Composite and spot sampling are two strategies for representing a stream, and the choice between them decides how well the sample matches the material that actually moved. A spot sample is a single grab taken at one moment; a composite is many small increments blended so the sample averages the stream over a batch or period. Which one you use is not a matter of preference but of what the measurement has to prove, and for custody or compliance the requirement is often specific. This guide compares the two strategies, explains flow-proportional versus time-proportional compositing, and shows how flow data drives a good composite.
Composite vs Spot Sampling in one line: Spot sampling takes a single grab of the stream at one instant, while composite sampling blends many small increments collected across a batch or time period into one averaged sample. A spot sample suits a stable stream or a quick check; a composite is used when the stream varies and the result must represent the whole transfer, which is why custody and compliance often require a composite.
A spot sample, also called a grab sample, is exactly what it sounds like: the operator captures a single portion of the stream at one point and one moment. Its virtue is simplicity and speed, and for a stream that is genuinely uniform, one grab can fairly represent the whole. Its weakness is that it sees only that instant, so if the stream's composition changes over the transfer, whether the grab happened to catch a representative moment is a matter of luck. A spot sample answers what the stream looked like right then, no more.
A composite sample is built to defeat that limitation by combining many small increments taken across the batch or period into a single container. Because the composite blends the stream from beginning to end, it averages out the variation that a spot sample would miss, and the resulting analysis represents the whole quantity that moved rather than one snapshot of it. The trade-off is complexity: a composite requires equipment or a routine to collect increments over time and blend them correctly, where a spot sample needs only one draw.
The decision between them follows from what the number is for. If a stream is stable and the question is a quick quality check, a spot sample is efficient and sufficient. If the stream varies and the result will be used to value or account for an entire transfer, a spot sample risks being unrepresentative in a way that matters, and a composite is the honest choice. This is why custody transfer and compliance measurements, where a whole batch is valued on one analysis, so often specify a composite rather than a grab.
Once a composite is required, the next question is how the increments are spaced, and there are two logics. Time-proportional compositing takes an increment at fixed time intervals, one every so many minutes regardless of how much product is flowing. It is simple and works well when the flow rate is roughly constant, because equal time then corresponds to equal volume. But when flow varies, time-proportional sampling over- or under-represents periods, taking the same number of increments during a trickle as during full flow, which can bias the composite.
Flow-proportional compositing fixes that by tying increments to volume rather than to the clock: an increment is taken every time a set quantity of product has passed, so more increments are collected when flow is high and fewer when it is low. The composite then weights each part of the transfer by how much product actually moved during it, which is precisely what a representative average of a varying stream requires. For custody transfer, where the sample must represent the volume that changed hands, flow-proportional compositing is the standard because a barrel that flowed fast and a barrel that flowed slow both get their fair share of the sample.
The distinction is not academic when flow swings during a transfer. A batch that starts slow, ramps to full rate, and tails off would give a time-proportional composite dominated by the slow start and finish, misrepresenting the bulk of the material that moved at full rate. A flow-proportional composite of the same batch tracks the volume and represents it correctly. This is why serious custody and batch sampling reaches for flow-proportional compositing whenever the flow is not steady.
A flow-proportional composite is only as good as the flow signal that paces it, because the sampler needs to know how much product has passed to decide when to take the next increment. That is a measurement-and-control task: the flow meter reports throughput, and the sampler grabs an increment each time the accumulated volume crosses a threshold. Without a reliable flow signal, flow-proportional sampling is impossible, and this is exactly where field measurement and sampling meet.
A cloud SCADA platform such as Merobix carries the flow measurement that drives compositing and records the sampling alongside it. The same metered throughput that accounts for the transfer can pace a flow-proportional sampler, and the platform logs the flow trend and the transfer volume so an engineer can confirm, from a browser, that the composite was collected in proportion to the product that actually moved. When a composite result is questioned, the flow history is the evidence that the sample fairly represents the batch.
This linkage also helps decide which strategy a given measurement warrants. Continuous flow trends show whether a stream is steady enough that a spot sample or a time-proportional composite is acceptable, or variable enough that only a flow-proportional composite will represent it. Rather than choosing a sampling strategy by habit, an operator can look at how the flow actually behaves and match the sampling to the stream. Composite versus spot remains a decision about representativeness, and reliable flow data is what turns that decision into a defensible one.
A spot sample is appropriate when the stream is stable and uniform or when you need a quick, single-point quality check rather than a representation of a whole transfer. If the composition does not change meaningfully over time, one grab fairly represents the stream. When the stream varies or the result must value an entire batch, a composite is used instead because a single grab could easily miss the true average.
Time-proportional sampling takes an increment at fixed time intervals, which represents the stream well only when flow is roughly constant. Flow-proportional sampling takes an increment each time a set volume has passed, so periods of high flow contribute more increments than periods of low flow. Flow-proportional sampling weights the composite by actual volume, which is why it is preferred for custody transfer when the flow rate varies.
In custody transfer a whole batch is valued on the quality determined from the sample, so the sample must represent the entire quantity that moved, not one instant of it. A composite, especially a flow-proportional one, blends increments across the transfer so the analysis reflects the average of everything that changed hands. A single spot sample could catch an unrepresentative moment and misvalue the batch.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.