Artificial Intelligence and Data Analysis Approach to Deliver 5800 BOPD from Cyclic Steam Stimulation Job in Krakatau Field

Abstract

Abstract Krakatau field was discovered in 1941 with an active steam flood operating in 1985. To maintain oil production, well intervention jobs become critical activities to perform at 6000 producers. Cyclic steam stimulation is one type of well intervention job that is a relatively low-cost and efficient method to improve well performance. The process involves injecting steam which allows oil heated by the steam to flow more easily. Started 2015, cyclic stimulation activity levels raised up from 50 jobs/month to 350 jobs/month. The type of cyclic stimulation performed is called short cyclic stimulation, in which the objective is cleaning the wellbore from asphalt, resin, and paraffin deposit. The review candidate process is done by Petroleum Engineer manually, which potentially delivers inefficiency and inconsistent processes among the Engineers. Having this condition, the team search for opportunities to deliver shorter cyclic time of reviews and improved job success rate by developing an Artificial Intelligence Tool using an Artificial Neural Network (ANN) algorithm. In 2022, cyclic jobs in Krakatau field successfully hit the highest execution at 525 jobs/month. The combination of an Artificial Intelligence Tool and visualization data tool in the form of Power BI for data analysis is very impactful to increase productivity and improve the job success ratio. The team will elaborate on the subsurface parameters to develop Artificial Intelligence and analyze data processes to deliver cyclic gain performance in level 5,800 BOPD.