Real time evaluation of sanding potential

More than 70% of the world oil and gas are domiciled in unconsolidated reservoir rocks with a high risk of sand production. To manage the sanding problem it is important to answer such questions as "will the reservoir fail?", "when will it fail?", "what volume of sand will it produce?" and "where will the sand come from - from the sand face or from the reservoir rock matrix?" In order to predict whether sand production will be problematic, the Sand Management Research Team at The Robert Gordon University has developed a three-fold approach the details of which are presented in this paper. Firstly, two alternative methodologies have been adopted for evaluating the failure potential of reservoir sand. Secondly, an estimate of sand production volumes and rates as a function of reservoir management policy, operational procedures and time is required with the caveat that sand failure is not equal to sand production. Thirdly, it is operationally optimal to have this done in real time, and over the life cycle of the field. A critical issue for accurate sanding quantification is the precise prediction of stress magnitudes and orientation with varying reservoir pressure depletion. The Mohr-Coulomb failure envelope is deployed in this tool to model rock failure based on the critical porosity concept. In the alternate approach, Neural networks have consequently been used to develop models for predicting both UCS and grain size distribution in real time. Critical model outputs include grain size distribution of failed sand, sand mass produced, maximum sand rate and time to first sand production, drawdown and depletion at first sand production. The pinnacle of this model is the derivation of a SAND PRODUCTION FORECAST. The real-time predictions help Sand Management teams develop optimal design and optimal productions strategies in relations to when and where to initiate sand control, effective topside management and real-time management of reservoirs with sanding problems.