ArticlePDF Available

Improving Project Budget Estimation Accuracy and Precision by Analyzing Reserves for Both Identified and Unidentified Risks

Authors:

Abstract and Figures

Project risk is a critical factor in estimating project budget. Previous studies on this topic have only addressed estimation methods that consider project budget reserves against identified risks. As a result, project managers still face the challenge of completing projects within given budgets but without the relevant tools to deal with unidentified risks. This study proposes an approach for estimating reserves for both identified and unidentified risks separately. The study also suggests using the three-point estimation technique and R-value determination for estimating risk costs, which can improve budget accuracy and precision. The construction of residential building projects in South Korea demonstrates the advantages of the proposed approach compared with previous methods.
Content may be subject to copyright.
Article
Improving Project Budget Estimation
Accuracy and Precision by Analyzing Reserves
for Both Identified and Unidentified Risks
Hyukchun Kwon
1
and Chang Wook Kang
1
Abstract
Project risk is a critical factor in estimating project budget. Previous studies on this topic have only addressed estimation methods
that consider project budget reserves against identified risks. As a result, project managers still face the challenge of completing
projects within given budgets but without the relevant tools to deal with unidentified risks. This study proposes an approach for
estimating reserves for both identified and unidentified risks separately. The study also suggests using the three-point estimation
technique and R-value determination for estimating risk costs, which can improve budget accuracy and precision. The con-
struction of residential building projects in South Korea demonstrates the advantages of the proposed approach compared with
previous methods.
Keywords
contingency reserve, management reserve, risk cost, budget estimation
Introduction
Cost is the most critical parameter (Becker, Jaselkis, & El-gafy,
2014; Ke, Ling, & Ning, 2013; Sweis, Sweis, Rumman,
Hussein, & Dahiyat, 2013) within the standard success criteria
of cost, schedule, and performance targets—often called the
“iron triangle” (Pfleeger & Atlee, 2006; Williams, 2016)—
when it comes to managing projects. However, in reality,
project cost overruns and scope creep are normal phenomena
in infrastructure and construction projects carried out in both
developed and developing countries (Bhargava, Anastrasopou-
lous, Labi, Shiha, & Mannering, 2010; Doloi, 2013; Enshassi,
Al-Najjar, & Kumaraswamy, 2009; Frimpong, Oluwoye, &
Crawford, 2003; Sambasivan & Soon, 2007; Smith, 2014).
Flyvbjerg, Holm, and Buhl (2002) found that 90%of construc-
tion projects underestimated costs, which resulted in cost over-
runs of between 50%and 100%.
Project cost overruns are significant problems in govern-
ment project management as well. In government project
management, projects are complex and larger. Thus, many
large-scale, complex systems development projects also expe-
rience persistent cost and schedule overruns (U.S. Government
Accountability Office [GAO], 2013). In 1983, the Nunn-
McCurdy Act was passed into law by the U.S. Congress to
prevent overruns. The law requires Department of Defense
(DoD) acquisition programs and other large-scale federal gov-
ernment projects to report to Congress when they exceed
certain established cost overrun thresholds (Schwartz, 2010).
This law has been amended many times over the years to reflect
evolving federal project management and reporting practices
(Adoko, Mazzuchi, & Sarkani, 2016).
While project managers have been managing projects to meet
budget, time, and performance targets, researchers have been
trying to identify the root causes of cost overruns and develop
accurate budget estimation methods to solve these problems.
According to previous researchers, risks are one of the major
reasons for cost overruns; thus, various budget estimation meth-
ods, including estimating project reserves against risks, have
been developed. However, previously developed project reserve
estimation methods were presented to estimate reserves against
identified risks only, even though budgets against unidentified
risks are also included in project reserves, which is a deficiency
of the budget estimation methods previously studied. Thus, proj-
ects still may suffer from cost overruns.
Project budgets are funds estimated during the planning
phase based on what the project is expected to cost at
1
Department of Industrial & Management Engineering, Hanyang Univ. ERICA
Campus, Korea
Corresponding Author:
Hyukchun Kwon, Department of Industrial & Management Engineering,
Hanyang Univ. ERICA Campus, Korea.
Email: hkwon21@hanyang.ac.kr
Project Management Journal
Vol. 50(1)1–15
ª2018 Project Management Institute, Inc.
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/8756972818810963
journals.sagepub.com/home/pmx
completion. It is very difficult to estimate project budgets accu-
rately before executing the projects due to lack of information
and risks. Thus, a project management plan including cost
estimation are developed at an early stage before projects are
constructed. So, it is very difficult to accurately estimate the
project budget due to lack of information or data (Creedy,
Skitmore, & Wong, 2010; Koushki, Al-Rashid, & Kartam,
2005; Oberlender & Trost, 2001), and is needed for justifica-
tion of projects on economic grounds and for efficient capital
planning and financing (Baccarini, 2006; Caron, Ruggen, &
Merli, 2013).
The purpose of this article is to develop an innovative bud-
get estimation method that includes estimating project budget
reserves against both identified and unidentified risks. This
article will discuss the advantages of the method presented
herein, along with recommendations for a simplified treatment
of correlations between past performance and future perfor-
mance. Application of this methodology in example projects
will also be presented. Twenty residential building construction
projects in South Korea were selected, and the variances
between budgets and actual costs were analyzed.
Literature Review
Researchers have investigated the root causes of cost overruns
and developed solutions to prevent them in different countries
and various ways. According to previous research, inaccurate
cost estimation and uncertainties are the major reasons for cost
overruns. Various estimation methods have been developed to
mitigate the additional costs resulting from uncertainties. How-
ever, the existing models have estimated project reserves for
identified risks only so that cost variances cannot be predicted
and controlled; thus, cost overruns are still a common occur-
rence in project management. The main drawback of the exist-
ing estimate reserve methods is their lack of consideration of
unidentified risks. Therefore, an innovative budget estimation
method that responds to unidentified risks as accurately and
precisely as possible is required in order to avoid cost overruns.
The Root Causes of Cost Overruns
Nawaz, Shareef, and Ikram’s (2013) work on cost performance
in Pakistan listed factors that are responsible for cost overruns.
These factors include corruption and bribery, political interests,
poor site management, delays in site mobilization, rigid atti-
tudes among consultants, extra work without approvals, and
frequent changes during execution. In Ghana, 75%of ground-
water construction projects exceeded the original project
schedule and budget. The main causes of the delays and cost
overruns were financial difficulties, poor resource manage-
ment, and unexpected natural events (Frimpong et al., 2003).
Sweis et al. (2013) analyzed different types of public construc-
tion projects in Jordan and found that 65%of them were not
finished on budget. The major factors that caused the cost over-
runs were governmental delays, followed by severe weather
conditions, and design changes. These three factors account for
73%of cost overrun causes. A survey of 104 public projects in
Singapore indicated that nearly two-thirds suffered from cost
overruns and more than half were delayed due to risks (Hwang,
Zhao, See, & Zhong, 2015; Ke et al., 2013). According to
Koushki et al. (2005), delays and cost overruns increased in
the construction of a private residential project in Kuwait due to
three main causes: contractor-related problems, material-
related problems, and owners’ financial constraints. In the
Nigerian construction industry, one of the causes of cost over-
runs was inadequate contingency allowance, where a 5%to
10%contingency allowance was a common practice. However,
the actual 17.34%contingency allowance estimated falls
within the 15%to 20%allowance recommended by the U.S.
Department of Energy (DOE) for budget estimates of new
buildings (Aibinu & Jagboro, 2002). Aziz (2013) presented
improper bidding/tendering methods, inaccurate cost estima-
tions, and unexpected risks as the major factors causing cost
variations for constructing wastewater projects in Egypt. Eco-
nomic factors such as interest rate, unit price for material and
labor, rental rate for equipment, and changes in planned works
were risk factors for building construction in South Korea (Cha
& Shin, 2011). More recently, other studies have shown some-
what similar results. Major factors affecting cost overruns in
public construction projects included: materials price fluctua-
tions, lack of experience among contractors, incomplete draw-
ings, government delays, incompetence, inaccurate estimates,
improper planning, and poor labor productivity (Doloi, Sawh-
ney, & Rentala, 2012; Kasimu, 2012; Memon, Rahman, & Azi,
2012; Tabish & Jha, 2011). Researchers have indicated that
these significant cost overruns are caused by uncertainties aris-
ing from risk events and inaccurate budget estimates (Banner-
man, 2008; Creedy et al., 2010; Elkjaer, 2000; Hullet, 2012;
Lai, Wang, & Wang, 2008). Uncertainty of cost items is an
important aspect in complex projects, and cost uncertainty
analysis aims to help decision makers understand and model
different factors that affect funding exposure and ultimately
estimate the cost of projects (Khodakarami & Abdi, 2014).
As a result of this research, two main causes of cost overruns
have been identified: One is managerial factors not related to
risks, such as regulations, contract methods, and political
issues, and so forth, which are beyond the scope of this article
because they could be improved by education, project team
member experience, and historical data. The other cause is
inaccurate cost estimation resulting from uncertainties—espe-
cially unrecognized or unexpected events—which are within
the scope of this research to help improve. The main objective
of this article is to present an innovative budget reserve estima-
tion method used to mitigate the impacts against both identified
and unidentified risks in order to minimize the cost variances.
Cost Estimation Methods
Estimating an accurate project budget is challenging for project
managers, because of the unpredictable risks concerning how
2Project Management Journal 50(1)
big the impacts on construction project results are and when
they will occur. Furthermore, budget estimation is conducted
during the planning phase (Caron et al., 2013; Koushki et al.,
2005; Sato & Hirao, 2013; Xenidis & Stavrakas, 2013), which
is an early stage of the project life cycle, when there is a lack of
data and information. Project managers need a budget estima-
tion method to respond to risks as accurately and precisely as
possible in order to prevent cost overruns. In response to uncer-
tainties and risks, Project Management Institute (PMI) (2013)
defined reserves that are composed of a contingency reserve for
identified risks (known-unknowns) and a management reserve
for unidentified risks (unknown-unknowns). Previous research-
ers have presented several methods to estimate cost reserves
but those are not sufficient to cover all types of risks. One
method is the traditional percentage model (Moselhi, 1997),
which is arbitrary and difficult to justify or defend (Thomson
& Perry, 1992); the other method is Monte Carlo simulations
(Barraza & Bueno, 2007; Clark, 2001; Eldosouky, Ibrahim, &
Mohammed, 2014); and a third method is the regression model
(Adoko et al., 2016; Kim, Kang, & An, 2004). These models
are used for estimating total project costs and are powerful
statistical tools used for analytical and predictive purposes in
forecasting the total final cost of the project. These methods
lack consideration of estimating risk costs because they are
used to estimate the total project budget without risk analysis.
If a single project is executed many times, the probability den-
sity function (p.d.f.) can be obtained. The probability of the
project cost can be calculated by the probability density func-
tion, so that the probability that actual cost can exceed the
established target cost can be obtained on the probability den-
sity function during project execution (Garvey, 2008; Zhu,
Zhang, & Wang, 2011). Thus, any types of risk can be esti-
mated to a monetary value, which can be involved in project
cash flow as well (Halawa, Adbelalim, & Elrashed, 2013). For
example, gamma distribution (Uzzafer, 2013) and a scenario-
based method without statistical concepts (Book, 2007; Gar-
vey, 2008) have been presented for estimating cost risk
provided by cumulative probability distribution with limited
predefined values of risk impacts for all risks, regardless of
whether they are identified or unidentified. Other models using
the fuzzy expert system (Carr & Tah, 2001; Dikmen, Birgonul,
& Han, 2007; Idrus, Nuruddin, & Rohman, 2011) and artificial
neural networks (Chenyun, 2012; Zhu et al., 2011) have been
used for the development of a project cost contingency estima-
tion model. These models are suitable for the nonlinear model-
ing of data, which contrasts with linear approaches using
regression (Baccarini, 2006), and may be used effectively in
the risk assessment for identified risks, but they are less effec-
tive for estimating cost contingency.
The results of previous research do not clearly estimate
reserves for unidentified risks, because unknown-unknown
risks were excluded from the research due to assumption by
the author (Baccarini, 2006) or unmanageable (Chapman,
2000). Furthermore, contingency resources estimated to han-
dle unknown risk events cannot be justified because these
could not be identified and estimated (Kitchenham & Link-
man, 1997). In addition, these are events not known to the
project team before they occurred or viewed as impossible in a
specific project situation. By definition, unknown-unknowns
are not foreseeable and thus cannot be dealt with proactively
(Smith & Merritt, 2002; Thamhain, 2013). Although risk cost
estimation methods for identified risks have been presented
by previous researchers, the reserves for unidentified risks
have not been sufficiently examined. The rare studies on esti-
mating reserves against unidentified risks cause large cost
variances and difficulty for sponsors or project managers in
making decisions properly in order to provide benefits from
the project results.
The Basic Terminologies of Budget
Compositions
Net cost, allowance, and point estimate (PE)
Project managers create a work breakdown structure (WBS)
and identify the applicable cost factors associated with project
work packages in order to develop the project budget. The
budget of each work package consists of the labor, materials,
and overhead costs. In this article, however, the overhead costs
are not included because they represent another cost factor that
can be variable depending on the organization’s management
level. The costs of work packages can be calculated by multi-
plying the quantity of material and/or labor by the unit cost of
material and/or labor. The net cost is the sum of the monetary
value of the resources for a fixed scope of work without any
tolerances or margins. Generally, the resources for a fixed
scope are described in the design or on drawings. The quantity
of materials and labor contains net quantity and any tolerance
or margin against mistakes and/or miscalculation, such as
human error. Additional funds for any margins or tolerances
is an allowance intended for specific subjects that have not or
cannot be fully specified (e.g., technical allowance, purchase
allowance, and weather allowance) (Bedarida & Conti, 2012).
The allowance can be changed depending on the project mem-
bers’ experience, educational level, knowledge, historical data
of the project, expert judgment, and lessons learned. The sum
of net cost and allowance becomes the PE of the element costs
across the project’s WBS without any adjustments for uncer-
tainty but including any allowances (Garvey, 2008). The sums
of each work package’s point estimate become the total point
estimate of the project without any risks.
Reserves and Budget Baseline
Once the PE has been reviewed and approved the next step is to
estimate reserves as a budget against risks. In addition to all the
work identified, projects will have some other unplanned work
that is the result of risks. The project budget consists of PE and
reserves. Although a PE not related to risks can be made as
accurately as possible based on education and experience,
reserves as risk cost related to risks can be estimated with
Kwon and Kang 3
probabilities and impact amounts. There are two categories of
reserves: the contingency reserve for identified risks and the
management reserve for unidentified risks. The former is the
budget for response actions taken against identified risks, and
the latter is the budget to cover other risks, such as unidentified
risks that include residual and secondary risks beyond identi-
fied risks. The project budget baseline can be described as the
sum of the PE from the project cost management process and
the contingency reserve for identified risks from the project
risk management process. The management reserve is not
included in the budget baseline, but it is included in the total
project budget (PMI, 2013). The relationship between the proj-
ect risk management process and the project cost management
process should, therefore, be considered in the estimate of the
total project budget.
Confidence Level
Historical data and experiences are aggregated to review and
determine an appropriate level of confidence. The confidence
level covers all of the actual costs, including PE and additional
costs from identified risks, unidentified risks, residual risks,
and secondary risks (Book, 2007; NASA, 2008). The additional
dollar amount beyond the budget baseline is considered and
executable within the level of confidence. Software programs
such as Crystal Ball and @Risk are used to determine the
confidence level. The confidence level is subject to change
depending on the projects’ features and categories, including
IT, construction, and R&D projects.
Analysis of Previous Project Cost
Performance
Project selection and analysis. For the research presented in this
article, 20 residential building construction projects were
selected for performance analysis to verify the insufficiency
of the existing estimation methods to cover risks. The purpose
of performance analysis on past projects is to develop the con-
fidence level and examine trends in the variances between
budgets and actual costs for future projects. All work packages
are assumed to be independent of each other so that there are no
positive and/or negative correlations between work packages,
because correlation is a very important aspect of combining
cost distribution. For example, if the cost of one work package
increases because of risks, then the cost of other work packages
neither increases nor decreases.
Data collection was conducted through interviews and con-
sulting with project managers and/or directors between the
years 2014 and 2015. The residential building projects ana-
lyzed were constructed between the years 2008 and 2015.
These projects were selected to maintain continuity and con-
sistency and to avoid the bias in data analysis caused by differ-
entiation of work breakdown structure, risk register, and
maturity level of project management. All the selected projects
were from one company. The company’s policy for the case
study to estimate reserves was to use a hybrid method with
traditional percentage and risk analysis to develop response
plans. Reserves for risks were determined to represent approx-
imately 10%of the total budget. The sum of the total budget of
20 projects are overrun by 37,693, whereas the point estimate
was 5,541 overruns, which were not caused by risks. Overruns
of reserves were 32,152, which were caused by inaccurate risk
budget estimation and unidentified risks. The 5,541 overruns of
the point estimate are caused by estimation errors or miscalcu-
lations. However, the 32,152 overruns of the reserves are
caused by the errors of the risk analysis, including risk identi-
fication and estimation risks. The details of cost overruns and
risk costs are analyzed in Table 1. The type of currency was not
identified for the purpose of company confidentiality.
Findings
Only the identified risks were recorded on the risk register.
Response plans were developed in response to those risks, but
the unidentified risks could not be recorded and calculated. It
was very difficult to determine accurate reserves, because proj-
ect managers or experts could not precisely forecast the number
of probabilities and the impact in the early stage, even though
some risks could be identified. In this article, the variances
between budgets and actual costs by each budget composi-
tion—such as PE, actual costs for identified risks, and for uni-
dentified risks—were analyzed to develop a new budget
estimation process. As a result of the above analysis, there must
be three cases regarding the relationship between risk costs and
actual costs. The first case is that there are budgets and actual
costs for risks; the second case is that there are budgets but
there are no actual costs; and the third case is that there are no
budgets, but actual costs are disbursed. The last case describes
a situation in which the costs are incurred by unexpected
events. As in the first case, the cost overruns of reserves for
identified risks were 17,542 between the budgets of 146,838
and actual costs of 164,380. In the second case, risk budgets for
exchange rate were estimated at 463, but there were no actual
costs. In the third case, there were actual costs of 15,073 with-
out risk budgets in some of the risk items such as cash flow
impact, delays due to excessive approval procedures, lack of
coordination among project participants, capability of the own-
er’s group, and others. The first and second cases are described
for the contingency reserves against identified risks, whereas
the third case was for management reserves against unidenti-
fied risks. In terms of project cost management, cost variances
between budgets and actual costs are inevitable. Thus, calcu-
lating the contingency reserve and management reserve sepa-
rately is necessary to control project costs by collecting data as
lessons learned for future projects in order to minimize cost
variances even when the total variances between risk budgets
and actual costs are underrun or zero. In this article, an inno-
vative estimation method for project risk budget—including
contingency reserves for identified and management reserves
4Project Management Journal 50(1)
for unidentified risks—have been developed separately to
improve project cost management.
Proposed Method for Estimating Project
Budget
Accuracy and precision are important factors to consider when
determining project budget to minimize cost variances by cost
overrun or underrun. The proposed project budget estimation
method is an innovative one that improves budget accuracy
using the probabilistic estimate (Book, 2007; Garvey, 2008),
and budget precision using the three-point estimation technique
and R-value determination.
Probabilistic Estimation
It is assumed that the actual cost ofa project is random variable Xi
with probability density function fi. Each project presents a p.d.f.
of the forecasted project cost if the total project cost is a lump sum
of many cost components, such as work packages. When the
number of WBS elements increases, the distribution of the total
cost of the WBS elements approximates the normal distribution
with mean mand variance s2based on the Central Limit Theorem
(Barraza & Bueno, 2007; Book, 2007; Eldosouky et al., 2014). A
three-point estimation technique for each work package on the
WBS is used to obtain this p.d.f. for project budgeting from the
Monte Carlo simulation (Clark, 2001). This is an improved
method over single-point activity cost estimates because it con-
siders uncertainty and risk better (Book, 2007; PMI, 2013). The
three-point estimation technique, which is assumed to follow
triangular distribution (Xenidis & Stavrakas, 2013), is adopted
to develop a p.d.f. with a cumulative S-curve. The main focus of
this research is to determine project budgets with a low probabil-
ity of overrun or underrun (accuracy) and small cost variances
(precision). The cumulative distribution function (c.d.f.) for proj-
ect iis also defined by the mean value (miÞand variance ðs2iÞ.
FiðxiÞ¼ Z
xi
0
fiðtÞdð1Þ
¼ðPE+aÞþðRC+eÞ¼ðPE þRCÞ+ðaþeÞð2Þ
Table 1. Risk Register and Risk Budgets versus Actual Costs
Category Risks
Probabilities
(%)
Overrun
Costs of PE
Risk Reserves
Budget
(A)
Actual
(B) A B
Natural and Environmental Weather impacts 60 54,230 68,240 –14,010
Political Regulation changes against constructors 40 320 45 275
Financial Exchange rate change 30 463 463
Capital funding impacts 40 255
Cash flow impacts 40 473 –473
High costs due to improper bidding parties 30 522 457 65
Poor estimating 30 4,530 4,133 5,802 –1,669
Increased labor costs 40 2,506 1,240 1,266
Increased material and equipment costs 40 90 2,236 –2,146
Technical Design changes 50 540 212 3,415 –3,203
New risks due to new technologies 40 727 1,250 –523
Failures in production equipment 30 4,221 5,413 –1,192
Scope change 70 24,450 23,984 466
Technology selection 30 373 438 –65
Implementation methodology 30 481 527 –46
Delay due to excessive approval procedures 30 746 –746
Managerial Quality management risk 40 32,548 21,498 11,050
Strikes of subcontractors 30 2,110 1,091 1,019
Poor communications 30 5,438 9,972 –4,534
Assigning unqualified project participants 40 57
Late making decisions 30 31
Lack of coordination between project participants 30 6,250 –6,250
Lack of professional pre-planning studies 30 5,632 6,814 –1,182
Capability of owner’s project group 40 6,890 –6,890
Contractor capability 30 478 528 –50
Vendor’s capability 30 8,367 11,430 –3,063
Others 30 128 714 –714
Total 5,541 147,301 179,453 –32,152
Kwon and Kang 5
aCosts of unplanned works by errors for PE
eCosts of unplanned works by risks for risk cost
0a;e
Unplanned work aand eare also assumed to be indepen-
dently and normally distributed random variables with mean
zero and variance s2
aand s2
e;respectively. The value of a,
which is the cost of unplanned works by errors or mistakes for
estimating PE including allowance, is reducible by the project
members’ experience, educational level, knowledge, histori-
cal data, expert judgment, and lessons learned. On the other
hand, the value of e, which is the cost of unplanned works by
unidentified risks, is unmanageable. Project managers have
often made efforts to calculate accurate risk costs for project
budgets by minimizing e. In reality, it is very difficult to make
aand evalue zero perfectly, so one technique to minimize e
and ais to develop a confidence level that can be obtained
from the c.d.f. of triangular distribution by using a three-point
estimation technique. The probability of the actual costs,
which come from analyzing previous project performance
results, becomes the confidence level of the firm, as described
in Figure 1.
Difference (RValue) Between the High and Low Values
in the Three-Point Estimation Technique to Estimate Risk
Costs
In this article, the three-point estimation technique has been
used to calculate the PE of work packages for estimating the
total project budget. Using a three-point estimation tech-
nique not only for estimating the PE of work packages, but
also for each risk response plan to estimate risk costs, is an
innovative application. Low, most likely, and high values of
the impact amounts are estimated in order to calculate the
expected monetary value (EMV) by multiplying probabil-
ities and each impact value together, respectively. The most
likely value is the cost of the activity based on a realistic
effort assessment for the required works and any predicted
expenses. The optimistic value (low value) is based on the
best-case scenario for the activity, and the pessimistic value
(high value) is based on the worst-case scenario for the
activity. Most project budgets are substantial, so two critical
factors are needed to control the project budget. One factor
is the accuracy to meet that project budget; the other factor
is precision to minimize the cost variance between the bud-
get and actual cost. The difference (RÞbetween the high and
low values can measure the precision for cost variation
between budgets and actual costs and is described as
follows:
R¼ðHigh value Low valueÞð3Þ
The smaller the Rvalue, the higher the confidence to
meet the budget with only small cost variances. Through
the three-point estimation technique, PE and risk cost can
be determined by selecting the most likely value or the
higher value between the mean value and the most likely
value, depending on the project management maturity level,
previous experiences, and historical data. In this article,
higher values were selected because of the low maturity
level of the company under evaluation. In addition, the
project risk probability and Rvalue matrix should be spec-
ified in the project risk management plan shown in Figure 2.
The dark gray area represents the high-risk response plans
that are required to analyze and develop preventive plans.
The light gray area represents moderate risk that may be
required for preventive plans, and the white area does not
require any analysis or additional actions.
The actual costs of the project can fall within the interval as
shown below:
ðPE aÞþðRC eÞ Actual costs ðPE þaÞþðRC þeÞ
ð4Þ
However, PE and an avalue can be minimized and esti-
mated as accurately as possible by developing a WBS and
using a three-point estimation technique. Therefore, the cost
variances between the budget and actual cost of PE are
assumed not to be critical if the scope of work has not
changed; however, the risk cost is changeable depending
on risk occurrence. It is a very challenging task to determine
the point estimate for risks. Probabilities and impact
amounts of risks can be estimated by expert judgment, his-
torical data, and experience. Therefore, cost variances can
be expected as follows:
Min :RC Cost variance Max :RC ð5Þ
Cost variances between the total project budget and
actual cost can be affected by risks; thus, making the gaps
between Min. RC and Max. RC smaller is a critical success
factor in determining the optimum budget with greater
precision.
1.0
0.5
0.0
Probabilities
Costs
Reserved for
Identified Risks
PE Confidence Level
(Expected actual costs)
Project Budget Baseline
Reserved for
Unidentified Risks
Figure 1. Confidence level on cumulative probability distribution.
6Project Management Journal 50(1)
Determine the Confidence Level
To determine the confidence level, the budget of each of the
selected 20 projects was re-estimated using a three-point esti-
mation method. According to the results of these re-estimates,
a cumulative S-curve of each project was obtained. The prob-
abilities of PE and the actual costs of each project can be
calculated on the cumulative S-curve and assumed to follow
normal distribution with means and variances after confirm-
ing a normality test. As with the results of the previous 20
projects’ performance, the mean probability of PE is 48.93%
with a variance of 12.0281, while the mean probability of
actual cost is 74.71%with a variance on 7.4237 on S-curve,
as shown in Table 2. Thus, the probability of actual cost
(74.71%) is determined as the confidence level on the cumu-
lative S-curve derived from the triangular distribution of each
project performance. The 90%confidence interval of the firm
for the residential projects falls between 70.2%and 79.19%,
which determined that the confidence level (74.71%)could
cover the whole cost, including identified and unidentified
risks as well as secondary and residual risks. However, the
confidence level can be adjusted by updating the cumulative
results of the project performance regularly.
Applying the Proposed Method to a Live
Project
Calculation of the Budget
Live projects from the company were selected to demonstrate
the budget estimating processes using this proposed method for
five ongoing projects, applied by this proposed method. The
comparison between the budget estimated by the traditional
methods that have been used to date and the re-budget by the
method proposed in this article are described to verify the
improvement of the new budget estimation process.
The Basic Terminologies of Budget Compositions
Net cost, allowance, and point estimate (PE). The net cost of each
work package from the WBS can be calculated as precisely
as possible by referring to the design drawings, and the
allowance can be determined based on the technical and
management levels. Thus, point estimate can be estimated
accurately if there is enough time to plan project manage-
ment; however, the project budget may not be estimated
accurately due to lack of information and uncertainties
within limited time. The three-point estimation technique
for the point estimate is typically used to cover the
unplanned works of ain a new method, because budget
estimation is required in the early planning phase before
executing the project. All work packages are assumed to
be independent of each other in terms of cost, but some
work packages are assumed to be interdependent in terms
of schedule. The estimated data of each work package are
shown in Table 3.
According to Equation (2), Planned PE ¼Actual PE +a.
The value acan be reduced by improving the project manage-
ment maturity level. Project managers should develop a
guideline or policy to determine PE considering the avalue.
The guideline or policy indicates which value is selected as
the PE among the mean, most likely, or a certain point on the
S-curve, because the confidence level depends on the com-
pany’s maturity level. In this article, the larger value between
mean value and the most likely value can be determined as the
PE for each work package, because the project maturity level
is low for this company.
PE ¼Max ðmean;most likely valueÞð6Þ
The S-curve is obtained using the three-point estimation
technique, and 6,172 becomes the PE in this case project
according to the above assumption. The obtained S-curve is
assumed to follow a triangular distribution. Therefore, PE ¼
Most likely value þa¼6,121 þ51¼6,172 at 52.12%on an
S-curve was obtained using the @Risk version 6.0.0 with
1,000 trials. When the avalue is included in the PE, the e
value only becomes the management reserve. Therefore, the
minimum project budget of 6,908 can be calculated by apply-
ing the confidence level (74.71%) on the cumulative S-curve,
as shown in Figure 1. The gap of 736 between CL (6,908) and
PE (6,172) becomes the risk costs including contingency and
management reserves.
Risk cost assessment. The project budget consists of two parts:
One is the PE not related to risks, and the other is risk cost (RC)
related to risks. Project risk cost has probabilities where actual
cost exceeds the established target cost during the project exe-
cution caused by project risks (Zhu et al., 2011). The identified
risks are recorded in the risk register in as much detail as
reasonable. The main purpose of risk management is to develop
risk response plans. These plans, which are additional pro-
cesses with cost and time, are variable depending on the prob-
ability of a risk’s occurrence and the risk’s amount of impact
50
30
25
Risk Probability
EMV/PE(%)
Figure 2. Project risk probabilities and EMV/PE matrix.
Kwon and Kang 7
Table 2. Three-Point Estimate and Confidence Level
Project Number
Preliminary Project Budget Re-estimate PE Actual Costs
Cumulative
S-curve (Probabilities)
PE RC Total L M H Mean PE RC Total PE CL
1001 60,214 5,955 66,169 21,256 60,214 93,328 58,266 60,034 7,892 67,926 54.05 72.96
1002 39,309 4,271 43,580 16,578 39,309 65,236 40,374 39,586 6,534 46,120 46.72 71.03
1003 26,143 2,492 28,635 9,784 26,143 45,265 27,064 27,678 4,756 32,434 46.11 75.73
1004 37,566 3,806 41,372 15,608 37,566 57,347 36,840 36,950 5,340 42,290 52.61 72.54
1005 12,357 1,236 13,593 3,346 12,357 21,882 12,528 12,128 2,790 14,918 48.61 72.53
1006 106,759 10,420 117,179 59,862 106,759 153,294 106,638 107,586 12,950 120,536 50.19 75.32
1007 60,006 6,079 66,085 42,677 60,006 83,270 61,984 60,023 8,378 68,401 42.69 76.59
1008 73,751 7,034 80,785 44,234 73,751 101,244 73,076 73,950 9,356 83,306 51.78 79.47
1009 49,453 4,783 54,236 33,890 49,453 70,619 51,321 48,941 6,823 55,764 42.37 71.61
1010 163,146 16,927 180,073 93,620 163,146 238,062 164,943 164,505 19,345 183,850 48.13 72.84
1011 48,124 4,935 53,059 31,702 48,124 68,899 49,575 48,234 8,376 56,610 44.15 80.46
1012 41,682 4,022 45,704 20,264 41,682 61,256 41,067 41,371 5,210 46,581 52.25 73.16
1013 148,585 15,055 163,640 107,053 148,585 199,983 151,874 148,120 17,239 165,359 44.69 74.90
1014 98,390 10,023 108,413 52,350 98,390 140,789 97,176 98,820 9,836 108,656 52.06 72.46
1015 167,892 18,912 186,804 102,985 167,892 236,894 169,257 168,495 19,127 187,622 48.47 73.73
1016 78,390 9,732 88,122 51,953 78,390 104,264 78,202 77,943 9,320 87,263 50.56 78.65
1017 76,551 7,295 83,846 40,392 76,551 114,453 77,132 78,254 8,137 86,391 48.82 71.95
1018 27,764 3,027 30,791 18,913 27,764 36,219 27,632 27,234 3,234 30,468 51.14 77.40
1019 49,764 4,948 54,712 24,814 49,764 75,635 50,071 49,990 7,460 57,450 51.93 76.25
1020 64,813 6,349 71,162 31,489 64,813 96,584 64,295 66,358 7,350 73,708 51.19 74.70
Total 1,430,659 147,301 1,577,960 822,770 1,430,659 2,064,523 1,439,317 1,436,200 179,453 1,615,653 48.93 74.71
Note. PE(Point Estimate), RC(Risk Cost), L(Low), M(Most Likely), H(High), CL(Confidence Level).
8
amounts; thus, the response plan should be added to the pre-
liminary project management plan. When general risks
occurred, project managers conducted risk response plans.
There are two steps in developing risk response plans: First,
project managers develop alternatives to respond to risks, if
possible, with impact amounts and probabilities of risk occur-
rence; then they select the best alternative among them. A
three-point estimate of the impact amounts should be esti-
mated the same as the method for the PE of work packages.
Second, the larger value between the mean and most likely
value of the impact amount among the estimated values using
the three-point estimate technique is determined to calculate
EMV by multiplying by probability and construct confidence
level to cover eas same method as PE. The lowest EMV of the
alternatives is selected as the response plan. Budgets for each
risk response plan can be estimated by calculating EMV, as
shown in Table 4.
The project budget baseline, excluding management
reserve, becomes 6,784 by adding 612 of contingency reserves
to 6,172 (PE), so that management reserves become 124, which
is described in Figure 3.
The second step is to analyze selected response plans for
improving budget precision. Risk response plans are categor-
ized into two types: preventive and adaptive plans (Sato &
Hirao, 2013). Preventive response plans should be contained
within the preliminary WBS, so that their additional costs are
included in the PE to mitigate the risks in advance. However,
the additional costs of the adaptive response plans are used as
contingency reserves when risks occur.
Generally, the prevention costs rather than the correction
costs save the total cost of quality; thus, the greater the pre-
ventive costs, the more the save costs and the adaptive costs
become lower and with greater precision, as shown in Figure 4.
However, project managers should consider that the sum of
preventive and adaptive costs cannot be higher than the pre-
liminary risk cost. The criteria of reassessment of the risk
response plan are developed in Figure 2.
EMV ¼ðProbability Impact amountsÞþPreventive costs ðPCÞ
ð7Þ
Several independent risks can be prevented by one risk
response plan that can cover some work packages that will
Table 4. Risk Response Plan and EMV
Probabilities Impacts
Mean EMV*
R*
Response Plan (%) L M H (H-L)
Improper estimate 30 78 129 174 127 39 29
Increased material and equipment costs 40 149 218 294 220 88 58
Scope changes 70 185 258 346 263 184 113
Design changes 50 121 192 238 184 96 58
Poor communication 30 181 243 298 241 73 35
Vendor’s capability 30 360 442 495 432 132 41
Table 3. Three-Point Estimate for PE
WBS Level 1 Level 2 L M H Mean PE
Design Design definition 29 42 54 42 42
Conceptual design 138 273 408 273 273
Preliminary design 226 360 521 369 369
Final design 195 394 588 392 394
Civil work Foundation 409 624 897 643 643
Roads 120 250 380 250 250
Landscape 113 220 350 228 228
Architecture work Steel fabrication 382 654 910 649 654
Steel erection 275 525 764 521 525
Pouring concrete 160 277 378 272 277
Internal finishing 151 236 342 243 243
External finishing 85 185 282 184 185
Mechanical work System 128 251 373 251 251
Facilities 160 260 355 258 260
Machines 165 259 376 267 267
Electrical work Rough in 188 382 545 372 382
Equipment 189 357 510 352 357
Installment 232 572 832 545 572
Sum 3,345 6,121 8,865 6,110 6,172
Kwon and Kang 9
be impacted by risks when they occur. The results of
analyzing and re-estimating risk response plans are shown
in Table 5.
Re-estimating the project budget. The final total project budget
becomes 6,819, which is reduced from the preliminary budget
of 6,908 with higher precision by decreasing the Rvalue of risk
from 333 to 220. The PE increases from 6,172 to 6,262 by
adding preventive costs of 90. On the other hand, risk
costs—including both contingency and management
reserves—decreased from 736 to 557 because the contingency
reserve was reduced from 612 to 433, but the management
reserve 124 remains unchanged.
Applying the New Method to Ongoing Projects
This method was applied to five ongoing projects to verify the
improvements of the budget estimation method between pre-
liminary budgets using traditional methods and the re-
estimated budget using the proposed method with the estimate
at completion (EAC). A comparison table of cost variance
percentage is shown in Table 6. One project was completed
with the new estimation method and the other four projects are
Costs
1.0
Probabilities
612
(Contingency Reserve)
Pre-PE
Re-PE
Budget
Baseline
Project
Budget
Low
Value
High
Value
Preventive
Cost R
Value
Adaptive Cost
(Contingency Reserve)
Management
Reserve
Figure 4. Project budget with re-estimated risk response plan.
Costs
1.0
Probability
6,942.5 (75.6%)
(High Value) Confidence Level
7,908 (74.71%)
Management
Reserve
(124)
6,172
(PE) 612
(Contingency Reserve)
333
(R Value)
6,784 (71.41%)
(Budget Baseline)
6,609.5 (66.4%)
(Low Value)
Figure 3. Preliminary budgets with risk cumulative distribution curve.
10 Project Management Journal 50(1)
still in progress. The completion percentage technique was
applied to estimate the EAC. These selected five projects are
very similar to the previous 20 projects so that the results of the
re-assessments could be consistent and comparable.
A¼ðPreliminary budget EACÞ
Preliminary budget 100;
B¼ðRe estimated budget EACÞ
Re estimated budget 100
Discussion and Conclusions
Application of the proposed method to real projects was carried
out to demonstrate improvements in project budget accuracy
and precision. Dual budgeting was conducted for the compar-
isons between budgets using the traditional and proposed meth-
ods. One project was completed with the new estimation
method and the other four projects are still in progress. The
completion percentage technique was applied to forecast the
EAC. While the budget accuracy can be calculated by dividing
the differences between the budget and actual costs, precision
can be calculated by the variances or the standard deviation of
the differences. The accuracy of PE, RC, and the total prelim-
inary budget against EAC is –1.72, 24.34, and 0.78, with var-
iance (precision) 1.85, 20.97, and 1.15, respectively, whereas
the re-budgets are 0.25, 0.24, and 0.23, with variance (preci-
sion) 1.02, 19.53, and 0.35, respectively. The results show that
budget accuracy and precision on risk cost improved by 24.10
and 1.44, while improving by 0.55 and 0.48 percentage points
in total, respectively.
The project budget has two types of reserves against risks:
One is the contingency reserve for identified risks as an event,
and the other is the management reserve for unidentified risks
as a variability. However, previous researchers have presented
various methods for estimating reserves to cover the risk as an
event that can only be identified as an expected uncertainty.
Thus, the planned response actions could not cover the risk as a
variability, which is an unexpected uncertainty that cannot be
identified. Thus, the reserves estimated by the previous meth-
ods were enough to cover all risks. The management reserve for
unidentified risks must be estimated and included in the project
budget, even though unidentified risks could not be managed
by the project management team. Because unforeseen work
due to unidentified risk is also within the scope of the project,
Figure 5 describes the scope of a project and budget. When
work breakdown structures are developed, all works must have
their own budgets. The project budget must cover all the scope
of work for constructing the project’s result. Project scope
generally includes both certain and uncertain events. While the
certain events become the basis for developing the WBS, the
uncertain events become the basis of risk response plans. Thus,
the PE is the most highly accurate feature of the budget,
whereas the management reserve is the lowest accuracy and
the contingency reserve is moderate. One of the best ways to
improve the accuracy of the project budget is to transfer uncer-
tain scopes of work to certain scopes by analyzing and
Table 5. Analyzed and Re-estimated Response Plans
Response Plan Preventive Cost
Probability
(%) L M H Mean Adaptive Cost
R
(H-L) EMV
Improper estimate 0 30 78 129 174 127 39 29 39
Increasing material and equipment costs 10 40 95 135 175 135 54 32 64
Scope changes 40 70 135 168 215 173 121 56 161
Design changes 15 50 95 143 184 141 72 45 87
Poor communication 10 30 105 165 195 155 50 27 60
Vendor’s capability 15 30 280 320 385 328 99 32 114
Sum 90 788 1,060 1,328 1,059 433 220 523
Note. Higher values as bold entries are selected because of the low maturity level of the company.
Table 6. Results of Application to Ongoing Projects
Project Number
AB
PE RC Total PE RC Total
201 0.62 19.19 2.60 2.04 –7.36 1.28
202 –1.82 20.18 0.29 –0.29 3.39 0.00
203 –2.19 25.70 0.18 0.00 0.57 0.03
204 –2.83 26.53 –0.05 –0.19 1.34 –0.08
205 –2.40 30.12 0.86 –0.33 3.26 –0.07
Mean –1.72 24.34 0.78 0.25 0.24 0.23
Var (s2) 1.85 20.97 1.15 1.02 19.53 0.35
SD 1.36 4.58 1.07 1.01 4.41 0.59
Kwon and Kang 11
quantifying the uncertainty; for example, the reserves against
risks is the budget for the uncertainty. The reserves can be
divided into the preventive action and the adaptive action by
analyzing risks. The preventive action becomes the certain
scope while the adaptive action remains in the uncertain scope.
Thus, the works of preventive action against risk should be
moved to the WBS.
The project budget should be estimated and determined to
the approximate actual costs in order to minimize cost var-
iances. The actual costs of each budget component such as the
PE, the contingency reserve against identified risks, and the
management reserve against unidentified risks can be recorded
separately for project performance information. Analyzing and
evaluating the variance between budgets and the actual costs of
each budget component are essential to improving project cost
management. However, it is very difficult to evaluate and
analyze the variances due to lack of budget estimation of each
budget component by previous budget estimation method. The
advantage of the estimation method proposed in this article is to
estimate reserves as the contingency reserve and the manage-
ment reserve against both identified and unidentified risks sep-
arately to improve budget accuracy. This method can be used to
analyze the differences between budgets and actual costs and
obtain feedback for future projects as lessons learned. Project
management performance can be improved by conducting
these processes iteratively and updating the best practice.
Furthermore, re-estimating risk costs using a three-point esti-
mation technique by evaluating the R-value is suggested in
order to minimize cost variances due to risks. That is another
way to improve budget precision. The PE is the budget for the
certain scope and the basis for developing a funding and pay-
ment schedule to construct the project result during project
Project
Project Budget
Confidence
Level
CDF
PDF
Identified Risks
Events
Develop Response Action
Risk Register
Unidentified Risks
Variability
Risks
Contingency Reserve Management Reserve
Develop Cost Risk
Adaptive ActionPreventive Action
Point Estimate
Three Point Estimate
WBS
Fixed Scope
Reserves
Residual,
Secondary Risks
NO
NO
YES
YES Response
Action?
Identified?
Figure 5. Scope of a project and budget.
12 Project Management Journal 50(1)
execution, because the budget and schedule of the uncertain
scope are almost unchangeable. On the other hand, the risk cost
is the budget for the uncertain scope of the project. Thus devel-
oping the funding and payment schedule for the uncertain
scope is inaccurate. Therefore, the full amount of the risk cost
should be kept in the project fund over the entire project period
to avoid the lack of cash. It also incurs the capital costs, so that
the less the risk cost, the fewer the additional capital costs. We
verified this by applying the proposed method to five ongoing
projects. The results demonstrated an improvement in budget
accuracy and precision, with smaller cost variances between
the re-budget and EAC than between the preliminary budget
and the actual.
This research can extend in several directions. Cash flow
management is one of the most important determinants of
the success of construction project management. Poor cash
flow may result in inadequate working capital and thus
undermine the sustainability of a project. Thus, the method
describedinthisarticlecanbe applied to improve project
cash flow management.
This research mainly emphasized estimating a budget for a
single project, especially estimating reserves against both iden-
tified and unidentified risks to minimize cost variances by
improving budget accuracy and precision. Budget composi-
tions are provided—including net cost, allowance, contingency
reserve, and management reserve—and are to be controlled by
different people or departments to save cost by analyzing the
variances between budgets and actual costs. This method is a
more efficient way of conducting project cost management
than the current methods used in project cost management.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship,
and/or publication of this article.
References
Adoko, M. T., Mazzuchi, T. A., & Sarkani, S. (2016). Developing a
cost overrun predictive model for complex systems development
projects. Project Management Journal,46(6), 111–125.
Aibinu, A. A., & Jagboro, G. O. (2002). The effects of construction
delays on project delivery in the Nigerian construction industry.
International Journal of Project Management,20(8), 593–599.
Aziz, R. F. (2013). Factors causing cost variation for constructing
wastewater project in Egypt. Alexandria Engineering Journal,
52(1), 51–66.
Baccarini, D. (2006). The maturing concept of estimating project cost
contingency—A review. In G. Runeson & R. Best (Eds.), Proceed-
ings of the Australasian University building educators association
annual conference 2006: 31st Annual Conference, 11–14 July
2006. Sydney, New South Wales, Australia: University of
Technology.
Bannerman, P. (2008). Risk and risk management in software projects:
A reassessment. The Journal of System and Software,81(12),
2118–2133.
Barraza, G. A., & Bueno, R. A. (2007). Cost contingency manage-
ment. Journal of Management in Engineering,23(3), 140–146.
Becker, T. V., Jaselskis, E. J., & El-gafy, M. (2014). Improving pre-
dictability of construction project outcomes through intentional
management of indirect construction costs. Journal of Construc-
tion Engineering and Management,140(6), 401–414. doi:http://dx.
doi.org:10.1061/(ASCE)
Bedarida, S., & Conti, S. (2012). Cost estimating process and quality
assurance. Cost Engineering-Morgantown,54(3), 27.
Bhargava, A., Anastrasopoulous, P. C., Labi, S., Shiha, K.C., &
Mannering, F. L. (2010). Three stage least squares analysis of time
and cost overruns in construction contracts. Journal of Construc-
tion Engineering and Management,136(11), 1207–1218.
Book, S. A. (2007, October 11). Allocating risk dollars back to indi-
vidual cost elements. Department of the Navy Cost Analysis Sym-
posium, Marine Corps Base Quantico, Quantico, VA.
Caron, F., Ruggen, F., & Merli, A. (2013). A Bayesian approach to
improve estimate at completion in earned value management.
Project Management Journal,44(1), 3–16.
Carr, V., & Tah, J. H. M. (2001). A fuzzy approach to construction
project risk assessment and analysis. Advances in Engineering
Software,32(10), 847–857.
Cha, H. S., & Shin, K. Y. (2011). Predicting project cost performance
level by assessing risk factors of building construction in South
Korea. Journal of Asian Architecture and Building Engineering,
10(2), 437–444.
Chapman, C. (2000, June 2–4). Project risk management—The
required transformations to become project uncertainty manage-
ment. PMI Research Conference, Paris, France.
Chenyun, Z. Y. (2012). The BP artificial neural network model on
expressway construction phase risk. Systems Engineering Proce-
dia,4, 409–415.
Clark, D. E. (2001). Monte Carlo analysis: Ten years of experience.
Cost Engineering,43(6), 40–45.
Creedy, G. D., Skitmore, M., & Wong, J. (2010). Evaluation of risk
factors leading to cost overrun in delivery of highway construction
projects. Journal of Construction Engineering and Management,
136(5), 528–537.
Dikmen, I., Birgonul, M. T., & Han, S. (2007). Using fuzzy risk assess-
ment to rate cost overrun risk in international construction projects.
International Journal of Project Management,25(5), 494–505.
Doloi, H. (2013). Cost overruns and failure in project management:
Understanding the roles of stakeholders in construction projects.
Journal of Construction Engineering and Management,139(3),
267–279.
Doloi, H., Sawhney, A., & Rentala, S. (2012). Analyzing factors
affecting delays in Indian construction projects. International
Journal of Project Management,30(4), 479–489.
Eldosouky, A. A., Ibrahim, A. H., & Mohammed, H. E. (2014).
Management of construction cost contingency covering upside
and downside risks. Alexandria Engineering Journal,53(4),
863–881.
Kwon and Kang 13
Elkjaer, M. (2000). Stochastic budget simulation. International Jour-
nal of Project Management,18(2), 139–147.
Enshassi, A., Al-Najjar, J., & Kumaraswamy, M. (2009). Delays and
cost overruns in the construction projects in the Gaza Strip. Journal
of Financial Management of Property and Construction,14(2),
126–151.
Flyvbjerg, B., Holm, S. M., & Buhl, S. (2002). Understanding costs in
public works projects: Error or lie? Journal of the American Plan-
ning Association,68(3), 279–295.
Frimpong, Y., Oluwoye, J., & Crawford, L. (2003). Causes of delay
and cost overruns in construction of groundwater projects in devel-
oping countries: Ghana as a case study. International Journal of
Project Management,21(5), 321–326.
Garvey, P. R. (2008). A scenario-based method for cost risk analysis.
Journal of Cost Analysis and Parametrics,1(1), 65–76.
Halawa,W.S.,Adbelalim,A.M.K.,&Elrashed,I.A.(2013).
Financial evaluation program for construction projects at the
pre-investment phase in developing countries—A case study.
International Journal of Project Management,31, 912–923.
Hwang, B. G., Zhao, X., See, Y. L., & Zhong, Y. (2015). Addressing
risks in green retrofit projects: The case of Singapore. Project
Management Journal,46(4), 76–89.
Idrus, A., Nuruddin, M. F., & Rohman, M. A. (2011). Development of
project cost contingency estimation model using risk analysis and
fuzzy expert system. Expert Systems with Applications,38(3),
1501–1508.
Kasimu, M. (2012). Significant factors that cause cost overruns in
building construction projects in Nigeria. Institute of Interdisciplin-
ary Journal of Contemporary Research in Business,3(11), 775–780.
Ke, Y. J., Ling, F. Y. Y., & Ning, Y. (2013). Public construction
project delivery in Singapore, Beijing, Hong Kong, and Sydney.
Journal of Financial Management of Property and Construction,
1(18), 6–15.
Khodakarami, V., & Abdi, A. (2014). Project cost risk analysis: A
Bayesian networks approach for modeling dependencies between
cost items. International Journal of Project Management,32(7),
1233–1245.
Kim, G. H., Kang, K. I., & An, S. H. (2004). Comparison of construc-
tion cost estimating models based on regression analysis, neural
networks, and case-based reasoning. Building and Environment,
34(2), 34–47.
Kitchenham, B., & Linkman, S. (1997). Estimates, uncertainty and
risks. IEEE Software,14(3), 69–73.
Koushki, P. A., Al-Rashid, K., & Kartam, N. (2005). Delays and cost
increases in the construction of private residential projects in
Kuwait. Construction Management and Economics,23(3), 285–294.
Lai, Y. T., Wang, W. C., & Wang, H. H. (2008). AHP and simulation-
based budget determination procedure for public building con-
struction projects. Automation in Construction,17(5), 623–632.
NASA. (2008). 2008-NASA-Cost-Handbook-FINAL_v6[1]. Retrieved
from https://www.nasa.gov/pdf/263676main_2008-NASA-Cost-
Handbook-FINAL_v6.pdf.
Nawaz, T., Shareef, N. A., & Ikram, A. A. (2013). Cost performance
in the construction industry in Pakistan. Industrial Engineering
Letters,3(2), 19–33.
Memon, A. H., Rahman, I., & Azi, A. (2012). Time and cost perfor-
mance in construction projects in southern and central regions of
peninsular Malaysia. International Journal of Advances in Applied
Sciences,1(1), 45–52.
Moselhi, O. (1997, July 13–16). Risk assessment and contingency
estimating. AACE transactions, Dallas, TX.
Oberlender, G. D., & Trost, S. M. (2001). Predicting accuracy of early
cost estimates based on estimate quality. Journal of Construction
Engineering Management,127(3), 173–182.
Pfleeger, S. L., & Atlee, J. M. (2006). Software engineering, theory
and practice (3rd ed., Pearson International ed.). Delhi, India:
Pearson Education India.
Project Management Institute (PMI). (2013). A guide to the project
management body of knowledge (PMBOK
®
guide) – Fifth edition.
Newtown Square, PA: Author.
Sambasivan, M., & Soon, Y. W. (2007). Causes and effects of delays
in the Malaysian construction industry. International Journal of
Project Management,25(5), 517–526.
Sato, T., & Hirao, M. (2013). Optimum budget allocation method for
projects with critical risks. International Journal of Project Man-
agement,31(1), 126–135.
Schwartz, M. (2010). The Nunn-McCardy act: Background, analysis,
and issues for Congress. Washington, DC: Congress Research
Service.
Smith, P. (2014). Project cost management—Global issues and chal-
lenges. Social and Behavior Sciences,119, 485–494.
Smith, P. G., & Merrit, G. M. (2002). Proactive risk management.
New York, NY: Productivity Press.
Sweis,G.H.,Sweis,R.,Rumman,M.A.,Hussein,R.A.,&
Dahiyat, S. E. (2013). Cost overruns in public construction
projects—The case of Jordan. Journal of American Science,
9(7), 134–141.
Tabish, S., & Jha, K. (2011). Important factors for the success of
public construction projects. 2nd International Conference on Con-
struction and Project Management IPEDR (15), IACSIT Press,
Singapore.
Thamhain, H. (2013). Managing risks in complex projects. Project
Management Journal,44(2), 20–35.
Thomson, P. A., & Perry, J. G. (1992). Engineering construction risks.
London, England: Thomas Telford.
U.S. Government Accountability Office. (2013). Reports to congres-
sional committees, NASA assessment of selected large-scale proj-
ects (GAO-13-276SP). Washington, DC: Author.
Uzzafer, M. (2013). A contingency estimation model for software
projects. International Journal of Project Management,31(7),
981–993.
Williams, T. (2016). Identifying success factors in construction
projects: A case study. Project Management Journal,47(1),
97–112.
Xenidis, Y., & Stavrakas, E. (2013). Risk based budgeting of
infrastructure projects. Social and Behavior Sciences,74,
478–487.
Zhu, B., Zhang, H., & Wang, X. (2011). Analysis and evaluation of
project cost risk based on BP algorithm. Systems Engineering Pro-
cedia,1, 264–270.
14 Project Management Journal 50(1)
Author Biographies
Hyukchun Kwon, PhD, is an adjunct professor in project man-
agement in the Graduate School of Engineering at Hanyang
University, Republic of Korea. He earned his master’s degree
in project management at the University of Alaska, a PhD
degree from Hanyang University, and he is also a CPA. He has
over 25 years of experience managing projects in various busi-
ness areas, including sales and marketing projects, manufactur-
ing and constructing projects for heavy industries and plants,
accounting services and financial consulting, and performing
business process reengineering projects for ERP implementa-
tion. He can be contacted at hkwon21@hanyang.ac.kr
Chang Wook Kang, PhD has been a professor in the Depart-
ment of Industrial and Management Engineering, Hanyang
University ERICA Campus, South Korea since 1991 and has
held the Hanyang University title of Distinguished Teaching
Professor since 2013. His fields of research are statistical pro-
cess control and project quality management. In 2006, Profes-
sor Kang created the master of science degree in project
management and served as chair of the Korean committee of
ISO21500/PC236. He served as the president of the Korea
Society of Industrial and Systems Engineering from 2002 to
2006, and from 2010 to 2012, he served as the founding pres-
ident of the Korean Society of Project Management. He served
as the dean of the College of Engineering Sciences from July
2016 to June 2018. He received his BS in industrial engineering
from Hanyang University in 1981 and his BA in statistics from
the University of Minnesota in 1984. He received MS and PhD
degrees in statistics from the University of Minnesota in 1988
and 1990, respectively. He can be contacted at cwkang57@
hanyang.ac.kr
Kwon and Kang 15
... A research conducted on Factors affecting material management in construction industry which aimed to identify and rank the most significant factors related to material management by concluded that major factors for effective material management were material requirements planning, proper cash flow control, identifying and selecting suppliers, organizing and scheduling the procurement, skilled negotiation with suppliers, non-delay of payments, skill and experience of material management team, planning delivery of materials to site, proper inspection and documentation of materials and site location and layout. A research by (Kang, 2018) showed delays and cost overruns increased in the construction of a private residential project in Kuwait due to three main factors: were contractor-related problems, material related problems, and owners" financial constraints. ...
... Contractor-related problems, material related problems, and owners" budget/financial constraints/ related problems were the three main factors for delay & cost overrun in the construction projects of private residential projects in Kuwait (Kang, 2018). Similarly A study by Nasir (2013), as cited by (Gurmu, 2018) stated materials purchasing procedure was found to be one of the significant practices for increasing the productivity of infrastructure projects in the US. ...
Thesis
Full-text available
This study aimed to examine factors that affect material management of construction projects in Gurage zone case of Wolkite town. An explanatory research design was employed in this study. Only primary source of data was used and structured questionnaire was applied to collect the data. The study population includes all permanent workers of contractors, Urban Development & Construction Department officials& experts and client office. The data was collected and analysed by the use of- Statistical Package for Social Sciences (SPSS V.20) for inferential & descriptive technique. The findings of the study reveals that, purchasing procedure from the bid up to material kept in the store, budget readiness for material procurement and inventory control have significant relationship on material management of construction projects in the study area at 5% significance level. In this study other variables such as availability of materials, qualified personnel and construction facilities have no significant relation with dependent variable. The study recommended that, the government bodies and contractors improve the proper purchase, budget and inventory level in the study area. For these significant variables have to be responsive for their individual responsibility for better material management of construction projects in the study area.
... calculate the project risk by adding the exposures of all the identified risks to draw up an efficient schedule and to effectively prepare budgets. Other recent works also resort to Monte Carlo Simulation to calculate the total project risk to determine time and cost contingencies(Allahi, Cassettari, & Mosca, 2017;Eldosouky, Ibrahim, & Mohammed, 2014;El-Kholy, Tahwia, & Elsayed, 2020;Kwon & Kang, 2019;Traynor & Mahmoodian, 2019). ...
Article
Full-text available
Obtaining a viable schedule baseline that meets all project constraints is one of the main issues for project managers. The literature on this topic focuses mainly on methods to obtain schedules that meet resource restrictions and, more recently, financial limitations. The methods provide different viable schedules for the same project, and the solutions with the shortest duration are considered the best-known schedule for that project. However, no tools currently select which schedule best performs in project risk terms. To bridge this gap, this paper aims to propose a method for selecting the project schedule with the highest probability of meeting the deadline of several alternative schedules with the same duration. To do so, we propose integrating aleatory uncertainty into project scheduling by quantifying the risk of several execution alternatives for the same project. The proposed method, tested with a well-known repository for schedule benchmarking, can be applied to any project type to help managers to select the project schedules from several alternatives with the same duration, but the lowest risk.
Article
The paper formulates methodological approaches to risk assessment in the field of financial support for the achievement of Russia's national development goals. Criteria of probability of realization of risk scenarios are analyzed and an assessment of the degree of their influence is given. Proposals have been formulated to improve the quality of work with risks and to ensure automation of detection and initial response to the occurrence of negative facts.
Chapter
Success of projects is measured in terms of achieving the projects’ objectives. In almost every project, the main objective includes compliance with the project planned duration and budget. Therefore, time and cost management of project activities is important in any project. Project activities can often be performed in one of several alternatives, also known as modes or options, each with its unique duration and cost requirements. A common practice in project time and cost management consists of balancing between activities’ completion time and cost to ensure the project’s timely completion within the project budget. For that reason, project managers look for a set of modes that result in a desirable tradeoff between the duration and cost of activities. The time–cost tradeoff problem (TCTP) is an optimization problem aiming to find such a tradeoff. The TCTP has been extensively studied, resulting in several variants, as well as many exact and nonexact solution techniques. In this chapter, the TCTP is reviewed from an optimization standpoint: Basic models are discussed, and major solution techniques from the literature are explained.
Article
Full-text available
Project is an activity or business that is temporary and has a certain period of time, unique and produces a product, service or both. Achieving project activity right on the schedule was an important thing for project owners or contractors. However, the schedule of the "X2" Well Drilling Project has delayed from the original plan due to several inhibiting factors. Therefore, the cost's project increased during project implementation. The factor analysis method is used to reduce the number of independent variables until new variables are found which are then examined for their relationship and a mathematical model is built using the regression analysis method. The location of this research was one company with 100 valid respondents. The results of this study indicate that there are 3 (three) new factors that have a positive effect and 1 (one) new factors that have a negative effect on the condition of the differences that occur. If a project can be completed before the work period ends, at other times if the owner wants to build a construction project again, there is no doubt that the owner will re-choose Oil Service Company to work on another Drilling Project.
Article
Full-text available
Defining "project success" has been of interest for many years, and recent developments combine multiple measurable and psychosocial factors that add to this definition. There has also been research into success factors, but little research into the causal chains through which success emerges. Following the multi-dimensionality of "success," this article shows how success factors combine in complex interactions; it describes factors contributing to project performance by a company working on two major construction programs and shows how to map and analyze paths from root causes to success criteria. The study also identifies some specific factors-some generic, some context-dependent-none of these is uncommon but here they come together synergistically.
Article
Full-text available
According to the feature of expressway construction phase risk, an evaluation index system of expressway engineering construction phase risk was proposed. The BP artificial neural network model can be effectively used in engineering risk evaluation was built. The BP artificial neural network evaluation of the basic principles and process of engineering risk evaluation model, including the establishment of neural networks, learning and training, test results and case analysis, was explained.
Chapter
On September 8, 1982, President Ronald Reagan signed into law the Department of Defense (DOD) Authorization Act, 1983 (P.L. 97-252), which included what has come to be known as the Nunn-McCurdy Act (10 U.S.C. § 2433). The Nunn-McCurdy Act requires DOD to report to Congress whenever a major defense acquisition program experiences cost overruns that exceed certain thresholds. The purpose of the act was to help control cost growth in major defense systems by holding the appropriate Pentagon officials and defense contractors publicly accountable and responsible for managing costs. A program that experiences cost growth exceeding any of the established thresholds is said to have a Nunn-McCurdy breach. There are two types of breaches: significant breaches and critical breaches. A "significant" breach is when the Program Acquisition Unit Cost (the total cost of development, procurement, and construction divided by the number of units procured) or the Procurement Unit Cost (the total procurement cost divided by the number of units to be procured) increases 15% or more over the current baseline estimate or 30% or more over the original baseline estimate. A "critical" breach occurs when the program acquisition or the procurement unit cost increases 25% or more over the current baseline estimate or 50% or more over the original baseline estimate. The Nunn-McCurdy Act has been statutorily amended a number of times over the years. One of the most significant changes to the reporting requirements occurred in the FY2006 National Defense Authorization Act (P.L. 109-163), when Congress added the original baseline estimate as a threshold against which to measure cost growth. The new standard prevents DOD from avoiding a Nunn-McCurdy breach by simply re-baselining a program. Another significant change occurred in the FY2009 Weapon Systems Acquisition Reform Act (P.L. 111-23), when Congress enacted a requirement that programs with critical breaches should be presumed terminated unless the Secretary of Defense certifies the program. For programs that are certified, DOD must (1) revoke the prior milestone approval, (2) restructure the program, and (3) provide Congress a written explanation of the root-cause of the cost growth. These changes were fueled in part over Congressional concern that programs with chronic cost growth and schedule delays were not being terminated and Congress was not being provided specific information explaining what caused the cost growth. Some analysts believe that the Nunn-McCurdy Act has been effective as a reporting mechanism for informing Congress of cost overruns in major acquisition programs. As a result of the Nunn-McCurdy process, Congress has increased its visibility into the cost performance of the acquisition stage of MDAPs. However, some analysts suggest that Nunn-McCurdy is not a sufficiently comprehensive reporting mechanism because program managers can sometimes take steps to avoid a breach and because Nunn-McCurdy does not apply to all elements of a weapon system's life-cycle costs, such as its operations, support, or disposal costs.
Article
While system complexity is on the rise across many product lines, the resources required to successfully design and implement complex systems remain constrained. Because financiers of complex systems development efforts actively monitor project implementation cost, project performance models are needed to help project managers predict their cost compliance and avoid cost overruns. This article describes recent research conducted by the authors to develop a cost overrun predictive model using five known drivers of complex systems development cost. The study identifies schedule and reliability, as the key determinants of whether or not a large complex systems development project will experience cost overrun.
Article
The objectives of this study are to identify the risks in green retrofit projects in Singapore; analyze their risk criticalities; compare the risk criticalities between conventional and green retrofit projects; and provide mitigation measures for the critical risks. Twenty risks and 37 mitigation measures were identified from a literature review. A questionnaire survey was performed with 30 professionals experienced in green retrofits, and five post-survey interviews were conducted. The results indicated “post-retrofit tenants’ cooperation risk” was the top risk, and that 19 risks were more critical in green retrofits than in conventional retrofits. Additionally, 28 mitigation measures obtained significant agreement.
Article
This article shows actual project results that support the validity of the Monte Carlo cost risk/opportunity method. Also presented are Honeywell's experiences applying it to capital projects. Actual cost results from 19 completed projects totaling $312 million show this method effective in identifying project risk/opportunity elements and quantifying contingency. The projects are Honeywell chemical process and plastic fiber/film making plants built in the USA and China in the last 10 years. All analyses were prepared before the projects were executed.