Project Cost Estimation

Project Cost Estimation:

 It can be described in various ways:

 It is “an evaluation of all the costs of the elements

of a project or effort as defined by an agreed-upon
scope.” R.E. Westney (1997)
 It is an assessment of the final cost of a project,
program, product, or process, based on specific facts
and assumptions.
 It is the process of developing an approximation
(estimate) of the costs of the resources needed to
complete project activities.

2
 Project cost estimation is both an art and a
science as it combines the use of the skills and
experience of individuals estimating costs and the
application of scientific methods such as
mathematical models or computer tools.

 Cost estimating involves developing an

approximation or estimate of the costs of all the
resources needed to complete a project – it implies
the necessity of resource planning.

3
 Cost estimates are used to:
1) Determine whether the project is worth doing.

 They are the chief means for assessing

project feasibility.

 A comparison of the cost estimates with the

estimates of revenues will enable the project
organization to determine if the project is
worthwhile to undertake.

4
 2. They help plan project financing/ funding:
 Along with project returns, they facilitate
decisions relating to project financing and
funding.
 3. They provide the mechanism for
managing cash flow during the course of
the project.
 4. They give the project manager a
framework for allocating scarce resources
as the project progresses.

5
 5. Develop time-phased budgets and establish the
project baseline, which is the basis for project
evaluation and control.

 Without reasonable cost estimation, project

budgets are essentially useless, and without
accurate budgeting, cost estimation is a
wasted exercise.
 Cost estimation and project budgeting are
inextricably linked

6
 It is a common knowledge that, to a
great extent, project success hinges on
accurate cost estimation.
 However, Cost overrun is typical of almost
all projects
 What are the causes or reasons for cost
overrun?

7
 Among the reasons leading to cost overrun
are:
 Low initial cost estimates
 Unanticipated technical difficulties
 Lack of or poor scope definition
 Specification changes
 Overlooking key project costs
 External factors such as inflation, political
instability, foreign exchange change. ---

8
 Accuracy of Cost Estimates Depends on :
1. The Availability of Information
 how much is known about the project, clarity of project
scope and quality of known information.
 Estimates become progressively more accurate as we
progress through the project life cycle (more information
and clarity comes as project progresses) – figure in the next
slide
2. The Amount of Time Available for Estimation.
 estimating accuracy increases (inaccuracy decreases) over
time (project phase), mainly because of increasing availability,
accuracy, and specificity of information.

9
3. Available Resources
 the number and qualifications of personnel deployed
 the type and number of equipment mobilized
 the amount of money available, and
 the sources of estimating data

10
High
Estimates become
progressively more
accurate as we progress
through
the project life cycle.

Low

11
A piece of advice:
 Since the initial cost estimation begins
during the project proposal
development stage, maximum effort
should be made to identify “all” relevant
costs that are likely to be incurred in the
project and include in the initial project
proposal.

12
 What are the major sources of project costs?
(1)Cost of labor: This involves hiring costs, salaries
and wages for the various human resources
associated with the project.
 Relatively complex to estimate: includes both
skilled and unskilled or different grades of labor.

• Should include related costs such as pension and

health benefits to project employees.

13
(2)Cost of materials: This is the cost of raw
materials, supplies, and parts needed to
complete project tasks.
 It varies from project to project: For instance,
material costs are significant in construction
projects whereas they are very small in software
development projects.

14
(3) Cost of equipment and facilities: includes
acquisition of major fixed assets including rental
charges for these items for “off-site” projects like
mining and construction projects.

(4) Other Sources: Other relevant sources of project

costs can include travel costs of team members, as
well as costs associated with subcontractors or
consultants.

15
 Project costs can be classified in different
ways.
◦ It helps us better understand the nature of project costs.
 Project costs can be classified as:
1. Direct or indirect
 Direct costs: directly traceable to the /charged
against/the project that can not be shared by other
projects: e.g. materials, labor and other direct costs.
 Indirect costs: costs of resources shared by
projects: allocated overhead costs such as taxes,
insurance & utilities, and selling and general
administrative costs.
16
2. Recurring or nonrecurring
 Recurring costs are repetitive costs , e.g.
labor and materials, are repeatedly incurred
throughout the project life cycle.

 Nonrecurring costs are one time costs,

e.g. marketing research costs occur at
beginning of the project and employees’ training
costs at the end of the project).

17
3. Fixed or variable
 Fixed costs: costs that do not change with
usage or throughout the project, e.g. the
cost of rented piece of equipment, cost of
hiring consultants, and so on.
 Variable costs: costs that vary directly
with usage or project activity level, e.g.
costs of material, labor or services needed
for the project.
18
4. Normal or expedited:
 Normal costs are incurred when project tasks are
performed /completed according to the original
planned duration.
 Expedited costs or crash costs are unplanned
costs incurred as a result of steps taken to
accelerate project completion, e.g. additional
overtime or costs of hiring additional workers to
speedup project completion.
 Note that many of these costs may fall in multiple
classifications, e.g. labor costs (D, R, V & N/E).

19
 One key approach is to cost out the project on a
disaggregated basis:
◦ First break the project down by deliverables and work
packages as a method for estimating task- level costs.
Next, these estimates can then be aggregated into an
overall project budget.
Cost Aggregation

Individual Project
Work
Project Deliverables Tasks: Budget
packages
Estimating
task-level
costs
Disaggregating
the Project

20
Disaggregating project activities to create
reasonable cost estimates:
Deliverable 1040—Site preparation
• Work Package (WP) 1041—Surveying …..Br 3,000
• WP 1042— Utility line installation ……….. 15,000
• WP 1043— Site clearing ……………….. 8,000
• WP 1044— Debris removal ………………. 3,500
Total cost for Deliverable 1040 ..…. Br 29,500
 It is typically more accurate to first identify the costs for
completing each work package individually and then create
a deliverable cost estimate and that of the project.

21
1. Ballpark estimates, also known as Rough order
of magnitude (ROM) Estimate
 ROM estimate provides a rough idea of what a project
will cost
 Often used when there is no sufficient information or
time available to develop more accurate or detailed
estimates
 It is used to make initial rough-cut estimates for project
selection decisions or making competitive bids for project
contracts.
 Estimates made without any detailed engineering
data.
22
2. Feasibility estimates—These estimates are
developed after preliminary project design work is
completed.
 Request for quotes (RFQs) are sent to suppliers and
subcontractors based on the initial project design.

 This kind of estimates are sometimes called

budgetary estimates.

23
3. Definitive estimates—These estimates can be
developed only after the completion of most design work.
 They are made where:
 the scope of the work is clearly known,
 no significant changes in specification are expected,
 a comprehensive project plan is in place,
 all activities and their sequences are identified,
 all major purchase orders are submitted based on known
prices and availability of materials and equipment.

24
There are four methods of cost estimation:
(1)Top-down or Analogous Approach
 Using the actual cost of a previous, similar project as the
basis for estimating the cost of the current project.
 Those historical costs are sometimes adjusted for known
differences in project complexity.
 Combines historical data and expert judgment.
 It is generally less costly but less accurate than other
methods.
 The worst it may generate a ballpark type of estimate.
 However, it can be the most reliable when:
a) the previous projects are similar in fact and not just in
appearance,
b) the individuals or groups preparing estimates have the
needed expertise.

25
(2) Bottom-up Approach
 Involving estimating individual work items costs and
summing them to get a project total.
 It is more accurate , a kind of definitive estimate.
 It is usually time-intensive (consuming) and
therefore expensive to use.
 The cost and accuracy is driven by the size and
complexity of the individual activity or work
package: smaller items increase both cost and
accuracy of the estimating process.
 The project management team must weigh the
additional accuracy against the additional cost.

26
(3) Parametric modeling
 Uses project characteristics (parameters) in a
mathematical model to predict project costs.
o Parametric estimation consists of two steps:
a) Identifying the features/parameters of an older or
well-known project that can be directly related to
the development cost of the current project; and
b) Defining the mathematical nature of that
relationship- develop a mathematical model.

27
o Parametric models may be simple or complex.
o Simple example: Model the cost of constructing a
residential home based on square footage of living space.
◦ Complex example: Model software development costs
using thirteen adjustment factors, each of which has five to
seven points.
o It may generate a budgetary type of estimate.
o Moreover, it can be the most reliable tool when:
1) the historical information used to develop the model was
accurate,
2) the parameters used in the model are readily quantifiable,
and
3) the model is scalable (expandable to accommodate more
variable as a need may arise).
28
The parametric method can be used in a variety
of project scenarios, including:
 Estimating costs before a full-scale design
 Quickly choosing from alternate design schemes to
initiate a project with little time for detail design and
cost estimation.
 Generating bids when time is limited to develop a full
cost proposal
 Setting design-to-cost targets (as in Target Costing)
 Evaluating the cost of changes to existing products
 Assessing the effects of new technology on product
development and pricing

29
 AMEC Inc., a major firm in the industrial construction
industry, was interested in developing a quick cost
estimate for the construction of a warehousing facility.
 In the recent past, the company completed
construction of ten similar warehousing projects with
similar general architecture, layout, and construction
materials.
 AMEC engineers analyzed the relationship between
material costs and several building parameters, and
developed a multiple regression model using the least
squares method.

30
 Based on existing historical data, AMEC understood that
overall material costs for a new warehouse were directly
related, in a linear manner, to the total floor space and
number of loading docks planned for the facility.
 Then, they developed, a multiple regression model
that relate material costs (Y) to:
 floor space (X1, in multiples of 10,000 square feet) and
 number of shipping docks (X2) in a facility as given below. That is,

Y = 315.41 + 47,319 X1 + 19,638 X2

31
 The regression model developed based on the data
from the previous ten projects would be a very
good predictor for the new project.

 Using the above model, therefore, engineers were able

to estimate the material costs for a new warehousing
facility with 500,000 square feet and five shipping docks
as follows:

Y = 315.41 + 47, 319 ∗ 50 + 19, 638 ∗ 5 = $2, 464, 455.41

32
(4) Computerized tools
o Computer based tools, such as spreadsheets
and project management software can be
employed to make working with different cost
estimates and cost estimation models easier.

33
Advantages of computerized project planning:
 Each project activity is examined in detail:
• resource required and associated tasks are identified
• more accurate cost estimates can be generated
• ratios and historical data can be used to improve
estimates accuracy further
 Many project planning software packages can generate
a time-phased cost schedule (budget) that can be
compared with actual costs for evaluating cost
performance.
 Enhances project control by linking the project work
plan with its cost plan.

34
Computerized methods have two important
shortcomings:
 First, a meaningful cost plan cannot be
developed unless a project plan with
considerable detail is available.

 Second, as the estimates of resources and

associated skills required for activities are often
based on subjective judgments of estimators,
the potential for inaccuracy is still high.

35
 The choice of project cost estimation approach by
a project organization depends on a number of
factors:
1) the knowledge of the firm’s industry (e.g., software
development versus construction),
2) its ability to account for and manage most project
cost variables,
3) its history of successful project management,
4) the number of similar projects completed in the
past,
5) the knowledge and resourcefulness of project
managers, and
6) the company’s budgeting requirements.

36
 Justification for Contingencies:
 When predicting future costs, there is always the
potential for error, and cost overruns.
 Therefore, additional allowances are needed to
ensure that the funds actually allocated to the project
will be sufficient.
 This allowance is known as Contingencies.
 We must know that:
 the greater the degree of uncertainty involved, the larger
the amount of contingency required, and
sometimes, significant problems in a project can cause
overspending of the contingency amount itself.

37
 Someof the benefits of including the contingency
element in the cost estimation process are:
First, the approach recognizes that the future is
uncertain, and that unexpected problems can increase
overall project costs.
Second, contingency estimates make explicit provisions
for potential and unexpected cost escalations in project
plans and budgets.
Third, the availability of a reasonable amount of contingency
funds enhances the viability and value of the cost
estimation process as a control tool.
Finally, the use of contingency funds provides an early
warning signal of potential cost overruns.

38
 Experience curves are based on the old adage that
practice makes perfect.
 Unlike the conventional view that work is done at a
constant rate, the learning curve view holds that
when a task involves a repetitive activity the
amount of time and cost necessary to complete the
activity successively declines.
 A product can always be manufactured better and
in a shorter time period not only the second time,
but each succeeding time.

39
 This concept is highly applicable to labor-
intensive projects, such as those in
manufacturing where labor forecasting has
been a tedious and time-consuming effort.
 In 1960, Personnel from the Boston
Consulting Group showed that each time
cumulative production doubled, the total
manufacturing time and cost fell by a
constant and predictable amount.

40
 Empirical studies on learning curve yielded
three conclusions on which the current theory
and practice are based:
 ● The time required to perform a task
decreases as the task is repeated.
 ● The amount of improvement decreases as
more units are produced.
 ● The rate of improvement has sufficient
consistency to allow its use as a prediction
tool.

41
42
 The labor hours per unit required to complete an
activity decreases by some fixed percentage called
learning rate each time the output doubles.
e.g., if a worker needed 100 hrs to finish the first
product and his learning rate is 85%, producing the
2nd product requires him only 85 hrs; producing 4th
unit requires 72.5 hrs. As production increased by
100%, production hrs consistently declined by 15%.
 A mathematical model is developed that can be used
to estimate the total time needed to complete a job of
repetitive activity.

43
 T n = an b
where
T n = the time required for the nth unit of output
a = the time required for the initial unit of output
n = the number of units to be produced, and
b = a factor derived from the slope of the learning curve
= expressed as log decimal learning rate/log 2
 E.g. if it takes 40 hrs to assemble the first product and it takes
only 34 hrs for the second, the learning rate is 85% or 0.85
(34/40).
 b= log0.85/log2= -0.235

44
 Slope of the curve. A percentage figure that
represents the steepness (constant rate of
improvement) of the curve.
 Using the unit curve theory, this percentage
represents the value (e.g., hours or cost) at a
doubled production quantity in relation to the
previous quantity.
 For example, with an experience curve having 80
percent slope, the value of unit two is 80 percent
of the value of unit one, the value of unit four is 80
percent of the value at unit two, the value at unit
1000 is 80 percent of the value of unit 500, and so
on.
45
 Unit one. The first unit of product actually
completed during a production run.
 Cumulative average hours. The average hours
expended per unit for all units produced
through any given unit.
 When illustrated on a graph by a line drawn
through each successive unit, the values form
a cumulative average curve.

46
 Unit hours. The total direct labor hours
expended to complete any specific unit.
 When a line is drawn on a graph through the
values for each successive unit, the values
form a unit curve.
 Cumulative total hours. The total hours
expended for all units produced through any
given unit.
 The data plotted on a graph with each point
connected by a line form a cumulative total
curve.
47
48
 Either cumulative average time (hour) or
cumulative total time information is needed
for total labor cost estimation based on
learning curve approach.
 Given the time to complete the initial unit and
the learning rate, cumulative average time
can be approximated through the following
formula.
 Note: Approximation would be more accurate
for more than 100 units cumulative
production level , i.e., when n > 100 unit
49
 T c = an b ÷ 1-/b/
 Where
 T c= cumulative average labor time for nth unit
 a = the labor time required for the initial unit of output
 n= cumulative units produced
 b= log decimal learning rate/log 2
 Once T c is known, cumulative total labor time can be
obtained by multiplying it by n and then cost per time
unit is applied to estimate total labor cost

50
 Example: A project involving the manufacture of aircraft
engines requires the production of 20 units of a complex
air-filtering device that the project organization has never
produced. One worker will be tasked to produce all 20
units of air filters. Previous studies indicate that it usually
takes 141.3 hours to complete production of the first unit
and the learning rate for workers undertaking this type of
repetitive activity is 0.85 or 85%.
 Required:

 1. Determine the time necessary to complete 20th unit

 2. Determine cumulative average production time

 3. Assuming the labor is paid a wage rate of $50/hour,

determine (estimate) total labor cost of this project.
 1.
 n=20
 a= 141.3
 b= log0.85/log2= -0.235
 T n = an b
 T 20 = 141.3* 20 -0.235
 T 20 = 70 hrs

52
 2. T c = ?
 T c = an b ÷ 1-/b/
 T c = 141.3 * 20 -0.235 ÷ (1- 0.235)
 T c = 70 ÷0 .765
 T c = 91.5 hrs
 3. Estimated total labor cost = cumulative total
labor hours * labor cost per hour
 cumulative total labor hours= cumulative average
labor hrs * cumulative production unit =91.5* 20 =
1830 hours (approximation)
 Estimated total labor cost = 1830 * 50 = $91,500

53
Thank You.

54