Managing hidden system threats for higher production regularity using intelligent technological solutions: A case study

Abstract

Identification and interpretation of hidden system threats on complex oil and gas production platforms has always been a challenge. These threats may gradually develop into failures/faults resulting in system shutdowns or eventually loss/reduction of production. Oil and gas industry is willing to test new technologies in managing uninterrupted, higher production regularity. In response to these challenges, a research project was initiated involving a leading oil company in Norway. A systematic investigative approach was adopted which incorporates domain experts‟ opinion and multiple information resources/databases. The paper attempts neural network modelling of a critical production loss-related scenario, based on real plant data from an offshore production facility. Analytical results captured symptoms of suboptimal performance from compressors installed in the gas compression system. This methodology could give plant operators an opportunity to early identify system‟s anomalies. As a result, unwanted shutdowns can be avoided, consequently improving overall plant‟s efficiency and productivity.