What is a representative sample?
How do you know if I have a representative sampler?
Assuming that the pipeline is sufficiently well mixed for sampling, then the representivity of a sample is largely determined by the following:
- Flow through the probe sampler or sampling bypass loop must be isokinetic to ensure that the fluids remain homogenous and representative of the pipeline profile. In a probe based system the head of the probe is usually streamlined or has an entry pitot to avoid bluff body effects biasing the flow entering the probe. In a bypass loop system, the loop size must be large enough to representative of the main pipeline (1" or above) and the velocity in the loop should be sufficiently high to maintain homogeneity (as per the mixing calculations in the standards). Both probe and bypass loop systems must be designed with no voids to water traps.
- The inlet to the sample probe or bypass loop should be 10 times the size of the water droplets created by the pipeline mixing and dispersion equipment.
- The individual sample grabs should be small enough so that enough of them are taken to be representative of the batch (for crude oil at least 10,000 - 1cc grabs (10 litres) is recommended). Each grab sample should be repeatable to +/-2% irrespective of any changes in pressure, viscosity, density or temperature. This points to the use of positive displacement sample extractors, as any device using line pressure for discharge is likely to be affected by process changes. How to ensure a system is working correctly
- The sampler should be operated flow proportionally. The water content changes during the batch and unless each grab sample represents a unit of volume (for example one grab every 50m3) the whole batch sample will be biased and not representative. The accuracy of the flow meter required to pace a sampling system is 10%.