Switching to heavy crude oil and synthetic crude oil derived from oil sands bitumen in refineries could double or triple refinery emissions and add 1.6-3.7 gigatons of carbon dioxide to the atmosphere annually from fuel combustion to process the oil, according to a new study by Greg Karras at Communities for a Better Environment (CBE), published in the ACS journal Environmental Science & Technology. This would increase total petroleum fuel cycle emissions by 14-33%. Extraction emissions would add to these percentages, Karras noted.
Karras compared refinery crude feed, processing, yield, and fuel data from four regions (PADD I, II, III and V, PADD=Petroleum Administration for Defense Districts) accounting for 97% of US refining capacity from 1999 to 2008 for effects on processing and energy consumption predicted by the processing characteristics of heavier, higher-sulfur oils. (Smaller, landlocked PADD IV, the Rocky Mountain states, refined non-diverse crude feeds.)
Preliminary estimates from fuel cycle analyses suggest that a switch to heavy oil and tar sands could increase the greenhouse gas emission intensity of petroleum energy by as much as 17-40%, with oil extraction and processing rather than tailpipe emissions accounting for the increment. This raises the possibility that a switch to these oils might impede or foreclose the total reduction in emissions from all sources that is needed to avoid severe climate disruption. Accurate prediction of emissions from substitutes for conventional petroleum is therefore critical for climate protection. However, estimates of the emissions from processing lower quality oils have not been verified by observations from operating refineries.
Crude oils are extremely complex, widely ranging mixtures of hydrocarbons and organic compounds of heteroatoms and metals. Refiners use many distinct yet interconnected processes to separate crude into multiple streams, convert the heavier streams into lighter products, remove contaminants, improve product quality, and make multiple different products in varying amounts from crude of varying quality. Factors that affect emissions from refinery process energy consumption include crude feed quality, product slates, process capacity utilization, fuels burned for process energy, and, in some cases, preprocessing of refinery feeds near oil extraction sites.
Estimates that construct process-by-process allocations of emissions among these factors have not been verified by observations from operating refineries in part because publicly reported data are limited for refinery-specific crude feeds and unavailable for process-level material and energy inputs and outputs. Research reported here distinguishes effects of crude feed quality on processing from those of the other factors using refinery-level data from multiple operating plants to estimate and predict the process energy consumption and resultant fuel combustion emissions from refining lower quality oil.—Karras 2010
The density of crude oils is proportional to the fraction of higher molecular weight, higher boiling point, larger hydrocarbon compounds in the oils that are distilled in a vacuum, then cracked into fuel-size compounds to make light hydrocarbon fuels, Karras wrote. The larger hydrocarbons have lower hydrogen/carbon ratios that require hydrogen addition to improve product quality and higher concentrations of sulfur and other catalyst poisons that are freed by cracking and bonded with hydrogen to remove them from the oil and protect process catalysts.
The hydrocracking and hydrotreating of gas oil and residua uses several times more hydrogen than does hydrotreating of lighter streams such as naphtha. These processing characteristics require increased capacity for vacuum distillation, cracking, and hydroprocessing of gas oil and residua in refineries designed to make light liquid products from heavier, higher sulfur crude oils.
Karras found that crude feed density and sulfur content could predict 94% of processing intensity, 90% of energy intensity, and 85% of carbon dioxide emission intensity differences among regions and years and drove a 39% increase in emissions across regions and years.
The fuel combustion energy for processing increased by approximately 61 MJ/m3 crude feed for each 1 kg/m3 sulfur and 44MJ/m3 for each 1kg/m3 density of crude refined. He also found that differences in products, capacity utilized, and fuels burned were not confounding factors.
This prediction applies to average CO2 emissions from large, multiplant refinery groups with diverse, well-mixed crude feeds and appears robust for that application. However, the method used here should be validated for other applications. If it is applied to different circumstances, the potential for significantly different product slates, poorly mixed crude feeds, synthetic crude oil impacts on refining, and effects on fuel mix emission intensity and hydrotreating resulting from choices among carbon rejection and hydrogen addition technologies should be examined.—Karras 2010
Greg Karras (2010) Combustion Emissions from Refining Lower Quality Oil: What Is the Global Warming Potential? Environmental Science & Technology 44 (24), 9584-9589 doi: 10.1021/es1019965