Rated range is a test-cycle number, and the US Department of Energy has quantified how far it travels from reality. Chassis dynamometer testing across EPA cycles found battery electric range falling 41% at 20°F against 10% for a comparable combustion vehicle, and roughly half at 0°F — with a spread running from 39% loss in highway driving to 59% in urban driving with ample idle time. That last figure is the one worth sitting with, because urban stop-start work is precisely the duty cycle electrification suits best. The conclusion follows directly: range is not a vehicle specification at all. It is an output of route, load, temperature and duty cycle together. Build a route energy model with us.
Your Worst Route Is the One Nobody Measured in February
What actually consumes energy on a route, why the vocations best suited to electrification lose the most range in winter, and how to build a usable energy model from five readings rather than a spreadsheet of assumptions.
What Temperature Actually Does
The headline is that cold costs range. The useful detail is how much, why, and which operations it hurts most — and on all three the published testing is unambiguous.
From the Department of Energy's Vehicle Technologies Office programme record, based on testing at Argonne National Laboratory's Advanced Mobility Technology Laboratory with cabin temperature held at 72°F. At 20°F battery electric range fell 41% against 10% for an internal combustion vehicle. At 0°F the average loss was around half, ranging from 39% in highway conditions to 59% in urban driving with ample idle time. Note the hot-weather result as well: at 95°F range fell 14%, which the record describes as "an identical proportional decrease to that of an ICEV in the same conditions." Heat is not an electric-specific problem. Cold is.
The mechanism matters because it is addressable. The dominant cause is cabin heating rather than battery chemistry, with HVAC drawing between 1.8 and 3.7 kW depending on technology and conditions — and a heat pump reduced total HVAC draw by 38% against resistance heating at 20°F. That makes heating technology a procurement decision with a direct range consequence, and preconditioning on shore power an operational one.
The Six Inputs That Decide a Real Number
Rated range accounts for roughly one of these. Swipe on mobile.
| Input | What it does | How to handle it |
|---|---|---|
| Ambient temperature | The largest single variable, and almost entirely cabin heating rather than battery chemistry. HVAC draws 1.8 to 3.7 kW depending on the technology and conditions. | Derate by season, not by annual average |
| Duty cycle | Stop-start urban work with idle time loses far more in cold than highway running — 59% against 39% at 0°F. The vocations best suited to electrification are the worst affected in winter. | Derate by route type, separately |
| Payload | Weight costs energy, and on a route with many stops it costs it repeatedly through acceleration. A truck assigned on empty-weight consumption will not finish a loaded day. | Model the loaded case, not the average |
| Terrain and grade | Climbing costs energy that regen partly returns on the way down — partly. A hilly route and a flat route of identical distance are not the same assignment. | Use actual route profiles |
| Auxiliary loads | Lift gates, refrigeration, PTO-driven bodies and compressors all draw from the same pack as propulsion. On a refuse or utility body this can be a substantial share. | Measure them; do not estimate |
| Capacity fade | The pack delivering the route today is not the pack that was commissioned. A margin that was comfortable in year one is the margin that strands a truck in year four. | Re-check assignments annually |
The second row deserves emphasis because it inverts the usual assumption. Urban stop-start duty is where electric drivetrains are strongest — regenerative braking recovers what friction brakes would waste, and idle costs nothing. It is also, per the same testing, where cold weather takes the largest bite. A municipal or delivery fleet can be entirely right about electrification and still be caught out in January if it planned on an annual average.
You probably need one measurement, not a model
Most fleets worried about range are worried about one or two specific routes. Run those, loaded, in the conditions you actually fear, and record kWh per mile. In thirty minutes we will set up the capture so the reading lands against the route rather than disappearing into a vehicle average.
Five Steps to an Assignment You Can Defend
None of this requires modelling software. It requires collecting readings somebody currently is not collecting.
- Measure, do not model, the baselineRun the route in the vehicle you intend to assign and record energy consumed and distance. One real reading beats any spreadsheet, and you only need the route you are actually worried about.
- Repeat it loaded and in winterThree readings make a usable picture: mild and empty, mild and loaded, cold and loaded. The third is the one that decides the assignment, and it is the one nobody collects because it is inconvenient to collect.
- Convert to kWh per mile, per routeNot per vehicle. Range is a property of the pairing, and the same truck on two routes produces two different numbers that are both correct.
- Set the margin deliberatelyDecide what reserve you require at the end of the worst case and write it down. A margin that exists as a feeling gets eroded quietly whenever dispatch is under pressure.
- Re-derive when anything changesNew route, new body, new season, another year of capacity fade. The assignment was correct under conditions that have now moved.
Setting a Margin That Survives a Bad Day
Every assignment carries an implicit reserve. Making it explicit is most of the discipline.
Start from the worst realistic case rather than the typical one: loaded, cold, with auxiliary loads running, on a pack that has aged. Then decide what state of charge you want remaining when the vehicle returns under those conditions, and write the figure down. A documented reserve can be argued about and adjusted. An undocumented one gets consumed the first time dispatch needs an extra drop.
Build in the things that will happen rather than the things that might. A detour, a charger out of service on return, a driver who runs the cab warmer than the test assumed, a route extended by one stop because somebody was off. None of these is exceptional and all of them draw on the same reserve.
Where the margin cannot be made to work, the honest answers are to shorten the route, add an opportunity charge, or assign a different powertrain to that run. Those are all legitimate. What is not legitimate is assigning on a rated figure and treating the first stranded truck as bad luck, since the testing above makes it entirely predictable.
When the Route Does Not Fit
A route that fails the margin test is useful information, not a failure. There are four responses and only one of them involves buying a bigger battery. Swipe on mobile.
| Option | What it involves | What it costs |
|---|---|---|
| Shorten the route | The cheapest fix and the one least often considered, because routes tend to be inherited rather than designed. Moving three stops to an adjacent run can bring a marginal route inside the margin without touching a vehicle or a charger. | No capital cost |
| Add an opportunity charge | A mid-shift top-up at a depot, a transfer station or a customer site. It converts a range problem into a dwell-time problem, which is usually easier to solve — but check what it does to your peak demand before committing, since a mid-day charge lands in the worst part of the tariff. | Capital, plus a demand charge consequence |
| Assign a different powertrain | Mixed fleets exist because no single powertrain suits every run. Putting the longest, coldest, heaviest route on diesel or CNG is not a failure of the electrification programme; it is the programme working correctly. | No capital cost if the vehicle exists |
| Specify differently next time | Heat pump rather than resistance heating, a larger pack where payload allows, or auxiliary loads moved off the traction pack. These only help at procurement, which is why the measurement should happen before the next order rather than after it. | Applies to the next purchase only |
The second option carries a hidden consequence worth naming. Mid-shift charging happens during the working day, which is exactly when a commercial electricity tariff is least forgiving — so a mid-day top-up can set a peak that costs considerably more than the energy it delivered. If opportunity charging is the answer, model its effect on peak demand before the charger is installed rather than after the first bill.
How FleetRabbit Handles This
Four things, and the first is the one that makes every other number on this page usable.
Because range is a property of the pairing. The same vehicle on two routes gives two valid answers, and only route-level data supports an assignment decision.
A summer figure and a winter figure for the same route, so the derate is measured from your own operation rather than borrowed from a test cycle.
Pack health only means something against the energy a route actually needs. Holding both together turns reassignment into a plan rather than a discovery.
A lift gate or a packer drawing from the same pack is a route cost. Recording it separately is what makes the next assignment accurate.
What Telematics Gives You, and What It Does Not
Most fleets already have the raw material for everything above, which is the good news. The gap is usually in how it is aggregated rather than in whether it exists.
Telematics will reliably give you energy consumed, distance travelled, and often state of charge over time. That is enough to derive kWh per mile, which is the number this entire page rests on. What it will not do on its own is attach that figure to a route rather than to a vehicle and a date, and that single aggregation choice decides whether the data answers an assignment question or merely describes last Tuesday.
The second gap is conditions. A consumption figure without the ambient temperature, the load and whether auxiliary equipment was running is a number you cannot compare to anything. Two readings from the same truck a month apart can differ by half for entirely legitimate reasons, and without the context you will either dismiss the variance as noise or draw the wrong conclusion from it.
The practical fix is small: record the conditions alongside the consumption, and group by route. Neither requires new hardware on a fleet that already runs telematics — it requires deciding that route is the unit of analysis rather than vehicle, which is a reporting decision rather than a technology one.
Questions Fleets Ask
How much range do electric trucks lose in winter?
DOE testing found 41% at 20°F against 10% for a comparable combustion vehicle, and around half at 0°F — between 39% in highway conditions and 59% in urban driving with idle time. Your own figure depends on duty cycle, heating technology and load, which is why one winter reading on your actual route is worth more than any published average. Log your winter baseline free.
Why is cold so much worse for electric than for diesel?
Because a combustion engine produces waste heat that warms the cab for free, and a battery electric vehicle has to generate that heat from the same energy it uses to move. HVAC draw of 1.8 to 3.7 kW is propulsion energy diverted. A heat pump cut total HVAC draw by 38% against resistance heating at 20°F, which makes it a specification worth checking at purchase. Review your heating spec.
Does hot weather hurt too?
Less than people expect, and not disproportionately. The same testing found a 14% range reduction at 95°F and described it as an identical proportional decrease to an internal combustion vehicle in the same conditions. Heat affects both roughly equally; cold does not. Record both seasons.
How do we account for lift gates and body loads?
Measure them separately rather than folding them into a vehicle average. A packer, a refrigeration unit or a lift gate draws from the same pack as propulsion, and on some bodies that share is large enough to decide an assignment. If it is not recorded separately, it shows up as unexplained variance between trucks doing supposedly identical work. Separate the auxiliary draw.
Should we just buy bigger batteries?
It is one answer and rarely the cheapest. More pack means more weight, which costs payload and some of the energy the extra capacity provided. Shortening a route, adding an opportunity charge, or assigning that run to a different powertrain often costs less. Work out what the specific route needs before specifying against the worst route in the fleet. Size it against real routes.
How often should assignments be reviewed?
At least annually, and whenever a route, body or season changes. Capacity fade means an assignment that was comfortable at commissioning narrows every year, and the failure mode is a truck returning early rather than a warning light. Treat it as a scheduled review rather than an event. Set the review cadence.
What if the margin does not work on our longest route?
Then that route is telling you something useful before it strands anybody. Shorten it, add a mid-shift charge, or run it with a different powertrain — all three are better outcomes than discovering the limit operationally. Mixed fleets exist precisely because no single powertrain suits every run, and the rest of your maintenance programme is identical across all of them. Match trucks to routes.
Measure the Route You Are Worried About
FleetRabbit records energy per mile against the route rather than the truck, holds summer and winter baselines separately, separates auxiliary draw from propulsion, and reads capacity trend against what each route actually demands.
Free for 3 vehicles · no credit card required
