payment

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uemj20

Engineering Management Journal

ISSN: 1042-9247 (Print) 2377-0643 (Online) Journal homepage: http://www.tandfonline.com/loi/uemj20

Integrated Risk of Progress-Based Costs and Schedule Delays in Construction Projects

Salahi Pehlivan & Ali Erhan Öztemir

To cite this article: Salahi Pehlivan & Ali Erhan Öztemir (2018) Integrated Risk of Progress-Based Costs and Schedule Delays in Construction Projects, Engineering Management Journal, 30:2, 108-116, DOI: 10.1080/10429247.2018.1439636

To link to this article: https://doi.org/10.1080/10429247.2018.1439636

Published online: 25 Apr 2018.

Submit your article to this journal

Article views: 129

View Crossmark data

Integrated Risk of Progress-Based Costs and Schedule Delays in Construction Projects

Salahi Pehlivan, Girne American University Ali Erhan Öztemir, Cyprus International University

Abstract: This article aims to explore the impact of the inte- gration of risk factors into delayed milestones for construction projects. A simulation model was developed to determine the impact of schedule variability on cost estimation. To generate random scenarios a Monte Carlo Simulation (MCS) technique was applied. The developed model computes the cost impact of delayed milestone in the expected budget. Using a risk integra- tion approach revealed the critical time frame that may lead to a budget deficit for a project. As a result, a number of cost- sensitive risk factors and schedule delays were identified for the critical time period where the risk of budget deficit increases. The method of integration proposed in this article highlights the priority of risk factors and schedule delays for construction contracts involving Payments at Event Occurrences (PEO). Con- sequently, the developed method can be useful for practitioners in anticipation of potential increase of costs, hence, prevention of failure due to budget deficit.

Keywords: Schedule Risk, Delay, Cost Estimate, Integrated Risk, Risk Prioritization

EMJ Focus Areas: Program and Project Management; Decision Making and Risk Management

C onstruction management and planning are challenging activities in the construction industry. The importance of planning cannot be underestimated for any project.

However, the actual duration of project activities often exceeds the planned duration. The reason behind these delays is the existence of internal and external risk factors that slow down or interrupt the progress of a project. As Mark, Cohen, and Glen (2004) state: “Risk is simply the potential for complications and problems with respect to the completion of a project task and the achievement of the project goal” (p. 1). The construction industry is strongly associated with various unexpected risks and uncertainties. One of the most common problems in construc- tion projects is the delay of project activities. Delays often mean that the project will not be completed on schedule, particularly if critical path activities are affected or if delays cause changes in the critical path. Numerous researchers have attempted to iden- tify risk factors in the construction industry that result from such delays. Studies by Assaf and Al-Hejji (2006) in Hong Kong, Sambasivan and Soon (2007) in Saudi Arabia and Lim and Mohamed. (2000) in Malaysia have focused on risk factors resulting from unexpected cost and scheduling problems. Iden- tification and evaluation of potential risk factors are important when performing risk analysis. Modeling of the cost risks relies on well-defined and consistent values (e.g. duration and cost of a land surveying) that are expected to minimize inaccuracy in risk analysis. The basic techniques, such as Program Evaluation and

Review Technique (PERT) and Critical Path Method (CPM) are capable of exploring critical path(s) and to estimate cost of a project.

PERT and CPM do not typically allow the study of risky delay factors, which may cause scheduling problems that also may lead to extension of project duration and/or additional expenditures. A number of shortcomings associated with PERT and CPM methods, emphasized by Adeli and Karim (1997) and Omar (2009), are worth considering when studying projects, particularly in the con- struction industry. Projects can fail to meet schedule goals due to unexpected costs, design changes, weather conditions, methods used, resources, late deliveries, and other problems. Risk factors can also vary with the project type. An increase in anticipated costs can create financial problems. Project cost estimation plays an important role in determining the cost risk of a project. Estimation of project costs may change depending on factors, such as location, financial uncertainties, rules and regulations, etc. Every factor that creates a fluctuation in actual project cost is a potential reason for a cost overrun that may lead to cost increases for a project. Evalua- tion of these risk factors should be taken into account by consider- ing the specifications of a project.

As one of the specifications, the contract type of a construction project plays an important role in payment schedules since the contractor has to pay the cost overrun . Contracted payment plans may lead to cash flow problems during the course of the project. To mitigate the risk of cash flow problems, a cost reimbursable approach contract is often used for construction projects. Lump- sum contracts can have progress payments (PP), which need to be specified in the contract. The main reason for this cost reimbur- sable contract is to minimize the financial burden on the contrac- tor. Consequently, the two parties must negotiate the payment method. There are three payment models that are based on the progress of the project: payments at activity completion times (PAC), and payments at event occurrences (PEO).

As one of the payment evaluation techniques, Earned Value Management (EVM) is often used to assess the performance of the project schedule and cost. EVM helps to identify the Estimate at Completion (EAC), which may differ from the Budget at Comple- tion (BAC), based on the performance of the project (Project Management Institute, 2013). To progress as planned, payments are extremely important for contractors in terms of performance. The payment model can vary with the agreement that determines time and amount of cash inflows received. According to Sweis, Sweis, Hammad, and Shboul (2008), “financial difficulties faced by the contractor” are ranked as the most important reason for delays. Therefore, delay of a milestone (event) can cause delayed payment for a project. From this perspective, the probability of delayed payments is crucially important for contractors in budget estima- tion, particularly for bid/no-bid decision-making processes. For this reason, a simulation of risks on milestones provides a more accurate framework to estimate a final budget. Moreover, the

Refereed Research Manuscript. Accepted by Associate Editor Gosavi.

108 Engineering Management Journal Vol. 30 No. 2 2018

integration of various approaches developed in this study can be used to predict the effect of risk factors on scheduling to minimize the cost risks by estimating the probable completion time of a construction project. It is not an easy task to integrate cost and schedule as noted by Cho, Hong, and Hyun (2010), where only repetitive project activities were considered for integration. Inte- grated approaches (Hulett & Nosbisch, 2012; Schatteman, Herroe- len, Vonder, & Boone, 2008) are based on proposing a range of outcomes using existing critical path(s). However, this study pro- poses to take into account all activities that may generate new critical paths, as well as potential delays of these activities to under- stand the potential cost impact.

The contributions of this research are twofold. First, an overall risk analysis where schedule delays are evaluated on milestone-based contracts involving PEO is developed. This work establishes the foundation for similar analyses of other types of contracts. Second, this work establishes that taking into account the impacts of both cost sensitivity of risk factors and probability of delays of milestones can be used as a frame- work for various cost simulations. Mitigated risk factors and the impact of prevented delays are identified for a project for which a cost-sensitivity ranking list was available. Milestone-based analysis helps to reveal the time frame in which a budget deficit may cause financial problems. Moreover, this approach also highlights the critical milestone(s) that cause budget deficits in a PEO-type payment model. The article is organized as follows. In the first section, risk modeling is introduced. Subsequently, schedule risk, integration of cost and schedule, and the integra- tion process are discussed based on a real-world case.

Risk Models in Construction Projects Risk analysis can be useful for projects where cost and schedule impacts are not experienced until construction begins . Identify- ing risk factors according to their influence on scheduling can help increase the likelihood of project success (Ökmen and Özta, 2008). Estimating the cost of schedule risks is essential for contingency planning. In order to minimize delays and budget deficits, risk management modeling has been recommended.

One approach to identify risk factors is to rank risk factors according to the severity of their potential influence on project activities and costs. Risk factors may have different levels of severity, and the risk they generate may affect critical path(s) and/or Cost Item(s) (CIs) of a project. As a result, new critical paths may be generated as the project progresses, which can increase the cost of the project and duration of completion (Chang, 2002). Focusing on a single path may lead to under- estimating the number of near critical and non-critical activities, thereby creating unexpected problems related to other existing activities of non-critical paths.

Schedule Risk Hulett and Nosbisch (2012) state: “Schedule risk has typically been ignored in assessments of cost risk” (p. 2). A common shortcoming of techniques used in the field of construction industry is the underestimation of schedule risk. A duration change in an activity may increase or decrease the cost and also the BAC of the project. However, it may not affect the completion time for the project. Scheduling is a key component in project success and must be controlled (Chang, 2002). There- fore, the efficient use of time is crucial to the project. Breaking down the project into activities helps to identify risk factors existing in schedules. However, considering every risk factor

may not be a practical approach to build a risk model. For this reason, a frequently applied practice is to consider three scenar- ios of duration: optimistic (minimum), most likely, and pessi- mistic (maximum) (Hulett, 2012). The three-point estimate of activity duration is a very practical approach to building a risk- analysis model. In practice, triangular distribution parameters (optimistic, most likely, and pessimistic) are easier to acquire than other probability distributions. In addition, the triangular distribution provides a good first approximation . A probability distribution for the duration of activities generates uncertainties on estimated costs for milestones. Risk analysis can highlight the activities that cause cost overruns. Moreover, this risk analysis can also reduce failure risk of a project. The ways to mitigate the risk of a project are presented on a real-world case and shown in the following section.

Data and Methodology This work proposes a model to integrate schedule and cost risks for PEO-type construction projects. For this study, CIs are defined as project events where payments are collected after the milestones are completed. This section is structured in two subsections. First, cost details and risk factors are given for the selected case, and second, the method of integration of these risk factors to the schedule is discussed.

Data The model is applied to a real-world case. The data for analysis are used with the permission of Ökmen (2008). In that study, a project consisting of 32 activities with the total cost broken down into 15 CIs was used. For the cost data, Turkish Lira (TL) exchange rate to USD was approximately one when the project was on track. Values were converted to USD using this exchange rate. The estimated total cost of the project for the optimistic scenario was $603,000. The most likely estimate was $796,000. The pessimistic scenario was $1,072,000. The CIs for each scenario are summarized in Exhibit 1.

The planned duration for the project was 300 calendar days. However, the actual duration was 514 days. The project involved ten different risk factors, which were assumed to have the potential to increase the risk of a cost overrun.

The risk factors were based on the characteristics of this construction project. This implies that the number of risk factors might be modified according to the needs of a project. The risk factors studied are enumerated in Exhibit 2.

Risk factors were taken directly from Ökmen (2008). Also, it was stated in the study of Ökmen (2008) that these factors were considered as potential risks to delay the project and increase the cost. Therefore, the model presented in this article is applied to the presented data to explore the impact of these risk factors.

Methodology The cost data used in this study are presented in Exhibit 3. The developed model uses the Monte Carlo Simulation (MCS) method to generate probable cost and schedule scenarios. The simulation was performed using 1,000 iterations, which is shown to be a reasonable number for construction projects. In terms of cost scenarios, the number of iterations was sufficient to draw a reasonable conclusion.

The simulation model used the Microsoft Excel 2016 (MS Excel) add-in @Risk, which computes the probability of activ- ity duration without altering the sequence of activities. The

Engineering Management Journal Vol. 30 No. 2 2018 109

simulation was performed in MS Excel, and the network diagram was generated by Microsoft Project 2016 (MS Pro- ject) (see Ökmen (2008) for project activities). Both MS Project and @Risk were used simultaneously for probability estimation in MS Excel. A new probabilistic schedule was created by MS Project for each iteration, and the schedule was transferred to MS Excel to estimate the cost of the project. This integration process involves some assumptions as listed below.

Assumptions of the model:

● Payments are based on the most likely case when milestone were completed

● If a milestone was delayed, there is a cost associated with the delay

● Risk factors increase only the cost of the project

● Delays are computed over the duration of the milestone.

In the model, simulated costs that are above the most likely costs shown in Exhibit 1 are considered a cost risk. Initially, optimistic (minimum) costs are assumed as the baseline cost. The total cost of the project increases due to risk factors as summarized in Equation 1,

TC ¼ Xn i¼1

CIi ¼ Xn i¼1

Oi þ ACið Þ ; i ¼ 1 ::: n (1)

Where, i denotes the index of the CI, and CIi, Oi and ACi represent the ith CI, the optimistic cost, and the additional cost value of risk factors, respectively.

The additional cost value of risk factors, ACi, is calculated as shown in Equation 2,

ACi ¼ Pi � Oið Þ � T (2)

where Pi represents the pessimistic cost of the i th CI, and T

represents a random variable using a triangular distribution with minimum, mean, and maximum values given by (0, 0.5, 1).

As the degree of influence of risk factors is not evaluated in Equation 2, a dynamic table was created to generate probabilistic costs as shown in Exhibit 1. Risk factors form a discrete prob- abilistic distribution ranging between optimistic and pessimistic.

Four categories for cost influence degrees are used: None, Low, Medium and High. As shown in Exhibit 3, influence degrees of the risk factors are used as input to the simulation. In each iteration, these influence degrees take different risk values and, hence assign new X values. The new X values are the weights of the risk factor and are used to generate random scenarios. The risk ratio of CIs and risk factors are evaluated as shown in Equation 3,

Exhibit 1. Cost Items (Milestones) for the Case Study (Ökmen, 2008)

Cost item description Optimistic cost ($) Most likelycost ($) Pessimistic cost ($)

1. Preliminary report 50,000 63,680 80,000

2. General layout plan 20,000 23,880 30,000

3. Upper plain plans and calculations 20,000 23,880 30,000

4. Pre-application projects 25,000 31,840 40,000

5. Post-application projects 150,000 199,000 300,000

6. Hydraulic structure projects 150,000 214,920 300,000

7. Operation and maintenance roads 14,000 15,920 18,000

8. Access road projects 5,000 7,960 9,000

9. Post-application projects of pump plans and elevation lines 25,000 31,840 35,000

10. Architectural and static projects of pumping stations 40,000 47,760 60,000

11. Channel cost estimation and measurement reports 30,000 39,800 50,000

12. Hydraulic structure and pumping station cost estimation and measurement reports 30,000 39,800 50,000

13. Green dossier, project reports 12,000 15,920 20,000

14. Submission, approval and reproduction of project originals 20,000 23,880 30,000

15. Final measurements and payment certificate 12,000 15,920 20,000

Total Project Cost 603,000 796,000 1,072,000

Exhibit 2. Risk Factors of the Real-World Case (Ökmen, 2008)

1. Design changes inside the firm

2. Design changes by the regional directorates

3. Design changes by the general directorates

4. Owner’s delay in approving the design drawings and reports

5. Reduced labor productivity

6. Temporary or permanent absence of the firm’s key designer staff

7. Bad weather conditions during the activities

8. Owner’s delay in payment

9. Disputes between the project participants

10. Delay in written project related communications within the owner’s organization

110 Engineering Management Journal Vol. 30 No. 2 2018

Risk ratio ¼ X1; X2; :::; Xnf g; p1; p2; :::; pnf gð Þ; (3)

where p denotes the probability of the occurrence of the risk factor. The risk ratio is a function of a set of X, which was generated by the @Risk model, along with associated p-values. In each category, the cost impact is directly proportional to the weight of the risk factor, X. However, the values that were generated from other categories also contribute to cost impact. The degree of influence takes values between [0, 10]. As shown in Equations 4–7, the range of X and p-values are assigned following Pehlivan and Öztemir (2015) and used to generate probabilistic risk factors.

No Risk ¼ X1 ¼ 0; P1 ¼ 1f g (4)

Low Risk ¼ X2 ¼ 1; P2 ¼ 0:3 X3 ¼ 2; P3 ¼ 0:4 X4 ¼ 3; P4 ¼ 0:3

8 < :

9 = ; (5)

Medium Risk ¼ X5 ¼ 4; P5 ¼ 0:2 X6 ¼ 5; P6 ¼ 0:3 X7 ¼ 6; P7 ¼ 0:3 X8 ¼ 7; P8 ¼ 0:2

8 >>< >>:

9 >>= >>;

(6)

High Risk ¼ X9 ¼ 8; P9 ¼ 0:3 X10 ¼ 9; P10 ¼ 0:4 X11 ¼ 10; P11 ¼ 0:3

8 < :

9 = ; (7)

Risk factors, RFij, which are calculated using Equations 4–7 are used for the additional cost value of risk factors ACRij of the ith CI of the jth risk factor as shown in Equation 8,

ACRij ¼ ðRFij= Xm k¼1

CIRFik Þ � Pi � Oið Þ � T; i ¼ 1 . . . n; j ¼ 1 . . . m (8)

where CIRFik represents the CI risk factor value of the i th CI of

the kth risk factor, Pi represents the pessimistic cost of i th CI, n

denotes the number of CIs, and m denotes the number of risk factors.

The model relies on the simulated cost risk values as calculated in Equation 8. This implies that the simulation model does not underestimate optimistic scenarios in the evaluation. Each ACRij value acquires a unique output given by the discrete probability distribution defined via Equations 4–7. To convert these risks into cost values, cost range of a CI is multiplied with the assigned continuous triangular probability distribution as shown in Equation 8, and risks are assigned a numeric value in terms of the cost impact.

Cost influence degrees are expressed numerically with Equation 3 and, in turn Equation 8 re-computes the cost impact as shown in Exhibit 4. Although the values in Exhibit 4 seem static, they are dynamic with respect to the analysis performed. The corresponding cell of “Risk #” and “CI” indicates the risk value of a specified cell. In addition to the impact of risk factors, project time should also be considered in cost estimation of CIs.

The duration uncertainty of project activities may lead to fluctuating costs. Therefore, MS Project is utilized to prepare a network diagram of activities. Initially, MS Project computes the project start and finish dates for each activity and CI. The assigned duration of a CI and the gap between the simulated duration is computed. The network duration of a CI is expressed as shown in Equation 9,

NDi ¼ Fdi � Sdi (9)

where NDi denotes the network duration of the i th CI, Fdi

denotes the finish date of the CI, and Sdi denotes the start date of CI. Further,

NDmi ¼ Fdmi � Sdmi (10)

NDsi ¼ Fdsi � Sdsi (11)

where NDmi, Fdmi and Sdmi represent the most likely net- work duration, the most likely end date and the most likely start date of the CI, respectively. In Equations 10 and 11, m stands for most likely duration. NDsi, Fdsi, and Sdsi repre- sent simulated network duration, simulated end date, and

Exhibit 3. Influence Degrees of the Risk Factors

RISK FACTOR

1 2 3 4 5 6 7 8 9 10

Preliminary report Medium High No High Medium High Medium Medium Medium High General layout plan High High No High Medium High Low High High High Upper plain plans and calculations Medium Medium No Low Medium Low Low Low Low High Pre-application projects High High No High Medium High Low Medium High High Post application projects Medium High High Medium Medium High High High High High Hydraulic structure projects High High Medium Medium Medium High Medium High High High Operation and maintenance roads Medium Medium Low Low Medium Low No Low Low Medium Access road projects Medium Medium Low Low Medium Low No Low Low Medium Post-application projects of pump plans and elevation Medium Medium Low Low Medium Low No Medium Low Medium Architectural and static projects of pumping stations Medium Medium Low Low Medium Low No Medium Low High Channel cost estimation and measurement reports Low No No No Medium Low No Low No Low Hydraulic structure and pumping station cost estimation and measurement reports Low No No No Medium No No Low No Low Green dossier, project reports No No No No Medium No No Low No Low Submission, approval and reproduction of project originals No No No Low Low No No Low Low No Final measurements and payment certificate No No No No Low No No Low No Low

COST ITEM

Engineering Management Journal Vol. 30 No. 2 2018 111

simulated start date of the ith CI, respectively. 00S00 stands for simulated duration.

The impact of duration variability on project cost needs to be computed simultaneously with the risk factors. How- ever, to complete this calculation, the uncertainty of activities must be analyzed. The network duration risk of a CI is expressed as shown in Equation 12,

Duration Risk ¼ Dri ¼ NDsi � NDmi þ

Pk¼1 i�1

Drk0 if ; Pk¼1 i�1

Drk<0

Dri ¼ NDsi � NDmi0 if ; Pk¼1 i�1

Drk>0

8 >>>< >>>:

(12)

where Dri represents the duration risk of the i th CI and Drk

represents the duration risk of the summed kth milestone. The simulated risk takes only negative Drk in calculating duration. This implies that the project was going ahead of schedule, which is calculated as cost decrease in the evaluation.

Schedule risk of the project is based on the delayed duration of activities. Furthermore, the extended duration of an activity increases cost, which may not be included in the most likely estimate. Delayed activities may have a serious impact on the corresponding CI. On the other hand, there could be a negative value (favorable) that will reduce the cost risk of CI. Since payments are based on the completion of milestones, the accu- mulated effect can increase or decrease the value of a CI. The cost of schedule risk of the ith CI is calculated as shown in Equation 13,

Si ¼ Xa i¼1

ðDri � CCIiÞ; i ¼ 1 . . . n ; (13)

where CCIi denotes the daily cost of delay for the CI. For estimating the delay cost of a CI on a daily basis, expected cost of a delay is divided by the duration of the activity as in Equation 14,

CCIi ¼ ðCIi � PeiÞ=Di (14)

where CIi denotes the estimated most likely cost, Pei denotes percentage cost value, and Di denotes duration (days) of asso- ciated activities of the ith CI.

The delay cost of a CI is based on the duration of related activities. For the sake of simplicity, all activities are assumed to have equal cost ratios in the simulation. However, in the model, expected cost impact of the delays can be adjusted by the given p-values. Integrating the cost and schedule risk determines the range of inherent cost risk of a project. Cost risk of a project is expressed with Equation 15.

IRC ¼ Xi¼1 n

IRCi ¼ Xi¼1 n

Oi þ Ci þ Sið Þ (15)

where IRC denotes the integrated risk of all CIs, Oi denotes the optimistic cost of the ith CI, Ci denotes the cost risk of the i

th CI, and Si represents cost of schedule risk of the i

th CI. Using Equation 15, the total cost risk is calculated to find

the probability of cost overruns. The risk of cost overrun is expressed as shown in Equation 16,

Budget Deficit ¼ Xn i¼1

IRCi � BAC (16)

where BAC represents BAC of the project. The BAC of a project is the sum of most likely values of CIs.

The difference between integrated cost risk and the most likely total cost indicates that there is additional budget that might be

Exhibit 4. Simulation Table of Cost Items and Risk Factors ($)

Cost Items Risk #1 Risk #2 Risk #3 Risk #4 Risk #5 Risk #6 Risk #7 Risk #8 Risk #9 Risk #10 Cost Risk

1. Preliminary report 1,154 2,308 0 2,308 1,154 2,308 1,154 1,154 1,154 2,308 63,680

2. General layout 649 649 0 649 325 649 130 649 649 649 23,880

3. Upper plain plans 714 714 0 286 714 286 286 286 286 1,429 23,880

4. Pre-application 1,042 1,042 0 1,042 521 1,042 208 521 1,042 1,042 31,840

5. Post application 4,412 8,824 8,824 4,412 4,412 8,824 8,824 8,824 8,824 8,824 199,000

6. Hydraulic structure 9,375 9,375 4,688 4,688 4,688 9,375 4,688 9,375 9,375 9,375 214,920

7. Operation and maintenance 333 333 133 133 333 133 0 133 133 333 15,920

8. Access road proj. 333 333 133 133 333 133 0 133 133 333 7,960

9. Pump plans and elevation lines 758 758 303 303 758 303 0 758 303 758 31,840

10. Architectural and static 1,316 1,316 526 526 1,316 526 0 1,316 526 2,632 47,760

11. Channel cost estimation 1,538 0 0 0 3,846 1,538 0 1,538 0 1,538 39,800

12. Hydraulic structure & pump 1,818 0 0 0 4,545 0 0 1,818 0 1,818 39,800

13. Green dossier, project reports 0 0 0 0 2,222 0 0 889 0 889 15,920

14. Submission, approval 0 0 0 1,250 1,250 0 0 1,250 1,250 0 23,880

15. Measurements and final payments 0 0 0 0 1,333 0 0 1,333 0 1,333 15,920

Total 23,443 25,652 14,607 15,730 27,750 25,117 15,289 29,977 23,675 33,260 796,000

112 Engineering Management Journal Vol. 30 No. 2 2018

required in pessimistic scenarios, that is, cost overrun. There- fore, the impact on cost figures can increase the schedule dura- tion or vice versa. Integrating schedule and cost risks in a single model is expected to give more realistic estimates than tradi- tional cost estimates. The developed model is applicable for all types of construction projects by adjusting the existing risk factors and their impact on the project. For non-construction projects according to specifications, some model modification may be required, such as impact of delay.

Results and Conclusions The total duration of the project was increased to 514 days since the duration of project completion was underestimated in the contract. The simulated total duration was similar to the actual total duration shown in study of Ökmen (2008). In addition, activities were investigated individually and the impact of activ- ities were gathered as given in Exhibit 5.

The probabilistic cost variability is shown in Exhibit 5 with its cumulative density function (CDF). As a result of applying the integrated risk approach, new pessimistic scenarios were produced to compare factors. In Exhibit 5, the probabilities of finishing the project under budget with cost risk and with integrated risk are shown with percentile values. Consequently, the shifted curve of risk factors indicates the cost impact of delays, which create cost overruns. The results from the simulated model for the cost over- run for each CI are summarized in Exhibit 6.

The findings of integrated risk model are realistic and compatible with the theoretical expectancies as well as actual costs as demonstrated by Ökmen (2008). In this regard, model- ing the cost uncertainty of the CIs helps to understand not only the factors affecting cost increases but also to generate realistic scenarios. In this study, selected p-values are used to present the level of risk for the potential cost impact of risk factors. Accord- ing to Hulett and Nosbisch (2012), the decision regarding per- centile values (p-values) depends on the risk management decision. Although some experts suggest their clients use 80% of the scenarios for selecting cost and schedule, some organiza- tions officially choose 50%, 65%, or 70% which are more risky than the suggested p-value. In some cases, this value may reach up to 90% to be more conservative in dealing with risk.

It is important to determine the dominant factor that results in budget deficits. In Exhibit 7 the cost impact of delays is compared with risk factors in terms of cost sensitivity. Results indicate that the delays have a stronger impact than the existing risk factors. The outcomes of the analysis are ranked with respect to the effect on total cost. The top ranking factors are “Approval of post-applica- tion of projects,” “Approval of hydraulic structure projects,” and “Approval of hydraulic structure and pumping station estimation.” The delay of these activities has a significant impact on total cost, when compared against other risk factors. High sensitivity for an activity duration also indicates where the project needs more attention. The cost impact in Exhibit 7 represents risk factors and delayed activities that have influence on total cost. This analysis identifies the crucial activities in which precautions, such as crash- ing activities, should be applied.

To assess the impact of schedule variability, the tested risk levels are 20%, 50%, and 80% Pe values. As a result, a significant increase is observed in costs, as summarized in Exhibit 8, with different risk levels. Values indicate that the total cost of the project is the same for various scenarios. There were cost differ- ences between various CIs. Likewise, the total cost of various project scenarios is the same at the completion of CI-15. For instance, the mean value of total cost in the 20% Pe scenario intersects with the minimum value of the total cost with the 50% Pe scenario. However, cost variability among the assigned sce- narios is significantly different at various times of the project. After Milestone 4 in particular, there is a substantial increase in the budget deficit. Moreover, Milestone 9 is another critical milestone, where the budget deficit significantly rises. Conse- quently, outcomes of the analysis are important in determining where the budget deficits could occur.

Scenarios were selected with uniformly changed Pe values for all activities. The model performs as expected by using Pe values to represent the impact of schedule delays. In Exhibit 8, two critical time frames are identified where the risk of a budget deficit significantly increased. These time frames cover more than one milestone. The beginning point for cost overruns is important in monitoring the progress of the project. Milestone (MS)-4 and Milestone (MS)-9 have a potential schedule risk for budget deficits. With the increase in state of Milestone (MS)-4, other Milestones are expected to be influenced by this cost

Exhibit 5. Total Cost of the Project

Engineering Management Journal Vol. 30 No. 2 2018 113

overrun. Thus, Milestone (MS)-9 is expected to cause the same problem.

A non-critical path may increase the cost without affecting completion times. The proposed method of integration was useful to examine the CIs individually in regards to cost increase. This integration process has a potential to be applied to other industries. To apply the method to different industries, the risk factors should be modified. To set the Pe values, risk factors should be identified and then the integration process can be implemented.

Implications for Engineering Managers The evaluation of a project from the perspective of a con- tractor is examined, and a model is proposed in this article. The integration process is suitable for PEO-type construction projects, where payments are based on the completion of events. Therefore, this method is suitable not only for con- struction but also other industries. To apply to other project types, cost and schedule of a project can be integrated in the same proposed way, but risk factors should be revised accord- ing to the risks involved in that project. Therefore, the sug-

Exhibit 6. Delayed Schedule and Integrated Cost Risk Effect on Cost Items

Most likely cost items ($) CIs with cost risk only ($) Cost of delays ($) Integrated(cost+schedule) risk ($) CIs with integrated risk ($)

Cost item 1 63,680 64,984 2,074 3,378 67,058

Cost item 2 23,880 25,003 5,799 6,923 30,803

Cost item 3 23,880 24,993 6,140 7,253 31,133

Cost item 4 31,840 32,492 7,177 7,829 39,669

Cost item 5 199,000 224,999 11,525 37,524 236,524

Cost item 6 214,920 224,919 8,562 18,561 233,481

Cost item 7 15,920 16,000 4,127 4,207 20,127

Cost item 8 7,960 7,020 2,236 1,296 9,256

Cost item 9 31,840 29,995 4,234 2,389 34,229

Cost item 10 47,760 50,013 16,699 18,953 66,713

Cost item 11 39,800 39,986 15,778 15,964 55,764

Cost item 12 39,800 39,869 15,857 15,926 55,726

Cost item 13 15,920 16,009 6,280 6,369 22,289

Cost item 14 23,880 25,034 6,940 8,094 31,974

Cost item 15 15,920 16,031 10,924 11,034 26,954

Total Cost 796,000 837,348 124,353 165,701 961,701

Exhibit 7. Impact of Risks on Total Cost ($)

114 Engineering Management Journal Vol. 30 No. 2 2018

gested model is practical, effective, and can be applied in various project types.

The sensitivity ranking of integrated cost risks is produced first to identify activities that have the potential to reduce both cost and schedule risks. This ranking also reveals non-critical activities. There- fore, the methodology presented in this article can assist engineering managers in understanding the causes of unexpected cost overruns. Total cost overrun causes can potentially be identified by the meth- odology presented in this study. Additionally, the sensitivity ranking identifies activities among which the reduction of total cost has the highest priority and should therefore be prioritized. Therefore, the developed model is expected to guide practicing engineering man- agers in decision making regarding the negotiation processes to convince the client by using simulated new budget allocations for each CI. Captured probabilistic costs can be compiled to discover break points which reveal the time frame for cost overruns. Further- more, the analysis reveals critical time frames where financial support may be required.

The developed simulation model also highlights potential cash flow issues for a contractor who has a limited budget or who runs several projects simultaneously. The simulation model may also be useful to practicing managers in estimating the probability of cost overruns in a particular project. Identifying the delayed project in advance may save the investment that might be lost for both parties (contractor and client) in terms of time, effort, and money. This may lead to disputes between the parties. For this reason, risk results of a project should be analyzed to find alternative solutions which can be discussed in advance of further stages in a project.

Acknowledgements The authors would like to thank the three anonymous reviewers and the editors of the Engineering Management Journal (EMJ) for their constructive feedback and suggestions. These suggestions have been extremely useful in improving the paper to its current version.

References Adeli, H., & Karim, A. (1997). Scheduling/Cost optimization

and neural dynamics models for construction. Journal of

Construction Engineering, 123(4), 450–458. doi:10.1061/ (ASCE)0733-9364(1997)123:4(450)

Assaf, S. A., & Al-Hejji, S. (2006). Causes of delay in large construction projects. International Journal of Project Man- agement, 24(4), 349–357. doi:10.1016/j.ijproman.2005.11.010

Chang, Andrew Shing-Tao. (2002). Reasons for cost and sche- dule increase for engineering design projects. Journal Of Management In Engineering, 18(1), 29–36. doi: 10.1061/ (ASCE)0742-597X(2002)18:1(29)

Cho, K., Hong, T., & Hyun, C. (2010). Integrated schedule and cost model for repetitive construction process. Journal of Management in Engineering, 26(2), 78–88. doi:10.1061/ (ASCE)ME.1943-5479.0000009

Cohen, M. W., & Palmer, G. R. (2004). ’Project risk identifica- tion and management’, AACE International Transaction, (pp. 1013-1015). Retrieved September, 10, 2006.

Hulett, D. (2012). Integrated cost-schedule risk analysis. Surrey, UK: Gower.

Hulett, D., & Nosbisch, M. R. (2012). Integrated cost and schedule using Monte Carlo simulation of a CPM model. Waste Man- agement (WM 2012) Conference, Phoenix, Arizona, USA.

Lim, C., & Mohamed., M. Z. (2000). An exploratory study into recurring construction problems. International Journal of Project Management, 18(3), 267–273. doi:10.1016/ S0263-7863(99)00016-2

Ökmen, Ö. (2008). Activity network scheduling and early cost estimation of construction projects under uncertainty: A risk analysis based modeling approach. Gaziantep, Turkey: Gaziantep University.

Ökmen, Ö., & Öztaş, A. (2008). Construction project network evaluation with correlated schedule risk analysis model. Jour- nal of Construction Engineering and Management, 134(1), 49– 63. doi:10.1061/(ASCE)0733-9364(2008)134:1(49)

Omar, A. (2009). Uncertainty in project scheduling–its use in PERT/ CPM conventional techniques. Cost Engineering, 51(7), 30–34.

Pehlivan, S., & Öztemir, A. E. (2015). Cost overrun estimation of construction project with significant risk factors. The Inter- national Journal of Engineering and Science, 4(7), 1–6.

Exhibit 8. Impact Scenarios of Delayed Milestones

-30,071.06

-92,732

-1,65,701

-1,42,729.30

-2,42,180.00

-3,62,203.60

-4,00,000

-3,50,000

-3,00,000

-2,50,000

-2,00,000

-1,50,000

-1,00,000

-50,000

0

50,000

T ot

al B

ud ge

t D ef

ic it

($ )

Milestones of the Project

Minimum 20 % Mean 20 % Maximum 20 %

Minimum 50 % Mean 50 % Maximum 50 %

Minimum 80 % Mean 80 % Maximum 80 %

Beginning of first budget deficit risk

The second budget deficit risk

Engineering Management Journal Vol. 30 No. 2 2018 115

Project Management Institute. (2013). A guide to the project management of knowledge (PMBOK guide). Newtown Square, PA: Author.

Sambasivan, M., & Soon, Y. W. (2007). Causes and effects of delays in Malaysian construction industry. International Journal of Project Management, 517–526. doi:10.1016/j. ijproman.2006.11.007

Schatteman, D., Herroelen, W., Vonder, S. V., & Boone, A. (2008). Methodology for integrated risk management and proactive scheduling of construction projects. Journal of Construction Engineering and Management, 134(11), 885– 893. doi:10.1061/(ASCE)0733-9364(2008)134:11(885)

Sweis, G., Sweis, R., Hammad, A., & Shboul, A. (2008). Delays in construction projects: The case of Jordan. International Journal of Project Management, 26, 665–674. doi:10.1016/j. ijproman.2007.09.009

About the Authors Salahi Pehlivan is a lecturer in the Industrial Engineering Department at Girne American University, Cyprus, where

he teaches modeling and optimization, and project manage- ment courses. He gained his PhD in Construction Manage- ment from Girne American University. He worked as an Investment Analyst/Project Manager for an international company for training programs (Investment Appraisal and Risk analysis, and Project Management) conducted for pro- fessionals in Africa, Asia, and Canada. His areas of interest are risk analysis, simulation, and cost estimation.

Ali Erhan Öztemir is a lecturer in the Construction Manage- ment/Civil Engineering Department at the Cyprus International University, Cyprus, where he teaches courses on Construction Management, Construction Engineering, and Civil Engineering Theory. He holds a PhD. in Civil Engineering from Arizona State University. Ali Erhan has conducted research in U.S. on construc- tion optimization. His main areas of interest include Construction Risk analysis. Ali Erhan’s professional experience includes work at various construction sites in Turkey and Cyprus.

Contact: Salahi Pehlivan, Girne American University, Faculty of Engineering, Industrial engineering department, Girne, North Cyprus, Mersin 10, Turkey; salahipehlivan@gau.edu.tr

116 Engineering Management Journal Vol. 30 No. 2 2018