Ignacio E. Grossmann

Center for Advanced Process Decision-making

Department of Chemical Engineering

Carnegie Mellon University

Pittsburgh, 15217, USA

grossmann@cmu.edu

The oil and gas industry has traditionally relied on mathematical optimization, by starting early on using liner programming for optimizing the planning of refineries. Much progress since then has taken place in the development of mathematical programming techniques since they can currently handle discrete variables, nonlinearities, nonconvex functions and uncertainties. As we will show in this presentation is that this has led to an increase of the scope that can be handled in the oil and gas industry. We first describe mixed-integer optimization models for the design and planning of offshore oil and gas infrastructures, and their extension to account for uncertainties in the size and deliverability of reservoirs, which gives rise to challenging multistage programming problems. We next describe how mixed-integer nonlinear programming models can be developed for the multiperiod planning of refineries with crude sequencing, to the scheduling of crude oil deliveries to refineries, and to multiperiod blending of final products. Finally, we also show that design and operating decisions for shale gas production are amenable to modeling with mathematical programming techniques. For the former, we describe a nonconvex optimization model for the design and planning of supply chains for shale gas that includes multiwall pads, pipelines, and natural gas plants. For the latter we describe a scheduling model for optimal water management in the fracturing operations.