OPTIMIZATION IN THE OIL &

GAS INDUSTRY
SEPTORATNO SIREGAR
Silvya Dewi Rahmawati

1
 RELATED SUBJECTS
• OPERATIONS RESEARCH
• MANAGEMENT SCIENCE
• QUANTITATIVE DECISION ANALYSIS
• MANAGEMENT DECISION ANALYSIS
• OPTIMIZATION TECHNIQUES

2
 WHAT IS OPTIMIZATION?
• OPTIMUM: PARAMETERS THAT WE CONTROL
TO GET A MAXIMUM OR MINIMUM
• MAXIMUM, MINIMUM: OUTCOME OF
CONTROLLING THE OPERATING PARAMETERS
WITHIN A RANGE OF APPLICATION /
CONSTRAINTS

3
 WHY DO WE OPTIMIZE?
• TO OBTAIN A DESIRED OBJECTIVE
• RESOURCES ARE LIMITED
• MANY INTERESTS , THUS MANY
CONSTRAINTS

4
 DECISION PROCESS
• MAKE A LIST OF ALTERNATIVES
• IDENTIFY THE SYSTEM CONSTRAINTS
• SET THE OBJECTIVE CRITERIA
• LIST OUR ASSUMPTIONS
• BUILD A DECISION MODEL
• FIND THE OPTIMUM SOLUTION
• PROPOSE THE SOLUTION TO HIGHER LEVEL
DECISION MAKERS

5
 MODELING

REAL
SYSTEM
PROPOSED
DECISION
SSSSSSIMIMSSS
rr
ASSUMED
SIMSSPS
REAL SOLUTION
MODEL
SYSTEM

6
 DECISION MODEL
• PROBABILISTIC OR DETERMINISTIC
• PROCESS MODELING:
1. DEFINE THE PROBLEM
2. MODEL CONSTRUCTION
3. FIND THE MODEL SOLUTION
4. RUN SENSITIVITY ANALYSIS
5. MODEL VALIDATION
6. IMPLEMENTATION
7. MODEL IMPROVEMENT (FROM FEEDBACK/UPDATE)

7
 ASSETS IN OIL & GAS INDUSTRY
• RESERVES: SIZE / VOLUME & ENERGY
• HUMAN RESOURCES
• FACILITIES:WELL, ROAD, PIPELINE
• INFORMATION (DATA), INCENTIVE, MARKET
DEMAND, TECHNOLOGY

8
 OPTIMIZATION EXAMPLES IN
OIL & GAS (1)
• DRILLING: RPM, WOB, MUD HYDRAULICS
• PRODUCTION: SEPARATOR (PRESSURE &
RATE)
• RESERVOIR: SPACING/NO OF WELLS, EOR/IOR
• HEAT MANAGEMENT IN STEAM INJECTION:
FRONTAL ADVANCE RATE VS GRAVITY
OVERRIDE
• POD STRATEGY: DEVELOP, REVITALIZE
• EKSPLORASI 9
 OPTIMIZATION EXAMPLES IN
OIL & GAS (2)
• GAS OR OIL PIPELINE NETWORK:
– MINIMUM SPANNING TREE
– SHORTEST ROUTE
– MAXIMUM FLOW
• OPTIMIZING COMPRESSOR POSITION
• OPTIMIZING PIPELINE DIAMETER
• MINIMIZING DEVELOPMENT PROJECT COSTS
• PRODUCTION ALLOCATION
10
 PARAMETERS USED AS INDICATORS
• INVESTMENT
• RESERVES
• ECONOMIC LIFE TIME
• PAYOUT TIME
• RATE OF RETURN (ROR)
• NET PRESENT VALUE (NPV)
• PROFIT TO INVESTMENT RATIO (PIR)
• DEVELOPMENT COSTS

11
 OPTIMUM PARAMETER TYPES
• DISCRETE, EX: NO OF WELLS THAT MAXIMIZE
PROFIT
• CONTINUOUS: FINDING AN OPTIMUM
SOLUTION FROM A CONTINUOUS FUNCTION
(LINEAR / NON LINEAR)

12
 OPTIMIZATION MODELS (1)
• TWO TYPES:
1. DETERMINISTIC
2. PROBABILISTIC
• LINEAR PROGRAMMING: GRAPHICAL
METHOD, ENUMERATION, SIMPLEX
• DECISION TREE ANALYSIS
• ALLOCATION: ASSIGNMENT,
TRANSPORTATION
13
 OPTIMIZATION MODELS (2)
• NETWORK ANALYSIS: MINIMAL SPANNING
TREE, SHORTEST ROUTE, MAXIMAL FLOW
• CRITICAL PATH METHOD (CPM), PROGRAM
EVALUATION & REVIEW TECHNIQUE (PERT)
USED IN PROJECT MANAGEMENT
(PLANNING, SCHEDULING, CONTROLLING,
REVISING, COST MINIMIZATION BY
ACCELERATION, RESOURCE LEVELLING)
• COST TIME TRADE OFF
14
 OPTIMIZATION MODELS (3)
• QUEUEING / WAITING LINE
• INVENTORY
• SIMULATION

15
 MATHEMATICAL / LINEAR
PROGRAMMING
• LP : INTRODUCED IN LATE 1940’S BY
G.DANTZIG
• APPLICATION: MILITARY, INDUSTRY,
AGRICULTURE, TRANSPORTATION,
ECONOMICS, HEALTH SYSTEMS, BEHAVIORAL
AND SOCIAL SCIENCES
• A TWO-VARIABLE MODEL : GRAPHICAL
SOLUTION

16
 A TWO-VARIABLE MODEL :
GRAPHICAL SOLUTION
EXAMPLE:
• A COMPANY PRODUCES 2 PAINT PRODUCTS
(INT & EXT)
• TWO RAW MATERIALS, A & B ARE BASIC
COMPONENTS OF THE PAINTS
• MAXIMUM AVAILABILITY OF A: 6 TONS/DAY
• MAXIMUM AVAILABILITY OF B: 8 TONS/DAY

17
DAILY REQUIREMENTS OF THE RAW MATERIALS PER
TON OF INT & EXT PAINTS:

TONS OF RAW MAXIMUM

RAW MATERIAL AVAILABILITY
MATERIAL PER TON OF
PAINT (TONS)

EXT INT
A 1 2 6
B 2 1 8

18
 MARKET SURVEY
• DAILY DEMAND FOR INT PAINT CANNOT
EXCEED THAT OF EXT PAINT BY MORE THAN 1
TON
• MAX DEMAND FOR INT PAINT : 2 TONS/DAY
• WHOSALE PRICE:
– EXT PAINT : $3000/TON
– INT PAINT : $2000/TON

19
 PROBLEM
• HOW MUCH INT & EXT PAINTS SHOULD THE
COMPANY PRODUCE DAILY TO MAXIMIZE
GROSS INCOME?

20
 MODEL CONSTRUCTION
BASIC COMPONENTS OF A MODEL:
1. VARIABLES (UNKNOWNS) OF THE PROBLEM
2. CONSTRAINTS IMPOSED ON THE VARIABLES
3. OBJECTIVE (GOAL) TO BE ACHIEVED

21
VARIABLES:
XE = TONS PRODUCED DAILY OF EXT PAINT
XI = TONS PRODUCED DAILY OF INT PAINT

OBJECTIVE FUNCTION:
MAXIMIZE Z = 3 XE + 2 XI

SUBJECT TO CONSTRAINTS:
1. RAW MATERIAL A: XE + 2 X I ≤ 6 (1)

2. RAW MATERIAL B: 2XE + XI ≤ 8 (2)

3. DEMAND RESTRICTIONS 1: -XE + XI ≤ 1 (3)

4. DEMAND RESTRICTIONS 2: XI ≤ 2 (4)

5. NON NEGATIVITY CONSTRAINTS: XE ≥ 0 ; X I ≥ 0 (5) , (6)

22
 GRAPHICAL SOLUTION
Unbounded criteria &
6 redundant constraint
XI
LOOK FOR PARALLEL LINE TO
REVENUE LINE THE REV LINE INTERSECTING
Z = 3 XE + 2 XI THE OUTER MOST POINT OF
THE SOLUTION SPACE
3
OPTIMUM SOLUTION:
XE = 3 1/3 TONS; XI=1 1/3 TONS
MAXIMUM REV = 12 2/3 THOUSAND $

G
H E D
K
4
F SOLUTION C
SPACE J XE 5
I A B
2 1

23
 ENUMERATION TECHNIQUE
• CALCULATE THE VALUE OF EACH CORNER
(EXTREME) POINT OF THE SOLUTION SPACE:
A: Z=0
B: Z = 3(4) + 2(0) = 12
C: Z = 3(3⅓) + 2 (1⅓) = 12 ⅔ THE HIGHEST VALUE
D: Z = 3(2) + 2(2) = 10
E: Z = 3(1) + 2(2) = 7
F: Z = 3(0) + 2(1) = 2

24
 SENSITIVITY ANALYSIS (1)
1. CHANGE ALLOWED IN THE OBJECTIVE
FUNCTION: Z = CEXE + CIXI

– SOLUTION WILL NOT CHANGE WITHIN THE

RANGE:
½ ≤ CE/CI ≤ 2

– FOR CI = 2, THE RANGE IS: 1 ≤ CE ≤ 4

– FOR CE = 3, THE RANGE IS: 3/2 ≤ CI ≤ 6

25
 SENSITIVITY ANALYSIS (2)
2. CHANGES ALLOWED IN THE RIGHT-HAND
SIDE OF THE CONSTRAINTS:
• INTERSECTION OF CONSTRAINTS 1 & 2 ( WITH
ZERO SLACKS):
4 ≤ AMOUNT OF RAW MATERIAL A ≤ 7 (POINTS B & K)
6 ≤ AMOUNT OF RAW MATERIAL B ≤ 12 (POINTS D & J)
• CONSTANTS OF PROPORTIONALITY:
Y1 = (13 -12)/(7-4) = 1/3 THOUSAND $ PER TON A
Y2 = (18-10)/(12-6)= 4/3 THOUSAND $ PER TON B
• Y1 & Y2 ARE WORTH PER UNIT OR DUAL OR
SHADOW PRICES
26
 SENSITIVITY ANALYSIS (3)
• INTERSECTION OF CONSTRAINTS (3) & (4), THAT
DO NOT PASS THE OPTIMUM POINT (WITH NON
ZERO SLACKS):
RHS OF CONSTRAINT (3) CAN BE INCREASED
INDEFINITELY WITHOUT CHANGING THE OPTIMUM,
SO THE WORTH PER UNIT IS ZERO:
-2 ≤ EXCESS OF INT OVER EXT PAINT ≤ ∞
RHS OF CONSTRAINT (4): SIMILAR AS (3):
4/3 ≤ MAX DEMAND FOR INT PAINT ≤ ∞

27
 LINEAR PROGRAMMING:
THE SIMPLEX METHOD (1)
• ESSENTIALLY SIMPLEX PUTS ALGEBRAICALLY THE
EXTREME POINT OR CORNER POINT OF THE
SOLUTION SPACE
• INITIAL STEP: EACH CONSTRAINT BE PUT IN A
SPECIAL STANDARD FORM (EQUATION) BY
AUGMENTING SLACK OR SURPLUS VARIABLES
• THE NO OF VRB USUALLY > NO OF EQS, SO THERE
CAN BE AN INFINITE NO OF SOLUTION POINTS
(COMPARE WITH GRAPHICAL METHOD)
• EXTREME POINTS ↔ BASIC SOLUTIONS
28
 LINEAR PROGRAMMING:
THE SIMPLEX METHOD(2)
• SIMPLEX IDENTIFIES A STARTING BASIC SOLUTION,
THEN MOVE TO OTHERS (ITERATION) THAT HAVE
POTENTIAL TO IMPROVE THE VALUE OF THE
OBJECTIVE FUNCTION, UNTIL THE ONE
CORRESPONDING TO THE OPTIMUM WILL BE
IDENTIFIED
• TWO VARIANTS OF SIMPLEX ALGORITHMS:
PRIMAL & DUAL. BOTH WILL RESULT IN THE SAME
OPTIMUM. PRIMAL MOVES IN THE FEASIBLE, DUAL
IN THE INFEASIBLE SPACE
29
 STANDARD LINEAR
PROGRAMMING (LP) FORM
1. ALL CONSTRAINTS ARE EQUATIONS (FOR
PRIMAL: RHS IS NON NEGATIVE)
2. ALL VARIABLES ARE NON NEGATIVE
3. OBJECTIVE FUNCTION MAY BE
MAXIMIZATION OR MINIMIZATION

30
 CONSTRAINTS
TYPE ≤ (≥) CONSTRAINT CAN BE CONVERTED
TO AN EQUATION BY ADDING A SLACK VRB
TO (SUBTRACTING A SURPLUS VRB FROM)
THE LEFT SIDE OF THE CONSTRAINT

X1 + X2 ≤ 6 BECOMES X1 + X2 + S1 = 6 , S1 ≥ 0

3 X1 + 2 X2 - 3 X3 ≥ 5 BECOMES 3 X1 + 2 X2 - 3 X3 – S2
= 5 , S2 ≥0
31
 EXAMPLE
OBJECTIVE FUNCTION:
MAXIMIZE Z = 3 XE + 2 XI

SUBJECT TO CONSTRAINTS:
1. RAW MATERIAL A: XE + 2 X I ≤ 6 (1)

2. RAW MATERIAL B: 2XE + XI ≤ 8 (2)

3. DEMAND RESTRICTIONS 1: -XE + XI ≤ 1 (3)

4. DEMAND RESTRICTIONS 2: XI ≤ 2 (4)

5. NON NEGATIVITY CONSTRAINTS: XE ≥ 0 ; X I ≥ 0 (5) , (6)

32
 STANDARD FORM
MAXIMIZE Z = 3XE + 2XI + 0S1 + 0S2 + 0S3 + 0S4

SUBJECT TO CONSTRAINTS:
1. XE + 2 XI + S1 =6 (1)

2. 2XE + XI + S2 =8 (2)

3. -XE + XI + S3 =1 (3)

4. XI + S4 = 2 (4)

5. XE , XI , S1, S2 ,S3, S4 ≥ 0 (5) , (6)

THE MODEL HAS m=4 EQUATIONS & n=6 VARIABLES,THUS THE NO.
OF NON BASIC (ZERO) VRB = 6-4=2
33
• IF WE CHOOSE X1=0 & X2=0, WE GET THE BASIC FEASIBLE
SOLUTION S1=6,S2=8,S3=1 & S4=2 (ORIGIN POINT A)
• THAT BASIC SOLUTION IS THE STARTING SOLUTION (OR
ITERATION) OF THE SIMPLEX METHOD
• THE OBJECTIVE FN: Z-3XE-2XI=0
• SINCE XE & XI AT POINT A = 0, THE VALUE OF Z IS GIVEN BY
THE RHS OF THE EQ (=0)
• THE NEXT STEP: MOVE TO A NEW BASIC SOLUTION
• PRINCIPLES:
– TO KEEP THE SOLUTION BASIC, BRING 1 NON BASIC VRB (0) AT A
TIME
– THE NEW NON BASIC VRB HAS TO IMPROVE THE VALUE OF THE
OBJECTIVE FN
• CHOOSE ENTERING VRB, WHICH GIVES THE MAXIMUM
CONTRIBUTION IF WE INCREASE BY 1 UNIT, BY CHOOSING
34
 THE VRB WITH THE MOST NEGATIVE COEFFICIENT (FOR
MAXIMIZING) IN THE OBJECTIVE FN, HOPING (HEURISTIC)
THAT THE SIMPLEX METHOD WILL MAKE THE LARGEST
JUMPS IN THE OBJECTIVE VALUE
• 1 UNIT INCREASE IN XE WILL INCREASE Z BY 3, 1 UNIT
INCREASE IN XI WILL INCREASE Z BY 2, SO CHOOSE XE AS THE
ENTERING VRB
• THE NEW BASIC SOLUTION MUST KEEP THE NO OF BASIC
VRB = m (=4), SO 1 BASIC VRB MUST LEAVE THE SOLUTION
(1 OF S1,S2,S3, OR S4)
• THE VALUE OF ENTERING VRB CORRESPONDS TO POINT B
(S2=0, SINCE XE=4,XI=0); ANY INCREASE BEYOND B WILL PUT
US OUTSIDE THE FEASIBLE SPACE
• LEAVING VRB CAN BE SELECTED DIRECTLY FROM THE
CONSTRAINT EQS BY COMPUTING THE NON NEGATIVE
35
 INTERCEPTS OF ALL THE CONSTRAINTS WITH THE XE AXIS
• THE SMALLEST INTERCEPT WILL IDENTIFY THE LEAVING
VRB (THE SAME STEP APPLIES FOR MINIMIZATION
PROBLEM)
• IN OUR MODEL, ONLY CONSTRAINTS (1) & (2) INTERSECT
THE XE AXIS IN THE POSITIVE DIRECTION, WHICH ARE
6/1=6 AND 8/2=4
• SINCE THE SMALLER INTERCEPT (=4) IS ASSOCIATED WITH
THE 2ND CONSTRAINT, THE BASIC VRB S2 MUST LEAVE THE
SOLUTION
• WE CAN AUTOMATE THE PROCESS OF SELECTING THE
LEAVING VRB BY COMPUTING ALL THE INTERCEPTS OF THE
CONSTRAINTS WITH THE XE AXIS AS THE RATIO OF THE RHS
TO THE CORRESPONDING CONSTRAINT COEF OF XE:

36
• CONSTRAINT 1 INTERCEPT = 6/1 = 6
CONSTRAINT 2 INTERCEPT = 8/2 = 4
CONSTRAINT 3 INTERCEPT = 1/-1 = -1
CONSTRAINT 4 INTERCEPT = 2/0 = INFINITY (PARALLEL TO
THE XE AXIS)
• THUS S2 MUST BE THE LEAVING VRB
• SELECTING THE ENTERING VRB: OPTIMALITY CONDITION
• SELECTING THE LEAVING VRB: FEASIBILITY CONDITION

37
 SUMMARY
OPTIMALITY CONDITION
THE ENTERING VRB IN MAXIMIZATION (MINIMIZATION) IS THE
NON BASIC VRB WITH THE MOST NEGATIVE (POSITIVE) COEF
IN THE OBJECTIVE Z EQ. A TIE MAY BE BROKEN ARBITRARILY.
THE OPTIMUM IS REACHED WHEN ALL THE NONBASIC COEF IN
THE Z EQ ARE NONNEGATIVE (NONPOSITIVE)

FEASIBILITY CONDITION
FOR BOTH MAX & MIN PROBLEMS,THE LEAVING VRB IS THE
CURRENT BASIC VRB HAVING THE SMALLEST INTERCEPT
(MINIMUM RATIO WITH STRICTLY POSITIVE DENOMINATOR IN
THE DIRECTION OF THE ENTERING VRB. A TIE IS BROKEN
ARBITRARILY
38
ITERATIVE STEPS OF PRIMAL SIMPLEX METHOD
• STEP 0:USING THE STANDARD FORM (WITH ALL
NONNEGATIVE RHS), DETERMINE A STARTING BASIC
FEASIBLE SOLUTION
• STEP 1:SELECT AN ENTERING VRB FROM AMONG THE
CURRENT NONBASIC VRBS USING THE OPTIMALITY
CONDITION
• STEP 2:SELECT THE LEAVING VRB FROM THE CURRENT BASIC
VRBS BY USING THE FEASIBILITY CONDITION
• STEP 3:DETERMINE THE VALUES OF THE NEW BASIC VRBS BY
MAKING THE ENTERING VRB BASIC AND THE LEAVING VRB
NONBASIC.
• STEP 4:GO TO STEP 1

39
• THE PROCESS OF SWAPPING A BASIC VRB WITH A NONBASIC
VRB IS EXACTLY EQUIVALENT TO MOVING BETWEEN
ADJACENT EXTREME POINTS ALONG THE EDGES OF THE
SOLUTION SPACE (COMPARE WITH THE FIGURE).
• THE PROCESS USED GAUSS-JORDAN METHOD
• THE MAIN GOAL OF THE GAUSS-JORDAN COMPUTATIONAL
PROCEDURE IS TO TRANSFORM THE EQS IN A MANNER
THAT ENABLES US TO OBTAIN THE NEW BASIC SOLUTION BY
ASSIGNING ZERO VALUES TO THE CURRENT NONBASIC VRBS

40
 EXAMPLE
TABLEAU 1 SUMMARY

BASIC Z XE XI S1 S2 S3 S4 SOLUTION
Z 1 -3 -2 0 0 0 0 0 Z EQUATION
S1 0 1 2 1 0 0 0 6 S1 EQUATION

S2 0 2 1 0 1 0 0 8 S1 EQUATION

S3 0 -1 1 0 0 1 0 1 S1 EQUATION

S4 0 0 1 0 0 0 1 2 S1 EQUATION

SI :
BASIC VRBS POINT A IN THE GRAPH
VALUES OF
XE &XI: BASIC VRBS
NONBASIC
VRB (=0)
41
 ENTERING VRB/PIVOT COLUMN

BASIC Z XE XI S1 S2 S3 S4 SOLUTION XE
INTERCEPTS
(RATIOS)
Z 1 -3 -2 0 0 0 0 0
S1 0 1 2 1 0 0 0 6 6/1=6
S2 0 2 1 0 1 0 0 8 8/2=4
S3 0 -1 1 0 0 1 0 1 -
S4 0 0 1 0 0 0 1 2 -

PIVOT EQUATION / PIVOT ELEMENT

ROW/LEAVING VRB

42
 NEW BASIC SOLUTION
• THE NEW BASIC SOLUTION MUST NOW INCLUDE S1, XE, S3
AND S4
• THIS IS ACHIEVED BY APPLYING THE GAUSS-JORDAN
ELIMINATION METHOD
• THE METHOD EFFECTS A CHANGE OF BASIS BY USING 2
TYPES OF COMPUTATIONS:
1. PIVOT EQUATION:
NEW PIVOT EQ = OLD PIVOT EQ : PIVOT ELEMENT
2. ALL OTHER EQUATIONS, INCLUDING Z:
NEW EQ = (OLD EQ) – (ITS ENTERING COLUMN COEFFICIENT) X (NEW PIVOT EQ)

43
 ENTERING VRB/PIVOT COLUMN

BASIC Z XE XI S1 S2 S3 S4 SOLUTION XE
INTERCEPTS
(RATIOS)
Z 1 -3 -2 0 0 0 0 0
S1 0 1 2 1 0 0 0 6 6/1=6
S2 0 2 1 0 1 0 0 8 8/2=4
S3 0 -1 1 0 0 1 0 1 -
S4 0 0 1 0 0 0 1 2 -

PIVOT EQUATION / PIVOT ELEMENT

ROW/LEAVING VRB

44
 TABLEAU 2 SUMMARY
BASIC Z XE XI S1 S2 S3 S4 SOLUTION XI
INTERCEPTS
(RATIOS)

Z 1 0 -1/2 0 3/2 0 0 12

S1 0 0 3/2 1 -1/2 0 0 2 2/(3/2)= 4/3

XE 0 1 1/2 0 1/2 0 0 4 4/(1/2)= 8

S3 0 0 3/2 0 1/2 1 0 5 5/(3/2)= 10/3

S4 0 0 1 0 0 0 1 2 2/1= 2

POINT B IN THE GRAPH

45
 TABLEAU 3 SUMMARY
BASIC Z XE XI S1 S2 S3 S4 SOLUTION

Z 1 0 0 1/3 4/3 0 0 12 2/3

XI 0 0 1 2/3 -1/3 0 0 4/3

XE 0 1 0 -1/3 2/3 0 0 10/3

S3 0 0 0 -1 1 1 0 3
S4 0 0 0 -2/3 1/3 0 1 2/3

THIS TABLEAU IS OPTIMAL BECAUSE NONE OF THE NONBASIC

VRBS HAVE A NEGATIVE COEF IN THE Z –EQUATION
(POINT C IN THE GRAPH METHOD)

46
 DECISION ANALYSIS (1)
• IN SOME CASES, PERFECT INFORMATION
(LEADING TO DECISION MAKING UNDER
CERTAINTY) IS NOT AVAILABLE
• PARTIAL OR IMPERFECT INFORMATION
LEADS TO:
1. DECISIONS UNDER RISK
2. DECISIONS UNDER UNCERTAINTY
• TWO EXTREMES: CERTAINTY AND
UNCERTAINTY
• IN BETWEEN SITUATION: RISK
47
DECISION ANALYSIS (2)
RISK ANALYSIS:
QUANTIFYING PROBABILITY

EXPECTED VALUE

DECISION MAKING BASED

ON MINIMUM EV COST OR
MAXIMUM EV PROFIT CRITERIA

48
 OIL WELL DRILLING EXAMPLE
I = 1 2
IF WE DRILL THE WELL SUCCESS FAIL

CONDITIONAL PROFIT (CP),$ 1,000,000 -500,000

PROBABILITY (P) 0.7 0.3
(EVP)I = PI X CPI , $ 0.7 X 1,000,000 = 700,000 0.3 X (-500,000) = -150,000

EVDA = ∑ (EVP)I = $700,000 +(- $150,000) = $550,000

BECAUSE EVDA OF DRILLING IS POSITIVE,THE DECISION IS: DRILL

(EVP)I = EXPECTED VALUE PROFIT OF EVENT I

EVDA = EXPECTED VALUE OF DECISION ALTERNATIVE

49
 DECISION TREE
• FOR MULTIPLE STAGE, IN WHICH DEPENDENT
DECISIONS ARE MADE IN TANDEM, THE
DECISION PROCESS CAN BE REPRESENTED
GRAPHICALLY

50
 DECISION TREE EXAMPLE
(TAHA SEC 12.2)
• A COMPANY HAS OPTIONS NOW: BUILD A
FULL OR SMALL SIZE PLANT THAT CAN BE
EXPANDED LATER
• IF FUTURE DEMAND IS HIGH: NOW BUILD
FULL SIZE PLANT; IF NOT, BUILD A SMALL
PLANT NOW AND DECIDE IN 2 YRS WHETHER
IT SHOULD BE EXPANDED
• THE DECISION INVOLVES 2 STAGES:
1. NOW: BIG OR SMALL SIZE PLANT?
2. IF BULD SMALL NOW, EXPAND TWO YRS LATER
51
 HIGH DEMAND
$1,000,000/YR
P=0.75
2
LOW DEMAND
$300,000/YR
P=0.25

HIGH DEMAND
$900,000/YR
P=0.75

1 5

LOW DEMAND
$200,000/YR
4 P=0.25

HIGH DEMAND
$250,000/YR
P=0.75
3 6
LOW DEMAND
$200,000/YR
P=0.25
STAGE 1: 2 YRS STAGE 2: 8 YRS
$200,000/YR
52
Characteristics of a decision tree
• A Decision Tree is a chronological
representation of the decision
process.
• The tree is composed of nodes
and branches.
A branch emanating from a
Chance decision node corresponds to a
node decision alternative. It includes
Decision P(S2)
a cost or benefit value.
node
A branch emanating from a state of
P(S2) nature (chance) node corresponds to a
particular state of nature, and includes
the probability of this state of nature.
53
 DECISION: FORWARD, CALCULATION:REVERSE
HIGH DEMAND
(EVDA)2=8,250,000 $1,000,000/YR
P=0.75
2
LOW DEMAND
$300,000/YR
P=0.25

HIGH DEMAND
$900,000/YR
(EVDA)5=5,800,000 P=0.75
3,250,000
1 5
1,300,000
1,600,000 LOW DEMAND
$200,000/YR
4 P=0.25
1,900,000
HIGH DEMAND
$250,000/YR
P=0.75
3 6
2,300,000 (EVDA)6 =1,900,000LOW DEMAND
$200,000/YR
P=0.25
STAGE 1: 2 YRS STAGE 2: 8 YRS
$200,000/YR
54
 OIL WELL DRILLING EXAMPLE: DEC .TREE (McCRAY)
• AN OPERATOR (PATRA) IS CONSIDERING DRILLING WELL A ON HIS LEASE
• COST OF A DRY HOLE: $90,000; COST OF A COMPLETED WELL: $150,000
• ANOTHER OPERATOR (IN ADJACENT LEASE) WILL DRILL WELL B, IF PATRA
CONTRIBUTES $15,000 FOR WELL B IN CASE IT IS DRY
• IF WELL B IS GOOD, PROSPECTS OF WELL A WOULD BE BRIGHTER
• QUESTION: SHOULD PATRA IMMEDIATELY DRILL WELL A, OR ENCOURAGE HIS
NEIGHBOR TO DRILL WELL B BY GUARANTEEING $15,000 DRY HOLE MONEY?
P=0.2 NPV,$
600,000
P=0.2 200,000
P=0.6 0

1
P=0.4 600,000
WELL B
GOOD P=0.4 200,000
DRILL WELL
A P=0.2 0
P=0.6 -15,000
55
 OIL WELL DRILLING EXAMPLE: DEC .TREE (McCRAY)
• AN OPERATOR (PATRA) IS CONSIDERING DRILLING WELL A ON HIS LEASE
• COST OF A DRY HOLE: $90,000; COST OF A COMPLETED WELL: $150,000
• ANOTHER OPERATOR (IN ADJACENT LEASE) WILL DRILL WELL B, IF PATRA
CONTRIBUTES $15,000 FOR WELL B IN CASE IT IS DRY
• IF WELL B IS GOOD, PROSPECTS OF WELL A WOULD BE BRIGHTER
• QUESTION: SHOULD PATRA IMMEDIATELY DRILL WELL A, OR ENCOURAGE HIS
NEIGHBOR TO DRILL WELL B BY GUARANTEEING $15,000 DRY HOLE MONEY?
P=0.2 NPV,$
46,000 600,000
1 P=0.2 200,000
P=0.6 0

P=0.4 600,000
63,800 WELL B 182,000
GOOD P=0.4 200,000
2 DRILL WELL A 3
P=0.2 0
P=0.6 -15,000
56
 ALLOCATION
• DETERMINE MINIMUM COST PLAN FOR
ALLOCATING SOME PEOPLE OR MATERIALS FROM A
NUMBER OF SOURCES TO A NUMBER OF
DESTINATIONS
• BASIC MODELS:
1. ASSIGNMENT
2. TRANSPORTATION
3. TRANSSHIPMENT

57
 ASSIGNMENT MODEL
• EXAMPLE: ASSIGNING N JOBS (OR WORKERS) TO N MACHINES
OBJECTIVE: TO ASSIGN THE JOBS TO THE MACHINES (ONE JOB PER
MACHINE) AT THE LEAST TOTAL COST
• IT IS A SPECIAL CASE OF ALLOCATION
• THE MODEL: N N
MINIMIZE Z   C
I 1 J 1
IJ X IJ

SUBJECT TO: N

X
J 1
IJ  1, I  1,2,........, N
N

X
I 1
IJ  1, J  1,2,........, N

X IJ  0 OR 1
WHERE: XIJ = 0, IF THE ITH JOB IS NOT ASSIGNED TO JTH MACHINE
XIJ = 1, IF THE ITH JOB IS ASSIGNED TO JTH MACHINE
58
 EXAMPLE OF ASSIGNMENT PROBLEM
MACHINE

1 2 3 ∑=
5 7 9
1 1 1
JOB

14 10 12
2 1 1
15 13 16
3 1 1
∑= 1 1 1

INITIAL SOLUTION: NORTHWEST CORNER RULE

59
 SOLUTION
1. TRIAL & ERROR
2. SIMPLEX: LINEAR PROGRAMMING PROBLEM
3. HUNGARIAN METHOD: EFFICIENT
5 7 9
14 10 12
15 13 16

0 2 4 CREATE MATRIX WITH ZERO ENTRIES, BY:

SUBTRACTING THE SMALLEST ELEMENT IN
4 0 2 EACH ROW (COLUMN) FROM THE CORRES-
2 0 3 PONDING ROW (COLUMN)

0 2 2 OPTIMAL ASSIGNMENT:1-1, 2-3, 3-2

4 0 0 MINIMUM COST = 5 + 12+ 13 = 30
2 0 1
60
 ANOTHER ASSIGNMENT EXAMPLE
1. CREATE MATRIX WITH ZERO ENTRIES
1 2 3 4
1 1 4 6 3 1 2 3 4
1 0 3 5 2
2 9 7 10 9
2 2 0 3 2
3 4 5 11 7 3 0 1 7 3
4 8 7 8 5 4 3 2 3 0

2. DRAW A MINIMUM NUMBER OF LINES

1 2 3 4 THROUGH SOME OF THE ROWS AND
1 COLUMNS SUCH THAT ALL THE ZEROS
0 3 2 2 ARE CROSSED OUT
2 2 0 0 2 3. IF THE NO OF LINES = NO OF ROW/COLUMN,
THE SOLUTIONIS OPTIMAL. OTHERWISE, GO
3 0 1 4 3 TO STEP 4
4. SELECT THE SMALLEST UNCROSSED
4 3 2 0 0 ELEMENT (=1 IN THIS TABLE).

61
 1 2 3 4 5. THIS ELEMENT IS SUBTRACTED FROM EVERY
UNCROSSED ELEMENT AND ADDED TO
1 0 2 1 1 EVERY ELEMENT AT THE INTERSECTION
OF TWO LINES. GO BACK TO STEP 2.
2 6. SINCE THE NO OF LINES = 4 (=NO OF ROWS/
3 0 0 2 COLUMNS), THE SOLUTION IS OPTIMAL.
3 0 0 3 2 7. THE OPTIMAL ASSIGNMENT (SOME-
TIMES MORE THAN 1 SOLUTION):
4 4 2 0 0 1-1, 2-3, 3-2, 4-4.
MINIMUM COST: 1+10+5+5 = 21

62
 TRANSPORTATION MODEL
LP MODEL REPRESENTATION:
M N
MINIMIZE Z   C IJ X IJ
I 1 J 1
SUBJECT TO
N

 X IJ  AI , I  1,2,3............M
J 1
SUM OF SHIPMENTS FROM A
SOURCE CAN’T EXCEED ITS SUPPLY

 X IJ  BJ , J  1,2,3.........N
SUM OF SHIPMENTS TO A DESTINA-
TION MUST SATISFY ITS DEMAND
I 1

X IJ  0, FOR  I AND J
63
 BALANCED TRANSPORTATION MODEL
• THE PREVIOUS MODEL IMPLIES THAT
TOTAL SUPPLY ≥ TOTAL DEMAND

 
M N
I 1
AI  B
J 1 J
• WHEN THE TOTAL SUPPLY EQUALS THE TOTAL DEMAND, IT IS A
BALANCED TRANSPORTATION MODEL. IF NOT, IT CAN BE BALANCED BY
ADDING DUMMIES, SO ALL CONSTRAINTS ARE EQUATIONS
N

X
J 1
IJ  AI , I  1,2,...., M

X
I 1
IJ  BJ , J  1,2,....., N

64
 BASIC STEPS OF TRANSPORTATION
TECHNIQUES
THE BASIC STEPS OF THE TRANSPORTATION TECHNIQUES ARE:
• STEP 1: DETERMINE A STARTING FEASIBLE SOLUTION
• STEP 2: DETERMINE AN ENTERING VRB FROM AMONG THE
NONBASIC VRBS. IF ALL SUCH VRBS SATISFY THE
OPTIMALITY CONDITION (OF THE SIMPLEX METHOD), STOP;
OTHERWISE, GO TO STEP 3
• STEP 3: DETERMINE A LEAVING VRB (USING THE FEASIBILITY
CONDITION) FROM AMONG THE VRBS OF THE CURRENT
BASIC SOLUTION; THEN FIND THE NEW BASIC SOLUTION.
RETURN TO STEP 2
• THE SOLUTION MUST INCLUDE (M+N-1) BASIC VRBS

65
 EXAMPLE OF TRANSPORTATION TECHNIQUE

DESTINATION

1 2 3 4 SUPPLY
10 0 20 11
1 X11 X 12 X13 X14 15
SOURCE

12 7 9 20
2 X21 X22 X23 X24 25
0 14 16 18
3 X31 X32 X33 X34 5

DEMAND 5 15 15 10

THE SOLUTION MUST INCLUDE (M+N-1) = 3 + 4 - 1 = 6 BASIC VRBS

THE UNIT TTRANSPORTATION COST CIJ IS IN DOLLARS
66
 STARTING SOLUTION (1)
• STARTING BASIC FEASIBLE SOLUTION PROCEDURES:
– NORTHWEST-CORNER RULE
– LEAST-COST METHOD
– VOGEL’S APPROXIMATION METHOD
• NORTHWEST-CORNER METHOD:
1 2 3 4
10 0 20 11
1 5 10 15
12 20
2 5 7 15 9 5 25

3 0 14 16 5 18 5

5 15 15 10

THE NO OF ALLOCATION: 6 (= 3 + 4 - 1)
THE ALLOCATION COST=5(10)+10(0)+5(7)+15(9)+5(20)+5(18)=$410
67
 STARTING SOLUTION (2)
• THE LEAST-COST METHOD: ASSIGN AS MUCH AS POSSIBLE TO THE VRB
WITH THE SMALLEST UNIT COST IN THE ENTIRE TABLEAU. TIES ARE
BROKEN ARBITRARILY

1 2 3 4
10 0 20 11
1 0 15 0 15

12 7 9 20
2 15 10 25

0 14 16 18
3 5 5

5 15 15 10

THE NO OF ALLOCATION MUST BE 6

THE ALLOCATION COST=0(10)+15(0)+0(11)+15(9)+10(20)+5(0)=$335

68
 STARTING SOLUTION (3)
• VOGEL’S APPROXIMATION METHOD:HEURISTIC, GENERALLY YIELD AN
OPTIMUM OR CLOSE TO OPTIMUM SOLUTION
• STEP 1: EVALUATE A PENALTY FOR EACH ROW (COLUMN) BY
SUBTRACTING THE SMALLEST COST ELEMENT IN THE ROW (COLUMN)
FROM THE NEXT SMALLEST COST ELEMENT IN THE SAME ROW
(COLUMN)
• STEP 2: IDENTIFY THE ROW OR COLUMN WITH THE LARGEST PENALTY,
BREAKING TIES ARBITRARILY. ALLOCATE AS MUCH AS POSSIBLE TO THE
VRB WITH THE LEAST COST IN THE SELECTED ROW OR COLUMN. ADJUST
THE SUPPLY AND DEMAND AND CROSS OUT THE SATISFIED ROW OR
COLUMN. IF A ROW AND A COLUMN ARE SATISFIED SIMULTANEOUSLY,
ONLY ONE OF THEM IS CROSSED OUT AND THE REMAINING ROW
(COLUMN) IS ASSIGNED A ZERO SUPPLY (DEMAND). ANY ROW OR
COLUMN WITH ZERO SUPPLY OR DEMAND SHOULD NOT BE USED IN
COMPUTING FUTURE PENALTIES (IN STEP 3)

69
 STARTING SOLUTION (4): VAM
• STEP 3:
a) IF EXACTLY ONE ROW OR ONE COLUMN REMAINS UNCROSSED OUT,
STOP
b) IF ONLY ONE (COLUMN) WITH POSITIVE SUPPLY (DEMAND) REMAINS
UNCROSSED OUT, DETERMINE THE BASIC VRBS IN THE ROW (COLUMN)
BY THE LEAST–COST METHOD
c) IF ALL UNCROSSED-OUT ROWS AND COLUMNS HAVE (ASSIGNED) ZERO
SUPPLY AND DEMAND, DETERMINE THE ZERO BASIC VRBS BY THE
LEAST-COST METHOD. STOP
d) OTHERWISE, RECOMPUTE THE PENALTIES FOR THE UNCROSSED–OUT
ROWS AND COLUMNS, THEN GO TO STEP 2.(NOTICE THAT THE ROWS
AND COLUMNS WITH ASSIGNED ZERO SUPPLY AND DEMAND SHOULD
NOT BE USED IN COMPUTING THESE PENALTIES)

70
 EXAMPLE

1 2 3 4 ROW PENALTY
10 0 20 11
1 5 10 15 10 11 11

12 7 9 20
2 10 15 25 10 2 2 13

0 14 16 18
3 5 0 5 0 14 - -

5 15 15 10

10 7 7 7
COLUMN
- 7 11 9
PENALTY
- 7 - 9

TOTAL ALLOCATION COST = 5(0)+10(11)+10(7)+15(9)+5(0)+0(18)=$315

(WHICH HAPPENS TO BE OPTIMAL)

71
 DETERMINATION OF ENTERING VRB (1)
• METHOD OF MULTIPLIERS
• MULTIPLIER: FOR ROW I:UI,FOR COLUMN J:VJ
• FOR EACH BASIC VRB XIJ IN THE CURRENT SOLUTION, THE
MULTIPLIERS UI & VJ MUST SATISFY THE FOLLOWING
EQUATION: UI+VJ=CIJ FOR EACH BASIC VRB XIJ
• THESE EQS YIELD (M+N-1) EQS (BECAUSE THERE ARE ONLY
M+N-1 BASIC VRBS) IN (M+N ) UNKNOWNS
• THE VALUES OF THE MULTIPLIERS CAN BE DETERMINED
FROM THESE EQS BY ASSUMING AN ARBITRARY VALUE FOR
ANY ONE OF THE MULTIPLIERS (USUALLY UI IS SET EQUAL TO
ZERO) AND THEN SOLVING THE M+N-1 EQS IN THE
REMAINING M+N-1 UNKNOWN MULTIPLIERS.

72
 DETERMINATION OF ENTERING VRB (2)
• APPLYING THIS PROCEDURE TO THE NONBASIC VRBS IN THE
CURRENT SOLUTION ALLOCATION TABLE (NW CORNER) , THE
EQS ASSOCIATED WITH THE BASIC VRBS ARE:
X11: U1+V1 = C11 = 10
X12: U1+V2 = C12 = 0
BY LETTING U1=0, THE VALUES OF
X22: U2+V2 = C22 = 7 THE MULTIPLIERS ARE SUCCESSIVELY
X23: U2+V3 = C23 = 9 DETERMINED AS V1=10, V2=0,U2=7,
X24: U2+V4 = C24 = 20 V3=2, V4=13,AND U3=5

X34: U3+V4 = C34 = 18

73
 DETERMINATION OF ENTERING VRB (3)
• THEN, THE EVALUATION OF EACH NONBASIC VRB XPQ IS
GIVEN BY:
CPQ = UP + VQ – CPQ, FOR EACH NONBASIC VRB XPQ
(THESE VALUES WILL BE THE SAME REGARDLESS OF THE ARBITRARY
CHOICE OF THE VALUE OF U1). THE ENTERING VRB IS THEN SELECTED AS
THE NONBASIC VRB WITH THE MOST POSITIVE CPQ(COMPARE WITH THE
MINIMIZATION OPTIMALITY CONDITION OF THE SIMPLEX METHOD)

• THE EVALUATIONS OF THE NONBASIC VRBS ARE THUS:

X13: C13 = U1+V3-C13 = 0+2-20 = -18
X14: C14 = U1+V4-C14 = 0+13-11= 2
X31 HAS THE MOST
X21: C21 = U2+V1-C21 = 7+10-12= 5 POSITIVE CPQ; SO IT IS
X31: C31 = U3+V1-C31 = 5+10-0 = 15 SELECTED AS THE ENTER-
X32: C32 = U3+V2-C32 = 5+0-14 = -9 ING VRB
X33: C33 = U3+V3-C33 = 5+2-16 = -9

74
 DETERMINATION OF LEAVING VARIABLE
(LOOP CONSTRUCTION) (1)
• THIS STEP IS EQUIVALENT TO APPLYING THE FEASIBILITY CONDITION IN
THE SIMPLEX METHOD
• SINCE ALL THE CONSTRAINT COEFFICIENTS IN THE ORIGINAL
TRANSPORTATION MODEL ARE EITHER ZERO OR 1, THE (POSITIVE)
RATIOS OF THE FEASIBILITY CONDITION WILL ALWAYS HAVE THEIR
DENOMINATOR EQUAL TO 1. THUS THE VALUES OF THE BASIC VRBS WILL
GIVE THE ASSOCIATED RATIOS DIRECTLY
• TO DETERMINE THE MINIMUM RATIO, WE CONSTRUCT A CLOSED LOOP
(EITHER CLOCKWISE OR COUNTERCLOCKWISE) FOR THE CURRENT
ENTERING VRB (X31 IN THE CURRENT ITERATION). THE LOOP STARTS AND
ENDS AT THE DESIGNATED NONBASIC VRB. IT CONSISTS OF SUCCESSIVE
HORIZONTAL AND VERTICAL (CONNECTED) SEGMENTS WHOSE END
POINTS MUST BE BASIC VRBS, EXCEPT FOR THE END POINTS THAT ARE
ASSOCIATED WITH THE ENTERING VRB. THIS MEANS THAT EVERY
CORNER ELEMENT OF THE LOOP MUST BE A CELL CONTAINING A BASIC
VRB (SEE NEXT TABLE).

75
 EXAMPLE OF LOOP FOR ENTERING VRB X31

1 2 3 4

1 5 10 15
- +

2 5 15 5 25
- +

3 X31 5 5
+ -

5 15 15 10

76
 DETERMINATION OF LEAVING VARIABLE
(LOOP CONSTRUCTION) (2)
• THE LEAVING VRB IS SELECTED FROM AMONGTHE CORNER
VRBS OF THE LOOP THAT WILL DECREASE WHEN THE
ENTERING VRB X31 INCREASES ABOVE ZERO LEVEL. THESE
ARE INDICATED BY THE VRBS LABELED BY MINUS SIGNS (-).
THE LEAVING VRB IS THEN SELECTED AS THE ONE HAVING
THE SMALLEST VALUE, SINCE IT WILL BE THE FIRST TO REACH
ZERO VALUE AND ANY FURTHER DECREASE WILL CAUSE IT
TO BE NEGATIVE (COMPARE THE FEASIBILITY CONDITION OF
THE SIMPLEX METHOD WHERE THE LEAVING VRB IS
ASSOCIATED WITH THE MINIMUM RATIO)
• X34 IS TAKEN AS THE LEAVING VRB

77
 EXAMPLE OF LOOPING IN THE 2ND ALLOCATION

1 V1=10 2 V2=0 3 V3=2 4 V4=13

10 0 20 11
1 U1=0 0 15 15
-18 +2
12 7 9 20
2 U2=7 0 15 10 25
+5
0 14 16 18
3 U3= -10 5 5
-24 -24 -15

5 15 15 10

78
 EXAMPLE OF LOOPING IN THE 2ND ALLOCATION) (2)
• CHECK FOR OPTIMALITY OF THIS BASIC SOLUTION BY
COMPUTING THE NEW MULTIPLIERS. THE NONBASIC VRB
X21(LARGEST POSITIVE CPQ THUS ENTERS THE SOLUTION)

79
 EXAMPLE OF LOOPING IN THE 2ND ALLOCATION) (3)

1 V1=10 2 V2=0 3 V3=2 4 V4=13

10 0 20 11
1 U1=0 0 15 15
- + -18 +2
12 7 9 20
2 U2=7 X21 0 15 10 25
+5 -
+
0 14 16 18
3 U3= -10 5 5
-24 -24 -15

5 15 15 10

80
EXAMPLE OF SELECTING THE LEAVING VRB (NEW ALLOCATION)
(4)
• THE CLOSED LOOP ASSOCIATED WITH X21 SHOWS THAT
EITHER X11 OR X22 CAN BE THE LEAVING VRB. WE
ARBITRARILY SELECT X11 TO LEAVE THE SOLUTION

81
 EXAMPLE OF LOOPING IN THE 3RD ALLOCATION

1 V1= 5 2 V2=0 3 V3=2 4 V4=13

10 0 20 11
1 U1=0 15 X14 15
-5 - -18 +2 +
12 7 9 20
2 U2=7 0 0 15 10 25
+ -
0 14 16 18
3 U3= -5 5 5
-19 -19 -10

5 15 15 10

ENTERING VRB: X14; LEAVING VRB: X24

82
 EXAMPLE OF OPTIMALITY CHECK IN THE 4TH ALLOCATION

1 V1= 5 2 V2=0 3 V3=2 4 V4=11

10 0 20 11
1 U1=0 5 10 15
-5 -18
12 7 9 20
2 U2=7 0 10 15 25
-2
0 14 16 18
3 U3= -5 5 5
-19 -19 -12

5 15 15 10

SINCE ALL CPQ ARE NONPOSITIVE, SOLUTION IS OPTIMAL

(COMPARE WITH THE MINIMIZATION OPTIMALITY CONDITION
OF THE SIMPLEX METHOD).TOTAL COST=$315 83
 PROJECT MANAGEMENT:PLANNING, SCHEDULING,
CONTROLLING (1)
• DEFINITION OF PROJECT: A COMBINATION OF INTERRELA-
TED ACTIVITIES THAT MUST BE EXECUTED IN A CERTAIN
ORDER BEFORE THE ENTIRE TASK CAN BE COMPLETED.
• THE ACTIVITIES ARE INTERRELATED IN A LOGICAL
SEQUENCE
• SOME ACTIVITIES CANNOT START UNTIL OTHERS ARE
COMPLETED
• AN ACTIVITY IN A PROJECT : A JOB REQUIRING TIME &
RESOURCES FOR ITS COMPLETION
• TWO PROJECT MANAGEMENT TECHNIQUES:
– CRITICAL PATH METHOD (CPM): DETERMINISTIC
– PROJECT EVALUATION AND REVIEW TECHNIQUE (PERT):
PROBABILISTIC

84
 PROJECT MANAGEMENT:PLANNING, SCHEDULING,
CONTROLLING (2)
• THE PLANNING PHASE:
– BREAKING DOWN PROJECT INTO DISTINCT ACTIVITIES
– DETERMINE TIME ESTIMATES FOR THESE ACTIVITIES
– CONSTRUCT A NETWORK (OR ARROW) DIAGRAM : EACH ARROW
REPRESENTS AN ACTIVITY
– THE ENTIRE ARROW DIAGRAM GIVES A GRAPHIC REPRESENTATION
OF THE INTERDEPENDENCIES BETWEEN THE ACTIVITIES OF THE
PROJECT
• THE SCHEDULING PHASE:
– CONSTRUCT A TIME CHART SHOWING THE START AND FINISH TIMES
FOR EACH ACTIVITY AND ITS RELATIONSHIP TO OTHER ACTIVITIES IN
THE PROJECT
– MUST PINPOINT THE CRITICAL ACTIVITIES (IN VIEW OF TIME) THAT
REQUIRE SPECIAL ATTENTION IF THE PROJECT IS TO BE COMPLETED
ON TIME. ALSO FOR CRASH/ACCELERATION/ COST-TIME TRADE OFF
– FOR NON CRITICAL ACTIVITIES, SLACK OR FLOAT TIMES MUST BE
SHOWN, TO BE USED ADVANTAGEOUSLY (USE LIMITED RESOURCES)

85
 PROJECT MANAGEMENT:PLANNING,
SCHEDULING, CONTROLLING (3)
• THE CONTROLLING PHASE:
– USE OF ARROW DIAGRAM AND TIME CHART FOR MAKING PERIODIC
PROGRESS REPORTS
– THE NETWORK MAY THUS BE UPDATED, ANALYZED AND REVISED
FOR THE REMAINING PORTION OF THE PROJECT IF NECESSARY

86
 ARROW (NETWORK) DIAGRAM
• ARROW: REPRESENTS AN ACTIVITY; ITS (ARC) HEAD
INDICATES THE DIRECTION OF PROGRESS IN THE PROJECT
• EVENT: REPRESENTS A POINT IN TIME THAT SIGNIFIES THE
COMPLETION OF SOME ACTIVITIES AND THE BEGINNING OF
NEW ONES. IT IS REPRESENTED BY A NODE
• THE BEGINNING AND END POINTS OF AN ACTIVITY ARE THUS
DESCRIBED BY TWO EVENTS USUALLY KNOWN AS THE HEAD
AND TAIL EVENTS
• THE ARROW DIAGRAM REPRESENTS THE INTERDEPENDEN-
CIES AND PRECEDENCE RELATIONSHIPS AMONG THE
ACTIVITIES OF THE PROJECT
• ACTIVITIES ORIGINATING FROM A CERTAIN EVENT CANNOT
START UNTIL THE ACTIVITIES TERMINATING AT THE SAME
EVENT HAVE BEEN COMPLETED

87
 RULES FOR CONSTRUCTING ARROW DIAGRAM (1)
• RULE 1: EACH ACTIVITY IS REPRESENTED BY ONE AND ONLY
ONE ARROW IN THE NETWORK. NO SINGLE ACTIVITY CAN BE
REPRESENTED TWICE IN THE NETWORK. HOWEVER, ONE ACTIVITY CAN
BE BROKEN DOWN INTO SEGMENTS, IN WHICH CASE EACH SEGMENT
MAY BE REPRESENTED BY A SEPARATE ARROW, EX: LAYING DOWN A PIPE
MAY BE DONE IN SECTIONS RATHER THAN AS ONE JOB
• RULE 2: NO TWO ACTIVITIES CAN BE IDENTIFIED BY THE
SAME HEAD AND TAIL EVENTS. USE DUMMY (D) ACTIVITY,
IF NECESSARY
A
2 2
A D D A

1 3 1 3
B B B
A C JOBS A&B MUST PRECEDE C
A C
D WHILE JOB E IS PRECEDED BY
B E B E JOB B ONLY

88
 RULES FOR CONSTRUCTING ARROW DIAGRAM (2)
• RULE 3: TO ENSURE THE CORRECT PRECEDENCE RELATION-
SHIP IN THE ARROW DIAGRAM, THE FOLLOWING QUES-
TIONS MUST BE ANSWERED AS EVERY ACTIVITY IS ADDED
TO THE NETWORK.
– WHAT ACTIVITIES MUST BE COMPLETED IMMEDIATELY BEFORE THIS
ACTIVITY CAN START?
– WHAT ACTIVITIES MUST FOLLOW THIS ACTIVITY?
– WHAT ACTIVITIES MUST OCCUR CONCURRENTLY WITH THIS
ACTIVITY?

89
 EXAMPLE
CONSTRUCT THE ARROW DIAGRAM COMPRISING ACTIVITIES
A,B,C,…..,L SUCH THAT THE FOLLOWING RELATIONSHIPS ARE
SATISFIED.
1. A,B AND C, THE FIRST ACTIVITIES OF THE PROJECT, CAN START
SIMULTANEOUSLY
2. A & B PRECEDE D
3. B PRECEDES E,F & H
4. F & C PRECEDE G
5. E & H PRECEDE I & J
6. C,D,F & J PRECEDE K
7. K PRECEDES L
8. I,G & L ARE THE TERMINAL ACTIVITIES OF THE PROJECT

90
 THE ARROW / NETWORK DIAGRAM OF THE EXAMPLE ABOVE

3 D 7
K
8
A D1 4 J L
D3
H
1 B E I
2 5 9
D2
C F G

91
 CRITICAL PATH CALCULATIONS (1)

• CPM-PERT: GIVES A SCHEDULE SPECIFYING THE START AND

COMPLETION DATES OF EACH ACTIVITY
• THE CALCULATIONS OF DATES ARE PERFORMED DIRECTLY ON
THE ARROW DIAGRAM BY SIMPLE ARITHMETIC
• CRITICAL ACTIVITY: IF THE DELAY IN THE START OF AN
ACTIVITY WILL CAUSE A DELAY IN THE COMPLETION DATE OF
THE ENTIRE PROJECT
• NON CRITICAL ACTIVITY: IF THE TIME BETWEEN ITS EARLIEST
START AND ITS LATEST COMPLETION DATE (AS ALLOWED BY
THE PROJECT) IS LONGER THAN ITS ACTUAL DURATION (IT
HAS SLACK OR FLOAT TIME)
• BENEFITS: 1. PROJECT MANAGEMENT
2. MINIMIZE TOTAL PROJECT COST
3. RESOURCE LEVELING
92
 CRITICAL PATH CALCULATIONS (2)
• TWO PHASE CALCULATIONS:
1. FORWARD PASS: CALCULATIONS BEGIN FROM THE
START NODE AND MOVE TO THE END NODE. AT EACH
NODE A NUMBER IS COMPUTED REPRESENTING THE
EARLIEST OCCURRENCE TIME OF THE
CORRESPONDING EVENT. THESE NUMBERS ARE
SHOWN IN THE FIGURE BELOW IN SQUARES
2. BACKWARD PASS: BEGIN CALCULATIONS FROM THE
END NODE AND MOVES TO THE START NODE. THE
NUMBER COMPUTED AT EACH NODE REPRESENTS
THE LATEST OCCURRENCE TIME OF THE
CORRESPONDING EVENT, SHOWN IN TRIANGLES
• ESJ = MAXI { ESI + DIJ}, FOR ALL (I,J) ACTIVITIES DEFINED
• LCI = MINJ { LCJ - DIJ}, FOR ALL (I,J) ACTIVITIES DEFINED
93
 CRITICAL PATH CALCULATIONS (3)
• WHERE: ES = EARLIEST START
LC = LATEST COMPLETION
D = DURATION
I = START OF AN ACTIVITY
J = END OF AN ACTIVITY
• AN ACTIVITY (I.J) LIES ON THE CRITICAL PATH IF IT SATISFIES THE
FOLLOWING THREE CONDITIONS:
THESE CONDITIONS ACTUALLY INDI-
1. ESI = LCI
CATE THAT THERE IS NO FLOAT OR
2. ESJ = LCJ SLACK TIME BETWEEN EARLIEST
3. ESJ – ESI = LCJ – LCI = DIJ START (COMPLETION) AND THE
LATEST START (COMPLETION) OF
THE CRITICAL ACTIVITY (ZERO FLOAT)

• IN THE DIAGRAM, THE NUMBERS IN AND ARE THE SAME AT

EACH OF THE HEAD AND THE TAIL EVENTS. THE DIFFERENCE BETWEEN
THE NUMBER IN ( OR ) AT THE HEAD EVENT AND THE
NUMBER IN (OR ) AT THE TAIL EVENT IS EQUAL TO THE
DURATION OF THE ACTIVITY

94
 DETERMINATION OF THE FLOATS
• TOTAL FLOAT FOR ACTIVITY (I,J): TFIJ IS THE DIFFERENCE
BETWEEN THE MAXIMUM TIME AVAILABLE TO PERFORM
THE ACTIVITY (=LC J- ESI) AND ITS DURATION (=DIJ):
TFIJ = LCJ – ESI – DIJ = LCJ – ECIJ = LSIJ – ESI
• FREE FLOAT (FF) IS DEFINED BY ASSUMING THAT ALL THE
ACTIVITIES START AS EARLY AS POSSIBLE. FFIJ FOR ACTIVITY
(I,J) IS THE EXCESS OF AVAILABLE TIME (= ESJ - ESI) OVER ITS
DURATION (=DIJ):
FFIJ = ESJ – ESI -DIJ

95
 6
6 CRITICAL ACTIVITY

5
0 3 2 7 19
0 3 19
0 3 6 6
2 5
START 13 END
13
2 3 3
2

ES
1 2 3 1
4 6 LC
2
6

96
 EARLIEST LATEST
START COMPLE- START COMPLE-
TION TION
ACTIVITY DURATI- TOTAL FREE
ON FLOAT FLOAT
(I,J) DIJ ESI ECIJ LSIJ LCJ TFIJ FFIJ
(1) (2) (3) (4) (5) (6) (7) (8)
(0,1) 2 0 2 2 4 2 0
(0,2) 3 0 3 0 3 0 0
(1,3) 2 2 4 4 6 2 2
(2,3) 3 3 6 3 6 0 0
(2,4) 2 3 5 4 6 1 1
(3,4) 0 6 6 6 6 0 0
(3,5) 3 6 9 10 13 4 4
(3,6) 2 6 8 17 19 11 11
(4,5) 7 6 13 6 13 0 0
(4,6) 5 6 11 14 19 8 8
97
(5,6) 6 13 19 13 19 0 0
 CALCULATION ON TABLE
D U R A T I O N (DIJ)
J
ES I 0 1 2 3 4 5 6
0
1
2
3
4
5
6
LC
98
 CALCULATION OF ES & LC ON TABLE

D U R A T I O N (DIJ)
FORWARD

J
ES I 0 1 2 3 4 5 6
0 0 2 3
2 1 2
3 2 3 2
4,6 3 0 3 2
6,5 4 7 5
9,13 5 6
8,11, 6
19
LC 0,2 4 4,3 17,10, 14,6 13 19 99
BACKWARD 6
 CONSTRUCTION OF THE TIME CHART (GANNT CHART)
DAY 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
I-J
0 –1 TF=2

0–2
1–3 FF=2
2–3
2–4 FF=1
3–4
TF=4
3–5
3–6
4–5
4–6
5–6

CRITICAL ACTIVITIES

100
 RESOURCE LEVELING
DAY 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
I-J
0 –1 0 TF=2
0–2 5
1–3 0
FF=2
2–3 7
2–4 3 FF=1
3–4
3–5 TF=4
2
3–6 1
4–5 2
4–6 5
5–6 6
10 10
10
NO OF WORKERS

9 WHAT SHOULD WE DO TO GET

LEVELED RESOURCES (YELLOW)?
7 7 CHECK THE NO OF MAN-DAY
6
5
5
2
101
0
 NO OF MAN-DAY
NEW MAN POWER
OLD MAN POWER
SCHEDULING
SCHEDULING
(AFTER LEVELING)
NO OF
DAYS
WORKERS
5 3 15 5 3 15
10 2 20 10 2 20
7 1 7 7 5 35
10 2 20 9 1 9
9 1 9 5 2 10
7 2 14 6 6 36
2 2 4
6 6 36
TOTAL 125 125

102
 COST-TIME TRADE OFF (PROJECT ACCELERATION)

• OBJECTIVE: FIND THE MINIMUM TOTAL PROJECT COSTS BY

OPTIMIZING PROJECT DURATION
• COSTS INCLUDE:
– DIRECT COSTS
– INDIRECT COSTS TOTAL COST
COST

MINIMUM TOTAL COST INDIRECT COST

DIRECT COST
OPTIMUM DURATION

CRASHED SCHEDULE NORMAL SCHEDULE DURATION

103
 EXAMPLE ON COST-TIME TRADE OFF (1)

8 10
2 18 COST=580
8 5 TIME =18
8 18
1 0 2 3
0
4 10
4 5 4
3 15
10

ACTIVITY NORMAL CRASH CRASH COST

CC - CN
(I,J) DURATION COST LIMIT DURATION COST SLOPE = D -D
C N
1-2 (CR) 8 100 2 6 200 50
1-3 4 150 2 2 350 100
2-4 2 50 1 1 90 40
2-5 (CR) 10 100 5 5 400 60
3-4 5 100 4 1 200 25
4-5 3 80 2 1 100 10
104
TOTAL DC = 580
 EXAMPLE ON COST-TIME TRADE OFF (2)

COST=630
TIME=17
7
2 10
5 17
7 7
8 17
0
1 FF=0 2
0 3
FF=5
4
FF=0 9
5 4
3 4
14
FF=0
9

105
 EXAMPLE ON COST-TIME TRADE OFF (3)

* : CRASH LIMIT COST=680

TIME=16
6
6* 2 10
16
7 6 5
8 16
0
1 FF=1 2
0 3
FF=4
4
FF=0 9
5 4
3 4
13
FF=0
8

106
 EXAMPLE ON COST-TIME TRADE OFF (4)

COST=920
6 TIME=12
6
6* 2 10 12
7 6 5
8 12
0
1 FF=1 2
0 3

4
9
5 4
3 4
9
4

107
 EXAMPLE ON COST-TIME TRADE OFF (5)

5* COST=990
6 TIME=11
6
6* 2 10 11
7 6 5
8 11
0
1 FF=1 2 2
0 3
4
9
5 4
3 4
9
4

108
 SUMMARY OF THE EXAMPLE ON COST-TIME TRADE OFF

CRASHED CRASH FF COMPRESSION PROJECT DIRECT COST,

ACTIVITY(IES) LIMIT, LIMIT, LIMIT, DAYS DURATION, $
DAYS DAYS MIN{CL,FFL} DAYS

NONE 18 580

1-2 (CR) 2 1 1 17 630(=580+50)

1-2 (CR) 1 1 1 16 680

2-5 (CR) 5 4 4 12 920

2-5 & 4-5(CR) 1 1 1 11 990

109
 COST , $ ASSIGNMENT : FIND THE INDIRECT
OPTIMUM DURATION FOR COST= $60 /
THE CASE IC= $60 / DAY DAY

1000

900

800

700

600
DIRECT COST

500

11 12 13 14 15 16 17 18 DAYS 110
 Network Models

Chapter 8 Operation Research An

Introduction (Hamdy A. Taha)
 Introduction (1)
• Previously, we learned about transportation (or distribution) problems
that deal with shipping commodities between sources and destinations at
minimum transportation costs.
• Consider the following situations:
– Design of offshore natural-gas pipeline network connecting wellheads in the
Gulf of Mexico with an onshore delivery point with the objective of minimizing
the cost of constructing the pipeline.
– Determination of the shortest route joining two cities in an existing network of
roads.
– Determination of the maximum annual capacity in tons of a coal slurry
pipeline network joining the coal mines in Wyoming with the power plants in
Houston.
– Determination of the minimum-cost flow schedule from oil fields to refineries
and finally to distribution centers. Crude oil and gasoline products can be
shipped via tankers, pipelines, and/or trucks. In addition to maximum supply
availability at the oil fields and minimum demands requirements at the
distribution centers, restrictions on the capacity of the refineries and the
modes of transportation must be taken into account
 Introduction (2)
• A Study of this representative list reveals that
network optimization problems can generally
be modeled by one of four models:
– Minimal spanning tree model
– Shortest-route model
– Maximum-flow model
– Minimum-cost capacitated network model
 Terminology of Graphs
• A graph (or network) consists of
– a set of points
– a set of lines connecting certain pairs of
the points.
The points are called nodes (or
vertices).
The lines are called arcs (or edges or
links).
• Example:
 Terminology of Graphs: Paths
• A path between two nodes is a
sequence of distinct nodes and edges
connecting these nodes.
Example:

a
b
 Terminology of Graphs:
Cycles, Connectivity and Trees

• A path that begins and ends at the same node is called a cycle.
Example:

• Two nodes are connected if there is a path between them.

• A graph is connected if every pair of its nodes is connected.
• A graph is acyclic if it doesn’t have any cycle.
• A graph is called a tree if it is connected and acyclic.
Example:
 Minimum Spanning Tree Problem
• Given: Graph G=(V, E), |V|=n
Cost function c: E  R .
• Goal: Find a minimum-cost spanning tree for V
i.e., find a subset of arcs E*  E which
connects any two nodes of V
with minimum possible cost.
• Example:
G=(V,E)
3 b 3 d
b d
2 8 2 8
a 3 5 a 3 5
4 7 4 7
c e c e
4 4
Red bold arcs are in E*

Min. span. tree:

G*=(V,E*)
 Algorithm for solving
the Minimum Spanning Tree Problem

• Initialization: Select any node arbitrarily,

connect to its nearest node.

• Repeat
– Identify the unconnected node
which is closest to a connected node

– Connect these two nodes

Until all nodes are connected
Note: Ties for the closest node are broken arbitrarily.
The algorithm applied to our example

• Initialization: Select node a to start.

Its closest node is node b. Connect nodes a
and b.
b 3
d
2 8 Red bold arcs are in E*;
a 3 5
thin arcs represent potential
4 7 links.
c e
4

• Iteration 1: There are two unconnected node closest to a

connected node: nodes c and d b 3 d
(both are 3 units far from node b). 2 8
Break the tie arbitrarily by a 3 5
connecting node c to node b. 4 7
c e
4
 The algorithm applied to our example

• Iteration 2: The unconnected node closest to a

connected node is node d (3 far from node b). Connect nodes
b and d.
b 3
d
2 8
a 3 5
4 7
c e
4
• Iteration 3: The only unconnected node left is node
e. Its closest connected node is node c b 3 d
(distance between c and e is 4). 2 8
Connect node e to node c. a 3 5
• All nodes are connected. The bold 4 7
c e
arcs give a min. spanning tree. 4
 Algorithm Characteristics
• Both Prim’s and Kruskal’s Algorithms work with undirected graphs

• Both work with weighted and unweighted graphs but are more

interesting when edges are weighted

• Both are greedy algorithms that produce optimal solutions (A

greedy algorithm is an algorithmic paradigm that follows the

problem solving heuristic of making the locally optimal choice at

each stage).
 Prim’s Algorithm

• dv records edge weights, not path lengths

 Walk-Through
Initialize
2
array
3
10
F C K dv pv
A 7 3 A F  
4
8
18 B F  
4
9
B D C F  
10
H 25
D F  
2
3 E F  
G 7
E F F  
G F  
H F  
 Start with any
2
node, say D
3
10
F C K dv pv
A 7
4
3 A
8 B
18
4
9
B D C
10
H 25
D T 0 
2
3 E
G 7
E F
G
H
 Update distances
of adjacent,
2
unselected nodes
3
10
F C K dv pv
A 7
4
3 A
8 B
18
4
9
B D C 3 D
10
H 25
D T 0 
2
3 E 25 D
G 7
E F 18 D
G 2 D
H
 Select node with
minimum
distance
2

3
10
F C K dv pv
A 7
4
3 A
8 B
18
4
9
B D C 3 D
10
H 25
D T 0 
2
3 E 25 D
G 7
E F 18 D
G T 2 D
H
 Update distances
of adjacent,
2
unselected nodes
3
10
F C K dv pv
A 7
4
3 A
8 B
18
4
9
B D C 3 D
10
H 25
D T 0 
2
3 E 7 G
G 7
E F 18 D
G T 2 D
H 3 G
 Select node with
minimum
2
distance
3
10
F C K dv pv
A 7
4
3 A
8 B
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E 7 G
G 7
E F 18 D
G T 2 D
H 3 G
 Update distances
of adjacent,
2 unselected nodes
3
10
F C K dv pv
A 7
4
3 A
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E 7 G
G 7
E F 3 C
G T 2 D
H 3 G
 Select node with
minimum
2 distance
3
10
F C K dv pv
A 7
4
3 A
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E 7 G
G 7
E F T 3 C
G T 2 D
H 3 G
 Update distances
of adjacent,
2
unselected nodes
3
10
F C K dv pv
A 7
4
3 A 10 F
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E 2 F
G 7
E F T 3 C
G T 2 D
H 3 G
 Select node with
minimum
2 distance
3
10
F C K dv pv
A 7
4
3 A 10 F
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H 3 G
 Update distances
of adjacent,
2 unselected nodes
3
10
F C K dv pv
A 7
4
3 A 10 F
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H 3 G

Table entries
unchanged
 Select node with
minimum
2 distance
3
10
F C K dv pv
A 7
4
3 A 10 F
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H T 3 G
 Update distances
of adjacent,
2
unselected nodes
3
10
F C K dv pv
A 7
4
3 A 4 H
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H T 3 G
 Select node with
minimum
2
distance
3
10
F C K dv pv
A 7
4
3 A T 4 H
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H T 3 G
 Update distances
of adjacent,
2
unselected nodes
3
10
F C K dv pv
A 7
4
3 A T 4 H
8 B 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H T 3 G

Table entries
unchanged
 Select node with
2
minimum
distance
3
10
F C K dv pv
A 7
4
3 A T 4 H
8 B T 4 C
18
4
9
B D C T 3 D
10
H 25
D T 0 
2
3 E T 2 F
G 7
E F T 3 C
G T 2 D
H T 3 G
 Cost of Minimum
Spanning Tree = 
2 dv = 21
3
F C K dv pv
A 4
3 A T 4 H
B T 4 C
4
B D C T 3 D
H D T 0 
2
3 E T 2 F
G E F T 3 C
G T 2 D
H T 3 G

Done
Kruskal’s Algorithm

Work with edges, rather than nodes

Two steps:
– Sort edges by increasing edge weight
– Select the first |V| – 1 edges that do not
generate a cycle
Walk-Through
3
Consider an undirected,
10
F C weight graph
A 4
4
3
8
6
5
4
B D
4
H 1
2
3
G 3
E
 Sort the edges by increasing
edge weight
3
10
F C edge dv edge dv

A 4 3 (D,E) 1 (B,E) 4
8 4
6 (D,G) 2 (B,F) 4
5
4
B D (E,G) 3 (B,H) 4
4
H 1
(C,D) 3 (A,H) 5
2
3 (G,H) 3 (D,F) 6
G 3
E (C,F) 3 (A,B) 8
(B,C) 4 (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
8 4
6 (D,G) 2 (B,F) 4
5
4
B D (E,G) 3 (B,H) 4
4
H 1
(C,D) 3 (A,H) 5
2
3 (G,H) 3 (D,F) 6
G 3
E (C,F) 3 (A,B) 8
(B,C) 4 (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3 (B,H) 4
4
H 1
(C,D) 3 (A,H) 5
2
3 (G,H) 3 (D,F) 6
G 3
E (C,F) 3 (A,B) 8
(B,C) 4 (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3 (A,H) 5
2
3 (G,H) 3 (D,F) 6
G 3
E (C,F) 3 (A,B) 8
(B,C) 4 (A,F) 10

Accepting edge (E,G) would

create a cycle
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3 (D,F) 6
G 3
E (C,F) 3 (A,B) 8
(B,C) 4 (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3 (A,B) 8
(B,C) 4 (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3  (A,B) 8
(B,C) 4 (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3  (A,B) 8
(B,C) 4  (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4 
4
8
6 (D,G) 2  (B,F) 4
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3  (A,B) 8
(B,C) 4  (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4 
4
8
6 (D,G) 2  (B,F) 4 
5
4
B D (E,G) 3  (B,H) 4
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3  (A,B) 8
(B,C) 4  (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4 
4
8
6 (D,G) 2  (B,F) 4 
5
4
B D (E,G) 3  (B,H) 4 
4
H 1
(C,D) 3  (A,H) 5
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3  (A,B) 8
(B,C) 4  (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
10
F C edge dv edge dv

A 4 3 (D,E) 1  (B,E) 4 
4
8
6 (D,G) 2  (B,F) 4 
5
4
B D (E,G) 3  (B,H) 4 
4
H 1
(C,D) 3  (A,H) 5 
2
3 (G,H) 3  (D,F) 6
G 3
E (C,F) 3  (A,B) 8
(B,C) 4  (A,F) 10
 Select first |V|–1 edges
which do not generate a
cycle
3
F C edge dv edge dv

A 3 (D,E) 1  (B,E) 4 
4
(D,G) 2  (B,F) 4 
5
B D (E,G) 3  (B,H) 4 
H 1
(C,D) 3  (A,H) 5 
2


}
3 (G,H) 3 (D,F) 6
G E (C,F) 3  (A,B) 8
not
consider
(B,C) 4  (A,F) 10 ed

Done
Total Cost =  dv =
21
 Minimum Connector Algorithms

Kruskal’s algorithm Prim’s algorithm

1. Select the shortest 1. Select any vertex

edge in a network
2. Select the shortest
2. Select the next edge connected to
shortest edge which that vertex
does not create a
cycle 3. Select the shortest
edge connected to
3. Repeat step 2 until any vertex already
all vertices have connected
been connected
4. Repeat step 3 until
all vertices have
been connected
Example
A cable company want to connect five villages to their network which
currently extends to the market town of Avonford. What is the minimum
length of cable needed?

Brinlei 5 Cornwel
gh l
3
4
8 6
8
Avonfor Fingle Donster
d 7 y
5
4 2

Edan
We model the situation as a network, then the
problem is to find the minimum connector for
the network

B 5 C

3
4
8 6
8
A F D
7
5
4 2

E
 Kruskal’s Algorithm
List the edges in
order of size:
B 5 C ED 2
AB 3
3 AE 4
4
8 6 CD 4
BC 5
8 EF 5
A D CF 6
7 F
AF 7
5 BF 8
4 CF 8
2

E
 Kruskal’s Algorithm

Select the
shortest
B 5 C edge in the
network
3
4
8 6 ED 2

8
A D
7 F

5
4 2

E
 Kruskal’s Algorithm
Select the next
shortest
B 5 edge which does not
C create a cycle
3 ED 2
4
8 6 AB 3

8
A D
7 F

5
4 2

E
 Kruskal’s Algorithm
Select the next
shortest
B 5 edge which does not
C create a cycle
3 ED 2
4
8 6 AB 3
CD 4 (or AE 4)
8
A D
7 F

5
4 2

E
 Kruskal’s Algorithm
Select the next
shortest
B 5 edge which does not
C create a cycle
3 ED 2
4
8 6 AB 3
CD 4
8 AE 4
A D
7 F

5
4 2

E
 Kruskal’s Algorithm
Select the next
shortest
B 5 edge which does not
C create a cycle
3 ED 2
4
8 6 AB 3
CD 4
8 AE 4
A D BC 5 – forms a cycle
7 F
EF 5
5
4 2

E
 Kruskal’s Algorithm
All vertices have
been
B 5 connected.
C
The solution is
3
4
8 6 ED 2
AB 3
8 CD 4
A D AE 4
7 F
EF 5
5
4 2 Total weight of tree:
18

E
 Prim’s Algorithm
Select any vertex

B 5 A
C
Select the shortest
3 edge connected to
4
8 6 that vertex

8 AB 3
A D
7 F

5
4 2

E
 Prim’s Algorithm
Select the shortest
edge connected to
B 5 any vertex already
C connected.
3 AE 4
4
8 6

8
A D
7 F

5
4 2

E
 Prim’s Algorithm
Select the shortest
edge connected to
B 5 any vertex already
C connected.
3 ED 2
4
8 6

8
A D
7 F

5
4 2

E
 Prim’s Algorithm
Select the shortest
edge connected to
B 5 any vertex already
C connected.
3 DC 4
4
8 6

8
A D
7 F

5
4 2

E
 Prim’s Algorithm
Select the shortest
edge connected to
B 5 any vertex already
C connected.
3 EF 5
4
8 6

8
A D
7 F

5
4 2

E
 Prim’s Algorithm
All vertices have
been
B 5 connected.
C
The solution is
3
4
8 6 AB 3
AE 4
8 ED 2
A D DC 4
7 F
EF 5
5
4 2 Total weight of tree:
18

E
 Some points to note

•Both algorithms will always give solutions

with the same length.

•They will usually select edges in a different

order – you must show this in your workings.

•Occasionally they will use different edges –

this may happen when you have to choose
between edges with the same length. In this
case there is more than one minimum
connector for the network.
Shortest – Route Problem
The Shortest Route Problem
Solution Method Summary
Select the node with the shortest direct route from the
origin.
Establish a permanent set with the origin node and
the node that was selected in step 1.
Determine all nodes directly connected to the
permanent set nodes.
Select the node with the shortest route (branch) from
the group of nodes directly connected to the
permanent set nodes.
Repeat steps 3 and 4 until all nodes have joined the
permanent set.

173
 Example: Shortest Route
• Find the Shortest Route From Node 1 to
All Other Nodes in the Network:
5
2 5

4 6
3
2
7 3 7
1 3

5 1 2
6
4 6
8
 Example: Shortest Route
• Iteration 1
– Step 1: Assign Node 1 the
permanent label [0,S].
– Step 2: Since Nodes 2, 3, and 4
are directly connected to Node 1,
assign the tentative labels of (4,1)
to Node 2; (7,1) to Node 3; and
(5,1) to Node 4.
 Example: Shortest Route
• Tentative Labels Shown
(4,1)
5
2 5

4 6
3 2
7 3 7
[0,S] 1 3 (7,1)

5 1 2
6

(5,1) 4 6
8
 Example: Shortest Route
• Iteration 1
– Step 3: Node 2 is the tentatively labeled node
with the smallest distance (4) , and hence
becomes the new permanently labeled node.
 Example: Shortest Route
• Permanent Label Shown
[4,1]
5
2 5

4 6
3 2
7 3 7
[0,S] 1 3 (7,1)

1
5 6 2

(5,1) 4 6
8
 Example: Shortest Route
• Iteration 1
– Step 4: For each node with a tentative label which
is connected to Node 2 by just one arc, compute
the sum of its arc length plus the distance value of
Node 2 (which is 4).
Node 3: 3 + 4 = 7 (not smaller than
current label; do not change.)
Node 5: 5 + 4 = 9 (assign tentative label to
Node 5 of (9,2) since node 5 had no label.)
 Example: Shortest Route
• Iteration 1, Step 4 Results
[4,1] (9,2)
5
2 5

4 6
3 2
7 3 7
[0,S] 1 3 (7,1)

1
5 6 2

(5,1) 4 6
8
 Example: Shortest Route

• Iteration 2
– Step 3: Node 4 has the smallest
tentative label distance (5). It now
becomes the new permanently
labeled node.
 Example: Shortest Route
• Iteration 2, Step 3 Results
[4,1] (9,2)
5
2 5

4 6
3 2
7 3
[0,S] 1 3 (6,4) 7

1
5 6 2

[5,1] 4 6
8
 Example: Shortest Route
• Iteration 2
– Step 4: For each node with a tentative label which
is connected to node 4 by just one arc, compute
the sum of its arc length plus the distance value of
node 4 (which is 5).
Node 3: 1 + 5 = 6 (replace the tentative
label of node 3 by (6,4) since 6 < 7, the current
distance.)
Node 6: 8 + 5 = 13 (assign tentative label
to node 6 of (13,4) since node 6 had no label.)
 Example: Shortest Route

• Iteration 2, Step 4 Results

[4,1] (9,2)
5
2 5

4 6
3 2
7 3
[0,S] 1 3 (6,4) 7

1
5 6 2

[5,1] 4 6 (13,4)
8
 Example: Shortest Route
• Iteration 3
– Step 3: Node 3 has the smallest tentative distance
label (6). It now becomes the new permanently
labeled node.
 Example: Shortest Route
• Iteration 3, Step 3 Results
[4,1] (9,2)
5
2 5

4 6
3 2
7 3
[0,S] 1 3 [6,4] 7

1
5 6 2

[5,1] 4 6 (13,4)
8
 Example: Shortest Route
• Iteration 3
– Step 4: For each node with a tentative label which
is connected to node 3 by just one arc, compute
the sum of its arc length plus the distance to node
3 (which is 6).
Node 5: 2 + 6 = 8 (replace the tentative
label of node 5 with (8,3) since 8 < 9, the current
distance)
Node 6: 6 + 6 = 12 (replace the tentative
label of node 6 with (12,3) since 12 < 13, the
current distance)
 Example: Shortest Route
• Iteration 3, Step 4 Results
[4,1] (8,3)
5
2 5

4 6
3 2
7 3 7
[0,S] 1 3 [6,4]
1
5 6 2

[5,1] 4 6 (12,3)
8
 Example: Shortest Route
• Iteration 4
– Step 3: Node 5 has the smallest tentative label
distance (8). It now becomes the new
permanently labeled node.
 Example: Shortest Route
• Iteration 4, Step 3 Results
[4,1] [8,3]
5
2 5

4 6
3 2
7 3 7
[0,S] 1 3 [6,4]

1
5 6 2

[5,1] 4 6 (12,3)
8
 Example: Shortest Route
• Iteration 4
– Step 4: For each node with a tentative label which
is connected to node 5 by just one arc, compute
the sum of its arc length plus the distance value of
node 5 (which is 8).
Node 6: 3 + 8 = 11 (Replace the tentative
label with (11,5) since 11 < 12, the current
distance.)
Node 7: 6 + 8 = 14 (Assign
 Example: Shortest Route
• Iteration 4, Step 4 Results
[4,1] [8,3]
5
2 5

4 6
3 2
7 3 7 (14,5)
[0,S] 1 3 [6,4]

1
5 6 2

[5,1] 4 6 (11,5)
8
 Example: Shortest Route
• Iteration 5
– Step 3: Node 6 has the smallest tentative label
distance (11). It now becomes the new
permanently labeled node.
 Example: Shortest Route
• Iteration 5, Step 3 Results
[4,1] [8,3]
5
2 5

4 6
3 2
7 3 7 (14,5)
[0,S] 1 3 [6,4]

1
5 6 2

[5,1] 4 6 [11,5]
8
 Example: Shortest Route
• Iteration 5
– Step 4: For each node with a tentative label which
is connected to Node 6 by just one arc, compute
the sum of its arc length plus the distance value of
Node 6 (which is 11).
Node 7: 2 + 11 = 13 (replace the tentative
label with (13,6) since 13 < 14, the current
distance.)
 Example: Shortest Route
• Iteration 5, Step 4 Results
[4,1] [8,3]
5
2 5

4 6
3 2
7 3 7 (13,6)
[0,S] 1 3 [6,4]

1
5 6 2

[5,1] 4 6 [11,5]
8
Example: Shortest Route
• Iteration 6
– Step 3: Node 7 becomes
permanently labeled, and hence all
nodes are now permanently
labeled. Thus proceed to
summarize in Step 5.
– Step 5: Summarize by tracing the
shortest routes backwards through
the permanent labels.
 Example: Shortest Route
• Solution Summary
Node Minimum Distance Shortest Route
2 4 1-2
3 6 1-4-3
4 5 1-4
5 8 1-4-3-5
6 11 1-4-3-5-6
7 13 1-4-3-5-6-7
 Dijkstra’s Shortest-Path Algorithm
• Iterative algorithm
– After k iterations, know least-cost path to k nodes
• S: nodes whose least-cost path definitively known
– Initially, S = {u} where u is the source node
– Add one node to S in each iteration
• D(v): current cost of path from source to node v
– Initially, D(v) = c(u,v) for all nodes v adjacent to u
– … and D(v) = ∞ for all other nodes v
– Continually update D(v) as shorter paths are learned

199
Maximal Flow Problem
 Maximum Flow Problem
• Given: Directed graph G=(V, E),
Supply (source) node O, demand (sink) node T
Capacity function u: E  R .
• Goal: Given the arc capacities,
send as much flow as possible
from supply node O to demand node T
through the network.
• Example: A
4

4
O B 6 D 5
5 T
4 4

C 5
 Towards the Augmenting Path
Algorithm
• Idea: Find a path from the source to the sink,
and use it to send as much flow as possible.
• In our example,
5 units of flow can be sent through the path O  B  D  T
;
Then use the path O  C  T to send 4 units of flow.
The total flow is 5 + 4 = 9 at this point.
• Can we send more?
A
4

4 5
O 5 B 5
D
4 5 6 5 T
4 4
4
C
5
 Towards the Augmenting Path
Algorithm
• If we redirect 1 unit of flow
from path O  B  D  T to path O  B  C  T,
then the freed capacity of arc D  T could be used
to send 1 more unit of flow through path O  A  D  T,
making the total flow equal to 9+1=10 .
• To realize the idea of redirecting the flow in a systematic way,
we need the concept of residual capacities.

A
1 4 1
4
5 45 4
O B D 5
4 5 6 5 T
4 4 1 5
4
C
5
 Residual capacities
• Suppose we have an arc with capacity 6 and current flow 5:
5
B D
6
• Then there is a residual capacity of 6-5=1
for any additional flow
through B  D .
• On the other hand,
at most 5 units of flow can be sent back from D to B, i.e.,
5 units of previously assigned flow can be
canceled.
In that sense, 5 can be considered as
the residual capacity of the reverse arc D  B .
• To record the residual capacities in the network,
we will replace the original directed arcs with undirected arcs:

B 1 5 The number at B is the residual capacity of BD;

D
the number at D is the residual capacity of DB.
 Residual Network
• The network given by the undirected arcs and residual capacities
is called residual network.
• In our example,
the residual network before sending any flow:

0 A 4

4 0
5 0
O 4 B 6 0 D 5 0
4 T
0
0 0
5
C
Note that the sum of the residual capacities on both ends of an arc
is equal to the original capacity of the arc.
• How to increase the flow in the network
based on the values of residual capacities?
 Augmenting paths
• An augmenting path is a directed path
from the source to the sink in the residual
network
such that
every arc on this path has positive residual
capacity.
• The minimum of these residual capacities
is called the residual capacity of the
augmenting path.
This is the amount
 Updating the residual network
by sending flow through augmenting
paths
Continuing with the example,
• Iteration 1: O  B  D  T is an augmenting path
with residual capacity 5 = min{5, 6, 5}.
• After sending 5 units of flow
through the path O  B  D  T,
the new residual network is: A
0 4

4 5 0 0
0 5 B 6
1 0
5
O
4 D 5
0 0
5 T
4
0
0 0
C 5
 Updating the residual network
by sending flow through augmenting
paths

• Iteration 2:
O  C  T is an augmenting path
with residual capacity 4 = min{4, 5}.
• After sending 4 units of flow
through the path O  C  T,
the new residual network is: 0 A 4
4 0
O 0 0 5 B 1 5
4 D 0 5 T
4 4
0
4
0 0
C 1 5
 Updating the residual network
by sending flow through augmenting
paths
• Iteration 3:
O  A  D  B  C  T is an augmenting path
with residual capacity 1 = min{4, 4, 5, 4, 1}.
• After sending 1 units of flow
through the path O  A  D  B  C  T ,
the new residual network is:

10 A 3
4
34 10
O 0 5 B 21 45
0 D 0 5 T
34
54
4 10
C 01
 Terminating the Algorithm:
Returning an Optimal Flow

• There are no augmenting paths in the last residual network.

So the flow from the source to the sink cannot be increased
further, and the current flow is optimal.
• Thus, the current residual network is optimal.
The optimal flow on each directed arc of the original network
is the residual capacity of its reverse
arc:
flow(OA)=1, flow(OB)=5, flow(OC)=4,
flow(AD)=1, flow(BD)=4, flow(BC)=1,
flow(DT)=5, flow(CT)=5.
The amount of maximum flow through the network is
5 + 4 + 1 = 10
(the sum of path flows of all iterations).
 The Summary of the
Augmenting Path Algorithm
• Initialization: Set up the initial residual network.
• Repeat

– Find an augmenting path.

– Identify the residual capacity c* of the path;
increase the flow in this path by c*.
– Update the residual network: decrease by c* the
residual capacity of each arc on the augmenting
path; increase by c* the residual capacity of each
arc in the opposite direction on the augmenting
path.
Until no augmenting path is left
• Return the flow corresponding to
the current optimal residual network
 Example: Maximal Flow
• Network Model
3
2 5
3
4 2 3 2
Source 3 4 Sink
4 3
1 4 7
3 1
3 5 1 5
3 6
6
 Example: Maximal Flow
• A problem that can be developed for the
maximal flow problem.
• We will add an arc from node 7 back to node 1
to represent the total flow through the
network.
• There is no capacity on the newly added 7-1
arc.
• We want to maximize the flow over the 7-1
arc.
 Example: Maximal Flow

• Modified Network Model

3
2 5
3
4 2 3 2
Source 3 4 Sink
4 3
1 4 7
3 1
3 5 1 5
3 6
6
 Maximal Flow Problem
• LP Formulation
(as Capacitated Transshipment Problem)
– There is a variable for every arc.
– There is a constraint for every node; the flow out
must equal the flow in.
– There is a constraint for every arc (except the
added sink-to-source arc); arc capacity cannot be
exceeded.
– The objective is to maximize the flow over the
added, sink-to-source arc.
 Maximal Flow Problem
• LP Formulation
(as Capacitated Transshipment Problem)
Max xk1 (k is sink node, 1 is source node)
s.t. xij - xji = 0 (conservation of
flow) i j

xij < cij (cij is capacity of ij arc)

xij > 0, for all i and j (non-negativity)
(xij represents the flow from node i to node j)
 Example: Maximal Flow

• LP Formulation
– 18 variables (for 17 original arcs and
1 added arc)
– 24 constraints
• 7 node flow-conservation constraints
• 17 arc capacity constraints (for original
arcs)
 Example: Maximal Flow
• LP Formulation
– Objective Function
Max x71
– Node Flow-Conservation Constraints
x71 - x12 - x13 - x14 = 0 (flow in & out of node 1)
x12 + x42 + x52 – x24 - x25 = 0 (node 2)
x13 + x43 – x34 – x36 = 0 (etc.)
x14 + x24 + x34 + x54 + x64 – x42 - x43 - x45 - x46 - x47 = 0
x25 + x45 – x52 – x54 - x57 = 0
x36 + x46 - x64 - x67 = 0
x47 + x57 + x67 - x71 = 0
 Example: Maximal Flow
• LP Formulation (continued)
– Arc Capacity Constraints
x12 < 4 x13 < 3 x14 < 4
x24 < 2 x25 < 3
x34 < 3 x36 < 6
x42 < 3 x43 < 5 x45 < 3 x46 < 1
x47 < 3
x52 < 5 x54 < 5 x57 < 5
x64 < 5 x67 < 5
 Example: Maximal Flow
• Optimal Solution
2
2 5
3 1 2
Source Sink
4 3
1 4 7
1
3 1 5
3 6
4

10
 INVENTORY MODELS (1)
• OBJECTIVE: FIND THE MINIMUM TOTAL INVENTORY COSTS
BY OPTIMIZING INVENTORY LEVEL
• CONSTRAINTS:
– STOCKS OF GOODS (PARTS, RAW MATERIALS, PRODUCTS) MUST BE
SUFFICIENT TO MAINTAIN A SMOOTH OPERATION OF A
PRODUCTION SYSTEM OR A BUSINESS ACTIVITY
– LIMITED SPACE (EX:OFFSHORE OPERATIONS).SINCE THE NEW
SYSTEM JUST-IN-TIME (JIT) WAS INTRODUCED IN JAPAN, THE
INVENTORY LEVEL COULD BE ALMOST “ZERO”. THIS IS SUITABLE
FOR HIGHLY REPETITIVE ASSEMBLY MANUFACTURING (“STEADY
STATE” FLOW), EX:AUTOMOBILE INDUSTRY.
• LEVELS OF INVENTORY CONTROL:
1. TIGHT (GROUP A):80% OF THE CUMULATIVE DOLLAR VALUE
CONTRIBUTED BY ONLY 20% OF ALL ITEMS
2. MODERATE (GROUP B):80-95% VALUE BY 25% OF ALL ITEMS
3. LOOSE (GROUP C): LARGER QUANTITY
 INVENTORY MODELS (2)
ABC CURVE / PARETO(?)
100
PERCENT OF TOTAL DOLLAR VALUE

80

60

40
A B C

20

20 40 60 80 100
PERCENT OF TOTAL ITEMS
 INVENTORY MODELS (3)
• TWO QUESTIONS IN THE DECISION PROCESS:
1. HOW MUCH TO ORDER? TERM: ORDER QUANTITY
2. WHEN TO ORDER?TWO TYPES OF INVENTORY SYSTEM:
• PERIODIC REVIEW: AT EQUAL TIME INTERVALS (WEEKLY, MONTHLY)
• CONTINUOUS REVIEW: A REORDER POINT IS USUALLY SPECIFIED BY
THE INVENTORY LEVEL AT WHICH A NEW ORDER MUST BE PLACED
• THE ORDER QUANTITY AND REORDER POINT ARE
NORMALLY DETERMINED BY MINIMIZING THE TOTAL
INVENTORY COST THAT CAN BE EXPRESSED AS A
FUNCTION OF TWO VRBS IN ITS PRINCIPAL COMPONENTS:
TOTAL SHORTAGE
PURCHASING SETUP HOLDING
INVENTORY = + + + COST
COST COST COST
COST
 INVENTORY MODELS (4)
• PURCHASING COST: BECOMES IMPORTANT WHEN THE COMMODITY
UNIT PRICE BECOMES DEPENDENT ON THE SIZE OF THE ORDER. IN CASE
OF QUANTITY DISCOUNT OR PRICE BREAK, WHERE THE UNIT PRICE OF
THE ITEM DECREASES WITH THE INCREASE OF ORDERED QUANTITY
• SETUP COST: FIXED CHARGE INCURRED WHEN AN ORDER IS PLACED.
THUS, MORE FREQUENT ORDERING OF SMALL QUANTITIES WILL RESULT
IN HIGHER SETUP COST THAN ORDERING LARGER ORDER
• HOLDING COST: COSTS OF CARRYING INVENTORY IN STOCK,EX: INTEREST
ON INVESTED CAPITAL, STORAGE, HANDLING, DEPRECIATION, &
MAINTENANCE; NORMALLY INCREASES WITH THE LEVEL OF INVENTORY
• SHORTAGE COST: A PENALTY INCURRED WHEN WE RUN OUT OF STOCK
OF A NEEDED COMMODITY. IT GENERALLY INCLUDES COSTS DUE TO LOSS
OF CUSTOMER’S GOODWILL AS WELL AS POTENTIAL LOSS IN INCOME
INVENTORY MODELS (5)
 SINGLE-ITEM STATIC MODEL (1)
• DEMAND IS CONSTANT OVER TIME
• INSTANTANEOUS REPLENISHMENT
• NO SHORTAGES

POINTS IN TIME AT WHICH ORDERS ARE RECEIVED

INVENTORY LEVEL

AVERAGE INVENTORY = Y/2

TO=Y/D
TIME
SAWTOOTH DIAGRAM
 SINGLE-ITEM STATIC MODEL (2)
• D = DEMAND RATE PER UNIT TIME
• Y = ORDER QUANTITY
• THE INVENTORY LEVEL REACHES ZERO LEVEL Y/D TIME UNITS AFTER THE ORDER
QUANTITY Y IS RECEIVED
• K = SETUP COST INCURRED EVERY TIME AN ORDER IS PLACED
• H = HOLDING COST PER UNIT INVENTORY PER UNIT TIME
• TCU = TOTAL COST PER UNIT TIME
TCU(Y) = SETUP / UNIT TIME + HOLDING COST / UNIT TIME

K Y
TCU (Y )   H( )
Y /D 2
• LENGTH OF EACH INVENTORY CYCLE IS TO=Y/D
• AVERAGE INVENTORY IN STOCK IS Y/2
 SINGLE-ITEM STATIC MODEL (3)

• THE OPTIMUM VALUE OF Y IS

OBTAINED BY MINIMIZING TCU(Y)
WITH RESPECT TO Y. THUS,
ASSUMING THAT Y IS A
CONTINUOUS VRB, WE HAVE:

dTCU (Y ) KD H
 2  0
dY Y 2
2 KD ECONOMIC ORDER QUANTITY
Y*  (WILSON’S ECONOMIC LOT SIZE)
H
TCU MIN  2 KDH MINIMUM TOTAL COST
Model 2: Jumlah Barang Ekonomis,
Siklus Produksi, Tanpa Kekurangan
Barang (1)
 q = Penyediaan barang, tp = waktu yang diperlukan untuk
menyediakan barang, p = laju produksi (barang/waktu).
 Tujuan: Mencari q yang Meminimumkan biaya
 p * tp = q
 Max. Inventory = tp * (p-u) = (q/p) * (p-u) = q * (1-(u/p))
 Jumlah tertinggi = tp * (p-u)
 Rata-rata stock = 0.5 * max Inventory
229
Model 2: Jumlah Barang Ekonomis,
Siklus Produksi, Tanpa Kekurangan
Barang (2)
 Biaya persiapan = Au / q
 Biaya angkutan / penempatan = cq / 2
 Biaya total = (Au/q) + ((cq/2) * (1 – (u/p)))

2𝐴𝑢 1
 Cari min 𝑞 ∗ = or 𝑐 ∗ = 2𝐴𝑢𝑐 1 − 𝑢/𝑝
𝑐 1− 𝑢/𝑝

230
Model 3: Jumlah Barang Ekonomis,
Pengadaan Barang Seketika, Boleh
Kekurangan Barang (1)
 q = Jumlah barang yang dipesan.
 S = Jumlah kekurangan max. dalam tiap siklus.
 Jumlah barang dalam stock = q – s
𝑞 −𝑆
 Jumlah rata-rata barang dalam inventory = dalam siklus
2

𝑡𝑑 𝑞−𝑠
fraksi waktu (td/t)  =
𝑡 𝑞

 t = td + ts
231
Model 3: Jumlah Barang Ekonomis,
Pengadaan Barang Seketika, Boleh
Kekurangan Barang (2)
𝑐 𝑞−𝑠 2
 Biaya penempatan =
2𝑞

𝑏𝑠 2
 Biaya kekurangan barang =
2𝑞

b = biaya pemesanan barang yang tidak dapat dipenuhi seketika.

𝐴𝑢 𝑐𝑞 𝑐𝑠 2 +𝑏𝑠 2
 Biaya total : 𝐶 = + − 𝑐𝑠 +
𝑞 2 2𝑞

232
Model 4: Jumlah Barang Ekonomis,
Siklus Produksi, Boleh Kekurangan
Barang
𝑞𝐴𝑢 1 𝑏+𝑐
 𝑞∗ =
𝑐 1− 𝑢/𝑝 𝑏

𝑏
 𝑐∗ = 2𝐴𝑢𝑐 1 − 𝑢/𝑝
𝑏+𝑐

2𝐴𝑢 𝑐
 𝑆∗ = 1 − 𝑢/𝑝
𝑏 𝑏+𝑐

233
 Dynamic Models
 The inventory level is reviewed periodically over a finite number
of equal periods.
 The demand per period, is dynamic, in the sense that it varies
from one period to the next.
 Two models of Dynamic model:
1. No-Setup Model
2. Setup Model

234
 No – Setup Model
This model involves a planning horizon with n equal periods. Each
period has a limited production capacity that can include several
production levels (e.g., regular time and overtime represent two
production levels).

A current period may produce more than its immediate demand to

satisfy demand for later periods, in which case an inventory holding
cost must be charged.

The general assumptions of the model are

1. No setup cost is incurred in any period.
2. No shortage is allowed.
3. The unit production cost function in any period either is constant or
has increasing (convex) marginal costs.
4. The unit holding cost in any period is constant.
235
 Setup Model
In this situation, no shortage is allowed and a setup cost is incurred
each time a new production lot is started.

𝒛𝟏 𝒛𝟐 𝒛𝒊 𝒛𝒊+𝟏 𝒛𝒏

𝒙𝟏 𝒙𝟐 𝒙𝒊 𝒙𝒊+𝟏 𝒙𝒏

𝑫𝟏 𝑫𝒊 𝑫𝒏

𝒛𝒊 = amount ordered 𝑫𝒊 = demand 𝒙𝒊 = inventory

236
 Discounts
Order quantity above certain amount (q ≥ q)

Discount (D%) from original price (k)

New price = (1 – D) k / item

Phenomena during discounts:

1. Savings due to discounts.
2. Savings on preparation costs.
3. Additional expenditure for greater inventory.

237
 Scheduling issue
job scheduling for each item.

Determine f (production demand / cycle)

qj = amount of items
uj = usage
qj = f * uj j = 1, 2, …, n

238
 Sequencing (1)
Models have been developed to find a sequence for Processing Jobs so that
the total elapsed time for all the jobs will be minimum.

The models also help to resolve the conflict between the objectives of
maximizing machines utilization and complying with predetermined
delivering rates.

work can be exchanged in sequence with other jobs.

Example: Two – Jobs, Two - Machines

M1 (time) M2 (time)
P1 2 7
P2 5 4

239
 Sequencing (2)
Gantt – chart for the problem:

P1 P2
M1 P2 await

P1 P2
M2
5 hours
0 13 hrs

P1 – P2
P2 P1 P1 await
M1

M2 P2
2 hours
0 16 hrs
240
P2 – P1
Thank You 

241