The Open Collections website will be unavailable July 27 from 2100-2200 PST ahead of planned usability and performance enhancements on July 28. More information here.

International Construction Specialty Conference of the Canadian Society for Civil Engineering (ICSC) (5th : 2015)

A time-cost-quality trade-off model for nuclear-type projects Shahtaheri, Maryam; Nasir, Hassan; Haas, Carl T.; Salimi, Tabassom 2015-06

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


52660-Shahtaheri_M_et_al_ICSC15_244_A_Time_Cost.pdf [ 1.18MB ]
JSON: 52660-1.0076417.json
JSON-LD: 52660-1.0076417-ld.json
RDF/XML (Pretty): 52660-1.0076417-rdf.xml
RDF/JSON: 52660-1.0076417-rdf.json
Turtle: 52660-1.0076417-turtle.txt
N-Triples: 52660-1.0076417-rdf-ntriples.txt
Original Record: 52660-1.0076417-source.json
Full Text

Full Text

5th International/11th Construction Specialty Conference 5e International/11e Conférence spécialisée sur la construction    Vancouver, British Columbia June 8 to June 10, 2015 / 8 juin au 10 juin 2015   A TIME-COST-QUALITY TRADE-OFF MODEL FOR NUCLEAR-TYPE PROJECTS Maryam Shahtaheri1,4, Hassan Nasir2, Carl T. Haas1 and Tabassom Salimi3 1 Department of Civil and Environmental Engineering, University of Waterloo, Canada 2 Department of Civil Engineering, King Abdulaziz University, Saudi Arabia 3 Department of Nanotechnology Engineering, University of Waterloo, Canada 4 Abstract: The purpose of this paper is to introduce a generalized multi-dimensional joint confidence level model for nuclear refurbishment planning, using the Darlington retube and feeder replacement (RFR) project (which is a multi-billion dollar effort) as a research platform. The intentions of this model are twofold: (1) determining the best set(s) of cost, schedule, and proxies for quality such as radiation expenditures by incorporating the variations of work-shift models on both the activity duration distributions and the logic of the work flow, (2) improve the expected reliability and predictability involved in the resource allocation system (schedule), by “efficiently” integrating the influential project factors, constraints, and labour shift models associated with the schedule. An effective Monte Carlo based time-cost-quality trade-off model is contributed for examining the performance and risk impacts of various work-shift designs and thus supports the choice of an optimal work-shift design. 1 INTRODUCTION 1.1 Background  The four CANDU nuclear reactors at the Darlington Nuclear Generating Station supply about 20 % of Ontario’s power needs.  These reactors require a multi-billion dollar Retube and Feeder Replacement (RFR) Project that will begin in 2016 and will continue for approximately 12 years. Research has clearly established that the greatest driver of success in such megaprojects is project definition, which includes engineering, project management, quality, schedule, target costs, and safety plans. This process should optimize productivity and manage resource allocation effectively so that the RFR project is successful in terms of cost, schedule, safety, quality, predictability, and participant satisfaction. In this project, a unique full-scale mock-up of the reactor’s fuel channels and feeders has been constructed and used for testing the functionality of tools, for training personnel, and for optimizing processes to achieve the stated objectives. With 960 calandria tubes, fuel channels, and inlet and outlet feeders, the RFR processes will be repeated cyclically in a manner that challenges conventional concepts of construction project scheduling and resource allocation. Unique constraints abound: radiation dosage limits, number of people allowed in the vault, and schedule milestones are among them. The varying impact on labour productivity of round-the-clock work-shift designs is only one of many additional challenges involved. The main question that will be addressed in this paper is the optimal allocation and scheduling of resources in such a situation. The primary objective is thus to develop a tool that will support determination of the best strategies based on cost, schedule, and proxies for quality for allocating 244-1 resources for nuclear retube and feeder replacement projects by: (1) identifying and including relevant constraints, such as the number of people allowed in the vault and the labour turnover that results from reaching radiation limits; (2) incorporating parameters such as estimated craft productivity and variations in process time; (3) addressing conflicting project objectives; (4) examining “what-if” shift-work designs and discovering possible systematic improvements; (5) estimating the impact of changes; and (6) providing an understandable and practical approach. The full-scale mock-up and training facility constructed in May 2014 helps enable the validation of this tool.  The methodology for this study will build on advances made by colleagues in the construction and industrial engineering and management fields in the area of hybrid models for construction trade-off problems (AbouRizk et al., 2011, El-Rayes & Moselhi 2001, Froese 2010, and Nasir et al., 2003).  For example, commercial software exists for incorporating stochastic modelling into critical path method (CPM) schedules for Monte Carlo type simulations and analysis, however the cyclical nature of much of the RFR work lends itself naturally to the logical branching capabilities of discrete event simulation systems.  Finding the most efficacious combination of these approaches considering the very unique objectives will be the focus of the next few sections of this paper. 1.2 Planning, Scheduling, and Estimating Complex Refurbishment Projects Managing a well-structured schedule involves satisfying project objectives, optimizing resource allocation, and mitigating uncertainties. This is in general the problem encountered in this study.  Some related insights and challenges are discussed below. 1.2.1 Association for the Advancement of Cost Engineering (AACE) Scheduling and Estimating Classification System As a recommended practice of AACE International, the Cost Estimate Classification System provides guidelines for applying the general principles of estimate classification to project cost estimates. As noted by the Recommended Practice No. 18-R (2005), the Cost Estimate Classification System maps the phases and stages of project cost estimating together with a generic maturity and quality matrix, which can be applied across a wide variety of industries. The Cost Classification System consists of five estimate classes which are defined based on the level of project definition (known as the primary characteristic). Secondary characteristics on the other hand include: typical estimate purpose, typical estimating method, and typical accuracy range are correlated to the level of project definition. Class 5 represents the lowest level of project definition (with a low range of -20% to -50% and a high range of +30% to +100%) and the Class 1 estimate (with a low range of -3% to -10% and a high range of +3% to +15%) represents the closest to complete project definition (100%). These Class estimates will be used in this study as mentioned below. 1.2.2 Project Operations and Modelling Structures Modelling project operations is one way to increase predictability and improve visualization prior to project execution (Russell et al., 2009, Mohamed et al., 2007, Zayed &Halpin 2004). Examples of modelling tools in construction are serious gaming environments, 3D and 4D visualization techniques, and discrete event simulation (DES) tools such as CYCLONE, COSYE, and STROBOSCOPE (AbouRizk et al., 2011). DES tools provide intuitive environments and functional elements that can accurately model and simulate construction operations (Puri & Martinez, 2012). However, validation of DES models is challenging. Challenges include: insufficient data causing high variance of output, project managers required to be knowledgeable about simulation tools and their workability, and the assumption that operations and time-slots involved with activities are independent and discrete, meaning that incorporating stochastic modelling for continuous sets of data results in further errors (Puri& Martinez, 2012; Rekapalli & Martinez, 2011). Scheduling and project operations modelling tools are often used for optimization of conflicting project objectives such as time and cost (Al-Hussein et al., 2005, El-Rayes & Moselhi 2001).  Approaches used tend to be heuristic or based on advanced algorithms such as genetic algorithms.  Linear and integer programing are almost never used for these problems because of the number of decision variables involved. 244-2 1.2.3 Uncertainty and Project Scheduling Critical Path Method (CPM) scheduling is useful for effective completion of most construction projects. Although CPM calculations have proven to be simple and straightforward, CPM-based scheduling is a challenging process, and it may fit reporting and documentation purposes more than decision support systems that reflect reality (Hegazy & Menesi, 2010). Some of these challenges include: problems with interference of constraints in a multi-constrained schedule, problems with multiple complex relationships, and inaccurate schedule calculations. While many studies have been conducted to tackle some of these problems, in 2010 Hegazy and Menesi argued that previous methods led to analysis at rough levels of detail that cause errors in calculations (Hegazy & Menesi, 2010).  In addition, most schedules are developed using a deterministic approach. Methods such as PERT are meant to address this problem by providing statistical distributions for activity durations that reflect uncertainty. Unfortunately, these methods have limitations. They assume that activities are independent, more effort is required to provide estimated values, and there is no recognition of critical path variations (Nasir et al., 2003). One common assumption of scheduling theory is a static environment which may lead to no formal justification for unexpected events. The result is deviation from the project plan (Ahmed at al., 2003), therefore, identifying and classifying uncertainty to model and later reduce it to an acceptable level is an important aspect of project planning (Song et al., 2005). Many models have been developed to classify, model, and reduce uncertainty using artificial neural networks, simulation models, heuristic approaches, and logic work flows. These methods are claimed to be more accurate compared to traditional scheduling approaches such as CPM and PERT (Shaheen et al., 2009; Song et al., 2005; AbouRizk et al., 2011) yet the gap between virtual and actual project schedule environments (i.e. simulated uncertainties versus actual uncertainties) requires more investigation.  1.2.4 Integration of Stochastic Modelling and Project Scheduling Techniques   Capturing the impact of probabilistic project input (e.g., activity durations and cost) on the project output (e.g., total cost and completion duration) can help project managers to execute a more reliable and predictable project. Many tools have been developed to map the stochastic nature of project objectives with the schedule. Among these tools, well-known ones are @ RiskTM (Palisade Corporation), Oracle and Primavera project portfolio management products such as Oracle Crystal BallTM and Oracle Primavera Risk AnalysisTM package. As @ RiskTM for MS ProjectTM is used for this study, details follow. The @RiskTM software is an add-on to MS Excel/ProjectTM and suggests ways to use probabilistic analysis and Monte Carlo Simulation to visualize and quantify the uncertainty in projects and have more accurate cost and schedule predictions. This tool is also used for cost contingency. The approach of this planning method starts by creating a three point estimate (minimum, most-likely, and maximum) or defining a distribution based on historical data for both duration and cost of every activity within the schedule. After subsequent steps including Monte Carlo simulation output in the form of histograms of simulated results, a cumulative probability density graph illustrating schedule variation and exposure analysis graphs must be interpreted. However, such tools are primarily applicable to high level schedule estimation and do not provide complete and descriptive accurate results for intense (e.g. cyclic operations) and complex (e.g. multi-constraint) projects.  1.2.5 Radiation Limits and Impact on Labour The estimation of radiation expenditures is one of the most important elements and the key constraint in this study, for the following reasons: (1) it has a direct impact on the health of the labourers while they are working at the reactor face, (2) reductions in the radiation expenditures can be achieved by altering the type of clothing (comfo, plastics, etc.) and through shielding in specific work areas which affects productivity; however, it cannot be eliminated, (3) the Canadian Nuclear Safety Commission (CNSC) Radiation Protection Regulations require the implementation of a managed system at nuclear sites in order to keep the amount of radiation absorbed by labourers (and members of the public) from radiation exposure as low as is reasonably achievable (ALARA). Radiation is measured in Roentgen Equivalent Man (REM) which is known as the dosage that will cause the same amount of biological injury as one rad of X rays or gamma rays. This requirement translates into specific radiation limits that are currently determined for each labourer for one-year (1 Rem-person) and five-year (5 Rem-person) time windows. 244-3 Once labourers reach either the one- or the five-year limit, they are assigned to non-radiated work areas, (4) while the initial cost and schedule estimations prior to project execution are set to zero, in this case, labourers do not walk in with a zero radiation dose but are expected to start the job with a “pre-existing” radiation rate that can be established and modelled as frequency distribution for a population of labourers, (5) the dose rate absorbed per labourer is reset on January 1st of each year. In this case, “reset” is defined as a radiation rate estimated for each labourer based on the average dose rate over a five-year time window, and (6) different activities are associated with different radiation rates because of varying distances from radiation sources.   1.2.6 Factors Associated with Labour Productivity To obtain a good level of accuracy with respect to cost and schedule estimated, two items should be considered: (1) identification of factors that drive labour productivity, such as: shift length, temperature, and congestion, and (2) definition of the baseline. The baseline is defined as the neutral work condition, meaning that no additional effort is required from labourers beyond that necessary for them to complete their work during a standard shift (i.e., 5-8s). For this specific project, if a multiplier of one is considered for the neutral work condition (i.e., based on assessed hours), multipliers (i.e., weights) for work conditions other than neutral are assessed.  For the proposed JCL model, factors are identified based on the Productivity handbook produced by CII (2014), which synthesizes over 50 years of quantitative research (CII 252-2d, 2014) and are assessed based on documents compiled by the Mechanical Contractors Association of America, Inc. (MCAA). 1.3 Joint Confidence Limit (JCL) Model A JCL is an integrated uncertainty analysis for cost and schedule estimation, first introduced by NASA HQ-Program Analysis and Evaluation Cost Analysis Division (2009). This model combines the cost, schedule, and uncertainty (using Monte Carlo analysis) associated with a project in order to identify the relative probability that the cost and schedule “jointly” fall within the targeted budget and schedule dates. This process helps update management with respect to the likelihood of the programmatic success of a project. A JCL can be constructed from either of two types of input: parametric models or probabilistic resource-loaded schedule (PRLS) estimates. If a JCL is constructed in the early estimation phase, parametric models can be used, and then as the project estimation phase advances, PRLS can come into play. Information required as part of the input for the JCL includes the recent cost data and project schedule. A final requirement is that statistics such as the mean and standard deviation of the cost and schedule be available (NASA HQ, 2009).  The multi-dimensional JCL introduced in this paper is an extension to the described 2 dimensional JCL model with the third dimension being the quality proxy (i.e., radiation expenditures). This section will be later discussed in section 3. 1.4 Scope of Analysis As vehicles for identifying the primary characteristics and interdependency of the three objectives (i.e., cost, schedule, quality proxy) within the RFR project, a work package known as the “feeder removal series” has been studied in detail. The rationale behind choosing this work package among many others include: (1) relatively high rates of radiation exposure at the workface; feeders are highly contaminated with toxic particles and labourers will be exposed to significant rates of radiation while removing the feeders from the reactor in the vault, and (2) labour-intensive work; feeders will be removed manually, leading to both expected and unexpected performance and productivity variations. To note that the combination of these two points creates a more challenging trade-off problem. Due to the wide range of variations that may possibly be caused by the two points, as a result of various work shift patterns and their interdependency, it becomes very interesting to explore the impact of these variations on the results of the Monte Carlo analysis and the MD-JCL model. 244-4 1.4.1 Feeder Description Feeders are an integral part of the primary heat transport system (PHTS). The function of a feeder is to transport the D2O coolant from the inlet feeders to the fuel channels (FCs) and from the FCs to the outlet headers. There are 960 feeders: one inlet and one outlet feeder for each of the 480 FCs. Feeders must be replaced when significant thinning of the outlet feeders occurs. Removing the feeders is expected to have the added benefit of reducing the overall radiation rates in the vault, which will expedite the fuel channel removal and installation series.  2 MODEL OBJECTIVES This section defines six interconnected sub-objectives related to finding good sets of solutions (defined mostly in this paper by crew arrangements) with an acceptable joint confidence limit for a complex schedule with the key objectives of minimizing variance while maximizing productivity. This is done through reasonable quantification and modelling of shift models, productivity factors, project constraints, and variations in the duration and cost of activities. The proposed multi-dimensional JCL model developed for the nuclear retube and feeder replacement and other similar projects must provide the following functions: 1. Identify and include relevant time and spatially dependent constraints such as the number of people allowed at the workface (i.e., vault), scheduling milestones, certified craft professionals available, etc. 2. Incorporate estimated probability distributions for craft productivity and process times. 3. Capture the impact of the uncertainty associated with activity durations and cost. 4. Define a quality proxy by measuring and reducing the impact of labour turn-over caused by labourers reaching radiation limits defined by radiation safety practices while being exposed to radiation sources at workface.  5. Address multiple objectives such as cost, schedule, and quality by developing a new multi-dimensional joint confidence level approach. 6. Be understandable, practical, and useful, with the potential to be adapted and applied broadly. 3 MODEL OVERVIEW 3.1 Duration and Cost In order to produce a more transparent and traceable set of outputs, it is crucial to assess uncertainty which is inherited in the duration of each planned activity. Uncertainty primarily refers to the variability in duration of the schedule activities and the values of the base cost estimates, with the amount of variability dependent on the degree of ambiguity and accuracy in the schedule and cost estimate data utilized. Uncertainty is embedded in the duration/cost values and transforms deterministic values into distributions. Examples of sources of uncertainty include: cost and schedule estimating assumptions, variable productivity rates, variable material costs, and mobilization problems.   For this study, factors such as variable skill sets and levels of experience, inconsistencies between individual workers at different times, lack of knowledge of/or failure to understand the scope definition of project specifications, and inaccurate assumptions made about “unknown unknowns” are considered and incorporated in the determination of ranges during construction estimations.   Currently the RFR project is within the Class 3 estimate. Based on the Recommended Practice No. 18R-97 (2010) the expected accuracy range for the majority of Class 3 estimates is -20 % to +30 % (AACE International, 2010). This range is referenced and assumed in the “Darlington Nuclear Generating Station (DNGS) RFR Project – Project Estimate Plan.” AACEi warns that the range is heavily dependent on the nature of the uncertainties and the technological complexity involved in a project. For this study, the accuracy range suggested by AACEi was modified for the development of the duration distributions for individual activities included in the JCL model.    244-5 In order to provide unbiased uncertainty ranges for cost lines and schedule durations, the following requirements have been deemed important for the development the MD-JCL model. 1. A high-quality project schedule 2. An estimate without contingency  3.2 Quality Proxy (Radiation Expenditures) Quality is an objective that can be defined and measured through different channels (e.g., process, procurements, and results) (Takim & Akintoye, 2002). Since labourers have an enormous influence on the quality of the work being executed, it is critical to determine the specifications of factors that impact the quality of their production. In this project, being exposed to radiation plays a crucial role in terms of advanced resource training programs (assurance of overlap between current and future set of labourers), learning curves, number of certified labourers, etc. For these reasons, the linkage between labourers and their exposure to radiation is used as a proxy to represent quality.  The preliminary estimation of the radiation rate is based on task duration, scope of work, and comprehensive work package documentation and incorporates as input the resource requirements as well as the time the labourers remain in the radiated work areas. Current radiation expenditure estimates are derived from historical measurements taken when the unit fuel channel contained fuel and the systems were full of D2O. For this study, the radiation rate per hour per person is used for calculations involving the feeder removal process. A rolled-up radiation rate is used for activities prior to feeder removal, and 25 % of the yearly limit (i.e., 0.25 Rem/person) represents the pre-existing radiation expenditure rate. To obtain more accurate results, the pre-existing radiation rate for all labourers over a five-year timeslot is based on a distribution of possible pre-existing dose rates rather than on a single-point estimate (i.e., the average).  The collective radiation expenditure (pre-existing dose rate+ hourly dose uptake) determines the number of labourers that reach the radiation limit as the project progresses. This measurement enables further advance evaluation that incorporates a determination of the number of labourers/resources who require training, which can take up to 2 weeks. This calculation is important because failure to train additional resources before the current resources reach radiation limits can create delays and lead to additional costs.  4 INITIAL STUDIES For demonstration purposes, eight different cases have been considered. The variation between these cases is a result of variation in productivity rates caused by factors such as including different work-shift designs per worker, which are the following: (A) 40 hours a week, (B) 50 hours a week, and (C) 60 hours a week, and different crew arrangements. The basis of shifts A and B are on a 24/7 schedule and the base of shift C is on work continuum from Monday to Saturday, and Sunday considered as the non-working day. This is shown in the Calendar column of Table 1, with Y being a working and N being a non-working day. The number of crew available in the vault is constant for every shift, however various arrangements of labourers at the workface has been considered and represented by three options: I (i.e., 3 workers remove one feeder), II (i.e., 2 workers remove one feeder), and III (2+1/2 workers remove one feeder) for the feeder removal package specifically. As shown in Table 1, the last row is the deterministic estimation of duration, cost, and proxy for quality (radiation expenditures), based on 50 hours of work per worker and crew arrangement I. In the next step, the deterministic estimations were used as the basis for the Monte Carlo analysis and uncertainty was then incorporated by various probability density functions using @Risk. The output of the Monte Carlo analysis using both MS project and @Risk are presented in Table 1 after 1000 simulation runs for each case.  As shown in Table 1, case 1 is the most suited option to be explored in detail.     244-6 Table 1: @ Risk Results for Eight Different Sets of Ranges Figure 1 represents the input distributions for duration and cost for the first two crew arrangements and the magnified figure shows the input distributions for the first case. The minimum, maximum, mean, 5th percentile, and 95th percentile of the distributions are also shown in this chart. The advantages of including the mentioned point estimates are twofold: (1) to comprehend and compare the variations of the outputs as the result of (defined) variations of the input sets, and (2) the ability to estimate the variance between the probabilistic objective values for a certain percentile (confidence level) and the deterministic estimates.         Figure 1: Output Results (Distributions) from the Monte Carlo Analysis using @Risk and MS Project Case # Crew Arrangement Shift Duration (Days) (mean) Cost ($) (mean) Quality (Rem) (mean) Calendar Sunday (Y/N) Finish Date Case 1 I A 33 $ 3,857,778 539 Y 11/6/16 Case 2 I C 43 $ 4,629,333 711 N 11/21/16 Case 3 II A 46.5 $ 4,667,911 770 Y 11/20/16 Case 4 II B 36 $ 5,516,622 593 Y 11/9/16 Case 5 II C 51 $ 4,629,333 840 N 11/30/16 Case 6 III A 39 $ 5,516,621 652 Y 11/12/16 Case 7 III B 39 $ 4,629,333 646 Y 11/16/16 Case 8 III C 42 $ 5,015,111 700 N 11/15/16 Deterministic (Start day: 10/3/16)  I  B 30 $ 3,500,000 400  Y  11/3/16     244-7  Figure 4: MD-JCL for Cost and Duration 5 DISCUSSION OF RESULTS As discussed earlier, primarily the focus was to look at different “possible” work-shift designs, and by incorporating unique factors and constraints, identify the ones that may result in better sets of project objectives. Since running @Risk resulted in a prolonged schedule, the following interpretations are made: (1) a shift schedule without buffers is unachievable, therefore the expected drift (variations from @risk results) and adherence to schedule can be shown by building in the buffers, and (2) since the drift results in the change of work stage (estimated vs. actual) at every shift (i.e. start and completion), therefore the relationship between the percentage of drift and productivity loss can be identified. However, none of the items above can relate to finding suited shift patterns among many, as it is only the “constant” flow of the labourers to the workface and a generator of a labour productivity number. Looking at alternative cases for labour balance can be one of many solutions to this issue.  Although this model has proven to be advantageous, it entails some limitations which are as follows: (1) validation of models that deal with prediction of (problem) objectives is always challenging and without the availability of proper historical data often impossible or misleading, (2) excessive time is required at this point in time to manually enter all inputs for any realistic project schedule (which can consist of several thousand activities), and finally (3) the quality of the results heavily depends on the quality of the input data.  6 CONCLUSIONS  Multi-objective trade-off problems are one of many challenging areas in the construction industry with many proposed solutions. Accurate Incorporation of various factors and constraints as well as considering the uncertainty around project objectives leads to a more challenging problem to tackle. The variation of estimated and actual project outcomes leads to enormous cost overruns and excessive delays in the planning phase of mega-projects. A good initiative toward solving such trade-off problems is to accurately identify and measure uncertainty around cost, schedule, and other project objectives based on expert judgment and historical data. The functionality and contribution of the proposed multi-dimensional joint confidence limit model is to serve as a simulation platform to produce a flexible Pareto-optimal solution set for possible work shift models by incorporating various related factors such as the impact of work-shift designs on labour productivity and constraints such as radiation limits within a non-deterministic framework. Duration (days) 244-9 References  AACE International. (2010). AACE International recommended practice No. 18R-97 Cost terminology.  AbouRizk, S., Halpin, D., Mohamed, Y., & Hermann, U. (2011). Research in Modeling and Simulation for Improving Construction Engineering Operations. Journal of Construction Engineering and Management, 843-852. AbouRizk, S., Halpin, D., Mohamed, Y., & Hermann, U. (2011). Research in Modeling and Simulation for Improving Construction Engineering Operations. Journal of Construction Engineering and Management, 843-852. Al-Hussein, M., Alkaas, S., & Moselhi, O. (2005). Optimization algorithm for selection and on site location of mobile cranes. Journal of Construction Engineering and Management , 131(5):779-590. El-Rayes, K., & Moselhi, O. (2001). Optimizing Resource Utilization for Repetitive Construction Projects. Journal of Construction Engineering and Management, 127, 18-27. El-Rayes, K., & Moselhi, O. (2001). Optimizing Resource Utilization for Repetitive Construction Projects. Journal of Construction Engineering and Management, 127, 18-27. Froese, T. (2010). The impact of emerging information technology on project management for construction. Automation in Construction, 19 (5), 531-538. Hegazy, T., & Menesi, W. (2010). Critical Path Segments Scheduling Technique. Journal of Construction Engineering and Management, 1078-1085. Mohamed, Y., Borrego, D., Francisco, L., Al-Hussein, M., Abourizk, S., & Hermann, U. (2007). Simulation-based scheduling od module assembly yards. Engineering, Construction and Architectural Management, 14(3), 293-311. NASA HQ – Office of Program Analysis and Evaluation (PA&E) Cost Analysis Division. (2009). Joint Confidence Level (JCL) FAQ. Nasir, D., McCabe, B., & Hartono, L. (2003). Evaluating Risk in Construction-Schedule Model. Journal of Construction Engineering and Management, 518-527. Puri, V., & Martinez, J. C. (2012). Modeling of Simultaneously Continuous and Stochastic Construction Activities for Simulation. Journal of Construction Engineering and Management. Rekapalli, P.V., & Martinez, J. (2011). Discrete-event simulation based virtual reality environments for construction operations: a technology introduction. Journal of Construction Engineering and Management, 137(3): 214-224. Research Group: 252-2d, C. (2013). The Construction Productivity Handbook. Austin, Texas: Construction Industry Institue. Russell, A., Staub-French, S., Tran, N., & Wong, W. (2009). Visualizing high-rise buidling construction strategies using linear scheduling and 4D CAD. . Automation in Construction, 18(2), 219-236. Shaheen, A. A., & Fayek, A. R. (2009). Methodology for integrating fuzzy ezpert systems and discrete event simulation in construction engineering. Canadian Journal of Civil Engineering, 1478-1490 Song, L., Al-Battaineh, H. T., & AbouRizk, S. M. (2005). Modeling uncertainty with an integrated simulation system. Canadian Journal of Civil Engineering, 533-542. Takim, R., & Akintoye, A. (2002). Performance Indicators for Successful Construction Project Performance. 18th Annual ARCOM Conference, 2, 545-555. Zayed, T., & Halpin, D. (2004). Simulation as a tool for pile productivity assessment . Journal of Construction Engineering and Management , 103(3), 394-404.   244-10 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items