Boyd Had Been Feeding It Batteries Since June

I was at Boyd’s kitchen table Saturday when something chirped down the hall.

He didn’t flinch. Didn’t look up. Went right on talking about his son-in-law’s boat.

Thirty seconds later it did it again, and I asked him about it.

He said it’s been at that since June. He’s put three batteries in it. Says he buys the nine-volts at the hardware store two at a time now, and that the fellow at the counter laughed at him the last time.

So I fetched the stepladder off his porch.

Boyd heard a chirp and thought battery. Anybody would. That’s the only thing a chirp has ever meant to any of us.

But a chirp isn’t one message. It’s several, and the two you’re most likely to meet sound near enough alike from the hallway.

Kidde spells its own out. A low battery chirps every thirty to forty seconds, for at least seven days, and a new battery ends it.

The other chirps every thirty seconds and means the alarm is finished. Kidde’s words: it’s time to replace the alarm.

A battery won’t stop that one. It isn’t meant to. The alarm is trying to tell you it’s done, and the only way to shut it up is to buy a new one.

Now the honest part. Kidde puts that end-of-life chirp on its sealed-battery models, and Boyd’s takes a nine-volt. So I can’t sit here and tell you which of the two he’d been listening to since June.

That’s rather the point. You can’t diagnose a ceiling from a hallway, and you don’t have to.

Boyd spent four months and three batteries on a noise he never once went up and looked at.

Twist one off the ceiling and turn it over. There’s a date printed on the back.

Not the date you put it up. The date it was built.

The US Fire Administration is plain about what that date is for. Smoke alarms, they say, need to be replaced ten years from the manufacture date.

Ten years from when it was made. Not from when you bought it, and not from when the builder screwed it in.

Boyd’s said 2011. He bought that house in 2013 and never gave the ceiling a thought after the walkthrough.

Same agency, same page, two more lines worth having. One inside and outside every bedroom, and one on every level. Test them monthly.

Here’s something I had wrong through forty years of owning houses.

There are two kinds of sensor in these things, and they aren’t equally good at the same fire.

NIST lays out how each one works. The ionization kind runs a tiny current through the air inside the chamber, and it’s quicker on the small black soot particles that come off a flaming fire. Curtains going up. Grease catching.

The photoelectric kind watches a light beam, and NIST says it’s more sensitive to particles that are larger and pale. The kind a smoldering fire puts out.

Now think about which of those two happens at two in the morning.

A cigarette in a cushion. A cord under a rug. A mattress. Things that put out smoke for a good while before they ever show a flame.

You can find people online who will tell you the ionization kind is worthless and ought to be banned. That argument has been running for years, and I’m not qualified to settle it.

So I’ll leave that fight alone and give you what the people who study it recommend instead.

The NFPA advises having both kinds in the house. The Fire Administration says the same thing, and adds the easy version: a dual-sensor alarm, which has both in one housing.

So this isn’t a purity question. It’s a coverage question.

Those NFPA numbers are worth reading slowly. Forty-one percent of home fire deaths happened where there was no alarm. Another sixteen percent happened where an alarm was on the wall and didn’t go off.

Add those and you’re near three out of five. And the leading reason the quiet ones stayed quiet wasn’t the sensor type at all. It was a missing battery, a dead battery, or a disconnected one.

Which means the sensor argument is the interesting problem, and the battery is the one actually killing people.

An hour, a stepladder, and a pencil. Get somebody to hold the ladder.

▸ Take every one down and find the date on it. Write that year on the back in pencil while you’ve got it in your hand, so the next man doesn’t have to climb up to find out.
▸ Anything older than ten, replace it. Not clean it, not test it. The date is the date, and no amount of working right now changes it.
▸ Count your doors while you’re up there. The Fire Administration wants one inside and outside each bedroom, and one on every level. Every house I’ve counted in has come up short.
▸ Buy dual-sensor when you replace. That’s both technologies in one housing, and it ends the argument in section four without you having to have an opinion.
▸ Put testing on the same day as something else. Monthly is the recommendation. Nobody remembers monthly on its own. Hang it on a day you already have a habit, whatever that happens to be.

That’s the job. Most of it is ladder time.

While you’re up there, look at the carbon monoxide alarm too. That one is worse, and almost nobody knows it.

Kidde says to replace those every seven to ten years depending on the model, and the date is on the back the same way.

Then comes the line I had to read twice.

Seven years after it first gets power, their CO alarm starts chirping every thirty seconds, and the chirp won’t stop until the unit is switched off. In that condition, in their words: “the alarm will not respond to carbon monoxide.”

So it’s on the wall, and it’s plugged in, and it isn’t watching a thing.

The Fire Administration wants CO alarms in a central spot outside each sleeping area, and on every level. They put the toll from accidental nonfire carbon monoxide at more than 150 Americans a year.

You can’t smell it, and the early part of it feels like a headache and a bad night. That’s the entire reason the thing on the wall exists.

The one in his hallway was fifteen years old, and the last four months of it were spent chirping at a man who kept answering with a battery.

We found four more in that house. The newest was 2016. The one over the stairs had no battery in it at all, and neither of us could tell you when that happened or who did it.

He bought six that afternoon, one more than he took down, and I held the ladder.

It isn’t the dates that bother me. It’s that Boyd heard a warning every thirty seconds for four months, and it turned into furniture.

That’s not a Boyd problem. That’s how anybody handles a noise that won’t explain itself.

I’m doing my own house this weekend. I’d be surprised if I like what I find.

— Walter