Page 203 - Water Loss Control

Contr olling Appar ent Losses—Customer Meter Inaccuracy 177

Davis describes an assessment conducted for the Metropolitan Domestic Water
Improvement District, a small water supplier serving communities northwest of
2
Tucson, AZ. The methodology included meter test-
ing on randomly selected and high-cumulative vol-
ume residential water meters. Meter accuracy was Targeting meter rotations
plotted versus cumulative volume for individual based upon cumulative mea-
low-, medium-, and high-flow rates. The best linear sured volume is similar to
fit of the data was determined and the weighted automobile maintenance,
meter accuracy was plotted versus the cumulative
where the 3,000 mi oil and
volume. Calculated lost revenue from meter inac-
curacy per year was plotted versus the cumulative filter change occurs not at any
volume and economic analysis was used to deter- set time, but only when the
mine the optimum cumulative volume for meter 3,000 mi odometer reading is
replacement. For the District, the optimum cumula- reached.
tive volume was determined to be 1,420,000 gal per
residential meter. Prior to the assessment, the district was replacing customer meters
at the relatively frequent interval of every 10 years. Many of the district’s customer
meters do not achieve 1,420,000 gal of cumulative volume in 10 years, therefore the
district was able to implement a meter rotation strategy that greatly improved the cost-
effectiveness of its customer meter management.

12.4 Customer Meter Sizing
Water meters must be properly sized in accordance with the actual customer con-
sumption patterns in order to accurately register the flows at all levels of consump-
tion. Historically, water utilities sized customer service connections and meters based
upon the peak flow rates that the meter was expected to encounter. Since peak flows
occur only on rare occasions, most of the time meters sized in this manner registered
flows in the low end of their design range. Many meter types are less accurate in the
low end of their flow range with very low flows not captured at all. Current wisdom
focuses on sizing the meter to accurately capture the flow range most usually encoun-
tered, not seldom-occurring peak flows. Many water utilities have recovered consid-
erable water and revenue by right-sizing oversized customer meters. Between 1990
and 1992, for example, the Boston Water and Sewer Commission’s meter downsizing
program recovered over 100,000 cubic feet of additional water per day in apparent
water loss, which translated into millions of dollars in subsequent additional billings
and revenue. 3
Data-logging technology and fixed network AMR technology (discussed in Chap. 13)
provide the means to obtain detailed customer consumption profiles in increments of
minutes or hours for periods of days, weeks, or months. By using this detailed data,
meters can be sized to fit the individual consumption profiles of customers. Applying
this user-specific approach can promote superior meter accuracy, particularly in large
water utilities with widely varying user classes. As described in the AWWA M22 pub-
lication Sizing Water Service Lines and Meters, accurate data-logging for meter sizing is
4
dependent on the resolution of the data. Data resolution is a function of the water
volume per pulse logged and the data storage interval. Both should be as small as pos-
sible so that actual flow rates are recorded, as opposed to just a collection of average
flow rates, which may not accurately reflect the consumption profile. Examples of