Earned Value Analysis

CIB W107 Construction in Developing Countries International Symposium
“Construction in Developing Countries: Procurement, Ethics and Technology ”
16 – 18 January 2008, Trinidad & Tobago, W.I..
EVA as a possible tool for effective project control in developing
countries
C. N. Masambaji1
and J. K Ssegawa2
1
Pula Consultants (Pty) Ltd, P. O. Box 2884, Gaborone, Tanzania. mchedi@yahoo.com
2
Department of Civil Engineering University of Botswana, P/Bag 0061, Gaborone,
Botswana. jssegawa@mopipi.ub.bw
Abstract
Construction projects in developing countries continue to experience delay and cost
overruns. While there are many problems associated with phenomenon, control of project
schedule and costs continues to be a significant factor. There are indications that the use
of poor control techniques throughout project implementation seem to be the root cause
of this problem. This paper reports the findings of a study which investigated the possible
use of earned value analysis (EVA) as a possible project control technique in Botswana.
A live project at construction phase was selected as case study for the exploratory
investigation. The project was a construction of a police station with its associated
ancillary facilities. Both the traditional and EVA were used simultaneously on the project
and data was collected for a period of nine months. Monthly, performance analysis was
carried out. It was found out that although, the EVA technique was involving, it was a far
superior technique in project control.
Keywords: project control, earned value analysis, project performance, public projects
Introduction
Construction projects in developing countries continue under-perform in all objectives of
functionality, cost, time and quality. The cause of this phenomenon has been a subject of
numerous studies for over decades, for example, Tah et al., (1993); Elinwa and Buba
(1993); Lewis ( ); Wachira (2000). A recent study by Ssegawa and Ngowi (2007) has also
indicated that only 8% of the projects in Botswana are completed within budget and time.
The need to reduce the risk of poor performance of projects is very crucial in developing
countries because most of the socio-economic programmes have an element of a
construction project. While many reasons have been forwarded for the poor performance
of construction projects, lack of effective control remains high on the list. This is firstly,
because project control is not a self contained activity but rather an integrated activity that
receives outputs from earlier phases of the project life-cycle processes (PMI, 2004).
Without a good quality plan, project control cannot be achieved. Secondly, project control
requires efficient, effective and vigilant processes and actions to ensure its successful
accomplishment.

2
In their illustrative paper, Rozens et al., (2006) reviewed literature relating to various
methods used to achieve project control. The major categories identified were the one-
dimensional and the multi-dimensional control systems. The former is where the control
objectives focus on one specific project dimension rather than the entire set of objectives.
These are normally simple systems to implement and understand. Even if two objectives
are used simultaneously, there is a lack of the necessary integration.
On the other hand multi-dimensional control systems integrate several project objectives
(Rozen et al., 2006) providing a multi-faceted reporting system. One of the most touted
multi-dimensional project control systems is the Earned Value Analysis (EVA). The
effectiveness of the system as a project control tool has been of much praise in project
management literature (e.g. Meredith and Mantel, 2003; Kezerner, 2005; Klastorin, 2005,
Anbari, 2003; Cleland and Ireland, 2007). However, there seems lack of any evidence
that the system has been tested or used on a construction project in a developing country
and hence the subject of discussion of this paper.
The paper therefore, presents the findings of a study conducted to investigate if EVA is a
suitable control technique for construction projects and in this in Botswana. It uses a
building construction project as an exploratory case study. The paper therefore, begins
with a brief background of the EVA system, followed by a brief description of the research
approach used to conduct the study. The third section looks at findings of the study and
ends with some conclusions, limitations and recommendation.
Project Control And Eva
A brief background of the nature of project control systems and the basic ingredients that
make up an EVA system is discussed in the next sections.
Nature of a control system
An effective project control system should provide a project manager with adequate and
timely information for decision making especially for taking corrective action to ensure
that a project stays within the planned parameters and hence meet its intended
objectives. To achieve this feat, the control system must have a quality project baseline
plan and an efficient and effective monitoring system for recording, measuring and
analysing data and reporting performance.
The EVA system.
The EVA incorporates scope, cost and time, by answering two basic questions
(Marchewka, 2003, PMI, 2004); Q1: How much work has been completed against what
should have been completed?; Q2: How much has been spent against what should have
been spent? To answer these questions the EVA is viewed as consisting of four basic
components namely baseline plan, monitoring process, performance analysis and
control, forecasting to update the baseline plan. Due to a plethora of literature on EVA
(e.g. Fleming and Koppelman, 2000; Anbari, 2003 Burke, 2003; and PMI 2004), only a
brief summary of its components is discussed next.
Development of baseline plan phase

3
To develop a baseline plan useful for an EVA systems a number of steps should be
followed namely; (i) identify the project scope by defining the major deliverables and work
packages by using a work breakdown structure (WBS); (ii) develop a project schedule by
identifying project activities, their duration, sequencing and dependencies; (iii) for each
activity, determine the methods of construction, identify the requisite resources their
quantity and estimate their costs; (iv) aggregate the cost of the activities making up a
work package, deliverables and project using the cost break down structure (CBS) and
associated cost codes to provide accurate and efficient data collection; (v) identify
periodic (monthly) valuations and produce a cashflow forecast.
The completion of a process to produce a baseline plan produces at least the following as
outputs namely, planned value (PV) per activity for each reporting period which is
monthly – this is however arranged cumulatively over the duration of the activity; the
schedule at completion (SAC) and the budget at completion (BAC) for each activity and
on aggregation for work package, deliverable and project.
Monitoring phase
On commencing the execution of a project the baseline plan is put into action through a
monitoring process of data collection and a comparison of the actual to plan. The first
task requires making two measurements, firstly recording the actual costs (AC) incurred
on the project from inception to the reporting date. This measurement may be done in
many ways including physical measurements, expert eye and rule based measurement
(for a more detailed analysis of the completion measurement techniques (please see
Ruskin, 2004). AC and PC are measured in a cumulative way to match the PV produced
in the baseline plan.
The measurement of PC enables the computation of the earned value (EV) to the time or
reporting through the use of the formula EV = PC * BAC.
The last part of the monitoring phase requires the generation of reports indicating the
performance of the project and these include performance variances and performance
efficiencies. The indicators answer the two questions already mentioned (Q1and Q2) in
section 2.2.
The variances (absolute or percentages) indicate the magnitude of deviation of project
performance from the baseline plan. The indices on the other hand indicate the
performance efficiency of the project as illustrated in Table 1.
Project control is achieved when the project variances and indices are interpreted using
the rules of thumb, for example, a negative variance means the project is over budget or
behind schedule (and vice versa) while an efficiency index of greater than one indicates
better than planned or scheduled performance (and vice versa). Corrective action can
then be taken depending on the direction of the performance indicators. Anabari (2003)
indicated that rule of thumb can be complemented with tolerances where, for example, a
deviation of above -5 or 10% signals some form of trouble. This of course does not
include critical activities.
Parameters Input Data Analysis

4
Table 1: EVA performance analysis
Forecasting and updating of baseline plan
A prudent and proactive project manager needs to have an indication of the most likely
state of affairs based on the past project experience and the most likely future scenario.
There is no need for the project manager to dream that the project will stay within the
planned state when it is very clear a deviation has occurred and is most likely to continue
into the future, provided no corrective action is taken. EVA empowers a manager to
change the baseline plan by forecasting the likely schedule and cost in order to update
the baseline plan. To accomplish this, the project manager must relate past performance
to the likely future performance by identifying the possible scenarios (assumptions) that
will prevail in future, for example, that the project still and will remain on schedule; or
future performance will be similar to what was planned despite the deviation of the past.
Once a forecast is made of the likely SAC and BAC of the package and indeed the
project, the baseline plan is updated.
The EVA background given provided a framework for the research design described in
next section.
Methodology
The case study methodology is described first, followed by a description of the nature of
construction building project selected. The section ends with a description of how EVA
was applied in the study.
The case study methodology
Two aspects made the case study methodology more suitable for the investigation than
other research approaches. Firstly, preliminary investigation from the construction
industry’s major stakeholders (e.g. clients, consultants and contractors) revealed that
EVA is an alien project control tool which has never been used on any project in
Botswana. Therefore, there was no need to use the survey method.
Performance -variances
Absolute
Cost Variance (CV)
Schedule Variance (SV)
AC, EV
EV, PV
CV = EV-AC
SV = EV – PV
Percentage
Costs variance percentage
(CVP)
Schedule variance percentage
(SVP)
CV, EV
SV, EV
CVP = CV/EV
SVP = SV/EV
Performance Efficiencies
Cost performance index (CPI)
Schedule performance index
(SPI)
EV, AC
EV, PV
CPI = EV/AC
SPI = EV/PV

5
Another issue which favoured the case method was that the issue under investigation
required the use of a live project to collect data as work progressed. Secondly, it would
be an expensive and enormous task to collect data from projects in several different
locations within the same period (Blumberg et al., 2005). The study was neither
establishing a causal relationship to warrant the use of the experiment method.
Selected project
An EVA baseline plan was developed for the project which consisted of a one storey
police administrative office with several ancillary facilities including, a horse stable,
hazardous store, external works and a water pump house. In addition, the project
included the normal external, electrical and mechanical trade works. A total of the
equivalent of US$ 2.9 million was contracted for the project for duration of 12 months
beginning June 2006.
Since the traditional procurement system (TPS) was used, naturally the TCM was used
as the basis for project control. The TCM is one-dimensional method which involves
reporting progress and construction associated problems in the monthly site meetings. All
relevant parties including client’s representatives (CoW and government officials),
consultants and contractor meet to discuss the project progress but without reference to
nay common data. In addition, a clerk of works (CoW) is appointed at the
commencement of the project to oversee the day-to-day project activities especially in
terms of workmanship and material brought to site. Any important issue relating to project
progress is quickly brought to the attention of the lead consultant by the CoW prior to site
meeting.
EVA information infrastructure
Due to lack of information migration from designer to contractor, a lot of information
pieces had to be done from scratch, for example, despite the fact that QS used a WBS to
produce a BoQ, the WBS is never passed on to the contractor. Thus to set up the EVA
infrastructure required the cooperation of the QS and the contractor to develop the
necessary pieces of EVA and also in the collection of the necessary data. A three phase
approach was needed to accomplish this task.
Phase I
Since the contractor did not use the EVA a parallel system as the one described in
Section 2.2.1 had to be set up. However the cooperation of the contractor was needed to
obtain necessary schedules of activity planned valuations to construct the necessary
PV’s SAC’s and BAC's for the various projects work package given in Appendix A.
Phase II
Phase II required taking measurements of the work completed. This was completed
sometimes by measuring physical dimensions e.g. length, area, volume or by counting
number of completed items. The data on AC was collected from quantity surveyors books
for each work package at each reporting period but the adjusted to remove the material
cost element normally included in the payment certificates. This was to remain only with
completed work in line with planned value (PV).

6
Phase III
Phase III required the computation of various performance variances and efficiencies
(mentioned in Section 2.2.2
Research schedule
The baseline infrastructure was set up two months before the project commenced.
However, data collection was allowed to continue up to nine months after the beginning
of the project at which point the project was about 70% complete. Data analysis and
forecasting was conducted every month.
Research Findings
As already indicated an effective project control system requires a good plan and
monitoring system. The findings are therefore discussed in these two themes, but in
conjunction with a comparison with the TCM which was being used as control system on
the project.
Planning
EVA approach requires a detailed plan involving hours of meticulous planning effort to get
the necessary PV’s SAC and BAC’s for the items listed in Appendix 1. It is also efficiently
achieved with a help of an IT system. Most of the research data was contained in
spreadsheet. Projects with repeatable elements make the use of the IT system more
efficiently and effectively. Moreover, this planning is best done during bidding and not
after so as the tender sum reflects a detailed estimation of the durations and costs of the
project’s work packages.
The last point was more contentious for the contractor of the selected project especially
after seeing the details required of the EVA system. He quipped that what if I put in so
many hours of work to make that EVA infrastructure and then I do not get the job and
perhaps even the next three’. The comment is a common feeling among contractors that
pre-contract estimation should not be as detailed because of the probability of not getting
the job (Smith, 1995). Some authors have even suggested that in practice contractors
would prefer to use the 20/80 rule of estimating 20% of the items that constitute 80% of
the tender value (Skitmore, 1994). Unfortunately EVA requires full commitment to
estimation to the last item and activity.
The quest for details also allowed the early discovery of missing, omitted and conflicting
information or items prior to construction. Had EVA been used at bidding time it would
have even made the information more complete.
It also became clear that for EVA to work and be useful to all project stakeholders, a
cooperative environment was needed between the client, design and construction team.
This requires not only changing the procurement regime but also the mindset and remove
the arms-length type of environment that characterises TPS and replace it with a team
work spirit in which information can be shared and migrated from one project phase to
another.

7
Monitoring and reporting
Though EVA was not the official system of control used on the project, study team shared
its output with the QS, contractor and architect. At the beginning of the there was
suspicion but later there was enthusiasm generated. As the months passed the team
was more anxious to see the EVA report such as that given in Appendix B for the Office
block. The study team was allowed in the site meetings as silent observers to the
proceedings.
EVA allowed the measurement of performance based on the same data and moreover in
an objective manner. This aspect contrasted with the TCM and in a classical incident,
during a site meeting the CoW reported that project progress was at 30% while the
contractor put it at 45%. Though the contractor had a programme, his assertions of
progress were not based on any measurement.
Furthermore, EVA system allowed data desegregations and aggregation to cater for work
package as well as overall project progress analysis and reporting. Table 2 illustrates a
report produced for the Office Block for the third month (see Appendix B for full data
analysis and report). Equivalent items picked in the site meeting are juxtaposed.
Table 2 Performance report
Project Status EVM Traditional
Parame
ter
Value Parameter No. or
Value
Scope/schedule performance
How much has been completed
How much should have been completed
EV
PV
US$107,
883
US$12,0
35
Scope
Nil
12%
Nil
Scope/schedule/cost performance
How much has been spent on work
done?
How much should have been spent on
the work done?
AC
EV
US$95,2
00
Gross
Costs
Nil
US$613,
022
Nil
Efficiency performance
Is the project ahead or behind
schedule?
SPI 9 Nil Nil
Is the project on or over budget? CPI 1.1 Nil Nil
From Table 1, TCM provided less information, for example, in the monthly meetings cost
is normally not an issue and may not reported unless project funds are exhausted. Even if
it were reported the figure would be that of payments made to the contractor for the entire
project which in this case was equivalent to about US$ 613,022. Progress is given an
arbitrary figure of 12%.

8
On the other hand EVA system is able to report progress of each work package of the
project as the report shows, the package is ahead of schedule and within budget as
shown by the performance indices (all greater than 1). This performance was repeated
throughout the duration of nine months, at least for this particular work package as Figure
1 illustrates. However, this performance contrasts with another package illustrated in
Figure 2 which shows that the work was behind schedule for most of the time (SPI was
less than 1) though from the third months is was on budget.
Figure 1: Graphice perfromance indices report for the Office Block (month 1-9)
There was agreement between the contractor, QS and the consultants that the reports
generated provided the project team or manager with clear, timely and objective
information on which to base decisions on how to control the project. This contrasted with
the rather ad- hoc, reactive and rather subjective regime used in the TCM.
Forecasting and updating the baseline
Project forecasts, if made diligently provide a project manager (team) the ability to take
proactive steps before the project gets into any significant slippage of time and cost. The
cooperative use of information provides the client and client representatives an objective
criterion to sanction changes to the baseline plan.
The TCM system is not to predict objectively any of the two parameters unless one relies
on gut feeling of the contractor or the CoW or lead consultant. In fact what actually
happened in the meeting was that there was no reference or desire to forecast or know
the likely duration and cost of the remainder of the work. Given this scenario, it takes long
before any party on the construction team realizes that more funds or time are needed for
the project. Secondly because of objectivity it is difficult to make a quick decision on such
a thing as to whether time extension should be approved or not.

9
Figure 2: Graphic perfromance indices report for the External works (month 1-9)
Conclusions
Projects continue to under perform in many parts of the world with a significant impact on
the developing world. Many reasons are always brought forward including project control,
a two stage process involving both planning and monitoring. In Botswana, most projects
are controlled by the TCM system. The system looks at one objective of a project at a
time, be it cost or schedule of a project, lacks integration, objectivity and relevant
information to control the project.
The EVA is a more effective system because of its ability to integrate scope, schedule,
and cost in its monitoring and reporting. EVA’s ability to monitor, analyze, report and
forecast in an integrated manner provides the project manager and other stakeholders
the capacity to take informed and objective decisions. Moreover, it ability aggregate and
disaggregate project data is a powerful tool for the project manager.
Limitations and Recommendations
There are some limitations related to the case study and the use of the EVA. First, the
study required a protracted negotiation between the QS, contractor and architect to
obtain relevant information to draw up the EVA baseline plan and thereafter to get
monitoring information. This in its self suggested that it may be difficult to use the EVA
unless there is a change of mindset and perhaps method of project procurement. The
TPS allows only particular information to migrate from the designer to the contractor is
what is found in the bid documents. Thereafter an arms length type of relationship ensues
throughout the construction phase.

10
Secondly, a quick survey of the consultants and contractors showed that there is little
knowledge about EVA in that some have not heard about it, while others have heard
about it but do not know how to use it.
Lastly this being an exploratory study it has not brought out a number of issues, for
example during the time project collection there were no scope changes yet these are
very common in the Botswana industry. Therefore, a further study is recommended
involving more cases of projects involving scope changes.
References
Al-Jibouri S H. (2003), Monitoring Systems and their effectiveness for project cost control
in construction. International Journal of Project Management, 21: 145-154
Anbari F T. (2003), Earned Value Project Management Method and Extensions. Project
Management Journal, 34(4): 12-23.
Assaf S A, Al-Khalil M and Al-Hazni, M (1995), Causes of delay in large buildings
construction projects, Journal of Management in Engineering, ASCE, 11(2), 45-50.
Burke R. (2003), Project Management Planning and Control Techniques, 4th
edition. West
Sussex: Wiley.
Charoenngam C & Sriprasert E (200), Assessment of cost control systems: a case study
of Thai construction organizations. Engineering, Construction and Architectural
Management, 8(5/6): 368-380.
Chimwaso D (2000), An Evaluation of Cost Performance of Public projects: Case of
Botswana, Proceedings of the 2nd
International conference of the CIB Task Group 29
on Construction in developing Countries, Gaborone November 15-17 2000,
Cleland and Ireland (2007), Project management: Strategic decision and implementation,
MacGrawHill, NY
Elinwa U and Buba S A (1993), Construction cost factors in Nigeria, Journal of
Construction Engineering and Management, ASCE, 119(4), 698-713.
Smith A J (1995), Estimating, Tendering and Bidding for Construction, Macmillan,
Basingstoke, UK
Kerzener H (2003), Project Management, A systems approach to planning, scheduling
and controlling, 8th
Edition,Wiley, NJ, USA.
Klastorin T (2004), Project management: Tools and trade-offs, Wiley and Sons, Danvers,
MA, USA
Koppelman J M & Fleming Q W (2001), Earned Value Management: An Introduction, in
Project Management for Business Professionals – A Comprehensive Guide, edited
by J Knutson. New York: John Wiley & Sons: 166-178.

11
Lewis T M (1984), A review of the recent problems in the construction industry in Trinidad
and Tobago, Construction Management and economics, 2,37-48.
Marchewka J K (2003), Information technology project management, Wiley, NJ, USA
Meridith and J R and Mantel S J (2003), Project Management: A managerial Approach,
5th edition, Wiley and Sons, NJ, USA
Mokotedi A (2006), Legislators want Mogae, Khama to crack whip. Daily News. 04
September 2006: 2.
Ofori G. (2000), Challenges of Construction Industries in Developing countries: Lessons
from various countries,, Proceedings of the 2nd International conference of the CIB
Task Group 29 on Construction in developing Countries, Gaborone November 15-17
2000.
Palalani K. (2000), Challenges Facing the Construction Industry: A Botswana
perspective, Proceedings of the 2nd
International conference of the CIB Task Group
29 on Construction in developing Countries, Gaborone November 15-17 2000.
PMI (2005), Earned value in public sector. Government of Punjab.
http://www.punjab.gov.pk/pmu/index.php?otion=com_content&task=view&id=35&I
temid=59 Accessed on 2006/08/10
Rozenes S, Vitner G & Spraggett S (.2006), Project Control: Literature Review. Project
Management Journal, 37(4): 5-14.
Ruskin A M.(2004), Two Issues Concerning the Use of Earned Value Measurements.
Engineering Management Journal, 16(3): 26-30.
Skitmore and Wilcock (1994), Estimating processes of smaller builders, Construction
Management and Economics, 12, pp 139-154
Shenhar A & Dvir D. (No date): Project Management Research: Challenges and
Opportunities.
http://howe.stevens.edu/CTMR/WorkingPapers/WP2005/Shenhar_PM%20Research
%20Challenges%202005 Accessed on 2006/09/19.
Shtub A, Bard JF and Globerson S. (2005): Project Management Processes,
Methodologies, and Economics, 2nd edition. New Jersey: Pearson Prentice Hall.
Ssegawa J K and Ngowi (2007) Causes of poor performance of public construction
projects in Botswana, BOCCIM-UB Sponsored Research Project, Department of
Civil Engineering, University of Botswana
Tah J H , Thorpe A and MacCaffer R (1993), Contractors project risk contingency
allocation using linguistic approximation, Computing systems in engineering, 4(2-3),
281-293.

12
Wachira I N (2000) , Labour Management in Kenya, Proceedings of the 2nd
International
conference of the CIB Task Group 29 on Construction in developing Countries,
Gaborone November 15-17 2000.
Wells J, Herd C and Hawkins J (2007), Modifying infrastructure procurement to enhance
social development, Engineers Against Poverty - Institute of Civil Engineers,

Earned Value Analysis

More Related Content

What's hot (20)

Similar to Earned Value Analysis (20)

Earned Value Analysis