The range number your EV displays at 11 pm is not a lie — it is an optimistic estimate based on warm batteries and recent driving conditions. The 190 km it shows at 7 am is not a malfunction — it is the battery's honest performance capability at 12°C after sitting still all night.
- Overnight range loss (10–40%) results from four overlapping effects: temperature-driven internal resistance increase, OCV relaxation after charging, vampire drain, and LFP-specific SOC estimation errors from the flat OCV curve.
- Temperature is the dominant contributor — internal resistance roughly doubles to triples between 25°C and 5°C for NMC, causing the BMS to estimate less accessible capacity and display lower range.
- LFP vehicles show more dramatic overnight swings than NMC because a 30 mV OCV relaxation corresponds to a ~43% SOC correction on LFP's near-flat curve versus only ~7% on NMC's steeper curve.
- The range display recovering as you drive in the morning is not a miracle of efficiency — it is the battery warming from its own heat generation, reducing internal resistance and unlocking previously inaccessible capacity.
- For range-critical trips in cold conditions, always plan using the cold-morning range estimate, not the warm-evening post-charge display.
If you charged your EV to full last night, the range display showed 300 km. This morning, without driving a single kilometre, it shows 190 km. You haven't lost 110 km of range. But you haven't gained it back either. What changed is a combination of four overlapping effects — temperature, OCV relaxation, vampire drain, and BMS algorithm behaviour — each of which causes the displayed number to be lower in the morning than the night before.
Effect 1: Temperature and Usable Capacity
Battery internal resistance increases significantly with decreasing temperature. For NMC cells at 5°C compared to 25°C, internal resistance roughly doubles to triples. The consequence is that the terminal voltage under load (the voltage the battery delivers when current flows) drops further at cold temperatures.
The BMS cuts off discharge when any cell reaches its minimum voltage threshold (typically 2.8–3.0V for NMC). At 25°C, the cell can deliver 95% of its nominal capacity before reaching this threshold. At 5°C with 2–3x higher resistance, the voltage under load drops steeply and the cutoff is reached earlier — at perhaps 80–85% of nominal capacity for a typical cell.
The range display implication: Your range estimate at 11 pm (35°C cell temperature from the day's driving and charging) assumed you could access 95% of the battery's stored energy. The range display at 7 am (12°C cell temperature after sitting overnight) is calculated with the BMS knowing that only 82% of stored energy is accessible at that temperature. 300 km × (82/95) = 259 km — already a 41 km reduction from temperature alone.
The temperature effect on range is real but temporary. Once you start driving and the battery warms up from its own internal heat generation, the accessible capacity increases back toward room-temperature levels. After 15–20 minutes of driving, the range estimate typically climbs from the cold-morning low. This 'range recovery during driving' confuses drivers who interpret the rising range number as driving efficiency gain — it is actually the battery delivering better performance as it reaches operating temperature.
Battery internal resistance (DCIR) increases with decreasing temperature due to slower ionic mobility in the electrolyte and electrodes — Arrhenius-type behaviour where each 10°C drop increases resistance by 20–50% depending on chemistry. For NMC cells, resistance roughly doubles to triples between 25°C and 5°C. Under load, the terminal voltage drops further at cold temperatures: V_terminal = V_OCV − I × R_internal. The BMS's minimum cutoff voltage is reached earlier in the discharge curve, meaning less total energy is delivered before cutoff. For a cell at 5°C versus 25°C, accessible capacity may drop by 10–18% — directly reducing the displayed range estimate, since the BMS continuously updates its accessible energy estimate based on measured cell temperature.
Effect 2: OCV Relaxation and Post-Charge Estimation
When charging stops, the battery's terminal voltage includes a polarisation component — residual voltage gradients inside the cell from the charge current. This extra voltage decays away over 30 minutes to 4 hours as ion concentrations equilibrate within the electrodes.
Immediately after charging ends:
- The terminal voltage reads high (includes polarisation)
- The BMS estimates SOC from this elevated voltage
- The BMS calculates range from this optimistic SOC
After 4 hours of rest (7 am in the morning):
- The terminal voltage has fully relaxed to the true OCV
- The true OCV corresponds to a slightly lower SOC
- The BMS corrects its SOC estimate downward
- Range estimate falls to reflect the corrected SOC
For NMC cells, the OCV relaxation effect is typically 20–50 mV, corresponding to a SOC correction of 2–5%. For a 300 km range EV, a 3% SOC correction is 9 km of range reduction.
Effect 3: The LFP Flat Curve Problem
LFP chemistry has a famously flat OCV-SOC curve. Between 20% and 90% SOC, the OCV changes by only 50–100 mV total — compared to 200–400 mV for NMC over the same range.
This creates a disproportionate OCV relaxation problem:
| Chemistry | OCV-SOC slope (20–80% SOC) | 30 mV OCV change = SOC change of... | Range error on 300 km EV |
|---|---|---|---|
| NMC 622 | ~4 mV / % SOC | 7.5% SOC | 22 km |
| NMC 811 | ~4.5 mV / % SOC | 6.7% SOC | 20 km |
| LFP | ~0.07 mV / % SOC | 43% SOC | 129 km |
A 30 mV OCV relaxation that creates a 7.5% SOC correction on NMC — already meaningful — creates a 43% SOC correction on LFP. In practice, OCV relaxation is smaller than 30 mV for LFP in the plateau region (the voltage is so flat that even a 20 mV relaxation produces extreme SOC uncertainty). This is why LFP-battery vehicles using OCV for SOC estimation perform especially poorly: the OCV contains almost no information about SOC in the plateau region.
Modern BMS firmware for LFP handles this through: Coulomb counting as the primary SOC estimator during operation, OCV correction only at the top and bottom of charge (where the LFP curve is steeper), and periodic recalibration at full charge.
LFP's OCV-SOC curve is nearly flat between 20% and 90% SOC — voltage changes by only 50–100 mV across this ~70% SOC range, compared to 200–400 mV for NMC over the same window. When OCV relaxation occurs after charging (typically 20–50 mV voltage drop as polarisation decays), the BMS must convert that voltage change into a SOC correction. On NMC, 30 mV corresponds to roughly 7% SOC change. On LFP, that same 30 mV corresponds to 30–40% SOC change because the slope dV/dSOC is so shallow. This is why Tata Nexon EV (LFP) owners commonly report dramatically larger overnight range swings than Hyundai Ioniq 5 (NMC) owners experiencing the same ambient temperature drop — the LFP voltage curve amplifies every small voltage change into a large displayed range correction.
Effect 4: Vampire Drain
Modern EVs do not fully power down when you park them. The BMS continues monitoring cell voltages and temperatures. Cellular connectivity stays alive for remote monitoring apps and over-the-air updates. Climate pre-conditioning systems may activate intermittently.
Typical overnight vampire drain:
- Budget Indian 2W/3W EV with basic BMS: 10–50 Wh (negligible)
- Mid-range Indian 4W EV: 100–250 Wh
- Premium connected EV (Tesla, Ioniq): 200–500 Wh
- EV with poor firmware optimisation and active connectivity: 500–800 Wh
At 5 km/kWh efficiency, 400 Wh of vampire drain equals 2 km of range. This is a real but minor contributor to overnight range loss — the 110 km overnight reduction in our example is not from vampire drain, which might account for 2–4 km.
Indian EV owners who leave their car parked for multiple days (e.g., international travel, extended leave) can experience significant vampire drain accumulation. A vehicle losing 300 Wh/day left for 10 days loses 3 kWh — potentially 15–20 km of range plus the risk of cells dropping into deep discharge if the pack was not full before parking. For extended parking, charge to 80% before leaving (not 100% — high SOC accelerates calendar aging, but not so low that vampire drain could reach the BMS low-voltage cutoff).
How Range Estimation Algorithms Handle All This
Most EV BMS range estimation systems use a weighted average of recent consumption history multiplied by estimated remaining usable energy:
Displayed range = Estimated remaining energy / Estimated consumption rate
Both inputs have temperature and history dependence:
- Remaining energy estimate adjusts for temperature (reduced accessible capacity at cold temperatures)
- Consumption rate estimate is typically a rolling average of recent trips — if you drove conservatively yesterday, your rolling average is optimistic for tomorrow's motorway driving
This is why range displays also vary based on recent driving style, not just temperature. A driver who spent the evening on quiet residential roads at 25 km/h will see a higher range estimate than one who drove at 120 km/h on a motorway, even with the same battery state.
For Indian EV drivers planning long-distance overnight stops in winter conditions (Rajasthan, Himachal Pradesh, where temperatures drop to 0–8°C overnight), the cold-morning range estimate should be treated as the planning number, not the warm-evening estimate. A driver who charged to full (300 km displayed at night) and plans a 250 km next-day trip may find only 200 km available at 7 am in cold conditions — an 83% confidence trip becomes a range anxiety situation. Always plan based on the cold-morning range estimate, or pre-condition the battery before departure.
Why It Gets Better As You Drive
After 15–20 minutes of driving, range estimates typically increase. Drivers often find this confusing ('I drove 20 km but my range only dropped 10 km'). The explanation:
- The battery warms from internal heat generation during discharge
- Warmer cells have lower internal resistance
- Lower resistance → less voltage drop under load → higher terminal voltage
- Higher terminal voltage → BMS estimates more energy is accessible
- Updated accessible energy estimate → higher range display
The battery is not recharging. The range that was hidden by cold temperature is becoming available as the battery reaches operating temperature. The display correctly reflects improving performance.
Key Takeaways
- Overnight range reduction of 10–40% is normal for EVs, caused by overlapping effects: temperature-driven resistance increase, OCV relaxation correcting an overstated post-charge SOC, and vampire drain. Temperature is the dominant contributor in most cases.
- At 10°C vs 35°C cell temperature, usable NMC capacity drops by 10–15% — a 300 km range becomes 255–270 km from temperature alone, before any OCV relaxation or drain effects are added.
- LFP vehicles show more dramatic overnight range swings than NMC because the LFP OCV-SOC curve is so flat that any voltage relaxation corresponds to a large SOC correction. LFP BMS systems using coulomb counting are less susceptible to OCV-relaxation errors but remain fully vulnerable to temperature effects.
- Range estimates recover as you drive and the battery warms up. The apparent range gain during early kilometres reflects improving accessible capacity as cell temperature rises, not improved driving efficiency.
- For range-critical trips in cold conditions, always plan using the cold-morning range estimate, not the warm-evening displayed range after charging completes.
Part of the deepdive Series
Frequently Asked Questions
Why does an EV's displayed range drop significantly overnight without driving anywhere?
What is OCV relaxation and why does it cause range estimation errors?
How much range does a typical EV lose overnight to vampire drain?
Why do LFP battery vehicles show more dramatic overnight range swings than NMC vehicles?
Can this overnight range loss be reduced, and how?
References
- Plett, G.L. — Battery Management Systems Volume I: Battery Modeling, Artech House, 2015
- Waag, W., Käbitz, S. and Sauer, D.U. — Experimental investigation of the lithium-ion battery impedance characteristic at various conditions and aging states, Applied Energy, 102, 2013
- Huria, T. et al. — High fidelity electrical model with thermal dependence for characterization and simulation of high power lithium battery cells, IEEE VPPC, 2012
- Dubarry, M. et al. — Best practices for incremental capacity analysis, Frontiers in Energy Research, 2020