Meters and AMI in practice

You didn't sign up to be a meter technician, but the questions land on your desk anyway. A stuck meter under-registers for months before anyone notices. A reading route that used to take two days now takes a week. And the board wants your opinion on an AMI sales pitch by Thursday.

None of that requires an engineering degree. It requires a few habits: run routes that catch problems before the bill goes out, know which meters to test first, and ask a vendor the right questions instead of nodding along with the demo.

This is the practice side of meters and AMI, not the theory. For meter types, sizing math, and the full sensor-signal breakdown, see the companion guide on water distribution meters and AMI.

What should a reading route actually confirm?

A reading route confirms two things: the number is right, and it belongs to the right meter. Before you record anything, verify the service address and the meter serial number match what's on file. A misread on the wrong meter creates two billing errors, not one.

Read routes in the same order every time. A consistent sequence makes month-to-month comparisons meaningful, and it makes a skipped or estimated read obvious instead of buried in the pile.

Before the bill run closes, compare total production, what the source meter measured, against total billed consumption. A widening gap is your earliest signal that something in the route, not just one customer, needs a second look. Any register that's fogged, cracked, or reading erratically gets written down on the spot, not remembered later.

How do you work a high-bill or no-read complaint?

Start with the boring explanations before you touch the meter itself. Most complaints trace back to something simple, not a broken instrument.

• High bill: check for continuous flow and compare against the customer's historical

• Low bill or "meter not reading": look for a coupling failure, a stuck measuring element,

• No water or low pressure: check for a clogged strainer, a partially closed valve, a

• Erratic reads: on electromagnetic or ultrasonic meters, suspect air in the line,

pattern. The usual culprits are irrigation, a pool fill, an open hydrant, a service-line break, or a stuck toilet flapper. a dead endpoint, reverse flow, or a bypass valve left open. failed pressure-reducing valve, or a collapsed service line. turbulence, or a grounding and wiring problem before you suspect the sensor itself.

Whatever you find, write down what it was and what you did about it. That record is what turns a dispute into a closed case instead of a repeat call next month.

Which meters should you test or replace first?

Weight your list by revenue and by wear, not by whichever complaint came in loudest this week.

Start with the meters that carry the biggest financial swing if they drift. That means the production meter and the largest commercial accounts first. A small percentage error on a high-volume service costs far more than the same error on a residential line.

Next, look at age and material exposure. Positive-displacement, single-jet, and compound meters all have a measuring element that wears with use. Grit in the water acts like sandpaper on that element, and the older the meter gets, the sooner it drifts low.

Any meter already flagged by a complaint or a field observation moves to the front of the line: a stuck disc, a slipping magnetic coupling, a fouled jet. It's already telling you something.

Electromagnetic and ultrasonic meters don't wear the same way; they have no moving parts to grind down. When one reads badly, check grounding, wiring, and whether the pipe is actually running full before you condemn the sensor. Most "bad meter" calls on these types turn out to be installation problems.

While you're testing, ask whether the meter still fits the account's current flow. A meter sized for construction-era demand can be too big for what the property uses today. An oversized meter quietly under-registers at everyday flows for years before anyone notices the pattern.

What should you check when you install or replace a meter?

Meter accuracy is largely an installation outcome, not just a hardware choice. A perfectly good meter set up wrong will lie to you from day one.

Confirm flow direction and orientation match the manufacturer's approval before you tighten anything down. Air is the enemy on more meter types than you'd expect.

Avoid high points in the piping, keep the run full, and vent where the layout calls for it. Trapped air can spin an impeller and cause over-registration, or throw erratic readings on electronic meters.

Give the meter room to breathe hydraulically. Follow the straight-run distance your manufacturer or AWWA recommends near elbows, pumps, and valves.

Don't set a meter immediately downstream of a tee or a partially open valve. The closer a meter sits to a disturbance, the less predictable its reads get, especially on velocity-based types.

Think past the meter itself: the pit or vault has to be safe to access, drain, and service. It needs protection from freezing, flooding, traffic loads, and vandalism. A meter that's technically accurate but sitting in a flooded pit isn't doing anyone any good.

What does AMI actually change for a small system?

AMI does not change what a meter measures. It changes how often you hear from it, and that gap is bigger than it sounds.

A manual or touch read gives you one number, once a month or every other month. Drive-by AMR collects that same single number faster, but it's still one number when you go get it. A fixed AMI network reports on a schedule, often every 15 or 60 minutes, plus alerts, with nobody driving anywhere.

A monthly total tells you how much water moved. Interval data tells you when it moved, how steadily, and what changed. That's the difference that surfaces leaks, pressure problems, and demand patterns a single number can't show.

It also settles billing disputes faster: an hour-by-hour usage record beats arguing over one lump total.

Small systems get an extra advantage here. A district metered area typically covers 1,000 to 2,000 customers, sometimes up to 5,000. Plenty of small systems are close to that size already.

That means your whole distribution area can function like a single leak-detection zone. Your overnight minimum flow becomes one of the most direct leak signals you have.

What should you ask an AMI vendor before you sign anything?

Ask about your terrain and your budget before you ask about features. The network technology has to fit both, or the rest of the pitch doesn't matter.

• Which network type fits your service area: dense neighborhoods suit mesh, moderate density

• How the system tells you when a meter goes quiet, and how fast you'd know about it.

• What data comes standard and what costs extra: interval reads, leak or continuous-flow

• How reads reach your billing and mapping systems, and whether that link is automatic or a

• What happens to your data and your hardware if you switch providers down the road.

with decent tower placement suits point-to-multipoint, rural or scattered sites suit cellular (with its own ongoing data fee), and infrequent reads over long range suit low-power radio with well-placed gateways. alerts, and acoustic or pressure add-ons aren't always bundled together. manual export somebody has to run.

A vendor who answers those five plainly, without steering you toward the most expensive tier, is one worth a second meeting.

None of this asks you to take anyone's word for it. Test the meters that carry the most risk. Document what you find on every route, and let a vendor's answers, not their slide deck, make the AMI call for you.

For what happens to a route's findings once they hit a customer's account, the field guide's entry on the billing cycle picks up from here.