National City City Council
May 16, 2023

94 SECTION 7. NCFD DATA ANALYSIS
This data analysis was prepared as a key component of the study of the National City Fire Department (NCFD), which provides fire protection service to the City of National City and surrounding communities. This analysis examines all calls for service between January 1, 2019, and December 31, 2020, as recorded in the regional computer-aided dispatch (CAD) system, with National Fire Incident Reporting System (NFIRS) data obtained from multiple sources . The analysis results are primarily presented for 2019; the results for 2020 are compared with those for the prior year in Attachment I.

This analysis is made up of five parts. The first part focuses on call types and dispatches. The second part explores the time spent and the workload of individual units. The third part presents an analysis of the busiest hours in the year studied. The fourth part provides a response time analysis of the studied agency’s units. The fifth and final part analyzes the total fire loss.

The NCFD is a multi-service fire department, primarily serving an area of approximately

9.1 square miles and 63,000 residents. It provides fire, rescue, and paramedic first responder emergency medical services (EMS) to the National City Fire District including the City of National City, Lower Sweetwater Fire Protection District, the Port of San Diego, and surrounding communities. The department operates out of three fire stations and utilizes two frontline engines, one fire truck, one squad unit, and one command unit (battalion chief) .

In 2019, the NCFD responded to 8,846 calls, of which 58 percent were EMS calls. The total combined workload (deployed time) for NCFD units was 3,105.6 hours. The average response time was 6.5 minutes. The 90th percentile response time was 8.7 minutes.

In 2020, the NCFD responded to 8,481 calls, of which 57 percent were EMS calls. The total combined workload (deployed time) for NCFD units was 3,895.8 hours. The average response time was 6.9 minutes. The 90th percentile response time was 9.3 minutes.

95 METHODOLOGY
In this report, CPSM analyzes calls and runs. A call is an emergency service request or incident. A run is a dispatch of a unit (i.e., a unit responding to a call). Thus, a call may include multiple runs.

We linked the CAD and NFIRS data sets. Then, we classified the calls in a series of steps. We first used the NFIRS incident type to identify canceled calls, motor vehicle accidents (MVA), and fire category call types. Calls identified by NFIRS as EMS calls along with any calls that lacked a matching NFIRS record were categorized using the CAD system’s incident descriptions. We describe the method in Attachment VII.

The analysis focuses on calls that involved a responding NCFD unit. We examine aid received by other fire departments within National City in Table 7-1 and provide greater detail in Attachment IV. We analyze American Medical Response’s (AMR) EMS calls within National City in a separate section.

We received records for a total of 23,415 calls in 2019 and 2020. We removed 3,150 calls that had no responding units. These calls were canceled, and their cancel reasons are summarized in Attachment VIII. We also removed 2,022 calls in National City where only AMR responded. In addition, we removed 21 calls outside National City that did not record a respondi ng NCFD unit.

Finally, we excluded one incident to which the NCFD’s administrative unit was the sole responder; however, the workload of administrative units is documented in Attachment II. The remaining 18,221 calls included in this analysis are summarized in Table 7-1.

The main analysis in the following sections focuses on the 8,846 calls in 2019 where NCFD responded inside and outside of its fire district (see the highlighted rows in Table 7-1). All calls outside NCFD’s fire district are identified as aid given. The detailed call types for aid given calls are presented in Attachment III. During the two years, NCFD received aid from other fire departments for incidents that occurred inside National City. This included 1,069 calls together with NCFD and 894 calls without a responding NCFD unit. Attachment IV provides further detail for aid received calls.

TABLE 7-1: Studied Calls by Location, Responding Agency, and Year Location Responding Agency 2019 2020 Total
Inside NCFD District
NCFD only 6,193 5,821 12,014
NCFD and FD agencies 499 570 1,069
NCFD Total 6,692 6,391 13,083
Other FD agencies only 452 442 894
Total 7,144 6,833 13,977
Outside NCFD District NCFD Total 2,154 2,090 4,244
Total 9,298 8,923 18,221
Observations:
■ Of all calls involving NCFD, 76 and 75 percent were inside the National City fire district in 2019 and 2020, respectively.
■ Of all calls within the National City fire district, outside agencies responded independently to 6 percent of calls in both years.

96 AGGREGATE CALL TOTALS AND RUNS
In 2019, NCFD responded to 8,846 calls, of which, 6,692 occurred inside and 2,154 occurred outside the National City fire district, respectively. During the year, there were 31 structure fire calls and 125 outside fire calls within the National City fire district.

Calls by Type
Table 7-2 shows the number of calls that NCFD responded to by call type, average calls per day, and the percentage of calls that fall into each call type category. Figures 7-1 and 7-2 show the percentage of calls that fall into each EMS (Figure 7-1) and fire (Figure 7-2) type category.

TABLE 7-2: Calls by Type
Call Type Total Calls Calls per
Day
Call
Percentage
Breathing difficulty 722 2.0 8.2
Cardiac and stroke 779 2.1 8.8
Fall and injury 999 2.7 11.3
Illness and other 1,344 3.7 15.2
MVA 407 1.1 4.6
Overdose and psychiatric 151 0.4 1.7
Seizure and unconsciousness 738 2.0 8.3
EMS Total 5,140 14.1 58.1
False alarm 318 0.9 3.6
Good intent 56 0.2 0.6
Hazard 48 0.1 0.5
Outside fire 125 0.3 1.4
Public service 121 0.3 1.4
Structure fire 31 0.1 0.4
Fire Total 699 1.9 7.9
Canceled 853 2.3 9.6
Aid given 2,154 5.9 24.3
Total 8,846 24.2 100.0

97 FIGURE 7-1: EMS Calls by Type

FIGURE 7-2: Fire Calls by Type

98 Observations:
■ In 2019, NCFD responded to an average of 24.2 calls per day, including 2.3 canceled (10 percent) and 5.9 (24 percent) mutual aid calls per day. ■ EMS calls for the year totaled 5,140 (58 percent of all calls), an average of 14.1 calls per day. □ Illness and other calls were the largest category of EMS calls at 15 percent of total calls (26 percent of EMS calls).
□ Motor vehicle accidents (MVA) made up 5 percent of total calls (8 percent of EMS calls). □ Cardiac and stroke calls made up 9 percent of total calls (15 percent of EMS calls). ■ Fire calls for the year totaled 699 (8 percent of all calls), or an average of 1.9 calls per day. □ False alarm calls made up 4 percent of total calls (45 percent of fire calls). □ Structure and outside fire calls combined made up 2 percent of total calls (22 percent of fire calls), or an average of 0.4 calls per day, or one call every 2.3 days.

99 Calls by Type and Duration
The following table shows the duration of calls by type using four duration categories: less than 30 minutes, 30 minutes to one hour, one to two hours, and two or more hours.

TABLE 7-3: Calls by Type and Duration
Call Type Less than
30 Minutes
30 Minutes
to One Hour
One to
Two Hours
Two or
More Hours Total
Breathing difficulty 633 84 5 0 722
Cardiac and stroke 662 81 34 2 779
Fall and injury 876 102 20 1 999
Illness and other 1,212 117 14 1 1,344
MVA 359 39 9 0 407
OD 136 13 2 0 151
Seizure and UNC 642 84 12 0 738
EMS Total 4,520 520 96 4 5,140
False alarm 289 25 4 0 318
Good intent 52 3 1 0 56
Hazard 34 12 2 0 48
Outside fire 84 29 10 2 125
Public service 95 18 5 3 121
Structure fire 18 5 4 4 31
Fire Total 572 92 26 9 699
Canceled 837 11 5 0 853
Aid given 1,883 210 44 17 2,154
Total 7,812 833 171 30 8,846
Observations:
■ A total of 5,040 EMS calls (98.1 percent) lasted less than one hour, 96 EMS calls (1.9 percent) lasted one to two hours, and 4 EMS calls (0.1 percent) lasted two or more hours. ■ A total of 664 fire calls (95.0 percent) lasted less than one hour, 26 fire calls (3.7 per cent) lasted one to two hours, and 9 fire calls (1.3 percent) lasted two or more hours. ■ A total of 113 outside fire calls (90.4 percent) lasted less than one hour, 10 outside fire calls (8.0 percent) lasted one to two hours, and two outside fire calls (1.6 percent) lasted two or more hours.
■ A total of 23 structure fire calls (74.2 percent) lasted less than one hour, four structure fire calls (12.9 percent) lasted one to two hours, and four structure fire calls (12.9 percent) lasted two or more hours.

100
Average Calls by Month and Hour of Day
Figure 7-3 shows the monthly variation in the average daily number of calls handled by NCFD in 2019. Similarly, Figure 7-4 illustrates the average number of calls received each hour of the day.

FIGURE 7-3: Calls per Day by Month

Observations:
■ EMS calls per day ranged from 12.7 in January 2019 to 15.8 in March 2019. ■ Fire calls per day ranged from 1.2 in April 2019 to 2.5 in September 2019. ■ Other calls per day ranged from 7.1 in July 2019 to 9.1 in January 2019. ■ Total calls per day ranged from 23.1 in July 2019 to 25.9 in March 2019.

101
FIGURE 7-4: Average Calls by Hour of Day

Observations:
■ Average EMS calls per hour ranged from 0.25 between 4:00 a.m. and 5:00 a.m. to

0.78 between 11:00 a.m. and noon.
■ Average fire calls per hour ranged from 0.04 between 5:00 a.m. and 6:00 a.m. to

0.13 between 3:00 p.m. and 4:00 p.m.
■ Average other calls per hour ranged from 0.13 between 3:00 a.m. and 4:00 a.m. to

0.55 between 10:00 a.m. and 11:00 a.m.
■ Average total calls per hour ranged from 0.44 between 4:00 a.m. and 5:00 a.m. to

1.41 between 3:00 p.m. and 4:00 p.m.

102
Units Arriving at Calls
In this section, we limit ourselves to calls where a unit from NCFD arrived. For this reason, there are fewer calls in Table 7-4 than in Table 7-2. For 2019, Table 7-4, along with Figure 7-5, detail the number of calls with one, two, three, and four or more NCFD units arriving at a call, broken down by call type.

TABLE 7-4: Calls by Call Type and Number of Arriving NCFD Units Call Type Number of Units Total
Calls One Two Three Four or More
Breathing difficulty 715 0 1 0 716
Cardiac and stroke 743 31 0 0 774
Fall and injury 975 3 0 0 978
Illness and other 1,298 22 1 1 1,322
MVA 319 61 12 1 393
Overdose and psychiatric 138 0 0 0 138
Seizure and unconsciousness 728 6 0 0 734
EMS Total 4,916 123 14 2 5,055
False alarm 233 63 4 5 305
Good intent 49 4 1 0 54
Hazard 41 2 3 2 48
Outside fire 92 14 9 9 124
Public service 111 6 1 2 120
Structure fire 6 5 2 18 31
Fire Total 532 94 20 36 682
Canceled 400 20 1 1 422
Aid given 1,513 53 23 14 1,603
Total 7,361 290 58 53 7,762
Percentage 94.8 3.7 0.7 0.7 100.0

103
FIGURE 7-5: Calls by Number of Arriving NCFD Units

Observations:

Overall
■ On average, 1.1 units arrived at all calls; for 94.8 percent of calls, only one unit arrived. ■ Overall, four or more units arrived at 0.7 percent of calls.

EMS
■ On average, 1.0 units arrived per EMS call.
■ For EMS calls, one unit arrived 97.3 percent of the time, two units arrived 2.4 percent of the time, three units arrived 0.3 percent of the time, and four units arrived less than 0.1 percent of the time.

Fire
■ On average, 1.4 units arrived per fire call.
■ For fire calls, one unit arrived 78.0 percent of the time, two units arrived 13.8 percent of the time, three units arrived 2.9 percent of the time, and four or more units arrived 5.3 percent of the time.
■ For outside fire calls, three or more units arrived 14.5 percent of the time. ■ For structure fire calls, three or more units arrived 64.5 percent of the time.

104
WORKLOAD: RUNS AND TOTAL TIME SPENT
The workload of NCFD’s units is measured in two ways: runs and deployed time. The deployed time of a run is measured from the time a unit is dispatched through the time the unit is cleared.

Because multiple units respond to some calls, there are more runs (10,239) than calls (8,846) and the average deployed time per run varies from the average duration per call.

Runs and Deployed Time – NCFD Units
Deployed time, also referred to as deployed hours, is the total deployment time of NCFD units deployed on all runs. Table 7-5 shows the total deployed time, both overall and broken down by type of run, for all non-administrative NCFD units in 2019. Table 7-6 and Figure 7-6 present the average deployed minutes by hour of day.

TABLE 7-5: Annual Runs and Deployed Time by Run Type
Run Type
Deployed
Minutes
per Run
Total
Annual
Hours
Percent
of Total
Hours
Deployed
Minutes
per Day
Total
Annual
Runs
Avg.

Runs
per
Day
Breathing difficulty 20.0 251.2 8.1 41.3 753 2.1
Cardiac and stroke 22.2 315.3 10.2 51.8 851 2.3
Fall and injury 20.1 349.7 11.3 57.5 1,046 2.9
Illness and other 17.9 433.4 14.0 71.2 1,451 4.0
MVA 17.2 160.3 5.2 26.4 558 1.5
OD 18.4 47.5 1.5 7.8 155 0.4
Seizure and UNC 20.5 267.1 8.6 43.9 782 2.1
EMS Total 19.6 1,824.5 58.8 299.9 5,596 15.3
False alarm 13.5 103.1 3.3 16.9 459 1.3
Good intent 15.8 17.9 0.6 2.9 68 0.2
Hazard 17.5 22.5 0.7 3.7 77 0.2
Outside fire 25.1 89.3 2.9 14.7 213 0.6
Public service 23.0 57.5 1.9 9.4 150 0.4
Structure fire 41.1 81.5 2.6 13.4 119 0.3
Fire Total 20.5 371.8 12.0 61.1 1,086 3.0
Canceled 7.0 123.6 4.0 20.3 1,060 2.9
Aid given 18.9 785.7 25.3 129.2 2,497 6.8
Other Total 15.3 909.3 29.3 149.5 3,557 9.7
Total 18.2 3,105.6 100.0 510.5 10,239 28.1
Note: OD=Overdose and psychiatric; UNC=Unconsciousness.

105
Observations:

Overall
■ The total deployed time for 2019 was 3,105.6 hours. The daily average was 8.5 hours for all NCFD units combined.
■ There were 10,239 runs, including 1,060 runs dispatched for canceled calls and 2,497 runs dispatched for aid given calls. The daily average was 28.1 runs.

EMS
■ EMS runs accounted for 59 percent of the total workload. ■ The average deployed time for EMS runs was 19.6 minutes. The deployed time for all EMS runs averaged 5.0 hours per day.

Fire
■ Fire runs accounted for 12 percent of the total workload. ■ The average deployed time for fire runs was 20.5 minutes. The deployed time for all fire runs averaged 1.0 minutes per day.
■ There were 332 runs for structure and outside fire calls combined, with a total workload of 170.8 hours. This accounted for 5 percent of the total workload. ■ The average deployed time for outside fire runs was 25.1 minutes per run, and the aver age deployed time for structure fire runs was 41.1 minutes per run.

106
TABLE 7-6: Deployed Minutes by Hour of Day
Hour EMS Fire Other Total
0 8.1 1.8 4.6 14.5
1 8.4 1.6 3.3 13.4
2 7.1 1.5 2.8 11.4
3 6.6 1.4 2.9 10.8
4 6.0 1.4 2.9 10.3
5 7.8 1.9 4.6 14.3
6 9.0 1.4 4.0 14.4
7 11.9 2.1 5.6 19.6
8 13.2 2.0 6.6 21.9
9 12.4 2.6 6.8 21.8
10 13.7 1.5 8.8 24.0
11 15.6 3.2 8.1 26.9
12 16.1 4.2 8.2 28.4
13 15.3 4.2 7.7 27.2
14 15.6 4.1 7.1 26.7
15 15.4 3.6 8.7 27.7
16 15.9 2.7 7.4 26.1
17 16.3 2.2 8.6 27.1
18 16.9 3.7 7.7 28.3
19 17.1 2.7 7.3 27.1
20 16.0 2.8 7.1 25.9
21 14.2 3.8 5.8 23.7
22 11.0 2.3 6.4 19.8
23 10.0 2.4 6.7 19.2
Daily
Avg. 299.9 61.1 149.5 510.5

107
FIGURE 7-6: Average Deployed Minutes by Hour of Day

Observations:
■ Hourly deployed time was highest during the day from 11:00 a.m. to 9:00 p.m., averaging more than 26 minutes per hour.
■ Average deployed time peaked between noon and 1:00 p.m., averaging 28.4 minutes. ■ Average deployed time was lowest between 4:00 a.m. and 5:00 a.m., averaging 10.3 minutes.

108
Workload by Unit
Table 7-7 provides a summary of each NCFD unit’s workload for the year. Tables 7-8 and 7-9 provide a more detailed view of workload, showing each unit’s runs broken out by run type (Table 7-8) and its daily average deployed time by run type (Table 7-9).

TABLE 7-7: Workload by Unit
Station Unit Unit Type
Deployed
Minutes
per Run
Total
Hours
Total
Pct.

Deployed
Minutes
per Day
Total
Runs
Runs
per
Day
31 NCE31 Engine 18.1 915.3 29.5 150.5 3,031 8.3
NCE231 Engine 12.3 0.6 0.0 0.1 3 0.0
Total 18.1 915.9 29.5 150.6 3,034 8.3
33 NCSQ33 Squad 20.2 742.2 23.9 122.0 2,201 6.0
34 B57 Battalion 18.9 145.2 4.7 23.9 462 1.3
NCE34 Engine 17.4 1,011.5 32.6 166.3 3,495 9.6
NCE234 Engine 648.0 10.8 0.3 1.8 1 0.0
NCT34 Truck 16.1 280.0 9.0 46.0 1,046 2.9
Total 17.4 1,447.5 46.6 237.9 5,004 13.7
Total 18.2 3,105.6 100.0 510.5 10,239 28.1

TABLE 7-8: Total Runs by Run Type and Unit
Station Unit EMS False
Alarm
Good
Intent Hazard Outside
Fire
Public
Service
Structure
Fire Cancel Aid
Given Total
31 NCE31 1,279 101 27 21 68 31 26 279 1,199 3,031
NCE231 2 0 0 0 0 0 0 1 0 3
Total 1,281 101 27 21 68 31 26 280 1,199 3,034
33 NCSQ33 1,773 76 12 13 29 37 23 229 9 2,201
34 B57 33 12 3 6 28 5 22 17 336 462
NCE34 2,008 181 22 29 65 60 22 428 680 3,495
NCE234 0 0 0 0 0 0 0 0 1 1
NCT34 501 89 4 8 23 17 26 106 272 1,046
Total 2,542 282 29 43 116 82 70 551 1,289 5,004
Total 5,596 459 68 77 213 150 119 1,060 2,497 10,239
Note: See Table 7-7 for unit type.

109
TABLE 7-9: Deployed Minutes per Day by Run Type and Unit Station Unit EMS False
Alarm
Good
Intent Hazard Outside
Fire
Public
Service
Structure
Fire Cancel Aid
Given Total
31 NCE31 66.1 3.3 1.1 1.0 4.8 1.3 2.9 4.7 65.1 150.5
NCE231 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1
Total 66.2 3.3 1.1 1.0 4.8 1.3 2.9 4.7 65.1 150.6
33 NCSQ33 107.0 2.6 0.3 0.7 1.8 1.7 2.3 5.4 0.2 122.0
34 B57 1.4 0.3 0.2 0.2 1.5 0.9 2.0 0.3 17.1 23.9
NCE34 104.1 7.7 1.1 1.4 5.0 4.3 2.9 8.1 31.7 166.3
NCE234 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.8 1.8
NCT34 21.2 3.0 0.3 0.4 1.5 1.3 3.3 1.8 13.3 46.0
Total 126.7 11.0 1.5 2.0 8.1 6.4 8.2 10.2 63.9 237.9
Total 299.9 16.9 2.9 3.7 14.7 9.4 13.4 20.3 129.2 510.5 Note: See Table 7-7 for unit type.

Observations:
■ Station 34 made the most runs (5,004 or an average of 13.7 runs per day) and had the highest total annual deployed time (1,447.5 or an average of 4.0 hours per day). □ EMS calls accounted for 51 percent of runs and 53 percent of total deployed time. □ Outside and structure fire calls accounted for 4 percent of runs and 7 percent of total deployed time.
■ Station 31 made the second-most runs (3,034 or an average of 8.3 runs per day) and had the second-highest total annual deployed time (915.9 or an average of 2.5 hours per day). □ EMS calls accounted for 42 percent of runs and 44 percent of total deployed time. □ Outside and structure fire calls accounted for 3 percent of runs and 5 percent of total deployed time.
■ Unit NCE34 made the most runs (3,495 or an average of 9.6 runs per day) and had the highest total annual deployed time (1,011.5 or an average of 2.8 hours per day). □ EMS calls accounted for 57 percent of runs and 63 percent of total deployed time. □ Outside and structure fire calls accounted for 2 percent of runs and 5 percent of total deployed time.
■ Unit NCE31 made the second most runs (3,031 or an average of 8.3 runs per day) and had the second-highest total annual deployed time (915.3 or an average of 2.5 hours per day). □ EMS calls accounted for 42 percent of runs and 44 percent of total deployed time. □ Outside and structure fire calls accounted for 3 percent of runs and 5 percent of total deployed time.

110
Workload by Fire District
Table 7-10 breaks down the agency’s workload by fire district. Table 7-11 provides further detail for the workload associated with structure and outside fire calls. Table 7-11 includes the aid given runs to outside and structure fires outside the National City fire district.

TABLE 7-10: Annual Workload by Fire District
District Calls
Pct.

Annual
Calls
Runs
Runs
Per
Day
Deployed
Minutes
Per Run
Annual
Hours
Pct.

Annual
Work
Deployed
Minutes
Per Day
National City 6,692 75.7 7,742 21.2 18.0 2,319.9 74.7 381.3 San Diego City 1,323 15.0 1,495 4.1 19.8 494.5 15.9 81.3 Chula Vista 699 7.9 864 2.4 15.6 225.1 7.2 37.0
San Diego
County 101 1.1 105 0.3 32.4 56.8 1.8 9.3
Imperial Beach 21 0.2 21 0.1 13.0 4.5 0.1 0.7
Coronado 7 0.1 9 0.0 29.0 4.4 0.1 0.7
Lemon Grove 3 0.0 3 0.0 9.3 0.5 0.0 0.1
Total 8,846 100.0 10,239 28.1 18.2 3,105.6 100.0 510.5

TABLE 7-11: Structure and Outside Fire Runs by Fire District District Structure
Fire Runs
Structure
Fires
Deployed
Min. per Run
Outside
Fire
Runs
Outside Fires
Deployed
Min. per Run
Hours for
Structure
and Outside
Fires
Pct. of
Structure and
Outside Fire
Workload
National City 119 41.1 213 25.1 170.8 51.3
San Diego 122 22.9 44 62.1 92.2 27.7
Chula Vista 75 34.3 36 22.0 56.1 16.8
Imperial Beach 12 17.6 0 NA 3.5 1.0
San Diego
County 3 53.9 3 78.6 6.6 2.0
Coronado 2 119.0 0 NA 4.0 1.2
Total 333 32.7 296 30.8 333.2 100.0
Note: All runs outside the National City fire district were mutual aid. The runs within National City match the number of runs described in Table 7-5.

111
Observations:

National City Fire
■ There were 6,692 calls or 76 percent of the total calls. ■ There were 7,742 runs or 21.2 runs per day.
■ Total deployed time for the year was 2,319.9 hours or 75 percent of the total annual workload.

The daily average was 381.3 minutes for all units combined.

San Diego Fire
■ There were 1,323 calls or 15 percent of the total calls. ■ There were 1,495 runs or 4.1 runs per day.
■ Total deployed time for the year was 494.5 hours or 16 percent of the total annual workload.

The daily average was 81.3 minutes for all units combined.

Chula Vista Fire
■ There were 699 calls or 8 percent of the total calls. ■ There were 864 runs or 2.4 runs per day.
■ Total deployed time for the year was 225.1 hours or seven percent of the total annual workload. The daily average was 37.0 minutes for all units combined.

Other
■ There were 132 calls or one percent of the total calls. ■ There were 138 runs or 0.4 runs per day.
■ Total deployed time for the year was 66.2 hours or two percent of the total annual workload.

The daily average was 10.9 minutes for all units combined.

112

ANALYSIS OF BUSIEST HOURS
In this analysis, we included all 9,298 calls that occurred inside and outside National City’s fire district in 2019. For these calls, there is significant variability in the number of calls from hour to hour. One special concern relates to the resources available for hours with the heaviest workload. We tabulated the data for each of the 8,760 hours in the year. Table 7-12 shows the number of hours in the year in which there were zero to six and more calls during the hour.

Table 7-13 shows the ten one-hour intervals which had the most calls during the year. Table 7-14 examines the number of times a call overlapped with another call in each station area in 2019.

TABLE 7-12: Frequency Distribution of the Number of Calls by Year Calls in an Hour Frequency Percentage
0 3,297 37.6
1 2,938 33.5
2 1,641 18.7
3 582 6.6
4 217 2.5
5 62 0.7
6+ 23 0.3
Total 8,760 100.0

TABLE 7-13: Top Ten Hours with the Most Calls Received
Hour Number
of Calls
Number
of Runs
Total
Deployed Hours
5/14/2019, midnight to 1:00 a.m. 10 10 3.1
3/5/2019, midnight to 1:00 a.m. 9 12 2.2
11/28/2019, 11:00 a.m. to noon 8 19 2.8
7/20/2019, 2:00 p.m. to 3:00 p.m. 8 13 4.9
6/3/2019, midnight to 1:00 a.m. 8 11 1.7
10/31/2019, 11:00 a.m. to noon 8 10 1.0
11/15/2019, 2:00 p.m. to 3:00 p.m. 7 7 2.0
4/12/2019, 2:00 p.m. to 3:00 p.m. 6 10 2.4
1/8/2019, 3:00 p.m. to 4:00 p.m. 6 9 2.6
12/28/2019, 3:00 p.m. to 4:00 p.m. 6 9 2.4
Note: Total deployed hours is a measure of the total time spent responding to calls received in the hour. The deployed time from these calls may extend into the next hour or hours. The number of runs and deployed hours includes all units from the studied agencies. Here we considered units from all responding agencies

113
TABLE 7-14: Frequency of Overlapping Calls
Station Scenario Number
of Calls
Percent of
All Calls
Total
Hours
31 No overlapped call 2,862 87.1 995.8
Overlapped with one call 380 11.6 65.9
Overlapped with two calls 41 1.2 4.8
Overlapped with three calls 3 0.1 0.5
34 No overlapped call 3,289 85.3 1,048.1
Overlapped with one call 505 13.1 87.6
Overlapped with two calls 55 1.4 7.6
Overlapped with three calls 7 0.2 0.6
Overlapped with four calls 2 0.1 0.0
Outside
No overlapped call 1,968 91.4 631.1
Overlapped with one call 173 8.0 34.3
Overlapped with two calls 13 0.6 1.3

Table 7-15 examines each NCFD station’s availability to respond to calls within its first due area.

At the same time, it focuses on calls where at least one unit (NCFD or another FD agency) eventually arrived and ignores calls where no unit arrived. While there were 7,144 calls within National City’s fire district (See Table 7-1, the fifth row of the “Total” column), there were 573 calls without an arriving unit.

TABLE 7-15: NCFD Station Availability to Respond to Calls Station Calls in
Area
First Due
Responded
First Due
Arrived
First Due
First
Percent
Responded
Percent
Arrived
Percent
First
31 3,063 1,430 1,347 1,270 46.7 44.0 41.5
34 3,508 2,700 2,639 2,588 77.0 75.2 73.8
Total 6,571 4,130 3,986 3,858 62.9 60.7 58.7
Note: For each station, we count the number of calls occurring within its first due area. Then, we count the number of calls to where at least one unit arrived. Next, we focus on units from the first due station to see if any of its units responded, arrived, or arrived first.

Observations:
■ During 23 hours (0.3 percent of all hours), six or more calls occurred; in other words, the department responded to six or more calls in an hour roughly once every 16 days. □ The highest number of calls to occur in an hour was 10, which happened once. ■ The hour with the most calls was from midnight to 1:00 a.m. on May 14, 2019. The hour’s 10 calls involved 10 individual dispatches resulting in 3.1 hours of deployed time. These 10 calls included three cardiac and stroke calls, two illness and other calls, two MVA calls, one breathing difficulty call, one fall and injury call, and one seizure and unconsciousness call.

114

RESPONSE TIME
In this part of the analysis, we present response time statistics for different call types. We separate response time into its identifiable components. Dispatch time is the difference between the time a call is received and the time a unit is dispatched. Dispatch time includes call processing time, which is the time required to determine the nature of the emergency and the types of resources to dispatch. Turnout time is the difference between dispatch time and the time a unit is en route to a call’s location. Travel time is the difference between the time en route and arrival on scene.

Response time is the total time elapsed between receiving a call to arriving on scene.

In this analysis, we included all calls within the National City fire district to which at least one non-administrative NCFD unit arrived. Units from non-NCFD agencies were also included. Also, calls with a total response time exceeding 30 minutes were excluded. In addition, non-emergency calls were excluded. Finally, we focused on units that had complete time stamps, that is, units with all components recorded, so that we could calculate each segment of response time.

Based on the methodology above, for 8,846 calls in 2019, we excluded 2,154 aid given calls (outside National City), 853 canceled calls, one non-emergency call, 43 calls where no units recorded a valid on-scene time, 85 calls with a total response time exceeding 30 minutes, and 56 calls where one or more segments of the first arriving unit’s response time could not be calculated due to missing or faulty data. As a result, in this section, a total of 5,654 calls are included in the analysis. Using the same method, we obtained 5,364 calls for the same analysis for 2020. 2020’s response time analysis is compared with that of 2019 in Attachment I.

Response Time by Type of Call
Table 7-16 breaks down the average dispatch, turnout, travel, and total response times by call type for all 2019 calls in the National City fire district, and Table 7-17 does the same for 90th percentile response times. A 90th percentile means that 90 percent of calls had response times at or below that number. For example, Table 7-17 shows an overall 90th percentile response time of 8.7 minutes, which means that 90 percent of the time, a call had a response time of no more than 8.7 minutes. Figures 7-7 and 7-8 illustrate the same information.

115
TABLE 7-16: Average Response Time of First Arriving Unit, by Call Type Call Type Time in Minutes Number of Calls Dispatch Turnout Travel Total Breathing difficulty 1.8 1.1 3.2 6.1 703
Cardiac and stroke 2.1 1.0 3.1 6.2 762
Fall and injury 2.1 1.0 3.5 6.7 979
Illness and other 2.2 1.0 3.3 6.6 1,300
MVA 1.2 1.1 3.7 6.0 377
Overdose and psychiatric 2.3 1.1 4.2 7.5 145
Seizure and unconsciousness 2.0 1.0 3.2 6.2 725
EMS Total 2.0 1.0 3.3 6.4 4,991
False alarm 1.7 1.2 3.5 6.4 300
Good intent 2.3 1.1 5.3 8.7 51
Hazard 1.7 1.2 4.0 6.9 47
Outside fire 1.7 1.3 3.6 6.5 123
Public service 2.3 1.1 4.1 7.5 112
Structure fire 2.2 1.0 2.6 5.8 30
Fire Total 1.9 1.2 3.7 6.8 663
Total 2.0 1.1 3.4 6.5 5,654

FIGURE 7-7: Average Response Time of First Arriving Unit, by Call Type – EMS

116
FIGURE 7-8: Average Response Time of First Arriving Unit, by Call Type – Fire

TABLE 7-17: 90th Percentile Response Time of Average Response Time of First Arriving Unit, by Call Type
Call Type Time in Minutes Number of Calls Dispatch Turnout Travel Total Breathing difficulty 3.1 1.9 4.8 8.0 703
Cardiac and stroke 3.4 1.8 4.8 8.3 762
Fall and injury 3.6 1.8 5.4 8.8 979
Illness and other 3.8 1.8 5.1 8.9 1,300
MVA 2.1 1.8 5.9 8.5 377
Overdose and psychiatric 3.9 1.8 6.8 10.4 145
Seizure and unconsciousness 3.5 1.7 4.8 8.2 725
EMS Total 3.5 1.8 5.2 8.6 4,991
False alarm 2.7 2.1 5.5 8.7 300
Good intent 4.7 1.7 10.6 13.8 51
Hazard 2.7 2.0 5.8 10.8 47
Outside fire 2.5 2.0 5.6 9.3 123
Public service 3.5 2.0 6.6 10.8 112
Structure fire 3.3 1.7 4.4 7.7 30
Fire Total 3.2 2.0 6.1 9.7 663
Total 3.5 1.8 5.3 8.7 5,654

117
Observations:
■ The average dispatch time was 2.0 minutes.
■ The average turnout time was 1.1 minutes.
■ The average travel time was 3.4 minutes.
■ The average total response time was 6.5 minutes.
■ The average response time was 6.4 minutes for EMS calls and 6.8 minutes for fire calls. ■ The average response time was 6.5 minutes for outside fires and 5.8 minutes for structure fires. ■ The 90th percentile dispatch time was 3.5 minutes.
■ The 90th percentile turnout time was 1.8 minutes.
■ The 90th percentile travel time was 5.3 minutes.
■ The 90th percentile total response time was 8.7 minutes. ■ The 90th percentile response time was 8.6 minutes for EMS calls and 9.7 minutes for fire calls. ■ The 90th percentile response time was 9.3 minutes for outside fires and 7.7 minutes for structure fires.

118
Table 7-18 shows the average response time by the time of day. The table also shows 90th percentile response times. Figure 7-9 shows the average response time by the time of day.

TABLE 7-18: Average and 90th Percentile Response Time of First Arriving Unit, by Hour of Day
Hour Dispatch Turnout Travel Response
Time
90th Percentile
Response Time
Number
of Calls
0 2.0 1.5 3.9 7.4 9.6 152
1 1.9 1.7 3.4 7.0 9.2 138
2 1.8 1.7 3.6 7.1 9.0 120
3 2.1 1.7 3.6 7.4 9.4 111
4 1.7 1.6 3.6 6.9 8.6 105
5 1.7 1.6 3.8 7.1 8.6 139
6 1.9 1.5 3.6 7.0 9.1 164
7 2.0 1.0 3.5 6.5 8.8 226
8 2.1 0.9 3.3 6.2 8.6 272
9 2.1 0.9 3.2 6.1 8.4 246
10 2.1 0.9 3.3 6.3 8.2 280
11 2.0 0.9 3.4 6.3 8.8 324
12 2.1 0.9 3.4 6.5 8.8 298
13 2.2 0.9 3.1 6.3 8.4 281
14 2.3 0.8 3.3 6.3 8.8 290
15 2.1 0.8 3.5 6.4 9.2 314
6 2.0 0.9 3.3 6.1 8.4 309
17 2.0 0.9 3.3 6.1 8.2 312
18 2.0 0.9 3.3 6.1 8.3 316
19 1.9 0.9 3.2 6.1 8.3 307
20 1.9 1.0 3.3 6.3 8.3 297
21 2.0 1.2 3.5 6.7 8.8 270
22 2.0 1.4 3.3 6.6 8.8 203
23 2.0 1.4 3.6 6.9 9.1 180
Total 2.0 1.1 3.4 6.5 8.7 5,654

119
FIGURE 7-9: Average Response Time of First Arriving Unit, by Hour of Day

Observations:
■ Average dispatch time was between 1.7 minutes (4:00 a.m. to 5:00 a.m.) and 2.3 minutes (2:00 p.m. to 3:00 p.m.).
■ Average turnout time was between 0.8 minutes (2:00 p.m. to 3:00 p.m.) and 1.7 minutes (2:00 a.m. to 3:00 a.m.).
■ Average travel time was between 3.1 minutes (1:00 p.m. to 2:00 p.m.) and 3.9 minutes (midnight to 1:00 a.m.).
■ Average response time was between 6.1 minutes (7:00 p.m. to 8:00 p.m.) and 7.4 minutes (midnight to 1:00 a.m.).
■ The 90th percentile response time was between 8.2 minutes (10:00 a.m. to 11:00 a.m.) and

9.6 minutes (midnight to 1:00 a.m.).

120
Response Time Distribution
Here, we present a more detailed look at how response times to calls are distributed. The cumulative distribution of total response time for the first arriving unit to EMS calls is shown in Figure 7-10 and Table 7-19. Figure 7-10 shows response times for the first arriving unit to EMS calls as a frequency distribution in whole-minute increments, and Figure 7-11 shows the same for the first arriving unit to outside and structure fire calls.

The cumulative percentages here are read in the same way as a percentile. In Figure 7-10, the 90th percentile of 8.6 minutes means that 90 percent of EMS calls had a response time of

8.6 minutes or less. In Table 7-19, the cumulative percentage of 84.7, for example, means that

84.7 percent of EMS calls had a response time under 8 minutes.

FIGURE 7-10: Cumulative Distribution of Response Time – First Arriving Unit – EMS

121
FIGURE 7-11: Cumulative Distribution of Response Timer – First Arriving Unit – Outside and Structure Fires

TABLE 7-19: Cumulative Distribution of Response Time – First Arriving Unit – EMS Response Time
(Minute) Frequency Cumulative
Percentage
1 2 0.0
2 24 0.5
3 79 2.1
4 260 7.3
5 709 21.5
6 1,177 45.1
7 1,183 68.8
8 794 84.7
9 386 92.4
10 177 96.0
11 80 97.6
12 31 98.2
13 26 98.7
14+ 63 100.0

122
TABLE 7-20: Cumulative Distribution of Response Time – First Arriving Unit – Outside and Structure Fires
Response Time
(Minute) Frequency Cumulative
Percentage
1 0 0.0
2 1 0.7
3 1 1.3
4 9 7.2
5 33 28.8
6 42 56.2
7 25 72.5
8 17 83.7
9 8 88.9
10 9 94.8
11 3 96.7
12 1 97.4
13 2 98.7
14+ 2 100.0
Observations:
■ For 85 percent of EMS calls, the response time of the first arriving unit was less than 8 minutes. ■ For 84 percent of outside and structure fire calls, the response time of the first arriving unit was less than 8 minutes.

123

FIRE LOSS
Table 7-21 presents the number of outside and structure fires, broken out by levels of fire loss.

Table 7-22 shows the amount of property and content loss for outside and structure fires inside the NCFD fire district in 2019.

TABLE 7-21: Total Fire Loss Above and Below $25,000
Call Type No Loss Under $25,000 $25,000 plus Total
Outside fire 108 16 1 125
Structure fire 16 11 4 31
Total 124 27 5 156

TABLE 7-22: Content and Property Loss – Structure and Outside Fires Call Type Property Loss Content Loss
Loss Value Number of Calls Loss Value Number of Calls
Outside fire $1,092,100 15 $3,200 5
Structure fire $287,200 13 $39,700 13
Total $1,379,300 28 $42,900 18
Note: The table includes only fire calls with a recorded loss greater than 0.

Observations:
■ 108 outside fires and 16 structure fires had no recorded loss. ■ 1 outside fire and 4 structure fires recorded losses above $25,000. ■ Structure fires:
□ The highest total loss for a structure fire was $155,000. □ The average total loss for structure fires with loss was $21,793. □ 13 structure fires recorded a content loss totaling $39,700. □ Out of 31 structure fires, 13 recorded a property loss totaling $287,200. ■ Outside fires:
□ The highest total loss for an outside fire was $1,000,000. □ The average total loss for outside fires with loss was $64,429. □ 5 outside fires recorded content losses totaling $3,200. □ Out of 125 outside fires, 15 recorded property losses totaling $1,092,100.

124
ATTACHMENT I: 2019 & 2020 COMPARISON
In this analysis, we compare portions of the previous analysis with similar records for 2020. We compare calls by type, unit workload, agency’s availability, and response times.

Calls Volume by Year
Table 7-23 shows the number of calls for both 2019 and 2020. Figure 7-12 shows the monthly variation in the number of calls per day for both years. Similarly, Figure 7-13 illustrates the average number of calls per hour for both years.

TABLE 7-23: Calls by Type and Year
Call Type
2019 2020
Total
Calls
Calls
per Day
Total
Calls
Calls
per Day
Breathing difficulty 722 2.0 674 1.8
Cardiac and stroke 779 2.1 740 2.0
Fall and injury 999 2.7 952 2.6
Illness and other 1,344 3.7 1,303 3.6
MVA 407 1.1 349 1.0
Overdose and psychiatric 151 0.4 171 0.5
Seizure and unconsciousness 738 2.0 620 1.7
EMS total 5,140 14.1 4,809 13.1
False alarm 318 0.9 216 0.6
Good intent 56 0.2 81 0.2
Hazard 48 0.1 33 0.1
Outside fire 125 0.3 162 0.4
Public service 121 0.3 139 0.4
Structure fire 31 0.1 29 0.1
Fire total 699 1.9 660 1.8
Canceled 853 2.3 922 2.5
Aid given 2,154 5.9 2,090 5.7
Total 8,846 24.2 8,481 23.2

125
FIGURE 7-12: Average Calls by Month and Year

FIGURE 7-13: Average Calls by Hour of Day and Year

126
Workload by Year
Table 7-24 compares the call volume, annual runs, and workload by fire district in 2019 and 2020.

Table 7-25 compares the annual runs and workload by NCFD station and unit during the two years. Figure 7-14 compares the average deployed minutes by the hour of the day in 2019 and 2020. Note that in Figure 7-14, the workload created by incident FMSC202350 was not included.

Unit NCE34 responded to this incident with a duration time of 752.9 hours (19 pe rcent of the annual workload). This is an outlier but has a significant influence on the workload in 2020.

TABLE 7-24: Annual Workload by District and Year
District 2019 2020
Calls Runs Hours Calls Runs Hours
National City 6,692 7,742 2,319.9 6,391 7,540 2,320.1
San Diego 1,323 1,495 494.5 1,328 1,525 541.6
Chula Vista 699 864 225.1 653 813 224.8
San Diego County 101 105 56.8 77 83 45.1
Imperial Beach 21 21 4.5 21 25 5.8
Coronado 7 9 4.4 10 13 5.6
Lemon Grove 3 3 0.5
Fresno County * 1 3 752.9
Total 8,846 10,239 3,105.6 8,481 10,002 3,895.8
Note: *2020 included responses to one wildfire (Creek Fire) recorded as incident number FMSC202350.

Unit NCE34 responded to this call from September 6, 2020, to October 7, 2020.

TABLE 7-25: Annual Workload by Station, Unit, and Year
Station Unit Unit Type 2019 2020
Hours Runs Hours Runs
31 NCE31 Engine 915.3 3,031 916.6 2,989
NCE231 Engine 0.6 3
Total 915.9 3,034 916.6 2,989
33 NCSQ33 Squad 742.2 2,201 696.3 2,098
34 B57 Battalion 145.2 462 182.8 460
NCE34* Engine 1,011.5 3,495 1,711.0 3,152
NCE234 Engine 10.8 1 113.3 368
NCT34 Truck 280.0 1,046 275.9 935
Total 1,447.5 5,004 2,282.9 4,915
Total 3,105.6 10,239 3,895.8 10,002
Note: *NCE34 includes 753 hours associated with one wildfire (Creek Fire) in 2020.

127
FIGURE 7-14: Average Deployed Minutes by Hour of Day in 2019 and 2020

Agency’s Availability by Year
Table 7-26 compares each NCFD station’s response availability to calls that occurred in its first due area in both years. We focused on calls where a unit eventually arrived and ignores calls where no unit arrived.

TABLE 7-26: NCFD Station Availability to Respond to Calls by Year Station
2019 2020
Calls
in
Area
Percent
Responded
Percent
Arrived
Percent
First
Calls
in
Area
Percent
Responded
Percent
Arrived
Percent
First
31 3,137 45.6 42.9 30.1 2,880 46.4 42.7 30.8
34 3,658 73.8 72.1 54.5 3,584 73.7 71.6 53.7
Total 6,795 60.8 58.7 43.2 6,464 61.6 58.7 43.5
Note: For each station, we count the number of calls occurring within its first due area. Then, we count the number of calls to where at least one unit arrived. Next, we focus on units from the first due station to see if any of its units responded, arrived, or arrived first.

128
Response Time by Year
Tables 7-27 and 7-28 compare the average and 90th percentile response times by call type in 2019 and 2020.

TABLE 7-27: Average Response Time of First Arriving Unit, by Call Type and Year Call Type
2019 2020
Time in Minutes Calls Time in Minutes
Calls Dispatch Turnout Travel Total Dispatch Turnout Travel Total False alarm 1.7 1.2 3.5 6.4 300 1.8 1.1 3.9 6.8 203
Good intent 2.3 1.1 5.3 8.7 51 2.0 1.1 4.4 7.6 75
Hazard 1.7 1.2 4.0 6.9 47 1.7 1.0 3.4 6.1 33
Outside fire 1.7 1.3 3.6 6.5 123 1.8 1.2 4.1 7.0 160
Public service 2.3 1.1 4.1 7.5 112 2.0 1.1 4.3 7.3 126
Structure fire 2.2 1.0 2.6 5.8 30 1.7 0.9 3.3 5.8 29
Fire Total 1.9 1.2 3.7 6.8 663 1.8 1.1 4.0 7.0 626
EMS Total 2.0 1.0 3.3 6.4 4,991 2.1 1.1 3.7 6.8 4,738
Total 2.0 1.1 3.4 6.5 5,654 2.1 1.1 3.7 6.9 5,364

TABLE 7-28: 90th Percentile Response Time of First Arriving Unit, by Call Type and Year
Call Type
2019 2020
Time in Minutes Calls Time in Minutes
Calls Dispatch Turnout Travel Total Dispatch Turnout Travel Total False alarm 2.7 2.1 5.5 8.7 300 2.9 2.0 6.1 9.4 203
Good intent 4.7 1.7 10.6 13.8 51 3.6 2.0 6.4 11.0 75
Hazard 2.7 2.0 5.8 10.8 47 3.0 1.5 5.0 8.4 33
Outside fire 2.5 2.0 5.6 9.3 123 3.0 2.1 6.2 9.4 160
Public service 3.5 2.0 6.6 10.8 112 3.9 2.0 7.3 10.8 126 Structure fire 3.3 1.7 4.4 7.7 30 2.4 1.8 5.1 8.2 29
Fire Total 3.2 2.0 6.1 9.7 663 3.1 2.0 6.2 9.4 626
EMS Total 3.5 1.8 5.2 8.6 4,991 3.6 2.0 5.5 9.3 4,738
Total 3.5 1.8 5.3 8.7 5,654 3.5 2.0 5.6 9.3 5,364

129
ATTACHMENT II: ADDITIONAL PERSONNEL
Table 7-29 illustrates the workload of NCFD’s administrative units in 2019 and 2020, respectively.

TABLE 7-29: Workload of Administrative Units by Year
Unit ID Unit Type
2019 2020
Annual
Hours
Annual
Runs
Annual
Hours
Annual
Runs
5701 Fire Chief 6.9 2 0.0 0
5703 Battalion Chief 0.0 0 0.3 2
5705 Fire Marshal 10.5 10 10.3 6
5706 Deputy Fire Marshal 9.7 6 12.4 9

130
ATTACHMENT III: CALLS OUTSIDE NATIONAL CITY FIRE DISTRICT From 2019 to 2020, NCFD responded to 4,244 aid-given calls outside of its fire district (see Table 7-23). Table 7-30 details these calls by call type and year. Of these, 241 were structure fire calls and 153 were outside fire calls. Figures 7-15 and 7-16 show the percentage of calls that fall into each EMS (Figure 7-15) and fire (Figure 7-16) type category by year.

TABLE 7-30: Calls Outside NCFD District by Call Type and Year Call Type
2019 2020
Total
Calls
Calls
per Day
Pct.

Calls
Total
Calls
Calls
per Day
Pct.

Calls
Breathing difficulty 173 0.5 8.0 176 0.5 8.4
Cardiac and stroke 209 0.6 9.7 192 0.5 9.2
Fall and injury 204 0.6 9.5 181 0.5 8.7
Illness and other 347 1.0 16.1 303 0.8 14.5
MVA 128 0.4 5.9 128 0.3 6.1
OD 26 0.1 1.2 36 0.1 1.7
Seizure and UNC 178 0.5 8.3 143 0.4 6.8
EMS Total 1,265 3.5 58.7 1,159 3.2 55.5
False alarm 80 0.2 3.7 81 0.2 3.9
Good intent 16 0.0 0.7 25 0.1 1.2
Hazard 25 0.1 1.2 35 0.1 1.7
Outside fire 67 0.2 3.1 86 0.2 4.1
Public service 31 0.1 1.4 37 0.1 1.8
Structure fire 135 0.4 6.3 106 0.3 5.1
Fire Total 354 1.0 16.4 370 1.0 17.7
Canceled 535 1.5 24.8 561 1.5 26.8
Total 2,154 5.9 100.0 2,090 5.7 100.0
Note: OD=Overdose and psychiatric; UNC=unconsciousness.

131
FIGURE 7-15: EMS Calls by Type and Year, Outside National City

FIGURE 7-16: Fire Calls by Type and Year, Outside National City

132
ATTACHMENT IV: AID RECEIVED WORKLOAD
This section focuses on aid received within National City’s fire district from ot her fire agencies.

From 2019 to 2020, there were 1,963 calls in National City where aid was received from other agencies. Of these, 1,069 calls involved a joint response with NCFD and 894 calls involved a response by other agencies alone (See Table 7-1).

Aid Received Calls by Type
Table 7-31 shows the number of calls to which other FD agencies responded, broken out by call type and year. The table also presents the annual runs and work hours for each type of call.

TABLE 7-31: Aid Received Workload by Type and Year, Inside National City Call Type
Total Annual
Calls
Total Annual
Runs
Total Annual
Hours
2019 2020 2019 2020 2019 2020
Breathing difficulty 62 66 65 72 30.1 40.7
Cardiac and stroke 75 86 85 94 43.9 41.4
Fall and injury 100 114 106 119 38.5 47.2
Illness and other 131 140 144 179 52.7 80.7
MVA 193 177 295 270 75.1 68.7
OD 19 22 22 24 8.6 12.4
Seizure and UNC 69 71 70 75 30.4 42.0
EMS Total 649 676 787 833 279.1 333.1
False alarm 67 52 104 94 15.1 15.2
Good intent 21 48 22 77 5.9 12.5
Hazard 13 18 32 46 9.9 6.2
Outside fire 36 57 94 127 23.5 45.6
Public service 25 31 32 54 12.1 8.8
Structure fire 24 25 106 135 31.7 48.8
Fire Total 186 231 390 533 98.2 137.1
Canceled 116 105 153 146 35.2 29.2
Total 951 1,012 1,330 1,512 412.4 499.4
Note: OD=Overdose and psychiatric; UNC=Unconsciousness.

133
Runs and Arrivals by Aid Agency
Tables 7-32 and 7-33 compare the number of aid-received runs and arrivals by different agencies in 2019 and 2020.

TABLE 7-32: Aid Received Runs by Agency, First Due Area, and Year Agency
2019 2020
First Due Area Total First Due Area Total 31 34 31 34
Bonita FD 104 0 104 94 1 95
Coronado FD 0 4 4 1 1 2
Chula Vista FD 136 196 332 182 240 422
Escondido FD 1 0 1 0 0 0
Federal FD 0 1 1 1 0 1
Lemon Grove FD 0 0 0 0 1 1
San Diego FD 522 364 886 569 421 990
San Miguel FD 1 1 2 1 0 1
Total 764 566 1,330 848 664 1,512

TABLE 7-33: Aid Received Arrivals by Agency, First Due Area, and Year Agency
2019 2020
First Due Area Total First Due Area Total 31 34 31 34
Bonita FD 75 0 75 61 1 62
Coronado FD 0 1 1 0 0 0
Chula Vista FD 95 131 226 121 159 280
Lemon Grove FD 0 0 0 0 1 1
San Diego FD 326 207 533 372 257 629
Total 496 339 835 554 418 972

134
ATTACHMENT V: LINCOLN ACRES
One area of particular interest is Lincoln Acres. While not officially part of National City, it is an unincorporated area that is entirely enclosed within National City’s boundaries. Up until this point, calls within Lincoln Acres were included as part of the National City Fire District. For this section, we used each call’s recorded latitude and longitude to locate the calls within Lincoln Acres.

Table 7-34 compares the volume of calls and the workload for this area for both years, broken down by call type. While Table 7-1 distinguishes calls without a responding NCFD unit, all calls within Lincoln Acres involved a responding NCFD unit. To better understand the workload within Lincoln Acres, we included runs and associated work for all fire agencies responding to calls within the area. Table 7-35 shows the average and 90 percentile response time to calls that occurred in this area. Due to the small sample size, we used all calls in two years in the analysis of response time. Table 7-36 examines the average and 90th response times of the first arriving units by the time of day (in four-hour intervals).

TABLE 7-34: Calls and Workload in Lincoln Acres by Call Type and Year Call Type 2019 2020
Calls Hours Runs Calls Hours Runs
Breathing difficulty 16 20.7 34 16 23.7 35
Cardiac and stroke 19 30.7 46 21 27.7 48
Fall and injury 16 23.9 35 15 24.4 34
Illness and other 23 31.4 54 31 42.6 67
MVA 23 30.4 74 31 30.4 93
OD 2 2.0 4 6 6.6 13
Seizure and UNC 14 19.7 29 15 23.2 31
EMS Total 113 158.8 276 135 178.6 321
False alarm 5 1.8 9 5 7.0 15
Good intent 3 2.6 5 6 5.1 24
Hazard 1 0.1 1 4 2.3 10
Outside fire 5 5.6 20 7 12.5 20
Public service 5 1.6 6 3 0.9 3
Structure fire 4 42.0 36 0 0.0 0
Fire Total 23 53.8 77 25 27.7 72
Canceled 28 23.7 77 41 34.9 100
Total 164 236.2 430 201 241.2 493
Note: OD=Overdose and psychiatric; UNC=Unconsciousness.

135
TABLE 7-35: Response Time in Lincoln Acres, by Call Type Call Type
Average Response Time (Minutes) 90 Percentile Response Time (Minutes) Call Count Dispatch Turnout Travel Total Dispatch Turnout Travel Total False alarm 2.7 0.7 3.3 6.8 12.3 1.7 7.2 12.6 8
Good intent 2.3 0.8 4.9 8.0 7.0 1.6 10.6 13.2 7
Hazard 2.5 1.1 4.2 7.7 7.8 1.5 5.6 9.1 5
Outside fire 1.7 1.0 3.9 6.6 2.7 1.8 6.6 8.7 12
Public service 3.2 1.1 4.5 8.8 10.8 2.0 7.9 15.5 7
Structure fire 3.6 0.6 1.6 5.8 7.4 1.2 2.0 9.3 3
Fire Total 2.5 0.9 3.9 7.3 7.0 1.7 6.6 11.5 42
EMS Total 2.0 1.0 3.9 7.0 3.4 1.8 6.2 9.6 240
Total 2.1 1.0 3.9 7.0 3.5 1.8 6.3 9.8 282

TABLE 7-36: Response Time in Lincoln Acres, by Time of Day Time Average, Minutes 90 Percentile, Minutes Call
Count Dispatch Turnout Travel Total Dispatch Turnout Travel Total 00:00 - 03:59 2.1 1.6 3.5 7.2 4.3 2.3 5.4 9.8 28
04:00 - 07:59 1.8 1.3 4.5 7.6 2.9 2.2 8.5 12.4 26
08:00 - 11:59 1.8 0.8 4.0 6.6 3.0 1.4 6.2 9.0 60
12:00 - 15:59 2.3 0.7 3.9 6.9 3.0 1.4 6.2 8.9 50
16:00 - 19:59 2.1 0.8 4.2 7.1 3.4 1.4 6.8 9.2 61
20:00 - 23:59 2.2 1.2 3.6 7.1 5.0 1.9 6.2 10.3 57
Total 2.1 1.0 3.9 7.0 3.5 1.8 6.3 9.8 282

136
ATTACHMENT VI: PARADISE HILLS
Another area of particular interest is Paradise Hills. Paradise Hills is a neighborhood within San Diego that is located close to National City. Calls into Paradise Hills are part of aid given calls measured in Table 7-10. As in the previous section, we used each call’s recorded latitude and longitude to locate calls within Paradise Hills. We compare the volume of calls and the workload for this area over two years. Table 7-37 presents the comparison, broken down by call type. Aid given workload only included calls, workload, and runs associated with NCFD units .

TABLE 7-37: Calls and Workload in Lincoln Acres by Call Type and Year Call Type 2019 2020
Calls Hours Runs Calls Hours Runs
Breathing difficulty 95 31.3 95 110 45.1 111
Cardiac and stroke 116 46.2 116 107 48.2 108
Fall and injury 91 31.6 94 99 36.2 102
Illness and other 120 47.6 128 127 48.2 128
MVA 17 8.3 20 23 7.5 28
OD 7 2.2 7 14 5.9 14
Seizure and UNC 93 39.9 94 73 28.8 73
EMS Total 539 207.3 554 553 219.9 564
False alarm 19 7.1 19 21 5.9 26
Good intent 2 0.4 2 7 1.4 7
Hazard 3 1.7 6 4 19.3 9
Outside fire 6 3.2 6 6 2.6 9
Public service 9 2.6 9 7 2.8 7
Structure fire 12 7.5 18 13 6.8 20
Fire total 51 22.5 60 58 38.8 78
Canceled 73 12.3 99 93 19.1 129
Total 663 242.0 713 704 277.9 771
Note: OD=Overdose and psychiatric; UNC=Unconsciousness.

Observations:
■ In 2019, there were 663 aid-given calls to Paradise Hills. This was 50 percent of aid-given calls (1,323) to San Diego.
■ In 2019, there were 713 aid-given runs to Paradise Hills. This was 48 percent of aid-given runs (1,495) to San Diego
■ In 2019, there were 242.0 aid-given work hours associated with calls in Paradise Hills. This was 49 percent (494.5) of aid-given work associated with calls in San Diego. ■ In 2020, call volume increased by 6 percent from 663 to 704. ■ In 2020, total runs increased by 8 percent from 713 to 771. ■ In 2020, the workload increased by 15 percent from 242.0 to 277.9.

137
ATTACHMENT VII: CALL TYPE IDENTIFICATION
When available, NFIRS data serves as our primary source for assigning call categories. In this work, for an MVA or fire call that had a matched NFIRS record, we used the NFIRS incident type to assign a call category. Otherwise, we used the CAD incident type and problem description to assign a call category. All EMS calls were categorized by the CAD incident type and problem description. Tables 7-38 and 7-39 specify the call categories identified by available NFIRS and CAD information, respectively.

TABLE 7-38: Call Type by NFIRS Incident Type Code
Call Type Incident
Type Code
Frequency
2019 2020
Canceled
611 1,357 1,421
621 1 0
622 38 64
False Alarm
700 296 217
710 2 1
713 1 0
715 1 1
730 3 0
733 3 1
735 3 4
736 4 2
740 1 0
743 2 0
744 3 2
745 4 11
746 1 10
Good
Intent
600 40 39
631 0 2
641 0 2
650 6 4
651 8 29
652 2 1
653 3 3
661 0 2
671 5 10
672 1 0

138
Call Type Incident
Type Code
Frequency
2019 2020
Hazard
223 1 0
400 4 7
410 1 1
411 3 1
412 6 12
413 2 0
420 0 1
421 2 1
423 0 1
424 2 1
440 7 4
441 3 2
442 1 1
443 1 0
444 7 1
445 5 2
460 1 1
461 1 1
480 4 7
481 0 1
MVA
322 464 392
323 7 5
324 2 9
352 1 1
Outside
Fire
100 6 8
130 29 38
131 0 1
140 26 46
150 101 126
151 5 3
161 0 1

139
Call Type Incident
Type Code
Frequency
2019 2020
Public
Service
500 14 22
510 13 9
511 20 10
512 1 1
520 7 7
521 2 1
522 5 5
531 11 24
540 1 1
541 0 1
542 2 1
550 7 6
551 6 8
552 4 9
553 6 9
554 25 13
561 7 23
571 0 1
812 1 0
813 1 0
900 3 3
911 0 1
Structure
Fire
111 51 59
113 22 15
Total 2,686 2,730

140
TABLE 7-39: Call Type by CAD Problem Description
Call Type Problem Frequency
2019 2020
Breathing
Difficulty
Breathing Problems 909 856
Choking 30 30
Cardiac
and Stroke
Cardiac / Respiratory Arrest 131 175
Chest Pain 563 512
Heart Problems 119 116
Stroke 213 169
Fall and
Injury
Assault/Rape 210 214
Drowning/Diving Accident 1 3
Electrocution 3 1
Falls / Back Inj 868 812
Stabbing/Gunshot 39 42
Traumatic Injuries Spec 26 19
Traumatic Injuries, Spec 115 108
False Alarm
Carbon Monoxide Alarm 3 3
Ringing Alarm 53 27
Ringing Alarm Coronado 2 0
Ringing Alarm Highrise 18 13
Vegetation 1st Alarm 18 17
Good
Intent
Noxious Odor 0 1
Odor of Chemical 0 2
Odor of Smoke 1 1
Smoke Check 21 43
Hazard
Nat Gas Leak Broken/Blowing 5 11
Natural Gas Leak/Odor-Inside 3 3
Natural Gas Odor - Outside 2 2
Electrical Short 2 1
Extinguished Fire 1 4
Fuel Spill 1 2
HazMat 1 0
HazMat Single Engine 0 2
Illegal Burn 12 2
Wires down 2 1

141
Call Type Problem Frequency
2019 2020
Illness and
Other
Confined Space/Trench Rescue 1 0
Abdominal Pain/Problems 60 63
Advised Incident* 7 0
Allergy/Hives/Med Rx/Stng 41 43
Animal Bites/ Attacks 13 11
Back Pain 33 19
Burns / Explosion* 3 4
C O / Inhalation/ Haz Mat* 2 2
CV Medical Aid 2 2
Diabetic Problems 147 141
Elevator Rescue 11 11
Eye Problems / Injuries 2 1
Headache 42 32
Heat / Cold Exposure 3 6
Hemorrhage / Lacerations 236 219
Illegal Burn* 3 0
Industrial Rescue 0 1
Lift Assist* 7 2
Medical Aid 16 13
Medical Alert Alarm 43 38
Miscellaneous Rescue 0 1
NC Medical Aid 0 1
Open Space Rescue 1 1
Poison Control 0 1
Preg/Birth/Miscarriage 30 25
Sick Person 806 763
Special Service* 10 5
Suspected COVID19 0 63
Traffic Accident* 90 70
Traffic Accident FWY* 5 3
Unknown Problem* 152 129
Vehicle Fire Freeway* 1 0
Vehicle Rescue 11 13
Vehicle vs. Pedestrian* 4 3
Water Rescue 0 2
MVA
Traffic Accident 120 122
Traffic Accident FWY 30 26
Vehicle vs Structure 3 5
Vehicle vs. Pedestrian 1 1
Note: *NRIFS incident type code is 321.

142
Call Type Problem Frequency
2019 2020
Outside
Fire
Boat Fire 1st Alm 0 1
Fence* 1 0
Pole Fire 0 1
Rubbish Fire 8 8
Tree* 0 1
Vegetation Initial Attack 10 13
Vehicle Fire 5 4
Vehicle Fire Freeway 5 7
Overdose
and
Psychiatric
OD/Ingestion/Poisonings 112 122
Psych / Suicide Attempt 78 100
Public
Service
Advised Incident 1 2
AID - ENGINE 0 1
Assist PD 1 1
Assist PD - Ladder Bldg 0 1
Investigate 1 0
Knocked Off Hydrant 3 4
Lift Assist 2 0
Lock in/out 3 9
Move Up 7 6
SNAKE REMOVAL 1 0
Special Service 4 8
Strike Team Type 1 1 3
Strike Team Type 3 1 1
Water Removal/Flooding CV/NC 2 0
yGT General Transport 1 0
Seizure and
UNC
Convulsions / Seizures 330 258
Unc/Fainting 634 549
Structure
Fire
Oven Fire 1 1
Structure Fire - Comm / Apt 38 29
Structure Highrise/Hospital 1 0
Structure Residential 53 31
Total 6,612 6,193
Note: *Level 2 fires; UNC = Unconsciousness.

143
ATTACHMENT VIII: REMOVED CANCELED CALLS
TABLE 7-40: Removed Calls by Cancel Reason and Year
Cancel Reason Frequency
2019 2020
Duplicate Call 754 793
Call complete / Available 425 485
CAD test 263 220
False Alarm 81 46
Caller refused ambulance 11 11
Patient not ready 8 9
Stand back cancellation 8 2
Canceled by PD/CHP on scene 2 6
Canceled/Turned 1 3
Change in level of service 0 2
Delayed in traffic 2 0
Private transport arranged 1 1
Wrong location 1 1
Level 4 triage 0 1
Canceled by first responder 1 0
NA 6 6
Total 1,564 1,586

144
SECTION 8. AMR DATA ANALYSIS
This data analysis was prepared as a key component of the study of the American Medical Response (AMR) ambulance service in the National City fire district. This analysis examines all calls for service between January 1, 2019, and January 1, 2021, as recorded in the regional computer-aided dispatch (CAD) system, and AMR’s EMS incident Reporting System.

This analysis is made up of five parts. The first part focuses on call types and dispatches. The second part explores the time spent and the workload of individual units. The third part presents an analysis of the busiest hours in the year studied. The fourth part provides a response time analysis of the studied agency’s units. The fifth and final part is an analysis of unit transports. The analysis results are primarily presented for the 2019 calendar year. The results for 2020 are compared with those for the prior year in Attachment I.

As the primary emergency medical service (EMS) provider within the National City fir e district, AMR works closely with the National City Fire Department (NCFD) to provide both advanced life support (ALS) and basic life support (BLS) services. In 2019, AMR responded to 7,328 calls. The total workload was 7,335.9 hours. The average response time to EMS calls was 8.0 minutes, and the 90th percentile response time was 13.2 minutes. In 2020, the AMR responded to 6,945 calls.

The total workload was 6,561.9 hours. The average response time to EMS calls was 8.3 minutes, and the 90th percentile response time was 13.5 minutes.

METHODOLOGY
In this report, CPSM analyzes calls and runs. A call is an emergency service request or incident. A run is a dispatch of a unit (i.e., a unit responding to a call). Thus, a call may include multiple runs.

This analysis studied AMR’s 9-1-1 EMS response. We received data from both the regional CAD system and the AMR’s EMS incident Reporting System. We first matched the two sets of data based on the available information of call time and location. The AMR data lacked information of incident type and unit transport times. Therefore the analysis was primarily conducted based on the CAD data that included the description of call nature and transport time stamps of AMR units. The method to categorize calls based on the call nature description is detailed in Attachment II. With the AMR data, we used the call received time for the analysis of AMR unit’s response time to calls and used the available unit time stamps to fill the missing unit time stamps in the CAD data.

Working independently or jointly with fire departments, AMR responded to 14,273 total calls in the National City fire district in 2019 and 2020. The following table summarizes these calls by responding agency and year. The main analysis in the following sections focuses on the 7,328 calls in 2019. The results for 2020 are presented along with the corresponding 2019 results in Attachment I for comparison.

TABLE 8-1: Studied Calls Responding Agency and Year
Responding Agency 2019 2020 Total
AMR only 1,036 986 2,022
AMR and FD agencies 6,292 5,959 12,251
Total 7,328 6,945 14,273

145
Observations:
■ Of all calls where AMR responded within the National City fire district, AMR responded jointly with FD agencies to 86 percent of calls in both years.

146

AGGREGATE CALL TOTALS AND RUNS
In 2019, AMR responded to 7,328 calls in the National City fire district. Of these calls, 99 percent were 9-1-1 EMS calls and one percent were the service calls for assisting fire or PD agencies.

Calls by Type
Th following table and figure show the number of calls by call type, average calls per day, and the percentage of calls that fall into each call type category for the 12 months studied.

TABLE 8-2: Call Types
Call Type Number of
Calls
Calls per
Day
Call
Percentage
Breathing difficulty 815 2.2 11.1
Cardiac and stroke 881 2.4 12.0
Fall and injury 1,296 3.6 17.7
Illness and other 2,453 6.7 33.5
MVA 677 1.9 9.2
Overdose and psychiatric 266 0.7 3.6
Seizure and unconsciousness 867 2.4 11.8
EMS Total 7,255 19.9 99.0
Fire & PD assist 73 0.2 1.0
Total 7,328 20.1 100.0

147
FIGURE 8-1: Calls by Type

Note: Other includes Canceled and Fire & FD assist calls.

Observations:
■ In 2019, AMR responded to an average of 20.1 calls per day. ■ EMS calls for the year totaled 7,255 (99 percent of all calls), an average of 19.9 calls per day. □ Illness and other calls were the largest category of EMS calls at 34 percent of total calls (34 percent of EMS calls) or an average of 6.7 calls per day. □ Cardiac and stroke calls made up 12 percent of total calls (12 percent of EMS calls) or an average of 2.4 calls per day.
□ Motor vehicle accidents made up 9 percent of total calls (9 percent of EMS calls) or an average of 1.9 calls per day.

148
Calls by Type and Duration
The following table shows the duration of calls by type using four duration categories: less than 30 minutes, 30 minutes to one hour, one to two hours, and two or more hours.

TABLE 8-3: Calls by Type and Duration
Call Type Less than
30 Minutes
30 Minutes
to One Hour
One to
Two Hours
Two or
More Hours Total
Breathing difficulty 103 202 477 33 815
Cardiac and stroke 118 207 533 23 881
Fall and injury 315 248 683 50 1,296
Illness and other 651 509 1,189 104 2,453
MVA 374 86 201 16 677
Overdose and psychiatric 81 63 113 9 266
Seizure and unconsciousness 161 174 493 39 867
EMS Total 1,803 1,489 3,689 274 7,255
Fire & FD assist 58 1 13 1 73
Total 1,861 1,490 3,702 275 7,328
Observations:
■ On average, there were 10.9 EMS calls per day that lasted more than one hour. ■ A total of 3,292 EMS calls (45 percent) lasted less than one hour, 3,689 EMS calls (51 percent) lasted one to two hours, and 274 EMS calls (4 percent) lasted two or more hours. ■ A total of 325 cardiac and stroke calls (37 percent) lasted less than one hour, 533 cardiac and stroke calls (60 percent) lasted one to two hours, and 23 cardiac and stroke calls (3 percent) lasted two or more hours.
■ A total of 460 motor vehicle accidents (68 percent) lasted less than one hour, 201 motor vehicle accidents (30 percent) lasted one to two hours, and 16 motor vehicle accidents (2 percent) lasted two or more hours.

149
Average Calls by Month and Hour of Day
Figure 8-2 shows the monthly variation in the average daily number of calls handled by AMR in 2019. Similarly, Figure 8-3 illustrates the average number of calls received each hour of the day over the year.

FIGURE 8-2: Average Calls by Month

Observations:
■ Average calls per day overall ranged from 18.5 in January 2019 to 22.6 in March 2019.

150
FIGURE 8-3: Calls by Hour of Day

Observations:
■ Average calls per hour overall ranged from 0.4 between 3:00 a.m. and 5:00 a.m. to

1.1 between 11:00 a.m. and noon.

151
Arriving Units
Table 8-4, along with Figure 8-4, detail the number of calls with one and two or more units arriving to a call, broken down by call type. In this analysis, we limit ourselves to calls where a unit from AMR arrives. For this reason, there are fewer calls in Table 8-4 than in Table 8-2.

TABLE 8-4: Calls by Call Type and Number of Units Arriving Call Type Number of Units Total Calls One Two
Breathing difficulty 780 7 787
Cardiac and stroke 848 4 852
Fall and injury 1,221 11 1,232
Illness and other 2,129 18 2,147
MVA 480 36 516
Overdose and psychiatric 227 4 231
Seizure and unconsciousness 818 7 825
EMS Total 6,503 87 6,590
Fire & FD assist 30 1 31
Total 6,533 88 6,621
Percentage 98.7 1.3 100.0

FIGURE 8-4: Calls by Number of Units Arriving

152
Observations:
■ On average, 1.0 units arrived at all calls
■ For 99 percent of calls, one unit arrived.
■ For 1 percent of calls, two or three units arrived.

153
WORKLOAD: RUNS AND TOTAL TIME SPENT
The workload of each AMR units is measured in two ways: runs and deployed time. Th e deployed time of a run is measured from the time a unit is dispatched through the time the unit is cleared. Because multiple units respond to some calls, there are more runs than calls and the average deployed time per run varies from the total duration of calls.

Runs and Deployed Time – All Units
Deployed time is the total deployment time of all units deployed on all runs. Table 8-5 shows the total deployed time, both overall and broken down by type of run, for all units in 2019.

Table 8-6 and Figure 8-5 present the average deployed minutes by hour of day.

TABLE 8-5: Annual Runs and Deployed Time by Run Type
Call Type
Deployed
Minutes
per Run
Annual
Hours
Percent
of Total
Hours
Deployed
Minutes
per Day
Annual
Runs
Runs
per
Day
Breathing difficulty 61.3 916.6 12.5 150.7 897 2.5
Cardiac and stroke 60.4 995.3 13.6 163.6 988 2.7
Fall and injury 54.7 1,342.3 18.3 220.7 1,472 4.0
Illness and other 50.8 2,395.0 32.6 393.7 2,826 7.7
MVA 35.8 480.6 6.6 79.0 805 2.2
Overdose and psychiatric 47.1 244.7 3.3 40.2 312 0.9
Seizure and unconsciousness 58.2 936.2 12.8 153.9 966 2.6 EMS Total 53.1 7,310.7 99.7 1,201.8 8,266 22.6
Fire & FD assist 18.9 25.1 0.3 4.1 80 0.2
Total 52.7 7,335.9 100.0 1,205.9 8,346 22.9
Observations:
■ The total deployed time for the year was 7,335.9 hours. The daily average was 20.1 hours for all units combined.
■ There were 8,346 runs. The daily average was 22.9 runs. ■ The average deployed time for EMS runs was 53.1 minutes per run. The deployed time for all EMS runs averaged 20.0 hours per day.

154
TABLE 8-6: Average Deployed Minutes by Hour of Day
Hour EMS Fire & FD
Assist Total
0 32.4 0.0 32.4
1 28.4 0.1 28.5
2 26.5 0.2 26.7
3 22.3 0.0 22.3
4 20.9 0.0 21.0
5 24.8 0.1 24.9
6 31.4 0.2 31.5
7 38.2 0.0 38.2
8 47.1 0.0 47.1
9 55.2 0.0 55.2
10 60.4 0.1 60.4
11 67.5 0.4 67.9
12 70.5 0.3 70.8
13 67.2 0.5 67.7
14 69.8 0.5 70.3
15 70.1 0.1 70.2
16 69.9 0.2 70.1
17 67.3 0.1 67.4
18 63.4 0.1 63.5
19 66.7 0.2 66.9
20 60.3 0.2 60.5
21 56.5 0.3 56.8
22 46.9 0.2 47.2
23 38.0 0.3 38.3
Daily Avg. 1,201.8 4.1 1,205.9

155
FIGURE 8-5: Average Deployed Minutes by Hour of Day

Observations:
■ Hourly deployed time was highest during the day from 9:00 a.m. to 9:00 p.m., averaging more than 65 minutes.
■ Average deployed time peaked between noon and 1:00 p.m., averaging 70.8 minutes. ■ Average deployed time was lowest between 4:00 a.m. and 5:00 a.m., averaging 21.0 minutes.

156
Workload by Unit
Tables 8-7 summarizes the overall workload of AMR’s ambulances in 2019. Tables 8 -8 and 8-9 provide a more detailed view of workload, showing each ambulance’s runs broken out by run type (Table 8-8) and the resulting daily average deployed time broken out by run type (Table 8-9). Here, we grouped the ambulances by SA and SD types. SA ambulances primarily responded to general 9-1-1 medic calls and SD ambulances primarily responded to BLS calls.

Additionally, we grouped together all SD ambulances that had less than seven total runs.

TABLE 8-1: Call Workload by Unit
Type Unit
Deployed
Minutes
per Run
Total
Hours
Total
Pct.

Deployed
Minutes
per Day
Total
Runs
Runs
per
Day
SA
AM254 24.1 11.6 0.2 1.9 29 0.1
AM255 13.0 7.8 0.1 1.3 36 0.1
AM256 82.4 30.2 0.4 5.0 22 0.1
AM257 26.3 40.4 0.6 6.6 92 0.3
AM401 57.6 98.9 1.3 16.3 103 0.3
AM402 51.4 14.6 0.2 2.4 17 0.0
AM411 53.0 210.4 2.9 34.6 238 0.7
AM412 52.0 246.4 3.4 40.5 284 0.8
AM413 38.7 87.0 1.2 14.3 135 0.4
AM414 44.3 280.5 3.8 46.1 380 1.0
AM415 48.3 286.4 3.9 47.1 356 1.0
AM416 49.2 557.9 7.6 91.7 680 1.9
AM417 54.1 2,218.3 30.2 364.7 2,460 6.7
AM418 55.9 2,012.5 27.4 330.8 2,160 5.9
AM419 49.3 109.3 1.5 18.0 133 0.4
AM420 49.9 185.6 2.5 30.5 223 0.6
AM492 45.6 49.4 0.7 8.1 65 0.2
AM493 58.0 238.0 3.2 39.1 246 0.7
AM494 82.0 5.5 0.1 0.9 4 0.0
AM495 53.4 188.8 2.6 31.0 212 0.6
AM496 56.8 225.5 3.1 37.1 238 0.7
AM980 55.9 42.8 0.6 7.0 46 0.1
AM985 55.9 24.2 0.3 4.0 26 0.1
Total 52.6 7,172.0 97.8 1,179.0 8,185 22.4
SD
AM202 60.8 7.1 0.1 1.2 7 0.0
AM205 81.2 10.8 0.1 1.8 8 0.0
AM231 57.7 7.7 0.1 1.3 8 0.0
AM239 37.2 4.3 0.1 0.7 7 0.0
Other* 61.3 133.9 1.8 22.0 131 0.4
Total 61.1 163.9 2.2 26.9 161 0.4
Total 52.7 7,335.9 100.0 1,205.9 8,346 22.9
Note: *“Other” is the group of SD ambulances that made less than seven total runs.

157
TABLE 8-8: Annual Runs by Run Type and Unit
Type Unit Breathing
Difficulty
Cardiac
and
Stroke
Fall
and
Injury
Illness
and
Other
MVA OD
Seizure
and
UNC
Fire &
FD
assist
Total
SA
AM254 2 4 3 11 4 2 3 0 29
AM255 2 11 3 9 5 2 4 0 36
AM256 2 4 3 9 2 1 1 0 22
AM257 11 7 16 36 8 2 11 1 92
AM401 0 2 1 12 1 0 1 0 17
AM402 21 37 45 71 30 11 23 0 238
AM411 22 27 51 90 39 9 41 5 284
AM412 11 17 27 42 20 3 14 1 135
AM413 36 52 65 124 44 12 47 0 380
AM414 43 41 69 99 42 10 47 5 356
AM415 76 79 118 213 65 21 102 6 680
AM416 268 300 425 877 211 102 250 27 2,460
AM417 267 259 387 687 214 74 250 22 2,160
AM418 22 10 23 42 13 6 16 1 133
AM419 28 25 35 75 20 9 28 3 223
AM420 6 6 16 21 3 3 9 1 65
AM492 15 33 47 76 28 7 39 1 246
AM493 0 2 2 0 0 0 0 0 4
AM494 21 22 41 69 20 7 29 3 212
AM495 24 23 48 80 22 3 36 2 238
AM496 4 9 14 11 4 0 4 0 46
AM980 3 2 6 5 5 2 2 1 26
AM985 2 10 10 68 1 8 3 1 103
Total 886 982 1,455 2,727 801 294 960 80 8,185
SD
AM202 0 0 0 7 0 0 0 0 7
AM205 0 0 1 7 0 0 0 0 8
AM231 0 0 1 6 0 1 0 0 8
AM239 2 0 1 2 1 1 0 0 7
Other 9 6 14 77 3 16 6 0 131
Total 11 6 17 99 4 18 6 0 161
Total 897 988 1,472 2,826 805 312 966 80 8,346
Note: OD=Overdose and psychiatric; UNC=Unconsciousness; “Other” is the group of SD ambulances that made less than seven total runs.

158
TABLE 8-9: Average Deployed Minutes by Run Type and Unit Type Unit Breathing
Difficulty
Cardiac
and
Stroke
Fall
and
Injury
Illness
and
Other
MVA OD
Seizure
and
UNC
Fire &
FD
assist
Total
SA
AM254 0.1 0.2 0.2 0.5 0.4 0.0 0.4 0.0 1.9
AM255 0.2 0.8 0.0 0.2 0.0 0.0 0.0 0.0 1.3
AM256 0.6 1.5 0.5 1.9 0.2 0.3 0.0 0.0 5.0
AM257 1.3 0.1 1.1 2.2 0.8 0.2 1.0 0.0 6.6
AM401 0.0 0.1 0.0 2.2 0.0 0.0 0.0 0.0 2.4
AM402 3.6 6.3 6.5 8.6 3.6 1.9 4.1 0.0 34.6
AM411 3.9 4.7 8.3 12.0 4.2 1.0 6.0 0.4 40.5
AM412 1.1 2.3 3.3 3.8 1.4 0.2 2.1 0.0 14.3
AM413 4.9 8.4 8.4 13.6 3.5 1.3 6.0 0.0 46.1
AM414 6.6 7.3 10.2 11.8 4.5 0.8 5.9 0.0 47.1
AM415 12.4 11.8 17.0 27.6 4.6 2.8 15.5 0.1 91.7
AM416 46.3 50.1 66.7 123.5 21.9 12.6 41.8 1.7 364.7
AM417 47.9 47.1 61.5 98.1 20.9 10.7 43.5 1.2 330.8
AM418 4.7 1.4 2.4 4.8 0.9 1.0 2.9 0.0 18.0
AM419 4.2 2.6 5.1 10.5 1.9 0.9 5.2 0.1 30.5
AM420 0.9 0.6 2.0 3.2 0.1 0.3 1.0 0.0 8.1
AM492 2.9 6.1 6.6 11.8 4.1 1.3 6.2 0.2 39.1
AM493 0.0 0.4 0.5 0.0 0.0 0.0 0.0 0.0 0.9
AM494 2.8 4.0 5.6 10.5 2.1 0.8 5.1 0.2 31.0
AM495 4.4 4.7 7.8 11.7 2.2 0.4 5.8 0.0 37.1
AM496 0.5 1.3 2.4 1.6 0.5 0.0 0.8 0.0 7.0
AM980 0.5 0.5 0.9 0.9 0.9 0.2 0.2 0.0 4.0
AM985 0.3 1.2 0.7 12.5 0.3 1.1 0.1 0.1 16.3
Total 150.1 163.4 217.8 373.6 78.9 37.7 153.5 4.1 1,179.0 SD
AM202 0.0 0.0 0.0 1.2 0.0 0.0 0.0 0.0 1.2
AM205 0.0 0.0 0.0 1.8 0.0 0.0 0.0 0.0 1.8
AM231 0.0 0.0 0.0 1.2 0.0 0.0 0.0 0.0 1.3
AM239 0.1 0.0 0.1 0.3 0.1 0.2 0.0 0.0 0.7
Other 0.5 0.2 2.8 15.7 0.0 2.3 0.4 0.0 22.0
Total 0.6 0.2 2.9 20.1 0.1 2.6 0.4 0.0 26.9
Total 150.7 163.6 220.7 393.7 79.0 40.2 153.9 4.1 1,205.9 Note: OD=Overdose and psychiatric; UNC=Unconsciousness; “Other” is the group of SD ambulances that made less than seven total runs.

159
Observations:
■ SA ambulances made 8,185 runs (22.4 runs per day) and had 7,172.0 hours of annual deployed time (19.6 hours per day).
■ SD ambulances made 161 runs (0.4 runs per day) and had 163.9 hours of annual deployed time (26.9 minutes per day).
■ Ambulance AM417 made the most runs (2,460, or an average of 6.7 runs per day) and had the highest total annual deployed time (2,218.3 hours or an average of 6.1 hours per day). ■ Ambulance AM418 made the second most runs (2,160, or an average of 5.9 runs per day) and had the second highest total annual deployed time (2,012.5 hours or an average of 5.5 hours per day).

160
Workload by District
The following table breaks down AMR’s annual workload by the servic e district of each NCFD fire station.

TABLE 8-10: Annual Workload by NCFD Station Service District NCFD
Station Calls
Pct.

Annual
Calls
Runs
Runs
Per
Day
Deployed
Minutes
Per Run
Annual
Hours
Pct.

Annual
Work
Deployed
Minutes
Per Day
31 3,350 45.7 3,785 10.4 55.1 3,477.3 47.4 571.6
34 3,978 54.3 4,561 12.5 50.8 3,858.5 52.6 634.3
Total 7,328 100.0 8,346 22.9 52.7 7,335.9 100.0 1205.9
Observations:

NCFD Station 31
■ There were 3,350 calls, or 46 percent of the total calls. ■ There were 3,785 runs. The daily average was 10.4 runs. ■ Total deployed time for the year was 3,477.3 hours or 47 percent of the total annual workload.

The daily average was 9.5 hours for all units combined.

NCFD Station 34
■ There were 3,978 calls, or 54 percent of the total calls. ■ There were 4,561 runs. The daily average was 12.5 runs. ■ Total deployed time for the year was 3,858.5 hours or 53 percent of the total annual workload.

The daily average was 10.6 hours for all units combined.

161

ANALYSIS OF BUSIEST HOURS
There is significant variability in the number of calls from hour to hour. One special concern relates to the resources available for hours with the heaviest workload. We tabulated the data for each of the 8,760 hours in the year. Table 8-11 shows the number of hours in the year in which there were zero to five or more calls during the hour. Table 8-12 shows the 10 one-hour intervals which had the most calls that AMR responded during the year. Table 8-13 examines the number of times a call overlapped with another call within the National City fire district.

TABLE 8-11: Frequency Distribution of the Number of Calls Calls in an Hour Frequency Percentage
0 3,928 44.8
1 3,025 34.5
2 1,266 14.5
3 419 4.8
4 101 1.2
5+ 21 0.2
Total 8,760 100.0

TABLE 8-12: Top 10 Hours with the Most Calls Received
Hour Number
of Calls
Number
of Runs
Total
Deployed Hours
11/15/2019, 2:00 p.m. to 3:00 p.m. 6 12 8.1
8/27/2019, 10:00 a.m. to 11:00 a.m. 6 7 9.0
6/21/2019, 5:00 p.m. to 6:00 p.m. 6 7 5.5
4/12/2019, 2:00 p.m. to 3:00 p.m. 6 6 6.7
3/10/2019, 4:00 p.m. to 5:00 p.m. 6 6 4.2
3/20/2019, 8:00 p.m. to 9:00 p.m. 5 7 9.4
4/23/2019, 5:00 p.m. to 6:00 p.m. 5 7 5.9
10/22/2019, 3:00 p.m. to 4:00 p.m. 5 6 12.0
5/28/2019, 5:00 p.m. to 6:00 p.m. 5 6 6.7
7/18/2019, 11:00 p.m. to midnight 5 6 5.9
Note: Total deployed hours is a measure of the total time spent responding to calls received in the hour. The deployed time from these calls may extend into the next hour or hours. The number of runs and deployed hours includes all AMR units.

162
TABLE 8-13: Frequency of Overlapping Calls
Scenario Number of
Calls
Percent of
All Calls Total Hours
No overlapped call 3,064 41.8 2,977.7
Overlapped with one call 2,540 34.7 1,274.6
Overlapped with two calls 1,177 16.1 390.3
Overlapped with three calls 393 5.4 98.8
Overlapped with four calls 123 1.7 22.7
Overlapped with five calls 24 0.3 4.3
Overlapped with six calls 5 0.1 1.3
Overlapped with seven calls 2 0.0 0.2
Observations:
■ During 21 hours (0.2 percent of all hours), five or more calls occurred; in other words, AMR responded to five or more calls in an hour roughly once every 17 days. □ The highest number of calls to occur in an hour was six, which happened five times.

163

RESPONSE TIME
In this part of the analysis, we present response time statistics for different call types. We separate response time into its identifiable components. Dispatch time is the difference between the time when AMR received a call and the earliest time an ambulance is dispatched. Dispatch time includes call processing time, which is the time required to determine the nature of the emergency and the types of resources to dispatch. Turnout time is the difference between the earliest dispatch time and the earliest time an ambulance is en route to a call’s location. Travel time is the difference between the earliest en route time and the earliest arrival time. Response time is the total time elapsed between receiving a call to arriving on scene.

In this analysis, with all calls that were responded by AMR within the National City fire district, we excluded the fire & PD assist calls. In addition, calls with a total response time of more than 30 minutes were excluded. Finally, we focused on units that had complete time stamps, that is, units with all components recorded, so that we could calculate each segment of response time.

Based on the methodology above, we excluded 73 fire & PD calls, four non-emergency calls, 659 calls where no units recorded a valid on-scene time, 30 calls where the first arriving unit’s response time was greater than 30 minutes, and 14 calls where one or more segments of the first arriving unit’s response time could not be calculated due to missing or faulty data. As a result, the analysis in this section included 6,548 calls for 2019. Using the same method, we obtained 6,214 calls for the same analysis for 2020. 2020’s response time analysis is compared with that of 2019 in Attachment I.

164
Response Time by Type of Call
Table 8-14 breaks down the average dispatch, turnout, travel, and total response times by call type for all calls that AMR responded within the National City fire district, and Table 8 -15 does the same for 90th percentile response times. A 90th percentile response time means that 90 percent of calls had response times at or below that number. For example, Table 8-15 shows a 90th percentile response time of 13.2 minutes, which means that 90 percent of the time, a call had a response time of no more than 13.2 minutes. Figure 8-6 illustrates the components of the average response time.

TABLE 8-14: Average Response Time of First Arriving Unit, by Call Type Call Type Time in Minutes Number
of Calls Dispatch Turnout Travel Total
Breathing difficulty 0.7 0.8 5.9 7.4 786
Cardiac and stroke 0.8 0.8 6.2 7.7 851
Fall and injury 0.9 0.7 6.4 8.0 1,227
Illness and other 1.0 0.8 6.7 8.6 2,125
MVA 1.0 0.8 6.2 8.0 508
Overdose and psychiatric 1.0 0.9 6.7 8.6 225
Seizure and unconsciousness 0.8 0.8 6.1 7.7 826
Total 0.9 0.8 6.4 8.0 6,548

FIGURE 8-6: Average Response Time of First Arriving Unit, by Call Type

165
TABLE 8-15: 90th Percentile Response Time of First Arriving Unit, by Call Type Call Type Time in Minutes Number
of Calls Dispatch Turnout Travel Total
Breathing difficulty 1.6 1.8 10.1 11.6 786
Cardiac and stroke 2.0 1.8 10.7 12.6 851
Fall and injury 2.1 1.8 10.8 12.8 1,227
Illness and other 3.1 1.8 11.6 14.9 2,125
MVA 2.4 1.7 10.9 12.8 508
Overdose and psychiatric 3.1 2.0 11.7 14.2 225
Seizure and unconsciousness 2.0 1.7 10.4 12.2 826
Total 2.4 1.8 10.9 13.2 6,548
Observations:
■ The average dispatch time was 0.9 minutes.
■ The average turnout time was 0.8 minutes.
■ The average travel time was 6.4 minutes.
■ The average total response time was 8.0 minutes.
■ The 90th percentile dispatch time was 2.4 minutes.
■ The 90th percentile turnout time was 1.8 minutes.
■ The 90th percentile travel time was 10.9 minutes.
■ The 90th percentile total response time was 13.2 minutes.

166
Response Time by Hour
The components of average response time by the time of day are shown in Table 8-16. The table also shows the 90th percentile response time. Figure 8-7 shows the same information.

TABLE 8-16: Average and 90th Percentile Response Time of First Arriving Unit, by Time of Day
Hour
Time in Minutes 90th Percentile
Response Time
Number
of Calls Dispatch Turnout Travel Response
Time
0 0.9 1.1 6.3 8.4 13.5 168
1 0.9 1.3 6.5 8.7 14.3 160
2 0.9 1.4 5.7 8.0 13.3 153
3 1.0 1.7 6.1 8.7 13.4 130
4 0.9 1.4 6.1 8.3 13.5 130
5 0.7 1.2 6.3 8.3 12.6 155
6 0.7 1.6 6.4 8.7 14.3 188
7 0.9 1.0 6.4 8.3 13.6 232
8 0.9 0.8 6.2 7.9 13.5 312
9 0.9 0.6 6.2 7.7 12.1 303
10 1.0 0.7 6.2 7.9 13.2 350
11 0.9 0.5 6.3 7.7 12.3 369
12 0.9 0.5 6.5 8.0 13.3 357
13 0.8 0.6 6.5 7.8 12.8 363
14 0.9 0.5 6.4 7.7 12.5 367
15 1.0 0.5 6.7 8.2 12.4 363
16 0.8 0.6 6.4 7.8 12.5 358
17 1.1 0.5 6.6 8.2 14.1 354
18 1.0 0.6 5.9 7.6 12.4 338
19 0.9 0.6 6.4 7.9 13.0 344
20 1.0 0.6 6.4 8.0 12.8 323
21 0.9 0.8 6.6 8.3 13.6 291
22 0.9 1.0 6.4 8.3 13.5 223
23 0.8 1.0 6.8 8.6 13.3 217
Total 0.9 0.8 6.4 8.0 13.2 6,548

167
FIGURE 8-7: Average Response Time of First Arriving Unit, by Hour of Day

Observations:
■ Average dispatch time was between 0.7 minutes (5:00 a.m. to 6:00 a.m.) and 1.1 minutes (5:00 p.m. to 6:00 p.m.).
■ Average turnout time was between 0.5 minutes (3:00 p.m. to 4:00 p.m.) and 1.7 minutes (3:00 a.m. to 4:00 a.m.).
■ Average travel time was between 5.7 minutes (2:00 a.m. to 3:00 a.m.) and 6.8 minutes (11:00 p.m. to midnight).
■ Average response time was between 7.6 minutes (6:00 p.m. to 7:00 p.m.) and 8.7 minutes (3:00 a.m. to 4:00 a.m.).
■ The 90th percentile response time was between 12.1 minutes (9:00 a.m. to 10:00 a.m.) and

14.3 minutes (1:00 a.m. to 2:00 a.m.).

168
Response Time Distribution
Here, we present a more detailed look at how response times to calls are distributed. The cumulative distribution of total response time for the first arriving unit is shown in Figure 8-8 and Table 8-17. Figure 8-8 shows response times for the first arriving unit as a frequency distribution in whole-minute increments.

The cumulative percentages here are read in the same way as a percentile. In Figure 8 -8, the 90th percentile of 13.2 minutes means that 90 percent of calls had a response time of 13.2 minutes or less. In Table 8-17, the cumulative percentage of 61.8 means that 61.8 percent of calls had a response time under 8 minutes.

FIGURE 8-8: Cumulative Distribution of Response Time – First Arriving Unit

169
TABLE 8-17: Cumulative Distribution of Response Time – First Arriving Unit Response Time
(minute) Frequency Cumulative
Percentage
1 13 0.2
2 33 0.7
3 150 3.0
4 360 8.5
5 721 19.5
6 918 33.5
7 990 48.6
8 863 61.8
9 586 70.8
10 430 77.3
11 315 82.1
12 272 86.3
13 221 89.7
14 152 92.0
15 118 93.8
16+ 406 100.0
Observations:
■ For 62 percent of calls, the response time of the first arriving unit was less than 8 minutes.

170

TRANSPORT CALL ANALYSIS
In this section, we present an analysis for unit activity that involved transporting patients, the variations by hour of day, and the average time for each stage of transport service. We identified transport calls by requiring that at least one responding unit had recorded both a “beginning to transport” time and an “arriving at the hospital” time. Based on these criteria, we note that eight non-EMS (fire & FD assist) calls that resulted in transports are included in this analysis.

Transport Calls by Type
Table 8-18 shows the number of calls by call type broken out by transport and non-transport calls.

TABLE 8-18: Transport Calls by Call Type
Call Type Number of Calls Conversion
Rate Non-transport Transport Total
Breathing difficulty 167 648 815 79.5
Cardiac and stroke 183 698 881 79.2
Fall and injury 458 838 1,296 64.7
Illness and other 846 1,607 2,453 65.5
MVA 422 255 677 37.7
Overdose and psychiatric 116 150 266 56.4
Seizure and unconsciousness 232 635 867 73.2
EMS Total 2,424 4,831 7,255 66.6
Fire & FD assist 65 8 73 11.0
Total 2,489 4,839 7,328 66.0
Observations:
■ 67 percent of EMS calls involved transporting one or more patients ■ On average, 13 EMS calls per day involved transporting one or more patients.

171
Average Transport Calls per Hour
Table 8-19 and Figure 8-9 show the average number of EMS calls received each hour of the day during 2019. In the table the conversion rate measures the percent of EMS calls in which one or more patients was transported.

TABLE 8-19: EMS Transport Calls per Hour, by Time of Day Hour EMS
Calls Transport EMS Calls
per Day
Transports
per Day
Conversion
Rate
0 190 122 0.5 0.3 64.2
1 183 121 0.5 0.3 66.1
2 164 102 0.4 0.3 62.2
3 141 85 0.4 0.2 60.3
4 143 98 0.4 0.3 68.5
5 171 123 0.5 0.3 71.9
6 208 140 0.6 0.4 67.3
7 260 171 0.7 0.5 65.8
8 348 245 1.0 0.7 70.4
9 340 241 0.9 0.7 70.9
10 379 268 1.0 0.7 70.7
11 405 291 1.1 0.8 71.9
12 393 262 1.1 0.7 66.7
13 398 266 1.1 0.7 66.8
14 403 267 1.1 0.7 66.3
15 399 261 1.1 0.7 65.4
16 401 264 1.1 0.7 65.8
17 400 243 1.1 0.7 60.8
18 379 257 1.0 0.7 67.8
19 365 241 1.0 0.7 66.0
20 366 239 1.0 0.7 65.3
21 318 212 0.9 0.6 66.7
22 258 148 0.7 0.4 57.4
23 243 164 0.7 0.4 67.5
Total 7,255 4,831 19.9 13.2 66.6
Note: The conversion rate is measured by dividing the number of EMS transports by the number of EMS calls. For example, between midnight and 1:00 a.m., there were 122 EMS transports out of 190 EMS calls. This gives a conversion rate of 122 / 190 = 0.642, or 64.2 percent.

172
FIGURE 8-9: Average Transport Calls by Hour of Day

Observations:
■ Hourly EMS calls per day were highest during the day from 8:00 a.m. to 9:00 p.m., averaging between 0.9 and 1.1 calls per day.
■ Average hourly EMS calls per day peaked between 11:00 a.m. and noon, averaging 1.1 calls per day.
■ Average hourly EMS calls per day was lowest between 3:00 a.m. and 4:00 a.m., averaging

0.4 calls per day.
■ Hourly transport calls per day were highest during the day from 8:00 a.m. to 8:00 p.m., averaging between 0.7 calls per day and 0.8 calls per day. ■ Average hourly transport calls per day peaked between 11:00 a.m. and noon, averaging 0.8 calls per day.
■ Average hourly transport calls per day was lowest between 3:00 a.m. and 4:00 a.m., averaging 0.2 calls per day.
■ Average hourly transport conversion rates per day peaked between 5:00 a.m. and 6:00 a.m., averaging 72 percent per day.
■ Average hourly transport conversion rates per day was lowest between 10:00 p.m. and 11:00 p.m., averaging 57 percent per day.

173
Calls by Type and Duration
The following table shows the average duration of transport and non-transport EMS calls by call type.

TABLE 8-20: Transport Call Duration by Call Type
Call Type
Non-transport Transport
Average
Duration
Number of
Calls
Average
Duration
Number of
Calls
Breathing difficulty 34.4 167 75.2 648
Cardiac and stroke 33.6 183 76.2 698
Fall and injury 29.0 458 79.2 838
Illness and other 23.2 846 76.3 1,607
MVA 16.2 422 78.6 255
Overdose and psychiatric 28.3 116 74.7 150
Seizure and unconsciousness 31.1 232 76.8 635
EMS Total 25.6 2,424 76.8 4,831
Fire & FD assist 12.0 65 82.2 8
Total 25.3 2,489 76.8 4,839
Note: Duration of a call is defined as the longest deployed time of any of the units responding to the same call.

Observations:
■ The average duration was 25.6 minutes for non-transport EMS calls. ■ The average duration was 76.8 minutes for EMS calls where one or more patients were transported to a hospital.

174
Transport Time Components
Table 8-21 gives the average deployed time for an ambulance on a transport call , along with three major components of the deployed time: on-scene time, travel to hospital time, and at- hospital time.

The on-scene time is the interval from the unit arriving on-scene time through the time the unit departs the scene for the hospital. Travel to hospital time is the interval from the time the unit departs the scene to travel to the hospital through the time the unit arrives at the hospital. At - hospital time is the time it takes for patient turnover at the hospital.

This table analyzes times by run. Normally, the number of runs will exceed the number of calls as a call may have multiple runs. In addition, average times may differ slightly from similar averages measured per call.

TABLE 8-21: Time Component Analysis for Ambulance Transport Runs by Call Type
Call Type
Average Minutes Spent per Run Number of
Runs On
Scene
Traveling
to Hospital
At
Hospital Deployed
Breathing difficulty 16.0 13.4 39.0 74.9 649
Cardiac and stroke 16.0 13.7 38.7 75.3 698
Fall and injury 17.9 15.3 38.1 78.5 842
Illness and other 16.5 13.8 37.9 75.8 1,612
MVA 13.8 16.1 39.4 76.6 279
Overdose and psychiatric 15.8 11.2 39.4 73.4 151
Seizure and unconsciousness 15.9 13.2 40.4 76.5 637
EMS Total 16.4 14.0 38.6 76.1 4,868
Fire & Other Total 17.0 16.8 42.8 82.0 8
Total 16.4 14.0 38.7 76.1 4,876
Note: Average unit deployed time per run is lower than average call duration for some call types because call duration is based on the longest deployed time of any of the units responding to the same call, which may include an engine or ladder. Total deployed time is greater than the combination of on-scene, transport, and hospital wait times as it includes turnout, initial travel, and hospital return times.

Observations:
■ The average time spent on-scene for a transport EMS call was 16.4 minutes. ■ The average travel time from the scene of the EMS call to the hospital was 14.0 minutes. ■ The average deployed time spent on transport EMS calls was 76.1 minutes. ■ The average deployed time at the hospital was 38.6 minutes, which accounts for approximately 51 percent of the average total deployed time for a transport EMS call.

175
ATTACHMENT I: 2019 & 2020 COMPARISON
In this analysis, we compare portions of the previous analysis with similar records for 2020. We compare calls by type, unit workload, response time, and transport workload over the two years.

Call Volume by Year
Table 8-22 shows the number of calls by call type for both 2019 and 2020. Figure 8-10 shows the monthly variation in the average daily number of calls in two years. Similarly, Figure 8-11 illustrates the average number of calls received each hour of the day in two years.

TABLE 8-22: Calls by Call Type and Year
Call Type Number of Calls Calls per Day
2019 2020 2019 2020
Breathing difficulty 815 758 2.2 2.1
Cardiac and stroke 881 864 2.4 2.4
Fall and injury 1,296 1,229 3.6 3.4
Illness and other 2,453 2,421 6.7 6.6
MVA 677 589 1.9 1.6
Overdose and psychiatric 266 286 0.7 0.8
Seizure and unconsciousness 867 726 2.4 2.0
EMS Total 7,255 6,873 19.9 18.8
Fire & FD assist 73 72 0.2 0.2
Total 7,328 6,945 20.1 19.0
Observations:
■ The call volume decreased five percent, from 7,328 in 2019 to 6,945 in 2020.

176
FIGURE 8-10: Calls per Day by Month and Year

FIGURE 8-11: Calls per Hour by Time of Day and Year

177
Workload by Year
Table 8-23 compares the runs and workload for AMR units in 2019 and 2020. In the table, all SD type units are grouped. Figure 8-12 compares the average deployed minutes by the hour of the day in 2019 and 2020.

TABLE 8-23: Workload by Unit and Year
Type Unit Total Runs Runs per Day Total Hours Deployed Minutes per Day
2019 2020 2019 2020 2019 2020 2019 2020
SA
AM254 29 49 0.1 0.1 11.6 26.4 1.9 4.3
AM255 36 35 0.1 0.1 7.8 14.8 1.3 2.4
AM256 22 19 0.1 0.1 30.2 28.5 5.0 4.7
AM257 92 113 0.3 0.3 40.4 66.4 6.6 10.9
AM401 103 78 0.3 0.2 98.9 84.5 16.3 13.8
AM402 17 16 0.0 0.0 14.6 15.4 2.4 2.5
AM411 238 189 0.7 0.5 210.4 171.6 34.6 28.1
AM412 284 232 0.8 0.6 246.4 182.2 40.5 29.9
AM413 135 187 0.4 0.5 87.0 117.3 14.3 19.2
AM414 380 396 1.0 1.1 280.5 301.0 46.1 49.3
AM415 356 326 1.0 0.9 286.4 266.3 47.1 43.7
AM416 680 641 1.9 1.8 557.9 514.0 91.7 84.3
AM417 2,460 2,352 6.7 6.4 2,218.3 1,983.9 364.7 325.2
AM418 2,160 2,097 5.9 5.7 2,012.5 1,713.1 330.8 280.8
AM419 133 280 0.4 0.8 109.3 221.1 18.0 36.3
AM420 223 267 0.6 0.7 185.6 191.1 30.5 31.3
AM492 65 56 0.2 0.2 49.4 38.7 8.1 6.3
AM493 246 166 0.7 0.5 238.0 141.7 39.1 23.2
AM494 4 16 0.0 0.0 5.5 10.9 0.9 1.8
AM495 212 99 0.6 0.3 188.8 87.1 31.0 14.3
AM496 238 175 0.7 0.5 225.5 164.6 37.1 27.0
AM980 46 14 0.1 0.0 42.8 10.1 7.0 1.7
AM985 26 0 0.1 0.0 24.2 0.0 4.0 0.0
Total 8,185 7,803 22.4 21.3 7,172.0 6,350.7 1,179.0 1,018.8 SD Total 161 208 0.4 0.6 163.9 211.2 26.9 34.6
Total 8,346 8,011 22.9 21.9 7,335.9 6,561.9 1,205.9 1,075.7 Observations:
■ The total runs decreased 4 percent from 8,346 in 2019 to 8,011 in 2020. ■ The total work hours decreased 11 percent from 7,335.9 hours in 2019 to 6,561.9 hours in 2020.

178
FIGURE 8-12: Average Deployed Minutes by Hour of Day in 2019 and 2020

179
Response Time Comparison by Year
Tables 8-24 compares the average and 90th percentile response times broken out by call type and year. Figure 8-13 compares 2019’s and 2020’s average response time by hour of day.

TABLE 8-24: Average Response Time of First Arriving Unit by Call Type and Year Call Type
2019 2020
Average 90th
Percentile Calls Average 90th
Percentile Calls
Breathing difficulty 7.4 11.6 786 7.8 12.6 727
Cardiac and
stroke 7.7 12.6 851 7.8 13.1 825
Fall and injury 8.0 12.8 1,227 8.2 13.1 1,131
Illness and other 8.6 14.9 2,125 8.9 14.8 2,145
MVA 8.0 12.8 508 8.1 13.1 454
OD 8.6 14.2 225 9.1 14.8 257
Seizure and UNC 7.7 12.2 826 7.8 12.3 675
Total 8.0 13.2 6,548 8.3 13.5 6,214
Note: OD= Overdose and psychiatric; UNC=Unconsciousness.

FIGURE 8-13: Average Response Time of First Arriving Unit, by Hour of Day and Year

180
Observations:
■ The response times in two years did not change significantly.

TRANSPORT COMPARISON BY YEAR
Table 8-25 compares the transport calls and workload in 2019 and 2020. Figure 8-14 compares the average number of EMS and transport EMS calls received each hour of the day over the two-year period.

TABLE 8-25: Transport Calls and Workload by Call Type and Year Call Type
2019 2020
Calls Runs
Average
Call Duration
(Minutes)
Calls Runs
Average Call
Duration
(Minutes)
Breathing difficulty 648 649 75.2 569 569 72.2
Cardiac and stroke 698 698 76.2 625 626 71.7
Fall and injury 838 842 79.2 701 704 73.8
Illness and other 1,607 1,612 76.3 1,516 1,522 75.8
MVA 255 279 78.6 206 232 74.4
OD 150 151 74.7 166 167 72.4
Seizure and UNC 635 637 76.8 493 493 72.1
EMS Total 4,831 4,868 76.8 4,276 4,313 73.8
Fire & FD assist 8 8 82.2 4 6 66.2
Total 4,839 4,876 76.8 4,280 4,317 73.7
Note: OD= Overdose and psychiatric; UNC=Unconsciousness

181
FIGURE 8-14: Average Transport Calls by Hour and Year

182
ATTACHMENT II: CALL TYPE IDENTIFICATION
TABLE 8-26: Call Type by CAD Problem Description
Call Type Problem Frequency
2019 2020
Breathing
Difficulty
Breathing Problems 781 723
Choking 34 35
Cardiac and
Stroke
Cardiac / Respiratory Arrest 110 142
Chest Pain 485 465
Heart Problems 112 114
Stroke 174 143
Fire & PD Assist
Burns / Explosion 7 4
.Nat Gas Leak Broken/Blowing 0 1
.Natural Gas Odor - Outside 0 1
AID - MEDIC 1 0
Assist PD 3 2
Carbon Monoxide Alarm 12 6
Electrical Short 1 1
Extinguished Fire 1 0
Fuel Spill 4 4
HazMat 1 0
HazMat 1st Alarm 0 1
HazMat Single Engine 3 2
Illegal Burn 2 0
Investigate 1 0
Knocked Off Hydrant 1 0
Lift Assist 1 0
Lock in/out 3 2
Odor of Chemical 0 1
Oven Fire 1 0
Ringing Alarm Highrise 0 1
Rubbish Fire 1 2
Safe Surrender 0 1
SNAKE REMOVAL 1 0
Special Service 2 1
Structure Collapse 1 2
Structure Fire - Comm / Apt 8 21
Structure Highrise/Hospital 0 1
Structure Residential 10 11
Vegetation Initial Attack 1 2
Vehicle Fire 4 1
Vehicle Fire Freeway 2 4
Wires down 1 0

183
Call Type Problem Frequency
2019 2020
Fall
and
Injury
Assault/Rape 227 238
Drowning/Diving Accident 1 1
Electrocution 4 1
Falls / Back Inj 855 787
Stabbing/Gunshot 34 36
Traumatic Injuries, Spec 175 166
Illness and
Other
Abdominal Pain/Problems 209 222
Allergy/Hives/Med Rx/Stng 43 48
Animal Bites/ Attacks 14 13
Back Pain 75 67
C O / Inhalation/ Haz Mat* 3 4
Diabetic Problems 151 139
Elevator Rescue 12 9
Eye Problems / Injuries 3 8
Headache 61 43
Heat / Cold Exposure 6 6
Hemorrhage / Lacerations 227 237
Industrial Rescue 0 1
Lift Assist* 1 1
Medical Aid 7 4
Medical Alert Alarm 95 76
Miscellaneous Rescue 0 1
NC Medical Aid 53 51
Poison Control 2 1
Preg/Birth/Miscarriage 29 29
Sick Person 1,246 1,158
Special Service* 0 3
Suspected COVID19 0 108
Unknown Problem* 189 162
Vehicle vs. Pedestrian* 5 7
Vehicle Rescue 22 22
Water Rescue 3 0 1
MVA
Traffic Accident 589 529
Traffic Accident FWY 74 50
Vehicle vs Structure 13 9
Vehicle vs. Pedestrian 1 1
Overdose and
Psychiatric
OD/Ingestion/Poisonings 123 113
Psych / Suicide Attempt 143 173
Seizure and
UNC
Convulsions / Seizures 285 227
Unc/Fainting 582 499
Total 7,328 6,945

184
Note: *NRIFS incident type code is 321; UNC = Unconsciousness.

- END -

NIST TechnicalN ote 1661
Jason D. Averill
LoriMoo re-Merrell
Adam Ba rowy
Ro bertSantos
RichardPeacock
Kathy A.No tarianni
Do ug Wissoker
E dited by BillRo binson
U.S.D epartmen tofCommerce
Gary Locke,Secretary
Na tionalInstituteofStandardsand Technology
Patrick D. Gallagher,Director
April2 010
1 APPENDIX CPage 910 of 1153
2 Page 911 of 1153
April2 010
NIST TechnicalNo te 166 1
Rep orton Residential
Fireground FieldExperiments
Jason D. Averill
LoriMo ore-Merrell
Adam Barowy
Robe rtSantos
RichardPeacock
Kathy A.Notarianni
Doug W issoker
Edited by BillRob inson
U.S.D epartmentofCommerce
Ga ryLocke,Secretary
Na tionalInstitute ofStandardsand Technology
Patrick D. Gallagher,Director
3Page 912 of 1153
Certaincomm ercialentities,equipment,ormaterialsmay
be identified inthisdocume ntinordertodescribe an
experimentalprocedureorconceptadequately.Such
identification isnotintended toimply recomm endationor
endorseme ntby the N ationalInstituteofStandardsand
Technology,norisitintended toimply thattheentities,
materials,orequipm entare necessarily thebestavailable
forthe purpose.
4 Page 913 of 1153
NationalInstituteofStandards and TechnologyTechnicalNote 1661 , 104 page s ( Ma rch 2010 )CODE N:

Produced w ith the Cooperation of
Mo ntgome ryCounty
Fire and Re scue
ChiefRichard B ow ers
Produced with the Cooperation of
Fairfax County
Fire and R escue
ChiefRo nald Ma stin
Funding provided through DH S/FE MA GrantProgram DirectorateforFY 2 008 Assistance to FirefightersGrantProgram – FirePreventionand SafetyG rants (EMW- 20 08-FP-016 03)
5Page 914 of 1153
6

TableofContents
Ab stract ................................................................................................ 9 Executive Sum mary .......................................................................... 10 Background ........................................................................................ 12 Problem .............................................................................................. 13 Review ofLiterature .......................................................................... 14 Purpose and Scope oftheStudy .................................................... 16 A BriefOv erview ofFireDepartmentFiregroundOperations .... 17 The Relation ofTime-to-T askCompletionand Risk .......................... 18 Standards ofResponseCover ............................................................ 18 Part1: PlanningfortheField Experiments .................................... 20 Part2:Time-to-Task Experime nts .................................................. 24 Field ExperimentMethods .............................................................. 21 FieldSite .............................................................................................. 21 O verviewofFieldE xperiments .......................................................... 22 Instrumentation .................................................................................... 22 SafetyProtocols .................................................................................. 23 CrewSize ............................................................................................ 24 DepartmentParticipation .................................................................... 24 CrewOrientation .................................................................................. 24 T asks .................................................................................................... 25 Data Collection:StandardizedControlMeasures .............................. 27 T ask Flow Chartsand CrewCue Cards ............................................ 27 Radiocomm unications ........................................................................ 27 T ask Timers .......................................................................................... 27 Video records ...................................................................................... 27 CrewAssignm ent ................................................................................ 28 ResponseTime Assumptions .............................................................. 28 Part3:Room and C ontentsFires .................................................... 29 FuelPackages fortheRoom and ContentsFires .............................. 29 ExperimentalM atrixforRoom and ContentsFires ............................ 30 Procedure forM inimizing the EffectofVariance inFireGrow th Rate .......... 31 Ana lysis ofExperimentalResults .................................................. 33 Time-to-taskAnalysis .......................................................................... 33 Da ta Q ueries ........................................................................................ 33 StatisticalM ethods -TimetoT ask ...................................................... 33 Regression Analysis ............................................................................ 33 M easurem entUncertainty .................................................................. 34 Ho w to InterpretTime-to-T askGraphs ................................................ 34 Time-to-T askG raphs .......................................................................... 35 Part4:FireM odeling ........................................................................ 43 Time to UntenableConditions:R esearchQuestions ........................ 45 Fire M odeling M ethods ........................................................................ 45 Fire G rowth R ates ................................................................................ 46 FractionalEffective D ose (FED) .......................................................... 47 Results from M odelingMethods ........................................................ 48 InteriorFirefighting Conditionsand DeploymentConfiguration ........ 49 PhysiologicalEffectson Firefighters:Co m parison by C rew Size ........ 5 0 Study Limitations .............................................................................. 51 Conc lusions ........................................................................................ 52 FutureResearch ................................................................................ 53 Acknow ledgme nts ............................................................................ 55 References .......................................................................................... 56 7Page 916 of 1153
8 Page 917 of 1153
S
ervice expectations placed on the fireservice,including Em ergency Me dicalServices (EMS),responseto natural
disasters,hazardous mate rialsincidents,and actsof
terrorism,have steadilyincreased.Ho wever,local
decision-makers arechallenged to balance thesecomm un ity service expectations with finiteresourceswithout a solidtechnical foundation for evaluating the impac tofstaffingand deploym ent decisions on thesafetyof the pub licand firefighters.

For thefirsttim e,thisstudyinvestigates the effectof varying crew size,firstapparatusarrivaltime,and response time on firefighter safety,overalltask com pletion,and interior residential tenability using realisticresidentialfires.This study isalsounique becauseofthearray of stakeholdersand the caliberof technical expertsinvolved.A dd itionally,the structureused in the field experime ntsincluded customi zed instrumentation;allrelated industry standards were followed; and robustresearch met hods we reused.The resultsand conclusions willdirectlyinform the NPFA 1710T echnicalCo mmi ttee,w ho isresponsiblefor
developing consensusindustry deploymen tstandards.

T his repo rtpresents the results ofmo rethan 60 laboratory and residentialfiregroun d experiments designed to quantify the effectsofvarious firedepartm entdeployme nt configurations on themo stcom mo n type offire— a low hazard residential
structurefire.For the firegroun d experim ents,a 2,000sq ft(186 m 2),two-story residentialstructure w as designed and builtatthe M on tgome ry C oun tyPub lic Safety Training A cademy in R ockville,MD .Firecrews from M on tgome ry C oun ty,M D and Fairfax C oun ty,V Aw eredeployed in responseto livefireswithin this facility.In addition to system atically controlling for the arrivaltim esofthe firstand subsequent fireapp aratus,crew size w asvaried to consider tw o-,three-,four-,and five-person staffing.

Each deploym entperform ed a seriesof22tasks thatw eretimed, whilethe therm aland toxic environ m entinside the structurewa s m easured.A dd itionalexperiments with largerfuelloads asw ellas firemo deling prod uced addition alinsight.Re portresultsquantify theeffectiveness ofcrew size,first-due engine arrivaltime,and apparatus arrivalstagger on the duration and tim e to com pletion of the key 22firegroun d tasks and the effecton occupa nt and firefighter safety.

Abs tract
9Page 918 of 1153
B
oththe increasing demands on thefireservice-suchasthe
grow ing numbe rofEmergency Medi calServices(EMS)
responses,challengesfrom naturaldisasters,hazardous
ma terialsincidents,and actsofterrorism — and previousresearch pointtothe need forscientifically based studiesoftheeffectof differentcrew sizesand firefighterarrivaltimeson theeffectivenessof thefireservicetoprotectlivesand property.T omeet thisneed,a researchpartnership oftheCo mmission on FireAccreditation International(CF AI),InternationalAssociationofFireChiefs(IAFC), InternationalAssociation ofFirefighters(IAFF),NationalInstituteof Standardsand T echnology (NIST),and W orcesterPolytechnic Institute (WPI)w asformed to cond ucta multiphasestudy ofthe deploym entofresourcesasitaffectsfirefighterand occupantsafety.

Starting inFY 2005,fund ingwas provided throughthe Departmentof Hom eland Security(DHS)/FederalEme rgencyManagement Ag ency (FEMA) GrantProgram DirectorateforA ssistanceto Firefighters GrantProgram — FirePrevention and SafetyGrants.In additionto thelow-hazardresidentialfireground experimentsdescribed inthis report,the mu ltiplephasesoftheoverallresearcheffortinclude developmentofa conceptualmod elforcommu nityriskassessment and deploymentofresources,impl eme ntation ofa generalizable departmen tincidentsurvey,and deliveryofa softwaretooltoquantify theeffectsofdeploymentdecisions on resultantfirefighterand civilian injuriesand on propertylosses.

Th e firstphaseoftheprojectwas an extensivesurvey of more than 400careerand combin ation (both careerand volunteer)fire departm entsintheUnitedStatesw ith theobjectiveof optimizinga fireserviceleader’scapability to deploy resourcestopreventor mitigateadverse eventsthatoccur inrisk-and hazard-filled environm ents.T he resultsof thissurvey arenot docu mented inthis report,which islimited tothe experimentalphaseof theproject.

Th e survey resultswillconstitutesignificantinputintothe developm entof a future softw aretooltoquantifytheeffectsof comm un ityrisks and associated deploymen tdecisionson resultant firefighterand civilianinjuriesand propertylosses.

T he following researchquestions guided the experimental designofthelow-hazard residentialfireground experimen ts do cumen tedin thisreport:

1.H ow do crew sizeand staggeraffectoverallstart-to-completion response timi ng?

2.H ow do crew sizeand staggeraffectthe timin gs of task initiation,task du ration,and task completion for each ofthe 22 criticalfireground tasks?

3.H ow does crew sizeaffectelapsed timesto achievethreecritical eventsthatare known to change firebehavior or tenability within thestructure:
a.Entryinto structure?
b.W ater on fire?
c.V entilation through windows (three upstairsand one back downsta irs win dow and theburn room window ).

4.Ho w does the elapsed time to achieve the nationalstandard of assem bling 15firefightersatthe scene vary between crew sizes of four and five?
In order to address the prima ry research questions,theresearch w asdivided into four distinct,yetinterconn ected parts:
/square6Part1— Laboratoryexperime ntsto design appropriatefuelload /square6Part2— Experiments to m easure the tim e for various crew sizes and apparatus stagger(intervalbetween arrivalof
various app aratus)to accomp lish key tasksin rescuing
occupa nts,extinguishing a fire,and protecting property /square6Part3— A dd ition alexperiments with enh anced fuelload that prohibited firefighter entry into the burn prop – a building constructed for the fireexperime nts
/square6Part4— Firem odeling to correlate time -to-task com pletion by crew size and stagger to the increasein toxicity ofthe atmo sph erein the burn prop for a range offiregrowth rates.

Th e experime nts we recon ducted in a burn prop designed to simulate a low -hazard 1 firein a residentialstructure described as typicalin NFPA 1710® O rganization and Dep loymen tofFire Suppression O perations,Em ergencyM edicalO perations,and Special O perationsto the Publicby C areerFire D epartme nts .NFPA 1710is theconsensus stand ard for careerfirefighter deploym ent, including requirem entsfor firedepartme ntarrivaltime ,staffing levels,and firegroun d responsibilities.

Limi tations ofthestudy include firefighters’advance kno wledge oftheburn prop,invariable nu mbe r ofapparatus,and lackof experime ntsin elevated outdoor tempe raturesor atnight.Further, theapplicabilityoftheconclusions from thisreporttocommercial structurefires,high-risefires,outside fires,terrorism/natural disaster response,H AZ MAT or other technicalresponseshasnot been assessedand should not be extrapolatedfrom thisreport.

Prim ary Findings
Of the 22firegroun d tasks m easured during the experime nts, results ind icated thatthe followin g factorshad the m ost significantimp acton the success offirefighting operations.All differentialoutcom esdescribed below arestatisticallysignificant atthe 95% confidence levelor better.

O verallScene Tim e:

Th e four-person crews operating on a low-hazard structurefire com pleted allthe tasks on the firegroun d (on average)seven m inutes faster— nearly 30% — than the tw o-person crews.The four-person crews com pleted the sam e numb er offireground tasks (on average)5.1m inutes faster — nearly25% — than the three-person crews.O n the low -hazard residentialstructurefire, adding a fifth person to the crews didno tdecreaseoverall firegroun d task time s.Ho w ever,itshould be noted thatthe 1A low-hazardoccupancy isdefinedin theNFPA Handbook asa one-,two-,or three-familydwellingand some smallbusinesses.Medium hazardsoccupanciesinclude apartments,offices,mercantileand industrialoccupanciesnot normally requiringextensiverescueor firefightingforces.High-hazardoccupanciesincludeschools, hospitals,nursinghomes, explosiveplants,refineries,high-risebuildings,and otherhighlifehazard or largefirepotentialoccupancies.

ExecutiveSummary
10 Page 919 of 1153
2NFPA Standard 1710-A.5.2.4.2.1…Other occupanciesand structuresin thecommunity thatpresentgreaterhazardsshould be addressedby additionalfirefighter functionsand additionalresponding personnelon theinitialfullalarm assignment. 3NFPA 1710Standard fortheOrganizationand Deployment of FireSuppressionOperations,Emergency Medical Operations,and SpecialOperations to thePublicby Career FireDepartments.Section5.2.1– FireSuppressionCapabilityand Section5.2.2Staffing. 4As definedin thehandbook, a fastfiregrows exponentiallyto 1.0MW in 150seconds.A medium firegrows exponentiallyto 1MW in 300seconds.A slow firegrows exponentiallyto 1MW in 600seconds.A 1MW firecan be thought-ofasa typicalupholsteredchairburning atitspeak.A largesofamight be 2to 3MWs. benefitof five-person crews has been documen tedin other evaluations to be significantform edium- and high-hazard structures,particularlyin urban settings,and isrecognizedin industry standards. 2
Time to Wa teron Fire:

Th ere wa s a 10% difference in the“wa teron fire”time between the two-and three-person crews.There was an additional 6% differencein the"water on fire"time between the three-and four-person crew s.(i.e.,four-person crews put wateron thefire 16% fasterthan two person crews).Th erewas an additional6% differencein the“w ateron fire”time between the four-and five-person crew s (i.e.five-person crews put water on the fire22% faster than two-person crews).

Ground Ladders and Ventilation:

Th e four-person crews op erating on a low-hazard structurefire completed laddering and ventilation (forlife safety and rescue) 30% fasterthan the two-person crews and 25% fasterthan the three-person crew s.

PrimarySearch:

Th e three-person crews startedand com pleteda prim ary search and rescue 25% fasterthan thetwo-person crews.Th e four-and five-person crew s startedand com pleteda prim ary search 6 % faster than the three-person crews and 30% faster than the two-person crew.A 10% difference was equivalentto justover one mi nute.

Hos e Stretch Time:

In comp aring four-and five-person crews to two-and
three-person crew s collectively,thetime difference to stretcha line was 76 seconds.In conduc tingmore specificanalysis com paring allcrew sizesto thetwo-person crews the differences aremo re distinct.Tw o-person crews took 57 seconds longerthan
three-person crew s to stretch a line.Two-person crews took 87 second s longerthan four-person crews to complete the same tasks.Finally,them ostnotablecompa rison wa s between
two-person crews and five-person crews — more than 2minut es (122seconds)difference in task com pletion time.

IndustryStandard Ach ieved:

As defined by NFPA 1710,the “industrystandard achieved” time started from the firstengine arrivalatthehydrant and ended wh en 15firefighters were assem bledon scene .3An effective response forcewas assembled by the five-personcrews three min utesfasterthan thefour-person crews.Based on the study protocols,mode led aftera typicalfiredepartment apparatus deployme nt strategy,thetotalnum ber of firefighterson scenein the two-and three-person crew scenarios neverequaled15and therefore thetwo-and three-person crews were unab leto
assembl e enou gh personnel to meet thisstandard.

Oc cupantRe scue:

T hreedifferent“standard”fireswe resimulated using the Fire Dynamics Simu lator(FDS)mo del.Ch aracterizedin the
Han dbook oftheSocietyofFireProtection Engineersasslow-, m edium -,and fast-grow th rate 4,the fires grew expo nentially with time. Th e rescue scenario w as based on a non-am bulatory occupantin an up stairsbedroom with the bedroom do or open.

Ind ependent offire size,there wa sa significant difference betwe en thetoxicity,expressed asfractionaleffective dose(FED),for occupantsatthe time ofrescue depending on arrivaltimes forall crew sizes.O ccupantsrescued by early-arriving crews had less exposure tocom bustion productsthan occupantsrescued by late-arriving crews.T he fire m odeling showe d clearlythat two-person crew scannotcomp leteessentialfiregroun d tasksin time to rescue occup antsw ithoutsubjecting them toan increasingly toxic atmos phere.For a slow -grow th rate fire with tw o-person crews,the FEDw asapproaching the levelatw hich sensitivepopulations,such aschildren and the elderly are threatened.Fora m edium-growth ratefire w ith two -person crew s,the FEDwa sfarabove that thresholdand approached the levelaffecting the generalpopulation.

Fora fast-grow th ratefire w ithtw o-person crews,the FEDwa sw ell above the m edian levelatwh ich 50% ofthe generalpopulation wou ldbe incapacitated.Largercrew sresponding toslow-growth ratefirescan rescue mo stoccupantsprior toincapacitation along w ith early-arriving largercrews responding tome dium-growth rate fires.T he resultforlate-arriving (two m inuteslaterthan early-arriving)largercrew sm ay resultin a threatto sensitive populationsform edium -growth ratefires. Statisticalaverages shouldnot,howe ver,m ask the factthatthereisno FEDlevelso low thateveryoccupant ineverysituation issafe.

Co nclusion:

Mo re than 60 full-scale fire experime ntswere conductedto determi ne the imp actofcrew size,first-due enginearrivaltime,and subsequentapparatus arrivaltime son firefightersafetyand effectivenessata low-hazard residentialstructure fire.This report quantifiesthe effectsofchangestostaffing and arrivaltimesfor residentialfirefighting operations.Whi leresource deployment is addressed in thecontextofa singlestructuretype and risk level,itis recognized thatpu blic policy decisions regardingthecost-benefitof specific deployme nt decisions are a functionof many otherfactors includ inggeography,localrisks and hazards,availableresources,as w ellascommuni tyexpectations.Thisreportdoesnot specifically addresstheseother factors.

T he results of these field experiments con tribute significant kno w ledge to the fireservice industry.First,the resultsprovide a qu antitative basis for the effectiveness offour-person crews for low-hazard respo nsein NF PA 1710 .Th e results also provide valid m easuresoftotaleffective respon se forceassembl y on scene for firegroun d operations,as w ellas the expected perform ance time-to-critical-task me asures for low -hazard structure fires.

Ad ditionally,the results provide tenability m easures associated with a range ofmo deled fires.

Future research should extend the findings of thisreportin order to quantify the effects of crew size and app aratus arrival times for mo derate-and high-hazard events,such as firesin high-risebuildings,com me rcialproperties,certain factories,or w areho use facilities,respo nses to large-scale non-fireincidents,or technicalrescue operations.
11 Page 920 of 1153
T
he fireservicein theUnitedStateshasa deservedlyproud
tradition of serviceto communi tyand country dating back hundr eds ofyears.As technology advances and thescope
of servicegrows (e.g.,m ore EM S obligations and growing response to naturaldisasters,hazardous m aterialsincidents,and actsofterrorism),the fireservice rem ains comm ittedto a core mission of protecting livesand property from theeffectsof fire.

Firefighting is a dan gerous business with substantialfinancial implications.In 2007,U.S.m un icipalfiredepartmen tsresponded to an estimated 1,557,500fires.The sefireskilled3,430civilians (non-firefighters)and contributedto 17,675reportedcivilianfire injuries.Directproperty dama ge wa s estimat ed at$14.6billion dollars(Karter,2008).In spiteof the vigorous nationwide efforts to prom ote firefighter safety,the num berof firefighter deathshas con sistently rem ained tragically high. In both 2007and 2008,the U.S.FireA dm inistration reported 118 firefighter fatalities (USFA 2008).

Al though no tallfirefighterdeaths occur on the fireground — accidents in vehicles and training fatalities add to thenumb ers— every statisticalanalysisof the fireproblem in the United States identifiesresidentialstructure fires asa key com ponent in firefighter and civilian deaths,asw ellasdirectpropertyloss.

C on sequen tly,comm un ity plann ersand decision-make rsneed toolsfor op tima lly aligning resources with the service com mi tm ents needed for adequate protection ofcitizens.

Back ground
12 Page 921 of 1153
D
espitethe mag nitude of thefireproblem in theUnited
States,thereareno scientificallybased toolsavailableto
commu nityand fireservice leadersto assesstheeffectsof
prevention,fixed sprinklersystems ,fire fighting equipme nt,or deployme nt and staffing decisions.Presently,com munit y and fire service leadershave a qualitative un derstanding of theeffectof certain resourceallocation decisions.For example,a decisionto dou blethenumbe r of firehouses,apparatus,and firefighters woul d likelyresultin a decreasein com munit y firelosses,w hile cutting the nu m ber of firehouses,apparatus,and firefighters woul d likelyyieldan increase in the commun ityfirelosses,both huma n and property.Howe ver,decision-makerslacka sound basis for quantifying the totalim pactofenhanced fireresources on the num ber offirefighter and civilian lives savedand injuries prevented.

Studies on adequatedeployme ntofresourcesareneeded to
enablefiredepartme nts,cities,coun ties,and firedistrictsto design an acceptable levelofresourcedeployme ntbased up on com mun ity risks and serviceprovision comm itm ent.The se stud ies willassistwith strategicplanning and munic ipaland state bud getprocesses.Ad dition ally,as resource studies refine data collection m etho ds and m easures,both subsequent research and imp roveme nts to resourcedeploym entm odels willhave a sound scientific basis.

Problem
13 Page 922 of 1153
R
esearch to datehas docume nted a consistentrelationship betwe en resourcesdeployed and firefighterand civilian
safety.Studiesdocum enting engine and ladder crew
perform ance in diversesimulated environm entsas w ellasactual responsesshow a basicrelationship between apparatusstaffing levelsand a range of impo rtantperforman ce variablesand outcom e me asurements such asme an on-scenetime,time-to-task completion,incidence of injuryamon g fireservicepersonnel,and costsincurred asa resultof on -scene injuries(Cush man 1981, McMan is1984,Morr ison1990,O ntario1991,Ph oenix 1991,
Ro berts1993) .

Re portsby fire service officialsand consulting associates review ing fire suppression and eme rgency response by firecrews in U.S.citieswere the firstpublications to describethe
relationshipbetwe en adequatestaffing levelsand response time, time to comple tion of various firegroun d tasks,overall effectivenessoffire suppression,and estima tedvalueof property lossfor a wide range ofrealand simulatedenvironmen ts.In 1980, the Co lumbu s Fire Division’sreporton firefightereffectiveness showed thatfor a predetermined nu m ber of personn elinitially deployed to thesceneofa fire,the proportion of incidentsin wh ichprop ertylossexceeded $5,000and horizontalfirespreadof more than 25sq ft(2.3m 2) wa s significantlygreaterforcrews wh ose number s fellbelow thesetthresholdsof 15totalfireground personn elatresidentialfiresand 23atlarge-riskfires(Backoff 1980).Th e following year,repeated liveexperimentsata
one-family residentialsiteusing modern apparatusand
equipme nt demo nstratedthat largerunitsperformed tasks and accomp lished kn ockdown mo requickly,ultimatelyresultingin a lowerpercentageof lossattributableto factorscontrolledby the firedepartmen t.T he authorsof this articlehighlightedthatthe firecomp any isthe firedepartm ent’sbasicworking unitand furtherem phasized the importance of establishing accurateand up -to-dateperforma nce m easuremen tsto help collectdata and developconclusive strategiesto im prove staffing and equipm ent utilization(Gerard 1981).

Subsequent reportsfrom the UnitedStatesFireAdmi nistration (USFA)and severalconsulting firms continued to provide
evidence for the effectsof staffing on firecrews’ability to complete tasks involvedin firesuppressionefficientlyand effectively.Ci ting a seriesof testsconducted in 1977 by theDallas FireDepartm ent thatmea sured the time ittook three-,four-,and five-person team s to advance a lineand put wa teron a simulated fireatthe rearof the thirdfloorof an oldschoo l,officialsfrom the USFA underscored thattime-to-task com pletionand finallevelof physicalexhaustion for crews m arkedlyimproved not after any one threshold,bu tw iththe additionof each new team membe r.

T his reportw ent on to outline thema nn erin which simulated testsexempl ifya clear-cutmea ns to recordand analyzethe resourcesinitiallydeployed and finallyutilized atfirescenes(NFA 1981).A later pu blication detailing m ore DallasFireDepartment simulations — ninety-one runs each fora private residentialfire, high-riseofficefire,and apartm enthouse fire — showed again thatincreased staffing levelsgreatlyenhanced thecoordination and effectivenessof crews’firesuppression effortsduring a finite time span (McMan isAss ociates1984).Numerous studiesoflocal departmen tshave supported thisconclusion using a diverse collection ofdata,including a reportby the NationalFire A cademy (NFA )on firedepartme nt staffing in smaller
com mun ities,w hich showe d thata com pany crew staffed with fourfirefighterscould perform rescue ofpo tentialvictim s approxim ately 80 % faster than a crew staffed with three firefighters(M orrison 1990).

During the sam e tim e period thatthe im pactofstaffing levelson fireoperation s w as gaining attention ,investigatorsbegan to qu estion w hether staffing levels could alsobe associatedwith the riskof firefighter injuriesand the costincurred asa resultofsuch injuries atthe firescene.Initialresults from the Col um bus Fire Division show ed that“firefighterinjuries occurred m oreoften whe n the totalnumb er ofpersonnelon the fireground was less than 15atresidentialfiresand 23atlarge-risk fires”(Backoff 1980),and mo unting evidence hasind icated thatstaffing levels area fun dame ntalhealth and safety issue for firefightersin addition to being a key determinantofim m ediaterespo nse capacity.O ne early analysis by the Seattle FireDepartm ent for thatcity’sExecutive Board reviewe d the average severityof injuries suffered by three-,four-,and five-person engine com panies,with the find ing that“the rate offirefighterinjuries expressed as totalhoursofdisability per ho ursoffireground exposure w ere54% greater for engine com paniesstaffedwith 3 person nelw hen com pared to those staffed with 4firefighters, whilecomp anies staffed with 5personnelhad an injury rate that w asonly one-thirdthatassociated with four-person comp anies” (Cus hma n 1981).A jointreportfrom the International
A ssociation of FireFighters(IA FF)and John s H opkins University con cluded,after a comp rehensive analysis ofthe m inimum staffing levelsand firefighter injuryrates in U.S.cities with pop ulations of 150,000or mo re,thatjurisdictions operating with crew s ofless than four firefightershad injury rates nearly twice thepercentage ofjurisdictions operating with crews of
four-person crews or m ore (IA FF,JH U1991).

Mo rerecentstud ieshave continued to suppo rtthe find ing that staffing per piece ofapparatus integrallyaffects the efficacy and safetyof firedepartm entpersonn elduring em ergency response and firesupp ression.Two studies in particulardemo nstrate the con sistency ofthese conclusion s and the increasing levelof detail and accuracy presentin the mo strecentliterature,by looking closely atthe discrete tasks thatcould be safely and effectively performe d by three-and four-person firecom panies.Af ter testing drills comp rised of a seriesof com m on fireground tasks atseveral firesim ulation sites,investigatorsfrom the Aust in Fire Departm entassessed the physiologicalim pactand injury rates amo ng the variably staffed firecrew s.In thesesimulations,an increase from a three-to four-person crew resulted in m arked imp roveme nts in tim e-to-task comp letion or efficiency for the two-story residentialfiredrill,aerialladder evolution,and high-risefiredrill,leading the researchersto con clude thatlossof lifeand property increasesw hen a sufficientnum ber ofpersonnel arenot available to conductthe required tasksefficiently, independ entoffirefighter experience,preparation,or training.

R eview s of injury reports by the A ustin FireDepartment furtherm orerevealed thatthe injury rate for three-person com paniesin the four yearspreceding the study was nearly one-and-a-halfthatof crew s staffed with four or m ore personnel (R oberts 1993).In a sequence ofsim ilar tests,the Office ofthe FireM arshalofO ntario,Ca nada likewise foun d thatthree-person Rev iew ofLiterature
14 Page 923 of 1153
firecomp aniesw ere unab leto safelyperform deployment of backup protection lines,interiorsuppression or rescueop erations, ventilationoperations thatrequired accessto the roofof the involvedstructure,use oflargehand-held hose lines,or establish a watersupplyfrom a staticsource without additionalassistance and w ithin thetim e limitsof the study.Following thesedata,Fire Ma rshalofficials no tedthatthree-person crews we realsoat increased riskforexhau stion due to insufficientreliefatfire scenesand made recomm endations forthe m inimum staffing levelsper apparatus necessaryforsuppression and rescue related tasks(Office of the FireMa rshalofO ntario1993).

Th e most com prehensivecontempo rarystudieson the
implicationsof fire crew staffing no w include m uch more accurate performan ce mea suresfortasksatthe fireground,in addition to thebasicmetric of response time. Th ey include environm entalm easuresof performan ce ,such astotalw ater supp ly,w hich expand thepotentialfor assessing the
cost-effectivenessof staffing not only in terms of fireground personn elinjury ratesbut also compara tiveresourceexpenditure required forfiresuppression.Severalexam plesfrom theearly 1990sshow investigatorsand independen tfiredepartment s beginning to gatherthekind ofspecific,com prehensivedataon staffing and fireground tasks such as those suggestedand outlined in concurrentlocalgovernm ent publications thatdealtwith man agem ent offire services(Co leman 1988).A reportby the Ph oenix FireDepartme nt laidout clearprotocolsforresponding to structure firesand response evaluation in terms of staffing, objectives,taskbreakdown s,and times in additionto outlining the responsibilitiesof responding firedepartment mem bersand the orderin which they should be accomplished fora full-scale simulationactivity (Phoenix1991 ).O ne attempt to devisea prediction mode lfortheeffectiveness of manual firesuppression similarlyreached beyond respon setime benchm arksto describe fireoperations and the step-by-stepactions offirefightersat incidentscenes by delineating the time-to-task breakdowns for size-up,water supply,equipm entselection,entry,locating thefire, and advancing hose lines,wh ilealso com paring the predicted time-to-taskvaluesw iththe actualtimes and totalresources (Men ker1994).Two separate studiesof localfiredepartment perform ance,on e from Taoyuan Count y in Taiwan and ano ther from theLon do n FireBrigade,have drawn ties between firecrews’ staffing levels and totalwa terdem and as the consequence of both response time and fireseverity.Fielddatafrom Taoyuan Count y for casesof firein comm ercial,business,hospital,and educational propertiesshow ed thatthe type of land use aswellasresponse time had a significantimpact on the watervolum e necessary for firesup pression,with the no tablequantitativefinding thatthe w ater supp lyrequired on-scene doubled wh en the firedepartm ent respon se increased by ten mi nu tes(C hang 2005).

R espon se tim e as a predictor of residentialfireoutcom es has received less stud y than the effectofcrew size.A Ra nd Institute study demo nstrated a relation ship betwe en the distance the respon ding com panies traveled and the physicalpropertydam age.

This study show ed thatthe fireseverityincreased with respo nse distance,and thereforethe ma gnitude of lossincreased
propo rtion ally (Ran d 1978).Using records from 307 firesin non residentialbuildings over a three-year period,investigatorsin theUnited K ingd om correspon dinglyfoun d response time to have a significant im pacton finalfirearea,w hich in turn wa s propo rtion alto totalw ater dema nd (Sardqvist2000).

R ecentgovernm entand professionalliterature continue s to dem on strate the need for m oredata thatw ould quantify in depth and illustratethe required tasks,eventsequences,and necessary respon se tim es for effective firesup pression in order to determ ine with accuracy the fulleffectsof eithera reduction or increase in firecomp any staffing (K arter 2008) .A reportprepared for NationalInstituteof Standardsand T echnology (NIST)stressed theon going need to elucidatethe relationship betw een staffing and personn elinjury rates,stating that“a scientificstudy on the relation ship betw een the num ber offirefightersper engine and theincidence ofinjuries w ould resolve a lon g-standing question con cerning staffing and safety” (TriData2005).W hile not addressing staffing levelsas a centralfocus,an ann ualreview of firedepartm entcallsand false alarm s by the NationalFire Protection As sociation (NFPA )exemp lified the need to capture not on lythe numb er ofpersonnelper app aratus for effective fire supp ression butalso to clarifythe dem ands on individualfire departm ents with resolution atthe station level(NFPA 2008).

In lightofthe existing literature,thererem ain un answ ered qu estions aboutthe relationships betw een fireservice resource deploym entlevels and associated risks.For the firsttime this study investigates the effectofvarying crew size,firstapp aratus arrivaltim e,and respon se tim e on firefighter safety,overalltask com pletion and interior residentialtenability using realistic residentialfires.T hisstudy isalso un ique because of thearray of stakeholdersand the caliberof technicaladvisorsinvolved.

Ad ditionally,the structureused in the field experime ntsincluded custom ized instrum entation for the experiments;allrelated industry stand ards w erefollow ed;robustresearch m ethods w ere used;and the results and conclusion s willdirectly inform the NFP A 1710 T echnicalC om m ittee,as w ellas public officials and firechiefs. 5
15 5 NFPA isa registeredtrademark of theNationalFireProtectionAssociation,Quincy,Massachusetts.NFPA 1710definesminimum requirementsrelatingto the organizationand deployment of firesuppressionoperations,emergency medical operations,and specialoperationsto thepublicby substantiallyallcareerfire departments.The requirementsaddressfunctionsand objectivesof firedepartment emergency servicedelivery,responsecapabilities,and resources.The purpose of this standardisto specifytheminimum criteriaaddressingtheeffectivenessand efficiencyof thecareerpublicfiresuppressionoperations,emergency medical service,and specialoperationsdeliveryin protectingthecitizensof thejurisdictionand theoccupationalsafetyand healthof firedepartment employees.At thetime of the experiments,the2004editionof NFPA 1710was thecurrentedition.

T
his projectsystemat icallystudiesdeploymen tof fire
fighting resourcesand the subsequent effecton both
firefightersafety and theability to protectcivilians and theirproperty.Itisintended to enablefiredepartmen tsand city/county m anagersto m ake sound decisions regardingoptimal resourceallocation to meet service commi tme ntsusing the results of scientificallybased research.Specifically,the residential fireground experiments providequantitativedataon theeffectof crew size,first-due engine arrivaltime, and subsequen tapparatus stagger on time-to-task forcriticalstepsin responseand fire fighting.

T he firstphase of them ultiphaseprojectwas an extensive survey of mo rethan 400careerand com binationfiredepartm entsin the UnitedStates with the objective of op timizing a fireservice leader’s capabilityto deployresourcesto preventor mi tigate adverseevents thatoccur in risk-and hazard-filledenvironment s.

Th e resultsof thissurveyarenot do cum ented in this report, wh ichislimited to the experimentalph ase of the project,bu tthey willconstitutesignificantinput into futureapplications of the datapresented in thisdocumen t.

T his reportdescribes thesecond phaseof theproject,divided into four parts:
/square6Part1— Laboratory experimen tsto design theappropriate fuelpackagesto be used in the burn facilityspecially
constructed fortheresearch project
/square6Part2— Fieldtests for criticaltime-to-task com pletionof key tasksin fire suppression
/square6Part3— Fieldtests with realfurniture(room and contents experime nts)
/square6Part4— Fire m odeling to applydata gatheredto slow-, medi um-,and fast-growthratefires
Th e scope of this study is limi ted to un derstand ing therelative influen ce ofdeploym entvariables on low-hazard,residential structurefires,simi larin m agnitud e to the hazardsdescribed in NFPA ® 1710,Standard forthe Orga nizationand D eploym entof FireSuppression O perations,Em ergencyMedicalOpe rations,and SpecialO perationsto the Pu blic by Ca reer FireD epartme nts .

The stand ard uses as a typicalresidentialstructure a 2,000sq ft (186 m 2)two-story,single-fam ily dwe lling with no basem ent and no exposures (nearby buildings or hazardssuch as stacked flamm able m aterial).

Th e lim itation s of the study,such as firefighters’advance kno w ledge ofthe facility constructed for this experiment, invariable numb er ofapparatus,and lack of experim entsin extrem e tempe raturesor atnight,willbe discussed in the Limitation s section ofthis repo rt.Itshould be notedthatthe applicability ofthe con clusion s from thisreportto comm ercial structurefires,high-rise fires,outside fires,and respo nse to hazardous m aterialincidents,actsofterrorism,and natural disastersor other technicalrespon seshas no tbeen assessed and sho uld no tbe extrapolated from this repo rt.

Purpose and Scope oftheStudy
16 Page 925 of 1153
R
egardlessof thesize of a structureon fire,firefighting
crews identifyfour priorities: lifesafetyof occupantsand firefighters,confinemen tof thefire,property conservation, and reduction of adverse environmen talimpact .Interdependent and coordinated activitiesof allfire fighting personnel are required to m eetthe priorityobjectives .

NF PA 1710 specifies thatthenumber of on -duty fire
supp ression personn elmust be sufficientto carryout the necessaryfire fighting op erations given theexpectedfirefighting cond itions.During each firegroun d experiment,thefollowing we redispatchedto the testfirebuilding:
/square6threeengine com pan ies
/square6one truckcompany
/square6a com mand vehicle with a battalionchiefand a comman d aide
Staffing num bers fortheengineand truck crews and response timeswe revaried forthepurposesof thetests.Addi tional
personn elavailableto ensure safety willbe describedlaterin this report.

Th e following narrativeaccount describesthe generalsequence of activitiesin part2oftheexperiments (time-to-task),wh en the fuelload permitted firefighterentry :

T he firstarriving enginecom pany conductsa size-upor
initiallifesafety assessmen tof the building to include signsof occupantsin the hom e,construction features ,and location of theoriginalfireand any extension to otherpartsof the
structure.T hiscrew laysa supplyline from a hydrant closeto thebuilding fora continuouswater supply.

T he truck compa ny usuallyarrivesin closeproximityto the firstengine compa ny.The truckcompa ny is responsiblefor gaining accessor forcingentryinto the buildingso thatthe engine com pany can advance thefirsthose lineinto the
building to locateand extinguish the fire.Usually,they assist theengine company in finding thefire.The NFPA and
OSH A 2In/2Ou t 6 crew is alsoassem bledprior to anyone entering an atmo spherethatisim media telydangerous to life or health (IDLH).Th isimportant safetyrequirement will
have a large impact on availability offirefightersto enter the building when sm allcrews aredeployed.

O nce a door isopened,theenginecrew advancesa hose line (attack line)toward the locationof thefire.At thesame time, m embers from thetruckcrew accomp any theenginecrew and assistin ventilating thebuilding to provide a mo re tenable atmo spherefor occup antsand firefighters.V entilation also helpsby imp roving visibilityin an otherw ise“pitch black” environment, bu titmu stbe coordinated with theattack line crew to ensureithelpscontrolthefireand does not contribute tofiregrowth.T he truck crew performs a systematicrapid searchoftheentirestructurestarting in theareawhere
occup antswo uld be in themo stdanger.The most dangerous areaisproximat e tothefire and theareasdirectlyabove
thefire.

Depending upo n the traveldistance,the battalion chiefand comm and aide willhave arrived on the scene and have taken comm and ofthe incidentand established a com ma nd post.

Th e role ofthe incidentcom ma nd eris to develop the action plan to mi tigate the incidentand see thatthoseactions are carried outin a safe ,efficient,and effectivem anner.The comm and aide is respo nsiblefor situationalassessm ent and comm un ication s,including comm unications with crew
officersto ensurepersonnelaccountability .

Depending on respon se tim e or station location ,thesecond (engine 2)and possibly the third engine com pany (engine 3) arrive.T he second arriving engine (engine 2)connec tsto the firehydrantw herethe firstengine (engine 1)laidtheir sup ply line.Engine 2pump s w ater from the hydrantthrough the
supply line to the firstengine for fire fighting operations.

Ac cording to N FPA 1710 ,w ater should be flowing from the supply line to the attack engine prior to the attackcrew’s entry into the structure.

Th e crew from the second engine advances a second hand line as a backup line to protectfirefightersoperatingon the inside and to preventfirefrom spreading to other partsof the structure.

Th e thirdengine crew is respo nsible for establishing a Rap id Intervention T eam (R IT) ,a rescue team staged ator near the comm and postor asdesignated by the Incident Co mm and er (in the front of the building)with allnecessary equipm ent needed to locate and/or rescue firefightersthatbecom e
trapped or incapacitated.T he R IT plans entry/exitpo rtals and remo veshazards ,iffoun d ,to assistinterior crews.

As the fire fighting,search and rescue,and ventilation
operation s arecontinuing,two m emb ersofthe truck
comp any aretasked with placing groun d laddersto wind ows and the roof to provide a m eans ofegressfor occupa ntsor firefighters.T he truck crew is respo nsible for controlling interior utilities such as gasand electricaftertheirventilation, search ,and rescue duties arecom pleted.

O nce the fireis located and extinguished and occup antsare A BriefOverview ofthe Fireground Operations
6 The “2In/2Out” policyispartof paragraph (g)(4)of OSHAs revisedrespiratoryprotectionstandard,29CFR 1910.134.This paragraph appliesto privatesector workers engaged in interiorstructuralfirefightingand to Federalemployees coveredunder Section19of theOccupationalSafetyand HealthAct. Statesthathave chosen to operateOSHA-approved occupationalsafetyand healthstateplansarerequiredto extend theirjurisdictionto includeemployees of theirstateand localgovernments.

These statesarerequiredto adopt a standardatleastaseffectiveastheFederalstandardwithinsixmonths.

OSHAs interpretationon requirementsforthenumber of workers requiredto be presentwhen conducting operationsin atmospheres thatareimmediately dangerous to lifeand health(IDLH)coversthenumber of personswho must be on thescenebeforefirefightingpersonnelmay initiatean attackon a structuralfire.An interior structuralfire(an advanced firethathas spreadinsideof thebuildingwhere high temperatures,“heat”and dense smoke arenormally occurring)would presentan IDLH atmosphere and therefore,requiretheuse of respirators.In thosecases,atleasttwo standby persons,in additionto theminimum of two personsinsideneeded to fight thefire,must be presentbeforefirefightersmay enterthebuilding.

Letterto Thomas N.Cooper, Purdue University,from Paula O.White, Directorof Federal-StateOperations,U.S.Department of Labor,OccupationalSafety& Health Administration,November 1,1995.
17 Page 926 of 1153
removed ,the incident comma nder reassessesthe situation and providesdirection to conducta very thorough secondary search of thebuildingto verify thatthefirehas not extended into voidspacesand thatitisfullyextinguished.(Ina
no nexp erim entalfire situation,salvageablepropertywould be covered or remov ed to min imize damag e.)
T hrougho ut theentireincident,each crew officer is
responsibleforthesafety and accountabilityof his or her personnelalong with airman agem ent.The location and
w ellnessof crew s istracked by the comma nd aidethrough a system of personalaccountabilitychecksconducted at
20-minuteintervals.

Following extinguishme nt of the fire,an on site review is con ducted to identify actionsfor improvem ent.Crews are mo nitored,hydratedand restedbeforereturning to work in thefirebuilding.

The RelationofTime-to-Task Complet ionand Risk
Delayed response,particularlyin conjunctionwith the
deployme nt of inadequate resources,reducesthelikelihood of controllingthefire in time to preventma jordam age and possible lossoflifeand increasesthe dangerto firefighters.

Figure1illustrates a hypotheticalsequence of eventsfor
response to a structure fire.Duringfiregrowth,the tempe rature of a typicalcom partmen tfirecan riseto over 1,000 o F (538 o C).

Wh en a firein partof a com partmen treaches flashover,therapid transitionbetween the grow th and the fullydeveloped firestage, flame breaks ou talmo statonce over the surfaceof allobjects in the comp artm ent,with results for occupants,even firefightersin fullgear,thatarefrequently deadly.

Successfulcontainm ent and controlof a firerequirethe
coordination ofma ny separate tasks.Firesuppression m ustbe coordinated with rescue operations,forcible entry,and utilities con trol.V entilation typicallyoccursonly after an attack line is in place and crews areready to m ove in and attack thefire. The incidentcom m ander needs up-to-the-mi nute kno wledge ofcrew activities and the status oftask assignm entsw hich could resultin a decision to change from an offensive to a defensive strategy .

Standards ofRes pon se Cov er
Developing a standard ofrespon se cover — the policies and procedures thatdeterm ine the distribu tion ,concentration,and reliability offixed and mo bileresources for respon seto fire(as we llasotherkinds oftechn icalrespon se)— related to service com mi tm ents to the comm un ity is a comp lex task. Fireand rescue departme ntsmus tevaluate existing (or proposed) resources againstidentified risk levels in the com m unityand againstthe tasks necessary to cond uct safe,efficientand effective firesup pression atstructures identified in thesevarious risk levels.

Leadersm ustalso evaluate geographic distribution and depth or con centration ofresourcesdeployed based on tim e parame ters.

Re cognition and repo rting ofa firesets offa chain ofevents beforefirefightersarrive atthe scene: callreceiptand processing, dispatchof resources,don ning protective gear,and travelto the scene.N FPA 1710 definesthe overalltim e from dispatch to scene arrivalas the totalresponsetime .Th e standard dividestotal Figure1: HypotheticalTimelineofFireDepartment Response toStructureFire 18 Page 927 of 1153
response time into a number of discretesegmen ts,ofw hich travel time — thetime intervalfrom the beginning oftravelto thescene to the arrivalatthescene — is particularlyim portant forthis stud y.

Ar rivalof a firefighting respon seforcemust be imm ediately followed by organization of the resources into a logical,properly ph ased sequen ce of tasks,some of wh ichneed to be performed simultaneou sly.Kn owing the time ittakesto accomplish each taskw iththeallotted nu mber ofpersonnel and equipment is critical.Ideallycrew s should arrive and intervene in sufficient time to preventflashover or spreadbeyon d the room of origin.

Decision-mak ing about staffing levelsand geographic
distribution ofresources must considerthose timeswhe n there willbe simultaneou s eventsrequiring resourcedeployment .

Th ere should be sufficientredundan cy or overlapin thesystem to allow for sim ultaneous calls and high volum e ofnear
simultaneous respon seswithou tcomp rom ising the safetyof the pub licor firefighters.

Policy m akershave long lacked studies thatquantifychanges in firegroun d performa nce based on apparatusstaffing levels and on-scene arrivaltime intervals.T hese experim entswe redesigned to observe the im pactofapp aratusstaffing levels and app aratus arrivaltim eson the time ittakesto executeessentialfireground tasks and on the tenabilityinside the burn prop for a fullinitial alarm assignme ntrespon se.Itis expected thatthe resultsof this study willbe used to evaluate the related perform ance objectives in N FPA 1710 .
19 Page 928 of 1153
LaboratoryExperiments
T he pu rposeof thefirstsegme nt,the laboratoryexperimen ts ,was to characterize theburning behavior ofthe wood palletsas a function of:
/square6numbe r of pallets and thesubsequent peak heatreleaserate (HR R)
/square6comp artmen teffectson burning of woo d pallets /square6effectofwin dow ventilationon the fire
/square6effecton fire grow th rateof the loading configuration of excelsior(slender w ood shavings typicallyused aspacking mat erial)
Cha racterization of thefuelpackage was criticalin order to ensure thatthefieldexperimen tswould not resultin a flashover cond ition,on e of the primary safety considerations in complying with theprotocolsin NFP A 1403:Standard on Livefire Training Evolutions .7 Ap pend ixA ofthisreportcontainsthe m etho ds and fullresultsfor thelaboratory experimen ts,w hich aresum marized below .Figure2shows a testburn of palletsin thelaboratory.

Results ofLaboratoryExperimen ts
Th e objective of the laboratory experimen tswas to quantifythe spread of heatand smok e throughout theplanned burn prop in orderto ensurethatthe fuelpackage would resultin a firelarge enough to generate heatand smo ke consistent with a residential structurefire,yetnot so largeasto transition to flashover.Th e fullresultsofthelaboratory experimen tsand modelin g areshown inAp pendix A and Ap pendix B.T o summ arizebriefly,a
four-palletconfiguration,wh ichproduced a peak of
approximat ely2 M W, wa s determined to be thelargestfuelload the room couldsupportwithout thethreatoftransitioning to flashover.The com partmen tproduced a negligibleeffecton the heatreleaserate of the firecompa red to open burning conditions.

Th e presenceof an open win dow in theburn room reduced the prod uction ofcarbon m onox ide and carbon dioxide gases, prim arilythrough enh anced oxygen availabilityand dilution, respectively.T he location and quan tity ofexcelsior had a significantimp acton the growth rate offire.More excelsior located nearer the bottom ofthe pallets resulted in a mo rerapid achievem entofpeak burn ing .

Th e results ofthe fuelload experim ents to inform thebuilding and experim entaldesign indicated developm entofuntenable con dition s in the fieldexperiments betw een 5m in and 15m in, depend ing up on severalfactors:firegrowth rate,ventilation con dition s,the totalleakage of heatinto the building and through leakage paths,and m anualfiresupp ression .T his tim e frame allow ed for differentiation ofthe effectivenessofvarious fire Part1: PlanningfortheFieldExperiments
Figure2: TestBurn ofPalletsinLaboratory
20 7 NFPA 1403containstheminimum requirementsfortrainingallfiresuppressionperonnel engaged in firefightingoperationsunder livefireconditions.

departmen tresponse characteristics.

I
n part2,fire experiment swere cond uctedina residential-scale burn prop attheMon tgomer y C ounty PublicSafetyT raining A cademy inRo ckville,MD .

FieldSite
Mon tgomer y Co un ty(MD)Fireand R escueDepartment
provided an open spaceto constructa tempora ry burn prop,with ready access to w aterand electricalutilities,atthe M on tgom ery Co unty Fireand Rescue Training Facility in Rock ville,M D.

Th e bu rn prop w asconstructedasa two-storyduplex with a commo n stairwelland mo vablewallsbetween the sections to allow for mu ltipleexperiments daily.Sym met ricallydividingthe structureabou tthe shortaxisallowedone sideof thetest
structureto cooland dry out after a firetestwith suppression.T he bu rn prop contained two mirror-image,two-storyun itseach totaling 2,000ft 2 (186 m 2),w itho ut baseme nt or nearby exposures — each therefore a typicalm odelofa low-hazard single-family residence identified in NFP A 1710 .An exterior view of the burn prop isshown in Figure 3.For each experiment there wa s a confirmed fire in the living room in thefirstfloor rearof one unit of thestructure.

Detailsand dim ensionareshow n
inthefloorplanin Figure4 .

Th e black linesin Figure4
indicateload-bearingreinforced
concretew allsand red lines
indicatethegypsum oversteelstud
partitionwalls.T he ceilingheight
was 94in(2.4m)throughou tthe
entire structure exceptintheburn
comp artments,w hereadditional
hardening wa sinstalledto protect
againstrepeatedexposuretofire
duringtheexperimen ts.Th is
additionalfire proofingslightly
reduced theceiling height .

Comp letedetailsaboutthe
buildingconstructionareincluded
inAppendix C .

Noncombu stiblefurniture (angle
iron and gypsum board
construction)was fashioned to
representobstaclesofrealisticsize
and locationforfirefighters
navigatingtheinteriorofthe
structure.The dimensions were
typicalofresidentialfurnishings.

Figure 5shows an example ofthe
noncombust iblefurnitureused in
thetime -to-taskexperime nts.

Part2: FieldExperimentMethods
Figure3:ExteriorView ofBurn Prop
Figure4:Dimensions oftheBurn Prop FloorPlan
21 Page 930 of 1153
Ov erview ofFieldExperime nts
In ordertoevaluate the performance representative ofa NFPA 1710 -com pliantfiredepartment,thefieldexperimentsconsistedof two parts(thesecond and thirdpartsofthefourdescribedinthis report).In thefirstofthe two partsofthefieldexperiments, firefighterparticipantsfrom Montgo mer y County (MD)and Fairfax Cou nty (V A)FireDepartmentssimulated an initialalarm assignment responsetoa structure described inNFPA 1710 asa low-hazard residentialstructuretowhich firefightersrespondon a regularbasis.

T he staffingleveloffire apparatuswas variedincrementallyfrom two tofivepersonnelperpiece.The intervalbetweenapparatuson-scene arrivaltimesw asvaried ateither60 sor120s.T rainedtimingstaff we reusedtorecord thestartand compl etiontimesof22tasks deeme d essentialform itigation ofa residentialfireincidentby the study’stechnicalexperts.The palletand excelsiorconfiguration chosenfrom thelaboratoryexperimentsrepeatablyproduced a consistentand realisticquantityofheatand smoke,similarto w hat firefightersencounterata residentialstructurefire.

Al though the firesource used in part2of the fieldexperiments created a realistic amoun tof heatand smoke, the requirementsof NF PA 1403 preventeduse of a firesource w hich could potentially reachflashoverwithin thestructure.Th erefore,part3ofthefire experime ntsw ascond uctedin orderto chan ge thefuelpackage to be representative of realisticfuelloadingthatcould be found in a living room in a residentialstructure(sleeper-sofa,uph olstered chairs,end tables,etc).Th e
intentof this partof thestudy
was to determine how the times
of firefighterinteractions,
averaged with respectto the
staffing and arrivalintervals,
impac ted theinteriortenability
cond itions.Fire fightingtactics
we reperforme d in a ma nne r
wh ichcomplied with N FPA
1403 ;ventilation w asperformed
with properpersonalprotective
equipme nt (PPE)and hand tools
from theexteriorof theburn
prop.Suppressionwas
perform ed with an interior
remo te suppressiondevice
operated from the exteriorof the
bu rn prop.

Instrumentation
Instrument ation to me asure
gas tempe rature,gas
concentrations,heatflux,visual
obscuration,video,and time
during theexperiments was
installedthroughou ttheburn
prop.Th e data w ere recordedat
1-secondintervalson a
compute r-based dataacquisition
system. Figure 6 presentsa
schemat icplan view ofthe
instrume ntation .Al l
instrume ntsw erewired to a
centralizeddata collection room
attached asa separatespace on
the we stsideofthebuilding,
wh ichisdescribed laterin this
reportensuring physicalseparation for the data collection person nelfrom the effectsof the fire,w hilemi nim izing thewire and tube lengths to the datalogging equipm ent.SeeA ppendixC foraddition aldetailsaboutthe instrum entation.

Figure5: N oncombus tible Furniture U sed in theTime-to-T ask Experiments Figure6: Instrumentationand FurnitureProp Layout
22 Page 931 of 1153
Safety Protocols
Firefighter safety wa s always a prim ary concern in conducting the research.Participantswere drawn from two departmen ts— Fairfax Coun ty,V Aand M ontgome ry Count y,M D — that
regularlyconductNFPA 1403compliant livefiretraining fortheir staffand recruits .

A safety officerw asassigned to the experiments by the
M on tgome ry C oun tyFireand R escue Departm entto assure com pliance with N FPA 1403 .Th e safety officer(Figure 7) participated in allorientation activities,daily briefings ,and firefighter gearchecks and w as alw ays actively involvedin overseeing allexperim ents. T he safetyofficer had fullauthorityto term inateany operation ifany safety violation wa s observed.In addition to the safety officer,a rapid intervention team (RIT), assigned from dedicated crew s notin the actualexperim ent,wa s in place for each experim ent ,and a staffed amb ulance wa s on standby atthe site.Rad iocomm unication wa s always available during the experiments should a “m ayday” em ergency arise.

Experiments w erestopp ed for any action considered to be a protocolbreach or safetyconcern.For exam ple,allladders— 24 ft(7.3m)or 28 ft(8.5m )— w ereto be raisedby two firefighters.

A s crew sizes w erereduced,some firefightersattempt ed to place ladd erssingle-handedlyin an effortto comp letethe task mo re qu ickly.T hisprocedure,w hile vividly illustrating how firefighters try to do m ore with lessin the field,is un safe and could potentially resultin strain or imp actinjuries.

A dditionalsafety features w erebuiltin to the field structure.

A deluge sprinkler system oriented to the known location ofthe fuelpackage could be remo tely activated for rapid fire supp ression.A llfirstfloor rooms had directaccessto theexterior of the building throug h either do orsor wind ows. The second story had an em ergency exitto the roofof the attached
instrume ntation room .

A closely related con cern to ensure firefighter safetyand readiness to repeatexperim ents with equivalentperforma nce wa s adequate rehabilitation (see Figure 8) .At the beginning and end of each day,crews comp leted a health and safety check.Th e imp ortance of staying w ell-hydrated beforeand during
experime nts wa s especiallyem ph asized.

Figure8: Crew Rehabilitation
Figure7: FiregroundSafetyOfficer
23 Page 932 of 1153
On -Scene FireDep artmen tTasks
T he on -scene firedepartment taskpartof thestudy focusedon the tasksfirefightersperform aftertheyarriveon thescene of a low-hazard residentialstructurefire.A number of nationally recognized fireserviceexpertsw ereconsultedduring the
developmen tof theon-scenefiredepartment tasks inorder to ensurea broad applicabilityand appropriatenessof thetask distribution. 8 The experiment scomp aredcrew performance and wor kloadfora typicalfire fightingscenariousingtwo-,three-, four-,and five-person crews. 24totalexperimentswere conduc ted toassessthetime ittook variouscrew sizestocompl etethesame taskson technicallysimilarfiresinthesame structure.In addition tocrew sizes,theexperimentsassessedtheeffectsof stagger between the arriving comp anies.Close staggerwas definedasa 1-mi nute time differenceinthearrivalof each responding com pany.Farstaggerwas definedasa 2-minutetime difference in thearrivalof each respond ingcompany .One-minu teand
two-minu tearrivalstaggertimeswere determined from analysisof deployment datafrom more than 300U.S.firedepartments
respond ingtoa survey of firedepartment operationsconduc tedby theInternationalAs sociationof Fire Chiefs (IAFC) and the InternationalA ssociationof FireFighters(IAFF).C onsidering both crew size and comp any staggertherewere eightexperiments conducted intriplicatetotalingtwenty-four tests ,asshown inthe fullreplicateblock in Table1.A fullreplicatewas completed ina randomiz ed order (determined by randomization software)before a testconfiguration w asrepeated.

Crew Size
For each experimen t,three engines,a ladder-truck and a battalion chiefand an aidew eredispatched to the scene of the residentialstructurefire.T he crew sizesstudiedincluded two-, three-,four-,and five-personcrews assigned to each engineand truck dispatched.R esultant on-scene staffing totalsfor each experime nt follow:(FF= firefighter)
/square6Two Person crew s = 8 FFs +Chi efand Aide = 10totalon-scene /square6T hreePerson crews = 12FFs +Ch iefand Aide= 14total on-scene
/square6Four Person crew s = 16FFs +Chie fand Ai de = 18total on-scene
/square6Five Person crews = 20FFs +Chie fand Aide = 22total on-scene 9
Dep artmentParticipation
T he experimen tsw ereconducted in M ontgome ry County, MD atthe M ontgom ery Coun tyFireR escue Training A cademy during the mon thsofJanuary and February2009.All experiments took placein daylightbetween 0800hours and 1500ho urs.

Experim entsw ere postponed for heavy rain,ice,or sno w and rescheduled for a later datefollowing otherscheduled
experime nts.

Mo ntgom eryCou nty (M D)and Fairfax Count y (V A)
firefightersparticipated in the fieldexperiments.Each day both departmen tscomm ittedthree engines,a laddertruck and
associated crew s,as we llas a battalion chiefto the experiments.

The two battalion chiefs,alternated betwe en the rolesof battalion chiefand aide.Firefightersand officersw ereidentified by participating departm ents and oriented to the experiments.Each experime nt included engine crews,truck crews and com m and officersfrom each participating departme nt.Participantsvaried with regard to age and experience.Crew s thatno rm ally operated togetherasa com pany w erekept intactfor the experimentsto assuretypicaloperation for the crew during the scenarios.

H ow ever,in allexperime nts crew s we reused from both departm ents,including engine crews,truck crews,and officers .

Th isallocation ofresources m ade itpossible to conduc t back-to-back experim entsby rotating firefightersbetween field wor k and rehabilitation areas.

Crew Orientation
Al lstudy participantsw ererequired to attend an orientation prior to the beginn ing of the experim ents (see Figure9,page 25).

The orientations w ereused to explain experiment procedures, task flow s,division oflabor betwe en crews,and mi lestone events in the scenario.

Daily orientations w ereconducted for allshifts to assure every participantattended.O rientations includ ed a description ofthe overallstudy objectives as we llas the actualexperim entsin which they w ould be involved.Per the requirem ents ofN FPA 1403 ,full disclosureregarding the structure,the fire,and the tasks to be com pleted w ereprovided.Cre w s we realso oriented to the firegroun d props,instrume ntation used for data collection,and thespecificscenarios to be con ducted.Every crew m em ber wa s provided a w alkthrough ofthe structureduring the orientation and each day prior to the startofthe experim ents.

Table1: PrimaryVariablesforTime-to-Task Experiments
8 T echnicalexpertsincludedDennisCompton, RussellSanders,William “Shorty”Bryson,VincentDunn,DavidRohr,Richard Bowers,Michael Clemens, James Walsh, LarryJenkinsand Doug Hinkle.More informationabout theexpertsispresentedin theAcknowledgments laterin thisreport. 9 Note thattheon-scenetotalsaccount foronly thepersonnelassignedto“work” thefire.Additionalpersonnelwere provided foran RIT team,a staffedambulance on site,and a safetyofficerspecificto theexperiments.The additionalpersonnelarenot includedin theestaffingdescribedabove.

Time-to-TaskExperiments
CC rreeww SSiizzee AAppppaarraattuuss SSttaaggggeerr 2 Person C lose Stagger (One minute)
3 Person C lose Stagger (One minute)
4 Person C lose Stagger (One minute)
5 Person C lose Stagger (One minute)
2 Person Far Stagger (Two minutes)
3 Person Far Stagger (Two minutes)
4 Person Far Stagger (Two minutes)
5 Person Far Stagger (Two minutes)
24 Page 933 of 1153
Figure 10: Ground Ladders Figure 11: Ventilation
Figure 12: Ground Level Window Breakage Prop
Figure 14: Door Forcible Entry Prop Figure 15: Crew Preparation and Cue Cards Figure 13: Second Story Window Breakage Prop
Figure 9: Crew Orientation and Walkthrough
25 Tasks
Tw enty-two fireground tasks
we re comple ted in each
experime nt. Meticulous
procedures gathered data to
mea sure key areas of focus,
such as individual task start
times, task completion times ,
and overall scenario
perform ance time s. Each task
was assigned a standardized
start and end marker, such as
crossing the threshold to enter
the building with a hose line or
touching a ladder to raise it to
a secon d story wind ow . The 22
tasks, with the events for
mea suring start and stop times,
are shown in Table 2 (page26) .

Figures 10 — 19 illustrate
firefighter activity in a number
of the tasks to complete
experime nts or prepare for the
next experiment .

For reasons of both safety and
cost efficiency, two tasks —
forcible entry of the front door
and ventilation of the windows
on the first and second stories
— required special procedures.

Th e study could not
accommodate replacing the
do ors and windows daily for
the fire suppression
experime nts. Before the start
of experim ents with the full
sequence of tasks, these two
tasks were measured in a
realistic manner using training
prop s constructed at the site of
the firegroun d experimen ts. As
with the overall experiments,
these two tasks w ere repeated in
triplicate and the tim es
averaged. The average time to
compl ete the tasks was then
used in the larger scale
experiment. As firefighters
came to the point of breaching
the door or windo ws, the timers
wou ld hold them for the time
designated by the earlier
experiments and then give them
the approval to open the door
or windows. The start and end
times were then recorded just as
other tasks were.

T able 2: T asks and Measurement Parameters

1. Stop at Hydrant , Wrap Hose STAR T - Engine stopped at hydrant
STO P - Firefighter back on engine
and wheels rolling

2. Position Engine 1 STAR T - Wheels rolling from
hydrant
STO P -Wh eels stopped at
structure

3. Conduc t Size-up STAR T - Officer off engine
(360- degree lap ), transmit STOP - Completes radio
report, establish command transmission of report

4. Engage Pump STAR T - Driver off engine
STO P - Driver throttles up pum p

5. Position Attack Line STAR T - Firefighter touches hose (Forward Lay) to pull it from engine
STO P - Flake, charge and bleed
com plete (hose at front door
prepared to advance)

6. Establish 2 In/2 Out Company officer announces – “2
In/2 O ut established” (4 persons
assembled on scene OR at the
call of the Battalion
Chief/Comp any Officer)

7. Supply Attack Engine STAR T - Firefighter touches
hydrant to attach line
STO P - Wa ter supply to attack
engine

8. Establish RIT Time that Company Officer
announces RIT is established

9. G ain/Force Entry STAR T - Action started
(HO LD time= 10 seconds)
STO P - Do or opened for entry
10 . Adv ance Attack Line STAR T – Firefighter touches hose STO P – Wa ter on fire
1 1. Advance Backup Line STAR T - Firefighter touches hose (stop time at front door) to pull from engine bed
STO P - Backup line charged to
nozzle

12. Adv ance Backup START - Firefighter crosses
Line/Protect Stairwell threshold
STO P - Position line for attack at
stairwell
1 3. Conduct Prima ry Search STAR T - Firefighters enter front door
STOP - Firefighters transm it
“search com plete ”
14 . G round Ladders in Place STAR T - Firefighter touches ladder to pull it from truck
STOP - 4 Ladders thrown: 3
ladders on the 2 nd -story windows
and 1 to the roof
1 5. Horizontal Ventilation STAR T- Firefighter at 1 st w indow to (Ground) begin ventilation (H OL D for 8
seconds)
STOP - Hold time com plete -
window open
1 6 . Horizontal Ventilation STAR T - Firefighter grabs ladder (2 nd Story) for climb . (Firefighter mu st leg lock
for ventilation. HO LD tim e at each
window is 10 seconds)
STOP - All 2 nd -story w indows open
-descend ladder - feet on ground.
1 7. Control U tilities (Interior) STAR T - Radio transm ission to control utilities
STOP - When firefighter
com pletes the task at the prop
1 8 . Control U tilities (Exterior) STAR T - Radio transm ission to control utilities
STOP - When firefighter
com pletes the task at the prop
1 9. Conduct Secondary Search STAR T - Firefighters enter front door
STOP - Firefighters transm it
“secondary search com plete ”

20. C heck for Fire Extension STAR T- Firefighters pick up (walls) check-for-extension prop
STOP - Completion of 4 sets total
(1 set = 4 in and 4 out)
This task may be done by more
than one person.
21 . Check for Fire Extension STAR T - Firefighters pick up (ceilings) check-for-extension prop
STOP - Comp letion of 4 sets total
(1 set = 3 up and 5 down)
This task may be done by more
than one person.

22. Mechanical Ventilation STAR T - Firefighters touch fans to remov e from truck
STOP - Fans in place at front door
and started
Tasks M easurem ent Parame ters Tasks Mea sureme nt Parameters 26 Page 935 of 1153
Data Collection:

Standardized Control
Meas ures
Several control measures were
used to collect data, including
crew cue cards, radio
commu nications, task timers ,
and video recording.

Performance was timed for each
task in each scenario including
selected milestone tasks such as
doo r breach, water-on-fire, and
individual wind ow ventilation.

Data were collected for crew
performance on each task, and
individual firefighter
perform ance was not considered .

Task Flow Charts and
Crew Cue Cards
T ask procedures were
standardized for each
experiment/scenario. T echnical
experts worked with study
investigators to break down crew
tasks into individual tasks based
on crew size. T ask flow charts
we re created and then
custom ized for the various crew
sizes. Th e carefully designed task
flow ensured that the same
overall workload was maintained
in each experime nt, but was
redistributed based on the
nu mb er of personnel available
for the work. See Append ix D
for additional details.

Al l tasks were included in each
scenario and cue cards were
developed for each individual
participant in each scenario. For
exam ple, a four-person crew
wo uld have a cue card for each
person on the crew including the
officer, the driver ,and the two
firefighters. Cards were color
coded by crew size to assure
proper use in each scenario.

Radio commu nications
Interoperability of radio equipment used by both participating departmen ts made it possible to use regular duty radios for commu nication during the experiments. Company officers were instructed to use radios as they would in an actual incident.

Mon tgomer y Coun ty Fire and Rescue Commun ications recorded all radio interaction as a means of data backup. Once all data quality control measure were complete, the records were then overwr itten as a routine procedure.

Task Timers
T en observers/tim ers, trained in the use of a standard stop watch with split-time feature, recorded time-to-task data for each field experime nt. T o assure understanding of the observed tasks, firefighters w ere used as time rs, each assigned specific tasks to observe and to record the start and end tim es.

T o enhance accuracy and consistency in recording times, the data recording sheets used several different colors for the tasks (see A ppend ix D). Each timer was assigned tasks that were coded in the same color as on the recording sheet. All timers wore high-visibility safety gear on the fireground (see Figure 20).

Video records
In addition to the tim ers, video docum entation provided a backup for tim ed tasks and for quality con trol (see Figure 21). No less than six cam eras w ere used to record fireground activity from varied vantage po ints. Obs erver/time r data were comp ared to video records as part of the quality con trol process .

Figure 16: Connecting to the Hydrant Figure 17: Crews Responding Figure 18: Ceiling Breach/Molitor Machine Figure 19: Incident Command Figure 20: Task Timers Figure 21: Video Recording for Quality Control 27 Page 936 of 1153
Crew Assignm ent
Cre ws from each department that regularly operated together we re assigned to work as either engine or truck companies in each scenario. Both Fairfax County and Montgomery County crews participated in each experim ent.

Cre ws assigned to each responding com pany position in one scenario were assigned to another responding compan y position in subsequent scenarios, with the objective of min imizing learning from one experimen t to another . For example, crews in the role of engine 1 in a morning scenario might be assigned to the engine 3 position in the afternoon , thus elimin ating learning from exact repetition of a task as a factor in time to completion.

Ad ditionally, participating crews from both Mo ntgomery Coun ty and Fairfax Cou nty were from three different shifts, further reducing opportunities for participant repetition in any one position .

Resp onse Time Assumpt ions
Re sponse time assump tions were made based on time objectives set forth in the N FPA 1710 . T im e stagger allocations were set by the project technical advisors in order to assess the impact of arriving unit tim e separation on task start and completion times, as well as the overall scene time.

Below are the values assigned to the various tim e segm ents in the overall respo nse time . T he total of the response time segme nts ma y also be referred to as the total reflex time .

1. Fire ignition = tim e zero

2. 60 s for recognition (detection of fire) and call to 9-1-1

3. 60 s for call processing/dispatch

4. 60 s for turno ut 10

5. Cl ose Stagger = 240 s travel tim e FIR ST engine with 60 s ladd er-truck lag and 90 s lag for each subsequent engine a. Truck arrives at 300 s from notification
b. Second engine at 330 s from notification
c. Th ird engine at 420 secon ds from notification

6. Far Stagger = 24 0 s travel tim e FIR ST engine with 120 s ladder-truck lag and 150 s lag for each sub sequent engine a. Truck arrives at 360 s from notification
b. Second engine arrives at 390 s from notification
c. T hird engine arrives at 540 s from notification.

Th e design of this part of the experiments allowed firefighter entry into the burn building. T he next part of the experiments required a mo dified me thodo logy.
28 10 After the experiments were complete, the NFPA 1710 technical committee released a new edition of the standard that prescribes 80 seconds for turnout time.

Part 3: Room and Contents Fires
A
s previously discussed,
N FPA 1403 prohibits
firefighters in a training
exercise from entering a
structure with sufficient fuel
load to result in room
flashover. But the value of the
data from the time -to-task
experime nts lies not just in the
duration and
time-of-completion statistics
for tasks, but also in measuring
the tenability of the
atmo sphere for occupants
urgently needing firefighter
assistance. Therefore Part 3 of
the experiments (room and
contents fires) used a larger
fuel load to focus on the seven
of the 22 tasks that cause a
change in the fire behavior
throug h ventilation or active suppression :

1. Forced entry of the front do or

2. Wa ter on fire

3. Second floor window #1 ventilated (burn room window)

4. Second floor window #2 ventilated (front window , near corner)

5. Second floor window #3 ventilated (front window , near front doo r)

6. First floor window #1 ventilated (wind ow beside the fire room)

7. First floor window #2 ventilated (self-ventilated at flashover) Because the fuel load was sufficient for flashover, all firefighter activity was conduc ted outside the building. T asks that in Part 3 required entry into the building, such as search or interior utility control, were factored into this part by delaying the next task for the average duration of the task from Part 2. Firefighters in full gear opened the door with a gloved hand or opened windows from the ground with a tool such as a pike pole or angle iron, again at the time specified by the averages from Part 2. Av erages were derived from the three iterations of each scenario. The different number of iterations in Part 3 will be explained later in this report.

Because firefighters could not enter the building, a nozzle controlled from the instrume ntation room was installed. The nozzle was placed in the room directly outside the burn room and oriented toward the burn room near the doorway in order to best emula te the nozzle location of live firefighter suppression (see Figure 22). The nozzle was encased with mineral wool and heavy-duty aluminum foil (bottom picture in Figure 22 )to protect the electronics and w iring from the intense radiation energy emitted by the fire. Blocks were used to anchor the nozzle against the lateral forces exerted by the momen tum of the water supp ly. The activation time for suppression was determined by the data from the tim e-to-task test results.

A 15 o spray pattern was directed toward the seat of the fire and swept horizontally from side to side. While the remotely controlled hose line knocked down the majority of the fire, it was not as effective as a live firefighter with a better view into the room of origin. T herefore, after the fire was dim inished, a supp leme ntal stream w as app lied through the burn room window in order to control the fire (see Figure 23). Al l personne l on the hose line we re in full turnou t gear and self-contained breathing apparatus during the exterior app lication of water.

Fuel Packages for the Room and Contents Fires
In order to m axim ize the repeatability of the fire developm ent, nom inally identical rooms of furniture of identical manufacturer, style, and age w ere used for each test. A plan-view schem atic of the furniture is shown in Figure 24 and pictures of the burn room prior to testing are shown in Figure 25. Ke y dim ensions, ma ss, and materials for com bustible furnishings are detailed in A ppend ix C .

The Tornado Remote Controlled
Monitor is Produced by Task
Force Tips, Valparaiso, Indiana,
USA. Permission to publish
courtesy of Task Force Tips
Figure 23: Supplemental Suppression Applied for Room and Contents Tests
Figure 22: Remotely Controlled Fire Suppression Nozzle for Room and Contents Fires
29 Page 938 of 1153
T he ignition source con sisted
of a cardboard book of 20
ma tches that w as ignited by an
electrically heated wire, often
referred to as an electric m atch.

T he electric m atch wa s placed
near the bottom of a 21 qt
(19.9 L) po lypropylene waste
container. T he height of the
w aste container wa s 15.5 in
(394 m m ) with interior
dime nsions at the top opening
of 14.5 in (368 m m) by 11.3 in
(287 m m ). Approxim ately 0.7
lbs (0.3 kg) of dry newspaper
w as added to the wa ste
container. T he m ajority of the
newspaper wa s folded flat, and
placed on edge alon g the sides
of the w aste con tainer. Four
sheets of new spaper, 22 in (559
mm ) by 25 in (635 m m ) we re
crum pled into “balls”
approxim ately 3.9 in (100 mm)
diam eter and placed on top of
the electric m atch in the center
of the w aste con tainer.

Experimental Ma trix for
Room and Con tents Fires
Sufficient am ounts of
furniture for 16 rooms we re
available for the room and
contents fires, so eight
experiment scenerios were
cond ucted — each with a
replicate. Because the tim e to
untenable cond itions wa s a
prima ry variable of interest in
the room and contents fires,
the arrival time of the first due
engine w as a paramo unt
consideration. Because the
effects of the subsequen t
apparatus stagger were
explored in the tim e-to-task
tests, the stagger was fixed at
the “close arrival” tim e.

Ad ditionally, a baseline
me asurem ent wa s required to
comp are the effectiveness of
respon se to the absence of a
fire departme nt response.

Th erefore, a five-person, later
arrival comb ination was
elim inated in favor of a
no -respo nse scenario (with
replicate) . Table 3 summ arizes
the 16 tests conducted.

Th e first due engine arrival
tim es we re determ ined using
the following assum ptions:
ignition of the fire occurs at
Figure 24: Configuration of Furnishings in Burn Room (Room and Contents Fires) Figure 25: Pictures of the Room and Contents Furnishings 30 Page 939 of 1153
time zero. Sm oke detector
activation and a call to 9-1-1
occurs at 60 seconds after the
fire starts. C all intake and
processing requires an
add ition al 90 secon ds. T he
firefighters take 60 second s to
comp lete their turnout at the
station and begin travel to the
scene. T hus travel tim e begins

3.5 mi nu tes into experime nt.

Th e two levels of arrival tim e
are then determ ined by two
different travel tim es: early
arrival assume s a three-m inute
travel tim e, while later arrival
assume s a five-mi nute travel
time. For all scenarios in the
room and contents experimen ts,
the close stagger (60 seconds)
between subsequen t apparatus
times was used.

Procedure for Minimizing the Effect of Variance in Fire G rowth Rate
Fires involving furnishings have inherent variance in burning behaviors. Factors such as humi dity and minor variations in m aterials (particularly wo rn furnishings that m ay have different foam com pression or fabric w ear patterns), can result in un certainty of 20 % or mo re, despite significant efforts to enh ance repeatability. Th e early growth period of fire developm ent is often associated with the greatest variance, since m inor factors (as discussed above) can influence the therm al environm ent mo re easily wh en the fire is sm all. Therefore, the room and con tents fires were norm alized to the 212 °F (100 °C) tempe rature near the ceiling in the burn room in order to m inim ize the variance of the room and contents fires. The tim e at which the burn room reached this temp erature (usually in approxim ately 180 seconds) rather than the actual ignition time , w as designated as the “zero tim e. ”
Figure 26 shows the time-tempe rature curves before and after no rmalizing at 100°C. This approach was implemented during the experime nts by watching the time temper ature data in real-time from the instrumentation room and announcing the “zero-time” over the fireground radio system. T he normal ization procedure did not negatively affect tenability measurements in the target room because when the fire is small, products of combustion do no t reach the room because of lack of momentum. Therefore, adjusting all room and contents tests to the same upper layer tem perature was an appropriate way to minimize variance.

M ilestone Time s for Critical Tasks
A s stated earlier , firefighters could not enter the burn building during the room and contents experiments because of the danger for potential flasho ver in an experimental scenario. Therefore, prescribed tasks were performed at specified times based on data from part 2. In this section we report on significant data gathered from instrume ntation and describe an additional part of the experiments designed to extend our und erstanding of the effect of crew size and stagger on the tenability of the atmo sphere in a burning structure.

Table 4 (page 32) identifies significant tasks selected as key m ileston es because of the wa y they affect fire behavior and atmos pheric tenability inside the structure.

CC rreeww SSiizzee FFiirrsstt DD uuee AArrrriivvaall TTiimm ee 2-Person Early Arrival of First Engine (6.5 min) – close stagger 3-Person Early Arrival of First Engine (6.5 min) – close stagger 4-Person Early Arrival of First Engine (6.5 min) – close stagger 5-Person Early Arrival of First Engine (6.5 min) – close stagger 2-Person Later Arrival of First Engine (8.5 min) – close stagger 3-Person Later Arrival of First Engine (8.5 min) – close stagger 4-Person Later Arrival of First Engine (8.5 min) – close stagger No Response (Baseline) N/A
Table 3: Experimental Matrix for Room and Contents Tests (Each Conducted in Replicate) Figure 26: Direct Comparison of Temperatures,
Before (Top) and After Adjustment (Bottom)
31 Page 940 of 1153
Table 4: Tasks That Affect Fire Behavior and Atmospheric Tenability 32 Page 941 of 1153
T
his section describes the analytic approaches used to
address the research objectives of the study. First the statistical methods used to analyze the fireground
tim e-to-task observations are presented. Then the time-to-task data and the room and contents data were combin ed to assess crew performan ce in relation to tenability within the structure.

Time-to-Task Analysis
Tim e-to-task data were compiled into a database and assessed for outliers and missing entries. Because all time-to-task experime nts were conducted in triplicate, missing data were apparent and were review ed via video and radio tapes. Missing data attributable to timer error were replaced by a time observed in the video. Where video and/or radio documen tation was not adequ ate, missing data were recoded to the mean of the task times from the other two experiments.

Data Queries
Th e statistical methods used to analyze the time-to-task data we re driven by a principal goal of this research project — to assess the effect of crew size, first-due engine arrival time, and subsequent apparatus stagger on time-to-task for critical steps in response and fire fighting. This research goal motivated the developme nt of fou r specific research questions (see Figure 27) that in turn pointed to specific statistical analyses for generating inference and insight.

Statistical Methods – Time -to-Task
Th e analysis of the time-to-task data involved a sequence of multiple linear regression s using Ordin ary Least Squares to generate and test the effects of staffing and stagger on timings.

Th e regressions were of the form:
wh ere the xik reflect factors such as stagger and crew size, and the y represents our dependen t/outcome variable.

Tim e-related outcom es (i.e., the dependen t variables in the regression equations) could include task duration, elapsed time to start the task ,and elapsed time until task completion, all mea sured in seconds. Table 5 (page 34) lists the time-related outcom es used to test the effect of crew size and stagger for the tasks in the field experiments.

T he effects of crew size and stagger were explored using indicator variables in the regression analyses. The coefficient for a given ind icator (for example, crew size of four relative to a crew size of two) indicated the number of seconds the larger crew size added or redu ce the timing outcom e of a task. Crew sizes were collapsed in some regressions to test whether the timings of “larger” crew sizes of four and five were significantly different than “smaller” crew sizes of two and three. Interaction terms were not assessed in these regression analyses because of the sma ll numbe r of experiments available for analysis.

Standa rd t-tests examined statistical significance (i.e., to see if the hypothesis of “no impact” could be rejected) to estimate the impac t of several specific configurations:
/square6crew sizes of three versus two
/square6crew sizes of four versus three
/square6crew sizes of five versus four
/square6(occasion ally) five versus two, and four versus tw o /square6larger (four & five com bined) versus sm aller (two & three comb ined) and
/square6stagger
T he specific tests for each task (regression analysis) are shown in the Appendix E. The actual coefficients of each regression and their correspond ing standard errors are presented in A ppend ix F .

T o infer imp act, significant tests were conducted at the 0.05 significance level. Only statistically significant contrasts of crew size and/or stagger are included in this section of the report.

G raphic expositions of relevant time/ task related findings are then presented as well . W here stagger was statistically significant, the effects are graphed separately. W here stagger was not statistically significant , the data for crew size were combine d.

Analysis of Experimental Results
Time -to-Task Research Questions
1) H ow do crew size and stagger (i.e., timi ng of between first engine and sub sequent apparatuses) affect overall (i.e., start to comp letion) response tim ing?
a. T o wh at extent do variations in crew size affect overall respo nse tim ing?
b. T o what extent do variations in both crew size and
stagger affect overall response tim ing?
2) Ho w do crew size and stagger affect the tim ings of task initiation , task duration ,and task com pletion for each of the tasks comp rising the suite of 22 tasks?
a. T o what extent do variations in crew size affect timi ngs across the suite of tasks?
b. T o what extent do variations in both crew size and
stagger affect respon se tim ings across the suite of
tasks?
3) Ho w does crew size affect elapsed tim es to achieve three critical events kn own to change fire behavior or
atm osph eric tenability for occup ants?
a. Entry into structure
b. Water on fire
c. V entilation of each window (three up stairs and one down stairs window and the burn room wind ow)
4) Ho w does the elapsed tim e to achieve the national stand ard of assem bling 15 firefighters at the scene
(me asured using “at hydrant” as the start time) vary by crew sizes of 4 and 5?
Figure 27: Research Questions for Time-to-Task Experiments 33 Page 942 of 1153
Regression analyses
Ap pendix F presents the
regression results for each task
and relevant outcome , along
w ith their correspon ding
stand ard errors. T he results of
cond ucting significance tests at
the 0.05 level of significance
are shown in Appendix E.

R ather than detailing each of
the lengthy lists of coefficients
foun d to be significant, on ly
the answers to the prima ry
research questions are
presented for each task.

Me asurem ent U ncertainty
The measurement s of length,
temp erature, mass, moisture
content, smoke obscuration,
and stopwatch timi ng taken in
these experime nts have uniqu e
comp onents of uncertainty that
must be evaluated in order to
determine the fidelity of the
data. App endix G summari zes
the uncertainty of key
measureme nts taken during the
experime nts. Importantly, the
magni tudes of uncertainties
associated with these
measureme nts have no imp act
on the statistical inferences
presented in this report.

How to Interpret
Time -to-Task Graphs
Figure 28 presents a sam ple
tim e-to-task analysis, in this
case results for venting tim e.

Each crew size has a column
graphic showing the start time
and com pletion time for the
task. Visually, colum ns starting
low er on the graph depict
deploym ent con figurations
that resulted in earlier start
tim es. T he height of the
column graph ic is a
visualization of the duration of
the task, taller colum ns
indicating longer tim es to task
comp letion. Tim e data are also
shown in a table below the
graph. Wh ere stagger was
statistically significant, the
effects are graphed separately.

Wh ere stagger was not
statistically significant, as in the
illustration, the data for crew
size w ere comb ined.

Table 5 : Dependent Variables Used in a Regression Analysis of the Effect of Crew Size and Stagger on Time-to-Task Outcomes
Figure 28: Example Time-to-Task Graph
34 Page 943 of 1153
Time-to-Task Graphs
Ov erall Scene Time ( Time to
Comp lete All 22 T asks)
Th e four-person crews
operating on a low-hazard
structure fire completed the
same number of tasks on the
fireground (on average) 7
min utes faster than the
tw o-person crews (see Figure
29). The four-person crews
completed the same number of
fireground tasks (on average)

5.1 minutes faster than the
three-person crew. The
four-person crew s were able to
com plete necessary fireground
tasks on a low-hazard
residential structure fire nearly
30 % faster than the
two-person crews and nearly
25 % faster than the
three-person crews. Although
on the low-hazard residential
structure fire, adding a fifth
person to the crews did not
show any additional decrease in
fireground task times , the
benefits of a five-person vs. a
four-person crew are significant
in other measureme nts,
particularly the “water-on-fire”
time . Additionally, the greater
need for five-person crew s for
me dium- and high-hazard
structures ,particularly in urban
settings ,has been documented
in other studies (Backoff et al.,
1980; Cushman, 1982;
McMa nis Associates et al.,
1984) and five-person crews are
required for areas that contain
me dium and high-hazard
structures in fire protection
consensus standards. 11
Figure 29: Overall Scene Time
11 NFPA 1710, Section 5.2.3.1.2 and Section 5.2.3.2.2: In jurisdictions with tactical hazards, high-hazard occupancies, high incident frequencies, geographical restrictions, or other pertinent factors as identified by the AHJ, these companies shall be staffed with a minimum of five or six on duty members. 35 Page 944 of 1153
Figure 30 b: Overall Scene Time-Four Person Crew
Ov erall Scene Time and
Crew Sizes
T he graphs in Figure 30 show
average times for each task by
crew size.

Figure 30 a: Overall Scene Time-FIve Person Crew
36 Page 945 of 1153
Figure 30 c: Overall Scene Time-Three Person Crew
Figure 30 d: Overall Scene Time-Two Person Crew
37 Page 946 of 1153
38 Adv ance Attack Line
Time (Hose Stretch Time)
Figure 31 meas ures the
interval from the start of the
task “Position Attack Line” to
the end of the task “ Advanc e
At tack Line . ” In compar ing
four -and five-person crews to
two and three-person crew s
collectively, the time difference
for this measure was
statistically significant at 76
seconds (1 minute 16 second s).

In condu cting more specific
analysis compari ng all crew
sizes to a two-person crew the
differences are more distinct. A
two-person crew took 57
seconds longer than a
three-person crew to stretch a
line. A two-person crew took
87 seconds longer than a
four-person crew to complete
the same task. Finally, the most
no table compari son was
between a two-person crew and
a five-person crew, with a
122-second difference in task
com pletion time. 12 ,13
Figure 31: Advance Line Time (Hose Stretch Time) by Crew Size 12 Apparatus stagger was not statistically significant, so the data for crew size were combined. 13 Where subtracting the start time from the end time yields a result that differs from the duration noted in the chart by one second, it is the result of rounding fractional seconds to the nearest whole second.

39 Time to Water on Fire
T here w as a 10% difference in
the “wa ter on fire” tim e
betwe en the two - and
three-person crews. There was
an addition al 6% difference in
the "w ater on fire" tim e
betwe en the three- and
four-person crews. (i.e.,
four-person crews put water on
the fire 16% faster than tw o
person crews). T here was an
addition al 6% difference in the
“w ater on fire” tim e betw een
the four- and five-person crews
(i.e. five-person crews put
w ater on the fire 22% faster
than two -person crews).

Adv ancing a Backup Line
A dvancing a backup line to
the door and stairwell was
started 16 % faster and
comp leted 9 % for replicates
with shorter staggers betw een
comp any arrivals. A dvancing a
backup line is typically a task
comp leted by the third arriving
engine on a full alarm
assignm ent and is critical to
the safety of firefighters already
in the building on the initial
attack line. For this task,
stagger of arrival wa s
statistically significant and is
an im po rtant consideration for
overall station location and full
alarm respon se capability. T he
differences can be seen in
Figure 33, w hich show s the
tim e from the start for the task
“Deploy Backup Line” to the
end of the task “A dvance
Backup Line. ”
Figure 32: Water on Fire Time by Crew Size and Stagger
Figure 33: Times to Advance Backup Line by Crew Size and Stagger

14 Stagger was not significant, so data from close and far were combined to increase statistical power. 40 Figure 34: Times to Conduct Primary Search by Crew Size

Prim ary Search
Figure 34 summ arizes the
tim es that crews took to start
the primary search. O n the
low -hazard, two-story
single-fam ily dw elling 2,000 sq
ft (186 m 2), the three-person
crew started a prima ry
search/rescue mo re than 25 %
faster than the tw o-person
crew . In the sam e structure,
the four- and five-person crews
started a primary search 6%
faster than the three-person
crew s and 30 % faster than the
two-person crew. Note that
there is no end time included
in this figure. Prima ry search
end tim es w ere reliant upo n
radio com mun ication by
firefighters inside the structure.

O n occasion this
comm un ication did no t occur
or w as delayed. Therefore data
reliability w as insufficient for
analysis of task duration and
end tim e. 14
Ladde ring and
Venting Time
A four-person crew op erating
on a low-hazard structure fire
comp leted laddering and
ventilation (for life safety and
rescue) 30 % faster than a
two-person crew and 25 %
faster than a three-person crew.

G round laddering time
started with the remo val of the
first ladder from the truck and
stopped at end time of the last
ladd er pu t in place. A total of
four ladders we re raised on
each experim ent.

Truck op eration s ventilation
tim e is the tim e from the start
tim e of ventilation of the first
win dow un til the last window
ventilation was com plete.

The differences in start times
and duration of the tasks can be
seen in Figure 35 and Figure 36.

Figure 35: Laddering Time by Crew Size
Figure 36: Ventilation Times by Crew Size 11 55
15 Stagger was not statistically significant, so the data for crew size were combined. 41 Page 950 of 1153
42 16 Stagger was not statistically significant, so the data for far and near stagger were combined.

Figure 37: Industry Standard Effective Response Force Assembly Time Industry Standard
Effective Respon se Force
Assemb ly Time
N FPA 1710 requires that a fire
departmen t have the capability
to deploy an initial full-alarm
assignme nt to a scene within
eight-minutes (480 seconds).

Th e number of people required
falls between 15 and 17,
depend ing on whether an
aerial app aratus is used, and/or
if two engines are being used to
provide a continuo us water
supp ly. In these experimen ts,
the measuremen t for an
effective response force
assembly time started from the
first engine arrival at the
hydrant and ended when 15
firefighters were assembled on
scene. Figure 37 reveals the
differences in assembly times
between the four and
five-person crew s. An effective
response force was assembled
by the five-person crews a full
three minutes faster than the
four-person crew s. It is
importan t to note that (by
definition), the two -and
three-person crew s were unab le
to meet this standard at any
time during the experime nts. 16

43 Part 4: Fire Modeling
I
n the room and contents
experiments conducted in
Part 3 of the study,
instrumentation measured
oxygen, carbon dioxide, and
carbon monoxide
concentrations . Data were
grou ped by the type of
experiment conducted with
respect to crew size and first
du e engine arrival time. As
previously shown in the
experimental matrix, each
grou p contained two replicate
tests. In each group of data the
results of the replicates were
averaged to simpl ify the data
for further comparison. Figure
38 and Figure 39 show the
typical concentration curves for
the experiments .

Th ese two graphs show the
ranges representative of those
found in the experimen ts.

Cha rts of gas curves for the
rema ind er of the experim ents
— for both the burn room and
the target room — can be
found in Appendix H.

Fire Modeling Methods
A prim ary goal of fire
department response is to
prevent civilian injuries and
deaths. Because the significant
ma jority of fire deaths in the
United States occur in
residences, a rapid fire service
response provides the last
line-of-defense against civilian
fire deaths. Further, because the
fire service is less likely to rescue
occupants intim ate with the
fire (i.e., inside the room of
origin where conditions
deteriorate rapidly), tenability
mea surements were taken in a
remote bedroom on the second
floor of the residential burn
structure. The gas and
temperature measurements were
taken at the 5 ft (1.5 m ) height
above the floor, 3 ft (0.9 m)
from the west wall in order to
simul ate a nonambulatory
occupant (e.g, some one asleep,
und er the influence of alcohol or
drugs, or otherwise mobility
impa ired ).

Figure 38: Representative Oxygen Concentration
Figure 39: Representative Carbon Monoxide and Carbon Dioxide Con centrations

C omp utational fire models used the average suppression timings obtained from the time -to-task experimen ts under specific deployme nt configurations as inputs to the model . Th is quantitative approach eliminated the experimen tal variance of the fire. The resulting “computa tional” fire is repeatable, and therefore, any differences in occupant exposure to toxic gases will be due to the intervention times associated with a specific deployme nt configuration rather than the rando m variation that naturally occurs from fire to fire.

Fire sim ulations w ere comp leted using the NIST Fire Dynam ics Simula tor (FDS). FDS is a com putational fluid dynami cs m od el of fire-driven fluid flow . Th e first version of the FDS was released in 2000. FDS has been extensively verified and validated (USNR C 2007). Since the initial release, num erous im proveme nts have been m ade and new features added. T his study used FDS version

5.4.2 (Sub -version #4957), w hich w as released on October 19, 2009. In order to calibrate the model, sim ulation s were performe d to replicate the experim ental results observed in the Figure 40: Measured vs. Predicted Temperature at the 2.1 m (6.9 ft) Thermocouple Location in the Burn Compartment
Figure 41: Measured vs. Predicted Temperature at the 1.8 m (5.9 ft) Thermocouple Location in the Burn Compartment
Figure 42: Measured vs. Predicted Temperature at the 1.5 m (4.9 ft) Thermocouple Location in the Burn Compartment
Figure 43: Measured vs. Predicted Temperature at the 1.2 m (3.9 ft) Thermocouple Location in the Burn Compartment
Figure 44: Measured vs. Predicted Temperature at the 0.9 m (2.9 ft) Thermocouple Location in the Burn Compartment
Figure 45: Measured vs. Predicted Temperature at the 0.6 m (1.9 ft) Thermocouple Location in the Burn Compartment
44 Page 953 of 1153
room -and-contents fires. On ce
the ability of the mod el to
replicate experim ental results was
established, the different fire
grow th rates and deploym ent
configurations were simulated to
characterize the effectiveness of
different responses relative to
different fire growth rates.

Th e occupant exposure to toxic
gases was assume d to occur un til
the occupant is rescued by the
truck crew (start time of prima ry
search plus one minute). Table 6
shows the “rescue time ” for the
variou s crew sizes that correspond
to the test matrix for the room
and contents experime nts.

Part 4 of the experiments used
fire mo deling to correlate response
time s to atmos pheric tenability in
a burning structure. In order to
calibrate the computer fire model,
simulations were performed to
replicate the experimental
results observed in the
room-and-contents fires.

Model inputs include bu ilding
geometry and material properties, ventilation paths (doors, windows, leakage paths), and heat release rate of the fuel package.

Wh ile the building geometry is easily measured and material properties (such as the therm al properties of drywall and concrete) are readily estimated, the heat release rate wa s not directly measured during the experiments. The heat release rate of the fuel package is the prima ry determinan t of the production rate of heat, smoke, and gas species (e.g., carbon dioxide, carbon mono xide).

Figures 40 through 45 compare the experimental and simulated bu rn room tempe ratures using the burn room therm ocouple tree.

Th e tree con tained thermo couples located at 0.6 m (1.9 ft),

0.9 m (2.9 ft), 1.2 m (3.9 ft), 1.5 m (4.9 ft), 1.8 m (5.9 ft), and

2.1 m (6.9 ft) above the floor. For additiona l informa tion about the instrume ntation type location, see Appendix C. The results for thermocou ples located in the hot gas layer show excellent agreemen t. The temp erature at the lower two therm ocouples show an overprediction of the hot gas layer depth in the computer simulation. A small difference in the location of the interface height (the steep temp erature gradient between the relatively cool lower gas layer and the hot upper gas layer), can result in significant predicted temp erature differences with relatively little effect on the bulk heat and mas s transport accuracy. This explanation is supported by the agreem ent of the temperatures in the remote bedroom.

Figure 46 compare s the experimen tal and predicted oxygen concentration levels in the upstairs bedroom (m easured at 5 ft (1.5 m) above the floor, centered above the bed). Figures 47 throug h 52 compare the experimen tal and simulated
tem peratures in the upstairs (target room ) bedroom. As expected, the tempe ratures are moderated by mixing (cool ambie nt air mix es with hot combustion gases during transport between the bu rn room and the target room ) and by therma l losses to the (cooler) surfaces between the two rooms .

O nce the mo del inp uts we re determ ined to agree with the experime ntal results, the inp ut heat release rate was changed to represent three fire growth rates representative of a range of fire hazard developm ent – slow , m edium , and fast, which are described in greater detail in the following section s .

Time to Untenable Conditions: Research Questions
In the real wo rld, fires grow at ma ny different rates – from very slow, sm oldering fires all the w ay to ultra-fast, liquid fuel or spray fires. In order to extend the applicability of the findings of this report beyond the one fire growth rate observed in part 3 of this report (residential room and contents fires), com puter fire m od eling was used to quantify the effectiveness of fire departm ent op erations in respo nse to an idealized range of fire grow th rates (characterized as slow , m edium , and fast). Based on the research questions shown in Figure 53, fire mod eling m etho ds were then selected to maxim ize the applicability of the times to task results.

Figure 46: Measured Versus Predicted Oxygen Levels in the Upstairs Bedroom at 5 ft (1.5 m) 1) H ow do performance times relate to fire growth as
projected by standard fire time /temp erature curves?
2 ) Ho w do these performa nce times vary by crew size, first due arrival time, and stagger?
3) H ow do crew size, stagger, and arrival time affect occupant tenability within the structure?
Figure 53: Research Questions for Time to Untenable Conditions 45 Page 954 of 1153
Figure 47: Measured vs. Predicted Temperature at the 2.4 m (7.8 ft) Thermocouple Location in the Bedroom
Figure 48: Measured vs. Predicted Temperature at the 2.1 m (6.8 ft) Thermocouple Location in the Bedroom
Figure 49: Measured vs. Predicted Temperature at the 1.8 m (5.9 ft) Thermocouple Location in the Bedroom
Figure 50: Measured vs. Predicted Temperature at the 1.5 m (4.9 ft) Thermocouple Location in the Bedroom
Figure 51: Measured vs. Predicted Temperature at the 1.2 m (3.9 ft) Thermocouple Location in the Bedroom
Figure 52: Measured vs. Predicted Temperature at the 0.9 m (2.9 ft) Thermocouple Location in the Bedroom
Fire Growth Rates
Th ree fire growth rates were used in the computer fire modeling to assess the effectiveness of different fire department deployment configurations in response to fires that were similar to, faster growing, and slower growing than the fires observed in the room-and-contents fires. The slow, medium, and fast fire growth rates are defined by the Society of Fire Protection Engineers according to the time at which they reach 1 megawatt (MW ) . A typical upholstered chair burning at its peak would produce a 1-MW fire, while a large sofa at its burning peak would produce roughly a 2-MW fire.

Th e grow th rate of fires is often approxim ated by sim ple correlation of heat release rate to the square of time. If a fire is not supp ressed before full-room involveme nt, the probability of spread beyon d the room of origin increases dram atically if there is nearby fuel load to sup port fire spread. If a nearby fuel load is available, the 12 ft (3.7 m) by 16 ft (4.9 m) compartment used in the fire experime nts would becom e fully involved at
approxim ately 2 M W. Table 7 shows the time in seconds at which 1-MW and 2 -MW (fully involved) fires in this com partm ent woul d be reached in the absence of sup pression .
46 Page 955 of 1153
A fire department rescue
operation is a race between the
deteriorating interior conditions
inside the structure and the
rescue and suppression activities
of the fire department. Each fire
growth rate was used as a
baseline heat release rate for the
simu lation. Intervention times
(window and door open ing times
and suppression time) from the time-to-task tests were
system atically input into the model to evaluate the effects on interior tenability conditions. The interior tenability conditions we re calculated in a remote upstairs bedroom (above the room of fire origin on the first floor) in order to maximize the opportunity for differentiation among different crew configurations.

Fractional Effective Dose (FED ))
In order to convert instantaneous measurements of local gas conditions, the fractional effective dose (FED) formulation published by the International Standards Organization (ISO) in document 13571 Life-threatening Com ponents of Fire – Guidelines for the Estimation of Time Available for Escape Using Fire Data (ISO 2007) we re used. FED is a probabilistic estimate of the effects of toxic gases on humans exposed to fire effluent. The formulation used in the simul ations accounts for carbon monoxide (CO), carbon dioxide (CO 2), and oxygen (O 2) depletion. Other gases, including hydrogen cyanide (HCN) and hydrogen chloride (HCl), were not accounted for in this analysis and may alter FED for an actual occupant.

T here are three FED thresholds generally representative of different expo sure sensitivities of the general population. An FED value of 0.3 indicates the potential for certain sensitive pop ulations to become incapacitated as a result of exposure to toxic com bustion products. Sensitive pop ulations may includ e elderly, youn g, or individua ls with comp romi sed immune system s. Incapacitation is the po int at wh ich occupants can no longer effect their own escape. An FED value of 1.0 represents the m edian incapacitating exposure. In other words, 50 % of the general population will be incapacitated at that exposure level.

Finally, an FED value of 3.0 represents the value where occupants who are particularly tolerant of comb ustion gas expo sure (extrem ely fit persons, for exam ple) are likely to becom e incapacitated.

T hese thresho lds are statistical probabilities, not exact m easurem ents. T here is variability in the way ind ividuals respond to toxic atm ospheric condition s. FED values above 2.0 are often fatal do ses for so-called typical occupa nts. There is no threshold so low that it can be said to be safe for every exposed occupant. 17 2-Person Early 12:47
3-Person Early 9:03
4-Person Early 9:10
5-Person Early 8:57
2-Person Late 14:47
3-Person Late 11:03
4-Person Late 11:10
Table 6: Rescue Time for Different Deployment Configurations De ploym ent
Co nfiguration
(All times with
close stagger adjusted
for early and late arrival
of first due engine)
R escue Tim e for
De ploym ent
Con figuration
(Min : Sec)
Fire G rowth Ra te Time in Seconds Time in Seconds to
Re ach 1 M W Reach to 2 MW
Slow 600 848
Me dium 300 424
Fast 150 212
Table 7: Time to Reach 1 MW and 2 MW by Fire Growth Rate In the Absence of Suppression Wh ere C iis the concentration of the ith gas and (C t)iis the toxic concentration of ith gas and Δt is the tim e increm ent.

Eq .1
47 17 See the following sections of ISO Document 13571:

5.2 Given the scope of this T echnical Specification, FED and/or FEC values of 1,0 are associated, by definition, with sublethal effects that would render occupants of average susceptibility incapable of effecting their own escape. The variability of human responses to toxicological insults is best represented by a distribution that takes into account varying susceptibility to the insult. Some people are more sensitive than the average, while others may be more resistant (see Annex A.1.5). The traditional approach in toxicology is to employ a safety factor to take into consideration the variability among humans, serving to protect the more susceptible subpopulations.

5.2.1 As an example, within the context of reasonable fire scenarios FED and/or FEC threshold criteria of 0,3 could be used for most general occupancies in order to provide for escape by the more sensitive subpopulations. However, the user of this T echnical Specification has the flexibility to choose other FED and/or FEC threshold criteria as may be appropriate for chosen fire safety objectives. More conservative FED and/or FEC threshold criteria may be employed for those occupancies that are intended for use by especially susceptible subpopulations. By whatever rationale FED and FEC threshold criteria are chosen, a single value for both FED and FEC must be used in a given calculation of the time available for escape.

Results from Mod eling Me thods
Table 8 shows the FED for slow-, medium-, and fast-growth rate fires correlated to rescue times based on crew size and arrival time in the study. A s with the room-and-contents fire in part 3, results in Table 8 included only the close-stagger rescue time data. The effect of far-stagger rescue times on occupant tenability shou ld be investigated in future studies. Values above 0.3 are shown in yellow , and those above the median incapacitating exposure of 1.0 are shown in red.

Figure 54 show s that with slow-grow th fires in the experime ntal residential structure, all crew configurations could achieve rescue time before FED reached incapacitating levels. Figure 55 illustrates the greater danger of
me dium-grow th fires, wh ere
the FED at rescue tim e for
two-person crew s is well above
the 0.3 level, and alm ost to that
level for the other crews.

Figure 56 (page 49) vividly
illustrates the extrem e danger
of fast-growth fires. By the
tim e a tw o-person crew is able
to facilitate a rescue, the FED
has far exceeded the median

1.0 level. For other crew sizes,
the FED has exceeded 0.3,
wh ich is a threshold level for
vulnerable population s.

Table 8: FED as a Function of Deployment Configuration and Fire Growth Rate Figure 54: FED Curves for Early Arrival for All Crew Sizes at Slow-Growth Fires
Rem ote Room Tenability for Slow Fires
Figure 55: Average FED Curves for Early Arrival for All Crew Sizes at Medium-Growth Fires
Re mot e Room Tenability for Me dium Fires
48 As with the room-and-contents fire in part 3,
results

Interior Firefighting Conditions and Deployment
Con figuration
The available time to control a fire can be quite small. Risks to firefighters are lower for smaller fires than larger fires because smaller fires are easier to suppress and produce less heat and fewer toxic gases. Therefore, firefighter deploym ent configurations that can attack fires earlier in the fire development process present lower risk to firefighters. The longer the duration of the fire development process without intervention, the greater the increase in risk for occup ants and responding firefighters. Th erefore, time is critical.

Stop ping the escalation of the event involves firefighter intervention via critical tasks performed on the fireground.

Cri tical tasks, as described previously, include those tasks that directly affect the spread of fire as we ll as the associated structural tenability.

T here are windows of opportunity to complete critical tasks. A fire in a structure with a typical residential fuel load at six min utes post-ignition is very different from the same fire at eight minutes or at ten minutes post-ignition. Some tasks that are deemed “important” (e.g., scene size-up) for a fire in early stages of growth become critical if intervention tasks are delayed. Time can take away opportunities. If too much time passes, then the w ind ow of opportunity to affect successful outcome s (e.g., rescue victim or stop fire spread) closes .

For a typical structure fire event involving a fire department respon se, there is an incident com m ander on the scene who determ ines both the strategy and tactics that will be em ployed to stop the spread of the fire, rescue occupants, ventilate the structure, and ultim ately extinguish the fire. Incident com ma nders m ust deal with the fire in the present and make intelligent com m and decisions based on the circums tances at han d upo n arrival. Additionally, arrival tim e and crew size are factors that con tribu te to the incident com ma nd er’s decisions and affect the capability of the firefighters to accomp lish necessary tasks on scene in a safe, efficient ,and effective manner.

Table 9 illustrates vividly the m ore dangerous cond itions sm all crew s face because of the extra tim e it takes to begin and com plete critical tasks (particularly fire suppression). In the two minutes m ore it took for the two -person crew (early arrival) than the five-person crew (early arrival) to get w ater on the fire, a slow grow th rate fire w ould have increased from 1.1 MW to 1.5 MW.

This growth w ould have been even mo re extreme for a
m edium -or fast-growth rate fire. T he difference is even m ore sub stantial for the tw o-person crew with late arrival as the fire almos t do ub led in size in the tim e difference between this crew and the five-person crew.

Based on fire m odeling for the low hazard structure studied with a typical residential fuel load, it is likely that medium- and fast-growth rate fires will move beyond the room of origin prior to the arrival of firefighters for all crew sizes. Note that results in T able 8 included only the close-stagger rescue time data. Th e effect of far-stagger rescue tim es on occupant tenability should be investigated in future studies. T herefore, the risk level of the event upon arrival will be higher for all crews w hich m ust be considered by the incident com ma nd er wh en assigning firefighters to on-scene tasks.

Figure 56: Average FED Curves for Early Arrival for All Crew Sizes at Fast-Growth Fires
Table 9: Fire Size at Time of Fire Suppression
Re mo te Room Tenability for Fast Fires
Dep loym ent Time to Water Fire Size at Time of
Co nfiguration on Fire Suppression for
Slow- G rowth Fires
2-Person, Late Arrival 14:26 2.1 MW
2-Person, Early Arrival 12:26 1.5 MW
3-Person, Late Arrival 13:24 1.8 MW
3-Person, Early Arrival 11:24 1.3 MW
4-Person, Late Arrival 13:11 1.7 MW
4-Person, Early Arrival 11:11 1.3 MW
5-Person, Late Arrival 12:33 1.6 MW
5-Person, Early Arrival 10:33 1.1 MW
49 (Min : Sec)

R
epo rts on firefighter fatalities consistently docum ent overexertion/overstrain as the leading cause of line-of-duty fatalities. There is strong epidemiological evidence that heavy physical exertion can trigger sudden cardiac events (Mittleman et al. 1993; Albert et al. 2000). Therefore, information about the effect of crew size on physiological strain is very valuable .

During the planning of the fireground experiments,
investigators at Skidmo re College recognized the opportunity to cond uct an independent study on the relationship between firefighter deployment configurations and firefighter heart rates.

With the approval of the Institutional Review Board of Skidm ore Co llege, they were able to leverage the resources of the field experime nts to conduct a separate analysis of the cardiac strain on fire fighters on the fireground.

For details, consult the com plete report (Sm ith 2009). Two imp ortant con clusion s from the repo rt reinforce the im po rtance of crew size:
/square6Av erage heart rates were higher for m emb ers of sma ll crews, particularly two-person crews.
/square6Danger is increased for sm all crews because the stress of fire fighting keeps heart rates elevated beyond the maxim um heart rate for the duration of a fire respo nse, and so the higher heart rates we re m aintained for sustained tim e intervals.

Physiological Effects of Crew Size on Firefighters
50 Page 959 of 1153
Study Limitations
T
he scope of this study is limited to un derstanding the relative influence of deployment variables to low-hazard, residential structure fires, similar in ma gnitude to the hazards described in NF PA 1710 . The applicability of the conclusions from this report to comme rcial structure fires, high-rise fires, outside fires, terrorism /natural disaster response, HA ZM AT or other technical responses has not been assessed and should not be extrapolated from this report.

Every attem pt was made to ensure the highest possible degree of realism in the experime nts while complying with the
requirem ents of NF PA 1403 , but the dynam ic environment on the fireground canno t be fully reproduced in a controlled experiment.

For exampl e, NF PA 1403 required a daily walkthrough of the burn prop (including identifying the location of the fire) before ignition of a fire that would produce an Im media tely Dangerous to Life and Health (IDLH) atm osphere, a precaution not available to responde rs dispatched to a live fire.

Th e number of responding apparatus for each fireground response was held constant (three engines and one truck, plus the battalion chief and aide) for all crew size configurations. The effect of deploying either more or fewer apparatus to the scene was not evaluated.

Th e fire crews who participated in the experim ents typically operate using three-person and four-person staffing. Therefore, the effectiveness of the two-person and five-person operations may have been influenced by a lack of experience in operating at those staffing levels. Stand ardizing assigned tasks on the firegroun d w as intend ed to m inim ize the imp act of this factor, w hich has an un known influen ce on the results.

T he design of the experiments con trolled for variance in performa nce of the incident comm ander. In other wo rds, a m ore-or less-effective incident com m ander may have a significant influen ce on the outcome of a residential structure fire.

A lthoug h efforts w ere ma de to m inim ize the effect of learning across experim ents, som e participants took part in more than one experime nt, and others did no t.

T he we ather cond ition s for the experiments were mo derate to cold . Frozen equipme nt such as hydrants and pump s was no t a factor. Ho w ever, the effect of very hot we ather conditions on firefighter perform ance w as not me asured.

A ll experim ents w ere cond ucted during the daylight hours.

Nighttim e operations could po se additional challenges.

Fire spread beyond the room of origin was not considered in the room and con tents tests or in the fire m odeling. Therefore, the size of the fire and the risk to the firefighter may be som ew hat un derestim ated for fast-growing fires or slow er-response con figurations.

T here is m ore than one effective wa y to perform many of the required tasks on the firegroun d. A ttem pts to generalize the results from these experime nts to individua l departments mus t take into accoun t tactics and equipm ent that vary from those used in the experiments.
51 Page 960 of 1153
Con clusions
M
ore than 60 laboratory and full-scale fire experiments were condu cted to determine the impact of crew size, first-due engine arrival time, and subsequent apparatus arrival
times on firefighter safety and effectiveness at a low-hazard residential structure fire. This report quantifies the effects of changes to staffing and arrival times for low-hazard residential firefighting operations. While resource deployment is addressed in the context of a single structure type and risk level, it is recognized that public policy decisions regarding the cost-benefit of specific deployment decisions are a function of many factors including geography, available resources, community expectations, as well as all local hazards and risks. Though this report contributes significant knowledge to community and fire service leaders in regard to effective resource deployment for fire suppression, other factors contributing to policy decisions are not addressed.

Th e objective of the experiments w as to determine the relative effects of crew size, first-due engine arrival time, and stagger time for subsequent apparatus on the effectiveness of the firefighting crews relative to intervention times and the likelihood of occupant rescue using a parametric design. Therefore, the experimental results for each of these factors are discussed below.

Of the 22 fireground tasks measured during the experiments, the following were determ ined to have especially significant impact on the success of fire fighting operations. Their differential outcomes based on variation of crew size and/or apparatus arrival times are statistically significant at the 95 % confidence level or better.

Ove rall Scene Time:

Th e four-person crews operating on a low-hazard structure fire comp leted all the tasks on the fireground (on average) seven min utes faster — nearly 30 % — than the two-person crews. The four-person crews completed the same number of fireground tasks (on average) 5.1 minutes faster — nearly 25 % — than the three-person crew. For the low-hazard residential structure fire, adding a fifth person to the crews did not decrease overall fireground task times. However, it should be noted that the benefit of five-person crew s has been documente d in other evaluations to be significant for med ium- and high-hazard structures, particularly in urban settings, and should be addressed according to industry standards. 18 Time to Water on Fire:

Th ere was a nearly 10 % difference in the “water on fire time” between the two and three-person crews and an additional 6 % difference in the “water on fire time” between the three- and four-person crews (i.e., 16 % difference between the four and two-person crews). The re was an additional 6 % difference in the “w ater on fire’” time between the four- and five-person crews (i.e., 22 % difference between the five and two-person crews).

Ground Ladders and Ventilation:

T he four-person crew operating on a low-hazard structure fire can comp lete laddering and ventilation (for life safety and rescue) 30 % faster than the two-person crew and 25 % faster than the three-person crew.

Primar y Search:

T he three-person crew started and completed a primary search and rescue 25 % faster than the two-person crew. In the same structure, the four- and five-person crews started and com pleted a prim ary search 6 % faster than the three-person crews and 30 % faster than the two-person crew . A 10 % difference was equivalent to just over on e minu te .

H ose Stretch Time:

In comp aring four-and five-person crews to two-and three-person crews collectively, the time difference to stretch a line was 76 seconds.

In conducting mo re specific analysis comp aring all crew sizes to a tw o-person crew the differences are more distinct. A two-person crew took 57 seconds longer than a three-person crew to stretch a line. A two -person crew took 87 second s longer than a four-person crew to compl ete the same tasks. Finally, the mo st notable compar ison was between a two-person crew and a five-person crew — more than 2 mi nu tes (122 second s) difference in task completion time.

Industry Standard Ach ieved:

The “industry stand ard achieved” tim e started from the first engine arrival at the hydrant and ended w hen 15 firefighters were assemb led on scene. 19 An effective respo nse force w as assem bled by the five-person crews three m inutes faster than the four-person crew s. A ccording to study deployme nt protocal, the two- and three-person crews we re un able to assem ble eno ugh personnel to m eet this standard .

O ccupant Rescu e :

Three different “standard” fires (slow-, medium-, and fast-grow th rate) were sim ulated using the Fire Dynamics Simulator (FDS) model.

T he fires grew exponentially with time . The fire modeling simulations demonst rated that tw o-person, late arriving crews can face a fire that is twice the intensity of the fire faced by five-person, early arriving crews.

T he rescue scenario was based on a nonambulatory occupant in an upstairs bedroom with the bedroom door open.

Independ ent of fire size, there was a significant difference between the toxicity, expressed as fractional effective dose (FED), for occup ants at the time of rescue depend ing on arrival times for all crew sizes. Occu pants rescued by crews starting tasks two minutes earlier had lesser exposure to comb ustion products.

The fire modeling showed clearly that two-person crew s cannot compl ete essential firegrou nd tasks in time to rescue occupants withou t subjecting either firefighters or occupants to an increasingly hazardou s atmo sphere. Even for a slow-growth rate fire, the FED was approaching the level at which sensitive populations, such as children and the elderly are threatened. For a me dium-growth rate fire w ith two-person crews, the FED w as far above that threshold and approached the level affecting the median sensitivity in general population. For a fast-growth rate fire, the FED was well above the med ian level at which 50 % of the general population wou ld be incapacitated. Larger crews responding to slow-grow th rate fires can rescue most occupants prior to incapacitation along with early-arriving larger crews responding to me dium-growth rate fires. The result for late-arriving (two mi nu tes later than early-arriving) larger crews may result in a threat to sensitive popu lations for me dium-growth rate fires. ” The new sentence is consistent with our previous description for two-person crews where we identify a threat to sensitive popu lations..

Statistical averages should no t, however, mask the fact that there is no FED level so low that every occupant in every situation is safe. 52 18 NFPA Standard 1710 - A.5.2.4.2.1 …Other occupancies and structures in the community that present greater hazards should be addressed by additional fire fighter functions and additional responding personnel on the initial full alarm assignment. 19 NFPA 1710 Standard for the Organization and Deployment of Fire Suppression Operations, Emergency Medical Operations, and Special Operations to the Public by Career Fire Departments. Section 5.2.1 – Fire Suppression Capability and Section 5.2.2 Staffing.

53 Summar y:

The results of these field experiments contribute significant knowledge to the fire service industry. First, the results establish a technical basis for the effectiveness of company crew size and arrival time in N FPA 1710 . The results also provide valid measures of total effective response force assembly on scene for fireground operations, as w ell as the expected performance of time-to-critical-task meas ures for a low-hazard structure fires. Additionally, the results provide tenability measures associated with the occupant exposure rates to the range of fires considered by the fire model.

54 I
n order to realize a significant reduction in firefighter line-of-duty death (LODD) and injury, fire service leaders must focus directly on resource allocation and the deploym ent of resources, both contributing factors to LODD and injury. Future research should use similar met hods to evaluate firefighter resource deployment to fires in medium- and high-hazard structures, including multiple-family residences and comm ercial properties. Additionally, resource deploym ent to
multiple-casualty disasters or terrorism events should be studied to provide insight into levels of risks specific to individua l commu nities and to recomme nd resource deployment
proportionate to such risk. Future studies should continue to investigate the effects of resource deployment on the safety of both firefighters and the civilian population to better inform pu blic policy.

Future Research

55 Acknowledgeme nts
A
project of this magnitude extends significantly beyond the capabilities and expertise of the report authors. The
follow ing individu als were instrumen tal in the success of the experiments:
/square6T echnical Experts — Dennis Com pton, Retired Ch ief from Mesa , A Z and consultant, IFSTA; Russell Sanders, Retired Ch ief from Louisville, KY and staff, NFPA; William “Sho rty” Bryson, Retired Chief of Miam i, FL and Past President of Metrop olitan Fire Chiefs; David Roh r, Ope rations Chief from Fairfax Cou nty Fire and Rescue; Richard Bow ers, Chief from Montgo me ry Cou nty Fire and Rescue Departme nt; Vincent Dunn, Retired from Fire Departmen t of New Y ork; M ichael Clem ens, Chief of Training from Mon tgom ery Cou nty Fire and Rescue Departme nt; James W alsh, Battalion Chief from Fairfax Cou nty Fire and Rescue; Larry Jenkins, Captain I from Fairfax Cou nty Fire and Rescue; Doug Hinkle, Training
Ca ptain from Mo ntgome ry Coun ty Fire and Rescue
Departmen t; and Paul Neal, Safety Officer for Montgome ry Co unty Fire and Rescue Departm ent.
/square6M ontgome ry County Fire Departmen t – Form er Chief T om Ca rr and Chief Richard Bowers — AND Fairfax Coun ty Fire and Rescue Services — Chief Ron ald Mastin for supporting this study over a period of years to an unp recedented degree. /square6NIST experim ental and m odeling personnel - Mi chael Selepak, R oy M cLane, A nthon y C hakalis, And rew Lock, Marco Fernandez, Ed Hn etkovsky, Jay M cElroy, Lauren DeLauter, Glenn Forney, Dan M urphy, and C raig van Norman.
/square6IA FF Staff/ Data Entry/ Time r Supervision /Heart Rate Mo nitors– Nicole Taylor, Ra ndy G oldstein, and Ro n Benedict /square6Skidmo re C ollege – Denise Sm ith and Polar Heart Ra te Mo nitors for supplem ental study to bolster the significance of the m ain study results.
/square6T im ers — DeW ayne Dutrow — Lead, Cliff Berner, Michael Flem ing , C olby Po ore , Ch ris M aple , Mi chael Tho rnton , Ro bert Daley , Ry an Loh er
/square6M ontgom ery County Supp ort Services — Joey Fuller III — Lead , C hris Hinkle , Doug Dyer , Joey Fuller IV
/square6T he dedicated Fire O fficers and Firefighters from Mo ntgom ery County Fire & Re scue and Fairfax County Fire & R escue, wh o perform ed the difficult w ork of structural fire fighting safely and courageously.

56 Re ferences
/square6A lbert CM, Mittleman MA, Chae CU, Lee IM, Hen nekens CH, Ma nson JE (2000). Triggering of sudden death from cardiac causes by vigorous exertion. N Engl J Med
343(19):1355-1361.
/square6Backoff, R. W.; et al. (1980). Firefighter Effectiveness - A Prelimin ary Report. Columbus Fire Division, The Ohio State University.
/square6Barnard RJ, Duncan HW [1975]. Hea rt rate and ECG respon ses of firefighters. J Occup Med 17: 247-250.
/square6Blevins, L. G. and Pitts, W. M. (1999). Modeling of Bare and Asp irated Thermo couples in Co mpart me nt Fires. Fire Safety Journal, Vol. 33, 239-259.
/square6Bryant, R. A., et al. (2004). The NIST 3 Megawatt Quan titative Hea t Release Rate Facility - Description and Procedure. Natl.

Inst. Stand. T echno l. NIST IR 7052
/square6Ce ntaur Associates. (1982). Report on the Survey of Fire Suppression Crew Size Practices.
/square6Ce nter for Public Safety Excellence. (2008.) CFA I:

STAN DA RDS OF COVER, FIFTH EDITION . Chantilly, Va.
/square6Cent er for Public Safety Excellence. (2009.) FIRE & EMERGE NCY SERVICE SELF-ASSESSM ENT MAN UAL.

Cha ntilly, V A.
/square6Ch ang, C. Huang, H. (2005). A Water R equirements Estima tion Model for Fire Suppression: A Study Based on Integrated Uncertainty Analysis, Fire T echnology, Vol. 41, NO. 1, Pg. 5.
/square6C oleman, Ronny J. (1988). MANA GIN G FIR E SERVIC ES, 2nd Edition, International City/Co un ty M anageme nt Ass ociation , Washington, DC.
/square6C ushman , J. (1982). Report to Executive Board, Minimum Ma nning as Health & Safety Issue. Seattle, WA Fire
Departmen t, Seattle, WA.
/square6Ge rard, J.C. and Jacobsen, A.T. (1981). Reduced Staffing: At Wh at Cost?, Fire Service T oday, Pg. 15.
/square6Fahy R (2005). U.S. Firefighter Fatalities Due to Sudden Ca rdiac Death 1995-2004. NFPA Journal. 99(4): 44-47.
/square6H all, John R. Jr. (2006). U.S Unintentiona l Fire Death Rates by State. National Fire Protection Association, Quincy, MA. /square6Hu ggett, C. (1980). Estimation of the Rate of Heat Release by Mea ns of Oxygen Consump tion. J. of Fire and Flamm ability, Vol. 12, pp. 61-65.
/square6International Association of Fire Fighters/John’s Hopkins University. (1991). “Analysis of Fire Fighter Injuries and M inim um Staffing Per Piece of App aratus in Cities With Po pulation s of 150,000 or Mo re, ” Decem ber 1991.
/square6ISO (2007). ISO 13571: Life-threatening Co mp onents of Fire — G uidelines for the Estima tion of Tim e Ava ilable for Escape Using Fire Data, Internation al Stand ards Organization, Ge neva.
/square6Janssens, M . L. (1991). M easuring Rate of Heat Release by Oxy gen C onsum ption., Fire T echno logy, Vol. 27, pp. 234-249. /square6Jones, W. W. (2000). Forney, G. P .; Peacock, R. D.; Reneke, P . A.

T echnical Reference for CF AS T: An Engineering T ool for Estima ting Fire and Smo ke Transport. Nation al Institute of Standards and T echnology, G aithersburg, MD. NIST TN 1431; 190 p. M arch 2000.
/square6Ka rter, M.J. Jr. (2008). U.S. Fire Loss for 2007. NFPA Journal, Septemb er/O ctober 2008.
/square6Mc Grattan, K. B. (2006). Fire Dynam ics Sim ulator (V ersion 4): T echnical R eference G uide. NIST Gaithersburg, M D. NIST SP 1018; NIST Special Pub lication 1018; 109 p. March 2006. /square6Mc Ma nis Associates and Joh n T. O’H agan and A ssociates (1984). “Dallas Fire Departm ent Staffing Level Study, ” June 1984; pp. I-2 & II-1 through II-7.
/square6Me nker, W.K . (1994). Predicting Effectiveness of Ma nua l Supp ression , M S T hesis, Wo rcester Polytechn ic Institute. /square6Me tro Chiefs/Internation al Association of Fire Chi efs (1992) “M etro Fire C hiefs - M inimum Staffing Position, ” May 1992. /square6Mi ttlem an M A, Ma clure M , T ofler G H, Sherwoo d JB, Go ldb erg R J, Mul ler JE (1993). Triggering of acute m yocardial infarction by heavy physical exertion . N Engl J M ed
329(23):167 7–1683.
/square6Mo rrison, R . C . (1990). Ma nn ing Levels for Engine and Ladder Co m panies in Sma ll Fire Departm ents National Fire Ac adem y, Emm itsburg, MD .
/square6NFA (1981). Fire Engines are Becomi ng Expensive Taxi Cabs:

Inadequate Ma nning. National Fire Academy , United States Fire A dmi nistration, Emm itsburg, MD .
/square6NFPA (2007). NFPA 1403: Standard on Live Fire Training Evolutions. Nation al Fire Protection As sociation, Qui ncy, MA .
/square6NFPA (2004). NFPA 1710: Standard for the Organization and Deploym ent of Fire Suppression O peration s, Emergency Me dical Operations, and Special Operations to the Pub lic by Ca reer Fire Departm ents. National Fire Protection
As sociation, Q uincy, MA .
/square6NFPA (2008). Fire Protection H and bo ok, 20 th Edition.

National Fire Protection A ssociation, Quincy, MA.

57 /square6O ffice of the Fire Marshal of O ntario. (1993). Fire Ground Staffing and Delivery Systems Within a Compr ehensive Fire Safety Effectiveness Model. Ministry of the Solicitor General, T oronto, Ontario, Canada.
/square6O mega Engineering, Inc. (2004). The T emperature Han db ook . 5th Edition.
/square6Parker, W. J. (1984). Calculations of the Heat Release Rate by Ox ygen-Con sum ption for Various Applications., Journ al of Fire Sciences, Vol. 4, pp. 380-395.
/square6Phoenix, AZ Fire Departme nt, ” Fire Departmen t Evaluation System (FIREDAP), ” December 1991; p. 1.
/square6Purser, D. (2002). “T oxicity Assessmen t of Com bustion Produ cts. ” In The SFPE Handboo k of Fire Protection
Engineering, 3rd Edition. DiNenno (Editor). Nation al Fire Protection Association, Quincy, MA.
/square6R and Institute. (1978). Fire Severity and Response Distance:

Initial Findings. Santa Monica, CA.R oberts, B.
/square6R omet TT, Frim J (1987). Physiological responses to firefighting activities. Eur J Appl Physiol 56: 633-638. /square6Sardqvist, S; Holms ted, G., Correlation Between Firefighting Ope ration and Fire Area: Analysis of Statistics, Fire
T echno logy, Vol. 36, No. 2, Pg. 109, 2000
/square6Smith DL, Petruzzello SJ, Kramer JM, Warner SE, Bone BG , Mis ner JE (1995). Selected physiological and psychobiological respon ses of physical activity in different configurations of firefighting gear. Ergonomi cs 38(10): 2065-2077
/square6Smith, D. Effect of Deployment of Resources on Ca rdiovascular Strain of Firefighters. ” DH S, 2009.
/square6T hornton, W. (1917). The R elation of Oxygen to the Heat of Co mbust ion of Organic Compoun ds., Philosophical
Ma gazine and J. of Science, Vol. 33.
/square6TriData Corporation. The Econo mic Co nsequences of Firefighter Injuries and Their Prevention, Final Re po rt.

National Institute of Standards and T echnology, U.S.

Departmen t of Comme rce, Gaithersburg, MD. 2005.
/square6USFA (2002). Firefighter Fatality Retrospective Study. United States Fire Admi nistration
/square6USFA (2008). Fatal Fires, Vol. 5-Issue 1, March 2005. USFA, Firefighter Fatalities in the United States in 2007. June 2008.

Prepared by C2 T echnologies, Inc., for U.S. Fire
Ad ministration , Contract Number EME-2003-CO-0282.
/square6USNR C (2007). V erification and Validation of Selected Fire Mo dels for Nuclear Pow er Plant Applications. Volum e 2:

Experim ental Uncertainty. Washington, DC : United States Nuclear R egulatory C om m ission. 1824.

58 Page 967 of 1153
59 APPEN DIX A: LaboratoryExperiments
T
he firesuppression and
resourcedeploym ent
experim ents consistedof
fourdistinctparts:laboratory
experime nts,tim e-to-task
experime nts,room and contents
experime ntsand fire modelin g.

Th e pu rposeof thelaboratory
experime ntsw asto assure a fire
in the field experim entsthat
woul d con sistentlymeet NFP A
1403 requiremen tsfor live fire
training exercises.The
laboratoryexperiments enabled
investigatorsto characterize the
bu rning behaviorofthewood
palletsasa function of:
/square6numbe r of pallets and the
subsequen tpeak heatrelease
rate
/square6comp artmen teffectson burning of woo d pallets /square6effectofwin dow ventilationon the fire
/square6effecton fire grow th rateof the loading configuration of excelsior(slender w ood shavings typicallyused aspacking mat erial)
Des ignand C onstruction
FigureA-1 shows theexperimentalconfigurationforthe
compar tment palletburns.T wo identically sized compartments (3.66m x 4.88 m x 2.44m) were connectedby a hallway (4m x 1 m x 2.4m).Ateach end of the hallway,a singledoor connectedthe hallway toeach of thecomp artments.Intheburn compar tment,a single window (3 m x 2 m) was coveredwith no ncombu stible board thatw asopened for some experiments and closedforothers.

At the end of test,itwas opened toextinguishthe remaining burning materialand toremov e any debrispriortothenexttest.In the second compartm ent,a singledoorw ay connected the
compar tment to therestof thetestlaboratory.Itw askept open throughout thetestsallowingthe exhausttoflow intothemain collectionhoo d form easurement of heatreleaserate.

The structurewas constructedof two layerof gypsum wallboard over steelstuds.The floor ofthe structurewas lined with two layersofgypsum w allboard directlyover the concretefloorofthe testfacility.Intheburn com partmen t,an additiona llining of cem ent bo ard w asplacedoverthegypsum wa llsand ceiling surfacesnearthe firesource to min imize firedam age to the structureafter m ultiple fireexperiments.A doorway 0.91m wide by 1.92 m tallconn ected the burn compart men tto thehallway and an open ing 1 m by 2 m connectedthe hallwayto thetarget compa rtmen t.C eiling height was 2.41 m through out the structure,exceptfor the slight variation in theburn room.

FuelSo urce
The fuelsource for allofthe testswa s recycledhardw ood pallets constructed of severallengthsof hardwo od boards no minally83 m m wide by 12.7mm thick.Lengths ofthe ind ividualbo ards ranged from nom inally 1 m to 1.3 m .The finished size of a single palletw asapproxim ately1 m by 1.3 m by 0.11 m .Figure A- 2 show s the fuelsourcefor one ofthe tests including six stacked palletsand excelsior ignition source.Fo r an ignition source, excelsior w asplaced within the pallets,with the amo unt and location depending on the ignition scenario.F igure A- 3 show s thepallets prior to a slow and a fastignition scenariofire.Table A- 1 details the totalma ss ofpallets and excelsior for each ofthe free burn and comp artme nttests.

Experime ntalC onditions
The experiments w erecond ucted in tw o series.In thefirst series,heatrelease m easurem ents w erem ade und erfreeburn con dition s beneath a 6 m by 6 m ho od used to collectcomb ustion gases and provide the heatrelease rate (H RR) m easurem ent.A secon d series oftestsw asconducted with the firein a
com partm ented structureto assess environm entalcond itions within the structure during the fires and determine the effectof thecomp artm entenclosure on the firegrowth.Table A-1presents a sum ma ry of the tests con ducted.

Figure A -1. C ompa rtm entCo nfiguration and Instrumentation forPalletTests Figure A -2. Palletsand Excelsior Ignition So urce
U sed as a FuelSource

60 M easurem entsC onducted
H eatrelease rate(HR R )w as m easured in alltests.H RR m easurem ents w ereconducted un der the 3 m by 3 m calorime ter atthe NISTL arge F ireR esearch L aboratory.Th e H RR
m easurem entw as based on the oxygen con sump tion calorimetry principle firstproposed by Tho rnton (Thornton 191 7)and developed further by H uggett(Hug gett1980) and Parker(Parker 1984 ).This m ethod assume s thata kno wn am ount ofheatis released for each gram ofoxygen con sume d by a fire.The m easurem entofexhaustflow velocityand gasvolume fractions (O 2,C O 2 and CO )we reused to determine the HRR based on the form ulation derived by Parker (Parker 1984 )and Janssens (Janssens 198 1).The com bined expand ed relativeuncertaintyof theHR R me asureme ntsw asestim ated at± 14 %,based on a propagation ofun certainty analysis (Bryant2004).

Fo r the comp artm entfiretests,gas tem peraturem easurem ents were ma de in the burn comp artm entand in the target
com partm entconnected by a hallw ay to the burn comp artm ent using 24 gauge bare-bead chrome l-alum el(type K )
thermo couples position ed in verticalarray.Therm ocouples we re located atthe center ofeach comp artm entatlocations 0.03 m ,

0.30m,0.61 m ,0.91 m ,1.22 m,1.52 m ,1.83m,and 2.13 m from theceiling.The expanded uncertaintyassociated with a type K thermo couple is approxim ately ± 4.4 o C.(O me ga 2004) Ga s species w erecontinuo usly m onitored in the burn
com partm entata level0.91 m from the ceiling ata location centered on the side w allofthe com partme nt,0.91 m from the w all.Ox ygen w asm easured using param agnetic analyzers.

C arbon mo no xide and carbon dioxide w ereme asuredusing non -dispersive infrared (ND IR) analyzers.Allanalyzersw ere calibrated with nitrogen and a known concentration of gasprior to each testfor a zero and span con centration calibration.The expanded relative un certainty ofeach ofthe span gas m olar fractions is estim ated to be ± 1 %.

T otalheatflux wa s me asured on the side w allof theenclosure at a location centered on the side w all,0.61 m from the ceiling level.

The heatflux gauges w ere6.4m m diam eter Schmi dt-Boelter type, w ater cooled gauges with emb edded type-K thermo coup les (see Figure A- 4).The m anufacturer reports a ± 3 % expanded un certaintyin the respo nse calibration (the slop e in kW/m 2/mV).

C alibrations atthe NIST facilityhave varied within an additional ± 3 % ofm anufacturer’s calibration.Fo r thisstudy,an uncertainty of ± 6 % is estim ated.

TableA-1. TestsConduc ted and Ambient C onditions
atBe ginning ofEach Test
No tes: PA L stands for“pallet”and CR A (“Comm un ityRisk Assessment”)is thedesignator forthe configuration of pallets bu rned in thecom partm ent.Efforts w erem ade to use the same amo un tofexcelsior ma ssforCR A 2 (~0 .8kg),but thevaluewas not m easured.

FigureA- 3 . Fueland ExcelsiorSource forSlow (top)
and Fast(bottom )IgnitionScenarios
FigureA- 4:He atFlux Ga uge w ithRad iation Shielding

61 Re sults
Table A- 2 show s the peak H R R and tim e to peak H RR for the free burn tests and for the com partm ent tests.F igureA- 5 includ es ima gesfrom the free burn experiments nearthe tim e of peak H RR for each ofthe experiments.F igure A- 6 illustrates the progression ofthe firefrom the exitdo orway looking down the hallway to the burn com partm entfor one ofthe tests.F igure A- 7 to F igure A- 10 presentgraph s ofthe heatreleaserate for allof the tests.F igure A- 11 throug h F igure A- 15 show s the gastem perature, m ajor gas species concentration s,and heatfluxin theburn com partm entand targetcomp artm entin the five comp artm ent tests.

TableA- 2. Peak Hea tRelease Rate DuringSeveralPallet
TestsinFree-burnand ina C ompartment
FigureA - 5. Free-Burn Expe rimentsNear Time ofPeak Burning Page 970 of 1153 62 FigureA- 6. Examp leFire Progressionfrom TestCRA 1
FigureA-7. H RR, Slow Ignition,Free Burn Scenario
Figure A -8 . HR R ,FastIgntion,Free Burn Scenario
Figure A -9 . HR R ,Slow Ignition,C ompartment Test
Figure A -10. H RR ,FastIgnition,Co mpartmen tTest

63 FigureA - 11 . Tempe rature,Gas Concentration,and Heat Flux Du ringTestC RA 1,6 Pallets,Slow Ignition Scenario

64 Figure A - 12 . Tempe rature,Ga s C oncentration,and Heat Flux Du ringTestC RA 2,4 Pallets,Slow Ignition Scenario

65 FigureA- 13 . Temper ature,Gas Concentration,and Heat FluxDuring TestCR A 3, 4 Pallets,Fast Ignition Scenario

66 FigureA- 14 . Temp erature,Gas Concentration,and Heat FluxDu ringTestC RA 4,4 Pallets,Slow Ignition Scenario (Replicate)

67 Figure A- 15 . Temper ature,Ga s C oncentration,and Heat Flux Du ring TestC RA 5,4 Pallets,Slow Ignition Scenario (O pen W indow Venting )

68 B
ased upon theresultsof thelaboratory experime nts,the
projectteam determined thatfour palletswo uld provide
both a realistic firescenario,aswellasa repeatableand
we ll-characterized fuelsource.Varyingtheplacement and
quantityof excelsior provided significantvarianceintherateof fire growth.Prior tofinalizationof thefuelpackage and
con struction specifications,mod elingw asused toensurethatthe com bination offueland residentialgeometry would resultin un tenableconditions throughou tthestructurew ithoutsubjecting thefirefighterstounsafetestingcond itions.Therefore,CF AST (theconsolidated fireand smoke transportmodel (Jones 2000)) and F DS (firedynami cssimu lator mod el(McG rattan2006))w ere used to predictthetempe raturesand toxicspecieswithinthe structureasa function oftheexperime ntallydetermined heat releaserates.The resultssummari zed below confirme d thatthe building geometry and fuelpackageproduced adequatevariation in tenabilitycond itions in theresidentialstructure and ensured thattheroom oforigin w ould no treach flashovercond itions(a key provision ofNFP A 1403 ).Me eting theseconditions provided thefound ation for experime ntstome etthetwo primary objectives offiredepartment response:preservation of lifeand property.

APPEN DIX B: D esigning FuelPackages forFieldExperiments FigureB-1:Time-dependent tem peraturecontours infield structure with fastgrowth fire Figure B-2:Time -dependentsm oke densitycontoursinfieldstructurewithfastgrow thfire Figure B-1 and B -2 show the therm aland sm oke
condition s in the residentialstructureatdifferenttime
period s using the fastgrow th,four palletfuelpackage.

The resultsofthe firemo deling ind icated developmen t
ofuntenablecond itionsin the field experime nts
betwe en 5 and 15 m inutes,depending upon several
factors:fire growth rate,ventilation cond itions,the total leakage ofheatintothe building and through leakage
paths,and firefighterintervention.Thistime frame
allow ed fordifferentiation ofthe effectivenessofvarious fire departm entdeploym entm odels.

69 T
hroug h thegenerosityofthe M ontgome ry Count y (MD),
an open spacew asprovided to constructa temporary burn
prop atthe Mon tgome ry Co un tyFireand Rescue Training Fac ilityin Rock ville,MD. The areahad readyaccess to waterand electricalutilities.A licensed generalcontractor wa s retained, including a structuralengineerforthe design of criticalceiling memb ers,and theburn prop was constructedover a several month period in late200 8.

Th e bu rn prop consistedof two 2,000 ft. 2 (186 m 2)floors totaling 4,000ft. 2 ( 372 m 2).An exteriorview of two sides of the bu rn prop isshown in Figure C-1 .

Ad ditionalpartitions were installedby NISTstaffto createa floor plan representativeof a two-story,186 m 2 (2,000 ft. 2)single familyresidence.No tethat the structuredoes not have a
basement and includesno exposures.Th e overalldimension s are consistentwith thegeneralspecifications ofa typicallow hazard residentialstructurethatm any firedepartm ents respond to on a regularbasis,asdescribed in NFP A 1710 .

Furtherdetailsabou ttypicalsinglefamily ho me designs arenot provided in thestandard .Th erefore,a floorplan representativeof a typicalsinglefami lyhom e wa s createdby the projectteam .

De tails and floor plan dimen sions areshown in Figure C-2 .

The black lines indicateload-bearing reinforcedconcrete walls and red linesind icate the gypsum oversteelstud partition wa lls.

The ceiling height,no tshown in F igure C -2,is94 in. (2.4 m ) throug ho utthe entirestructure exceptin the burn comp artm ents, whe rethe ceiling height is93 in.(2.4 m ) .The purp ose ofthe partition w alls w as to symm etrically divide the structure about theshortaxis in orderto allow on e side ofthe teststructureto cooldown and dry-outaftera firetestwith suppression w hile con ducting experiments on the otherside .

The concrete w alls originalto the burn prop w ere8 in. (204 mm )thick steelreinforced poured
concrete and the floorson the first
leveland secon d levelsw ere4 in.
(102 m m )thick poured concrete.

Th e supp ortstructurefor the
second floor and the roof
consisted of corrugatedm etalpan
we lded to open w eb steeljoists.

Th e dim ensions ofthe joistsare
shown in F igure C- 3.The ceiling
wa s constructed from ½ in. (13
mm )thick cem entboard fastened
to the bottom chord of the steel
joists.Partition w alls w ere
constructed from 5/8 in. (17 mm )
thick gypsum panels attached to
20 gauge steelstuds fastened to
steeltrack,spaced 16 in. (407
mm )on center.

Add ition alconstruction wa s
im plem ented in the burn
comp artm ents to addresstherma l
loading and hose stream
im pingem entcon cerns.Spray-on
fireproofing wa s applied to the
steeljoistsprior to fastening the
ceiling,as shown in F igure C - 4.

Th e ceilings w ereconstructed
with three layersof½ in.(13 m m )
cem ent bo ard,asopp osed to on e
layer construction in the restof
the building.Ea ch layer was
fastened in a differentdirection so
thatseams ofadjacentlayersran
ortho gonally.Th e difference in
ceiling heights previously
APPEN DIX C: Temporary BurnProp C onstruction and Instrumentation FigureC-1:View oftwo sidesoftheburn prop
FigureC-2:Dimensions oftheBurn Prop FloorPlan

70 ment ioned isthe resultof the two additionalsheetsof cement board.Th e bu rn comp artmen twa llsw ereconstructedfrom a singlelayer of ½ in. (13 m m) cem ent board over a singlelayerof 5/8 in.( 16 mm) gypsum board,attached to 7/8in. (22 m m) offset me talfurringstrips.Particular carewa s takenso thatallceiling and partitionwallseams were filledwith chem ically-setting type joint com poun d to preventleakage into the interstitialspace between theceiling and thefloor above.After construction of the ceiling was comple te,a dry-standpipe delugesystem was installed with on e head in each burn room to provideemer gency
supp ression.Durin g an experiment,a 2.5 in.(104 m m) ballvalve fitting was attached and charged from a nearby hydrant.Figure C - 5 w astaken during the process ofreplacing “w orn out” ceiling panels and shows the add itionalconstruction im plem ented in the burn room asw ellasthe deluge sprinkler head.

W indow s and exterior doorsw ereconstructed to be
non -com bustible.Wi ndow s w erefabricated from 0.25 in.(10 m m) thick steelplateand the exterior doorsw ereof prefabricated hollow -core steeldesign.Th e window s on the firstfloor w ere30 in. (0.76 m) width x 36 in. (0.91 m) heightand 36 in. (0.91m) width x 40 in. (1.02 m) heighton the second floor.E xterior doors were 35.8 in.(0.8 8 m )width x 8 0 .5 in. (2.03 m )height.There were no do orsattached to the doorw ays inside the structure.

Figure C - 6 shows the con struction ofthe burn prop wind ow s as we llasthe NF PA 1403 -comp liantlatchm echanism.Figure C- 7 is a pictureofthe interior ofthe burn prop taken justoutside the burn comp artm ent,showin g the construction of the ceiling, interior doorw ay construction,gypsum wing walland the joint com po und used to sealseam s in the ceiling and walls.

Instrume ntation
A fterconstruction,theinstrumentation to measure the
propagation ofproductsofcombu stion was installedthrou ghout theburn prop.The instrumentation plan was designed tomeasur e gastemp erature,gasconcentrations,heatflux,visualobscuration, video ,and time during theexperime nts.The dataw ererecorded at intervalsof1 son a comp uter based dataacquisitionsystem.A schematic plan view oftheinstrumentation arrangement isshown in F igure C- 8.

TableC- 1 givesthe locations ofallof the instrume nts.

M easureme nts taken prior to
the comp artm entfire
experime ntswe relength,wo od
m oisturecontent,fuelma ss
and we ather conditions
(relative hum idity,
tem perature,wind speed and
direction).G as tem peratures
w ereme asured w ith tw o
different constructs of type K
(C hrom el-A lum el)
therm ocoup les.A ll
therm ocoup les outside the
burn com partm ents w ere
fabricated from 30 gauge
glass-wra pp ed therm ocoup le
wire.V erticalarraysofthree
therm ocoup les w ereplaced
near the front do or on the
north side and south sides of
the stairwellon the firstfloor.

O n the secon d floor,vertical
arraysofeighttherm ocouples
w ereplaced near the center of
each targetroom .Inside the
burn com partm ents,seven 3.2
m m (0.125 in.)exposed
junc tion thermo couplesand

0.76 m (30 in.)SUP ER
O M EGA C LA D XL ® sheathed
therm ocoup leprobeswe re
arranged in a floor-to-ceiling
array .F igure C - 9 show s the
verticalarray in the burn
FigureC-3: StructuralSteelD imensions
FigureC- 4:Fireproofingadded tostructuralsteel FigureC- 5:Additionalconstructionofburn room wallsand ceilingand deluge sprinklerhead.

FigureC- 6:Window & Latch Construction FigureC- 7:InteriorView ofBurn Prop

71 comp artm ent.Type K
therm ocouple probeswe re
chosen because oftheir ability to
withstand high tempe rature,
mo isture and physicalabuse
resulting from physicalcontact
with hose stream s and
firefighters.T o protectthe
extension wireand con nectors
from the effects ofheatand
wa ter,throug h-holesw eredrilled
in the burn com partme ntwa lls
and the sheaths w erepassed
through from the adjacent
comp artm ent.T o preventleakage
through the holes,allvoid spaces
we retightly packed with mi neral
wo ol.Inside the burn
comp artm entthe end of each
probe w aspassed through an
angle iron stand,and fastened to
the floor and ceiling to provide
addition alprotection from
physicalcon tactwith firefighters
and to ensure thatthe
me asureme ntlocation rem ained
fixed throug ho utthe
experiments.Incon sideration of
the risk associated with heating
the open w eb steeljoists,
addition althermo coupleswe re
placed above each burn
comp artm entto mo nitor the
temp erature ofthe interstitial
space.

FigureC-8: Instrumentation& FurnitureProp Layout

Gas concentration s were samp ledatthesame location in each targetroom .Bo th gasprobes w ereplumbed to the sam e analyzer and isolatedusing a switch valve;gaswas on lysam pledatone location during any giventest.The gas sam plingpoints were located in thecenterof the Wes twa ll(C Side)of both room s,1.5 m (5ft.)above thefloor.The sampling tubeswere connected to a diaph ragm pum p wh ichpulledthegas sam plesthrough stainless steelprobesinto a sample cond itioning system designed to eliminatem oisture in the gas sample.Th e dry gas sam ple was then piped to thegasanalyzersetup.Inallof theexperiments, oxygen was m easured using a parama gneticanalyzerand carbon mono xide and carbon dioxidewere mea sured using a
non-dispersive infrared(ND IR)analyzer.On e floor-to-ceiling therm ocou ple arraywas alsoco-locatedwith each sample port inlet.

Schmid t-Boelterheatfluxgauges were placedin theNo rthburn room .On e gauge wa s located1.0m (3.3ft.)above the floor and was oriented toward s thefireorigin (wastebasket).Th isheatflux gauge wa s placed to characterizetheradiativeheatflux attheface piecelevelthatwo uld be experienced by a firefighterinsidethe room .A second flux gauge was placed on thefloor in orderto characterizethe radiativeheatfluxfrom the upper layer and to mak e an estimate of how close the room was to flashing over with respectto time from ignition(using the com mon criteriaof flashover occurring at~2 0kW/m 2 atthe floorlevel).The heatflux gaugeswere co-locatedwith thethermo couple probe array.

Alllength me asurem ents we rem ade using a steelme asuring tape.Wo od m oisturecontent me asurem entsw eretaken using a non -insulated-pin type w oo d m oisturem eter.Fue lm asswa s m easured prior to each experimentusing a platform-style heavy duty ind ustrialscale.M ass w asnotm easuredafter each
experime nt because ofthe absorption offiresuppression wa ter.

Pub liclyaccessible D avisVantage Pro2 w eather instrum entation (availablevia http://ww w .wun derground .com )located
approxim ately tw o m iles from the experimentation sitewa s used to collectw eather data in five m inuteintervalsfor the each day thatthe experiments w ereconducted. F igure C - 10 isa
pho tograph of the W estw allof the N orth targetroom ,showing thetherm ocouple array,the smo ke obscuration me ter ,and a gas sampl ing probe used during the phase two experiments.The layoutis identicalto thatin the South targetroom .

No n-com bustible“prop” furniture wa s fabricated from angle iron stock and gypsum wa llboard.The purp ose ofthe furniture w astwo fold.The furniturew as placed inside the burn prop to simulate realistic obstacles w hich obscure the search pathsand hose stream advancem ent.The second use for the furniturewas so thatm easurem ent instrum entation could be strategically placed within the fram e ofthe furniture.This served to protect instrume ntation from physicaldam age as a resultof contactwith firefightersand their tools.Figure C - 11 show s an examp le of a table placed outside the burn room .

Allinstrum ents w erewired to a centralized data collection room, shown in Figure C -12,w hich w as attached as a separate space on one side of the building.This ensured physicalseparationfor the data collection person nelfrom the effectsof the fire,while m inim izing the wire and tube lengths to the data logging equipm ent.N ote thatthe roof ofthe instrum entroom was designed to serve as an add itionalme ans of escape for personnel from the second floor of the burn prop through a m etaldo or.A railing w as installed in order to m inim ize the fallrisk in theevent thatthe em ergency exitw asrequired.

FigureC- 9 :Burn Room Thermocouple Array FigureC- 10 :TargetRoom InstrumentCluster FigureC-11: Non-combustible“Prop”Table
72 Page 981 of 1153
73 FigureC- 12 :InstrumentationRoom
Outside Inside

74 Time-to-TaskData CollectionChart
Da te ______________ StartTim e __________ End Time (alltask com plete) __________ Time r Na me _____________________________________________________ ___________ APPEN DIX D: DataCollection and Com pany ProtocolsforTim e-to-Task Tests

75 Tasks/Compa ny
Arrive on Scene
-A rrive/stop athydrant
-Po sitionengine
______________
-Layout report
-On-scene report
-C ond uct size-up – 360 o
lap– incidentaction plan – offensive
– detailincident (situationreport)
-T ransmit size-up torespond ingunits
-Transfer com man d to chief
EstablishSupplyline
-Hydra nt-Drop line (wra p)
-Po sitionengine
-Pum p engaged
-4”straight lay
-----------------
-Supply attackengine
Po sition attackline
-Flake
-Cha rge
-Bleed
-----------------
-Advance
Establish-2 in – 2 out
(InitialRIT)
EstablishRIT
(D edicated)
Company Protocols:Crew Sizeof2
(10totalpersonnelon scene)
PLUS 4 RIC – 1403 = total14 needed
Engine 1/ 2
Driver
O fficer
-
Driver/O
Driver/O
Driver/O
O fficer– (No t
interior—j ust
frontdoor)
O fficer
Truck 1/2
-Arrive
-360 o lap
Po sition Truck
O fficer
O /D
O /D (perform s
allR ITduties)
Engine 2/2
-D ry L ay – 2nd
engine takes
hydrant
-Ch arged
hydrant
– Supply attack
engine
D river
Battalion C hief/Aide
-A rrives
-A ssum es C omm and
-Ev aluates R esources
-Establishes
C om ma nd post
-Ev aluates expo sure
problem s
-D irects hose
po sitioning
-C oordinatesUnits
-Transm its
Progress reports
-C hanges strategy
-O rders,records,and
transm its results of
prima ry and
secondary searches
-Dec lares fireund er
control
Engine 3 /2

76 Tasks/Compa ny
Gain/Fo rceEnt ry
Advance Line
- scan search fireroom
- suppression
De ployBack-up Line and protect
stairwell
Co mplete Primary Search
(incomb o with Fire Attack)
Search Fire Floor
Search other Floors
V entilation
(vent for fireor ventfor life)
-Hor izontal
-V entilation
Ground Laddering – 2nd story
windows,front and side,for
firefighter m eans ofegressand for
verticalventilation – 24’/28’and
roofladderin caseof verticalvent.

Co ntrolUtilities
(Interior and exterior)
Co nd uct Secondary Search
-Search FireFloors
-Search otherFloors
Chec k for F ire E xtension
Ope n ceilingwallsnearfireon fire
floor
Chec k floor above forfire
extension
-wallbreech
-ceiling breech
Mecha nicalV entilation
Engine 1/ 2
O fficer
(ifofficer commit s
then he mu stpass
comma nd)
Officer
Officer
Truck 1/2
O/D
Driver/Officer
Driver /Officer
Driver/Officer
Engine 2/2
O fficer
O fficer
O fficer
Battalion C hief/ A ide Engine 3 /2
O /D
O /D
Driver/Officer
O /D

77 Tasks/Compan y
A rrive on Scene
-A rrive/stopathydrant
-Position engine
______________
-Layout report
-O n-scene report
-C ondu ctsize-up – 360 o
lap – incidentaction plan – offensive
– detailincident(situation report)
-T ransm itsize-uptorespondingunits
-Transfercom man d to chief
E stablishSupply line
-Hy drant-Drop line (wrap)
-Position engine
-Pump engaged
-4”straightlay
-----------------
-Sup plyattackengine
Positionattack line
-Flake
-Ch arge
-Bleed
-Adv ance
E stablish-2 in – 2 out
(InitialRIT)
E stablishR IT
(D edicated)
Company Protocols:Crew Sizeof3
(14totalpersonnelon scene)
PLUS 4 RIC – 1403 = total18 needed
Engine 1/ 3
Driver
Officer
-
Driver
Driver
Driver
D/R B
Truck 1/3
-Arrive
-360 degree lap
PositionTruck
O/R B
Engine 2/3
D ry Lay – 2nd
engine takes
hydrant
C harged
hydrant–
Supply attack
engine
D river
O/R B— advance
by footto getto
point ofentry –
performs allRIT
duties
Battalion C hief/ A ide
-Arrives
-Assume s C omm and
-E valuates R esources
-Establishes
C om m and post
-E valuates expo sure
problem s
-D irects hose
po sition ing
-C oordinatesUnits
-Transm its
Progress reports
-C hanges strategy
-O rders,records,and
transm its results of
prim ary and
secon dary searches
-Dec lares fireun der
control
Engine 3 /2

78 Tasks/Compan y
Ga in/For ce En try
Advance Line
-scan search fire room
-sup pression
D eployBack-up Line and
protectstairwell
C om plete Prim ary Search
(in com bo with Fire Attack)
Search FireFloor
Search other Floors
V entilation
(vent for fire or ventfor life)
-H orizon tal
-V entilation
Ground Laddering – 2nd story
w indows,frontand side,for
firefighterme ans of egressand for
verticalventilation – 24’/28’and
roof ladder in caseof verticalvent.

C on trolUtilities
(Interiorand exterior)
C on duct Secondary Search
-Search F ire Floors
-Search otherFloors
Ch eck for FireEx tension
O pen ceiling walls nearfireon fire
floor
Ch eck floorabove for fire
extension
-wa llbreech
-ceiling breech
Mech anicalV entilation
Engine 1/ 3
O/R B
(ifofficercommi ts
then he mu stpass
comm and)
O/R B
Truck 1/3
O/RB
O/ R B
-
Driver
Driver
D river (exterior)
O/RB (Interior)
O/RB
Driver
Engine 2/ 3 Battalion C hief/ A ide Engine 3 /3
O /RB
D river
D river
D river
(exterior)
O /RB
D river

79 Tasks/Compan y
A rrive on Scene
-A rrive/stopathydrant
-Position engine
______________
-Layout report
-O n-scene report
-C ondu ctsize-up – 360 o
lap – incidentaction plan – offensive
– detailincident(situation report)
-T ransm itsize-uptorespondingunits
-Transfercom man d to chief
E stablishSupply line
-Hy drant-Drop line (wrap)
-Position engine
-Pump engaged
-4”straightlay
-----------------
-Sup plyattackengine(1 3/4”)
Positionattack line
-Flake
-Ch arge
-Bleed
-Adv ance
E stablish-2 in – 2 out
(InitialRIT)
E stablishR IT
(D edicated)
Company Protocols:Crew Sizeof4
Totalon scene = 18
PLUS 4 RIC – 1403 = total22 needed
Engine 1/ 4
Driver
Officer
-
Driver
Driver
Driver
RB/N ozzle
LB/Flake
Both advance line
forfire attack
Truck 1/4
-Arrive
-360 degree lap
PositionTruck
D/L B
Engine 2/4
-D ry L ay – 2nd
engine takes
hydrant
C harged
hydrant–
Supply attack
engine
D river
O /LB/RB—
advanceby foot
togettopointof
entry– performs
allR IT duties
Battalion C hief/ A ide
-Arrives
-Assume s C omm and
-E valuates R esources
-Establishes
C om m and post
-E valuates expo sure
problem s
-D irects hose
po sition ing
-C oordinatesUnits
-Transm its
Progress reports
-C hanges strategy
-O rders,records,and
transm its results of
prim ary and
secon dary searches
-Dec lares fireun der
control
Engine 3 /4

80 Tasks/Co mpan y
Ga in/For ce En try
Advance Line
-scan search fire room
-sup pression
De ployBack-up Line and
protectstairwell
Com plete Prim ary Search
(in com bo with Fire Attack)
Search FireFloor
Search other Floors
V entilation
-H orizontal
-V entilation
Ground L addering – 2nd story
w indows,frontand side,for
firefighterme ans of egressand for
verticalventilation – 24’/28’and
roof ladder in caseof verticalvent.

Con trolUtilities
(Interiorand exterior)
Con duct Secondary Search
-Search Fire Floors
-Search otherFloors
Ch eck for FireEx tension
O pen ceiling walls nearfireon fire
floor
Ch eck floor above for fire
extension
-wa llbreech
-ceiling breech
Mech anicalV entilation
Engine 1/ 4
RB/L B
O fficer – no ton line
(ifofficer commi ts
then he m ustpass
command)
O/R B
Truck 1/4
O/R B
Officerand R B
-
Driverand L B
Driver/LB
Driver/LB
(controlexterior)
O/R B
(controlinterior)
O/R B
D/LB
Ba ttalion C hief/ A ide Engine 3 /4
O /RB
D /LB

81 Tasks/Compa ny
Arrive on Scene
-A rrive/stopathydrant
-Po sitionengine
______________
-Layout report
-On-scene report
-Lo cateFire
-C ond uct size-up – 360 o
lap– incidentaction plan – offensive
– detailincident (situationreport)
-T ransm itsize-up torespond ingunits
-Transfer com man d to chief
EstablishSupplyline
-Hydra nt-Drop line (wra p)
-Po sitionengine
-Pum p engaged
-4”straight lay
-----------------
-Supply attack engine (1 3/4”)
Po sition attackline
-Flake
-Cha rge
-Bleed
-Advance
Establish-2 in – 2 out
(InitialRIT)
Company Protocols:Crew Sizeof5
D/O/LB/RB/CB Totalon scene = 22
PLUS 4 RIC – 1403 = total26 needed
Engine 1/ 5
Driver
O fficer
-
Driver
Driver
Driver
R B/Nozzle
LB/Flake
CB /C ontrol
---------------
Ad vance line for
fire attack
----------------
The Officer
responsibilityis
to sup erviseho se
stretch /mon itor
safety and
continuallysurvey
the scene
Truck 1/5
-Arrive
-360 degree
Size up .

Po sition Truck
D /LB
Engine 2/5
-Dr y L ay – 2nd
engine takes
hydrant
C harged
hydrant–
Supply attack
engine
D river
Battalion C hief/Aide
-A rrives
-A ssum es C omm and
-Ev aluates R esources
-Establishes
C om ma nd post
-Ev aluates expo sure
problem s
-D irects hose
po sitioning
-C oordinatesUnits
-Transm its
Progress reports
-C hanges strategy
-O rders,records,and
transm its results of
prima ry and
secondary searches
-Dec lares fireund er
control
Engine 3 /4

82 Tasks/Co mpan y
E stablishR IT
(D edicated)
Ga in/For ce En try
Advance Line
-scan search fire room
-sup pression
Insuresfirstline flowingwater—
De ployBack-up Line and protect
stairwell(1 ¾ ”)
Com plete Prim ary Search
(in com bo with Fire Attack)
Search FireFloor –
Search other floors-
V entilation(ventforfireorventforlife)
- H orizontal
- V ertical
Ground L addering – 2nd story
w indows,frontand side,for
firefighterme ans of egressand for
verticalventilation – 24’/28’and roof
ladderin caseof verticalvent.

Con trolUtilities aftersearch,force
entry,venting and fireextinguished
(Interiorand exterior)
Con duct Secondary Search
-FireFloor
-Prima ry and secondary search of
entirefloor above
Ch eck for FireEx tension
O pen ceiling walls nearfireon fire
floor
Ch eck floor above for fire
extension
wa llbreech
ceiling breech-
Mech anicalV entilation
Engine 1/ 5

RB/L B/CB
Officer– not on
line (ifofficer
commi tsthen he
mus tpass
comma nd)
O/R B
Truck 1/5

O /RB/CB
Officerand
RB/C B
Driverand L B
Driver/LB
Driver/LB
(controlexterior)

O/R B/CB
(controlinterior)
D/LB
Engine 2/ 5
O /L B/R B—
advance by foot
to getto point
ofentry –
perform s all
RIT duties
Ba ttalion C hief/ A ide Engine 3 /5

O /RB/CB
D /LB

O /RB/CB

O /RB/CB

83 A ppendix E: StatisticalA nalysis ofTim e to Task TestD ata Identifying Statistically SignificantD ifferences in C rew Size and Staggeron a N um berofT ask Tim ings U sing R egression Analyses ofTim es (Start, End and D uration) on C rew Size and Stagger

84 App endixF: AllRe gression C oefficients
Regression M odelsofTime toTask (inSeconds) as a Func tion ofCrew Size and Stagger (Standard ErrorsareinParentheses und erneathcoefficients)

85 AllR egressionC oefficients(CONTINUED )
Regression M odelsofTime toTask (inSeconds) as a Func tion ofCrew Size and Stagger (Standard ErrorsareinParentheses und erneathcoefficients)

86 Regression M odelsofTime toTask (inSeconds) as a Func tion ofCombin ed Crew Sizeand Stagger(StandardErrorsappearinParentheses)

87 Regression M odelsofTime toTask (inSeconds) as a Func tion ofCombin ed Crew Sizeand Stagger(CONTINUED) (StandardErrorsappearin Parentheses)

88 T
he me asureme ntsof length,temperature,mas s,moisture
content,smo ke obscuration ,and time taken inthese
experimentshave unique comp onentsof un certaintythat
mu stbe evaluatedinorder todeterminethe fidelityof thedata.

Thesecompo nentsof uncertainty can be grouped into two
categories:Type A and Type B.Type A uncertaintiesarethose evaluated by statisticalmethod s,such ascalculatingthestandard deviationof themean of a setof measurements.Type B
un certaintiesarebased on scientificjud gment usingallavailable and relevantinformation.Usingrelevantinformation,theuppe r and lower limitsoftheexpected valueareestimated so thatthe probabilitythatthem easurement fallswithintheselimitsis essentially10 0 %.Afterallthecomp onent uncertaintiesof a measu rement have been identifiedand evaluated itisnecessaryto use them tocompute thecombine d standarduncertaintyusingthe law of propagation of uncertainty(the “rootsum of squares”).

Althou gh thisexpressesthe uncertainty of a given measurement,it ismore usefulin a firemode lvalidation exercisetodefinean intervalforwhich themeasurement w illfallwithina certainlevelof statisticalconfidence.Thisisknown asthe expanded uncertainty.

Th e currentinternationalpractice istomu ltiplythecombined standard uncertainty by a factoroftwo (k=2),givinga confidence of 95 %.

Length m easurements of room dimensions,openingsand
instrument locations were taken usinga steelmeasu ringtapewith a resolutionof 0.02 in(0.5 mm).H owev er,measurement errordu e to un even and unlevelsurfacesresultsinan estimated uncertaintyof ±

0.5% forlength me asurement staken on thescale of room
dime nsions.The estimated totalexpanded uncertainty forlength measu rements is± 1.0%.

Th e standard uncertainty of thethermocou ple wireitselfis1.1°C or 0.4% of themeas ured value,whicheverisgreater(Omega 2004).

Th e estimatedtotalexpanded uncertainty associatedwith type K therm ocoup lesisapproximately ± 15 %.Previouswork done at NISThas shown thattheuncertainty of the environment
surrounding thermo couplesina full-scale fireexperiment has a significantlygreaterun certainty(Blevins 1999) than the un certaintyinherent with thermocoup le design.F urthermo re, wh ile a verticaltherm ocoup learray givesa good approximationof the tem peraturegradientwith respecttoheight,temperatures cannotbe expected tobe uniform acrossa planeatany height becauseof thedynam icenvironm entina compartment fire.

Inaccuraciesofthermocoup lemeasurements ina fireenvironment can be causedby:
/square6R adiative heating or coolingof the thermocouple bead /square6Soot deposition on thethermoc ouple bead which change its m ass,em issivity ,and thermalcond uctivity
/square6H eatconduction along thermoc oup lewires
/square6Flow velocityoverthethermoc ouple bead
T oreducethese effects,particularlyradiativeheatingand cooling, therm ocoup leswith sm allerdiameterbeadswere chosen.Thisis particularlyimpo rtantforthermocoup lesbelow theinterface becausetheradiativetransferbetween thesurrounding room
surfaceswillbe significantly lessuniform than ifthethermocouple we reinthehot gaslayer.Itissuggestedin[Pitts]thatitmay be possibleto correctforradiative transfergiven enough sufficient knowledge about thermoc oup lepropertiesand theenvironment.

Ho wev er,me asureme ntsoflocalvelocityand theradiative environment we renot taken.Ad ditionally,theprobeswere located awa y from theburn comp artment wallsin order to avoidthe effects ofwallsand corners.

The gasme asureme nt instrume ntsand sampling systemused in thisseriesofexperime ntshave been demons trated tohave an expanded (k= 2)relative uncertainty of± 1 % when compar ed with span gasvolume fractions (Matheson).Given thelimitedsetof samp ling pointsin these experime nts ,an estimateduncertaintyof ± 10 % isbeing applied to theresults.

The potentialfor sootdeposition on theface of thewater-cooled totalheatflux gaugescontributessignificantuncertaintyto the heat flux me asureme nts.C alibration ofheatflux gaugeswas comple ted atlower fluxesand then extrapolated tohigher valuesand this resulted in a higher uncertainty in theflux measu rement.

C omb ining allofcomp onent uncertaintiesfor totalheatflux resulted in a totalexpand ed uncertainty of -24 % to +13 % forthe flux me asureme nts.

Prior to experime ntation,ten of the wo oden palletsused inthe fuelpackageswe rerando mly selected for meas urement .T wo me asureme ntswe re taken,mo isture contentand mass.Moisture content was me asured using a pin-type moisture meter w itha mo isture meas ureme nt range of 6 % to 40% and an accuracyof < 0.5% of the me asured value betwe en 6 % and 12 % moisture content.M assme asureme ntsw erema de w ithan industrialbench scalehaving a range of0kg to 100 kg,a resolutionof 0.1 kg and an uncertainty of± 0.1kg.

Alltiming staffw ere equipped w ith thesame mod elof digital stopwatch with a resolution of 0.01seconds and an uncertaintyof ± 3 second sper24 hours;the uncertainty of the timingmec hanism in the stopw atchesissmallenough overthedu ration of an experiment thatitcan be neglected.Therearethree comp onentsof uncertainty whe n using peopleto time fire fighting tasks.First,timersmay have a biasdepending on whether they record thetime inanticipation of,or reaction to an event.A second compone nt existsbecause multipletimer swe reused torecord alltasks.The thirdcomponent isthemo de ofthe stimu lus to which thestaffisreacting : audible (firefightersanno uncing taskup datesovertheradio)or visual (timing staffseesa task startor stop).

M ilestone eventsin these experiment sw ere recorded both audibly and visually.A testseriesdescribed in the N IST R ecomme nded PracticeG uide forStopwatch and Tim erC alibrations found the reaction time sfor thetwo mo desofstimul us tobe approximately the same ,so thiscompo nent can be neglected.Because of thelack ofknowl edge regarding theme an biasofthetimers,a rectangular distribution was assume d and thewo rstcasereaction time biasof 120 ms w asused,giving a standard deviationof 69 ms.The standard deviation ofthereaction time was assumed to be the wo rstcaseof230 ms.The estimat ed totalexpanded uncertaintyof task time sme asured in these experime ntsis240 ms .

An additionalcomp onent ofuncertainty existsfor thetime me asureme nt oftheapplication ofwater on thefire.In orderto me asure thistime ,timi ng staffwe re required tolistenforradio confirmation thatsuppressing waterhad been appliedby the interior attack crew .Thisprocessrequired a member of theinterior crew to find and mani pu late theirmi crophone,waitfortheradioto accessa repeater,and transmittheme ssage.Because of thelackof APPEN DIX G: Measurement Uncertainty