Linear Programming Refining Transportation

Solving a Refinery Problem with Excel Solver

Type of Crude or Process Product
A B C1 C2 D Demand
Profits on Crudes 10 20 15 25 7
Products Product Slate for Crude or Process
G 0.6 0.5 0.4 0.4 0.3 170
H 0.2 0.2 0.3 0.1 0.3 85
F 0.1 0.2 0.2 0.2 0.3 85
L 0.0 0.0 0.0 0.2 0.0 20
Total Crude 100 100 C1+C2 = 200 100

The objective function to maximize for this problem comes from multiplying each product by its
profits and summing up or
π = 10*A + 20*B + 15*C1 + 25*C2 + 7*D.
The total gasoline from crude A is 0.6*A, from B it is 0.5*B, etc. To make sure we satisfy the
gasoline market then our gasoline constraint must be
170 < 0.6*A + 0.5*B + 0.4*C1 + 0.4*C2 + 0.3*D.
Similarly our constraints for heating oil, fuel oil and lubes are
85 < 0.2*A + 0.2*B + 0.3*C1 + 0.1*C2 + 0.3*D,
85 < 0.1*A + 0.2*B + 0.2*C1 + 0.2*C2 + 0.3*D, and
20 < 0.2*C2.
In addition, we can not use more crude oil than we have at our disposal, so the crude oil
constraints are
A < 100,
B < 100,
C1 + C2 < 200,
D < 100.
With this more complicated problem, we will solve using a computer. We will use Solver in
Excel for MS Office 2000, which is quite user friendly and widely available. (Directions for
other versions may vary slightly.) To start, go to a new worksheet in Excel and enter the
following profit values in C1 to G1 as follows: (Note the bold faced values are addresses and
should not be typed in the cell)
C1: 10
D1: 20
E1: 15
F1: 25

Ch 14 Page 1
 Linear Programming Refining Transportation

G1: 7
Put a profit label in cell A2
A2: Profit =
Create the formula for the profit function in B2 as follows. Be sure to type the = sign so Excel
will know you are typing a formula. Then
B2: = C2*C1 + D2*D1 + E2*E1 + F2*F1 + G2*G1
The solution values to maximize the formula in B2 will be in C2 to G2. Start out with some
sample values. A good starting point might be to put 100 in each cell, which assumes that all
crudes are run at capacity or
C2: 100
D2: 100
E2: 100
F2: 100
G2: 100
Put a label in A3. Product constraints =
Then type the labels and constraints for crude and products as
B4: G
C4: 170
D4: = 0.6*C2+ 0.5*D2 + 0.4*E2 + 0.4*F2 + 0.3*G2
B5: H
C5: 85
D5: = 0.2*C2+0.2*D2 + 0.3*E2 + 0.1*F2 + 0.3*G2
B6: F
C6: 85
D6: = 0.1*C2 + 0.2*D2 + 0.2*E2 + 0.2*F2 + 0.3*G2
B7: L
C7: 20
D7: = 0.2*F2
A8: The crude oil constraints are
B9: A
C9: = C2
D9: 100
B10: B

Ch 14 Page 2
 Linear Programming Refining Transportation

C10: = D2
D10: 100
B11: C
C11: = E2 + F2
D11: 200
B12: D
C12: = G2
D12: 100
Last we need to ensure that the model does not come up with negative crude values, so we will
need 5 more constraints. If you solved the above model without the non-negativity constraints
you will find that Excel will make one of the crude runs negative. Enter the negativity
constraints as follows:
A13: Nonzero Constraints
B14: 0<=A
B15: 0<=B
B16: 0<=C1
B17: 0<=C2
B18: 0<=D
C14: 0
C15: 0
C16: 0
C17: 0
C18: 0
D14: =C2
D15: = D2
D16: = E2
D17: = F2
D18: = G2
Alternatively if all your choice variables have to be positive you can go to >Options and
check Assume Nonnegative. Also click Assume Linear Model to speed up the solution to the
problem. Note all the constraints are set up to be less than or equal to constraints. Solver tends
to work better if all the inequality constraints are arranged in the same direction. When you are
finished your work sheet should look as follows. The bold faced letters and numbers along the
edge of the spread sheets are address labels. The shaded cells can be changed if your profits or
constraints change but the refining technology stays the same.

Ch 14 Page 3
 Linear Programming Refining Transportation

Table 14.7 Worksheet for Refinery Optimization Problem

A B C D E F G
1 10 20 15 25 7
2 Profit = 8,700 100 100 0 200 100
3 Product constraints are
4 G 170 220
5 H 85 90
6 F 85 100
7 L 20 40
8 Crude oil contraints are
9 A 100 100
10 B 100 100
11 C 200 200
12 D 100 100
13 Non-zero Constraints
14 0<=A 0 100
15 0<=B 0 100
16 0<=C1 0 0
17 0<=C2 0 200
18 0<=D 0 100

To solve this model move to cell B2 and then go to Tools, Solver. The following menu in
Figure 14.5 should come up.

Figure 14.5 Excel Solver Menu

Ch 14 Page 4
 Linear Programming Refining Transportation

If Solver is not in the menu, you will have to go to >Tools >Add-Ins and install it. If it is not in
the Add-Ins you will have to reinstall the data analysis pack that came with your software. Once
in the above Solver menu, the target cell should say $B$2. Make sure Equal to is set at Max.
Click below By Changing Cells and type in C2.G2. Click on Add. The following menu in
Figure 14.6 should come up, which allows you to add in all your constraints.
Figure 14.6 Excel Solver Constraint Menu

Click on the empty box under Cell Reference. Type in or click on cell C4. Make sure the sign
to the right of the cell reference is <=. If it is not, click next to it and select <=. Click below
Constraint: and type in or select D4. Click on OK. Repeat the process for cells C5 and D5 and
so on until all constraints are entered. Note it does not matter what order you enter your
constraints. Your Solver parameter menu should look something like Figure 14.7.

Ch 14 Page 5
 Linear Programming Refining Transportation

Figure 14.7 Solver Menu for Refinery Problem

Once you have your model all set up, click on Solve. Solver will tell you if it can solve the
problem or not. If it solves, the optimal crude use will be in cells C2.G2. How much product is
produced will be in cells D4.D7.

Ch 14 Page 6