Human reliability incidence

Reinartz, S. (1989). Activities, Goals and Strategies Team Behavior during Simulated Nuclear Power Plant Incidents. In Conference Proceedings, Human Reliability in Nuclear Potver. London International Business Commurucations. 

Failure sequence modeling techniques such as fault tree analysis or event tree analysis are used to estimate die likelihood of incidents in facilities where historical data is unavailable, or is inadequate to accurately estimate die likelihood of the liazardous incidents of concern. Odier modeling tecluiiques may be required to consider die impact of external events (eardiquakes, floods, etc.), common cause failures, and liunian factors and human reliability. 

The committee s analysis was conducted on several levels. First, members investigated the causal factors for each of the seven events listed in Tables 2-1 and 2-2. They then developed a notional causal tree for each of the two events in Table 2-2 that were analyzed in depth. For illustrative purposes, a causal tree developed by the committee for the December 3-5,2000, incident at JACADS appears at Appendix F. The tree is a standard tool in reliability analysis and is particularly useful in human reliability analysis where operator actions contribute either positively or negatively to an incident. Lastly, the committee provides a series of general and specific observations about the events. 

A direct prediction of immunogenicity potential in humans (the incidence and/or severity) based on animal studies using a therapeutic dosing regimen with a biopharmaceutical (not designed to invoke a strong immune response) is very challenging. However, there is the need for a more reliable way to... 

In conducting the OCM risk assessments, it is important to consider previous incidents, upsets, and exercises to determine where significant risks have been identified. Human reliability needs to be considered in the risk assessment, particularly human failures that could result from issues mentioned in the checklists above. 

Analyses of specific incidents, especially when systematic and thorough, can illuminate systemic weaknesses and help us understand how things go wrong. We have seen how there is frequently a chain of events leading to an incident and a variety of contributing factors. Having understood these principles, we are now able to approach the examination of system weaknesses from a different perspective. Rather than take a case, analyse it and see where it leads us, an alternative approach is to begin with a process of care and systematically examine it for possible failure points. This is the province of human reliability analysis. 

Literature has many theories and concepts discussing human reliability and associated human error causal factors that always trigger incidents and accidents within safety-critical systems. The main - by definition - characteristics of such safety occurrences are their randonmess, rare predictability, sophisticated, yet vague sequence of propagation. Such characteristics can basically allow for the retrospective analysis of these occurrences and their causes at various sectors and levels within industry such that re-occurrence margins are reduced if not totally eliminated. The major drawback of such reactive treatment... 

In a PSA for a nuclear power plant (NPP), various models have to be applied to obtain predictions for the frequencies of damage states after undesired initiating events. Fault-tree and event-tree models account for possible events during an incident or accident scenario and the corresponding prohahihstic assessment physical models evaluate the thermal-hydraulic consequence and the release of radionuchdes for differing initial and boundary conditions and human reliability models account for human actions and their prob-abihstic assessment. Of course, there may be large uncertainties in the predictions of these models due to diverse uncertainty sources which are involved in their calculations. 

There have been a number of recent and well publicised accidents in which human error has played a prominent part. At Three Mile Island a combination of poor display design, bad maintenance and operator error led to the reactor core becoming uncovered and the release of radioactivity into the atmosphere. At the Ekofisk oil field incident a blowout preventer was installed upside down which resulted in an accident, following which widespread environmental pollution ensued. In both these cases the overall losses have been estimated in tens and hundreds of millions of pounds. These and similar incidents have brought sharply into focus the need to include an appraisal of human reliability and the factors which affect it when considering the safety performance of both large and small scale systems of work. ... 

Reliability and Efficiency—Identify and resolve facility, business work process, and human reliability and efficiency concerns that may cause significant incidents or performance gaps. 

The HFACS method was initially developed to avoid mishaps in naval aviation, and has later been applied in several other domains similar to the space domain. Hence, it seems well suited to ESA. However, it should be adapted to the space environment. It consists of a comprehensive analysis of previous accidents, also identifying lacks in barriers that are more remote from the incident itself, as management or supervision or organisational causes Thus, taxonomy such as HFACS could be adapted and implemented in order to have an empirical background to implement and work with tools such as Human Reliability Analysis (HRA). 

This paper indicates that the use of taxonomies from a bi-dimensional approach such as CREAM to classify data is a potential solution to produce meaningful information from three different types of source using the same framework (i) historical data, as demonstrated in this research, plus (ii) incident investigations and (ii) prospective analysis, as in the original application of CREAM HRA. The common framework to conduct human reliability predictions as well as retrospective analysis of events during Human Reliability Analysis in a specific facility or industry is perfectly able to interface with the proposed classification scheme for past accidents, considering that they basically share the same taxonomy. 

Discussion of and research into human error focusses on slips, mistakes, incidents and accidents and their internal and external mechanisms. Research on accident liability centers on work- or traffic-related accidents emphasizing personal variables as an explanation of the occurrence of accidents. Systems variables such as workplace design or work environment are not considered systematically. Modern human factors approaches examine the joint effects of personal, organizational and technical factors on human reliability, employing accident-independent as well as accident-centered safety analysis methods. 

The Chemical Process Industry (CPI) uses various quantitative and qualitative techniques to assess the reliability and risk of process equipment, process systems, and chemical manufacturing operations. These techniques identify the interactions of equipment, systems, and persons that have potentially undesirable consequences. In the case of reliability analyses, the undesirable consequences (e.g., plant shutdown, excessive downtime, or production of off-specification product) are those incidents which reduce system profitability through loss of production and increased maintenance costs. In the case of risk analyses, the primary concerns are human injuries, environmental impacts, and system damage caused by occurrence of fires, explosions, toxic material releases, and related hazards. Quantification of risk in terms of the severity of the consequences and the likelihood of occurrence provides the manager of the system with an important decisionmaking tool. By using the results of a quantitative risk analysis, we are better able to answer such questions as, Which of several candidate systems poses the least risk Are risk reduction modifications necessary and What modifications would be most effective in reducing risk ... 

However, this option presents some difficulties for radionuclides, because studies of radiation effects in human populations have focused on cancer fatalities as the measure of response and probability coefficients for radiation-induced cancer incidence have not yet been developed by ICRP or NCRP for use in radiation protection. Probabilities of cancer incidence in the Japanese atomic-bomb survivors have been obtained in recent studies (see Section 3.2.3.2), but probability coefficients for cancer incidence appropriate for use in radiation protection would need to take into account available data on cancer incidence rates from all causes in human populations of concern, which may not be as reliable as data on cancer fatalities. Thus, in effect, if incidence were used as the measure of stochastic response for radionuclides, the most technically defensible database on radiation effects in human populations available at the present time (the data on fatalities in the Japanese atomic-bomb survivors) would be given less weight in classifying waste. 

---