–19-12-2020

CVL 245
IIT DELHI
CIVIL ENGINEERING MATERIALS.

( Lecture 11)
Project management: planning

B. Bhattacharjee
CIVIL ENGINEERING DEPARTMENT
IIT DELHI

B. Bhattacharjee
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

General Outline
™Work breakdown structure
™Network AOA
™AON
™CPM
™PERT

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

2

–1
 –19-12-2020

ES,EF
Early start (ES) of an activity is the earliest time
it can start. For all starting activity it is usually
zero
Duration of activity i-j say is Dij ,
ES of activity [(A)/(i-j)]=Max of EF (early finish)
of all predecessor activities
EF of activity [(A)/(i-j)]=ES [(A)/(i-j)]+Dij
Project duration T=Max (EF time of all activities)
ES of event j= max [ESi + Dij] for i=1,2…j-1
Forward pass calculation

B. Bhattacharjee
3
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

LS, LF
Late finish(LF) of an activity is the latest time it
can complete. For all finishing activity it is T
the project duration
LF of activity [(A)/(i-j)]=Min of LS (late start) of
all successor activities
LS of activity [(A)/(i-j)]=LF [(A)/(i-j)]-Dij
Total Float=LS (activity)-ES(activity)=LF (activity)-
EF(activity)
Free float= min (ES of all successor activity)-EF
(activity
Backward pass calculation
B. Bhattacharjee
4
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–2
 –19-12-2020

LS, LF
For critical activity: ES(activity)-LS(activity) =0;
EF(activity)-LF(activity) =0; i.e, total float=0

Total Float TFij=LFj-ESi-Dij

Free float FFij= ESj-ESi-Dij

B. Bhattacharjee
5
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

A Automated distribution center

Request for
proposal

B. Bhattacharjee
6
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–3
 –19-12-2020

AON Network

4 Design Layout
5 Implement Layout
ES=10 EF=17
1 Obtain RFP D=7 ES=17 EF=26
2 Select Vendor LS=10 LF=17 D=9
ES=0 EF=6 ES=6 EF=10 LS=17 LF=26
D=6 D=4
LS=0 LF=6 LS=6 LF=10
3 Order Systems
ES=10 EF=16
D=6
LS=20 LF=26
10 Train user
ES=35 EF=37 7 Configure software
D=2
LS=35 LF=37 ES=29 EF=32 8 Install equipment
D=3
LS=29 LF=32 6 Install computers
ES=26 EF=31
ES=26 EF=29 D=5
END D=3 LS=27 LF=32
LS=26 LF=29
9 Test System
ES=32 EF=35
D=3
LS=32 LF=35

B. Bhattacharjee
7
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

Cont…

B. Bhattacharjee
8
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–4
 –19-12-2020

Cont…Gantt Chart (Bar chart)

B. Bhattacharjee
9
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

PERT
Project Evaluation and Review Techniques
(PERT)
So far we assumed activity durations are known
with certainty.
Where past experience is inadequate and idea
about durations are scanty, such as a research
project,
PERT considers the uncertainty involved in
estimation of activity duration and resulting
uncertainty in project duration/completion.
PERT deals with such situations.

B. Bhattacharjee
10
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–5
 –19-12-2020

PERT
Uncertainty in activity duration arises out of
variation of durations of component work
elements of the activity
General pattern of variation of activity duration can
be visualized by considering a situation where
same activity is repeatedly performed several times
i.e., sufficiently large number of times.
For a given activity which is repeated large
number of say the minimum duration is Dmin and
Maximum duration is Dmax, one can divide the
range Dmax, - Dmin in to several class interval
{NCI=3.3log(n)+1}, & plot a histogram.

B. Bhattacharjee
11
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

PERT (activity duration)

Common experience shown in histogram
plot
250

200
Most often
Frequency

150

100 Series1

50

0
5 6 7 8 9 10 11 12 13 14 15 16 17
Duration

Frequency Histogram Dmax =17, Dmin =5

B. Bhattacharjee
12
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–6
 –19-12-2020

PERT (activity duration)

Common experience shown in histogram plot can
be represented by 2 parameter β distribution
Most often Mean

Beta distribution
two parameter α=2 β=5
B. Bhattacharjee
13
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

PERT (activity duration)

X axis is (D-Dmin)/(Dmax-Dmin), Y axis is
probability density
1<α <β

Beta distribution
two parameter α=2 β=4

β function
B. Bhattacharjee
14
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–7
 –19-12-2020

PERT (activity duration)

X axis is (D-Dmin)/(Dmax-Dmin), Y axis is
probability density 1<α <β

0.25×(17-5)+5=8=mode
1/3×(17-5)+5=9=mean Beta distribution
two parameter α=2 β=4

β function

B. Bhattacharjee
15
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

PERT duration
Thus expected (mean) activity duration is
calculated from 3 time information in PERT,
pessimistic duration tp, most likely duration tm,
and optimistic duration to.

Network and subsequent calculations follows same

way as done for CPM using expected activity
duration
The project duration calculated is expected
project duration.
PERT deals probabilistic project duration.

B. Bhattacharjee
16
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–8
 –19-12-2020

NATURE OF VARIATIONS
(uniform distribution)
0.18

0.16

0.14 ¾ The variations

Relative Frequency

0.12

0.1
Leads to a uniformly
0.08
Distributed population
0.06

0.04

0.02 fi
0 relativefrequncy = n
N = ∑ fi
1 2 3 4 5 6
Outcomes

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

NATURE OF VARIATIONS
(triangular distribution)
0.18

0.16 ¾ Two dies

0.14
The variations
Relative frequency

0.12

0.1 Leads to a triangular

0.08

0.06
Distributed population
fi
relativefrequncy =
0.04

0.02 n
0
1 2 3 4 5 6 7 8 9 10 11
N = ∑ fi
outcomes 1
1 2 3 4 5 6 7 8 9 10 11
2 3 4 5 6 7 8 9 10 11 12

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–9
 –19-12-2020

NATURE OF VARIATIONS
(nonlinear distribution)
0.06

0.05
¾ Three dies
0.04 The variations
0.03 Leads to a nonlinear
0.02
Distributed population
0.01 fi
relativefrequncy = n
N = ∑ fi
0
1 2 3 4

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

NATURE OF VARIATIONS (more

activities)

a= minimum duration
b=maximum duration

¾ Three activities in series: The variations Leads

to a nonlinearly Distributed population

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–10
 –19-12-2020

NATURE OF VARIATIONS (more

activities)

¾ Three parallel activities ending on a node: The

variations Leads to a nonlinearly Distributed
population
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

NATURE OF VARIATIONS

¾ For large number of

Frequency

Factors, i.e. activities

¾The variations
Leads a normally
Distributed population

Project duration
All critical activities finishes at
All critical activities finishes at
pessimistic duration
optimistic duration

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–11
 –19-12-2020

NATURE OF VARIATIONS

x−μ ¾ Area under the curve

Z= ;
σ is 1
⎛ x− μ ⎞
2
−⎜
⎝ σ ⎠
⎟ x−μ
e = 0 for = −∞ or + ∞
σ

1 ⎛ 1 ⎛ x − μ ⎞2 ⎞
f ( x) = exp ⎜ − ⎜
σ 2π ⎜ 2 ⎝ σ ⎟⎠ ⎟⎟
⎝ ⎠

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

NATURE OF VARIATIONS

ƒAverage represents mean

ƒSD measures
dispersion

Mean= Σx/n
s2=Σ(xi-x¯)2/(n-1)
Measure of
σ=s
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI
24

–12
 –19-12-2020

NATURE OF VARIATIONS

Mean= Σx/n
s2=Σ(xi-x¯)2/(n-1)

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

25

PERT duration
SD of activity duration.

Average is a measure of Mean & varies linearly

n
f i xi
mean μ = ∑ n
i =1
N = ∑ fi
1
For single dice experiment mean =
6
i (6 × 7)
mean μ = ∑ = = 3.5
i =1 N =6 2
For two dice 6
mean μ = ∑
( i + 1) i + 11 ( i + 1) (12 − i) = 7 = 3.5 + 3.5
: ∑
i =1 N = 36 i =7 36

B. Bhattacharjee
26
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–13
 –19-12-2020

PERT duration
SD of activity duration.

Mean all activity duration is sum of means of

individual activity durations
mean μ = ∑ μ j ; j s tan dsforall criticalactivities
j

SD cannot be added as it involves square terms

SD square of activities can be added to obtain

project SD

B. Bhattacharjee
27
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

PERT duration
SD of activity duration.

Using the mean and standard deviations probability

of project completion for a given duration can be
calculated
x−μ
Z= ;
σ
Normal distribution table or graph given in the
next slide can be used.

B. Bhattacharjee
28
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

–14
 –19-12-2020

PERT duration
x−μ
Z= ;
σ
99.9% of time project
would be completed
by Mean+3×SD of
project duration, i.e.
Z=3
(95% of project will be
completed
Mean+1.65×SD of
project duration, i.e.
Z=1.65
In general p(D) for
D=μproj +Z×σproj from
graph or table

B. Bhattacharjee
29
DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

THANK YOU
FOR HEARING

DEPARTMENT OF CIVIL ENGINEERING, IIT DELHI

30

–15