Page |1

PROGRAM EVALUATION AND REVIEW TECHNIQUE (PERT)

Planning and scheduling with CPM require an accurate knowledge of time

and cost for each activity because it is a deterministic approach. However,
in construction projects, most of the time, it is not possible to estimate
activity duration accurately. The variability in the estimate of the durations
may arise from work process itself or from the estimator having
insufficient data to characterize the work process exactly.

PERT was developed in the late 1950’s for the U.S. Navy’s Polaris Missile
program. While it is similar to CPM and many of the rules for network
manipulation are the same, the techniques were developed independently.
PERT allows for the estimates of an activity duration to be treated as a
probabilistic quantity. Basically, the only difference between the two
methods is that CPM uses fixed duration for each activity, while PERT
uses a probability distribution.

PERT introduces uncertainty into the time estimates for activity and
project durations. It is therefore well suited for those situations where there
is either insufficient background information to specify accurately utility
data or where project activates require research and development.

PERT uses an activity duration called “the expected mean time (te)”,
together with an associated measure of uncertainty of this activity duration.
This uncertainty can be expressed either as the “standard deviation (σte)”
or the “variance (vte)” of the duration. In order to find these values, the
analyst has to specify three time estimates for the activity duration.

1) The optimistic time (to)= is the time estimate where a particular

activity may be completed if everything goes well and exceptionally
good luck is experienced. However, this is not the crash time.
2) The most likely time (tm)= is the time estimate in which a particular
activity can most often be completed under normal conditions.
3) The pessimistic time (tp)= is the time estimate in which particular
activity may be completed under an adverse situation, such a having
unusual and unforeseen complications. This time estimate should
 Page |2

not take into account major hazards such as flood, fire, earthquake,
etc.

In analysis with PERT, main assumptions is that activity durations can be

modeled by Normal Distribution, which is unimodal and can cover both
symmetric and skewed cases.

Peak or mode at most likely time

Expected mean time (divides area

under curve into half approx.)
Probability

Optimistic
time
Pessimistic
time

Time
to t m te tp
0
Activity completion with in this
range is anticipated

Start of activity

The expected mean time is derived from the following equation.

t0 +4tm +tp
 te =
6
ASSUMPTION: Mean duration is taken as the arithmetic mean of the mid-
point between t 0 and t p and the most likely time (tm), weighed as double.
t𝑜 +tp
+2tm t𝑜 +4tm +tp
 te = 2
=
3 6
The standard deviation (the statistical measure of uncertainty being the
spread of the distribution curve about the mean value) could be written as
follows:
tp −to
 σte=
6
The variance is defined as the square of the standard deviation and is
expressed by the following equation:
tp −to 2
 vte= (σte)2=( )
6
 Page |3

99% of area
68,3% of area

95,5% of area

-3 -2 -1 xത +1 +2 +3

With these parameters, an activity duration is characterized with its t e and

σte. There is a 99% chance that the duration will stay in (to,tp) range, and
there’s 68,3% chance that duration lies in te + ̅ σte while there’s 95,5%
̅ 2σte ranges.
chance that it lies in te+

With this curve, the probability of meeting the desired scheduled time (TS)
is obtained by determining the percentage of the area cut off by this time
from the total area beneath the normal distribution curve.

probability
peak or mode at expected mean
event time

normal distribution curve

σTE σTE

time
TS TE

shaded area
Probability of TS = [ ]%
area under curve

Instead of plotting a normal distribution curve each time, the practical

approach is to use standard probability tables prepared for normal
distribution functions. To use this approach, the difference between
scheduled and expected mean times for this event is divided by the standard
deviation to give a factor Z.
TS −TE
Z=
σTE
 Page |4

This factor Z gives directly from the table, the probability of meeting TS.

An equivalent form of the same equation, enables the scheduled time of an

event to be determined, based on given risk level:

TS = TE + ZσTE

The probability of completing an activity by any specific time TS, can be

determined by using the Z-Table.

TS −TE
 Z=
σTE
where,
Z=number of standard deviations from the mean
TE=critical path(project) mean
σTE=critical path(project) standard deviation.
Ts=any date you choose

By adopting activity expected mean times, the critical path calculations

proceed as before. The timing of events thus computed will be expected
mean event time, and consequently subject to doubt. The measure of this
uncertainty requires the derivation of event standard deviations.

Once the expected mean time (TE) for an event and its standard deviation
(σTE) are determined, it is possible to calculate from probability theory the
chances of meeting a specific event scheduled time (TS). To achieve this,
the event completion time is considered to have a normal probability
distribution with the mean value (TE) and a standard deviation (σTE) as
illustrated in the above graph.

PRINCIPLES:
1) Calculate EEET (Earliest Expected Event times) as in CPM (forward
pass)
2) Add variances along the forward pass
3) Calculate LEET (Latest Expected Event times) as in CPM
(backward pass)
 Page |5

4) Add variances along the backward pass

VL=at event 1=VE at event 7
VE=measure of uncertainty up to that event.
VL=measure of uncertainty after that event.
5) Difference between EEET and LEET is called event slack (total float
in CPM). An event with zero slack is on the expected critical path.
6) VL + VE on the critical path must always be equal to VE of the last
event

Central Limit Theorem implies that; distribution for T (project time) is

approximately normal with E(T) and VAR(T).

Therefore, unlike CPM, the total project time is not a single figure but
expressed with a NORMAL DISTRIBUTION. This means that much more
information about the project time can be obtained. For example, given a
target date T*, probability that the project time will be greater than T* can
be calculated. Probability that T is less/greater than a certain value T* can
be evaluated from standard normal distribution function tables. (N(0,1))

T∗ −E(T)
P(T<T*)=P(z < )
√VAR(T)
 Page |6

Z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0.0 0.0000 0.0040 0.0080 0.0120 0.0160 0.0199 0.0239 0.0279 0.0319 0.0359

0.1 0.0398 0.0438 0.0478 0.0517 0.0557 0.0596 0.0636 0.0675 0.0714 0.0753

0.2 0.0793 0.0832 0.0871 0.0910 0.0948 0.0987 0.1026 0.1064 0.1103 0.1141

0.3 0.1179 0.1217 0.1255 0.1293 0.1331 0.1368 0.1406 0.1443 0.1480 0.1517

0.4 0.1554 0.1591 0.1628 0.1664 0.1700 0.1736 0.1772 0.1808 0.1844 0.1879

0.5 0.1915 0.1950 0.1985 0.2019 0.2054 0.2088 0.2123 0.2157 0.2190 0.2224

0.6 0.2257 0.2291 0.2324 0.2357 0.2389 0.2422 0.2454 0.2486 0.2517 0.2549

0.7 0.2580 0.2611 0.2642 0.2673 0.2704 0.2734 0.2764 0.2794 0.2823 0.2852

0.8 0.2881 0.2910 0.2939 0.2967 0.2995 0.3023 0.3051 0.3078 0.3106 0.3133

0.9 0.3159 0.3186 0.3212 0.3238 0.3264 0.3289 0.3315 0.3340 0.3365 0.3389

1.0 0.3413 0.3438 0.3461 0.3485 0.3508 0.3531 0.3554 0.3577 0.3599 0.3621

1.1 0.3643 0.3665 0.3686 0.3708 0.3729 0.3749 0.3770 0.3790 0.3810 0.3830

1.2 0.3849 0.3869 0.3888 0.3907 0.3925 0.3944 0.3962 0.3980 0.3997 0.4015

1.3 0.4032 0.4049 0.4066 0.4082 0.4099 0.4115 0.4131 0.4147 0.4162 0.4177

1.4 0.4192 0.4207 0.4222 0.4236 0.4251 0.4265 0.4279 0.4292 0.4306 0.4319

1.5 0.4332 0.4345 0.4357 0.4370 0.4382 0.4394 0.4406 0.4418 0.4429 0.4441

1.6 0.4452 0.4463 0.4474 0.4484 0.4495 0.4505 0.4515 0.4525 0.4535 0.4545

1.7 0.4554 0.4564 0.4573 0.4582 0.4591 0.4599 0.4608 0.4616 0.4625 0.4633

1.8 0.4641 0.4649 0.4656 0.4664 0.4671 0.4678 0.4686 0.4693 0.4699 0.4706

1.9 0.4713 0.4719 0.4726 0.4732 0.4738 0.4744 0.4750 0.4756 0.4761 0.4767

2.0 0.4772 0.4778 0.4783 0.4788 0.4793 0.4798 0.4803 0.4808 0.4812 0.4817

2.1 0.4821 0.4826 0.4830 0.4834 0.4838 0.4842 0.4846 0.4850 0.4854 0.4857

2.2 0.4861 0.4864 0.4868 0.4871 0.4875 0.4878 0.4881 0.4884 0.4887 0.4890

2.3 0.4893 0.4896 0.4898 0.4901 0.4904 0.4906 0.4909 0.4911 0.4913 0.4916

2.4 0.4918 0.4920 0.4922 0.4925 0.4927 0.4929 0.4931 0.4932 0.4934 0.4936

2.5 0.4938 0.4940 0.4941 0.4943 0.4945 0.4946 0.4948 0.4949 0.4951 0.4952
 Page |7

SHORTCOMINGS OF PERT

1) How accurate can be to, tp, tm ?

In construction projects usually there is no enough data related with
activity durations. Therefore, accuracy depends on the ability of the
estimator.

2) It is based on many assumptions.

e.g. activity times, in fact, cannot be modeled by continuous functions all
the time. Different probability distributions may be more appropriate.
(Simulation technique may solve this problem)

3) Critical path may change, then, joint probabilities must be used.

Generally PERT is not commonly used in construction operations because
probabilistic approaches are not popular. However, a construction-
planning problem is a probabilistic problem although analytically it may
not be treated as such. On the other hand, many writer criticize PERT
because it has computational difficulties and at the end, obtained results
don’t worth all those time consuming computations
 Page |8

Example-Question

DURATIONS (DAYS)
ACTIVITY OPTIMISTIC (to) MOST LIKELY(tm) PESSIMISTIC (tp)
A 2 4 7
B 5 8 14
C 4 6 8
D 2 2 2
E 7 10 21

A B C D E

 Compute the following nine values:

The probability that the project will finish by the end of day 32.
The probability that the project will finish by the end of day 34.
The probability that the project will finish before day 30.
The probability that the project will finish on the 32nd day.
The probability that the project will finish no later than the 35th day.
The probability that the project will finish at least 2 days early.
The probability that the project will finish at least 2 days late.
The probability that the project will finish on the 32nd day±1 day

 Calculate te, ơet, and Vet for individual activities and TE, VE for
entire path.
 Find ơE =√8.833=2.972 days

DURATIONS (DAYS) EXPECTED STANDARD VARIANCE

ACTIVITY OPTIMISTIC MOST PESSIMISTIC DURATION DEVIATION (Vte = ơte2)

(to) LIKELY(tm) (tp) (te) (ơte)

A 2 4 7 4.167 0.833 0.694

B 5 8 14 8.500 1.500 2.250
C 4 6 8 6.000 0.667 0.444
D 2 2 2 2.000 0 0
E 7 10 21 11.333 2.333 5.444
TE =32 VE
=8.833
 Page |9

The probability that the project will finish by the end of day 32.
o Z= ( TS－TE)/ơE = (32-32)/2.972=0
o From Z table, we find that Pr (TS≦32)=0.5=50%

28 30 32 34 36

The probability that the project will finish by the end of day 34.
o Z= ( TS－TE)/ơE = (34-32)/2.972=0.67
o From Z table, we find that Pr (TS≦34)=0.749=74.9%

28 30 32 34 36

The probability that the project will finish before day 30.
o This probability is the same as the probability of finishing by
the end of 29 days.
o Z= ( TS－TE)/ơE = (29-32)/2.972=-1.01
o When Z＜0, take the probability that corresponds to the
positive value of Z, then subtract it from 1.0.
o Pr (TS≦29)=1-0.844=15.6%

28 30 32 34 36
 P a g e | 10

The probability that the project will finish on the 32nd day.
o Pr (TS＝32)= Pr (TS≦32)－Pr (TS≦31)
o Pr (TS≦32)＝0.5＝50%
o Pr (TS≦31),
o Z= ( TS－TE)/ơE = (31-32)/2.972=-0.34
o Pr (TS≦31)=1-0.633=0.367=36.7%
o PR (TS＝32) = 50%－36.7% = 13.3%

28 30 32 34 36

The probability that the project will finish no later than the 35th
day.
o This probability is the same as the probability that the
project will finish by the end of day 35:
o Z= ( TS－TE)/ơE = (35-32)/2.972=1.01
o Pr (TS≦35)=0.844=84.4%

28 30 32 34 36

The probability that the project will finish at least 2 days early.
o This probability is the same as the probability of finishing on
the 30th day or earlier, or the same as the probability that the
project will finish by the end of day 30:
o Z= ( TS－TE)/ơE = (30-32)/2.972=-0.67
o Pr (TS≦30)=1-0.749=0.251=25.1%
 P a g e | 11

28 30 32 34 36

The probability that the project will finish at least 2 days late.
o This probability is the same as the probability of finishing by
the 34th day or later, or the same as the probability that the
project will finish after day 33:
o Pr (TS＞32)=1－Pr (TS≦33)
o For Pr (TS≦33), Z= (33-32)/2.972=0.34
o Pr (TS≦33)=0.633=63.3%
o Pr (TS＞32)=1－Pr (TS≦33)
=1-0.633=36.7%

28 30 32 34 36

The probability that the project will finish on the 32nd day±1
day
o This means finishing on day 31, 32 or 33.
o Pr (TS＝31, 32, 33)= Pr (30＜TS≦33)
= Pr (TS≦33)－Pr (TS≦30)
o Pr (TS＝31, 32, 33)= 0.633-0.251=0.382=38.2%

28 30 32 34 36
 P a g e | 12

Example:

ACTIVITY DESCRIPTION to tm tp te σte vte

1-2 A 8 8 8 8 0 0
1-3 C 7 9 11 9 0.67 0.45
1-4 B 8 10 24 12 2.67 7.13
2-4 Dummy 0 0 0 0 0 0
2-5 D 3 5 13 6 1.67 2.79
3-6 E 12 13 32 16 3.33 11.09
4-5 F 6 7 8 7 0.33 0.11
4-7 G 10 11 18 12 1.33 1.77
4-6 H 8 10 36 14 4.67 21.81
5-7 I 8 8 20 10 2.00 4.00
6-7 K 8 8 8 8 0 0
Given Calculate by using
formulas

t +4tm +tp
te= o
6

t −to Calculate by
σte= p using these
6
formulas
vte= (σte)2

1) Calculate te, σte, and vte from above formulas

8+4∗8+8
e.g. A te= =8
6
0
σte= =0
6
vte=0
When to=tp=tm  t is deterministic ! (fixed duration)

7+4∗9+11
C te= =9 days
6
11−7
σte= =0.67 day
6
vte=(0.67)2=0.45 day2
 P a g e | 13

2) Construct the network just like in CPM

Early Late
Time EEET LEET EEET= early expected event time
Variance Var.E Var.L LEET= late expected event time

1 2
te,vte
EEET LEET EEET LEET
Var.E Var.L Var.E Var.L

Network:

8 12 19 24
0 21.81 D 7.24 4
2 6,2.79 5
A I
F

0 0 B G 34 34
0 28.94 1 4 7 28.94 0
12,7.13 12,1.77
12 12
7.13 21.81 H K
C

E
3 16,11.09 6

9 10 26 26
0.45 11.09 28.94 0

3) Calculate EEET (earliest expected event times as in CPM)

4) Add variances along the forward pass
5) Calculate LEET (latest expected event times as in CPM)
6) Add variances along the backward pass
 VL at event 1=VE at event 7
VE=measure of uncertainty up to that event
VL=measure of uncertainty after that event
 P a g e | 14

7) Difference between EEET and LEET is called event slack. (Total

float in CPM). An event with zero slack is on the expected critical
path.

The project has a total expected duration of 34 days, and a variance of

28.94.(Normal distribution with N(34,5.38)) (σ=√28.94=5.38 days)

This information is used to define a probability distribution curve for the

project duration. As explained before, this is a normal distribution.

It is therefore possible to determine the probabilities of meeting various

target dates from the table (z table).

Suppose Target Date= 30 days,34 days, 35 days

30−34
P(T>30)=1-P(z < )=1-z(-0.74)=1-0.23=0.77
5.38

P(T>34)=1-P(z < 0)=0.5

P(T>35)=1-P(z < 0.37)=1-0.64=0.36

Exp. Time Stand. Target Probability of Meeting

Event # z
(TE) Dev. (σte) Date (TS) Target TS (%)
7 34 5.38 30 - 0.74 23%
32 - 0.37 35%
34 0.0 50%
36 + 0.37 64%
38 + 0.74 77%
 P a g e | 15

Example:

1 A 2 B 4 D 5 F 6 G 7

̅ 1 σ, +
Calculate the times for event 7 that correspond to + ̅ 2 σ, and +
̅3 σ
from the mean value of event 7’s time

to tm tp te σte vte
A 3 4 6 4.17 0.5 0.25
B 4 6 7 5.83 0.5 0.25
C 3 3 4 3.17 0.17 0.03
D 5 6 7 6 0.33 0.11
E 8 10 12 10 0.67 0.44
F 11 12 14 12.17 0.5 0.25
G 3 3 3 3 0 0
Estimated times Calculate by using
formulas
 P a g e | 16

Solution:

7.34 10
0.28 0.36
3

C
0 0 4.17 4.17 3.17,0.03
16 16 28.17 28.17 31.17 31.17
0 0.94 0.25 0.69 0.69 0.25 0.94 0 0.94 0
1 A 2 B 4 D 5 F 6 G 7
4.17,0.25 5.83,0.25 6,0.11 12.17,0.25 3,0

10 10
0.5 0.36
E
10,0.44

E(T)=31.17
V=0.86 σ=0.93

̅ 0.93%68.3
31.17+
Ranges ̅ 1.85%95.5
31.17+
̅ 2.79%99
31.17+