Risk management in industrial projects using structural equation modeling

Article history: Received January 5, 2016 Received in revised format June 15, 2016 Accepted July 19, 2016 Available online July 21, 2016 This paper presents an empirical investigation to study the effects of different factors influencing on accomplishment of projects in Iranian oil industry. The proposed study designs a questionnaire consists of 50 questions in Likert scale with seven factors including sanctions, economy, scheduling, contractor management weaknesses, cultural/social, force majeure and contractee. The study considers the effects of these factors in three categories; namely risk of project scheduling, risk in project cost and risk in management weakness. Using structural equation modeling, the study confirms that all three factors influence on the success of oil projects. In other words, The results have indicated that budgeting as well as cost accounting is the most important factor in accomplishment of oil projects followed by weakness in management and having an appropriate scheduling. © 2016 Growing Science Ltd. All rights reserved.


Introduction
During the past few years, there have been tremendous efforts on detecting important risks involved in executing big industrial projects (Craighead et al., 2007;Olson & Dash Wu, 2010).Risk identification often generates nothing more than a long list of risks, which can be difficult to manage.The list is normally prioritized to detect, which risks should be considered first, but this does not give any insight into the structure of risk on the project (Hillson, 2003).Tang and Musa (2011) investigated the research development in supply chain risk management (SCRM).The study identified and classified the potential risk related to various flows, namely material, cash and information flows.Thun and Hoenig (2011) performed an empirical analysis of supply chain risk management practices based on a survey with 67 manufacturing firms conducted in the German automotive industry and identified supply chain risks by analyzing their likelihood to happen and their potential effects on the supply chain.According to Qazi et al. (2016) "Project complexity has been extensively explored in the literature because of its contribution towards the failure of major projects in terms of cost and time overruns".Qazi et al. (2016) proposed a method, which helps capturing interdependency among project complexity, complexity induced risks and project objectives.Cervone (2006) shed light on the issues associated with risk management in digital library projects as well as techniques for mitigating risk in these projects.ALNabhani et al. (2016) discussed the relative importance of public participation in legislation of TENORM risk management in the oil and gas industry.Salazar-Aramayo et al. (2013) shed light on factors, which influence Exploration and Production (E&P) project management success and corporate financial performance using structural equation modeling (SEM) methodology for a case study in a large Brazilian oil company.Marcella and Rowley (2015) presented an exploration of the extent to which project management tools and techniques could be implemented across creative industries through a comprehensive investigation of their application in the fashion industry in the North East of Scotland.Bjerga and Aven (2015) discussed management of risk for a case of large uncertainties and the adaptation of risk management in such circumstances.They used a case from the oil and gas industry to gain insights into how adaptive risk management could be used when giving due attention to the knowledge and uncertainty aspects of risk.

The proposed study
This paper presents an empirical investigation to study the effects of different factors influencing on accomplishment of projects in oil industry.The proposed study designs a questionnaire consists of 50 questions in Likert scale with seven factors including sanctions, economy, scheduling, contractor management weaknesses, cultural/social, force majeure and contractee.Details of the questionnaire is given in Table 8 in Appendix.The study considers the effects of these factors in three categories; namely risk of project scheduling, risk in project cost and risk in management weakness.The current research is a practical and applicable type of research and it is descriptive and inferential in terms of collecting data which explains the features of the sample and then generalizes these features to the statistical population.In this study, the descriptive research has been applied in survey type, because the means of collecting data (questionnaires) has been distributed and then collected in the samples of statistical population.The methods of collecting data in this research is of library and field operation kind.Thus, it can be generally said that this study is a practical descriptive survey research of correlation type analyzed by using exploratory function analysis for the first and second time and by the helping of SPSS and factor analysis together with Amos application and afterwards, by utilizing the factor analysis, the structural model was presented.Due to the fact that this is a case study and related to Derik energy company, the statistical population of this study is of accessible type in sampling method including the present managers, experts and employees and the volume of the samples (X) for explorative function analysis (5×q<x<15×q) and "q" stands for the number of questions in questionnaires Kline, Paul. (1994).First, the effective variables of the structures in the first step of Delfi were identified by the experts and then the effective variables were added from the sources including articles and the books after being collected.The common items of sources and elites were omitted and in the second step of Delfi elites, 38 effective variables of structures were ranked between the scales of 1 and 10 and 80 percent of the variables with high ranking were kept and the others were deleted and ultimately, "q" as the number of questions in questionnaires was designed equal to 31 effective variables in Likert scale between 1 and 5 and its content validity was confirmed the experts and the thesis supervisors and thesis advisors.The sample volume n=172 is considered between the minimum of 150 and maximum of 453.The questionnaires are distributed among the experts and managers and employees and in the final step, the questionnaires were collected and the entering of information and reforming the lost data by means of the median method were carried out (Draycott, & Kline, 1994).Fig. 2 demonstrates personal characteristics of the participants.

Fig. 2. Personal characteristics of the participants
As we can observe from the results of Fig. 2, most participants in our survey are middle aged with good university educations.The proposed study of this paper uses structural equation modeling (SEM) to measure the effects of three mentioned risk factors on accomplishment of industrial projects.Fig. 3 demonstrates the preliminary results of the implementation of the proposed study.The SEM has been extensively used for accomplishing different projects (Seyedaliakbar et al., 2016;Seyedaliakbar & Zaripour, 2017).

Fig. 3. The results of the implementation of structural equation modeling
As we can observe from the results of Fig. 3 and relationships among internal variables.Fig. 4 demonstrates the revised model after the changes have been made (See Appendix for more details of the method used).As we can observe from the results of Table 2, all statistics are within desirable limits.Chi-Square value is more than the desirable value, and root mean square error of approximation (RMSEA) is less than 10%.Other statistics such as Goodness of fit index (GFI), Adjusted goodness of fit index (AGFI), Non-normed fit (NFI) index, etc. have maintained values of greater than 0.9, which confirm the overall model.Therefore, We can examine the hypotheses of the survey based on the results.3 shows the results of our survey.

Conclusion and discussion
Based on the results of the survey, we are now ready to discuss of testing different hypotheses of the survey.The first hypothesis of the survey investigates the relationship between the risk in industrial projects and an appropriate scheduling of the projects.Table 4 shows the results.

Table 4
The results of testing the relationship between risk in industrial projects and an appropriate scheduling of the projects Correlation Sig.Result Appropriate scheduling → Risk in project 0.755 0.000 Confirmed According to the results of Table 4, we observe a positive and significant correlation between having an appropriate scheduling and risk in project.In other words, when there is an appropriate scheduling for accomplishment of a project, one may expect a reduction on risk involved in industrial projects.Table 5 demonstrates the results of testing the effect of cost involved in accomplishment of a project and the cost and budget.

Table 5
The results of testing the relationship between risk in industrial projects and cost of the projects Correlation Sig.Result Budgeting the project → Risk in project 0.75 0.000 Confirmed The results also confirm that there was a positive and meaningful relationship between these two variables and we can confirm the second hypothesis of the survey.Thus, we may expect to reduce the risk involved in project by appropriately assigning cost and budget.Finally, the last hypothesis of the survey studies the relationship between weakness in management and risk involved in accomplishment of projects, which are summarized in Table 6 as follows,

Table 6
The results of testing the relationship between risk in industrial projects and weakness in management Correlation Sig.Result Weakness in management → Risk in project 0.14 0.023 Confirmed The results of the survey have confirmed the relationship between these two variables and confirm the last hypotheses.Finally, Table 7 shows the results of standard coefficients.2013), for instance, presented similar implementation by using a conceptual model for project management of exploration and production in the oil and gas industry in Brazil and reported that the model could contribute substantially to the firm because it was a global representation of the main factors for improving E&P project management.Capolei et al. (2015) reported that in oil production optimization, we normally plan to maximize a deterministic scalar performance index such as the profit over the expected reservoir lifespan.However, when there are some uncertainty in the existing parameters, the profit results in a random variable, which could assume a range of values depending on the value of the uncertain parameters.They considered the concept of risk and explored how to handle the risk by applying appropriate risk measures.

Table 8
The questionnaire of the survey  Table 12 shows the results of Bootstrap selection procedure.Table 13 compares the results of Bootstrap with maximum likelihood estimator (MLE).

Fig. 1 .
Fig. 1.The structure of the proposed study

Fig. 4 .
Fig. 4. The revised model Fig. 5.The results of standard coefficients

Table 1
The results of statistical observations

Table 2
The results of statistical observations

Table 3
The results of comparison of details of the fitting model Table 7 have indicated that budgeting & cost accounting is the most important factor in accomplishment in projects involved in oil industry followed by weakness in management and having appropriate scheduling.The results of this survey are consistent with other findings reported earlier in the literature.Salazar-Aramayo et al. ( The assessment of normality test is executed in Table9.According to the results of Table9, some of the variables have maintained Kurtosis values of less than 2.58 and we can observe that 18 variables are not normally distributed.To handle this problem, we use maximum likelihood estimator and Bootstrap.Table10shows the Mahalanobis distance for observations farthest from the centroid.

Table 11
The summary of Mahalanobis distance for observations farthest from the centroid