|
|
Dr. Ignacio E. Grossmann Center for Advanced Process Decision-making Department of Chemical Engineering Carnegie Mellon University Pittsburgh, PA, 15213, USA |
Abstract |
|
Natural gas is an abundant energy source and the cleanest-burning fossil fuel. It is extracted from dense shale rock formations, and has become the fastest-growing fuel that could become a significant new global energy source. In addition, shale gas contains significant amounts of light hydrocarbons (e.g. ethane, propane, butane) providing lower cost raw materials to chemical industry. The long-term planning and development of the shale gas supply chain network around each play is a very relevant problem, that to the best of our knowledge, has not been addressed before through the use of advanced and comprehensive optimization models and tools. An additional critical aspect in shale gas production is water management. Shale gas production is a highly water-intensive process, with a typical well requiring between 3-5 million gallons of water to drill and fracture. The vast majority of this water is used during the fracturing process, with large volumes of water pumped into the well with sand and chemicals to facilitate the extraction of the gas. Typically only 15-30% of this water returns as flowback water. Therefore, a long-term planning model for the development of shale gas fields should account for water availability. Furthermore, given the short periods of time in which very large amounts of water are used in the fracking of shale gas (e.g. 3 months), the scheduling of the fracturing of the wells and coordination of the logistics for transporting the water is a very important problem. This presentation provides an overview of recent optimization models for shale gas production. We first describe a new mixed-integer optimization model for the design of shale gas infrastructures. The model is aimed at optimizing the selection of the number of wells to drill, size and location of new gas processing plants, location and length of pipelines for gathering raw gas, delivering dry gas, and natural gas liquids, location and power of gas compressors, and planning of freshwater consumption from available reservoirs for well drilling and fracturing. The goal of this model is to maximize the net present value. As will be shown this problem gives rise to a nonconvex large scale mixed-integer nonlinear programming for which we have developed a Branch-Refine-Optimize (BRO) algorithm that is based on successive MILP piece-wise linearizations that provide increasingly tighter upper bounds on the net present value (NPV) that are coupled with MINLP subproblems in which the goal is to determine the drilling strategy and pipeline diameters for the predicted supply chain structure, as well as a lower bound for the NPV. We also describe a detailed operational model to optimize water use life cycle for well pads. The objective is to minimize transportation cost, treatment cost, freshwater cost, and additional infrastructure cost while also accounting for the credit of the production of shale gas within the specified time horizon. This time horizon must be at least one year to capture the seasonal availability of water. Assuming that freshwater sources, river withdrawal data, location of well pads and treatment facilities are given, the goal is to determine an optimal fracturing schedule, recycling ratio, additional impoundment capacity, and treatment unit installation. We show that this problem can be formulated as a mixed-integer linear programming (MILP) based on a discrete-time representation that can be solved with reasonable computational expense. Results with both models for long-term infrastructure design and short-term scheduling of water management show the importance of applying optimization tools to these problems. |
 |
 |
 |
 |
 |
 |
 |
 |