Case 8

Western Pharmaceuticals (A)

Overview

Western Pharmaceuticals (A) is a supply chain network design case placed in the setting of
a merger of two firms, Western Pharmaceutical and Atlantic Medical. While their product lines
don’t overlap, there is definite overlap in their distribution systems. Both the demand and cost
data is provided in spreadsheets to facilitate a total cost analysis. This case provides students
with a comprehensive and realistic distribution center location problem that can be solved using a
spreadsheet analysis.
While the analysis can be completed using a supply chain design planning tool such as
CAPS, Supply Chain Planner, Supply Chain Navigator, or Logic*Net, it is often useful to have
the students completed the analysis using a spreadsheet so they develop some insight regarding
the logic and dynamics of such planning tools. This case can also be used to illustrate and apply
the Logistics Planning Process discussed in Chapter 16.
This case is an effective tool for students understand the process of developing a total cost
model and then using it to determine relevant total cost curves includes the major logistics cost
components.

Solution

The first step in solving the case is to define the critical assumptions for completing
analysis. Specific assumptions include:

1. All demand is assigned to a single point within each state. A representative point is the
capitol or the largest city;
2. All demand is shipped to customers in either less-than-truckload (LTL) or truckload (TL)
volumes and each state receives a combination of both;
3. Service level is measured as the percent of volume within 750 miles of distribution centers.
In effect the demand point in each state for 95 percent of the total volume must be within
750 miles of the servicing distribution center;
4. The analysis scope includes movements from plants to distribution centers to markets;
5. Analysis costs include fixed costs of distribution centers, variable costs at distribution
centers, inbound transportation cost (plants to distribution centers), outbound
transportation cost (distribution centers to markets), and inventory carrying cost
6. Market demand is assigned to the distribution center with the lowest total cost of service;
7. Projections regarding changes in inventory carrying costs are estimated using the “Square
Root of N” approach;
8. All products in a state are served from a common distribution center so it is not necessary
to consider alternate sources for each product. This is a common practical assumption to
minimize complexity and provide consistency for customers
9. The product mix for all states is the same as provided in Table 3 from the case
10. For determining service distances, the student either look up miles such a by using
Mapquest (rather tedious) or use a disk that equates to a 750 mile scale for the map they
are using; and
11. The fixed and variable distribution center rates are independent of throughput volume.

The second step is to develop a total cost model in concept and then apply it to a
spreadsheet. In concept the objective is to minimize

TC = Σ Fi + Σ( Σj ((C i,j,tl + Hi) * D j,tl) + Σj ((C i,j,ltl + Hi) * D j,ltl)) + Σ (Pk,i * (D j,tl + D j,ltl))
Where:
TC = Total logistics cost;
Fi = Fixed cost for facility i;
C i,j,tl = Per pound transportation cost for truckload movement from distribution center i to
state j;
Hi = Per pound handling or variable cost for distribution center i;
D j,tl = Truckload demand for market j;
C i,j,ltl = Per pound transportation cost for LTL movement from distribution center i to
state j;
D j,ltl = LTL demand form market j; and
P k,i = Per pound transportation cost from plant k to distribution center i.

The third step is to determine the inventory carrying cost implications for the varying
number of distribution center locations. Although there are currently eight distribution
centers, there is an obvious redundancy in Atlanta that would be removed when the systems
are combined. So there are currently seven distribution centers from an inventory
perspective. “Case 8 – Inventory Calculator for Western Pharmaceuticals” provides the
details of the inventory analysis concerning the impact of the number of stocking locations.
This applies the “Square Root of N” approach discussed in Chapter Fifteen. Row 2 of the
spreadsheet contains the dollar sales for each company (Total Pounds * $5/pound * Percent of
Sales for division). Row 3 calculates the inventory for each company based on annual turns.
Row 4 contains the number of current distribution centers. Row 5 contains the typical
replenishment cycle for each company (bi-weekly from the case). The cycle and safety stock
factors are calculated for each company based on the values of Rows 2-5. The factors are
different because the turn rates are different. The lower part of the spreadsheet contains the
aggregate inventory levels for a range in the number of distribution centers. Column B
contains the number of distribution center locations. Column D contains the projected
inventory for Western Pharmaceuticals for the corresponding number of locations in Column
B. Column D is calculated using the cycle and safety stock factors and the number of
stocking locations. Column F contains the Western Pharmaceutical inventory turns based on
the annual sales and projected inventory level. Columns H and J contain the projected
inventory level and turn rates for Atlantic Medical using the Atlantic Medical factors and
number of stocking locations. Columns L and M contain the combined inventory and turn
rates for both systems if there were aggregated into a common distribution system assuming
current inventory management practices.

The fourth step is to cost out each combination of each distribution system scenario.
While there are numerous combinations that could be considered, there are a relatively few
that are logical and feasible. Table 8-1 summarizes the feasible scenarios tested. Table 8-2
summarizes the results of these scenarios. Scenario 1 represents the base case using all seven
current locations. Scenarios 2 through 7 are the six distribution center scenarios without
considering the closing of Atlanta. Atlanta is not considered for closures since it really
doesn’t have an alternative site that can service the major populations in the Southeast within
the 750-mile limit. Scenarios 8 and 9 consider five and four distribution center scenarios by
closing Mechanicsburg, Sparks, and then Indianapolis. While there are other combinations
of four and five distribution centers, the combinations considered are the most logical due to
service and demand densities. While one could consider fewer than four distribution centers,
it is not possible to meet management’s service objective with any fewer locations.

Once the minimum cost number of distribution centers is determined, the final step is to
complete some sensitivity analyses by consolidating production of key products. Since A
and F represent the highest volume products and they are produced on the geographic
extremes of the U.S., the sensitivity analysis considers the relocation of production to a more
centralized plant location such a Omaha. While there would likely be some reduction in
annual operating costs, there is a substantial one-time cost to shift the volume.
Table 8-2 summarizes the results of each scenario. The base case with the existing
assignments has a total cost of $80.197 million with 98.14 percent of demand within 750 miles
of a distribution center. The total cost of the six DC scenarios ranges from $75.9 to $77.0
million. They all meet the service objective except the scenario without Omaha. The five
and four DC scenarios continue to reduce cost as the increases in transportation costs are
more than overcome by decreases in fixed and inventory carrying cost. The minimum
operational cost scenario (transportation, variable, and fixed costs) is five distribution centers
while the minimum total cost scenario (including inventory carrying cost) is four DCs
( Atlanta, Los Angeles, New Brunswick, and Omaha). The total cost for the minimum total
cost four DC scenario is $68.053 million which represents a savings of $12.144 million from
the base.
The sensitivity analyses (Scenarios 10 and 11) shift production of one or two product
groups to the more central facility in Omaha. This reduces backhaul from the respective
plants. This results in substantial additional savings with no impact on service. The shift of
product A results in an annual savings of $5.2 million for a one-time cost of movement of
$500 thousand. The shift of product F yields an additional $1.367 million for an additional
one-time cost of $500 thousand.
 Table 8-1
Supply Chain Scenarios Tested
Atlanta Indianapolis Los Mechanicsburg New Omaha Sparks
Scenario Description Angeles Brunswick
Current 7 DCs,
X X X X X X X
1 6 Plants
X X X X X X
2 6 DCs (no NB)
6 DCs (no
X X X X X X
3 Mech)
6 DCs (no
X X X X X X
4 Sparks)
X X X X X X
5 6 DCs (no LA)
6 DCs (no
X X X X X X
6 Omaha)
6 DCs (no
X X X X X X
7 Indianapolis)
5 DCs (no
X X X X X
8 Mech, Sparks)
4 DCs (no
Mech, Sparks, X X X X
9 Indy)
4 DCs (No
Mech, Sparks,
X X X X
Indy), Product A
10 to Omaha
4 DCs (No
Mech, Sparks,
Indy), Product A X X X X
and F to
11 Omaha
 Table 8-2
Western Pharmaceuticals (A)
Scenario Total Cost Analysis
($000)

Scenario Scenario Fixed Variable Inbound Outbound Total Inventory Total Capacity Service Savings
Number Trans. Trans. Less Carrying Cost Shift Level
Inventory Cost
No Capacity Shifts
1 Current 7 DCs, 6 Plants 2,100 1,369 12,786 10,448 26,703 53,453 80,197 0 98.14%
2 6 DCs (no NB) 1,800 1,369 12,984 10,609 26,762 49,696 76,496 0 98.14% $3,701
3 6 DCs (no Mech) 1,800 1,369 12,571 10,472 26,212 49,734 75,946 0 98.14% $4,251
4 6 DCs (no Sparks) 1,800 1,369 12,772 10,462 26,403 49,734 76,137 0 96.36% $4,060
5 6 DCs (no LA) 1,800 1,369 12,824 10,874 26,867 49,734 76,601 0 98.14% $3,596
6 6 DCs (no Omaha) 1,800 1,369 13,002 10,333 26,504 49,734 76,238 0 87.99% $3,959
7 6 DCs (no Indy) 1,800 1,369 13,166 10,958 27,293 49,734 77,027 0 98.14% $3,170

8 5 DCs (no Mech, Sparks) 1,500 1,369 12,557 10,486 25,912 45,645 71,557 0 96.36% $8,640
9 4 DCs (no Mech, Sparks, Indy) 1,500 1,369 12,950 11,111 26,930 41,123 68,053 0 96.36% $12,144

Capacity Shifts
10 4 DCs (No Mech, Sparks, Indy) 1,200 1,369 7,965 11,111 21,645 41,123 62,768 500 96.36% $17,429
Product A to Omaha
11 4 DCs (No Mech, Sparks, Indy) 1,200 1,369 6,598 11,111 20,278 41,123 61,401 1,000 96.36% $18,796
Products A and F to Omaha