards are also not capable of quantifying the behavior at boundaries. The most appropriate basis for estimating more gradual transitions in ground movement is historical evidence. This type of information is typically limited to areas of high seismic activity where a historical record of observed ground deformations exits. If the basis for a more gradual transition is judgment, the sensitivity analyses should include an abrupt transition in ground deformation pattern.

C4.6 Sensitivity Analyses
Sensitivity analyses are recommended to provide some indication of the sensitivity of computed pipeline strains to various input parameters. The primary reason for performing sensitivity analyses is to assess the impact of uncertainty regarding definition of the earthquake hazard and soil loading parameters.

The recommended variations in these guidelines are based upon the premise that the soil spring definitions in these guidelines tend to overestimate soil loads. Some determination of the expected variability in modeling parameters should be included in the scope of any assessment. While higher soil loads generally produce more severe pipeline distortion, there are certain exceptions. A simple example illustrates this point.

Consider a pipeline alignment shown in Figure C4.6 subjected to right-lateral strike-slip faulting. The pipeline has an induction bend west of the fault and a 90° 3R elbow on the east side of the fault approximately 350 m from the induction bend. Given the shallow crossing angle, the response of the pipeline is governed by axial loads between the two bend locations. With this configuration, the two bends have the potential to behave as anchors. Overestimating the axial soil loads on the pipeline will tend to focus high strains at the induction bend. As the pipeline strains axially, the soil loads gradually reduce and may be insignificant at the 3R elbow if the total axial strain is sufficient to allow no relative displacement at the elbow. If the soil loads are underestimated, the soil will tend to slide along a much greater length of pipeline because the strains induced by the low soil loads are not sufficient to match the ground displacement. In this case, the strains at the induction bend may be reduced but one leg of the 3R elbow will be forced by high lateral soil loads to match the difference between the ground displacement and PR-268-9823
2261
SECTION C4: Pipeline Analysis Procedures C4-7
the axial pipe elongation. In effect, soil reaction forces developed at the 3R elbow will be pulling the pipe between the bends to accommodate the displacement mismatch. This condition can shift the most critical location of pipeline straining from the induction bend to the 3R elbow.

This example demonstrates that possible variations in soil spring definition can not only define the variability in computed strains but may also indicate a potential for alternate critically loaded portions of the pipeline alignment. When selecting variations of soil springs for sensitivity analyses, the potential for such behavior should be considered.

This may involve a more extreme variation in soil spring definition and a separate assessment to determine the likelihood of experiencing such an extreme variation.

PR-268-9823
2262
SECTION C4: Pipeline Analysis Procedures C4-8
0 0.2

0.4 0.6

0.8 1

1.2 1.4

1.6 0 25 50 75 100 125 150
Undrained Shear Strength, kPa
Adhesion Factor
Recommended in Guidelines
ASCE (1984)
Tomlinson (1957)
Rizkalla et al. (1996)
Sladen (1992) for effective stress of 20 kPa
Rizkalla et al. (1996) Pipeline Data
Rizkalla et al. (1996) Test Specimen Data
Paulin et al. (1998) Data
Honegger (1999) Data (3 locations)
Honegger (1999) Data (1 location)
Figure C4.1 Plotted Values for the Adhesion Factor, α
PR-268-9823
2263
SECTION C4: Pipeline Analysis Procedures C4-9
Figure C4.2 Values of Nqh and Nch (from Hansen, 1961)
PR-268-9823
2264
SECTION C4: Pipeline Analysis Procedures C4-10
Figure C4.3 Ranges for Values of Nqv and Ncv (from Trautman and O’Rourke, 1983) PR-268-9823
2265
SECTION C4: Pipeline Analysis Procedures C4-11

5.14 1.00
0 10
20 30
40 50
60 70
80 0 10 20 30 40 5
φ, degrees
Nc
Nq
Nγ
0 Figure C4.4 Plotted Values of Bearing Capacity Factors (Nq, Nc, and Nγ) PR-268-9823
2266
SECTION C4: Pipeline Analysis Procedures C4-12

0.0 0.2

0.4 0.6

0.8 1.0
0 0.2 0.4 0.6 0.8 1
x/Ws
δ
n = 2
n = 4
n = 10
n = 20
n = 50
n = 500
a) Cosine Function Representation of Slide Displacement

0.0 0.2

0.4 0.6

0.8 1.0
0 0.2 0.4 0.6 0.8 1
x/Wf
δ
n = 2
n = 4
n = 6
n = 10
n = 20
n = 40
b) Cosine Function Representation of Fault Displacement Figure C4.5 Cosine Function for Representing Variation in Ground Displacement Patterns
PR-268-9823
2267
SECTION C4: Pipeline Analysis Procedures C4-13

INDUCTION
BEND
90° 3R ELBOW

IDEALIZED AXIAL

STRAIN VARIATION

FOR LOW AXIAL

FRICTION

IDEALIZED

AXIAL STRAIN

VARIATION FOR

HIGH AXIAL

FRICTION

DIRECTION OF
FAULT

MOVEMENT
FAULT
NORTH
Figure C4.6 Example of Pipeline Configuration Illustrating Consequences of Variability in Soil Spring Forces
PR-268-9823
2268
SECTION C5: Mitigation Options C5-1
C5.0 MITIGATION OPTIONS
As with pipeline design and assessment guidelines, the benefits of implementing mitigation measures have generally not been demonstrated by experience in actual earthquakes or in controlled tests. A notable exception being the limited field testing performed to validate the concept of dual layers of geotextile wrapping. The presentation of mitigation options is limited to a general discussion of various techniques that have been implemented or considered in the past. For this reason, no additional discussion of potential mitigation options is provided in this commentary.

PR-268-9823
2269
SECTION C6: References C6-1
C6.0 REFERENCES

1. Altaee, A., B.H. Fellenius and H. Salem, 1996. “Finite Element Modeling of Lateral Pipeline-Soil Interaction,” Proceedings of the 15th International Conference on Offshore Mechanics and Arctic Engineering, American Society of Mechanical Engineers, Vol. V, pp. 333-341.

2. American Gas Association, 1990. “Collapse of Offshore Pipelines,” seminar notes, Houston, TX, February 20.

3. American Petroleum Institute, 1990. Specification for Line Pipe, API Specification 5L.

4. American Petroleum Institute, 1993. “Design, Construction, Operation, and Maintenance of Offshore Hydrocarbon Pipelines,” API Recommended Practice 1111, Second Edition, November.

5. American Petroleum Institute, 1999. API Recommended Practice 1104, Nineteenth Edition, September.

6. American Petroleum Institute (1993). Bulletin on Formulas and Calculations for Casing, Tubing, Drill Pipe, and Line Pipe Properties, API Bulletin 5C3.

7. American Society of Civil Engineers (ASCE), 1984. “Guidelines for the Seismic Design of Oil and Gas Pipeline Systems,” Technical Council on Lifeline Earthquake Engineering, Committee on Gas and Liquid Fuel Lifelines, New York.

8. American Society of Mechanical Engineers, 1991. Manual for Determining the Remaining Strength of Corroded Pipelines, ASME B31G-1991.

9. American Society of Mechanical Engineers, 1992. Pipeline Transportation Systems for Liquid Hydrocarbons and Other Liquids, ANSI Standard B31.4, New York.

10. American Society of Mechanical Engineers, 1995. Gas Transmission and Distribution Piping Systems, ANSI Standard B31.8, New York.

11. Bartlett, S.F. and T.L. Youd, 1992. “Empirical Analysis of Horizontal Ground Displacement Generated by Liquefaction-Induced Lateral Spreads,” NCEER-92- 0021, National Center for Earthquake Engineering Research, State University of New York at Buffalo, August.

12. Bowkamp, J.G. and R.M. Stephen, 1973. “Large Diameter Pipe Under Combined Loading,” Journal of the Transportation Engineering Division, ASCE, vol. 99, no.

TE3, pp. 521-536.

13. British Standards Institute, 1993. Code of Practice for Pipelines, BS 8010.

14. Canadian Standard Association (CSA), 1996. Oil and Gas Pipeline Systems, Z662-

96. 15. Cappelletto, A., R. Tagliaferri, G. Giurlani, G. Andrei, G, Furiani and G. Scarpelli, 1998. “Field Full Scale Tests on Longitudinal Pipeline-Soil Interaction,” Proceedings of the International Pipeline Conference, Calgary, Alberta, American Society of Mechanical Engineers, pp. 771-778.

PR-268-9823
2270
SECTION C6: References C6-2

16. Denys, R.M., 1992. “Fracture Behavior of Girth Welds Containing Natural Defects, Comparison with Existing Workmanship Standards,” Final Report to the Welding Supervisory Committee, PR-202-009.

17. Denys, R.M., 1993. “The Effect of Weld Metal Matching on Girth Weld Performance, “ Final Report to the Welding Supervisory Committee, PR-202-922.

18. Denys, R.M., 1994. “Fracture Behavior of Large Diameter Pipeline Girth Welds:

Effect of Weld Metal Yield Strength and Defect Interaction,” Final Report to the Welding Supervisory Committee, PR-202-011.

19. Det Norske Veritas, 2000. Rules for Submarine Pipelines.

20. Dorey, A.B., J.J.R. Cheng and D.W. Murray, 2001. “Critical Buckling Strains for Energy Pipelines,” Structural Engineering Report 237, Department of Civil Engineering, University of Alberta, Canada.

21. Foeken, R.J. and A.M. Gresnigt, 1998. “Buckling and Collapse of UOE Manufactured Steel Pipes,” TNO-Report 96-CON-R0500, prepared for PRC International.

22. Gresnigt, A.M., 1986. “Plastic Design of Buried Steel Pipelines in Settlement Areas,” HERON, vol. 31, no. 4.

23. Hansen, J.B., 1961. “The Ultimate Resistance of Rigid Piles Against Transversal Forces,” Bulletin 12, Danish Geotechnical Institute, Copenhagen, Denmark.

24. Honegger, D.G., 1999a. “Assessment of Wave Propagation Effects on a 1925 Pipeline with Oxyacetylene Girth Welds,” Proceedings of the 5th National Conference on Lifeline Earthquake Engineering, American Society of Civil Engineers, August.

25. Honegger, D.G., 1999b. “Field Measurement of Axial Soil Friction on Buried Pipelines,” Proceedings of the 5th National Conference on Lifeline Earthquake Engineering, American Society of Civil Engineers, August.

26. Honegger, D.G., 2000. “Wave Propagation Versus Transient Ground Displacement as a Credible Hazard to Buried Oil and Gas Pipelines,” Proceedings of the International Pipeline Conference 2000, Calgary, Canada.

27. Horsley, D.J., A.G. Glover and R. Denys, 1997. “An Assessment Technique for Defects in Under and Over Matched Pipeline Girth Welds,” Proceedings of the 11th Biennial PRCI-EPRG Joint Technical Meeting, Arlington, VA, April 7-11, Paper 30.

28. International Conference of Building Officials, 2000. International Building Code (IBC), Whittier, California.

29. Ishihara, K., 1990. “Evaluation of Liquefaction Potential and Consequent Deformations in Sand Fills,” Proceedings of a Workshop on Seismic Issues, Port of Los Angeles, San Pedro, CA, March 21-23.

30. Ishihara, K., 1985. “Stability of Natural Deposits During Earthquakes,” Proceedings of the Eleventh Conference on Soil Mechanics and Foundation Engineering, San Francisco, Vol. 1.

PR-268-9823
2271
SECTION C6: References C6-3

31. Jones, D.J., G.S. Kramer, D.N. Gideon and R.J. Eiber, 1986. “An Analysis of Reportable Incidents for Natural Gas Transmission and Gathering Lines 1970 through June 1984,” Report to Line Pipe Research Supervisory Committee of the American Gas Association, Battelle, Columbus, Ohio.

32. Kennedy, R.P., A.W. Chow and R.A. Williamson, 1977. “Fault Movement Effects on Buried Oil Pipeline,” Journal of the Transportation Engineering Division, ASCE, vol. 103, no. TE5, pp. 617-633.

33. Larkin, T.J. and E.J. Marsh, 1982. “Two Dimensional Nonlinear Site Response Analysis,” Proceedings Pacific Conference on Earthquake Engineering, Vol. 3, Aukland, New Zealand, pp. 217-227.

34. Lee, M.K.W. and W.D.L. Finn, 1978. “DESRA-2: Dynamic Effective Stress Response Analysis of Soil Deposit with Energy Transmitting Boundary Including Assessment of Liquefaction Potential,” Department of Civil Engineering, University of British Columbia, Vancouver, Canada.

35. Mohareb, M.E., A.E. Elwi, G.L.Kulak and D.W. Murray, 1994. “Deformational Behavior of Line Pipe,” Structural Engineering Report 202, Department of Civil Engineering, University of Alberta, Canada.

36. Murphey C.E. and C.G. Langner, 1985. “Ultimate Pipe Strength Under Bending, Collapse, and Fatigue,” Proceedings of the Offshore Mechanics and Arctic Engineering Conference, pp. 467-477.

37. Nelson, I. and M.L. Baron, 1981. “Earthquakes and Underground Pipelines - An Overview,” Grant Report No. 17, prepared for the National Science Foundation by Weidlinger Associates, PFR 78-15049.

38. Newmark, N.M., and W.J. Hall, 1975. “Pipeline Design to Resist Large Fault Displacement,” Proceedings of the U.S. National Conference on Earthquake Engineering, Ann Arbor, Michigan, pp. 416-425.

39. Newmark, N.M., 1965. “Effects of Earthquakes on Dams and Embankments,” in Selected Papers by Nathan M. Newmark, W.J. Hall ed., American Society of Civil Engineers, pp. 631-652.

40. Nyman, D.J. and W.J. Hall, 1991. “Scenario for Improving the Seismic Resistance of Pipelines in the Central United States,” Proceedings, Third U.S. Conference on Lifeline Earthquake Engineering, ASCE, Los Angeles, California.

41. O’Rourke, M., 1988, "Seismic Analysis Procedures for Welded Steel Pipeline," Proceedings of 1st Japan - United States Workshop on Liquefaction, Large Ground Deformation and their Effects on Lifeline Facilities, Tokyo, Japan, November 16-18.

42. O’Rourke, T.D., 1991. “An Overview of Geotechnical and Lifeline Earthquake Engineering,” Geotechnical Earthquake Engineering and Soil Dynamics III, American Society of Civil Engineers, pp. 1392-1426.

43. Parnes, R. and P. Weidlinger, 1979. “Dynamic Response of an Embedded Pipe Subjected to Periodically Spaced Longitudinal Forces,” Grant Report No. 13, PR-268-9823
2272
SECTION C6: References C6-4
prepared for the National Science Foundation by Weidlinger Associates, PFR 78- 15049.

44. Paulin, M.J., R. Phillips and R. Boivin, 1995. “Centrifuge Modelling of Lateral Pipeline/Soil Interaction - Phase II,” Proceedings of the 14th International Conference on Offshore Mechanics and Arctic Engineering, American Society of Mechanical Engineers, Vol. V, pp. 107-123.

45. Paulin, M.J., R. Phillips and R. Boivin, 1996. “An Experimental Investigation Into Lateral Pipeline/Soil Interaction,” Proceedings of the 15th International Conference on Offshore Mechanics and Arctic Engineering, American Society of Mechanical Engineers, Vol. V, pp. 313-323.

46. Paulin, M.J., R. Phillips, J.I. Clark, A. Trigg and I. Konuk, 1998. “A Full-Scale Investigation into Pipeline/Soil Interaction,” Proceedings of the International Pipeline Conference, Calgary, Alberta, American Society of Mechanical Engineers, pp. 779-787.

47. Pease, J.D. and T.D. O’Rourke, 1997. “Seismic Response of Liquefaction Sites,” Journal of Geotechnical and Geoenvironmental Engineering, American Society of Civil Engineers, pp. 37-45.

48. Rizkalla, M., A. Trigg and G. Simmonds, 1996. “Recent Advances in the Modeling of Longitudinal Pipeline/Soil Interaction for Cohesive Soils,” Proceedings of the 15th International Conference on Offshore Mechanics and Arctic Engineering, American Society of Mechanical Engineers, Vol. V, pp. 325-332.

49. Roberts, K.A. and R.J. Pick, 1998. “Correction for Longitudinal Stress in the Assessment of Corroded Pipe,” Proceedings of the International Pipeline Conference, Calgary, Alberta, American Society of Mechanical Engineers, pp. 553- 561.

50. Rosenfeld, M.J., 1996. “Serviceability of Corroded Girth Welds,” report to Pipeline Research Committee International, Catalog No. L517452.

51. Rowe, R.K. and E.H. Davis, 1982a. “The Behavior of Anchor Plates in Sand,” Geotechnique, vol. 32, no. 1, pp. 25-41.

52. Rowe, R.K. and E.H. Davis, 1982b. “The Behavior of Anchor Plates in Clay,” Geotechnique, vol. 32, no. 1, pp. 9-23.

53. Seed, H.B. and I.M. Idriss, 1982. “Ground Motions and Soil Liquefaction During Earthquakes,” Earthquake Engineering Research Institute Monograph.

54. Seismological Society of America, 1997. Seismological Research Letters, Vol. 68, No. 1.

55. Southwell, R.V., 1914. Philosophical Transactions, Royal Society, London, Series A. vol. 213.

56. Stephens, D.R., R.J. Olson and M.J. Rosenfeld, 1991. “Pipeline Monitoring - Limit State Criteria,” Battelle Memorial Laboratory, AGA NG-18 report no. 188, September.

PR-268-9823
2273
SECTION C6: References C6-5

57. Tawfik M.S. and T.D. O’Rourke, 1985. “Load-Carrying Capacity of Welded Slip Joints,” Journal of Pressure Vessel Technology, Vol. 107, February, American Society of Mechanical Engineers.

58. Trautmann, C.H. and T.D. O'Rourke, 1983. “Behavior of Pipe in Dry Sand Under Lateral and Uplift Loading,” Geotechnical Engineering Report 83-6, Cornell University, Ithaca, New York.

59. Tseng, W.S. and C.H. Lee, 1995. “Structural Performance Criteria for Fitness-forService Evaluations of Underground Natural Gas Pipelines,” Proceedings of the Second National Conference on Advances in Underground Pipeline Technology, J.K.

Jeyapalan and M. Jeyapalan, ed., American Society of Civil Engineers, pp. 536-547.

60. U.S. Department of Transportation (DOT), 1998a. Transportation of Natural and Other Gas by Pipeline, Title 49, Code of Federal Regulations (CFR), Part 192.

61. U.S. Department of Transportation (DOT), 1998b. Transportation of Hazardous Liquids by Pipeline, Title 49, Code of Federal Regulations (CFR), Part 195.

62. Varnes, D.J., 1978. “Slope Movement Types and Processes,” Landslides Analysis and Control, Special Report 176, Transportation Research Board, National Academy of Sciences, pp. 11-33.

63. Wang, Y.Y., J.R. Gordon, M.T. Kirk, T. McGaughy, M.J. Rosenfield and I.

Roytman, 1996. “Plastic Deformation Capacity of Girth-Welded and Structural Sleeve Connections,” Pipeline Research Committee International, PR-185-9211.

64. Wang, W., M.Q. Smith, C.H. Popelar and J.A. Maple, 1998. “A New Rupture Prediction Model for Corroded Pipelines Under Combined Loads,” Proceedings of the International Pipeline Conference, Calgary, Alberta, American Society of Mechanical Engineers, pp. 563-572.

65. Wang, Y-Y, 2003. “Guidelines on Tensile Strain Limits,” report prepared by Engineering Mechanics Corporation of Columbus (EMC2
) under GRI Contract 8656
for the Design, Construction and Operations Technical Committee of Pipeline Research Council International, Inc., June.

66. Warke, R.W. and W.E. Amend, 1997. “Seismic Fitness of Girth Welds in and Early 1930s-Vintage Gas Transmission Pipeline,” Proceedings of the 1997 Pipeline Risk Management & Reliability Conference, Houston, TX.

67. Warke, R.W., C. Ferregut, A.G. Glover and D.J. Horsley, 1997. “A ReliabilityBased Method for Assessing the Fitness for Service of Pipeline Girth Welds,” Proceedings of the 11th Biennial PRCI-EPRG Joint Technical Meeting, Arlington, VA, April 7-11, Paper 29.

68. Wells, D.L. and K.J. Coppersmith, 1994. “New Empirical Relationships among Magnitude, Rupture Length, Rupture Width, Rupture Area, and Surface Displacement,” Bulletin of the Seismological Society of America, Vol. 84, No. 4, pp. 974-1002.

PR-268-9823
2274
SECTION C6: References C6-6

69. Wilson R.C. and D.K. Keefer, 1985. “Predicting Areal Limits of EarthquakeInduced Landsliding,” in Evaluating Earthquake Hazards in the Los Angeles RegionAn Earth Science Perspective, U.S. Geological Survey, Professional Paper 1360, pp. 317-345.

70. Yoosef-Ghodsi, N., G.L. Kulak and D.W. Murray, 1994. “Behavior of GirthWelded Line Pipe,” Structural Engineering Report 203, Department of Civil Engineering, University of Alberta, Canada.

71. Yoshizaki, K., H. Ando and N. Oguchi, 1998. “Large Deformation Behavior of Pipe Bends Subjected to In-Plane Bending,” Proceedings of the International Pipeline Conference, Calgary, Alberta, American Society of Mechanical Engineers, pp. 733- 740.

72. Youd, T.L., 1984. “Geologic Effects – Liquefaction and Associated Ground Failure,” Proceedings of the Geologic and Hydrologic Hazards Training Program, Open-file Report 84-760, U.S. Geological Survey, Menlo Park, California.

73. Youd, T.L., C.M. Hansen and S.F.Bartlett, 2002. “Revised Multilinear Regression Equations for Prediction of Lateral Spread Displacement,” Journal of Geotechnical and Geoenvironmental Engineering, American Society of Civil Engineers, Vol. 128, No. 12, pp. 1007-1017.

74. Youd, T.L., C.M. Hansen and S.F.Bartlett, 1999. “Revised MLR Equations for Predicting Lateral Spread Displacement,” Proceedings of the Seventh U.S.-Japan Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures Against Soil Liquefaction, MCEER-99-0019, Multidisciplinary Center for Earthquake Engineering Research, pp. 99-114.

75. Youd, T.L. and I.M. Idriss (eds.), 1997. Proceedings of the NCEER Workshop on Evaluation of Liquefaction Resistance of Soils, NCEER-97-0022, National Center for Earthquake Engineering Research.

76. Youd, T.L. and D.M. Perkins, 1978. “Mapping Liquefiable Ground Failure Potential”, Journal of the Geotechnical Engineering Division, ASCE, 104, 4, 433- 445.

77. Zimmerman, T.J.E, M.J. Stephens, D.D. DeGreer and Q. Chen, (1995). “Compressive Strain Limits for Buried Pipelines,” Proceedings of the 1995 Offshore Mechanics and Arctic Engineering Conference, American Society of Mechanical Engineers, Volume V, pp. 365-378.

PR-268-9823
2275
This Page Intentionally Left Blank
2276
Appendix A: Considerations for Establishing Performance Goals A-1

CONSIDERATIONS FOR ESTABLISHING
PERFORMANCE GOALSExisting pipeline standards emphasize minimum requirements for normal design, construction, and operation. Many pipeline companies exceed the requirements stipulated by the standards for safety and reliability. With regard to seismic hazard mitigation, the appropriate level of seismic design is typically determined on a project-by-project basis using site-specific information to define the seismic hazard exposure and the current state-of-practice. In practice, this often results in the definition of seismic performance goals for the pipeline being based largely upon criteria adopted from other non-pipeline projects, with similar consequences to the public.

A fundamental question to be answered in performing a seismic assessment for a new or existing pipeline is what level of performance is necessary. These guidelines assume a definition in terms of an annual probability of exceeding an unacceptable condition such as excessive pipe deformation or loss of pressure boundary integrity. This probability is then equated to the probability of occurrence of the earthquake.

A.1 Objectives
The main objective of seismic design for natural gas transmission and liquid hydrocarbon pipelines is to provide life/safety protection from the effects of stipulated design earthquakes distributed along the pipeline route. A secondary, but important, objective is to minimize capital loss and disruptions to operations.

The goal for seismic design of natural gas transmission and liquid hydrocarbon pipelines is to achieve a balanced design to withstand the effects of earthquakes and other loads which is both safe and economically feasible. The design should take into account the nature and importance of the project, cost implications, and risk assessment centering around such items as public safety, loss of product or service, and damage to property and the environment. The treatment of the seismic hazards should be consistent with the treatment of other natural and manmade hazards.

The concept of a balanced design may be difficult to achieve because many of the parameters necessary for making rational decisions are either unknown or not well defined. The goal of a balanced design is for all major components of a pipeline system to have failure risks and consequences that are consistent with system performance objectives. Pipelines and support facilities should be designed to withstand the effects of earthquake ground shaking and permanent ground movements that have a reasonable probability of occurrence along the route during their operational life.

A.2 Dual-Level versus Single-Level Earthquake Hazard
Definition
The use of dual-level earthquake design criteria is an outgrowth of practices developed for the nuclear power industry and other projects (e.g., offshore structures) where seismic PR-268-9823
2277
Appendix A: Considerations for Establishing Performance Goals A-2 damage has the potential for severe safety or environmental consequences. The origin of the two-level design earthquake was primarily intended for the design of buildings and surface structures, with their contained electrical and mechanical systems and aboveground piping systems. The first major pipeline project to adopt dual-level earthquake criteria was the Trans-Alaska pipeline, but design implementation was restricted primarily to aboveground pipeline segments.

Current practice for the seismic design of major pipeline facilities (new construction) often follows the precedent set for offshore oil production platforms, Department of Energy critical facilities, and nuclear power plants in that two levels of earthquake hazard are selected for design. In this approach, the low level event generally has a return period on the order of 200 to 500 years. The higher level event is generally one that has a return period on the order of 1,000 years or more, depending upon the nature of the facility.

The lower level event is often viewed as a threshold for continued operation of the pipeline within accepted safety margins and little if any permanent deformation of the pipeline is permitted. The pipeline response to higher level event is typically required to prevent a threat to safety, although significant structural damage could occur. Specific criteria for the response of the pipeline to these events, including the amount of permissible damage, varies according to the type of structure or component and its function. The selection of earthquake return periods and performance objectives should be established through a project-specific risk assessment.

There is no standard nomenclature for referring to the two earthquake levels. The lower level event is often referred to as the "probable design earthquake" (PDE), the “design basis earthquake” (DBE), or the “strength level earthquake” (SLE). The higher level event is commonly referred to as the "contingency design earthquake" (CDE), “maximum credible earthquake” (MCE), “safe shutdown earthquake” (SSE), or “ductility level earthquake” (DLE).

For the design or assessment of buried pipelines it is not clear whether a two-level design is necessary or practical in many cases. The fundamental difference in the condition expected of buried pipelines and typical structures and equipment following the low-level earthquake should be considered if a dual-level earthquake design is specified.

Establishing a lower level earthquake for structures and equipment generally assures that the post-earthquake condition is essentially the same as the pre-earthquake condition for the less severe earthquake. This is expected because their primary earthquake loading condition is a transitory inertial excitation. However, the primary earthquake load on buried pipelines is permanent ground displacement. The pipeline’s total strain capacity to resist ground displacement in the future is reduced because of the strains produced by the low-level earthquake displacements. Also, it is possible that landslide and lateral spread hazards are not direct functions of the level of ground shaking necessary to trigger large displacements. These factors should be considered with respect to the purpose of specifying a dual-level earthquake to provide an increased likelihood of normal operation for moderate events and prevent loss of pressure integrity for major events.

PR-268-9823
2278
Appendix A: Considerations for Establishing Performance Goals A-3 The decision on whether or not to specify a dual-level earthquake design for a buried pipeline may also be influenced by considering measures that may be required following an earthquake that produces permanent ground displacement. Following the earthquake, a decision will be necessary regarding the suitability of the pipeline for continued shortterm and long-term service. Considerations impacting this decision include the consequences if the pipeline performance is compromised by future ground displacement, the likelihood for future earthquake generated ground displacement during the life of the pipeline, the remaining capacity of the pipeline to withstand future ground displacement, and the consequences to customers if pipeline operations are interrupted.

Given that this decision process generally occurs regardless of the level of ground displacement experienced by the pipeline, it is not clear what benefits are to be gained by specifying two levels of earthquake hazard. This is especially true for cases where analysis provides information on pipe strains as a function of ground displacement such that the pipeline response for lower level events is understood.

Another consideration relates to the assessment of permanent ground displacement patterns after an earthquake when the ground displacement was insufficient to lead to loss of pipe contents. With the exception of well-defined abrupt surface expressions of displacement, it is often necessary to compare the positions of surface features following the earthquake that were known before the earthquake in order to evaluate the extent of ground deformation. Examples of such surface features include roads, sidewalks, fence lines, and survey monuments. In the case of pipelines within a liquefiable soil experiencing lateral spread movement, the problem is compounded by the fact that the ground movements within the liquefiable layer may not be well represented by surface ground displacement. In most earthquakes, detailed surveys are necessary to determine the amount of ground displacement. In remote locations, it is not unusual to have no preexisting features from which to ascertain ground displacement. In such cases, assessment of pipeline condition is often qualitative and relies on limited excavation of the pipeline to observe any signs indicative of large strains (e.g., wrinkles, damaged coating, gaps between the soil and the pipe, ovaling of the pipe). It may also be decided to use an instrumented pigging device to identify any abnormalities in the pipe wall or alignment.

In any case, assessing pipeline response to ground displacement following an earthquake may take several days or weeks.

These problems can be reduced somewhat by performing surveys to locate surface features or to install survey monuments prior to the earthquake in the vicinity of areas expected to experience ground displacement. Post-earthquake surveys using GPS equipment can then provide a relatively rapid assessment of the amount of ground displacement that has occurred in the vicinity of the pipeline. This entails some longterm costs for verifying the survey and maintaining survey monuments.

Whether a pipeline is to be constructed onshore or offshore may dictate the need to account for differences in the consequences of seismic damage when establishing performance goals. Seismic damage to an offshore pipeline would result in a more difficult repair situation than for an onshore pipeline, particularly in an extreme climate or deep water, or in cases where the pipeline is deeply buried to protect against anchor PR-268-9823
2279
Appendix A: Considerations for Establishing Performance Goals A-4 dragging, ice gouging, etc. The increased difficulty is due mainly to mobilizing offshore equipment and gaining access to the pipeline. Considering the differences in damage consequences for offshore and onshore pipelines and production platforms, less frequent seismic hazards (larger earthquake hazards) should be considered for offshore pipelines than for onshore pipelines. This can be achieved by selecting annual probabilities of earthquake hazards for offshore pipelines that are less than what would be selected for onshore pipelines (i.e., longer average return periods for seismic events).

If life safety is the primary criteria, the annual probability for experiencing a design earthquake hazard for an offshore pipeline may not need to be as low as for an offshore production platform, because pipeline damage does not generally lead to direct life-safety consequences. Pipeline failures tend to be localized (at one or more locations) and repairs might be implemented in a relatively short time frame if the water depth is not great. Such disruptions would have limited adverse consequences with respect to life safety. A very important exception is offshore oil pipelines and pipelines located in very deep water for which the environmental consequences and costs associated with pipeline repairs may be very severe. In these cases, the design criteria may be more stringent that what is necessary to meet established life-safety performance objectives.

A.3 Perspective on Seismic Risk
The principal consequence of the rupture of a natural gas transmission pipeline is the potential for release and subsequent ignition of gas in areas where there is a potential for injury or significant property damage. Other consequences include service interruption and the cost of repair and cleanup. Rupture of liquid hydrocarbon pipelines may have severe environmental consequences that are of equal or greater concern to the pipeline owner and govern the conservatism in defining the earthquake hazard. One of the most important and difficult issues associated with earthquake risk mitigation for natural gas and liquid hydrocarbon pipelines is the determination of the appropriate level of investment in seismic hazard mitigation versus the benefit that would be derived. Costs and benefits are difficult to assess due to the limited knowledge of earthquake recurrence, the uncertainty involved in projecting pipeline damage, and the quantification of the impact of pipeline service interruptions.

Pipelines are subject to risk of damage due to a number of causes totally unrelated to earthquakes. Among these are corrosion, outside force, material failure, and construction defects. An analysis of reportable incidents for natural gas transmission and gathering lines in the U.S. (Jones et al., 1986) indicates that approximately 685 failures, classified as propagating ruptures, punctures, blowouts or tears, occurred over a 14.5-year period.

The total length of natural gas transmission pipelines during this period was about 500,000 km. This leads to an average of approximately 1 x 10-4 failures per kilometer of pipeline per year. For a postulated 50-year service life, this failure rate equates to approximately 5 x 10-3 failures per kilometer or one failure per 200 km (Nyman and Hall, 1991).

PR-268-9823
2280
Appendix A: Considerations for Establishing Performance Goals A-5 If one considers the hypothetical case of an earthquake with an estimated recurrence interval of 500 years causing damage over a 100-km segment of the pipeline route, five such pipeline failures would be comparable to the statistics for normal operation. This failure rate probably represents an upper bound for most welded steel pipelines.

Admittedly, this analogy is rough and approximate, but it does illustrate the order of magnitude of the relative threat posed by strong earthquakes compared to normal operation.

In light of the foregoing example, it seems appropriate that improvement of seismic resistance through capital expenditures should be tempered by recognition of the hazards associated with day-to-day operations. While mitigation of the seismic threat through design is usually the preferred alternative, it may be more reasonable in certain cases to focus attention on minimizing the consequences of destructive earthquakes. In particular, the safety impact of extensive pipeline damage can be reduced through improved line break isolation, earthquake contingency plans, repair plans, and seismic hardening of control systems and communications.

PR-268-9823
2281
This Page Intentionally Left Blank
2282
Appendix B: Seismic Design Provisions in Existing Standards B-1

SEISMIC DESIGN PROVISIONS IN EXISTING STANDARDS
Requirements for liquid hydrocarbon pipelines in the United States are set forth in Title 49 of the Code of Federal Regulations (CFR), Parts 192 and 195 (U.S. Department of Transportation, 1998a, 1998b), which regulate the design, construction, operation, and maintenance of natural gas and crude oil pipelines, respectively. ASME B31.4 and ASME B31.8 are consensus standards developed in accordance with the requirements of the American National Standards Institute (ANSI) which have been developed and maintained by technical committees of the American Society of Mechanical Engineers (ASME) (1998, 2000). The ANSI standards are supplemented by industry standards such as API Recommended Practice 1111 for the design construction, operation, and maintenance of offshore hydrocarbon pipelines (American Petroleum Institute, 1993).

In Canada, liquid hydrocarbon and natural gas pipelines are designed in accordance with Canadian Standards Association (CSA) Z662, Design of Oil and Gas Pipeline Systems.

CSA Z662 covers both onshore and offshore pipelines.

In Europe, two standards are commonly used to control the design and operation of offshore pipelines: British Standard BS 8010, Code of Practice for Pipelines, Part 3 (British Standards Institute, 1993), and Rules for Submarine Pipeline Systems published by Det Norske Veritas (DNV) (2000).

None of the aforementioned standards provide specific requirements for seismic design of crude oil and natural gas pipeline systems. Typically, these standards acknowledge that consideration should be given to the various types of seismic hazards, but they do not prescribe specific criteria or methodology. Reference is often made to “giving consideration to,” “providing reasonable protection,” or “taking reasonable precautions,” but the level of design treatment is undefined. The responsibility for assuring that adequate provisions have been made to address seismic hazards is delegated to the pipeline designer and the appropriate regulatory body.

The references made by the above-listed standards to seismic design is summarized below.

B.1 ANSI B31.4, Pipeline Transportation Systems for Liquid Hydrocarbons and Other Liquids
ANSI B31.4 is the U.S. standard for the design, construction, operation, and maintenance of liquid hydrocarbons pipelines, both onshore and offshore. ANSI B31.4 makes only limited reference to seismic design by acknowledging the need to consider certain situations involving earthquake-induced loads. No specific criteria, requirements or guidelines are offered. The major objective is to create a general awareness of the need to evaluate seismic hazards.

1. Paragraph 401.5.3, Earthquake (relating to dynamic effects). “Consideration in the design shall be given to piping systems located in regions where earthquakes are PR-268-9823
2283
Appendix B: Seismic Design Provisions in Existing Standards B-2 known to occur.” This paragraph acknowledges the need for seismic design, but offers no guidance or specific requirements.

2. Paragraph 402.1, General (relating to design criteria). “… the design engineer shall provide reasonable protection to prevent damage to the pipeline from unusual external conditions which may be encountered in river crossings, offshore and inland coastal water areas, bridges, areas of heavy traffic, long self-supported spans, unstable ground, vibration, weight of special attachments, or forces resulting from abnormal thermal conditions.” This provision goes on to state that protective measures, such as increasing the wall thickness, may be used. This provision simply acknowledges the possibility of extreme events for which some special protection is warranted. No specific requirements are set forth for any of these load conditions, nor is there any delineation of what constitutes “reasonable protection.”

3. Paragraphs 402.3.3, Limits of Calculated Stresses Due to Occasional Loads, and 419.6.4, Stress Values. “The sum of the longitudinal stresses produced by pressure, live and dead loads, and those produced by occasional loads, such as wind or earthquake, shall not exceed 80% of the specified minimum yield strength of the pipe…” This limit is intended for aboveground piping or elevated pipeline spans and is not applicable to extreme situations involving large ground movements due to faulting or liquefaction.

B.2 ANSI B31.8, Gas Transmission and Distribution Piping Systems
ANSI B31.8 is the U.S. standard for the design, construction, and operation of onshore and offshore gas transmission and distribution systems. In a manner similar to the ANSI standard for liquids lines (B31.4), the ANSI B31.8 standard for gas transmission and distribution lines makes very limited reference to seismic design of onshore pipelines, but has more extensive treatment of seismic hazards (without specifics) for offshore pipelines (B31.8, Chapter 8).

The B31.8 provisions that mention seismic design are as follows:

1. Paragraph 833.4(c), relating to the calculation of longitudinal stress. External loads such as weight and wind that contribute to longitudinal bending stress for aboveground piping are mentioned. This section applies to primary seismic inertial loads but not to secondary loads on buried piping.

2. Paragraph 841.13, Protection of Pipelines and Mains from Hazards, Item (a). “When pipelines and mains must be installed where they will be subject to natural hazards, such as washouts, floods, unstable soil, landslides, earthquake related events (such as surface faulting, soil liquefaction, soil and slope instability characteristics), or other conditions which may cause serious movement of, or abnormal loads on, the pipeline, reasonable precautions shall be taken to protect the pipeline, such as increasing the wall thickness, constructing revetments, preventing erosion, and installing anchors.” This paragraph effectively identifies the potential for seismic hazards but does not PR-268-9823
2284
Appendix B: Seismic Design Provisions in Existing Standards B-3 prescribe any specific criteria or requirements for design or operational precautions or what constitutes “reasonable precautions.”

3. Paragraph A835, Anchorage for Buried Piping. “When a submerged pipeline is to be laid across a known fault zone, or in an earthquake-prone area where new faults are a possibility, consideration shall be given to the need for flexibility in the pipeline system and its components to minimize the possibility of damage due to seismic activity.”

4. Paragraph A841.1, Design Conditions, Item (f). Seismic activity is identified as a prospective offshore design condition.

5. Paragraph A841.33, Design Environmental Conditions, Item (d). Seismic events are identified as a prospective offshore environmental load condition.

6. Paragraph A842.23, Alternate Design for Strain. This paragraph permits design to strain criteria in excess of yield strain for situations where the pipeline experiences a predictable, noncyclic displacement of its support and provided the consequences of yielding are not detrimental to the integrity of the pipeline. The paragraph goes on to say that permissible strain criteria should be determined on the basis of available ductility and buckling behavior. This provision of B31.8 is quite important in that it offers confirmation of the appropriateness of using strain limits as the design criteria for pipelines subjected to seismic ground movements, albeit such limits are undefined.

7. Paragraph A843, On-Bottom Stability. “Pipeline design for lateral and vertical stability is governed by seafloor bathymetry, soil characteristics, and by hydrodynamic, seismic, and soil behavior events having a significant probability of occurrence during the life of the system. … The pipeline system shall be designed to prevent horizontal and vertical movements, or shall be designed so that any movements will be limited to values not causing design strength to be exceeded. …Typical factors to be considered in the stability design include: …(c) liquefaction, (d) slope failure.” This paragraph states that offshore pipelines should be designed to withstand seismic effects without exceeding design strength, but there is no link back to the provision in Paragraph A842.23 for design on the basis of strain limits.

8. Paragraph A843.5, Soil Liquefaction. “Design for the effects of liquefaction shall be performed for areas of known or expected occurrence. …. Seismic design conditions used to predict the occurrence of bottom liquefaction or slope failure shall have the same recurrence interval as used for the operating design strength calculations for the pipeline.” This paragraph calls for the design of offshore pipelines to mitigate the effects of liquefaction and slope failure, but no specific criteria or methodology are defined.

PR-268-9823
2285
Appendix B: Seismic Design Provisions in Existing Standards B-4 B.3 API Recommended Practice 1111, Design, Construction, Operation, and Maintenance of Offshore Hydrocarbon
Pipelines
API Recommended Practice (RP) 1111 sets out criteria for the design of offshore pipelines. It is intended to be used in conjunction with ANSI B31.4 and B31.8 standards for liquids and gas transmission pipelines. The procedures presented in API RP 1111 are essentially the same as that prescribed by the ANSI standards, and the allowable stress provisions from B31.4 and B31.8 are incorporated by reference. The mention of seismic requirements are quite limited as summarized below:

1. Section 2.1.4, Dynamic Forces. A general statement is made that the design should consider dynamic forces imposed on the pipeline and the resulting stresses, and that these stresses may be caused by a number of possible loading conditions, one of which is seismic activity. The provision serves only as an alert to consider seismic events as a source of dynamic loads.

2. Section 2.4.2.1, On-Bottom Stability. The discussion of considerations for on-bottom stability includes the effects of natural phenomena. The potential for earthquakes to cause the liquefaction of sea-bottom sediments and associated settlement or floatation of a pipeline is specifically mentioned as a potential situation requiring consideration.

B.4 BS 8010, Code of Practice for Pipelines, Part 3, Pipelines Subsea: Design, Construction and Installation
British Standard BS 8010, Part 3, provides recommendations for the design, construction, installation, testing, and commissioning of subsea pipelines. BS 8010 contains numerous references to seismic activity as an environmental load condition that must be considered.

The provisions of BS 8010 that relate to seismic design are listed below:

1. Section 4.2.2.5, Functional and Environmental Loads. Earthquakes are identified as an environmental load condition requiring consideration, and cross-reference is made to Appendix B of BS 8010 for further discussion.

2. Section 4.2.5.4, Equivalent Stress. This section specifies that equivalent stress (computed in accordance with the von Mises’ stress criterion) shall be less than 0.96 times the specified minimum yield stress for environmental load conditions.

3. Section 4.2.6, Alternative Design for Strain. This section states that the limit on equivalent stress set forth in Section 4.2.5.4 may be replaced by an allowable strain limit of 0.1%.

4. Section 4.5.4.2, Soil Instability, Item (a). This section states that consideration should be given to stabilization enhancement methods in locations where the ground along the pipeline route might become unstable due to a number of reasons, including seismic activity.

PR-268-9823
2286
Appendix B: Seismic Design Provisions in Existing Standards B-5

5. Section 4.12.1, Possible Causes of Damage, Item (I). Seismic activity is named as a possible cause of damage that requires consideration.

6. Section B.1.10, Seismic Action. This section calls for consideration of the effects of seismic ground motions along the pipeline route. Specific mention is given to the possibility of soil liquefaction. No guidance is offered on the appropriate level of treatment.

7. Section B.2.7.5, Liquefaction, Item (b). This section calls for consideration of soil liquefaction, and seismic action is identified as a potential cause of liquefaction.

8. Section C.1.3, Axial Compression, and Section C.1.4, Bending. These two subsections under Appendix C, Buckling, provide stress limits for local buckling.

B.5 CSA Z662-96, Oil and Gas Pipeline Systems
The Canadian standard for oil and gas pipeline systems provides stress design requirements for operating pressure, thermal expansion ranges, temperature differential, and sustained force and wind loadings. Other load conditions are not specifically addressed. The Canadian standard provides for the design of pipelines using limit state criteria defined in its non-mandatory Appendix C, although the prescribed load conditions for which the limit state criteria are to be applied are the same as in the mandatory sections of the standard.

B.6 DNV Rules for Submarine Pipeline Systems
The DNV Rules for submarine pipeline systems specify requirements for the design, construction, installation, operation, and maintenance of subsea pipelines. The DNV Rules contain several references to seismic activity as a load condition that must be considered. These provisions are summarized below.

1. Section 2.1.1, Environmental Phenomena. This section requires consideration of all environmental phenomena that impair pipeline system function or reduce system reliability. Seismic activity is listed as an environmental phenomenon that should be considered.

2. Section 3.3.1.3, relating to environmental loads. This section states that the return period to be taken for environmental loads shall be greater than 100 years.

3. Section B.1, Local Buckling. This section of Appendix B provides methodology for determining critical buckling stress based upon a Ramberg-Osgood representation of the stress-strain curve for steel.

PR-268-9823
2287
This Page Intentionally Left Blank
2288
Appendix C: Probabilistic Seismic Hazard C-1

PROBABILISTIC SEISMIC HAZARD
Nearly all modern earthquake design codes rely on probabilistic seismic hazard analyses to define the level of ground shaking. The basic procedures for estimating the ground shaking hazard in a probabilistic hazard analysis are similar to those used in a deterministic approach where a specific earthquake magnitude and location are specified.

The main difference in a probabilistic approach is that multiple earthquakes are considered to contribute to the level of ground shaking likely to be experienced at a particular location.

The general process for performing a probabilistic seismic hazard analysis includes the following steps, illustrated graphically in Figure C.1:

1. Characterize the type and location of earthquake source zones.

2. Define the frequency of occurrence of earthquakes of varying magnitude within the source zones.

3. For a particular location, use attenuation relationships to compute the level of ground shaking from each earthquake from each source zone and weight the resulting ground shaking by the likelihood of earthquake occurrence.

Earthquake source zones are typically characterized as point, line, or area sources (Figure C.2). A point source represents an earthquake at a single location and is typically incorporated into line-source or area-source models. A line source, often used to represent earthquakes occurring along faults, can have an earthquake point source centered at any portion of the line source considered to rupture in an earthquake. An area source is often used to model regions with no known active faults and point sources are assumed to occur at random locations within the defined area.

The rate of occurrence of earthquakes is estimated from regional earthquake records, regional strain rate information, or specific fault investigations that provide estimates of average fault slip rates. A particular concern in establishing earthquake occurrence rates is to maintain consistency between regional deformation associated with the model and deformation related to regional strain measurements.

Two common methods used to estimate recurrence rate for earthquakes of varying magnitude include the use of the Gutenberg-Richter relationship and a characteristic model of earthquake recurrence. The Guttenberg-Richter relationship assumes a continuous distribution of earthquake magnitudes between zero and the maximum magnitude for the particular fault. The form of this relationship is as follows:
log ( ) N M a bM =− (C-1)
where:

N(M) = earthquake moment magnitude [6.0 < M < 8.0]
PR-268-9823
2289
Appendix C: Probabilistic Seismic Hazard C-2
a, b = constants determined from regional earthquake catalogs The a parameter in (C-1) corresponds to a magnitude between –0.05 and +0.05. In probabilistic ground motion hazard analyses, it is customary to truncate the relationship at an earthquake magnitude between 5.00 and 6.50. This truncation is necessary to avoid overestimating regional deformation produced by a large number of very small earthquakes. For areas with known active faults, the maximum earthquake magnitude given by the Gutenberg-Richter equation is generally limited to the magnitude computed based upon empirical relationships (e.g., Wells and Coppersmith, 1994) between magnitude and fault length or fault area assuming the entire fault ruptures.

The concept of a characteristic earthquake is generally accepted as an appropriate means to address hazards from individual faults. The characteristic earthquake approach is based upon the hypothesis that earthquakes generated by a particular fault tend to have similar magnitudes instead of the continuous magnitude distribution assumed by the Gutenberg-Richter model. The characteristic earthquake is typically related to an earthquake that ruptures an entire fault segment.

For area sources, earthquake occurrence rates are nearly always based upon a GutenbergRichter recurrence model with the maximum earthquake magnitude estimated using available information and considerable judgment.

In addition to weighting the ground motions at a site produced by each earthquake source, the uncertainty in selection of attenuation relationship can be incorporated into the probabilistic hazard. This may be accomplished by selecting multiple attenuation relationships, with a weighting factor based upon judgment or by accounting for the variability of a particular attenuation relationship directly.

Probabilistic hazard studies produced for areas where there are no well-defined active faults (i.e., modeled entirely with area sources) have a common characteristic: ground motion estimates increase with decreasing probability of occurrence and these motions are associated with smaller magnitude earthquakes occurring closer to the site of interest.

For estimating wave propagation strains or assessing landslide hazard, the ground shaking estimates from a probabilistic hazard analysis can be used directly. For the assessment of liquefaction or lateral spread displacement, it is necessary to define an earthquake magnitude and distance. While this information is generally available as a by-product of the probabilistic hazard analysis, it is often not reported. Deaggregation studies extract information from probabilistic hazard analyses on the contribution of individual point sources within line and area source models. There is little guidance on how to select an appropriate combination of magnitude and distance from a deaggregation study. Past project experience indicates that selecting a combination that captures 50% to 84% of the total seismic hazard may be appropriate. This is illustrated in the Commentary.

The results of probabilistic hazard studies are of no value for estimating fault displacement. However, if probabilistic hazard studies are available for areas with PR-268-9823
2290
Appendix C: Probabilistic Seismic Hazard C-3
known fault activity, there will be information on specific faults used in the hazard study (e.g., fault length, fault width, slip rate, dip angle, type of faulting).

It is recommended that assistance from individuals with special expertise in defining earthquake surface fault hazards be used to fault motions for assessing pipeline performance. However, a very approximate estimate of the likelihood of an earthquake of the maximum magnitude on a fault can be made based upon the reported fault slip rate and fault dimensions. The key assumption in this approach is that all geologic slip is associated with the estimate of the maximum earthquake that can occur on a fault. This assumption overestimates the occurrence rate since it does not account for slippage from smaller, more frequent earthquakes. Nonetheless, the approach can be used to compare relative risk from numerous faults and is often adequate for performing preliminary engineering assessments.

The slip rate can be used to estimate the rate of accumulation of seismic moment using the definition of seismic moment rate, Mr:

M RLD FW r =⋅ ⋅ µ ⋅sr (C-2)
where:
a, b = constants determined from regional earthquake catalogs Mr = moment rate
µ = shear modulus of earth’s crust (typically taken as 3x1010 N/m2 ) RLD = fault rupture length at depth
FW = fault width
sr = fault slip rate
The moment released, Me by an earthquake of magnitude M is computed as Me
M = + 1015 9 .( ) (C-3)
The earthquake magnitude, M, can be estimated using empirical relationships from Wells and Coppersmith (1994), one of which is provided below:

M = +⋅ 4 38 149 . . log(RLD) (σ = 0.26 units of magnitude) (C-4) The mean recurrence period, RP, for an earthquake of magnitude M can be estimated as Me/Mr which is given by the expression below.

RP
RLD FW sr
M
= ⋅ ⋅⋅
+ 1015 9 .( )
µ
(C-5)
By assuming that an earthquake of a selected magnitude is responsible for all of the fault slip, the simplified approach always overestimates the probability of occurrence of a PR-268-9823
2291
Appendix C: Probabilistic Seismic Hazard C-4
particular earthquake. Comparisons with published recurrence intervals for many faults in the United States indicate that the simplified approach typically overestimates earthquake probability by less than a factor of two. Since the simple approach always overestimates the likelihood of an earthquake of a specific magnitude, it will generally significantly underestimate the earthquake magnitude for a defined return period. This is the reason the simplified approach is limited to consideration of the maximum earthquake magnitude for a particular fault.

If the maximum magnitude earthquake on a particular fault is not judged to be appropriate or a more refined estimate of return period is needed, assistance from individuals with expertise in seismology is required.

PR-268-9823
2292
Appendix C: Probabilistic Seismic Hazard C-5
Figure C.1 Probabilistic Seismic Hazard Process (from ASCE, 1984) PR-268-9823
2293
Appendix C: Probabilistic Seismic Hazard C-6
Figure C.2 Common Types of Earthquake Source Models (from ASCE, 1984) PR-268-9823
2294
Appendix D: Example Calculations D-1

EXAMPLE CALCULATIONS
This example calculation is provided to serve several purposes. First, the example illustrates the application of various equations provided in the guidelines. Second, the example provides a basis for understanding the magnitude of numerical values (e.g., ground displacements, soil spring forces) associated with estimating parameters for a seismic assessment of a buried pipeline. Finally, the example illustrates the processes involved in performing a seismic assessment of a buried pipeline. To assist in presenting the calculations, the example has been framed in the context of a hypothetical pipeline alignment. While parts of the example are extracted from actual past project experience, the results are for illustrative purposes only.

Problem Description:

The site in question is the portion of an NPS 20 pipeline alignment at the south bank of the Flowing River in the western United States. The NPS 20 pipeline crosses the river in a straight alignment. The site is being evaluated for potential liquefaction hazards. The plan and profile of the south side of the river crossing is shown in Figure D.1. The ground is typically flat south of the crossing (less than 0.5% slope) until the pipeline reaches a hill approximately 650 ft (200 m) south of the riverbank. Surface geology maps indicate the hill is an outcropping of weathered shale. The pipeline in the vicinity of the river crossing is classified as a Class 2 location in accordance with ASME B31.8.

The performance requirement established for the NPS 20 pipeline is defined as a mean annual probability of pipeline rupture from earthquake causes less than 0.2%. This is equivalent to a 500-year average return period. A review of available geologic maps and regional hazard information available from the government indicate two potential earthquake sources. One source is the Shakem fault, located 19.3 miles (31 km) from the river crossing. The other source is the Enditall fault, located more than 30 miles (48 km) from the site. Both faults are characterized by strike-slip fault movement with surface fault expression. Seismic parameters available for the two faults in question are summarized in Table 1. The variability in fault length and slip rate is for illustration purposes only. In actual practice, there is considerable variability in all fault descriptive parameters that are usually related to the particular interpretation of subsurface data by different geologists and seismologists.

Table 1: Summary of Fault Descriptive Parameters
Fault Length, km
(range)
Depth
km
Slip
mm/yr
Sediment
Depth
km
Shakem 80 (70 - 120) 11 5±2 1
Enditall 175 (100 - 250) 13 1±0.5 1
PR-268-9823
2295
Appendix D: Example Calculations D-2
The pipeline, installed in 1960, is fabricated from API 5L Grade X42 seamless pipe with a coal tar coating. The pipe wall thickness is typically 0.280 inches (7.1 mm) except at the river crossing where the thickness is 0.375 inches (9.5 mm). The river crossing portion of the pipeline also has 3 inches (76 mm) of concrete coating for buoyancy control. The maximum allowable operating pressure for the pipeline is 825 psi (5,700 kPa) although the typical operating pressure is below 500 psi. No internal inspection of the pipeline has been performed to confirm the condition.

Liquefaction is considered possible based upon regional seismic hazard maps and knowledge of past subsurface investigations for other construction projects near the pipeline crossing. Given this information, subsurface investigations to collect SPT (one location) and CPT data (three locations) were performed at the sites indicated in Figure D.1. SPT and CPT data were collected at a common location to provide a direct correlation between the SPT and CPT data. In addition to the SPT data, SPT drilling provided an opportunity to collect soil samples. Samples were collected from each SPT using sample liners and later evaluated to determine density and grain size distribution.

Some results from the subsurface investigation are provided in Table 2 and Figures D.2 through D.4.

Table 2: Selected Data from Subsurface Investigations
SPT 1 CPT 3
Sample Depth 20 ft Test Depth 20 ft
SPT Value 8 qc (bars) 50
% Fines 19 fs (bars) 0.2
Average Density (pcf) 110
Step 1: Define Earthquake for Performing the Seismic Assessment In most cases, the earthquake hazard will be defined by geologists or seismologists familiar with the tectonic setting of the site being assessed. The following approach, based upon methods described in Section 2 and Appendix C is provided to illustrate the process for obtaining an approximate quantification of the seismic hazard with limited data.

The information on fault length, depth, slip rate, and direction of slip are used to estimate the risk of an earthquake and the ground shaking levels in the event of an earthquake.

The calculations necessary to perform this estimate for the Shakem fault are summarized below.

Earthquake Magnitude:

Use the Wells and Coppersmith (1994) relationship for unknown or unspecified type of faulting
MR =+ =+ = 4 38 149 . . log . . log b LDg 4 38 149 80 7 2 b g .

PR-268-9823
2296
Appendix D: Example Calculations D-3
Mean Recurrence Interval:
µ = 3(10)10 N/m2
Me = N-m 1.5( ) 9 1.5(7.2) 9 19 10 10 6.31(10) M ++ ==
Mr = ( ) ( )( )( ) ( ) N-m/yr 10 17
µ ⋅ ⋅ ⋅= RLD FW sr 3 10 80,000 11,000 .005 1.32 10 =
Mean recurrence interval estimated to be no more frequent than M M
e
r
= 478 years or
approximately 500 years.

Amount of Fault Movement:

Although the pipeline does not cross a fault, the design fault dispalcements are computed for the purpose of the example problem. The pipeline is in a Class 2 location so the design fault displacement will be based upon the mean average fault displacement. For the Shakem fault:
log( ) 0.88 1.43 0.88log(80) 1.43 0.24 AD SRL = −= −=
AD = 1.8 m
The potential variability in the above estimates of fault movement and return period can be estimated by examining the results for possible combinations of fault rupture length and slip rate. The results are summarized in the following tables.

Table 3a: Earthquake Magnitude from Fault Rupture Length Fault Rupture Length, km
70 80 120

7.1 7.2 7.5
Table 3b: Mean Fault Displacement from Fault Rupture Length Fault Rupture Length, km
70 80 120

1.6 m 1.8 m 2.5 m
PR-268-9823
2297
Appendix D: Example Calculations D-4
Table 3c: Earthquake Recurrence Interval (years)
Slip Rate Fault Rupture Length, km
mm/yr 70 80 120
3 710 840 1,390
5 430 500 830
7 306 360 600
Similar values for the Enditall fault are provided below:

Magnitude: 7.4 - 8.0
Fault Movement: 2.1 - 4.8 m
Recurrence Interval: 1,120 - 14,430 years
For the best estimate fault parameters, the Enditall fault average return period is estimated to be 5,610 years compared to the approximately 500 years for the Shakem fault. If fault crossing evaluations were to be performed for the Enditall fault, the design fault displacements would be obtained by multiplying the mean average fault displacement by 0.18, the performance goal (500 years) divided by half of the average return period of the fault (2,805 years). This would result in design fault displacements of 0.38 m to 0.86 m. Since both faults are strike-slip, there is no need for a correction to derive fault displacement components in three orthogonal displacements.

It is important to note the considerable variability in the estimates of the earthquake hazard. The level of uncertainty indicated by the results in Table 3 is typical of actual experience in estimating the earthquake hazard. It is possible, but not certain, that additional geologic investigations including trenching across the fault, deep borings, or seismic refraction studies could better define the basic fault descriptors.

PGA at Site:

Use Campbell attenuation relationship for a soil site:
0 0 031
0 1 12 90
SR HR
T SB
FSSR
HHH α
= = ==
= = == D

1.01 0.32 1.01 0.32(7.2) 10 10 16 M W −+ −+ == =
[] [] 11 sin( ) 12 0 16 3.85 22
d H HW H S BT = + − ≥ = + − =− α S
PR-268-9823
2298
Appendix D: Example Calculations D-5
dH SS ==1
22 2 2 1 33 31 R dR sS = += + = km
[] [
[]
]
2 2 0.647
1 2 2 0.647(7.2)
ln( ) 3.512 0.904 1.328ln 0.149

1.125 0.112ln( ) 0.0957 0.440 0.171ln( )

0.405 0.222ln( ) ( ) ( )

3.512 0.904(7.2) 1.328ln 31 0.149
S
M
SS
S HR A
PGA M R e

R MF R SSR
R S fD E x
e
−
=− + − + ⎡⎤ ⎣⎦
+− − + −
+ − + +Φ
=− + − + ⎡⎤ ⎣⎦
PGA = 0.18 g
The range of PGA, adopting the variability in earthquake magnitude from Table 3 and not including variability in the Campbell attenuation relationship, is 0.17 - 0.22 g. Similar PGA values for the Enditall fault are 0.13 g to 0.20 g.

Although the levels of ground shaking are comparable for the Shakem and Enditall faults, the Shakem fault is determined to be the controlling earthquake hazard because the recurrence interval is more compatible with the stated performance goal.

For the remainder of this example, the central fault description values (length = 80 km, slip rate = 5 mm/yr) are used.

Step 2: Assess Liquefaction Potential for Subsurface Soils Calculations to assess the potential for liquefaction are performed for the same depth at location 1 where SPT and CPT data were collected. For the SPT calculations, hammer energy was assumed to be 60% and the tests were performed with a standard diameter sampler. A gradation curve for a sample retrieved at the depth being assessed for location 1 is provided in Figure D.4.

Depth = 20 ft = 6.1 m
Depth to water table = 6 ft (1.8 m) at time of test
Dry density = 110 pcf
Total overburden stress = σ vo = 20 110 2 200 b g = , psf = 1.1 tsf = 105 kPa Effective stress = σ ′
vo vo =− σ 62 4 20 6. b − g = 1,326 psf =0.663 tsf = 63.5 kPa Pa == = 14 5 100 2 088 ., psi kPa psf
z = D/m = 6.1
PR-268-9823
2299
Appendix D: Example Calculations D-6
r
zz z
zz z
d = − ++
− +− +
1 0 4113 0 04052 0 001753
1 0 4177 0 05729 0 006205 0 001210
15 15 2
. ..
. .. .
.
. z
rd = −+ +
−+ − +
1 0 4113 61 0 04052 61 0 001753 61
1 0 4177 61 0 05729 61 0 006205 61 0 001210 61
15 15 2
. .. . . .
. .. . . . . .
.
.
bg bg
bg bg bg
= 0.96
() 2,200 0.65 0.65 0.18 0.96 0.19
1,326
vo
d
vo
CSR PGA r σ
σ
⎛⎞ ⎛⎞ == ⎜⎟ ⎜⎟ ′ ⎝⎠ ⎝⎠ =
SPT Assessment:

N = 8
C P
N
a
vo
= ′ == σ
2 088
2 200
125 ,
, .

C
z
R =+ − =+ − 0 75 0 25 = 4
6 0 75 0 25 61 4
6 .. .. 084 (. ) . b g
FC = 19%
α
β
= ==
=+ =+ =
−−
ee
FC
FC
1 76 190 1 76 190
19 15 15
22 323
0 99
1000
0 99 19
1000
107
..
..
.
.. .
.
.
bN NC C C C C 1g60 = ⋅ ⋅ ⋅ ⋅ ⋅ = ⋅ ⋅⋅⋅ ⋅= N EBRS 8 126 1 1 084 1 8 4 ..

NN 1 60FC 1 60 b g =+ = + = αβb g 323 107 8 4 12 2 . .. b g MSF
M === 10 1072
105
3 81
4 42
3 81
4 42
.
.
.
. . .

Kσ is 1.0 since the effective stress is less than the reference pressure, Pa 23 4 2 5 3
13 2 4 3 6 4
23 4 2 5 3
1 4.8(10) 4.721(10) 6.136(10) 1.673(10)
1 1.248(10) 9.578(10) 3.285(10) 3.714(10)

4.8(10) 4.721(10) 12.2 6.136(10) 12.2 1.673(10) 12.2
1 1.248(10) 12.2
xx x CRR MSF K
xx x x σ
−− − −
−− − −
−− − −
−
⎛⎞ −+ − =⋅ ⎜⎟ ⎝⎠ −+ − +
−+ − = −+ 32 4 3 6 4 1.05 1.0

9.578(10) 12.2 3.285(10) 12.2 3.714(10) 12.2

0.14 −−−
⎛⎞ ⋅ ⎜⎟ ⎝⎠ −+
=
PR-268-9823
2300
Appendix D: Example Calculations D-7
Since the CRR of 0.14 is less than the CSR of 0.19, liquefaction is expected for 0.18 g PGA.

CPT Assessment:
qc = 50 bar = 5,000 kPa
fs = 0.2 bar = 20 kPa
F = 20 100% 100% 0.41%
5000 100
s
ca
f
qP
⎡⎤ ⎡⎤ ⎢⎥ == ⎢⎥ ⎣⎦ −− ⎣⎦
Q1 = 5000 105 100 77.1
100 63.5
c vo a
a vo
qP
P
σ
σ
−− ⎛⎞ ⎛⎞ ⎜⎟ == ⎜⎟ ⎝⎠ ′ ⎝⎠
I1 = ( )( ) ( )( ) 22 2 2 3.47 log 1.22 log 3.47 log 77.1 1.22 log 0.41 1.79 − ++ = − ++ = QF 1 Q0.5 =

0.5 0.5 100 kPa 5000 105 100 61.4
100 kPa 100 63.5
c vo
vo
q σ
σ
−− ⎛⎞ ⎛⎞ ⎜⎟ == ⎜⎟ ⎝⎠ ′ ⎝⎠
I0.5 = ( )( ) ( )( ) 22 2 2 3.47 log 1.22 log 3.47 log 61.4 1.22 log 0.41 1.88 − ++ = − ++ = QF 0.5 n = 0.5

0.5 100 1.26

63.5 n
a
Q
vo
P C
σ
⎛⎞ ⎛⎞ == = ⎜⎟ ⎜⎟ ⎝⎠ ′ ⎝⎠
K IIII cn =− + − + − n n n
=− + − + −
=
17 88 33 75 2163 5581 0 403
17 88 33 75 188 2163 188 5581 188 0 403 188
116
23 4
23 .. . . .
. . (. ) . (. ) . (. ) . (. )
.
4 () 1
5,000 1.26 1.16 73
100
c
cN c Q FC
a
q q KC P
⎛⎞ ⎛⎞ =⋅ = ⋅ = ⎜⎟ ⎜⎟ ⎝⎠ ⎝⎠
CRR MSF qc N FC == = F
H
G I
K
J +
L
N
M
M
O
Q
P
P
F
H
G I
K
J +
L
N
M
M
O
Q
P
P
93 1000
0 08 93 73
1000
0 08 1
33 105 012 () .. ..

Since the CRR of 0.12 is less than the CSR of 0.19, liquefaction is expected for 0.18 g PGA.

PR-268-9823
2301
Appendix D: Example Calculations D-8
From similar calculations for the other boring locations, the extent of liquefiable material for the site is estimated. The cross-section of the site is provided in Figure D.1 that indicates the extent of potentially liquefiable material at the riverbank.

Step 3: Assess Potential for Lateral Spread Movement
The site is representative of a free face condition. For this example, thickness of saturated sand deposits with (N1)60 less than 15 is taken as 75% of the average liquefiable thickness of 15.3 m or 11.5 m, the average fines content of the material with (N1)60 less than 15 is 5%, and D5015 is 0.10 mm. In practice, this determination is made based upon a review of all SPT and CPT data.

Lateral spread displacement for the onshore portion of the crossing can be estimated as a function of the distance from the crest of the bank using a free-face condition.

M = 7.2
R = 31 km
Ro = 0.89 5.64 0.89(7.2) 5.64 10 10 5.86 M −− ==
R* = R +Ro = 36.86
T15 = 11.5
F15 = 5
D5015 = 0.10
H = 20
For sloping ground conditions:
*
15 15 15 0.1)
( ) 16.213 1.532 1.406

0.012 0.338 0.540

3.413 (100 ) 0.795 ( 50
Log LSD M LogR
R LogS LogT
Log F Log D +
=− + −
−+ +
+ −−
For free-face conditions:
*
15 15 15 0.1)
( ) 16.713 1.532 1.406

0.012 0.592 0.540

3.413 (100 ) 0.795 ( 50
Log LSD M LogR
R LogW LogT
Log F Log D +
=− + −
−+ +
+ −−
Lateral spread displacement is computed for various distances from the toe of the riverbank as summarized in Table 4. The maximum distance of interest coincides with a transition between river deposits and the weathered shale hillsides bounding the river.

Lateral spread displacements on the slope of the bank are estimated using the ground PR-268-9823
2302
Appendix D: Example Calculations D-9
slope equation even though the slope exceeds the 0% to 6% recommended range in slope for applicability of the equation. This is likely a conservative assumption considering the much larger displacements estimated by the ground slope equation compared to the freeface equation. The surface ground displacement profile assumes free-face conditions beyond the crest of the riverbank since the ground profile in this area is relatively flat.

This assumption is consistent with the formulation of data used in the Youd, Hansen, and Bartlett equations. The resulting surface ground displacement profile is illustrated in Figure D.5.

Table 4: Computed Lateral Spread Displacements
Station
(m)
Distance from Toe
(m)
W & S
(%)
Slope LSD
(m)
Free-Face LSD
(m)
2355 0 N/A N/A N/A
2596 241 8.3 2.7 N/A
2605 250 7.8 2.6 1.4
2655 300 6.7 2.5 1.3
2705 350 5.7 2.4 1.2
2739 384 5.2 2.3 1.1
Step 4: Estimate Relative Lateral Spread Displacement Affecting the Pipeline Typically, the ground displacement resulting in loads on the pipeline is less than that indicated in the plot of surface displacements for portions of the pipeline within the liquefiable material. The ground displacements at the pipeline depth are estimated as a function of the ratio of the liquefiable layer depth below the pipe to the total depth of the liquefiable layer. Alternatively, a more conservative approach is to assume that the pipe is subjected to the same horizontal ground displacements as the surface. This assumption is consistent with block-like movement of the liquefiable layer over the deeper nonliquefiable soil.

Step 5: Compute Vertical Settlement from Pore Water Dissipation The amount of vertical settlement related to pore water dissipation is estimated using the average computed factor of safety against liquefaction and Figure 2.4. For the location previously assessed, the factor of safety against liquefaction is CRR/CSR = 0.13/0.19 =

0.7. Figure 2.4 was developed for clean sands. The use of (N1)60FC = 12.2 instead of (N1)60 is assumed to account for the presence of fine-grained material. From Figure 2.4, the post-liquefaction volumetric strain at the location where liquefaction is estimated as approximately 3%. The vertical settlement that the pipeline will experience is estimated as 3% of the depth of liquefiable soil beneath the pipeline and is plotted in Figure D.5.

The maximum differential vertical displacements that the pipeline is exposed to is less than 15 cm over 35 m of pipe. This is sufficiently small to be neglected in the assessment of pipeline performance.

PR-268-9823
2303
Appendix D: Example Calculations D-10
Step 6: Compute the Soil Spring Properties Along the Pipeline Soil spring properties vary with soil material and pipeline burial depth. Within the same soil material, soil spring values are assumed to vary linearly between two different burial depths. The following example computes soil springs near stations 2453 and 2749.

Station 2453:

This location is within the zone of soil expected to liquefy. The soil strength is treated as if the soil were a cohesive material (φ = 0) with an undrained shear strength equal to 20% of the effective overburden stress.

Soil cover = 8.4 m = 27.6 ft
Water table elevation = -2.75 m = -9 ft
Ground surface elevation = -13.8 m = -45.3 ft
Diameter = D+2(3inches) = 26 inches = 2.167 feet = 0.660 m Material = liquefied soil
H = C + D/2 = 8.73 m = 28.6 ft
H/D = 13.2
Effective stress = 28.6(110-62.4) = 1,361 psf = 65 kPa
c = 0.2(1,361) = 272 psf = 0.272 ksf = 13 kPa
The primary direction of soil movement is in the axial direction. Therefore, assume all lateral (horizontal and vertical) movement is confined to the soils immediately surrounding the pipeline.

Axial Soil Spring:
α =− − ++
+
=− − ++
+ =
0 608 0123 0 274
1 0 695
1 0 608 0123 0 272 0 274
0 272 1
0 695
0 272 1
10 23
23 .. ..
. . (. ) .
(. )
.
(. ) .
c
cc
T D c DH K u
o =+ F +
H
G I
K πα π γ δ π J == 1
2 tan( ) ( . )( . )( ) , 2 167 10 272 1 851 27 lb
ft
kN
m =
Horizontal Soil Spring:

PR-268-9823
2304
Appendix D: Example Calculations D-11
23 23

11.063 7.119 6.752 0.065
11 11.063 7.119 6.752 0.065(13.2) (13.2 1) (13.2 1)

7.6 ch
H N
D HH
DD
⎛⎞ =+ − + ⎜⎟ ⎝⎠ ⎛ ⎞⎛ ⎞ ⎜ ⎟⎜ ⎟ ++ ⎝ ⎠⎝ ⎠
=+ − +
++
=
P N cD N HD u ch qh =+ = = = γ 7 6 272 2167 4 480 654 . ( )( . ) , . lb ft
kN
m
∆p H D=+ F
H
G I
K
J = ⋅+ F
H
G I
K 0 04 J = 2
0 04 12 28 6 22 167
2 . .. 14 . inches = 36 cm
Vertical Bearing Soil Spring:
tan( 0.001) 2
tan(0.001) 2
tan( 0.001) 2 tan(0.001) 2

0.18 2.5

0.001 cot( 0.001) tan 45 1
2 0.001 cot(0.001) tan 45 1 5.14 2

0.001 0.001 tan 45 tan 45 1
22 c
q
Ne
e
Ne e
Ne e
πφ
π
πφ π
ϕ
γ
φ φ
φ
+
+
−
⎡⎤ ⎛⎞ +
=+ ⎜⎟ + − ⎢⎥ ⎣⎦ ⎝⎠
⎡⎤ ⎛⎞ =+ ⎜⎟ − = ⎢⎥ ⎣⎦ ⎝⎠
⎛⎞ ⎛ +
= += + ⎜⎟ ⎜ ⎝⎠ ⎝
== 2.5 0 − ≈
⎞ =⎟
⎠
2 2
lb kN 5.14(272)(2.167) (110 62.4)(2.167)(28.7) 5,990 87.4 ft m dc q
D Q N cD N HD Nγ =+ + γγ
=+ − ==
∆qd == = 0125 0125 26 3 3 . . ( ) . inches D = 84 cm
Vertical Uplift Soil Spring:

N H
D cv = F
H
G I
K
J ≤
==
28 2 9 5 19 8 ( .) >
Ncv = 8
Q N S D N HD u cv u qv =+ = = = γ 8 272 2167 4 715 688 ( )( . ) , . lb ft
kN
m
PR-268-9823
2305
Appendix D: Example Calculations D-12

0.12 ( ) 0.1 ( ) in 0.1 0.1(28.7ft) 12 34 inches ft
qv cv
qu
cv qv
H N HD H N cD
H
N cD N HD
γ
γ
+ ⎛⎞ ∆= = = ⎜⎟ = + ⎝⎠ = 86 cm
Station 2749:

Soil Cover = 4.9 ft (1.50 m)
Diameter = D = 20 inches = 1.667 ft = 0.508 m
Material = granular (c = 0)
H = C + D/2 = 4.9(12)+.5(20) = 68.8 inches = 5.733 ft = 1.75 m H/D = 68.8/20 = 3.4
The internal friction angle is taken to be 35°.

The pipe is coated with fusion bonded epoxy. Use δ = 0.6φ = 0.6(35) = 21° The primary direction of soil movement is in the axial direction. Therefore, assume all lateral (horizontal and vertical) movement is confined to the soils immediately surrounding the pipeline.

Axial Soil Spring:
1 tan( ) 2
1 0.5 lb kN (1.667)(5.733)(110) tan(21 ) 950 13.9 2ft
o
u
K T D c DH πα π γ δ
π
⎛⎞ +
=+ ⎜⎟ ⎝⎠
⎛⎞ + == ⎜⎟ ⎝⎠
D
m =
Horizontal Soil Spring:
23 4
23 6.816 2.019 0.146 0.00765 1.683(10)

6.816 2.019(3.4) 0.146(3.4) 0.007651(3.4) 1.683(10) (3.4)

12.3 qh

HH H H N
DD D D
−
−
⎛⎞ ⎛⎞ ⎛⎞ =+ − + − ⎜⎟ ⎜⎟ ⎜⎟ ⎝⎠ ⎝⎠ ⎝⎠
=+ − + −
=
4 4 4
P N cD N HD u ch qh =+ = γ 12 3 110 5733 1667 12 930 189 . ( )( . )( . ) , = = lb ft
kN
m
∆p H D=+ F
H
G I
K
J = ⋅+ F
H
G I
K 0 04 J = 2
0 04 12 5 733 20
2 . . . 32. inches = 8 cm
PR-268-9823
2306
Appendix D: Example Calculations D-13
Vertical Bearing Soil Spring:

Ne e q =+ F
H
G I
K
J =+ F
H
G I
K
J = πφ tan( ) φ π tan( ) tan tan 235 2 45
2 45 35
2 33
Ne c =+ + Nq
F +
H
G I
K
J − L
N
M O
Q
P =− + cot( . ) tan . cot( ) tan( . ) φ πφ φ 0 001 45 0 001 2 1 35 1 0 001 2 = 46
(0.18 2.5)
(0.18(35) 2.5) 45
Ne
e
φ
γ
−
−
=
==
2 2
2 (1.667) 33(110)(5.733)(1.667) (45)(110) 2
lb kN 41,570 607 ft m
dc q
D Q N cD N HD Nγ =+ + γγ
=+
==
∆qd == = 0125 0125 20 2 5 . . ( ) . inches D = 5 cm
Vertical Uplift Soil Spring:

N H
D qv = F
H
G I
K
J
F
H
G I
K
J = F
H
G I
K tan( ) φ tan( ) . . J = φ
44 35 35
44 34 19
Q N S D N HD u cv u qv =+ = γ 162 110 4 833 1667 1 431 1 430 . ( )( . )( . ) , , = ≈ lb ft
∆qu
qv cv
cv qv
H N HD H N cD
N cD N HD = H
+
+ == = 012 01
012 012 58 7 0 . ( ) .( ) . .() . γ
γ
inches
Step 7: Perform Analysis
The analytical model for assessing pipeline response is provided in Figure D.6. In plan, the pipeline is straight with the pipe anchored against movement at the center of the river crossing and at a point approximately 1,000 ft south of the ground deformation boundary.

Anchoring the pipeline at the center of the river assumes that that north bank may experience a similar ground failure and the center of the river is a point of symmetry. If no instability of the north bank occurs, this assumption will lead to an overestimate of the pipe strains at the center of the river and an underestimate of the pipe strains at the onshore ground failure boundary. The southerly anchorage point was determined based upon the recommendation in Section 4.4.

PR-268-9823
2307
Appendix D: Example Calculations D-14
(26)(0.375)52,000 758 ft 231 m
2,100
y
anchor
u
Dt
L
T
π σ π == = =
The length of the pipe elements is one pipe diameter within 100 feet of the ground displacement boundaries. Between the zones with element length of 1 pipe diameter, the pipe elements are allowed to be up to 10 feet long.

No horizontal or vertical soil springs are specified more than 100 feet from the ground displacement boundaries as the pipeline is only exposed to axial loading. Lateral movement (horizontal and vertical) is prevented for this portion of the analytical model.

The ground displacement pattern is applied in 120 equal increments to a maximum displacement that is about 50% greater (4 m) than the estimates from Step 4 (2.9 m).

This allows the behavior of the pipeline to be evaluated for alternate assumptions that might increase the amount of applied ground displacement.

The amount of axial ground movement relative to the pipe is provided in Figure D.7a. As can be seen in Figure D.7a, the pipeline response is nearly separated between the head and toe regions of the lateral spread. The onshore portion of the pipeline acts to restrain the pipe from ground displacements creating tension in the pipe at the head of the lateral spread. Near the toe of the lateral spread, the ground displacement produces compression in the pipe that is resisted by the pipe outside of the zone of ground displacement. Figure D.7b illustrates the variation of maximum strain magnitude (tension or compression) along the pipeline at a maximum ground displacement of 4 m. The location of maximum tensile and compressive longitudinal strains are also indicated in Figure D.7b.

The variation in maximum longitudinal strain as a function of maximum ground displacement magnitude is shown in Figure D.8. For a specific ground displacement value, the strains plotted in Figure D.8 represent the largest total strain at any of 8 equidistant circumferential locations around the pipe cross-section at any location in the pipe model. For the ground displacement estimated in Step 4, the maximum strain is approximately 0.8% tension and 0.45% compression.

Step 8: Assess Nominal Pipeline Performance for Computed Strain Levels An assessment will first be performed assuming the pipeline construction meets the requirements of Section 3.1. For a performance requirement that pressure integrity be maintained, the allowable longitudinal tensile strain limit is assumed to be 4%. The corresponding longitudinal compressive strain limit is computed as follows:

For the concrete-coated pipe:
.375 1.76 1.76 3.3%
20 cp
t
D
ε == =
For the onshore portion of the pipeline (t = 0.280 inches) PR-268-9823
2308
Appendix D: Example Calculations D-15
.28 1.76 1.76 2.5%
20 cp
t
D
ε == =
The pipe strain computed for 2.9 m of maximum ground displacement is far less than the above strain limits. On this basis, it is clear that the pipeline is not at risk for the ground displacements modeled in the analysis.

Additional analyses that should be performed, but are not included in this example problem, include the following modifications:

1. Variation in the soil springs to account for uncertainties in the estimated values.

2. Variation in the ground displacement pattern that might include assessing the pipeline assuming the surface ground displacements occur at the pipeline depth.

3. Extension of the pipeline model to represent a condition at the center of the river in which the pipeline is not anchored.

Step 9: Assess Potential for Landslide Hazard South of River Crossing In addition to the lateral spread displacement, the slope south of the river is examined to determine if a more detailed geotechnical investigation of potential slope instability is necessary. The pipeline is located within the near-surface granular soils (Group B in Figure 2.8) that overly the weathered shale (Group C in Figure 2.8) of the hill. Two potential slide mechanisms are considered:

1. A shallow disruptive slide consisting of failure at the interface between surface soils and the underlying weathered shale.

2. A deep coherent slide within the weathered shale.

The weathered shale is assumed to provide a low-permeability boundary. Therefore, the interface between the surface soil and the weathered shale is assumed to be “wet” while the weathered shale is considered to be “dry.” Noting that the slope of the hillside is 10%, the critical acceleration for sliding is estimated from Figure 2.8 to be 0.24 g for the disruptive slide and 0.20 g for the coherent slide.

The ground motions at the hillside are better represented by using soft rock conditions in the attenuation relationship (Ssr = 1). In the Campbell attenuation relationship, this results in the following additional factor:

0.44 0.17 ln 0.44 0.17 ln(40) 0.187 − = − =− R
The following correction factor is applied to the PGA:

0.187 PGA e g 0.18 0.83 0.15 g − ⋅ = ⋅=
PR-268-9823
2309
Appendix D: Example Calculations D-16
Since the PGA is less than the critical acceleration, there are no concerns for potential landslide movements.

PR-268-9823
2310
Appendix D: Example Calculations D-17
Weathered Shale
Flood Plain
-50
-40
-30
-20
-10
0 10
20 2300 2350 2400 2450 2500 2550 2600 2650 2700 2750 2800 2850 2900 STATION, m
GEODETIC ELEVATION, m
NPS 20
Ground Surface
CPT 1 CPT 2
CPT 3
SPT 1
Liquefiable
Figure D.1 General Plan and Profile of NPS 20 Pipeline River Crossing PR-268-9823
2311
Appendix D: Example Calculations D-18 Depth, ft
Description
Strata Sample No. Blows
0 Ground Surface
Soft, brown silt, some organics,
trace sand (TOPSOIL)
5 Soft to firm, gray SAND 3
10 1 6
Loose, gray SILT, some SAND to
SILT and SAND
15 2 5
20 3 8
Loose to compact, grey SAND,
trace to some silt. Contains
thin SILT lenses below 24 ft
depth.
25 4 13
Loose to moderate dense gray
30 SAND 5 16
Figure D.2 SPT Boring Results
PR-268-9823
2312
Appendix D: Example Calculations D-19
qc fs
Figure D.3 Log from CPT 3
PR-268-9823
2313
Appendix D: Example Calculations D-20
SPT 1
Depth: 10.2 m
Figure D.4 Gradation Curve from SPT Sample
PR-268-9823
2314
Appendix D: Example Calculations D-21
0 0.5
1 1.5
2 2.5
3 3.5
2300 2350 2400 2450 2500 2550 2600 2650 2700 2750 2800
STATION, m
Horizontal Surface
Displacement, m
Horizontal Displacement
at Pipe, m
Vertical Settlement
Beneath Pipe, m
-50
-40
-30
-20
-10
0 10
20 30
40 50
2300 2400 2500 2600 2700 2800 2900 3000 3100 3200 3300 3400 3500 STATION, m

0.500 wall
Concrete

0.375 wall
Modeled as fixed against
translation and rotation
Modeled as fixed
against translation
and rotation
Regions modeled with
shorter pipe elements
1 0.1
Figure D.5: Estimated Liquefaction-Induced Ground Displacement Figure D.6 Schematic Representation of Model Used to Analyze Pipeline Response PR-268-9823
2315
This Page Intentionally Left Blank
2316
Appendix E: Guidelines Reviewers E-22
Maximum Compression
Maximum Tension
Maximum Tension
Maximum Compression
a) Relative axial soil displacement along pipeline at 4-m maximum ground displacement b) Maximum strain magnitude in the pipe at 4-m maximum ground displacement Relative Axial
Ground
Movement
(+ to right)
Maximum
Longitudinal
Strain Magnitude
Figure D.7 Maximum Total Compression Strain at 4-m Maximum Ground Displacement
PR-268-9823
2317
Appendix E: Guidelines Reviewers E-23
-0.01
-0.008
-0.006
-0.004
-0.002
0 0.002

0.004

0.006

0.008

0.01 012345
Maximum Ground Displacement (m)
Maximum Longitudinal Strain (mm/mm)
Tension
Compression.

Maximum Applied Ground Displacement
Figure D.8 Variation of Strain with Maximum Ground Displacement PR-268-9823
2318
Appendix E: Guidelines Reviewers E-1
Reviewer
(Subject Area)
Background
Jean Audibert
(Soil Mechanics)
Dr. Jean Audibert is currently Engineering Department Manager for Fugro-McClelland in Houston, Texas . His responsibilities include overall supervision, coordination of engineers, and technical direction and management of offshore geotechnical projects. Dr. Audibert has over 30 years of professional engineering and research experience related to the oil and gas industry and is the author or co-author of over 40 technical papers.

He is recognized for his research work on soil-pipeline interaction and has applied his unique knowledge to the design of offshore pipelines subjected to seafloor instabilities and earthquake strong motions or faulting.

David Murray
(Pipe Strain Limits)
Dr. David W. Murray received his B.Sc. in Civil Engineering from the University of Alberta, Canada, his M.Sc. from Imperial College, England, and his Ph. D. from UC Berkley, USA. He was a
professor of civil engineering at the University of Alberta for 35 years and has been Professor Emeritus there since 1992. For the past 15 years he has been working primarily on predicting the physical behavior of pipelines using extensive laboratory testing of full-sized specimens and computational numerical analysis.

James Hart
(Analysis Procedure)
Dr. James Hart is the President of SSD, Inc. He has extensive experience in pipeline engineering problems with an emphasis on evaluation of pipelines subjected to extreme loading such as thaw settlement, frost heave, imposed fault movement and windinduced vibration. He has hands-on familiarity with all aspects of stress and deformation analysis of pipelines. He has been closely involved with numerous pipeline serviceability evaluations, experimental evaluations of operating pipelines, and full-scale pipe testing programs. He is a member of ASCE and ASME, and is a registered civil engineer in the States of Nevada, California and Alaska.

William Lettis
(Geology & Seismology)
Dr. William Lettis is the President and Principal Geologist of William Lettis & Associates, Inc. He specializes in the assessment of seismic hazards for engineered structures and has worked on critical facilities throughout the world. Dr. Lettis investigated the impact of surface fault rupture to pipelines following the 1999 Tukey and Taiwan earthquakes, and has evaluated existing and new pipelines to develop mitigation strategies at fault-pipeline crossings along the San Andreas, Hayward, and Calaveras faults in California.

PR-268-9823
2319
Appendix E: Guidelines Reviewers E-2
Reviewer
(Subject Area)
Background
Mike Rosenfeld
(Fracture Mechanics,
Welding Procedures, &
Pipe Strain Limits)
Mr. Mike Rosenfeld is currently the President of Kiefner & Associates, Inc. Mr. Rosenfeld received his B.S degree in mechanical engineering from the University of Michigan in 1979 and his M.S. degree in mechanical engineering from CarnegieMellon University in 1981. He specializes in structural mechanics, stress analysis by finite element and traditional methods, component design, and pressure vessel and piping design. Mr.

Rosenfeld’s project experience includes developing guidelines for in-service pipeline relocation, welding procedure qualification, and research related to criteria development for fitness-for-service applications and the acceptance of dents and corrosion defects on pipelines.

Robert Warke
(Fracture Mechanics &
Pipe Strain Limits)
Mr. Robert Warke (currently Assistant Professor, Welding Engineering and Engineering Technology at Le Tourneau
University) reviewed the guidelines while he was a Senior Engineer at Southwest Research Institute. He has over fifteen years of consulting, research and industrial experience emphasizing structural integrity issues. He holds a BS in Welding Engineering from LeTourneau University and an MS in
Metallurgical and Materials Engineering from Illinois Institute of Technology. Prior to joining SwRI, Mr. Warke held materials engineering positions at the Edison Welding Institute, Case Corporation and Packer Engineering, Inc. His particular areas of expertise include weldment defects, properties and failure mechanisms, fractographic interpretation, root cause failure analysis, and the application of probabilistic structural mechanics to the assessment of pipeline reliability.

Thomas Zimmerman
(Analysis Methods &
Pipe Strain Limits)
Dr. Thomas Zimmerman, has twenty-two years of experience in structural engineering design and research. Working at C-FER Technologies for the last seventeen years, he has been
responsible for research and development activities related to pipelines, offshore structures, pressure vessels, and conventional bridge and building systems. In 1994 he was the project manager for a major C-FER research project studying compressive strain limits for buried pipelines. He chaired the Canadian Standards Association (CSA) Task Force on Limit States Design of Pipelines, and co-authored of the first draft of that document. He is currently a member of ISO Working Group 12, which is in the process of developing a reliability-based design standard for onshore pipelines.

PR-268-9823
2320
This Page Intentionally Left Blank
2321
Copyright, 2005
All Rights Reserved by Pipeline Research Council International, Inc.

PRCI Reports are Published by Technical Toolboxes, Inc. 3801 Kirby Drive, Suite 340
Houston, Texas 77098
Tel: 713-630-0505
Fax: 713-630-0560
Email: [email protected]
2322
City of American Canyon—Giovannoni Logistics Center Project Draft EIR
FirstCarbon Solutions
Appendix F:

Hazardous Materials Supporting Information
2323

THIS PAGE INTENTIONALLY LEFT BLANK
2324
220161 Giovannoni HH Memo 12-3-2020 1
BALANCE HYDROLOGICS, Inc.

MEMO
To: Andrea Bellanca – Carlson, Barbee, and Gibson, Inc.

From: Johnny Caspers, Ed Ballman
Date: December 3, 2020
Subject: Preliminary Hydrologic and Hydraulic Modeling
for the Proposed Giovannoni Logistics Center
Introduction & Project Description
The Giovannoni project (Project) is located on a 208-acre site in the City of American Canyon.

The site sits to the west of Highway 29, north of Green Island Road, and south of the developing Napa Logistics Park. The existing property consists of open space, with a total elevation variance of roughly 30 feet; two highpoints to the southeast and southwest of the Project area slope gradually north to a seasonal wetland that discharges into No Name Creek. Along with the Project area, the creek collects runoff from roughly 280 acres of vineyards, open space and previously developed industrial land uses, the majority of which lies to the east of Highway 29.

The first phase of the Project, referred to as the Giovannoni Logistics Center East and covered herein, proposes the construction of two warehouses (gross square footage of the two buildings is roughly one million square feet) along with parking lots to accommodate semi-trucks, trailers, and worker vehicles on roughly 70 acres. The runoff from the developed area will be managed by multiple stormwater facilities on the northern side of the Property designed to meet flood control requirements set forth by the City of American Canyon Department of Public Works and stormwater treatment and flow-duration control requirements outlined in the state-wide MS4 General Permit.

Although not yet constructed, the City of American Canyon plans to extend Devlin Road from its current terminus at the southern end of the Napa Logistics Park, south to intersect with Green Island Road (see Figure 1). This planned extension will serve as the western boundary for the first phase of the Project. Stormwater runoff that, prior to the construction of the road, would flow from east to west through the northern portion of the site will now be directed through four 2-ft by 4-ft concrete box culverts located beneath the roadway. A hydrologic and hydraulic analysis for the new roadway, entitled Devlin Road and Vine Trail Extension, was completed in August 2020 by GHD and will be referred to herein as the “Devlin Road Analysis”.

Purpose of the Study
This study seeks to model, assess, and outline how the changes in existing, and proposed land cover, along with the proposed construction of Devlin Road, will impact the hydrology and 2325
220161 Giovannoni HH Memo 12-3-2020 2
hydraulics of the creek directly downstream of the Project area and along the reach to its confluence with the Napa River.

Standards & Methods
In compliance with American Canyon Department of Public Works Engineering Standards and Specifications for Public Improvements (1995) where stormwater detention is utilized, Contra Costa County Flood Control (Flood Control) methodology was used to create the hydrologic model of the pre- and post-project conditions1. The following documents were relied upon to inform the modeling and design process:
 Engineering Standards and Specifications for Public Improvements, American Canyon Department of Public Works, May 1995.
 BASMAA Post-Construction Manual, Bay Area Stormwater Management Agencies Association, January 2019.
 HEC-HMS Guidance for Flood Control, Unit Hydrograph Method, March 2018  HEC-HMS Template Model, Flood Control, June 2016
 HYDRO-6 Program Parameters, Flood Control, March 2018  Mean Seasonal Isohyets Compiled from Precipitation Records (1900-1960), File:
isohyetal_cnty, Napa County GIS, January 2011.

Background and Assumptions
For pre-project conditions, the Project site topography (in the proposed development area) has a maximum elevation of approximately 50 feet in the southeast corner and a minimum of approximately 22 feet in the northwest. The post-project grading will utilize this variance to facilitate runoff.

Soil data was obtained from the NRCS’s Soil Survey online database. The database stated that the Project area is comprised of three main soils: Haire loam, Haire clay loam, and Clear Lake clay. These soils, characterized by low rates of infiltration, all fall into the NRCS Hydrologic Soil Group “D”.

Infiltration rates for pre- and post-project conditions were ascertained from Table 1 of the Flood Control HYDRO-6 Program Parameters. The lower bound of the infiltration rates provided was used due to the “D” Soil Group classification outlined above; a rate of 0.17 inches/hour was assumed for existing conditions, and 0.03 inches/hour was assumed for proposed industrial conditions. For watersheds that contained multiple land use categories a weighted average was calculated.
1 Based upon Sections 4.02.A and 4.02.C of the City of American Canyon Department of Public Works Engineering Standards and Specifications for Public Improvements Manual, it is our understanding that the Unit Hydrograph Method outlined by the Contra Costa County Flood Control District is the preferred design method for detention basin analyses.
2326
220161 Giovannoni HH Memo 12-3-2020 3
N-values were determined using Figure 1 (for areas less than 200 acres) of the HYDRO-6 Program Parameters; for pre-project areas a value of 0.065 was assumed, characterizing the site as currently undeveloped, and for post-project areas N-values ranging from 0.029 to 0.04 were estimated using the area-based extrapolation method.

The mean annual precipitation (MAP) for the Project area is 20 inches. This value was used in the GHD Devlin Road Analysis and cross-referenced with an isohyet GIS layer available on the Napa County GIS online database. Precipitation falling on the Project footprint currently sheet flows from southwest to northeast where it enters the wetland area and, ultimately, the unnamed tributary to the Napa River.

Facilities Being Modeled
Detention Upstream of Devlin Road
The construction of Devlin Road will impede the flow of water to No Name Creek during large storm events; therefore, to accurately represent the post-development conditions a detention element was included in the model to account for the active storage provided by the elevated roadway during large flow events. Separate storage relationships were developed for the pre- and post-project conditions to assure that any changes in the open space detention volume east of Devlin is accounted for. The stage-storage relationships used in the modeling are included in Table 1. Specifics of the culvert design, provided in the Devlin Road Analysis, were used to inform the modeled culvert discharge.

Table 1. Devlin Road Elevation-Storage Relationships
Drainage Management Areas
The proposed design divides the Project area into two drainage management areas (DMAs) comprised of a mix of parking lot, roof area, and pervious landscaping. The tentative mapping shows that the eastern DMA A is 31.4 acres in size while DMA B to the west is 27.2 acres

2. In
conjunction with Flood Control standards, the DMAs are modeled under the industrial land use classification (see Attachment 1 for model parameters). 2 Modeled DMA acreages shown in Attachment 1 and Figure 1 are slightly larger than what is provided in the tentative mapping to comply with peak flow modeling requirements; in contrast to water quality, peak flow modeling in HMS requires that rain falling on the footprints of stormwater management facilities is accounted for in basin sizing and flow reduction.
2327
220161 Giovannoni HH Memo 12-3-2020 4
Stormwater Facilities
To meet the flood control requirement outlined by the City of American Canyon in their Engineering Standards and Specifications Manual (post-project peak flows must be reduced to 90% of the pre-project peak) three detention basins are proposed on site. DMA A will have two detention basins and a dedicated bioretention area for water quality, while DMA B will be serviced by one basin sized to comply with both flood control and water quality treatment standards. The bioretention facilities were sized based on the “4% Rule” outlined in the BASMAA Post-Construction Manual. Furthermore, detention basins and bioretention facilities will also be designed to comply with the 2-year hydromodification requirement required for the regional MS4 permit.

Model and Calculation Parameters
To assess the performance of the proposed stormwater facilities during large storm events, a hydrologic model was developed using the U.S. Army Corps of Engineers’ HEC-HMS software and modeling protocols outlined in Flood Control’s HEC-HMS Guidance Manual.

Input parameters for the HEC-HMS model were compiled from site-specific topography and precipitation information, preliminary Project plans, soil survey information, and additional input recommendations from Flood Control’s HEC-HMS Guidance Manual. Attachment 1 shows land use parameters that were used to calculate lag time (in hours) for both modeled conditions.

Model runs were developed for three conditions: existing conditions, existing conditions with the addition of Devlin Road, and the proposed Project conditions (per current CBG design). For the three conditions, model runs were developed for the 2-, 10-, and 100-year design storm recurrence intervals using a 24-hour storm duration3. These storm events were based on a mean annual precipitation of 20 inches and design storm coefficients and rainfall totals specified by Flood Control methodology.

Model Results
Prior to assessing the effect of the projected development, it is important to understand how the construction of Devlin Road will change the hydrology of the local watershed at large. Table 3 shows a comparison of the model results for the watershed prior to, and after, the construction of Devlin Road.

Table 3. Summary of HMS Modeling Results for Devlin Road Construction 3 The 10- and 100- year design storms utilized the Contra Costa County Flood Control 24-hour storm distribution curve with a peak percent runoff of 11.5%, whereas the 2-year storm was modeled using the Flood Control’s 21.2% curve. The higher peaking storm distribution was utilized in the smaller storm event to more accurately assess hydromodification requirements.
2328
220161 Giovannoni HH Memo 12-3-2020 5
Based on the modeled conditions, the construction of the roadway reduces the peak flow for the 2-, 10-, and 100-year 24-hour storm events by 22%, 27%, and 35%, respectively. These results are wholly dependent on the construction of Devlin Road: water is detained behind the elevated roadway and metered out through the box culverts. Beyond the nominal increase in impervious area incurred by the surface of the roadway, these results reflect no change in land use from the pre-project, no development condition. When the post-development land use changes are added to the Devlin Road scenario in HMS the results show an increase in peak flow shown below in Table 4.

Table 4. Summary of HMS Modeling Results for All Scenarios While these results show a net increase in the peak flow after the development of the Project area, it is important to understand what factor is driving this change. Focusing on the results from the project site provides a better understanding of the cause for this increase (see Table 5).

Table 5. Summary of HMS Modeling Results for Project Site Only As indicated by the table above, the proposed stormwater facilities go far beyond the required level of peak flow reduction for the design storm events. These results, when viewed in conjunction with the data in Table 3, attest to the impact that the reduced detention behind Devlin Road has on the system at large. The roughly thirty percent decrease at all elevations in the stage-storage relationship (refer back to Table 1) creates a scenario where the culvert is experiencing a higher water surface elevation for each storm event, despite a marked reduction in peak flows leaving the project site before and after development.

As Table 4 indicates, the construction of Devlin Road is predicted to greatly reduce the preproject peak flow rate as water is backed up to the east of the roadway and metered out through the four 2-foot by 4-foot box culverts. However, through further analysis one can see that this benefit is largely derived from the detention provided by the undeveloped Giovannoni property.

Conclusion
Based upon the outcomes for the three modeled conditions, multiple changes from the preproject scenario are evident. First off, the construction of Devlin Road will largely constrict the flowrate from the watershed to No Name Creek; the 100-yr discharge drops to 65% of the pre2329 220161 Giovannoni HH Memo 12-3-2020 6
project peak flow with the addition of the roadway. Secondly, the peak flow increases to 68% of the pre-project condition for the 100-year storm event after the development of the Project area, but still remains well within the flood control requirements set forth by the City of American Canyon. By viewing this proposed development on the scale of the regional watershed the combined construction of Devlin Road and the Project area will not create adverse effects related to flood control or hydromodification downstream of the development. 2330
Attachment 1. Hydrologic Model Subbasin Parameters for Giovannoni Project American Canyon, Napa County, California.

High Low
acres (in/hr) (feet) (feet) (feet) (feet) (feet) (hours) Watershed 1A 186.2 0.17 0.065 3880 2208 275 60 293 0.339 Watershed 1B 96.1 0.10 0.040 3511 506 54 34 30 0.177
Watershed 1C 64.2 0.17 0.065 4369 2434 54 25 35 0.550
Watershed 2 68.4 0.17 0.065 4505 2336 54 25 34 0.551
Watershed 3 52.5 0.17 0.065 3327 1665 44 22 34 0.431
Watershed 4 19.6 0.17 0.065 1293 484 44 18 104 0.167
Watershed 1A 186.2 0.17 0.065 3880 2208 275 60 293 0.339 Watershed 1B 96.1 0.10 0.040 3511 506 54 34 30 0.177
Watershed 1C* 91.1 0.17 0.065 2767 1434 47 32 29 0.550
Watershed 3* 94.5 0.17 0.065 3269 1665 44 22 34 0.414
Watershed 4 19.6 0.17 0.065 1293 484 44 18 104 0.167
DMA A 37.8 0.03 0.028 2620 1060 48 44 8 0.188
DMA B 31.4 0.03 0.028 2000 1410 40 34 16 0.167
Watershed 1A 186.2 0.17 0.065 3880 2208 275 60 293 0.339 Watershed 1B 96.1 0.10 0.040 3511 506 54 34 30 0.177
Watershed 1C** 29.0 0.17 0.065 2840 1050 47 32 96 0.550 Watershed 3* 94.4 0.17 0.065 3269 1382 44 26 34 0.431
Watershed 4 19.6 0.17 0.065 1293 484 44 18 64 0.167
Modeled
Condition
Lc
Elevation Slope Lag2,3
No
Development
Devlin Road
Only
Shed1 Area Infiltration N
L
PostDevelopment

1. Refer to Figure 1 for location. Asterisks denote that a shed was changed to comply to changes in project conditions; however, the general location of the sheds are maintained across the scenarios.

2. Contra Costa Flood Control District Methodology Lag Time Equation: 24 x N x (L*Lc / S^0.05)^0.38.

3. A minimum Lag Time of 0.167 hours was assumed during modeling. 2331
¬«29
W:\Projects\220161 CAD\220161 GIS\220161_MAP_Overview.mxd Figure 1. Giovannoni Modeling Overview
American Canyon, Napa, California
© 2020 Balance Hydrologics, Inc.

SHED 1B SHED 1A
SHED 1C
SHED 3
DMA B

GREEN ISLAND ROAD DEVLIN ROAD
DMA A
Point of Comparison
SHED 4
No Name Creek
Culvert Location
f
^)
0 290 580 870
Feet ¯
Legend
Longest Flowpath
Proposed Devlin Road
Drainage Management Areas
Offsite Watersheds
2332
City of American Canyon—Giovannoni Logistics Center Project Draft EIR
FirstCarbon Solutions
Appendix G:

Noise Supporting Information
2333

THIS PAGE INTENTIONALLY LEFT BLANK
2334
Mechanical Equipment Noise Calculation
Receptor: Closest Residence - northwest of Project Noise Level Calculation Prior to Implementation of Noise Attenuation Requirements No. Equipment Description Lmax Lmax Leq
1 Commercial grade mechanical ventilation equipment 60 1 100 1470 0.5 0 24.6 15.8 37.71869728 2 Commercial grade mechanical ventilation equipment 60 1 100 1500 0.5 0 24.4 15.5 35.86095691 3 Commercial grade mechanical ventilation equipment 60 1 100 1550 0.5 0 24.2 15.2 33.0385346 4 Commercial grade mechanical ventilation equipment 60 1 100 1600 0.5 0 23.9 14.8 30.51757813 5 Commercial grade mechanical ventilation equipment 60 1 100 1650 0.5 0 23.6 14.5 28.25791804 6 7
8 9
10 Notes: Lmax[4] 25 Leq 22
[1] Percentage of time activity occurs each hour
[2] Soft ground terrain between project site and receptor. [3] Shielding due to terrain or structures
[4] Calculated Lmax is the Loudest value.

Ldn Calculations
Time Hourly Leq Leq' 0.1*Leq antiLog
Night 12:00 AM 22.2 32.2 3.218518923 1653.93685
1:00 AM 22.2 32.2 3.218518923 1653.93685
2:00 AM 22.2 32.2 3.218518923 1653.93685
3:00 AM 22.2 32.2 3.218518923 1653.93685
4:00 AM 22.2 32.2 3.218518923 1653.93685
5:00 AM 22.2 32.2 3.218518923 1653.93685
6:00 AM 22.2 32.2 3.218518923 1653.93685
Day 7:00 AM 22.2 22.2 2.218518923 165.393685
8:00 AM 22.2 22.2 2.218518923 165.393685
9:00 AM 22.2 22.2 2.218518923 165.393685
10:00 AM 22.2 22.2 2.218518923 165.393685
11:00 AM 22.2 22.2 2.218518923 165.393685
12:00 PM 22.2 22.2 2.218518923 165.393685
1:00 PM 22.2 22.2 2.218518923 165.393685
2:00 PM 22.2 22.2 2.218518923 165.393685
3:00 PM 22.2 22.2 2.218518923 165.393685
4:00 PM 22.2 22.2 2.218518923 165.393685
5:00 PM 22.2 22.2 2.218518923 165.393685
6:00 PM 22.2 22.2 2.218518923 165.393685
7:00 PM 22.2 22.2 2.218518923 165.393685
8:00 PM 22.2 22.2 2.218518923 165.393685
9:00 PM 22.2 22.2 2.218518923 165.393685
Night 10:00 PM 22.2 32.2 3.218518923 1653.93685
11:00 PM 22.2 32.2 3.218518923 1653.93685
Sum 17366.33692
Sum/24 723.5973717
Log10(Sum/24) 2.859496981
10*Log10(Sum/24) 28.59496981
24 Hour Ldn 29
Shielding
(dBA)[3]
Calculated (dBA)
Energy
Reference (dBA)
25 ft
Quantity
Usage
factor[1]
Distance to
Receptor
Ground
Effect[2]
2335
Parking Lot activity Reference Noise Level Calculation
No. Equipment Description Lmax Lmax Leq
1 parking lot activity 70 8 1 1360 1 0 41.3 16.0 39.75422349 2 parking lot activity 70 10 1 1400 1 0 41.1 16.6 45.55393586 3 parking lot activity 70 20 1 1450 1 0 40.8 19.1 82.00418221 4 parking lot activity 70 20 1 1500 1 0 40.5 18.7 74.07407407 5 6
7 8
9 10
Notes: Lmax[4] 41 Leq 24
[1] Percentage of time maximum noise levels occur each hour [2] Soft ground terrain between project site and receptor. [3] Noise reduction factor due to terrain, landscaping, or structures Ldn Calculations [4] Calculated Lmax is the Loudest value. Time Hourly Leq Leq' 0.1*Leq antiLog Night 12:00 AM 23.8 33.8 3.382712826 2413.864156
1:00 AM 23.8 33.8 3.382712826 2413.864156
2:00 AM 23.8 33.8 3.382712826 2413.864156
3:00 AM 23.8 33.8 3.382712826 2413.864156
4:00 AM 23.8 33.8 3.382712826 2413.864156
5:00 AM 23.8 33.8 3.382712826 2413.864156
6:00 AM 23.8 33.8 3.382712826 2413.864156
Day 7:00 AM 23.8 23.8 2.382712826 241.3864156
8:00 AM 23.8 23.8 2.382712826 241.3864156
9:00 AM 23.8 23.8 2.382712826 241.3864156
10:00 AM 23.8 23.8 2.382712826 241.3864156
11:00 AM 23.8 23.8 2.382712826 241.3864156
12:00 PM 23.8 23.8 2.382712826 241.3864156
1:00 PM 23.8 23.8 2.382712826 241.3864156
2:00 PM 23.8 23.8 2.382712826 241.3864156
3:00 PM 23.8 23.8 2.382712826 241.3864156
4:00 PM 23.8 23.8 2.382712826 241.3864156
5:00 PM 23.8 23.8 2.382712826 241.3864156
6:00 PM 23.8 23.8 2.382712826 241.3864156
7:00 PM 23.8 23.8 2.382712826 241.3864156
8:00 PM 23.8 23.8 2.382712826 241.3864156
9:00 PM 23.8 23.8 2.382712826 241.3864156
Night 10:00 PM 23.8 33.8 3.382712826 2413.864156
11:00 PM 23.8 33.8 3.382712826 2413.864156
Sum 25345.57364
Sum/24 1056.065568
Log10(Sum/24) 3.023690883
10*Log10(Sum/24) 30.23690883
24 Hour Ldn 30
Calculated (dBA)
Energy
Reference (dBA)
50 ft
Quantity
Usage
factor[1]
Distance to
nearest Property
Line
Ground
Effect[2]
Shielding
(dBA)[3]
2336
Loading/Unloading
Receptor: Closest receptor - school south of the site Noise Level Calculation Prior to Implementation of Noise Attenuation Requirements No. Equipment Description Lmax Lmax Leq
1 Loading/Unloading - surface level 70 10 5 1600 1 0 39.9 21.8 152.5878906 2 Loading/Unloading - surface level 70 10 5 1650 1 0 39.6 21.4 139.1323705 3 Loading/Unloading - surface level 70 10 5 1700 1 0 39.4 21.0 127.2135152 4 Loading/Unloading - surface level 70 10 5 1750 1 0 39.1 20.7 116.6180758 5 Loading/Unloading - docks 70 7 5 1800 1 0 38.9 18.8 75.01714678 6 Loading/Unloading - docks 70 7 5 1850 1 0 38.6 18.4 69.09758553 7 Loading/Unloading - docks 70 7 5 1900 1 0 38.4 18.0 63.78480828 8 Loading/Unloading - docks 70 7 5 1950 1 0 38.2 17.7 59.00301758 9 10
Notes: Lmax[4] 40 Leq 29
[1] Percentage of time activity occurs each hour
[2] Soft ground terrain between project site and receptor. [3] Shielding reduction from proposed warehouse structure [4] Calculated Lmax is the Loudest value. Ldn Calculations Time Hourly Leq Leq' 0.1*Leq antiLog
Night 12:00 AM 29.0 39.0 3.904420368 8024.544103
1:00 AM 29.0 39.0 3.904420368 8024.544103
2:00 AM 29.0 39.0 3.904420368 8024.544103
3:00 AM 29.0 39.0 3.904420368 8024.544103
4:00 AM 29.0 39.0 3.904420368 8024.544103
5:00 AM 29.0 39.0 3.904420368 8024.544103
6:00 AM 29.0 39.0 3.904420368 8024.544103
Day 7:00 AM 29.0 29.0 2.904420368 802.4544103
8:00 AM 29.0 29.0 2.904420368 802.4544103
9:00 AM 29.0 29.0 2.904420368 802.4544103
10:00 AM 29.0 29.0 2.904420368 802.4544103
11:00 AM 29.0 29.0 2.904420368 802.4544103
12:00 PM 29.0 29.0 2.904420368 802.4544103
1:00 PM 29.0 29.0 2.904420368 802.4544103
2:00 PM 29.0 29.0 2.904420368 802.4544103
3:00 PM 29.0 29.0 2.904420368 802.4544103
4:00 PM 29.0 29.0 2.904420368 802.4544103
5:00 PM 29.0 29.0 2.904420368 802.4544103
6:00 PM 29.0 29.0 2.904420368 802.4544103
7:00 PM 29.0 29.0 2.904420368 802.4544103
8:00 PM 29.0 29.0 2.904420368 802.4544103
9:00 PM 29.0 29.0 2.904420368 802.4544103
Night 10:00 PM 29.0 39.0 3.904420368 8024.544103
11:00 PM 29.0 39.0 3.904420368 8024.544103
Sum 84257.71308
Sum/24 3510.738045
Log10(Sum/24) 3.545398426
10*Log10(Sum/24) 35.45398426
24 Hour Ldn 35
Shielding
(dBA)[3]
Calculated (dBA)
Energy
Reference (dBA)
50 ft
Quantity
Usage
factor[1]
Distance to
Receptor
Ground
Effect[2]
2337
TABLE Existing-01

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Paoli Loop Road - south of Green Island Road NOTES: Giovannoni Logistics Center Project - Existing
______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 3000 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 59.92 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 55.4 118.7
______________________________________________________________________

2338
TABLE Existing-02

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - Paoli Loop Road to future Devlin Road
NOTES: Giovannoni Logistics Center Project - Existing
______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 4800 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 61.97 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 75.5 162.2
______________________________________________________________________

2339
TABLE Existing-03

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - west of future Devlin Road NOTES: Giovannoni Logistics Center Project - Existing
______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 4800 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 61.97 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 75.5 162.2
______________________________________________________________________

2340
TABLE Existing-04

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: South Kelly Road - SR 29 to Devlin Road NOTES: Giovannoni Logistics Center Project - Existing
______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 1400 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 56.62 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 0.0 71.6
______________________________________________________________________

2341
TABLE Existing-05

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: SR 29 - South Kelly Road to Green Island Road NOTES: Giovannoni Logistics Center Project - Existing
______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 30900 SPEED (MPH): 65 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 24 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 73.84 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------
128.4 272.8 585.9 1261.2
______________________________________________________________________

2342
TABLE Existing plus Project-01

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Paoli Loop Road - south of Green Island Road NOTES: Giovannoni Logistics Center Project - Existing plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 3700 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 60.84 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 63.6 136.4
______________________________________________________________________

2343
TABLE Existing plus Project-02

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - Paoli Loop Road to future Devlin Road
NOTES: Giovannoni Logistics Center Project - Existing plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 6400 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 63.22 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 91.4 196.5
______________________________________________________________________

2344
TABLE Existing plus Project-03

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - west of future Devlin Road NOTES: Giovannoni Logistics Center Project - Existing plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 4800 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 61.97 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 75.5 162.2
______________________________________________________________________

2345
TABLE Existing plus Project-04

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: South Kelly Road - SR 29 to Devlin Road NOTES: Giovannoni Logistics Center Project - Existing plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 2500 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 59.13 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 0.0 105.2
______________________________________________________________________

2346
TABLE Existing plus Project-05

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: SR 29 - South Kelly Road to Green Island Road NOTES: Giovannoni Logistics Center Project - Existing plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 31300 SPEED (MPH): 65 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 24 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 73.90 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------
129.5 275.2 590.9 1272.1
______________________________________________________________________

2347
TABLE Future without Project-01

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Paoli Loop Road - south of Green Island Road NOTES: Giovannoni Logistics Center Project - Future without Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 3700 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 60.84 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 63.6 136.4
______________________________________________________________________

2348
TABLE Future without Project-02

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - Paoli Loop Road to future Devlin Road
NOTES: Giovannoni Logistics Center Project - Future without Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 5900 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 62.86 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 86.6 186.1
______________________________________________________________________

2349
TABLE Future without Project-03

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - west of future Devlin Road NOTES: Giovannoni Logistics Center Project - Future without Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 5900 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 62.86 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 86.6 186.1
______________________________________________________________________

2350
TABLE Future without Project-04

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: South Kelly Road - SR 29 to Devlin Road NOTES: Giovannoni Logistics Center Project - Future without Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 2400 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 58.96 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 0.0 102.3
______________________________________________________________________

2351
TABLE Future without Project-05

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: SR 29 - South Kelly Road to Green Island Road NOTES: Giovannoni Logistics Center Project - Future without Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 52300 SPEED (MPH): 65 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 24 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 76.13 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------
180.8 386.7 831.8 1791.0
______________________________________________________________________

2352
TABLE Future plus Project-01

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Paoli Loop Road - south of Green Island Road NOTES: Giovannoni Logistics Center Project - Future plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 4300 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 61.49 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 70.2 150.8
______________________________________________________________________

2353
TABLE Future plus Project-02

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - Paoli Loop Road to future Devlin Road
NOTES: Giovannoni Logistics Center Project - Future plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 7400 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 63.85 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 100.6 216.5
______________________________________________________________________

2354
TABLE Future plus Project-03

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: Green Island Road - west of future Devlin Road NOTES: Giovannoni Logistics Center Project - Future plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 5900 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 62.86 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 86.6 186.1
______________________________________________________________________

2355
TABLE Future plus Project-04

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: South Kelly Road - SR 29 to Devlin Road NOTES: Giovannoni Logistics Center Project - Future plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 3500 SPEED (MPH): 40 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 6 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 60.59 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------

0.0 0.0 61.3 131.5
______________________________________________________________________

2356
TABLE Future plus Project-05

FHWA ROADWAY NOISE LEVEL ANALYSIS
RUN DATE: 08/25/2021
ROADWAY SEGMENT: SR 29 - South Kelly Road to Green Island Road NOTES: Giovannoni Logistics Center Project - Future plus Project ______________________________________________________________________ * * ASSUMPTIONS * *
AVERAGE DAILY TRAFFIC: 52700 SPEED (MPH): 65 GRADE: .5

TRAFFIC DISTRIBUTION PERCENTAGES

DAY NIGHT
--- -----
AUTOS

88.08 9.34

M-TRUCKS

1.65 0.19

H-TRUCKS

0.66 0.08
ACTIVE HALF-WIDTH (FT): 24 SITE CHARACTERISTICS: SOFT
______________________________________________________________________ * * CALCULATED NOISE LEVELS * *
Ldn AT 50 FT FROM NEAR TRAVEL LANE CENTERLINE (dB) = 76.16 DISTANCE (FEET) FROM ROADWAY CENTERLINE TO Ldn
70 Ldn 65 Ldn 60 Ldn 55 Ldn
------- ------- ------- -------
181.7 388.7 836.0 1800.2
______________________________________________________________________ 2357
2358
City of American Canyon—Giovannoni Logistics Center Project Draft EIR
FirstCarbon Solutions
Appendix H:

Transportation Supporting Information
2359

THIS PAGE INTENTIONALLY LEFT BLANK
2360
Balancing Functionality and Livability since 1995
w-trans.com

TRAFFIC ENGINEERING

TRANSPORTATION PLANNING
July 22, 2021
Prepared for the City of American Canyon
Submitted by
W-Trans
Traffic Impact Study for the
Giovannoni Logistics Center
2361
2362
i Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Table of Contents
Executive Summary .................................................................................................................................................................. 1 Introduction................................................................................................................................................................................. 2 Transportation Setting............................................................................................................................................................. 4 Capacity Analysis....................................................................................................................................................................... 7 Vehicle Miles Traveled .......................................................................................................................................................... 20 Alternative Modes .................................................................................................................................................................. 24 Access and Circulation .......................................................................................................................................................... 25 Parking ........................................................................................................................................................................................ 26 CEQA Checklist ........................................................................................................................................................................ 27 Conclusions and Recommendations ............................................................................................................................... 29 Study Participants and References ................................................................................................................................... 30 Figures

1. Study Area and Existing Lane Configurations...................................................................................................................... 3

2. Existing Traffic Volumes.............................................................................................................................................................10

3. Future Traffic Volumes................................................................................................................................................................12

4. Site Plan............................................................................................................................................................................................14

5. Project Traffic Volumes...............................................................................................................................................................16

6. Existing plus Project Traffic Volumes....................................................................................................................................17

7. Future plus Project Traffic Volumes.......................................................................................................................................18 Tables

1. Collision Rates for the Study Intersections............................................................................................................................ 5

2. Bicycle Facility Summary .............................................................................................................................................................. 6

3. Intersection Level of Service Criteria........................................................................................................................................ 8

4. Existing Peak Hour Intersection Levels of Service............................................................................................................... 9

5. Future Peak Hour Intersection Levels of Service ..............................................................................................................11

6. Trip Generation Summary.........................................................................................................................................................13

7. Trip Distribution Assumptions.................................................................................................................................................15

8. Existing and Existing plus Project Peak Hour Intersection Levels of Service.........................................................15

9. Future and Future plus Project Peak Hour Intersection Levels of Service..............................................................19

10. Vehicle Miles Traveled Analysis Summary ..........................................................................................................................20

11. Estimated Employee VMT Reduction....................................................................................................................................23

12. Parking Analysis Summary........................................................................................................................................................26

13. XVII. Transportation/Traffic......................................................................................................................................................27 2363
ii Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Appendices
A. Collision Rate Calculations
B. Traffic Count Data
C. Intersection Level of Service Calculations
2364
1 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Executive Summary
The proposed project includes the construction of 2.2 to 2.4 million square feet of warehouse which would be built concurrently with the planned extension of Devlin Road between Green Island Road and South Kelly Road in the City of American Canyon. This project is expected to generate an average of 3,888 equivalent passenger car trips per day, including 240 trips during the a.m. peak hour, and 264 during the p.m. peak hour.

The study area for the operational analysis includes the intersections of Devlin Road/South Kelly Road, Devlin Road/Green Island Road and Paoli Loop Road/Green Island Road. Analysis indicates that the unsignalized intersections operate acceptably under Existing Conditions and would continue to do so under Future Conditions and upon the addition of project-related traffic volumes. SR 29/South Kelly Road would operate at LOS F under Future Conditions without or with the project, but this operation was considered acceptable since the intersection is exempt from the City’s LOS standard and Caltrans has no operational standard. As required by the City in their Traffic Impact Fee Program, the project should pay a fee toward the cost of the planned future improvements to SR 29 to address cumulative effects of project traffic.

Although the project would be expected to have a less-than-significant transportation impact on vehicle miles traveled (VMT), the implementation of TDM measures such as use of ridesharing, offering secure bicycle parking, and educational materials would further lower the anticipated project VMT and are therefore recommended.

The project includes the addition of sidewalks along the Devlin Road extension and a multi-modal path along Green Island Road in the area adjacent to the project. Bicycle facilities serving the project site are generally adequate, though space should be provided for bicycles to be stored on-site. The project includes an extension to the Napa Valley Vine Trail along Delvin Road which improves local and regional bicycle connectivity. Existing transit stops are not within the immediate vicinity of the site. Thisis typical for this area and is therefore considered an acceptable condition.

Vehicles would access the project via multiple driveways on Green Island Road and the future Devlin Road extension. Sight distances at the project driveways for both entering and exiting drivers is expected to be adequate assuming placement of buildings, signs and landscaping are outside the areas needed for clear lines of sight. Assuming a maximum project size of 2.4 million square feet, City Code requires a minimum of 1,230 parking spaces should be provided.
2365
2 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Introduction
This report presents an analysis of the potential traffic impacts that would be associated with the proposed development of 2.2 to 2.4 million square feet of warehouse space between Green Island Road and Devlin Road in the area west of SR 29 in the City of American Canyon. The traffic study was completed in accordance with the criteria established by the City of American Canyon and is consistent with standard traffic engineering techniques.

Prelude
The purpose of a traffic impact study is to provide City staff and policy makers with data that they can use to make an informed decision regarding the potential traffic impacts and effects on traffic operation of a proposed project, and any associated improvements that would be required in order to mitigate these impacts and effects to an acceptable level under CEQA, the City’s General Plan, or other policies. Impacts relative to access for pedestrians, bicyclists, and to transit are addressed in the context of the CEQA criteria. Consistent with SB 743, the project’s transportation impacts were analyzed using VMT. While no longer a part of the CEQA review process, vehicular traffic service levels at key intersections were evaluated for consistency with General Plan policies by determining the number of new trips that the proposed use would be expected to generate, distributing these trips to the surrounding street system based on anticipated travel patterns specific to the proposed project, then analyzing the impact the new traffic would be expected to have on the study intersections.

Project Profile
The project as proposed includes 2.2 to 2.4 million square feet of high cube warehouse space. The site is situated between Green Island Road and Devlin Road in the area west of SR 29 in the City of American Canyon and would be accessible from either roadway. Approximate boundaries of the site are illustrated in Figure 1. 2366
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 1 – Study Area and Lane Configurations
Study Intersection
LEGEND
2 3 Future Devlin Rd 4
1 2367
4 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Transportation Setting
Operational Analysis
Study Area and Periods
The study area consists of the following intersections selected through consultation from City staff:

1. SR 29/South Kelly Road

2. Devlin Road/South Kelly Road

3. Devlin Road/Green Island Road (Future Intersection)

4. Paoli Loop Road/Green Island Road
Operating conditions during the weekday a.m. and p.m. peak periods were evaluated to capture the highest potential impacts for the proposed project as well as the highest volumes on the local transportation network.

The morning peak hour occurs between 7:00 and 9:00 a.m. and reflects conditions during the home to work or school commute, while the p.m. peak hour occurs between 4:00 and 6:00 p.m. and typically reflects the highest level of congestion during the homeward bound commute.

Study Intersections
SR 29/South Kelly Road is a signalized four-legged intersection with protected left turns on all approaches. There are crosswalks on the west, north and east legs.

Devlin Road/South Kelly Road is an all-way stop-controlled intersection with crosswalks on the east and south legs.

Devlin Road/Green Island Road (Future) is a future tee intersection with stop control on the southbound Devlin Road approach. Crosswalks are planned at the north leg.

Paoli Loop Road/Green Island Road is a tee intersection with stop control on the northbound Paoli Loop Road approach. There are no crosswalks.

The locations of the study intersections and the existing lane configurations and controls are shown in Figure 1.

Collision History
The collision history for the study area was reviewed to determine any trends or patterns that may indicate a safety issue. Collision rates were calculated based on records available from the California Highway Patrol as published in their Statewide Integrated Traffic Records System (SWITRS) reports. The most current five-year period available is October 1, 2015 through September 30, 2020.

As presented in Table 1, the calculated collision rates for the study intersections were compared to average collision rates for similar facilities statewide, as reported in the publication 2016 Collision Data on California State Highways, California Department of Transportation (Caltrans). These average rates statewide are for intersections in the same environment (urban, suburban, or rural), with the same number of approaches (three or four), and the same controls (all-way stop, two-way stop, or traffic signal). A more detailed evaluation is provided for study intersections where the collision rates were higher than the statewide average. The collision rate calculations are provided in Appendix A.
2368
5 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Table 1 – Collision Rates for the Study Intersections
Study Intersection Number of
Collisions
(2015-2020)
Calculated
Collision Rate
(c/mve)
Statewide Average
Collision Rate
(c/mve)

1. SR 29/South Kelly Road 48 0.82 0.58

2. Devlin Road/South Kelly Road 2 0.32 0.55

3. Devlin Road/Green Island Road 1 - - -

4. Paoli Loop Road/Green Island Road 2 0.20 0.16
Note: c/mve = collisions per million vehicles entering 1 Collision rate was not evaluated at this future intersection Of the 48 reported collisions that occurred at the intersection of SR 29/South Kelly Road, 30 were rear-end collisions with the primary cause being unsafe speeds. This type of crash is common at signalized intersections where there is congestion, especially during peak periods. Because of the regional nature of the traffic that causes the congestion, and as this location is under the jurisdiction of Caltrans, there is little that the City can do to address this condition though increased enforcement and analysis of the potential to improve signal timing to reduce congestion could be requested of the Highway Patrol and Caltrans, respectively.

The collision rate at Paoli Loop Road/Green Island Road is higher than the statewide average, with one of the two reported collisions being a head-on collision and the other being a broadside collision. The broadside collision resulted from “traffic signal and sign” violations, and the head-on collision was attributed to unsafe speed. With no clear pattern and given the very low number of crashes, no remedial actions are apparent.

Alternative Modes
Pedestrian Facilities
Pedestrian facilities include sidewalks, crosswalks, pedestrian signal phases, curb ramps, curb extensions, and various streetscape amenities such as lighting, benches, etc. In general, sidewalk coverage is limited in the area surrounding the project site. Existing gaps and obstacles along the connecting roadways impact convenient and continuous access for pedestrians and present safety concerns in those locations where appropriate pedestrian infrastructure would address potential conflict points. Within the study area, new sidewalks are planned along the future extension of Devlin Road.

Bicycle Facilities
The Highway Design Manual, Caltrans, 2017, classifies bikeways into four categories:
• Class I Multi-Use Path – a completely separated right-of-way for the exclusive use of bicycles and pedestrians with cross flows of motorized traffic minimized.
• Class II Bike Lane – a striped and signed lane for one-way bike travel on a street or highway. • Class III Bike Route – signing only for shared use with motor vehicles within the same travel lane on a street or highway.
• Class IV Bikeway – also known as a separated bikeway, a Class IV Bikeway is for the exclusive use of bicycles and includes a separation between the bikeway and the motor vehicle traffic lane. The separation may include, but is not limited to, grade separation, flexible posts, inflexible physical barriers, or on-street parking.

In the project area, Class II bike lanes exist on Devlin Road between Middleton Way and South Kelly Road. Bicyclists ride in the roadway and/or on sidewalks along all other streets within the project study area. Table 2 summarizes 2369
6 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
the existing and planned bicycle facilities in the project vicinity as contained in the Napa Countywide Bicycle Plan, Napa Valley Transportation Authority, 2019.

Table 2 – Bicycle Facility Summary
Status
Facility
Class Length
(miles)
Begin Point End Point
Existing
Devlin Road II 0.09 Middleton Way South Kelly Road
Planned
S Kelly Rd I 0.20 SR 29 Devlin Rd
Vine Trail (along Devlin Rd) I 1.62 Middleton Way Watson Ln Green Island Rd I 0.33 Vine Trail Commerce Blvd
Green Island Rd II 0.25 Northern Intersection of Green
Island Rd and Mezzetta Ct
Vine Trail
Source: Napa Countywide Bicycle Plan, Napa Valley Transportation Authority, 2019 Transit Facilities
Valley Intercity Neighborhood Express (VINE) Transit provides fixed route bus service throughout Napa County.

American Canyon Transit (ACT) is an on-demand, door-to-door, transit service for persons with disabilities who cannot independently use regular fixed-route transit services. Neither VINE nor ACT maintains stops near the project site.

On-demand private taxi services are available in the study area 24 hours a day. Taxis can be used for trips within the local Planning Area and farther destinations, including nearby airports. Other ride-hailing applications are also available in study area and provide transportation throughout the Bay Area. 2370
7 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Capacity Analysis
Intersection Level of Service Methodologies
Level of Service (LOS) is used to rank traffic operation on various types of facilities based on traffic volumes and roadway capacity using a series of letter designations ranging from A to F. Generally, Level of Service A represents free flow conditions and Level of Service F represents forced flow or breakdown conditions. A unit of measure that indicates a level of delay generally accompanies the LOS designation.

The study intersections were analyzed using methodologies published in the Highway Capacity Manual (HCM), 6th Edition, Transportation Research Board, 2018. This source contains methodologies for various types of intersection control, all of which are related to a measurement of delay in average number of seconds per vehicle.

The Levels of Service for the intersections with side street stop controls, or those which are unsignalized and have one or two approaches stop controlled, were analyzed using the “Two-Way Stop-Controlled” intersection capacity method from the HCM. This methodology determines a level of service for each minor turning movement by estimating the level of average delay in seconds per vehicle. Results are presented for individual movements together with the weighted overall average delay for the intersection.

The study intersections with stop signs on all approaches were analyzed using the “All-Way Stop-Controlled” Intersection methodology from the HCM. This methodology evaluates delay for each approach based on turning movements, opposing, and conflicting traffic volumes, and the number of lanes. Average vehicle delay is computed for the intersection as a whole and then related to a Level of Service.

The study intersections that are currently controlled by a traffic signal, or may be in the future, were evaluated using the signalized methodology from the HCM. This methodology is based on factors including traffic volumes, green time for each movement, phasing, whether the signals are coordinated or not, truck traffic, and pedestrian activity. Average stopped delay per vehicle in seconds is used as the basis for evaluation in this LOS methodology.

For purposes of this study, delays were calculated using signal timing obtained from Caltrans.

The ranges of delay associated with the various levels of service are indicated in Table 3. 2371
8 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Table 3 – Intersection Level of Service Criteria
LOS Two-Way Stop-Controlled All-Way Stop-Controlled Signalized A Delay of 0 to 10 seconds. Gaps in
traffic are readily available for
drivers exiting the minor street.

Delay of 0 to 10 seconds. Upon
stopping, drivers are immediately
able to proceed.

Delay of 0 to 10 seconds. Most
vehicles arrive during the green
phase, so do not stop at all.

B Delay of 10 to 15 seconds. Gaps
in traffic are somewhat less
readily available than with LOS A,
but no queuing occurs on the
minor street.

Delay of 10 to 15 seconds. Drivers
may wait for one or two vehicles to
clear the intersection before
proceeding from a stop.

Delay of 10 to 20 seconds. More
vehicles stop than with LOS A, but
many drivers still do not have to
stop.

C Delay of 15 to 25 seconds.

Acceptable gaps in traffic are less
frequent, and drivers may
approach while another vehicle is
already waiting to exit the side
street.

Delay of 15 to 25 seconds. Drivers
will enter a queue of one or two
vehicles on the same approach and
wait for vehicle to clear from one or
more approaches prior to entering
the intersection.

Delay of 20 to 35 seconds. The
number of vehicles stopping is
significant, although many still
pass through without stopping.

D Delay of 25 to 35 seconds. There
are fewer acceptable gaps in
traffic, and drivers may enter a
queue of one or two vehicles on
the side street.

Delay of 25 to 35 seconds. Queues
of more than two vehicles are
encountered on one or more
approaches.

Delay of 35 to 55 seconds. The
influence of congestion is
noticeable, and most vehicles have
to stop.

E Delay of 35 to 50 seconds. Few
acceptable gaps in traffic are
available, and longer queues may
form on the side street.

Delay of 35 to 50 seconds. Longer
queues are encountered on more
than one approach to the
intersection.

Delay of 55 to 80 seconds. Most, if
not all, vehicles must stop, and
drivers consider the delay
excessive.

F Delay of more than 50 seconds.

Drivers may wait for long periods
before there is an acceptable gap
in traffic for exiting the side
streets, creating long queues.

Delay of more than 50 seconds.

Drivers enter long queues on all
approaches.

Delay of more than 80 seconds.

Vehicles may wait through more
than one cycle to clear the
intersection.

Reference: Highway Capacity Manual, 6th Edition, Transportation Research Board, 2018 Traffic Operation Standards
City of American Canyon
The City of American Canyon General Plan Circulation Element specifies minimum level-of-service standards for all streets and intersections in the City’s jurisdiction. In section 4.1.6, the City establishes the following performance standards for acceptable LOS:

Achieve and maintain a Multimodal LOS D or better for roadways and intersections during peak hours where possible for as long as possible. However, recognizing that LOS D may not be achievable or cannot be maintained upon full buildout of the General Plan, due to traffic generated from sources beyond control of the City, the City Council shall have the discretion to only require feasible mitigation measures that may not achieve LOS D, but will reduce the impact of any development use or VMT planned for in the Land Use Element of the General Plan.

The locations that may not achieve or maintain LOS D are as follows and will be exempt from the LOS D policy:
2372
9 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
• State Route 29 through the City
• American Canyon Road from SR 29 to Flosden Road-Newell Drive • Flosden Road south of American Canyon Road
Caltrans
The study intersection of SR 29/South Kelly Road is located on the State Transportation Network (STN) and is therefore under jurisdiction of Caltrans. It is noted that Caltrans does not currently have a standard of significance relative to intersection operation as this is no longer a CEQA issue. The new vehicle miles traveled-focused Transportation Impact Study Guide (TISG), published in May 2020, replaced the Guide for the Preparation of Traffic Impact Studies, 2002. As indicated in the TISG, the Department is transitioning away from requesting LOS or other vehicle operations analyses of land use projects and will instead focus on Vehicle Miles Traveled (VMT).

Existing Conditions
The Existing Conditions scenario provides an evaluation of current operation based on existing traffic volumes during the a.m. and p.m. peak hours. This condition does not include project-generated traffic volumes. Volume data was collected on various dates in February, May and November 2019 while local schools were in session and prior to the outbreak of COVID-19 and the resulting change in travel patterns. It is noted that travel patterns within American Canyon vary between the a.m. and p.m. peak hours and individual drivers may use different routes for their morning and evening commutes. Copies of the traffic count data sheets are provided in Appendix B.

Intersection Levels of Service
Under existing conditions, all study intersections are operating acceptably during both peak hours evaluated. The existing traffic volumes are shown in Figure 2. A summary of the intersection Level of Service calculations is contained in Table 4, and copies of the calculations are provided in Appendix C.

Table 4 – Existing Peak Hour Intersection Levels of Service Study Intersection
Approach
AM Peak PM Peak
Delay LOS Delay LOS

1. SR 29/S Kelly Rd 34.6 C 15.9 B

2. Devlin Rd/S Kelly Rd 7.8 A 8.0 A

3. Devlin Rd/Green Island Rd 1 - - - -
Southbound (Devlin Rd) Approach - - - -

4. Paoli Loop Rd/Green Island Rd 4.9 A 3.1 A
Northbound (Paoli Loop) Approach 13.3 B 13.5 B
Notes: Delay is measured in average seconds per vehicle; LOS = Level of Service; Results for minor approaches to two-way stop-controlled intersections are indicated in italics 1 Level of Service was not evaluated at this future intersection Future Conditions
Future (2040) a.m. and p.m. peak hour volume projections for SR 29/South Kelly Road were derived from a buildout analysis which is contained in the Napa Junction III Transportation Impact Analysis Report, Omni-Means, LTD, 2373
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 2 – Existing Traffic Volumes
(xx) PM Peak Hour Volume
xx AM Peak Hour Volume
Study Intersection
LEGEND
100(10)
0 (1)
26 (4)
(0) 1
(13)44
(1) 0
2 350(106)
(374)114
3 Future Intersection
193(42)
33 (42)
(155)59
(219)55
4 24(5)
16(1)
88(95)
(57)19
(18)23
(45)17
1 2374
11 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
2011; this scenario represents cumulative traffic conditions that would be expected upon build out of the land uses identified in the General Plan. Although some of the anticipated development included in this previous effort may already be complete and occupied, to provide a conservative estimate of future operation the incremental increase in trips associated with build out of the City of American Canyon under its current General Plan was added to current volumes to determine Future (year 2040) operating conditions without the project. A growth rate was then created for the intersection of SR 29/South Kelly Road based on the volumes for this location in the Napa Junction III Transportation Impact Analysis Report.

Because the west side of SR 29 is already nearly built out other than the project site, so would be expected to experience considerably less of an increase in traffic compared to SR 29, a growth rate of one-percent per year was used to develop future a.m. and p.m. peak hour volumes for the study intersections located west of SR 29.

Under the anticipated Future volumes, the study intersections are expected to operate acceptably during the a.m. and p.m. peak hours except for SR 29/South Kelly Road, which is expected to operate at LOS F during both peak hours. Future volumes are shown in Figure 3 and operating conditions are summarized in Table 5.

Table 5 – Future Peak Hour Intersection Levels of Service Study Intersection
Approach
AM Peak PM Peak
Delay LOS Delay LOS

1. SR 29/S Kelly Rd 107.4 F 84.3 F

2. Devlin Rd/S Kelly Rd 8.0 A 8.0 A

3. Devlin Rd/Green Island Rd 1 - - - -
Southbound (Devlin Rd) Approach - - - -

4. Paoli Loop Rd/Green Island Rd 5.2 A 3.1 A
Northbound (Paoli Loop) Approach 14.3 B 13.4 B
Notes: Delay is measured in average seconds per vehicle; LOS = Level of Service; Results for minor approaches to two-way stop-controlled intersections are indicated in italics 1 Level of Service was not evaluated at this future intersection While the intersection of SR 29/South Kelly Road is projected to operate at LOS F during each peak hour, this operation was considered acceptable since SR 29 is exempt from the City’s LOS standard and Caltrans no longer applies an LOS standard. The City of American Canyon – Broadway District Specific Plan Draft EIR, First Carbon Solutions, 2017, states that there are future plans for SR 29 to have three through lanes in each direction through American Canyon. Because funding has not been identified for this capacity enhancement, it was conservatively assumed that the current configuration with only two through lanes in each direction would remain, and this configuration was used for the analysis.

Furthermore, the future year analysis results for facilities along SR 29 (specifically, the intersection of SR 29/South Kelly Road) is deemed to be more conservative than what was presented for this intersection in the Watson Ranch Specific Plan Environmental Impact Report (September 2018) which indicated LOS C and D operation for the a.m. and p.m. peak hours, respectively.

Project Description
The project includes the construction of 2.2 to 2.4 million square feet of warehouse which would be built concurrently with the planned extension of Devlin Road between Green Island Road and South Kelly Road in 2375
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 3 – Future Traffic Volumes
(xx) PM Peak Hour Volume
xx AM Peak Hour Volume
Study Intersection
LEGEND
234(51)
40 (51)
(188)72
(265)67
4 424(129)
(453)138
3 Future Intersection
121(13)
0 (2)
32 (5)
(0) 2
(16)54
(2) 0
2 37 (8)
25 (2)
137(161)
(96)30
(30)36
(76)27
1 2376
13 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
American Canyon. It should be noted that the extension of Devlin Road is identified in the City’s General Plan as a circulation improvement and its completion would have been pursued by the City with or without the involvement of this project. The project would be completed in two phases. Phase 1 includes the construction of the Devlin Road extension and two warehouse buildings along with supporting surface parking lots and circulating roads on the east side of the Devlin Road extension. A supply of 858 parking spaces is presently planned for Phase 1. Phase 2 would consist of three or four warehouse buildings and supporting parking lots and circulating roads to the west of the Devlin Road extension.

The extension of Devlin Road would be approximately 3,000 feet long and include bicycle lanes and a sidewalk along its eastern edge. The project also includes the construction of a multi-modal path along the northern side of Green Island Road at the project frontage.

The proposed project site plan for Phase 1 is shown in Figure 4. A site plan for Phase 2 is currently being developed and was not available for use in this study.

Trip Generation
The anticipated trip generation for the proposed project was estimated using standard rates published by the Institute of Transportation Engineers (ITE) in Trip Generation Manual, 10th Edition, 2017 for a “High-Cube Transload and Short-Term Storage Warehouse” (Land Use #154). The project would be comprised of multiple warehouse buildings with a combined size of between 2.2 and 2.4 million square feet. To be conservative, the maximum size of 2.4 million square feet was used to estimate the trip generation. The project is not anticipated to generate any internal capture trips, pass-by trip credits or any other trip reductions. The number of truck trips associated with a High-Cube Warehouse was estimated using rates published in the Trip Generation Manual and validated using local vehicle classification counts conducted in June 2021. Copies of the counts and truck percentage derivation are provided in Appendix B.

As shown in Table 6, the project is expected to generate an average of 3,888 net-new passenger car equivalent trips per day, including 240 trips during the a.m. peak hour and 264 during the p.m. peak hour. These new trips represent the increase in traffic associated with the project compared to existing volumes. To account for the effect of heavy vehicles (such as tractor trucks), a heavy vehicle adjustment factor was applied to convert truck trips to an equivalent passenger car trip total. The passenger car equivalency factor (PCE) for heavy vehicles is assumed to be 2.0 (i.e., each tractor truck has the effect of two passenger cars due to longer start up times at intersections and when making turns). Thus, the number of truck trips per hour was multiplied by 2.0 to determine the equivalent passenger car trips per hour.

Table 6 – Trip Generation Summary
Land Use Units Daily AM Peak Hour PM Peak Hour
Rate Trips Rate Trips In Out Rate Trips In Out
High-Cube Warehouse 2,400 ksf
Vehicles (Trucks + Pass. Cars) 1.40 3,360 0.08 192 148 44 0.10 240 67 173 Trucks 0.22 528 0.02 48 37 11 0.01 24 7 17
Pass. Cars - 2,832 - 144 111 33 - 216 60 156
Trucks (pass. car equivalents) - 1,056 - 96 74 22 - 48 14 34 Total Passenger Car Trips 3,888 240 185 55 264 74 190
Note: ksf = 1,000 square feet, Total Passenger Car Trips = Passenger Cars plus Passenger Car Equivalent Units of Trucks 2377
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 4 – Site Plan
Source: First Carbon Solutions 2/2
2378
15 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Trip Distribution
The pattern used to allocate new project trips to the street network was determined by reviewing likely routes for employees, visitors, and deliveries. The directionality experienced on SR 29 during the morning and evening commute periods was considered in developing the proposed assumptions. Based on the assumptions shown in Table 7, the following distribution was applied. Project traffic volumes are shown in Figure 5.

Intersection Operation
Existing plus Project Conditions
Upon the addition of project-related traffic to the Existing volumes, all study intersections are expected to continue operating at LOS C or better. These results are summarized in Table 8. Existing plus Project traffic volumes are shown in Figure 6.

Table 8 – Existing and Existing plus Project Peak Hour Intersection Levels of Service Study Intersection
Approach
Existing Conditions Existing plus Project
AM Peak PM Peak AM Peak PM Peak
Delay LOS Delay LOS Delay LOS Delay LOS

1. SR 29/S Kelly Rd 34.6 C 15.9 B 34.7 C 16.7 B

2. Devlin Rd/S Kelly Rd 7.8 A 8.0 A 8.4 A 7.9 A

3. Devlin Rd/Green Island Rd - - - - 0.5 A 1.4 A
Southbound (Devlin Rd) Approach - - - - 12.1 B 12.7 B

4. Paoli Loop Rd/Green Island Rd 4.9 A 3.1 A 7.5 A 4.0 A Northbound (Paoli Loop) Approach 13.3 B 13.5 B 17.9 C 18.1 C Notes: Delay is measured in average seconds per vehicle; LOS = Level of Service; Results for minor approaches to two-way stop-controlled intersections are indicated in italics
Finding – The study intersections are anticipated to operate at LOS C or better with project-generated traffic added.

Future plus Project Conditions
Upon the addition of project-generated traffic to the anticipated Future volumes, all unsignalized study intersections are expected to operate acceptably. SR 29/South Kelly Road would continue to operate at LOS F during both peak hours. Future plus Project traffic volumes are shown in Figure 7 and operating conditions are summarized in Table 9.

Table 7 – Trip Distribution Assumptions
Route AM PM
Percent Trips Percent Trips
To/from North on SR 29 50% 120 55% 145
To/from South on SR 29 50% 120 45% 119
TOTAL 100% 240 100% 264
2379
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 5 – Project Traffic Volumes
(xx) PM Peak Hour Volume
xx AM Peak Hour Volume
Study Intersection
LEGEND
23(11)
(27) 7
(86)28
4 71(26)
3 Future Intersection
71(31)
2 (78)21
1 2380
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 6 – Existing Plus Project Traffic Volumes
(xx) PM Peak Hour Volume
xx AM Peak Hour Volume
Study Intersection
LEGEND
100(10)
0 (1)
97 (35)
(0) 1
(13)44
(1) 0
2 71 (26)
350(106)
(374)114
3 Future Intersection
216(53)
33 (42)
(182)66
(305)83
4 24(5)
16(1)
88(95)
(57)19
(18)23
(123)38
1 2381
Traffic Study for the Giovannoni Logistics Center
aca033.ai 4/21
Not to Scale
29 Project Site
1 2
3 4
Figure 7 – Future Plus Project Traffic Volumes
(xx) PM Peak Hour Volume
xx AM Peak Hour Volume
Study Intersection
LEGEND
121(13)
0 (2)
103(36)
(0) 2
(16)54
(2) 0
2 71 (26)
424(129)
(453)138
3 Future Intersection
257(62)
40 (51)
(215)79
(351)95
4 37 (8)
25 (2)
137(161)
(96)30
(30)36
(154)48
1 2382
19 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Table 9 – Future and Future plus Project Peak Hour Intersection Levels of Service Study Intersection
Approach
Future Conditions Future plus Project
AM Peak PM Peak AM Peak PM Peak
Delay LOS Delay LOS Delay LOS Delay LOS

1. SR 29/S Kelly Rd 107.4 F 84.3 F 107.0 F 88.7 F

2. Devlin Rd/S Kelly Rd 8.0 A 8.0 A 8.5 A 7.6 A

3. Devlin Rd/Green Island Rd - - - - 0.4 A 1.4 A
Southbound (Devlin Rd) Approach - - - - 13.1 B 13.9 B

4. Paoli Loop Rd/Green Island Rd 5.2 A 3.1 A 7.8 A 3.7 A Northbound (Paoli Loop) Approach 14.3 B 13.4 B 19.3 C 16.7 C Notes: Delay is measured in average seconds per vehicle; LOS = Level of Service; Results for minor approaches to two-way stop-controlled intersections are indicated in italics
As mentioned previously, there are future plans for SR 29 to have three travel lanes in each direction through American Canyon. As required by the City in their Traffic Impact Fee Program, the project should pay a proportional share fee toward the cost of this planned future infrastructure improvement. As specific building projects move forward, each should contribute to the funds needed for the planned improvements to SR 29 based on the City’s fee schedule.

Finding – The study intersections will continue operating acceptably with project traffic added, with the exception of SR 29/South Kelly Road, which would operate at LOS F with or without the addition of project traffic, though the project’s effect would be considered acceptable since SR 29 is exempt from the City’s LOS standard.

Recommendation – As required by the City in their Traffic Impact Fee Program, the project should pay a fee toward the cost of the planned future improvements to SR 29. 2383
20 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Vehicle Miles Traveled
Senate Bill (SB) 743 established a change in the metric to be applied for determining traffic impacts associated with development projects. Rather than the delay-based criteria associated with a Level of Service analysis, the increase in Vehicle Miles Traveled (VMT) because of a project will be the basis for determining impacts. The City of American Canyon has not yet established thresholds of significance related to VMT. The Napa County travel demand model is not currently available for use as a source for VMT analysis. In lieu of an established local methodology, the project-related VMT impacts were quantitatively assessed based on guidance provided by the California Governor’s Office of Planning and Research (OPR) in the publication Transportation Impacts (SB 743) CEQA Guidelines Update and Technical Advisory, 2018 and approved by City of American Canyon staff.

Based on a review of established policies currently used by OPR, Sacramento County and the City of San Jose, a VMT impact would be identified at an industrial project if the project VMT per employee is higher than the regional average VMT per employee. For this study, the regional average VMT is defined as the nine county Bay Area average. As reported by the City of Vallejo in their CEQA Transportation Impact Analysis Guidelines (dated October 2020), the Nine-County Bay Area Average VMT per employee is 23.00 miles per employee. According to Statewide Travel Demand Model estimates, this project is located within a Traffic Analysis Zone (TAZ) with a projected VMT per employee of 16.24 miles. Because this per capita VMT rate is lower than the significance threshold of 23.0 miles, the project would be considered to have a less-than-significant VMT impact. A summary of the VMT findings is provided in Table 10.

Table 10 – Vehicle Miles Traveled Analysis Summary
VMT Metric Regional Average/
Significance Threshold
Project
VMT Rate
Resulting
Significance
Employment VMT per Capita 23.00 16.24 Less-Than-Significant Note: VMT Rate is measured in VMT/Capita, or the number of daily miles driven per employee Finding – The project would be expected to have a less-than-significant transportation impact on vehicle miles traveled.

Transportation Demand Management (TDM) Measures
The following section describes a potential Transportation Demand Management (TDM) program for employees of the logistics center. While the project’s potential impact is expected to be less than significant, implementation of a TDM program would potentially lower the vehicle miles traveled associated with this project and thus have a regional benefit.

According to SB 1128, adopted in 2016, all employers with 50 or more full-time employees within the nine-county Bay Area are subject to SB 1128, which requires employers to offer their employees one of four options – pre-tax benefits, transit or vanpool subsidies, employer-provided transit, or an alternative benefit of equal effectiveness.

The following TDM program would fulfill this legislative requirement.

Ridesharing Program
Carpooling is one of the most common and cost-effective alternative modes of transportation and one that commuters can adopt part-time. Given the delivery station’s auto-oriented location, it can also be one of the most effective alternatives to driving alone. There are numerous benefits to ridesharing. Carpooling can reduce peakperiod vehicle trips and increase commuters’ travel choices. Further, it reduces congestion, road and parking facility costs and pollution emissions. Carpooling tends to have the lowest cost per passenger-mile of any 2384
21 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
motorized mode of transportation since it makes use of a vehicle seat that would otherwise be empty. Carpooling also provides financial savings for the consumer by decreasing fuel and parking costs. Financial incentives can be an effective way to encourage employees to carpool. The tenant should provide a monetary incentive to employees who agree to carpool to work a minimum of 50 percent of the time. This program should be offered to all employees of the project.

Ridematching
The greatest barrier to workplace carpooling is often simply being able to identify and travel with other employees who live nearby. There are many services that can assist in pairing employees within the same organization or across organizations. The most basic publicly available service is 511.org’s free ridematching service, Merge. There are also various private ridematching providers (e.g. Zimride, RideAmigos, Via, Scoop) that can effectively create carpool networks while making them safe and convenient for their users. Information on a variety of programs is offered through 511.org.

Recommendation: The tenant should market the Merge program to employees and assist on an as-needed basis with ridematching questions or concerns. A monetary incentive may also be offered to employees who carpool to and from work.

Vanpooling
Employer-provided vans can provide several advantages. Vans are defined as vehicles able to carry at least six adults and can reduce the cost of commuting for employees by removing the need for workers to put mileage on their own vehicles and, depending on the level of subsidy, they may not need to pay for gas. For tax purposes, employers may be able to deduct the costs of vans as a “qualified transportation fringe benefit.” Recommendation: The tenant should market the vanpool program to all employees. Identify which employees would be the most suitable to participate in vanpools based on their home locations and reach out directly to those employees to determine if vanpools of at least six persons each are feasible. If so, provide one or more fully subsidized vans for those employees with the condition that they must use it for their commute.

Priority Parking
An easy way of promoting ridesharing is to designate priority employee parking spaces for carpools and vanpools.

The cost of this improvement is limited to the paint or sign used and can help market ridesharing as a priority in the organization.

Recommendation: Up to five parking spaces for rideshare vehicles should be designated on the plans for the project.

Electric Vehicle Charging Stations
While the provision of dedicated parking for Electric Vehicles (EVs) and charging stations do not result in trip or VMT reductions, they can help reduce the total GHG emissions of a project.

Recommendation: Electric vehicle parking stalls and charging stations consistent with California Green (Cal Green) building standards should be provided for each building.

Guaranteed Ride Home (GRH)
Guaranteed Ride Home (GRH) is a program that provides a “back-up” ride to employees who use transit, carpool, biking/walking, or other alternatives as their commute mode; in Napa County, it is provided by the NVTA via their V-Commute Program. If an employee who carpools to work, so does not have their own vehicle, needs to leave work for an emergency, such as a sick child or other unexpected need, they will be redeemed for up to four GRH 2385
22 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
trips per calendar year. This is an important supportive measure to encourage employees to not drive alone to work and often goes as a welcome, but unused benefit.

Recommendation: The tenantshould market the GRH program to employees and assist with sign-ups as needed.

Bicycle Benefits
Bicycle Parking and Maintenance Tools
The provision of both short-term and long-term bicycle parking is important. Secure long-term parking (e.g. bike lockers) is a critical component in encouraging employees to bike to work as the lack of secure parking is often cited by employees as a deterrent. Short-term parking (e.g. bike racks) can be used by employees or visitors.

In addition to providing bicycle parking, some businesses are now encouraging bicycle use by providing employees with the basic tools necessary to maintain their bicycles and on-site shower and clothes locker facilities.

Often, these tools can be kept in bicycle storage areas and include simple items such as a bike pump and tire patches that are essential for bike travel, but also inexpensive.

Financial incentives can also be an effective way to encourage employees to ride a bicycle to reach the site. In addition to those who carpool, the applicant should provide a monetary incentive to employees who agree to walk or bicycle to work a minimum of 50 percent of the time.

Recommendation: The tenant should provide both short-term and long-term bicycle parking as well as basic bicycle maintenance tools and on-site shower facilities. Offer a monetary incentive to employees who walk or ride a bicycle to work.

Education, Outreach & Marketing
Welcome Packet for New Employees
New employees should be provided with a welcome packet containing relevant transportation information. The packet should include information on all available TDM measures as well as information from Vine Transit, the Napa County Bicycle Coalition and material regarding 511.org ride-matching services.

Monitor Performance
It is important to continually monitor the performance of a TDM program and adjust measures as necessary to ensure its success. The tenant should conduct mode split and vehicle miles traveled (VMT) surveys each year both to determine if adjustments are needed and to use as marketing material. Employee satisfaction surveys are also an effective way of ensuring a quality TDM program.

Transportation Coordinator
The presence of a staff person dedicated part-time to overseeing and managing the TDM program would be helpful in ensuring the ongoing success of these programs. This would not be a distinct position, but instead would be a role that is integrated into the duties of an on-site manager. The duties can include:
• Create and distribute resident transportation information welcome packets • Maintain and update the transportation information kiosk board • Distribute Napa Bicycle Coalition maps
• Monitor bicycle parking to ensure that there is sufficient space • Assist in selling or distributing transit passes
• Market the TDM program and transit service
• Advertise the unbundled parking program
2386
23 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
• Promote 511.org’s ride-matching program
Recommendation: An on-site manager for each company occupying the buildings or a representative of the property management organization should be designated as the Transportation Coordinator and tasked with providing appropriate information to employees and both monitoring and tailoring the TDM program as necessary.

Summary of TDM Reductions
The expected VMT reductions associated with the various TDM measures were estimated based on information published in the California Air Pollution Officers Association (CAPCOA) report Quantifying Greenhouse Gas Mitigation Measures, CAPCOA, 2010, the location of the project site, and knowledge of transportation characteristics of the area. Packaged together as a required commute trip reduction program, the TDM strategies listed above are projected to result in a VMT reduction potential of 12.0 percent. The estimated VMT reduction calculations are summarized in Table 11.

Table 11 – Estimated Employee VMT Reduction
TDM Measure VMT Reduction (%)
Project Estimate
Ridesharing Program 8.5
Education, Outreach, and Marketing 4.0
Bicycle Benefits Supportive
Emergency Ride Home Supportive
Total Potential VMT Reduction 1 12.0
Notes: VMT = vehicle miles travelled; TDM = transportation demand management 1 The reductions shown consider the relative effectiveness of reductions when individual TDM strategies are paired with each other, which is not the sum of the individual measures.

Finding – Successful implementation of the TDM measures identified in this report, such as use of a ridesharing program, offering secure bicycle parking, and providing educational materials strategies, have the potential to reduce project-related travel by 12.0 percent.

2387
24 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Alternative Modes
Pedestrian Facilities
Given the nature of the study area and the surrounding industrial land use, it is reasonable to assume that very few project patrons and employees will desire to walk to reach the project site. There may, however, be a desire by employees to walk in the area for recreational purposes during break times or to reach nearby buildings. Upon completion of the project, sidewalks will be provided along Devlin Road between Green Island Road to South Kelly Road. The project also includes the construction of a multi-modal path along Green Island Road which would be available for use by pedestrians.

Finding – Pedestrian facilities serving the project site are adequate and would be further improved upon completion of the sidewalks along Delvin Road and the multi-modal path along Green Island Road.

Bicycle Facilities
Existing bicycle facilities, including Class II bike lanes on Devlin Road between Middleton Way and South Kelly Road together with shared use of minor streets provide adequate access for bicyclists. The project would include bicycle lanes along the Devlin Road extension and a multi-modal path along the north side of Green Island Road.

The planned Class I and II bicycle facilities on South Kelly Road and Green Island Road, as well as the Napa Valley Vine Trail along Devlin Road, would improve bicycle connectivity near the project site.

Finding – Bicycle facilities serving the project site are adequate and would be further improved upon completion of the planned bicycle facilities in the study area.

Bicycle Storage
The project site plan does not identify the provision of bicycle parking or storage facilities.

Finding – The adequacy of the bicycle parking facilities at the site are deemed to be inconclusive since the current site plan does not include a sufficient level of detail.

Recommendation – Bicycle parking serving the project site should be designed to comply with City of American Canyon municipal code requirements and to support the TDM Plan for the project.

Transit
Existing stops are not within an acceptable walking distance of the site. Should an employee need to use transit, they could ride a bicycle along Devlin Road to the nearest transit stop at the intersection of Airport Boulevard/Devlin Road.

Finding – The lack of existing transit service within an acceptable walking distance of the project site is typical for such remote locations and is therefore considered acceptable, though employees could use a bicycle to reach nearby transit stops.
2388
25 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Access and Circulation
Site Access
Vehicular access to Phase 1 would be provided via four driveways on Green Island Road and four driveways on the future extension of Devlin Road. Access to Phase 2 is anticipated to be provided via numerous full access driveways with connections to the Devlin Road extension. All driveways and internal roadways would be designed to current City standards to accommodate heavy vehicles and so can be expected to accommodate the access requirements for both emergency and passenger vehicles.

Sight Distance
A substantially clear line of sight should be maintained between the driver of a vehicle waiting at a driveway and the driver of an approaching vehicle. Sight distances along Green Island Road from the project driveways were evaluated based on sight distance criteria contained in the Highway Design Manual published by Caltrans. The recommended sight distance for driveway approaches is based on stopping sight distance using the approach travel speed as the basis for determining the recommended sight distance. Based on the posted speed limit of 40 mph, the minimum stopping sight distance required is 300 feet; a review in the field shows that sight distances at the project driveway locations on Green Island Road would be adequate, provided any vegetation or buildings are sited to ensure maintenance of adequate sight lines. The sight lines for driveways on the Devlin Road connection are expected to be adequate based on a review of the site plans.

Finding – Adequate sight distance would be available at each driveway to accommodate all turnsleaving the site.

To maintain minimum sight distance standards, it is recommended that vegetation along Green Island Road and the Devlin Road extension be trimmed and maintained by the property management firm. 2389
26 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Parking
The City of American Canyon Municipal Code, Chapter 19.21.030; Vehicle parking requirements, states that warehousing and storage projects are required to provide one parking space per 1,000 square feet of gross floor area for the first 20,000 square feet of space, plus one parking space per each 2,000 square feet of gross floor area thereafter. Phase 1 of the project includes two buildings totaling 1,069,904 square feet of warehouse space with 858 parking spaces (428 at Building A and 430 at Building B). The site plan for Phase 2 is currently under development and the size and arrangement of the buildings as well as the specific number of parking spaces has not yet been established. Assuming a conservative total project size of 2,400,000 square feet, Phase 2 would be comprised of approximately 1,330,100 square feet of warehouse. Based on these parameters, the project would be required to provide a total of 1,230 parking spaces (555 with Phase 1 and 675 with Phase 2). Phase 1 of the project as proposed would provide 858 parking spaces, which is greater than the City requirement for that portion of the project and represents a 303-parking space surplus. A summary of the City parking requirements and proposed supply is provided in Table 12.

Table 12 – Parking Analysis Summary
Land Use Units Supply
(Parking
Spaces)
City Code Requirements
(ksf) Rate Parking Spaces
Required
Warehousing and Storage (Phase 1
– Building A) 601.4 428 1 per 1,000 square feet for
the first 20,000 square feet,
plus 1 per 2,000 square feet
thereafter
311
Warehousing and Storage (Phase 1
– Building B) 468.5 430 244
Warehousing and Storage (Phase 2) 1,330.1 TBD 675
Total 2,400.0 TBD 1,230
Notes: ksf = 1,000 square feet; TBD = To Be Determined; Finding – Based upon City requirements and a maximum project size of 2.4 million square feet, the project would be required to provide 1,230 parking spaces.
2390
27 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
CEQA Checklist
This section details discussion and results of the analysis of issues identified in the CEQA checklist for potential transportation/traffic impacts. The results are contained in Table 13 below and a discussion of each criterion follows.

Table 13 – XVII. TRANSPORTATION/TRAFFIC
Would the Project: Potentially
Significant
Impact
Less than
Significant
Impact with
Mitigation
Less than
Significant
Impact
No
Impact
a) Conflict with a program, plan, ordinance, or policy
addressing the circulation system, including transit,
roadway, bicycle, and pedestrian facilities?
X
b) Conflict or be inconsistent with CEQA Guidelines §
15064.3, subdivision (b)? X
c) Substantially increase hazards due to a geometric
design feature (e.g., sharp curves or dangerous
intersections) or incompatible uses (e.g., farm
equipment)?
X
d) Result in inadequate emergency access? X
Discussion of Results
a. Would the Project conflict with a program, plan, ordinance, or policy addressing the circulation system, including transit, roadway, bicycle, and pedestrian facilities?
According to the City of American Canyon General Plan, numerous objectives, policies, and programs have been established supporting the need for all modes of travel to be accommodated by the transportation system. This is demonstrated through the following General Plan Guiding Policies. • Policy 1.6 Achieve and maintain a Multimodal LOS D or better for roadways and intersections during peak hours where possible and as long as possible. The locations that may not achieve or maintain LOS D are as follows and will be exempt from the LOS D policy: State Route 29 through the City, American Canyon Road from SR 29 to Flosden Road-Newell Drive, and Flosden Road south of American Canyon Road. • Policy 1.9 Use of existing facilities. Make efficient use of existing transportation facilities and improve these facilities as necessary in accordance with the Circulation Map. • Policy 1.11 Reduce Vehicle Miles Traveled. Through layout of land uses, improved alternate modes, and provision of more direct routes, strive to reduce the total vehicle miles traveled by City residents. • Policy 2.1 Promote walking and bicycling. Promote walking and bike riding for transportation, • recreation, and improvement of public and environmental health. • Policy 3.1 Promote safe, efficient, and convenient public transportation. Promote the use of public transportation for daily trips, including to schools and workplaces, as well as other purposes. 2391
28 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Roadway Facilities
The proposed project would not cause any study intersections to which it applies to operate below the City’s standard of LOS D. The project’s roadway impacts would therefore be less-than-significant, and no mitigation measures are required. It should be noted that while SR 29/South Kelly Road would operate at LOS F under Future Conditions without or with the project, this intersection is exempt from the City’s LOS standard and Caltrans no longer has an operational standard.

Pedestrian Facilities
The proposed project would provide pedestrian connectivity infrastructure on-site as well as sidewalks along Devlin Road. With these improvements, pedestrian impacts would be reduced to less-than-significant.

Bicycle Facilities
Shared use of minor streets combined with the Class II facility on Devlin Road provides adequate access for bicyclists. Therefore, bicycle impacts would be less-than-significant.

Transit Facilities
The Valley Intercity Neighborhood Express (VINE) Transit provides fixed route bus service in American Canyon.

VINE does not currently serve any bus stops within an acceptable walking distance of the site. Should an employee wish to use transit, they could ride a bicycle along Devlin Road to the nearest transit stop north of the project at Airport Boulevard. The lack of transit service within an acceptable walking distance of the project is typical within this area and would therefore be considered a less-than-significant impact. b. Would the Project conflict or be inconsistent with CEQA Guidelines § 15064.3, subdivision (b)? CEQA Guidelines §15064.3, Subdivision (b) indicates that land use projects would have a significant impact if the project resulted in vehicle miles traveled (VMT) exceeding an applicable threshold of significance. It further notes that if existing models or methods are not available to estimate the VMT for the project being considered, a lead agency may analyze the project’s VMT qualitatively. The City of American Canyon currently has no thresholds of significance related to VMT, and as of the date of the analysis, a regional travel demand model for Napa County is under development but not yet available for VMT analysis. The project was assessed quantitatively using the Statewide Travel Demand Model and determined to have a less-than-significant VMT impact. c. Would the Project substantially increase hazards due to a geometric design feature (e.g., sharp curves or dangerous intersections) or incompatible uses (e.g., farm equipment)? As part of the project, a new segment of Devlin Road would be constructed resulting in a continuous section of roadway running between South Kelly Road and Green Island Road. The project itself as well as the new segment of roadway would be designed and constructed to meet currently applicable codes and requirements, so would not be expected to result in any increased hazards. Project driveways would be expected to have adequate sight lines. The project is therefore expected to have a less-than-significantimpact regarding geometric design features or incompatible uses.
d. Would the Project result in inadequate emergency access? The project site would have adequate emergency access and would improve emergency access and response times to the area near Devlin Road with the provision of a new street connection; therefore, the project would result in a less-than-significant and in fact beneficial impact related to emergency access. 2392
29 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Conclusions and Recommendations
Conclusions
• The proposed project is expected to generate an average of 3,888 passenger-car equivalent trips per day, including 240 trips during the weekday a.m. peak hour and 264 during the p.m. peak hour. • Under Existing Conditions all study intersections operate acceptably, and they are expected to continue doing so upon the addition of project-related traffic.
• Under anticipated Future volumes the unsignalized study intersections are expected to operate acceptably without or with the addition of project-related traffic. SR 29/South Kelly Road would operate at LOS F either without or with project traffic, which was considered acceptable since SR 29 is exempt from the City’s LOS policy.
• The proposed project would have a less-than-significant transportation impact on vehicle miles traveled.

Successful operation of the recommended Transportation Demand Management Plan is expected to further lower project-related VMT.
• Pedestrian access would be adequate upon completion of the sidewalks and trail that are part of the proposed project.
• The shared use of minor streets along with the planned facilities outlined in the Napa Countywide Bicycle Plan would provide adequate access for bicyclists, though the project site plan does not identify the provision of bicycle parking.
• The lack of existing transit service within an acceptable walking distance of the project site is typical for this location and is therefore considered acceptable, though employees could use a bicycle to reach nearby transit stops.
• The project’s driveways and internal roadways should be designed to current City standards and are therefore expected to accommodate the access requirements for both emergency and passenger vehicles. • Although sight lines at the project driveways are or would be adequate for the approach speeds, it is recommended that vegetation be trimmed and maintained along Green Island Road and the Devlin Road extension to maintain the minimum sight distance requirements. • Based upon City requirements, the project is required to provide 1,230 parking spaces.

Recommendations
• As required by the City in their Traffic Impact Fee Program, the project should pay a fee toward the cost of the planned future improvements to SR 29.
• Landscaping, buildings, and signage along the project frontage should be sited to ensure that adequate sight lines are maintained.
• The project should incorporate TDM measures such as a ridesharing program, offering secure bicycle parking, and educational materials to further reduce their carbon footprint. • Secure bicycle storage should be provided, if not already planned as part of on-site amenities. 2393
30 Traffic Impact Study for the Giovannoni Logistics Center July 22, 2021
Study Participants and References
Study Participants
Principal in Charge Dalene J. Whitlock, PE, PTOE
Traffic Engineer Kenny Jeong, PE
Assistant Engineer Kimberly Tellez
Graphics Cameron Wong
Editing/Formatting Hannah Yung-Boxdell, Cameron Wong
Quality Control Dalene J. Whitlock, PE, PTOE
References
2016 Collision Data on California State Highways, California Department of Transportation, 2018 American Canyon Municipal Code, Quality Code Publishing, 2017 City of American Canyon – Broadway District Specific Plan Draft EIR, First Carbon Solutions, 2017 City of American Canyon General Plan Circulation Element, City of American Canyon, 2018 City of San Jose Transportation Analysis Handbook, City of San Jose 2018 City of Vallejo CEQA Transportation Impact Analysis Guidelines, City of Vallejo, 2020 Transportation Analysis Guidelines, County of Sacramento, 2020 Highway Capacity Manual, 6th Edition, Transportation Research Board, 2018 Highway Design Manual, 6th Edition, California Department of Transportation, 2017 Napa Countywide Bicycle Plan, Napa Valley Transportation Authority, 2019 Napa Junction III Transportation Impact Analysis Report, Omni-Means, 2011 Watson Ranch Specific Plan Environmental Impact Report, September 2018 Quantifying Greenhouse Gas Mitigation Measures, California Air Pollution Control Officers Association (CAPCOA), 2010
Statewide Integrated Traffic Records System (SWITRS), California Highway Patrol, 2015-2020 Technical Advisory on Evaluating Transportation Impacts in CEQA, Governor's Office of Planning and Research, 2018
Trip Generation Manual, 10th Edition, Institute of Transportation Engineers, 2017 Vehicle Miles Traveled-Focused Transportation Impact Study Guide, California Department of Transportation, 2020 VINE Transit, http://www.ridethevine.com
ACA033
2394
A Traffic Impact Study for the Giovannoni Logistics Center July 2021
Appendix A
Collision Rate Calculations
2395
2396
Date of Count:

Number of Collisions: 48
Number of Injuries: 23
Number of Fatalities: 1
Average Daily Traffic (ADT): 31900
Start Date:

End Date:

Number of Years: 5
Intersection Type: Four-Legged
Control Type: Signals
Area: Rural
48 x
31,900 x x 5
Study Intersection 0.82 c/mve
Statewide Average* 0.58 c/mve
Notes
c/mve = collisions per million vehicles entering intersection * 2016 Collision Data on California State Highways, Caltrans Date of Count:

Number of Collisions: 2
Number of Injuries: 0
Number of Fatalities: 0
Average Daily Traffic (ADT): 3400
Start Date:

End Date:

Number of Years: 5
Intersection Type: Four-Legged
Control Type: 4 Way Stop
Area: Rural
2x
3,400 x x 5
Study Intersection 0.32 c/mve
Statewide Average* 0.55 c/mve
Notes
c/mve = collisions per million vehicles entering intersection * 2016 Collision Data on California State Highways, Caltrans Giovannoni Logistics Center TIS
Wednesday, May 29, 2019
Tuesday, November 05, 2019

33.2%
Intersection Collision Rate Worksheet
October 1, 2015
September 30, 2020
Intersection # SR 29 & South Kelly Road
Collision Rate = 1,000,000
Devlin Road & South Kelly Road

38.0%
ADT = average daily total vehicles entering intersection October 1, 2015
365
Intersection #
September 30, 2020
Number of Collisions x 1 Million Collision Rate =
1:

Collision Rate Injury Rate

0.0%
Collision Rate Fatality Rate
Collision Rate = 365
2:

Number of Collisions x 1 Million

1.0%
Collision Rate =
ADT x Days per Year x Number of Years

47.9%
1,000,000
Injury Rate
Fatality Rate

2.1%
ADT x Days per Year x Number of Years

0.0%
ADT = average daily total vehicles entering intersection

0.8%
ns
7/22/2021