Grant: CCI Applied Research and Development Grants

Oil sands ore mined from the Athabasca oil sands are commonly transported in the form of slurry for further processing in hydrotransport pipelines. If flow conditions are insufficient for the complete suspension of coarse sand particles, the heavier grains may settle to form a sliding bed at the bottom of the pipeline. If these conditions continue, the entire hydrotransport pipeline may become plugged by deposited coarse grains in what is known as a “sanding event”. Such plugging events, and the subsequent loss of production and cleanup, can cost operators $2-5 million per incident. Currently, there are no reliable methods to prevent sanding events, and flowrate and pressure measurements in typical operations detect these events at a very late stage when pipeline plugging becomes hard or impossible to prevent. Recent developments of velocity profilers for non-intrusive flow measurement provide a new opportunity. Laboratory testing has shown promising results, however, pilot deployment in a mining environment has not been successful due to a lack of methodologies in addressing significant signal noise and interference from other instrumentations on site. There is also a lack in understanding of how to correlate flowrate measurements to sanding events and signatures or trends associated with sand bed formation on actual mining applications of hydrotransport pipelines. In this project, work on advanced processing, simulation, modeling, and predictive analytics is planned to bridge that gap and help us understand how sanding events develop. The present proposal is a multidisciplinary, collaborative effort between two technology access centres at the Northern Alberta Institute of Technology (NAIT) and Imperial Oil Ltd. The research aims to understand the mechanism of sand deposition and subsequent pipeline clogging through the large set of sensor data. A non-intrusive velocity profiler (SANDTracTM) will be deployed in a hydrotransport line at a mining site of Imperial Oil Ltd. Data obtained from the velocity profiler, along with real-time data of nearly fifty parameters of the flow process, will be used for rigorous data analytics and machine learning for predictive modeling of field-deployed data to better understand how sanding events develop. The project will benefit industry by reducing operation costs, minimizing process upsets, optimizing the hydrotransport process, and an overall reduction in greenhouse gas emissions.