| Enterprise safety management innovation and research has been accompanied by advances in industrial technology, industrial structure adjustment, population and labor force changes and evolution, as well as subtle changes in social relations and the times. From research on facilities to workers who operate the machine, then human-machine-environment based on systematic point of view, currently enterprise safety management showing characteristic of the times of micro particles, real-time, ubiquitous technology, easy to use and so on, that has a close relationship with cloud computing, uncertainty calculation and big data technologies. These features and times are calling us the key that must be solved in the modern enterprise safety management is:how to use the network formed by information and information technology to precisely regulate substance of "people","machine","environment", information and energy in enterprise safety management.Offshore oil industry is a high input, high-risk, high-tech industry, decision-making and operational mistakes may cause huge economic and social losses. Because of particularity of natural environment, production processes, work space and complexities of the job content, the HSE risk in offshore oil industry has differences, severity, diversity, variability, hidden features. In this particular context, unsafe behaviors among offshore oil industry employees take on what kind of surging rules and characteristics, how to prevent and regulate unsafe behaviors among offshore oil industry employees, how to ensure the lives and health of majority employees fighting at home and abroad on offshore platforms is glorious mission of safety manager and researchers engaged in safety science and technology. Based on this view, unsafe behavior characteristics of employees among offshore oil industry is taken as topic, closely integrated safety management trends, safety science as the theoretical basis, around the actual operating characteristics of offshore oil industry, analytical methods of the outside to the inside are used to launch a series of researches on unsafe behavior characteristics of employees among offshore oil industry. Finally oriented big data’s employee unsafe behavior tendentious pre-control system is put forward, which plays a reference function on innovating and improving safety technology and management level of offshore oil industry.From four perspectives, according to frame mode of the outside to inside, around the reality characteristics of offshore oil industry, the paper carries out a series of studies rely on environment impact on safe operation and dependency of safety production accidents associated with human-machine-environment, staff unsafe behavior subjective, tendency of unsafe behavior caused by objective stress. The main contents are as follows:1. Due to the offshore oil production facilities long setting in waters without cover, suffering from typhoons, tropical storms, hurricanes and other severe and complex special marine environment, furthermore waves, flow, etc. caused by extreme sea conditions creating enormous damage on offshore platforms, thus, the marine meteorological environment has been a important factor that affects safety issues of offshore oil industry. Doing research into typical safe operation on offshore platforms to discuss the influence that marine meteorological environment on platform device unit and safe operation of offshore oil production facilities, thus fuzzy inference of rules assessment model are refined to construct fuzzy inference framework. Multi-level assessment and decision-making ideas as well as fuzzy logic inference system are used to establish three fuzzy inference evaluation model of meteorology-environment-performance that affects safe operation of offshore oil production facilities. Simulation experiment testing has proved that evaluation results are basically consistent with the actual. The fuzzy inference evaluation model provides new ideas and techniques for employee safety operational performance and efficiency of equipment and facilities under a variety of complex weather conditions of marine environment, which has certain theoretical value on protecting employee safety and enhancing equipment and facilities integrity management.2. On the basis of statistical analysis of safety production accident in China’s offshore oil. from practical point of view to improve collection, storage, transmission mode of safety production accident data, the accident analysis of safety science as theoretical basis, descriptive and inference methods as well as association rules Apriori algorithm are used for data mining, which has conducted two aspects of research about analysis techniques of safety production accident, one is correlation study on safety production accidents with different times, the other is correlation between category of safety production accidents and human, machine, environment factors.3. The paper discusses occurrence time distribution of unsafe behavior, distribution of daily number of unsafe behavior in detail. These statistical results show that data from different employees in these areas have basically same statistical characteristics, namely unsafe behavior of employees is in the presence of various forms of non-Poisson distribution, including distribution of interval time, frequency of unsafe behavior that are showing scaling characteristics. These distributions not only involve time statistical characteristics, but also relate to statistical properties of occurred intensity, which having carried out analysis on these rules of statistics properties, at the same time comparing with other human behavior unsafe behavior of employees are found with certain particularity.4. In the work process of offshore oil production facilities, the presence of extreme natural environment and restrictions on the level of existing technological processes make the work on offshore drilling platforms especially difficult. Despite the provision of adequate safety measures, safety and health of employees are still in jeopardy. Based on "demand-control-support" job stress model by Karask and Johnson JV, expert assessment questionnaire method is used to carry out feasibility study on unsafe behavior under certain circumstances due to degrees of interference of physical and mental aspects among offshore drilling workers generated permanent pressure. On the basis of operating risks on C platform of an oil field in Bohai Bay caused objective and subjective psychological pressure of employees, a pressure tendency analysis about unsafe behavior is carried out. By a fall accident of drilling workers on the platform to analyze the limitations of six major safety tools that current use in the team’s, thus tendencies pre-control system of staff unsafe behavior on offshore platform is proposed for large data.An engineering company of offshore oil industry as an example, the paper has a research on characteristics of unsafe behavior of offshore oil industry workers. Main contents and results are as follows:1. Objective assessment of Meteorological environment safeguards is important prerequisite for offshore oil exploration and development, improving safety operation performance. In recent years, theoretical research and practice summary of offshore oil production facilities clearly state that adequate information acquisition of marine meteorological environment, accurate information analysis processing, optimized assessment decision support have become an assistor for safety operation performance. Conducting an in-depth study on typical safe operation of offshore oil production facilities and profoundly discussing effect of marine meteorological environment, equipment and facilities unit of offshore oil production facilities and safe operation, thus refine fuzzy inference rules of assessment model and construct fuzzy inference framework. The idea of multi-level assessment and decision-making, fuzzy logic inference system are adopted to establish three fuzzy inference evaluation model of meteorological environment affecting safe operation of offshore oil production facilities. Simulation test on specific offshore oil platforms of Bohai Sea, East China Sea, South China Sea shows that the assessment results basically are consistent with objective reality, which proves validity of the evaluation model and feasibility of the technical approach. The fuzzy inference evaluation model provides new theoretical basis and technical support for assessing employee safety operational performance under a variety of complex weather conditions and energy efficiency of equipments and facilities, it will help protect employee safety and upgrade integrity management level of equipments and facilities.2. On the basis of statistical data of safety production accidents occurred2004-2011in a company of Offshore oil production facilities, according to principles of system engineering, the method of mathematical statistics application is applied to statistical analysis for the company’s production safety situation, which focuses on three reasons of "person","machine","environment" related to safety production accidents conducting a detailed statistical analysis and demonstrates the safety situation of offshore oil production facilities. These conclusions are:(1) Safety production accidents shows some regularity from the perspective of time of occurrence that the maximum number of accident in2008, the minimum in2004in the company. The company has consistently implemented Production Safety Law that was enacted in2003, thus safety management is enhanced and employee safety awareness is improved. Employee safety awareness has declined over time, aggravating task making a sudden increase in production safety accidents in2008, but the reduction of production safety accidents in2009and beyond shows the company has adjusted safety management ideas to strengthen safety management in2008. Production safety accidents of the company are less in February, August, October in a year, they are more in the3rd, the16th and the24th in each month.(2) From the perspective of hours of work and training time, tasks have grew in an effective and reasonable way while production safety accidents have also been effectively controlled ensuing seven years after the state promulgated Production Safety Law in2003. Meanwhile, safety management shows systematization, pertinence with the ideology of safety manager first enhanced, so to active suppression of production safety accidents based on safety training present thriving trend, production safety accidents decreased year by year in the case of without increasing amount of safety training especially after2006. It is explained that the current safety management is relatively effective.(3) Mechanical damage with the largest proportion as example, the law of its occurrence is that mechanical injuries occur more frequently in March, April, May, June, July, especially2,3,14,15, 21,22,26in a month.(4) Incident categories and hazardous factors of people, machines have a close relationship according to correlation between accident categories and hazardous factors. Accident categories have not significantly close ties with environmental factors. By Apriori association analysis can get to know that "mechanical injury" has strong association with human’s "operator error", objects "protection flawed", objects "defective tools accessories, equipment and facilities" respectively. The analysis on correlations between accident categories and factors of accident investigation indicates that accident categories are closely related to human unsafe behavior, physical insafety and environmental insafety. By Apriori association analysis can get to know that "mechanical injury" has strong association with lack or defective of protective safety devices and lights, defective tools accessories, equipment and facilities of physical insafety respectively.(5) Man, machine and environment are analyzed on the basis of multiple mechanical injuries and fall accidents of Offshore Oil Company, meanwhile control measures aiming at identification of hazardous factors and man, machine and environment status are put forward.3. Based on non-Poisson characteristics of time statistical properties of human behavior, from an empirical perspective, the paper carries on a comparative study about statistical regularities of a variety of random behavior in human activities and unsafe behavior in the industrial production in time and other aspects, it also explores law of dynamics of employees unsafe behavior, which provide a theoretical basis for effectively preventing various accidents caused by unsafe behavior.(1) In the case of distinguishing any details about data of unsafe behavior, examining its overall statistical characteristics can obtain overall unsafe behavior of staff during the operation. Time interval distribution of unsafe behavior occurred and frequency distribution of daily unsafe behavior are discussed in detail,,the results show that data from different employees in these areas have basically the same statistical characteristics, namely various distributions have obvious fat-tailed.(2) Research is conducted at the individual level due to the small sample size; even so, common characteristic coming from statistical results of unsafe behavior of different employees still reflects a quantity of universal properties of unsafe behavior to a certain extent.(3) Time intervals distribution among similar events is one object of study what human dynamics focuses on. The Paper discusses time distribution of three employees occurring unsafe behavior, which is close to a power law distribution of P(τ)∝τ-α.In the curve, it is obtained that time intervals of unsafe behavior in successive around10days has a small climax, and the occurrence of employees unsafe behaviors have proved accuracy of ebbinghaus forgetting curve.(4) The research finds that various forms of non-Poisson distribution in the presence of employees’ unsafe behavior, including interval time of unsafe behavior, frequency distribution of unsafe behavior showing scaling property, these distributions not only involve time statistical characteristics, but also involve strength statistical properties. These statistical property suggest unsafe behavior of employees do not occur random. Overall, these distributions imply apparent clustering relations between human characteristics and behaviors, which have similar vitality with undesired unsafe behavior of employee. Obviously, users groups watch online depending on movie content and interest, followed by inference, employees are prefer to unsafe behavior more depends on job content and job characteristics. Speaking from the individual level of employee, statistical properties between unsafe behavior of individual employees and SMS communication behavior have obvious similarities, it is not apparent statistical bias comparing to other human behavior. So we can know that other human behavior has the same preferences, it can be considered that unsafe behavior of employees cannot change absolutely only relatively by safety system, safety education, safety training, consequently management of employees’unsafe behavior should focus on controlling the serious consequences caused by unsafe behavior rather than simply controlling the absolute number of unsafe behavior.This article firstly takes operating activities process on C platform of an oilfield in Bohai Bay as an example, then it carries on an analysis about major risk factors affect employees that including changes of wells data, variability of sea conditions and other natural conditions, non-fixity of workplace, work space limitations, hazardous substances, the risk of operating process and so on. On the basis of this,48staffs from a team on C platform of an oilfield in Bohai Bay adopt expert assessment questionnaire method to proceeding subjective and objective stress analysis, based on job stress model of "demand-control-support" by Karask and Johnson JV to carry out feasibility study of occurrence of offshore drilling workers’ unsafe behavior under certain circumstances due to different degrees of physical and mental interference. By analyzing limitations of the team’s six safety tools that current use rely on a fall accident of drilling workers on the platform, it proposes tendencies pre-control system of staff unsafe behavior on offshore platforms for large data.(1)Objective stress analysis apply job evaluation questionnaire by B. Dudek to getting higher pressure on unpleasant working conditions, job complexity, danger, conflict and responsibilities on the drilling platform. This platform organization does not attach importance to the uncertainty arising from work, rush, physical and competition. Objective pressure is higher according to expert score of objective working conditions on the drilling platform.Psychological threat may be directly or indirectly affect the mental and physical health by the pressure, and then affect safety operation. Employees feel pressure when request do not meet their knowledge and abilities or their needs in work-related case. All subjective psychological pressure values of employees on the platform oscillate around the mean. Staffs on the platform feel more significant pressure caused by they are eager to gain support from co-worker and superiors, they feel more pressure in which involved in uncharted territory and lack of opposition role of unsafe behavior during operation.(2)Correlation analysis between individual stress factors of employees on the platform consist of need, control, support and tendency of unsafe behavior on C platform of an oilfield in Bohai Bay shows that the staff is more demand for remains character strong and cheerful in long-term sea platform life along with have someone and a certain procedure to stop unsafe behavior in demand; in control, staff makes him not concentrate at work because he has difficulty in cognitive control thus minimizes job seriously; in support, the staffs attention is significantly influenced by the common operator’s support. Due to the role of three factor of demand-control-support towards driller produces some work pressure, so tendency of unsafe behavior is more apparent.(3)The fall accident of a driller who is Mr.Tian triggered the thinking that tendentious control measures of unsafe behavior are inadequate on the platform, thus deficiencies of six tools including JSA job safety analysis, BBS observable behavior, STOP authorization, prompt cards of "five thought five quit", work permits are analyzed current.(4)The operator is subject, the target is to prevent accidents in accident prevention system of offshore platform operating in the theory of unsafe behavior tendencies, adaptability screening for platform operating, safety quality training and behavioral defects corrected, high-risk environment controlled have constituted control system of tendency comprehensive preventing unsafe behavior. The system is mainly based on "perceived behavioral state, abnormal detected", real-time data mining as a means in big data environment, it provide comprehensive, timely, accurate and dynamic safety information for enterprise employees at all levels in order to enterprise employees at all levels conduct accurate decisions on the job. |
PDF Full Text Request