Sep-2024
Increase compressor reliability: Combine digital maintenance and AI-based monitoring (RI 2024)
Over the past decades, the reliability and performance of key components such as valves, rings, and packings for reciprocating compressors has improved significantly.
Gunther Machu and Magnus Terner
Hoerbiger
Viewed : 149
Article Summary
It is not uncommon, for instance, for valves to run for 16,000-32,000 hours without a failure. In many cases, this increase in component reliability translates into higher compressor uptime and availability. This extra uptime is beneficial to operations and production, as well as for scheduling repair and maintenance activities. The improvements have led to a situation where operators can, in principle, now run their machines for several years without stopping.
There is limited insight, however, on the health, performance, and remaining useful life of these components while the compressor is in operation. Because of this lack of transparency, most operators, still stick to scheduled maintenance, mostly following the original equipment manufacturer (OEM) recommendations. They are, therefore, unable to reap the full benefits of reliability, efficiency, and extended maintenance intervals made possible by the improved components.
For maintenance departments, replacing and repairing compressor components such as valves, packings, and cylinder rings is a necessity – and often a costly major intervention. The parts cost money to buy and store, and they can be hard to keep track of accurately. At best, they need to be replaced at scheduled maintenance intervals. At worst, they fail unexpectedly and cause downtime. Research by the European Forum for Reciprocating Compressors (EFRC) suggests that the most critical parts that fail on a reciprocating compressor include valves, packings, piston rings, and rider bands (Figure 1). Valves are projected to have a criticality that is 3.3 times higher than the criticality of a crankshaft.
To improve the reliability of a compressor, it therefore makes sense to focus on these components first, managing them intelligently and learning as much as possible from their operational history. In today’s world, that means a fully digital approach to tracking components in service: recording details of their performance, carrying out maintenance, repairing and refurbishing components, and managing spares inventory and unit and component history. The activities extend into using the component know-how and live data combined with artificial intelligence (AI) to make predictions about the future performance and failures of named components. Across the industry, there is a common understanding that digitalisation for maintenance activities is not only important but essential for successful business operations in the long run.
Digital management of repair and maintenance
Managing parts intelligently is key to increasing the efficiency of maintenance activities. Hoerbiger has developed a digital solution called VISTRA that streamlines the repair process and supports compressor maintenance activities.
VISTRA is a cloud-based system that helps operators proactively manage their reciprocating compressor spare parts inventory and maintenance processes. The system also provides important insights to help identify problems or ‘bad actor’ components more quickly and, hence, reduce compressor downtime. The foundation of this offering is the serialisation of every component (Figure 2).
Serialisation offers several important benefits. It enables the system to connect parts with compressor configurations, thus helping to guide the maintenance crew that is working on the compressor. It also collects data on where the parts have been used and for how long. This information is used to present key performance indicators to maintenance and reliability departments (Figure 3).
Customers of Hoerbiger in the natural and process gas markets across North America have deployed VISTRA for large reciprocating compressor fleets. Since the first onboarding in 2020, there are presently close to 1,000 units in the system, with more than 32,000 serialised components being tracked last year.
Customers report positive effects on stockholding: it is easier to find the right parts, but also to handle issues like obsolescence and installing the right parts configured for a particular unit. Additionally, tracking and collecting data around the repair of parts helps operators identify troublemakers and act swiftly. By analysing the data and looking at the mean time between repairs (MTBR) over different time periods, the system helps users take action by making informed decisions.
Benefits of a digital repair process
VISTRA offers multiple benefits for end users. It can eliminate maintenance errors, provide full transparency on inventory and fleet status, facilitate ordering of spare parts, and eliminate maintenance delays.
The system helps end users shorten repair times, maximise uptime, and save money by tracking down the exact parts when needed. It also presents key performance indicators (KPIs) such as MTBR, even down to the component level.
The benefits for the end user are the ability to run compressors for the optimum time, avoid unplanned downtime, plan maintenance activities, and ultimately reduce running costs.
Another benefit is the centralisation of repair and component records. Once installed, the system frequently replaces a scattered series of maintenance, repair, and inventory records that some operators maintain – the typical ‘Excel on the side’. In short, by using VISTRA, operators of reciprocating compressors can proactively manage potentially bad components and increase unit reliability, thus increasing compressor uptime.
Power of know-how combined with AI
Triggering increases in compressor efficiency and reliability is not limited to a single solution. A multitude of possibilities are available, and the rise of new technologies brings even more ways to support operators of reciprocating compressors. By utilising the data collected by VISTRA and combining it with sensor signals – both slow and fast – from the process, we can now use AI as a powerful tool to help understand both the present and future performance of compressors.
Hoerbiger is now working on a research project utilising ‘first-principles’ AI to build an intelligent system to monitor component lifetime and predict compressor conditions.
This will allow operators of reciprocating compressors to switch to condition-based maintenance while retaining a transparent insight into the safe and economic operation of their assets. The benefit of this approach is the combination of reciprocating compressor know-how with AI. Without a proper understanding of the complex workings of a reciprocating compressor, the AI cannot contribute to its full potential. The new Hoerbiger prediction tool will do just that, analysing the live data stream based on exceptions. If all is well, the customer will not be bothered. When there is an exception or an anomaly, they will be informed and guided into a three-step approach.
The first step is a heatmap of the system showing which cylinder is indicating a problem (Figure 4) and what is happening at the component level. The system shows a prediction of the component’s remaining lifetime, the probability of failure, and the time remaining to take action.
This short article originally appeared in the 2024 Refining India Newspaper, which you can VIEW HERE
Categories:
Add your rating:
Current Rating: 3