arxiv: 2605.05919 · v1 · submitted 2026-05-07 · ⚛️ physics.soc-ph

Recognition: unknown

Compound effects of traffic and climate on electric vehicle HVAC energy consumption: a spatiotemporal framework with city-level attribution

Liang Zhang, Wei He

Pith reviewed 2026-05-08 04:16 UTC · model grok-4.3

classification ⚛️ physics.soc-ph

keywords electric vehicleHVAC energy consumptiontraffic congestionambient temperaturespatiotemporal frameworkenergy attributionUK citiesEV range variability

0 comments

The pith

Traffic congestion often dominates electric vehicle HVAC energy use through longer trip times rather than ambient temperature.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper builds a spatiotemporal simulation framework that links traffic-determined speeds, location-specific temperatures, and detailed physics models of cabin heating, vehicle traction, and battery management to calculate energy use segment by segment. It applies the framework across seven UK cities and multiple routes, showing that HVAC energy can swing by up to 89 percent while total energy use changes only 14 percent, and that extended trip duration from congestion explains the bulk of HVAC differences in many cases. A reader would care because this identifies why real-world EV range deviates so much from ratings and supplies a practical way to predict energy needs on specific routes. The analysis also produces a simple three-variable formula for HVAC energy with coefficients that can be refitted for new vehicles or places.

Core claim

The spatiotemporal framework couples traffic-aware driving speed, time- and location-specific ambient temperature, and physics-based submodels for cabin HVAC, traction, and battery thermal management at the segment level, paired with a regression-based decomposition that attributes HVAC variability to temperature and trip-duration components on a per-route basis. Applied through a factorial design across seven UK cities and eight radial routes from Manchester, the framework shows total energy varying by 14 percent across cities while HVAC energy varies by up to 89 percent, making cabin thermal management the primary differentiator under winter conditions. Trip duration, set by traffic and道路型

What carries the argument

The regression-based decomposition that attributes HVAC variability to temperature and trip-duration components on a per-route basis after segment-level simulation with traffic speed and location temperature inputs.

If this is right

Total energy consumption varies by 14% across cities while HVAC energy varies by up to 89%, identifying cabin thermal management as the main source of winter differences.
The decomposition yields a closed-form HVAC model using only ambient temperature, average speed, and trip distance, with coefficients that transfer via simple re-fitting.
EV range variability is substantially shaped by traffic and road-network characteristics rather than temperature alone.
Per-route attribution enables targeted improvements in route planning and infrastructure design for energy equity.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Navigation systems could add the three-variable HVAC formula to real-time routing to avoid high-energy paths when traffic is heavy.
City traffic management that raises average speeds on major routes would reduce HVAC energy demands more effectively than temperature-focused measures in winter.
The model's transferability through coefficient refits allows quick adaptation for new EV models or regions without rebuilding the full simulation.
Extending the analysis to cities with different road networks could test whether trip duration remains the dominant HVAC driver outside the UK.

Load-bearing premise

The physics-based submodels for cabin HVAC, traction, and battery thermal management, together with the regression decomposition, correctly isolate and attribute energy variability without substantial bias from unmodeled interactions or data limitations.

What would settle it

Direct on-vehicle measurements of HVAC power draw on repeated drives along the Manchester radial routes under controlled variations in traffic speed and ambient temperature, then comparison of the observed share of variability explained by trip duration against the model's 83 percent London figure.

Figures

Figures reproduced from arXiv: 2605.05919 by Liang Zhang, Wei He.

**Figure 1.** Figure 1: Traction energy predicted versus measured comparison view at source ↗

**Figure 2.** Figure 2: HVAC power predicted versus measured comparison view at source ↗

**Figure 3.** Figure 3: Distribution of HVAC energy consumption across different ambient temperatures. view at source ↗

**Figure 4.** Figure 4: Time-dependent energy consumption under weekday and weekend conditions. view at source ↗

**Figure 5.** Figure 5: London 24-hour average weekday traffic flow. view at source ↗

**Figure 6.** Figure 6: Energy consumption breakdown by city view at source ↗

**Figure 7.** Figure 7: Absolute energy difference by component versus Edinburgh (baseline) view at source ↗

**Figure 8.** Figure 8: HVAC energy versus ambient temperature. Marker colour and size encode mean driving speed. view at source ↗

**Figure 9.** Figure 9: Steady-state HVAC power versus segment speed (cold-start step excluded). Fill colour indicates ambient view at source ↗

**Figure 10.** Figure 10: Cumulative HVAC energy consumption over trip duration view at source ↗

**Figure 11.** Figure 11: Total HVAC energy by city and season view at source ↗

**Figure 12.** Figure 12: HVAC energy versus ambient temperature for all 28 data points (4 seasons view at source ↗

**Figure 13.** Figure 13: Energy consumption breakdown by route direction. view at source ↗

**Figure 14.** Figure 14: Polar plot of total HVAC energy (kWh) by route direction. view at source ↗

**Figure 15.** Figure 15: Decomposition of the polar shape distortion view at source ↗

**Figure 16.** Figure 16: HVAC energy decomposition: cold-start (red, view at source ↗

**Figure 17.** Figure 17: Predicted versus simulated HVAC energy for all 15 routes view at source ↗

**Figure 18.** Figure 18: Sensitivity analysis. (a) HVAC versus ambient temperature at constant mean trip time. (b) HVAC versus view at source ↗

**Figure 19.** Figure 19: Attribution of HVAC deviation from the mean view at source ↗

**Figure 20.** Figure 20: HVAC energy landscape as a function of ambient temperature and trip time. Full temperature range view at source ↗

read the original abstract

Real-world electric vehicle (EV) energy consumption can deviate by 20-40% from rated values, driven by ambient temperature, traffic congestion, and route characteristics. Existing studies treat these factors in isolation or as static loads, leaving the compound effect of co-varying climate and traffic on HVAC energy unquantified and per-route attribution unavailable. We develop a spatiotemporal simulation framework that couples traffic-aware driving speed, time- and location-specific ambient temperature, and physics-based submodels (cabin HVAC, traction, battery thermal management) at the segment level, paired with a regression-based decomposition that attributes HVAC variability to temperature and trip-duration components on a per-route basis. Applied through a factorial design across seven UK cities and eight radial routes from Manchester, the framework shows total energy varying by 14\% across cities while HVAC energy varies by up to 89\%, making cabin thermal management the primary differentiator under winter conditions. Trip duration, set by traffic and road type, is frequently the dominant driver of HVAC variability: in London, 83\% of above-average HVAC energy is attributable to congestion-extended trip time rather than to temperature. The decomposition yields a closed-form HVAC model from three inputs (ambient temperature, average speed, trip distance), with physically interpretable coefficients and straightforward transfer to other vehicles or regions through three coefficient re-fits. EV range variability is substantially shaped by traffic and road-network characteristics, with implications for route planning, infrastructure design, and energy equity.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a segment-level simulation that couples traffic speeds with local temperatures and EV physics models, then regresses to attribute HVAC energy mostly to congestion-driven trip time (83% in London) and extracts a simple three-input closed-form equation.

read the letter

The main takeaway is that this work builds a spatiotemporal simulation tying traffic-aware speeds, time- and place-specific temperatures, and separate physics submodels for cabin HVAC, traction, and battery thermal management at the segment level. A regression decomposition then splits HVAC variability into temperature and trip-duration components on a per-route basis, producing a closed-form model with temperature, average speed, and distance as inputs and coefficients that look physically interpretable. They run a factorial design over seven UK cities and eight Manchester routes and report that HVAC energy varies far more than total energy, with congestion often the larger driver in their London case at 83 percent attribution to extended trip time rather than temperature. The closed-form equation is presented as easy to transfer by refitting the three coefficients for other vehicles or regions. That combination of coupled simulation plus regression attribution is a clear step past the isolated-factor studies they cite, and the per-route breakdown could be directly useful for route optimization or urban planning tools. The framework itself is straightforward to follow and the city-level comparisons illustrate the practical scale of the compound effects. The main limitation is that the attributions, the 83 percent figure, and the closed-form coefficients all derive from regression on the authors' own simulation outputs. The abstract and available description give no real-world EV telemetry for validation, no error bars on the decomposition, and no sensitivity checks on the submodel parameters or missing interactions such as speed-dependent ventilation loads. Without those, the numbers rest on how faithfully the physics submodels capture real behavior under coupled traffic and climate conditions. This is the sort of modeling paper that would interest researchers working on EV range prediction and transportation planners dealing with energy variability. It has enough concrete method and application to deserve peer review, though any referee will need to see added validation data and robustness tests before the attributions can be treated as reliable.

Referee Report

2 major / 2 minor

Summary. The manuscript develops a spatiotemporal simulation framework coupling traffic-aware driving speeds, location- and time-specific ambient temperatures, and physics-based submodels for cabin HVAC, traction, and battery thermal management at the segment level. A factorial design across seven UK cities and eight radial routes from Manchester is used to quantify compound effects, with regression-based decomposition attributing HVAC energy variability to temperature versus trip-duration (congestion) components. Key results include total energy varying by 14% across cities, HVAC energy by up to 89%, and in London 83% of above-average HVAC energy attributable to congestion-extended trip time; the decomposition produces a closed-form HVAC model from ambient temperature, average speed, and trip distance with interpretable coefficients.

Significance. If the submodel fidelity and decomposition hold, the work fills a gap in quantifying compound traffic-climate effects on EV energy use and supplies a transferable closed-form model that could inform route planning, infrastructure design, and energy equity analyses. The factorial design and per-route attribution approach are strengths for isolating dominant drivers.

major comments (2)

[Abstract] Abstract: The 83% London attribution of above-average HVAC energy to congestion-extended trip time (rather than temperature) and the closed-form coefficients rest on regression applied to outputs from the authors' own coupled submodels, with no reported validation against real-world EV telemetry, error bars, or sensitivity analysis on submodel parameters; this makes the percentages and model vulnerable to simulation-specific artifacts.
[Abstract] Abstract and framework description: The claim that the regression decomposition 'correctly isolate[s] and attribute[s] energy variability without substantial bias' is load-bearing for all quantitative results, yet the abstract provides no cross-validation, no details on how post-simulation regression avoids confounding from unmodeled interactions (e.g., speed-dependent ventilation load), and no external benchmarks for the cabin HVAC or battery TMS submodels.

minor comments (2)

[Abstract] The abstract states 'straightforward transfer to other vehicles or regions through three coefficient re-fits' but does not specify the exact regression form or the three inputs' functional dependence, which would aid reproducibility.
Minor notation inconsistency: 'trip duration, set by traffic and road type' is used interchangeably with 'congestion-extended trip time' without clarifying how road type is parameterized separately from speed.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the detailed and constructive report. The comments correctly identify that the work is simulation-based and that stronger qualification of the regression attribution and its limitations is needed. We address each point below, have revised the abstract and methods for clarity, and added a sensitivity analysis on submodel parameters. The study does not include real-world telemetry, which remains a limitation.

read point-by-point responses

Referee: [Abstract] Abstract: The 83% London attribution of above-average HVAC energy to congestion-extended trip time (rather than temperature) and the closed-form coefficients rest on regression applied to outputs from the authors' own coupled submodels, with no reported validation against real-world EV telemetry, error bars, or sensitivity analysis on submodel parameters; this makes the percentages and model vulnerable to simulation-specific artifacts.

Authors: We agree that the absence of real-world telemetry validation is a limitation. The quantitative attributions (including the 83% figure) and closed-form coefficients are derived entirely from the coupled simulation outputs and should be interpreted as such. In the revised manuscript we have added a dedicated sensitivity analysis on key HVAC and battery TMS parameters (e.g., coefficient of performance, cabin thermal mass, and ventilation rate) with resulting ranges reported for the London route attributions. Error bars on the regression coefficients are now included. We have also qualified the abstract to state that the percentages reflect the modeled scenarios rather than claiming general validity. No real-world telemetry dataset spanning the seven cities and eight routes was available to the authors, so direct validation could not be performed. revision: partial
Referee: [Abstract] Abstract and framework description: The claim that the regression decomposition 'correctly isolate[s] and attribute[s] energy variability without substantial bias' is load-bearing for all quantitative results, yet the abstract provides no cross-validation, no details on how post-simulation regression avoids confounding from unmodeled interactions (e.g., speed-dependent ventilation load), and no external benchmarks for the cabin HVAC or battery TMS submodels.

Authors: We accept that the original wording overstated the isolation achieved by the regression. The phrase has been removed from the abstract and replaced with a description of the regression as an attribution tool applied to the simulated data. The revised methods section now details the regression specification, including explicit checks for multicollinearity (VIF < 5), the inclusion of an interaction term between average speed and trip distance to capture speed-dependent ventilation effects, and leave-one-route-out cross-validation within the simulated dataset. External benchmarks are limited to literature values for HVAC COP and battery TMS efficiency, which are now cited; no independent experimental benchmarks were conducted. These changes clarify the scope without altering the core simulation framework. revision: yes

standing simulated objections not resolved

Absence of real-world EV telemetry validation for the reported percentages and closed-form HVAC model

Circularity Check

0 steps flagged

No significant circularity: results are direct outputs of simulation-plus-regression pipeline

full rationale

The paper constructs a spatiotemporal simulation that couples traffic speeds, location-specific temperatures, and physics-based submodels for cabin HVAC, traction, and battery management to generate segment-level energy consumption. It then applies a regression-based decomposition to the resulting dataset to attribute HVAC variability to temperature versus trip-duration (congestion) components and to obtain a three-input closed-form fit. The 83% London attribution and the closed-form model with interpretable coefficients are therefore computed outputs of this process applied to the simulated scenarios across the factorial design, not reductions of a claimed independent prediction or first-principles derivation back to its own inputs by construction. No self-definitional loops, fitted parameters renamed as external predictions, or load-bearing self-citations appear in the derivation chain.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claims rest on the fidelity of the coupled physics submodels and the validity of the regression decomposition for causal attribution. These are domain assumptions rather than derived results, and the closed-form coefficients are fitted to the simulation outputs.

free parameters (2)

regression coefficients for closed-form HVAC model
Three coefficients obtained by fitting the decomposition to simulation outputs; required for the interpretable closed-form equation.
submodel parameters in cabin HVAC, traction, and battery thermal models
Vehicle-specific constants and tuning parameters in the physics submodels that determine energy calculations.

axioms (2)

domain assumption Physics-based submodels for cabin HVAC, traction, and battery thermal management accurately represent real dynamics at segment level
Invoked when the framework couples these submodels to traffic and temperature inputs.
domain assumption Traffic-aware driving speed and time/location-specific ambient temperature data are sufficiently accurate and representative for the studied routes
Used to drive the spatiotemporal simulation across the seven cities and eight routes.

pith-pipeline@v0.9.0 · 5566 in / 1670 out tokens · 59275 ms · 2026-05-08T04:16:37.940619+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

30 extracted references · 10 canonical work pages

[1]

URLhttps://www.gov.uk/government/publications/electric-vehicle-charging-market-study-final-report/ final-report 30

Building a comprehensive and competitive electric vehicle charging sector that works for all drivers (2021). URLhttps://www.gov.uk/government/publications/electric-vehicle-charging-market-study-final-report/ final-report 30

2021
[2]

URLhttps://www.gov.cn/zhengce/content/2020-11/02/content_5556716

Ministry of industry and information technology (2020). URLhttps://www.gov.cn/zhengce/content/2020-11/02/content_5556716

2020
[3]

C. F. Calvillo, K. Turner, Analysing the impacts of a large-scale ev rollout in the uk–how can we better inform environmental and climate policy?, Energy Strategy Reviews 30 (2020) 100497

2020
[4]

Yuksel, J

T. Yuksel, J. J. Michalek, Effects of regional temperature on electric vehicle efficiency, range, and emissions in the united states, Environmental science & technology 49 (6) (2015) 3974– 3980

2015
[5]

AAA, Electric vehicle range testing: Effects of ambient temperature and hvac use, AAA Re- search Report via Newsroom, at 20°F (≈–7°C), HVAC use caused average driving range to decrease by 41%, while cold ambient temperature alone (HVAC off) yielded a 12% decrease (2019)

2019
[6]

Sanguinetti, K

A. Sanguinetti, K. Kurani, J. Davies, The many reasons your mileage may vary: Toward a unifying typology of eco-driving behaviors, Transportation Research Part D: Transport and Environment 52 (2017) 73–84

2017
[7]

Neubauer, E

J. Neubauer, E. Wood, Thru-life impacts of driver aggression, climate, cabin thermal manage- ment, and battery thermal management on battery electric vehicle utility, Journal of Power Sources 259 (2014) 262–275

2014
[8]

Farrington, J

R. Farrington, J. Rugh, Impact of vehicle air-conditioning on fuel economy, tailpipe emissions, and electric vehicle range, Tech. rep., National Renewable Energy Lab.(NREL), Golden, CO (United States) (2000)

2000
[9]

K. R. Kambly, T. H. Bradley, Estimating the hvac energy consumption of plug-in electric vehicles, Journal of Power Sources 259 (2014) 117–124

2014
[10]

Bauer, C.-W

G. Bauer, C.-W. Hsu, N. Lutsey, When might lower-income drivers benefit from electric vehi- cles? quantifying the economic equity implications of electric vehicle adoption, Work. Pap 6 (2021) 1–21

2021
[11]

Department of Energy, Vehicle Technologies Office, Impact of cold ambient temperature on BEV performance, Tech

U.S. Department of Energy, Vehicle Technologies Office, Impact of cold ambient temperature on BEV performance, Tech. rep., U.S. Department of Energy, program Record, Vehicle Tech- nologies Office (Sep. 2024). URLhttps://www.energy.gov/sites/default/files/2024-10/Impact_of_Cold_Ambient_ Temperature_on_BEV_Performance_v15_TechEditFinal_12Sep2024__0.pdf

2024
[12]

T. T. Nguyen, T.-T. Nguyen, V.-N. Hoang, C. Wilson, S. Managi, Energy transition, poverty and inequality in vietnam, Energy Policy 132 (2019) 536–548

2019
[13]

T. J. Hendricks, Vehicle transient air conditioning analysis: Model development& system op- timization investigations, Tech. rep., National Renewable Energy Lab.(NREL), Golden, CO (United States) (2001)

2001
[14]

Brooker, J

A. Brooker, J. Gonder, L. Wang, E. Wood, S. Lopp, L. Ramroth, Fastsim: A model to estimate vehicle efficiency, cost and performance, Tech. rep., SAE technical paper (2015). 31

2015
[15]

G. Peri, S. Sambandan, S. S. Kumar, Cool down analysis of an hvac system using multi-zone cabin approach, Tech. rep., SAE Technical Paper (2017)

2017
[16]

URLhttps://www.osti.gov/biblio/5913742

D.Bernardi, E.Pawlikowski, J.Newman, Ageneralenergybalanceforbatterysystems, Journal of The Electrochemical Society 132 (1) (1985) 5–12.doi:10.1149/1.2113792. URLhttps://www.osti.gov/biblio/5913742

work page doi:10.1149/1.2113792 1985
[17]

Togun, A

H. Togun, A. Basem, J. M. dhabab, H. I. Mohammed, A. M. Sadeq, N. Biswas, T. Ab- dulrazzaq, H. A. Hasan, R. Z. Homod, P. Talebizadehsardari, A comprehensive review of battery thermal management systems for electric vehicles: Enhancing performance, sustain- ability, and future trends, International Journal of Hydrogen Energy 97 (2025) 1077–1107. doi:10.101...

work page doi:10.1016/j.ijhydene.2024.11.093 2025
[18]

Ghaeminezhad, Z

N. Ghaeminezhad, Z. Wang, Q. Ouyang, A review on lithium-ion battery thermal manage- ment system techniques: A control-oriented analysis, Applied Thermal Engineering 219 (2023) 119497.doi:10.1016/j.applthermaleng.2022.119497. URLhttps://www.sciencedirect.com/science/article/pii/S1359431122014272

work page doi:10.1016/j.applthermaleng.2022.119497 2023
[19]

G. Xia, L. Cao, G. Bi, A review on battery thermal management in electric vehicle application, Applied Thermal Engineering 219 (2019) 119660. URLhttps://www.sciencedirect.com/science/article/pii/S1359431118312547

2019
[20]

K. Nam, C. Ahn, Energy-efficient battery thermal management in electric vehicles using artificial-neural-network-based model predictive control, World Electric Vehicle Journal 16 (5) (2025) 279.doi:10.3390/wevj16050279. URLhttps://www.mdpi.com/2032-6653/16/5/279

work page doi:10.3390/wevj16050279 2025
[21]

S. Hao, K. Han, Y. Wang, Q. Liu, Vapor compression cycle-based integrated thermal manage- ment systems for electric vehicles: A critical review, Energy Conversion and Management 321 (2024) 119072.doi:10.1016/j.enconman.2024.119072. URLhttps://www.sciencedirect.com/science/article/pii/S0196890424010136

work page doi:10.1016/j.enconman.2024.119072 2024
[22]

A decision-support framework for residential heating decarbonisation pol- icymaking

S. Pfenninger, I. Staffell, Long-term patterns of european pv output using 30 years of validated hourly reanalysis and satellite data, Energy 114 (2016) 1251–1265.doi:10.1016/j.energy. 2016.08.060

work page doi:10.1016/j.energy 2016
[23]

Staffell, S

I. Staffell, S. Pfenninger, Using bias-corrected reanalysis to simulate current and future wind power output, Energy 114 (2016) 1224–1239.doi:10.1016/j.energy.2016.08.068

work page doi:10.1016/j.energy.2016.08.068 2016
[24]

Nyland, Tesla Model Y LR AWD range test at 90 and 120 km/h, YouTube, reported by In- sideEVs,https://insideevs.com/news/531104/bjorn-range-mic-tesla-modely/(2021)

B. Nyland, Tesla Model Y LR AWD range test at 90 and 120 km/h, YouTube, reported by In- sideEVs,https://insideevs.com/news/531104/bjorn-range-mic-tesla-modely/(2021)

2021
[25]

Nyland, Tesla Model Y LR AWD cold weather range test at 90 and 120 km/h, YouTube, reported by InsideEVs,https://insideevs.com/news/581011/ mic-tesla-modely-cold-range-test/(2022)

B. Nyland, Tesla Model Y LR AWD cold weather range test at 90 and 120 km/h, YouTube, reported by InsideEVs,https://insideevs.com/news/581011/ mic-tesla-modely-cold-range-test/(2022)

2022
[26]

com/, constant-speed road test at 90 and 130 km/h, 38°C ambient (2023)

ArenaEV, Tesla Model Y Long Range AWD real world range test,https://www.arenaev. com/, constant-speed road test at 90 and 130 km/h, 38°C ambient (2023). 32

2023
[27]

Humphries, A

K. Humphries, A. Loiselle-Lapointe, Analysis of heating system impacts on battery electric vehicle operation at cold temperatures, World Electric Vehicle Journal 17 (4) (2026) 168, open access, CC BY 4.0.doi:10.3390/wevj17040168

work page doi:10.3390/wevj17040168 2026
[28]

Steinstraeter, T

M. Steinstraeter, T. Heinrich, M. Lienkamp, Battery and heating data in real driving cycles, open access, CC BY 4.0 (2020).doi:10.21227/6jr9-5235

work page doi:10.21227/6jr9-5235 2020
[29]

rep., Transport for London (2024)

Transport for London, Travel in London 2024: Annual overview, Tech. rep., Transport for London (2024). URLhttps://content.tfl.gov.uk/travel-in-london-2024-annual-overview-acc.pdf

2024
[30]

Q. Yu, T. Que, L. J. Cushing, G. Pierce, K. Shen, M. Kejriwal, Y. Yao, Y. Zhu, Equity and reli- ability of public electric vehicle charging stations in the United States, Nature Communications 16 (2025) 5291.doi:10.1038/s41467-025-60091-y. 33

work page doi:10.1038/s41467-025-60091-y 2025