Research on Delay Risks of EPC Hydropower Construction Projects in Vietnam

In recent years, in Vietnam, economy has been developing rapidly. To ensure rapid and sustainable economic growth, strong support from the energy sector is required. Governments in Vietnam have invested in numerous hydropower projects, many of which employ the EPC (Engineering, Procurement and Construction) contract. However, the EPC general contractors are facing many difficulties, resulting in schedule delays and considerable losses. This research is conducted to highlight the main risk factors in the delays of hydropower construction projects in Vietnam. The research employs the method of statistical calculations and risk analysis to obtain feedback from experts participating in similar projects. The research outcomes are as follows: identifying the risks that can cause delays in EPC hydroelectric construction projects in Vietnam; calculating and classifying the degree of impact of each risk to the progress of the construction. The practical significance of this study is to ensure the timely completion of projects, benefits for the investors, and the EPC general contractors.

To identify these factors, the author analyzes the characteristics of hydropower projects combined with the opinions of experienced experts with hydropower projects in Vietnam.On this basis, hypotheses about risk models are developed.
In recent years in the Vietnam, more attention has been given to risk management of hydropower projects.For instance, Zhao Juelong (2008) [1] studied cases of EPC hydropower projects in Vietnam, proposed risk factors, and suggested ways to minimize risks and proposed management measures.Li Wei (2012) [2], through the research of the Con River hydropower station in Vietnam, showed risks in project procurement, contract construction, material purchases, risks of delays in the project, and the increased expenses in construction, Jixin Wei and Liujian Zhe, through the "The whole process of overseas engineering project risk management" [3] studied about project risks.The most general characteristics of hydropower projects following the EPC in Vietnam as follows: 1) The use of EPC in Vietnam is relatively new, and project management is poor; 2) At the construction sites, the people's culture standard is low, causing various difficulties; 3) Resettlement, land withdrawal and handover for the construction contractors are complex; 4) Hydropower equipment for the projects must be imported from abroad with complex procedures, difficult shipment, and slow assembly; 5) In Vietnam currently keep high inflation rates, which affect the purchase of required materials, machines, and equipment; 6) Natural conditions such as climate, hydrology, topography, and geological conditions lead to further complications; 7) The sub-contractors' construction capacity is poor; the domestic construction technology has low productivity, and is not up to standard; 8) The infrastructure and traffic facilities for transport are poor; machine and equipment transportation encounter many difficulties, leading to delays, etc.
In the above mentioned literature, the author finds that research on risks in hydropower projects in Vietnam is still limited.With the reality of tardy construction projects and progress delays, the author deems it urgent to conduct research on risks involved in delaying the construction progress of the hydropower project using EPC in Vietnam.
The research employs the method of statistical calculations and risk analysis to obtain feedback from experts participating in similar projects.The research outcomes as follows: identifying the risks that can cause delays in EPC hydroelectric construction projects in Vietnam; calculating and classifying the degree of impact of each risk to the progress of the construction.

Project Orientation
Using the public information on the Internet, television, newspapers and other documents, the author carried out on-site interviews with experts and officers participating in EPC projects.On the basis of these opinions, the author hypothesized the risk factors, and calculated statistical with SPSS and AMOS software to analyze and complete the objective: research on delay risks of EPC hydropower construction projects in Vietnam.

Research Structure
The structure of this research includes three main parts: 1) The risk hypothesis and the impacts of risks on construction schedule; 2) Calculation and inspection of risk; 3) Controlling and limiting risks.

Risk Variables
Through the analysis of information and consultation of experts' opinions, we summarize the characteristics of the hidden risks leading to delays in the construction progress of the hydropower projects.Based on these characteristics, the main reasons leading to the construction progress delays can be divided into the following groups: Risk from contracts (B1), Risk from politics and law (B2), Risk from technology (B3), Risk from natural conditions and social environment (B4), Risk from economy (B5), Risk from management (B6), Risks from EPC general contractors (B7).Table 1 is systematic table of risk factors.

Selection of Variables for Risk Calculation Models
Based on the above hypothesis of risks, the author summarized and proposed the hypothesis of the risk model affecting progress in Figure 1.

Data and Supporting Software
From the hypothesis of risks in Table 2, the author did an investigation using slips with 5 levels of risk assessment as follows.

Calculate and Analyze the Discovery Factors
Before performing the SEM model simulation, it is necessary to conduct the calculation and analysis of the discovery factors, investigate the main factors, including the observation variables (survey questions), and test the reliability as shown in Table 4.In the factor analysis of SPSS, the factor deduction method "Principal Axis Factoring" and the horizontal rotation method, Promax, were used.
The results shown in Table 4, the KMO test coefficient features the value of 0.705 (>0.5), and the coefficient with the Bartlett statistical meaning of (Sig < 0.05).This proves the survey results have reliability; the question hypotheses are reasonable; the survey data is proper, and objective.The data is sufficient for conducting analysis in the following steps.Additionally, each variable features the factor loading coefficient larger than 0.5; Jabnoun & Al-Tamimi (2003) [7] providing that the factor loading coefficient of the variables is not less than 0.3, Gerbing & Anderson (1988) [8] clarifies the percentage of variance higher than 50%.Initially, the author used 18 variables, based on the standard of the factor loading coefficient larger than 0.5.The author gradually deleted the variables b3.2, then the factors analysis was conducted.Seven factors were chosen, B1, B2, B3, B4, B5, B6, B7, whose percentage of variance reached 56.5%, higher than the standard value of 50%, as shown in Table 5.

Analyze and verify the combination of factors
The author used the AMOS20.0software for 8 assumption factors and 19 assumption risk variables to calculate the standardized factor loading coefficient of the 19 assumption risk variables in the interval of 0.501 to 1.038 (Table 6).In accordance with the standard factor loading coefficient >0.5, which shows the assumption risk variables for the groups of combined factors in a close relationship; the hypothesized risk variables have the largest effect on the factors group, as pointed out in the model.

Calculate verify the efficiency of the factors
The reliability value of the CR combination of the minimum factor is 0.75.All values are larger than the standard coefficient of 0.5 [4], proving that the assumption variables compared with the assumption variables models is highly consistent.The author calculated the Average Variance Extracted, AVE, found the abnormal average values, and conducted the confirmation of convergence of assumption variables in the model.The result showed the AVE value is 0.51 to 0.74, All values are larger than the standard coefficient of 0.5 [9], proving the assumption variables compared with the factors with good convergence.
Verify the proposed model Shown in Figure 2 and from the following Table 7, it is possible to conclude that the assessment result is

Review the parameters of the model
According to the parameters of the regression model given in Table 8, the values (p) of the assumption items are also less than 0.05, which explains the reliability level of over 95%.The risk factors strongly affected the extension of the construction progress.

Figure 1 .
Figure 1.Specifies the assumptions of the risk system model affecting the progress.

Figure 2 .
Figure 2. The structure of SEM model and assessment result.

Table 1 .
Hypothesis of risk group.

Table 2 .
Investigation using slips with 5 levels of risk assessment.

Table 6 .
Average variance extracted and AVE values.