Three rounds of 2026 data. Lap times, sector splits and tyre deg from FastF1. A battery simulation built from FIA regulation constraints. A pace-based prediction for Miami.
The 2026 rules are the biggest power unit change since 2014. Electric power now matches the combustion engine. The MGU-H is gone. And the team that best manages its battery across 57 laps has an advantage no aerodynamic tweak can close.
Everything here is from FastF1. SC, VSC and pit laps are filtered out before any analysis — only clean, representative laps make it in.
New to F1 data? These are the key terms used throughout this analysis.
The FIA caps energy deployment at 4 MJ per lap. This model runs the battery physics across a full race distance for three deployment strategies — showing how thermal limits become just as constraining as the energy budget itself.
The 2026 F1 Energy Store (the battery) is governed by one hard rule: you can only change the battery's charge level by 4 megajoules per lap, in either direction. That cap is set by the FIA in Article 5.2.6 of the 2026 power unit regulations.
This model simulates what happens to the battery over a full race distance under three strategies — aggressive (deploy as much as possible each lap), conservative (hold charge back), and adaptive (deploy proportionally based on current charge level). It tracks two things lap by lap: how much charge is left, and how hot the battery is getting.
Every time you charge or discharge a battery, it generates heat. The harder you push — deploying and recovering energy aggressively — the faster heat builds up. F1 Energy Stores have a warning threshold at 55°C and a hard limit at 60°C. Above that, the team has to back off deployment to protect the pack.
The thermal model here uses a standard engineering approach: heat input proportional to charge/discharge current squared (I²R losses), cooled by a dielectric cooling system. It's the same framework used to design battery thermal management in electric vehicles.
Batteries wear out. Every charge/discharge cycle degrades the cell chemistry slightly — reducing how much energy it can hold. The capacity fade chart models this across 20 race weekend cycles using the Wang 2011 empirical formula for lithium-ion degradation.
Two factors drive it: depth of discharge (DoD) — how much of the battery you're using each cycle — and temperature. High DoD at high temperature is the worst case. Low DoD at low temperature barely degrades the cell at all. The three curves show just how much the operating regime changes the degradation trajectory.
Under aggressive deployment, battery temperature approaches the 55°C warning threshold before the energy budget runs out. The constraint is heat, not charge.
Deploying proportionally to current charge level keeps temperature stable and maintains usable SoC throughout the race — the best compromise of the three strategies.
The difference between high and low stress scenarios compounds over race weekends. After 20 cycles, the high-stress cell has lost meaningfully more capacity — and the gap widens each weekend.
What's real: the 4 MJ SoC cap, 350 kW MGU-K limit, and 8.5 MJ/lap recovery target are all from the FIA 2026 technical regulations. The capacity fade formula is from published Li-ion research (Wang et al., 2011).
What's estimated: the thermal mass of the battery pack (~20 kg), the cooling conductance of the dielectric system (~150 W/K), and the starting SoC (3.2 MJ). These aren't public — actual F1 Energy Store specs are proprietary. The model demonstrates the methodology and relative behaviour of strategies, not absolute temperature values.
What's not modelled: lap-to-lap variation in recovery (braking zones vary by lap), MGU-K override activations, tyre-induced load changes on the rear axle affecting regen efficiency.
Ranked by recency-weighted pace across the first three rounds. No circuit adjustment — Miami hasn't been raced under 2026 rules yet, so any track-specific weighting would be made up. What you see is purely who has been fastest, weighted toward the most recent race.
Green = field best (0.000s Δ) · Red = furthest from best · Sorted by weighted pace