Alexandria Engineering Journal 78 (2023) 88–103

Contents lists available at ScienceDirect

Alexandria Engineering Journal

journal homepage: www.elsevier.com/locate/aej

Original Article

Appropriate budget contingency determination for construction projects:

State-of-the-art
Taher Ammar *, Mohamed Abdel-Monem, Karim El-Dash
Faculty of Engineering at Shoubra, Benha University, Postcode 11629, Egypt

A R T I C L E I N F O A B S T R A C T

Keywords: Contingency is a critical component in the cost estimation process for any construction project. The contingency
Construction project reserve considers potential costs related to risks and uncertainties associated with construction projects. It is
Cost estimation usually assumed to damp any resulting uncertain monetary impact and to prevent project cost overrun. Many
Cost contingency
contingency calculation methods for construction projects proposed in literature ranged from simple percentage
Estimation method
to complex mathematical methods. Deciding the optimum contingency method for a given project at a given
phase represents the main challenge in cost estimation process. This study presents a comprehensive compilation
of all contingency calculation methods and divided them into three main groups: deterministic, probabilistic, and
modern mathematical methods which have been discussed in details. Appropriate method for estimating con­
tingency amount depends on many criteria such as project peculiarity, complexity, ease of method used, and the
accuracy level of the estimates. This study proposed a practical guidance approach for construction agencies to
choose their appropriate cost contingency method.
This research is expected to help agencies/owners in the budget development stage to allocate a contingency
budget for their construction projects.

1. Introduction project are tricky obstacles in calculating an accurate cost estimate.

Therefore, a contingency reserve is usually added to the total project
At the appraisal stages, the cost is a crucial component of a project budget to accommodate the impact of such risks and uncertainties [8].
budget. Risk management is one of the most important tasks that faces Cost contingency estimation is an essential phase in risk manage­
planners in the construction field. Risk management is the process of ment, especially for stochastic performance regimes [9]. Project man­
maximizing positive possibilities and reducing negative possibilities in agers and their hierarchical heads are mainly responsible for
order to improve the chances of project success. It includes the following determining cost contingencies [10]. The contingency reserve is the
processes; risk planning, risk identification, risk analysis, risk response, amount of time, money, or budget required in excess of the project cost
and risk monitoring [1,2]. Cost contingency is one primary tool to estimate to reduce the risk of overruns of the project goals [1].
mitigate/consider risks for construction projects [3]. Contingencies are generally applied to reduce cost overruns in a
Highly accurate cost estimation is vital in the early phases of a project which implemented under differnt unexpected circumstances,
construction project and is a critical success criterion [4,5]. Estimating and contingency provisions usually varies according to the project
construction costs and predicting price increase are the major challenges [11,12].
facing project owners [6]. Owners are usually keen to get accurate early Contingencies consider identifiable risks and unexpected events that
estimates of the cost of their construction projects to provide adequate may arise during different project phases. At the beginning of a project,
budgets [7].Risks and uncertainties associated with any construction the contingency allowance is relatively large owing more uncertainties

Abbreviations: AHP, analytical hierarchy process; ANN, artificial neural network; BBN, Bayesian belief network; CBR, case-based reasoning; EJ, expert judgment;
EV, expected value; MCS, Monte Carlo simulation; MLR, multiple linear regression; MOM, method of moments; NN, neural network; PSO, particle swarm optimi­
zation; R2, coefficient of determination; RA, regression analysis; RBF, radial basis function; RCF, reference class forecasting; RE, range estimating; SVM, support
vector machine; TOC, theory of constraints.
Peer review under responsibility of Faculty of Engineering, Alexandria University.
* Corresponding author.
E-mail addresses: [email protected] (T. Ammar), [email protected] (M. Abdel-Monem), [email protected] (K. El-Dash).

https://doi.org/10.1016/j.aej.2023.07.035
Received 5 November 2022; Received in revised form 27 June 2023; Accepted 15 July 2023
Available online 22 July 2023
1110-0168/© 2023 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria University. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Table 1
Main contingency estimation methods [20].
Groups/Classes Methods Researcher

Deterministic Traditional / Percent [3,10,19,24,25,26]

Expert Judgment [20,27,28,29]
Probabilistic Range EstimatingSimulation [3,8,20,30]
(Monte Carlo)
Integrated Model for Cost and Schedule
[3,38,31,32,33,34,35]

Expected Value [3,8,31,36,37,38]

Methods of Moment [3,39,40]

Probability Tree [20,41]

Program Evaluation and Review Technique

[3,35,42,43]

Parametric Estimating
[16,44]

Regression Analysis
[3,6,20,28,35,45,46,47,48,49]
Analytical Hierarchy Process
[3,28,35,50,51,52,53,54,55]

Reference Class Forecasting

(Optimism Bias Uplifts) [28,55,57,58,59]

Modern Mathematical Fuzzy Techniques

[3,28,35,60,61,62,63,64,65,66,67,68]

Artificial Neural Network

[2,6,28,69,70,71,72,73,74,75,76,77,78,79]

and is gradually reduced to adapt the costs incurred during project policies of most construction organizations, the project contingency
phases. There are two fundamental types of contingencies exist in any percentage is often set by estimators and the management department
construction project; (1) design contingencies assigned for design changes [19]. Many previous studies have proposed several methods, models,
and modifications, considering some factors, such as data, lack of clarity and sliding scales considering risks and uncertainties in construction
of the scope, and inaccurate estimation methods [13], and (2) Con­ projects [20]. Some of these models and methods have been charac­
struction contingencies consider changes during the construction process terized by the complexity in their implementation in real-world projects.
wherein unresolved design issues before awarding the contract are In addition, their techniques and knowledge acquisition process are
incorporated into the estimated project cost [14]. time-consuming. However, rationality, flexibility, swiftness, ease of use,
Accurate cost estimations are also required by construction agencies and accuracy are different criteria for selected method. They also focus
prior to contract engagements to reach the best decision during the on the best prediction of the tender price to produce consistent results
feasibility stage of the proposed project. Flyvbjerg and Dirk [15] re­ [20].
ported that 90% of infrastructure projects experienced actual cost Recently, many agencies have begun conducting formal probabilistic
overruns in terms of estimated costs, and road projects were subject to risk assessments to estimate a contingency budget rather than adopting
an average cost escalation of 20%. To add contingency reserves to their deterministic approaches [21]. To develop a contingency budget, a
project budgets, owners must learn how to select contingencies and company must make a tremendous effort to allocate an improved budget
assess and evaluate their amounts. estimate, which becomes more important when dealing with concurrent
Contingency reserves are added to the total project budget, which projects simultaneously. On the other hand, allocating an excess budget
consists of two separate parts: base cost and contingency cost. The for any project consumes parts of the money allocated for other projects.
former does not include contingency. It has been considered that the The objectives of this study are threefold:
difficulty of accurate cost estimation stems from the desire to be able to
predict a somewhat uncertain future state [16]. (a) present the contingency calculation methods most commonly
A risk assessment process can be used to determine the likelihood of used by many agencies in different construction projects;
adverse events occurring, as well as the potential impact of these events (b) review, compile, and discuss the advantages and disadvantages of
during project construction. Therefore, the estimated impact can be the main methods; and
added as a contingency amount to the project base estimate. Tolerance (c) introduce previously applied methods in studies to help estima­
in the specifications, budget money, and schedule float are the three tors select the appropriate method for their own construction
basic types of general contingencies in construction projects [17]. projects.
Cost contingency for a construction project is an estimate of the costs
associated with risks and identified uncertainties and added to the base
estimate to obtain the total project cost estimate [18]. According to the

89
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

2. Literature review entire test sample.

In USA, Diab et al. [49] developed regression model to analize the
In planning construction projects, contingency funds and manage­ dependency relationship between predetermined cost contingency
ment reserves are assets held as reserves to pay for mandatory and amounts for highway projects and risk drivers’ perceived ratinus
voluntary changes initiated by the owner or contractor after awarding assessed by professionals. The study concluded that inadequate con­
the construction contract [22,23]. Many methods were introduced for structablity reviews had a significant impact on the determination of
contingency estimation and allocation. The main contingency estima­ owner’s contingency amount, whereas changes by owner’s request
tion techniques are divided into three main classes, as shown in Table 1 impacted contingency amount for both owner and contractor. Whereas,
[20]. Lam et al. [85] used regression analysis model for the estimation and
Lhee et al. [80] proposed an artificial neural network (ANN) model management of contingency funds in the public works in Zambia. They
to estimate the optimal contingency costs for transportation construc­ proposed a model to identify the risk factors at the construction phase
tion projects. Their method improved the owner’s budgeting decisions that cause project cost overruns, and to establish a reliable method for
and reduced the risk of overcommitment or underutilization of funds estimating the contingency amount. Regression analysis demonstrated
[77]. that the contingency amount was correlated positively to the estimated
In Kenya, Shrestha et al. [71] developed an ANN model to estimate contract budget. The findings of the qualitative interview and Pearson
cost contingency for a road maintenance p as a mathematical model. correlation analysis also showed that the total contingency amount and
Their model predicts contingency costs for each road maintenance ac­ project complexity tended to be highly correlated. They also proved that
tivity included in the contract based on historical records of change the type of work is not a significant direct factor.
order data. Since this method calculates contingency based on the pre­ Regression analysis is the most powerful applied method in cost
vious historical cost weight of the maintenance activities, the calculated estimation studies for construction projects [28].
contingency for a contract was more accurate and scientific than Additionally, there are other types of cost estimation methods called
adopting the traditional approach. Based on the cost weighting value of “Mixed and Hybrid “, SVM [86], BBN [87], CBR [88], Factor Rating [89],
each maintenance activity, the total cost contingency amount of the Controlled Interval and Memory [90], Influence Diagram [91], and TOC
project was determined. [92].
Idrus et al. [81]estimated contingency costs using risk analysis and a Also, there are differen cost contingency methods mainly focusing on
fuzzy expert system as a logical and flexible approach to accommodate manufacture of construction products [16,27,69, and 93] including;
contractors’ subjective judgment. Compared to the actual cost contin­ function costing, feature costing, generative costing, activity-based
gencies, their model prediction had about 20% accuracy. Salah & costing, group technology, knowledge-based engineering system, top-
Moselhi [60], Nawar et al. [61], Afzal et al. [35] and Alfaggi & Naimi down processing, bottom-up processing, operation-based approach,
[67]used fuzzy technique to estimate project costs and cost contingency breakdown approach, tolerance-based cost models, decision support
for many types of construction projects in Canada, Egypt, Chaina and systems, rule-based systems. All these methods will discussed in the
Libya respectively. follwing sections.
Many models were developed to manage and estimate project cost The selection of the most appropriate method of contingency
contingency. For example, Barraza and Bueno [17]provided a method calculation for a specific project is depending on many factors such as
for calculating and managing contingency costs based on Monte Carlo the project’s type, size; uncertainty level, criticality; and complexity.
simulations (MCSs) to control target costs and keep actual project costs Other factors such as the time available for the bidding period and the
below them. Their method provided a performance status rating that can experience of the estimators also plays an important role in selecting the
be used as a practical guide for decision-making. Although the method appropriate contingency calculation method [94].
can be used for any project, it does not indicate how contingency should Other criteria for selecting best method including flexibility, ratio­
be allocated to each individual activity [82]. nality, accuracy, swiftness, and ease of use should be considered to get
Hammad et al. [82] presented MCS model for a real bridge main­ the best bid price indicators and achieve consistent results [20].
tenence project to estimate cost contingency during the planning stage For example, Salah and Moselhi [60] suggested four main factors
as well as managing cost contingency during the constructuin stage by affect the strategy of cost contingency selection method by construction
assigning the probability distribution to all project activities and organizations:
selecting the appropriate confidence interval. New measures have been
introduced into their proposed model to control expenditure, balance of (1) Project type which is a key factor since contingency management
contingency costs for activities, and to estimate future needs. Findings could be different.
showed that the proposed cost contingency model is adaptable and (2) Project complexity can be related to the type of project, but it can
robust enough to accommodate the typical disturbances in the project be also based on technical difficulties, fast-track system, or
environment. However, El-Kholy et al. [3] stated that the use of simu­ modern technology.
lation in most cases is cumbersome, difficult, and rarely used in the (3) Risk impact is considered a crucial factor-affecting contingency.
construction field. (4) Risk type which can be financial, political, schedule, resources, or
EI-Touny et al. [83] estimated cost contingency using analytical hi­ country-based.
erarchy processes (AHPs) to identify the impact of risk factors during the
bidding stages. They developed a reliable model for estimating the ex­ Islam et al. [95] stated that the selection of the appropriate contin­
pected contingency costs for Egyptian highway construction projects. gency method for construction projects based on:
Their model demonstrated that the predicted cost contingency matches
96.31% of the average estimated contingency for real case projects, and (1) Careful project’s variables study such as project type, size, and
the model showed robust results. contract.
Alfred et al. [84] proposed a model to identify critical factors that (2) Constraints such as data availability and limitation of the
may serve as indicators of potential cost overruns. They used data of method.
USAF construction projects across a full range of project types and (3) Level of use such as bidding, cash flow analysis, or progress
project scopes. They used multiple linear regression (MLR) for data monitoring.
analysis and compared the proposed model to the common practice of
assigning contingency funds. Applying the MLR model reduced the Therefore, construction agencies recommend that the peculiarity of
average contingency budget error from − 11.2% to − 0.3% over the each project, complexity, ease of use for the contingency calculation

90
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Fig. 1. Research methodology.

Fig. 2. Initial and basic cost estimation techniques.

method, and the needed accuracy level of the estimates should be the contingency estimating methods in construction projects. A compre­
basis for choosing the appropriate method for estimating contingency hensive literature survey was performed to evaluate the previously re­
amount [37]. ported results for examining and assessing various contingency
estimation methods. This study is developed in two main phases. The
3. Research methodologies first phase is exploratory, intended to understand the phenomenon of
cost contingency estimate. It consists of an extensive literature review of
This study deelpy analyzed the previous literature reviews regarding most cost contingency techniques used in the construction industry. This

91
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

assessment of contingency amount after risk assessment evaluation;

and
post-hazard contingency modeling management plan.

Tayefeh et al. [28] based on the analysis of previous applied methods

in last thiry years concluded that analytical approach is the least used
method for cost estimation (1%). Intuitive (3%), Parametric (28%), and
Analogous (68%) are the most used approaches respectively as shown in
Fig. 3.
They also categorized each cost estimation method according to
project type in the construction industry as shown in Table 2. The most
common methods that have been used in these projects were ANN and
RA as shown in Table 2.
The three main groups for contingency calculation methods are
Fig. 3. Distribution of cost estimation method by approach types [28].
deterministic, probabilistic, and modern mathematical methods [20].

includes key aspects, attributes, and challenging ideas provided by 3.1.1. Deterministic methods
previous researchers for allocating, managing, and estimating contin­ Deterministic techniques are the simplest and most commonly used
gency methods for construction projects. The second phase highlights the methods to establish a contingency budget for a project [19]. Deter­
strengths and weaknesses of each method and technique, offering more ministic technique is a traditional method, which has been mostly used
extensive and in-depth qualitative data from previous studies applied to to estimate contingency risks for the construction projects. It is usually
construction projects using different cost contingency methods. Fig. 1 express risk contingency in terms of a certain percentage of the contract
illustrates the processes and relevant methods used in this study. The budget [94].
research methodology includes literature review, contingency esti­ DOI [40] found that deterministic contingency calculation methods
mating methods, advantages and disadvantages of some methods, pre­ include deterministic range-based, factor-based, and reference class
viously applied studies using some methods, and results and discussion. forecast methods. These methods typically use judgment or statistical
analysis based on previous assets, experts, or project experiences. The
3.1. Main contingency estimation methods deterministic technique can be summarized into two main categories:

Construction agencies usually need an accurate early cost estimate to A. Predefined Percentages (Fixed/Line Items) or Traditional Method/
provide an adequate project budget. Uncertainties and risks associated Percent
with the project present impediments to obtaining an accurate cost es­
timate. Niazi et al. [27] divided the project cost estimation techniques In this method, contingency is usually calculated as an inclusive
into qualitative and quantitative, whereas Hueber et al. [16] and Curran additional percentage over base estimation [19]. This method usually
et al. [69] classified the basic cost estimation techniques into three makes use of an across-the-board addition on a base estimate by his­
groups: analogous, parametric, and bottom-up, as shown in Fig. 2. torical data, previous experience, or mere intuition [10].
A contingency budget is added to the total project budget to
accommodate the cost impact of expected (or not) risk factors. Many D. Expert Judgment (EJ) / Engineering Judgment
estimation techniques were used to calculate cost contingency for con­
struction projects [82]. The calculation of contingency allowance should In this method, an expert or a group of experts in risk analysis and
align with the project risk profile, complexity, and benchmarks based on risk management defines the percentage of contingency for the project
past project cost performances and the lifecycle stage of the investment [20].
[96]. Agencies seek an appropriate method for their projects, especially According to Niazi et al. [27], it is a qualitative/intuitive decision
at appraisal stages, depending on previous studies, historical data, and support system. EJ is an expression of opinion made by a knowledgeable
professional experts. and experienced person or group to find solutions. It uses many tools
Islam et al. [95] classified the contingency modeling techniques into such as the Delphi technique, interviews, historical data, and
three main categories: brainstorming.

prediction models using ANN, MCS, probability distribution, and 3.1.2. Probabilistic methods
regression analysis (RA) without risk analysis; The main difference between deterministic and probabilistic
methods is that uncertainties are explicitly modeled in probabilistic

Table 2
Distribution of cost estimation method by project type [28].
Methos

ANN Fuzzy SVM CBR PSO RBF RA Other Methods

Application area
Buildings 42% 5% 3% 2% 1% 1% 23% 23%
Highways 55% 0% 0% 0% 0% 0% 18% 27%
Public Projects 20% 0% 0% 0% 0% 0% 40% 40%
Water-related projects 0% 0% 0% 0% 0% 0% 0% 100%
Roadways 43% 0% 14% 0% 0% 0% 14% 29%
Tunnels 33% 0% 0% 0% 0% 0% 33% 34%
Railways 33% 0% 0% 0% 0% 0% 33% 34%
Hydropower 33% 0% 0% 0% 0% 0% 67% 0%
Power Plants 0% 0% 0% 0% 0% 0% 0% 100%

92
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

methods using appropriate statistical distributions. Probabilistic method • Method of Moments (MOM)
deals with assigning project cost components to probability distribution
functions PDF and thus generating a PDF of the overall project cost This is a method of estimating population parameters such as mean,
through an aggregate process [94]. median, and variance by equating the unobserved population moments
Probabilistic cost estimation techniques focus on risks and un­ with the sample moments and then solving those equations for the
certainties involved in the project. They attempt to quantify the project quantities to be estimated [39]. MOM is an analytical non-simulation
cost variability based on one or more parameters. Moreover, these probabilistic approach that may be performed either by hand or using
techniques address the concerns regarding the chance of exceeding a Excel sheets relatively quickly with only a few additional steps [40].
particular cost in the range of possible costs, possible amount of the cost MOM takes a step further than the expected value by calculating the
overrun, and different types of uncertainties and how they drive cost standard deviation for each risk with the average and maximum risk
[97]. Probabilistic methods are divided into two main categories: values. In this technique, the final cost is expressed as a continuous
simulation and non-simulation methods. probability distribution, contrast to the expected value method which
expresses it as a fixed figure [94].
A. Simulation Methods (Monte Carlo)
• Probability Tree
In MCSs, EJ and numerical methods are used together to obtain a
probabilistic output using a simulation routine. MCS is an advantageous A probability tree provides a systematic method to convert individ­
mathematical technique for analyzing uncertain scenarios from a ual risks with the impact of a conditional expected value and their
probabilistic angle [39]. MSC has been identified as the most effective probability of occurrence into the overall probability and expected value
than other methods as it is understandable, easy to use, and simple. It is [20].
also analyzes opportunities, uncertainties, and threats [94,98]. MCS can
be a useful tool for risk managers and can be used to estimate con­ • Program Evaluation and Review Technique
struction project costs [42,99].
Historical data can be used for MCS to give the planning team an idea Program evaluation and review technique (PERT) is an analytical
about the variation in time and cost. MCS can also be used in the risk solution for activities with uncertain durations, which estimates the
management process by providing help to the planning team to make probability of their completion. This method is rarely used in con­
better decisions on risk mitigation strategy [100]. struction projects, but it has the potential to solve problems efficiently
through correction and improvement [42]. It is also known as the three-
3.2. Range estimating (RE) point-estimate technique and assumes a Beta distribution. This tech­
nique requires optimistic, most likely, and pessimistic values as inputs
In this method, critical cost items are identified, and their corre­ for each item [101]. It has been used for estimating and planning large
sponding deterministic estimate is considered the most likely value. projectsconsidering uncertainties which is one of its most powerful
Then, the maximum and minimum values of the critical items are concepts [102]. Recentlly, PERT has been used in several useful appli­
defined. Finally, the cumulative distribution function of the total cost is cations in manufacturing, research management, construction field, as
calculated using MCSs [20]. RE may include simple “best-case” and well as in product development and information technology [43].
“worst-case” point estimates, or they may be shown graphically with a
probability curve or probability mass function. • Parametric Estimating

• Integrated Models for Cost & Schedule Parametric estimating is a cost estimation methodology using sta­
tistical relationships between historical costs and other program vari­
Although cost and schedule are somehow related, both probabilistic ables such as performance and physical characteristics of the system,
methods for cost and schedule estimates are performed separately in the employee loading, or contractor output metrics. The parametric model is
absence of a direct link. However, some models integrating schedules a helpful tool for preparing early conceptual estimations in case of few
and costs have been developed [33]. Cho et al. [32] developed an in­ technical data or engineering deliverables for providing a basis for using
tegrated schedule and cost model to enable planning and controlling of more detailed estimation methods [16,44].
repetitive construction operations, assist the project manager in the
actual construction site. The model will allow project managers to more • Regression Analysis (RA)
effectively integrate timelines and cost of repetitive high-rise construc­
tions buildings. It will also support the decision making process of RA is a qualitative analogical statistical process to estimate the
project managers. relationship between variables [103]. It is considered one of the most
common methods for estimating costs for construction projects [45].
B. Non-Simulation Methods Linear regression technique is a simple method that is usually used in
modeling the relationship between numerical variables; and calculating
In this category, analytical methods are used, in which risk assess­ the trend of the data to predict the contingency amount [6]. The key
ment and contingency calculation are conducted without the use of objective in regression analysis technique is providing the necessary
simulation software packages. tools for modeling methodology that consider the in-depth analysis of a
dependent variable [104]. It is more effective for the early cost estimate
• Expected Value (EV) / Individual Risks or Estimated Risk Analysis when sufficient details about the project are unavailable [20]. RA is an
estimation technique used for calculating project cost contingency,
EV corresponds to the mean of the project risks probability distri­ which is a key component of a project budget [46].
bution. Risks involved in a project may be classified into two types: fixed The stepwise regression modeling was used to analyze the de­
risk, where events will either happen during the project or never at all, pendency relationship between predetermined cost contingency
and variable risk, where events will occur but to uncertain extents amounts in highway transportation projects and the perceived ratings of
[31,94]. The EV assumes that the project’s individual risks have been risk drivers assessed by project professionals. The model based on linear
identified beside their impact value and their probability of occurrence regression showed better prediction accuracy than the fuzzy-based
[8]. model in predicting cost overruns for highway construction projects in

93
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Egypt [47]. handle nonlinear problems as well [69,115]. NNs are computer systems
simulating the learning method of the human brain [76]. ANNs methods
• Analytical Hierarchy Process (AHP) can generate the necessary weight adjustment factors for the input data
and can handle most kind of dataset that could have a best fit equation,
AHP is a multicriteria method for decision-making analysis and such as logical, logarithmic, binomial, linear, polynomial, binomial, and
support, which can be used to solve complex decision problems in exponential [71,116].
evaluation, estimation, and decision-making [50,105]. A key aspect of Because of its exceptional characteristics of self-learning, fault
the AHP is that decision criteria are evaluated with respect to their tolerance, non-linearity, adaptability, and progression of input to output
relative importance allowing trade-offs between them [54,106]. It helps planning; ANNs have become the most widely used for universal func­
decision-makers to set their priorities and to choose the best alternatives tion approximation in numerical models [73,117]. NN techniques have
when both quantitative and qualitative aspects are considered [107]. been widely applied in many fields of industry including the construc­
The AHP consists of three steps: hierarchy formation (the decision goal), tion sector [70].
pairwise comparisons (decision-makers), and verification of consistency Meijuan [75] found that ANNs can produce more accurate estimating
(EJs) [51,108]. AHP is a mathematical tool that helps the decision- models. ANNs systems were inspired by the information processing
makers to divide the complex problem into simple criteria [55,109]. behavior of the human brain, where computation relies on inter­
The capability to examine and reduce the inconsistency of expert connected information processing units, which receive inputs and
judgments is the main advantage of AHP [110]. convert them into desired outputs [118]. Barakchi et al. [72] found that
Eskandar [53] used AHP method for normalizing the uncertainty ANN and unit cost are the two most commonly used methods across
estimates and ranking the likelihood of risk which occurs during bidding transportation infrastructure.
and construction phases in both Egypt and Saudi Arabia. Findings
showed that risk from financial, design risk, political and construction 3.3. Mixed and hybrid contingency methods
risk are the most likelihood occurrence from project stakeholders point
of view. Most of the following methods and approaches are mixed or hybrid,
integrating deterministic and probabilistic methods. Some of them
• Optimism Bias Uplift / Reference Class Forecasting (RCF) require special software packages, whereas others focus on
manufacturing construction products.
Optimism bias uplift or RCF is a non-simulation probabilistic method
developed by Flyvbjerg et al. [56] for the British Department for 3.3.1. Support vector Machines (SVMs) / kernel ridge regression
Transport to deal with the optimistic bias of estimating the capital SVMs are a very successful and accessible set of techniques for
project costs. RCF takes an outside view of the planned actions of a classification, which can be used to predict the parametric cost estima­
project by setting forecasts regarding the actual performance of a tion [86]. SVM is a binary classification model that can create a hyper-
comparative project rather than focusing only on the project in progress level division of sample data and achieve structural risk reduction based
[56]. on the principles of maximum margin [119,120]. It does not require any
assumption on data distribution and it can effectively deal with multi­
3.2.1. Modern mathematical methods variate small sample data, thus, it is suitable for project cost estimation
Modern mathematical methods deal with applied mathematics and [121].
its associated finite methods.
3.3.2. Bayesian belief network (BBN)
A. Fuzzy Techniques BBNs, also known as belief networks, belong to the family of prob­
abilistic graphical models. BBN is a particular type of diagram (called a
Fuzzy set ranking methods are generally based on the fuzzy set directed graph) with an associated set of probability tables [87]. BBNs
theory developed by Lotfi Zadeh [111]. Because of the uncertainty of have gained attention because they provide a powerful mechanism for
people’s perception of things, to better solve the problem of multi- performing cause-and-effect analysis with both qualitative and quanti­
attribute decision-making using fuzzy information, the concept of tative traits, and thus, for supporting decision-making under uncertainty
fuzzy set appears [112]. [87,122]. Mathematically, BBNs are classified as directed acyclic
The fuzzy set is an intuitive qualitative comparative analysis tech­ graphs, where the vertices contain information about events, and the
nique and decision support system that identifies meaningful cases; it is arches describe causal relationships between the vertices [123]. In terms
located midway between exploratory research and hypothesis testing of accuracy, Bayesian network is superior to other prediction methods
[68]. Fuzzy set approach enables modeling and processing of data which [124].
are difficult to be quantified based on probability theory and mathe­
matical statistics [65,113]. 3.3.3. Case-Based reasoning (CBR)
The fuzzy set provides a flexible link between qualitative and CBR is an intuitive qualitative process for solving existing problems
quantitative characteristics where continuous variables are reinter­ based on previous solutions to similar problems [88]. It is an interactive
preted as presence or absence of a particular feature, while discrete activity to solve everyday life problems based on past personal experi­
variables are transformed into membership or a continuous degree of ences and not only a dominant method for computer reasoning. CBR can
belonging [62]. play its best role only when the most similar cases to new problems
One of the most important advantages of fuzzy theory is the ability to through some retrieval method are found [88]. Compared to traditional
deal with uncertain, incomplete, and imprecise data associated with methods, the CBR approach only considers information similarity. It can
complex construction projects [114]. The fuzzy logic is a powerful avoid missing important correlated information by making inferences
solution-finding model for high overall quality, lowest cost incurred, from multiple sources of information [125].
and shortest project duration time [60,66].
3.3.4. Factor Rating (FR)/estimate quality
B. Artificial Neural Networks (ANNs) FR is a technique or procedure for evaluating multiple alternatives
based on a number of selected factors. The FR method is used when
Artificial neural networks (ANNS) systems are qualitative/analogical historical building and component data are available from similar works
techniques, particularly useful in uncertain conditions and adaptable to [89]. Scaling relationships of existing component costs are used to

94
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Table 3
Main advantages and limitations of the most used contingency methods.
Methods/Usage Advantages Disadvantages / Limitation

Traditional • Consistency and ease of use [46]. • Unscientific technique and a reason for over budget for many projects.
More reliable and rational. Difficult to justify the reason for the used percentage [3].
(All project types) Subjectivity is expelled from the procedure [74].

Expert Judgment • Flexible. • Difference experts use different mechanisms.

Quick produce [28]. Susceptible to bias.
(Most project types) Unstructured [93].

Range Estimating (Monte • A practical method when a work package has a significant level of • Applied only to the capital cost estimate.
Carlo) uncertainty. More difficult than the traditional approach [3].
Optimistic and pessimistic ends of estimation [30].
(The substantial capital
projects)

Integrated Model • Increase the flexibility of the risk analysis and reduce the amount • More time is needed to develop the model.
(Monte Carlo) of manual coordination. More costly than other estimating tools.
The ability to capture schedule-related costs impacts [82]. There are no satisfactory results if there is disbelief among project
(Most project types) Improved productivity [3,17]. team members [34].

Methods of Moment • Quick and provides physical interpretations [39]. • Mechanical equivalent equations do not give the same safety index.
Little iteration for convergence. Not reasonable for complex and mega infrastructure [74].
(Infrequently) Allows a probability distribution of cost to be developed in Excel
sheets [40].
Expected Value • Providing a contingency estimate without using (MCS). • Identifying variable and fixed factors of the project is [3].
More versatile in that it can be applied to any project plan of any The positive risks maybe missed.
(All project types) class or quality [37].
Probability Tree • Considers both numerical and categorical input data. • Becomes impractical as the number of results increases significantly
Considers non-linear relationships and correlated variables and with the number of risks.
(Infrequently) ignores useless variables [20]. Complicated and time-consuming [41].
PERT • Enables to predict the time and cost in advance. • Inaccurate in the case of large and complex projects [20].
Suggests areas for increasing efficiency and reducing cost. More expensive compared with other methods.
(Infrequently) Shows the critical path of the project [43].
Parametric Estimating • Objective and repeatable. • Excellent for early results.
Non-technical expert can apply it. Difficult to develop.
(Most project types) Allows scope for quantifying risk [16]. Relationship is not easily understandable.
Loses predictive credibility outside its relevant data range [16].
Regression Analysis • Powerful for predictive and analytical purposes of examining the • Needs much time to gather all historical data.
contribution of variables for overall estimate reliability [28]. Does not produce an accurate model if detailed project data is not
(Most project types) Focuses on the best predictors of tender Price [49]. available. [37].
Analytical Hierarchy Process • Capability to check • Requires experienced data, judgment and maybe used as a decision
(AHP) Reduce the inconsistency of expert judgments. support tool not as the means for deriving the final answer [3].
Providing group decision-making.
(Most project types) Effortless, flexible and Usability [83].
Reference Class Forecasting • Replaced assumptions with data [56].By taking the • Needs a cumulative sample of similar projects with large sample size.
(RCF) (Outside View), itigates strategic representation and optimism Impractical and does not provide reliable results [69]
Optimism Bias Uplifts bias [59].

(Transportation)

Artificial Neural Network • A powerful and practical tool for solving many problems [57,71]. • Requires a large case base to be effective.
(Infrequently) Can be used for predicting cost overruns [80]. Complex and Logic not visible.
Only suitable for non-linear modelling. [74]
Fuzzy Techniques • Robust system where no precise inputs are required. • Hard to develop a model.
Deal with problems with uncertainties and imprecise Difficult in real practice.
(Infrequently) information [35]. Different models based on different expert opinions [3].
Support Vector Machines • Provides a higher performance than traditional learning machines. • Need specialized experts.
(SVM) A powerful tool for solving regression and classification Different experts use different inputs [120]
problems [86]
(Road projects)
Bayesian Belief Network • Deal with case-effect relationships and support decision-making • Needs a lot of effort for data collection
(BBN) under uncertainty. Expensive to build.
Supports modeling of qualitative and quantitative aspects of Need expert to use
software development [87]. Hard to construct compared to other networks [128].
(Residential projects)
Case-Based Reasoning (CBR) • Solving existing problems based on previous similar problems. • Complex to handle complex cases.
Effective in the absence of bias. Only considers information similarity compared to traditional
Easy for interpretation [88]. methods [125].
Factor Rating • The ease of use. • If the estimate is inaccurate, it may lead to a high contingency.
(FR) Predicting the cost range according to the desired confidence Choosing of the four determinants is inadequate enough for the
Estimate Quality level [89]. accuracy of estimation [37]
(continued on next page)

95
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Table 3 (continued )
Methods/Usage Advantages Disadvantages / Limitation

(Mega capital projects)

Controlled Interval Memory • Indicative importance for the scientific management of networks • Limited practicability [90].
[90]. Hard to use.
(Rarely used) Accommodate the different views perspectives of the project Experts think it’s hard to understand [74].
[74].
Influence Diagram (Causal • Suitable for logical reasoning situations and making decisions • Becomes impractical with enormous number of risks, as the number of
Network Modeling) under uncertainty [91,126]. outcomes increases exponentially with the number of risks [91,127].

(Multifaceted projects)
Theory of Constraints • Organizes the project management process. • Assumes that cost items are independent of each other which is not
(TOC) Improve communication applicable in real-life [37].
Uses thinking process logic tools [92].
(Rarely used) Produces an immunized and achievable plan based on the
critical path that rules the project through to its execution [74].

predict the cost of similar new work. It is also known as the “equipment methodology deals with identifying the cost-related features of the
factor” estimation method. Oberlender & Trost [89] in their study pre­ product and identifying the associated costs.
dicted the required amount of cost contingency based on previous his­ • Generative Costing: It evaluates the cost of manufacturing a product
torical cost data. The quality of cost estimate depending on four based on its complexity.
determinants; participants in preparing the estimate, how the estimate • Activity-Based Costing: It consists of a quantitative analytical system
was prepared, known information about the project, and any other that focuses on calculating the costs spent on performing necessary
considered factors while preparing the estimate. activities to manufacture a product. The implementation of this
method requires a knockdown of the respective templates in sub-
3.3.5. Controlled interval and Memory assemblies and individual components.
This method has limited practicability due to the need for special- • Group Technology: The estimating system is based on the similarity
purpose software for its application. It additionally requires many between the product to be estimated and the products within the
simplifying defaults, which makes it less used. However, its process may classification.
be successfully applied to simple estimations. Zhang et al. [90] found • Knowledge-Based Engineering System: It is an attempt to imitate the
that controlled interval and memory is of indicative significance for the expert reasoning methods to estimate costs.
scientific management of gas transmission and distribution networks,
which efficiently prevents accidents. Hueber et al.[16], Niazi et al.[27], and Curran et al.[69] added seven
more cost estimation techniques and systems:
3.3.6. Influence diagram / causal network modeling
Influence diagram or causal network modeling is suitable for logical • Top-Down Processing: It is a divergent method that identifies the main
reasoning situations and making decisions under uncertainty, as the construction project objectives and tasks needed to support them.
model is based on a probabilistic theory [126]. Influence diagrams are a The newly defined activities are then broken down into smaller tasks
graphical tool for mapping the interaction of different elements of de­ and spread across different team members to begin their develop­
cision making [91]; ment. In that sense, top-down processing is a diverse approach.
• Bottom-Up Processing: It is a convergent method where smaller goals
• Rectangles usually represent decisions, are initially defined, and activities start from the lower organiza­
• Events or uncertainties are represented with circles or ovals, tional levels of the company [69]. At a later stage, these smaller goals
• Inputs and outputs are represented with rounded rectangles, and will be connected to the primary project goals and strategies.
• Triangles represent values or results. • Operation-Based Approach: It is an analytical quantitative technique,
generally used in the final stages of design because of the required
Pearl [127] stated that the graphical models are distinguished in information, and it is one of the first attempts for estimating
reasoning and decision-making stems as they have four main compu­ manufacturing costs [27].
tational features; modular representation of probabilities, systematic • Breakdown Approach: It is an analytical quantitative method [16],
methods of construction, explicit coding of independence, and the which estimates the total product costs by summing up all costs
effective procedures for inference. incurred during the production cycle, including direct costs and
overheads.
3.3.7. Theory of constraints (TOC) • Tolerance-Based Cost Models: These are analytical quantitative tech­
The TOC is regarded as the philosophy of continuous improvement, niques to estimate the cost of a product considering design tolerances
which has expanded its methodological base over time. The TOC uses as a function of its cost [27].
thinking process logic tools to identify the core constraint [92]. • Decision Support Systems: These are intuitive qualitative systems
Evans et al [93] studied more than six cost estimation methods and useful for evaluating design alternatives. The main purpose is to
systems as follows: assist estimators to improve the decision-making process at different
levels of the estimation process using the experts’ stored knowledge
• Function Costing: It is based on the concept that the cost of a proposed in the same field [16].
system or product can be directly calculated from the performance • Rule-Based Systems: These are based on process and cost calculation
specifications of this system or product. of potential operations from a set of available operations for
• Feature Costing: Feature-based cost estimation is a quantitative manufacturing a part of the product based on design or
analytical technique developed from the activity-based approach. It manufacturing constraints [16].
is differentiated by determining the manufacturing times. Its

96
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Table 4
Case studies applications based on contingency methods.
Method Researchers Country Project Technique and Purpose Results
type

Traditional Baccarini [46] Australia Calculates contingency through the Australian The average construction contingency provision was 5.24% of the
(Road networks) government road authority for highways and roads awarded contract value, while the average value of contract
projects. variations was 9.92%.
Expert Olumide et al. [131] USA Using the Delphi technique as an expert judgment Provided a consistent rationale for estimating contingency.
Judgment (Highway) method to predict contingencies of sliding scale for Estimators are encouraged to identify and document risks as
highway construction projects. justifications for contingency values applied to a project. For a most
complex projects, contingency value was found to be within 40% to
100%, whereas for a non-complex projects a range between 22% and
50% was found to be suitable.
AHP E I – Touny et al. [83] Egypt Developed a reliable model to identify the impact of risk The model demonstrated that the predicted cost contingency
(Highway) factors during the bidding stages and to estimate the matches 96.31% of the average estimated contingency for real case
expected contingency costs. projects, and the model showed robust results.
Range Para et al. [130] Spain Proposed an extension for Monte Carlo simulation model The proposal introduced in this study allows managers to obtain a
Estimating (Shipbuilding) to calculate the likelihood of complying with the more optimal estimate of the contingency reserve, therefore reducing
calculated budget for shipbuilding projects. economic risks. They found the contingency between 4.3% and 4.9%.
Regression Ogungbile et al. [132] Developed cost models for predicting the initial R2 of the developed models ranged from 0.85 to 0.99 indicating
Analysis Nigeria (Roads) estimation for road projects at the early stage using linear consistency between the expected values from the prediction models
and multiple regression techniques. and the actual data
Fuzzy Idrus et al. [81] Malaysia Estimated contingency costs using a logical and flexible Compared with the actual cost of the contingency, the predictions
Techniques (Buildings and approach and accommodate contractors’ subjective made by the system were within an accuracy range of 20%.
Infrastructures) judgment based on risk analysis. The average actual cost contingency amount was 6.37%, where the
predicted value was 5.7%.

Artificial Lhee et al. [80] USA To estimate the optimal contingency to fund The proposed models can improve the owner’s budgeting decisions
Neural (Transportation) transportation construction projects. and reduce the risk of either increased commitment to funds or
Network underutilization.
The range of cost contingency was 15% from one-step model, and
13% from the two-step model.

collecting data or making assumptions about variables.

It is difficult to study or list all risks related to a construction project
activity. It is also difficult to collect all the exact data associated with the
impact of risks on the overall project costs. Most of these models would
be more suitable in case of similar previously implemented construction
projects [84].This is mainly due to the collection of accurate data from
these projects, which will result in a more accurate calculation of con­
tingency [15].
Most of previous studies, whether in developing or developed
countries studied cost contingency estimates in most types of construc­
tion projects, a few of them were about managing cost contingency and a
little about combining estimating and managing cost contingency.
Lhee et al. [80], Idrus et al. [81], Hammad et al. [82] and Lhee, 2014
[129] listed different methods to estimate cost contingency. Some
require the project manager to have prior knowledge, such as fuzzy logic
and ANNs, others assign the probability distribution to all activities and
Fig. 4. Distribution of cost estimation method by project type. select an appropriate confidence interval. While others require data
collection from previous projects and using some subjective opinions.
4. Advantages and disadvantages of contingency estimation Para et al. [130] presented a method to calculate and manage cost
methods contingency, based on Monte Carlo simulation.
Table 4 shows a selected sample of applied case studies for cost
Based on previous literature studies,Table 3 shows the advantages contingency calculation, explaining the purpose and results of the
and disadvantages of the most used contingency estimation methods. application cases, based on the most common methods, project type
(building, transportation, infrastructure), development phase (bidding,
5. Results and discussion execution) and the project’s geographical location.
Based on leterature review and previous studies [3,17,28,82], Fig. 4
5.1. Results of previous models and applications for contingency methods samarizes the distribution of cost estimation method by project type.

This section summarizes the applications of the previously developed 5.2. Discussion
models and case studies considering different cost contingency methods.
Some methods can be applied for smaller construction projects, while Accurate cost estimates are important ingredients for a successful
other methods are preferable for large-scale projects. construction project delivery. Cost contingency is a vital component of
In summary, previously reported studies focused either on estimating cost estimation, which considers risks and uncertainties associated with
cost contingency or combining cost contingency estimation and man­ construction projects to enable the completion of the project within the
agement. For some models, the estimator’s prior knowledge is a pre­ assigned budget. On the one hand, high contingency costs may result in
requisite, such as in fuzzy techniques, whereas other models require poor cost management, lack of available funds for other projects, and
uneconomic completion of a project. On the other hand, low

97
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Table 5
Recommanded cost contingency methosds.
Approach/ Project type Benefits/Contributions References
Technique

ANN Buildings. Transportation. ANN is a better solution for modeling complex nonlinear relationships and reducing errors [3,6,20,27,28,37,50,51,53,54,55,80]
Public projects. and human efforts.
A helpful tool in resolving uncertainty problems such as cost estimation at the conceptual
phase.
Obtain an accurate cost prediction for insufficient information/data in the preliminary
stage of design process or tender phase.Has the ability to predict tender price, performance
of construction budget, construction demand, and project cash
flows.It
can calculate cost contingency for projects considering the fluctuating material prices,
thus, facilitates incorporating risk contingency into construction projects.

The ANN method is superior than the traditional and RA methods in predicting simulated
cost contingency compared to actual contingency value, thereby, ANN model is more
powerful in theoretical modeling of cost contingency simulation.

Fuzzy Buildings. The fuzzy method provides an approach for estimation and allocation that allows users to
Infrastructure. monitor and control contingency funds over project life cycle. [28,37,60,62,62,65,66,81,112]
The fuzzy method overcomes the limitations of probabilistic methods and provides
flexibility for experts to express their knowledge based on their intuition and experience.
Historical data records are not required to generate probability density functions (PDF)
for individual cost elements and does not require a large number of simulations. It has
proven to be effective not only in estimating contingency but also in estimating range for
cost or time.
Provides useful and future information about project contingencies during all project
phases, such as bidding, design, implementation, and delivery.
Providing a flexible unified procedure adopting the policies and procedures of the
companies based on their intuition, lessons learned, management and technical skills, and
their guidelines and respective regulations.

Generates a contingency baseline that allows project managers to monitor and control the
contingency at any milestone and prevents the contingency surplus of non-occurred risks.
It also allows to properly re -allocate the surplus of the contingency fund.
Compared to commonly used methods, Fuzzy method is an accurate and effective tool.

AHP Transportation Building AHP is an essential technique used for assessing and analyzing construction project risks [28,37,50,51,53,54,55,83]
Infrastructure through bidding and construction stages and to overcome traditional techniques
limitations.AHP is a powerful tool for decision-making techniques and deals with
uncertainties that may arise in the construction procedures and assist in solving complex
problems.It
prepares more solutions for making a decision for problems and assists evaluation’s
measures by examine the consistency of evaluations to reduce all inconsistencies in
decision-making.
Supports the Project Manager with a structured procedure to model, analyze and control
the risks impact; It also improves risk analysis and prioritization results to support project
managers in making decisions, such as planning more effective mitigation actions and re-
assigning risk ownership.
AHP method redresses the effective calculation of the total construction cost and considers
the tentative mitigation costs required against each risk
impact.It
identifies the financial implications during bidding phase that affect cost contingency and
can estimate the expected contingency of construction projects with reliable predictive
accuracy of cost estimation and guides decision-makers in the early planning stages.

(continued on next page)

98
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

Table 5 (continued )
Approach/ Project type Benefits/Contributions References
Technique

RA Transportation Regression models are widely used as predictive techniques to calculate the contingency [6,20,28,37,45,46,47,48,49,132]
Building costs of a project and to examine the best-fit parameters from available system data
Regression model can be helpful prior-to-tender stage for owners to predicted cost
contingency and to overcome the potential cost overruns during project life cycle.
Linear regression-based models accurately estimate expected cost escalation for any road
projects more accurately than fuzzy-based models. Therefore, it is recommended to use
linear regression-based models in predicting cost contingencies for transportation
projects.
MCS Infrastructure Easy to use as it does not require complex equations or timely data collection, also does not [3,28,37,50,51,53,54,55,82,130]
Transportation require the project manager to have prior knowledge of any other simulation techniques.
Provides an early warning for uncompleted activities by calculating the available cost
contingency reserve to spend on future activities, as such, it would be a very powerful tool
to warn overspending and to show the acceptable overspending range.
Supports the project managers for carefully planning activities based on spending pattern
for past, current, and future activities.

Allows the project manager to determine the required confidence level through its ability
to absorb turbulence of the project environment.

contingency costs could lead to an unrealistic financial environment, complex projects in the planning phase, a range between 22% and
inadequate funding for project execution, and unsatisfactory perfor­ 50% was found to be suitable.
mance outcomes. • To achieve reliability, 32 experts in risk assessment and cost esti­
This emphasizes the paramount importance for owners to choose the mation were participated in the study through three iterative rounds
most appropriate method for estimating contingency costs at the early with controlled feedback. Experts concluded that Sliding scale
stages of any construction project. method provides consistent rationale for contingency estimation.
The application of the most used models in Table 4 for estimating and
managing project cost contingency is discussed as follows: In Egypt, E I – Touny et al. [83] used AHP to predict cost contingency
Baccarini [46] used data available from 48 road construction pro­ for Highway projects during the bidding stages.
jects from the Australian road authority for quantitative analysis of the
estimation of construction contingency. • They found that the average contingency cost of real projects was
5.97%, which is close to the value obtained from their developed
• The average construction contingency provision was 5.24% of the model of 5.75%.
awarded contract value, while the associated variations were 9.92%, • To verifiy the developed model, real data from five highway projects
resulting in a contingency deficit of 4.68%. were collected from previous projects, which include actual and
• The contract variation values were not reflected in the construction estimated cost.
contingency. • The predicted contingency matches 96.31% of the average estimated
• There were no significant correlations between project variables and contingency of the real projects.
cost contingency that could be used for contingency predictions. This • Their model can be reliable and robust in predicting cost contingency
is a somehow surprising observation since it was expected that the values.
amount of contingency and variance would be strongly correlated
because the former is estimated to cater to the latter. Para et al. [130] presented a study with estimating methods of
• Therefore, it was concluded that the weak correlation would indicate project contingency reserve at the early budget calculation stages for a
that contingency estimating methods need to be evaluated for ways ferry construction project in Spain.
to improve accuracy.
• The budget contingency reserve estimate is usually performed based
Olumide et al. [131] in USA, performed a contingency evaluation for on shipyard experience.
highway projects using Delphi technique that relies on expert judgment. • Problems arise when the shipyard does not have sufficient experi­
ence in the concrete type of vessel to be built.
• The developed various scenarios as a framework for contingency • The use of an extension of the triangular Monte Carlo distribution
assessment to solve a complex problem that created consistency in model was proposed to calculate the likelihood of complying with
the definition of contingency and estimation and developed sliding the calculated budget.
scales that provides a consistent method to be used for assessing • They found the contingency amount between 4.3% and 4.9%.
contingencies in the projects. • To verify the use of the Monte Carlo method, real data for a series of
• Contingencies sliding scale were developed according to the three ferry-type vessels were used. The results with triangular dis­
complexity of the project. tribution covers a smaller range of values, which reflects a better
• Contingency amount varies based on the development of the project definition of the model, therefore, the model has been considrered to
phases. be more reliable.
• For a most complex projects in the planning phase, contingency
value was found to be within 40% to 100%, whereas for a non-

99
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

• Thus, a contingency reserve that provides sufficient security to • Finally, they claimed that the two-step model could improve owner
execute the project within the limits of the economic risks identified budgeting decisions and reduce the risk of either overcompliance or
by the company can be calculated. underutilization.
• The proposal presented in the study allows managers to obtain an
optimal estimate of contingency reserves and thus reduce economic Based on deep analysis of the literature and previous studies, the
risks. following approaches and techniques are recommended for construction
agencies to choose the appropriate cost contingency method during the
Ogungbile et al. [132] proposed regression models for predicting the different stages of a project based on the peculiarity of each project, ease
initial cost estimation of road construction projects in Nigeria. of use for each method, and the needed accuracy level of the estimates;
as shown in Table 5.
• They identified the quantitative and qualitative factors that affect Cost contingency and cost estimation methods in construction pro­
road project construction costs. jects have been reviewed in the published articles over the past twenty
• They used the procedures leading to the formulation of linear and years. Almost in all cases, estimates at the early stage of a project have
multiple regression models to determine the cost of functional ac­ the major concern.
tivities of roads as a function of project characteristics. Most of the proposed techniques have attempted to predict cost
• The dovloped model was validated using data from twenty contingencies fro the projects by developing models for application at
completed road projects in southwestern Nigeria. preliminary stages to help the decision-making process of organization.
• R2 value of the developed models ranged from 85% to 99%, which Essentially, effective cost factors must be thoroughly explored and
indicates that the expected values from the prediction models are examined. In addition, the effective cost factors should be studied, but
compatible with real-life data. also the factors affecting the cost model accuracy must be reviewed in
• The importance of estimating an accurate budget in the bidding depth.
stages is essential to be able to compete under the best One of the cost factors that have been noted repeatedly is the
circumstances. regional geographical factor (Project country location), which shows the
importance of differentiating between projects with diverse origins.
Idrus et al. [81] proposed an estimating cost contingency method for Also, the model’s overall scalability and applicability to new cases
building and infrastructure projects in Malaysia using a flexible and are of high importance.
rational approach based on risk analysis technique and fuzzy expert ANN and RA are the most common methods applied in prvious
system. literature reviews, respectively.

• A risk analysis technique was applied as a key concept for identifying 6. Conclusions
and assessing the risk level for each risk factor.
• The fuzzy expert system was used as a method for assessing the level Construction agencies require a budget estimate in the early stages of
of risk in the step of risk assessment. a project to manage costs that consist of two main components, a base
• Five stages were used to describe the developed model: development estimate, and a contingency cost, which together represent the esti­
of the conceptual model of cost contingencies, identifying their risk mated final cost of the project. The contingency reserve usually includes
factors, developing the fuzzy expert system, model testing, and risks and uncertainties associated with construction projects to contain
finally, tuning and validation. uncertainties-induced monetary impact and avoid project cost overruns.
• The project cost contingency was formed as a function of the level of Therefore, the contingency cost is included within the budget estimates
risk factors, which was determined as the magnitude of risk. to represent the project agency’s total financial commitment. Thus, ac­
• The average actual cost contingency amount was 6.37%, where the curate estimation of the contingency cost is of critical importance to the
predicted value was 5.7%. project.
• Three case studies projects were used to perform the validation test, Construction contingency is the key component of the project budget
and the results showed that the minimum error was 4.44%, while the for a construction contract to meet contract variation requirements that
maximum error was 18.00%. could arise during the project implementation phase. The cost perfor­
• The predicted contigency amount was within an accuracy range of mance of a construction project is a critical success criterion intimately
20%. dependent upon accurate estimations of cost contingency. Therefore, to
keep the project within budget, an accurate contingency calculation is
Lhee et al. [80] developed an ANN-based model for estimating essential. Hence, it is essential for any agency to be aware of different
optimal contingency costs for transportation construction projects in the contingency calculation techniques and select the most appropriate one
USA. based on the project characteristics.
The essential contribution of this study is to guide agencies toward
• The model was developed using the contingency rate as an output appropriate contingency methods for their construction projects.
variable of the ANN to predict the owner contingency costs for Through in-depth analysis of different methods of estimating contin­
construction projects. gency, the advantages and limitations of each method have been sum­
• As a first step, performance comparison with the ANN-based model marized. Moreover, different application case studies were identified to
was performed. conclude the criteria for selecting each method.
• In the second step, mean squared error, average error, correlation In this study, most of the cost contingency calculation methods were
values, and statistical analysis of error rates were used as “net sta­ collected and classified. Overall, 34 different methods and techniques
tistics.” It was found that the predictive performance of the second- were identified and discussed under these categories.
step model is better than that of the first-step model. The study provided several guidelines for cost contingency methods
• This model was validated by comparing the performance with the in the construction industry in terms of;
ANN-based one-step model.
• The range of cost contingency was 15% from one-step model, and • Selecting the model that matches the project type and objectives.
13% from the two-step model. • Creating a data set priority and checking its adequacy and efficiency.
• Using the proposed model at the preliminary stages of the project to
achieve the best results.

100
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

The study proposed a practical guidance approach for construction [18] D. Baccarini, P.E.D. Love, Statistical characteristics of cost contingency in water
infrastructure projects, J. Constr. Eng. Manag. 140 (3) (2014).
agancies to choose their appropriate cost contingency method. This
[19] D. Baccarini, The maturing concept of estimating project cost contingency-a
work is expected to form a good resource for agencies in the budget review, in: 31st Australasian University Building Educators Association
development stage and those who want to allocate a contingency budget Conference, (AUBEA), Australia, 2006.
for their construction projects. [20] P. Bakhshi, A. Touran, An overview of budget contingency calculation methods in
construction industry, Proc. Eng. 85 (2014) 52–60.
For future research the review for schedule, quality, and budget [21] M. Gabel, M. Sujka, Z.W. Davis, A.E. Keizur, Performance of risk-based estimating
contingency could be extended together to examine their combined for capital projects, Transport. Res. Rec. 2677 (1) (2023) 1059–1070.
impact on the construction project budget. [22] R.L. Kimmons, C. Reserves, A management control tool, Proj. Manage. (2017)
67–79.
Conflict of interest. [23] L. Keith Bea, Cheryl Runyon, Kae M. Warnock, Nevada Emergency Management
The authors declare that they have no known competing personal and Homeland Security Statutory Authorities Summarized, in: Congressional
relationship or financial interest that could have appeared to influence Information Service, Library of Congress, Washington, DC, 2004.
[24] G. Rothwell, Cost contingency as the standard deviation of the cost estimate, Cost
the work reported in this study. Eng. 47 (7) (2005) 22–25.
[25] O. Akinradewo, C. Aigbavboa, A. Oke, H. Coffie, Appraisal of risk contingency
planning for construction projects, IOP Conf. Ser.: Mater. Sci. Eng. 640 (1) (2019)
Declaration of Competing Interest 012019.
[26] M. Donn, A. Dykman, N. Bakshi, Interpreting Traditional Cost Contingency
The authors declare that they have no known competing financial Methods in the Construction Industry, July 2021, https://doi.org/10.26686/
wgtn.14984772.v1.
interests or personal relationships that could have appeared to influence
[27] A.J. Niazi, S. Dai, S. Balabani, L. Seneviratne, Product cost estimation: technique
the work reported in this paper. classification and methodology review, J. Manuf. Sci. Eng. 128 (2) (2006)
563–575.
[28] S. Tayefeh Hashemi, O.M. Ebadati, H. Kaur, Cost estimation and prediction in
Acknowledgment construction projects: a systematic review on machine learning techniques, SN
Appl. Sci. 2 (10) (2020).
We extend our sincere thanks to the Department of Civil Engineering, [29] M. Jorgensen, Practical guidelines for expert-judgment-based software effort
estimation, IEEE Softw. 22 (3) (2005) 57–63.
Faculty of Engineering at Shoubra, Benha University, Egypt. Special [30] D. Curto, F. Acebes, J.M. González-Varona, D. Poza, Impact of aleatoric,
thanks to Nile Engineering Consulting Bureau (NECB). stochastic and epistemic uncertainties on project cost contingency reserves, Int. J.
Prod. Econ. 253 (2022) 108626.
[31] D. Baccarini, Understanding project cost contingency-A survey, in: Conference
Funding proceedings of the Queensland University of Technology (QUT) research week
international conference, 4–8 July, 2005.
[32] K. Cho, T. Hong, C. Hyun, Integrated schedule and cost model for repetitive
This research received no external funding. construction process, J. Manage. Eng. 26(2) (2010) 78–88.
[33] David Todd Hulett, Monte Carlo simulation for integrated cost-schedule risk
References analysis, in: Handbook of Research on Leveraging Risk and Uncertainties for
Effective Project Management, 2017, pp. 29–60.
[34] P.M. Diaz, Analysis of benefits, advantages and challenges of building
[1] Project Management Institute, A Guide to the Project Management Body of
information modelling in construction industry, J. Adv. Civil Eng. 2(2) (2016) 1-
Knowledge: (PMBOK Guide) - 7th Ed, and the standard of project management
11.
(English) Project Management Institute, 2021.
[35] F. Afzal, S. Yunfei, D. Junaid, M.S. Hanif, Cost-risk contingency framework for
[2] A.M. Keshk, I. Maarouf, Y. Annany, Special studies in management of
managing cost overrun in metropolitan projects: using fuzzy-AHP and simulation,
construction project risks, risk concept, plan building, risk quantitative and
Int. J. Manag. Projects Bus. 13(5) (2020) 1121–1139.
qualitative analysis, risk response strategies, Alex. Eng. J. 57 (4) (2018)
[36] A. Touran, Owners risk reduction techniques using a CM. CMAA Research Report,
3179–3187.
Construction Management Association of America, Washington, DC, 55pp., 2006.
[3] A.M. El-Kholy, A.M. Tahwia, M.M. Elsayed, Prediction of simulated cost
[37] R.A. Hamid, F.J. Kehinde, Choosing an appropriate contingency sum estimating
contingency for steel reinforcement in building projects: ANN versus regression-
methods for highway construction projects in Nigeria: a literature review, Plan.
based models, Int. J. Constr. Manage. 22 (9) (2020) 1675–1689.
Malaysia 15 (1) (2017, 2017,) 13–20.
[4] G.A. Silva, S.K. Warnakulasooriya, B.N.F., and Arachchige, B., Criteria for
[38] J.-P. Paquin, P.-P. Morin, A. Lambert, T. Koplyay, Assessing project contingency
construction project success: A literature review, Project & Program management
reserves with the expected cost overrun risk measure, J. Constr. Eng. Manag. 148
e Journal (2016).
(10) (2022).
[5] S. Sarkar, M.S. Bhaskar, K. Uma Rao, P. V, D. Almakhles, U. Subramaniam, Solar
[39] K. Binder, D.W. Heermann, Monte Carlo Simulation in Statistical physics: An
PV network installation standards and cost estimation guidelines for smart cities,
Introduction, sixth ed., Springer, 2019.
Alex. Eng. J. 61 (2) (2022) 1277–1287.
[40] Department of Infrastructure, Transport, Cities and Regional Development.
[6] Y. Elfahham, Estimation and prediction of construction cost index using neural
Annual Report 2017-18, 2018 <https://www.infrastructure.gov.au/about-us/
networks, time series, and regression, Alex. Eng. J. 58 (2) (2019) 499–506.
corporate-reporting/annual-reports>.
[7] F. Tayyari, Cost and Analysis for Engineers and Scientists, First ed., CRC press,
[41] Y. Fu, Z. Yin, M. Su, Y. Wu, G. Liu, Construction and reasoning approach of belief
2022 <http://www.taylorfrancis.com/ books/9780429432163>.
rule-base for classification base on decision tree, IEEE Access 8 (2020)
[8] I.A. Eldosouky, A.H. Ibrahim, H.E. Mohammed, Management of construction cost
138046–138057.
contingency covering upside and downside risks, Alex. Eng. J. 53 (4) (2014)
[42] A. Claudius, V. Dorca, A. Radu, M. Radu, Risk consideration and cost estimation
863–881.
in construction projects using Monte Carlo Simulation, Management 10 (2)
[9] F. Allahi, L. Cassettari, M. Mosca, Stochastic risk analysis and cost contingency
(2015) 163–176.
allocation approach for construction projects applying Monte Carlo simulation,
[43] A. Aribisala, A. Otenaike, O. Balogun, L.O. Ofusori, Program evaluation and
in: Proceedings of the World Congress on Engineering, London, UK, 5–7 July
review technique analysis of an engineering project, J. Eng. Technol. 2 (2017).
2017, vol. I, pp. 385–391.
[44] M. Relich, A. Świć, Parametric estimation and constraint programming-based
[10] J.I. Ortiz, E. Pellicer, K.R. Molenaar, Determining contingencies in the
planning and simulation of production cost of a new product, Appl. Sci. 10 (18)
management of construction projects, Proj. Manag. J. 50 (2) (2019) 226–242.
(2022) 6330.
[11] G. Jackson, Contingency for cost control in project management: a case study,
[45] T.P. Williams, Predicting final cost for competitively bid construction project
Const. Econ. Build. 3 (1) (2012) 1–12.
using regression models, Int. J. Proj. Manage. 21(8) (2003) 593-599.
[12] D. Salem, A. Bakr, Z. El Sayad, Post-construction stages cost management:
[46] D. Baccarini, Accuracy in estimating project cost construction contingency – a
sustainable design approach, Alex. Eng. J. 57 (4) (2018) 3429–3435.
statistical analysis, in: Cobra: RICS International Construction Conference,
[13] N.J. Smith, T. Merna, P. Jobling, S. Thompson, Managing Risk in Construction
Responding to Change, 7–8 September, 2004, Headingley Stadium, UK.
Projects, 3rd ed., Wiley Blackwell, 2014.
[47] A.E. Abu El-Maaty, A.M. El-Kholy, A.Y. Akal, Modeling schedule overrun and cost
[14] P.E. Love, L.A. Ika, Making sense of hospital project (Mis) performance: over
escalation percentages of highway projects using fuzzy approach, Eng. Constr.
budget, late, time and time again—why? And what can be done about it?
Archit. Manag. 24 (5) (2017) 809–827.
Engineering 12 (2021) 183–201.
[48] G. Polat, B.N. Bingol, A.P. Gurgun, B. Yel, Comparison of ANN and MRA
[15] B. Flyvbjerg, D.W. Bester, The cost-benefit fallacy: why cost-benefit analysis is
approaches to estimate bid mark-up size in public construction projects, Proc.
broken and how to fix it, J. Benefit-Cost Anal. 12 (3) (2021) 395–419.
Eng. 164 (2016) 331–338.
[16] C. Hueber, K. Horejsi, R. Schledjewski, Review of cost estimation: methods and
[49] M.F. Diab, A. Varma, K. Panthi, Modeling the construction risk ratings to estimate
models for aerospace composite manufacturing, Adv Manuf. Polym. Compos. Sci.
the contingency in highway projects, J. Constr. Eng. Manag. 143 (8) (2017).
2 (2016) 1–13.
[50] K. Dobi, J. Gugic, D. Kancijan, AHP as a decision support tool in the multicriteria
[17] G.A. Barraza, R.A. Bueno, Cost contingency management, J. Manag. Eng. 23 (3)
evaluation of bids in public procurement, in: Proceedings of the ITI, 32nd
(2007) 140–146.

101
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

International conference on information technology interfaces, 2010, [84] A.E. Thal, J.J. Cook, E.D. White, Estimation of cost contingency for Air Force
pp. 447–452. construction project, J. Constr. Eng. Manag. 136 (11) (2010) 1181–1188.
[51] R. Velmurugan, S. Selvamuthukumar, R. Manavalan, Multi criteria decision [85] T.Y.M. Lam, N. Siwingwa, Risk management and contingency sum of construction
making to select the suitable method for preparation of nanoparticles using an projects, J. Financ. Manage. Proper. Constr. 22 (3) (2017) 237–251.
analytical process, Pharmazie 66 (11) (2011) 836–842. [86] N.I. El-Sawalhi, Support vector machine cost estimation model for road projects,
[52] Y.M.R. Aboelmagd, Decision support system for selecting optimal construction J. Civil Eng. Architect. 9 (9) (2015).
bid price, Alex. Eng. J. 57 (4) (2018) 4189–4205. [87] A.L.R. de Sousa, C.R.B. de Souza, R.Q. Reis, A 20-year mapping of Bayesian belief
[53] R.F. Eskander, Aziz., Risk assessment influencing factors for arabian construction networks in software project management, IET Softw. 16 (1) (2021) 14–28.
projects using analytic hierarchy process, Alex. Eng. J. 57 (4) (2018) 4207–4218. [88] Y. Wang, L. Fei, Y. Feng, Y. Wang, L. Liu, A hybrid retrieval strategy for case-
[54] A. Darko, A.P.C. Chan, E.E. Ameyaw, E.K. Owusu, E. Pärn, D.J. Edwards, Review based reasoning using soft likelihood functions, Soft Comput. 26 (7) (2022)
of application of analytic hierarchy process (AHP) in construction, Int. J. Constr. 3489–3501.
Manag. 19 (5) (2019) 436–452. [89] G.D. Oberlender, S.M. Trost, Predicting accuracy of early cost estimates based on
[55] S. Panchal, A.K. Shrivastava, Landslide hazard assessment using analytic estimate quality, J. Constr. Eng. Manage. 127 (3) (2001) 173–182.
hierarchy process (AHP): a case study of national highway 5 in India, Ain Shams [90] X. Zhang, Y. Bai, L. Lv, Application of the controlled interval and memory model
Eng. J. 13 (3) (2022) 101626. in the risk assessment of city gas transmission and distribution networks, ICPTT.
[56] B.H. Flyvbjerg, C, & Fok, WH., Reference class forecasting for Hon Kong’s major (2012).
roadworks projects, Proc. Inst. Civil Engin.: Civil Engin. 169 (6) (2016) 17–24. [91] B. Kandi, H. William, S. Marjorie, G. Kathleen, Sustainable Land Development
[57] H. Liu, C. Jiang, Y. Liu, M. Hertogh, X. Lyu, Optimism bias evaluation and and Restoration: Decision Consequence Analysis, Elsevier Science Publisher,
decision-making risk forecast on bridge project cost based on reference class Oxford, 2010.
forecasting: evidence from China, Sustainability 10 (11) (2018) 3981. [92] J.F. Cox, J.G. Schleier, Theory of Constraints Handbook, first ed., McGraw-Hill,
[58] T. Servranckx, M. Vanhoucke, T. Aouam, Practical application of reference class 2010.
forecasting for cost and time estimations: Identifying the properties of similarity, [93] D. Evans, D.J. Lanham, D.R. Marsh, Cost estimation method selection: matching
Eur. J. Oper. Res. 295 (3) (2021) 1161–1179. user requirements and knowledge availability to methods, University of West of
[59] A. Chadee, H. Martin, S. Gallage, U. Rathnayake, Reducing cost overrun in public England, Systems Engineering and Estimation for Decision Support (SEEDS)
housing projects: a simplified reference class forecast for small island developing Group, Bristol.
states. Buildings 13 (4) (2023) 998. [94] O. Akinradewo, L. Ngwenya, C. Aigbavboa, W. Thwala, L. Mphela, Improving the
[60] A. Salah, O. Moselhi, Contingency modelling for construction projects using efficacy of cost contingency plans for construction projects in South Africa, IOP
fuzzy-set theory, Eng. Constr. Archit. Manag. 22 (2) (2015) 214–241. Conf. Ser.: Mater. Sci. Eng. 640 (1) (2019) 012030.
[61] S. Nawar, O. Hosny, K. Nassar, Owner time and cost contingency estimation for [95] M.S. Islam, M.P. Nepal, M.S. Alam, M.d. Arifuzzaman, Cost contingency
building construction projects in Egypt, Construct. Res. Congr. (2018) 367–377. modelling for construction projects; insight from the literature, IEEE (2019) 1–5.
[62] S. Kraus, D. Ribeiro- Soriano, M. Schussler, Fuzzy set qualitative comparative [96] C.M. Harper, K.R. Molenaar, S. Anderson, C. Schexnayder, Synthesis of
analysis (fsqca) in entrepreneurship and innovation research- the rise of a performance measures for highway cost estimating, J. Manage. Eng. 30 (3)
method, Int. Entrep. Manag. J. 14 (1) (2018) 15–33. (2014).
[63] M.S. Islam, M.P. Nepal, M. Skitmore, R. Drogemuller, Risk induced contingency [97] M. Modarres, Risk Analysis in Engineering: Techniques, Tools, and Trends, CRC
cost modeling for power plant projects, Autom. Constr. 123 (2021) 103519. Press, 2016, 10.1201/b21429.
[64] J. Xu, M. ye, Construction project cost estimation model cost dependent on multi- [98] F.S. Al-Duais, R.S. Al-Sharpi, A unique Markov chain Monte Carlo method for
objective fuzzy optimization calculation, J. Phys. Conf. Ser. 1904 (1) (2021) forecasting wind power utilizing time series model, Alex. Eng. J. 74 (2023)
012001. 51–63.
[65] E. Plebankiewicz, K. Zima, D. Wieczorek, Modelling of time, cost and risk of [99] M. Ali Elfarra, M. Kaya, Estimation of electricity cost of wind energy using monte
construction with using fuzzy logic, J. Civ. Eng. Manag. 27 (6) (2021) 412–426. carlo simulations based on nonparametric and parametric probability density
[66] D.-T. Nguyen, L. Le-Hoai, P. Basenda Tarigan, D.-H. Tran, Tradeoff time cost functions, Alex. Eng. J. 60 (4) (2021) 3631–3640.
quality in repetitive construction project using fuzzy logic approach and [100] S. Züst, M. Huonder, W.S. Shaun, O. Stoll, Life cycle-oriented risk assessment
symbiotic organism search algorithm, Alex. Eng. J. 61 (2) (2022) 1499–1518. using a Monte Carlo simulation, Appl. Sci. 12 (1) (2021) 8.
[67] W. Alfaggi, S. Naimi, An optimal cost estimation practices of fuzzy ahp for [101] A.M. El-kholy, Time-cost tradeoff analysis considering funding variability and
building construction projects in Libya, Civil Eng. J. 8 (6) (2022) 1194–1204. time uncertainty, Alex. Eng. J. 52 (1) (2013) 113–121.
[68] S. Kumar, S. Sahoo, W.M. Lim, S. Kraus, U. Bamel, Fuzzy-set qualitative [102] R.F. Aziz, RPERT: repetitive-projects evaluation and review technique, Alex. Eng.
comparative analysis (FsQCA) in business and management research: a J. 53 (1) (2014) 81–93.
contemporary overview, Technol. Forecast. Soc. Chang. 178 (2022) 121599. [103] E. Tosun, K. Aydin, M. Bilgili, Comparison of linear regression and artificial
[69] R. Curran, S. Raghunathan, M. Price, Review of aerospace engineering cost neural network model of a diesel engine fueled with biodiesel-alcohol mixtures,
modelling: the genetic causal approach, Prog. Aerosp. Sci. 40 (8) (2004) 487–534. Alex. Eng. J. 55 (4) (2016) 3081–3089.
[70] G. Kim, J.M. Shin, S. Kim, Y. Shin, Comparison of school building construction [104] A. Khan, M. Amanullah, H.M. Aljohani, S.A.M. Mubarak, Influence diagnostics for
costs estimation methods using regression analysis, neural network, and support the Poisson regression model using two-parameter estimator, Alex. Eng. J. 60 (5)
vector machine, J. Build. Construct. Plan. Res. 1 (1) (2013) 1–7. (2021) 4745–4759.
[71] K.K. Shrestha, P.P. Shrestha, A contingency cost estimation system for road [105] H.-A.-A.-F. Haroun, A.F. Bakr, A.-E.-S. Hasan, Multi-criteria decision making for
maintenance contracts, Proc. Eng. 145 (2016) 128–135. adaptive reuse of heritage buildings: Aziza Fahmy Palace, Alexandria, Egypt,
[72] M. Barakchi, O. Torp, A.M. Belay, Cost estimation methods for transport Alex. Eng. J. 58 (2) (2019) 467–478.
infrastructure: a systematic literature review, Proc. Eng. 196 (2017) 270–277. [106] A.H. Ibrahim, M.A. Shaker, Sustainability index for highway construction
[73] O.I. Abiodun, A. Jantan, A.E. Omolara, K.V. Dada, N.A. Mohamed, H. Arshad, projects, Alex. Eng. J. 58 (4) (2019) 1399–1411.
State-of-the-art in artificial neural network applications: a survey, Heliyon 4 (11) [107] P.I. Cakmak, E. Cakmak, An analysis of causes of disputes in the construction
(2018) e00938. industry using analytical hierarchy process (AHP), AEI (2013).
[74] R.E. Adaurhere, I. Musonda, C.S. Okoro, Construction contingency determination: [108] I.M. Mahdi, A.M. Ebid, R. Khallaf, Decision support system for optimum soft clay
a review of processes and techniques, Adv. Sci., Technol. Innov. (2021) 271–277. improvement technique for highway construction projects, Ain Shams Eng. J. 11
[75] C. Meijuan, Application of ANN technique to predict the thermal conductivity of (1) (2020) 213–223.
Nanofluids: a review, J. Therm. Anal. Calorim. 145 (4) (2021) 2021–2032. [109] I.M. Mahdi, H. Arafat, A.M. Ebid, A.F. El-Kadi, Decision model to identify the
[76] S. Yun, Performance analysis of construction cost prediction using neural network optimum retaining wall type for restricted highway projects sites, Int. J. Constr.
for multioutput regression, Appl. Sci. 12 (19) (2022) 9592. Manage. 23 (8) (2021) 1387–1394.
[77] W. Ou, S. Xiao, C. Zhu, W. Han, Q. Zhang, An overview of brain-like computing, [110] S. Aminbakhsh, M. Gunduz, R. Sonmez, Safety risk assessment using analytic
Architect. Appl. Future Trends Front. Neurorob. 16 (2022). hierarchy process (AHP) during planning and budgeting of construction projects,
[78] A. Bonini Neto, D.A. Alves, C.R. Minussi, Artificial neural networks: multilayer J. Safety Res. 46 (2013) 99–105.
perceptron and radial basis to obtain post-contingency loading margin in [111] Z.J. Al-Araji, S.S.S. Ahmad, N. Kausar, A. Farhani, E. Ozbilge, T. Cagin, Fuzzy
electrical power systems, Energies 15 (21) (2022) 7939. theory in fog computing: review, taxonomy, and open issues, IEEE Access 10
[79] F.T. Kristanti, Z. Safriza, Dwi Fitrizal Salim, Are Indonesian construction (2022) 126931–126956.
companies financially distressed? A prediction using artificial neural networks, [112] B. Ning, H. Wang, G. Wei, C. Wei, Several similarity measures of probabilistic
Invest. Manage. Fin. Innov. 20 (2) (2023) 41–52. dual hesitant fuzzy sets and their applications to new energy vehicle charging
[80] S.C. Lhee, I. Flood, R.R. Issa, Development of a two- step neural network- based station location, Alex. Eng. J. 71 (2023) 371–385.
model to predict construction cost contingency, J. Inform. Technol. Construct. 19 [113] D.-H. Tran, M.-Y. Cheng, A.-D. Pham, Using fuzzy clustering chaotic-based
(2014) 399–411. differential evolution to solve multiple resources leveling in the multiple projects
[81] I. Arazi, M.F. Nuruddin, M. Arif Rohman, Development of project cost scheduling problem, Alex. Eng. J. 55 (2) (2016) 1541–1552.
contingency estimation model using risk analysis and fuzzy expert system, Exp. [114] P.K. Goyal, S. Sharma, Evaluating cost contingency for construction projects: a
Syst. Appl. 38 (3) (2011) 1501–1508. fuzzy risk analysis approach, Lect. Notes Netw. Syst. (2022) 607–622.
[82] M.W. Hammad, A. Abbasi, M.J. Ryan, Allocation and management of cost [115] A.I. Taloba, An artificial neural network mechanism for optimizing the water
contingency in projects, J. Manage. Eng. 32 (2016) 04016014. treatment process and desalination process, Alex. Eng. J. 61 (12) (2022)
[83] EI–Touny, A.S., Ibrahim, A.H., and Amer, M.I., Estimating cost contingency for 9287–9295.
highway construction projects using analytic hierarchy processes, Int. J. Comp. [116] K.C. Onyelowe, J.S. Effiong, A.M. Ebid, Predicting subgrade and subbase
Sci. Iss. 11(6):73 (2014). California bearing ratio (CBR) failure at Calabar-Itu highway using AI (GP, ANN,

102
T. Ammar et al. Alexandria Engineering Journal 78 (2023) 88–103

and EPR) techniques for effective maintenance, Artif. Intell. Mach. Learn. Smart [125] X. Jiang, S. Wang, J. Wang, S. Lyu, M. Skitmore, A decision method for
City Plan. (2023) 159–170. construction safety risk management based on ontology and improved CBR:
[117] A.S. Fahmy, M.-E.-T. El-Madawy, Y. Atef Gobran, Using artificial neural networks Example of a subway project, Int. J. Environ. Res. Public Health 17 (11) (2020)
in the design of orthotropic bridge decks, Alex. Eng. J. 55 (4) (2016) 3195–3203. 3928.
[118] B. Nagy, D.L. Galata, A. Farkas, Z.K. Nagy, Application of artificial neural [126] K. Jeet, V. Mago, B. Prasad, R.S. Minhas, An influence diagram-based approach
networks in the process analytical technology of pharmaceutical for estimating staff training in software industry, J. Intell. Syst. 18 (4) (2009)
manufacturing—a review, AAPS J. 24 (4) (2022). 267–284.
[119] G. Ma, Z. Chao, Y.e. Zhang, Y. Zhu, H. Hu, The application of support vector [127] J. Pearl, Influence diagrams- historical and personal perspectives, Decis. Anal. 2
machine in geotechnical engineering, IOP Conf. Ser:. Earth Environ. Sci. 189 (4) (2005) 232–324.
(2018) 022055. [128] E. Lee, Y. Park, J.G. Shin, Large engineering project risk management using a
[120] S. Petrusheva, D. Car-Pušić, V. Zileska-Pancovska, Support vector machine based Bayesian belief network, Exp. Syst. Appl. 36 (3) (2009) 5880–5887.
hybrid model for prediction of road structures construction costs, IOP Conf. Ser.: [129] S.C. Lhee, Finding significant factors to affect cost contingency on construction
Earth Environ. Sci. 222 (2019), 012010. projects using ANOVA statistical method -focused on transportation construction
[121] M. Xu, B. Xu, L. Zhou, L. Wu, Construction project cost prediction based on projects in the US-, Architect. Res. 16 (2) (2014) 75–80.
genetic algorithm and least squares support vector machine, in: Proceedings of [130] L. Para-González, C. Mascaraque-Ramírez, A.E. Madrid, Obtaining the budget
the 5th International Conference on Civil Engineering and Transportation, 2015. contingency reserve through the Monte Carlo method: Study of a ferry
[122] S.M.A. Aljeddani, M.A. Mohammed, Parameter estimation of a model using construction project, Brodogradnja. 69 (3) (2018) 79–95.
maximum likelihood function and Bayesian analysis through moment of order [131] A.O. Olumide, S.D. Anderson, K.R. Molenaar, Sliding- scale contingency for
statistics, Alex. Eng. J. 75 (2023) 221–232. project development process, Transport. Res. Rec. J. Transport. Res. Board. 2151
[123] A. Leśniak, F. Janowiec, Application of Bayesian Networks in construction (1) (2010) 21–27.
engineering, Civil Environ. Eng. Rep. 30 (2) (2020) 221–233. [132] A.J. Ogungbile, A.E. Oke, K. Rasak, Developing cost model for preliminary
[124] S.-S. Leu, C.-Y. Lu, P.-L. Wu, Dynamic-Bayesian-network-based project cost estimate of road projects in Nigeria, Int. J. Sust. Real Estate Constr. Econ. 1 (20)
overrun prediction model, Sustainability 15 (5) (2023) 4570. (2018) 182.

103