Page 282 - Design of Simple and Robust Process Plants

268 Chapter 7 Optimization of an Integrated Complex of Process Plants
nearly full, and so the upstream process is switched off. In case of a process failure,
the processes in the production chain are set at 60% capacity. The above discussed
level controller affects the capacity controllers of the processes. The level controller
might be designed in a more sophisticated fashion, but for the purpose of the exam-
ple it was considered sufficient. A qualitative histogram of storage tank level and the
related inlet and outlet flows is shown as result of failures in Figure 7.6.
The assumptions for this study are that:
. the process runs for five years between turn arounds, which is equal to the
selected mission time.
. an intermediate process stop is planned every two-and-a-half years for 1
week.
. utility availability is 1.0.
. the availability of raw material is not a constraint.
. product delivery does not constrain the production.
. the storage capacities under evaluation (expressed in hours of maximum
throughput) are: 0, 10, 18, 20, 24, and 36 hours.
. the set-point range for storage is 45±55%.
. the capacity throughput alternatives as a percentage of maximum capacity
are 0, 60, 95, 100, and 105%.
. the number of simulation runs were limited to 60 histories, and this causes
some variation in the results. An increase in the number of simulations
would have narrowed down the variability.
. the minimum repair times were set at 12 hours based on organisational con-
straints. All small rotating equipment was reparable or exchangeable within
12 hours.

7.6.3
Simulation and Results

With the tank and its controller implemented in the RDB (reliability block diagram)
flowsheet between both processes, we now are able to simulate the whole configura-
tion. The processes may be described as a combination of individual components, or
as overall plant failure rates deduced from its components. The system is now sub-
ject to probabilistic simulations with a Monte Carlo generator for the variables (as
was described for the individual process reliability studies in Section 6.4 of Chapter
6). In the example, the upstream and downstream processes are described as a com-
bination of serial and parallel components. This explains, in combination with the
limited number of runs, the variation in mechanical failure results (number of stops
and down-time).
The results can be presented in different formats and might emphasize:
. The number of process stops per year.
. The down-time in hours per year (both the down-time and number of stops
can be specified per component and totalled).