EV range calculator: Wh/km on a drive cycle, the battery for a range, and the ADVISOR 45 km check

An EV's range is its usable battery energy divided by the energy it uses per km on a drive cycle. Give the mass, rolling resistance, drag, efficiency, regen and accessories, pick MIDC, WLTC, UDDS, HWFET or a constant speed, and you get Wh/km at the wheels and from the battery, where the energy goes, and the km until the battery is empty. Free, runs in your browser.

The second tab rebuilds NREL ADVISOR's EV_defaults_in from its data files for the Week-3 ADVISOR challenge in Skill-Lync's hybrid electric vehicle course: 25 modules and 500 kg cargo give about 41 km on FTP, so the answer is no, and 28 modules is the smallest pack that reaches 45 km. Skill-Lync read 935 public reports on it. The third tab sizes the battery for a target range.

CMVR TAP-115/116 (MIDC) · EPA Test Car List 2023-24 · NREL ADVISOR 2003 data files · checked against a separate solution (753 checks at build) · Updated 22 Nov 2026

For the EPA presets the efficiency and regen share are the one pair that fits all four cars' EPA numbers (wall socket to wheels, so the charger is in it), and the result is energy from the wall. Their battery kWh is not in the EPA records used: type yours.

Range254 km23.77 MIDC cycles
From the battery102.4 Wh/kmincl. accessories
At the wheels94.6 Wh/kmtraction only
Usable battery energy left against distance, repeated MIDC0255075100050100150200250300 kmusable energy left %
  • One MIDC pass: 1180 s, 10.67 km. At the wheels the car needs 94.6 Wh/km to drive; the brakes take 27.1 Wh/km.
  • Regen recovers 60% of the braking energy at the wheels; 14.3 Wh/km gets back to the battery.
  • From the battery: 102.4 Wh/km. 26 kWh usable lasts 23.77 cycles = 253.6 km.
Where the battery energy goeskWhShare
Rolling resistance0.36333.3%
Aerodynamic drag0.35732.7%
Friction brakes (kinetic energy not recovered)0.11510.6%
Drivetrain losses: motor, inverter, gearbox (1 − η)0.15814.5%
Accessories0.0989.0%
Total from the battery1.092100%

Can EV_defaults_in travel 45 km on FTP with 500 kg cargo?

No. ADVISOR stops at about 41.4 km. The 500 kg of cargo takes the test mass from 1,144 kg to 1,508 kg, and the 25 lead-acid modules, about 7.7 kWh, run flat in the third FTP cycle. This calculator, built from ADVISOR's own data files, gets 40.7 km and stops after 5,475 s. With the default 136 kg cargo it gets 48.0 km, against 48.2 km in ADVISOR.

ADVISOR EV_defaults_in state of charge against distance on repeated FTP cycles, 25 modules, 500 kg cargo025507510001020304050 kmSOC %45 km

The fewest modules that reach 45 km is 28, in ADVISOR and here: 27 modules give 44.2 km (44.8 km in ADVISOR) and 28 give 46.0 km (46.3 km). Each extra module adds 11 kg and 0.3 kWh. Between 25 and 35 modules each one buys 1.54 km on average. The return per module falls slowly: ten more modules add 40% of energy but also 110 kg, so the energy per km rises 1.6% (188 to 191 Wh/km at the cells).

Range on repeated FTP cycles against the number of battery modules, 500 kg cargo: this calculator against ADVISOR01020304050607010152025303540ESS_PB25 moduleskm45 kmthis calculatorADVISOR, as reports state it

This calculator against ADVISOR, from the reports

ModulesCargoMassADVISOR (reports)This calculatorDiff.Example reports
15500 kg1,398 kg24.8 km23.90 km−3.6%1, 2
20500 kg1,453 kg33.5 km32.78 km−2.2%1, 2
25500 kg1,508 kg41.4 km40.71 km−1.7%1, 2
26500 kg1,519 kg42.7 km42.17 km−1.2%1, 2
27500 kg1,530 kg44.8 km44.16 km−1.4%1, 2
28500 kg1,541 kg46.3 km45.99 km−0.7%1, 2
29500 kg1,552 kg47.7 km47.48 km−0.5%1, 2
30500 kg1,563 kg49.1 km48.98 km−0.3%1, 2
35500 kg1,618 kg56.5 km
53.2 km on 3 cycles
56.09 km
53.23 km on 3
−0.7%1, 2 · 3 cycles: 1, 2
25136 kg1,144 kg48.2 km47.99 km−0.4%1, 2

ADVISOR distances as the reports state them, 500 kg cargo unless marked. Both columns run CYC_FTP until SOC < 0.001. Two values appear for 35 modules because three cycles (53.3 km) end before the battery does.

This calculator runs 1.3% short of ADVISOR on average and at most 3.6%. The gap is largest for small packs, which spend longest near empty. Nothing was tuned to close it. What differs:

  • Trace misses near empty. As SOC falls, the pack's internal resistance rises (ESS_PB25: 12 to 41 mΩ per module) and its power limit drops. ADVISOR's driver then lags the trace; here a lagging second is solved forward from the torque that is left, and the next second chases the trace again. With 25 modules and 500 kg, 192 seconds miss by more than 2 mph. ADVISOR's driver model and Simulink solver are not reproduced exactly.
  • Tyre slip uses the static front-axle load (60% of the weight). ADVISOR's traction block adds weight transfer.
  • Battery and motor temperature are left out. ESS_PB25's tables are the same at 0, 22 and 40 °C, so this changes nothing for the battery.
  • The accessory is off during the 600 s key-off soak, as the key-on signal switches the accessory block in ADVISOR's BD_EV diagram.

How the ADVISOR model works

Test mass is the sum of rounded component masses, as gui/recompute_mass.m does it: glider 592 kg, motor and controller 91 kg, final drive 50 kg, wheels 0, 11 kg per module, plus cargo.

Every second the vehicle needs m·a, 0.009·mg rolling resistance and ½ × 1.2 × 0.335 × 2.0 m² × v² of drag, at the mean speed of the second. The wheel adds its inertia (3.264 kg·m², 41 kg equivalent at r = 0.282 m) and an axle loss of 1.6 to 9.6 N·m with mass. When braking, the share of braking that goes through the motor is 0 at 0 mph, 50% at 10 mph and 80% from 60 mph, per WH_SMCAR; friction brakes take the rest. TX_1SPD sets the single gear ratio:

and its efficiency comes from tx_eff_mapper's loss equation on output speed, torque and power, saturated at the map's edges. The motor's electrical input comes from MC_AC75's map (losses from the 0.70 to 0.92 efficiency map, mirrored for regen), with the torque limit times the 1.8 overtorque factor and the 480 A current limit. The motor is off when the car is stopped. Then the battery, 25 Ah modules in series with Voc and R from SOC:

Coulombic efficiency 0.9 applies to charging current only, as ADVISOR's Rint documentation describes. The run starts at SOC 1, the 700 W accessory draws from the bus whenever the key is on, and it stops when SOC < 0.001.

What did the 935 ADVISOR reports find?

Skill-Lync read all 935 public "ADVISOR Tool" reports on 22 Nov 2026. 543 write 41.4 km. 638 write 17.7 or 17.8 km, one FTP cycle, and 253 write 35.5 km, two cycles: in many of them that is where the run ended with charge left. 177 write 53.2 or 53.3 km, three cycles. 55 write 48.2 km, the default-cargo result. Each count is of reports that write that distance followed by "km" anywhere in the text, so a report can be in several.

The usual slips, which "Check my numbers" in the second tab looks for: a distance that is a whole number of cycles (raise the cycle count), 48.2 km when the question asks for 500 kg (the cargo didn't reach the vehicle), 41.4 km after changing the modules (the change didn't take) and 11.99 km (UDDS, not FTP).

  • 41.4 km is what ADVISOR gives for 25 modules and 500 kg cargo in the reports Skill-Lync read. This calculator gets 40.7 km.

What is the formula for EV energy consumption per km?

Battery Wh/km is the road load times distance, divided by the drivetrain efficiency, less what regen returns, plus accessories. The road load, as NPTEL's hybrid and electric vehicles course (IIT Guwahati, after Ehsani) writes it, with δ the rotating-mass share:

or, for an EPA test car, the coastdown coefficients from the Test Car List with the equivalent test weight as the mass:

The calculator steps through the cycle one second at a time:

η is battery to wheels, r the share of the braking energy at the wheels that regen recovers, below the cut-off speed and above the power cap none. The cycle repeats until the usable energy is used, the last cycle second by second:

At a constant speed it reduces to a hand formula, which the verifier checks exactly:

The illustrative car on each cycle

CycleOne cycleWheelsBatteryRange, 26 kWhNo regen
MIDC (India, Part 1 x 4 + Part 2, 1,180 s)10.67 km, 1,180 s
top 90.0 km/h
94.6 Wh/km102.4 Wh/km254 km223 km
WLTC class 3b (WLTP)23.27 km, 1,800 s
top 131.3 km/h
124.1 Wh/km130.4 Wh/km201 km178 km
UDDS (EPA city)11.99 km, 1,369 s
top 91.2 km/h
103.9 Wh/km103.7 Wh/km251 km204 km
HWFET (EPA highway)16.51 km, 765 s
top 96.4 km/h
101.4 Wh/km114.1 Wh/km227 km218 km
FTP-75 (no soak)17.77 km, 1,874 s
top 91.2 km/h
106.7 Wh/km106.8 Wh/km244 km200 km
US06 (EPA aggressive)12.89 km, 600 s
top 129.2 km/h
164.6 Wh/km167.5 Wh/km155 km136 km

1,250 kg, δ 5%, Crr 0.010, Cd 0.32, A 2.1 m², η 88%, 60% regen, 300 W accessories: typed values for illustration, not a real car.

Which drive cycle is used for EV range in India?

The Modified Indian Driving Cycle (MIDC). CMVR TAP-115/116 defines it for four-wheelers as four elementary urban cycles (ECE-15 with 2 s gear changes, 195 s each, 50 km/h top) and an extra-urban part of 400 s with a 90 km/h top, the low-powered EUDC, not the 120 km/h one in UN R101. The calculator builds it second by second from the official tables: 10.667 km in 1,180 s, against 4 × 1.013 + 6.594 = 10.647 km stated (the gear-change rows are held at constant speed here, which adds 0.2%).

AIS-040 (2003) ends the range test when the car can't follow the trace up to 85% of the cycle's or the car's top speed, whichever is less. MG quotes the Windsor EV's range on MIDC Part 1 + Part 2. Other claims don't always say which parts were run. Claimed ranges and the battery Wh/km they imply:

CarBatteryClaimed rangeBasisImplied
MG ZS EV50.3 kWh461 kmcertified range (maker page)109 Wh/km
MG Windsor EV (Exclusive Pro)52.9 kWh449 kmMIDC Part 1 + Part 2118 Wh/km
MG Comet EV17.4 kWh230 kmARAI-certified76 Wh/km
Tata Tiago.ev 24 kWh24 kWh315 kmMIDC (launch press release, 2022)76 Wh/km
Tata Tiago.ev 19.2 kWh19.2 kWh250 kmMIDC, estimated (launch press release, 2022)77 Wh/km

Battery kWh divided by claimed km, as stated on the makers' pages (MG Motor India, seen 22 Nov 2026; Tata Motors press release, 28 Sep 2022). Gross capacity, so the real figure from the usable part is lower.

How close is the model to EPA's measured energy use?

Within 5.1% on all four cars, with one efficiency and one regen share for all of them. The EPA Test Car List gives each car's coastdown coefficients and test weight, and fueleconomy.gov its unadjusted city (UDDS) and highway (HWFET) MPGe, measured from the wall socket. At 33.7 kWh per gallon, those become Wh/mi. The model's traction T and braking B energy at the wheels then have to satisfy:

Solved for each car, η is 82, 84, 82, 77% and ρ 87, 79, 95, 107% (Leaf, Bolt EUV, Model 3, Kona). Fitted to all eight numbers at once: η = 81.1%, ρ = 91.8%, and the predictions land here:

CarCycleWheels: traction / brakingEPA, wallPredictedDiff.
2023 Nissan Leaf (147 hp)UDDS134 / 62 Wh/km119.2 Wh/km
175.6 MPGe
118.6 Wh/km−0.5%
HWFET129 / 13 Wh/km148.3 Wh/km
141.2 MPGe
149.1 Wh/km+0.6%
2023 Chevrolet Bolt EUVUDDS135 / 65 Wh/km117.5 Wh/km
178.2 MPGe
117.9 Wh/km+0.3%
HWFET126 / 14 Wh/km140.5 Wh/km
149 MPGe
144.5 Wh/km+2.8%
2023 Tesla Model 3 RWDUDDS138 / 70 Wh/km113.0 Wh/km
185.3 MPGe
117.4 Wh/km+3.9%
HWFET112 / 17 Wh/km123.1 Wh/km
170.1 MPGe
125.3 Wh/km+1.7%
2024 Hyundai Kona Electric Long RangeUDDS129 / 67 Wh/km113.8 Wh/km
184 MPGe
109.5 Wh/km−3.7%
HWFET118 / 15 Wh/km142.0 Wh/km
147.5 MPGe
134.7 Wh/km−5.1%

The Kona's own fit gives ρ above 100%, which is not physical: one constant efficiency can't describe a real drivetrain at every load. The label values are about 0.7 times the unadjusted ones (Leaf 0.700): 40 CFR 600.115-11 lets makers multiply 2-cycle results by 0.7. The 2023 Bolt EV is left out because its Test Car List values don't match fueleconomy.gov.

A published worked example, reproduced

x-engineer.org's EV energy example: a 1,741 kg car with an 80 kg driver and a 1.05 rotating-mass factor on the kerb mass (1,908.05 kg), Cd 0.36, A 2.42 m², air at 1.202 kg/m³ and Crr 0.011, on the WLTC class 3 cycle, with no losses and full regen. They give 3,205.39 Wh and 137.8 Wh/km. With their time-stepping (the acceleration of each second from the speed before it, at the sample speed) the calculator gets 3,205.39 Wh and 137.8 Wh/km. With its own mean-speed steps it gets 3,072.28 Wh (−4.2%): same physics, different integration of the same trace.

How do you size an EV battery for a target range?

Multiply the target km by the Wh/km, then divide by the usable SOC window, and repeat with the battery's mass added until it settles. The illustrative car at 1,050 kg without its battery, 7 kg per kWh and a 90% window needs 31.3 kWh usable for 300 km on MIDC: 34.8 kWh gross, 244 kg of pack, 1,294 kg in all, 104.4 Wh/km. It settles in 10 steps:

If the battery's mass costs more energy than it stores, the iteration runs away. With lead-acid ESS_PB25 modules (36.7 kg per kWh) and an 80% window, 3,000 km on MIDC never converges: each kWh adds about 46 kg and pushes the Wh/km up faster than the extra energy covers. Use the battery pack calculator to turn the kWh into cells in series and parallel.

What this calculator does not cover

  • Cold, heat and air conditioning beyond a constant accessory load, and battery ageing.
  • Efficiency maps for your own motor. The first tab uses one efficiency; only the ADVISOR tab uses a map (MC_AC75).
  • The Prius gradeability question of the same challenge: use the gradeability calculator.
  • The AIS-040 Rev.1 end-of-test rule. Its text could not be read from ARAI's site; the 2003 rule is quoted above.
Learn EV modelling

Running ADVISOR is the easy part; knowing why a pack falls short of 45 km, and what one more module buys, is what EV powertrain teams look for. Vehicle energy on drive cycles and ADVISOR are taught in Skill-Lync's hybrid and electric vehicle courses.

Frequently asked questions

How do you calculate the range of an electric vehicle?

Work out the battery energy per km on a drive cycle, then divide the usable kWh by it. Per second, road load times speed is the wheel power; divide by the drivetrain efficiency, take off what regen returns, add accessories. The illustrative 1,250 kg car here uses 102 Wh/km on MIDC: 254 km from 26 kWh.

Can the ADVISOR EV_defaults_in vehicle travel 45 km on FTP with 500 kg cargo?

No. With 25 ESS_PB25 modules and 500 kg of cargo the test mass is 1,508 kg and ADVISOR stops at about 41.4 km, when SOC falls below 0.001. This calculator, rebuilt from ADVISOR's data files, gets 40.7 km. 28 modules is the smallest pack that reaches 45 km.

How many battery modules does EV_defaults_in need for 45 km?

28 modules, in ADVISOR and in this calculator. Each ESS_PB25 module adds 11 kg and 0.3 kWh at 12 V and 25 Ah. With 500 kg cargo, 27 modules give 44.2 km here (44.8 km in ADVISOR) and 28 give 46.0 km (46.3 km). Run enough FTP cycles, or the run ends before the battery does.

Why does ADVISOR show 17.77 km or 35.5 km for the EV range?

Because the run ended with charge left. One CYC_FTP is 2,477 s and 17.77 km; ADVISOR stops at the number of cycles you set. 35.5 km is two cycles and 53.3 km is three. Only when SOC drops below 0.001 before the last cycle is the distance the range.

What is the MIDC drive cycle?

India's Modified Indian Driving Cycle for cars: four 195 s urban cycles (ECE-15 with gear changes, 50 km/h top) and one 400 s extra-urban cycle with a 90 km/h top. CMVR TAP-115/116 gives 1.013 km per urban cycle and 6.594 km for the extra-urban part, about 10.65 km in 1,180 s.

What is a good Wh/km for an electric car?

On India's claimed MIDC ranges, from about 76 Wh/km of battery capacity for a small city car (MG Comet EV, 17.4 kWh, 230 km) to 118 Wh/km for a larger car (MG Windsor EV, 52.9 kWh, 449 km on MIDC Part 1 + 2). Real driving uses more. EPA's unadjusted city figure for the 2023 Leaf is 119 Wh/km from the wall.

Related on Skill-Lync

Sources

  • NREL ADVISOR 2003 (open-source release, GitHub mirror geraldo-reboucas/adv-vehicle-sim): saved_vehicles/EV_defaults_in.m; data files VEH_SMCAR, WH_SMCAR, ESS_PB25, MC_AC75, TX_1SPD, ACC_HYBRID, PTC_EV, CYC_FTP; gui/recompute_mass.m, gui/tx_eff_mapper.m, init_conds_hot.m; models BD_EV.mdl and lib_energy_storage, lib_electric_machine, lib_transmission, lib_wheel, lib_accessory; documentation ess_rint, motor_controller, mc_cont, wheel_axle.
  • MoRTH / CMVR / TAP-115/116 (Issue 4), Part 14 Chapter 3, Tables II and III: the Modified Indian Driving Cycle. ARAI AIS-040 (2003), Battery operated vehicles, method of measuring the range, clause 5.4; AIS-039 (2003). UNECE Regulation No. 101 Rev.3, Annexes 7 and 9.
  • US EPA, Test Car List data files for model years 2023 and 2024 (Equivalent Test Weight, Target Coef A, B, C); fueleconomy.gov web services, vehicle records 46016, 45750, 46206 and 47449 (UCity, UHighway, label MPGe, range). 40 CFR 600.115-11 (2015 edition): the 0.7 factor. EPA dynamometer schedules UDDS, HWFET, US06; WLTC class 3b from UNECE GTR 15.
  • NPTEL 108103009, Introduction to Hybrid and Electric Vehicles (IIT Guwahati), Module 2, Lectures 3-4: road load, rolling resistance table, rotating mass factor (after M. Ehsani et al., Modern Electric, Hybrid Electric and Fuel Cell Vehicles).
  • x-engineer.org, "EV design – energy consumption": the WLTC worked example (3,205.39 Wh, 137.8 Wh/km).
  • MG Motor India product pages (ZS EV, Windsor EV, Comet EV), seen 22 Nov 2026; Tata Motors, Tiago.ev launch press release, 28 Sep 2022.
  • Skill-Lync, all 935 public ADVISOR Tool reports, read 22 Nov 2026. scripts/verify/ev_range.py re-implements both models with different numerical methods (root-finding for the battery current and the trace-miss speed, bisection for the battery size) and the page checks itself against it at build time.
Free counselling

From an ADVISOR run to an EV powertrain job.

Explaining with numbers why a pack falls short, and what it takes to reach the range, is the kind of answer EV teams look for. A Skill-Lync counsellor can map the HEV or EV course that fits the role you want.

Talk to an EV course counsellor

Free. We call you back, usually the same day.

By submitting you agree to our Terms and Privacy Policy.

Talk to a career counsellor

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.