How We Calculate Real Range and Cost

Every range, cost, and journey figure on HonestRange is calculated using a single physics model based on each car's measured parameters, not WLTP values or flat percentages. This page details the equations, the constants and their sources, current validation status, and the individual responsible for the model.

Validation status

Simulink grid

3.33%avg error

9-point grid compared to ANL-validated Simscape model. 8 out of 9 points within 5%.

ANL dyno: Model 3

3.1%avg error

7 laboratory data points, −10 to +27 °C at 65 mph steady cruise.

Operating points

20data points

Across 3 vehicles, using 2 validation methods and real-world temperatures.

Simulink grid: 9 operating points

The power model is validated against a 9-point grid (30, 50, 70 mph at −5, 10, 20 °C) exported from the author's ANL-validated Simscape/Simulink Tesla Model 3 LR AWD powertrain model (v26, ≤2.2% BMS-net error on UDDS, HWFET, US06, and WLTP drive cycles). The comparison removes the model's HVAC term and normalizes auxiliary load.

Average error

3.33%

Worst point

5.12%

Within 5%

8 of 9

Within 6%

9 of 9

The 5.12% error at 50 mph and 20 °C occurs because the model applies a fixed 90% drivetrain efficiency, while the Simscape model accounts for speed-dependent motor and inverter losses. This is a known structural simplification.

ANL dynamometer: real lab data

The model is also validated against actual Argonne National Laboratory (ANL) D3 dynamometer measurements, using real current and voltage traces from depletion runs on a chassis dynamometer. Three vehicles were tested at 65 mph steady cruise under real ambient temperatures.

Tesla Model 3 LR AWD

7 data points, −10 to +27 °C. Average error 3.1%, maximum 6.2%. Validation performed at this speed.

Nissan Leaf e+

3 data points, average error 13.6%, maximum 14.2%. Residual error is primarily due to the unmodeled linear road-load term (B=3.18 N·s/m, approximately 2.7 kW at 65 mph).

Chevrolet Bolt EV

4 data points, error range 4–13%. This validation-only vehicle, not sold in the UK, highlights the model's missing linear road-load term.

HVAC model: measured heater channels

The model's PTC heater ramp (180 W/°C, 4 kW cap) is supported by measured cabin-heater channel data from the ANL Bolt EV's PTC system. The resulting cold-weather penalty aligns with the Tesla's observed total versus aerodynamic surplus at low ambient temperatures. See fixtures/anl_d3/hvac_findings.md for the full analysis.

The model does not currently include a hot-weather cooling term; approximately 1 kW AC compressor draw measured on the Bolt at 40 °C is not modeled.

The equations

The power required to move the car at a steady speed is the sum of four road loads, plus cabin and auxiliary electrical consumption. Speed v is measured in metres per second.

Aerodynamic drag

Paero = ½ · ρ(T) · Cd · A · v³

Rolling resistance

Proll = Crr · m · g · v(×1.1 in rain)

Drivetrain loss

Pdt = (Paero + Proll) · (1/ηdt − 1)

Efficiency

η = vmph / (Ptotal / 1000)mi/kWh

Air density is calculated using the ideal gas law, so colder, denser air reduces range. Cabin heating uses a resistive (PTC) demand of clamp((16 − T) · 180, 0, 4000) watts. A heat pump performs the same function at 45% of that above -5 °C.

Battery degradation

State of health follows an exponential approach to a minimum value: SoH = floor + (1 − floor) · e−k·age, with k adjusted so the first-year loss matches the calibrated annual figure for each chemistry and cooling type. Age and mileage are weighted 70/30.

Journeys & cost

Energy use is integrated for each journey segment; a rapid charge (2080%, approximated at 75% of the car's peak DC rate) is included before any segment that would drop below 10%. Running cost includes energy + VED + (from 2028/29) 3p/mile, compared to petrol at 40 mpg.

Constants & sources

Every constant in the model is sourced. Items marked provisional are modeling assumptions or pending expert review and are not presented as measured facts.

ConstantValueSource
Gravity g9.80665 m/s²CODATA / ISO 80000 standard value
Sea-level pressure101325 PaISO 2533 Standard Atmosphere
Specific gas constant287.05 J/(kg·K)CIPM-2007 dry-air molar mass
Rain rolling ×× 1.1Wet-road Crr uplift, provisional
HVAC setpoint16 °CSPEC §5.1, indoor-comfort reference
PTC heater ramp180 W/°C, cap 4000 WANL Bolt EV PTC heater data
Base auxiliary load320 WModelling assumption, provisional
Heat-pump fraction45% above -5 °CModelling assumption, provisional
Preheat HVAC ×× 0.5SPEC §5.1, provisional
Degradation blend70 / 30Age/mileage, provisional
Annual mileage ref8000 mi/yrDfT National Travel Survey, provisional
Journey SoC floor10%Model rule (SPEC §5.3)
DC-curve de-rate× 0.75Flat approximation, provisional
2028/29 eVED3p/mileAnnounced per-mile EV charge
EV standard VED£195/yrVED reform from 1 April 2025
Default petrol mpg40 mpgSPEC §5.4 user-selectable default
UK gallon4.54609 LUK Weights & Measures Act
Cost condition38 mph, 10 °C, dryBlended UK urban + A-road, provisional

Degradation & tariff data

1

Battery degradation profiles: Approximated from the Geotab 2024 fleet study and Nissan Leaf-specific fleet telemetry; provisional and pending author review.

2

Journey ambient temperature: UK monthly averages (Met Office 1991–2020); a per-location Open-Meteo refinement is planned.

3

Energy tariffs and pump prices: Indicative UK figures, updated periodically. Each value is individually sourced.

Who stands behind the model

F S Khan

Electric Vehicle Engineer

Every calculation on HonestRange is based on a physics model I developed and validated against real laboratory data. I specialize in modeling electric vehicle energy use, building detailed computer simulations, and verifying them against measurements from Argonne National Laboratory, a leading EV testing facility.

My Tesla Model 3 simulation has been verified to within 2.2% of real-world laboratory measurements across four international test cycles. The model used on this site is exported directly from that simulation, ensuring the figures presented are based on the same data used by research laboratories.

For full transparency, every equation, constant, and data source is published on this page. You can review exactly how each figure is calculated and assess the methodology yourself.

See the model in action

Every figure on each car page is calculated using this model. Explore real range and cost for popular used EVs.

Browse models