Meter Testing, Repair, and Replacement

Incoming revenue defines the viability and sustainability of any functioning utility.  Water sales revenue is directly linked to customer meter reading (and billing). Water loss detection and management is also tied to meter reading and data management.  Therefore, maintaining accurate and reliable meters in place is critical to overall utility operations.

Water meters are the basic link between a water company and its customers.  Meter reading, testing, replacement, and data management are key responsibilities that every water company or utility must address.  Meter reading and data management are discussed in the previous BMP on customer water use data collection and management.  This BMP will focus on testing, repair and replacement of meters.

Typically meters are tested in place in the field or in a laboratory by removing the customer meter from the yoke and using a test bench.  Field testing methods typically involve using calibrated flow meters to monitor flow through a hose bib or similar and comparing that measured flow rate with the one recorded by the customer’s meter.  Field testing can be accurately performed on meters from 5/8-inch to 8-inch (http://www.ejprescott.com/media/reference/SensusMeterTestEquiH-25.pdf).  Laboratory testing requires removing the meter and replacing it with a temporary meter, while the meter is gone. Many laboratories can perform testing quickly using overnight shipping to minimize disruption in service and metering. Laboratory test benches can be configured to handle any size meter, however due to size and weight concerns typically laboratory bench testing is limited to meters that are 2-inches and less.

Implement a program to test all distribution system meters at regular intervals. A protocol based on age of meter can be established for the testing and replacement frequency best suited for the utility. In accordance with the guidelines provided by the American Water Works Association (AWWA) Manual M6, 95 percent of meters scheduled for tests on a periodic basis should be actually tested. In addition, at least 95 percent of the meters actually tested should register results within the accuracy limits established for both normal and minimum test-flow rates.

The AWWA recommends that meters in service be tested, on average, as follows:

Meter sizes 5/8 in. to 1 in = every 10 years

Meter sizes 1 in. to 4 in. = every 5 years

Meter sizes 4 in. and larger = every year

It is advisable to provide for more frequent tests of large meters on the basis that an error in their registration has a greater effect on customer equity, utility credibility and on revenue issues. Older meters and those registering the largest volume should be given priority, since they generally read low. Mechanical drive meters require more frequent maintenance and show increased wear tendency compared to magnetic drive meters.

As water meters age, they can lose accuracy due to worn parts (e.g., impellers worn down by sand and grit) or chemical scaling or other causes, natural and man caused.  Although many meters can be repaired, if the housing cracks or the impellers are damaged, it is often more cost effective to simply replace the meter, especially for meters that are 1.5-inches or smaller (1.5-inch meters cost in the range of $500, whereas 2-inch meters can cost closer to $1,000 or more; 5/8-inch meters can cost less than $100).  Due to the number of different types and sizes of meters, it is typically in the best interest of the utility to have standard practices for meter repair versus replacement programs.  In this way, a consistent set of customer meters can be installed over time.

The water utility will be required to replace water meters due to age, damage, and malfunction such that customer water use can be accurately measured and billed.  Because meter technology is always improving, new types and models of meters are frequently available in the marketplace.  Utilities often buy meters in bulk, to receive price breaks, such that one type of meter is installed for a group of new meter installations, until a new bulk purchase is made.  In this way, meters within a single organization may be associated with many different manufacturers. 

One important consideration that utilities should consider, especially for large meters, is their ability to maintain accurate readings over a large range of flow rates.  All meters have an optimum range for flow detection and measures, however mechanical flow meters typically do not measure low flows accurately, which in large diameter meters can be a significant amount of water.  In addition, residential meters, low flow (< 0.1 gpm, which is equivalent to nearly 150 gallons per day) can be confounding, since small leaks in residential fixtures (i.e., faucet aerators or toilets) could draw flows in this range.   Given that many utilities have a great percentage of residential customers, not tracking or detecting low flows can create a substantial apparent water loss.

New meters can track lower volumes of water movement (typically down to 0.25 to 0.5 gpm) (which is equivalent to 350 to 720 gallons per day).  To improve on this performance, there are unmeasured flow reducers (UFR) that can be added to meter yokes that will batch the water related to low flows, allowing it to be measured and not lost.  The equipment for a 5/8-inch connection is about $65, not including labor to install.

Meter replacement should be conducted in a manner that is stepwise and purposeful.  In other words, meter replacement should be budgeted for and implemented each year for some specific group of customer meters, with the replacement program based on data collection and verified with testing and analysis, if possible. 

For example, meter replacement for residential customers could be based on tracking of water deliver for each meter, with replacement scheduled whenever a meter reaches 2 million gallons of water delivery (as well as in the case of mechanical failure, which may be budgeted as a percentage of total meters in place).   Residential customer meter replacement could also be developed as 10% per year, such that every 10 years every customer would receive a new meter.   In these two examples, no meter testing would be required, in that data tracking either by water delivery or meter age is used to signal meter replacement.  It is a better management practice to replace a working meter erroneously than to leave an inaccurate meter in place given the negative consequences of providing water to customers for free.

For larger meters, replacement intervals should be shorter, if age is tracked to signal meter replacement.  Replacing and/or testing the meters should follow the AWWA timing listed above.  For very large meters (4-inch or greater) it is a best practice to test yearly, and be in position to replace and/or repair the meter in accordance with this testing schedule.

 

Resources

American Water Works Association Publications on Meters