Three weeks back I was in a private gallery off Chapel Street in Prahran, checking a client’s storage room before a revaluation. Standard smoke alarm on the ceiling, ten years old, battery half flat. The owner told me it was “basically the same thing” as the aspirating system I’d quoted for the vault next door. It is not the same thing. Not close.
i get this question a lot, usually after someone’s read a spec sheet and assumed all smoke detection does the same job at different price points. It doesn’t. A standard smoke alarm and an early-warning aspirating system solve different problems, and if you’ve got anything irreplaceable in a room, the difference is the whole point.
What a standard smoke alarm is actually built to do #
Your bog-standard photoelectric or ionisation smoke alarm, the kind bolted to a hallway ceiling in every house in the country, is designed to detect visible smoke once it’s already accumulated enough to trigger the sensor. That’s fine for a house. The goal there is simple: wake people up before a fire becomes unsurvivable. It’s a life-safety device, not a loss-prevention one.
The problem is timing. By the time a point detector like this trips, you’ve usually got visible smoke and often some heat. For a canvas, a lacquered frame, works on paper, or anything in a climate-controlled storage rack, that’s already too late. Paper scorches and discolours from heat and smoke exposure long before flame reaches it. Varnish blisters. Silver tarnishes from sulphur compounds in smoke within minutes. A standard alarm buys you seconds to evacuate people. It doesn’t buy the artwork anything.
Early-warning detection works on a completely different principle #
Aspirating smoke detection, the VESDA-style systems most fire engineers will know by name even though it’s technically a category not a brand, actively draws air through a network of small sampling pipes into a central detection chamber. That chamber uses laser-based particle counting sensitive enough to pick up combustion particles at concentrations thousands of times lower than what a point detector needs to trigger.
In practice that means these systems can flag a problem at the pre-combustion stage, when something is overheating or smouldering but hasn’t produced visible smoke yet. I’ve had one go off in a client’s server-adjacent storage room in Richmond because a power supply unit was cooking itself quietly for two days before it would have shown any smoke a normal alarm could catch. Nobody smelled anything. The system did.
The standard that actually governs this in Australia #
Fire detection design in commercial and institutional settings in Australia sits under AS 1670.1, which covers detection, warning and intercom systems. It doesn’t mandate aspirating detection specifically for art storage, that’s a client decision driven by risk appetite and what’s being protected, but it does set the framework for how any detection system needs to be designed, installed and maintained if you want it to count for anything on paper. If a fire engineer or insurer ever asks whether your system is compliant, this is the standard they mean.
Museums and public collecting institutions have their own layer on top of this. I’ve covered what institutions are actually required to do around museum security obligations elsewhere, and fire detection sits inside that same compliance conversation, not separate from it.
Response time, in actual numbers not marketing language #
A conventional point smoke alarm typically needs smoke obscuration levels of somewhere around 4 to 8 percent per metre before it triggers. An aspirating system can be configured to alert at obscuration levels of a fraction of a percent, some as low as 0.005 percent, well before anything is visible to the eye. In real terms that’s the difference between catching an issue while it’s still an electrical fault you can switch off at the wall, versus catching it once there’s an actual fire load in the room.
I’ll be blunt about my own view here: for anything stored off-display, in a vault, a plan chest room, or a climate-controlled rack facility, a standard smoke alarm on its own is not adequate protection for the value sitting in that room. It’s adequate for the humans if anyone’s in there, which is rarely the case in a storage vault at 2am. That’s a mildly unpopular thing to say to a client who’s just spent a fortune on a system they think is fine, but it’s true and I’d rather say it before the loss than after it.
Where the money actually needs to go #
Not every room needs aspirating detection. That’s the other half of this conversation people skip. If you’ve got a gallery floor with staff present during opening hours and good general ventilation, a well-maintained conventional system paired with proper monitored alarm response is often a reasonable call, particularly where the room is inspected daily anyway.
Where I push clients toward early-warning is anywhere unattended for long stretches: off-site storage, vault antechambers, roof cavities above hanging spaces, and anywhere with electrical plant nearby, HVAC units, pumps, lighting transformers. Those are the spaces where a slow smoulder can run for hours before a human notices, and they’re exactly the spaces that tend to hold the highest-value pieces because nobody wants a Sidney Nolan hanging next to a server rack.
This is the same logic that applies to vault design more broadly, you’re building for the scenario where nobody’s watching, not the scenario where someone’s standing right there.
What this has to do with your insurance, and what it doesn’t #
I’ll say this plainly because I get asked constantly and I want to be straight about where my job ends. We don’t sell insurance, arrange it, or advise on policy wording, that’s not what a locksmith or security fitter does and anyone in this trade telling you otherwise should be treated carefully. What I can tell you from twenty-odd years doing this is that insurers increasingly ask about detection specifications during underwriting, particularly for high-value collections, and it’s worth having your broker involved in that conversation early rather than after a claim.
What I won’t tell you, because it isn’t true, is that missing a security or detection condition means an automatic denied claim. Under the Insurance Contracts Act, an insurer generally has to show the failure actually caused or contributed to the specific loss, not just that a box wasn’t ticked. That’s a legal question for your broker or solicitor, not for me. If you want a starting point on that side of things, I’d point you to finding a fine art insurance broker before you need one, and to what actually happens if you let a policy or valuation lapse without reviewing it, which we’ve written about separately.
A fire nobody was watching for #
The 2018 fire at Brazil’s National Museum in Rio de Janeiro destroyed roughly 90 percent of its 20 million-item collection, and investigations pointed to inadequate fire prevention infrastructure, including detection, as a central factor. It’s an extreme case and Australian institutions operate under a different regulatory and funding environment entirely, but it’s the example I bring up when a client asks why any of this matters beyond ticking a compliance box. Detection systems don’t get credit when they work quietly in the background for fifteen years. They get noticed the one time they don’t exist.
If you’re not sure which category your storage rooms fall into, get someone out to actually look at the room rather than guessing off a floor plan. We run free security assessments for exactly this reason, and half the time the answer isn’t “replace everything”, it’s “upgrade this one room and leave the rest alone.” That’s usually the honest answer, and it’s cheaper too.