Multi-Compressor Capacity Optimization
Proceedings Publication Date
Presenter
Roy Milum
Presenter
Company
Author
Roy Milum
Part of the proceedings of
Abstract
In compressor stations where centrifugal compressors operate in parallel, operators are tasked with maximizing the station’s uptime and overall throughput with reduced personnel. In operating scenarios where parallel compressors are packing the pipeline, a stronger compressor can drive a weaker compressor towards a surge condition. Operation at or near a surge condition causes inefficient recycling or even unwanted compressor trips, thus reducing the station’s throughput flow in periods of high demand. In operating scenarios where the discharge pressure has reached capacity controller’s setpoint, it is desirable to operate both compressors at equal-distant to margins from surge while reducing their throughput flow. This type of operation allows both compressors to enter recycling simultaneously as the pipeline becomes packed and enables the operator to plan unit shutdowns based on the anticipated demand scenario. In the past, many capacity control systems have only addressed specific operating scenarios and did not properly manage capacity control requirement across multiple station condition. In these scenarios, station operators were required to utilize personnel to manually maintain each compressor’s share of the station load. A comprehensive capacity control system must enable both compressors to remain online during peak demands and provide a controlled capacity reduction as demand decreases. This technical paper describes the deployment of a centrifugal compressor capacity control system that manages the load sharing of parallel compressors of varying head and flow capacities in all compressor station conditions. This paper also provides an analysis of the cost savings realized by the operator and describes other benefits discovered during off peak operation.

To view the video or download the paper please register here for free

You already have access? Sign in now.