CHAPTER 4
PERT AND ACTIVITY TIME ESTIMATES
The second technique under the network analysis model for managing project is the PERT. The
major distinction between PERT and CPM is that while CPM uses only one time estimate for
each activity, PERT makes use of 3 times estimate for each activity. The methodology
presupposes that it is not possible (realistic) to estimate the exact time element to perform an
activity with 100% accuracy. Thus, it uses a set of assumptions and together with statistical
analysis, to determine the most appropriate time for the execution of project activities.
Though PERT and CPM differs slightly in their terminology, the objectives are
basically the same. Furthermore, their project analysis technique is very similar. Probabilities are
attached to each of the times which in turn are used for completing expected values and potential
variation for project activity time.
The 3-time estimate specify for each activity in PERT are:
1.Optimistic time (to) : It is the shortest time taken to complete an activity. It means that if
everything goes well then there is more chance of completing within this time.
2. pessimistic time(tp) : It is the longest time an activity will take to be completed . It is the
worst time an activity would take if on unexpected problems are faced.
3. Most probable or most likely time (tm ) – this is the best estimate of what should
normally occur. The most likely time is the time in which the probability for computing project task is
high
The difference between the 3-time elements tells the relative uncertainty involved in timing
of project activities. The expected time (te) adopted under the PERT analysis can be calculated using the
following formular.
tp+ 4 tm+¿
Mean (expected time) (te) = ( )
6
Where;tp = pessimistic time
to = optimistic time
tm = most likely time.
te = expected time to perform project activities.
The multiple time estimated approach is better because it provides and improves estimate of the
expected time to complete Project activity. A basic reason for estimating more than one time
element is to provide management with information that may be used to calculate the probability
that a particular activity or the entire project may be completed within a particular date.
Projects are most likely to suffer due to variation in time estimate. Variation is
usually possible when time element is not well estimated. The two most important statistical
properties for determining the time dispersion (variation) are the variance and the Standard
Deviation. The variance for project activity (V) can be estimated as follows.
tp−¿ 2
Variance (V) =( )
6
Similarly, the standard deviation (S) can be calculated as:
tp−¿
S = √(
6
)2 = √V
Where v= variance for project task and S= Standard Deviation for project task.
The variance for the entire project also known as the variance of the critical path can be
estimated as follows.
tP−¿ 2 tP−¿ 2 tp−¿ 2
V=( )A +( ) B ----------------- + ( ) N
6 6 6
Where: A,B ---------------N are activities along the critical path.
N =last activity along the critical path.
S= (√ tP−¿
6 ) √ 6 √ 6 )… … N
A+ (
tP−¿
) B + (
tP−¿
When using PERT. the expected time (te) is calculated first from the 3 values of activity time.
The values of these expected time are then use exactly as in CPM. Also, this expected time can
help management undertake complete project analysis by calculating the EST, EFT, LST. LFT
and activities slack, the critical path etc.
Application 1:
The table below shows the activities and activity relationship relating to a project design to
extend PAMOL plantation in the South West Region.
Activities Preceding act. Optimistic Most likely Pessimistic
time (to) time (tm) time (tp)
A - 6 6 24
B - 6 12 18
C - 12 12 30
D A 6 6 6
E B 12 30 46
F C 12 30 42
G D,E 12 30 54
Required;
(1) Determine the expected duration (te) and variance (V) for each project activity
(2) What is the expected length of the project. (Critical path)
(3) Estimate the variance and standard deviation of the entire project (critical path).
PROJECT PROBABILITY ANALYSIS
Once the expected completion time and variance of project of activities are known, it is possible
to determine the probability that such activities may be completed within a specific or range of
dates. The assumption is usually made the distribution of activities and completion dates follows
that of a standard normal distribution. An activity is said to be normally distributed, if it has a
mean value of zero (0) and a variance of 1 i.e X N(0,1).
Any activity whose mean and variance are not equal to 0 and 1 respectively, are not
normally distributed. However, if they are standardized, they can be approximated to follow the
normal distribution.
Standardization of project Activities.
In standardizing project activities, we seek to know how much the activity deviate from the
mean . Deviation is said to be normal if they do not disperse very far from the mean. This is
usually achieved by subtracting the mean from a particular value of the activity and dividing the
result by the standard deviation as follows.
Suppose X N ( μ , δ 2)
x−μ
Then Z = ( ) N(0,1)
δ
Where Z = Standard value
X = project activity
μ = The mean value
δ 2 = Standard deviation
δ = Variance.
After standardization, the standard normal distribution table can be used to read the probability
for any activity that follows a normal distribution. Once an activity has been standardized, its
distribution becomes symmetrical about the mean.
