The construction of large infrastructure can make a significant contribution to economic development. Misunderstanding the impacts of construction on the affected communities in countries that are developing their regulatory frameworks can lead to unintended consequences. These can manifest as risks for the project due to protests or outbreaks of disease. By articulating these risks due to social harm to engineering decision makers policies can be created that not only minimise risk but also increase benefits for the communities surrounding a project.
This report aims to investigate whether taking an Engineering Systems approach can provide more rigour to this process than using standard probabilistic analysis using a risk register; in collaboration with the consultancy MonkeyForest Consulting and the Nghi Son Refinery and Petrochemical LLC.
The study adapts Systems-Theoretic Process Analysis and Bayesian Network Analysis to study the social risks to the construction of the Nghi Son Refinery and Petrochemical complex in the Thanh Hoa Province of Vietnam as it increases operations to peak construction. This methodology allows the inclusion of expert opinion, behavioural theory of decision making, corporate social responsibility literature, and project data to build a high level analysis of project risks manifesting from social hazards.
The analysis shows that project risks emerge as a result of changes in components of the socio-technical system surrounding the project. These changes propagate through the system from social impacts by the project to social risks to the project.
The components form subsystems and if the rate of change of social subsystems is faster than the reaction of project subsystems risks will emerge. These changes can be reacted to or anticipated if the feedback mechanisms between different subsystems of the whole system are working correctly.
Two significant flaws were found. First of all the management structure of the project caused delays between the Social Team learning of a social grievances and implementing a solution with construction teams. Secondly, the tendency of the project to use the grievance reporting mechanism as a process to appease the lenders rather than a tool for organisational learning was hindering the project’s ability to respond to social changes that will manifest project risks.
Finally, the analysis of the causal and controlling structures of the socio-technical system were used to inform a simple and subjective Bayesian Network Analysis of the risk of project delays due to protest. The model showed the logical connection between the probability of conflict with the affected community and the subsequent cost of delays assuming three different levels of community engagement and organisational learning. This showed a fivefold increase in costs and increase in probability of conflict if community engagement and organisational learning was not conducted sufficiently.
In conclusion this study shows that Engineering Systems approaches do provide significant benefits to the analysis of social risks and forms a strong base from which further work can be developed. Further work could develop a stronger analysis of component and subsystems interactions to determine how specific social risks emerge. This methodology could then be used across multiple projects, whilst ensuring project anonymity, to analyse general trends in social risk emergence and further support the development of policy to increase social benefits.