A common question in data center operations: after several years of service, a routine inspection finds one swollen cell in a UPS battery string. Its voltage is low, but the other cells look normal. Can you replace just the faulty unit, or do you need to replace the entire string?
The answer is almost always: replace the whole string.
Why a Single Failed Cell Often Dooms the Entire String
Lead-acid battery strings are connected in series. In a series circuit, current is the same through every cell, but voltage, internal resistance, and capacity are never perfectly matched across cells. That is the root of the problem.
Internal Resistance Difference
A new, small-capacity 12V lead-acid battery typically has an internal resistance between 0.015 and 0.018Ω. A battery that has been in service for three to five years can see its internal resistance rise above 0.05 Ω – often much higher.
When old and new cells are charged together in series, the charging voltage distributes according to internal resistance. The old cell, with higher resistance, receives a higher voltage. As a result, the new cell may not be fully charged while the old cell is already overcharged – overheating, losing water, swelling, and accelerating its own death. During discharge, the situation reverses: the old cell’s higher internal resistance causes its voltage to drop faster. The entire string shuts down early because of the old cell’s low voltage, leaving the new cell’s capacity completely unused.
Voltage Mismatch
Standards for battery strings specify maximum allowed differences in open‑circuit voltage between cells. The voltage difference between an old cell and a new cell is far beyond that limit. From a compliance perspective, mixing old and new cells is simply not recommended.
Capacity Degradation
The usable capacity of a battery string is determined by its weakest cell. After three to five years, an old cell’s capacity may have fallen to 70–80% of its rated value, while a new cell is at 100%. During discharge, the old cell empties first, triggering string protection. The new cell still has energy to give, but it cannot deliver it because the string has already shut down.
For all these reasons, industry practice is clear: if a lead‑acid battery string has been in service for more than three years and one cell fails, replace the whole string. Replacing just one cell may cost only a fraction of a full replacement, but the resulting risks – more failures, higher maintenance costs, and potential downtime – far outweigh the short‑term savings.
How Is a Nickel‑Zinc Battery Different?
What if the data center uses a Nickel‑Zinc battery instead? Does the same problem apply?
The first key difference is service life.
A Gerchamp Nickel‑Zinc battery （8XNFZ38） has a design life of 10–15 years, compared with only 3–5 years for conventional lead‑acid batteries.
A longer service life means more stable performance over time. After five or six years of operation, a Nickel‑Zinc string is still in the first half or middle of its lifecycle. Cell‑to‑cell consistency in internal resistance, voltage, and capacity is far better than that of lead‑acid batteries of the same age. The probability of a single cell failing badly is much lower.
Second, the failure mode is different. As lead‑acid batteries age, internal resistance rises sharply and capacity drops suddenly. When one cell fails, the whole string collapses soon after. In contrast, a Nickel‑Zinc battery near the end of its life mainly shows a gradual increase in internal resistance. It does not suddenly open‑circuit or short‑circuit, nor does it swell, leak acid, or release corrosive gases like lead‑acid batteries. Even if one cell’s performance degrades, the rest of the string can continue working – only total capacity is reduced. Maintenance teams have ample time to schedule a planned replacement instead of being called in for an emergency repair in the middle of the night.
Of course, mixing old and new cells is not recommended for any battery chemistry, Nickel‑Zinc included. But the key point is: Nickel‑Zinc batteries maintain much better long‑term consistency. They do not develop the large differences seen in lead‑acid strings after three or four years. This means the “one bad cell” scenario occurs much later – and much less often – in the lifecycle of a Nickel‑Zinc UPS battery.
Practical Takeaways for Data Center Operators
So, back to the original question: one bad UPS battery cell – what should you do?
If you are using lead‑acid batteries and the string is more than three years old: replace the whole string. This is not about driving up costs; it is a rational choice based on real‑world experience. The few hundred dollars you save by replacing only one cell will likely be paid back many times over in higher operational costs – or worse, a preventable outage.
If you are evaluating a different battery technology for your data center or AIDC facility: consider Nickel‑Zinc. Its design life is two to three times longer than lead‑acid. Over the life of the data center, the number of battery replacements drops dramatically.
From a total cost of ownership perspective, a Nickel‑Zinc battery may have a higher upfront cost than lead‑acid, but it avoids two or three complete string replacements – eliminating not only material and labor costs but also the downtime coordination headaches. For any facility that values high availability and low maintenance burden, Nickel‑Zinc is an option worth a serious look.