Using a Clamp-On Flow Meter for Leak Detection and Non-Revenue Water

A water utility pumps a certain volume into its network and manages to bill for a good deal less. The gap — water that was produced and paid for but never generated revenue — is called non-revenue water, and in many systems it runs between twenty and forty percent of everything put into the pipes. Some of that gap is metering error and unbilled legitimate use, but a large share is simply water leaking out of the ground through cracks, failed joints, and corroded mains. Closing that gap is worth real, recurring money. The obstacle is that finding a leak means measuring the network segment by segment, and you cannot shut down a live transmission main to bolt in a meter every time you want to check a section.

This is where a portable clamp-on flow meter becomes a leak-hunting instrument. This article covers the method, the one specification that determines whether you can find small leaks at all, the overnight measurement that does most of the work, and the specific pipe material that can defeat the whole approach if you are not watching for it.

The method: metering a live network without touching it

The core idea is simple. A leak is a volume of water that enters a segment of pipe and does not come out the far end. If you can measure the flow into and out of a section — or compare what a section should be carrying against what it actually is — the discrepancy points you toward the leak. The trouble is that installing meters on a live network is exactly the intrusive, shut-it-down operation you are trying to avoid.

A clamp-on meter removes that obstacle entirely. It straps to the outside of a live main, reads the flow through the pipe wall, and moves on — nothing cut, nothing shut down, no service interruption. That means a survey crew can walk a network, metering successive sections, hunting for the section that loses more than its neighbours, the flow that should be zero and is not, the delivered volume that does not reconcile with the supplied volume. The crew moves as fast as they can prep pipe and take a reading, and the network never knows they were there.

The one spec that matters: low-velocity resolution

Here is the specification that separates a meter that can find leaks from one that cannot, and it is not the headline accuracy figure people fixate on. It is the minimum resolvable velocity — the smallest flow the meter can distinguish from standing still.

The reason is straightforward. The flows you are hunting are often small. A leak might be a slow, steady loss; a night-time flow in a healthy zone should approach zero. To detect a small leak you must be able to measure a small velocity, and a meter that cannot resolve low velocities will report “zero” or noise across exactly the range where the useful signal lives. You will look at a leaking segment and see a number that looks like no flow, because the instrument could not resolve the small real flow that was there.

This is why, for leak-detection work, the specification to interrogate is low-velocity sensitivity. An instrument such as the Ultraflux UF801-P resolves velocity down to roughly ±0.03 feet per second, which is what lets it distinguish a genuine small leak from true zero flow. A meter without that sensitivity is not a leak-detection tool no matter what its accuracy class says, because it goes blind precisely where you need it to see.

Night-flow analysis: where the leak actually reveals itself

The single most productive measurement in leak detection is the overnight flow, and understanding why explains most of the discipline.

A district metered area — a discrete, hydraulically bounded zone of the network — has a daily rhythm. It runs hard in the morning as people wake and use water, tails off through the day, and drops to its lowest point in the small hours of the night when almost no one is drawing water. In a sound zone, that minimum night flow approaches zero, because legitimate demand has essentially stopped. In a leaking zone, the minimum night flow does not approach zero — because the leak does not sleep. Water is still escaping at 3am at the same rate it escapes at noon, and with legitimate demand stripped away, that persistent overnight flow is the leak, laid bare.

So the crew measures the minimum night flow in each zone and compares it against what legitimate overnight use should be. A zone whose night flow sits well above the expected legitimate minimum is a zone with a leak, and the size of the excess estimates the size of the leak. It is the cleanest signal in the whole discipline, and it is only accessible to an instrument that can resolve the small overnight flows — which loops back to low-velocity resolution and to turndown.

Turndown earns its importance here. Turndown is the ratio between the largest and smallest flows a meter can measure accurately. A meter with narrow turndown handles the busy daytime flow well and then loses the plot at the low overnight flow — reporting zero or noise across the very measurement that matters most. For night-flow work you want wide turndown so the meter stays honest from the daytime peak all the way down to the overnight trickle.

The pipe that defeats this: cement-lined ductile iron

Now the honest caveat, because a large fraction of water distribution mains are made of exactly the material that can stop a clamp-on meter cold: cement-mortar-lined ductile iron.

When the mortar lining is intact and bonded to the pipe wall, these pipes usually read fine. But on old mains the mortar frequently delaminates — it separates from the iron, leaving a thin air gap between the liner and the wall. Ultrasound cannot cross an air gap; the impedance mismatch destroys the signal. And this gap is on the inside of the pipe, unreachable and unfillable from the outside. On a segment where the liner has delaminated, a clamp-on meter simply will not read, and no adjustment fixes it.

The practical consequence for a leak survey is that you cannot assume uniform coverage across a network built from this material. Some segments will read perfectly and some will give you nothing, depending on the state of their liner — and you will not know which until you try. This is the textbook case where renting and testing before buying a fleet is the intelligent move. Take a meter, try it on a representative sample of your actual mains, and find out what fraction of your network you can actually survey before you commit capital to a program. Discovering the limitation during a rental is cheap; discovering it after buying twenty meters is not.

Building a survey program

Put together, an effective clamp-on leak-detection program looks like this. Divide the network into district metered areas you can measure at their boundaries. Survey each with a portable clamp-on meter chosen for low-velocity resolution and wide turndown, and lean heavily on minimum-night-flow measurements as the primary signal. Prioritise the zones whose night flow most exceeds their expected legitimate minimum, because those hold the biggest recoverable losses. And before committing to the program, prove the meter on your actual pipe material — especially if your network is heavy on cement-lined ductile iron — so you know going in what you can and cannot survey.

The payoff is that a non-invasive instrument turns leak detection from a series of disruptive dig-and-check operations into a walking survey of a live network. The water you recover is water you already paid to produce, so every leak found and fixed goes straight to the bottom line, year after year.

More on water & utility applications → · The UF801-P portable →