Monte Carlo Simulation for Enhancing the Schedule Completion Forecast of Jakarta Central Railway Station Construction Project

Abstract

Some construction projects in Indonesia have been experiencing frequent major delays, and there is an urgent need to perform a comprehensive schedule forecast to anticipate them. This study aims to enhance the forecast of the project-completion schedule of a major railway construction project in Indonesia, with considerations of the identified project risks. Using structured interviews of practitioners and experts, this study has identified key risks in the Phase II Construction project. The @RISK V8.5.1 software was used to run Monte Carlo Simulation to calculate the scheduling forecast more accurately with the PERT method. The results of the scheduling simulation on the inherent key risks that affect key activities in the critical path show that there are additional days due to pre-mitigation during the construction of the project, ranging from 130.81% to 136.25% of the initial contract completion duration to approximately 95.82% to 102.23% of the original contract completion duration. This research provides a way not only to identify and prioritize the risks but also to quantify them to estimate the project completion schedule more up to 95% accuracy using Monte Carlo Simulation, where the model can be used to plan and monitor the risks of future projects.

1. Introduction

The infrastructure development program is the main target of the Indonesia government, which aims to grow the economy and national connectivity. Limitations and delays in infrastructure development in Indonesia make it difficult for Indonesia’s economy to develop. Infrastructure investment in Cameroon heavily influences economic and private-sector growth [1]. Transportation infrastructure is important in supporting economic growth in the United Kingdom. Improving connectivity, and facilitating trade and investment are two of the goals of building a strong infrastructure supporting economic development in the United Kingdom [2]. By influencing production and consumption patterns and promoting sustainable economic development through higher tourism and private consumption, rail infrastructure impacts economic growth in China [3]. Transportation infrastructure, particularly railways, significantly affects economic growth by improving connectivity, reducing transportation costs, and facilitating trade. Investment in railway projects can increase efficiency in the movement of goods and people, stimulating the local economy and attracting investment [4]. According to [5], delays in railway projects significantly hamper economic growth, as timely implementation of transport networks is critical to development. In developing countries such as Vietnam, economic growth is one of the significant obstacles, one of which is the problem of postponing the railway project schedule [6]. Despite the performed studies in emerging markets, they lack evidence in the context of complex construction work, such as central station building, due to interdependencies with several factors, such technical, social, interface-management, and any other related factors.

One of the ways to support the government’s program in growing the economic sector in Indonesia is to revitalize the Jakarta Central.

Jakarta Central Railway Station Construction Project Phase II, one of the largest and busiest central stations in Indonesia, has a very complex location, and the implementation of the project has experienced a very significant delay. According to the initial contract, the construction of the Railway Station Phase II project began on 11 October 2019 and is planned to be completed on 30 September 2021.

Based on Figure 1, which was obtained from the author’s research results through project archive studies, the construction of Railway Station Phase II should have been completed on 30 September 2021; however, the actual progress achieved on that day was merely 42.08%, resulting in a delay deviation of −57.92% [7]. However, significant delays led to the project’s completion on 31 January 2024, a delay of approximately 2.5 years where the risks are identified and qualitatively assessed, as shown in Figure 1.

The delayed construction of the station has negatively impacted the project owner, causing increased operational challenges and financial losses. Furthermore, the commuters were disappointed, and it led to negative perceptions and complaints that spread across the train users. Project owners face reputational risks, while service providers struggle [8].

The delay problem occurs because, in managing the project and the estimated schedule and controlling the project schedule with ineffective risk management, minimal risk management will have an impact on risk. Risk analysis and probabilistic methods are used for scheduling simulations, ensuring timely completion and minimizing delays [9].

This research aims to identify the risks in the project causing the delay, assess them using quantitative risk analysis, and enhance the project completion forecast based on the prioritized risks with PERT and Monte Carlo Simulation. It examines several occurrences of delays in global railway projects, using the Jakarta Central Railway Station Phase II Construction in Indonesia as a case study. Furthermore, to mitigate the unpredictability of scheduling risk, the PERT analysis approach is employed to estimate the time of essential tasks via three-point estimation. This study uses the Monte Carlo Simulation probabilistic method to assess the influence of risk, estimate variations in activity length, and predict project completion time with greater precision.

This research is grounded in a case study of the Phase II Development of the Jakarta Central Railway Station, with a focus on three main problem formulations:

(a)

What are the indicators that cause the risk of delay in completing the construction project (RQ1)?

(b)

How can the risk-based simulation of the project implementation scheduling be established to optimize the project’s completion time (RQ2)?

(c)

What are the appropriate risk response strategies and recommendations to respond to the risks that have a major impact on the construction schedule to optimize the project completion time (RQ3)?

This study’s primary benefit is quantifying the risk of project delays, which can be used as lessons to prevent future occurrences, particularly in railway station construction projects.

2. Literature Review

A theory is defined as a collection of logically coherent statements regarding a phenomenon that effectively encapsulates current empirical understanding, systematically arranges this knowledge through precise relational statements among variables, offers an explanation for the phenomenon, and serves as a foundation for predicting behavior [10]. In risk management, theory provides the foundation for understanding, predicting, and managing risk based on basic principles and assumptions. Some of the key theories that are relevant to explaining the concept of risk and risk analysis include probability theory, rational model theory, utility theory, and prospect theory. Theory used in this study is utility theory. Utility theory is used in risk analysis to evaluate decision scenarios under uncertainty by selecting options that maximize expected utility. This theory is also applied in risk modeling and simulation to simulate decision-making behavior in complex systems and test various scenarios [11]. The use of utility theory on EMV (Expected Monetary Value) helps to measure not only the expected number of risky decisions but also to consider the expected satisfaction based on risk preferences. EMV is a risk analysis tool that measures the financial impact of potential events by multiplying their probabilities and impacts, helping project managers make data-driven decisions to choose the best risk mitigation strategies.

2.1. Program Evaluation and Review Technique (PERT)

From some of the literature reviews mentioned above, it can be concluded that the delay of railway projects is a serious problem in the world that needs to be solved, including in Indonesia. To address the issue above, one approach is to employ a Monte Carlo Simulation to incorporate uncertainty and generate a more precise range of scheduling solutions, enhancing the accuracy of the entire scheduling-estimation process. Simultaneously, the Program Evaluation and Review Technique (PERT) offers a probabilistic method for estimating activity durations through a three-point estimation methodology [12]. The framework will employ Monte Carlo Simulation to estimate fluctuations in activity duration and anticipate project completion time with greater accuracy. The simulation-based methodology enhances project schedule risk analysis by including novel equations for quantifying duration risk in PERT method analysis. This approach facilitates improved decision-making in ambiguous contexts [13]. Efficient scheduling is crucial for railway projects to minimize delays and ensure timely completion. PERT methodology and Monte Carlo Simulation assess project-completion probability and identify time-critical paths.

The Program Evaluation and Review Technique (PERT) method is a dependable analytical approach for managing delays in railway project scheduling. The PERT method facilitates time estimation and project monitoring through a systematic methodology [14,15,16]. PERT offers systematic scheduling and resource estimation methodology, aiding in managing identified risks among uncertainty. According to [17], the delineation of PERT constitutes a methodology for approximating project duration through a weighted mean of optimistic (O), pessimistic (P), and most likely (ML) to estimate activity durations in the presence of ambiguity regarding individual activity assessments. The PERT equation comprises three elements (O, P, and ML) that employ a weighted mean of the three approximated values, presupposing a beta probability distribution for the temporal estimations, as delineated by [16].

$$T_e={\displaystyle \frac{O+4ML+P}{6}}$$

(1)

where $T_e$ represents the expected time (in days); optimistic (O), in days, denotes the minimum duration required to accomplish the task; most likely (ML), in days, signifies the duration with the greatest likelihood of completion; and pessimistic (P), in days, shows the maximum duration necessary to finish the task. According to [18], the timely completion of a project by the planned likelihood is assessed with a normal distribution, utilizing the subsequent formula:

$$Z={\displaystyle \frac{D-T_e}{\sqrt{{\sigma }_{cp}^2}}}$$

(2)

where Z is the standard measurement of the deviation from the actual value to the mean value, ${\sigma }_{cp}^2$ (in days) shows the variation in individual activities on the critical path, and D (in days) indicates the likelihood that the period will be fulfilled by the deadline.

The studies of [5] stated that the PERT method was used to analyze the delays in the scheduling of railway projects. Meanwhile, ref. [13,19] utilized a combination of AHP and PERT in optimizing the schedule-completion forecast in their study. Previous studies have been conducted using various techniques, which have resulted in differing risk variable assessments, as each project’s geographical conditions will vary according to its cultural context. According to a study by [20], focus-group discussions (FGDs) serve to accurately evaluate the risks and impacts of project activities, as respondents are selected among experts engaged in the relevant case studies. Subsequently, data analysis will employ critical path data and the Critical Path Method (CPM) on historical project scheduling data, along with PERT, to forecast duration uncertainty and mitigate project risks [13,21]. Meanwhile, ref. [22] used both PERT and M-PERT to compare the scheduling result for a construction project.

Furthermore, some studies employed the Monte Carlo technique with @Risk software as simulation method, which successfully analyzes the probability of project completion. Refs. [12,13,23], applied Earned Value Management, CPM, and PERT in estimating construction project duration and costs. A risk-based approach to estimate the duration of the project schedule will be used in performing the calculation using the probabilistic method PERT with help from Monte Carlo Simulation.

2.2. Monte Carlo Simulation (MCS)

Quantitative risk analysis is a mathematical method used to predict the likelihood and consequences of risks, identifying primary reasons for construction project delays by analyzing literature data and assessing their impact levels. According to [13,24], qualitative risk analysis evaluates potential project hazards through a systematic methodology employing numerical and static methodologies. This method aims to assess the potential impact of diverse risks to enhance decision-making and resource-allocation efficiency. According to Nguyen, qualitative analysis can integrate findings with quantitative methods such as Monte Carlo Simulation (MCS) to evaluate the overall risk value. The impact of risk and the development of effective risk management strategies can create a better understanding of risk [6]. We propose that employing a novel equation for total risk value alongside Monte Carlo Simulation enables the estimation of activity duration variability. This approach is very effective for assessing the likelihood of train delays. This strategy can enhance the accuracy of project completion time predictions by adhering to deadlines [13]. This is an illustration of a Monte Carlo probability distribution, as suggested by [25].

Figure 2 illustrates an example of Probability Density Distribution for Input and Cumulative Probability Distribution derived from the study of [25]. The procedure and process for Monte Carlo Simulation analysis is as follows: (1) Probability distribution of many random variables and stochastic experimental values is sampled. (2) Input sample data are processed to yield computed results inside a deterministic model. (3) The selected outcome values for this experiment, including the time and cost for completion, are recorded in a file. (4) To achieve a substantial quantity of experimental results and the desired precision, revert to step 1. (5) Preserving analytical outcomes. Repeated and uniform distribution sampling will yield an estimate of the expected time. If the value is mathematically proximate, then the outcomes of random sampling will be acquired:

$$V=\underset{-\infty }{\overset{\infty }{\int }}X\left(fx\right)dx$$

(3)

where V is the expected value (in days), x represents the outcome value (in days), f(x) is the probability density function, and $dx$ represents the infinitesimal elements of random variables $x$ (in days). The straightforward yet sophisticated simulation process conducts integration on our behalf. The integration of X(fx) is highly significant in most pertinent assessment issues. According to [26], the following formula determines the number of iterations:

$$\sigma =\sqrt{{\displaystyle \frac{{\sum }_{i=1}^k{(x_i-\overline{x})}^2}{k-1}}}$$

(4)

$$\epsilon ={\displaystyle \frac{\overline{x}}{\left(\frac{1}{RE}\right)}}$$

(5)

$$n={\left({\displaystyle \frac{3\sigma }{\epsilon }}\right)}^2$$

(6)

where σ (in days) is the standard deviation, x_i (in days) is the data value of each population member, $\overline{x}$ (in days) is the overall data mean, k is the number of data, ε (in days) is the absolute standard error, RE is the maximum error value, and n is the number of iterations.

This study applies the PERT technique to refine the schedule completion forecast using the Monte Carlo Simulation, with estimated duration based on the identified risks using the structured risk from previous study. Confidence levels of 80%, 90%, and 95% were chosen to increase the likelihood and enhance the schedule completion forecast based on the calculated duration estimates. The refinement of the risk-based calculation method shall be the major contribution of this study.

3. Research Methodology

According to [27], there are three conditions that determine the research strategy: the form of the research question (RQ), control over the events being researched, and focus on contemporary events (Contemporary/historical events). A research strategy was established for each research question in this study. For the RQ1 research process, the methodology is based on a literature review, archival analysis, and historical data from case studies. Subsequently, expert validation is performed by qualitative analysis utilizing the Delphi method and focus-group discussions (FGDs). This research method is to identify the indicators associated with the reasons for delays in the completion of the Railway Station Phase II Construction Projects. Meanwhile, to answer RQ2, a literature review, archival analysis, and historical data examination were conducted. Additionally, qualitative analysis encompassing the pre-risk mitigation phase was performed throughout the focus group-discussion phase. Finally, to answer RQ3, PERT techniques and Monte Carlo Simulation with @Risk Software from Lumivero for quantitative analysis were established to create a simulation for scheduling the implementation of Phase II of the Railway Station Development Project, encompassing both post-risk mitigation phases.

The Critical Path Method (CPM) quantitative analysis regarding essential scheduling activities for RQ2 and RQ3 was conducted utilizing archive data from a case study involving MS Project V2016 software. The calculated values from RQ2 and RQ3 will be treated as inputs to estimate the total duration, which is calculated using MS Project by entering the calculated PERT values, theoretical values, and simulated values. Those three types of values are compared, and deviations with the initial baseline based on the contract are also calculated. The calculation is performed based on 22 of the initial 51 risk factors for railway project delays, derived from various literature surveys and research findings from the previous research results [7]. The qualitative analysis is employed to validate the data of 51 variables related to railway project delay risk, where 22 risks were validated, while the rest were eliminated due to insignificance of the delay, as per the findings of that study. Furthermore, confidence levels (α) of 80%, 90%, and 95% are suggested according to various previous studies, delineating the confidence level (α) value and the number of iterations from various literature reviews on railway project delays [6,13,28,29]. This study differentiated the confidence level (α) according to theoretical and simulated values. This study also suggested 10000 iterations in order to have much better results.

The criteria for survey respondent profiles are essential for guaranteeing the reliability and validity of survey outcomes. These criteria include several elements, such as demographic attributes, recruitment techniques, and response behaviors, which collectively influence the quality of the obtained data. The respondent experts consist of five individuals who are both physically and mentally healthy; possess at least a bachelor’s degree in civil engineering; and have academic or practical experience in railways and the construction of buildings, bridges, or railway operations facilities, with a minimum of ten years of relevant work experience. Five respondents were selected: four of them are from contractors, and one of them is an academic, with experience ranging from 18 to 37 years.

4. Discussion

4.1. Data Collection for, and Analysis and Discussion of Research Question 1 (RQ 1)

The initial step of RQ1 data collecting is derived from the findings with additional data based on the research conducted by [7]. The data collection process at this stage begins with the gathering of both primary and secondary data from the case studies, and this process involves locating various literature articles concerning railway project delays globally and compiling historical and archival data from these case studies.

The qualitative analysis was performed to identify the most significant risk in each critical scheduling action. A focus-group discussion with five railway specialists was performed to identify the primary variables influencing the completion time of a project work activity. This study employed focus-group discussions (FGDs) to identify the primary risk indicators influencing scheduling, following these steps: (1) Experts were briefed on the overarching objectives, background, and research subjects, and allowed a question-and-answer session to synchronize research perspectives. (2) Experts were allotted time to review and complete the distributed questionnaire. (3) Discussions ensued regarding the responses from all experts. (4) Each expert elucidated their questionnaire responses and engaged with feedback from fellow experts. (5) All experts concurred on the outcomes regarding the predominant hazards associated with each essential activity of the railway station project phase II.

Table 1. Validation results of delay indicators by experts for the Jakarta Central Railway Station project phase II.

Risk ID	Risk Factor	Risk Categories	Reference
R1	Land acquisition delays	(I) Technical risk	[30,31]
R2	Delay in handover of work land
R3	Many utilities are not detected at the start of work.
R4	Difficult and limited project road access
R5	The project owner’s time-consuming decision-making process	(II) Project owner risk	[30,32,33,34,35]
R6	Delay in issuing work permits
R7	Changes in scope of work
R8	Inaccuracy in predicting the budget
R9	Project delay by the owner
R10	Sequencing work activities must be adjusted to train operating patterns (switch over)	(III) Construction risk	[24,32]
R11	Inconsistencies in the order of schedule priorities that can affect project execution
R12	Many changes to the design of drawings, specifications, and scope-of-work packages during the project by the project owner/external	(IV) Project design and methods risk	[5,30,32,35,36,37]
R13	Incomplete design from the planner
R14	Request for changes in work method/design from clients/external parties during the construction period
R15	Design inconsistencies and ambiguity
R16	Project complexity (design)
R17	Lots of additional work	(V) Contract risk	[32,36,38]
R18	High indirect costs	(VI) Financial risk	[5,30,36]
R19	There are complicated procedures for permitting construction projects from various parties.	(VII) External risks	[5,24,30,32,33,34]
R20	Interfaces, dependencies between phases, and work packages
R21	High-risk project location	(VIII) HSE and environmental risks	[24,30,35]
R22	Working on active train tracks

Source: [7] and author’s research.

shows the 22 validated risks from the previous study.

According to the expert validation results in Table 1, there are eight risk category groupings derived from 22 validated risk variables about delays in the Jakarta Central Railway Station Project Phase II ([7] and author’s Research), specifically (I) technical risk, (II) project owner risk, (III) construction risk, (IV) project design and methods risk, (V) contract risk, (VI) financial risk, (VII) external risks, and (VIII) HSE and environmental risks. Two risk category groupings emerged from the expert validation results, characterized by the highest number of risk indicator variables: project owner risk, which has five risk indicators: (R5) the project owner’s time-consuming decision-making process, (R6) delay in issuing work permits, (R7) changes in scope of work, (R8) inaccuracy in predicting the budget, and (R9) project delay by the owner. A primary factor contributing to project delays is the deficiency in owner competencies and strategies for managing mega projects [30]. The owner’s contribution is the primary cause of the project delay [33]. Each cause attributed to the owner significantly impacts project delays, either moderately or severely [35]. There are five risk indicators in the Design and Methods Risk Project: (R12) many changes to the design of drawings, specifications, and scope-of-work packages during the project by the project owner/external; (R13) incomplete design from the planner; (R14) request for changes in work method/design from clients/external parties during the construction period; (R15) design inconsistencies and ambiguity; and (R16) project complexity (design). As stated by [39], design risk is significantly affected by the designer’s critical role in comprehending and fulfilling the project owner’s wishes and needs. Risks during this phase are governmental bureaucracy, design inaccuracies, and modifications initiated for clients and contractors. Additionally, the final improper work method can cause project delays, as stated by [5].

4.2. Data Collection for, and Analysis and Discussion of Research Question 2 (RQ 2)

The preliminary phase of data collection is based on the results of the study conducted by [7] and other author’s research. The data collecting process at this stage initiates with the gathering of information about research question 1 (RQ1), specifically focusing on 22 indicators of delay risk that may influence the delay of the Jakarta Central Railway project phase II. Utilizing historical and archival data from the case-study project schedule, each essential activity that may impact project delays is identified by the essential Path Method (CPM) in the MS Project. The case-study scheduling identifies nine essential tasks, as detailed in

Table 2. The dominant risk in critical work activities.

No. Activity	Critical Work Activity	Duration (in Days)	Risk ID	Dominant Risk
A	Demolition Zone 1	50.00	R10	Sequencing work activities must be adjusted to train operating patterns (switch over)
B	Building Work Zone 1	290.00	R6	Delay in issuing work permits
C	Demolition Zone 2	25.00	R20	Interfaces, dependencies between phases, and work packages
D	Track Work Zone 2	122.00	R22	Working on active train tracks
E	Demolition Zone 3	29.00	R6	Delay in issuing work permits
F	Track Work Zone 3	114.00	R22	Working on active train tracks
G	Demolition Zone 4-Sub A	44.00	R6	Delay in issuing work permits
H	Building Work Zone 4-Sub A	196.00	R5	The project owner’s time-consuming decision-making process
I	Safety Assessment	29.00	R19	There are complicated procedures for permitting construction projects from various parties.

Source: [7] and author’s research.

: Demolition Zone 1, Building Work Zone 1, Demolition Zone 2, Track Work Zone 2, Demolition Zone 3, Track Work Zone 3, Demolition Zone 4-Sub A, Building Work Zone 4-Sub A, and Safety Assessment.

Qualitative analysis is essential for identifying the most significant risk in each critical scheduling action. A focus-group discussion with five railway specialists was performed to identify the primary variables influencing the completion time of a project work activity. This study employed focus-group discussions (FGDs) to identify the primary risk indicators influencing scheduling, following these steps: (1) Experts were briefed on the overarching objectives, background, and research subjects, and allowed a question-and-answer session to synchronize research perspectives. (2) Experts were allotted time to review and complete the distributed questionnaire. (3) Discussions ensued regarding the responses from all experts. (4) Each expert elucidated their questionnaire responses and engaged with feedback from fellow experts. (5) All experts concurred on the outcomes regarding the predominant hazards associated with each essential activity of the Jakarta Central Railway Project Phase II. Furthermore, the experts were asked to identify the critical work activities, and those were paired with identified risks; hence, they were categorized as dominant risks, as shown in Table 2.

According to Table 2, the predominant risk in each important work activity of the case study (Jakarta Central Railway Station Phase II) indicates, based on the FGD results, that there are six primary risk indicators in each critical schedule activity, with the most significant risk being R10. In (R1), there is one critical activity: Demolition Zone 1. In (R6), there are three critical activities: Building Work Zone 1, Demolition Zone 3, and Demolition Zone 4-Sub A. In (R20), there is one critical activity: Demolition Zone 2. In (R22), there are two critical activities: Track Work Zone 2 and Track Work Zone 3. In (R5), there is one critical activity: Building Work Zone 4-Sub A. In (R19), there is one critical activity: Safety Assessment.

Following the study in Table 2, a subsequent FGD was performed with experts to ascertain the additional days required due to the predominant risk associated with each activity, utilizing quantitative PERT analysis. Each expert delivered an evaluation and elucidation of the delay in project completion (in days) attributable to the predominant risk, utilizing the PERT distribution approach across the pessimistic (P), most-likely (ML), and optimistic (O) categories. All experts concurred with the definitive conclusion, including the inclusion of days attributable to inherent risk (pre-risk mitigation).

Adjusting the estimated results from PERT requires a conversion of the confidence level (α). The estimated value at each confidence level varies according to its corresponding time value. Furthermore, the theoretical value is derived from the PERT number formula, whereas the simulation value is determined by the number of iterations conducted in the model.

The simulation results are chosen and presented for the activity that had a substantial increase in duration. Utilizing Risk software for nine activities, A to I, as per Table 2, the Monte Carlo Simulation was performed, and the comprehensive simulation results are presented in detail in

Table 3. Quantitative analysis of Monte Carlo Simulation using @Risk software with PERT distribution approach on inherent risk conditions (pre-risk mitigation).

No. Activity	Duration (Days)	Risk ID		Additional Days (Inherent Risk)			PERT (Day)	Theoretical Value (in Days)			Simulation Value (*) (in Days)
				P (in Days)	ML (in Days)	O (in Days)		α = 80	α = 90	α = 95	α = 80	α = 90	α = 95
A	50.00		R10	720.00	700.00	680.00	700.00	706.94	710.13	712.44	706.94	710.13	712.43
B	290.00		R6	45.00	30.00	30.00	32.50	34.13	35.53	36.76	34.13	35.53	36.76
C	25.00		R20	45.00	30.00	30.00	32.50	34.13	35.53	36.76	34.13	35.53	36.76
D	122.00		R22	40.00	35.00	30.00	35.00	36.73	37.53	38.11	36.73	37.53	38.11
E	29.00		R6	45.00	35.00	30.00	35.83	37.78	38.31	39.67	37.78	38.31	39.67
F	114.00		R22	45.00	35.00	30.00	35.83	38.31	39.67	40.70	38.31	39.67	40.70
G	44.00		R6	45.00	35.00	30.00	35.83	37.78	38.31	39.67	37.78	38.31	39.67
H	196.00		R5	50.00	40.00	30.00	40.00	43.47	45.07	46.21	43.47	45.07	46.21
I	29.00		R19	45.00	35.00	25.00	35.00	38.47	40.07	41.21	38.47	40.07	41.21
Ʃ (day)	899.00						982.50	1007.73	1020.16	1031.52	1007.73	1020.16	1031.52
Ʃ (Month)	29.97						32.75	33.59	34.01	34.38	33.59	34.01	34.38

Source: author’s research. (*) The Monte Carlo Simulation is performed using @Risk Software with 10000 iterations.

.

Following the implementation of various distribution models based on data collected in stage 2, modeling was performed according to the number of iterations and confidence levels utilized in this study. Subsequently, the scheduling model was constructed using the Gantt chart approach in the MS Project scheduling case study. Multiple schedule estimates were conducted, using the number of additional days attributable to inherent risk (pre-risk mitigation), alongside the project’s initial length and simulation findings, to derive an estimate for the completion time of the Jakarta Central Railway Station Phase II, as presented in

Table 4. Comparison of duration and project completion time based on the Railway Station Phase II Project Implementation Schedule Model under inherent risk conditions (pre-risk mitigation).

No.	Schedule Model	Duration (in Days) 1	Completion Time (Pre-Risk Mitigation)	Comparison with Initial Duration (%)
1	Initial contract	674.00 days	30 September 2021	-
2	PERT value	1555.66 days	16 April 2024	130.81%
3	Theoretical value (confidence level 80%)	1575.70 days	6 May 2024	133.78%
4	Theoretical value (confidence level 90%)	1584.78 days	15 May 2024	135.13%
5	Theoretical value (confidence level 95%)	1592.33 days	23 May 2024	136.25%
6	Simulation value (confidence level 80%)	1575.70 days	6 May 2024	133.78%
7	Simulation value (confidence level 90%)	1584.78 days	15 May 2024	135.13%
8	Simulation value (confidence level 95%)	1592.33 days	23 May 2024	136.25%

Source: Author’s research. ¹ The total duration is calculated using MS Project by entering the number of related activities identified in Table 2.

.

Table 4 shows the duration and project completion time comparison of the Station Phase II Project Implementation Schedule Model under inherent risk conditions (pre-risk mitigation). First of all, it indicates that the initial completion length is 674.00 days. Under PERT conditions, the value includes an additional period of 1555.66 days or 130.815% of the original schedule duration.

Under theoretical value conditions (confidence level 80%), there is an additional length of 1575.70 days or 133.78% of the initial timeframe. Under theoretical value conditions (confidence level 90%), there is an additional length of 1584.78 days or 135.13% of the starting period. Under theoretical value conditions (confidence level 95%), there is an additional length of 1592.33 days, equating to 136,25% of the initial period. In the simulation value condition (confidence level 80%), there is an additional length of 1575.70 days, representing 136.25% of the initial timeframe.

In the simulation value condition (confidence level 90%), there is an additional period of 1584.78 days or 135.13% of the initial timeframe. In the simulation value condition (confidence level 95%), there is an additional duration of 1592.33 days or 136.25% of the initial timeframe. The findings of the additional time from all scheduling models under the inherent risk condition (pre-risk mitigation) range from 130.81% to 136.25%.

4.3. Data Collection for, and Analysis and Discussion of Research Question 3 (RQ 3)

In the third step of data collection, the preliminary phase is informed by the findings of the RQ 2 study conducted by [7] and the author’s research, specifically identifying the most prevalent risk in every critical activity. The literature study, archival data, and historical data in case studies are employed to conduct a cause-and-effect analysis of the dominant risk in each activity. A focus-group discussion (FGD) was performed to ascertain the causal relationships associated with each dominant risk in the activities examined in the case study. FGD was undertaken to ascertain the suitable risk response approach by identifying preventative and corrective actions for each dominant risk associated with critical activities. The same approach was used, where confidence levels (α) of 80%, 90%, and 95%, and 10000 iterations were selected as a basis for the simulation. The comprehensive simulation results are presented in detail in

Table 5. Quantitative analysis of Monte Carlo Simulation using @Risk software with PERT distribution approach on residual risk conditions (post-risk mitigation).

No. Activity	Duration (Days)	Risk ID	Additional Days (Residual Risk)			PERT (Day)	Theoretical Value			Simulation Value (*)
			P (in Days)	ML (in Days)	O (in Days)		α = 80	α = 90	α = 95	α = 80	α = 90	α = 95
A	50.00	R10	550.00	500.00	470.00	503.33	516.80	523.85	529.15	516.80	523.85	529.15
B	290.00	R6	30.00	25.00	20.00	25.00	26.73	27.53	28.11	26.73	27.53	28.11
C	25.00	R20	35.00	25.00	20.00	25.83	28.31	29.67	30.70	28.31	29.67	30.70
D	122.00	R22	30.00	25.00	25.00	25.83	26.38	26.84	27.25	26.38	26.84	27.25
E	29.00	R6	35.00	25.00	20.00	25.83	28.31	29.67	30.70	28.31	29.67	30.70
F	114.00	R22	40.00	35.00	30.00	35.00	36.73	37.53	38.11	36.73	37.53	38.11
G	44.00	R6	35.00	25.00	25.00	26.67	27.75	28.70	29.51	27.75	28.70	29.51
H	196.00	R5	40.00	30.00	25.00	30.83	33.31	34.67	35.70	33.31	34.67	35.70
I	29.00	R19	30.00	25.00	20.00	25.00	26.73	27.53	28.11	26.73	27.53	28.11
Ʃ (day)	899.00				723.33	751.07	765.99	777.33	751.07	765.99	777.33
Ʃ (Month)	29.97				24.11	25.03	25.53	25.91	25.03	25.53	25.91

Source: Author’s research. (*) The Monte Carlo Simulation is performed using @Risk Software with 10000 iterations.

.

The simulation results were collected and put into MS Project, and scheduling calculations were performed. Again, multiple schedule estimates were conducted, using the number of additional days attributable to residual risk (post-risk mitigation), alongside the project’s initial length and simulation findings, to derive an estimate for the completion time of the Railway Station Phase II, as presented in

Table 6. Comparison of duration and project completion time based on the Railway Station Phase II Project Implementation Schedule Model under residual risk conditions (post-risk mitigation).

No.	Schedule Model	Duration (in Days) 1	Completion Time (Post-Risk Mitigation)	Comparison with Initial Duration (%)
1	Initial contract	674.00 days	30 September 2021	-
2	PERT value	1319.82 days	24 August 2023	95.82%
3	Theoretical value (confidence level 80%)	1342.26 days	16 September 2023	99.15%
4	Theoretical value (confidence level 90%)	1345.10 days	28 September 2023	99.57%
5	Theoretical value (confidence level 95%)	1363.02 days	7 October 2023	102.23%
6	Simulation value (confidence level 80%)	1342.26 days	16 September 2023	99.15%
7	Simulation value (confidence level 90%)	1354.10 days	28 September 2023	100.91%
8	Simulation value (confidence level 95%)	1363.02 days	7 October 2023	102.23%

Source: Author’s research. ¹ The total duration is calculated using MS Project by entering the number of related activities identified in Table 4.

.

The inherent risks are further discussed and analyzed with respective respondents (experts). Using cause–effect analysis and further preventive, as well as corrective, measures, the following risk response strategies are proposed as shown in Table 6. Referring to Table 6, showing a duration and project completion time comparison based on the Station Phase II Project Implementation Schedule Model under residual risk conditions (post-risk mitigation), the initial completion length is 674.00 days.

Using PERT conditions, the value includes an additional period of 1319.82 days or 95.82% of the original timeframe. Under theoretical value conditions (confidence level 80%), there is an additional length of 1342.26 days or 99.15% of the initial timeframe. Under theoretical value conditions (confidence level 90%), there is an additional length of 1345.10 days or 99.57% of the starting period. Under theoretical value conditions (confidence level 95%), there is an additional length of 1363.02 days, equating to 102.23% of the initial period. In the simulation value condition (confidence level 80%), there is an additional length of 1342.26 days, representing 99.15% of the initial timeframe. In the simulation value condition (confidence level 90%), there is an additional period of 1345.10 days or 135.13% of the initial timeframe. In the simulation value condition (confidence level 95%), there is an additional duration of 1592.33 days or 99.57% of the initial timeframe. The findings of additional time from all scheduling models under the residual risk condition (post-risk mitigation) range from 95.82% to 102.23%.

The risk responses are suggested and validated by the experts in the form of preventive and corrective actions.

Table 7. Cause–effect analysis, preventive–corrective actions, risk response strategies (Source: current research work).

Act.	Duration (Days)	Risk ID	Risk Causes	Code	Additional Days (Inherent Risk)			Code	Preventive Actions	Code	Risk Response (Inherent)
					P (in Days)	ML (in Days)	O (in Days)
A	50.00	R10	Land acquisition	C1	720.00	700.00	680.00	IR1	Creating a land acquisition mitigation plan	PA1	Mitigate
			Interface, dependencies between packages	C2					Create integration of work stages together	PA2	Transfer
			Many design changes	C3					Check all DED drawings before starting work.	PA3	Mitigate
B	290.00	R6	Poor communication between parties	C4	45.00	30.00	30.00	IR2	Creating a communication management plan	PA4	Mitigate
			Complicated bureaucracy	C5					Prepare a licensing flowchart	PA5	Accept
			Licensing is complicated by various parties	C6					Developing a licensing plan strategy	PA6	Mitigate
C	25.00	R20	Waiting for other work packages to complete	C7	45.00	30.00	30.00	IR3	Make a joint planning schedule	PA7	Transfer
D	122.00	R22	Limited job access	C8	40.00	35.00	30.00	IR4	Create the right working method	PA8	Mitigate
			High-risk project location	C9					Creating an HSE work plan	PA9	Mitigate
			Work during the window time	C10					Socialization to passengers	PA10	Accept
E	29.00	R6	Inadequate communication among parties	C11	45.00	35.00	30.00	IR5	Creating a communication management plan	PA11	Mitigate
			Complicated bureaucracy	C12					Prepare a licensing flowchart	PA12	Accept
			Licensing is complicated by various parties	C13					Developing a licensing plan strategy	PA13	Mitigate
F	114.00	R22	Limited job access	C14	45.00	35.00	30.00	IR6	Create the right working method	PA14	Mitigate
			High-risk project location	C15					Creating an HSE work plan	PA15	Mitigate
			Work during the window time	C16					Socialization to passengers	PA16	Accept
G	44.00	R6	Inadequate communication among parties	C17	45.00	35.00	30.00	IR7	Creating a communication management plan	PA17	Mitigate
			Complicated bureaucracy	C18					Prepare a licensing flowchart	PA18	Accept
			Licensing is complicated by various parties	C19					Developing a licensing plan strategy	PA19	Mitigate
H	196.00	R5	Poor communication between parties	C20	50.00	40.00	30.00	IR8	Creating a communication management plan	PA20	Mitigate
			Complicated bureaucracy	C21					Creating a communication flowchart	PA21	Accept
			Project owner’s inconsistency in schedule priority order	C22					Develop a plan for the method and appropriate stages of work	PA22	Mitigate
I	29.00	R19	Complicated bureaucracy	C23	45.00	35.00	25.00	IR9	Creating a communication flowchart	PA23	Accept
			Incomplete administrative files	C24					Create a document monitoring plan	PA24	Mitigate
			Poor communication between parties	C25					Creating a communication management plan	PA25	Mitigate
Act.	Duration (Days)	Risk ID	Risk Effect	Code	Additional Days (Residual Risk)			Code	Corrective Action	Code	Risk Response (Residual)
					P (in Days)	ML (in Days)	O (in Days)
A	50.00	R10	Delay in work time	E1	550.00	500.00	470.00	RR1	Coordinate with parties to accelerate land acquisition	CA1	Transfer
			Changes to the implementation drawing	E2					Conduct joint meetings between packages to integrate drawings	CA2	Transfer
			Changes in the scope of work	E3					Conduct design reviews with experts	CA3	Mitigate
B	290.00	R6	Work is hampered	E4	30.00	25.00	20.00	RR2	communicate well with all parties	CA4	Accept
			Delay in work time	E5					Write officially regarding any problems	CA5	Transfer
			Administrative approval is delayed	E6					Complete the necessary administration	CA6	Mitigate
C	25.00	R20	Delay in completion time	E7	35.00	25.00	20.00	RR3	Hold regular meetings together	CA7	Transfer
D	122.00	R22	The safety of passengers, workers, and train operations is disrupted	E8	30.00	25.00	25.00	RR4	Create a mutually agreed standard operating procedure (SOP)	CA8	Mitigate
			Delay in work time	E9					Conducting socialization of labor hazards	CA9	Transfer
			Increase in overtime costs for manpower	E10					Make monitoring of work activities	CA10	Mitigate
E	29.00	R6	Work is hampered	E11	35.00	25.00	20.00	RR5	Communicate well with all parties	CA11	Accept
			Delay in work time	E12					Write officially regarding any problems	CA12	Transfer
			Administrative approval is delayed	E13					Complete the necessary administration	CA13	Mitigate
F	114.00	R22	The safety of passengers, workers, and train operations is disrupted	E14	40.00	35.00	30.00	RR6	Create a mutually agreed standard operating procedure (SOP)	CA14	Mitigate
			Delay in work time	E15					Conducting socialization of labor hazards	CA15	Transfer
			Increase in overtime costs for manpower	E16					Make monitoring of work activities	CA16	Mitigate
G	44.00	R6	Work is hampered	E17	35.00	25.00	25.00	RR7	Communicate well with all parties	CA17	Accept
			Delay in work time	E18					Write officially regarding any problems	CA18	Transfer
			Administrative approval is delayed	E19					Complete the necessary administration	CA19	Mitigate
H	196.00	R5	Delay in issuing work permits	E20	40.00	30.00	25.00	RR8	Write officially regarding any problems	CA20	Transfer
			Lots of work reworked	E21					Monitor any changes in scope	CA21	Mitigate
			Cost increase	E22					Submitting a claim for a fee increase	CA22	Transfer
I	29.00	R19	Delay in handover of work	E23	30.00	25.00	20.00	RR9	Conduct a joint planning meeting	CA23	Transfer
			Delay in administrative approval	E24					Complete the necessary administration	CA24	Accept
			Delay in issuing safety assessment permits	E25					Immediately follow up on the safety assessment test findings	CA25	Accept

Source: Current research work.

shows how the respective risks are responded using risk response strategies as defined by the [17].

5. Conclusions

After performing data analysis, this research has revealed the answers to the proposed research questions. The construction project risks were validated, and the results showed that 22 out of 51 risk indicators related to the Jakarta Central Railway Station project phase II were selected as major inherent (pre-mitigated) risks, with 29 remaining rejected items [7]. The next finding is that there were nine critical activities (identified as activities A to I in this study) in the project that are significant to the project schedule, where they are correlated with six most dominant risks in each critical activity from the qualitative analysis of the FGD. The results of the simulation and calculated values on inherent risk showed additional days due to residual risk ranging from 130.81% to 136.25% of the initial contract completion duration. The risks were further analyzed, discussed, and validated with designated risk response strategies using preventive actions. After the implementation of preventive and corrective measures, and risk response strategies via focus-group discussions, the scheduling simulation indicated that residual risk under post-risk mitigation conditions resulted in an extension of approximately 95.82% to 102.23% of the original contract completion duration. The R10 is the biggest contribution of the extended duration to the project. This is because of the complexity of the operational activities (life train operations) that have to be taken into account during the project’s construction. Priority to the train operations is given as priority to construction activities. The impact was calculated as being bigger, as sometimes there were clashes of train traffic peak with the construction activities that were within the critical path.

This research shows that by using proper project risks management process and Monte Carlo, Analysis using a rich-featured software, as well project management software, a reduction in project durations can be calculated. By knowing this, project organizations (hereinafter, the project managers and project sponsors) can adjust their expectations and make further strategic decisions that might lead to major changes in the projects. This is much better than a normative and subjective decision that might further cause a much bigger impact on the project cost as in implication to the deviation of the schedules.

This study has provided further enhancement to the project risk management studies, especially in railway construction. The results prove the utility and prospect theories that the risk impact to the schedule can be quantified and somehow can be accepted due to the respective risk attitudes and their thresholds. The research result also has enriched the development of the current risk studies, especially where the risks of a complex railway project can be managed using proper processes, tools, and inputs to the project risks, especially in quantifying the deviation of schedule as a form of risk impact. Furthermore, this study has also provided insight that better risk response strategies can be proposed if the risk impacts are quantified.

From a practical point of view, this research has provided insightful results, where identified risks, processes, and tools will be treated as a significant information of lessons learned repository for future projects. The information will be used in estimating a more realistic project schedule. The identified and analyzed risks are also considered as significant inputs not only in schedule, but also to other project areas, such as project cost, project contract, resources estimates, and any other related aspects of the railway construction projects. The most important aspect that is taken into consideration by the construction project practitioners, especially of railways, is that project risk quantification requires powerful project management tools (risk management and project management software). The findings could also be considered a construction policy for the railway regulator and/or can be applied to the contract clauses; meanwhile, the result can also be considered as project management governance for the civil contractors.

The authors are aware that there are limitations to this study. Despite conducting a deep and comprehensive analysis of the railway construction using Monte Carlo analysis, this study refers only to the railway station construction project in Indonesia. Furthermore, the limited number of five panel reviewers could also potentially lead to biases in the justification and assessment of the risks. The uniqueness of the project could limit the generalization of the result and its applicability to other construction projects. Some aspects, such as project governance, organization capabilities (that might relate to project management maturity), and social environment, could also deviate the result of this study from its applicability.

This study also employed several data-gathering strategies in qualitative risk analysis to ascertain the impact of these methods on qualitative risk outcomes. In addition to PERT, various other distribution models can be employed to discern the distinctions among other distribution models not utilized in this study.

Furthermore, this research also employs various ways to collect data on risk responses, enabling the assessment of their impact on the development of a strategy model. Further study can be conducted by analyzing the time–cost trade-off. This study also proposed additional research advancement, where the risk indicators for the railway construction project could be expanded to include a wider range of factors and use different case-study projects.