Fire Alarm Batteries: What Does Your Test Actually Prove?

An engineer opens a fire alarm panel, puts a meter across the batteries and gets a plausible voltage. The mains supply is removed, the panel stays on, and the service sheet gets a tick.

What has actually been established?

The batteries supplied the load that was connected, for the time they were observed. That is useful information. It does not tell us whether they can support the required standby period and then operate the alarm.

Battery testing gets confusing when several different checks are all recorded as the same thing: “batteries tested”. A voltage reading, a conductance result, a calculated ampere-hour reading and a discharge test provide different evidence. Knowing which one you have carried out matters as much as the number on the display.

This article concerns the valve-regulated lead-acid batteries commonly used in fire alarm panels and associated power supplies. Other battery technologies need their own procedures.

Charged, healthy and correctly sized are different things

There are three questions to answer.

1. Is the battery adequately charged? A serviceable battery can be partly discharged after a power cut or previous testing.
2. Is it still in suitable condition? Age and deterioration can reduce the capacity available even when the battery has been charged.
3. Is the installed capacity sufficient for this system? A brand-new battery can be too small for the load and required operating duration.

A useful maintenance assessment covers all three. Recharging does not reverse every form of deterioration, and fitting new batteries does not correct an undersized design.

What a voltage reading tells you

First establish the conditions in which you took it.

A reading with the charger connected is a charging-system measurement. An open-circuit reading is taken with the battery disconnected from both the charger and the load. A reading under load shows how the battery behaves while supplying current. Those values are not interchangeable.

Open-circuit voltage can help assess state of charge, but the manufacturer's resting conditions matter. For example, EnerSys specifies a resting period and temperature for its DataSafe HX state-of-charge assessment. It is not a reading taken immediately after pulling a charging lead.

Even a satisfactory rested voltage does not establish how many ampere-hours the battery will deliver. You have measured electrical potential, not run the battery through its required duty.

Keep the voltage reading. Just record whether it was on charge, off charge or under load, rather than leaving the next engineer to guess.

Removing the mains proves transfer, not endurance

A controlled mains-failure test helps establish that the power supply transfers to batteries and that the connected equipment continues operating. Depending on the test scope, it can also verify the relevant fault indications.

But consider a panel left on batteries for two minutes with the sounders disabled. That observation cannot demonstrate its performance through a prolonged outage followed by an alarm condition.

The load matters. So does the time.

If the alarm load was not applied, say so. If only the panel supply was checked, do not extend that result to separate power supplies elsewhere in the building.

These tests need planning with the person responsible for the premises, including arrangements for monitoring, connected equipment and any temporary loss of protection. Battery circuits can deliver substantial fault current even with the mains isolated. Follow the equipment procedure and use appropriate competence, tools and precautions.

What handheld testers actually measure

Portable testers are valuable because a lengthy discharge test is often impractical during a routine visit. However, different instruments work differently.

Conductance and impedance

These measurements provide evidence of battery condition. They are particularly useful when compared with an appropriate reference or previous results taken consistently.

The GS Yuasa GYT117, for example, reports conductance in siemens and a percentage of a selected reference. Its instructions explain how to establish that reference and account for temperature. A displayed percentage of reference is not automatically the percentage of standby time remaining.

Megger makes a similar distinction in its explanation of battery impedance baselines: impedance is useful condition information, but does not directly determine capacity.

A result becomes much less useful if nobody knows which reference was selected or whether the previous engineer used the same method.

Pulse testers displaying ampere-hours

Some instruments apply a short pulsed load and calculate an Ah result. ACT describes its CHROME tester as simulating a C20 discharge test in seconds. That gives a practical capacity assessment under the instrument's stated conditions; it is not a record of the battery physically supplying the installed system for 20 hours.

ACT's instructions also explain why clean terminal contact, sound test leads, temperature and state of charge matter. A poor connection can produce a low or erratic reading.

Use the correct instrument for the battery type and rating, follow its connection and isolation instructions, and maintain its accuracy as the manufacturer specifies. Do not put a tester intended for an individual 12 V battery across a complete 24 V pair.

The useful question is what the displayed number represents and how it supports the decision to retain, recharge, investigate or replace the battery.

A discharge test answers a more direct question

A controlled discharge test measures what the battery supplies over time to a defined endpoint. Its meaning depends on the starting charge, temperature, discharge rate and end voltage. Megger distinguishes this from condition testing in its battery testing guidance.

It still needs a proper test plan. Letting a live fire alarm run until it shuts down is not a sensible substitute for one. Protection during testing, stopping criteria, recharge and the return to service all need to be addressed.

A completed test also consumes stored energy. Reconnecting the mains starts recovery; it does not instantly restore the charge used.

Check	What it helps establish	What it does not establish on its own
Visual inspection	Physical defects, connection problems and evidence of deterioration	Available capacity
Voltage measurement	Voltage under the recorded charging or load conditions	Required operating duration
Brief operation on batteries	Transfer and operation at the tested load	Full standby and alarm endurance
Conductance or impedance	Condition relative to a suitable reference or trend	Installed-system runtime
Pulse test with an Ah result	Capacity assessment using the tester's method	Completion of the system's actual duty
Controlled discharge test	Delivered performance under defined test conditions	Suitability for a different load or untested conditions

The Ah label comes with conditions

A battery marked 7 Ah is not promising 7 A for one hour under all circumstances.

Manufacturers specify capacity at a particular discharge rate, temperature and endpoint. Yuasa's NP7-12 data, for example, lists 7 Ah at the 20-hour rate to 1.75 V per cell at 20°C, and 6.4 Ah at the 10-hour rate to 1.8 V per cell at 20°C. Both the duration and endpoint differ.

Those conditions matter when comparing a tester result with a label or selecting a battery for a particular duty. Use the relevant discharge data and design method rather than assuming the headline capacity is available at every load.

There is another basic point worth checking: two 12 V, 7 Ah batteries connected in series form a nominal 24 V, 7 Ah supply. They do not become a 24 V, 14 Ah supply. The voltage adds; the same current flows through both batteries.

Record the individual battery results as well as the pair voltage. A total reading alone can hide an imbalance between them.

Healthy batteries can still be too small

Imagine a hypothetical system with these requirements:

• standby current: 0.30 A;
• standby duration: 24 hours;
• total alarm current: 1.20 A;
• subsequent alarm duration: 30 minutes.

The basic charge demand is:

Standby: 0.30 A x 24 h = 7.20 Ah
Alarm:   1.20 A x 0.5 h = 0.60 Ah
Total:                    7.80 Ah

These durations are illustrative, not a universal requirement. This arithmetic is also not a complete battery-sizing calculation: the applicable design method, allowances, discharge performance, temperature and equipment limits still need to be applied.

It does show why a good test result on a 7 Ah pair cannot settle the question. The assumed duty already calls for more than its nominal rating before those other considerations are addressed.

Establish the required standby and alarm periods for the actual installation. For non-domestic systems, use the applicable recommendations in BS 5839-1:2025, the design information and the manufacturer's instructions.

If devices or auxiliary loads have been added, revisit the calculation. “We fitted the same size again” only describes the replacement.

Check the charger as well

Replacing batteries without investigating the charging arrangement can leave the underlying problem untouched.

Check the specified charging voltage, any temperature compensation, the permitted battery range and the current available for recharging while the supply serves its load. A larger battery is not automatically an acceptable fix just because it fits in the enclosure.

For example, C-TEC's BF562-1 instructions specify a supported battery range, charging performance and different output-current conditions. They also explain its battery-resistance monitoring. These are equipment-specific details, not settings to copy to another panel.

The absence of a battery fault means the supply has not detected a condition outside its monitoring criteria. It is not a certificate that the installed batteries will meet every required standby duty.

Age and environment belong in the assessment too. A printed design life is not a guaranteed replacement interval for every installation. Record the installation date, inspect for damage or swelling, and investigate abnormal heat, repeated discharge or charging problems. Follow the battery and system manufacturers' replacement guidance, including how a series pair should be replaced.

Write down enough to support the decision

“Batteries tested OK” leaves out almost everything another engineer needs.

A useful record identifies the panel or PSU, both batteries, their ratings and dates, the instrument and test method, the readings and the conditions. It should identify the acceptance criteria used and link to the current capacity calculation.

Here is a fictional example of an incomplete test recorded honestly:

Panel 1: two 12 V, 7 Ah batteries in series. Battery operation observed for two minutes at quiescent load. Alarm outputs were disabled under the agreed test arrangements. This check did not assess available Ah capacity or full standby duration. Capacity assessment and review of the load calculation remain outstanding and have been raised with the client. Mains restored and temporary disablements removed; final system status recorded in the visit report.

That is a limitation requiring follow-up, not a completed battery assessment. It is still far more useful than an unsupported pass.

On a completed visit, retain the actual individual results, the action taken and the restored condition. If a battery was replaced, preserve the failed result as part of the record. This is the same principle discussed in The Golden Thread in Fire Safety: keep enough evidence for the next person to understand what was found and what changed.

The purpose of testing is to support a decision about the standby supply. A sensible voltage, a green light on a tester and a panel that stays on briefly can all contribute evidence. None should be asked to prove something it never measured.

The manufacturer examples describe particular products. Their test limits, settings and procedures should not be transferred to other equipment without checking the relevant instructions.