Over the past couple years of energy market turbulence, pretty much everyone has come to acknowledge that the U.S. — and the rest of the world — will continue to require refineries and refined products for decades to come. It’s also likely, though, that U.S. refiners, like their European counterparts, will be required to do more to reduce the volumes of carbon dioxide (CO₂) and other greenhouse gases (GHGs) generated during the process of breaking down crude oil and other feedstocks into gasoline, diesel, jet fuel and other valuable products. And, thanks to new federal incentives, it might even make sense for refineries to capture and sequester at least some of the CO₂ they can’t help but produce. In today’s RBN blog, we begin a series on refinery CO₂ emission fundamentals, the differing policies that are applied here in the U.S. and abroad, and how those policies might ultimately influence refining competitiveness.

It will come as no surprise that the refining industry generates significant volumes of GHGs, including CO₂, from both the refining processes themselves and the fossil fuel consumption needed to power them — just consider the vast amounts of heat that need to be generated for distillation and other reactions that need to occur. It’s equally unsurprising that the refining industry — not to mention the separate-but-related transportation sector, which depends heavily on refined fuels — has been coming under increasing scrutiny regarding GHGs.

As we’ll discuss in more detail later in this series, the regulation of GHG emissions from refineries is still a work in progress here and in other parts of the world where strategies aimed at mitigating emissions vary widely in scope and efficacy. As we’ve recently detailed in previous blogs on the topic, the U.S. federal government has mainly attempted to reduce GHG emissions through indirect methods: the Renewable Fuel Standard (RFS) and various energy-efficiency programs as well as the adaption of prior legislation — the Clean Air Act and Corporate Average Fuel Economy (CAFE) standards, among them — most of which are aimed at reducing demand for refined products and, by extension, reducing the amount of refined products that need to be produced. The one federal regulation that directly impacts refiners is the GHG Reporting Program (GHGRP), which since 2010 has required about 8,000 industrial and other entities that emit over 25,000 metric tons (MT) annually to report their emissions. This regulation, which helps quantify the scope of emissions, is a likely first step toward more informed rulemaking.

For now, at least, state-level regulations are probably the biggest issue for some refineries. For example, for several years California has overseen a cap-and-trade program that it uses to help ratchet down GHG emissions from refineries and other emitters, and a few other states have been following suit. Those will be the subject of a future blog in this series.

Before we dive deeper into federal and state GHG regulation — current and prospective — we need to review the sources of emissions so that we can understand how effective those regulations might be. So, today, we’ll look at (1) how the refining industry’s CO₂ and other GHG emissions fit into the broader picture, (2) which parts of refining account for most of the industry’s GHG generation, and (3) what refineries can do to reduce their GHG emissions.

The sourcing of GHG emissions in the U.S. is usually broken down into seven categories: Electric Power, Transportation, Industry, Agriculture, Commercial, Residential, and Land Use (see Figure 1 below). As a result of big reductions in coal-fired generation by the Electric Power sector over the past few years (gray line), Transportation (orange line) is now the most significant source of GHG emissions, followed by power generators and capital-I Industry (yellow). These three biggest sector emitters represent about 75% of gross U.S. emissions. [You might notice the figure below only shows six categories, the missing category is managed lands (officially called Land Use, Land-Use Change, and Forestry), which results in an emissions sink (reduction) of around 750 MMT CO₂e (CO₂ equivalent) annually. So, we get a little help from nature.] Total emissions peaked in 2007 and have generally been trending downward since that time — again, thank coal-to-gas switching for most of that decline.

Figure 1 U.S. Gross GHG Emissions, 1990-2021. Source: EPA

GHG emissions from Industry (blue line in Figure 1 and sum of colored layers in Figure 2) peaked in the late 1990s but stopped declining around 2012. While many industrial subsectors continue to decline slowly, these declines are offset by increased emissions from oil & gas and chemicals (gold and dark gray layers, respectively). Refining emissions represent just under 3% of total gross emissions within the U.S., or just over 15% of industrial emissions (dark blue layer). But the products produced by refineries are responsible for around 30% of gross U.S. emissions in the Transportation sector. Another way to look at this, for every metric ton of emissions a refinery puts out, refinery fuel consumers — cars, SUVs, trucks, locomotives, airplanes, etc. — emit 10 MT.

Figure 2 U.S. Industrial Sector GHG Emissions, 2010-2021. Source: EPA

We can further break down refinery emissions by the sources within the refinery, as shown in Figure 3 below. Just over 60% of refinery emissions result from combustion to provide heat in process units and raise steam (dark blue layer and left axis). About a quarter of the GHGs released by refineries are linked to refining processes (aside from heat from furnaces; green layer), including flaring, coking, catalytic cracking and any methane that escapes from storage tanks and leaks. Finally, 10% of refinery emissions result from hydrogen production (from steam methane reformers, or SMRs; blue layer) and cogeneration units (dark gray layer). The number of refineries in the U.S. has declined steadily over the past several years, from 150 in 2010 to 137 by 2021. This 9% decline is not reflected in total refinery emissions because the shutdown of mostly smaller refineries has been offset by capacity expansions at larger refineries still in operation. Lower refinery utilization rates during the pandemic (gold line and right axis) appear to be the primary contributor to the drop in emissions in the last two years.

Figure 3 U.S. Refining Subsector GHG Emissions and Gross Inputs, 2010-2021. Sources: EPA, EIA

Next, we will look at the primary factor in determining the emissions intensity of a refinery. Figure 4 below highlights the overall CO₂ intensity calculated by dividing daily CO₂ refinery emissions by refinery crude capacity for U.S. refineries expressed as MT CO₂/bbl. On the left, we plot intensity against refinery capacity; on the right, we plot intensity versus refinery complexity. You might expect larger refineries to gain some scale and have a lower emissions intensity, but the data show this is not the case. On the right, we can see a better correlation between CO₂ intensity and refinery complexity, as measured by the Baker & O’Brien Replacement Cost Index, or RCI. RCI is a measure of the capital cost of upgrading units and their throughput at a refinery compared to the atmospheric crude distillation unit (CDU). A CDU would get an RCI value of 1 and a refinery’s complexity would move up as it adds more upgrading capacity, with the most complex refineries getting an RCI value of just over 20. More-complex refineries tend to also be more profitable, as they will create a larger slate of gasoline and low-sulfur distillates. Less-complex refineries tend to produce more heavy products like residue — often sold as feedstock to more-complex refineries or used in shipping as fuel — and asphalt for roads, ultimately retaining carbon.

Figure 4 CO₂ Intensity vs. Crude Capacity and Refinery Complexity in the U.S.
Source: Baker & O'Brien PRISM Refinery Model

We can look at refinery complexity in a more detailed way to highlight where emissions come from within the refinery by processing units. In Figure 5 below, we show how emissions are broken out using four U.S. Gulf Coast refineries of various complexity using Baker & O’Brien’s PRISM refinery model, with bars to the left in both charts being the least complex and bars to the right being the most complex. The chart on the left shows emissions by major units within the refinery and the chart on the right shows emissions by the function being provided. Refineries with fluid catalytic crackers (FCC), SMRs and coking units have much higher emissions per barrel than less-complex plants.

Figure 5 GHG Emissions by Type of Refinery (100 Mb/d capacity running WTI).
Source: Baker & O’Brien PRISM Refinery Model

Refining is, by necessity, a carbon-intensive industry, and refiners have only limited, incremental approaches that they can deploy to reduce emissions. Furnace efficiency improvements, i.e., finding ways to optimize the usage of heat, are generally the first thought in reducing the CO₂ footprint. However, most of the low-hanging fruit here has likely already been captured by attempts to reduce costs, especially during periods of high natural gas prices (e.g. 2003-08), and competitive pressures have squeezed out smaller, inefficient plants. Fuel switching for refineries — like switching from natural gas to biomass/biofuels, hydrogen or electricity — is another technique that is gaining attention, but that solution comes with various hurdles (e.g., infrastructure requirements) and costs. [It’s worth noting that including hydrogen in the industrial fuel mix is a major part of hydrogen hub development plans, like the Leading in Gulf Coast Hydrogen Transition (LIGH₂T) proposal.] Switching electrical power imports to renewable energy sources has similar potential as elsewhere in the economy but also has similar challenges around reliability, which is paramount for refiners. Additionally, purchased electricity does not result in site emissions and, therefore, is not ultimately under the refiner’s control.

A final emissions-reduction approach involves changes in the crude slate. In Figure 6 below, we run several crudes through an 8.3 RCI refinery with a coker and FCC to show the potential change in CO₂ emissions. We ran Eagle Ford condensate (59.8 API gravity and 0.02% sulfur), WTI (42.1 API gravity and 0.21% sulfur), Mars (28.9 API gravity and 1.84% sulfur) and WCS (20.3 API gravity and 3.43% sulfur), which can all be found in Gulf Coast markets, through the model. Crudes with higher API gravities and less sulfur are known to be higher quality, which also correlates with them producing much less emissions to upgrade. The big difference is that the lower the quality of the crude, the more the FCC and coker must be involved to refine that crude. The tradeoff is that lower quality crudes, while requiring more processing, are normally also sold at much cheaper prices than higher quality crudes, bringing better profitability to those refineries that can run them. By focusing on emissions during the crude-selection process, refineries could conceivably make changes through the crude slate while they optimize for profits, too (if carbon emissions came at a cost). But we are likely talking only a few percentage points here. We will look at the magnitude in crude oil value for different carbon prices in a future blog.

Figure 6: Emissions from Superlight, Light, Medium and Heavy Crudes Run Through a Refinery. Source Baker & O’Brien PRISM Refinery Model

To make significant changes to the emissions profile of a refinery, we are ultimately talking carbon capture. There are three basic techniques that can be implemented: post-combustion, pre-combustion and oxyfuel combustion. Without getting into the technical details and nuances of the various capture technologies, post-combustion capture (processing furnace flue gas through a solvent to recover the CO₂) is probably the most likely path forward for any refinery, at least as it relates to furnaces. However, a better target for CO₂ capture is one where higher concentrations or higher pressures of CO₂ streams are available. For those refineries that have a SMR hydrogen production unit, that is the best place to start as it offers the possibility of pre-combustion capture (at high pressure) or post-combustion capture at higher concentration than flue gas. The catch is, only the most complex refineries own their own SMRs (orange segments in left chart in Figure 5). Most SMRs (especially in the Gulf Coast region) instead are owned by large industrial gas companies that supply hydrogen to refineries through long-term agreements and pipeline networks.

Refining carbon-capture options are scarce and expensive. This means that regulation of refinery CO₂ emissions will result in higher costs. In the future, a refinery’s CO₂ “footprint” will likely become a key strategic and tactical consideration when considering expansion, closure, and crude selection.

In the next blog in this series, we’ll look at the policies that national governments in Europe and Canada — and state governments like California — have implemented to persuade refineries to invest in these options, and how the U.S. government’s revised 45Q tax credit and the incentives it provides for refineries — and other major GHG emitters — may lead them to consider carbon capture and sequestration.

Note: The article was authored by Alex Hardman of Baker & O’Brien and published on RBN Energy’s Daily Energy Post on March 17, 2023.

“Over the Hills and Far Away” was written by Jimmy Page and Robert Plant and appears as the third song on side one of Led Zeppelin’s fifth studio album, Houses of the Holy. Page and Plant wrote the song at Bron-Yr-Aur, a small cottage they rented in the Welsh countryside after finishing a massive North American tour with Led Zeppelin in 1970. The tune was originally called “Many, Many Times.” The intro section is played by Page on acoustic guitars, utilizing Eastern-influenced pull-offs in the key of G that Page is fond of. The midsection of the song is led by the band and guitar-driven riffs, followed by a quiet outro featuring Page on guitar and pedal steel guitar. The song was released as the first single from the album in May 1973 and went to #51 on the Billboard Hot 100 Singles chart. Personnel on the record were: Robert Plant (vocals), Jimmy Page (guitars, pedal steel), John Paul Jones (bass, piano, organ, Mellotron, synthesizer), and John Bonham (drums).

Houses of the Holy was recorded between December 1971-August 1972 with The Rolling Stones Mobile Studio at Headley Grange and Stargroves, and at Island and Olympic studios in London, with Jimmy Page producing and Eddie Kramer engineering. The album was released in March 1973 and went to #1 on the Billboard 200 Albums chart. It has been certified 11x Platinum by the Recording Industry Association of America. Two singles were released from the LP.

© 2023 Baker & O’Brien, Inc. Publication of this article without the express written consent of Baker & O'Brien, Inc., is prohibited.