Optimum Maintenance Strategy Recommendation of Compressor Based on OREDA Failure Rate

Abstract

Grunnleggende vedlikeholdsintervall i olje- og gasselskaper er kanskje ikke så godt undervist. Som en kommende doktorgradsstudent er dette grunnleggende kalkulerende vedlikeholdsintervallet fra historiske data levert av OREDA ganske viktig før du går til det mer komplekse og prediktivt baserte vedlikeholdet. I denne skrivingen vil den elektrisk drevne skruekompressoren bli analysert slik at resultatet blir de numeriske verdiene for sannsynlighet for feil som bruker eller klient vil bestemme sin vedlikeholdsstrategi fra.

In the oil and gas company, usually there will be preventive maintenance activities that rely on the interval time like daily, weekly, monthly, etc. This makes the writer wonder on how do they decide on that. How do they see the criticality and the probability of the equipment so that they will have their own maintenance interval? From the datasheet that is available to be accessed, the writer decide to narrow down the equipment to electric driven screw compressor. This steps are usually done in the design phase of a project where the equipment is not yet running. This uses many historical data that are given on books and documentations. There are also some international standards that give the procedure and the equation how to calculate the needed probability of failure for this analysis. The calculation resolves around the failure rate given from historical data and also the alpha and beta values for each failure mode that is related to the equipment where each equipment can have more than one failure modes. Then, each failure mode has maintainable items that needed to be taken care of. This will allow the user/company/engineer to decide what main- tainable item should have a certain maintenance interval. Many maintainable items are listed on the result later, but one of the more important result of this writing is that the pressure and temperature instrument are really probable in terms of probability of failure. They needed to be maintained daily to remain at remote probability of failure category. Some of the results would be to proceed with the calculated maintenance strategy, and to focus to only one or two sample, control unit and pressure and temperature instrument are cho- sen, due to the several failure modes that control unit has and also the significant probability of failure that pressure and temperature instrument have. The decision would be decided by looking at the probability of failures table in which the failure probability with colour to the red is likely to be frequent.