GE401 - Engineering Economics

Design Optimization (Present Economy Studies)

Design optimization
Here cost depends on a design variable
Problem is to find best value
Present economy studies
Preview of what well be doing
Time element missing: here costs are all in the present
Example: passenger jet
Operating cost = knv3/2, k = const
n = trip length (km), v = velocity (km/h)
At 400 km/h, average cost is S.R.300/km
Suppose passengers time costs S.R.300K/hr
What velocity will minimize total cost?
The 2 cost types behave differently with v
So we have a tradeoff situation
What are we trying to do?
Minimize total cost
So first write the total cost
T = knv3/2 + (300000)(n/v)
Whats k? Use average cost info
At v = 400, (knv3/2)/n = 300 (S.R./km)
So k = 300(400-3/2) = 300/8000 = .0375 [dimensions are
S.R.(hrs3/2)/km5/2]
Find a critical point
Go back to total cost
T = .0375nv3/2 + (300000)(n/v)
This is linear in n: we can analyze T/n (cost/km)
 Find critical point with respect to v
Derivative must be zero there
So 0 = (3/2)(.0375)v1/2 300000v-2
.05625v5/2 = 300000
So v = 490.7 km/h (only critical point)
But is this a maximize or a minimize?
Check if it does the job
Derivative is .05625v1/2 300000v-2
Second derivative will be (1/2).05625v-1/2 + 600000v-3
This is positive (no need to compute!)
So 490.7 km/h is a minimize
No other minimize (no other critical point)
See spreadsheet for graph of cost
Notice what we did here
We had infinitely many possible designs
One for each value of v
Out of all these we found the one giving least total cost
Calculus was necessary we cant look at each design
With finite number, we could look at each
Two basic kinds of problems
Finitely many alternatives
Can list them and compare
Infinitely many alternatives
Necessary to have a cost model or formula
Then apply mathematical analysis
Much more of this in GE401
Present economy studies
Here were concerned with selecting a best alternative
All costs/revenues occur at same time
 So time value of money doesnt matter
Basic method: careful comparison of
Profitability (typical in industry), or
Other measure of effectiveness (MOE) (government or nonprofits)
Example: 2 machines
Machine 1
Produces 150 parts/hr (8 hr/day)
Reject rate 6%
Good parts sell for S.R.10 each
Operating cost
Operator S.R.16.00/hr
Traceable overhead S.R.6.25/hr
Material costs S.R.4/part
Machine 2
Produces 200 parts/hr (7 hr/day)
Reject rate 12%
Good parts sell for S.R.10 each
Operating cost
Operator S.R.18.00/hr
Traceable overhead S.R.5/hr
Material costs S.R.4/part
How to figure profit?
Profit = revenue expense
Here we look at profit generated by this machine
Dont consider overhead other than traceable
Why?
Well compare figures for the two machines
Select the one generating better profit
Machine 1
Good parts/day will be (8)(150)(1 - .06) = 1,128
Revenue/day = (10)(1128) = S.R.11,280
Costs
Materials (8)(150)(4) = S.R.4,800
Operation (16 + 6.25)(8) = S.R. 178
Profit S.R.11,280 S.R.4,978 = S.R.6,302/day
Machine 2
Good parts/day will be (7)(200)(1 - .12) = 1,232
Revenue/day = (10)(1232) = S.R.12,320
Costs
Materials (7)(200)(4) = S.R.5,600
Operation (18 + 5)(7) = S.R. 161
Profit S.R.12,320 S.R.5,761 = S.R.6,559/day
Conclusion
Machine 2 is slightly more profitable
About 4% better
If there are no external considerations, probably pick Machine 2
External considerations could include
Reliability
Space required
Etc.
Make-or-buy decisions
Payroll section of agency
5 payroll specialists
Supervision by business manager
Consultant looking at agencys staff
Considers possibility of using outside payroll service for a fee
Consultants analysis
 Yearly cost of payroll section
5 specialists: S.R.145,200
of business manager: S.R.16,500
Supplies used: S.R.4,800
Office space (firms building): S.R.24,000
Share of SG&A overhead: S.R.23,400
Total cost S.R.213,900/year
Consultants recommendation
Outside payroll service will contract to do the work for S.R.170,000
Shut down payroll department and save S.R.43,900/year
Firm is considering this idea
What should it do?
Basic comparison principle
Look at what costs will be with payroll department, and without it
Cost without department:
Fee: S.R.170,000
business manager: S.R.16,500
Office space (used by firm): S.R.24,000
Share of SG&A: S.R.23,400
Total: S.R.233,900
Cost with department: S.R.213,900
Cost without department: S.R.233,900
Very different from consultants S.R.170,000
What happened?
Many of the costs here would be paid by company anyway (e.g.
space, part of manager, SG&A).
No savings from these
So actual difference came down to
Fee of S.R.170,000 vs.
 Salaries and supply cost of S.R.150,000
Cheaper to keep the department going
Key error here was treating fixed costs as if they were variable costs
What Weve Covered
Design optimization
Cost depends on parameters
Can handle infinitely many possibilities
Present economy studies
No time element
Careful comparison of costs
Essential to understand whats fixed and whats variable
Again . Costs and Design Economics
Identifying costs:
Fixed, variable and incremental costs
Recurring and non-recurring costs
Direct, indirect and overhead costs
Cash costs and book costs
Sunk costs
Opportunity costs
Life-cycle costs

A riding problem:
400 kilometers each way. Annual kilometerage 15,000.
Cost Element Cost Per Kilometer
Gas S.R. 0.120
Oil 0.021
Tires 0.027
Depreciation 0.150
Insurance and Tax 0.024
Repairs 0.030
Garage 0.012
Total: 0.384

Sunk Costs:
Mr. Hassen bought a bad second hand
mower machine for S.R.100, hoping to
spend additional S.R.160 on accessories and
repair it. Then he would be able to sell it for
S.R.500. However, after spending S.R.200,
he found that he would still need additional
S.R.250 to finish the repairing. Shall he
continue or stop?
The sunk cost in this case is S.R.300.

An idealized situation:
Demand as an affine function of the price:
p = a - bD
where 0 D a / b with a > 0 and b > 0.

The total revenue is:

TR(D) = p(D)D = aD - bD2.

The maximum of TR is obtained at

D = a / 2b.

Suppose that the cost function is

C(D) = Cf + CvD.
Then we have
Profit(D) = TR(D) - C(D)
= -Cf +(a - Cv)D - bD2.
To maximizing the profit we get an optimal
demand size:
D* = (a Cv) / 2b.
Notes:
(1) We must have a > Cv to guarantee a real
profit;
(2) Is D* a good solution?

Break-even analysis
In the engineering economics studies it is
often important to know what is the
boundary.

Interested in the D value such that TR(D) =

C(D), i.e.
aD' - bD' 2 = Cf +CvD'

D' = (-(a - Cv) (a Cv)2 - 4bCf )/-
2b

Example:
A company producing electronic timing
switch.
Per month: Cf = S.R.73; 000 and Cv =
S.R.83.
Moreover, p = S.R.180 - 0:02(D).
Then,
D* = (a Cv) / 2b.
= 2, 425
units per month.

The total optimal profit is

[180(2, 425) - 0:02(2,
425)2]
-[73, 000+83(2, 425)]
= S.R .44, 612
and the break-even points are
D'1 = 932
and
D'2 = 3, 918

Average unit cost function:

Cu = Total Costs / Demand Volume

In case of affine cost structure we have

Cu = (Cf +CvD) / D = Cv +Cf / D.:
Cost-driven design optimization:
Operating cost of a jet-powered airplane is
Co = knv3/2
where k is a constant, n is the distance
covered and v is the velocity. It is known
that the average cost is S.R.300 per
kilometer if operated at 400 kilometer/hour.
Assume that the disutility of one hour on
board is S.R.300,000. Decide the optimal
velocity.

First we decide the value of the constant k:

Co/n = k(400)3/2 = 300
yielding
k = 0.0375.
The total cost is:
Ct(v) = knv3/2 +300, 000 n/v.
By setting dCt(v)/dv to zero we solve
V* = 490.68 kilometer/hour.

Care must be taken to analyze the cause of

costs/benefits.

Example: Make-versus-buy.

Dept A of a company produces a product X.

The daily production is 576 units of X.
Cost accounting record shows:

Labor: 4 hours at S.R.22.5/hr

Manager S.R.30 p.d.
S.R.120.00
Material: S.R.86.40
Overhead: floor area rental
S.R.82.00
Total cost: S.R.288.40

An outside company sells X at S.R.0.35 per

unit.
This means that the purchase price will be
S.R.0.35 x 576 = 201.60.
Should X be produced or purchased?

Present Economy Studies.

So far, an important assumption is implicitly
made, i.e. the decisions are for one period
only.
There is no time value involved in costs and
benefits.
This is important as many applications fall
into this category.
It is restrictive too, simply because life is a
dynamic process. Time and timing play an
important role in decision making.

Exercises:
1. A company produces circuit boards used
to update outdated computer equipment.
The fixed cost is S.R.42,000 per month and
the variable cost is S.R.53 per circuit board.
The selling price per unit is p =
S.R.1500:02D. Maximizing output of the
plant is 4,000 units per month. a. Determine
optimum demand for this product. b. What
is the maximum profit per month?
c. At what volumes does break-even occur?
d. What is the companys range of profitable
demand?

2. The cost of operating a large ship varies

as the square of its velocity; specifically C =
knv2 where n is the number of the trip length
in kilometers, and k is a constant. It is
known that at 12 kilometers/hour the
average cost of operation is S.R.100 per
kilometer. The owner of the ship wants to
minimize the cost of operation, but it must
be balanced against the cost of the
perishable cargo cost, which the customer
has set
as S.R.1,500 per hour. At what velocity
should
the trip be planned to minimize the total
cost?