3

A Better Way to Measure Climate Footprints

THE WIDELY DIVERGING climate footprints of various oils and gases (especially unconventional ones) make clear how vital it is to find ways to accurately measure these differences. In 2013, I hosted a symposium on unconventional oil at the Carnegie Endowment for International Peace, where I was a senior associate in the Energy and Climate Program. That morning, I stood before a packed room in the nation’s capital to tee up the conversation on uncovering the unknowns about oil.1 I arranged several panels and keynote addresses with experts from industry, government, academia, and nongovernmental organizations (NGOs). Top energy journalists served as moderators.2

Everyone in the room asked the same burning question: What does the transforming oil and gas landscape mean for climate change? The answer is complicated. The climate risks of hydrocarbons are not a one-size-fits-all proposition. As the preceding chapters have shown, their differences are much greater than their similarities.

We do know that long-held assumptions that a barrel of oil or a cubic foot of gas emits a set amount of carbon dioxide (CO2) are not valid.3 The varying climate footprints of oil and gas can amount to much more than the carbon contained in these raw materials. Lifecycle greenhouse gas (GHG) emissions—from generating energy and leaking pollutants in the early stages of extracting and processing oil and gas to the later stages of delivering and consuming a multitude of petroleum byproducts—could be just as divergent as the resources themselves are. The question remained, then, are the differences between the GHG emissions of different types of oil and gas large enough to matter?

To answer this question, I assembled a team of academic research partners. Together, we would use scientific methods, collect empirical data, and gather measurable evidence to quantify and compare the GHG emissions of twenty-first-century oil and gas in a systematic way.

In this chapter, I describe the development, design, data inputs, emission drivers, and uncertainties of the tool we developed to accomplish this task. The Oil Climate Index plus Gas (OCI+) is a first-of-its-kind assessment tool for estimating the lifecycle GHG emissions from different oil and gas resources.4 Its findings decisively overturn the widely held (but mistaken) beliefs that oils (and gases) all have basically the same climate impacts and that the transportation sector (mainly motorists) is responsible for essentially all petroleum sector emissions. In reality, some barrels of oil and cubic feet of gas pose far greater climate risks than others and needless amounts of GHGs are emitted and leaked by the oil and gas industry itself (supply-side Scope 1 and 2 emissions5) before fuels even reach customers (Scope 3 emissions).6 Armed with this knowledge, climate actions can focus on those oil and gas assets with large industrial climate footprints while we work to pivot the entire market to a durable net-zero emission energy transition.

Barreling Ahead

Immediately following the 2013 Carnegie symposium, I approached one of the panelists—Adam Brandt, a professor of energy resources at Stanford University.7 We agreed that the petroleum sector’s total climate impacts could not be estimated by simply counting the carbon contained in the oil or gas itself. Take, for example, a barrel of heavy Midway-Sunset oil (California’s most produced crude). According to the Environmental Protection Agency’s (EPA’s) calculations, one barrel is estimated to emit 448 kilograms of carbon dioxide equivalent (CO2e) in emissions when counting all of the carbon contained in that one barrel of Midway-Sunset oil.8 In reality, a barrel of Midway-Sunset oil emits an estimated 765 kilograms CO2e (an increase of over 70 percent) when all emissions from its production, refining, shipping, and wide-ranging end uses are considered.9

Adam’s research identified vast differences in upstream oil GHG emissions (from wellhead to refinery gate),10 which he projected using a novel GHG-estimating model.11 The California Air Resources Board formally adopted this tool to implement the state’s low-carbon fuel standard (LCFS).12 This policy encourages the use of alternative transport fuels with lower GHG emissions by imposing on oil refiners mandatory emission cuts with tradeable GHG credits.13

Such “product-centric”14 policies, however, can easily overlook GHG emissions from system leakage and petroleum coproducts, emissions that do not stem directly from consuming transport fuels.15 So-called well-to-wheel analyses do not offer comprehensive, process-level details, and they omit portions of the oil and gas supply chain.16 Such product-centric analyses make consequential errors that tend to underestimate the variation in total petroleum emissions.

Poor boundary choices that focus on select transport fuels miss the climate effects of coproducts like petroleum coke, heavy residual fuels, and petrochemical feedstocks whose GHG emissions and market values do not correlate well with those of gasoline and diesel.17 Moreover, using averages rather than more detailed assessments of representative practices does not capture the full range of observed variability in emissions levels.18

A more comprehensive method of assessing the differences between oil and gas resources is called for. Adam agreed that a resource-centric, barrel-forward approach was needed to count all the emissions in the whole barrel to highlight the GHG savings potential in the entire oil and gas supply chain.

Assembling the OCI+ Team

As we began adapting Adam’s upstream production GHG model, the next step was to model the GHG emissions of oil refineries. We did not have to search long for a project partner. Joule Bergerson from the University of Calgary and her colleagues were in the early stages of developing a model that indicated that emissions of the refining process varied as much as production-stage GHG emissions.19 Joule was eager to join the OCI+ research team.

Jonathan Koomey was our fourth and final partner.20 Jon and I had gone to graduate school together; worked at the Lawrence Berkeley National Laboratory under the same mentor, Art Rosenfeld; and had successively taught at the Yale School of Forestry and Environmental Studies (since renamed the Yale School of the Environment). Although we pursued different energy paths—mine paved with hydrocarbons and his amid electrons and data—two decades later, we would be reacquainted. It was clear that Jon’s broad-based energy and environmental expertise would greatly benefit our OCI+ team. With all the researchers in place, we set out to develop a new tool to estimate the total lifecycle emissions of the oil and gas sector.

Parsing oils by their climate impacts allows multiple stakeholders, each with their own objectives, to consider climate risks in prioritizing the development of future oils and gases and the adoption of greater policy oversight over today’s oils and gases. While stakeholders’ priorities vary, all actors would be better served by accurate, transparent measures of climate risk associated with different oil and gas resources.

What follows is an explanation of the various models underpinning the OCI+, how they work, the data they use, and what drives GHG emissions in each stage of the oil and gas lifecycle.

Constructing the OCI+

The OCI+ is a metric that takes into account the total lifecycle GHG emissions of individual oils—from upstream extraction to midstream refining to downstream end uses. It offers a powerful yet user-friendly way to compare oils and gases and assess their particular climate impacts before development decisions are made and once operations are underway.

The OCI+ uses the following open-source tools to evaluate actual emissions levels associated with an individual resource’s supply chain. The underlying OCI+ models currently run on an Excel-based platform, but future versions could be programmed to speed up execution and enhance functionality.21 The index’s main three components are Adam’s model for measuring upstream emissions from exploration until the petroleum resources enter the refinery, a model Joule developed for gauging the midstream emissions of the refining process, and a model Jon and I constructed to assess the downstream GHG emissions of the shipping and end uses of the resulting petroleum products.

The Oil Production Greenhouse Gas Emissions Estimator (OPGEE): Led by Adam Brandt, this model estimates upstream emissions from oil and gas exploration, drilling, production, separation, processing, and transport to the refinery inlet and gas distribution system.22

The Petroleum Refinery Lifecycle Inventory Model (PRELIM): Spearheaded by Joule Bergerson, this tool estimates the emissions and petroleum product yields of refining crude oil.23 It is the first open-source refinery model that estimates energy and GHG emissions associated with various crudes processed by different refinery configurations using different processing equipment.

The Oil Products Emissions Module (OPEM): Led by me and Jonathan Koomey, this module estimates the emissions from the transport and end use of all petroleum products yielded by a given oil or gas. An overriding goal of OPEM is to include (and thereby avoid) carbon leakage from petroleum coproducts.24

Putting these pieces together, Figure 3.1 illustrates a flowchart of the OCI+’s three underlying emission estimation models. Details on each model follow.

image

image

FIGURE 3.1 Simplified Process Flowchart of the Oil Climate Index + Gas (OCI+) Model

Notes: OPGEE assumes that any NGLs removed upstream along with the processed gas are not sent to the refinery. Any remaining liquid hydrocarbons that cross the field boundary, however, make their way to the refinery and are modeled in PRELIM. NGLs, natural gas liquids; OPGEE, Oil Production Greenhouse Gas Emissions Estimator; PRELIM, Petroleum Refinery Lifecycle Inventory Model; SLCPs, short-lived climate pollutants.

Source: Author’s depiction.

The OCI+ analyzes total GHG emissions using a mass-balance approach, which means that all the carbon that goes in must come out. Therefore, all coproducts are counted, and none are lost or hidden. Two functional units (the underlying metrics or bases) enable different comparisons. Lifecycle GHG levels can be measured per barrel of oil equivalent (BOE) of crude and gas produced. Alternatively, emissions can be assessed per megajoule of energy content for all final petroleum products. These units can then be converted to estimate emissions per dollar value of all petroleum products sold. The adaptability of this model was essential to presenting its findings in multiple ways, particularly ones that convey the economic factors at play.

Modeling Upstream GHG Emissions

Unearthing oil and gas deposits and preparing them for shipment to a refinery, petrochemical plant, or gas distribution system is the first step in the petroleum value chain. The processes involved differ depending on the hydrocarbons extracted. These are the upstream processes that OPGEE seeks to forecast. It is an open-source model that is free to download, modify, and use, as long as users make their assumptions known if they publish the results.25 Figure 3.2 lays out a simplified schematic of the OPGEE model.

image

FIGURE 3.2 Simplified Schematic of the Oil Production Greenhouse Gas Emissions Estimator (OPGEE) Model

GHG, greenhouse gas. Source: Author’s adaptation based on OPGEE 3.0aBETA Documentation.

The upstream GHG impacts of different oils and gases vary significantly. OPGEE has been run on nearly 9,000 global crudes from ninety countries, using data from nearly 800 references, including government sources, scientific literature, and publicly available technical reports.26 Proprietary databases can be used to supplement these data when information is unavailable in the public domain.27 OPGEE contains more upstream oil runs than any other modeling effort to date.

In the most recent version, OPGEE 3.0a, it is possible to estimate upstream GHG emissions from gas as well as oil production.28 When gases are separated from liquids (oil and water), they are then treated, gathered, dehydrated, decontaminated (removing excess CO2), and demethanized (removing methane). Heavier gases or natural gas liquids (NGLs) may be recirculated to lift further resources out of the ground or shipped to industrial customers. Methane (or marketable natural gas) is compressed and shipped to various end users. Throughout the various upstream processes, computations are made for gases (like methane, ethane, and other volatile organic compounds [VOCs]) that are flared, vented, and released as fugitive emissions.

What Drives Upstream Emissions?

GHG emissions result at every upstream stage due to combustion and leakage, which are influenced by several factors. The characteristics of a given hydrocarbon resource—such as gas content, water content, carbon content, and field age—determine the extraction techniques and surface processing required. GHG-sensitive resources include those with high gas-to-oil ratios, flaring-to-oil ratios, water-to-oil ratios, and low American Petroleum Institute (API) gravities (heavy oils). Upsets and nonroutine operations—such as blowouts, well workovers, and poorly maintained flares—can significantly elevate GHG levels. Emission drivers include pumping and compressing fluids and gases, generating steam, removing water and contaminants, leaking methane, and flaring and venting associated gases.

Where oil and gas deposits are located—the geography and surrounding ecosystem (whether it be desert, Arctic tundra, jungle, forest, or offshore)—influences how disruptive extraction is to land use. When resource development changes land use, this can affect the land’s biological (soil and plants) capacity for carbon storage. The more naturally stored carbon that is released, the more GHGs that are emitted. OPGEE accounts for land use–related GHG emissions from oil and gas development based on the land’s carbon richness and development intensity. When appropriate, data for a particular region can be input to overwrite the model’s fixed factors.

Once the oil and gas are extracted, they require various degrees of additional processing, whether on-site or at another nearby or distant facility. An oil field’s location, its distance from transport hubs, and refinery selection determine how the crude is shipped; the resulting transport emissions can be estimated accordingly. A gas field’s location, the availability of on-site processing equipment and types of byproducts that result, existing infrastructure in place, and purchasing arrangements determine its transport emissions that can be estimated. Sometimes pipelines, railroads, or trucks ship the oil and gas over land. In other instances, barges move oil over inland waterways, and seaborne shipments use marine vessels. The model uses standard routes as default values, which users can change for particular distances and modes. Figure 3.3 displays the relative share of highly variable upstream emission drivers for producing a range of different oil and gas assets.

image

FIGURE 3.3 Drivers of Upstream GHG Emissions for Different Oils and Gases

EOR, enhanced oil recovery; GHG, greenhouse gas; LNG, liquefied natural gas. Source: OCI+ Preview Web Tool, 2020.

Modeling Midstream GHG Emissions

Once the oil and gas has been extracted and moved to a refinery, the midstream stage of the petroleum value chain begins. Refining used to be a simple process that involved heating up and boiling oil to separate out its main components. But the changing nature of oils demands corresponding changes in refinery operations.

By adjusting for various refinery configurations and processing techniques, the PRELIM model analyzes how crude quality and refinery units affect energy use and GHG emissions. PRELIM can run numbers on a single crude or a blend of oils, and when combined with OPGEE, the model sheds light on the midstream emissions in the broader lifecycle of oils and the liquids contained in gases. The model influences the OCI+ in two important ways. It estimates midstream GHG emissions, and it outputs a given refinery’s petroleum product yield.29 The type and volume of products vary depending on the input crude and the refinery’s design.

Oils vary significantly in their midstream GHG emissions. PRELIM has been run on 343 crude oils processed in 478 refineries located in eighty-three countries, representing 93 percent of global crude oil refining throughput in 2015.30 Public oil assays, analyses of crude oil composition at various distillation temperatures designed to simulate the inner workings of a refinery, are used wherever possible, and proprietary databases are used to supplement these data when public information is unavailable. This model represents more refining oil runs than any other similar modeling effort to date.

Matching Oils to Refineries

Matching oil characteristics with refining infrastructure to meet end-use product demand is the ultimate goal of all refiners. Every refinery is unique in terms of the combination of equipment it uses, the blends of crudes it is optimized for, and ultimately the products it sells.

Figure 3.4 illustrates the three major refinery configurations, from simplest to most complex, assessed by PRELIM—hydroskimming, medium conversion, and deep conversion. PRELIM also contains a total of eleven processing unit combinations within these basic refinery categories. One deep-conversion refinery configuration, for example, might employ a coking unit to reject high levels of carbon in the form of petcoke, while another may use hydrotreating to add hydrogen to crude to remove sulfur and other contaminants and manufacture gasoline and other petroleum products.

image

FIGURE 3.4 Simplified Refinery Configurations in the Petroleum Refinery Lifecycle Inventory Model (PRELIM)

Source: Deborah Gordon and Eugene Tan, “Different Oils, Different Climate Impacts,” Carnegie Endowment for International Peace, 2015, https://carnegieendowment.org/files/DifferentOils_Print.pdf

The characteristics of individual process units incorporated into the PRELIM model were obtained from published literature and input from industry experts. Technically, each crude can be blended and processed in many different refinery configurations, but in practice crude oils are best matched to certain configurations. PRELIM selects the default refinery configuration that best suits a given crude oil based on its properties (API gravity and sulfur content). As such, light and sweet (low-sulfur) crudes are best processed in simpler refineries, and heavy and sour (high-sulfur) crudes are ideally directed to complex deep-conversion refineries. The following rules of thumb illustrate the basic principles at play:

Deep-conversion refinery: heavy crude under 22 degrees API with any sulfur level

Medium-conversion refinery: medium, sweet crude (22 to 32 degrees API, with less than 0.5 percent sulfur content by weight); medium, sour crude (22 to 32 degrees API with more than 0.5 percent sulfur content by weight); and light, sour crude (over 32 degrees API with more than 0.5 percent sulfur content by weight)

Hydroskimming refinery: light, sweet crude over 32 degrees API and less than 0.5 percent sulfur content by weight

Many experts think that a crude oil’s API gravity and sulfur content are reliable predictors of a refinery’s GHG emissions. This, however, is a fallacy that has long hampered the collection of the full range of data needed to model refining emissions. While API gravity and sulfur are good indicators of the GHG emissions stemming from a default refinery type and end use, they alone are not sufficient for conclusively determining midstream refinery GHG emissions.

What Drives Midstream Emissions?

PRELIM reveals that a number of factors lead to elevated GHG levels during midstream petroleum operations. Crude quality, the selected process units employed (the refinery configuration), and the energy efficiency of the process units all play important roles in determining the energy requirements and resultant emissions of an individual crude (or a crude blend). Individual oils are rarely refined in isolation, and PRELIM’s GHG-level estimates can be proportionally assigned to individual oils in a mixture of crudes fed into a refinery.

While inputs of heat, steam, and electricity influence refinery emissions, the amount of hydrogen required to process each crude is also a major driver of refinery GHG emissions. How hydrogen is generated largely determines the relative share of this emissions driver. The heavier the crude, the more hydrogen is usually utilized during refining. Figure 3.5 displays the relative share of highly variable, midstream emissions drivers for refining different crude oil assets.

image

FIGURE 3.5 Drivers of Midstream GHG Emissions for Refining Different Oils

GHG, greenhouse gas; SMR, steam methane reforming. Source: OCI+ Preview Web Tool, 2020.

Modeling Downstream GHG Emissions

Downstream GHG emissions encompass the marketing (transport) and consuming of natural gas and petroleum products.31 Transport can involve ocean vessels, pipelines, railcars, barges, or tanker trucks. The consumption of petroleum products meets demand for billions of kinds of goods and services in every economic sector. Transport movements and end uses are difficult to precisely track. The OCI+ sets default values that users can modify.

Assumptions about Downstream Transport

OPGEE models the transport of crude oil to the refinery entrance and processed gas on to the distribution hub. It also accounts for petcoke transport, when it is removed via upstream upgrading. OPEM models all petroleum product transport including petcoke when it is produced in the refinery. Default transport distances are set in OPGEE and OPEM. It is assumed that all crude oil is transported to Houston refineries by pipeline (US crudes) or marine vessel (imported crudes)32 and that gas is transported via pipeline within each continent to a major regional hub.33 (Natural gas converted to liquefied natural gas [LNG] for intercontinental transport employs ocean vessels and liquefaction and regasification operations.) Volumes transported are provided by PRELIM for refined products and OPGEE for coproducts removed upstream, such as some amounts of petcoke and NGLs.

Multiple modes are typically used to deliver petroleum products, which users can input in OPEM. For example, exported gasoline can move from the refinery via pipeline to a marine terminal, be exported by ocean vessel, be put into a pipeline at its destination, and then be moved by a tanker truck to refueling stations. Alternatively, domestic gasoline can be put in a pipeline and shipped to another city where it is then loaded onto a tanker truck to various gasoline stations. There are different modes and distances that petroleum products can follow, but the highest GHG emission intensities tend to occur in the final miles due to the low fuel efficiency of heavy-duty trucks. Current default assumptions in OPGEE and OPEM result in estimated transport emissions that minimally amount to an estimated 1 to 2 percent of oil lifecycle GHG emissions.34

Assumptions about Downstream End Use

While transport emissions are typically minor relative to those stemming from other parts of the petroleum lifecycle, GHG emissions derived from end-use consumption can dominate the lifecycle of oil and gas production. Prior calculations assessing the petroleum lifecycle have historically compared oil to alternative transport fuels and compared GHG emissions predominantly on gasoline and diesel yields. However, end-use GHG emissions are highly variable when one fully accounts for the consumption of all petroleum products, including petrochemical feedstock and bottom-of-the-barrel byproducts like petcoke, fuel oil, bunker fuel (known as bunker C), and asphalt. Figure 3.6 illustrates a simplified flowchart of the OPEM model for downstream emissions.

image

FIGURE 3.6 Simplified Schematic for the Oil Products Emissions Module (OPEM)

Note: The transportation of crude oil from the field to the refinery and natural gas and upstream NGL shipping GHGs are calculated in Oil Production Greenhouse Gas Emissions Estimator (OPGEE), but they are reconciled in OPEM. PRELIM, Petroleum Refinery Lifecycle Inventory Model.

Source: OCI+ Preview Web Tool, Methodology, OPEM Model.

What Drives Downstream Emissions?

Four variables drive GHG emissions from the transport of petroleum products: mode, distance, fuel used, and the mass of the product. Depending on the fuel that is used, the typical range of transport options for GHGs from the least emissions intensive to the most starts with ocean vessels, pipelines, trains, and barges before ending with tanker trucks. The distance traveled, delays incurred, and fully loaded return journeys also drive the transport share of GHG emissions.

However, it is difficult to ascertain precise modes, distances, fuels, and other logistical details for these figures. There is no global agency or group to collect and audit data that involves multiple private actors controlling highly dynamic logistics systems driven by changing supply and demand. This lack of centralized oversight matters because there is great variance in many of these variables, including the routes taken. For example, there are five different sea routes—ranging from 9,500 to 16,940 nautical miles—that a tanker of liquefied petroleum gas (LPG) can take from Texas to Japan alone (the largest importer of US products), each running through a different global choke point with its own geopolitical challenges.35 This multiplicity of fluctuating variables is true of nearly every route and transport permutation, a level of complexity that compounds the challenge of tabulating accurate downstream emissions totals.

End-use GHG emissions are driven by fuel type, the uses various petroleum products are put to, specific fuel quality, vehicles’ fuel efficiency and level of maintenance, and local conditions such as traffic congestion. Many, but not all, petroleum products derive their highest value by being burned. Table 3.1 charts the different emissions levels generated by combusting equivalent volumes of different petroleum products, as well as the differences generated by different fuels relative to gasoline.

Table 3.1 End-Use Emissions Factors of Refined Petroleum Products

Petroleum Product

Combustion GHG Emissions (kilograms of CO2e/gallon)

GHG Emissions Relative to Gasoline (+/− %)

Petrochemical Feedstocksa

1.3

86%

LPG

5.7

35%

NGLs

6.3

29%

Natural Gasb

7.5

15%

Gasoline

8.8

0

Jet fuel

9.8

11%

Diesel

10.2

16%

Fuel Oil

11.0

25%

Liquid Heavy Ends (Residual Fuels)

11.3

28%

Petroleum Coke

14.7

67%

a GHGs for petrochemical feedstocks consider emissions from converting ethane to ethylene.

b Natural gas combustion emissions are listed in liquid terms (per BOE).

BOE, barrel of oil equivalent; GHG, greenhouse gas; LPG, liquefied petroleum gas; NGLs, natural gas liquids.

Source: Environmental Protection Agency, Emission Factors for Greenhouse Gas Inventories, 2018,  https://www.epa.gov/sites/production/files/2018-03/documents/emission-factors_mar_2018_0.pdf

Calculating Carbon Dioxide–Equivalent GHG Emissions

Having calculated more accurate estimations of the emissions profiles of the full spectrum of petroleum products, the next step is to ascertain their respective impacts on the climate. To sum up the effects of different GHGs with varying climate forcing properties,36 the Intergovernmental Panel on Climate Change (IPCC) uses global warming potentials (GWPs) over set timeframes. Table 3.2 identifies current GWP multipliers for aggregating long-lived (CO2) and short-lived climate pollutants (SLCPs) over periods of 20 and 100 years.

The OCI+ and each of its three underlying models estimate CO2e levels of GHG emissions; it currently tabulates estimates for CO2, methane, and nitrous oxide. The model provides sliders to assess the lifecycle emissions of GHGs using the GWPs of these two timeframes. Future versions of the OCI+ could incorporate others like black carbon, carbon monoxide, nitrogen oxides, and VOCs—four additional GHGs that are prevalent in the oil and gas sector.

Table 3.2 Global Warming Potentials of GHGs from Oil and Gas Systems

GHGs Emitted

Oil and Gas Sources

Estimated Lifetime

100-Year GWP100

20-Year GWP20a

Carbon Dioxide (CO2)b

All oil and gas combustion

n/a (centuries)

1

1

Methane (CH4)b

Natural gas systemwide leakage; oil and gas combustion

12 years

34

86

Volatile Organic Compounds (VOCs)

Condensates, gas, light oil, petrochemicals

5 hours to 60 days

5

14

Ethane (C2)

Same as VOCs

58 days

13

n/a

Propane (C3)

Same as VOCs

13 days

12

n/a

Butane (C4)

Same as VOCs

1 week

8

n/a

Nitrous Oxide (N2O)b

All oil and gas combustion

121 days

298

268

Nitrogen Oxides (NOx)

All oil and gas combustion

22 days

−11

19

Ozone (O3)

Formed by VOCs + NOx and intermediary reactions

22 days

n/a

65

Carbon Monoxide (CO)

Incomplete oil and gas combustion

45 days

5

19

Black Carbon (BC)

Extra-heavy oil, fuel oil, diesel, and petcoke

100 years

460–900

1,600–3,200

CH2F2 (R32)

Industry refrigerants

5 years

677

2,430

CHF2CF3 (R125)

Industry refrigerants

28 years

3,170

6,090

Hydrogen (H2)

Future transition fuel

2.5 years

6

n/a

Water Vapor (H2O)

All fossil fuel combustion

Several days

<1c

<1c

a The GWPs cited consider climate-carbon feedbacks. Without these feedback values, the respective 100- and 20-year methane GWPs are 28 and 84 and the N2O GWP is 264. Note that GWP for methane in particular have been increased over time, starting at 21 in 1995.

b These GHGs and their 100- and 20-year GWPs are currently used in the OCI+ webtool.

c The GWP100 for near-surface emitted water vapor is in the range [−0.001, +0.0005] and the GWP20 is [−0.004, +0.002].

Notes: The GWP values are rounded to the next whole number.

GHG, greenhouse gas; GWP, global warming potential; OCI+, Oil Climate Index plus Gas.

Sources: Gunnar Nyhre and Drew Shindell, IPCC AR5, https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_Chapter08_FINAL.pdf; “Anthropogenic and Natural Radiative Forcing,” Tables in Appendix 8A, 2013, https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_Chapter08_FINAL.pdf; Climate and Clean Air Coalition; Olvind Hodnebrog et al., “Lifetimes, Direct and Indirect Radiative Forcing, and GWPs of Ethane, Propane, and Butane,” 2018, https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/asl.804; Leide Timma et al., “Temporal Aspects in Emissions Accounting,” Energies, 2020, https://www.mdpi.com/1996-1073/13/4/800/pdf; Aiswarya Rogothaman and William Anderson, September 2017, https://www.researchgate.net/publication/319911732_Air_Quality_Impacts_of_Petroleum_Refining_and_Petrochemical_Industries; Richard Derwent et al., “Global Environmental Impacts of the Hydrogen Economy,” 2006, https://www.geos.ed.ac.uk/~dstevens/Presentations/Papers/derwent_ijhr06.pdf; Steven Sherwood et al., “The Global-Warming Potential of Near-Surface Water Vapour,” 2018, https://iopscience.iop.org/article/10.1088/1748-9326/aae018/pdf

The IPCC has increased these multipliers over time as more data becomes available.37 Researchers now think that the current GWP approach of uniform multipliers masks the true behavior of much more potent SLCPs. As such, alternative approaches are under development.38 Increasing methane’s GWP will elevate OCI+ estimates for gas and light oil with high leakage potential. Given the high GWP for black carbon, adding this pollutant to future versions of the OCI+ is expected to raise the GHG emissions of heavy oils. In sum, while it was once thought that CO2 dominated the incidence of global warming, the more scientists learn about the climate-forcing nature of short-lived GHGs, the greater the role they are thought to play in climate change. As such, given the key role methane, VOCs, black carbon, and other non-CO2 emissions play in the oil and gas sector, the current OCI+ likely undercounts GHG emissions. The OCI+ estimates will be updated over time as new GWP assumptions come to the fore.

Model Data and Uncertainty

Whether global oil and gas production and consumption return to record levels, wane, or fluctuate in the future, the increasingly complex and unconventional nature of these hydrocarbons necessitates a better understanding of the oil and gas supply chain, especially as it transforms in the future. To make its assessments, the OCI+ models require trustworthy, open-source, standardized data. The more data that is gathered and input into the OCI+, the less uncertain its estimates are. And as new and updated information is gathered, the OCI+ can be rerun to reassess results.

To date, the oil industry has generally not been forthcoming with detailed information about their operations and future plans for their expansion. Of course, data can also be purchased from firms like Wood Mackenzie and IHSCera when modeling all oil and gas operations worldwide, but these input data are pricey and cannot be made public. General public data can be obtained from news sources, industry announcements, and academic journals. And more specific oil and gas data on companies and specific operations is expanding as investors, regulators, and the public demand greater transparency and remote sensing data is relayed from satellites and other sources. Remote sensing is also working to improve government inventories where open-source data is housed and the OCI+ can access these government records.

The OCI+ also uses remote sensing data inputs. Tower-based measuring stations, drive-by detection, and flyover techniques using satellites, aircraft, and drones are collecting data on GHGs. Governments, companies, and private citizens are involved, and some of these data are open source.39 For example, the National Aeronautics and Space Administration's (NASA’s) Carbon Monitoring System (CMS) shares data with stakeholders from its satellites, flyovers, and other missions.40 CMS is a multi-million-dollar initiative established following a 2011 pilot program to use NASA satellites to support national and international policy, regulatory, and management activities.41 In addition to measuring methane, CO2, and other GHGs, satellites also monitor gas flares, as discussed later.42 Satellite data can also provide intel about installed equipment holding out the possibility of using machine learning to identify and count various oil and gas emission sources like compressors, storage tanks, or heat exchangers from space.

If they are open source, other data streams that the OCI+ could use—like asset-grade and blockchain data—are coming online too. These data are obtained directly from their respective sources, and this information is verifiable and immutable, so any alterations can be detected. While not all of these asset-grade data are publicly available, some new firms are planning to share their data with civil society actors.43 These collections of data could someday offer near-instantaneous recordkeeping, greater transparency, and increased security.44 Once they are made public, such advances in data capture can spur new thinking on calibrated regulatory approaches, streamlined management practices, and asset investment strategies.45

OPGEE Data

OPGEE utilizes up to sixty data inputs, from simple entries like the country where an oil or gas field is located to hard-to-obtain information like a given field’s productivity index (expressed in daily production per unit pressure).46 Key variables include steam-to-oil and water-to-oil ratios, flaring rates, venting rates, system pressures, crude gravity, gas composition, gas-to-oil ratio, and production rates. Users do not need to identify every input value for a given inquiry because OPGEE functions with limited data. The model has comprehensive, smart defaults that fill in missing data. Reasonable estimates based on empirical analyses, technical references, and published journals are assigned to fill in for missing data.

The largest source of uncertainty in OPGEE is the lack of public information on global oil fields and their contents. Many operators in regions around the world are not required to formally publish such data. Such omissions mean that data are not consistently and routinely reported year over year, introducing uncertainty with respect to updating emissions over time. Imprecise data reporting introduces additional uncertainty. Errors in applying the model can lead to further uncertainty. In its open-source terms of use, OPGEE requires that all users that publicly release results publicly share their input data so that use of the model can be validated and any user errors can be identified and corrected.

PRELIM Data

Refineries use heat to separate oil and gas into various hydrocarbon components. To model this process, PRELIM requires crude oil assay data that is arranged into nine standardized temperature cuts to mirror real-world refining operations.47 Assays are typically published only when a crude variety is widely marketed, a condition that applies to only a small subset of global oils that trade globally in large volumes. The majority of crude assays are not made public, especially when oils are used domestically and not exported. Assays are also required for wet gas and condensate assets that contain some oil.

Data availability and quality are ongoing issues because PRELIM is sensitive to an oil’s composition. Uncertainty ensues when assays are outdated, unreliable, or not standardized. For example, PRELIM transforms assays that do not contain certain temperature cuts, and the model uses a proxy assay when one is missing altogether. As new assays are publicly reported, PRELIM can be rerun to update the results. A global library of open-source oil assays would make it possible to run PRELIM on every oil resource (current and prospective) worldwide.

OPEM Data

OPEM data inputs entail a detailed product yield (in volume or mass) per barrel of processed oil and gas. Product volumes are provided by OPGEE and PRELIM (as discussed earlier), and emission factors for shipping and end use are obtained from US government sources.48 Default assumptions are entered for distances traveled to end users.49

The main OPEM uncertainties involve uncertainties about product output derived from PRELIM, combustion emission factors in locations beyond the United States, and transport specifications for marketing petroleum products.

All models as well as measurement systems always contain a degree of uncertainty. This does not necessarily diminish their value in decision-making. Instead, in an effort to reduce uncertainty, models are refined and new versions are developed. This is the case with OPGEE, PRELIM, and OPEM, which continue to be updated to better reflect the engineering systems they model using newly gathered data. Opportunities to fine-tune assumptions and reduce uncertainty also arise in response to publishing OCI+ estimates. For example, operators reach out to provide better data, scientists conduct comparative analyses with their satellite measurements, governments collect additional data, and civil society actors focus their attention on higher emitting assets. In other words, despite their uncertainty, OCI+ estimates promote ongoing efforts to zero in on GHG emissions.

The Role of Remote Sensing

The data limitations in the OCI+ improve as more open-source data is gathered and published. Since operators are unlikely to be the guaranteed source of these data, especially if policymakers do not require data transparency, other sources are needed. Here, government scientists can fill the void with a growing stream of remotely sensed data from a variety of instruments that can observe industry practices without their involvement.

In November 2019, I was invited to present the OCI+ to a large group of NASA scientists at their workshop on CMS applications.50 For three days, scientists interacted with researchers, like me, who are using their data on GHGs. Their reaction to the OCI+ confirmed for me that the model has added value in the panoply of GHG-estimating tools. Top-down remote sensing, bottom-up reporting, regional inventories and assessments, and lifecycle GHG models like the OCI+ are part of a multipronged mitigation approach. Taken together, these methods mutually reinforce accuracy and knowledge on climate risks from the oil and gas sector.51

Between satellites, planes, drones, and ground monitors, GHG remote sensing is on the rise worldwide. Table 3.3 offers a sampling of past, present, and future remote-sensing missions to measure and attribute methane and other GHG emissions.

Table 3.3 Future, Present, and Past GHG-Measuring Satellites and Aircraft (as of 2021)

Institution Satellite Name

Country of Origin

Type (# of) Instruments

Spatial Resolution (Pixel Area)

GHGs Detected

Revisit Rate

Launch Dates

Public-Private-NGO Partnership

Carbon Mapper

US

Satellite ecosystem (2+)

30 m x 30 m

Methane, CO2

Daily to weekly

2023

Non-Governmental (NGO)

GeoCARB

US

Satellite

5 to 10 km

Methane, CO, CO2

Daily

2022

MethaneSat

US

Satellite

400 m x 10 0m

Methane, others?

7 days or less

2022

Commercial

Bluefield

US

Satellites, aircraft

20 m x 20 m

Methane

As ordered

2019, ongoing

GHGSat

Canada

Satellites (>3), aircraft

50 m x 50 m, 25 m x 25 m

Methane, CO2

As ordered

2015, 2019, 2020, ongoing

DigitalGlobe

US

Satellite

3.7, 30 m?

Methane, petcoke

As ordered

2014, ongoing

Government

MERLIN

Germany /France

Satellite

120 m

Methane

28 days

2025

CO2M (Sentinel 7)

EU

Satellite

2 km x 2 km

CO2, NO2, Methane

Global coverage in 5 days

2025

GOSAT-GW

Japan

Satellite

1-3 km x 10 km

CO2, methane, ozone, NO2

3 days

2023

Feng Yun 3G

China

Satellite

<3 km

CO2, Methane, NO2, CO

n/a

2023

TanSat-2

China

Satellites (constellation, 6)

< 2 km x 2 km

CO2, CH4, CO

3-5 days

2022

MetOp

EU

Satellite

7.5 km x 7.5 km

Methane, CO2, NO2

Daily, 2 days

2021

Microcarb

EU

Satellite

2 km x 2 km

CO2

21 days

2021

EnMAP

Germany

Satellite

30 m x 30 m

CO2, methane, ozone

27 days

2021

TROPOMI

Netherlands

Satellite

7 km x 7 km, 5.5 km x 3.5 km

Methane, ozone, NO2, ammonia

Daily

2018

GOSAT-2

Japan

Satellite

9.7 km

CO2, methane, ozone, NO2

3 days

2018

Gao-Fen 5

China

Satellite

10.3 km

CO2, methane

Daily

2018

Feng Yun 3D/F

China

Satellite

10 km

CO2, Methane, NO2, CO

Monthly

2017, 2023

TanSat

China

Satellite

1 km x 2 km, 2 km x 2 km

CO2

16 days

2016

OCO-3/OCO-2

US

Satellites (2)

2.25 km x 1.29 km

CO2

16 days

2014, 2019

AVIRIS-NG

US

Aircraft

4 m - 20 m

Methane

As ordered

2012, ongoing

VIIRS

US

Satellite

375 m, 750 m, resampled to 500 m and 1 km

Gas flares, fires, lights

4 days

2011

GOSAT

Japan

Satellites (3)

10.5 km

CO2, methane, ozone, NO2

16 days

2009

TES-Aura

US

Satellite

0.53 km x 0.53 km

CO2, methane, ozone, NO2, N2O

16 days

2004

SciSat-1

Canada

Satellite

>500 m

CO2, methane, N2O

Annual

2003

AIRS-Aqua

US

Satellite

13.5 km

CO2, methane, ozone

16 days

2002

EnviSat

EU

Satellite

30 km x 60 km

Methane, CO2

35 days

2002

Hyperion

US

Satellite

30 m

Methane

16 days

2000

Source: Carbon Mapper, Personal Communication, June 19, 2021.

Notes: m=meter; km=kilometer.

A remote-sensing network, in theory, can pick up any emission signal anywhere on the planet at any time. In practice

e, these sensors are best suited to detecting regional hot spots, quantifying emissions rates, and guiding ground-based follow-up studies.52 These satellite-based sensors are limited by the instruments onboard, their positioning, cloud cover, signal strength over natural background emissions in the atmosphere, array of sources clustered together, and an adequate inventory of existing emission sources. Current techniques rely on complex atmospheric inversion models to decipher and estimate sources of transport-prone, rapidly dispersed emissions amid background GHG concentrations (the natural and manmade emissions that are present).53 Future detection methods are under development, including some designed to use methane isotopes,54 ethane,55 and other tracer gases.56

Private firms are acquiring and analyzing raw government satellite data, and this can introduce errors because raw data must first be carefully cleaned up to remove data gaps, remove outlying data, eliminate mistaken observations over the earth’s reflective surface, and use extreme care in complex atmospheric computations. For example, in May 2020, the European Space Agency released a global map with elevated methane readings compared to the year prior, and a private data firm set about analyzing and selling the data.57 When I asked a renowned expert about the validity of the private firm’s data, he said that companies advertise methane satellite data, but buyers have to take care because firms’ estimates can be off by a factor of three or more.58 In response, data providers are improving their analytics and scientists are automating methods that speed up satellite data analysis with a user-friendly inversion tool for cleaned-up satellite data.

The OCI+ currently uses satellite data and other remote-sensing data as model inputs, including from a satellite for flaring data shared by NASA and the National Oceanic and Atmospheric Administration (NOAA) and another satellite called the Tropospheric Monitoring Instrument (TROPOMI) that maps methane hot spots. The NASA-NOAA satellite known as the Visible Infrared Imaging Radiometer Suite (VIIRS) was originally launched to observe light sources on Earth like city lights and forest fires. In 2015, scientists discovered that VIIRS also detects the radiant emissions from gas flares used by petroleum systems.59 The TROPOMI satellite was developed jointly by the Dutch and the European Space Agency. In addition to using satellite data as inputs, the OCI+ is being used to improve remote sensing by improving GHG inventories of existing oil and gas emissions sources that scientists require in their inversion models. And the OCI+ is also being used to affirm satellite readings during oil and gas events like blowouts.60 Thus, not only is remote sensing being used to swap out new remote-sensing data for model inputs that were previously filled in using defaults, but also the OCI+ is being used to calculate and verify the remote-sensing data.

Spotting Flares from Space

Gas flaring has a unique heat signature that can be detected via the VIIRS satellite.61 Nearly 7,500 individual flare sites were detected worldwide in 2012, burning an estimated 143 billion cubic meters of gas (nearly 4 percent of global gas production).62 Despite efforts to curtain this wasteful practice (discussed in chapter 7), in both 2018 and 2019, global gas flaring was up 3 percent a year.63 Flares in the United States, Russia, Iraq, and Iran account for the greatest volumes of natural gas being burned worldwide.64 Increased fracking (in the United States) and political unrest in regions around the world may be the impetus of increased flaring when normal operations and shipping arrangements are upended. Flaring can indicate a state in crisis or an economic decision to prioritize more profitable oil over less profitable gas production.

The data obtained by VIIRS provides the flaring-to-oil ratio input that OPGEE uses. Overlaying VIIRS data on a map of OCI+-modeled GHG emissions creates a visualization of the climate risks generated by flaring.65 NASA now conducts an ongoing monitoring, reporting, and verification (MRV) system of global flaring sites through CMS using VIIRS to remotely sense time-series data.66 Work is underway to arrange for these data to offer site-specific tracking of flares, including estimates for the volume of gas burned and the ability to distinguish routine versus nonroutine flaring.67

Mapping Methane Hot Spots

The OCI+ also benefits from data generated by space-based sensors that can be used to measure methane hot spots. A few years ago, I received a call from a gas company that had used an early version of the OCI+ to show how much lower their GHG emissions were than their competitors’. (While I was pleased that they found the model useful, I voiced my concern because they omitted methane in the OCI+’s application to rank their own and others’ operations—a big measurement oversight.). In our meeting, I was told that the company’s competitors were covertly releasing methane. After BP’s Deepwater Horizon oil spill in 2010, regulations were adopted limiting the amount of gas that companies could flare and vent from US offshore platforms.68

But detecting violations of this law proved challenging. A Coast Guard boat equipped with Light Detection and Ranging (LiDAR) equipment, remote sensing that uses infrared lasers to detect methane, would make a single daily pass to record visible flaring and invisible methane from each platform.69 But some operators wanted to discard more gas than permitted. To avoid a citation, as soon as the coast was clear, they turned off the flare’s pilot ignition and expelled invisible gas into the air, effectively avoiding detection. Breaking this law saved companies real money. Not only could they produce more oil but also they avoided paying royalties on the gas they emitted.

In part to remedy this problem, instead of relying on one-off LiDAR passes, regulators now use satellites to map methane hot spots, while the OCI+ is mapping these hot spots alongside global oil and gas assets.70 The TROPOMI satellite began reporting in 2018.71 For example, TROPOMI reported leakage from a gas well blowout in Ohio in early 2018, and together with a tracer transport simulation, scientists quantified the emission rate and total methane release from the accident.72 But TROPOMI’s coverage is course. Like a wide-angle camera lens, it cannot zoom in and pinpoint equipment leaks. Other satellites in this emerging remote sensing ecosystem, like Carbon Mapper, have “portrait” lens capabilities with higher resolution that can measure individual leaks.73

Historically, problems existed quantifying methane located over water. The ocean absorbs sunlight, preventing satellites from detecting the light reflecting off the earth’s surface. Scientists have developed new “glint mode” observations—satellites equipped with sensors that point to bright spots over oceans where solar radiation is directly reflected off the earth’s surface.74 This means that one-third of current oil and gas production, most refineries, and all shipping conducted offshore and near coastlines are now detectable.

Tracking GHGs from Above

In addition to satellites, low-flying, high-resolution, access-limited instruments can be used to verify oil and gas emissions. This localized approach can help reconcile top-down with bottom-up GHG estimations.75 Using aircraft to fly over target sites and drones to collect data can police super emitters in targeted regions.76 However, time and resources limit researchers’ abilities to locally examine widespread areas. Regional conflicts, political tensions, and other reasons may also prevent flyovers in regions that contribute most heavily to GHG emissions. Still, pairing flyover data with satellite data, emissions inventories, and OCI+-type modeling offers a fairly reliable way to see the big picture.

Top-down sensors, bottom-up reported data, and GHG-estimating models can work in tandem to assess oil and gas emissions. It will take a network of different measurement, reporting, and computational tools, each with their different capabilities, to quantify and attribute GHG emissions to different sources. None of these methods are foolproof and each has associated uncertainties. Taken together, however, various emission detection and assessment methods can be mutually reinforcing. The OCI+ model not only employs reported and remote sensing data in its calculations but also can be used to fill data gaps, provide intel in areas where data transparency is lacking, and run scenarios (where data does not yet exist) to identify future sensor and monitoring opportunities that may arise.

Jointly Modeling GHG Emissions and Other Air Pollutants

The OCI+ was originally developed to account for the lifecycle emissions of GHGs. But many GHGs simultaneously pollute the air and pose serious health and social justice risks, especially for low-income individuals. These copollutants increase mortality rates, raise health care costs, damage crops, lower agricultural yields, erode infrastructure, and impair visibility. Likewise, many air pollutants contribute to climate change. These two manmade atmospheric concerns—air pollution and climate change—are inextricably linked and exacerbate one another.77

Reducing air pollution can also protect the climate, just as reducing GHG emissions can improve air quality. For example, particulate matter (PM) from the partial burning of diesel fuel in engines is a known carcinogen that sickens and kills people.78 PM (in the form of black carbon79) is also a powerful GHG that circulates in the atmosphere, ending up in the North and South Poles, landing on ice and snow, darkening surfaces, and causing less sunlight to be reflected back into space, thus contributing to global warming. As darkened snow and ice melt, such warming increases. Wetter, warmer weather in turn spurs plant growth and casts shadows that further darken the earth’s surface, leading to additional warming. Reducing PM levels, then, lessens morbidity and mortality from air pollution as well as GHG emissions.

The oil-fueled petroleum sector reportedly contributes the most to net global warming effects over a twenty-year timeframe, due to high emissions levels of CO2, methane, ozone-forming air pollutants (like VOCs and nitrogen oxides), and black carbon (PM).80 Compared to other economic sectors, like power generation, oil and gas have lower emissions of pollutants that cool the atmosphere in the short term, including sulfates, aerosols, and organic carbon. In other words, oil and gas have the largest effects on both air quality and climate change in the short term.

Comanaging climate change and air quality is a high priority, especially in Asia where regional pollution is driving the policy agenda more so than climate change. The 2008 Olympics in Beijing spotlighted the city’s intense air pollution, and the government adopted Los Angeles’s 1984 strategy to simply halt driving, leading to less smog and blue skies during the games. But the pollution returned thereafter.81 Today, China’s cities suffer some of the world’s worst air pollution, which compete closely with India’s metropolitan centers.82 Elsewhere, the Middle East (particularly Saudi Arabia and Kuwait), with their major oil and gas operations, have major air pollution problems. Developing a unified front to jointly reduce air pollutant and GHG levels could be more effective than focusing on climate change alone.

Several air pollutants are already included in one or more of the OCI+’s underlying models. Once a pollutant is computed in all of the three models, the OCI+ can model out corresponding estimates (see Table 3.4).

Table 3.4 Air Pollutants in OCI+ Models (as of 2020)

Air Pollutant Emitted

Oil and Gas Sources

OPGEE

PRELIM

OPEM

Carbon Dioxide

Oil and gas combustion and system leakage

Y

Y

Y

Methane (CH4)

Gas leakage and oil and gas combustion

Y

Y

Y

All VOCsa

Condensates, gas, light oil, and petrochemicals

Y

Y

N

Nitrous Oxide (N2O)

All oil and gas combustion

Y

Y

Y

Nitrogen Oxides (NOx)

All oil and gas combustion

N

Y

N

Ozone (O3)c

Photochemical reaction of VOCs and NOX, with methane as intermediary

N

Y

N

Carbon Monoxide (CO)

Incomplete oil and gas combustion

Y

Y

N

PMb

Extra-heavy oil, fuel oil, diesel, and petcoke

N

Y

N

Sulfur Dioxide (SO2)

High-sulfur oil and gas

Y

Y

N

Hazardous Air Pollutants (HAPs)d

Impurities in oil and gas

N

Y

N

Ozone Depletion (CFC-11)

Refrigerants in oil and gas processing and insulation in oil and gas systems

N

Y

N

a VOCs include ethane, propane, butane, and a long list of other hydrocarbons contained in oil and gas.

b Includes PM2.5 and PM10.

c Photochemical ozone-forming potential is considered since ozone is not directly emitted and is formed via photochemical reactions of VOCs and NOx.

d Includes several heavy metals, aromatics, and other chemical compounds that are known and suspected carcinogens and ecotoxins that pollute air and water (also known as air toxins).

OCI+, Oil Climate Index plus Gas; OPGEE, Oil Production Greenhouse Gas Emissions Estimator; OPEM, Oil Products Emissions Module; PRELIM, Petroleum Refinery Lifecycle Inventory Model.

Sources: Author’s inputs based on documentation from OPGEE, PRELIM, and OPEM models. For more information, see Congressional Research Service, “Methane and Other Air Pollution Issues in Natural Gas Systems,” September 17, 2020, https://fas.org/sgp/crs/misc/R42986.pdf

Massive Mismatch in Climate Risks

The OCI+ highlights three central facts. First, massive tonnes of CO2e emissions will be emitted from consuming oil and gas in the decades ahead, much more than the atmosphere can safely accommodate, as shown in Figure 2.2. Second, different barrels of oil and cubic feet of gas have widely varying GHG emissions, and it is the industry’s responsibility to reduce the GHG emissions intensity from the various hydrocarbons they produce, process, refine, and ship. Third, petroleum’s varying climate impacts are not currently recognized or priced into the market value of competing crudes, natural gas, or their products. As such, we need to both consume less oil and gas overall and reduce the emissions intensity of each barrel and cubic foot we continue to use.

Analysis of thousands of oil and gas resources modeled to date reveals that their emission differences are far greater than currently acknowledged. Large emissions ranges exist whether values are calculated per BOE, per megajoule of products, or per dollar value of products, and these emissions ranges are expected to grow as new, unconventional oils are identified.83

Large variations in GHG emissions exist upstream, where the oil with the highest emissions intensity has approximately fifteen times the emissions of the lowest-intensity oil.84 These differences are also evident in upstream gas production and processing emissions, which vary by a significant order of magnitude.85 Midstream refining emissions exhibit large variations too, as the oil with the highest emissions intensity produces approximately fourteen times the emissions of the lowest-intensity oil.86 The differences stemming from the downstream shipping of GHGs are estimated at a factor of two for oil and compressed gas and significantly higher for LNG.87

These supply-side climate impacts are the responsibility of the oil and gas industry.88 In the World Energy Outlook 2018, the International Energy Agency (IEA) used the OCI+ and its underlying models to estimate supply-side emissions intensities for all global oil and gas resources. Figure 3.7 illustrates the broad range of climate risks posed by the oil and gas industry.

image

FIGURE 3.7 Estimated Ranges of Currently Modeled Emissions Intensities of Global Oil and Gas Supplies

Note: Estimates assume 100-year GWP. Values increase substantially using 20-year GWPs. See OCI+ Web Tool for details.

Sources: International Energy Agency, World Energy Outlook 2018, Table 11.1, Figure 11.6 (for global average), https://www.iea.org/reports/world-energy-outlook-2018/oil-and-gas-innovation; OCI+ Preview BETA Web Tool, (for lowest- and highest-emission-intensity oil and gas). BOE, barrel of oil equivalent; CO2e, carbon dioxide equivalent; GHG, greenhouse gas.

Although it is outside the direct control of the oil and gas industry, the OCI+ estimates a significant variation in the end-use (Scope 3) GHG emissions of various oils, gases, and related products. The oil with the highest emissions intensity from end-use consumption has approximately 65 percent more emissions than the lowest-intensity oil does, and for gas the corresponding variation is over 100 percent.89

From a Theoretical Model to Real-World Impact

The OCI+ traces the overall GHG footprints for different oils and gases, which are large and could get bigger over time as unconventional oil and gas development grows. Several companies have engaged in OCI+ development and have helped improve the underlying models, including Chevron and Exxon. Other companies have put the OCI+ into action. For example, Saudi Aramco, Norway’s Equinor, and Texas-based Southwestern have used the OCI+ models for strategic planning. While it is unclear to what extent individual companies have developed their own models, global energy consultancies—such as Rystad, Baker Hughes, and Wood Mackenzie—are using the OCI+ to advise their clients in the oil and gas industry.

OCI+ applications by industry, government, and civil society actors affirm the need to better assess and differentiate oil and gas GHG emissions. The most climate-intensive oils and gases, covered in the next chapter, require special attention from operators and investors, government policymakers, and civil society alike. Assessing the climate footprints of oil and gas resources is essential to finding a way to balance the enormous economic value that petroleum delivers with the equally massive threats these hydrocarbons pose to the global climate.

If you find an error or have any questions, please email us at admin@erenow.org. Thank you!