Patrick Talbot

Expose Patrick Talbot

We do not disagree with the need to extract minerals using processes such as horizontal drilling, commonly known as fracking. Oil and gas operations assure our energy independence. We do, however, see the danger in residential fracking. This book quantifies the risks of fracking in residential neighborhoods. The key message is that even though yearly risks associated with a single well are nearly negligible, the cumulative risks to health safety, welfare, and the environment for large fracking operations spanning many decades are very significant. Unique benefits of this risk analysis are that requirements are carefully defined, a numerical impact scale is provided, and a robust technique for converting qualitative risks to quantitative risks is elaborated. Calculations are based on referenced data sources, risks that the Operator identified to the Securities Exchange Commission are quantified, Best Management Practices are analyzed to determine which actually mitigate risk, a risk update network is defined, and the idea of background risks is described. The result of this risk analysis is that while individual risks for a single well are near negligible, the cumulative risk for the proposed nearly 100 wells over a period of 30 years is substantial. Best Management Practices, as stated, are deemed to be largely ineffective in reducing risk. Predictive analytics provides a truly innovative contribution to the practice of risk analysis. We develop a semantic network - a network with meaning - that mathematically combines uncertain evidence and propagates it upward in the network to identify risk by category, quantify the effects of mitigations and produce a measure of overall risk. This predictive analysis technology, sometimes referred to as a deep belief network or more generally as deep learning has found advocates in many domains.