Sample stream mixing is the deliberate homogenizing of a stratified or non uniform stream upstream of a sample probe, so the probe draws from an even mixture rather than a separated one. Left alone, free water and sediment settle to the bottom of a pipe and lighter product rides on top, so a probe sees whatever happens to be at its elevation, not the true average. Mixing disperses the phases across the cross section so a single probe location becomes representative. This page explains why upstream mixing is needed, the mixing options available, and how poor mixing invalidates water content and composition results.
Sample Stream Mixing in one line: Sample stream mixing homogenizes a stream before it reaches the sample probe, dispersing free water, sediment, and any stratified phases evenly across the pipe cross section so the probe draws a representative mixture. Without it, heavier phases settle to the bottom and the probe over- or under-samples them depending on its position, biasing water content and composition.
Liquids of different density do not stay mixed on their own. In a pipeline carrying crude with free water and sediment, the water and solids are heavier than the oil, and given low velocity or a horizontal run they migrate toward the bottom of the pipe while the lighter oil rides above. The stream becomes stratified: a cross section through it shows water rich fluid low and oil rich fluid high, not a uniform blend. A sample probe sits at one place in that cross section and can only draw from the layer in front of it, so what it captures depends entirely on where its tip happens to be.
That makes a stratified stream fundamentally unsamplable by a single fixed probe. Put the probe low and it over samples water and reports the stream far wetter than it is; put it high and it under samples water and reports it far drier. Neither reading is the true average water content of the whole stream, and there is no single probe position that gives the right answer for a stratified flow, because the correct answer is a blend of all the layers and no one layer contains it. The probe is honest about the layer it sees and wrong about the stream.
Because water content, the water and sediment fraction, directly sets net oil in custody transfer, stratification is not a minor sampling nuisance but a direct threat to measurement fairness. A stratified stream sampled by a badly placed probe produces a water content that is systematically wrong, and that error flows straight into the accounted net oil. The purpose of mixing is to remove the stratification before the probe, turning a stream that no single probe could sample correctly into one where a single well placed probe captures the true average.
The most common device is a static mixer, a set of fixed vanes or elements installed in the pipe that force the flow to fold and cross over itself as it passes through. Because it has no moving parts, a static mixer is simple and reliable, and it draws its mixing energy from the pressure drop the flow gives up crossing it. Its effectiveness depends on flow: at good velocity it homogenizes the stream well, but at very low flow there may not be enough energy to fully disperse heavy water, so a static mixer is sized for the flow range it will actually see and the probe is placed just downstream where the mixing is freshest.
A power mixer, or motor driven mixer, adds its own energy rather than relying on the stream's pressure drop, using a driven impeller to actively stir the fluid. This makes it far less dependent on flow rate, so it can keep a stream homogeneous even at low velocities where a static mixer would struggle, which suits streams with heavy or variable water that would otherwise separate. The trade is complexity: a power mixer is a rotating machine that consumes power and needs maintenance, so it is chosen when the mixing duty is genuinely beyond what a passive element can guarantee. Jet or recirculation mixing, where a portion of the stream is pumped back in as a high velocity jet to stir the pipe, is another active option used in some installations.
Whatever the mixing element, its placement relative to the probe governs whether it does any good. Mixing has to happen upstream of the probe and close enough that the stream has not re stratified by the time it reaches the tip, because a homogeneous mixture left to travel a long, slow horizontal run will simply separate again before it is sampled. The probe therefore sits just after the mixer, in a section where velocity keeps the mixture together. Getting the pairing right, adequate mixing plus a probe positioned in the well mixed zone, is what actually delivers a representative sample; a good mixer with a badly placed probe, or a probe after too long a settling run, still samples a separated stream.
Poor mixing produces a recognizable failure: water content and composition results that do not match independent checks and that vary in ways the true stream should not. Water contents that swing with flow rate are a classic sign, because a static mixer loses effectiveness as flow drops, so the stream re stratifies at low flow and the probe reads differently even though the crude has not changed. Results that read persistently high or low compared to reconciliation point to a probe drawing from a stratified layer rather than a mixed stream. A power mixer that has failed or a static mixer that is fouled or undersized shows up as water content that has become erratic or biased without any real change in the product.
Diagnosis starts by separating mixing problems from the other things that bias water content, because the symptoms overlap. Poor mixing and non isokinetic withdrawal both bias the heavy phase, so the first step is to confirm the stream is actually being homogenized: is the mixer intact and, if powered, running, is it sized for the flow it is seeing, and is the probe positioned in the well mixed zone rather than after a long settling run. Only once mixing is confirmed sound does it make sense to look at withdrawal velocity, sample handling, or the analysis itself. The tell that points at mixing specifically is the dependence on flow rate, since that is the variable a passive mixer is most sensitive to.
Flow rate is exactly the variable that governs static mixer performance, and it is measured, which makes the mixing risk visible through monitoring. A cloud SCADA platform such as Merobix reads the metered flow and, where the mixer is powered, its run status, along with the water content the sampling and analysis produce, so operators can trend water content against flow across their custody points. A sample point whose reported water content tracks flow rate, or a power mixer that has stopped running, reveals a stream that is stratifying and being sampled unrepresentatively, and that pattern surfaces in the trends before it becomes a settlement dispute. The mixing happens in the pipe, but the evidence that the stream reaching the probe is actually homogeneous lives in the monitored data across the fleet.
Because heavier phases like free water and sediment settle to the bottom of the pipe, a stream can stratify into water-rich fluid low and oil-rich fluid high. A single fixed probe then draws only from the layer in front of it and reports whatever is at its elevation, not the true average. Mixing disperses the phases evenly across the cross section so a single well-placed probe captures a representative sample.
A static mixer is a fixed set of vanes that homogenizes the stream using the pressure drop of the flow itself, with no moving parts, so it is simple and reliable but loses effectiveness at very low flow. A power mixer adds its own energy through a driven impeller, so it stays effective even at low velocities and heavy water, at the cost of being a rotating machine that consumes power and needs maintenance. The choice depends on how demanding the mixing duty is across the flow range.
If the stream is not homogenized, it stratifies and the probe draws from an unrepresentative layer, so water content reads too high if the probe is low or too low if it is high, and neither reflects the true average. A telltale sign is water content that swings with flow rate, because a static mixer loses effectiveness as flow drops. Since water content sets net oil in custody transfer, this bias flows directly into the accounted volume.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.