The gas an analyzer is measuring right now was drawn from the pipeline some time ago, because it had to travel from the sample tap through the conditioning system to the detector. That travel time is the sample lag, and on a steady gas nobody cares. But when composition is changing quickly, during a blend change, a well swing, or an upset, the analyzer is reporting the past while the flow meter is measuring the present, and multiplying them together without accounting for the lag books energy against the wrong volume. Understanding sample lag is understanding one of the subtler ways energy accounting drifts off.
Analyzer Sample Lag Time in one line: Analyzer sample lag time is the transport delay between the sample tap on the pipeline and the moment the gas reaches the analyzer detector, caused by the volume of the sample line and conditioning system and the rate gas moves through it. It matters when composition is changing, because the heating value the analyzer reports corresponds to gas that was in the pipe some time ago, not the gas flowing past the meter now. Uncompensated, this misalignment multiplies the wrong heating value against the current volume and distorts energy accounting.
When an analyzer draws a sample, that gas does not arrive at the detector instantly. It travels from the sample probe in the pipeline, through the sample line, through filters, regulators, and any conditioning, and into the analyzer, and all of that has volume the gas must sweep through before a new parcel reaches the measurement. The time this takes is the sample lag, and it is set by the volume of everything between the tap and the detector divided by the flow rate through it, plus the analyzer's own cycle time if it is a batch instrument like a chromatograph. A long sample line or a slow flow makes the lag longer.
Sample systems often use a fast loop precisely to fight this lag. A fast loop pulls a larger continuous flow from the pipeline past the analyzer take-off point and returns it, so the gas at the analyzer's tap is kept fresh and close to what is currently in the line, and only a small slip stream branches off into the analyzer itself. The fast loop shortens the transport delay for the bulk of the run, but there is still residual lag in the branch to the detector and, for a chromatograph, in the analysis cycle, so a fast loop reduces sample lag but does not eliminate it.
It helps to separate the two kinds of delay involved. Transport lag is pure dead time: the gas simply has not arrived yet, and nothing about the measurement changes until it does. Analyzer cycle time is different, because a chromatograph reports a result only once per cycle, so even after fresh gas arrives, its composition may not be reported until the next analysis completes. Both add up to how stale the reported composition is relative to the gas flowing past the meter, and both matter when the gas is changing fast enough that staleness turns into error.
Energy measurement multiplies volume by heating value: the flow computer takes the volume the meter measured over an interval and multiplies it by the heating value of that gas to get energy, typically in BTU or MMBTU. That multiplication is only correct if the heating value and the volume refer to the same gas. On a steady gas they always do, because the composition is the same now as it was when the sample was drawn, so the lag is invisible. The problem appears only when the composition is changing, because then the heating value the analyzer reports belongs to gas that has already passed the meter or has not yet reached it.
Consider a blend change that raises heating value. If the analyzer lags the meter by some minutes, then during the transition the flow computer is multiplying the current, higher-flowing or richer volume by an older, lower heating value, or the reverse, depending on the direction of the swing. The energy booked during the transition is wrong, sometimes over and sometimes under, and while a single short swing may be small, frequent swings, or a facility that blends or cycles often, accumulate a real error in the energy that gets billed. The faster and larger the composition changes, the more the uncompensated lag matters.
The fix is time alignment: delay the heating value so that when the flow computer multiplies, it uses the composition that actually corresponds to the volume flowing at that moment. If the sample lag is known, the composition can be held back by that amount so it lines up with the gas the meter is currently measuring, which re-couples heating value to the right volume through the transition. This is why sample lag is not just an analyzer curiosity but a number that has to be quantified and applied, because getting the alignment right is what keeps energy accounting honest when the gas is moving.
The practical machinery for handling sample lag lives in the flow computer and the SCADA layer, and it starts with knowing the lag. Operators characterize the sample system's transport delay, often by introducing a known change and timing how long it takes the analyzer to see it, and that measured lag becomes the alignment the system applies. With the lag known, the composition can be delayed before it is combined with volume, so the energy calculation uses time-aligned inputs rather than whatever the analyzer happens to have reported most recently.
Time-stamping is what makes this auditable rather than approximate. When each analysis carries the time it represents rather than the time it was reported, the SCADA layer can align composition to volume deliberately and can also show operators how stale the current composition is, which is exactly the information needed during a fast transition. A composition that is several minutes old on a rapidly swinging gas is a warning that the instantaneous energy number is uncertain, and surfacing that staleness lets an operator treat transition-period numbers with appropriate caution rather than trusting them blindly.
A cloud monitoring platform such as Merobix helps by keeping the time-stamped composition, the flow, and the resulting energy together, so the alignment can be reviewed and the effect of a swing can be seen after the fact. Because the platform holds the history with real timestamps, an operator investigating a disputed energy total during a blend change can see how the composition and volume lined up through the transition and whether the lag alignment held, rather than trying to reconstruct it from separate instruments with their own clocks. That turns sample lag from an invisible source of accounting drift into something the monitoring record can account for.
Energy is calculated by multiplying the measured volume by the heating value of that gas, which is only correct if both refer to the same parcel. On a steady gas the composition now is the same as when the sample was drawn, so the lag makes no difference. When composition is changing, the heating value the analyzer reports belongs to gas that was in the line earlier, so multiplying it by the current volume books energy against the wrong composition and distorts the total.
A fast loop pulls a larger continuous flow from the pipeline past the analyzer take-off point and returns it, keeping the gas at the analyzer tap fresh and close to what is currently in the line. Only a small slip stream branches off into the analyzer itself. This shortens the transport delay for most of the run, though residual lag remains in the branch to the detector and, for a chromatograph, in the analysis cycle, so a fast loop reduces sample lag but does not remove it entirely.
You first quantify the lag, often by introducing a known change and timing how long the analyzer takes to see it, then time-align the composition to the volume by delaying the heating value by that amount before the flow computer multiplies. Time-stamping each analysis with the moment it represents rather than the moment it was reported makes the alignment deliberate and auditable. The goal is to ensure the heating value used in the energy calculation matches the gas actually flowing past the meter.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.