Petroleum Industry Contingency Planning Using Auditing Theories and Inferential Statistics

Petroleum Industry Contingency Planning Using Auditing Theories and Inferential Statistics

Copyright: © 2022 |Pages: 14
DOI: 10.4018/978-1-6684-5279-0.ch002
(Individual Chapters)
No Current Special Offers


This chapter demonstrates how to apply a novel Bayesian auditing statistical technique with a predictive probability distribution model to predict petroleum accidents. The goal was to first develop the petroleum oil spill patterns for a given population, develop a predictive distribution model, and validate that using Bayesian audit statistical techniques. Sample data were taken from a U.S.-based state and fixed interval random sampling was used to select the records for the audit. A beta distribution was generated illustrating the materiality parameters along with the misstatement results.
Chapter Preview


This chapter investigates if risk management techniques used in the auditing discipline could help identify and or prevent petroleum accidents. First the severity of the petroleum spill problem is briefly reviewed to establish the rationale for the current study. Next the relevant literature is reviewed to inform the methods. The methods are explained and a model is developed using sample data collected in the U.S.

Oil spill disasters happen more often around the world than environment conservationists would find acceptable (Liang, Ryvak, Sayeed, & Zhao, 2012; Mendelssohn et al., 2010; Strang & Nersesian, 2014). In the U.S. for example, there have been over 24,000 oil spills documented during the last two decades. In New York State there were at least 1,000 oil spills documented during the decade ending 2013 (Strang, 2015). Figure 1 depicts the frequency of oil spills in the USA during 2000-2020 based on the author’s analysis of data obtained from the U.S. Bureau of Transportation Statistics (USBTS, 2022). Figure 1 suggests the oil spill trend is decreasing since 2000, and as expected, the incidents were low during COVID-19 (2019-2020), which was as much data that were available at the time of the current study. Paradoxically and alarmingly, the analysis of the current study reveals that the volume of oil spilled actually increased dramatically to 714,416 gallons, just prior to the COVID-19 pandemic, which was the largest amount of annual oil spilled in almost two decades, since 2000. Overall, the data from figure 1 demonstrate that there were 24,245 known oil spills during the last two decades (2000-2020) resulting in a cumulative 3,781,384 gallons of oil contributing to pollution and critical resource wastage in the U.S.

Petroleum disaster prevention has been studied by several researchers, using inferential statistics and qualitative case studies. For example, Strang (2015) retrospectively analyzed 1005 oil spill accidents recorded by the New York State Department of Environment Conservation, in an attempt to develop an empirically-grounded environmental planning risk management model to prevent the frequency and severity of future human-caused petroleum-related accidents. Interestingly, Strang (2015) found gas service stations were likely have 4-8 times more petroleum accidents, and more so, repeat-offenders were common: If a gas station in New York had experienced a petroleum accident then it was 40% more likely to have a repeat occurrence. Nersesian and Strang (2013) described how a discrete probability distribution could be developed and tested with sample data to identify potential oil spills. Other researchers including Mendelssohn et al. as well as Matlow, Wray and Richardson (2012) used qualitative approaches to examine this oil disaster problem. The common finding across those studies was that novel techniques were proposed but there were no complete practice-based solutions for how to reduce petroleum accidents given the large geographic expanse of the earth.

Thompson (2012) made a salient point which was the impetus for the current study. In the U.S. he reviewed the problems arising from hydraulic fracturing of bedrock deep under the surface (fracking), with a finding that in addition to more accidents there were health implications including potable water table contamination and structural cracks in building foundations of nearby populations. His salient point was that more environmental audits were needed to prevent fracking problems. The current study takes his idea further, to examine if auditing theory, models and or procedures could be applied to detect petroleum accidents. The research question (RQ) in the current study became: Could risk-based statistical techniques used in the auditing discipline identify potential petroleum accidents?

This chapter makes a contribution to the petroleum contingency management body of knowledge by reviewing the literature and illustrating how to apply a novel Bayesian auditing statistical technique with a predictive probability distribution model to predict petroleum accidents. Prediction provides an important benefit in the contingency management practice because, if a decision maker can project a potential risk, then a contingency plan may be created to eliminate or mitigate any future occurrence. A statistical model can facilitate creating a contingency plan, but not all managers know how to do that. This chapter presents one possible approach.

Figure 1.

Oil spill disasters across the U.S. during 2000-2020

(collected and analyzed by the author)

Complete Chapter List

Search this Book: