Home / Insights / Ultrasonic water meters
Water & Compliance · Technology guide 7 min read Published 24 August 2026

Ultrasonic water meters: what they fix, what they cost you to ignore, and when mechanical still wins

By Frank Guo · Technology & Product Leadership, addanode

TL;DR — An ultrasonic water meter measures flow with sound waves instead of a spinning mechanism: no moving parts, so nothing to wear out, jam with grit, or slow down with age. That solves the two failure modes that quietly create non-revenue water in South African networks — under-registration as mechanical meters wear, and blindness to low flows (the dribble a leaking toilet or an almost-closed bypass produces). Ultrasonics also ship telemetry-ready, which is why smart-metering rollouts standardise on them. The honest caveats: they cost more upfront, they need batteries or power, and on a clean, high-flow bulk line a good mechanical meter with a pulse output — telemetered — still earns its keep. Choose per installation, not per fashion.

How an ultrasonic meter actually works

Two transducers sit in (or clamp onto) the pipe, sending ultrasonic pulses diagonally through the water — one with the flow, one against it. Sound travels faster downstream than upstream; the transit-time difference is directly proportional to flow velocity, and velocity × pipe area = flow rate. No impeller, no gears, no register wheel: the measurement is electronic end to end, which is also why the meter can timestamp, log and transmit every reading.

The three problems ultrasonics solve in South African networks

  • Wear-driven under-registration. A mechanical meter's moving parts slow down over years of grit and water hammer — and a slow meter bills less than it delivers, converting revenue into apparent "loss". Fleet-wide, this is one of the largest commercial-loss mechanisms utilities face. An ultrasonic meter's accuracy doesn't decay with mechanical wear.
  • Low-flow blindness. Mechanical meters have a starting flow below which the impeller simply doesn't turn. Leaking cisterns, dripping taps and throttled illegal connections live below that threshold — measured as zero. Ultrasonics register these trickles, which is exactly the consumption profile that reveals leaks on the customer side.
  • Manual reading, with all its costs. Ultrasonics are natively electronic, so AMR/AMI comes built in or bolts on cleanly: readings over LoRaWAN, 4G or drive-by, into the same dashboards as your water monitoring. The meter reader's route — and the estimated readings when the route fails — disappear.
Wear and a starting flow are two different losses Two mechanisms make a mechanical meter under-read, and they are not the same problem. Wear slows the moving parts, so the meter bills less than it delivers and the difference appears as apparent loss. A starting flow sets a floor below which consumption is recorded as zero — which is exactly where a leaking cistern or a throttled illegal connection sits. A mechanical meter An ultrasonic meter Accuracy over years Moving parts slow with grit and water hammer, so the meter bills less than it delivers No mechanical wear for accuracy to decay with What a slow meter does to the loss figure Turns revenue water into apparent loss on the books Nothing, because there is no mechanical decay to turn A leaking cistern, or a throttled illegal connection Below the starting flow, so it is measured as zero Registered — and that trickle is the profile that reveals a customer-side leak Reading it A meter reader's route, and estimated readings when the route fails Natively electronic: AMR or AMI over LoRaWAN, 4G or drive-by Where the other one still wins Cost, on a connection where none of the above matters Not automatically the right answer everywhere
Wear and a starting flow are two different losses Wear makes it bill less than it delivers; a starting flow records a trickle as zero. Accuracy over years Mechanical Moving parts slow with grit and water hammer, so the meter bills less than it delivers Ultrasonic No mechanical wear for accuracy to decay with What a slow meter does to the loss figure Mechanical Turns revenue water into apparent loss on the books Ultrasonic Nothing, because there is no mechanical decay to turn A leaking cistern, or a throttled illegal connection Mechanical Below the starting flow, so it is measured as zero Ultrasonic Registered — and that trickle is the profile that reveals a customer-side leak Reading it Mechanical A meter reader's route, and estimated readings when the route fails Ultrasonic Natively electronic: AMR or AMI over LoRaWAN, 4G or drive-by Where the other one still wins Mechanical Cost, on a connection where none of the above matters Ultrasonic Not automatically the right answer everywhere
The three problems above, against where the other technology still wins. Wear and a starting flow are two different losses — one makes a meter bill less than it delivers, the other records a trickle as zero.

Where mechanical meters still make sense

An honest selection guide has to say this part out loud:

  • Clean, steady, high-flow bulk lines — a well-sized mechanical bulk meter with a pulse output, telemetered through a gateway, delivers most of the operational value at a lower unit price. Many of our DMA installations run exactly this way.
  • Sites with no power case — ultrasonics run on long-life batteries (typically several years), but battery replacement across thousands of buried chambers is a real lifecycle cost that must be planned, not discovered.
  • Tight capital, wide rollout — if the budget covers ultrasonics for 30% of the network or mechanical-plus-telemetry for 90%, coverage usually wins: you can't analyse the zone you didn't meter.

Ultrasonic, AMR, AMI — untangling the jargon

These get sold as one bundle but are three separate decisions. The meter technology (ultrasonic vs mechanical) decides measurement quality. AMR — automated meter reading — means readings collect themselves (walk-by, drive-by or network) instead of via a clipboard. AMI — advanced metering infrastructure — means a permanent two-way network with interval data flowing into your systems continuously. You can telemeter a mechanical meter (AMR on old iron) and you can hand-read an ultrasonic (waste of its talents). Smart-metering rollouts in the region — including Nairobi's ultrasonic programme north of us — pair the technologies deliberately: ultrasonic accuracy where revenue is at stake, network reading everywhere.

What this looks like in practice

Our role is the telemetry and analytics layer, vendor-neutral on the meter itself: we read ultrasonic meters' digital outputs, mechanical meters' pulses, and everything into the addaNet platform — where zone balances, night-flow analysis and billing reconciliation actually happen. The meter is the sensor; the loss programme is the point.

Ultrasonic vs mechanical vs electromagnetic — side by side

Mechanical (velocity / volumetric)UltrasonicElectromagnetic
Moving partsYes — wear and driftNoneNone
Low-flow accuracyPoor as it ages; under-reads dripsGood — starts near zeroGood
Dirty water / solidsClogsTolerant (clamp-on especially)Tolerant; needs conductive fluid
PowerNoneBattery, 10–15 yr typicalMains or larger battery
TelemetryBolt-on pulse or cameraNative — built for AMR/AMINative
Typical roleLegacy consumer metersConsumer and small bulk, remote sitesBulk and DMA inlets
CostLowestMidHighest

Frequently asked questions

What is the advantage of an ultrasonic water meter?

No moving parts — so accuracy doesn't decay with wear, low flows that stall a mechanical impeller still register, dirty water can't jam it, and it's tamper-resistant and natively telemetry-ready. For revenue metering, the wear point matters most: a meter that stays accurate protects billing for its whole life.

Are ultrasonic water meters worth the extra cost?

Where under-registration or low-flow losses are suspected — older meter fleets, high-value connections, prepaid schemes — usually yes: recovered billing pays the premium. On clean bulk lines a telemetered mechanical meter is often the pragmatic choice. Run the arithmetic per installation class, not for the fleet as one decision.

How long do ultrasonic meter batteries last?

Typically several years (manufacturers commonly rate 6–15 years depending on model and transmission frequency) — but treat battery replacement as a planned lifecycle cost across the fleet, because thousands of buried meters reaching end-of-battery together is a programme, not an errand.

Can existing mechanical meters be made "smart" instead?

Often, yes — many mechanical bulk meters accept a pulse or register-mounted sensor, which we telemeter into the same dashboards as native smart meters. It's the fastest, cheapest first step for zone metering; meter replacement then follows where measured drift proves it's needed.

Do ultrasonic meters help reduce non-revenue water?

They attack the commercial-loss half: under-registration and unread low flows. The physical-loss half — leaks and bursts — is found by zone balancing and night-flow analysis, which need telemetered zone meters of either technology. The winning rollouts do both, which is why meter choice and telemetry design belong in one conversation.

Related reading

Solution

Non-revenue water reduction

Where the meter data goes to work.
Guide

Smart metering vs NRW

What metering realistically recovers.
Price guide

Smart metering project costs

What an NRW project costs in SA.

Primary sources

Choosing meters? Design the telemetry first.

Send us your network layout and metering questions. We'll tell you honestly where ultrasonics pay, where mechanical-plus-telemetry wins, and how the data becomes a loss programme.

Talk to our Water Team NRW solution →