Reciprocating machinery analysis has been around for more than 25 years and has traditionally been considered an art. Analysts are valued for their experience and their ability to look at the pressure curves and determine if there are issues with the machine, the root cause of a fault, and/or predicting when the machine should be removed from service. The collected performance data must be thoroughly analyzed in order to properly identify impending issues. Many in the industry agree, it takes skill to decipher the information, and can be a very time-consuming process.
Traditional performance data plots taken from reciprocating compressors are good for performance but are dependent on many factors. HolizTech LLC has introduced a patent-pending system designed to change all of that, moving condition monitoring from an art to a science by providing portable and online systems with built-in pattern recognition technology among other features that alarms the end-user to changes in the engine or compressor health.
“Reciprocating compressor valves are the heart of the machine, yet today, they are ignored by online monitoring systems,” said Troy Broussard, managing partner of Houston, USA, based HolizTech. “Nearly 75% of all unscheduled shutdowns on existing compressors can be associated with compressor valves and valve related issues. Valves fail for a variety of reasons and are the most critical part of a piston compressor. They are key to its overall availability. HolizTech was designed to help determine machinery condition before it impacts performance.”
The heart of the HolizTech system is a precise diagnostic “machine learning” software that analyzes actual engine/compressor mechanical conditions. HolizTech’s mechanical analysis module incorporates real-time valve diagnostics and modified algorithms which give an accurate and detailed mechanical analysis of reciprocating compressors, pumps, and engines.
“Our valve-specific monitoring approach provides detailed information about each valve and its dynamic operating condition,” said Michael Boken, managing principal at HolizTech and developer of the technology. “The pattern recognition has been proven to transfer from like machinery, making the big data and the industrial internet of things (IIoT) approach to monitoring a fleet of machinery possible today.
This article appears in the March 2019 issue of Gas Compression Magazine.
Want your own copy of Gas Compression Magazine delivered directly to you each and every month? Subscribe for free and choose print or digital delivery (or both!).
Free to subscribe. Free to renew. Don’t miss an issue of Gas Compression Magazine!