When gas quality varies over a day, the single heating value used to bill for that day has to represent the whole day fairly, and the only fair way to do that is to weight it by how much gas actually flowed at each quality. A flow-weighted average heating value does exactly that: it gives more weight to the heating value that applied while a lot of gas was flowing and less weight to the value that applied while flow was low. The intuitive but wrong alternative is a simple time average, which treats every hour equally regardless of how much gas moved in it. Because flow rate and gas quality both change over a day, and often change together, a naive time average can be meaningfully off from the true energy, which is why flow computers accumulate energy interval by interval rather than averaging heating values by the clock.
Flow-weighted average heating value in one line: A flow-weighted average heating value is a daily or monthly average heating value in which each interval's heating value is weighted by the volume of gas that flowed during that interval, so periods of high flow count more than periods of low flow. It is the correct basis for billing because it reflects the energy actually delivered, unlike a simple time average that weights every hour equally regardless of flow. A flow computer achieves it by accumulating energy interval by interval and dividing total energy by total volume, which avoids the error naive time-averaging introduces when flow and quality both vary.
It is tempting to compute a daily heating value by simply averaging all the heating values measured during the day, treating each measurement equally. That is a time average, and it is wrong whenever the flow rate is not constant, because it gives the same weight to a heating value that applied while barely any gas was moving as to one that applied during peak flow. The heating value only matters in proportion to the gas it is applied to, so counting a low-flow hour and a high-flow hour equally misrepresents how much energy each contributed. The average heating value becomes a description of the readings rather than of the delivered energy.
The error becomes concrete when flow and quality vary together, which they often do in practice. Suppose a stream delivers most of its volume during a period of one heating value and only a trickle during a period of a different heating value. A simple time average would let the low-volume, off-quality period pull the daily number toward a value that had almost nothing to do with the energy actually delivered, because the vast majority of the gas moved at the other quality. The larger the swing in both flow and quality, and the more they are correlated, the further a time average drifts from the truth. This is not a rounding nuance; it can bias the billed energy in a consistent direction.
The fundamental point is that billing is about energy, and energy is heating value multiplied by volume, so the correct daily heating value is the one that reproduces the total energy when multiplied by the total volume. Only a flow-weighted average has that property. A time average has no guaranteed relationship to total energy at all once flow varies, so using it introduces an error that is entirely avoidable. The whole reason to weight by flow is to make the average heating value consistent with the energy that actually passed through the meter.
A flow computer sidesteps the averaging problem by not averaging heating values at all in the primary calculation. Instead it works interval by interval: in each short accounting interval it takes the volume that flowed during that interval and multiplies it by the heating value that applied during that interval to get the energy for that interval, then adds that energy into a running total. Over the day it accumulates both a total volume and a total energy this way, each built up from the same small intervals. Because the energy for each interval already carries the correct volume weighting, the accumulation is inherently flow-weighted.
The flow-weighted average heating value then falls out naturally as a derived number rather than being computed directly: it is simply the accumulated total energy divided by the accumulated total volume for the period. Because both totals were built interval by interval, that ratio is exactly the heating value that, applied to the whole volume, reproduces the whole energy, which is the definition of a proper flow-weighted average. The flow computer does not have to average heating values by the clock and hope; it computes energy where it belongs, at the interval, and lets the average emerge from the totals.
This interval-by-interval accumulation is also what makes the result robust to how wildly flow and quality move within the day. Whether the gas surges and quality shifts many times or holds steady, each interval contributes exactly its own energy, and the totals capture all of it correctly. Shortening the interval improves how faithfully rapid changes are captured, but the principle holds at any interval length: energy is accumulated at the interval and the weighted average is derived from the totals. That structure is precisely why electronic flow measurement is built around accumulating energy rather than averaging heating values.
For a monitoring and measurement operation, the practical lesson is to make sure the number that reaches billing is the flow-weighted average the flow computer accumulated, not a convenience average computed somewhere downstream by time. It is surprisingly easy to reintroduce the error by, for example, pulling the heating value readings into a database and averaging them by timestamp for a report, which quietly discards the flow weighting the flow computer worked to preserve. The discipline is to carry the accumulated energy and volume totals through to billing and derive the average from them, rather than re-averaging the instantaneous heating values.
A cloud SCADA platform such as Merobix supports this by trending and retaining the accumulated energy and volume totals from each meter, not just the instantaneous heating value, so the flow-weighted average can be reconstructed from the right quantities for any period. Keeping the energy and volume totals available means a daily or monthly heating value shown for billing or reconciliation is derived from the same accumulation the flow computer used, which keeps the monitoring view consistent with the custody record. It also lets an analyst check a billed average against the totals directly, rather than trusting a re-averaged figure that may have lost its flow weighting.
The stakes are commercial, because the flow-weighted average heating value is what turns a month of volume into a month of billed energy, and the difference between a proper flow-weighted number and a naive time average can be a real quantity of energy on a large stream. On a stream where flow and quality both swing, that difference is not noise; it is a systematic bias that favors one party over the other depending on how the swings line up. Getting the weighting right, and being able to show from the accumulated totals that it was right, is part of making the energy calculation both accurate and defensible, which is exactly what custody measurement demands and what a monitoring platform should preserve rather than undermine.
Because billing is about energy, and energy is heating value multiplied by volume, so the heating value only matters in proportion to the gas it is applied to. A time average weights every hour equally regardless of how much gas moved, letting a low-flow period count as much as a high-flow one, which misrepresents the delivered energy. A flow-weighted average gives more weight to the quality that applied while more gas flowed, so it reflects the energy actually delivered.
It does not average heating values directly. In each short interval it multiplies the interval's volume by the interval's heating value to get that interval's energy, and accumulates both total energy and total volume over the period. The flow-weighted average heating value is then simply total energy divided by total volume, which by construction is the value that reproduces the total energy when applied to the total volume.
It depends on how much flow and quality vary and whether they vary together. If flow is steady, a time average and a flow-weighted average nearly agree, but when most of the volume moves at one quality and only a trickle at another, a time average can drift meaningfully from the true energy because it ignores the volume weighting. On a large stream with correlated swings, that difference is a systematic bias in billed energy, not just rounding noise.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.