A probabilistic model for risk assessment and predicting the health risk of occupational exposure to pesticides in agriculture

Abstract

Introduction. The main point is the influence of a complex of chemical and physical stressors on agricultural machine operators. The processes of occurrence and interaction of harmful factors are probable. Markov processes are a convenient model that can describe the behaviour of physical processes with random dynamics. Purpose of the work: was to develop a probabilistic model of risk assessment for agriculture workers during the application of pesticides based on Markov processes’ theory and evaluate with the help of the developed model the probability of occurrence, the degree of severity and the prediction of the different influence of adverse factors on the operator. Materials and methods. The mechanized treatment of pesticide is presented in the form of a system, the states of which are ranked according to the degree of danger to the operator: from non-dangerous to dangerous. The transition occurs under the influence of negative factors and is characterized by the probability of pij transition. Based on the marked graph of the system states, a stochastic matrix P[ij] of transition probabilities was constructed in one step. There are formulas by which it is possible to calculate the state of systems in k steps for a homogeneous and non-homogeneous Markov chain. Results. Based on Markov chains’ theory, the system’s behaviour is modelled when using single-component preparations based on imidacloprid for rod spraying of field crops. Received vector of probabilities of possible hazardous conditions for the employee after each hour of spraying within 10 hours. After 6 hours of working, the probabilistic risk for the operator to stay in a non-dangerous state is about 50 %, and the probability risk of going into a dangerous - at 24 %. The stationary probability distribution results show the inevitability of the transition to a hazardous state of the system if enough steps have been taken. Conclusion. With this model, you can supplement the operator’s health risk assessment system, analyze, compare and summarize the results of years of research. The calculated statistical probabilities can be used in the development of new hygiene regulations with using pesticides.