| In the oil sands industry, screening is a critical part of the mining process. Syncrude Canada Ltd., an oil sand company, uses screens to separate oversized lumps from the oil-rich sand before it enters the extraction process. However, under unknown screening conditions, some of the sand will pass over the screens, resulting in unexplained variations in screening performance. To investigate these variations, multiple linear regression is used on data from historical databases to identify water and geological variables that affect screening performance. A prediction model, developed using partial least squares regression, is compared to a simple linear model that uses only the oil sand feed rate. Results show an average 25 percent reduction in RMS error over a feed-rate-only model. This is the first known study to identify plant variables, other than the feed rate, that provide insight into oil sand screening behaviour. |
