Water level sensors: what they save is a trip, not money

Some links on this page are affiliate links. HornCore earns a commission on qualifying purchases, at no extra cost to you.

Research. Nothing here is monitored yet. What follows is what the three sensing methods measure, why the radio is the hard part on a rural place, and whether an alert arrives early enough to matter. Once a monitor has been through a freeze here, this page gets updated with the date.

A tank monitor does not save water and does not save money. It saves a walk, and occasionally a tank.

What it actually saves

Price the walk honestly. A tank sixty yards off the back door, checked twice daily on the way to something else, is not worth $200 to $400 of electronics and a plan. A tank half a mile up a lane that goes soft in February is different arithmetic.

A monitor is information, not action. It does not close a valve, stop a leak or start a pump. Shutoff is a separate purchase with its own failure mode, and a mechanical float valve does the ordinary job with no battery and no radio.

Float, pressure, ultrasonic

Float switch. A magnet in a float closes a contact at a set height. Cheap and hard to confuse. One or two setpoints and no trend — full or not full, nothing about how fast. Fails by sticking: crud on the stem, or ice.

Submersible pressure transducer. Sits on the bottom and reads the head of water above it — a continuous level, and the accurate option. The detail nobody prints: gauge-referenced transducers carry a vent tube up the inside of the cable to reference atmospheric pressure. Let that vent get damp or plugged and readings drift with the weather.

Ultrasonic from above. Nothing in the water, which is the appeal. It times an echo through air, and the speed of sound moves with temperature — an uncompensated head reads differently at 20 F than at 90 F. It also has a dead band: it cannot read the top four to twelve inches, so mounting high costs range at the full end.

Closed cistern: ultrasonic or pressure. Open stock tank: float. An ultrasonic head over an open tank will happily measure a heifer.

The radio is the hard part

Prove the link before you buy the sensor. Backwards is how people end up owning a gateway they never budgeted for.

Wifi is barely in this conversation. It dies inside 200 feet, and leaf-on trees eat 2.4 GHz.

LoRa and other sub-GHz are built for this and quote kilometres — in line of sight. Terrain is the real spec, not distance, and a ridge between gateway and tank cuts it hard. It needs a gateway at the house with power and internet of its own.

Cellular works where the network is at that spot. Bars at the house prove nothing about the bottom of a hollow. Walk down with a phone on the same carrier, leaf-on, and check that data moves, not that a call connects. Then add $5 to $15 a month, usually per device.

Power and winter

Battery figures assume a slow reporting interval and mild weather. A sensor quoted at two years reporting hourly at 70 F is not a two-year sensor reporting every fifteen minutes at 20 F. Primary lithium cells hold capacity in cold far better than alkaline. Size solar for December, not for the year.

The freeze failure is the ugly one. A stuck float reports the last thing it saw and keeps reporting it, so the alarm you trust says all is well while the tank sits empty under ice. A pressure transducer keeps reading the water below the cap, which is better behaviour but not the answer. The answer is a tank heater.

Will the alert arrive in time

Work out how long your tank goes from fine to empty. A 100-gallon tank with ten head in July, or any tank with a split fitting, can be gone in two hours. If the threshold sits at a quarter full and the tank drains in ninety minutes, the alert buys twenty minutes and you are at work.

So set the threshold high and alarm on the wrong thing deliberately. "Not refilling after a drawdown" is an earlier signal than "nearly empty." The failure that costs most is the opposite one anyway — a float valve stuck open, running the well into a tank that reads full throughout.

Then test the notification path, and test it again after the next phone update. A push notification is silent at three in the morning.

What we are buying

This one stalls on a question that has nothing to do with electronics: whether the walk a monitor would save is worth $200 to $400 of hardware and a plan on ground like ours. The radio is the next gate, and it has to be proved at the tank rather than from the house — a walk we have not made. A purchase records the sensing method, the radio, what a gateway added on top of the sensor, and the date it went in. The rest is slower than a receipt. Whether a threshold we choose raises the alert early enough to be any use, and whether a float still reads true with ice on it, are answered by weather. The Research tag stays until a monitor here has been through a freeze.

Before you buy

• Prove the radio link at the tank before you own the sensor. Terrain, not stated range.
• Float in an open stock tank. Pressure or ultrasonic in a closed cistern.
• Alarm on filling behaviour, not on nearly empty. Nearly empty is already too late.

— Chris D., McEwen, Tennessee