RELIABILITY ANALYSIS OF CRUDE OIL PIPELINE WITH API 5L GRADE B STEEL MATERIAL: COMPARISON OF MONTE CARLO AND API RP 579 FITNESS FOR SERVICE METHOD

• Main Author: HUNTADI NUGRAHA (NIM 12104033), ALFONSUS
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/9879
• Institution: Institut Teknologi Bandung

The Duri-Dumai crude oil pipelines with API 5L grade B steel material 30 and 36 inch in diameter, had been operating since 1970 and 1958. The operation age had caused both pipelines 50 kilometers each in length, suffers quality degradation, due to external corrosion on pipe wall. To determine whether both pipeline can be operate safely for some intended periods, reliability analysis is needed. The common reliability analysis methods are probabilistic and deterministic method.<p>The comparison between both reliability analysis methods: probabilistic with "Monte Carlo Simulation", and deterministic with API RP 579 "Fitness-for-Service" standard, carried out on both pipelines. Both analysis methods used data gathered from in-line inspection (ILI) by intelligent pigging, which carried out in 2006 and adopting magnetic fluks leakage (MFL) methods. The Monte Carlo simulation is based on statistical approach with iteration that evaluates limit state function or condition where failure is expected to happen. Failure on both pipelines is expected to happen due to wall thinning by corrosion, which leads to leaking. Evaluation of the limit state function is based on random variables that Extremely Value (Gumbel) distributed. Some locations on both pipelines with highest probability of failure based on Monte Carlo simulation analysis are being analyzed furthermore by Fitness for Service method. The Fitness for Service method assesses the pipeline reliability by evaluates pipe wall materials remaining strength that had been corroded. There are two evaluation steps carried out by Fitness for Service method, they are level 1 and level 2 evaluation. Each of those evaluation levels has specific data requirements and analysis conservatism.<p>The remaining life prediction by Monte Carlo simulation on 30 inch pipeline is 36 years after ILI, and 27 years after ILI for 36 inch pipeline. In the other hand, remaining life prediction by Fitness for Service method on 30 inch pipeline is 12-20 years after ILI, and on 36 inch pipeline is 20 – 30 years after ILI. The minimum remaining life of both pipelines are 20 years after ILI for 36 inch pipeline on km 10,8-10,89, and 12 years after ILI for 30 inch pipeline on km 8,4-8,49. The remaining life prediction differences between those methods are caused by: the limit state function, the undetected part due to sleeves, the rate of corrosion, and the corroded part dimensions. The Fitness for Service method has higher analysis accuracy and more data requirements and calculations than Monte Carlo simulation method. Monte Carlo simulation method is suitable for screening step, that evaluate probability of failure from the whole pipeline, before the detail and accurate analysis, Fitness for Service method, carried out. <br /> <br />