Quantitative Analysis Of HAZOP For LNG Filling Station

LNG fillling stations are usually built on the edge of cities and countries which are in planning, and it will cause serious consequences once accidents happen. So it is important to take risk recognition technology to analysis the potential danger of LNG filling stations, and then puts forward effective solution to the danger to prevent accidents. Hazard and operability analysis (HAZOP) has the advantages of systematic, comprehensive, structured and detailed among many risk identification method. So the author use HAZOP for risk analysis of LNG filling atations.The HAZOP analysis technology has been videly used in various fields as a qualitative risk analysis method once the beginning, while it has a lot of lake. For instance, deviations that used in the process of HAZOP analysis have great ambiguity and uncertainty, and the results of HAZOP analysis are lack of the pertinences,even the staff can’t found the weak joints and risk rating of the system. Therefore, this article puts forward that HAZOP quantitative analysis should be used for the LNG stations’risk analysis.The HAZOP quantitative analysis method presented in this paper was carried out on the basis of qualitative analysis.So the principle and steps of HAZOP qualitative analysis was introduced at first. Then a LNG filling station in Sichuan was analysis as an example. In order to overcome the big disadvantage of the fuzziness of the deviation, a method to define a deviation critical value was proposed. The value could be defined according to the process parameters control targets of the LNG filling station. For the value that could not be directly determined through process indexes, it could be defined by other parameters through using the HYSYS software to dynamically simulate the impact on the system when the deviation occurred. To make the result of the HAZOP analysis to be targeted, and to make it clear that which links were weak and the main hazards of the LNG filling station, the article proposed to combine the risk matrix to determine the risk level of deviation. As the SDG (signed digraph) had similar reasoning mechanism with HAZOP, a natural link existed between the two methods, SDG and HAZOP was combined in this paper. First of all, the SDG model was drawn according to the process of LNG filling stations. Then, selected the node where the parameter deviation occurred as object and pull the parameter much more deviationist (deviated the critical value of the deviation), and found out the propagation path of the deviation. In this paper, the reason event probability was obtained by expert-based fuzzy method, and the probability of the propagation variables was calculated by SDG probability calculation rule based on Bayesian network, then the probability of the deviation occurred could be determined. Given the special nature of the process of LNG filling station (different processes contain similar equipment), the deviation of the HAZOP analysis could not be determined in a general value and the consequences of different deviations were different, so the deviation consequences in this paper were divided into two categories:the process consequences (impact the production process) and safety consequences (deviation caused accidents). When there was a security consequence, then considering the security consequence, and conducted security consequence severity evaluation. When the deviation was not enough to cause security consequence but lead to process consequences, then conducted process consequence severity evaluation. Reference to relevant regulations in domestic, combined with the scale and characteristics of LNG filling station, an accident consequence severity classification was established to determine the level of the deviation consequences.Finally, the author chosen a LNG filling station in Sichuan as a study object and carried out HAZOP quantitative analysis, identified that the overpressure of the LNG storage tank was the biggest potential hazard of this LNG filling station, which in accordance with the actual situation and it verified the practicality of HAZOP quantitative analysis proposed in this paper.