Predict centrifugal compressor performance in off-design condition

摘要

Centrifugal compressors are vital components within process plants.The capability to detect early indicators of malfunctions in these machines,and to identify and understand their causes and possible remedies in a timely way helps minimize overall operational costs.This trend is known as predictive maintenance.For common centrifugal machines,predictive techniques are applied to the vibrational and structural dynamic aspects of machine rotor operation.On the operation floor,machine vibrational parameters are measured and compared to acceptable limits and used mainly for alarm triggering.Predictive maintenance for compressors.Predictive strategies based on performance analysis are a tool for diagnosing and evaluating machine health status during operation.For centrifugal pumps,for instance,the use of machine modeling to compare measured performance and design performance is relatively easy and straightforward.For centrifugal compressors,however,the same process is more complex due to the dependency of compressor performance on the gas mix composition and the operational inlet conditions(i.e.,pressure and temperature).An approach based on performance assessment for centrifugal compressors requires a more complex machine model.Analyzing the operating performance of centrifugal compressors requires the ability to adjust the compressor performance map to actual inlet conditions.Expected performance is described in graphs of discharge pressure,discharge temperature,polytropic head,efficiency and absorbed power.However,inlet conditions in the field maybe different from the specifications defined on machine data sheets.