Developing prescriptive environmental protection models from descriptive human accident behavior

Purpose – This study aims to examine human-made oil–gas disasters to illustrate how a prescriptive model could be developed. Resilience to human-made disasters, such as oil or gas spills, can be improved by using prescriptive models developed by analyzing past behavior. This type of study is useful for urban planning and monitoring, as there is a higher probability of human triggered disasters in densely populated areas. Design/methodology/approach – This study examined 10 years of more than 1,000 oil–gas disasters that were caused by humans in the upstate New York area to illustrate how a prescriptive model could be developed. Findings – A statistically significant predictive model was developed that indicated humans in certain industry categories were approximately six times more likely to have an oil–gas accident resulting in environmental pollution. Research limitations/implications – A prescriptive environmental protection model based on human accident behavior would generalize to all levels of government for policy planning, and it would be relevant to environmental protection groups in any region with a large population of humans using oil and gas (that covers most countries on earth). Originality/value – The empirical risk management literature was reviewed to identify factors related to environmental accident prediction with the goal of developing an explanatory model that would fit the oil–gas human accident data.