A PREDICTION MODEL FOR CONSTRUCTION PROJECT TIME

CONTINGENCY
Dalia Mohamed1, Florida Srour1, Wael Tabra1, and Tarek Zayed2
1

M.A.Sc. Student, Department of Building, Civil and Environmental Engineering,

Concordia University, Montreal, Quebec, Canada.
2
Associate Professor, Department of Building, Civil and Environmental Engineering,
Concordia University, Montreal, Quebec, Canada.
ABSTRACT
Construction projects often suffer from high level of uncertainty in many fronts: time,
cost, quality, safety, etc. Most of the conducted research focuses on estimating cost
contingency. Even though construction project scheduling has received extensive
attention of researchers, time contingency was not treated well in the literature. In
order to meet the deadline of a project, an accurate scheduling should be sought. Due
to the nature of construction projects, scheduling should be flexible enough to
accommodate changes without negatively affecting the overall duration of the project.
As such, the objectives of the research presented in this paper are to assess the factors
that affect scheduling contingency and to develop a simple model that can be used in
estimating the expected time contingency of a construction project. A survey was
conducted on sixteen construction companies in Montral, Qubec, Canada. The data
obtained from the survey was then processed using Analytic Hierarchy Processes
(AHP) to develop a time contingency model. Results show that the predicted time
contingency matches with 87% the estimated contingency for real projects. The
developed time contingency model showed robust results.
INTRODUCTION
The main goals of any successful construction project management system(s) are to
complete the project on time, within the planned budget, and with the required quality
limits. The three goals are inter-related where each of them is affecting, and being
affected by, the others. In order to meet the time deadline of a project, an accurate
scheduling should be sought. Due to the unique nature of construction projects, time
contingency and project uncertainty are essential for accurate scheduling, which
should be flexible enough to accommodate changes without negatively affecting the
overall duration of the project. It is essential to allocate a contingency value to both
cost and time (Touran, 2003). Yet, there are situations where there could be delays in
activities that result in a delay in the overall project duration. These delays will

736
Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

737

consequently have a negative impact on the quality and budget of the project.
Therefore, estimating time contingency is seen as a major factor for achieving a
successful construction project. Although several industrial sectors developed and
used software for estimating time and cost contingencies in order to minimize delays
and over budget, yet limited efforts are reported in the literature in the area of
predicting time contingency in the construction sector. Therefore, the objective of the
presented research in this paper is to identify the factors that affect schedule (time)
contingency and develop a model that predicts the expected contingency of a
construction project.
LITERATURE REVIEW
There is no standard definition of contingency in which it could mean different
meanings to different estimators, contractors, and owners organizations (Moselhi,
1997). Contingency is probably the most misunderstood, misinterpreted, and
misapplied word in project execution (Patrascu, 1988). It is an amount of money or
time (or other resources) added to the base estimated amount to achieve a specific
confidence level or allow for changes where experience shows obligation (AACE,
2000). It can also be defined as the budget that is set aside to cope with uncertainties
during construction (Touran, 2003) or the amount of money or time needed above the
estimate to reduce the risk of overruns of project objectives to an acceptable level to
the organization (PMI, 2000). Treasury (1993) identified two major categories of
contingency for construction projects:
(i) Design Contingency it addresses the changes during the design process for
factors, such as incomplete scope definition and inaccuracy of estimating methods
and data (Clark & Lorenzoni, 1996).
(ii) Construction Contingency it addresses the changes during the construction
process. Under a traditional procurement arrangement, the contract typically
contains a variation clause(s) to allow for changes and provide a mechanism for
determining and valuing variations (Staugas, 1995).
There are many definitions for contingency in the literature; most of them focused on
cost contingency. Contingency has different meanings to different people. Despite its
importance, estimating time contingency was not thoroughly addressed in the
literature. Prior to reviewing the estimating methods for project contingency, there are
different attributes that affect contingency. Yer and Jha (2006) identified 55 factors
affecting the performance of the project schedule. They observed seven factors that
have the most significant impact on the schedule outcome and divide them into two
main categories. The first include factors that posses the capability to improve
performance level, such as owners competence as well as commitment and conflict
among project members. The second include factors that tend to retain the schedule at
its existing level, such as coordination among project members, lack of knowledge
and skills for the project managers, hostile socioeconomic environment, and
uncertainty of project members. In addition, Touran (2003) indicated that the effect of
change orders increased the original cost and schedule since they modified the
original contract.
There are many factors that impact time contingency. Sukumaran et al. (2006) stated
that it was not only important to identify the factors but evaluate their impact on the
project duration as well. They developed a model in highway work zones that

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

738

identified different factors, which influence the construction schedule. Most

engineers, planners, and agencies depend on their experience to estimate cost and
time contingency (AbouRizk 2005, Touran 2003). The contractors contingency was
represented as a fixed percentage of the contract value (Smith, and Bohn, 1999) or as
a percentage of total project cost or duration (Touran 2003). Smith and Bohn (1999)
estimated the contingency as 5-10% of the contract value. On the other hand, Touran
(2003) estimated two values of cost contingency: (1) 15% for underground
construction activities and tunneling and (2) 7.5% for the rest of the project. Time
contingency has been used to assure the completion time of a project and provide a
degree of confidence that the planned duration can be accomplished successfully
(Park, and Pea-Mora 2004, Mulholland and Christian 1999). The completion date of
a project was often missed due to uncertain events, which their impacts were difficult
to predict because of the uniqueness of construction projects (Thompson and Perry
1992). Kanoglu (2003) stated that overruns of the project time is common in the
construction projects. Despite this fact, the construction industry did not recognize the
criticality of the problem. Arif (2005) studied an alternative way to estimate the
contingency by using other methodologies, such as correlation, regression analysis,
and neural networks. Park and Pea-Mora (2004) estimated time contingency as 20%
of the project duration. AbouRizk and Boskers (2005) presented a methodology that
incorporates network analysis and duration uncertainty in project time analysis. They
studied the effect of various factors on time contingency using a linear equation.
Illsley (2006) stated that in the industrial sector, there are several project scheduling
software such as Primavera and Microsoft Project; however, the most popular is Risk
Expert. This is because it provides quantitative and qualitative analyses of project
information, which is used to give a clearer picture of the true cost and time scale of
any project considering risk, penalties, and complex scheduling variables.
RESEARCH METHODOLOGY
Factors that affect time contingency were identified through literature review as
shown in Figure 1. A survey was conduct to collect the impact of each factor on time
contingency. The research methodology was performed using two different
approaches: deterministic and stochastic. In the deterministic approach, the impact
and likelihood for each factor were obtained from the survey results. Using the AHP
method, the weight of each factor was obtained and time contingency can be
determined. The stochastic approach is not presented in this paper because of size
limitations. The developed deterministic AHP model was validated and a sensitivity
analysis was conducted to determine the effect of these factors on time contingency.
FACTORS AFFECTING TIME CONTINGENCY
Based on literature and practitioners, several factors that are thought to have a direct
effect on time contingency were determined. These factors were divided into three
major categories: project, management, and environmental conditions. Table 1 shows
the description of these categories and factors.
DATA COLLECTION
A survey was developed in the form of a questionnaire to collect project data from
Montral contractors. The questionnaire was designed using the significant factors
identified from the literature as shown in Table 1. The factors were identified by
analyzing several research works, such as Moselhi (1993), COBRA (2004), Method

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

739

Engineering (1996), COBRA (2006), www.interface-consulting.com, and Iyer and

Jha (2006). The questionnaire included three parts. Part one included the respondent
person general information: years of experience, total value of completed works, type
of company partnership, type of the projects, project duration. In part two, a
comparison between criteria and sub-criteria was developed to obtain the ratio among
parameters of time contingency.

Figure 1. Research methodology

Part three included 13 questions mainly regarding the impact of factors on time
contingency. In order to facilitate the answers for the reviewers, a scale consisting of
numbers from 1-9 is used in which 1 means extremely ineffective and 9 means
extremely effective. Physical and telephone interviews, with senior managers of
sixteen construction management teams, were conducted. The construction projects
were located in Montral, Quebec, Canada. The surveyed companies had an
experience history ranged from 3 to 60 years. They worked in variety of projects,
such as buildings, hospitals, university, and bridges rehabilitation. The budget values
of these projects ranged from $25 to $500 Million Canadian dollars and the duration
ranged from 6 months to 3 years.
DEVELOPMENT OF TIME CONTINGENCY MODEL
Analytic hierarchy process (AHP) has been widely used and applied in different fields
of theory and practice (Al-Barqawi, 2006; Saaty, 1982 & 1991). It has been applied in
multi-criteria decision making, planning and resource allocation, conflict resolution,
and prediction problems (Saaty, 1982 & 1991). Therefore, the AHP is used in the
presented research to assess the weights of various factors that affect time
contingency through pair-wise comparison matrices. These matrices have several
important characteristics as shown in Table 2.

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

740

Environmental Conditions

Management Conditions

Project Conditions

Criteria

Table 1. Factors Affecting Time Contingency

Factor
Description
Project Location Location may influence the amount of risk and therefore
the level of contingency.
Project size
It affects planned schedules negatively. In the large projects
(Design
there are many various activities that required many
different resources and involved many parties. All of these
complexity)
variables are interfering together which may cause delay in
the project duration.
Equipments
Technology requirements comprising of method of
availability
construction equipments Issue of renting equipments,
(Construction
damages that may occur will increase contingency.
technology)
Material
This factor is related to the site condition and storage area.
availability
Transport material from the supplier to the site is time
(Market)
consuming which required a prior arrangement.
Amount of
Any project could have stop work order because of the
interference
owner or engineer interference, this is happened due to lack
(Skills)
of knowledge and experience from all the participants. If
the amount of interference increased the delay of schedule
will be increase until they make their decision.
Number of
Change of orders or extra work order usually requires long
change orders
process of redesigning or modifying specification. In
addition the extra work may force the contractor to
accelerate work which could cause loss of the labour
productivity and then caused the delay.
Payments
Any delay of payment may cause delay of supplying
(Delays)
resource to the project which will affect the planned
schedule.
Time to make a Owner is the main responsible in this case until he/she
decision
makes the decision of change or not. The contractor should
record this delay against the owner in case of claim.
Productivity of Losing of labor productivity that caused by acceleration or
labor and
extra work will affect the project schedule. Any damage of
equipments
equipment will cause serious delay on the current activity
consequently causes delay in project schedule.
Weather conditionWeather in some countries has the highest impact on the
schedule delay.
Soil condition
Some unforeseen soil conditions in the site cause delay in
the schedule.
Labor strike
This stoppage will cause delay in project schedule.
Shortage of
Some unforeseen events like work accident, sickness,
human resources social, psychological and other unpredicted event may
cause labor pain, or absenteeism. Hence, this factor will
cause delay in the activity and schedule.

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

741

Table 2. Typical Pair-Wise Comparison Matrix for Different Factors

Factors
a1
a2
a3
a4

a1
1
a21
a31
a41

a2
a12
1
a32
a42
a11, i =1, j = 1

a3
a13
a23
1
a43

a4
a14
a24
a34
1

1. The diagonal elements are all equal to one because they represent the comparison
of a criterion against itself.
2. The lower triangle values are the reciprocal of the upper triangular values (i.e. aij
= 1/aji).
3. All numbers in the matrix are positive.
4. Square matrix (i.e. number of rows equal to number of columns)
The weight of each factor is determined using Equation (1) as follows:
j= n
ai
(1)
W x =  nn j
j=1 
ai
i =1

Where, aij represents the matrix elements for i row and j column; n represents the
pair-wise comparison matrix dimension; and Wx represents the weight of factor x.
After determining the weights of each factor in the hierarchy, the time contingency
(CD) is developed using the model shown in Equation (2):
n

CD =  Wi * Si * Pi

(2)

i =1

Where, Wi represents the general weight of factor i; Si represents the score of each
factor in a specific project; Pi represents the probability of occurrence of factor I; and
CD is the time contingency.
TIME CONTINGENCY MODEL IMPLEMENTATION
The following are the steps that were used to develop time contingency model using
the collected data.
1- Determine the relative weight of each major category; i.e. project conditions,
management conditions and environmental conditions.
2- Determine the weights (Wi) of the sub factors relative to the weight of its
category.
3- Calculate the factors score (Si) for each of the thirteen factors (using a 1 - 9
scale) in which one represents the most ineffective and nine represents the
most effective to the contingency value.
4- Calculate the Probability of occurrence average (Pi) for each of the thirteen
factors.
5- Multiply the three values Wi * Si * Pi.
6- Sum all the values of multiplication, which constitute the time contingency
value CD.

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

742

After implementing the abovementioned steps to the collected data, Table 3 shows
the weights and relative weights of categories and factors, respectively. In addition, it
shows the time contingency calculation based upon average values of scores and
probabilities. It also shows that time contingency represents 36.78% of project
duration. Therefore, the time estimator has to consider almost 36% increase in project
duration due to the effect of the contingency factors. This value accounts for the
occurrence of all factors; however, the case might be different if only few of them are
expected to occur. In this case, the contingency value might be lower.
Sensitivity Analysis
Sensitivity analysis was performed to determine the most sensitive factors that affect
time contingency rate using @Risk software. The sensitivity analysis was done based
on the change in percentile of distributions 1%, 5%, 25%, 50%, 75%, 95% and 99%.
After the analysis was done, the factors were put in order of their effect on time
contingency as shown in Figure 2. It is clear from the chart that materials availability
factor is the most sensitive factor, followed by project size and 3 resource availability.
Therefore, the developed model proved to be sensitive to all considered factors.

Table 3. Time Contingency Calculation

Relativ
Weig
e
Score
Factors
hts Weight (Si)
s (Wi)
Size
0.234 0.076 0.638
8
0
Location
0.219 0.071 0.613
8
2
0.3238
Equipment
0.274 0.089 0.750
availability
9
0
Materials availability 0.270 0.087 0.750
4
6
Amount of
0.184 0.067 0.775
interferences
1
6
Number of change
0.160 0.058 0.656
orders
1
8
Time required for
0.164 0.060 0.600
0.3673 decisions
2
3
Payments (delays)
0.177 0.065 0.750
3
1
Equipments condition 0.142 0.052 0.520
8
5
Productivity
0.171 0.063 0.700
uncertainty
4
0
0.3089 Soil condition
0.220 0.068 0.613
9
2

Enviro
nmenta

Management Conditions

Project Conditions

Criteri Weight
a
s

Copyright ASCE 2009

Time
Probabili
Contingenc
ty (Pi)
y (CD)
0.533

0.0258

0.542

0.0237

0.550

0.0367

0.583

0.0383

0.567

0.0297

0.675

0.0260

0.600

0.0217

0.458

0.0224

0.400

0.0109

0.617

0.0272

0.592

0.0247

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

743

Weather conditions
Strike
Site shortage of
resources

0.264 0.081 0.738

4
7
0.264 0.081
0.738
4
7
0.250 0.077
0.700
3
3

0.533

0.0321

0.267

0.0161

0.600

0.0325

CD = 

Wi * Si * Pi

0.3678

i =1

Figure 2. Sensitivity Analysis for Time Contingency Model

MODEL VERIFICATION
In order to check the accuracy of estimated average time contingency (0.3678), data
were collected from experts through telephone calls regarding the actual delays in
their previous projects, which include the starting and planned finish dates in addition
to actual dates. Delays for the project, in months, were calculated by subtracting the
planned from actual finish dates, then, divide this delay over the planned project
duration to obtain time contingency. Table 4 includes the collected data and its
analysis for seven projects. It is noticed that time contingency ranged from 0.167 to
0.778 out of project duration. Thus, the average time contingency of the seven
projects is 35.4%, which is close to the value obtained from the developed model.
Based on Zayed and Halpin (2005), two equations are used to verify the developed
time contingency model as shown in Equations (3) and (4).

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

744

AIP = { 1 - (Ei /Ci )}/n *100

(3)

AVP = 100 AIP

(4)

i=1

Where, AIP = Average Invalidity Percent

AVP = Average Validity Percent
Ei = Estimated/Predicted Value
Ci = Actual Value
Applying these equations to the collected cases shows that AIP = 0.127 and AVP =
0.87. The values of AIP and AVP show that the developed model is robust in
predicting the values of time contingency.
Table 4. Delay and Contingency Analysis for Several Construction Projects
Project Start Scheduled end Actual end Planned
Delays in Contingency
date
date
date
project
months
duration
1
Aug-06
Feb-07
Mar-07
6
1
1/6=0.167
2
Sep-04
May-05
July-05
8
2
2/8=0.200
3
Aug-05
April-07
Sep-07
20
5
5/20=0.250
4
Feb-05
Feb-07
Aug-07
24
6
6/24=0.250
5
July-03
Oct-04
Mar-05
15
5
5/15=0.330
6
Feb-02
Feb-04
Feb-05
24
12
12/24=0.500
7
Sep-01
Mar-03
May-04
32
14
14/32=0.0.778
Average
0.354
SUMMARY AND CONCLUSIONS
Estimating scheduling (time) contingency is a major factor in achieving a successful
schedule for construction projects. In this research, a survey was conducted on many
construction companies to assess the factors that affect time contingency. The data
obtained from the survey was then processed using Analytic Hierarchy Process
(AHP) in order to evaluate the weight of each factor and estimate the time
contingency value. The results show that the average time contingency value is
estimated as 36.78%. The developed model is verified using seven cases studies,
which shows robust results (87%). This value shows that the obtained results are
fairly good and acceptable.
REFERENCES
AACE (American Association of Cost Engineers) (2000). AACE Internationals
Risk Management Dictionary, Cost Engineering, 42(4), pp. 28-31
Al Barqawi, H. A. (2006). Condition Rating Models for Underground Infrastructure:
Sustainable Water Mains, M.A.Sc. thesis, Concordia University, Montreal,
Quebec.
Brinkkemper, S., Lyytinen, K., and Welke, R. (1996). Method Engineering:
Principles of Method Construction and Tool Support, Kluwer Academic
Publishers, MA, USA.

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib

745

COBRA (2006). The construction and Building Research, Conference of the Royal
Institution of Chartered Surveyors, University College London, September 7-8.
Clark, F. and Lorenzoni, A. (1996). Applied Cost Engineering, M. Dekker, CRC;
3rd edition, New York, USA.
Dikmen, I., Birgonul, M., and Kiziltas, S. (2005). Prediction of Organizational
Effectiveness in Construction Companies, J. of Construction Engineering and
Management, ASCE, February, 131(2), pp. 252-261.
Illsley, R. (2006). Business Intelligence Pretmaster, Risk Expert, v7.8, Butler
Group Subscription Services, New Jersey, USA.
Interface Consulting International Inc. (2007). Interpreting Force Majeure in the
Wake of Disaster, <www.interface-consulting.com> (Surfed in November 2007)
Iyer, K., and Jha, K. (2006). Critical Factors Affecting Schedule Performance:
Evidence from Indian Construction Projects, J. of Construction Engineering and
Management, ASCE, August, 132(8), pp. 871-881.
Kanoglu, A. (2003). An Integrated System for Duration Estimation in Design/Build
Projects and Organizations, J. of Engineering, Construction and Architecture
Management, Emerald Group Publishing Limited, 10(4), pp. 272-282.
Moselhi, O. (1993). Applied Earned Value for Project Control, CIB W-65
conference Proceedings, Trinidad, WI, USA.
Moselhi, O. (1997). Risk Assessment and Contingency Estimating", Transactions of
the American Association of Cost Engineers, 41, A.06.1.
Park, M., and Pea-Mora, F. (2004). Reliability Buffering for Construction
Projects, J. of Construction Engineering and Management, ASCE,
September/October, 130(5), pp. 626637.
Patrascu, A. (1988). Construction Cost Engineering Handbook, M. Dekker, New
York, USA.
Saaty, T. (1982). Decision Making for Leaders: The Analytic Hierarchy Process for
Decision in a Complex World. Lifetime Learning Publications, Belmont,
California, USA.
Saaty, T. (1991). Decision Making with Dependence and Feedback: The Analytic
Network Process. RWS Publications, Pittsburgh, PA, USA.
Staugus, J, (1995). Variations, Building and Construction Law, 11(3), 156-158.
Sukumaran, P., Bayraktar, M.E., Hong, TH., and Hastak, M. (2006). Model for
Analysis of Factors Affecting Construction Schedule in Highway Work Zones, J.
of Transportation Engineering, ASCE, June, 132(6), pp. 508-517.
PMI (2000). A Guide to the Project Management of Knowledge (PMBOK Guide),
Project Management Institute, PA, USA.
Treasury, H. (1993). CUP Guidance: No.41Managing Risk and Contingency for
Works Projects, Central Unit on Procurement, London, UK.
Zayed, T. and Halpin, D. (2005). Pile Construction Productivity Assessment, J. of
Construction Engineering and Management, ASCE, June, 131(6), pp. 705714.

Copyright ASCE 2009

2009 Construction Research Congress

Downloaded 13 Sep 2011 to 132.205.59.19. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelib