Complex system failure

There are a variety of ways to express absolute QRA results. Absolute frequency results are estimates of the statistical likelihood of an accident occurring. Table 3 contains examples of typical statements of absolute frequency estimates. These estimates for complex system failures are usually synthesized using basic equipment failure and operator error data. Depending upon the availability, specificity, and quality of failure data, the estimates may have considerable statistical uncertainty (e.g., factors of 10 or more because of uncertainties in the input data alone). When reporting single-point estimates or best estimates of the expected frequency of rare events (i.e., events not expected to occur within the operating life of a plant), analysts sometimes provide a measure of the sensitivity of the results arising from data uncertainties. 

Perry SJ, Wears RL, Cook RI. 2005. The role of automation in complex system failures. J Patient Saf 1 56. 

FIGURE 10.1. COMPLEX SYSTEM FAILURE CYCLE OF ERRORS... 

Cook, R. I. Brief Look at the New Look in Complex System Failure, Error, and Safety. Chicago Cognitive technologies Laboratory, Department of Anesthesia and Critical Care,... 

APPENDIX TEN. A BRIEF LOOK AT THE NEW LOOK IN COMPLEX SYSTEM FAILURE, ERROR, AND SAFETY... 

A brief look at the New Look In complex system failure, error, safbty, and resilience... 

Control by flooding of condensate introduces another variable — change in heat transfer area. Condensate flooding also reduces the available pressure drop because the hydrostatic head is reduced. Again, the entire mode of operation is altered. Also, condensate flooding often results in improper venting — which adds another variable to a complex system. Failure to vent inerts may lead to corrosion, especially with steam. 

Johnson, S.B. 2004. White Paper on Engineering Culture and Complex System Failure. Space Studies DepartmenL University of North Dakota, June 3. 

On complex systems, which are repaired as they fail and placed back in service, the time between system failures can be reasonably well modeled by the exponential distribution (14,15). 

The event" list, across the top of the event tree, specifies events for which the probability of failure (or success) must be specified to obtain the branching probabilities of the event tree. Events that are the failure of a complex system may require fault tree or equivalent methods to calculate the branching probability using component probabilities. In some cases, the branching probability may be obtained directly from failure rate data suitably conditioned for applicability, environment and system interactions. 

Tliis cliapter is concerned willi special probability distributions and tecliniques used in calculations of reliability and risk. Tlieorems and basic concepts of probability presented in Cliapter 19 are applied to llie determination of llie reliability of complex systems in terms of tlie reliabilities of their components. Tlie relationship between reliability and failure rate is explored in detail. Special probability distributions for failure time are discussed. Tlie chapter concludes with a consideration of fault tree analysis and event tree analysis, two special teclmiques lliat figure prominently in hazard analysis and llie evaluation of risk. 

Analysis of Esso Longford as well as analysis in the UK Health and Safety Executive (HSE) investigation report into petrochemical complex major incidents all show that common underlying causes are often repeated. The Longford incident clearly illustrates the multiple root cause concept. A number of PSM system failures occurred either in... 

While the Challenger disaster was not a process incident in the strictest sense, the nature of the failure was similar to many piping system failures that typically occur in the process industries. More importantly, organizational failure was a fundamental cause of the incident. This case study serves as a classic example of the type of loss that can occur in a large complex organization if management systems are not effective. 

There are thousands of sets of signal proteins, like insulin and its receptor. These systems of paired protein-control molecules and their receptors have evolved over millions of years and provide the very precise orchestration of thousands of different chemical reactions that are required to keep our bodies alive and working. Although illness that develops because of a failure in these systems may seem like a terrible betrayal of how things should work, the fact that so many complex systems in so many... 

The positive therapeutic effect of the complex treatment involving Ovosorb was achieved in 405 patients and required 2-7 hemosorption sessions. In 46 patients involution of the destructive process in the pancreas was observed after a single procedure of hemosorption. It should be noted that in the process of complex therapy there were no cases in which signs of acute organ system failure appeared... 

The hardest part of engineering risk assessment has turned out to be the prediction of the modes of failure. Serious accidents at nuclear installations, such as those at Three Mile Island or at Chernobyl, have been caused by modes of failure that had not been analysed at all. For example, the report of the Presidents Commission on the Accident at Three Mile Island (Presidents Commission, 1979, p9) highlighted that the concentration of the assessment process on more obvious large break scenarios meant that the eventual mode of failure, which was a result of a chain of a number of more minor events, was not even considered. Despite the use of significant resources in the design process, the risk assessment had been unable to characterize the complex system adequately, a system that was totally human-made and defined. In particular, the risk assessment process had not been able to identify modes of failure caused by humans involved in the operations of the reactor behaving in unexpected ways. 

Sir Samuel F. Edwards (Cavendish Laboratory. University of Cambridge noted (1987). "Liquids are everywhere in our lives, in scientific studies and in our everyday existence. The study of their properties, in terms of the molecules of which they arc made, has been the graveyard or many theories put forward by physicists and chemises, Hie modern student of liquids places his laith in Hie computer, and simulates molecular motion with notable success, but this still leaves a void where simple equations should exist, as are available for gases and solids. There is a powerful reason for the failure ol analytical studies of liquids, i.e.. the difficulty experienced in rinding simple equations for simple liquids. We can explain the origin of the trouble and show lhai it docs not apply lo wlul at first might seem a much more Complex system, that of polymer liquids where, instead of molecules like HjO or C(,H(,. one has systems of molecules like H lCHi)iu no or H (CHC H(,i .ni i which behave like sticky jellies and yet have complex properties that can he predicted successfully. ... 

FMEA is a quantitative risk analysis for complex systems (Fig. 6). As this approach involves assessment of occurrence probabilities, detection of failures, and judgment as to the severity of a failure, it should only be chosen if some practical experience with the technical system is available. Each of the three values will be assigned a number from 1 to 5. Multiplying these values results in the risk priority number. This number indicates the priority of the assessed failure. The pure version of the FMEA is seldom practiced in the pharmaceutical industry. 

Reason J. 1990. The contribution of latent human failures to the breakdown of complex systems. Philos Trans R Soc Lond B Biol Sci h27. 

Moreover, such a judgemental system might itself fail in the eyes of some other judgemental system, a possibility that is well understood by the legal system, with its hierarchy of courts. So, there is a (recursive) notion of failure , which clearly is a relative rather than an absolute notion. So then is the concept of dependability, at any rate when one is dealing with hugely complex systems. 

A description of a system s structure identifies its component systems, their observable boundaries and functions, and their means of interaction. Careful identification of system boundaries is fundamental to understanding and distinguishing between failures, faults and errors, so as to analyze possible or actual failure situations and find means of reducing their likelihood, especially in complex systems. System structuring thus plays a central role in dependability. 

The successful design and deployment of any complex system that interacts directly with humans thus calls for socio-technical as well as technical expertise. One particular problem is that of how best to partition an overall task between humans and computers so as (i) to reduce the probability of failures due to misunderstandings across the human-machine interface, and (ii) to make best use of the greatly differing abilities that humans and computers have with respect to following complicated detailed sequences of instructions, and recognizing that a situation is both novel and potentially dangerous. 

The need for minimal function reinforces the irreducible complexity of the system. Imagine you were adrift in a life raft on a stormy sea, and by chance a box floated by that contained an outboard motor. Your joy at the hope of deliverance would be short-lived if, after you affixed it to the boat, the outboard propeller turned at a rate of one revolution per day. Even if a complex system functions, the system is a failure if the level of performance is not up to snuff. 

---