W4_Juhaina_Estimation of Duration
to Tender Site Studies Contracts Using PERT
1- Problem identification:
As part of the Request
for Proposals Package (RFP) related to power and water projects, it is an
essential requirement to include site studies related to the allocated site in
order to provide the bidders with enough level of information on the site, and
therefore reduce the site risks and related costs. OPWP usually hires third
party contractors to perform these studies prior to releasing the RFP through a
competitive tender process.
As is it is essential
to issue the RFP as schedules, it is important to understand the most expected duration
it takes to tender and hire a contractor to perform the site studies. This blog
aims to estimate the most expected duration.
2- Feasible Alternatives
The current project schedules
allow for a standard 90 day period to hire a contractor, however the PERT analysis
(Program Evaluation and Review Technique) can be used to identify the most
suitable period which is the most common method used in three-point estimation
technique.
3- Development of the outcome of the alternatives
Using
the PERT analysis on the historical data, we can provide the alternatives of
the duration as per the following:
1.
Best Case
2.
Worst Case
3.
Most Likely
4- Selection criteria
The selection criteria
will be based on a 95% comfort level to achieve the process within the duration
estimated.
5- Analysis and Comparison of the alternatives
The historical data of
the tender duration for the most recent 10 projects has been collected and
presented in the following table:
#
|
Project
|
Tender Duration
(days)
|
1
|
Barka IWP
|
72
|
2
|
Sohar IWP
|
89
|
3
|
Salalah IWP
|
75
|
4
|
Sharqiyah IWP
|
75
|
5
|
Khasab IWP
|
71
|
6
|
Duqm IWP
|
71
|
7
|
Sohar IPP
|
93
|
8
|
Ibri IPP
|
105
|
9
|
Salalah 2 IPP
|
84
|
10
|
Mussandam IPP
|
159
|
Using the statistical Process Control, we first eliminate any
outliers from the data using -/+ 3 sigma. The average is 89.4 days (by
calculating the sum of total days by 10). Sigma here is calculated as (highest
number of days – lowest number of days)/6 and the result is 14.66. Therefore
the following is calculated:
-
Upper Control Limit is 89.4 + 14.66 = 104.06
-
Lower Limit is 89.4 – 14.66 = 74.74
The data is
accordingly modified and only the following data set is considered:
#
|
Project
|
Tender Duration
(days)
|
2
|
Sohar IWP
|
89
|
3
|
Salalah IWP
|
75
|
4
|
Sharqiyah IWP
|
75
|
7
|
Sohar IPP
|
93
|
9
|
Salalah 2 IPP
|
84
|
According to the data
above, the following table provides the most probable scenarios:
Most Optimistic (O)
|
Most Likely (M)
|
Most Pessimistic (P)
|
75
|
83.2
|
93
|
Using the Pert Formula below, we can calculate the weighted
average:
(O + 4M + P)/6
= (75 + (4*83.2) + 93)/6
= 83.5 days
-
Mean Duration = 83.5 days
-
Standard Deviation = (P – O)/6 = (93 – 75)/6 = 18/6 = 3
-
Variance = (3)2 = 9 days variance
Using the Z table, the P95 is calculated as
follows:
-
P95 = Mean + (Standard Deviation * 1.65) = 83.5 + (3*1.65) =
88.45 days
6- Selection of the preferred alternative
From the pert analysis
demonstrated above, it is shown that for a 95% comfort level, the tenders are expected
to be achieved within 88.45 days. This is very close to the standard set in the
schedules and this show that our standard is based on a good estimation. It is
recommended to adjust the timeframe to be 88.45 days accordingly.
7- Performance monitoring and the post evaluation
results
In order to further
verify the above, it is suggested to look into the different activities within
the process and identify the critical path and possibly replan the schedule to
make it more time efficient.
8- References
1-
Use PERT Technique for More Accurate Estimates,
by Tom Mochal, June 25, 2007 (visited on 30 November 2017)
2- What is a PERT Chart? By Eileen
O’Loughlin (Visited on 30 November 2017)
3- PERT Estimation Technique, (Visited on 30
November 2017)
AWESOME case study BUT....... how SUCCESSFUL were these projects in terms of TIME or COST? In other words IF these projects were NOT "successful" then is this a valid basis to do your analysis?
ReplyDeleteJust a suggestion but why not find projects that were SUCCESSFUL and use only that data, but first normalize it based on some parameter like duration per KWH or duration per MLiters of water?
My concern is if you don't first validate your INPUT data then you run the risk of "Garbage In/Garbage Out" effect.
I would urge you to do another posting but this time look at the relative "success" or "failure" of the project, the theory being that there is a correlation between how long you allow and how "successful" the project is? Maybe the answer is no but I would sure recommend you at least explore that avenue just to see?
BR,
Dr. PDG, Jakarta
Usually I never comment on blogs but your article is so convincingthat I never stop myself to say something about it.
ReplyDeletebest telemedicine platforms