화학공학소재연구정보센터
Fuel, Vol.246, 187-195, 2019
Data-driven approach to fill in data gaps for life cycle inventory of dual fuel technology
Life cycle assessment (LCA) is a powerful tool and schematic model to evaluate an emerging technology in the oil and gas industry. In the oilfield operations such as drilling and hydraulic fracturing, dual fuel (DF) diesel engines utilize natural gas and diesel fuel simultaneously, thereby reducing diesel fuel consumption and offering certain emissions reductions. Emissions from DF engines can vary greatly depending on fuel consumption and how the engine is operated. In this study, linear regression and cross validation are applied to analyze field testing data of dual fuel engines. This study enables rapidly prediction of emissions and filling of data gaps for fuel efficiency, substitution ratio and other parameters to optimize DF engine operations. Greenhouse Gas (GHG) emissions are predicted by linear regression with uncertainty of 2.7% based on power, methane loss, natural gas and diesel consumption. Exhaust after-treatment system (ATS) adds complexity and will significantly increase prediction uncertainty for carbon monoxide (CO) and non-methane hydrocarbon plus oxides of nitrogen (NMHC+ NOx). This model could potentially be integrated into an engineering-based LCA tool-such as the Oil Production Greenhouse Gas Emissions Estimator (OPGEE) to predict environmental impacts for a range of energy consumptions. This is an innovative application of multiple linear regression and cross-validation to analyze dual fuel technology for the purpose of LCA.