Computer risks

Caution must be observed in such an interpolation, however, and all the relevant data must be used. In at least one case—vinyl chloride-careless use of the procedure could have underestimated the risk, Le., in rats exposed to inhaled vinyl chloride monomer, the dose-inddence curve was observed to satiuate at the dose at which 20% of rats had angiosarcomas (Maltoni, 1977). Hence, if experimental points at lower doses had not been available, a simple interpolation between the zero dose incidence and the inddence at the maximum dose would have underestimated the computed risk at low doses, as illustrated in Fig uie 8.3. 

Fig. 8.3 Incidence of liver angiosorcouia in rats versus concentration of vinyl chloride monomer in air inhaled for 4 hours/day, 5 days/week. The line connecting the origin to the highest measured point obviously gives too low a computed risk at the lower dose levels. Also, shown are int polations indicating the number of cases per ppm inhaled (fixun Maltoni, 1977).

Lloyd, I.J. and Simpson, M.J. (1997), Computer Risks and Some Legal Consequences, in Safety and Reliability of Software Based System, Springer-Verlag, New York. 

More complicated 3D effects were studied in Refs. 6 and 7 with the help of 3D Monte Carlo digital simulation performed with a rather powerful computer (RISK System/6000). Sedimentation FFF with different breadth-to-width channel ratios and both codirected and counterdirected rotation and flow were studied. Secondary flow forming vortexes in the y-z plane is generated in the sedimentation FFF channel, both due to its curvature, and the Coriolis force caused by the centrifuge rotation. The exact structure of the secondary flow was calculated by the numerical solution of the Navier-Stokes equations and was used in the Monte Carlo simulation of the movement of solute molecules. 

ENVIRONMENTAL AND ECOLOGICAL TOXICOLOGY COMPUTATIONAL RISK ASSESSMENT... 

In the risk evaluation we need to see beyond the computed risk picture in the form of the summarising probabilities and expected values. Traditional QRAs often fail in this respect. This observation was the starting point for the development of the semi-quantitative... 

Another example of risk analysis follows a process of computing risk and cost for controls. The final results give estimates of return on investment. The procedure will analyze processes, activities or equipment. The first step is hazard identification, followed by estimates of frequency and severity of losses for each hazard. There are one or more controls associated with each hazard. The procedure determines the... 

Risk recovery can be included in supply chain optimization models. One can first create risk scenarios based on previously computed risk values (using techniques described in Section 7.9). Then, assuming certain inventory levels, the p parameter for each scenario can be calculated and the sum of p parameters of each supplier can be minimized. 

This is the method used by the commercial software packages Crystal Ball and RISK . The method is ideally suited to computers as the description of the method will reveal. Suppose we are trying to combine two independent variables, say gross reservoir thickness and net-to-gross ratio (the ratio of the net sand thickness to the gross thickness of the reservoir section) which need to be multiplied to produce a net sand thickness. We have described the two variables as follows ... 

A Monte Carlo simulation is fast to perform on a computer, and the presentation of the results is attractive. However, one cannot guarantee that the outcome of a Monte Carlo simulation run twice with the same input variables will yield exactly the same output, making the result less auditable. The more simulation runs performed, the less of a problem this becomes. The simulation as described does not indicate which of the input variables the result is most sensitive to, but one of the routines in Crystal Ball and Risk does allow a sensitivity analysis to be performed as the simulation is run.This is done by calculating the correlation coefficient of each input variable with the outcome (for example between area and UR). The higher the coefficient, the stronger the dependence between the input variable and the outcome. 

The overall objective of research under way as of ca 1997 is to develop a system of sale by description for fine and medium wools whereby the buyer is presented only with measured data on the principal characteristics of the raw wool, as well as an assessment of the less important characteristics by an independent skilled appraiser (8). A scheme for assessing the risk of the presence of colored fiber content in greasy wool has been proposed which depends on production parameters and on the age and sex of the sheep (5). Instmmentation and computer algorithms for the measurement of style and handle... 

Virtual Sources The previous equations apply to point source releases. Real releases, such as a boiling pool of liquid or a streaming jet of flashing hquid, involve a more complex geometiy One approach Guideline.s for Chemical Proce.s.s Quantitative Risk Analy.sls, AIChE, 1989, p. 87) is to define a virtual source upwind from the ac tual source such that the computed plume matches the real plume. However, to achieve this, a concentration at a centerhne point directly downwind must be known. 

IRI. Recommended Good Practice for the Protection of Electronic Data Processing and Computer Controlled Industrial Processes. Industrial Risk Insurers, Hartford, CT. 

The QRA project team can select the appropriate technical approach once you specify the study objectives, and together you can define the scope. A variety of modeling techniques and general data sources (discussed in Section 3.2) can be used to produce the desired results. Many computer programs are now available to aid in calculating risk estimates, and many automatically give more answers than you will need. The QRA team must take care to supply appropriate risk characteristics that satisfy your study objectives—and no more. 

The use of computers is essential in probabilistic design (Siddal, 1983). However, research has shown that even the most complete computer supported analytical methods do not enable the designer to predict reliability with sufficiently low statistical risk (Fajdiga et al., 1996). Far more than try to decrease the statistical risk, which is probably impossible, it is hoped that the approach will make it possible to model a particular situation more completely, and from this provide the necessary redesign information which will generate a reliable design solution. 

VIEW is the quantification module. All minimal cutsets are stored in the speciiic libraries for the fault trees, supercomponents and sequences. VIEW recalculates the point estimates. It computes and displays the Fussel-Vesely importance, risk increase and risk reduction measures. 

RISKMAN is an integrated Microsoft Windows , personal computer software system for [H. i forming quantitative risk analysis. Used for PSAs for aerospace, nuclear power, and chemical [iroccsses, it has five main modules Data Analysis, Systems Analysis, External Events Analysis, Event Tree Analysis, and Important Sequences. There are also modules for software system maintenance, backup, restoration, software updates, printer font, and page control. PEG has also integrated the fault tree programs CAFTA, SETS, NRCCUT, and IRRAS into RISKMAN. 

The confluence of sharply rising Operations and Maintenance (O M) costs. NRC requested Individual Plant Examinations (IPEs) and increased personal computer capabilities gave rise to the R R Workstation. Its uses and maintains-current PSA models and databases for individual plants to perform O M planning and scheduling, and uses the PSA in IPE models to identify plant design, procedure and operational vulnerabilities. The Risk and Reliability Workstation Alliance was organized by EPRI to support the R R Workshop in order to achieve O M cost reduction, plant productivity and safety enhancement through risk-based, user-friendly, windowed software louls (Table 3.6 8). The Alliance, initiated in 1992, includes 25 U.S. utilities and four international partners from Spain, France, Korea, and Mexico. SAIC is the prime contractor for the R R Workstation, with participation of five other PSA vendors. 

Safety Monitor is an interactive computer program (Stamm, 1996) that performs real-time assessments of configuration-specific plant accident risk. Originally used at Southern California Edison s San Onofre station in 1994, further development was sponsored by three nuclear utilities and EPRI, to include shutdown operation and expanded user features. It will be enhanced to calculate large early release frequency (a Level 2 risk). 

Having completed the risk analyses, computed the uncertainties, and identified critical systems by importance measures (which also identifies valuable systems improvements having low costs), the PSA results must be presented. An executive summary compares the risk of operations that were analyzed with the risks of similar operations. It identifies and explains the main contributors to the risk to people untrained in PSA and statistical methods. Figure 6.3-5 shows two pie-charts that show the risk contributions of various initiators for PWRs and BWRs. A chart similar to one of these would be an effective way of showing the risk contributions in simplified form. 

---