| Ammonia is widely used in many fields,such as chemical industry,light industry,chemical fertilizer,pharmaceutical industry,synthetic fiber and so on.With the demand for ammonia increasing,the ammonia production in China is also increasing.The ammonia is generally pressurized or cooled to a liquid ammonia to store,and normal pressure cooling is generally used for large storage tanks with more than 1000m3.Ammonia is a toxic,flammable and explosive gas.Once the ammonia leak in a large tank farm will cause serious consequences.So it is very important to control the risk of tank farm.Risk early warning is an effective way of risk control,that can forewarn potential danger,eliminate and control unsafe factors from the root,minimize the occurrence of accidents and reduce the losses caused by accidents,and ensure the safety of personnel and equipment in liquid ammonia tank areaAccording to the actual situation of the liquid ammonia tank,analyzing and comparing the advantages and disadvantages of fuzzy comprehensive evaluation model,grey relational analysis model,artificial neural network model and extension theory early warning model,the extension theory early-warning model is used to calculate the risk early warning in this paper.After analyzing the accident of liquid ammonia tank area,combining with field research and expert advice,a scientific and systematic early warning index system for liquid ammonia storage tank area was established,including 4 first level early warning indicators of human,machine,environment and management,and 25 two level early warning indicators.For the two level indicators ofhuman factors and management factors,because of its sensitivity to weak,the static weight of the combination of expert scoring and AHP is adopted.For the two tank level indicators of mechanical equipment and environmental factors,because of its high sensitivity,the dynamic weight of the combination of extension theory and AHP is adopted.On this basis,the risk early warning model of liquid ammonia tank area based on extension theory is established.The model divides each warning index into four grades:Grade Ⅰ,Ⅱ,Ⅲ and Ⅳ.It can not only calculate the early warning index of the tank farm,but also predict the indicators early,so as to take control measures.The model is applied to the liquid ammonia tank area of an enterprise,and the data collected from the field is input into the model calculation.It is concluded that the tank farm index is a II warning,and it can produce normally,but there are some risk factors.From a single indicator,we can see that the noise level of tank farm reaches the Ⅳ warning,the implementation rate of safety system and the proportion of full-time safety management has reached the Ⅲ warning.For these three indicators,corresponding measures should be taken to reduce the early warning level.Through the application of examples,it is verified that the early warning model established in this paper can forewarn the risk of liquid ammonia tank area,and provide a feasible theoretical basis for intelligent management of liquid ammonia tank area. |
