Leak Busters

Sept. 1, 2010

The days of free or unmetered water, or flat-rate billing for it, are gone or fast fading; now, perhaps, looms the demise of another water utility icon: the boxy, gray, analog-dialed mechanical positive-displacement utility meter. Though, for the moment, the positive displacement (PD) meter rules an estimated 80 to 85% of the residential market, sundown is coming—that is, if newly arriving, high-performance solid-state competitors have their way.

Within this year and last, several next-generation breakthroughs have hit this market. These solid-state, more technologically sophisticated boxes reportedly will last longer, provide significantly improved precision, add new capabilities, and generate higher revenues, all for more efficient resource management.

Also arriving in 2010 will be the definitive, long-unfolding, exhaustive study on this topic, done by the Water Research Foundation (WRF), working under the American Water Works Association (AWWA). For the past five years, a team of lab techs have been scrutinizing, logging, and tracking the performance qualities of no fewer than about 800 domestic US water meter models. Publication of the results will surely bring a tremendous impact on water utility management for years to come—especially as the best of the new meter design succeed in phasing-out the old.

Now you may wonder: Who out there cares so much about water measurement to such an exacting degree, and why?

Dripping Taps, Floating Flappers
To answer, the following brief discussion on the underlying fiscal calculus of the water industry may enlighten and surprise you.

One ardent and high-profile leak hunter, Tom Kelly of the Washington Suburban Sanitary Commission (WSSC) (450,000 water/sewer customer accounts) sums up the root problem in this concisely: “The 24/7 factor is an absolute killer,” he says.

The meter-mounting flange bolts were poured with the concrete, making the meter unmovable.

What he means, of course, is that it’s not so much leaks per se that are the concern, it’s the prospect of their indefinite persistence. It’s their permanence over time. Just think of the accumulating water wastage and cost that are sustained from tens of thousands of endpoint leaks never detected and fixed.

As an illustration, consider these ramifications: Normal water consumption at a home (typical at WSSC and elsewhere) might be 16,000 gallons per quarter. Now, aggregating a rash of bad leaks at this home might add up to one-fourth of a gallon per minute. Multiply that by 60 minutes in an hour, times 24 hours a day, times three months of a yearly quarter. This nets an astounding 32,000 gallons!

In other words, steady leaks might precisely double a home’s water usage; in effect, the leak transforms one account into two more phantom, hard-to-detect ones, and difficult—at first—even to believe in. It begins sounding like a water utility horror movie: They’re ghostly and hidden—maybe silent or making faint tapping sounds—but nobody seems to notice. They’re ghosts getting water, but not paying for it.

The real issue isn’t leaks, or even water loss: It’s money—over the years, thousands, then millions, in lost revenues and unrecovered costs spent treating, pumping, and sewering.

Improperly installed or faulty, damaged meters often result in loss of revenues.

Granted, this quart-size leak example makes a perhaps unrealistic case. In order to aggregate one-fourth of a gallon per mile (gpm) it would take a large houseful of dysfunctional tank flappers and dribbling spigots. But here’s the thing: Even if the total leakage were just a fraction of this, the enormous multiplier-effect of a city full of drips will mean, all told, maybe tens of thousands of phantom freeloaders.

Also note: The one-fourth-gpm leakage scenario is not arbitrary; it’s the rate that represents the gentle seepage that can occur within a meter—a threshold level, below which many meters cannot register movement. The water is lost. It is rendered unmeasurable, not only because in-home leaks are shrugged off, but because meters aren’t sensitive enough to detect them.

Low-threshold leakage is a well-known, perennial challenge to leak-hunters, says Kelly, whose job title is Strategic Coordinator, and who also sits on the AWWA’s meter standards committee. He notes that a bit of internal seepage is inevitably suffered by most meters; the AWWA’s M6 standard specifies, in the Manual for Meter Selection, Installation, Testing, and Maintenance, that 95% of flow must be measured; this means 5% is allowed to squeeze through uncounted—even more, if the meter is worn.

Of course, mechanical meters do wear. One of Kelly’s strategic tasks is to sample and examine WSSC’s huge inventory of PD meters for wear, by putting them through low-, medium-, and high-flow testing rigors. He minutely scrutinizes physical tolerances within their chambers, tolerances which must be tight enough to prevent internal seepage under pressure, yet loose enough to allow parts to move and to record flows.

Photos: Craig Hannah of Johnson Controls, Inc.
Trash dumped in meter vault

Meter has been disabled.

Meters newly arriving from the manufacturer are also spot-tested by him. A fresh batch, he says, “starts out having a tight tolerance of maybe five-thousandths of an inch.” Then, “over a long period of time, it loosens by another eight-, ten-, and twenty-thousandths. That doesn’t seem that much, but with a PD meter you have to have a pressure differential between the upstream and downstream sides in order to cause those measuring components to move and registration to occur. Below a certain flow rate, water can essentially slide through those tolerances without causing a pressure differential.”

Components that don’t move, don’t register. And each meter truly represents, as it were, a utility cash register. Customers thus get useless “free water,” dripping down the drains.

In the nearby state of Pennsylvania, one of Kelly’s leak-hunting colleagues, Robert Bissey, serves as distribution services manager of the York Water Company (61,000 accounts, all PD-metered). Bissey is also current president of the Easter Meter Management Association (EMMA): They’re a regional group of water company managers—leak hunters—who together explore the nuances of meter quality, performance, and optimization—the applied science of meterology.

Bissey’s meter regimens are similar to Kelly’s, scaled-down for a small town. For example, in 2009 York Water routinely replaced about 6,000 aging Neptune meters. The retired units were then pried open, to examine the wear and tear. From this inspection Bissey could extrapolate and approximate the remaining meter service life in other delivery sections.

Technicians found only about 50 of the 6,000 had big problems—less than 1%. Bissey attributes their good condition to the York-area’s soft and grit-free water. On this point, he observes, “Conditions of water really plays a part as to meter accuracy” and wear.

Localities with hard water and sandy soils tend to suffer markedly worse meter erosion and inaccuracy. York’s customers will likely keep their the current one in service at least 15 years, says Bissey, and maybe up to 30.

York Water also does macro-scale, twice-yearly accuracy audits, using a regimen encouraged by AWWA. The procedure usually goes as follows:

First, he explains, “an entire pumped area is measured with a propeller-type master meter in the pump houses.” From this, “We know exactly how much water has passed through that pump house and into that given repump zone.”

Meanwhile, every customer’s meter has been coded, “to match it with that booster re-pump zone—so that we can tally all the individual meters in the customers’ homes and come up with a cumulative gallonage.”

This figure is then matched against the meter tally in the pump house. Discrepancies, if any, point to problems, such as the presence and severity of likely losses, either from cumulative leaks or people opening fire hydrants and stealing; or from major breaks or a serious meter inaccuracy at the booster station.

Seeing anomalous numbers, says Bissey, “We can go back, look at history and say, ‘Wow, this month it’s really up!’”—or down, as the case may be. Then crews go out to troubleshoot leaks accordingly.

Bissey also raises a critical point on accuracy relating to meter sizes—be they the sub-inch pipes delivering water to homes, or the larger two-, three-, four- and six-inch diameters

Multi-jet meters must be installed level and horizontal to be accurate

for commercial sites. Larger meters wear more rapidly than home-scale ones, and—because each represents so much money coming in—they must be changed about every four or five years.

Large meters also work harder and prove less durable. “When they start to go south, they go real quick,” he says. “As they begin slowing down, by the time we get to them and pull them out, and get them back to the shop and tested, they’re normally down to just 80 to 85% accuracy. So we are losing 15 to 20%”—in other words, unacceptably big bucks.

It’s quite common for a disproportionate share of revenues to come from a small slice of accounts: In the case of nearby WSSC, about 40% of the agency’s receipts derive from maybe 4% or 5% (20,000 meters) of the total base, notes Kelly, who’s also active in EMMA.

More frequent testing of large-pipe units is thus desirable, to assure continued good performance at varied flows. In some meters, tens of thousands of dollars may be at stake in a single location.

So: Is One Meter the Most Accurate?
Along with the pervasive PD classic, all of the remaining half-dozen mechanical home types “live up to the AWWA standards, pretty much” answers Ken Molli. He is an independent water utility consultant who specializes in helping agencies assess meter performance and manage replacement programs; he also spent six years with a meter manufacturer.

Although performance differs subtly between, say, velocity-type meters (a bit wider variances) and PDs, the answer really depends, again, on local water hardness and particulates. “PDs, may not work as well in gritty [water], but an inferential meter may have less problem, because only part of the rotor is in the stream, and it doesn’t get blocked up with that kind of material,” he says. (Besides the PD, the other mechanical types include single- and multijet, fluidic oscillator, piston, nutating disks, turbine, and propeller designs.)

Also, how correctly the meter gets installed can play a huge role. “If it’s not perfectly level, front-to-back and side-to-side for an inferential meter, it may cause bearings to wear faster, and that meter becomes inaccurate sooner,” says Molli.

Actually, the “Which meter is best?” question is of such great moment to the industry, that the quest for answers was undertaken (as noted above) in an exhaustive five-year study, just now being completed by the AWWA’s WRF. One participant was associate research professor Michael Johnson, of Utah State University’s Water Research Laboratory (USUWRL), where most of the WRF study was done.

To attempt to find the very best meters, USUWRL researcher collected, he says, “upwards of 800 meters, mostly residential,” sized five-eighths, three-fourths, and 2 inches, from which dozens of specific data-points were harvested.

“We tried to cover everything available in the US that claimed to meet the AWWA,” he recounts.

Products were put through accelerated erosion tests, forced to handle hard-water grit, and subjected to the all-important low-, medium-, and high-flow assessments, to see at what point they fail to register adequately.

The voluminous final report, titled Accuracy of In-Service Water Meters at Low and High Flow Rates, is set for release in October.

What conclusions or surprises can Johnson reveal?

He answers, guardedly that, “Performance is so water-dependent it is hard to say.” But some meters “are really good” for having passed the sand-particle test, which others failed. In some cases, results will depend on the installation.

All in all, he says, “There are a lot of good products and lot of good choices, and the technology is still advancing.” Results are being revealed through various AWWA media.

Residential Metering: Three Breakthroughs
The story of the year in meter accuracy, surely, is the launching of several dramatically innovative new products. What’s happened, in essence, is that designs, which were long used in large-diameter, higher-end electromagnetic and ultrasonic fluid metering (especially for oil pipelines), have been scaled down and adapted for home water meters.

Two industry leaders, Sensus and Badger, are now in the throes of introducing products based on the two key enabling technologies: Respectively, these are Sensus’ electromagnetic iPERL, and Badger’s ultrasonic eMeter, says Craig Hannah, of Johnson Controls Inc. (JCI), of Milwaukee, WI. Along with the US firms, global leader Elster AMCO Water recently brought out its SmartMeter and an evoQ4 Mag Meter (electromagnetic) with fire service UL listing.

Hannah, a development engineer for JCI’s municipal utilities solutions team, visits industrial plants where he constantly encounters the larger, high-end predecessors to the new home meters. For the latter use, he thinks the impact of electromagnetic and ultrasonic technology is going to be “really exciting,” for several reasons. Above all is the capability of capturing minuscule flows that PDs miss. “Today, traditional [PD] meters are designed to register minimum flow-rates at about one-sixteenth gpm when new,” he says. “The new electronic meters register flows as low as one-twentieth gpm.

“So, you’ve got phenomenal low-flow accuracy,” he adds.

Another benefit will be the ability to detect and measure backflow. Transient reverse-flows sometimes occur, for instance, any time a pump motor turns on or off or valves open and close quickly; dampeners of the resulting waves cause pressures which might push water the wrong way. In this scenario, “with a traditional PD or multi-jet meter—due to their design and a totalizing register on them—you really can’t measure backflow. But with these new electronic meters, you’re going to have instant notification of a reverse-flow event,” says Hannah.

Sensus’ iPERL, the first and so far only home-scaled “mag meter,” as they’re nicknamed, was introduced at the AWWA show last year. Sensus executive Doug Neely describes how they accomplish measurement by developing a magnetic field across a known diameter. “This captures a lot more low-flow data than the typical mechanical meters can,” he says. In the larger-diameter sizes, sensitivity reaches as low as one-sixteenth of a gpm. AWWA’s standard only requires one-eighth gpm.

Inside the meter chamber, a stainless-steel liner renders it fairly impervious to erosion from sandy grit or particles, notes Neely, who was Sensus’ vice president of sales for North America during much of the iPERL rollout; since March, he’s been a company consultant. He also participates on the Water Advisory Group at Utilimetrics, the automated metering trade association.

“The increases in low-flow registration are going to offset the higher cost of the meter within a few months of service,” he believes. “If you have small leakage in a toilet or sink, you’re probably not being billed. But the new electronic meter will detect and bill for that usage.”

Being solid-state, the meter can sniff not only trace flows, but can associate usage with time; this will enable utilities to begin billing, someday, at tiered water rates that are seemingly destined to come, in some towns. Revenues will probably soar with these money-makers.

Higher bills often do follow soon after the implementation of any meter, notes Peter Sanburn, senior product marketing analyst at Itron, the automatic meter reading/ Advanced Meter Infrastructure (AMR/AMI) industry leader in North America. (Several sources for this story agree, and two suggested that the billing impact after a changeout approaches 100% of ratepayers.) The reason is that new meters are simply more accurate. Billing spikes do come as a shock to recipients, so this may necessitate planning a public relations campaigns to explain to ratepayers what’s happening. A consultant, Molli, who has participated in dozens of meter change-outs from the customer side, notes that the billing surprise, indeed, “is one of the biggest issues” with meter upgrades.

But on the positive side, Molli suggests that a revenue boost of a few percent can bring payback of the incremental premium over the cost of a cheaper conventional meter, in 10 to 20 years. When doing calculations, he adds, don’t forget to include sewer rates if linked to water, and figure that a solid-state meter will save on maintenance.

Photo: Aclara
AMR aids in the collection and transmission of water use measurements.

On this last point, though, Hannah finds a rather different scenario occurring with large-bore, solid-state meters. “We’ve seen a lot of customers that never bother to test and calibrate their electronic meters,” he says. Negligence results in a drifting effect. Meters “will lose their accuracy, and consequences vary all over, from a few percentage points to 10 or 15%. A number of factors cause this. Drift is very common, and it absolutely occurs if electronic meters are not calibrated and recalibrated.”

AMR: A Passive Bystander Here
Automated meter reading, which produces so much buzz these days, plays no role in meter accuracy—as “the sole purpose of AMR is to collect the measurement off the register and transmit it,” says Sanburn.

What about corrupted AMR transmission? He’s unaware, he answers, of a single instance “in which Itron products have been involved in and proven inaccurate in transmission of collected read back to utility.” Radio transmissions are so well-encoded that interference would be prohibitively tough to do. Moreover, the technology has steadily improved since its introduction about 15 years ago; not only is the streaming data more refined, but the signal also sends tamper alerts and weak battery alarms. The battery can store and resend data for many weeks, if need be. Reliable battery life has lengthened to about 10 or 20 years.

One possibility for overcoming transmission issues involves the use of multiple, time-synched meter endpoints. By utilizing system synchronization, utilities can compile real-time use data. Aclara’s 3300 Water Meter Transmission Unit (MTU), for example, takes hourly, time-stamped reading over the STAR Network system.

However, as Bissey points out, there are occasional signal strength challenges and issues. York has used AMR for years, and is getting good reads monthly with a drive-by Neptune R900 1-way system. Failures occur less than 1% of the time. Signals are sometimes demanded from deep subterranean meters positioned behind thick walls, making reads very difficult, but, ultimately, not insurmountable. In his repair shop, he sometimes discovers an extremely rare inaccuracy condition in which a meter wheel gets stuck, say, in the “two” position, at which, “for some reason the radio picks it up as a three.” But any wildly anomalous readings like these get kicked out by the billing system, so that’s really little problem here.

New Ultrasonic Meters…and a High-Performance PD Renaissance
The second of the two solid-state innovations being fielded for homes is an ultrasonic one, as noted from Badger. Along with it, a line of new ultrasonics for larger water-pipe sizes is also being phase-in, from Master Meter.

Ultrasonic technology uses transducers and sound energy to measure flows through the pipe; again, as in the electromagnetic meter, the absence of moving parts means little or no wear of parts, longer life, and reduced maintenance.

Taken together as a one-two combination on the market, the technologies “will give you very similar performance characteristics,” asserts Master Meter’s vice president of marketing Ian MacLeod.

Master Meter has long been known to the industry for its multi-jet meter, a distant competitor with PD; his firm wanted to grab more of the emerging solid-state opportunity says MacLeod, so it broke from its traditional product and launched into R&D (research and development) some years ago. The decision to go with ultrasonic over electromagnetic came down to a lower retail figure.

Regardless of pricing, though, any solid-state meter is going to be a bargain, MacLeod says, “because it’s not going to wear out, and it’s extremely accurate, having the ability to measure in ranges from a quarter-gallon to a thousand gallons.” Master Meter’s ultrasonic product, called the Octave, is now launching abroad in ISO measurements, and domestically will come in 6-, 4-, 3-, and 2-inch sizes, starting in a few months.

As for residential three-fourths- and five-eighths-inch use (where Badger lately introduced the eMeter), MacLeod thinks the high added cost is hard to justify; so, his company decided to zero-in on designing instead, a “new and better” PD meter. Vastly improved accuracy and wear-resistance for the old classic were achieved at computer-aided design drafting desks, he says, by applying unprecedented sophisticated computer modeling and simulation of flow dynamics. Resulting product performance was then verified in the above-described WRF study: In the new PD meters’ best trials in Utah, it retained new meter accuracy, he says, even after four million gallons of normal tap flushed through in, an accelerated test; that’s equivalent to 35 or 40 years of use.

Summing up: Whatever meters are going to be installed, utilities will find the newcomers more accurate and more sensitive to registering low flows; also, quieter, longer lasting, and able to reduce pressure losses (the last feature translating into reduced pumping resistance and expense).

Sensus’ Neely sees a bright future for water resource management, speaking with an optimism that undoubtedly is shared by all of these leak busters. He says, “I can’t think of a better, more exciting time to be part of this industry, and to see the benefits of this type of technology getting into the marketplace.”

About the Author

David Engle

David Engle specializes in construction-related topics.