Math 208 - Tutorial 8

Mina Moeini
Mina moeini@sfu.ca
March 11, 2025

1 Plant Location (Problem 7.2)

The Spencer Shoe Company manufactures a line of inexpensive shoes in one plant located in
Pontiac and distributes them to five main distribution centers: Milwaukee, Dayton, Cincin-
nati, Buffalo, and Atlanta. From these distribution centers, the shoes are shipped to retail
stores. Distribution costs include freight, handling, and warehousing costs.
To meet increased demand, the company has decided to build at least one new plant with
a capacity of 40,000 pairs per week. Potential locations for the new plant include Cincinnati,
Dayton, and Atlanta. While Atlanta has lower production costs, its distribution costs are
relatively high compared to the other two locations. The given data are as follows:

To Distribution Centers Pontiac Cincinnati Dayton Atlanta

Milwaukee 0.42 0.46 0.44 0.48
Dayton 0.36 0.37 0.30 0.45
Cincinnati 0.41 0.30 0.37 0.43
Buffalo 0.39 0.42 0.38 0.46
Atlanta 0.50 0.43 0.45 0.27

Table 1: Distribution Costs per Pair

Distribution Center Demand (pairs/week)

Milwaukee 10,000
Dayton 15,000
Cincinnati 16,000
Buffalo 19,000
Atlanta 12,000

Table 2: Demand at Distribution Centers

1
 Location Capacity (pairs/week) Production Cost/pair (USD) Fixed Cost/week (USD)
Pontiac 32,000 2.70 7,000
Cincinnati 40,000 2.64 4,000
Dayton 40,000 2.69 6,000
Atlanta 40,000 2.62 7,000

Table 3: Capacity, Production Cost, and Fixed Costs per Location

(a) Selecting One New Plant Location

Determine which of the three new alternatives (Cincinnati, Dayton, or Atlanta) leads to the
lowest total cost, including production, distribution, and fixed costs.

(b) Optimal System Design

Assuming that Spencer Shoe Company can start from scratch and operate any combination
of the four plants, determine the plant locations that minimize total cost. Compare the
results to part (a) and determine the potential weekly cost savings with the optimal system
design.

Solution
Problem Parameters
The problem parameters include transportation costs, demand at each distribution center,
plant capacities, production costs per pair, and fixed costs per week at each plant.
Let:

• P = {P, C, D, A} be the set of plants: Pontiac, Cincinnati, Dayton, and Atlanta.

• D = {M, D, C, B, A} be the set of distribution centers: Milwaukee, Dayton, Cincin-

nati, Buffalo, and Atlanta.

• ri,j , i ∈ P, j ∈ D be the transportation cost of shipping from plant i to distribution

center j.

• CAPi , i ∈ P be the capacity of plant i.

• pi , i ∈ P be the production cost per pair at plant i.

• f ci , i ∈ P be the fixed cost for plant i.

• dj , j ∈ D be the demand at distribution center j.

2
Problem Decision Variables
The decision variables define the quantity of shoes to be produced and shipped from each
plant to each distribution center.
Let:

• xi,j , i ∈ P, j ∈ D be the quantity of shoes produced at plant i and shipped to

distribution center j.

• Ii , i ∈ P be a binary variable indicating whether plant i is built (Ii = 1) or not

(Ii = 0).

Constraint:
Ii ∈ {0, 1}, ∀i ∈ P (1)

Problem Constraints
The problem constraints include:

Demand Satisfaction
Each distribution center must receive at least its demand:
X
xi,j ≥ dj , ∀j ∈ D (2)
i∈P

Capacity Constraints
The production at each plant cannot exceed its capacity when the plant is active:
X
xi,j ≤ CAPi × Ii , ∀i ∈ P (3)
j∈D

Objective Function
The objective is to minimize the total transportation cost, production cost, and fixed cost:
XX X X X
min ri,j xi,j + pi xi,j + f ci Ii (4)
i∈P j∈D i∈P j∈D i∈P

3
Overall Model
XX X X X
min ri,j xi,j + pi xi,j + f ci Ii (5)
i∈P j∈D i∈P j∈D i∈P
X
s.t. xi,j ≥ dj , ∀j ∈ D (6)
i∈P
X
xi,j ≤ CAPi × Ii , ∀i ∈ P (7)
j∈D

Ii ∈ {0, 1}, ∀i ∈ P (8)

The optimized objective value is 227, 210. The decision variables is as follows:

To Distribution Cost From

Pontiac Cincinnati Dayton Atlanta
Milwaukee 0 0 0 10000
Dayton 0 15000 0 0
Cincinnati 0 16000 0 0
Buffalo 0 9000 0 10000
Atlanta 0 0 0 12000
Indicator 0 1 0 1

Table 4: Decision Variables for Distribution Cost

4
Landfill Location (Problem 7.3)
The Metropolis city council is examining four landfill sites as candidates for use in the city’s
solid waste disposal network. The monthly costs per ton have been estimated for operating
at each site and for transportation to each site from the various collection areas. In addition,
the amortized monthly cost for the facility at each proposed site has also been estimated.
The data are shown in the table below.

Collection Area L1 L2 L3 L4 Monthly Tons

A $14 $16 $10 $8 500
B $12 $11 $12 $14 700
C $13 $8 $9 $11 1500
D $10 $15 $14 $12 1000
E $8 $12 $10 $11 1800
F $11 $10 $8 $6 1200

Table 5: Transportation Costs per Ton to Landfill Sites

Landfill Site L1 L2 L3 L4
Operating Cost per Ton $8 $10 $9 $11
Fixed Cost per Month $1000 $800 $700 $900

Table 6: Operating and Fixed Costs for Each Landfill Site

What is the optimal configuration and the minimum system monthly cost?

Solution
Problem Parameters
The problem parameters include the transportation cost per ton to each landfill site, the
operating cost per ton for each landfill, the fixed cost per month for each landfill, and the
monthly tons of garbage produced at each collection area. Let S = {A, B, C, D, E, F }
represent the set of collection areas, and let L = {L1, L2, L3, L4} represent the set of landfill
sites. The transportation cost per ton from collection area i to landfill j is denoted by ri,j ,
where i ∈ S and j ∈ L. The monthly tons of garbage produced in each collection area,
referred to as the demand, is given by di , where i ∈ S. The operating cost per ton for each
landfill is represented by oj , where j ∈ L, and the fixed cost per month for each landfill is
denoted by fj , where j ∈ L.

5
Problem Decision Variables
The decision variables represent the amount of garbage sent from each area to a landfill:

• Let xi,j , where i ∈ S, j ∈ L, be the amount of garbage sent from area i to landfill j.

• Since there is a fixed cost for each landfill we choose to operate, we introduce a binary
indicator variable Ij , where j ∈ L, which is defined as:
(
1, if landfill j is used
Ij =
0, otherwise

Problem Constraints
We must ensure that all garbage from each area is collected:
X
xi,j ≥ di , ∀i ∈ S (9)
j∈L

We also introduce linking constraints to determine whether a landfill is used. We define

a large constant M , which is an upper bound greater than the sum of all garbage produced
(in this case, M > 6700 tons):
X
xi,j ≤ M Ij , ∀j ∈ L (10)
i∈S

The binary constraint for landfill usage is:

Ij ∈ {0, 1}, ∀j ∈ L (11)

Problem Objective
The objective is to minimize the total transportation cost, the operating cost of used landfills,
and the fixed cost of each landfill:
! ! !
XX X X X
min ri,j xi,j + oj xi,j + fj Ij (12)
i∈S j∈L j∈L i∈S j∈L

6
Overall Model
XX X X X
min ri,j xi,j + oj xi,j + fj Ij (13)
i∈S j∈L j∈L i∈S j∈L
X
subject to xi,j ≥ di , ∀i ∈ S (14)
j∈L
X
xi,j ≤ M Ij , ∀j ∈ L (15)
i∈S
Ij ∈ {0, 1}, ∀j ∈ L (16)

The optimal cost is minimized at $119,400. The optimal decision variables are as follows:
The optimal allocation of garbage to landfill sites is shown in the table below:

Collection Area L1 L2 L3 L4
A 0 0 500 0
B 700 0 0 0
C 0 0 1500 0
D 1000 0 0 0
E 1800 0 0 0
F 0 0 1200 0

Table 7: Optimal Garbage Allocation to Landfills

The binary indicator variables for landfill operation are:

Landfill Indicator Decision (Ij )

IL1 1
IL2 0
IL3 1
IL4 0

Table 8: Indicator Variables for Landfill Selection

From this table, we see that landfills L1 and L3 are selected for funding.