Abstract: LiFePO4 home batteries degrade to roughly 70–80% of original capacity by end of warranty, but the loss is not linear — the first year drops about 2–4%, then the curve flattens. The rule buyers and installers miss: size for the end-of-warranty capacity , not the initial nameplate. If a home needs 5 kWh per night and the battery will be at 70% by year 10, the install needs at least 7.2 kWh of initial usable capacity. Most customers who size for today's load and ignore degradation are fine for 5–7 years, then quietly lose their full-night coverage.
What degradation actually looks like
LFP cells fade faster in the first year, then settle. A typical calendar-plus-cycling curve for a quality 16S pack:
| Year | Remaining capacity (relative to initial) |
|---|---|
| 1 | 96–98% |
| 3 | 91–94% |
| 5 | 86–90% |
| 8 | 78–84% |
| 10 (EOL) | 70–80% |
These numbers assume 25°C ambient, one full cycle per day, and 90% DoD. The warranty EOL — the point where the manufacturer guarantees capacity stops — is usually 70% (some brands use 80%). That is the number to design against.
The sizing rule: work backward from end-of-warranty
Do not ask "how much capacity do I need today?" Ask "how much will be left in year 10, and is that still enough?"
Worked example:
- Home's maximum overnight need: 5 kWh (winter, heating, family of four)
- Warranty EOL factor: 70% (0.70)
- Required initial usable capacity: 5 ÷ 0.70 = 7.14 kWh
- Usable is 90% of nameplate → required nameplate: 7.14 ÷ 0.90 = 7.94 kWh
So this home needs an 8 kWh-class battery (a 10 kWh unit with 90% DoD), not the 5 kWh unit the nameplate math suggests.
Check the same home with a 5 kWh install:
| Year | Capacity left | Still covers 5 kWh? |
|---|---|---|
| 1 | 4.7 kWh | No, short from day one |
| 5 | 4.3 kWh | No |
| 10 | 3.5 kWh | No |
A 5 kWh battery never covers a 5 kWh night once usable capacity (90% DoD) is applied. That is why sizing must start from the end-of-warranty usable number , not the nameplate.
When you do NOT need degradation margin
The extra-capacity rule is not universal. Skip the degradation allowance when:
- The battery already has 2× margin. A 10 kWh battery for a 3 kWh/night home: even at 70% EOL it delivers 6.3 kWh — still double the need. No allowance needed.
- The customer will expand within 5 years. Stackable systems (like the HZEB-LCT stackable range) allow adding modules later; buying the extra module now just wastes cycles and capital.
- The tariff is flat. If self-consumption savings are small, the cost of unused capacity never pays back. Size to the load, accept the degradation.
When you DO need the margin
Add the allowance in four situations:
- The load sits close to the usable capacity. Within 20% of the usable number → size up one step (e.g., 5 → 10 kWh).
- Hot climates. Ambient temperature above 30°C accelerates calendar aging by roughly 1.5–2×. Design against 65–70% EOL instead of 70–80%.
- Deep daily cycling. Customers who run 90% DoD every single day age the pack faster than the datasheet curve. Use the 70% factor, not 80%.
- The buyer wants 10+ years without touching the system. Contractual "no expansion" requirement → build in the year-12 margin now.
The cost of over-sizing
Over-sizing is not free. A 10 kWh battery costs roughly 30–40% more than 5 kWh, and the extra capacity cycles shallowly, which actually extends cycle life — but the capital sits idle for years. The decision is a trade-off between:
- Buy margin now → higher first cost, no rework later
- Buy the minimum + expand later → lower first cost, but module availability and compatibility must still exist in 5 years
For most homes, the middle path wins: one size step above the nameplate math (5 → 10, 10 → 16), which covers degradation without paying for a second cabinet.
What to check on the datasheet before ordering
When comparing quotes, look for three numbers that determine the real degradation story:
- EOL definition — does the warranty guarantee 70% or 80%? A 70% EOL pack needs more initial margin.
- Cycle life at the DoD you will run — "6,000 cycles @ 90% DoD" vs "6,000 @ 80%" changes the year-8 capacity materially.
- Operating temperature range and derating — a pack rated to 45°C without derating ages faster than one with active thermal management.
A note on the "measured vs published" rule
The curves above are manufacturer published data under lab conditions. Real sites run warmer, cycle irregularly, and drift from the datasheet. Treat the numbers as design guidance, not guarantees — and if the supplier publishes both a nameplate and a tested degradation curve, ask for the tested one.
Ruibit sizes Dawnice HZEB-LCT systems by working backward from the customer's end-of-warranty need: we take the largest overnight load, divide by the EOL factor and DoD, and quote the next standard size up. Send us your winter load and climate, and we return the size that still covers the night in year 10.
FAQs
Q: How much capacity does a home battery lose per year? LFP batteries lose roughly 2–4% in the first year and 0.5–1% per year after, in normal 25°C conditions. Hot climates and deep daily cycling accelerate this.
Q: Should I buy a bigger battery just for degradation? Only if your load is close to the usable capacity (within ~20%) or you need 10+ years without expansion. Otherwise the extra capital sits idle.
Q: What does 70% EOL mean on a battery warranty? It means the manufacturer guarantees the battery retains at least 70% of original capacity at the warranty's end. Below that, a capacity claim applies.
Q: Does a shallower daily cycle slow degradation? Yes. Running 50–70% DoD instead of 90% roughly doubles cycle life. Over-sizing the battery creates shallower cycles, which is why "too big" batteries often age slower.
Q: Can I add battery modules later instead of buying margin now? Yes, if the system is stackable and the modules remain available. This is the lower-first-cost path, but confirm module compatibility and availability with the supplier before relying on it.
