An analysis method of influencing factors of crude oil carbon footprint based on XGB-CE: a case study of an onshore oil production area in Shengli Oilfield, China
摘要
The production of crude oil products will produce a large number of carbon emissions, and many production factors in the production process have complex nonlinear effects on the emission results. In order to determine the key influencing factors of crude oil carbon footprint, the carbon footprint of the whole crude oil production process in an oil production area of Shengli Oilfield was calculated based on the life cycle assessment method. A method for analyzing the importance of influencing factors of crude oil carbon footprint based on XGBoost and Copula Entropy is proposed: XGB-CE. It establishes an accurate regression model through XGBoost and uses Copula Entropy to separately quantify the nonlinear correlation information between the input factors and output results in the model, and the importance of the influencing factors in the model was obtained. Finally, the influence law of key factors is analyzed by using partial dependence diagram. The results of carbon footprint calculation show that extraction process has the largest contribution to the carbon footprint of crude oil products, accounting for 79.52%; the link with the largest contribution is the water and steam injection link, accounting for 29.06%; the type of emission that contributes the most is electricity consumption, accounting for 57.67%. This is due to the use of a variety of injection methods to improve the oil recovery rate in the oil production area, which consumes a lot of electric energy. The analysis results of the influencing factors of carbon footprint show that the liquid measure is the most important factor affecting the emission of electricity consumption, and the inlet temp has the greatest impact on the emission of fuel combustion. At the same time, it also shows the influence laws of these key factors and puts forward emission reduction suggestions, which can provide reference for enterprises' future emission reduction decisions.
Graphical Abstract