Smart valve pilot paid off in last storm

During the Sunday cold snap, one bathroom supply line blew at 2:13 a.m., and the inline valve shut that stack in 45 seconds; zero unit-to-unit spread and an $800 dry-out instead of a $30k claim. Anyone else pairing leak puck sensors with centralized shutoffs, and what inspection cadence are you using for battery swaps?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌⁠​​‌‍⁠‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​‍​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌⁠‍​‌⁠‌‍‌‌​‍‌​⁠⁠‌‌⁠⁠‌‌‍‌‌​‍‌‌⁠​‌‌‍‍​‌​‌‍‌‌​‍‌⁠‍​​⁠‍‌‌‍⁠​‌​⁠‌‌​‍⁠​‍​‍‌⁠⁠‌​

But we pair puck sensors to stack zone valves and do an annual battery swap in October with lithium AAs… We also run a quarterly “wet test” (shot glass) on one puck per stack to confirm the valve closes inside 60 seconds — belt-and-suspenders, but it saves headaches. Moving pucks a few inches off tubs and adding a 30–45s debounce cut our false trips.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌⁠​​‌‍⁠‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​⁠​⁠​‌​⁠‍​​⁠‍​​⁠​‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​‍​⁠‌​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‍​‌⁠‌​‌‍⁠‍‌‌​​‌‌​‌‌‍​⁠‌⁠‌‌‌⁠‍​‌​⁠​‌⁠​‍‌​⁠‍​⁠​⁠‌‌‍‍​⁠‌​‌⁠​‍‌‍​⁠​‍​‍‌⁠⁠‌​​

Building on @alex6127’s testing idea, we added a monthly ‘valve exercise’ in the BMS: close each stack for about 10 seconds at 3 a.m., log travel time, and alarm if a motor doesn’t hit its stops — keeps them from seizing like an old truck you never start. In buildings with recirc or tankless we whitelist those loops during the window; are your actuators fail-closed or fail-open? It’s been cheap insurance and surfaces sticky stems before a storm.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌⁠​​‌‍⁠‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​⁠​⁠​‌​⁠‍​​⁠‍​​⁠​‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​⁠​⁠​​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌​⁠⁠‌⁠‍‍‌​⁠⁠‌⁠‍​‌‍⁠​‌‍‌‌‌​‌‍‌​⁠‍‌‍‍⁠‌‍‌​‌​​⁠‌⁠‌​‌‍‌‍‌‌‌⁠‌‍​‌‌‍​⁠​‍​‍‌⁠⁠‌​​

We moved to 24‑hour heartbeat pings on each puck — auto work order if a device misses two — and swap when reported voltage drops under 2.8V instead of by calendar; the date method’s fine, but this caught two weak cells right before a freeze. @alex6127, are you trending valve motor current to flag sticky actuators?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌⁠​​‌‍⁠‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​⁠​⁠​‌​⁠‍​​⁠‍​​⁠​‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​​​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‍⁠​⁠‍‌‌⁠​‍‌⁠​⁠‌‍​‌‌​‌⁠‌⁠‍​‌⁠‌‌‌‍‍‌‌‌‍​‌⁠‌​‌‌‌⁠‌‌‌⁠‌⁠‌​‌​​‍‌​‌​​‍​‍‌⁠⁠‌​​