# Fuel consumption, fuel accounting & lift-and-coast

> Know fuel per lap, laps per tank and the lap-time cost of fuel mass; measure fuel by mass with temperature correction; when you must save, lift-and-coast at the end of fast straights is usually the cheapest technique.

Area: Engine & Hybrid · Starting level: L1 Foundation · Updated 2026-11-16
Source page: https://race-engineer-academy.vercel.app/lesson/engine-fuel-consumption-lift-coast

## L1 · Foundation

Three questions every race engineer must answer before the race (Segers §16.1):

1. **Fuel per lap** — in racing we think in litres (or kg) per lap, not L/100 km.
2. **Laps per tank** — sets stint length, pit-stop count and pit windows.
3. **Weight penalty** — how much each kg of fuel costs in lap time, and how the balance changes as the tank empties (CG height drops; if the tank is not central, weight distribution moves).

**Where the number comes from.** The ECU knows total injector opening time; the logger/dash multiplies it by an injector flow-rate scalar to give fuel used. This calculation must be **calibrated** against reality: weigh the fuel in and out (with temperature) and correct the **fuel-used calculation scalar** — not the ECU's fuelling injector characterisation, which sets pulse width and would change lambda. Segers' GT example: 107.1 L actually used, the logged fuel-used figure said 106.2 L (99.1 %) — the scalar is adjusted so that the dash fuel figure is trustworthy for the race.

**Why weigh, not measure volume?** Fuel density changes with temperature, roughly 0.1 % per °C. Fuel at 15 °C vs 40 °C differs by ≈ 2.5 % in volume for the same mass: on a 100 L tank that's 2.5 L — about one lap of a GT car.

**Typical numbers.**
| Car | Fuel per lap (≈ 4–5 km lap) |
|---|---|
| GT3 | 2.5–4 L |
| GT4 | 2.5–3.5 L |
| LMP2 | 2–2.7 L |
| F3 | 1.5–2 L |

**Lap-time cost of fuel**: a useful scaling is that lap time grows by roughly 0.1–0.3 × the relative mass increase. For a ≈ 1300 kg GT car on a 2-minute lap that is ≈ 0.01–0.03 s per kg per lap; for a ≈ 800 kg formula car on a 90 s lap ≈ 0.03 s/kg. Carrying 10 kg more fuel than needed therefore costs ≈ 0.1–0.3 s every lap — get your car's exact figure from the lap sim.

**Lift-and-coast (L&C)**: the driver lifts off the throttle some distance *before* the normal braking point, coasts, then brakes a bit later/softer. Fuel is saved during the coast (injectors cut on overrun) at the moment where the car is fastest, accelerates least and is about to brake anyway — so it costs very little lap time per litre saved.

## L2 · Practitioner

**Fuel accounting (calibrating the fuel-used calculation).** Procedure from Segers:
1. Weigh fuel added before and during the session and fuel drained after (scale accurate to ±10 g), and measure temperature each time.
2. Fuel mass used = in − out (mass is temperature-independent).
3. Convert to volume with the temperature-dependent density, compare with the logged fuel-used total.
4. Correct the **fuel-used calculation scalar** in the logger/dash (not the ECU fuelling injector characterisation, which would change lambda): $k_{new} = k_{old} \cdot V_{real}/V_{logged}$.

Remember in/out laps use less fuel; extract race-pace consumption per lap only once the total is validated.

**Fuel needed for a stint/race:**

$$V_{fuel} = N_{laps}\, \bar V_{lap} \cdot (1 + s) + V_{reserve}$$

with a safety margin $s$ (1–3 %), plus the regulation sample/reserve (e.g. ~1 L or a minimum mass must remain for scrutineering — check rules) and formation/safety-car laps (SC laps use ≈ 40–60 % of a race lap).

**Fuel-saving techniques, in order of time cost per litre (cheapest first, usually):**
1. **Lift-and-coast** before the big braking zones, especially from high speed.
2. **Leaner fuel map** (if permitted): λ 0.88 → 0.95 saves several % fuel with ~1 % power loss.
3. **Short-shifting** (upshift earlier): engine runs at lower rpm, lower friction/pumping; on a flat-topped (restricted) power curve almost free as long as post-shift rpm stays on the plateau.
4. **Higher gear in slow corners** — costly if traction-limited exits suffer.
5. **Lower apex speeds / earlier braking** — the most expensive, avoid.

**Lift-and-coast in practice.** Define it as *coast distance* (m) or *lift point marker* before each target braking zone. Measure saved fuel per lap with the ECU fuel channel; measure time lost per lap with a delta. Typical outcome: coasting 50–100 m before two or three big stops saves 2–5 % fuel at a cost of 0.1–0.3 s per lap.

**Data**: overlay throttle and speed; L&C appears as throttle = 0 with brake = 0 before the braking point. Check consistency lap to lap (the dash "fuel target / delta" helps the driver).

**Balance effect**: L&C reduces brake temperatures and moves weight transfer earlier — some drivers find entry balance changes; also tyre temperatures drop, which can matter in cold conditions.

## L3 · Advanced

**Why lift-and-coast is cheap.** Let the car be at speed $v$ near the end of a straight with full-throttle acceleration $a_{ft}$ (small near terminal speed) and coasting deceleration $a_c$ (drag + engine braking + rolling, large at high speed). Coasting for a short distance $\Delta x$ before the braking point:
- **Fuel saved** ≈ $\dot m_{f,ft}\, \Delta x / v$ (full-throttle fuel flow × time coasted, overrun fuel cut ≈ 0).
- **Time lost** ≈ the integral of the speed deficit, $\Delta t \approx \int \Delta v / v^2 \,dx$. The speed deficit grows with $(a_{ft} + a_c)$; but at the braking point the car is slower, so the subsequent braking distance shrinks — the driver brakes later, recovering part of the loss.

Because $\Delta t$ scales as $\Delta x^2$ for small coast distances (deficit grows linearly from zero) while fuel saved scales as $\Delta x$, **the first metres of coast are almost free** and the marginal cost rises with coast length. The optimum distributes coasting over several straights so the marginal cost $\partial t/\partial m_f$ is equal everywhere — the same principle as energy-per-stint optimisation.

Numbers (GT3, 400 kW, BSFC ≈ 260 g/kWh): full-throttle flow ≈ 104 kg/h ≈ 0.029 kg/s ≈ 0.039 L/s. At 250 km/h, coast deceleration ≈ 2–3 m/s². One second of coast (≈ 70 m) saves ≈ 0.04 L and costs on the order of 0.02–0.04 s — roughly **0.5–1 s per litre saved**. Short-shifting usually costs more per litre on non-restricted engines; leaner mapping can be cheaper where available.

**Fuel mass effect on lap time.** Sensitivity $\partial t_{lap}/\partial m$ comes from the sim or from two runs with known fuel difference (correcting for tyre state). It is not linear with lap length alone: tracks with many slow corners and long acceleration phases are more mass-sensitive. Use it to:
- pick qualifying fuel (minimum needed + margin),
- compute fuel-corrected lap times for degradation analysis: $t_{corr} = t - k_m (m_{fuel} - m_{ref})$.

**Density.** Gasoline thermal expansion coefficient ≈ 0.00095–0.0012 /K. With ρ15 ≈ 0.75 kg/L: $\rho(T) = \rho_{15}[1 - \beta(T - 15)]$. For energy-limited formulas, the sensor reads energy via fuel flow × calorific value; fuel batch LHV variations (≈ ±0.5 %) are part of the regulator's tolerance.

## Key formulas

**Laps per tank**

$$N = \left\lfloor \frac{V_{tank} - V_{reserve}}{\bar V_{lap}} \right\rfloor$$


**Temperature-dependent fuel density**

$$\rho(T) = \rho_{15}\,[1 - \beta (T - 15)], \quad \beta \approx 0.001\,\text{K}^{-1}$$


**Fuel-used scalar correction**

$$k_{new} = k_{old} \frac{V_{real}}{V_{logged}}$$


**Fuel-corrected lap time**

$$t_{corr} = t_{lap} - k_m (m_{fuel} - m_{ref})$$

- k_m: lap-time sensitivity to mass (s/kg)
- m_fuel: current fuel mass (kg)

## Trackside takeaways

- Measure fuel by mass and temperature; calibrate the logger/dash fuel-used scalar against the drain (never the fuelling injector data).
- Fuel volume changes ≈ 0.1 %/°C — 25 °C difference on 100 L is about a GT lap of fuel.
- Each extra kg costs ≈ 0.01–0.03 s/lap in a GT car (≈ 0.03 s in a light formula car); don’t carry more than the stint needs.
- Lift-and-coast at the end of fast straights is usually the cheapest saving; first metres are nearly free.
- Spread fuel saving over several straights — equal marginal cost everywhere.
- Safety-car laps use ~40–60 % of a race lap: include them in the plan.

## By car category

### GT3

Endurance: stint length set by tank (≈ 100–120 L) and/or BoP-defined refuel time or stint energy. Sprint: fuel load matters for qualifying and mandatory minimum fuel at the end.

### Hypercar (LMH/LMDh)

Energy per stint replaces "litres" as the binding constraint; lift-and-coast and hybrid regen are tuned to the energy target per lap shown on the dash.

### LMP2

Fuel-limited stints at Le Mans; lift-and-coast and fuel maps are routine to extend a stint by a lap.

### IndyCar

Fuel saving is a core strategy tool: teams target a fuel number per lap to make a pit window; drivers use lift-and-coast and mixture/map settings, with push-to-pass as a counter-weight.

### FIA F2

No refuelling: start fuel must cover the race distance; drivers may need lift-and-coast in feature races if the car is filled close to minimum to save weight.

### Super Formula

Fuel-flow-limited engine; fuel strategy interacts with the Overtake System time budget.

## References

- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §16.1 Fuel Consumption (fuel-used calculation, weighing with temperature correction, flow-rate scalar, laps per tank)
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §16.2.1 Fuel Load Variation (fuel mass effect on lap time)
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §4.5 The Importance of Corner Exiting Speed (speed deficit vs time along a straight)
