How Long Does a Lithium Battery Last? A Procurement View on Solar Storage

Every week, someone asks me how long does a lithium battery last? It sounds like the practical entry point for solar storage. After six years of buying inverters and batteries for a 40-person solar installer, I think it's the wrong first question.

The better question is: when does the system stop being worth the money? That is not the same as the warranty period, the cycle count, or the chemistry spec.

What everyone asks about solar storage

Our company gets calls from commercial clients who read renewable energy storage news today and suddenly want a battery. The first question is always the same: how long does a lithium battery last? They want a number. I understand the impulse. But a number without operating conditions is marketing, not engineering.

I remember a quote from 2023 that said 'lifetime: 10 years' on a lithium iron phosphate battery. It sounded great. The fine print defined lifetime as 80% of initial capacity at 25 degrees Celsius and 0.5C charge/discharge. Our installs live in metal rooms where summer temperatures hit 35 degrees, and the charge current sometimes sits at 1C for an hour. The real lifetime was different. Worse than expected.

The battery itself wasn't the problem. The problem was the gap between the lab spec and the field.

The deep causes: Why battery life is not one number

Calendar aging is the real tax

Most buyers focus on cycle life. The question everyone asks is: how many cycles can it do before it degrades? The question they should ask is: what is the calendar aging rate at the state of charge where the system will sit?

A lithium battery that cycles once a day is idle for 23 hours. If it sits at high state of charge, especially above 90%, it ages even with no electrons moving. The National Renewable Energy Laboratory has published degradation research showing that calendar aging can dominate cycle aging in lightly cycled storage systems. That's counterintuitive. It also means a battery used every day may last longer in real years than one used twice a week and left full in between.

This is why 'how many cycles' is not the right filter. A 6,000-cycle battery could end up financially dead at year 8, not because of cycling, but because of heat and rest voltage.

Inverter integration is the ignored variable

Here is something vendors won't tell you: the same battery can behave differently depending on the inverter. Compatibility is more than a checkbox.

We have standardized on the SMA Solar SB7.7-1SP-US-40 for most of our single-phase residential and small commercial projects. Not because it's cheap. Because it's predictable. Its published datasheet, as of early 2025, includes the AC coupling details we need, and the inverter plays cleanly with our monitoring stack. But we didn't get there by reading a brochure. We got there after a third-party 'compatible' inverter decided to stop responding to the battery management system. The battery was fine. The communications stack was not.

That truck roll cost us a day and a half. The client did not care which component was at fault. They cared that the system didn't work. From a procurement view, the source of the failure is irrelevant. The cost is the cost.

The cost of using a spec sheet as a business plan

When I audited our 2023 spending, I noticed a pattern in storage service tickets. The battery chemistry was rarely the root cause. The root causes were firmware conflicts, wrong Modbus registers, missing temperature derate settings, and a wallbox that ignored the solar signal.

Let me give you a concrete example. We installed a 15.8 kWh battery that was supposed to last 10 years. Within seven months, it throttled every hot afternoon. That's exactly when the client wanted solar energy for their EV charger. We logged four service visits before we found the problem: the inverter's charge profile was fighting the battery's internal temperature limit. The total labor cost was $1,150. The replacement inverter cost another $1,400. The 'cheap' battery package had saved us $900 up front. We lost that and more.

What I mean by that is simple: the total cost of ownership is not the battery price. It is the price of the battery plus the cost of making it operate correctly in a specific building, with a specific inverter, under a specific load pattern, for the entire time it's installed.

I have mixed feelings about extended warranties. On one hand, they cover some hardware failures. On the other, most early failures are installation and integration issues that a warranty does not cover. The money is better spent on commissioning and on testing the communication chain from inverter to battery to wallbox.

What I actually recommend now

Stop asking how long a lithium battery lasts. Ask what degradation curve is assumed, what operating temperature is assumed, and what happens at 80% capacity. Then put every quote into a cost model that includes commissioning, a spare comms gateway, an inverter replacement at year 10 to 12, and at least one service visit. That model is the only answer that matters.

The battery should have a UL 9540A listing, and the inverter should be on the AHJ-approved list. If it's not, the project can stop before the first electron moves.

For charging, if you're evaluating a wallbox mit solar, the first question is not the charger's price. It is whether the wallbox can receive a real-time power setpoint from the inverter. If it can't, the 'solar' charging is a schedule, not a load-following system. Ours once pulled grid power at 2 AM and called it solar because the overnight rate was low. That's not solar charging. That's a billing accident.

For hardware, we now choose the lowest predictable total cost, not the lowest quote. The SMA Sunny Boy SB7.7-1SP-US-40 has become our default because it's boring in the right way. And SMA support is part of the calculation. When we needed the Modbus register map for a custom energy management project, SMA support answered with the exact document and a wiring note. That response has a dollar value.

The lithium battery itself can last a long time. The technology is good enough for real projects. But if you buy on a number written by a marketing department, you'll miss the hidden costs that decide whether storage ever pays for itself.

So the honest answer to 'how long does a lithium battery last?' is: longer than the cheap quote, shorter than the brochure, and only if the system around it is designed to let it live that long.