Controlling Sand, In and Out of the Laboratory

Abstract

Guest editorial Laboratory work can be depressing business, particularly in the realm of sand control. The definitive answers that engineers require are not easy to deliver. Unlike pipes and widgets, rocks and reservoir fluids are not readily defined, and there are many uncertainties in sand failure and control in the field. Sand retention tests are dogged by experimental artifacts partly because the field processes cannot be replicated in the laboratory. We therefore have to rely on particle size distribution data, sometimes augmented with retention tests that are designed to be reproducible (rather than representative) so that different screens can be ranked for a particular application. Even something as apparently simple as particle sizing can be fraught with difficulties and unexplained discrepancies that have only recently become more widely recognized within the oil industry. Sieving is time-consuming and laborious when done meticulously, and inexperienced personnel coupled with imposed time constraints can result in poor final results. Use of a laser particle size analyzer (LPSA) is quicker and less prone to operator effects, but it also relies heavily on careful sample preparation and subsampling. It is common to get different results from different labs even if they use the same instrument; conversely, sometimes very comparable results can be obtained from different labs with different instruments. Even when the sample preparation variabilities are removed by measuring cleaned, disaggregated sands, there are sometimes distressingly large variations between labs. Laboratory experts are well aware of the complications and vagaries of testing. But engineers often require definitive answers in a short time (at minimum expense). With grain size measurements, this can mean that a single LPSA measurement is often made on a sand. Lack of duplicate analyses may mean that measurement problems are missed, especially if there is not any (carefully performed) sieve analysis data for comparison. So what should the industry do? In recent years, the range of sand control options has increased from a choice between wire wrap screen or gravel pack; nowadays a choice needs to be made between different designs of conventional screens, expandable screens, and gravel packs. Tools have been developed to allow the engineer to make decisions with limited information and within the constraints of current industry understanding. While they form a good starting point, the guidelines based on sand “quality” are necessarily conservative and need not always be applied as hard and fast rules. Additional data such as application specific laboratory testing or comparison with analogous field performance data can be used to challenge these guidelines while still managing risk. For example, the uniformity coefficient (defined as D40/D90) is a measure of the range of grain sizes in a sand, but the D90 is one of the most difficult parameters to measure reliably. Slight variations in the D90 calculated from LPSA measurements, which would be considered insignificant when viewed in cumulative distributions, can have a big impact on the uniformity.