From:
To: Emi Sugiyama
Subject: [City of Cupertino] Assignment: Technical Concern: Cumulative Evacuation System Failure on the Linda Vista Corridor (re: Evulich Court Project)
Date: Monday, March 2, 2026 7:52:28 AM
Attachments: Technical Rebuttal to Hexagon Transportation Analysis.pdf Letter to Assistant Chief Hector Estrada.pdf
260224 PC Public Hearing - Emergency Evacuation, Evulich Project .pdf Evacuation Traffic Model and Analysis.pdf
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
You have been assigned to this ticket (#156808).
DerChang Kau
Feb 27, 2026, 3:23 AM PST
To: Assistant Chief Hector Estrada, Santa Clara County Fire Department
CC: Cupertino City Council & Clark, Cupertino Planning Commissioners & division, Mercury Newspaper, Alice Lin and Frank Swanson
From: DerChang Kau
Date: Feb. 27, 2026
Subject: Technical Concern: Cumulative Evacuation System Failure on the Linda Vista Corridor (re: Evulich Court Project)
Dear Assistant Chief Estrada,
I am DerChang, a 3-decade Cupertinian. I am writing to you as a resident of the Linda Vista corridor. I know you are a hands-on leader who prioritizes public safety and operational readiness over theoretical paperwork.
This is why I am bringing a critical "cumulative impact" safety concern to your attention regarding the proposed 51-unit townhome development at 10857 Linda Vista Drive.
I am seeking your advocacy regarding the proposed 51-unit development at Evulich Court. Currently, your department’s "Finding of Consistency" is being used by the developer to bypass CEQA environmental review. However, that finding appears limited to the interior
of the project site (hydrant pressure, turn radii, and hose- pull).
We are deeply concerned that the “cumulative impact” on the Linda Vista "Spine" during a wildfire has not been modeled and the risk is not properly assessed. The implications for life safety are catastrophic.
The Tactical Reality: A Single-Point-of-Failure
Imagine a wind-driven wildfire encroaching from the South/Southwest (Stevens Creek County Park). The tree canopy over Linda Vista
becomes a "fuel tunnel." Residents have only one exit: North, funneling toward East bound of McClellan Road.
In this scenario, the Fire Department’s ability to defend the neighborhood depends on the residents being able to clear the road.
Our analytical modeling—based on the current VHFHSZ classification— shows that at the proposed R3 density, the addition of 51 high-density units will push the road Utilization Ratio of 110%.
Mathematically,
this creates a "System Failure." Once the evacuation surge begins, cars will be trapped in a standing queue.
From an operational
standpoint, this is a significant "trap hazard." If the road capacity is exceeded, the bottleneck at the intersection will prevent not just residents from leaving, but emergency vehicles from efficiently entering the corridor to stage for structure protection or rescue operations.
The Hexagon Report Gap: The developer's traffic study models only 30 cars per hour.
The Evacuation Reality: We are facing a surge of 250+
vehicles attempting to exit in a 30-minute window.
The Result: A mathematical "System Failure" where the exit rate at McClellan is slower than the arrival rate. This creates a standing queue that stretches the length of the spine.
Why the Current Assessment is Incomplete
The developer’s counsel argues that because the project meets internal Fire Code, it is exempt from further CEQA review. We believe this is an operational oversight.
1 . Merging Shockwaves: Concentrating 51 units at one entry point creates a "forced merge" scenario. Our models show this adds 28 minutes
of stutter-delay to the uphill residents already
in the queue.
2 . Order-to-Safety Time: Total evacuation time jumps from a survivable less than 30 minutes to 90+ minutes
. In a fire
moving 200 acres in 20 minutes, a 90 plus-minute bottleneck is a non-survivable event.
Our Request to the Fire Department
We believe this project cannot be adequately reviewed through a standard site-plan exemption. To ensure the safety of our neighborhood, we respectfully ask the Fire Department to:
1. Expand the Scope:
Review the cumulative impact of this
density on the entire Linda Vista evacuation corridor, rather than just the project interior.
2 . Model the "Surge": Evaluate whether the egress route can maintain a "Total Order to Safety" time within the 30-minute critical fire-spread window (200 acres in 20 minutes) as established by recent regional wildfire events (eg. 2025 Pacific Palisade Fire).
If our neighborhood’s evacuation capacity is mathematically compromised, we believe this project does not meet the "safety
To: Assistant Chief Hector Estrada, Santa Clara County Fire Department CC: Cupertino City Council & Clark, Cupertino Planning Commissioners & division From: DerChang Kau
Date: Feb. 27, 2026
Subject: Technical Concern: Cumulative Evacuation System Failure on the Linda Vista Corridor (re: Evulich Court Project)
Dear Assistant Chief Estrada,
I am DerChang, a 3-decade Cupertinian. I am writing to you as a resident of the Linda Vista corridor. I know you are a hands-on leader who prioritizes public safety and operational readiness over theoretical paperwork.
This is why I am bringing a critical "cumulative impact" safety concern to your attention regarding the proposed 51-unit townhome development at 10857 Linda Vista Drive.
I am seeking your advocacy regarding the proposed 51-unit development at Evulich Court. Currently, your department’s "Finding of Consistency" is being used by the developer to bypass CEQA environmental review.
However, that finding appears limited to the interior of the project site (hydrant pressure, turn radii, and hose-pull).
We are deeply concerned that the “cumulative impact” on the Linda Vista "Spine" during a wildfire has not been modeled and the risk is not properly assessed. The implications for life safety are catastrophic.
The Tactical Reality: A Single-Point-of-Failure
Imagine a wind-driven wildfire encroaching from the South/Southwest (Stevens Creek County Park). The tree canopy over Linda Vista becomes a "fuel tunnel." Residents have only one exit: North, funneling toward East bound of McClellan Road.
In this scenario, the Fire Department’s ability to defend the neighborhood depends on the residents being able to clear the road. Our analytical modeling—based on the current VHFHSZ classification— shows that at the proposed R3 density, the addition of 51 high-density units will push the road Utilization Ratio of 110%.
Mathematically, this creates a "System Failure." Once the evacuation surge begins, cars will be trapped in a standing queue. From an operational standpoint, this is a significant "trap hazard." If the road capacity is exceeded, the bottleneck at the intersection will prevent not just residents from leaving, but emergency vehicles from efficiently entering the corridor to stage for structure protection or rescue operations. • The Hexagon Report Gap: The developer's traffic study models only 30 cars per hour. • The Evacuation Reality: We are facing a surge of 250+ vehicles attempting to exit in a 30-minute window.
• The Result: A mathematical "System Failure" where the exit rate at McClellan is slower than the arrival rate. This creates a standing queue that stretches the length of the spine.
Why the Current Assessment is Incomplete
The developer’s counsel argues that because the project meets internal Fire Code, it is exempt from further CEQA review. We believe this is an operational oversight.
1. Merging Shockwaves: Concentrating 51 units at one entry point creates a "forced merge" scenario.
Our models show this adds 28 minutes of stutter-delay to the uphill residents already in the queue.
2. Order-to-Safety Time: Total evacuation time jumps from a survivable less than 30 minutes to 90+ minutes. In a fire moving 200 acres in 20 minutes, a 90 plus-minute bottleneck is a non-survivable event.
Our Request to the Fire Department
We believe this project cannot be adequately reviewed through a standard site-plan exemption. To ensure the safety of our neighborhood, we respectfully ask the Fire Department to:
1. Expand the Scope: Review the cumulative impact of this density on the entire Linda Vista evacuation corridor, rather than just the project interior.
2. Model the "Surge": Evaluate whether the egress route can maintain a "Total Order to Safety" time within the 30-minute critical fire-spread window (200 acres in 20 minutes) as established by recent regional wildfire events (eg. 2025 Pacific Palisade Fire).
If our neighborhood’s evacuation capacity is mathematically compromised, we believe this project does not meet the "safety requirements" necessary for a CEQA exemption. We cannot allow "paper compliance" to override the physical reality of a wildfire. We would welcome the opportunity to walk the corridor with you or a member of your team to show you the physical bottlenecks that our models have identified.
Again, our ask is your review of Emergency Evacuation Corridor Analysis for the entire "Spine," not just the project interior. Specifically, we request that the Fire Department clarify to the Planning Commission that:
• Internal site compliance does not equate to neighborhood evacuation safety. • A CEQA exemption is inappropriate without a Cumulative Evacuation Surge Study that accounts for the 155-unit total load on the Linda Vista bottleneck.
Attached you will find 4 documents
1. Detailing the evacuation model and assumption
2. Our plead to the public hearing of Cupertino City Planning Commission.
3. Technical Rebuttal to Hexagon Transportation Analysis
4. This letter
Thank you for your dedication to the safety of our community. We are eager to provide you with our full analytical data package to assist in your review.
Respectfully,
DerChang Kau
Intel Fellow (Ret)
Adjunct Professor, National Taiwan University
Wildfire Accident Risk Overview
Hyper-Local Risks for 10807–11197 Linda Vista
The "Choke Point": The lower part of Linda Vista Drive (closer to McClellan) can become congested. If you receive an Evacuation Warning, it is highly recommended to leave early, especially if you have trailers, pets, or large vehicles.
Vegetation Overhang: This stretch of road has significant tree canopy. In a high-wind fire event, "spotting" (embers jumping ahead of the fire) can ignite trees along the road, potentially blocking the only exit. Dead-End Cul-de-Sacs: If you live on a side street off this stretch (like Baxley Court or various private drives), you must be aware that you have zero "rear" exits.
Evacuation Density Risks
The density is a major point of concern for fire safety because:
Funneling: All 150+ households (plus residents on side streets like Rae Lane and Baxley Court ) must use Linda Vista Drive to reach McClellan Road.
The Park Choke Point: The entrance to Linda Vista Park is expected to be the sole entry/exit for the new Vista Heights development, adding concentrated traffic to the middle of your evacuation route. Vehicle Load: With an average of 2.9 to 3.1 people per household in Monta Vista, a full evacuation of just this street segment could involve 450+ people and 250+ vehicles attempting to exit simultaneously onto a winding two-lane road.
3 Evacuation Plan Analysis
Current Neighborhood (Approx. 70 Homes)
— At an average of 1.6 vehicles per household (typical for emergency evacuations where families try to consolidate), there are roughly 120 vehicles that must exit this specific stretch. — Flow Rate: A single lane of a winding residential road under stress typically handles 600–800 vehicles per hour.
— Physical Exit Time: Technically, 120 cars could pass a single point in about 12–15 minutes. — The "Human Factor" Delay: In reality, total evacuation time includes "mobilization time" (packing, hitching trailers, getting pets). For 70 homes, the neighborhood would likely be clear in 30 minutes from the time the order is issued, assuming no road blockages. Expanded Neighborhood (Planned ~150+ Units)
— With the addition of the Evulich Court (51 units) and Vista Heights (35 units) projects, the vehicle count jumps to approximately 240–260 vehicles.
— The Bottleneck: The main issue isn't just the number of cars; it’s the intersection of Bubb Road and McClellan. If traffic lights are out, this intersection becomes a four-way stop, which can drop flow rates by 50%.
— Estimated Time: Under these conditions, the physical exit time doubles. Adding in the complexity of more neighbors merging from high-density driveways, the total time to clear the street would likely stretch to 90+ minutes.
5 Utilization Ratio – ρ = Demand / Capacity
Traffic is like running water in the pipe. The Utilization Ratio is the pressure in that system. When ρ exceeds 1.0, the "pipes" burst—meaning the line of cars stops moving, and the time to exit becomes theoretically infinite unless the demand is reduced.
Demand – λ
30-minute requirement is the safety leave window from the learning of Pacific Palisade Fire. R1 Case: λbaseline = 70 Units × 1.6 vehicles/household / 0.5 hours = 224 cars / hour R3 Case: λplanned = 155 Units × 1.6 vehicles/household / 0.5 hours = 496 cars / hour.
Capacity – Ideal flow rate, µideal = 700 cars/hour.
The intersection of Bubb Road and McClellan is the only exit in an "Emergency State”. — Power is out, the traffic signals are dead and the intersection functions as a 4-Way Stop. Higher density creates a "standing queue” and degrades 4-way stop intersection efficiency from ~85% to ~60% — R1 Case: µbaseline = 600 cars / hour
— R3 Case: µplanned = 450 cars / hour.
Metric Baseline (R1 status) Planned (R3 status) Analytical Meaning Demand – λ 224 cars/hr 496 cars/hr The "Pressure"
Capacity – µ 600 cars/hr 450 cars/hr The "Pipe Size"
Utilization Ratio – ρ 37% 110% Result
Queuing Delay (DQ)
Queue will be formed when transit time between cars in traffic flow is shorter than Queuing Delay Transit time (tT) is inverse of flow rate at the demand capacity (λ) Queuing delay (DQ) @ 50% utilization ratio (ρ) is defined as intrinsic delay Time (tintrinsic), 8 seconds is used for this analysis • Queuing delay is not linear, DQ = ρ / (1- ρ) x tintrinsic R1 scenario – the system is ”Quiet." There are plenty of gaps between cars. tT,basel i ne = 1 / λbaseline = 16 sec/car vs. DQ, baseline = 4.6 sec/ cars tT,baseline > Dq, baseline è the "backlog" never builds up ~2.5 minutes to account for the very tail end of the line (~30 cars) R3 scenario – the system is "Over-Saturated." Cars are arriving faster than the intersection can process them. "Standing Queue" forms. 248 cars want to leave in 30 mins. vs. µplanned@ 450 cars/hr flow rate, the intersection can only handle 225 cars in 30 mins 23 cars are physically unable to leave the street during the safe window.
Those 23 cars, plus the massive slow-down caused by the 225 cars ahead of them crawling at a 110% saturation rate, results in a queue that stretches nearly the entire length of the Linda Vista "spine." In "Over-Saturation” traffic simulation, the average wait time for the last car in line jumps to 45+ minutes just to reach the stop sign.
Baseline (112 cars , R1) Planned (248 cars, R3)
Utilization 37% 110%
Wait per Car ~5 Seconds Infinite/Growing
Wait Time Logic You wait for 5-10 cars. You wait for 247 cars to clear a 1-lane exit.
Resulting Delay 2.5 minutes 45+ mins
Merging Delay (DM)
Based on Gap Acceptance Theory in traffic engineering – the difficulty of a car find a ‘gap’ entering a moving stream of traffic.
Merging time (tm): A driver pulls up to their driveway, waits an average tm (3.5 seconds used) for one car to pass, and merges.
Critical Gap time (tg): if cars on the road are closer than tg, (6 seconds used) apart, you cannot merge Force Merging time (tf): A driver pulls up to their driveway, forced merge their way into main road due to bumper-to-bumper traffic.
This causes the car on the main road to slam on its brakes creating a shockwave. An average of tf (30 seconds used) for one driveway to successfully "force merge" into the saturated line, pass and merges.
R1 scenario – Because the utilization is low (37%), there are ample gaps between cars Across the 70 homes, the "stutter" caused by people pulling out is minimal. Not subject to tg impact The total "friction" added to the system is only about 4 minutes (3.5 seconds * 70 driveway) for the entire neighborhood to clear.
R3 scenario – the "spine" is saturated (Utilization 110%). This creates a "Wall of Cars." A massive concentration at R3 Evulich Court try to exit one single point onto Linda Vista which is already at 110% capacity.
There are zero natural gaps. The "forced merge" effect take place at Evulich Court.
The total forced merging delay time = tf x 51 units = 25.5 minutes Additional 3 minutes delay need to be added for the rest of cars from R1 homes è 28.5 minutes of merging delay is used Baseline (R1) Planned (R3) Analytical Assessment
Merging Type Distributed Concentrated 51 units create a massive "conflict point." Gap Probability High (80%) Low (<5%) High density eliminates natural gaps.
Avg. Wait per Merge 3.5 Seconds 30.0 Seconds "Forced merging" stops the main flow.
Total Merging Delay 4.0 Mins 28.5 Mins [Wait per Merge * No. of Units].
Mathematically Rigorous, Objective Results with Life -and-Death Consequences 3 Traffic Utilization Ratio Matters
— Baseline @ 37%: The system is "Under-Saturated." No queue; you drive straight to the stop sign and leave.
— Planned @ 110% : The system is "Over-Saturated." a
"Standing Queue" forms; the line of cars gets longer.
Delays to Safety
— Queuing Delay: Because the R3 Utilization is 110%, the queue never clears. You are now waiting for the cars
ahead of you to clear a "bottleneck" that is physically too small for the crowd. This is a mathematical proof that the road cannot physically clear the neighborhood in under 30 minutes.
— Merging Delay: The 51-unit project creates a Concentrated Conflict Point that adds 28.5 minutes of 'Shockwave Delay' to the evacuation. This is not a subjective preference; it is a mathematical consequence of forcing high-density
merging into a saturated, single-lane 'spine' during a surge event."
All conclusions derived from transparent assumptions, validated models, and reproducible calculations Baseline Planned Logic & Boundary
Conditions
Total Units (N) 70 Units 155 Units Includes Evulich &
Vista Heights.
Vehicle Load (V) 112 Cars 248 Cars 1.6 vehicles/HH
Capacity (μ) 600 Cars/Hr 450 Cars/Hr Road capacity.
Utilization (ρ) 37% 110% Demand / Capacity.
Queuing Delay
(DQ) 2.5 mins 45.0+ mins Exponential delay at
saturation.
Merging Delay
(DM) 4.0 mins 28.5 mins "Critical Gap
Acceptance” model
Human Factor
(DH) 20.0 mins 20.0 mins Time from alert to
vehicle in gear.
Total Evacuation 26.5 mins 93.5+ mins Sum of: [DQ+DM+DH].
Critical Safety Window: 30 minutes, based on a 20 -minute fire spread rate of 200 acres (Pacific Palisade Fire Data) The Asks
Consistency Review:A formal analysis by the Fire Marshal to determine if 51 units is consistent with the2025 Health and Safety Element Updateand its "Evacuation-Constrained" criteria. Supplemental CEQA Review:A re-evaluation of the project’s impacts on emergency evacuation plans, given the change in the
environmental hazard baseline as of March 2025.
Consideration of Mitigation:We propose that a lower-density, R1- aligned project of12 single-family homesserves as the only feasible mitigation to preserve the "Vista" in Linda Vista while maintaining a safe, 30-minute evacuation window.
4 Technical Rebuttal to Transportation Analysis by Hexagon Transportation Consultants Inc.
What the Hexagon Report Analyzes (Daily Operations)
— Trip Generation: 51 units, 375 daily vehicle trips, only 26 and 30 trips during the AM and PM peak hour correspondingly. — Intersection Delay: the "project-generated traffic" add < 1 second of delay to the Linda Vista/McClellan intersection. — Safety Assessment: limited to whether a garbage truck or fire truck can physically turn around on the new Evulich Court The Flaw in the Developer’s Study (Hexagon Memo)
— Metric Mismatch: Hexagon models 30 vehicles/hour (Daily Commute) — Disaster Reality: Wildfire evacuation requires modeling a 250+ vehicle surge in a 30-minute window. — Omission: The study ignores "Threshold 4" safety impacts during power outages such as the Linda Vista/McClellan intersection defaults to a 50% capacity 4-way stop
A Quantifiable Safety Violation
— The Hexagon report addresses "Daily Operations" but fails to provide a "Project-Specific Evacuation Plan". — The Hexagon report evaluates routine only traffic operations but not covers emergency surge evacuation performance. — We urge SummerHill Homes consulting with reputable agencies in wildfire evacuation analysis and protection planning.
Learn more on "Wildfire Evacuation Modeling and Traffic Analysis" — Youtube: FSEG evacuation simulation animations, National Fire Protection Association and PTV Group. — Research & Academic Presentations: "Simulating wildfires before they happen" (Google Research), How technology is transforming wildfire response in the WUI (Lund University/GHD) 5 Why standard "Peak Hour" metrics fail to protect the Linda Vista High Fire Hazard Corridor Subject: Re: Feb/24/2026 Planning Commission Public Hearing – Technical Rebuttal to “Transportation Analysis for the Proposed Townhomes at 10857 Linda Vista Drive in Cupertino, CA” by Hexagon Transportation Consultants, Inc. https://cupertino.legistar.com/gateway.aspx?M=F&ID=8a56f691-e5d1-472e-9ba9-753651a0bd2e.pdf To: Cupertino City Planning Commissioners
CC: Planning Staff, City of Cupertino
From: DerChang Kau, Cupertino
Dear Chair and Honorable Commissioners,
The "Safety Assessment" on page 22 of the Hexagon report concludes that the project "would not worsen existing geometric hazards". We contend that this finding is based on incomplete data and fails to address the unique life-safety risks of the Linda Vista "Spine."
1. Failure to Model Emergency Surge Conditions
• The Error: Hexagon’s analysis is based on "Peak Hour" trips, estimating only 30 vehicles will exit the site during the busiest hour of a normal day. • The Reality: In a wildfire evacuation, the demand is a simultaneous surge. Our integrated model shows that 248 vehicles must exit the spine in a 30-minute window. • The Impact: Hexagon ignores the Utilization Ratio of 110%, where demand exceeds road capacity, causing total system failure.
2. Ignoring Intersection Degradation (McClellan Choke Point) • The Error: The report assumes the Linda Vista/McClellan intersection functions at a standard "Level of Service".
• The Reality: Disaster events (fire/earthquake) often lead to power outages, turning the signal into a 4-way stop.
• The Impact: This condition drops flow rates by 50%, doubling the physical exit time and trapping residents in the "tunnel of fire".
3. Omission of Merging Conflict and Shockwave Delay
• The Error: Hexagon evaluates the new cul-de-sac for basic garbage truck turning templates but ignores "merging friction".
• The Reality: Concentrating 51 units into one entry point on a saturated single-lane road creates "forced merging".
• The Impact: As shown in our analytical model, this creates a 28-minute "Shockwave Delay" that stalls the entire evacuation route for all 155 households.
Conclusion: A Quantifiable Safety Violation
The Hexagon report addresses "Daily Operations" but fails to provide a "Project-Specific Evacuation Plan". Because the project pushes the total "Order to Safety" time from less than 30 minutes (current R1) to more than 90 minutes (planned R3), it creates a "specific, adverse impact upon public safety" that cannot be mitigated by standard street design.
The General Plan Public Safety Element requires the City to reduce wildfire risk and ensure adequate emergency access and evacuation capability for existing and future residents. The Mobility Element likewise requires that the circulation system function safely and reliably, not only under daily peak conditions but for emergency response. The Hexagon report evaluates routine traffic operations; it does not evaluate emergency surge evacuation performance on the single-access Linda Vista spine.
Because the Hexagon report only focuses on "daily traffic", I strongly recommend city planning department urge SummerHill Homes consulting with reputable agencies in wildfire evacuation analysis and protection planning, such as Reax Engineering for evacuation modeling and Dudek for CEQA fire protection planning to produce flawless and adequate wildfire evacuation risk analysis and mitigation plan.
Thank you.
Sincerely,
DerChang Kau
Cupertino, CA 95014
From:
To: City of Cupertino Planning Dept.; City of Cupertino Planning Commission; City Council; City Clerk Cc:
Subject: The Cupertino Wildfire Evacuation and Public Safety Critique Date: Tuesday, March 10, 2026 8:52:26 PM
Attachments: The Cupertino Wildfire Evacuation and Public Safety Critique.pdf Feedback to Cupertino Health and Safety Element - General Plan Community Vision 2015-2040 - v3.pdf Importance: High
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
March 10, 2026
TO: City of Cupertino Planning Staff and Planning Commission CC: Mayor and City Council Members
FROM: DerChang Kau, Resident of Cupertino
RE: Feedback to the Health and Safety Element Update (General Plan: 2015–2040 Community Vision)
Dear Planning Staff and Commissioners,
As a retired engineer, university professor, and long-time resident of Cupertino, I am writing to express my grave concerns regarding the current wildfire evacuation analysis integrated into the City’s Health and Safety Element. While the existing analysis provides a useful "planning level" starting point, it is fundamentally insufficient for a community that is rapidly densifying and facing modern wildfire risks.
My review of the provided documents reveals four critical gaps that must be addressed before the General Plan can be considered adequate to protect residents through 2040:
Reliance on Outdated Data: The analysis utilizes 2023–2025 population data that does not account for the major density increases triggered by the July 2024 rezoning. For example, the evacuation demand for Evulich Court is modeled on its previous 4-unit density rather than the R3 up-zoned density
. Consequently, the city is planning for 2040 safety using 2023 conditions.
Inadequate "Planning-Level" Methodology: The City’s own consultants acknowledge that constrained areas require Dynamic Traffic Assignment (DTA) to accurately simulate traffic flow. While peer jurisdictions like Marin County have already adopted these advanced tools, Cupertino continues to rely on broad evaluations that do not reflect real-world emergency behavior.
Neglect of High-Risk Fire Zones: Critical evacuation zones (CUP-016, 017, 029, and 032) lie within Moderate to Very High Fire Hazard Severity Zones (FHSZ), yet they lack street-level modeling. Notably, CUP-029 generates 38% of all evacuation trips, yet its primary outlets— McClellan Road and Foothill Boulevard—are projected to reach V/C ratios of 4.2 and 4.0 respectively. These are not mere "constraints"; they represent total system failure.
Misleading Safety Metrics: The use of Volume-to-Capacity (V/C) ratios dramatically understates the danger. A V/C ratio of 1.0 indicates an over-saturated system where standing queues form and the network fails. Framing a V/C > 2 as merely "more than one hour" of delay obscures the reality that residents may be unable to reach safety in time.
Proposed Action:
I urge the City to move beyond high-level assumptions and commit to a "Total Evacuation to Safety" framework. This must include street-level dynamic modeling, realistic capacity adjustments for factors like signal outages and poor visibility, and evacuation time estimates that reflect the density of the 2040 General Plan.
The detailed critique and recommendation is attached to this correspondence for your review and consideration.
Cupertino is a center of global innovation; we should be a leader, not a laggard, in data-driven public safety. Our residents deserve an evacuation plan that remains functional during the exact emergency it was designed to manage.
Respectfully,
DerChang Kau, Resident of Cupertino
FEEDBACK TO THE HEALTH AND SAFETY ELEMENT
OF THE CUPERTINO GENERAL PLAN: 2015 -
2040 COMMUNITY VISION
March 10, 2026
DerChang Kau
Retired Engineer, University Professor and Resident of Cupertino High-Risk Zones Demand Street-Level Evacuation-to-Safety Analysis Outdated census data and
planning-level methods
cannot manage modern wildfire risk
1 Preface
Cupertino’s current wildfire evacuation analysis relies heavily on planning-level assumptions, dated population data, and static traffic metrics that do not reflect real-world emergency conditions. Across all four sections below, a consistent theme emerges: the City’s Health & Safety Element is not aligned with the evacuation realities of a denser, more wildfire-exposed Cupertino in 2030–2040.
The documents show that:
● The analysis uses 2023–2025 data, ignoring the major density increases triggered by the July 2024 rezoning. As one section notes, it “primarily reflects existing households” rather than future conditions.
● The methodology remains planning-level, even though the City’s own consultant acknowledges that dynamic traffic assignment is required for constrained areas. ● Critical evacuation zones—especially CUP029, which generates 38% of all evacuation trips—are located within Fire Hazard Severity Zones, yet the City has not modeled evacuation at the street level.
● The use of V/C ratios dramatically understates evacuation risk. At V/C ≥ 1, “the system is oversaturated and no longer functioning,” yet the report treats V/C > 2 as merely “more than one hour.”
Taken together, these findings show that Cupertino’s current evacuation analysis cannot support long- range planning through 2040. The City must adopt dynamic, data-driven, future-oriented evacuation modeling to protect residents and meet the intent of the General Plan’s Health & Safety Element.
Section 1 — Dated Data Source
“Outdated Inputs: Why the Current Evacuation Analysis Cannot Support 2040 Planning” The following data sources were used to estimate vehicle evacuation demand for residents, employees, students, and visitors in “Evacuation Route Capacity Report” ● 2023 American Community Survey (ACS) 5-year Estimates: Used for population, household data, and vehicle ownership distribution.
● Genasys Protect: Provides population, household estimates, and designated evacuation zones. ● StreetLight Data (2024): Used to evaluate and compare traffic volumes between school and non-school periods.
This analysis represents a snapshot of the city’s current evacuation capacity based on the most recent available census and municipal data from 2023–2025. It does not incorporate the increased residential densities anticipated under the July 2024 rezoning. As noted in the original text, “using 2023 ACS data and July 2025 Genasys data, it primarily reflects existing households.”
For example, Evulich Ct. is evaluated using its existing 4-unit R1 zoning, rather than the 51-unit R3 density under the rezoning. Consequently, Figure HS-3 in Chapter 7 of the Health and Safety Element | General Plan (Community Vision 2015–2040) does not identify this area as required under Policy HS-
2.13, which highlights residential parcels with fewer than two ingress/egress routes.
2 Because the analysis relies on pre-rezoning conditions, the document does not adequately support long- range planning through the year 2040. The document is therefore inadequate to support a general plan to the year 2040.
Section 2 — Traffic Flow Methodology
“Planning-Level Methods Are Not Enough: Cupertino Needs Advanced Evacuation Modeling” The referenced sources describe the current “planning-level” assessment as a broad evaluation of transportation system capacity rather than a definitive evacuation plan. While this type of assessment is useful for identifying major bottlenecks and informing policy decisions, it also explicitly acknowledges that more detailed, higher-resolution studies—such as dynamic traffic assignment—are the appropriate next step for areas with significant constraints.
As Fehr & Peers note, community-scale wildfire evacuation assessment is an emerging field. However, general guidance does exist, including ISO 22315 and NIST’s ESCAPE framework. Given Cupertino’s substantial wildfire exposure and its position at the center of global science and technology innovation, the city is uniquely positioned to take a leadership role in advancing rigorous, data-driven evacuation assessment and mitigation planning. Several jurisdictions, including Marin County, have already begun such modeling efforts to identify evacuation bottlenecks and evaluate mitigation strategies ‡.
Leadership in this area could include:
● Scenario-based wildfire evacuation simulations
● Traffic network evacuation time analysis
● Population departure behavior modeling
● Redundant evacuation route and shelter analysis
● Quantitative evacuation performance metrics (clearance time, route saturation, etc.) A forward-looking evacuation assessment framework would not only strengthen public safety for current residents but also provide a valuable foundation for evaluating the cumulative evacuation impacts of future development.
‡:https://www.marinwildfire.org/news/the -state-of-the-science-in-wildfire-evacuation-modeling-a-literature-review?utm source=chatgpt.com Section 3 — Prioritize Local Level Evacuation Route with FHSZ Map “High-Risk Zones, High Stakes: Street-Level Modeling Required for FHSZ Evacuation” This analysis is based on the evacuation routes released by the City of Cupertino in August 2025. Every one of those designated routes lies outside the mapped Fire Hazard Severity Zones (FHSZ). Yet four zones—CUP-016, CUP-017, CUP-029, and CUP-032—are fully or partially within Moderate, High, or Very High FHSZ. Given this reality, it is no longer acceptable to rely on a “planning-level” review of the main evacuation route shown in Figure 2. The City must move beyond high-level assumptions and conduct street-level dynamic traffic assignment within these at-risk zones. Anything less leaves thousands of residents exposed.
3 The findings in CUP-029 alone make the case unmistakable. CUP-029 represents 38% of all evacuation trips in Scenarios A and C—the single largest share of evacuation demand. This is not a marginal zone; it is the primary choke point in a wildfire emergency. In Scenario A, McClellan Road—serving CUP-029 and CUP-032—reaches a V/C ratio of 4.2, while Foothill Boulevard—serving CUP-017—reaches 4.0. A V/C ratio of 1.0 already means a road is over capacity. These corridors are projected to operate at four times their hourly capacity during an evacuation. That is not a “constraint.” That is a system failure waiting to happen.
Fehr & Peers themselves state that for “more constrained and congested conditions, higher-effort studies may involve modeled analysis of evacuation scenarios using dynamic traffic assignment to simulate traffic flow and evacuation times when street and freeway networks are at capacity.” Cupertino’s conditions are exactly what this warning describes. The City cannot continue to treat this as optional.
To protect residents, Cupertino must adopt dynamic, data-driven tools immediately. This includes the use of real-time traffic management systems capable of providing dynamic route guidance and monitoring to divert evacuees away from collapsing corridors. Furthermore, phased evacuation strategies must be developed not only for the main arterials but also for the high-demand FHSZ zones themselves, where evacuation surges originate. Without this, the extreme congestion projected on McClellan Road and Foothill Boulevard will materialize as modeled or worse, as real-world behavior is never as orderly as a static model assumes.
Cupertino has the data, the expertise, and the responsibility. The City must elevate this work beyond planning-level analysis and commit to the advanced modeling that wildfire-exposed communities across California are already pursuing. Residents deserve evacuation planning that matches the severity of the risk—not a framework that collapses the moment it is needed most.
Section 4 — Improvement to Evacuation-to-Safety Metric
“Beyond V/C Ratios: A Total Evacuation-to-Safety Framework Is Essential” The current assessment fails to correct roadway capacity for emergency conditions. The Volume-to- Capacity (V/C) ratio used throughout the report is an over2023simplified metric, and the accompanying text is misleading. For example, on page 24 the report states:
“V/C ratios greater than 2… indicate that these roads would require more than one hour to accommodate the evacuation demand from the evacuation area.”
This framing dramatically understates the severity of the problem. A V/C ratio above 1.0 does not simply mean “more than one hour.” It means the system is over-saturated and no longer functioning. Cars are arriving faster than the roadway can process them. A standing queue forms. The system fails.
1. Capacity Must Reflect Realistic “Adjusted Evacuation Route Capacity” The “Adjusted Evacuation Route Capacity” used in this study relies on planning-level hourly capacities from the Highway Capacity Manual (2000). These values do not represent real-world evacuation conditions and fail to incorporate:
● Spatial Adjustments: Increased residential density from the Housing Element requires downward adjustments to reflect degraded capacity on constrained terrain and limited visibility. 4 ● Temporal Adjustments: School pickup periods and short-term curbside parking can severely reduce effective capacity during critical hours.
● Operational Limitations: The analysis assumes normal signal operations and does not account for “red-flash mode” during power outages—an extremely likely condition during wildfire events.
2. Demand Must Reflect Realistic Future Conditions
The study is a snapshot of Cupertino’s evacuation capacity based on 2023–2025 data. It does not account for the substantial density increases triggered by the July 2024 rezoning.
Planning for wildfire evacuation using backward-looking data is planning through a rear- view mirror. It guarantees underestimation of future evacuation demand .
3. Moving Beyond V/C: “Total Evacuation to Safety” Must Be the Standard Fehr & Peers acknowledge that V/C ratios only identify bottlenecks; they do not measure total evacuation time. In traffic engineering and management science¶, V/C is a well-established tool for quantifying queuing and merging delays. When V/C ≥ 1, the system is not merely “congested”—it is over-saturated, and two critical delays dominate:
¶: Stanford University,
Department of Civil and Environmental Engineering
Department of Management Science and Engineering
● Queuing Delay: As V/C approaches 1, delay increases non-linearly.
𝐷𝑒𝑙𝑎𝑦 = 𝑉/𝐶
1 − 𝑉/𝐶 × 𝑡𝑟𝑎𝑣𝑒𝑙 𝑡𝑖𝑚𝑒
When V/C ≥ 1, queues grow indefinitely. The roadway cannot clear vehicles. ● Merging Delay: When traffic density eliminates safe gaps, merging becomes impossible without forced maneuvers that further destabilize flow.
Consider CUP-029 at 8PM during a late-summer, wind-driven wildfire approaching from the south. A quickly encroaching fire from the foothills would paint the sky red. The tree canopy along Linda Vista becomes a “fuel tunnel.” Residents have only one exit: northbound toward McClellan Road. Roughly 80 vehicles from 51 homes on Evulich Court must merge into an already over-saturated Linda Vista Drive.
This is not a theoretical model—it is a life-and-death situation at the choke point.
Wildfires such as the Pacific Palisades Fire have spread at rates of 200 acres in 20 minutes. Under these conditions, seconds matter. Residents need to be aware not just of potential evacuation bottlenecks, but how much time is required to reach safety. This requires the city to both plan and communicate a comprehensive and up-to-date evacuation plan that considers realistic models of queuing, merging, human behavior, and departure delays.
Human factors—delays in loading children, pets, and essential belongings—further slow departures.
Without quantifying these delays, the City cannot provide residents with the knowledge they need to prepare for an evacuation that may unfold in minutes, not hours.
5 Fehr & Peers states that the V/C metric is only a baseline tool. They identify dynamic traffic assignment (DTA) and phased departure timing as the necessary next steps for an objective, safety-driven evacuation assessment.
In conclusion, Cupertino cannot rely on planning-level V/C ratios to evaluate wildfire evacuation safety.
The City must adopt a Total Evacuation to Safety framework grounded in dynamic modeling, realistic capacity adjustments, and future population density. Anything less leaves residents with a false sense of security and exposes the community to unacceptable risk.
Closing Summary
“Why Cupertino Must Elevate Its Evacuation Planning Now” Cupertino’s current wildfire evacuation analysis provides a useful starting point, but it is fundamentally constrained by outdated data, planning-level methodologies, and metrics that do not reflect real- world emergency behavior. Across all four sections, the evidence is clear: the City’s Health & Safety Element does not yet meet the level of rigor required to protect residents in a rapidly densifying, wildfire-exposed community.
The analysis relies on 2023–2025 population and traffic data, even though the July 2024 rezoning dramatically increases residential density. As the document itself notes, it “primarily reflects existing households,” leaving future evacuation demand unaccounted for. This gap alone makes the current assessment insufficient for long-range planning through 2040.
Methodologically, the City continues to rely on broad, planning-level evaluations despite its consultant’s own acknowledgment that dynamic traffic assignment and scenario-based evacuation modeling are necessary for constrained areas. Peer jurisdictions across California have already adopted these advanced tools; Cupertino should not lag behind.
The risk is most acute in the Fire Hazard Severity Zones, from which evacuation surges originate. Zones such as CUP029 and CUP032 generate the largest share of evacuation demand, yet are absent from the City’s street-level evacuation models behavior. Projected V/C ratios of 4.0–4.2 on McClellan Road and Foothill Boulevard indicate not congestion, but system failure under evacuation conditions.
Finally, the continued reliance on V/C ratios obscures the true danger. At V/C ≥ 1, “the system is oversaturated and no longer functioning,” however the City’s report frames V/C > 2 as merely “more than one hour.” Residents need to understand time to safety, not just where bottlenecks occur. That requires modeling queuing, merging, human behavior, and departure delays—none of which are captured in the current framework.
In summary, Cupertino has the data, the expertise, and the responsibility to adopt a modern, dynamic, and future-oriented evacuation planning framework. The City needs urgently to meet the intent of the General Plan and to safeguard residents through 2040 and beyond. Cupertino must move beyond planning-level assumptions and commit to advanced, data-driven evacuation modeling that reflects the realities of wildfire risk and future population growth.
Outdated Inputs
The City’s evacuation analysis is built on pre-rezoning data and cannot support long-range planning.
Uses 2023 ACS, 2025 Genasys, and 2024 StreetLight data — “Primarily reflects existing households” — not future density Fails to identify parcels with single ingress/egress as required by California Government Code Section 65302(g)(5)¶
— Evulich Ct. modeled as 4 units, not 51 units allowed under rezoning
A 2040 General Plan cannot rely on 2023 conditions
The analysis is outdated the moment it is published.
2 https://leginfo.legislature.ca.gov/faces/codes displaySection.xhtml?lawCode=GOV§ionNum=65302.&article=5.&hig hlight=true&keyword=Fire%20Hazard%20Severity%20Zone#:~:text=(5)%20Upon%20the%20next%20revision,revision %20of%20the%20safety%20element.
¶
Planning-Level Methods Are Not Enough
Cupertino must move beyond broad, planning-level assessments to dynamic, high-resolution evacuation modeling.
Current method is a “broad evaluation,” not a plan
— Fehr & Peers: constrained areas require dynamic traffic assignment (DTA) DTA has been adopted widely
— National guidance exists: ISO 22315, NIST ESCAPE
— Peer jurisdictions (e.g., Marin County) already use advanced modeling Needed tools:
— Scenario-based simulations, Evacuation time analysis, Departure-behavior modeling, Redundant route & shelter analysis, Quantitative performance metrics
Cupertino should be a leader—not a laggard—in evacuation planning. 3 High-Risk Zones Require Street-Level Modeling
Four evacuation zones lie within Fire Hazard Severity Zones, yet the City has not modeled their street-level evacuation behavior. All designated evacuation routes lie outside FHSZ
But CUP016, CUP017, CUP029, CUP032 lie inside Moderate–Very High FHSZ — CUP029 alone = 38% of all evacuation trips
— McClellan Road V/C = 4.2; Foothill Blvd V/C = 4.0
— These are not “constraints”—they are system failures
City must adopt:
— Street-level DTA, Real-time traffic management, Phased evacuation strategies for FHSZ zone
Thousands of residents remain exposed until
Cupertino models evacuation where the risk actually exists. 4 Beyond V/C Ratios: A Total Evacuation-to-Safety Framework V/C ratios hide the true danger.
Cupertino needs a Total Evacuation-to-Safety model.
V/C > 1 means oversaturation — “the system fails”
Current study uses planning-level capacities from year 2000 Missing adjustments:
— Spatial (density, terrain, visibility), Temporal (school pickup, curbside parking), Operational (signal outages, red-flash mode)
Real evacuation requires modeling:
— Queuing delays, Merging delays, Human behavior, Departure lag Residents need to know time to safety, not just where bottlenecks occur. 5 ©2026, DerChang Kau. All rights reserved. 7
15’ after Red Flag Warning issued5’ after Red Flag Warning issued
©2026, DerChang Kau. All rights reserved. 8
45’ after Red Flag Warning issued30’ after Red Flag Warning issued
Summary: Cupertino Must Elevate Its Evacuation Planning The current evacuation analysis is not sufficient for
a wildfire-exposed, rapidly densifying Cupertino.
The City must adopt modern, data-driven methods to meet 2040 safety needs. Dated data source
Over simplified methods despite known constraints and consultant guidance High-risk zones (CUP016, 17, 29, 32) require street-level modeling V/C ratios mask true danger — oversaturation means system failure, not delay Residents need time-to-safety modeling, not bottleneck identification Peer jurisdictions already use dynamic traffic assignment & scenario-based tools To protect residents and meet the intent of the Health & Safety Element, Cupertino must move beyond planning-level assumptions and commit to advanced, future-oriented evacuation modeling now. 10 March 10, 2026
TO: City of Cupertino Planning Staff and Planning Commission CC: Mayor and City Council Members
FROM: DerChang Kau, Resident of Cupertino
RE: Feedback to the Health and Safety Element Update (General Plan: 2015–2040 Community Vision)
Dear Planning Staff and Commissioners,
As a retired engineer, university professor, and long-time resident of Cupertino, I am writing to express my grave concerns regarding the current wildfire evacuation analysis integrated into the City’s Health and Safety Element. While the existing analysis provides a useful "planning level" starting point, it is fundamentally insufficient for a community that is rapidly densifying and facing modern wildfire risks.
My review of the provided documents reveals four critical gaps that must be addressed before the General Plan can be considered adequate to protect residents through 2040:
• Reliance on Outdated Data: The analysis utilizes 2023–2025 population data that does not account for the major density increases triggered by the July 2024 rezoning. For example, the evacuation demand for Evulich Court is modeled on its previous 4-unit density rather than the R3 up-zoned density. Consequently, the city is planning for 2040 safety using 2023 conditions. • Inadequate "Planning-Level" Methodology: The City’s own consultants acknowledge that constrained areas require Dynamic Traffic Assignment (DTA) to accurately simulate traffic flow.
While peer jurisdictions like Marin County have already adopted these advanced tools, Cupertino continues to rely on broad evaluations that do not reflect real-world emergency behavior.
• Neglect of High-Risk Fire Zones: Critical evacuation zones (CUP-016, 017, 029, and 032) lie within Moderate to Very High Fire Hazard Severity Zones (FHSZ), yet they lack street-level modeling. Notably, CUP-029 generates 38% of all evacuation trips, yet its primary outlets— McClellan Road and Foothill Boulevard—are projected to reach V/C ratios of 4.2 and 4.0 respectively. These are not mere "constraints"; they represent total system failure. • Misleading Safety Metrics: The use of Volume-to-Capacity (V/C) ratios dramatically understates the danger. A V/C ratio of 1.0 indicates an over-saturated system where standing queues form and the network fails. Framing a V/C > 2 as merely "more than one hour" of delay obscures the reality that residents may be unable to reach safety in time.
Proposed Action: I urge the City to move beyond high-level assumptions and commit to a "Total Evacuation to Safety" framework. This must include street-level dynamic modeling, realistic capacity adjustments for factors like signal outages and poor visibility, and evacuation time estimates that reflect the density of the 2040 General Plan.
The detailed critique and recommendation is attached to this correspondence for your review and consideration.
Cupertino is a center of global innovation; we should be a leader, not a laggard, in data-driven public safety. Our residents deserve an evacuation plan that remains functional during the exact emergency it was designed to manage.
Respectfully,
DerChang Kau, Resident of Cupertino
From:
To: Luke Connolly; Emi Sugiyama
Subject: FW: Can you send me the letter from County Fire that was used to exempt the Linda Vista Project from CEQA? Date: Wednesday, March 4, 2026 7:34:17 AM
Importance: High
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
Hi Luke,
Per our conversation last night, please email all fire department communications pertaining to the exemption.
Thanks,
Rhoda
From: Rhoda Fry
Sent: Monday, March 2, 2026 8:45 AM
To: '[email protected]' <[email protected]>
Subject: Can you send me the letter from County Fire that was used to exempt the Linda Vista Project from CEQA?
Hi Emi,
Can you send me the letter from County Fire that was used to exempt the Linda Vista Project from CEQA?
letter:
Santa Clara County Fire Department. Fire Prevention Plan Review Comments. January 27, 2026.
I would like to see that letter.
Thanks for your help,
Rhoda Fry