Why the Battery Fails First in Solar Street Lights

The arithmetic behind "it worked for months"

Take the classic configuration from the field: a 200 W panel, a 60 W LED load, and a 120 Ah 12 V battery. Run the numbers honestly.

A 60 W LED on a 12-hour night draws about 720 Wh . A 120 Ah bank at 12 V stores 1,440 Wh nominal , and if you cycle it to a realistic 80% depth of discharge (DoD) you have roughly 1,150 Wh of usable energy . That is 1.6 nights of full load with zero recharge. Three consecutive cloudy days — the autonomy most specs ask for — is out of reach from day one.

So why does the system look fine at commissioning? Because at commissioning the bank starts every night at 100% and the first cloudy stretch has not happened yet. The failure is designed in; it just takes one bad weather week to surface. When a degraded, partially charged bank discharges, its voltage sags fast and the controller's low voltage disconnect (LVD) trips — the light cuts out early or never starts. That is the "works in summer, dead in the rainy season" call we get.

Why we refuse second-life EV cells

Some quotes get cheap by using repurposed EV cells. They will pass an initial capacity test — capacity at delivery is not the problem. The problem is everything you cannot see: unknown cycle history, unknown fast-charge exposure, and cells that were pulled because they failed matching. In a street light they live at partial state of charge (SoC) for years, where an uneven pack hides its degradation until one cell drags the whole string through the BMS. Our rule for our own builds and our advice to every buyer: street-light banks come with batch capacity test reports and a 3–5 year performance warranty, or they do not go on the pole.

Diagnose before you replace

The diagnostic sequence we use, in order:

1. Rest, then measure open-circuit voltage (OCV). After several hours of rest, a 12 V bank reading well below nominal (for example, under 11 V) is a failed or failing battery. One caveat that matters more every year: on a lithium iron phosphate (LiFePO4) bank, OCV is a weak SoC indicator because the discharge plateau is flat. A lithium bank at 30% and at 70% can read nearly the same volts. For lithium, trust the BMS coulomb count or a load test; reserve the OCV shortcut for lead-acid.
2. Swap test with a known-good battery of equivalent specification. If performance returns to normal, the original bank is conclusively the fault. This is still the fastest field verdict there is.
3. Check the charging source under full sun: panel open-circuit voltage (Voc) and short-circuit current (Isc) against nameplate. Panel output typically fades around half a percent per year in normal conditions — a number well below nameplate means the array, not the bank, is the problem.
4. Check the controller settings against the chemistry. Bulk/absorption/float voltages and the LVD threshold set for the wrong chemistry quietly shorten life. Then walk the cable runs: undersized conductors and corroded connections show up as voltage drop that both wastes charge energy and wastes discharge energy on the same compromised bank.

The sizing rule we build to

• Size for the cloudy-day sequence, not the average night. Three consecutive cloudy days is the common spec; five for critical sites. Do the night-load arithmetic first, then pick the bank.
• Keep daily DoD shallow. Minimizing daily depth of discharge is the single most effective lever on cycle life — a bank cycled to 50% daily lasts dramatically longer than one driven to 90%, at the same total energy delivered.
• Choose the chemistry for the job. LiFePO4 cells from reputable brands typically deliver 2000+ cycles with far better thermal stability and thermal-runaway resistance than other lithium-ion variants. For transport they must carry UN 38.3 test evidence, and for stationary use IEC 62619 is the safety benchmark we ask suppliers to meet.

Read the warranty as an engineering spec

A 3–5 year performance warranty is the normal benchmark for a quality bank. Our recommendation: make the warranty state a capacity-retention figure at a defined year (we ask for at least 80% retention at year three, in writing). A warranty that covers "failure" but says nothing about capacity is covering the wrong risk — capacity is what actually fails.

What this page does not cover

This page covers battery health, sizing, and procurement for DC-coupled solar street lights. It does not cover lead-acid chemistry selection (see gel, AGM and flooded comparison), two-battery configurations (see dual-battery fault analysis), or temperature effects on the bank (see battery environment temperature). For a complete unit with the bank, controller, and panel matched at the factory, see the all-in-one solar street light line.