Sei sulla pagina 1di 4

H84ACM ADVANCED COMPUTATIONAL METHOD

Question 2: Steady State Optimizer

Name

: Videsh A/L Sagindharan

Student ID

: 009497

Lecturer

: Dr. Chan Yi Jing

Videsh A/L Sagindharan (009497)

1.0

Question 2

1.1

Converged Base Case Simulation

Figure 1: Base Case Simulation Flowsheet

1.2

Sensible Limits

The limits that have been chosen for T4, T5 & Tee-100 Flow Ratio 2 are shown in Table 1. First off, the
reason why the limits for T4 is 0C to 40C is chosen is because according to case study , Figure 2, the
minimum cost for the total cost lies between 5C to 30C but in order to give allowance to the lower and
upper boundaries, 0C to 40C is chosen. Besides, inlet temperatures of T4 and T5 are 40C, hence 40C
should be chosen as the upper boundary for both T4 and T5. As for the lower boundary of T5, it is due to
the process where it needs to be cooled down to -20C, therefore -20C is selected as the lower boundary
of T5. The lower boundary for T4 should not be set below 0C as temperature cross may occur because
the inlet temperature of the refrigerant is 0C. The higher limits for flow ratio 2 is set at 0.99 and not 1
because it causes hysys to become over specified and result in error, hence 0 to 0.99 were chosen.

Table 1: Optimizer Boundaries for three primary variables


Variables
T4
T5
Flow Ratio 2 Tee-100

1.3

Lower Boundary
0C
-20C
0

Upper Boundary
40C
40C
0.99

Sensitivity Analysis

Sets of Iterations
1
2
3
4
5
6
7
8

T4 (C )
5
5
5
10
10
15
15
20

Before Optimized
T5 (C ) FR2
Cost (RM )
5
0.87 6.7481E+18
10
0.84 6.7615E+18
0
0.9 6.8437E+18
-5
0.93 7.1698E+18
0
0.9 7.2768E+18
15
0.9 9.9183E+18
15
0.93 9.9194E+18
30
0.18 1.7611E+19

T4 (C )
6.679
6.814
6.661
6.662
6.748
6.794
6.725
6.664

After Optimized
T5 (C )
FR2
Cost (RM )
-4.297 0.924 6.3533E+18
-4.602 0.926 6.3560E+18
-4.298 0.924 6.3535E+18
-3.747 0.922 6.3529E+18
-5.472 0.928 6.3566E+18
-6.823 0.933 6.3638E+18
-4.916 0.926 6.3547E+18
-3.941 0.922 6.3530E+18

Table 2: Sets of Iteration for Initial Guess based on Case Study

1|Page

Videsh A/L Sagindharan (009497)

Figure 2: Comparison between initial guess variables and total cost

Figure 3: Comparison between initial guess variables and optimized values

Optimized total cost against a few sets of initial guesses of T4, T5 & Tee-100 Flow Ratio 2 was plotted as
shown in Figure 2 to represent the sensitivity analysis. Different initial guesses would give different
optimized total cost; hence, a case study was performed by varying the three primary variables and to
observe the change in total cost. From the case study, a few sets of primary variables was selected based
on the minimum cost and those values were used to run the optimizer. The optimized total cost is RM
6.3529 1018/year which was obtained from optimized variables, T4 = 6.662C, T5 = -3.747 C and FR2 =
0.922. The results of the sensitivity analysis are displayed in Table 2 above and the optimized results are
plotted in Figure 2 & 3. From Figure 2, the optimized total cost graph isnt a linear graph as expected, this
is because, optimized total cost is highly sensitive to a change in T5 value. The bigger LMTD corresponds
to a smaller area for heat exchanger and vice versa which affects the cost of heat exchanger. Comparing
between Figure 2 & Figure 3, the optimized total cost fluctuates steeply when there is a sudden change in
T5 whereas the optimized total cost doesnt fluctuate when there are changes in T4 & FR2. Therefore,
optimized total cost is highly dependent on T5 and not towards T4 & FR2.

1.4

Reliability of Solution

The reliability of optimum solution is proved by conduction a case study. Initially, a case study was done
by varying one variable, for example T4, while fixing the other two variables and the same thing is done
for the other two variables, T5 & FR2. Figure 4, 5, and 6 shows how each variable affect the total cost. The
optimum solution obtained is within the range of boundary limits. To take it further, another case study
by varying all three variables, T4, T5 & FR2, randomly to determine the minimum optimum total cost.
Figure 7 shows the case study conducted to obtain minimum optimized total cost by varying all three
variables and the final solution is obtained where the minimum optimized total cost is RM 6.3529
1018/year at T4 = 6.662C, T5 = -3.747 C and FR2 = 0.922.

2|Page

Videsh A/L Sagindharan (009497)

Figure 4: Case Study for T4 against Total Cost

Figure 5: Case Study for T5 against Total Cost

Min Cost = RM 6.7481 x 10 /year

Figure 6: Case Study for FR2 against Total Cost

1.5

Figure 7: Case Study of Total Cost against T4 , T5 & FR2

Constraint Problem

Figure 8: The added constraint

Figure 9: Total Cost against FR2

The constraint imposed had caused the optimization problem to be more tightened and the solver faced
higher difficulties in reading a solution. The right combination of initial guesses is required for the solver
to return the solution, possibly the combination that will lead to solver to the direction where the
solution lies. Therefore, the previous case study was used with the addition of Q-100 heat flow variable.
The combinations for initial guess which converged the optimizer and also produced the lowest cost is T4
= 10C, T5 = 25C and FR2 = 0.64. The optimized solution obtain for the 3 variables after optimization are
T4 = 6.0430C, T5 = 25.0351C and FR2=0.5607 with the optimized cost of 9.8518 x 1018 kJ/h. This is
attributed to all the variables take higher values compared to previous case. The constraint causes the
process stream into E-100 to have higher temperature, so the cooler needs more energy to reduce the
additional energy which directly results in higher total cost. The lower the FR2 the higher the total cost
and vice versa as shown in Figure 9.

3|Page

Potrebbero piacerti anche