A critical compressor system located in a refinery starts to log elevating running temperature trend whereby the inter-connecting shaft from the compressor to gearbox side has been observed experiencing higher intensively worn out as compared to shaft connecting from motor to gearbox side. This uneven worn-out on both ends of the gearbox raised an alert to the Operation and Safety department to plan for the next action.
To lower the running temperature of the compressor gearbox to a safe zone, the loading of the unit may need to be adjusted which will lower the output from the plant. However, the wear and tear of the compressor may still continue to generate excessive heat back to unsafe zone. Weighing the options to prolong the operation versus potential risk & safety concern, our client was eager to conduct a comprehensive investigation on the unit, targeting to turnaround the unit at the shortest time with the root causes identified and eradicated.
To optimise the check during shutdown period, several groups of subject method experts were being called upon to gather the statistic and information on the gearbox’s condition. Metal Machine was honoured to be engaged for conducting dimensional check on the gearbox housing to gather the GD&T data such as flatness, straightness and parallelism etc. API laser tracker was chosen for this job as it is battery operated and have smaller physical profile which enables easy adaption to site constraints.
The gearbox housing was dismantled and two survey sessions were performed on the upper and lower gearbox housing respectively. Data from both upper and lower housings were then studied and analysed using Spatial Analyser software to affirm that the GD&T are within limits and thus rule out the possibilities that the overheating is caused by any casing deformation.
Laser tracker is capable of capturing a large number of survey points efficiently on large objects with the required temperature compensation when conducting outdoor measurement. By incorporate Spatial Analyser software into the workflow to process the data, the whole process took less than 2 days to complete.