Table 4.18 Comparative analysis of preliminary design data and simulation output data for simu-lation model sector 2 Assembly Design flow Model flow Model min.. Table 4.19 Acceptance cri
Trang 1Table 4.18 Comparative analysis of preliminary design data and simulation output data for
simu-lation model sector 2
Assembly Design flow Model flow Model min Model max.
vol vol flow vol flow vol.
Mill discharge tank 1 1,119 1,120 1,110 1,130
Mill discharge tank 2 1,119 1,120 1,110 1,130
Mill discharge tank 3 1,119 1,120 1,110 1,130
Mill discharge tank 4
Classifier feed pump (S)
Screen feed pot 4
e) Evaluation of Simulation Model Sector 3
A major characteristic of the process flow diagram (PFD) of sector 3 is that it depicts
continuous fluid flow and indicates how inputs are transformed by each assembly into outputs that, in turn, become modified logical flow inputs to the next assembly,
as depicted in the preliminary design data given in Table 4.20 The PFD is sys-tematically examined to analyse deviations in process flow and system performance and, in this case, to determine mass fluid flow balance through integrated assem-blies Each assembly is graphically represented in the simulation model by a virtual
prototype process equipment model (PEM).
Each of the assemblies of the PFD depicted in Fig 4.60, consisting of four pro-cessing tank systems containing 23 assemblies (four double tank feeder chutes, four processing tanks plus one standby, four sets of three-up parallel pumps, and one pump/condensate assembly), is a process equipment model
A fluid mass-flow balance is the application of conservation of mass to the anal-ysis of physical systems By accounting for materials (solids or fluids) entering and leaving a system, mass flows can be identified from one system, or assembly, to the next The exact mass-balance theory used in the analysis of the system depends on
Trang 2Fig 4.59 Simulation output for simulation model sector 2
the context of the design problem, specifically where the theory is used to analyse alternative processes
Process design specifications Each PEM contains selected model components that
are configured in such a way that the design specifications of each assembly are met through the component’s attributes The model component’s attributes for the four double tank feeder chutes convert the chutes’ output by modifying the compo-nent’s inputs through a selection of statistical functions based on feed specifications The model component’s attributes for each of the processing tanks’ pumps convert
a pump’s output by modifying the inputs through a selection of statistical functions representing the appropriate pump delivery characteristics
Figure 4.61 illustrates the application of Petri net (PN)-based optimisation
algo-rithms in dynamic systems simulation The optimisation algorithm is a model com-ponent inherent to the processing tank PEM and determines process flow pressure surge through the tank
Output performance results The fluid mass that enters a system must, by
conser-vation of mass, either leave the system or accumulate within the system Basically, the fluid mass-flow balance equation for a system without internal chemical reac-tions is: input= output + accumulation In the absence of a chemical reaction,
Trang 3Table 4.19 Acceptance criteria of simulation output data, with preliminary design data for
simu-lation model sector 2
Assembly Design min Design max Model min Model max Yes/no
vol 2.5% tol vol 2.5% tol vol. vol at 99%
Mill discharge tank 4
Classifier feed pump (S)
Screen feed pot 4
the logical fluid flow in and out of a system or assembly will be the same To per-form a balance, the boundaries of the system must be well defined Fluid mass-flow balances can be taken over physical systems at multiple scales, taking into consider-ation flow surges, and can be simplified with the assumption of steady state, where the accumulation term is zero
Figure 4.62 illustrates a typical output document showing performance results of the processing tank PEM These performance variables relate to assembly contents, input and output flow quantities, as well as flow surges The flow surge gives an indication of deviations from steady-state flow The plotted graph shows the trend
of flow from start-up to steady state
f) Conclusion of Simulation Model Sector 3 Evaluation
Table 4.21 gives the values of a comparative analysis of preliminary design data and simulation output data for simulation model sector 3
Figure 4.63 shows the simulation model’s output for simulation model sector 3
As with simulation model sectors 1 and 2, the range or variance of the model’s
Trang 4Table 4.20 Preliminary design data for simulation model sector 3
Assembly Code Flow vol Mass flow Liq Solids
Slurry splitter box 1 L026031 1,250 2,136 1,197 938
Slurry splitter box 2 L026041 968 1,642 928 721
Slurry splitter box 3 L026051 968 1,642 928 721
Slurry splitter box 4 L026061 1,250 2,136 1,197 938
Slurry forwarding pump 1 P026011 1,250 2,136 1,197 938
Slurry forwarding pump 2 P026021 968 1,642 928 721
Slurry forwarding pump 3 P026031 968 1,642 928 721
Slurry forwarding pump 4 P026041 968 1,642 928 721
Slurry forwarding pump 5 P026051 968 1,642 928 721
Slurry forwarding pump 6 P026061 1,250 2,136 1,197 938
Table 4.21 Comparative analysis of preliminary design data and simulation output data for
simu-lation model sector 3
Assembly Design flow Model flow Model min Model max.
vol vol flow vol flow vol.
Slurry splitter box 1 1,250 1,250 1,245 1,255
Slurry splitter box 4 1,250 1,250 1,245 1,255
Slurry forwarding pump 1 1,250 1,250 1,240 1,260
Slurry forwarding pump 6 1,250 1,250 1,240 1,260
Trang 5Fig 4.60 Process flow diagram for simulation model sector 3
output data is compared to acceptable lower and upper confidence limits within
a specified exact probability The design specification is again used as the mean, and the allowable design tolerance of ±2.5% of the mean is used as the square
root of the variance, namely the standard deviation, in the t-distribution, to deter-mine a confidence range or interval with lower tolerance limit (LL) and an upper tolerance limit (UL) at a 99% level of confidence for ten simulation runs The mini-mum and maximini-mum values of the simulation model’s output data are similarly com-pared against this confidence range or interval The last column of Table 4.22 indi-cates whether the model’s output is acceptable in meeting the design criteria within
a 99% level of confidence As can be seen, all the assemblies have a flow volume variance that is acceptable within the 99% confidence interval as set by the design criteria
Trang 6Fig 4.61 Design details for simulation model sector 3: process design specifications
4.4.3 Application Modelling Outcome
Verification of the process simulation model with the PEM blocks included the spec-ification of model components as well as the formulation of functional relationships, all of which are inherent in the dynamic systems simulation blackboard model that
is used to control the design knowledge sources and integrate the knowledge-based
design applications In contrast to model verification, the validity of the simulation model depended on the ability of the model to predict the results of the model’s behaviour However, validation of the simulation model was not based on a
corre-lation of the mean values of the model’s output data and the specified design flow volumes for each PEM, due to possible problems of autocorrelation and the lim-ited number of simulation model runs not being large enough to justify statistical spectral analysis of the output data Rather, statistical inference was applied to de-termine whether the range of the model’s output data fell between acceptable lower and upper confidence limits within a specified exact probability
In order to determine a confidence range or interval with a lower tolerance limit (LL) and an upper tolerance limit (UL), the specified design flow volume was used
as the mean, and the allowable design tolerance of±2.5% of the mean was used as
Trang 7Fig 4.62 Design details for simulation model sector 3: output performance results
standard deviation in the statistical t-distribution at a 99% level of confidence for ten simulation runs The minimum and maximum values of the simulation model’s out-put data were then compared against this confidence range or interval to determine whether the model’s output was acceptable in meeting the design criteria
As indicated in Tables 4.16, 4.19 and 4.22, not all of the assemblies listed met the required design criteria, indicating that the simulation model failed at a 99% level
of confidence specifically for those assemblies However, the statistical approach of determining confidence intervals with the t-distribution was repeated for 95% and 90% levels of confidence Close on 85% of the simulation model’s output data was
found to meet the required design criteria at a 95% level of confidence, and all of
the simulation model’s output data met the required design criteria at a 95% level
of confidence This implies that the process simulation model with the PEM blocks
is capable of predicting process output within a 10% margin of error for each PEM Due to the fact that the model simulates a complex integrated continuous process flow, a 90% level of confidence is acceptable for the preliminary design phase of the engineered installation
Trang 8Fig 4.63 Simulation output for simulation model sector 3
4.5 Review Exercises and References
Review Exercises
1 Discuss cost modelling for design availability and maintainability
2 Explain economic loss and the cost of dependency
3 Give a brief account of life-cycle analysis and life-cycle costs
4 Consider life-cycle cost elements in engineering design
5 Describe present value calculations for life-cycle costs
6 Discuss trade-off measurement for life-cycle costs
7 Give a brief account of availability modelling based on system performance, considering process capability, process characteristics and functional effective-ness
8 Explain the concept of sizing maximum or design capacity
9 Define inherent availability (Ai)
10 Discuss inherent availability modelling with uncertainty
11 Discuss the significance of the application of the exponential function for deter-mining inherent availability
12 Describe confidence determination of inherent availability predictions
Trang 9Table 4.22 Acceptance criteria of simulation output data, with preliminary design data for
simu-lation model sector 3
Assembly Design min Design max Model min Model max Yes/no
vol 2.5% tol vol 2.5% tol vol. vol at 99%
13 Discuss preliminary maintainability modelling
14 Give a brief account of Markov modelling for design availability and maintain-ability with regard to the two-state Markov model, and the multi-state Markov model
15 Define Markov model supplementary variables
16 Define achieved availability
17 Discuss achieved availability modelling subject to maintenance
18 Consider maintainability assessment with maintenance modelling
19 Discuss the impact of maintenance assessment on systems design
20 Describe maintainability measures and maintenance assessment
21 Discuss maintenance strategies and cost optimisation modelling
22 Give a brief account of the basic principles of maintenance
23 Describe a model of preventive maintenance physical checks
24 Describe a model of preventive maintenance replacement shuts
25 Define maintenance strategy
26 Explain the concepts of reliability, availability and maintainability in mainte-nance strategy and discuss their differences
27 Give a brief account of the three principles of a maintenance strategy
28 Discuss establishing maintenance strategies for engineering design
29 Describe maintenance cost optimisation modelling
30 Define dependability modelling
Trang 1031 Discuss the significance of dependability modelling for design availability and maintainability
32 Define operational availability (Ao)
33 Discuss operational availability modelling with logistic support
34 Consider a general approach for evaluating operational availability
35 Give a brief account of system availability evaluation considerations
36 Discuss maintainability evaluation and built-in or non-destructive testing (BIT)
37 Describe maintainability evaluation indices
38 Give a brief account of diagnostic systems and built-in testing
39 Explain basic system and BIT concurrent design and evaluation
40 Discuss the evaluation of BIT systems
41 Consider application modelling of availability and maintainability in engineer-ing design
42 Define equivalent availability (EA)
43 Discuss and compare the equivalent maintainability measures of downtime and outage
44 Describe outage measurement with the ratio of ER over EM
45 Discuss system performance measures and limits of capability
46 Describe performance parameters for system integrity and their significance in engineering design
47 Discuss analysis of the parameter profile matrix
48 Discuss the significance of the design checklist
49 Explain integrity prediction of common items of equipment
50 Give a brief account of a design review of performance parameters for system integrity
51 Discuss the significance of reliability and maintainability checklists
52 Describe system performance analysis and simulation modelling in engineering design
53 Consider different types of system performance models
54 Briefly describe the significance and contribution of system simulation mod-elling in engineering design
55 Discuss uncertainty in system performance simulation modelling
56 Explain propagation of the effect of uncertainties
57 Describe the extreme condition approach for uncertainty analysis
58 Describe the statistical approach for uncertainty analysis
59 Give an explanation for mitigating the effect of uncertainty
60 Describe maximising design availability using Petri net models
61 Discuss Petri net theory and its application in engineering design
62 Define the basic Petri net model and compare it to the definitions of stochastic Petri nets as well as Markovian stochastic Petri nets
63 Briefly explain the process of generating reachability graphs
64 Discuss the measures of Markovian stochastic Petri nets
65 Define stochastic reward nets and non-Markovian stochastic Petri nets
66 Consider designing for availability using Petri net modelling
67 Describe numerical computations for the availability Petri net model