Extended Data Fig. 5: Correlation matrix plot for destination country-level refining CI, quality of crude refined, and volume shares of different types of refinery. | Nature Climate Change

Extended Data Fig. 5: Correlation matrix plot for destination country-level refining CI, quality of crude refined, and volume shares of different types of refinery.

From: Carbon intensity of global crude oil refining and mitigation potential

Extended Data Fig. 5

All values are volume-weighted-average of each destination country. Numbers in top left of each subplot are Pearson correlation coefficients. A negative correlation (R = −0.2) exists between refining CI and the volume-weighted-average crude API gravity of each destination country. This implies that countries that refine crudes with higher API gravity would generally result in lower emissions. Destination refining CI is negatively correlated with the volume share of hydroskimming (R = −0.71), and moderately positively correlated with the volume shares of deep conversion coking (R = 0.33) and hydrocracking (R = 0.44). One could derive a first approximation of a country’s refining CI solely based on its refining capacity of hydroskimming, even though hydroskimming only has a 7% global volume share. As counterintuitive as this might seem, hydroskimming refineries generally have, if volume-weighted-average values are taken, lighter crude inputs and thus significantly lower refining CI (17.3 kg CO2eq bbl−1) as compared to the global average (CI = 40.7 kg CO2eq bbl−1) as well as other refinery types (Supplementary Table 6). Therefore, countries that solely rely on hydroskimming tend to have refining emissions that are significantly lower than others.

Back to article page