A new method to determine the optimum performance of natural gas processing plants has been developed. This methodology reduces the overall plant material balance equations into a linear form using the volatility of components and product specifications. Simulator response modeling relates key process variables to plant performance satisfying the remaining unknown information from the material balance equations. Rigorous economics are subsequently applied to the process model. This technique adequately combines contractual terms, product prices, and process information to calculate the optimum set of operating conditions for the plant offline. It is also a valuable tool to analyze the economic impact of processing additional streams and investigating new potential contract scenarios.

This paper describes an application of advanced multivariable control on a natural gas plant and compares its performance to the previous conventional feed-back control. This control algorithm utilizes simple models from existing plant data and/or plant tests to hold the process at the desired operating point in the presence of disturbances and changes in operating conditions. The control software is able to accomplish this due to effective handling of process variable interaction, constraint avoidance, and feed-forward of measured disturbances. The economic benefit of improved control lies in operating closer to the process constraints while avoiding significant violations. The South Texas facility where this controller was implemented experienced reduced variability in process conditions which increased liquids recovery because the plant was able to operate much closer to the customer specified impurity constraint. An additional benefit of this implementation of multivariable control is the ability to set performance criteria beyond simple setpoints, including process variable constraints, relative variable merit, and optimizing use of manipulated variables. The paper also details the control scheme applied to the complex turboexpander process and some of the safety features included to improve reliability.