Picture this: You’re a new maintenance engineer in a petrochemical facility that contains vent gas compressors that are generally reliable, but do fail on occasion.
The system consists of three parallel compressors that have had a remarkable number of valve failures and, on occasion, some instrumentation connections have broken off, with piping that appears inadequate on both the suction and discharge. You questioned the staff about the operation and discovered that many studies have been performed and the design contains pulsation bottles on the suction and discharge to “snuff out” any acoustically driven vibration problems.
In this scenario, the typical pulsation studies have been conducted to meet API 618. Everything appears to have been done correctly, but the compressors are still failing. To solve a problem of this nature, one has to consider the following approach.
1. Historical review: It is important to look at the original design basis. As with the given problem above, the functional operation of the machine did not align with the analysis performed. Acoustical analysis is different from many forms of engineering analysis in that the observed response has to be evaluated based on the actual operating conditions. Sometimes, there is a misconception that a safety factor must be applied for the acoustical analysis, but that is not the case. Also, it is important the final piping configuration be evaluated based on how it is built in the field. The suction and discharge pulsation bottles must be acoustically tuned to the actual piping system for the best results.
2. Field study: A field study should be conducted on a multichannel system that measures vibration and dynamic pressure. It is important to place the pressure transducers in an optimal location for detecting pressure pulsation. If the location is at a node point in the domain, one may be unable to detect actual pulsation. In addition, the location of the dynamic pressure transducer may be too far from the flow field. It’s also imperative that the process conditions are recorded at the time of the test. Appropriate planning involving the proper placement of transducers can achieve the best results in the field study.
3. Structural dynamics/digital pulsation study: This study involves the development of a numerical model that duplicates the response in the field study. The idea is to recreate the conditions in the field and determine an effective solution. There are numerical models for acoustical response and mechanical structural response. Once the model is validated to match the field data, the appropriate solution can be determined based on a valid numerical model. Every process scenario should be evaluated, including the variable speed range. Generally, unless the supports are approximately 10-times stiffer than the piping, they should be considered part of the domain of the structural model. Within the acoustic model, it is equally important to determine whether the side branches for instrumentation are functioning as side branch resonators. In variable systems, the pulsation bottles are often designed to “snuff out” everything from the compressor, especially to avoid interaction problems.
4. Design review: The proposed design should be reviewed with plant team members from operations, process engineering, mechanical and controls. The design basis should be discussed and approved by the team with an exhaustive overview reviewing the design in detail.
5. Installation: Once the modification is installed in the field, measurements should be taken to validate the response.
The problems are not that remarkable with the right approach that adequately addresses the details. Unfortunately, the quick and easy study often proves to be invalid due to oversight of important details. The resulting analysis should be reviewed and approved by a professional engineer.
KnightHawk has evaluated and performed many field studies on reciprocating compressors. What we bring to the table is full field study capability, process, controls, mechanical, and metallurgical and materials analysis. In short, we are a one-stop shop. No need to retain additional contractors as we have full capability in all areas.
For more information, visit www.knighthawk.com or call (281) 282-9200.
