Probabilistic approach to the investigation of the causes of emergencies at the gas pipeline facilities

When developing any technical process, there is a problem of its safety. The solution of these problems is an integral part of such a process. In this work we study the modeling of the dynamics of causeeffect relationships of the origin of accidents at the gas pipeline facility. Publications have been analysed on the solution of the security problem at the gas pipeline facilities, approaches to solve this problem, efficiency of practical implementation on the production site. Based on the analysis, positive aspects of the existing developments were taken into account, and also general significant shortcomings were revealed. This study proposes improvements of security solutions based on the probabilistic approach, allowing the use of quantitative and qualitative criteria in the algorithm of the prevention of security measures and obtaining the possibility of the prevention of emergency situations on gas pipeline objects. The obtained results can be adapted to other branches of production.


Introduction
Currently, in the leading organizations of oil and gas production and processing [1,2,3], the European API implements the life-cycle management system of the facility in accordance with the international standard API RP 581 (American Petroleum Institute). [4] The Risk Management is carried out in all phases of the life cycle. The Risk Based Inspection (RBI) methodology is widely used at the international level, especially in the oil and gas industry. [5,6] In the literature, several RBI projects are reported, demonstrating the strengths of the approach to risk management and optimization of inspection schedules. The API RP 581 is a well proven methodology for conducting risk-based inspections (RBIs) in the processing industry. The third edition of the standard was published only in April 2016.
The article proposes an expanded solution to security problems based on the probabilistic approach, which makes it possible to use quantitative criteria in preventing security measures and obtaining the possibility of prevention of emergency situations.
Presented, improved, the system ensures the effective operation of facilities and resources in conditions of limited financial resources, where the main goal is to achieve a balance between costs, risks and required asset performance. [7] The procedure for improving security measures entail the possibility of prevention of emergencies at a hazardous production facility (hereinafter referred to as "GRO"). The safety of the GRO is related to the risk of an emergency. In order to prevent, in the future, to forecast emergency situations at the site, it is necessary to learn how to manage risks.
However, the importance of risk from the point of view of management is not determined by one parameter, which is related to its probabilistic nature. Obviously, the risk, which in the case of implementation carries great losses, can be considered dangerous and requiring management. But if the probability of this risk is extremely small, then it can be neglected. Accordingly, and vice versa: risk with a small potential loss, but realized very often, will lead eventually to a significant total damage. Accordingly, it is necessary to characterize each specific risk with the help of two of its main parameters: the probability of occurrence and the impact of possible damage.
Note that although the effects of risks are not only financial, but also causing loss of life and health, the European systems accept financial, material as the main ones. This is due to the fact that in most cases non-financial losses can be expressed, although with a certain degree of conventionality, in value terms, and also because in the economic activities these kinds of losses are of the greatest importance.

Methodology
For the first time in Russia, the concept of risk-oriented management of oil, gas and related production and processing facilities appeared at the end of the XIX century. Various methods are used to assess security risks, including statistical analysis, analysis of species and consequences of failures, analysis of the tree of events, and others. As a data source, for statistical analysis of accidents on the gas pipeline, information was used. [8] As a criterion for the acceptability of risk, the ALARP principle is applied (As Low As Reasonably Practicable -the risk is so low as it is reasonable). [9] The acceptable level of risk in accordance with the principle of ALARP is a level of risk for which the costs of achieving it are economically justified. The results of the risk assessment are presented in the form of a risk matrix, which is constructed in accordance with the methodology. [4] Table 1 presents the types of possible accidents at the pipeline with the probability of occurrence and damage for the period 2014-2017. In Fig. 1 and Fig. 2 static information on accidents of various types in the period 2014-2017is presented, formed on the basis of the reporting documentation. [8,7]   In accordance with the statistical data on accidents on the gas pipeline for the period 2014-2017, Table 1, Fig. 1, Fig. 2, we form the risk matrix of Fig. 3. [11] Risk matrix -a picture of risks in the form of a table, where the columns show the magnitude of the damage from the implementation of risks, and by lines -the gradation of the probabilities (or frequencies) of their implementation. Each cell of the matrix contains information about the probability of this event and the amount of possible financial damage. The results of the analysis and risk assessment are ranked by importance in order to establish the optimal balance between costs, risks and performance indicators. Risk matrix which a certain number of rows and columns, and ultimately consists of the areas of four categories: 1) a low risk area (the probability value is from 0 to 0.1, the damage is from 0 rubles to 1000000000 rubles); 2) the area of an average risk (probability from 0.1 to 0.4, the amount of damage from 1000000000 rubles to 6000000000 rubles); 3) high risk area (probability value from 0.4 to 0.7, the amount of damage from 6000000000 rubles to 2 billion rubles); 4) area of extremely high risk (the probability value is from 0.7 to 1, the amount of damage from 20000000000 rubles and above). It should be noted that the presented risk matrix, depending on the time period, can undergo changes.