SIR-based oil tanks leak detection method

Leakage from oil tanks occurs unexpectedly, which may lead to serious environmental threats and economic loss. There is increasing demand to develop a method that can detect leakage accurately with a low probability of false alarm and a high probability of detection. As one of the methods of measuring the quantity of oil in the tank, automatic tank gauges give a low performance on determining a threshold which can be used to identify the leakage accurately. In order to solve the problem mentioned above, a statistical inventory reconciliation (SIR) method is proposed for leak detection in this paper. The proposed method utilizes a sophisticated statistical calculation to analyze a long period of data in a host computer, and then presents a leak detection report with calculated leak rate, threshold, probability of false alarm and probability of detection clearly. The experimental results conducted using an induced leak in the tanks with different dimensions show the validity of the proposed method. Restricted by the precision of sensors, this method need more time to give a conclusive result in large tanks. A novel low-range differential pressure (LRDP) device with high precision can be used to shorten the minimal detectable time of this method, which is of great value in further research.