Home, heating & renovation

Product overview

Welcome to New Zealand’s trusted, independent source of practical consumer information. Join us now to access all our information and Consumer advisers when you need them.  

Smoke alarms and batteries

14jun smoke alarms and batteries hero default

Which alarms and batteries rated best?

Smoke is toxic – and breathing it can kill. So you need an alarm that gives you early warning and more time to escape. We tested photoelectric, dual-sensor and ionisation smoke alarms. We also tested 9-volt batteries, specifically for their performance in smoke alarms.

From our test


Choose what’s right for you with confidence

Join today and get instant access to all test results and research.

About our tests

We tested 20 smoke alarms and 19 9-volt batteries.

Smoke alarms test

Our test was conducted in the fire laboratory of the Building Research Association (BRANZ). We purchased 3 examples of each model. Our results are the average of the performances of the 3 examples.

The alarms were fixed to a ceiling panel in a mocked-up hallway adjacent to the door opening of a fireproof room (which was about the size of a single-car garage).

  • The flaming fire test: 3 pieces of 40mm-square 300mm-long sticks were set alight using a small quantity of methylated spirits. The resulting smoke built up in the room and flowed into the mock hallway.
  • The smouldering fire test: A 60W electric soldering iron was pre-heated and laid horizontally on to a 1.2 kg piece of 70mm-thick polyurethane upholstery foam. An electric fan was used to evenly distribute the resulting smoke out into the mock hallway.

In both tests, the smoke density was measured at the location of the alarms using a laser light-source and detector.

Smoke alarm batteries test

The smoke alarm batteries test was specifically designed to obtain information about the batteries’ performance when powering a smoke alarm. The results aren’t an indication of their performance in other devices.

Smoke-alarm batteries have a challenging task to perform. They sit in the alarm over a long time – and can be called on to power the siren when they’re likely to be in a semi-discharged state. What’s more, it could be a cold night when they’re needed (batteries don’t function as well in the cold as they do when warm).

There were 4 steps in our testing:

  • To tailor our battery tests to smoke alarms, we took 3 different alarms and measured their electrical characteristics both on standby and when the alarms were sounding.
  • We measured the capacity (stored electrical energy) of each battery by monitoring the time it took when we deliberately discharged each from its initial 9 volts to 5.
  • We took another set of our test batteries and half-discharged them using the data from step 2.
  • These half-discharged batteries were refrigerated to 6°C and then set up to supply the electric current equivalent to a sounding smoke alarm. We measured how well they performed this task.

Which type of smoke alarm?

It could be a deadly mistake thinking you’re protected by the most common type of smoke alarm.

Smoke is not just smoke. Smoke from flaming wood or cooking-oil fires is different from that produced by the cooler smouldering of upholstery foam, bedding or the plastic components in electrical equipment.

Ionisation alarms sense the volatile combustion products from hot flaming fires. But they don’t do as good a job sensing the smoke from smouldering fires because fewer combustion products are present.

Photoelectric alarms shine a light beam across a chamber and detect if the air in the chamber becomes partially obscured. So they can detect smoke from both types of fires.

Our test findings

In 2006 and 2010 we tested how well photoelectric and ionisation alarms reacted to smoke from a flaming wood fire or a smouldering fire. The two alarms operate differently in detecting the presence of smoke (see above). We found that photoelectric smoke alarms gave you significantly more protection than the more common ionisation models – particularly for smoke from non-flaming smouldering fires.

Our new test reinforces our earlier findings. This time we tested 9 photoelectric, 3 dual-sensor and 8 ionisation models – 20 in total. We tested them against a flaming wood fire and smouldering upholstery foam, both with steadily increasing smoke levels.

The dual-sensor models and all the photoelectric models gave the best protection from both types of fire.

The ionisation models were slightly better overall at detecting flaming fires – but they were hopeless for smouldering fires. None of them sounded at all during our smouldering fire test-runs. That failing is potentially fatal.

Using dual-sensor alarms is the deluxe solution. But they’re not that much of an advantage over our recommended photoelectric models.

The key to your protection is to have several good photoelectric alarms in working order. See 'Code requirements' below for advice on where to place the alarms.

Which type have you got?

Ionisation models require a tiny amount of radioactive material to make them work. So all of them will have the radiation symbol somewhere on the plastic body.

Tip: If your house has ionisation alarms, supplement (or replace) them with our recommended photoelectric models.

Code requirements

Along with a raft of other changes as a result of the leaky building crisis, the building code was changed in 2003 to require smoke alarms to be fitted to all residential buildings.

This doesn't mean that a council inspector will come knocking on your door to check if you have fitted the alarms. But it does mean that a code compliance certificate won't be issued - for any building work that requires a permit - until the alarms are fitted.

The rules

According to rules set down by the Building Code, smoke alarms must be "type 1" units. This means they must have a hush button - to allow nuisance alarms to be cleared without removing the battery - and also a test button.

As well, they must comply with at least one of the following standards:

  • UL 217
  • CAN/ULC S531
  • AS 3786
  • BS 5446:Part 1

Where to fit alarms

According to the rules, smoke alarms must be located in escape routes on all levels within the household unit. On levels containing sleeping spaces, they must be installed within the sleeping space, or within 3 metres of every sleeping space door. And they must be audible through closed doors.

Alarms must be installed on or near the ceiling, and in accordance with manufacturer's instructions.

You will need several alarms for large houses or multi-level houses. If you buy the inter-connectable type and wire them up, all will go off together. Tests have shown an upstairs alarm in the stairwell is likely to respond before one fitted downstairs, even when the fire is downstairs.

The NZ Fire Service also has advice on where to place alarms.

Smoke alarms for the hearing-impaired

The New Zealand Fire Service says the ideal alarm for the deaf and hearing-impaired is a hard-wired series of interconnected photoelectric smoke alarms fitted with bed-shakers, strobe lights, or pagers.

These systems can be expensive, but if you meet certain criteria you may qualify for funding for a system to be installed by the Ministry of Health or Housing New Zealand. Contact Deaf Aotearoa to learn more.

Alternatively, the Fire Service provides a list of smoke alarm equipment available for deaf or hearing impaired people.