← Back to blog

When to replace a solar battery (and when you really don't have to)

"My solar battery feels weaker than before. Should I replace it?" — this is one of the most frequent questions we get from owners of residential storage systems, especially those installed before 2022. The honest answer, in nine cases out of ten, is no. What feels like a dying battery is usually a configuration drift, a sensor miscalibration, or simply a change in consumption patterns. The actual cells inside are typically far from end-of-life.

This article walks through the real signals that a solar battery does need replacing, the lifespan you should expect from modern LiFePO4 chemistry (the standard for residential storage in 2026), and how to use data — not gut feeling — to decide whether to keep, recalibrate, or genuinely swap your battery.

The "weak battery" illusion

Most modern solar batteries report their State of Charge (SoC) based on a model, not a direct measurement. The model integrates current over time and applies corrections based on voltage. Over months and years, small errors accumulate. Result: the battery seems to discharge faster, or to hit "low" sooner than expected, even when the actual capacity is unchanged.

Symptoms commonly attributed to "the battery dying" that are actually not end of life:

  • SoC seems to drop quickly from 80% to 20% — usually an SoC reset issue, not capacity loss
  • The system trips into low-battery shutdown more often — almost always a voltage threshold misconfiguration after a firmware update
  • You feel you need to charge it more often — increased household consumption (new heat pump, electric car) is the cause 80% of the time
  • Cycle counter shows a high number — most cycle counters in residential systems are derived metrics with very loose definitions, often inflated 3-5× compared to real equivalent full cycles

The first step before considering replacement is always verification with real data: how much energy actually went in and out over the last 30 days, and how does that compare to the rated capacity?

What replacement really means with LiFePO4

If your battery is LiFePO4 (lithium iron phosphate) — the dominant chemistry in residential storage since around 2020 — the replacement conversation looks completely different from older lead-acid or NMC systems.

Modern LFP cells from grade-A manufacturers are rated for 6,000 to 9,000 equivalent full cycles at 80% depth of discharge before reaching 80% State of Health (SoH). At one cycle per day, that's 16 to 25 years of daily operation before performance even starts to degrade noticeably. At 0.7 cycles per day (typical residential pattern), it stretches to 23-35 years.

In practice, this means that a LFP battery installed in 2022 is statistically far from needing replacement in 2026. Most are at 95-98% of their original capacity, and the real limits are not cell degradation but:

  • BMS firmware: the management software may need updates, not the cells
  • Connections: oxidized terminals, loose busbars and degraded cables explain many "lost capacity" cases
  • Imbalance: when one cell drifts in voltage relative to the others, the BMS limits the whole pack — but rebalancing recovers most of the apparent loss

None of these require buying a new battery. They require data, diagnosis, and in the worst case a service intervention on one specific component.

The signs that genuinely point to replacement

So when does a solar battery actually need replacing? Four real signals:

1. Measured capacity below 70% of nameplate. Not estimated, not derived from SoC behaviour. Measured by a controlled discharge from full to cutoff, tracked over multiple sessions. If your battery rated 15 kWh consistently delivers less than 10.5 kWh from full to empty, the cells are tired. This is the gold-standard test.

2. Cell-level imbalance that doesn't recover. A healthy LFP pack keeps cells within 20-50 mV of each other when at rest. When one cell systematically drifts to 200-300 mV from the others and active balancing can't bring it back over weeks of operation, that specific cell is at end of life — and if your battery is a fixed-cell topology, often the whole module needs replacement.

3. Recurring overvoltage or undervoltage events. The BMS logs these. If they appear with increasing frequency despite no change in usage, the cells are no longer behaving uniformly under load.

4. Visible damage or thermal anomalies. Swollen cells, persistent temperature differentials greater than 10 °C between cells under similar load, or any sign of casing deformation are clear red flags.

None of these can be diagnosed without proper telemetry. That's the catch: most batteries on the market today have a BMS that knows all this, but none of the data is recorded over time, so when the question comes up, the owner has no history to look at.

Why this matters for buyers in 2026

The European Battery Regulation 2023/1542 will require all stationary batteries above 2 kWh to carry a digital passport from February 2027, including continuously updated SoH and life expectancy data. This is not optional, and it changes how the replacement conversation will work in coming years.

Buyers will be able to verify battery health rather than trust a manufacturer's claim. Resale and second-life markets will function on this data. And replacement decisions will move from "I feel it's weak" to "the passport says SoH is 78% — let's plan a swap in 18 months." We've written about what this means in detail in our blog — and we've been building our systems to this standard for years, before the regulation existed.

If you're buying a battery today, the most important question is no longer "how big" or "how much" — it's "will it be able to tell me how it's doing in five years?". A battery that can't answer that question won't be repairable, won't be resaleable, and won't have a defendable second life. A battery that can answer it — like the systems we manufacture — protects your investment for the full 20-25 year horizon that modern LFP actually delivers.

The short answer

If you're asking "should I replace my solar battery" in 2026 and it's LFP installed after 2020: almost certainly not. Get someone to measure real capacity, check cell balance, and update the firmware. The cells are very probably fine.

If you're buying a new system: don't think in terms of "when will I replace it". Think in terms of "what will I know about it in 10 years". The answer to that question is the actual quality of the product you're buying.

Looking at a new solar storage system, or wondering whether your existing battery actually needs replacing? Talk to us — we'll help you make the call based on data, not pressure.