The aim of this research is 1 to identify the critical claims affecting the time performance of Vietnam construction projects, 2 to model the cause-effect relationships among these criti
INTRODUCTION
Problem Statement
It is undeniable that the global economy is booming compared to the last ten years, except for 2020 It is obvious that GDP is highest in industrialized countries Nonetheless, Vietnam is also one of the countries that spurred rapid economic growth with the 2,7 times GDP per capita increase from 2002 to 2018 Despite the severity of the COVID-19 pandemic in 2020, the Vietnam economy growth is unstoppable It is predicted that Vietnam economy will not face a recession, instead, estimated that GDP growth rate in the first half of 2020 will increase 1.8 percent, and expected the increase up to 2.8 percent for the whole year of 2020 It is also forecasted that the Vietnamese economy will rebound in 2021 [1]
&RQVWUXFWLRQ LQGXVWU\ LV RQH RI WKH VLJQLILFDQW VHFWRUV FRQWULEXWLQJ WKH 9LHWQDPảV economic growth Vietnam Construction Industry (VCI) has been rising up steadily with the annual growth rate of 15% for the last ten years In 2002, 39% of the Vietnam
*'3 :LWK 9LHWQDP JRYHUQPHQWảV promotion of industrialization and instilling of foreign investment capital through the Official Development Assistance (ODA) program, VCI has been thriving in a steady and flourishing state in recent years [2]± [4]
Despite the blooming development and growth of the VCI, the impact of claim and disputes on VCI are also escalating year after year In 2019 report of the leading arbitration center in Vietnam, the Vietnam International Arbitration Centre (VIAC), VIAC resolves 274 disputes in 2019 with total values of around 6.7 thousand billion VND (290 million dollars), which is an increase of 52.2 % compared to previous year Among these disputes, the construction industry accounted for 12%, as shown in Figure 1.1 and Figure 1.2 [5]
Figure 1.1 Disputes settled at VIAC (1993-2019) (Source: https://www.viac.vn/en/statistics/2019-statistics-s30.html)
Figure 1.2 Types of disputes settled at VIAC ± 2019 (Source: https://www.viac.vn/en/statistics/2019-statistics-s30.html)
In Vietnam most Vietnam construction projects cannot deliver the project in time and have been facing schedule delays frequently [6] One of the primary causes of schedule delays is due to construction claims However, it is been acknowledged that claims are unavoidable in construction projects cannot be solved at all easily [7] Managing the claims in construction projects is always challenging, as it calls for understanding the contractual terms, maintaining the transparent communication among different parties, and comprehending the causes of claims Nonetheless, the difficulty of managing the claims procedure could be reduced significantly if the reasons as to why the claims occurred could be found out and assessment of claims can be speculated precisely
To execute the projects successfully, schedules are indispensable Without a project schedule, the difficulty in coordinating and managing the various construction project activities dramatically increased [8] Because of the wild nature of construction projects, it is unavoidable to encounter unexpected problems, (e.g claims, risks, etc.)
As a consequence, schedule delays are quite regular in many construction projects It is been acknowledged that construction project schedule is indispensable in managing the project due to its impact on project success Delays in construction projects repeatedly can results in project stakeholders suffering severe financial losses As the VD\LQJ JRHV ààD SUREOHP ZHOO LGHQWLILHG LV D SUREOHP KDOI-VROYHG´ [9] Therefore, predicting potential schedule delays effectively can elevate the prospects of success of construction projects Hence, it is vital to quantify the likelihoods of construction schedule delays.
Research Objectives
The followings are the objectives of this study:
1 To identify the critical claims causing schedule delays in Vietnam construction projeccts
2 To model the cause-effect relationships among these critical claims
3 To quantify and assess the severity in schedule delays based on the occurrence of those critical claims
4 To validate the effectiveness of the BBNs model and provide a frame work in claim assessment of Vietnam construction projects.
Scope of Study
This study concentrates on identifying the critical claims causing schedule delays in Vietnam construction projects and quantifying and assessing their impact on the project schedule The methodologies involved are relative importance index (RII), mean value and Bayesian Belief Networks (BBNs), of which the former two methods are utilized for qualitative study and the latter for quantitative study of construction claims Three consecutive steps of data collection & analyses were applied in this study
Firstly, the lists of construction claims affecting the time performance of construction projects were obtained through literature review and expert discussion questionnaire type 1 was designed and surveyed based on those lists of claims, to identify the critical claims causing schedule delays in Vietnam construction projects
In the second step, using the list of critical claims from the first step, questionnaire survey (type 2) and expert interviews were conducted to determine the cause-effect relationship among these critical claims
In the final step, utilizing the significant relationship pairs from second step, BBNs models were constructed Three different case studies were applied to evaluate the effectiveness of BBNs models Inputs from experts were obtained through face-to- face interviews with experts working in these three projects to construct CPT tables of BBNs models There are two limits in constructing BBNs model The first one is that all claim variables have two different status/ states only (YES/ NO) except for the final outcoPH QRGH àVFKHGXOH GHOD\ả WKUHH VWDWHV 7KH VHFRQG LV OLPLWLQJ WKHPD[LPXPQXPEHURISDUHQWQRGHVWRWKHFKLOGQRGHWRàả7KHVHOLPLWVDUHVHWLQ order to ease the burden of eliciting probability tables from the experts based on the previous studies and researches in other construction fields
The scope of questionnaire surveys and interviews are limited to industrial practitioners in Vietnam construction industry The respondents involved people, working for main contractors, owners/ clients, consultants and subcontractors & suppliers Experts in this study are middle-level and upper-level management, and have experiences of dealing with construction claims The location of the case studies is limited to Ho Chi Minh City to relive burden of data collection process.
Contribution to Academic and Practical Fields
This study contributes not only to the qualitative study of the major claims and their causes causing schedule delays in Vietnam construction projects but also to the quantitative study of claims assessment based on the BBNs model It also works as a framework for claim assessment of Vietnam construction projects b Practically:
The research topic aims to understand the crucial claim factors and their root causes to handle the claims effectively and reduce their negative impacts on the project schedule Moreover, with BBN's models, the quantitative claim assessment can be done, and the contractor can run the scenario analysis and estimate impact of claims on project schedule in advance for each claim's occurrence, which will reduce the schedule delays, improve the efficiency of the business, and increase the success of the projects.
LITERATURE REVIEW
Construction Claims
Construction industry has always been a benchmark to evaluate the growth of both national and global economic situations Nonetheless, construction projects these days are aggressive, risky, complicated, and large scale, involving of diverse, possibly multicultural stakeholders It is often impracticable to complete the project without alternating anything from the initial plan, and hence, changes and claims occur [10], [11]
,Q-DSDQWKHWHUP³FODLP´LVPDLQO\XVHGLQFDVHVUHODWHGWRGLVSXWHVDULVLQJIURP defects after delivering the construction project In Europe and North America, àFODLPả LV NQRZQ DV FRPSHQVDWLRQ IRU XQH[SHFWHG FDVHV DV D VROXWLRQ DPRQJ WKH concerned parties in construction, such as proposals for design changes, requests for time extension etc [12]*HQHUDOO\àFRQVWUXFWLRQFODLPVảLVGHILQHGDVWKHVWLSXODWLRQ for the reparation of extra money and/ or perpetuation of project duration for the additional work out of the initial contract It has a serious impact on the time and cost performance of the construction projects [11], [13], [14]
Generally, claims are inevitable in the development process, and good claim management and administration strategies are equally essential as managing quality, risk and safety, etc on construction sites The provisions are outlined for most types of contract, under which the contractor and client can claim against each other for any losses suffered if the progress of work is affected by specific causes [15]
Most business in developing countries like Vietnam, are currently encountering a serious recession Construction companies in Vietnam have experienced miserable business failure because of horrible financial troubles, as well as overdue project delivery As construction projects become larger, complex and mega scale, the involvement and interaction between project stakeholders also dramatically increased, by which exposure to changes and disputes also arise [11] Hence, claim management, especially, claim assessment framework plays a key role in enhancing the success of the project As a result, it is necessary for Vietnam construction companies to pursue a solid and steady improvement in handling the construction claims
This study aims to identify the crucial claim factors, their cause-effect relationship and model a claim assessment framework using Bayesian Belief Networks (BBNs), by which striving to improve the claim management system in the Vietnam construction industry.
General Process of Claim Procedures
Generally, there are six processes in the practices of claim management in the Vietnam construction industry, based on Hai D T [11] They are (i) claim identification, (ii) claim notification, (iii) claim examination/ assessment, (iv) claim documentation, (v) claim presentation, and (vi) claim negotiation This study focusses RQWZRVWHSVàFODLPLGHQWLILFDWLRQảDQGàFODLPDVVHVVPHQWả7KHLGHQWLILFDWLRQDQG assessment of construction claims affecting the schedule delays were conducted by literature review, questionnaire surveys and expert interviews
The first and foremost task in claim management is to identify potential subjects or causes that might lead to claims and disputes Since construction projects are generally operated in chaotic and complex environment, investigating problems that cause construction claims is not simple Therefore, on-site people, involving in daily activities, should have appropriate awareness and cognition of the claim procedure Moreover, construction contract is the most crucial one in dealing with any claims [16]
Enshassi A [17] investigated the causes of claims in Palestinian construction projects IURPFRQWUDFWRUVảSHUVSHFWLYHVE\FRQGXFWLQJDTXHVWLRQQDLUHVXUYH\DQGLQWHUYLHZLQJ Palestinian contractors in the Gaza strip Relative importance index (RII) was utilized to analyze the data The findings were that awarding bids to the lower bidder, border FORVXUHV UHVLGHQWVả LQIHUHQFH GXULQJ SURMHFW LPSOHPHQWDWLRQ URDG EORFNDJH DQG difficulties in passing between cities and governorate were the main causes of claims
Sang Hee C [18] aimed to identify the cause of claims and disputes in Korean construction projects from the aspects of construction companies using FIDIC red books through FMEA and Delphi method The results indicated that delay in work progress by the contractor was one of the top 10 dispute risk factors
On the other hand, Zaneldin [19] studied the types, causes, and severity of claims in UAE construction projects through the analysis of RII The results showed that changes and extra-work types of claims have the most occurrence, while contract ambiguity was ranked last For the FDXVHVRIFODLPVàFKDQJHVYDULDWLRQRUGHUVảZDV the most frequent, while delay caused by owner and oral change orders by owner rank second and third, respectively
The same author, Zaneldin [7], conducted another study for construction claims in UAE in 2020 The results were similar to the previous one, except for types of claims, changes and delay are the most frequent ones, and for the causes of claims, the first two rankings are the same, excluding changes in material and labor cost which ranks third
Similar studies about construction claims have also been conducted in ASEAN countries Hardjomuljadi [20] focused on the casual factors of construction claims in WKHFRQVWUXFWLRQRIK\GURHOHFWULFSURMHFWVLQ,QGRQHVLDIURPHDFKSDUW\ảVYLHZSRLQW involved in project execution The results revealed that the number-one causal factor of construction claims was inefficiency and disruption
Recently, Kongchasing [21] examined the factors causing construction delays from ERWK 7KDLODQG FRQWUDFWRUVả DQG FRQVXOWDQWVả DVSHFWV WKURXJK WKH 'HOSKL WHFKQLTXH The results were that shortage of qualified labor, change orders made by owners, and insufficient financial liquidity on the part of the contractor were three vital factors affecting the construction schedule to both the contractors and the consultants
In Vietnam, Hai D T [11] intended to investigate the claim practices in the VCI of both contractors and consultants The analysis was done by RII, and the results indicated that time extensions claim, change order claim, and design-related claims are three critical types of claims with frequent occurrence in VCI
Recently, Khoa D V [10] conducted a questionnaire survey among the industry practitioners of Ho Chi Minh City, Vietnam, with 117 observations to understand the causes of claims in VCI Through the technique of RII and exploratory factor analysis, the results revealed that diversity of working style among the parties; reluctance to work; and poor teamwork were top three causes of claims
In this study, identification of claims was first done by literature review through the recent studies & researches of claims in the construction industry Nonetheless, those lists of claims obtained from literature review are bloated and imperfect Some claims have similar meaning and can be combined to represent a broader range, whereas, some claims labels inclined towards academic phrase, and not easily understood by industrial practitioners and fail to indicate the actual on ground situations Therefore, prior discussion is first conducted with the experts to reduce the number of claims without damaging their essence Finally, by combining the literature review and expert inputs, 47 claims were determined and it is shown in Table 2.1
Table 2.1 Identification of 47 claims from literature review
Code Claims Affecting the Delays of Construction
C3 Late giving of possession to the site by the owner 3 3 3 3 3 3 3 7
C7 Bad quality of contractor's works 3 3 3 3 3 3 6
C8 Delays in executing the changes of the works 3 3 3 3 3 3 6
C9 Price fluctuation of construction materials 3 3 3 3 3 3 6
C10 Different Description of item in BOQ than what was mentioned in specifications 3 3 3 3 3 3 6
C11 Pricing of extra items not included in the contract of
C14 Poor planning and resourcing by contractor/ consultants 3 3 3 3 3 5
C15 Inadequate/ incomplete specifications and drawings 3 3 3 3 3 5
C16 Different interpretations of the contract provisions 3 3 3 3 4
C19 Financial failure of the contractor 3 3 3 3 4
C20 Low contract price due to high competition 3 3 3 3 4
C21 Delays in work progress by the contractor 3 3 3 3 4
C22 Inadequate investigation before bidding by the contractor 3 3 3 3 4
Code Claims Affecting the Delays of Construction
C25 Delay in obtaining approvals of documents and drawings 3 3 3 3
C29 Lack of understanding & agreement in contract document 3 3 3 3
C30 Slow decision making of the owner 3 3 3 3
C31 Unrealistic expectations by the owner 3 3 3 3
C32 Nominated subcontractors/ suppliers/ subcontracting problems 3 3 3 3
C33 Variations due to design errors 3 3 3 3
C35 Technical inadequacy of the contractor 3 3 3 3
C36 Difficulty in justifying the ambiguity of drawings 3 3 3 3
C41 Acceleration and stop-and-go operation by the owner 3 3 2
C42 Poor controlling and monitoring of the owner to his supervision team 3 3 2
C43 Mistakes by contractor during construction stage 3 3 2
C44 Waiting time for approval of tests and inspection 3 3 2
C46 Delays in choosing the type of contract 3 1
C47 Complex drawing preparation and approval process 3 1
There are two different approaches for the assessment of claims; qualitative and quantitative analyses At the start of the project, the project's usefulness is inversely proportional to the accuracy of the project Therefore, qualitative analysis is better suited for the earlier stage; a quantitative approach should be conducted after the detailed design comes out The two components of quantitative analysis are; (1) to calculate the probability of an unexpected event and (2) to estimate its impacts, where the impacts are in terms of functionality to the decision-maker, not the physical units (dollars, months, etc.) [31] The accuracy of estimation for the quantitative analysis framework is precise, but a detailed evaluation is needed to implement in the projects Project managers can figure out whether the project is in danger of not reaching its requirements and whether or not to minimize the risk with the quantitative study Although industry practitioners have some knowledge about quantitative analysis, it is rarely applied in the actual construction projects
Previous studies about construction claims based on their approaches are presented in Table 2.2 Vo K D [10] and Hai D T [11] performed qualitative study of construction claims in Vietnam using relative importance index (RII) Zaneldin E K [7], Illankoon I M C S [30], and Enshassi A [17] also conducted the qualitative study of construction claims in UAE, Sri Lanka and Palestine, respectively, using relative importance index (RII) Moreover, Le-Hoai L [29], Parchami Jalal M [28], Sibanyama G [23], and Love P [22] studied the causes of construction claims qualitatively using mean value, Ishikawa diagram, percentage presentation, and case laws & interviews, respectively
On the other hand, Enshassi A [27] and Parikh D [24] conducted both qualitative and quantitative study on construction claims by developing the claim prediction models with the help of neural network It was found that most studies focus on qualitative study of construction claims rather than quantitative one This might be due to the fact that quantitative studies are expensive, vulnerable to the personal bias of the experts and difficult to determine the accuracy of the results In this study, Bayesian Belief Networks (BBNs) is applied to overcome the weakness of quantitative study
Bayesian Belief Networks (BBNs) are graphical tools with a probability distribution They are useful for the conjecture of uncertain states when the amount of information available is insufficient Although there are still some disadvantages, BBNs have the following paramount advantages compared to other methodologies: (1) belief networks are capable of greater flexibility in terms of receiving input and yielding output [32]; (2) in belief networks, a variable can be provided as a known input or can be generated as an output of the system by evaluating its likelihood [32]; (3) BBNs can easily compute the probability of events before and after the entry of evidence and restore its analysis or speculation [33]; (4) In the case of lacking historical data, belief networks can be constructed using expert opinion [32]; (5) in belief networks, variables can be added or removed without any significant impact on the rest of the network because adjustments to the network is secluded [32];and (6) BBNs provide effortless discernment into relationships among variables of the process due to its graphical display [9]
RESEARCH METHODOLOGIES
Research Processes
There are 3 research processes in this study;
(i) identification of the critical claims in VCI
(ii) determining the cause-effect relationship among those critical claims
(iii) case studies application of BBNs model
The research flowchart for overall processes is shown in Figure 3.1
3.1.1 Identification of the Critical Claims in VCI
The intent of this phase is to identify the critical claims variables causing schedule delays in Vietnam construction projects A list of claim variable was identified from the literature review and prior discussion with experts This process yielded 47 construction claims, presented in Table 2.1, engendering the schedule delays on Vietnam construction projects
Questionnaire (Type 1) was established based on those claims 5-point Likert scale was used for this survey scale Respondents were required to choose one out of five options (from ³QRHIIHFW´WR ³YHU\VWURQJHIIHFW´WRGHWHUPLQHWKHLPSDFWRI FODLPVHQJHQGHULQJWKHFRQVWUXFWLRQGHOD\V,QWKLVVWXG\àVFKHGXOHGHOD\VảLVGHILQHG as time-overrun of a construction project over the completion date stated in the contract The sample questionnaire (Type 1) form is presented in appendices A pilot survey was carried out to determine the reliability of the preliminary questionnaire design The reliability of preliminary questionnaire design was tested using
&URQEDFKảVDOSKDFRHIILcient After that, the final questionnaire form was developed The questionnaire survey was conducted in large-scale among the industrial practitioners in VCI The questionnaires were distributed in both physical copies and soft copies via email The feedback from the respondents were stored in the Google Form The respondents were working as or under the main contractors, clients/ owners, consultants, subcontractors, and suppliers
Conduct large scale survey of questionnaire (1)
Using mean to evaluate and rank
Determine significant cause and effect relationship pairs
Determine parent and child node
Assign the conditional probabilities to each node based on experts' opinions
Develop BBNs model to quantify the outcome of claims
Assess the impact of claims on construction project schedule
Validate model by comparing with case studies Develop preliminary questionnaire & conduct pilot survey
Cronbachảs alpha coefficient Test the reliability of preliminary questionnaire design
&URQEDFKDOSKDảVFRHIILFLHQWLVXVHGWRYDOLGDWHWKHUHOLDELOLW\RITXHVWLRQQDLUHGHVLJQ and RII and mean value is used to rank the claims based on their degree of impact on project schedule
RII is a non-parametric technique commonly used in construction and facilities management researches to analyze the structured questionnaire responses for data involving ordinal measurement of beliefs or insights [37] In the previous studies, RII is used to rank the critical claim factors based on their ranking of importance [7], [17], [20], [38] The equation used in this study is similar to the one used in research paper [38] and shown in Eq (3.1):
Where RII = relative importance index; W = weighting given to each factor by respondents; A = highest weight; and N = total number of respondents The RII value had a range of 0 to 1 (0 not inclusive); the higher the RII value, the more significant was that claim factor In this study, RII is implemented to determine critical claims in VCI
According to similar experiment conducted by Kim S Y., construction claims with mean value higher than 3.5 are considered to be critical claims and it is adequate to proceed to the next phase using these critical claims [9] In this study, 23 construction claims (RII > 0.7, Mean > 3.5) were identified as critical claims and proceed to next stage to investigate the significant cause-effect relationship pairs among them
3.1.2 Determining Cause-effect Relationship Pairs
The purpose of this phase is to investigate the significant cause-and-effect relationship pairs among the critical claims identified in previous phase 3.1.1 Questionnaire (Type 2) is developed to investigate the significant relationship pairs among these 23 identified critical claims The questionnaire was organized in a matrix form where the 23 claims in the left column embodied the causes and the similar 23 FODLPVSOXVILQDORXWFRPHQRGHàVFKHGXOHGHOD\ảWRWDOO\IDFWRUVODEHOHGWKURXJK the top embodied the effects If every expert were required to assess every single combination pair of variables in the matrix, they had to evaluate 276 relationships, which was unfeasible, since most experts have congested schedule and less amount of available time Therefore, in this study, the author applied two small steps: (1) matrix form questionnaire (Type 2) survey first, and (2) face to face interview with experts utilizing prototype BBNs graphical model illustrated with the identified significant relationship pairs obtained from step (1) in order to ease the burden of DFTXLULQJWKHH[SHUWVảMXGJHPHQWV7HQH[SHUWVSDUWLFLSDWHGLQWKHZKROHSURFHVVWR determine the cause-effect relationship pairs, and the sample size is adequate to analyze the data based on similar experiment run by Nasir D with nine experts [35]
In the first step, matrix form, questionnaire (Type 2) was sent to five experts to collect their opinions on cause-effect relationship among 23 critical claims Experts were UHTXHVWHG WR HYDOXDWH WKH IDFWRUV XVLQJ WKH IROORZLQJ QXPEHUV àYHU\ VWURQJ UHODWLRQVKLSả àVWURQJ UHODWLRQVKLSả àVRPHZKDW UHODWLRQVKLSả àZHDN UHODWLRQVKLSả àQR UHODWLRQVKLSả 7KH VDPSOH TXHVWLRQQDLUH 7\SH IRUP LV presented in appendices To obtain conditional dependence relationships among pairs, this study developed nine logical tests for five-point Likert scale, using two statistical values, that is, the average and the skewness, based on Nasir D experiment [35] The skewness statistic represents the importance of the opinions where a positive skewness means that a major part of the experts provided a low score but a sparse amount of them provided a higher score and vice versa The weak (W) criterion is the number of experts that rated the relationship with the score of 2 and below, and strong (S) is the number of experts that rated the relationship with the score of 3 and above The difference between these two numbers, (W-S), indicates WKH JDXJH RI WKH LPSRUWDQFH RI H[SHUWVả RSLQLRQV /RJLFDO 7HVWV ) WR ) DUH elimination tests while Tests F4 to F7 are approval tests as shown in Table 3.1 A relationship that was determined as a false value for Tests F1 to F7 is proceeded to Tests F8 and F9 concurrently These two tests are subjective tests in reliance on the H[SHUWảVRSLQLRQV7HVW)DSSURYHVWKHUHODWLRQVKLS if the data tends to lean towards
4 and Test F9 eliminates the relationship if the data tends to lean towards 0 Eighteen cause-effect relationship pairs are obtained from this step
Test Conditions If condition true
3 Average 2.99 and skewness = negative Approve relationship
7 No 0 scores, experts see some relationship Approve relationship
8 Scores are leaned toward 4 Approve relationship
9 Scores are leaned toward 0 Eliminate relationship
In the second step, a graphical BNNs model was constructed using eighteen relationship pairs identified from questionnaire (Type 2) The prototype model is illustrated in Figure 3.2 This graphical model was then presented to the experts during interviews, and experts were encouraged to add new relationships or remove old relationship pairs to fix imperfect models due to limitations of questionnaire survey (Type 2) Finally, with the combination of questionnaire survey and interviews, seventeen cause-effect relationship pairs were determined among the critical claims
1LQHROGUHODWLRQVKLSSDLUVIURPTXHVWLRQQDLUHVXUYH\àYDULDWLRQRUGHU± schedule GHOD\ả[6-y), (àSRRUTXDOLW\RIGHVLJQ± VFKHGXOHGHOD\ả[19-\àYDULDWLRQRUGHU ± GHOD\LQREWDLQLQJDSSURYDOVRIGRFXPHQWVDQGGUDZLQJVả[6-x12àODWHJLYLQJ possession of site by the owner ± VFKHGXOHGHOD\ả[8-\àSRRUTXDOLW\RIGHVLJQ ± time extHQVLRQ(27ả[19-x20àGHOD\VLQZRUNSURJUHVVE\WKHFRQWUDFWRU- bad FRPPXQLFDWLRQ EHWZHHQ SDUWLHVả [4-x22 àYDULDWLRQ RUGHU ± incomplete/
LQDGHTXDWH VSHFLILFDWLRQV DQG GUDZLQJVả [6-x17 àEDG TXDOLW\ RI FRQWUDFWRUảV works ± delays in wRUNSURJUHVVE\WKHFRQWUDFWRUả[5-x4DQGàSRRUSODQQLQJ and resourcing by contractor/ consultants - mistakes by contractor during
FRQVWUXFWLRQVWDJHả[13-x2ZHUHUHPRYHG(LJKWQHZUHODWLRQVKLSSDLUVàWLPH extensions/ EOT ± VFKHGXOH GHOD\ả (x20-\ àGHOD\ LQ REWDLQLQJ DSSURYDOV RI documents and drawings ± YDULDWLRQRUGHUả[12-x6àGHOD\VLQZRUNSURJUHVVE\ the contractor ± VFKHGXOHGHOD\Vả[4-\àVORZGHFLVLRQPDNLQJRIWKHRZQHU±
EDGFRPPXQLFDWLRQEHWZHHQSDUWLHVả[7-x22àSRRUSODQQLQJDQGUHVRXUFLQJE\ contractor/ consultants ± GHOD\VLQZRUNSURJUHVVE\WKHFRQWUDFWRUả[13-x4àSRRU quality of design ± YDULDWLRQRUGHUả[19-x6àEDGFRPPXQLFDWLRQEHWZHHQSDUWLHV ± delays in work progress by the FRQWUDFWRUả[22-x4DQGàEDGFRPPXQLFDWLRQ between parties ± YDULDWLRQRUGHUả[22-x6) were added from interviews with experts
Figure 3.2 BBNs prototype from questionnaire Compared to prototype model and old relationship pairs from the questionnaire survey (Type 2), there are a lot of changes in the relationship pairs after expert interviews The reason why there are such differences is that in questionnaire survey, experts have to visualize the relationship pairs model in their mind, provide scores for 276 relationship pairs and they have to determine one out of five options from (0- 4) for each 276 relationship pairs This is a complicated process, consumes huge amount of time Hence, there might be some negligence and lack of concentration while answering the survey This might lead to misjudgments and inputs during questionnaire survey There might also be lack of clarity or misunderstanding of questionnaire items by the expert which lead to ambiguous conclusion It was found that with the graphical network aid, the data collection process became easier, experts could provide better judgements and they were more willing to provide their opinions compared to matrix type questionnaire survey The detailed information about these seventeen pairs were mentioned in next chapter
3.1.3 Case Studies Application of BBNs Model
In the 1970s, Stanford University developed Bayesian belief networks (BBNs), which are also known as belief networks BBNs utilizes graphical models to display the cause-effect relationships among variables It comprises of nodes, indicating variables, and arcs, indicating dependence relationships between the nodes [32] Figure 3.3 represents a straightforward belief network, with the node at the arrow's WDLOWKHSDUHQWQRGHàWHFKQLFDOLQDGHTXDF\RIWKHFRQWUDFWRUảZKLFKGLUHFWO\DIIHFWV WKHQRGHDWWKHDUURZVWLSWKHFKLOGQRGHàEDGTXDOLW\RIFRQWUDFWRUảVZRUNVả$Q arrow or an arc is generally used to portray the cause-and-effect relationship between the parent node and the child node Child nodes are conditionally dependent on their parent nodes
BBNs are supported by chance theory, known as Bayes' theorem, which was discovered by Thomas Bayes during the late 1700s [9] The formula used for BBNs model is explained in Eq (3.2):
Where P(A) is the probability of event A, P(B) is the probability of event B, and P(A|B) is the probability of event A given that event B has occurred
Figure 3.3 Example of a basic belief network in construction claims
Based on seventeen cause-effect relationships among construction claims from the section 3.1.2, the free downloadable computer program, MSBNX, was utilized to build the BBNs model The devised model then used three different types of construction projects as the case studies to certify the model Face to face interviews with experts, working in case study projects were conducted to quantify the impact of claims on project schedule Detailed procedures about this is stated in next chapter.
DATA ANALYSIS AND FINDINGS
Critical Claims in Vietnam Construction Industry
Through literature review and prior discussion with experts, 47 claims engendering the schedule delays in Vietnam construction projects were obtained, and a preliminary questionnaire (Type 1), mentioned in previous section 3.1.1, was designed based on these claims To verify the appropriateness and clarity of the factors in preliminary questionnaire, a pilot survey was conducted among six experts in VCI (four experts working for contractor, one as a consultant and one owner/ client representative) with no less than ten years of working experience in construction claims fielG&URQEDFKảVDOSKDFRHIILFLHQWZDVDSSOLHGWRFRQILUPWKHUHOLDELOLW\RI the preliminary questionnaire design by computing internal consistency among factors The result of the test was 0.918, which falls under the acceptable range of 0.5 to 0.95, thus certifying the reliability of the preliminary questionnaire item at 5% significance level according to the assertion of Ursachi G [39]
The questionnaire design was finalized and large-scale survey was conducted among the Vietnam construction industry practitioners Google Form was employed to gather and store the responses of participants 158 observations were received Out of which, 23 were eliminated because of the incomplete and inconsistent answers In total, 135 observations were considered to be qualified for the data analysis To determine the reliability of final questionnaire (Type 1), another test was conducted XVLQJ&URQEDFKảVDOSKDFRHIILFLHQWDJDLQ7KHUHVXOWZDVZKLFKLVPRUHWKDQ 0.5, but less than 0.95, thus confirming the reliability of the questionnaire item at the 5% significance level, as stated in Ursachi G [39] The requisite information of the respondents comprised type of organization, work experience, knowledge about construction claims, and experience in handling with construction claims, as shown in Table 4.1
There were 71 (52.6%) contractors, 28 (20.7%) consultants, 31 (23%) owners and 5 (3.7%) subcontractors & suppliers for the types of organization the respondents work as or for For working experience, 56 (41.5%) of the respondents had less than 5 years of working experience, 52 (38.5%) had 5 to 10 years of experience, 17 (12.6%) had
11 to 15 years of experience, 10 (7.4%) had over 15 years of experience Additionally,
128 (94.8%) of the respondents had knowledge about construction claims and, 83 (61.5%) had experience in working for construction claims, and 52 (38.5%) had no experience in handling construction claims Based on these data, respondents could deliver reliable judgments about the impact of construction claims on schedule delays as most of them are familiar with construction claims
Table 4.1 General information about 135 respondents
The RII ranking of 47 claim attributes engendering the construction schedule delays is indicated in Table 4.2 The top five claim attributes were: payment delays (RII 0.825, Mean = 4.126, R = 1), mistakes by contractor during construction stage (RII 0.801, Mean = 4.007, R = 2), financial failure of the contractor (RII = 0.799, Mean 3.993 R = 3), delays in work progress by the contractor (RII = 0.79, Mean = 3.948 R DQGEDGTXDOLW\RIFRQWUDFWRUảVZRUNV5,, 0HDQ 5
Table 4.2 Ranking of claim attributes engendering construction schedule delays
Code Construction claims RII Item
C43 Mistakes by contractor during construction stage 0.801 4.007 2 C19 Financial failure of the contractor 0.799 3.993 3 C21 Delays in work progress by the contractor 0.790 3.948 4 C7 Bad quality of contractor's works 0.776 3.881 5
C30 Slow decision making of the owner 0.764 3.822 7
C3 Late giving of possession to the site by the owner 0.757 3.785 8
C35 Technical inadequacy of the contractor 0.754 3.770 9
C41 Acceleration and stop-and-go operation by the owner 0.750 3.748 10
C25 Delay in obtaining approvals of documents and drawings 0.738 3.689 12
C14 Poor planning and resourcing by contractor/ consultants 0.732 3.659 13
C22 Inadequate investigation before bidding by the contractor 0.729 3.644 14
C33 Variations due to design errors 0.729 3.644 15
C15 Inadequate/ incomplete specifications and drawings 0.717 3.585 17
C44 Waiting time for approval of tests and inspection 0.710 3.548 21 C26 Bad communication between parties 0.704 3.519 22
C32 Nominated subcontractors/ suppliers/ subcontracting problems 0.695 3.474 24
C47 Complex drawing preparation and approval process 0.692 3.459 26
C31 Unrealistic expectations by the owner 0.683 3.415 28C8 Delays in executing the changes of the works 0.679 3.393 29C20 Low contract price due to high competition 0.679 3.393 30
Code Construction claims RII Item
C16 Different interpretations of the contract provisions 0.659 3.296 36
C42 Poor controlling and monitoring of the owner to his supervision team 0.656 3.281 39
C10 Different Description of item in BOQ than what was mentioned in specifications 0.653 3.267 40
C29 Lack of understanding & agreement in contract document 0.649 3.244 42
C11 Pricing of extra items not included in the contract of BOQ 0.627 3.133 43
C9 Price fluctuation of construction materials 0.619 3.096 45
C36 Difficulty in justifying the ambiguity of drawings 0.599 2.993 46
C46 Delays in choosing the type of contract 0.593 2.963 47
7KHGHILQLWLRQRIàSD\PHQWGHOD\VảLVWKDWWKHRZQHULVLQFDSDEOHRIPDNLQJSD\PHQWV to the contractor in time, leading to the delay of the work progress by the contractor and schedule delays of the construction in VCI This was ranked first Most Vietnam construction companies have been encountering dreadful business failure due to severe financial difficulties [11] When the payment from the owner is delayed, contractors, particularly from the small and medium construction companies, might not have adequate financial liquidity to proceed the project, hence, resulting in construction schedule delays à0LVWDNHV E\ FRQWUDFWRU GXULQJ FRQVWUXFWLRQ VWDJHả LQGLFDWHV WKDW WKH ZRUNHUV DQG engineers under the contractors executed some human errors such as misconstruction
RIGHVLJQE\VLWHHQJLQHHUVPLVLQWHUSUHWDWLRQRIHQJLQHHUVảLQVWUXFWLRQVE\ZRUNHUV hence, additional work and rework have to be executed and extra time and money will be wasted from the contractor [7], [19] This was a second rank claim variable, because, in developing countries, alike to Vietnam, there is usually a lack of skilled labors and engineers Due to inadequate skillful labor force, human errors came off during the time of construction, bringing on rework and extra work, and finally, schedule was delayed [21] à)LQDQFLDO IDLOXUH RI WKH FRQWUDFWRUả UHSUHVHQWV WKDW WKH FRQWUDFWRU GLG QRW KDYH sufficient financial capacity and failed to proceed the project before the payment was done by the owner [7], [19], [21] This was ranked third One of the major issues LQIOXHQFLQJ WKH FRQWUDFWRUảV ODFN RI ILQDQFLDO OLTXLGLW\ LV WKH FRQWUDFWRUảV ZRUN performance [40] When the contractor executed the unsatisfactory quality of work, the consultant evaluated the work below standard, and hence, the owner might not SD\WKHFRQWUDFWRU7KLVKDVDVLJQLILFDQWLPSDFWRQWKHFRQWUDFWRUảs finance, and the contractor might be in jeopardy of not meeting the financial requirements necessary to proceed the project, and in severe cases, there might be stopping of work, resulting in construction claims between owner and contractor à'HOD\V LQ ZRUN SURJUHVV E\ WKH FRQWUDFWRUả PHDQV WKDW WKH FRQWUDFWRU IDLOHG WR execute the work in compliance with the allocated schedule [7], [10], [19] This was ranked fourth Majority of the Vietnam construction projects are behind schedule due rudimentary management techniques [9] Owner might claim against the contractor if there is a serious delay in work progress, resulting in more wasted time and budget of the project à%DGTXDOLW\RIFRQWUDFWRUVZRUNVảLVGHILQHGDVcontractor executed the poor quality of work, hence, consultant evaluated the work below standard [7], [19], [41] This factor ranked fifth Majority of Vietnam construction projects have been suffering significant schedule delays due to the problem of poor workmanship and reworks [9],
[29] Poor quality of work may result in being rejected as unqualified from the consultant Rework has to be done and hence, it results in delaying the project deliver time
,WZDVIRXQGWKDWà3ULFLQJRIH[WUDLWHPVQRWLQFOXGHGLQWKHFRQWUDFWRI%24ả5,, 0HDQ 5 àLQFOHPHQWZHDWKHUFRQGLWLRQả5,, 0HDQ
5 àSULFHIOXFWXDWLRQRIFRQVWUXFWLRQPDWHULDOVả (RII = 0.619, Mean 5 àGLIILFXOW\LQMXVWLI\LQJWKHDPELJXLW\RIGUDZLQJVả5,, 0HDQ
5 DQGàGHOD\VLQFKRRVLQJWKHW\SHRIFRQWUDFWả5,, 0HDQ 2.963, R = 47) were in the last rankings and these claims were considered to have least impact on the construction project schedule in VCI Among these five claims, àLQFOHPHQWZHDWKHUFRQGLWLRQảDUHDOVRLQWKHOLVWRIODVWILYHFODLPUDQNVLQWKHVLPLODU experiments run by Kim S Y [9] and Le-Hoai L [6] in Vietnam construction industry Therefore, it can be concluded that Vietnamese industrial practitioners SUHVXPHGWKHLPSDFWàLQFOHPHQWZHDWKHUFRQGLWLRQảFODLPRQSURMHFWVFKHGXOHDVWKH least significant and have enough strategies to overcome this claim.
Comparison of Top Five Claims with Other Countries
To affirm the list of critical claims in this study, the top five critical claim variables engendering the schedule delays in Vietnam construction projects, identified in the previous section, were further investigated by making a comparison with the results from previous five studies in other nations, as presented in Table4.3 Even though these studies have different research objectives and methodology, their findings can be assimilated to discern the aspects of crucial claims in other countries
Based on the findings from Table 4.3 àSD\PHQW GHOD\Vả is the most critical claim variable engendering the schedule delays in Vietnam construction projects This claim variable is also the most dominant one in the Egypt construction industry (1st rank) and quite crucial in other construction industries, too; UAE (2006 ± 4th rank),
UAE (2020 ± 5th rank), and India (8th rank) Therefore, it can be concluded that àSD\PHQWGHOD\VảLVRQHRIWKHVLJQLILFDQWFODLPYDULDEOHVRFFXUUHGLQPRVWFRXQWULHVả construction projects à0LVWDNHVE\FRQWUDFWRUGXULQJFRQVWUXFWLRQVWDJHảLVFRQVLGHUHGDVRQHRIWKHFULWLFDO claim variables in Vietnam (2nd rank in this study) However, it is not dominant in UAE (2006 ± 18th rank), UAE (2020 ± 17th rank), and India (39th rank) Compared to those countries, Vietnam is a developing country, lacking in skilled labors and engineers, full of technically incompetent contractors, and hence, human errors and PLVWDNHV RFFXU IUHTXHQWO\ GXULQJ FRQVWUXFWLRQ à'HOD\V LQ ZRUN SURJUHVV E\ FRQWUDFWRUảLVRQHRIWKHPDMRUFODLPYDULDEOHVLQ9&,UDQk 4th) Nonetheless, it is not dominant in UAE (2006 ± 13th rank), UAE (2020 ± 23th rank), and India (7th UDQNàEDGTXDOLW\RIFRQWUDFWRUảVZRUNảUDQNHGILIWKLQ9&,1RQHWKHOHVVLWLVQRW crucial in UAE (2006 ± 10th rank), UAE (2020 ± 18th rank), and India (37th rank) Compared to other countries, most Vietnam contraction companies fail to deliver the project on time This might be due to frequent occurrence of human error and mistakes during construction stage, hence resulting in bad quality of works and ILQDOO\ OHDGLQJ WR VFKHGXOH GHOD\V 7KHVH WKUHH FODLP DWWULEXWHV àPLVWDNHV E\ FRQWUDFWRUGXULQJFRQVWUXFWLRQVWDJHảàGHOD\VLQZRUNSURJUHVVE\FRQWUDFWRUảDQG àEDGTXDOLW\RIFRQWUDFWRUảVZRUNảDUHWKUHHXQLTXHW\SHVRIFODLPYDULDEOHVIRU9&,
IntULJXLQJO\àILQDQFLDOIDLOXUHRIWKHFRQWUDFWRUảLVRQHRIWKHFULWLFDOFODLPYDULDEOHV in Vietnam (3rd rank in this study) and Thailand (3rd rank), but not dominant in in UAE (2006 ± 10th rank), UAE (2020 ± 19th rank), and India (18th rank) This is because Thailand and Vietnam are both developing countries with comparable geographical positions in South East Asia, and therefore, there might be some VLPLODULWLHVLQWKHRSLQLRQVRIWKHVHWZRFRXQWULHVảLQGXVWULDOSUDFWLWLRQHUV
2YHUDOOH[FHSWIRUàSD\PHQWGHOD\VảDQGàILQDQFLDOIDLOXUHRIWKHFRQWUDFWRUảFODLPV it was found that there were some differences in the ranking of top five claims between Vietnam and other countries around the world This might be because of cultural differences, socio-economic conditions, and different geographical locations
Table 4.3 Comparison of top five claim variables between this study and previous studies
Top 5 claim factors from the survey
Vietnam UAE UAE Thailand Egypt India
Mistakes by contractor during construction stage 2 18 17 - - 39
Financial failure of the contractor 4 10 19 3 - 18
Delays in work progress by the contractor 3 10 18 - - 7
Identifying Cause-effect Relationship Pairs
After obtaining 23 critical claims from questionnaire (Type 1), the second step is to determine conditional dependence relationships between these claims As mentioned LQVHFWLRQWRHDVHWKHEXUGHQRIDFTXLULQJH[SHUWVảMXGJPHQWVDXWKRUDSSOLHV two small steps: (1) matrix form questionnaire (Type 2) survey first, and (2) face to face interview with experts utilizing prototype BBNs graphical model illustrated with the identified significant relationship pairs obtained from step (1)
For the first step, questionnaire (Type 2) was developed based on the 23 critical claims Sample questionnaire (Type 2) form is presented in appendices Five experts participated in this process Nine logical tests using two statistical values, that is, the average and the skewness were applied to analyze and evaluate the relationship pairs The details about this procedure was mentioned in section 3.1.2 In the second step, a graphical BBNs model was constructed based on the relationship pairs obtained from step 1 The model, illustrated in Figure 3.2, was then presented to the experts to validate those significant pairs Five more experts were interviewed This resulted in seventeen significant cause-effect relationship pairs, as shown in Table 4.4 The mean YDOXHRIQHZUHODWLRQVKLSSDLUVZLWKDVWHULVNLVDVVXPHGDVàảIRUWKHVDNHRIEHLQJ presentable in the Table 4.4
2QHLQWHUHVWLQJWKLQJZDVWKDWWZRUHODWLRQVKLSSDLUVàWLPHH[WHQVLRQV(27± VFKHGXOH GHOD\ả [20-\ DQG àGHOD\ LQ REWDLQLQJ DSSrovals of documents and drawings ± YDULDWLRQRUGHUả[12-x6KDGWKHLQLWLDOPHDQYDOXHJUHDWHUWKDQàảLQWKH questionnaire (Type 2) survey process However, in the second process of generating a prototype graphical model, these two relationship pairs, were rejected due to nine logical test and the rule of limiting the maximum number of parent nodes to four to ease the burden of acquiring CPT tables from experts, based on the similar experiment conducted by Nasir D [35] Nonetheless, during the time of expert interviews, all experts determined that these two relationship pairs should be added to generate a more generalized theoretical BBNs model for claim assessments in VCI Therefore, these two relationship pairs have mRUH WKDQ PHDQ YDOXH RI àả FRPSDUHG WR RWKHU newly added relationship pairs (mean value = 3) based on expert interviews
7KHGDWDDQDO\VLVVKRZVWKDWWKHUHODWLRQVKLSEHWZHHQàPLVWDNHVE\FRQWUDFWRUGXULQJ FRQVWUXFWLRQVWDJHảDQGàGHOD\LQZRUNSURJUHVVE\WKHFRQWUDFWRUảx2-x4) has been ranked highest with the mean value 3.6 (Table 4.4) by the experts It was mentioned in previous section 4.2, that these two claims were considered among the significant construction claims leading to project schedule, only in Vietnam construction industry, compared to other countries This is because compared other countries, Vietnam is a developing country, experiencing inadequate skillful labor force and engineers, technical incompetency of contractors, and hence, human errors and mistakes happen regularly during construction [21] Due to occurrence of errors and mistakes during the time of construction, it resulted in rework and extra work, and hence, leading to schedule delays in most Vietnam construction projects [9] Owner might claim against the contractor if there is a serious delay in work progress, resulting in more wasted time and budget of the project
7KHUH LV D VLJQLILFDQW UHODWLRQVKLS EHWZHHQ àYDULDWLRQ RUGHUả DQG àWLPH H[WHQVLRQ (27ả x6-x20), with the mean value of 3.4 Time extensions/ EOT means that the whole construction project or a part of it fail to complete within the allocated time as specified in the contract [41] This can occur due to changing requirements from the RZQHUDQGRZQHUảVIDLOXUHWRPDGHLQWHULPSD\PHQWRQWLPH+HQFHLWLVUHDVRQDEOH WRFRQFOXGHWKDWWKHUHODWLRQVKLSEHWZHHQàYDULDWLRQRUGHUảDQGàWLPHH[WHQVLRQ(27ả is important
7KH UHODWLRQVKLS EHWZHHQ àEDG FRPPXQLFDWLRQ EHWZHHQ SDUWLHVả DQG àGHOD\ LQ REWDLQLQJ DSSURYDOV RI GRFXPHQWV DQG GUDZLQJVả KDV D PHDQ YDOXH RI DQG LV considered significant A project involves several people from diverse backgrounds in construction with the skills and experience assembled to work together It is important to have a transparent communication among multiple stakeholders of the construction projects and trust each other to lead the project to success [10] Deterioration of the relationship between main contractor and owner can lead to generate several difficulties during the time of construction, like frequent changes in requirements by owner and several delays in obtaining approvals of documents and drawings from the owner and his/ her consultant In severe cases, owner might request main FRQWUDFWRUảVFRPSDQ\WRFKDQJHSHUVRQLQFKDUJHRIWKHSURMHFWDQGNLFNKLP out This will lead to serious delays in construction projects Therefore, it can be FRQFOXGHG WKDW WKHUH LV D VLJQLILFDQW UHODWLRQVKLS EHWZHHQ àEDG FRPPXQLFDWLRQ EHWZHHQSDUWLHVảDQGàGHOD\LQREWDLQLQJDSSURYDOVRIGRFXPHQWVDQGGUDZLQJVả
([SHUWVLQGLFDWHGWKDWSDUHQWQRGHVàWLPHH[WHQVLRQ(27ảx4-y DQG à'HOD\ LQ ZRUNSURJUHVVE\WKHFRQWUDFWRUảx20-y) are directly affecting the project schedule (Table 4.4) This is a reasonable conclusion since these two claims are reported to cause significant schedule delays [26], [29], [41], especially time extension is a fundamental type of claims regularly occurring in the construction projects, leading to schedule delays [11]
Table 4.4 Seventeen cause-effect relationship pairs
Rank Cause-Effect Relationship Mean
*new relationship pairs from expert interviews
Case Studies Application of BBNs- based Model
4.4.1 Construction of Theoretical BBNs Model
Based on the identified critical claims engendering the construction schedule delays and the cause-and-effect relationships among those claims, a theoretical BBNs model was constructed The BBNs model was developed utilizing the free downloadable software MSNBX from Microsoft Inc, and is illustrated in Figure 4.1 In a BBNs, each variable has four distinct traits: (1) the variable name; (2) the variable status/ state; (3) its relationships (parent node) with other variables (child nodes); (4) its data WDEOHFRQGLWLRQSUREDELOLWLHVWDEOH&376LQFHWKHGHWHUPLQDWLRQRIDYDULDEOHảVVWDWH is crucial for computerized applications, each variable is designated its relevant status to reflect its conditions The state of the variables is shown in Table 4.5 The final
RXWFRPHYDULDEOHàVFKHGXOHGHOD\ảLVGHVLJQDWHGWKUHHVWDWHVWKDWLVàảà± ả DQG à!ả EHFDXVH WKH JHQHUDO WLPH-overrun of a construction project in
Vietnam is between 10% and 20% of the primary duration based on the research of
Figure 4.1 Theoretical BBNs model àảà±ảDQGà!ảDUHGHILQHGDVWLPH-overrun durations which are less than 10%, between 10% and 20%, and more than 20% in comparison with the primary duration stated in the contract, correspondingly The remaining variables have two distiQFWVWDWHVà