Acoustic Sand Detector Virtual Calibration: Methods and Validation

Acoustic Sand Detectors (ASDs) have been widely used on topsides and subsea flowlines to measure sand production for the last 25 years. Essentially, ASDs are filtered microphones, strapped onto a bend or tee, that detect ultrasonic noise generated by impacting sand. ASDs are typically used either to detect the presence of sand or to quantify the sand rate. In the former case the well will typically be choked back when sand is present. This approach has the disadvantage that production may be curtailed even when sand production is too low to be problematic. Alternatively, the ASD could be calibrated by means of a sand injection. Once calibrated the ASD can output the sand rate in grams per second, allowing the operator to judge whether the sand rate justifies choking back the well. This strategy can significantly boost production rates. However, sand injection is not always feasible for topsides ASD installations and has never been done subsea (to our knowledge). This paper summarises work that attempts to address this limitation. A correlation was developed that calculates ASD calibration coefficients without the requirement to perform a sand injection. It works on the assumption that sand noise increases when fluid noise increases, with additional corrections for various multiphase flow effects. The correlation was developed using historical data from laboratory tests and from offshore sand injections. It was then validated against additional historical topsides injection data from platforms in the North Sea. In this initial assessment the correlation was validated against approximately 800 data points. One of the main potential applications of the correlation is to help estimate sand rates using data from subsea ASDs. Unfortunately, no suitable subsea data was available that could confirm the validity of the correlation, but it was shown that ASDs respond differently when installed subsea. Subsequently, laboratory flow loop testing was performed with ASDs installed on a pipe submerged in a pool. This confirmed that the correlation has a similar accuracy for both topsides and subsea ASD installations. The correlation is currently being trialed by an operator with a view to using it as a tool within their overall sand management strategy. ASDs are widely installed, but their output is often poorly understood or misused. This paper outlines the correlation methodology and identifies the possibilities and limitations of other approaches. The validation and testing exercises also flag a range of other issues when using ASDs; the effects of poor installation; how subsea acoustic attenuation affects ASDs; and important considerations when performing sand injection calibrations.