APPLICHEM
Case Analysis Presented to Professor D. Krishna Sundar Indian Institute of Management, Bangalore
On December 20th, 2005
In Partial Fulfilment of the Requirements of the course
Operations Management Submitted By
Abhishek Singh Rana (0511071) Ajit Phadnis (0511074) Alok Pande (0512003) Bhanu Pathak (0511081) Rahul Mehta (0511112)
Section B
2
Applichem
Introduction Applichem is a manufacturer of speciality chemicals, one of its unique products being Release- ease. The company has 6 manufacturing plants: 3 in North America (Gary, Canada and Mexico), 1 in Western Europe (Frankfurt), 1 in Latin America (Venezuela) and 1 in Pacific and Rest of the world (Sunchem). Each of these plants has unique characteristics of number of product lines manufactured, number of packaging varieties, capacities of plants, plant redesigns and regional laws. Due to these differences, the overall performances of the plants differ in terms of average yield of raw material and profitability. In this report we have analysed the effects of these differences between the plants and we recommend a model that Applichem can use to minimise its total costs.
Figure 1: Current Production status in Plants Market Gary Canada Mexico
North America Western
Designed capacity 18.5 3.7 22
1982 Production 14 2.6 17.2
Idle capacity 4.5 1.1 4.8
Yield 94.7% 91.1% 91.7%
Last update in Equipment 1964 1955 1978
Frankfurt Venezuela
Europe Latin America Pacific and
47 4.5
38 4.1
9 0.4
98.9% 90.4%
1974 1964
Sunchem
ROW
5
4
1
98.8%
1969
Process Flow The figure below illustrates the process flow followed by Applichem for manufacturing Release- ease.
3
Figure 2: Process flow for Production of Release-Ease
Case Analysis As per the case facts, we have listed the following factors affecting the performance of the manufacturing plants: •
Laws in Japan which increase the number of employees on the plant
•
Worker productivity potential which can be assumed to be a result of educational levels and training of staff
•
Higher number of product lines (other than Release- ease) may result in sharing of resources both physical and financial
•
Number of packaging types which results in additional setup costs.
•
Improvements in plants and machinery/ Process redesign completed/ Automation of processes/ Good Maintenance of plant and machinery
•
Emphasis on Quality
Factors influencing the performance of plants: There are two factors which have an overwhelming influence on overall performance of two plants Gary and Sunchem:
4
•
Gary was designed to manufacture prototype samples for customers and thus for development purposes spent 0.97 U.S. dollars per hundred pounds (second highest) of Release- ease(as indicated in Exhibit 2 of the case) and has the highest Number of people working for development as a percentage of total people of 3.77 %.
•
Due to Japanese regulations Sunchem has a very large number of direct and indirect labour (from Exhibit 3 of the case) i.e. 310 much more than the second largest 86.1 of Frankfurt even though Frankfurt produces nearly 10 times the amount of Sunchem.
(a)Labour Productivity
We have used the level of education, skill and training as a measure of labour productivity. A highly skilled labour is more likely to be paid a higher wage. In line with this data we have analysed the wage/per hour in USD of the plants as per 1982 data.
Plant Average
Gross
Money
Mexico
Canada
Venezuela
Frankfurt
Gary
Sunchem
1.03
8.33
3.34
6.15
8.50
6.06
Wages/
hour in USD as per 1982 figures
Figure 3: Wages at Various Plants
We find that as expected, due to lower skilled labour in Mexico and Venezuela their corresponding wages are also lower. Lower skilled labour can be one of the causes of lower yield in plants such as Mexico and Venezuela. As opposed to them, the Frankfurt plant has skilled labour which can be a reason for its high yield.
(b)No. of package types
With 80 package sizes and a cost of 13.78 US dollars per hundred pounds of Releaseease, Gary’s number of packages could have an effect on yield. As mentioned in the case “Changing the size of a bag in the packaging line frequently took a day.” With such high setup costs for packaging, the yield is likely to reduce.
(c)Volumes/ Capacity
5
Canada has the lowest annual design capacity of 3.7 million pounds and a capacity utilisation of 70.27 %. On the other hand the Frankfurt plant has the largest capacity of 47 million pounds and a capacity utilisation of 80.85 %. Therefore the larger the capacity of the plant and its utilisation, the higher the average yield for the plant (Given in the case). Deducing from this, the Frankfurt plant should have a higher yield than average while the Canada plant will have a lower than average yield.
(d)Plant Redesign/ Maintenance and Automation
Among the plants Mexico, Canada, Venezuela and Frankfurt; Canada spends the highest i.e. 2.75 USD per hundred pounds of Release- ease. Also processes in the Frankfurt plant have a high level of automation (mentioned in the case). A high level of maintenance and automation will have a positive effect on the overall yield.
(e)Emphasis on Quality
As mentioned in the case, the Canada plant was well-regarded for the quality of its products. Also the large expenditure of the Canada plant of 1.30 USD per hundred pounds of the product shows an emphasis on quality. An increase in quality control requires more number of checks in the process thereby reducing overall yield.
The effects of the number of product lines, though present are not substantial to change the yield.
We have summarised all the factors that affect the average yield and overall performance of the firm in figure 4 (Following page).
Each of these characteristics has been rated as Low (L), Medium (M) and High (H) relative to the other plants:
6
Plant
Mexico Canada Venezuela Frankfurt Gary Sunchem
Worker Productivity/ Training L M L M M H
No. of product lines M M L H H L
No. of package types L L L L H M
Volumes/ Capacity
Plant Redesign/ Maintenance M H L H M H
M L L H M L
Product Development
Emphasis on Quality
Laws
L M L M H H
M H H M H H
M L M M M H
Figure 4: Factors affecting average yield and overall performance
Measures of Performance On the basis of our evaluation of the parameters we can separate the Sunchem plant and the Gary plant in our comparative analysis. They should have a lower measure of performance due to a batch operation for research and development in Gary and the Japanese laws in Sunchem. Among the four other plants, Frankfurt is the most efficient, followed by Mexico (high capacity), Venezuela or Canada
We have taken two parameters to evaluate our analysis:
Plant No. of pounds/ per worker Average Yield
Mexico 1061 94.7%
Canada 257 91.l%
Venezuela 470 91.7%
Frankfurt 1209 98.9%
Gary 658 90.4%
Sunchem 35 98.8%
Figure 5: Efficiency at the plants
Based on the above analysis, we find that: Factors
within the
control
of management-No
of package
types, Plant
Redesign/Maintenance and Automation, Emphasis on quality Factors
outside
the
control
of
management-Labour
productivity,
Volumes/Capacity
HANDLING THE OVER CAPACITY AT DIFFERENT PLANTS The learning from Goldratt’s book “The Goal” is that having idle capacity is preferable to locking money in inventory. We find that in the present case, the six plants are producing exactly as per the demand and therefore there is no finished goods inventory. Also, since demand for Release ease is expected to remain constant for the next five years,
7
we recommend that the plants with excess capacity may explore the possibility of exporting and meeting demands of other plants where the cost of production is higher than the production plus the transportation costs of these plants. The table below brings out the cost of producing and importing from plants with excess capacity.
COSTS OF IMPORTING RELEASE-EASE VS. LOCAL PRODUCTION
From / To Mexico Canada Venezuela Frankfurt Gary
Mexico 95.0 173.4 197.3 138.7 180.7
Rank 1 3 5 2 4
Canada 106.4 93.3 126.3 88.2 108.9
Rank 3 2 5 1 4
Venezuela 153.0 159.5 116.3 133.8 170.9
Rank 3 4 1 2 5
Frankfurt 116.1 119.2 141.6 76.7 123.7
Rank 2 3 5 1 4
Gary 110.8 108.0 132.4 91.8 102.9
Rank 4 3 5 1 2
Sunchem 115.6 117.0 138.5 95.4 122.4
Rank 2 3 5 1 4
Sunchem
268.5
6
166.8
6
249.5
6
184.0
6
174.3
6
153.8
6
Figure 6: Cost of Importing vs Producing at various plants (In cents)
Plants which are the cheapest to procure from have been ranked as 1.
POSSIBILITY 1: Assumption: Frankfurt has the packaging facility required to pack half a kg and one kg material.
The cost of production at Sunchem is high due to its high overhead costs and energy costs. From the above table, we can observe that it is cheaper to produce and export from Frankfurt rather than produce at Canada, Gary and Sunchem. The differential features of the Frankfurt plant that make it such a low cost of production have been outlined earlier in the report.
We can also observe that it is cheaper to produce at any plant and export to Sunchem than produce there. Frankfurt, which has an excess capacity of 9 Mn pounds, is best suited to export to Sunchem. However, shifting of production to Frankfurt would lead to additional packaging costs for half a kg and one kg packs, which sizes are currently not available in Frankfurt. Our analysis does not incorporate these costs for want of data from the case.
8
POSSIBILITY 2: Assumption: Production process is such that packaging capability in the plants is limited and hence half a kg and one kg are either produced or imported from plants which have packaging facility of the desired size
The sales in Pacific are 11.9 Mn pounds while the production in Sunchem is only 4 Mn pounds. The balance is being imported from other plants. Since Sunchem and Gary are the only plants which have 0.5 and 1 kg packages, it can be inferred that Gary exports the remaining (11.9 less 4 Mn pounds) to the Pacific region.
The total exports of North America are 14.2 out of which 7.9 is exported to Pacific. The balance cannot be exported to Europe whose requirements are already being satisfied by the Frankfurt plant. Hence, the balance 6.3 Mn pounds are exported to Latin America which has an excess demand of 11.9.
The remaining import requirements to the tune of 5.6 Mn pounds of Latin America are met by Europe (Frankfurt plant)
Total From / To
North Am.
W Europe
Latin Am.
Pacific
production
North Am.
19.6
0
6.3
7.9
33.8
W Europe
12.4
20
5.6
0
38
Latin Am.
0
0
4.1
0
4.1
Pacific
0
0
0
4
4
Total sales
32
20
16
11.9
79.9
Figure 7: Consumption and Export by plants
POSSIBILITY 3: Assumption: Production process is such that each plant has the capability of producing different varieties of release ease and in every possible package size and also there is no additional cost involved in such packaging then what the plants are currently spending. In such a case we have developed a Linear Programming model as follows: A linear programming model was made minimizing the total Cost of production across all plants such that the total demand across all plants is satisfied. The Cost function
9
includes the cost of production in each plant as well as the cost of exports from one plant to another. For example suppose Frankfurt is producing 100 units while exporting 30 to Mexico. The cost function will include the cost of producing 100 units in Frankfurt and also the cost of exporting 30 units to Mexico. This is done with all possible combinations of productions and exports among the plants and the cost function is minimized.
While developing the Linear programming model, certain assumptions were made: 1. All the different varieties of release ease can be manufactured in any of the six plants without any significant increase in the total cost of production. This is based on the assumption that the existing plants and machinery at every location can be used to manufacture Release Ease of different formulations and packaging without a major redesigning. 2. The exchange rates and inflation does not affect the production costs significantly. 3. Production costs are independent of capacity utilization.
The LP formulation is as follows: VARIABLES
Cap i = The capacity of plant i
( i =
1 for Mexico, 2 for Canada, 3 for
Venezuela, 4 for Frankfurt, 5 for Gary, 6 for Sunchem)
Pi = Amount of release ease produced in pounds at Plant i
Tij = Amount of release ease exported in pounds from plant i to Plant j.
Cpi= Cost of production at Plant i
CTij= Cost of transportation from plant i to Plant j.
Cdj= Duty Cost at Place j
Binary Variables are: Xi=
Yij=
1
if Release Ease has to be produced at plant i
0
If no Release Ease need to be produced at plant i
1
If Release Ease has to be exported from plant i to plant j
0
if no Release Ease needs to be exported from plant I to plant j amount is being exported from plant i to place j
OBJECTIVE FUNCION: Min (Total production cost + Total Transportation cost) 10
= Min. ( ∑ [Pi - { ∑ (Tij * Yij)}] Pi*Xi + ∑ Tij*(CTij + CPij) * (1+Cdj) * Yij) (For i=1 to 6) For j=1 to 6)
(For i,j=1 to 6)
CONSTRAINTS
Production constraint: Total production <= The capacity of the plant
Therefore,
Pi <= Cap i
Demand constraint: The demand for Release ease in any region is equal to (Total production in the region + the total import to the region – the total export form the region) The demand constraint for Western Europe is formulated below for illustration, X4*P4 + ∑ (Ti4 * Yi4) - ∑ (T4j * Y4j) = 20, 000,000. Similarly, we formulated demand constraints for each region.
Export/Import Constraints o
Total export from a plant cannot exceed the total production at that plant. For example the constraint for Mexico plant is: ∑ T1j <= P1
o
A plant cannot both export as well as import from the same plant Therefore, Yij + Yji <= 1
When, i≠j
Yij + Yji = 0
When, i=j
We have simulated the linear programming model to get minimum costs. This cost includes Production cost, transportation cost as well as duty cost. The results of the linear program are part of the Appendix.
Exchange rates and their impact on make/buy decision:
11
USD has been steadily appreciating vis-à-vis the yen since 1978. The exchange rate was 194.6 yen / one USD in 1978 which has increased to 235 yen / one USD in 1982. This has had a negative impact on the attractiveness of exporting from Gary to Sunchem. This corroborates the statement made by Tom Schultz.
As we observed from figure 6, the cost incurred by Sunchem in importing from Gary is 1.224 USD and the corresponding cost of manufacturing in Sunchem is 1.538 USD.
In yen terms, this works out to be 1.224*235 = 287.536 for importing from Gary along with transportation costs and 1.538*235 = 361.43 for production at Sunchem.
Hence under this condition, it is advisable to import from Gary than to produce at Sunchem.
FUTURE STRATEGY FOR THE GARY PLANT
Our above analysis brings out that if the Frankfurt plant is not capable of meeting Sunchem’s packaging requirements, the Gary plant can be used to produce for Sunchem. This would ensure that out of the excess capacity if 4.5 Mn pounds at Gary, 4 Mn pounds will be utilized leaving only 0.5 Mn pounds of excess capacity.
However, if Frankfurt plant meets the requirements of Sunchem, other alternatives need to be evaluated for Gary plant:
Shifting Gary’s production facility for Release ease to Frankfurt
This option does not appear viable to us for the following reasons:
•
Frankfurt has excess capacity of only 5 Mn pounds (after meeting Sunchem’s demands) whereas the demand from the Gary plant is 14 Mn pounds.
12
•
The research and Development facility will also have to be shifted to Frankfurt. Due to its R&D facility the efficiency is lower and labour requirement is higher at Gary than Frankfurt.
•
There is also an issue of relocating labour involved in production of Release-ease from the Gary plant.
Recommendations:
•
The management needs to decide on the operability of Sunchem, as operating there does not seem viable at current costs Improvement in processes and Technology
•
Have a centralized R&D team which works on product improvement, rather than having a team in each area
•
Improve
the
information
exchange
between
plants
&
enable
the
implementation of best practices in process execution across plants. –
The
process
improvements made
in Frankfurt
plant should be
implemented everywhere. •
Frankfurt uses computer control in reaction step, this should be used everywhere.
•
Also, Frankfurt does extensive solids recovery & waste treatment, this should be taken everywhere.
•
According to the management’s expectations, any machinery in this field has an expected life of 20 years. However there are cases where machinery installed in 1959 continues to be used in 1982. This is impacting the productivity and efficiency of these plants.
•
The yield increases with volume s, small plants should be scaled up.
•
The volumes at many plants (e.g. Sunchem) were constrained by low drier capacity, such bottlenecks should be removed
13
Appendix
Costs per plant Mexico
Canada
Fixed Costs (per hundred pounds) Production in 1982 Total Fixed Cost
17.58 17,200,000 3,023,760
24.55 2,600,000 638300
Variable Costs (cents per pound)
77.43
72.8
Venezuela
Frankfurt
Gary
Sunchem
25.02 4,100,000 1025820
20.34 38,000,000 7729200
25.67 14,000,000 3593800
57.38 4,000,000 2295200
91.32
56.35
77.26
96.42
Analysis with Linear Programming on Excel Transport data
Ctij Mexico Canada Venezuela Frankfurt Gary Sunchem
Mexico 0.0 11.0 7.0 10.0 10.0 14.0
Duty Cost 1+ Cdj
Mexico
Plant Variable cost Fixed Costs
Mexico
Production Decision Xi=
Optimal Prod. Pi
Plant Capacity Capi Capacity Constraint
Canada 11.4 0.0 10.0 11.5 6.0 13.0
Canada
11.0 6.0 10.4 11.2 0.0 13.0
1.045
(percentage) Sunchem 1.060
Gary
1.000
Frankfurt 1.095
72.8 638,300
Venezuela 91.32 1,025,820
Frankfurt Gary Sunchem 56.35 77.26 96.42 7,729,200 3,593,800 2,295,200
Venezuela
Frankfurt 0.573412838
Canada
Mexico
Frankfurt 11.0 11.5 13.0 0.0 10.0 14.2
Venezuela 1.500
1.600
77.43 3,023,760
Venezuela 7.0 9.0 0.0 12.5 11.0 12.5
(cents/ pound) Sunchem 14.0 13.0 14.3 13.3 12.5 0.0
Canada
Gary
Sunchem
1
1
22000022
Canada 3700000
Venezuela Frankfurt Gary Sunchem 4500000 26950403.39 18500018.5 5000000
22,000,000
Canada 3,700,000
Venezuela 4,500,000
(pounds) Frankfurt Gary Sunchem 47,000,000 18,500,000 5,000,000
22,000,000
3,700,000
4,500,000
Mexico
Mexico
1
Gary
26,950,403
1
18,500,000
14
1
5,000,000
Optimal export Tij Mexico Canada Venezuela Frankfurt Gary Sunchem Total Imports Total Exports
Mexico
Canada
0 145847117.9 136161311.6 138928704.9 153883385 629234575.5 1204055095 1211710927
0 0 0 19181572.35 0 0 19181572.35 214842903.3
Venezuela Frankfurt Gary Sunchem 614360264.5 344112673.6 93443884.92 159794103.5 5371.683933 5371.683933 64550204.69 4434837.382 0 0 92808062.84 2274780.514 0 0 106110533 9387069.369 7477340.783 0 0 4434837.381 0 0 0 0 621842977 344118045.3 356912685.4 180325628.2 231244154.9 273607879.6 165795563.2 629234575.5
Cost of Production and Transportation of Exports Mexico Canada Mexico 77.4 88.8 Canada 83.8 72.8 Venezuela 98.3 101.3 Frankfurt 66.4 67.9 Gary 87.3 83.3 Sunchem 110.4 109.4
Venezuela 84.4 81.8 91.3 68.9 88.3 108.9
Constraints Demand Demand Constraint
WE 97,460,569 20,000,000
LA Pacific 395,098,822 -443,908,947 16,000,000 11,900,000
Canada
Venezuela
Plant wise Costs Costs Same region Exports Objective Function
NA 32,000,000 32,000,000
Mexico -92119315344 1.35248E+11
Frankfurt 88.4 84.3 104.3 56.4 87.3 110.6
Gary 88.4 78.8 101.7 67.6 77.3 109.4
Sunchem 91.4 85.8 105.6 69.7 89.8 96.4
Frankfurt Gary Sunchem -1.5371E+10 20706276226 13899148782 -1.138E+10 60188697771 25275127268 31539742080 24233494932 22896463091 1.11168E+11
136,711,312,363
15
