Water-Meter Pulse Totals: Check Litres per Pulse Before Comparing Usage

A water-use report may contain accurate timestamps and perfectly consistent arithmetic while still representing the wrong quantity. If its conversion factor treats every pulse as one litre when the documented channel represents ten litres, the reported volume will be only one-tenth of the intended result. Before interpreting a change as a consumption trend, reconcile the count, the meaning of the count and the reporting period.

This is a read-only data-review method. It uses exported records and existing documentation to test a calculation. It does not describe changing a meter, modifying an acquisition channel or making connections. A discrepancy should be documented and referred to the responsible metering team with the supporting evidence.

1. Establish what the channel actually records

A field named “pulses” is not a complete unit. Identify the meter, channel, output type, documented volume per pulse and any transformation already performed by the logging system. Determine whether the exported field is a cumulative raw count, an interval count, an already converted volume or a flow-rate value derived from a frequency signal. Multiplying an already converted volume by the pulse factor again creates a second scaling error.

The ONICON FT-3400 manual distinguishes frequency-related parameters from scaled volume pulses and identifies different output meanings, including forward volume, reverse volume and status functions. This supports a basic review principle: the channel’s documented meaning must be established before its numerical values are treated as a water quantity. The example below is illustrative and is not a claim about any unit’s present settings.

Keep a small evidence record beside the analysis: asset identifier, source document revision, exported field name, factor, factor units and reporting period. A factor written as “10” is ambiguous. Ten litres per pulse and ten pulses per litre are reciprocal descriptions and lead to different operations.

2. Convert a documented count into volume

Assume an exported cumulative counter reads 8,200 pulses at 08:00 and 9,450 pulses at 10:00 on the same day. Assume the count is monotonic within this interval, no reset or rollover occurs, all relevant pulses are captured, and the documented scale is 10 L/pulse. Both readings refer to the same channel and time basis.

• Interval count = 9,450 − 8,200 = 1,250 pulses.

• Interval volume = 1,250 pulses × 10 L/pulse = 12,500 L.

• Cubic-metre volume = 12,500/1,000 = 12.50 m³.

• Elapsed time = 2 hours = 7,200 seconds.

• Interval-average flow = 12,500/7,200 = 1.736 L/s, or 6.25 m³/h.

The pulse unit cancels when multiplied by litres per pulse. If documentation instead gives a factor F in pulses per litre, the correct volume is count/F. Always perform this unit cancellation explicitly before copying an equation across a large dataset.

Original read-only reconciliation example. Average flow uses elapsed time; the pulse count does not reveal peak flow.

The average flow describes the entire two-hour interval. It does not show the maximum flow, the minimum flow or whether consumption occurred continuously. The same 12.50 m³ could be distributed evenly or concentrated into a shorter period. A cumulative total cannot recover that missing time pattern by itself.

3. Quantify a scaling discrepancy correctly

Now suppose a separate report used 1 L/pulse for these same 1,250 counts. It would report 1,250 L, or 1.25 m³. The difference from the documented-factor result is 11.25 m³. Relative to 12.50 m³, the report is low by 11.25/12.50 = 90%.

Be precise with the denominator. Correcting 1.25 m³ to 12.50 m³ is a tenfold result, or a 900% increase relative to the erroneous number. That does not mean actual water use rose by 900%. It means the quantity calculation changed. Label corrected historical values as recalculated data and retain the original report so the change is auditable.

A factor-of-ten mismatch is a useful clue, not proof of its cause. The wrong volume unit, duplicated conversion, a copied channel definition or a different reporting interval could produce related symptoms. Confirm the documented factor and data lineage before assigning an explanation.

4. Match time boundaries before comparing registers

Suppose the meter display was photographed at 10:04 while the exported period ends at 10:00. Even a correctly scaled count may disagree because the display includes four additional minutes. Compare the difference between two register readings with the pulse accumulation across the same pair of timestamps, rather than comparing a lifetime register with one day’s pulse total.

Record the timezone and whether a timestamp marks the start or end of an interval. Daylight-saving changes, delayed uploads and local-versus-server timestamps can all complicate reporting periods. Use actual elapsed time for average-flow calculations. A nominal “daily” row is not automatically evidence of an uninterrupted 24-hour observation.

If a reading is missing at the exact boundary, disclose any interpolation and the assumption behind it. Do not silently assign an entire interval to a different day or treat an absent record as zero usage. A short gap during a variable-flow period can matter more than a long gap during a documented idle period.

5. Treat counter discontinuities as review items

A negative difference between successive cumulative counts may indicate a reset, rollover, channel replacement or a different counting convention. It is not sufficient evidence of negative water use. Review the counter’s documented range and event history before applying any reconstruction. A rollover correction is valid only when the actual counter behavior and number of rollovers are established.

Direction also matters. Forward-only, reverse-only and net quantities answer different questions. The difference between forward and reverse totals can be calculated only when both values cover the same period and share compatible units. An unexplained reversal or discontinuity should remain visibly flagged rather than being smoothed into a plausible chart.

6. Separate data resolution from meter accuracy

With a 10 L/pulse factor, one pulse represents 0.010 m³. Over the example’s two-hour interval, one count corresponds to 10/7,200 = 0.00139 L/s in interval-average flow. Over a much shorter period the same count represents a larger average-flow increment. This is a consequence of reporting resolution, not a statement of meter accuracy.

The ONICON manual also discusses pulse timing and overrun conditions. Existing alarm and event records can therefore be relevant when reviewing whether a recorded pulse total is complete. A mathematically correct conversion cannot restore unrecorded events or establish that the source measurement itself was accurate.

A defensible reconciliation ends with four linked items: the original counts, the documented unit factor, identical time boundaries and an exception list. Once those agree, usage comparisons become meaningful. Until then, present the chart as provisional and describe the unresolved data issue in plain language.