A closed-loop integration of scheduling and control for hydraulic fracturing using offset-free model predictive control

Abstract

As the development of shale gas resources has to be accelerated to satisfy the increasing global natural gas demand, many optimization approaches have been developed to handle the complexity associated with the shale gas system and discern the optimal design and operation strategy. However, it is worth mentioning that the essential hydraulic fracturing operation has not been considered as a dynamic process in the previous studies, which may lead to suboptimal solutions. Motivated by this consideration, we first develop an integrated model to simultaneously consider the scheduling and control of hydraulic fracturing operations to enhance the economic performance of the shale gas system. In particular, a reduced-order model is developed and integrated with the scheduling model to reduce the model complexity. To cope with the plant-model mismatch of the reduced-order model, we propose an online integrated framework with two feedback loops. In the outer loop, the integrated model is solved to determine the scheduling decisions and controller references. Then, an offset-free model predictive control system is designed to track the references online in the inner loop. By utilizing the offset-free model predictive control scheme, the tracking performance is enhanced and the plant-model mismatch is compensated for, which helps avoid dramatic changes in the solutions obtained by re-solving the integrated problem online. The effectiveness of the proposed online integrated framework is demonstrated by considering a hypothetical case study based on Marcellus Shale Play. It shows that the undesirable performance degradation induced by the plant-model mismatch can be decreased by the developed offset-free model predictive control system, and the overall economic performance of the shale gas system is improved with the proposed approach.