Car Battery Cold-Start Doctor

Use when the car cranks slow or won't start in cold weather (or before winter) — converts measured battery voltage + temperature into CCA health verdicts, predicts no-start temperature from battery age/capacity loss, separates battery vs starter vs alternator failure signatures with 3 no-tool tests, and gives jump-start plus cold-start protocol order.

Overview

Cold mornings kill batteries in a specific, predictable way: at −18 °C a lead-acid battery delivers ~40% of its rated cranking power while the engine simultaneously demands ~2.5× more current (thick oil). A battery that started the car fine all summer can cross the no-start line on the first hard freeze — not randomly, but at a temperature you can compute from resting voltage, age, and one cranking observation.

scripts/battery_doctor.py turns three numbers anyone can get (resting voltage, temperature, battery age in years) into a health verdict, a predicted no-start temperature, three no-tool tests that separate battery / starter / alternator failures, and the right cold-morning protocol — including when a jump is safe and when the battery is done regardless.

When to Use

• Car cranks slower than usual, especially mornings / after sitting.
• Won't crank at all: click / crank-no-start / nothing decision tree.
• Pre-winter check: will this battery survive the season?
• After a jump-start: is the battery recoverable or replace-now?
• Deciding battery tester readings at a shop (CCA measured vs rated).

Don't use for: EV traction batteries, hybrids' NiMH packs, AGM start-stop systems with intelligent battery sensors (different diagnostics), or 6 V classics — the model is 12 V lead-acid.

How It Works

1. Resting voltage → state of charge. 12.7+ V ≈ 100%, 12.4 ≈ 75%, 12.2 ≈ 50%, 12.0 ≈ 25%, <11.9 = discharged/possibly dead cell (a reading under ~10.5 V at rest = shorted cell, replace).
2. Age → capacity fade. Lead-acid loses ~5–8% of capacity per year average climate (faster with heat, deep discharges, short trips). Combined capacity × cold derate gives effective cold-cranking amps.
3. Cold derate physics. Available power falls roughly with temperature: 100% at 27 °C, ~65% at 0 °C, ~40% at −18 °C. Cranking demand rises ~2× by −18 °C (oil viscosity). The script finds the temperature where effective CCA < demand → your no-start line.
4. Failure signature triage. Three no-tool tests:
   • Headlights-bright-but-click → starter circuit/solenoid or cable.
   • Cranks-strong-but-no-start → fuel/ignition, NOT the battery.
   • Cranks-slow-then-dies → battery (or big parasitic drain). Plus dash voltmeter alternator check: 13.8–14.6 V running = charging.
5. Verdict ladder. OK / WATCH (this winter, likely) / REPLACE NOW (predicted no-start above your local design temperature, resting voltage <12.2 after a full charge, or dead-cell signature).
6. Protocols. Jump order (positive donor → positive dead → negative donor → negative ground on dead engine block), never let clamps touch, donor running; cold-start protocol (all loads off, crank 5 s bursts, 30 s rest, 3 attempts max); after-jump: 30+ min driving to put back real charge (idling barely charges).

Using the Script

bash

# morning no-crank: 12.1V rest, -5°C outside, 5-year-old battery, 550 CCA rated
python3 scripts/battery_doctor.py --resting-v 12.1 --temp-c -5 --age-years 5 --cca-rated 550

# pre-winter check on a healthy-reading battery
python3 scripts/battery_doctor.py --resting-v 12.55 --temp-c 10 --age-years 3 --cca-rated 600 \
  --design-low-c -25

# no voltmeter? age + temperature only (coarse mode)
python3 scripts/battery_doctor.py --age-years 6 --temp-c -12 --cca-rated 520 --no-meter

Outputs: SOC from resting voltage, effective CCA at current temperature, demand estimate for your engine class, the temperature where this battery stops starting the car, verdict (OK / WATCH / REPLACE NOW), the failure-signature triage guide, and jump/cold-start protocol reminders.

Common Pitfalls

1. Testing right after driving. Surface charge reads 13+ V for hours. True resting voltage: headlights on 30 s, off, wait 5 min, key off — or measure in the morning before anything.
2. Reading voltage while cranking is the real test. A healthy battery holds >10.5 V while cranking at −10 °C; below 9.5 V under load = insufficient. Resting voltage can look fine and still fail this (sulfation: voltage present, no current delivery).
3. Blaming the battery when it's the starter or cables. One click with bright headlights = starter circuit, not battery. The triage tests prevent a $180 battery that doesn't fix the problem.
4. Assuming a jump "fixed" it. A jump gets you going; it does not recharge a deeply discharged battery. 30+ minutes of driving or a proper charger — and if it deep-discharged twice, it's compromised.
5. Ignoring parasitic drain. Battery fine, alternator fine, but dies after sitting 3 days = something staying on (trunk light, dash cam, aftermarket alarm). Battery replacement won't fix it.
6. Buying CCA you don't need, or too few. Match the group size and meet/exceed OEM CCA. In cold climates +15–20% CCA headroom is cheap insurance; the script shows your real margin.
7. Cranking until it grinds. 5-second bursts, 30-second rests, three attempts. Long cranking overheats the starter and floods modern engines.

Verification Checklist

• Resting voltage taken correctly (no surface charge; morning, key off).
• Battery age from the date code stamped on the case (letter+digit, e.g. D6 = April 2016-style; not the purchase receipt).
• Temperature = actual local low, not the daily high.
• Failure signature triage done BEFORE buying anything.
• If voltage-under-crank available: >10.5 V pass, <9.5 V fail noted.
• After any jump: 30+ min drive or charger session; retest next morning.
• Verdict REPLACE NOW → replaced before the first hard freeze, not after.