Project Planning, Scheduling, and Control

ENMG 632

Program Evaluation Review Technique

(PERT)
Lesson 1: Basic Concepts
 Introduction
 Scheduling is the basis for the management of time on a construction project.

 The estimates for the various activities in a project form the basis on which the success
of the schedule relies.

 It has been so far assumed in CPM that the estimates of activity durations were
accurate and assigned to activities with confidence.

 A prudent scheduler should have a clear understanding of the uncertainty actually

associated with duration estimates.

 Even if a deterministic approach is used for duration estimates in CPM, it is important

to constantly recognize that uncertainty is an inherent characteristic of virtually every
activity.
ENMG 632: Project Planning, Scheduling, and Control 3
 Introduction (cont’d)
 Uncertainty in scheduling is considered in the method referred to as the Program
Evaluation Review Technique (PERT).

 PERT was developed in the late 1950s in collaboration between the U.S. Navy, Booz-
Allen Hamilton, and Lockheed Corporation for the creation of the Polaris missile
program.

 PERT was originally used in research and development (R&D), or other undertakings
where insufficient experience or historical data were available for estimating the
durations of individual activities in a project.

ENMG 632: Project Planning, Scheduling, and Control 4

 Introduction (cont’d)
 PERT, developed independently at the same time as the regular (deterministic) CPM,
is less commonly used in the construction industry and differs from CPM primarily in
the assumptions it makes about estimating activity durations.

 Like CPM, PERT uses logic networks to calculate the completion date of a project or
the date of any other event in the schedule.

 In PERT, a probability (likelihood) is associated with any event date.

 This probability depends on uncertainty in the durations of the activities that lead to
the desired event (e.g., project completion).

ENMG 632: Project Planning, Scheduling, and Control 5

 Introduction (cont’d)
 Accordingly, PERT realizes that actual durations vary from those assigned, so it
attempts to compensate for this variation with a ‘‘time range’’ during which activity
durations may realistically occur.

 This time range is defined by three values: (1) the activity’s most likely duration or the
mode (Tm), (2) the activity’s most pessimistic duration (Tp), and (3) the activity’s most
optimistic duration(T0).

 For these time estimates, probability distributions can be used that are either
symmetrical (the normal distribution) or asymmetrical (the beta distribution).

ENMG 632: Project Planning, Scheduling, and Control 6

 Introduction (cont’d)
 A normal distribution implies that the probability of an event taking the most likely
time is one that is centered on the mean of the distribution.

 Because pessimistic and optimistic values are estimated at the 95% confidence level
from either end of the distribution, they will cancel each other out, leaving the most
likely value as the mean or expected duration time for the activity (Te).

Tm /Mean= Te

ENMG 632: Project Planning, Scheduling, and Control 7

 Introduction (cont’d)
 In real life, it is extremely rare to find examples in which optimistic and pessimistic
durations are symmetrical to each other about the mean.

 In project management, it is more common to see probability distributions that are

asymmetrical; these are referred to as beta distributions.

 The asymmetry of the probability distribution suggests we recognize that certain events
are less likely to occur than others.

Mean= Te

ENMG 632: Project Planning, Scheduling, and Control 8

 Introduction (cont’d)
 The beta distribution suggests that the calculations for deriving Te and s are shown as:

ENMG 632: Project Planning, Scheduling, and Control 9

 Introduction (cont’d)

ENMG 632: Project Planning, Scheduling, and Control 10

Lesson 2: Uncertainty in Activity,
Activity Chain, and Project Duration
Estimates
 Uncertainty in Activity Duration Estimates (cont’d)

Duration= Mean+X*SD
 X= Duration-Mean/SD
= 30-24/8= 0.75

P(D<30)=P(X<0.75)
=0.7734
Normal Distribution Showing the Probability of Completing the Excavation Work in Less than 30 days

ENMG 632: Project Planning, Scheduling, and Control 12

 Uncertainty in Activity Duration Estimates (cont’d)

𝟒𝟎−𝟐𝟒
P(D<40)=P(X< )
𝟖
= P ( X <2)

= 0.9772

Normal Distribution Showing the Probability of Completing the Excavation Work in Less than 40 days

ENMG 632: Project Planning, Scheduling, and Control 13

 Uncertainty in Activity Duration Estimates (cont’d)
P(D<100)
P(D>100)= 1-P(D<100)
𝟏𝟎𝟎−𝟕𝟖
= 1- P(X< )
𝟏𝟔
=1-P(X<1.375)

= 1- 0.9155= 0.0845
Normal Distribution Showing the Probability of Completing an Activity in More than 100 Days

ENMG 632: Project Planning, Scheduling, and Control 14

 Uncertainty in Activity Duration Estimates (cont’d)
P(90<D<100)= P(D<100)- P(D<90)
𝟏𝟎𝟎−𝟕𝟖 𝟗𝟎−𝟕𝟖
= P (X < ) – P (X < )
𝟏𝟔 𝟏𝟔
= P( X < 1.375) – P ( X < 0.75)
= 0.9155 – 0.7734
= 0.1421
Normal Distribution Showing the Probability of Completing an Activity in More than 90 and Less than 100 Days

ENMG 632: Project Planning, Scheduling, and Control 15

 Uncertainty in Activity Chain Duration Estimates

Simple Precedence Network Using Three-Time Estimates

ENMG 632: Project Planning, Scheduling, and Control 16

 Uncertainty in Chain Duration Estimates (cont’d)

ENMG 632: Project Planning, Scheduling, and Control 17

 Uncertainty in Project Duration Estimates

Probability as Influenced by the Standard Deviation

ENMG 632: Project Planning, Scheduling, and Control 18

Lesson 3: PERT- Example I
 Example I - Given
What is the probability (in %) of completing the project in 55 days ? In 60
days? In 65 days?

ENMG 632: Project Planning, Scheduling, and Control 20

 Example I- Solution

ENMG 632: Project Planning, Scheduling, and Control 21

 Example I- Solution (cont’d)
Duration= Mean+X*SD
 X= Duration-Mean/SD

P(X< -a)= 1- P(X<a)

P(X<-1.09)= 1- P(X<1.09)

P(D<55)= P(X<-1.09)
=P(D>72.34)
= 1- P( D< 72.34)
𝟕𝟐.𝟑𝟒−𝟔𝟑.𝟔𝟕
= 1- P ( X<
𝟕.𝟗𝟑
)
= 1-P (X<1.09)

ENMG 632: Project Planning, Scheduling, and Control 22

 Example I- Solution (cont’d)

ENMG 632: Project Planning, Scheduling, and Control 23

 Example I- Solution (cont’d)

ENMG 632: Project Planning, Scheduling, and Control 24

Lesson 4: PERT- Example II
 Example II- Given

ENMG 632: Project Planning, Scheduling, and Control 26

 Example II- Solution
 With regular CPM calculations (using Most Likely durations), the critical path is
ABDHI with 55 days followed by paths ABDG and ABEHI at 54 and 53 days;
respectively.

ENMG 632: Project Planning, Scheduling, and Control 27

 Example II- Solution (cont’d)
 Now, let’s do PERT calculations… Path ABDHI

To find the duration with a minimum level of

confidence of 95%, find the X value associated with
a probability of 0.95.

From the table, this X value is 1.645, and

Duration= Mean + X* SD or TS = TE + X*σE
= 55.5 + 1.645* 1.77= 58.41 days
ENMG 632: Project Planning, Scheduling, and Control 28
 Example II- Solution (cont’d)
Path ABDG

For a minimum level of confidence of 95%,

the X value is 1.645 and

Duration= Mean + X* SD or TS = TE + X*σE

= 56.33 + 1.645 * 4.91 = 64.41 days

ENMG 632: Project Planning, Scheduling, and Control 29

 Example II- Solution (cont’d)
Path ABEHI

For a minimum level of confidence of 95%,

the X value is 1.645 and

Duration= Mean + X* SD or TS = TE + X*σE

= 54.83 + 1.645 * 3.53 = 60.64 days

ENMG 632: Project Planning, Scheduling, and Control 30

 Example II- Solution (cont’d)
 Discussion of results
CPM PERT

Path ABDHI: Expected Duration = 55.50 days

Path ABDG: Expected Duration = 56.33 days

Path ABEHI: Expected Duration = 54.83 days

--------------------------------------------------------------------------------

Path ABDHI: Duration (σE & 95% CL) = 58.41 days

Path ABDG: Duration (σE & 95% CL) = 64.41 days

Path ABEHI: Duration (σE & 95% CL) = 60.64 days

ENMG 632: Project Planning, Scheduling, and Control 31

Lesson 5: Concluding Remarks
 Conclusion
 In CPM, when durations are assigned to activities, a best-guess number is generally
used.

 This single number assignment of a duration makes it easy to perform CPM network
calculations.

 Duration assignments are guesses and that considerable uncertainty may actually
exist in the value used  the errors of overestimation and underestimation will
cancel out each other.

ENMG 632: Project Planning, Scheduling, and Control 33

 Conclusion (cont’d)
 However, the use of PERT is more accurate in its depiction of a project as the
scheduler is made very aware of the uncertainty that is associated with each activity
duration estimate and is not only making best guesses.

 Despite being less commonly used in the construction industry than CPM, it is
worthwhile to have a good understanding of the PERT technique.

 PERT makes it possible for the scheduler to make more informed decisions about the
probability of achieving stated project durations and can thereby help in keeping
schedules more realistic.

 Manual PERT can be replaced with computerized Monte Carlo Simulation.

ENMG 632: Project Planning, Scheduling, and Control 34
Thank You