Water Quality Wire

The conversation around advanced metering infrastructure has moved. A Water Finance and Management piece on AMI frames the current question as one of data usage rather than data collection. Its summary notes that the benefits of the technology include the fact that utility staff "no longer have to walk meter routes," along with more accurate billing and earlier leak identification. The summary does not address the harder question of what a utility should do with the reads once it has them, and that question depends on settings that are easy to overlook.

The Timestamp Is a Data Field, Not a Given

Each interval read is only as useful as the time attached to it. A meter records consumption against its own clock. The communication path and head-end system may then assign a receipt time, a measurement time, or a re-stamped time, depending on how the system was configured. If a clock drifts, or if the head end labels reads by arrival, an hourly spike can land in the wrong hour. For a monthly bill, that offset rarely matters. For a district metered area balance or a pressure-zone analysis, a one-hour shift can move a flow signature across an operating boundary. Staff should confirm which time basis the vendor uses before comparing reads with SCADA logs, pump run records, or sampling times recorded in the lab.

Interval Length Decides the Question You Can Answer

Daily, hourly, and fifteen-minute intervals answer different questions. Daily totals confirm billing trends and show gross changes in demand, but they smooth away short overnight events, such as a running toilet or a slow leak that disappears inside a daily sum. Fifteen-minute reads expose the night minimum flow, which is a common place to look for leakage in a district. They also multiply storage requirements and make routine noise harder to separate from signal. Hourly intervals are often a practical compromise. The tradeoff is less about technical elegance than about which decision the utility is prepared to make and whether the chosen interval can support it.

Register Units and Pulse Weighting

Electronic registers, pulse outputs, and older mechanical registers do not always share the same resolution. A pulse may represent a fixed volume that varies by meter size or by the scaling chosen at programming. When a utility merges reads from more than one meter generation, rounding and pulse weighting can differ between populations. A fleet in which some registers count in one-unit steps and others in tenths will show step changes that have nothing to do with consumption. Before any analysis that depends on small flows, staff should record the pulse value, the register unit, and the firmware revision for each meter population.

Where the Dealer Fits

Data usage depends on a service chain that often runs through a provider between the manufacturer and the utility. Independent regional dealers such as Jones Air & Water frequently handle commissioning, programming, and field service, which makes them a natural place to capture the programmed interval, the time source, and the register configuration for each site. A written commissioning record gives the utility a baseline. Without one, staff inherit a dataset whose assumptions were never documented, and the first anomaly becomes an argument about which setting was active on which date.

A Short Test Before the Next Phase

A utility planning to rely on interval data can run a limited check before expanding analytics. Install test flows on a small sample of meters with known volumes, then compare the recorded timestamps and volumes against a reference. Confirm the interval length and pulse weighting for each population. Then document what the head-end system does with a missing read: whether it interpolates, records zero, holds the last value, or flags a gap. Gap-filling defaults change the outcome of leak detection, and a default that nobody wrote down becomes a silent input to every report that follows.

The usage shift described by Water Finance and Management is real, and it rewards staff who treat the clock, the interval, and the register as part of the data rather than as background. Collection produced the numbers. Usage depends on knowing what those numbers were told to mean when they were recorded.