Battery-Powered Sensors and the Maintenance They Create

Why this matters

Every battery sensor you install is a small maintenance commitment that nobody wrote down. The box says the battery lasts two years. In a heated hallway it might. In an unheated garage or a crawlspace in January, the same sensor can go quiet in its first winter, and a dead freeze sensor or leak sensor fails in the worst possible way: silently, until the night it was supposed to speak. The customer remembers who installed it.

This card explains why the battery life on the box rarely matches the field, which battery chemistries hold up where, how to read a battery percentage without being fooled, and how to turn all of that into a schedule a shop can sell and keep.

What the battery life on the box actually measures

A claimed battery life is a lab figure. It assumes three things, usually without saying so:

• Room temperature. Battery capacity falls as temperature drops, and how much depends on chemistry (below).
• A good signal. A sensor that reaches its hub on the first try transmits briefly. A sensor at the edge of range retries, and each retry is another radio burst out of the same battery. A sensor that needs three attempts per report spends roughly three times the transmit energy of one that does not. Signal at the sensor is as much a battery question as a connectivity one.
• A typical event rate. A door contact on a back door used a few times a day reports far less than the same sensor on the garage-to-house door used forty times a day. A motion sensor in a hallway fires constantly.

So read the claim as "under good conditions," and treat each of those three as a reason a particular sensor will fall short of it. A sensor that is cold, at the edge of range, AND busy can fall a long way short, because the three effects stack.

The chemistries, and where each one belongs

Three practical rules fall out of the table:

Use only the chemistry the device's manual lists. A sensor built for a 3 V coin cell cannot take anything else. A sensor that takes AA cells usually accepts either alkaline or lithium AA; check the manual, because a few specify one. Rechargeables are generally a poor choice unless the manufacturer lists them, because their 1.2 V reads as low to a device expecting 1.5 V.

In cold spaces, choose lithium AA over alkaline where the device allows it. The garage, the crawlspace, the attic, the unheated outbuilding, the pump house: these are the places freeze and leak sensors exist to watch, and they are the places alkaline cells do worst.

Coin cells have a cold problem capacity does not explain. A coin cell has high internal resistance, which rises further in the cold. When the radio transmits, it draws a short burst of current, the coin cell's voltage sags under it, and if it sags far enough the sensor's chip resets. The battery may still have most of its capacity. The symptom is a sensor that reports, drops off, reappears, and logs repeated restarts, often on cold nights only. Fresh coin cells reduce it; moving the sensor closer to its hub (fewer retries, fewer bursts) reduces it more.

Reading a battery percentage without being fooled

The percentage in the app is almost never measured capacity. It is the device's estimate from the battery's voltage, and that estimate is only as good as the voltage curve it assumes.

Lithium cells hold a nearly flat voltage for most of their life and then fall quickly. A sensor reading 80 percent on a lithium cell can be weeks from dead, because the curve gives little warning until the end. A device whose percentage estimate was written for alkaline cells, which decline more gradually, can be especially misleading on lithium.

Temperature moves the reading too. A cold battery shows a lower voltage than the same battery warm, so a garage sensor may read 40 percent on a January morning and 70 percent in the afternoon. A "low battery" alert that clears by itself is often that, and it is still a warning worth acting on, because it is telling you the cell cannot hold voltage on the coldest night.

The working rule: treat a percentage as a trend and an alert, not a fuel gauge. Act on the first low-battery alert from any sensor that protects against water or freezing, and do not plan replacements by waiting for 0 percent.

The hazard in the replacement itself

Coin and button cells are a child-safety hazard, and the act of replacing them puts loose cells on counters and floors in homes with small children. A swallowed coin cell can cause severe internal burns within about two hours. If a child may have swallowed one, the response is immediate: go to an emergency room, and call Poison Control at 1-800-222-1222 on the way. Do not wait for symptoms.

The preventive half is in your hands at every replacement:

• never set fresh or used coin cells down where a child can reach them, and do not leave the used ones behind in a trash can;
• tape the terminals of used lithium cells before they go in your bag, so they cannot short against each other;
• re-close the battery door fully, including its screw if it has one. The federal standard for consumer products containing button and coin cells (16 CFR Part 1263, from the law often called Reese's Law) requires those compartments to resist a child opening them, and a door you leave unscrewed defeats that.

Dispose of used lithium cells through a battery recycling route according to your state and local rules; many do not allow them in household or business trash.

Building the replacement into a maintenance plan

A shop that installs sensors and never schedules their batteries is outsourcing reliability to customer memory. The fix is a short, recurring service item.

Inventory every sensor at install. For each one: location, battery chemistry and count, date fitted. This belongs on the device record; see related: Leaving a Device Record the Next Tech Can Use. Without the inventory, a battery visit is a house search.

Replace on a schedule, not on failure, for protective sensors. A common starting point:

• sensors in conditioned space: replace at the first low-battery alert, or at the annual visit, whichever comes first;
• sensors in cold or damp unconditioned space: fit lithium AA where the device allows, and replace every sensor there at an annual visit timed before the first hard freeze;
• smoke and carbon monoxide alarms with sealed long-life batteries: these are not battery jobs; the whole unit is replaced by its date. Smoke alarm replacement intervals come from NFPA 72 in the edition your jurisdiction has adopted and from the manufacturer; CO alarm life comes from the manufacturer's stated end-of-life date.

Tune the schedule after a season of records. If the cold-space sensors still die before the annual visit, move that visit earlier or improve their signal.

Test after every replacement. A new battery in a sensor that has dropped its pairing is still a sensor that does not report. Trigger each one (wet a leak sensor's probes, open the door, warm or cool the temperature sensor as its manual describes) and confirm the alert arrives on the customer's phone, not just on the hub.

Price it in time. Opening, swapping, re-securing and testing typically runs 3 to 5 minutes per sensor once the inventory is accurate, more for sensors under appliances or in crawlspaces. Sold as part of a maintenance agreement, it is a visit with a fixed scope and a customer who never has to remember it.

Worked example: the nine-sensor house after its first winter

An install in October: a water shutoff with five leak sensors (two in the crawlspace, one at the water heater in the garage, one under the kitchen sink, one behind the washer), two door contacts (front door, and the garage-to-house door), and two temperature sensors watching for freeze (garage and crawlspace). Every AA device got alkaline cells from the box. The coin-cell door contacts got the cells they shipped with. The box claimed about two years for each sensor.

By late February, the crawlspace temperature sensor had stopped reporting, one crawlspace leak sensor logged "low battery," and the garage-to-house door contact had restarted repeatedly on cold nights. Three of nine sensors (a third) were in trouble in about the first five months, against a two-year claim.

Reading each against the three conditions:

• Crawlspace temperature sensor: cold, and at the far edge of the hub's range under a floor. Alkaline in cold, plus retries. Two conditions stacked.
• Crawlspace leak sensor: cold, closer to the hub. One condition.
• Garage-to-house door contact: a coin cell, a busy door (dozens of reports a day), and a garage that drops near freezing. The resets were the coin-cell sag under transmit bursts, not an empty cell; it still read 70 percent.

The interior sensors, all in heated space with good signal, still read above 85 percent.

The plan the shop put on the agreement:

1. All four AA sensors in cold space (two crawlspace leak, the crawlspace temperature sensor, and the garage temperature sensor) switched to lithium AA, after confirming the manuals allowed it. The garage water-heater leak sensor took lithium too, making five.
2. A powered repeater for the hub's radio placed near the crawlspace access, to cut the retries. The far sensor's reported signal went from the app's lowest band to its middle band.
3. The garage-to-house door contact, which cannot take anything but a coin cell, was swapped for a model that takes a CR123A, a cell far better at delivering a transmit burst in cold.
4. A fixed annual battery visit in October, before the first hard freeze, covering all nine sensors: inventory check, replacement for the five cold-space sensors on schedule and for any interior sensor below its alert threshold, re-secure, test each to the customer's phone. Time booked: about 45 minutes, which is nine sensors at about 5 minutes each including the crawlspace crawl.

The next winter, nothing dropped. The shop's record for that house showed the lithium cells in the five cold-space sensors still above the alert threshold at the October visit, and they were replaced anyway on schedule, because a flat lithium curve means a healthy-looking reading is not evidence of another winter's margin.

What would have changed the plan: in a mild-climate house where the garage never approaches freezing, the cold-space rules mostly fall away and the annual visit can stretch, with replacement at first alert doing most of the work. In a house where the customer will not allow anyone in the crawlspace, the honest move is to say the crawlspace sensors cannot be maintained on schedule and to put that in writing.

References

• 16 CFR Part 1263, CPSC safety standard for button and coin cell batteries and consumer products containing them
• Battery manufacturers' published datasheets for alkaline and lithium AA cells (operating temperature ranges)
• NFPA 72, National Fire Alarm and Signaling Code, in the edition your jurisdiction has adopted (smoke alarm replacement)
• See related: Leaving a Device Record the Next Tech Can Use
• See related: Setting Up Freeze and Power-Loss Alerts for a Vacant Home