OCT Recovery Plan FAQs, by the folks who brought you Deep Decarbonization Pathways

350 PPM PATHWAYS

FOR THE UNITED STATES May 8, 2019

DEEP DECARBONIZATION PATHWAYS PROJECT

350 PPM Pathways for the United States U.S. Deep Decarbonization Pathways Project

Prepared by

Ben Haley, Ryan Jones, Gabe Kwok, Jeremy Hargreaves & Jamil Farbes

Evolved Energy Research

James H. Williams

University of San Francisco
Sustainable Development Solutions Network

May 8, 2019

Version 1

© 2019 by Evolved Energy Research 2

Table of Contents

Table of Contents……………………………………………………………………………………………………………. 3 List of Terms…………………………………………………………………………………………………………………… 4 Executive Summary…………………………………………………………………………………………………………. 6

1. 2.

3.

Introduction …………………………………………………………………………………………………………… 17 Study Design ………………………………………………………………………………………………………….. 23

1. 2.1.  Scenarios ………………………………………………………………………………………………………… 23
2. 2.2.  Modeling Methods and Data Sources…………………………………………………………………. 25
   1. 2.2.1.  EnergyPATHWAYS ………………………………………………………………………………………………… 25
   2. 2.2.2.  Regional Investment and Operations (RIO) Platform ………………………………………………… 27
   3. 2.2.3.  Key References and Data Sources…………………………………………………………………………… 27

Results…………………………………………………………………………………………………………………… 29

1. 3.1.  Emissions ………………………………………………………………………………………………………… 29
2. 3.2.  System Costs……………………………………………………………………………………………………. 32
3. 3.3.  Energy Transition …………………………………………………………………………………………….. 35
4. 3.4.  Infrastructure ………………………………………………………………………………………………….. 38
   1. 3.4.1.  Demand-Side Transformation………………………………………………………………………………… 38
   2. 3.4.2.  Low-Carbon Generation………………………………………………………………………………………… 40
   3. 3.4.3.  Biofuels Production ………………………………………………………………………………………………. 41
   4. 3.4.4.  Electricity Storage ………………………………………………………………………………………………… 44
   5. 3.4.5.  Electricity Transmission ………………………………………………………………………………………… 45
   6. 3.4.6.  Hydrogen Electrolysis……………………………………………………………………………………………. 48
   7. 3.4.7.  Direct Air Capture ………………………………………………………………………………………………… 49

Discussion ……………………………………………………………………………………………………………… 52

1. 4.1.  Four Pillars ………………………………………………………………………………………………………. 52
2. 4.2.  Regional Focus…………………………………………………………………………………………………. 53
3. 4.3.  Electricity Balancing …………………………………………………………………………………………. 55
4. 4.4.  Sector Integration ……………………………………………………………………………………………. 60
5. 4.5.  Circular Carbon Economy ………………………………………………………………………………….. 62

Conclusions …………………………………………………………………………………………………………… 63

5.1. Key Actions by Decade ……………………………………………………………………………………… 64 2020s……………………………………………………………………………………………………………………….. 65 2030s……………………………………………………………………………………………………………………….. 67 2040s……………………………………………………………………………………………………………………….. 68

Bibliography …………………………………………………………………………………………………………………. 69 Technical Supplement……………………………………………………………………………………………………. 73 Appendix ……………………………………………………………………………………………………………………… 81

© 2019 by Evolved Energy Research 3

List of Terms

1.0oC – One degree Celsius (1.8oF) of global warming over pre-industrial temperatures. 1.5oC – One-and one-half degrees Celsius (2.7oF) of global warming over pre-industrial temperatures, an aspirational goal in the Paris Agreement climate accord.

2oC – Two degrees Celsius (3.6oF) of global warming over pre-industrial temperatures. The Paris Agreement states the intention of parties to remain “well under” this upper limit.

350 ppm – An atmospheric CO2 concentration of 350 parts per million by volume
80 x 50 – A target for reducing CO2 emissions used in U.S. states and in other countries, referring to an 80% reduction below 1990 levels by 2050.

AEO – The Annual Energy Outlook, a set of modeled results released annually by the U.S. government that forecasts the energy system under current policy for the next three decades. AZNM – eGRID region comprising most of Arizona and New Mexico
Base Case – The primary deep decarbonization pathway with all technologies and resources available according to best scientific estimates.

Baseline – A scenario derived from the U.S. Department of Energy’s Annual Energy Outlook projecting the future evolution of the energy system given current policies

BECCS – Bioenergy with carbon capture and geologic sequestration
BECCU – Bioenergy with carbon capture and utilization of that carbon somewhere in the economy
Bioenergy – Primary energy derived from growing biomass or use of organic wastes
CAMX – eGRID region comprising most of California
CCE – Circular carbon economy, a term that refers to the capture and reuse of CO2 within the energy system
CCS – Carbon capture and storage (also called carbon capture and sequestration)
CCU – Carbon capture and utilization (for economic purposes)
CO2 – Carbon dioxide, the primary greenhouse gas responsible for human caused warming of the climate
DAC – Direct air capture, a technology that captures CO2 from ambient atmosphere
DDPP – Deep Decarbonization Pathways Project
DOE – U.S. Department of Energy
EER – Evolved Energy Research, LLC.
eGRID – Emissions & Generation Resource Integrated Database maintained by the Environmental Protection Agency. eGRID divides the country into regions used in this study that are relevant for electricity planning and operations
EnergyPATHWAYS – An open-source, bottom-up energy and carbon planning tool for use in evaluating long-term, economy-wide greenhouse gas mitigation scenarios.
EPA – U.S. Environmental Protection Agency
ERCOT – Electricity interconnection and balancing authority comprising most of Texas

Gt(C) – Gigatons (billions of metric tons) of carbon

GW – Gigawatt (billion watts)
GWh – Gigawatt hour (equivalent to one million kilowatt hours)
IAM – Integrated Assessment Model, a class of model that models the energy system, economy, and climate system, to incorporate feedback between the three.
Intertie – Electric transmission lines that connect different regions
IPCC – Intergovernmental Panel on Climate Change, an international organization mandated to provide policy makers with an objective assessment of the scientific and technical information available about climate change.
Land NET – Negative CO2 emissions as the result of the update of carbon in soils and terrestrial biomass
Low Biomass – A scenario that limits the use of biomass for energy
Low Electrification – A scenario with a slower rate of switching from fuel combustion technologies to electric technologies on the demand-side of the energy system
Low Land NETs – A scenario with a lower uptake of carbon in land sinks, resulting in a more restricted emissions budget for the energy system.
MMT – Million metric tonnes
N-1 – A test to determine the reliability of a system by ensuring any single component of the system can fail without jeopardizing the system as a whole
NET – Negative emissions technology, one that absorbs atmospheric CO2 and sequesters it Net-zero – A condition in which human-caused carbon emissions equal the natural uptake of carbon in land, soils, and oceans such that atmospheric CO2 concentrations remain constant. No New Nuclear – A scenario that disallows new nuclear construction
No Tech NETs – A scenario that disallows use of the specific technologies of biomass with carbon capture and geologic sequestration and direct air capture with geologic sequestration NWPP – Northwest power pool
Pg(C) – Peta (1015) grams
ppm – parts per million
ReEDS – Renewable Energy Deployment System – a capacity planning and dispatch model build by the National Renewable Energy Laboratory
RFC – Three separate eGRID regions in the mid-Atlantic and extending west through Michigan RIO – Regional Investment and Operations Platform, an optimization tool built by Evolved Energy Research to explore electricity systems and fuels
SDSN – Sustainable Development Solutions Network
SR – eGRID region composing all of the Southeastern United States outside of Florida
TBtu – Trillion British thermal units, an energy unit typically applied to in power generation natural gas
Tech NET – Negative emission technologies composed of either biomass with carbon capture and sequestration or direct air capture with sequestration.
TX – Transmission
VMT – Vehicle miles traveled
WECC – Western electricity coordinating council

Executive SummaryThis report describes the changes in the U.S. energy system required to reduce carbon dioxide (CO2) emissions to a level consistent with returning atmospheric concentrations to 350 parts per million (350 ppm) in 2100, achieving net negative CO2 emissions by mid-century, and limiting end-of-century global warming to 1°C above pre-industrial levels. The main finding is that 350 ppm pathways that meet all current and forecast U.S. energy needs are technically feasible using existing technology, and that multiple alternative pathways can meet these objectives in the case of limits on some key decarbonization strategies. These pathways are economically viable, with a net increase in the cost of supplying and using energy equivalent to about 2% of GDP, up to a maximum of 3% of GDP, relative to the cost of a business-as-usual baseline. These figures are for energy costs only and do not count the economic benefits of avoided climate change and other energy-related environmental and public health impacts, which have been described elsewhere.1

This study builds on previous work, Pathways to Deep Decarbonization in the United States (2014) and Policy Implications of Deep Decarbonization in the United States (2015), which examined the requirements for reducing GHG emissions by 80% below 1990 levels by 2050 (“80 x 50”).2 These studies found that an 80% reduction by mid-century is technically feasible and economically affordable, and attainable using different technological approaches. The main requirement of the transition is the construction of a low carbon infrastructure characterized by high energy efficiency, low-carbon electricity, and replacement of fossil fuel combustion with decarbonized electricity and other fuels, along with the policies needed to achieve this transformation. The findings of the present study are similar but reflect both a more stringent emissions limit and the consequences of five intervening years without aggressive emissions reductions in the U.S. or globally.

1 See e.g. Risky Business: The Bottom Line on Climate Change, available at https://riskybusiness.org/ 2 Available at http://usddpp.org/.

The 80 x 50 analysis was developed in concert with similar studies for other high-emitting countries by the country research teams of the Deep Decarbonization Pathways Project, with an agreed objective of limiting global warming to 2°C above pre-industrial levels.3 However, new studies of climate change have led to a growing consensus that even a 2°C increase may be too high to avoid dangerous impacts. Some scientists assert that staying well below 1.5°C, with a return to 1°C or less by the end of the century, will be necessary to avoid irreversible feedbacks to the climate system.4 A recent report by the IPCC indicates that keeping warming below 1.5°C will likely require reaching net-zero emissions of CO2 globally by mid-century or earlier.5 A number of jurisdictions around the world have accordingly announced more aggressive emissions targets, for example California’s recent executive order calling for the state to achieve carbon neutrality by 2045 and net negative emissions thereafter.6

In this study we have modeled the pathways – the sequence of technology and infrastructure changes – consistent with net negative CO2 emissions before mid-century and with keeping peak warming below 1.5°C. We model these pathways for the U.S. for each year from 2020 to 2050, following a global emissions trajectory that would return atmospheric CO2 to 350 ppm by 2100, causing warming to peak well below 1.5°C and not exceed 1.0°C by century’s end.7 The cases modeled are a 6% per year and a 12% per year reduction in net fossil fuel CO2 emissions after 2020. These equate to a cumulative emissions limit for the U.S. during the 2020 to 2050 period of 74 billion metric tons of CO2 in the 6% case and 47 billion metric tons in the 12% case. (For comparison, current U.S. CO2 emissions are about 5 billion metric tons per year.) The emissions in both cases must be accompanied by increased extraction of CO2 from the atmosphere using land-based negative emissions technologies (“land NETs”), such as reforestation, with greater extraction required in the 6% case.

3 Available at http://deepdecarbonization.org/countries/.
4 James Hansen, et al. (2017) “Young people’s burden: requirement of negative CO2 emissions,” Earth System Dynamics, https://www.earth-syst-dynam.net/8/577/2017/esd-8-577-2017.html.
5 Available at https://www.ipcc.ch/sr15/.
6 Available at https://www.gov.ca.gov/wp-content/uploads/2018/09/9.10.18-Executive-Order.pdf.
7 Hansen et al. (2017).

p. 7

Figure ES1 Global surface temperature and CO2 emissions trajectories. Hansen et al, 2017.

We studied six different scenarios: five that follow the 6% per year reduction path and one that follows the 12% path. All reach net negative CO2 by mid-century while providing the same energy services for daily life and industrial production as the Annual Energy Outlook (AEO), the Department of Energy’s long-term forecast. The scenarios explore the effects of limits on key decarbonization strategies: bioenergy, nuclear power, electrification, land NETs, and technological negative emissions technologies (“tech NETs”), such as carbon capture and storage (CCS) and direct air capture (DAC).

Table ES1. Scenarios developed in this study

p. 8

The scenarios were modeled using two new analysis tools developed for this purpose, EnergyPATHWAYS and RIO. As extensively described in the Appendix, these are sophisticated models with a high level of sectoral, temporal, and geographic detail, which ensure that the scenarios account for such things as the inertia of infrastructure stocks and the hour-to-hour dynamics of the electricity system, separately in each of fourteen electric grid regions of the U.S. The changes in energy mix, emissions, and costs for the six scenarios were calculated relative to a high-carbon baseline also drawn from the AEO.

Relative to 80 x 50 trajectories, a 350 ppm trajectory that achieves net negative CO2 by mid- century requires more rapid decarbonization of energy plus more rapid removal of CO2 from the atmosphere. For this analysis, an enhanced land sink 50% larger than the current annual sink of approximately 700 million metric tons was assumed.8 This would require additional sequestration of 25-30 billion metric tons of CO2 from 2020 to 2100. The present study does not address the cost or technical feasibility of this assumption but stipulates it as a plausible value for calculating an overall CO2 budget, based on consideration of the scientific literature in this area.9

8 U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2016, available at https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-2016
9 Griscom, Bronson W., et al. (2017) “Natural climate solutions.” Proceedings of the National Academy of Sciences 114.44 (2017): 11645-11650; Fargione, Joseph E., et al. (2018) “Natural climate solutions for the United States.” Science Advances 4.11: eaat1869.

p. 9

Figure ES2 Four pillars of deep decarbonization – Base case

Energy decarbonization rests on the four principal strategies (“four pillars”) shown in Figure ES2: (1) electricity decarbonization, the reduction in emissions intensity of electricity generation by about 90% below today’s level by 2050; (2) energy efficiency, the reduction in energy required to provide energy services such as heating and transportation, by about 60% below today’s level; (3) electrification, converting end-uses like transportation and heating from fossils fuels to low-carbon electricity, so that electricity triples its share from 20% of current end uses to 60% in 2050; and (4) carbon capture, the capture of otherwise CO2 that would otherwise be emitted from power plants and industrial facilities, plus direct air capture, rising from nearly zero today to as much as 800 million metric tons in 2050 in some scenarios. The captured carbon may be sequestered or may be utilized in making synthetic renewable fuels.

Achieving this transformation by mid-century requires an aggressive deployment of low-carbon technologies. Key actions include retiring all existing coal power generation, approximately doubling electricity generation primarily with solar and wind power (why just doubling?) and electrifying virtually all passenger vehicles and natural gas uses in buildings. It also includes creating new types of infrastructure, namely large-scale industrial facilities for carbon capture and storage, direct air capture of CO2, the production of gaseous and liquid biofuels with zero net lifecycle CO2, and the production of hydrogen from water electrolysis using excess renewable electricity. The scale of the infrastructure buildout by region is indicated in Figure ES3.

Figure ES3 Regional infrastructure requirements (Low Land NETS scenario)

p. 11

Figure ES4 shows that all scenarios achieve the steep reductions in net fossil fuel CO2 emissions to reach net negative emissions by the 2040s, given a 50% increase in the land sink, including five that are limited in one key area. This indicates that the feasibility of reaching the emissions goals is robust due to the ability to substitute strategies. At same time, the more limited scenarios are, the more difficult and/or costly they are relative to the base case with all options available. Severe limits in two or more areas were not studied here (so is all WWS and storage considered a limited use case?) but would make the emissions goals more difficult to achieve in the mid-century time frame. (is this true for WWS though?)

Figure ES4 2020-2050 CO2 emissions for the scenarios in this study

Figure ES5 shows U.S. energy system costs as a share of GDP for the baseline case and six 350 ppm scenarios in comparison to historical energy system costs. While the 350 ppm scenarios have a net cost of 2-3% of GDP more than the business as usual baseline, these costs are not out of line with historical energy costs in the U.S. The highest cost case is the Low Land NETs scenario, which requires a 12% per year reduction in net fossil fuel CO2 emissions. 

p. 12

By comparison, the 6% per year reduction cases are more closely clustered. The lowest increase is the Base scenario, which incorporates all the key decarbonization strategies. These costs do not include any potential economic benefits of avoided climate change or pollution, which could equal or exceed the net costs shown here.

Figure ES5. Total energy system costs as percentage of GDP, modeled (R.) and historical (L.)

A key finding of this study is the potentially important future role of “the circular carbon economy.” This refers to the economic complementarity of hydrogen production, direct air capture of CO2, and fuel synthesis, in combination with an electricity system with very high levels of intermittent renewable generation. If these facilities operate flexibly to take advantage of periods of excess generation, the production of hydrogen and CO2 feedstocks can provide an economic use for otherwise curtailed energy that is difficult to utilize with electric energy storage technologies of limited duration. 

p. 13

These hydrogen and CO2 feedstocks can be combined as alternatives for gaseous and liquid fuel end-uses that are difficult to electrify directly like freight applications and air travel. While the CO2 is eventually emitted to the atmosphere, the overall process is carbon neutral as it was extracted from the air and not emitted from fossil reserves. A related finding of this work is that bioenergy with carbon capture and storage (BECCS) for power plants appears uneconomic, while BECCS for bio-refineries appears highly economic and can be used as an alternative source of CO2 feedstocks in a low-carbon economy.

There are several areas outside the scope of this study that are important to provide a full picture of a low greenhouse gas transition. One important area is better understanding of the potential and cost of land-based NETs, both globally and in the U.S. Another is the potential and cost of reductions in non-CO2 climate pollutants such as methane, nitrous oxide, and black carbon. Finally, there is the question of the prospects for significant reductions in energy service demand, due to lifestyle choices such as bicycling over cars, structural changes such as increased transit and use of ride-sharing, or the development of less-energy intensive industry, perhaps based on new types of materials.

“Key Actions by Decade” below provides a blueprint for the physical transformation of the energy system. From a policy perspective, this provides a list of the things that policy needs to accomplish, for example the deployment of large amounts of low carbon generation, rapid electrification of vehicles, buildings, and industry, and building extensive carbon capture, biofuel, hydrogen, and synthetic fuel synthesis capacity.

Some of the policy challenges that must be managed include: land use tradeoffs related to carbon storage in ecosystems and siting of low carbon generation and transmission; electricity market designs that maintain natural gas generation capacity for reliability while running it very infrequently; electricity market designs that reward demand side flexibility in high-renewables electricity system and encourage the development of complementary carbon capture and fuel synthesis industries; coordination of planning and policy across sectors that previously had little interaction but will require much more in a low carbon future, such as transportation and electricity; coordination of planning and policy across jurisdictions, both vertically from local to state to federal levels, and horizontally across neighbors and trading partners at the same level; mobilizing investment for a rapid low carbon transition, while ensuring that new investments in long-lived infrastructure are made with full awareness of what they imply for long-term carbon commitment; and investing in ongoing modeling, analysis, and data collection that informs both public and private decision-making. These topics are discussed in more detail in Policy Implications of Deep Decarbonization in the United States.

Key Actions by Decade

This study identifies key actions that are required in each decade from now to mid-century in order to achieve net negative CO2 emissions by mid-century, at least cost, while delivering the energy services projected in the Annual Energy Outlook. Such a list inherently relies on current knowledge and forecasts of unknowable future costs, capabilities, and events, yet a long-term blueprint remains essential because of the long lifetimes of infrastructure in the energy system and the carbon consequences of investment decisions made today. As events unfold, technology improves, energy service projections change, and understanding of climate science evolves, energy system analysis and blueprints of this type must be frequently updated.

2020s

• Begin large-scale electrification in transportation and buildings
• Switch from coal to gas in electricity system dispatch
• Ramp up construction of renewable generation and reinforce transmission
• Allow new natural gas power plants to be built to replace retiring plants
• Start electricity market reforms to prepare for a changing load and resource mix
• Maintain existing nuclear fleet
• Pilot new technologies that will need to be deployed at scale after 2030
• Stop developing new infrastructure to transport fossil fuels
• Begin building carbon capture for large industrial facilities2030s
• Maximum build-out of renewable generation
• Attain near 100% sales share for key electrified technologies (e.g. EVs)
• Begin large-scale production of bio-diesel and bio-jet fuel
• Large scale carbon capture on industrial facilities
• Build out of electrical energy storage
• Deploy fossil power plants capable of 100% carbon capture if they exist

© 2019 by Evolved Energy Research 15

Maintain existing nuclear fleet

2040s

• Complete electrification process for key technologies, achieve 100% stock penetration
• Deploy circular carbon economy using DAC and hydrogen to produce synthetic fuels
• Use synthetic fuel production to balance and expand renewable generation
• Replace nuclear at the end of existing plant lifetime with new generation technologies
• Fully deploy biofuel production with carbon capture

© 2019 by Evolved Energy Research 16

1. Introduction

This report describes the changes in the U.S. energy system that, in concert with related actions in land use, will be required to reduce U.S. carbon dioxide (CO2) emissions to a level consistent with returning atmospheric concentrations to 350 parts per million (350 ppm) in 2100, achieving net negative CO2 emissions by mid-century, and limiting end-of-century global warming to 1°C. This study builds on previous work, Pathways to Deep Decarbonization in the United States (Williams et al. 2014) and Policy Implications of Deep Decarbonization in the United States (Williams, Haley, and Jones 2015) which examined the requirements for reducing GHG emissions by 80% below 1990 levels by 2050 (“80 x 50”).10

In the 1980s, with atmospheric CO2 concentrations climbing rapidly, the U.S. government recognized the need to establish a safe CO2 target and determine what would be required to reach it. By 1991, at the request of Congress, both the U.S. EPA and the Office of Technology Assessment had issued roadmaps for maintaining CO2 concentrations near the then-current level of 350 ppm (Lashof and Tirpak 1990; Office of Technology Assessment 1991). Over the last decade, as CO2 concentrations have risen toward and then passed 400 ppm, the question of what constitutes a “safe” concentration relative to dangerous anthropogenic impacts on the climate system has become increasingly urgent. A recent report by the Intergovernmental Panel on Climate Change evaluated the increased risks of 2°C of warming compared to exceeding 1.5°C, and of 1.5°C of warming compared to present warming, and found that a temperature rise of 1.5°C is “not considered ‘safe’ for most nations, communities, ecosystems, and sectors and poses significant risks to natural and human systems as compared to current warming of 1°C (high confidence)” (Intergovernmental Panel on Climate Change 2018). The U.S. Government’s Fourth National Climate Assessment similarly documents an acceleration of climate change impacts already underway (U.S. Global Change Research Program 2017). Studies using global climate models and integrated assessment models (IAMs) indicate that limiting

10 Available at http://usddpp.org/.

© 2019 by Evolved Energy Research 17

warming to a peak of 1.5°C will require reaching net-zero emissions of CO2 globally by mid- century or earlier (Intergovernmental Panel on Climate Change 2018). Reflecting these findings, a number of jurisdictions around the world have already announced more aggressive emissions targets, for example California’s recent executive order calling for the state to achieve economy-wide carbon neutrality by 2045 and negative net emissions thereafter (State of California 2018).

Several well-known climate studies have concluded that the best chance of avoiding the most catastrophic climate change impacts requires CO2 concentrations to be reduced to 350 ppm or less by the end of the 21st century ( Hansen et al. 2008; Veron et al. 2009; Hansen et al. 2013; Hansen et al. 2016; Hansen et al. 2017). The emission trajectories associated with reaching 350 ppm have lower allowable emissions (“emissions budgets”) in the 21st century than comparable trajectories that would peak at 2.0 or 1.5°C. These trajectories are intended to minimize the length of time the global temperature increase remains above 1°C to prevent the initiation of irreversible climate feedbacks indicated by paleoclimate evidence. In a recent article, Hansen and colleagues describe several possible trajectories for fossil fuel emission reductions that, in combination with specified levels of atmospheric CO2 removal, could achieve 350 ppm by 2100 (Hansen et al. 2017).

In this study we have modeled pathways – the sequence of technology and infrastructure changes – for the United States that result in net negative CO2 emissions before mid-century and that follow a global emissions trajectory consistent with a return to 350 ppm globally by 2100 (Figure 1). The cases modeled are a 6% per year and a 12% per year reduction in net fossil fuel CO2 emissions after 2020. These equate to a cumulative emissions limit for the U.S. during the 2020 to 2050 period of 74 billion metric tons of CO2 in the 6% case and 47 billion metric tons in the 12% case. (For comparison, current U.S. CO2 emissions are about 5 billion metric tons per year.) The emissions reductions in both cases must be accompanied by increased extraction of CO2 from the atmosphere. The 6% reduction case requires a global removal of 153 Pg.(C) incremental to the current global CO2 sink from 2020 to 2100, and the 12% reduction case requires an incremental removal of 100 Pg(C) during the same period. In our scenarios, the removal of the 100 Pg(C) or 153 Pg(C) is assumed to be accomplished through land-based negative emissions technologies (“land NETs”). These numbers imply an increase in the current

© 2019 by Evolved Energy Research 18

global land sink of about 40% and 60%, respectively (Le Quéré et al. 2018). Additional extraction of atmospheric CO2 using technological negative emissions technologies (“tech NETs”), meaning direct air capture (DAC) and bioenergy with carbon capture and storage (BECCS), is deployed in some of our cases. DAC is the removal of diffuse CO2 directly from the air, while BECCS involves capture of concentrated streams of CO2 from the effluent at industrial facilities that use biofuels; in both cases, the captured CO2 is stored in geologic structures.

Figure 1 Global surface temperature and CO2 emissions trajectories11.

The goal of this study is to understand what realistic 350 ppm-compatible cases would mean concretely for changes in the U.S. energy system and industrial fossil fuel use. Our study differs from recent IAM studies of 1.5°C in that it has a tighter emissions budget, concentrates on a single country, and provides a greater level of technical detail on the transformation to a low carbon economy, including sectorally detailed treatment of costs (Rogelj et al. 2015). The principal research questions addressed by this study are the following:

11 The solid blue line in (b) illustrates a 350 ppm trajectory based on 6% per year reduction in net fossil fuel CO2 emissions combined with global extraction of 153 PgC from the atmosphere. Reprinted from Hansen, ESD, 2017.

© 2019 by Evolved Energy Research 19

1. Is it technically feasible to achieve a 350 ppm trajectory within the U.S. energy system, given realistic constraints?
2. Is a 350 ppm trajectory robust against the absence of key carbon mitigation technologies, i.e. are there multiple technically feasible pathways?
3. What is the cost of achieving a low-carbon energy system on a 350 ppm trajectory in the U.S?

To answer these questions, we have developed future scenarios using two new models built for this purpose, EnergyPATHWAYS and RIO. These are sophisticated analysis tools with a high level of sectoral, temporal, and geographic granularity. We use these tools to rigorously assess the technical feasibility and cost of rapidly reducing CO2 emissions through the deployment of low carbon technologies and NETs, year by year from the present out to 2050. Changes in energy mix, technology stocks, emissions, and costs for the 350 ppm scenarios were calculated relative to a high-carbon baseline drawn from the Department of Energy’s Annual Energy Outlook (AEO), the U.S. government’s official long-term energy forecast.

The first research question above, regarding technical feasibility, was addressed through the development of a scenario called the 350 ppm base case, which uses currently available technologies to decarbonize the energy system while providing all the same energy services needed to support the U.S. economy and daily life forecast in the AEO. This scenario draws on objective, nationally recognized studies for both current data and future forecasts of performance and costs for each kind of fuel and technology used, in both energy supply and end use. The analysis in EnergyPATHWAYS and RIO was designed to address all major feasibility concerns, ranging from energy balances at a variety of scales, to the inertia of infrastructure stocks, to the hour-to-hour dynamics of the electricity system, separately in each of fourteen electric grid regions of the U.S.

The second research question is based on the observation that since even the best studies cannot perfectly predict the future decades ahead, it is important to understand what options exist if some key decarbonization technology or strategy does not materialize. This was addressed by the simulation of five additional 350 ppm pathways that remove or limit five key strategies used in the base case, either because in the future they do not meet current expectations for performance or cost, or because they are otherwise unable to be deployed at

© 2019 by Evolved Energy Research 20

scale, for example because they do not achieve social acceptance. The five scenarios start with the base case and then apply the following constraints separately: (i) limiting the availability of biomass for energy, (ii) limiting the rate of electrification of end uses, (iii) eliminating new nuclear plant construction, (iv) eliminating tech NETs, and (v) limiting the availability of land NETs.

In order to answer our third question, we calculate the costs of implementing this transition in the United States over the next three decades, with detailed year-by-year modeling of the energy economy. The 350 ppm-consistent scenarios are compared to a high-carbon case based on the AEO. This comparison is made “apples-to-apples” by ensuring that the energy services provided in the 350 ppm scenarios are the same as those provided in the AEO, and that the cost analysis reflects the differences in capital and operating costs for the low carbon technologies used in the 350 ppm scenarios relative to the business-as-usual technologies in the AEO.

The temporal, spatial, and sectoral detail in our modeling provides unique insights into how energy is supplied and used, and how carbon is, managed throughout the U.S. economy on a 350 ppm pathway. It improves current understanding of how energy and carbon removal interact technically, and how fossil fuel emissions, land NETs, and tech NETs trade off economically. Interactions between these different components of the energy-and-emissions system become increasingly important with tighter emissions constraints, so we account for them separately to avoid confusion and double-counting. Each of the scenarios demonstrates a different mode of utilizing infrastructure, balancing the electricity grid, and producing fuels as a single interactive system for least cost energy production. They also demonstrate how a 350 ppm-compatible energy system differs from one designed to achieve 80% reductions in CO2e below 1990 levels by 2050 (“80 x 50”), such as the pathways previously developed for the U.S. (Williams et al. 2014). 80 x 50 pathways are generally considered to be consistent with emissions scenarios (RCP 2.6) in the IPCC’s Fifth Assessment Report that give a 66% chance of not exceeding 2°C.

This study does not model land NETs, instead stipulating the global 100 Pg(C) and 153 Pg(C) cases mentioned above as boundary conditions for our scenarios. Some credible global evaluations indicate that achieving 153 Pg(C) of land-based C sequestration is potentially

© 2019 by Evolved Energy Research 21

feasible (Griscom et al. 2017) Achieving this level of sequestration will require changes in current policy and practices that not only improve carbon uptake but address such concerns as indigenous land tenure and competition with food production. Recent assessments of U.S. land- based negative emission potential indicate that a significant share of the required global land NETs, 20 Pg(C) or more of additional land sinks in the 21st century, is possible in the U.S. (Fargione et al. 2018).

For this analysis, an enhanced land sink in the United States on average 50% larger than the current annual sink of approximately 700 million metric tons was assumed.12 This would require additional sequestration of 25-30 billion metric tons of CO2 from 2020 to 2100. The present study does not address the cost or technical feasibility of this assumption, but stipulates it as a plausible value for the purpose of calculating an overall CO2 budget, subject to revision as better information becomes available.

The costs calculated in this study include the net system cost of the transformation in the supply and end use of energy, including tech NETs. They do not include the cost of land NETs or the mitigation of non-CO2 greenhouse gases. Macroeconomic effects are not explicitly considered. There are a variety of other benefits (“co-benefits”) of avoided climate change that are not within the scope of this study, including impacts on human health, ecosystems, and economic productivity. Such co-benefits are addressed in other studies.

The remainder of this report is organized as follows: Chapter 2, Study Design, including descriptions of the EnergyPATHWAYS and RIO modeling platforms, key data sources used, and the scenarios studied; Chapter 3, Results, including emissions, energy supply and demand, infrastructure, bioenergy use, carbon capture, and costs; Chapter 4, Discussion, addressing regional differences, electricity balancing challenges and solutions, cross-sector integration, and the circular carbon economy; and Chapter 5, Conclusions, including key actions by decade. The Appendix describes the scenarios and modeling methodology in detail.

12 U.S. EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2016, available at https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-2016

© 2019 by Evolved Energy Research 22

2. StudyDesign

2.1. Scenarios

This analysis explores the technical feasibility and cost of achieving a 350 ppm-compatible trajectory in the United States, transforming the energy system and achieving zero net emissions by mid-century. This is accomplished by developing a set of scenarios, subject to a variety of constraints (required outcomes and allowable actions), in the EnergyPATHWAYS and RIO models. In total we developed six 350 ppm- compatible scenarios: a core scenario called the Base Case, which is the least constrained, and five variants on this scenario to address concerns about the robustness of the results against implementation failures in key areas. The variants were designed specifically in response to criticisms of the assumptions made in other deep decarbonization analyses, typically conducted with global integrated assessment models. By doing this, we demonstrate the robustness of the result in a manner similar to “N-1” concept in electricity system planning, in which electricity systems must be designed with redundancy, so that even if the largest generator or transmission line fails, the system remains reliable. The contingencies that could threaten achieving 350 ppm are listed (1-5) below, then the business- as-usual baseline and the six 350 ppm-compatible scenarios are described quantitatively in Table 1.

Constraints

1. Restricted availability of zero-carbon primary biomass resources. This could be a result of inaccurate technical assessments of resource potential, unexpected difficulties in developing a bio-energy economy, or discovery of unintended impacts on other land- uses that lead to restrictions on biomass resource development.

2. Low rates of electrification. Direct electrification of end-uses such as light-duty vehicles and space heating is a key strategy of energy system decarbonization. Slower than expected rates of electrification would challenge a low carbon transition as it

© 2019 by Evolved Energy Research 23

would result in higher residual fossil fuel emissions that must be mitigated in other ways.

3. No new nuclear plants. Based on current cost forecasts for advanced (4th generation) nuclear facilities, it is expected that they would play a role in energy system decarbonization, especially in regions with limited renewable resource potential. Restricting new nuclear plant construction means that their role in a low carbon generation portfolio must be accomplished by carbon capture power plants or renewables. In this scenario we assume that nuclear plants already in operation will be operated and retired based on the schedule in the 2017 AEO.

4. No technological negative emissions technologies. “Tech NETS” includes biomass facilities (either fuel production or power generation) with carbon capture and sequestration or direct air capture with sequestration. Both of these technologies remove CO2 from the atmosphere, helping to offset any residual fossil fuel use in the economy. They are heavily relied on in many integrated assessment modeling (IAM) studies of deep decarbonization, which has drawn criticism. This scenario can be interpreted as a contingency test of what happens in the case of technological failure or social refusal of these approaches. While this scenario doesn’t employ Tech NETS, it does employ both carbon capture and carbon sequestration in other forms.

5. Low land NETS. Land NETS are strategies that use land-use management practices to increase terrestrial carbon sequestration. In this study we employ estimates of land NETS potential based on the literature to determine the remaining emissions budget for energy and industrial CO2. This scenario is used to assess whether the necessary energy and industrial CO2 mitigation can be achieved in the event that changes in land management practices can only produce 100 PgC of carbon sequestration globally by 2100. The changes to land NETS result in a reduced cumulative energy and industrial CO2 target and a net negative CO2 emissions target in 2050.

© 2019 by Evolved Energy Research 24

Table 1 Scenario definitions and emissions limits

2.2. Modeling Methods and Data Sources

This section summarizes the modeling methods used in this analysis. Further detail on all modeling tools and data sources is available in the Technical Appendix to this report.

2.2.1. EnergyPATHWAYS

EnergyPATHWAYS is a bottom-up energy sector scenario planning tool. It performs a full accounting of all energy, cost, and carbon flows in the economy and can be used to represent both current fossil-based energy systems and transformed, low-carbon energy systems. It

© 2019 by Evolved Energy Research 25

includes a granular technology representation with over 380 demand-side technologies and 100 supply-side technologies in order to represent all producing, converting, storing, delivering, and consuming energy infrastructure. It also has very high levels of regional granularity, with detailed representations of existing energy infrastructure (e.g., power plants, refineries, biorefineries, demand-side equipment stocks) and resource potential. The model is geographically flexible, with the ability to perform state-level and even county-level analysis. For this report, the model was run on a customized geography based on an aggregation of the EPA’s eGRID (U.S. Environmental Protection Agency 2018) geographies, as shown in Figure 2. The aggregation was done for computational purposes to reduce the total number of zones to a manageable number. EnergyPATHWAYS and its progenitor models have been used to analyze energy system transformations at different levels, starting in California (Williams et al. 2012) then expanding to U.S. wide analysis (Williams et al. 2014; Risky Business Project 2016; Jadun et al. 2017) and other state and regional analyses. The model has also been used internationally in Mexico and Europe. In each context, it has been successful in describing changes in the energy system at a sufficiently granular level to be understood by, and useful to, sectoral experts, decision makers, and policy implementers.

Figure 2 Regional granularity of analysis.

© 2019 by Evolved Energy Research 26

2.2.2. Regional Investment and Operations (RIO) Platform

EnergyPATHWAYS, described in the previous section, focuses on detailed and explicit accounting of energy system decisions. These decisions are made by the user as inputs to the model in developing scenarios. The Regional Investment and Operations (RIO) platform operates differently, finding the set of energy system decisions that are least cost. The rationale for using two models in this study is that energy demand-side decisions (e.g. buying a car) are typically unsuited to least cost optimization, because they are based on many socioeconomic factors that do not necessarily result from optimal decisions and are better examined through scenario analysis. However, RIO’s strength is in optimization of supply-side decisions where least cost economic frameworks for decision making are either applied already (e.g., utility integrated resource planning) or are regarded as desirable in the future. RIO is therefore complementary to EnergyPATHWAYS. We use RIO to co-optimize fuel and supply-side infrastructure decisions within each scenario of energy demand and emissions constraints. The resulting supply-side decisions are then input into EnergyPATHWAYS for energy, emissions, and cost accounting of these optimized energy supplies. RIO is the first model we are aware of to integrate the fuels and electricity directly at a highly resolved temporal level, resulting in a co- optimization of infrastructure that is unique and critical for understanding the dynamics of low- carbon energy systems.

RIO works with the same geographic representation as EnergyPATHWAYS. Each zone contains: existing infrastructure; renewable resource potentials and costs; fuel and electricity demand (hourly); current transmission interconnection capacity and specified expansion potential and costs; biomass resource supply curves; and restrictions on construction of new nuclear facilities.

2.2.3. Key References and Data Sources

The parameterization of EnergyPATHWAYS and RIO to perform U.S. economy-wide decarbonization analysis requires a wide variety of inputs and data sources. We describe the full breadth of these data sources in the Appendix. There are, however, a few principal sources that are central to understanding and contextualizing our results. First and foremost, we utilized the 2017 Annual Energy Outlook (U.S. Energy Information Administration 2017), which

© 2019 by Evolved Energy Research 27

includes detailed long-term estimates of economic activity, energy service demand, fuel prices, and technology costs. This allows us to compare our results to the principal energy forecast provided by the United States Government. Renewable costs and resource potentials are derived from National Renewable Energy Laboratory sources including the 2017 Annual Technology Baseline (National Renewable Energy Laboratory 2017) and input files to their ReEDS Model (Eurek et al. 2017). Biomass resource potential and costs are taken from the U.S. Department of Energy’s Billion Tons Study Update (Langholtz, Stokes, and Eaton 2016). In all cases we have sought to use thoroughly vetted public sources, which tend to be conservative about cost and performance estimates for low-carbon technologies.

© 2019 by Evolved Energy Research 28

3. Results

3.1. Emissions

Emissions trajectories for energy and industrial CO2 emissions are shown below for all 350 ppm scenarios. (For net emissions including the negative emissions from Land NETS (enhanced sink), see Table 1). In all scenarios, we find it to be technically feasible, from the standpoint of a reliable energy system that meets all forecast energy service demand, to reach emission levels consistent with the 350 ppm target (Figure 3).

Figure 3 CO2 emissions trajectories

© 2019 by Evolved Energy Research 29

All scenarios show broadly similar emissions trajectories, but they also contain differences in timing and carbon reduction magnitude. Scenarios that reduce the cost-effective mitigation options available before 2035 (i.e., Low Biomass and Low Electrification) show delayed emissions reductions, compensated by deeper reductions in later periods to hit the same cumulative CO2 budget. Those scenarios that remove options that are critical in the post-2035 time frame (i.e. No New Nuclear and No Tech NETS) show the opposite trend—larger reductions in near term CO2 reductions to accommodate the higher cost of deeper emissions cuts in the long-term. Scenario-specific findings include:

• The Low Land NETS scenario, with a smaller cumulative CO2 budget by mid-century, requires a steeper and deeper trajectory of emissions reductions from energy and industry than do the other scenarios, and thus requires higher levels of mitigation. This scenario requires that energy and industrial emissions become net negative after 2040, reaching -200 MMT CO2 per year in 2050, and remaining at that level through the rest of the century in order to meet the cumulative budget for the whole 2020-2100 period.
• The Low Electrification scenario shows the slowest rate of emissions reduction through 2035, with few cost-effective options for achieving the rate of transformation seen in the other cases with higher electrification rates. Post-2035, electrification levels catch up and the scenario employs more direct air capture than other scenarios to accelerate the mitigation trajectory.
• The Low Biomass scenario also shows a slower rate of emissions decline, as the biomass resources needed to displace fossil fuels directly are not available in sufficient quantity. The alternative strategy of electric fuels does not become cost-effective as a mitigation option until later in the period, when renewable penetrations increase and the electric fuel load can contribute to electricity balancing. The scenario uses direct air capture in the later periods as well, in order to accelerate the mitigation trajectory.
• The No New Nuclear scenario sees emissions decline faster than in other scenarios. This is because displacing residual fossil fuel on the electricity system is cheaper than attempting to achieve the same levels of electricity decarbonization in 2050 without the availability of nuclear. This scenario therefore finds a cost-effective route to have a slightly steeper slope but reach a less deep level by 2050.

© 2019 by Evolved Energy Research 30

• The 350 – No Tech NETS scenario also has a steeper initial trajectory relative to the Base scenario, because without Tech NETS, it becomes more expensive to reduce emissions in the long-term. Thus, the model trades higher near-term emission reductions for additional emissions budget in 2050.

Figure 4 shows the cumulative 2020 – 2050 emissions path of each mitigation scenario. The shape of cumulative emissions show how critical early action is for achieving 350 ppm goals, even with very aggressive climate mitigation. The first five years, 2020-2025, consumes over one-third (25 MMT) of the Base scenario budget, and the first decade, 2020-2030, consumes almost two-thirds.

Figure 4 Cumulative CO2 emissions trajectories

© 2019 by Evolved Energy Research 31

3.2. System Costs

Cost assessment of the 350 ppm scenarios is critical for assessing the potential economic and societal impacts of achieving a 350 ppm-compatible pathway, even if the technical feasibility of the pathway can be demonstrated. We apply a few different cost metrics to assess the economic feasibility of such a transition. First, we find the net cost of decarbonizing energy and industry to be consistent with results from other analyses of this type, using the metrics of incremental costs ($ per year) and incremental costs as a percentage of GDP per year (Figure 5). Incremental costs are calculated by comparing the cost of producing and using energy in each scenario compared to the baseline scenario derived from the AEO, which has no carbon constraint. Incremental cost includes the capital and operating costs of all low carbon energy supply infrastructure and demand-side equipment (e.g. electric vehicles and heat pumps) in comparison to the cost of the less efficient or carbon emitting reference technology that it replaces.

In all but one case these costs peak at less than 2% of the forecast GDP in 2040 (approximately $600B) annually. The Low Land NETS case is the only case that exceeds this value, with a 2040 value approaching 3% of GDP. This result emphasizes the value of negative emissions from land-use in managing the costs of decarbonizing the U.S. energy economy. All cases show the same peak in 2040, with continued cost declines of low-carbon technologies (renewables, electric vehicles, etc.) reducing the incremental cost compared to the fossil fuel baseline by 2050. We make no assessment of the human and environmental co-benefits (including, for example, avoided costs of climate impacts, national security benefits, and health benefits of improved air quality) associated with these emissions reductions, as such an assessment is outside of the scope of this analysis. However, such assessments have been made elsewhere (Risky Business 2015).

© 2019 by Evolved Energy Research 32

Figure 5 Annual net system costs in $2016 and as % of GDP

Second, we assess the total spending on the energy system (including carbon capture costs) as a share of GDP and compare that to historical levels of spending on energy. Incremental demand-side costs, such as the cost premium to purchase a high efficiency appliance, are assessed as an energy resource in this context, so that the incremental costs of electrification and efficiency are also treated as spending on energy. Figure 6 shows the results for the six 350 ppm scenarios and the baseline scenario relative to historical U.S. energy spending as a % of GDP going back to 1970. In all cases, the spending on energy in 350 ppm-compatible scenarios is lower than historical peaks. Even in the highest cost case, the peak is only equivalent to 2009 spending levels as a % of GDP. This is a measure of the economic feasibility of energy system transformation, a result arising from cost declines in renewables and electric vehicles (batteries), the continued transition of the U.S. towards a service economy, and the expected continuation of low natural gas prices which helps to manage overall energy system costs even in the 350 ppm-compatible scenarios.

© 2019 by Evolved Energy Research 33

Figure 6 Total energy system costs as % of GDP – modeled and historical

Third, while the overall system costs of 350 ppm pathways are within the range of historical values for the U.S., the way that money flows within the energy economy changes substantially. In the low-carbon economies represented by the 350 scenarios, low-carbon technology investments are substituted for fossil fuels. This transformation is shown in Figure 7 with large new investments in biofuels, demand-side equipment, electric fuels, the electricity grid, and low-carbon generation being offset by dramatically reduced spending on coal, natural gas, and oil, especially the refined oil products gasoline, diesel, and jet fuel.

© 2019 by Evolved Energy Research 34

Figure 7 Components of net energy system costs for all cases

3.3. Energy Transition

Transformation of the U.S. energy system occurs on both the demand and supply side of the system. Final energy consumption rapidly transitions away from direct combustion of fossil fuels towards the use of electricity (e.g. from gasoline powered vehicles to EVs) and other low carbon energy carriers, accompanied by a supply-side transition from primarily fossil sources of energy towards zero-carbon sources such as wind, solar, biomass, or uranium. Figure 8 shows these simultaneous transitions, with the left-hand side showing primary energy supply and the right-hand side showing final energy demand.

© 2019 by Evolved Energy Research 35

Figure 8 Primary and final energy demand for all cases from 2020 – 2050

Figure 9 shows the transition of the energy mix over time, as reflected on both the supply and demand sides of the system. The three columns show energy divided into the main energy carrier types (liquids, gases, and electricity). The top row shows the transition in final energy demand over time, broken down by sector. The use of liquids and gases falls dramatically over

© 2019 by Evolved Energy Research 36

time as a result of electrification, while electricity use increases for the same reason. The second row shows the evolving mix of energy types used to meet the final demand shown in the first row. The third row shows the average emissions intensity of the energy supply mix in the second row, which declines over time as lower carbon sources are used. The bottom row shows the total emissions over time from each of the main energy carriers, the product of the total amount of each used times its emissions intensity.

Figure 9 Components of emissions reduction for liquids, gas, and electricity in the 350 – Base case

© 2019 by Evolved Energy Research 37

3.4. Infrastructure

Accomplishing the transformation of the energy sector requires significant investments in low- carbon infrastructure and a transition away from the extraction of fossil fuels. We’ve noted the costs of these investments, and this section details the scale of required infrastructure in the key areas of demand-side equipment, low-carbon electricity generation, biofuels production, electricity storage, electricity transmission, hydrogen electrolysis, and direct-air capture facilities.

3.4.1. Demand-Side Transformation

In addition to employing many efficiency measures, primarily in electric-only end-uses like lighting, ventilation, and household appliances, the demand-side undergoes a large transformation in end-uses where there are direct electric alternatives to fuel combustion. Transitions to electric technologies results in efficiency gains as well as a reduction in the amount of fuels that need to be displaces by bio-based or electric alternatives, reducing the overall cost of achieving emissions reduction goals. Figure 10 shows this transition for a variety of residential, commercial, productive, and transportation end-uses. Transportation electrification is the most critical sector to achieve these electrification goals in due to the volume of liquid fuels that it currently consumes.

© 2019 by Evolved Energy Research 38

Figure 10 Electric Technology Stock Shares

© 2019 by Evolved Energy Research 39

3.4.2. Low-Carbon Generation

All 350 ppm-compatible scenarios result in the addition of 2000 to 3000 gigawatts of renewable electricity capacity by mid-century – in comparison to total electricity generating capacity of all kinds of about 1000 gigawatts today – because renewables are the lowest-cost zero carbon resource available (Figure 11). This capacity takes the form primarily of new wind resources through 2030, since the majority of U.S. electricity demand (and population) is located in areas with better wind than solar resources, and thus provides better economics for wind. Wind generation is also able to reach higher shares of total generation (renewable penetration) than solar before encountering significant balancing challenges, because the production profile of wind is more evenly distributed throughout the day, in contrast to the concentration of solar generation within a narrow band of hours. Penetrations of solar beyond a certain percentage requires complementary balancing resources, such as energy storage or flexible loads, to avoid curtailment and enable full utilization of the resource.

New renewables built after 2040 are primarily solar PV because: (1) the supply of new low-cost wind resources is exhausted by this point; (2) solar costs continue to decline; and (3) the system’s growing electric fuels production capacity can utilize larger quantities of daytime solar electricity production. Offshore wind is used as a resource in the Low Land NETS and No New Nuclear scenarios, primarily in the Northeast (New York and New England). Higher penetrations of offshore wind would be seen if onshore wind becomes difficult to site, or cost declines for offshore wind turn out to be greater than anticipated.

The scenarios in which new nuclear generation is permitted to be built also see an expansion of nuclear, though the importance of new nuclear is less critical than some of the constraints in other scenarios, as the No New Nuclear scenario shows. A relatively modest increase in the deployment of new renewables above the level in the Base 350 ppm scenario compensates for the constraint imposed by No New Nuclear.

Carbon capture and storage (CCS) is much less important in power generation in all scenarios than it is for capturing the CO2 streams from biofuel refining and other industrial activities. High

© 2019 by Evolved Energy Research 40

penetrations of renewables, which result in frequent surpluses of low marginal-cost energy, mean that CCS generators don’t achieve high capacity utilization. This makes CCS an expensive option for limiting emissions from electricity due to its high capital cost spread over a limited number of hours. CCS electricity generation is generally found in regions with restricted new nuclear build and within regions that have limited wind resources where it can provide a consistent source of off-peak power.

Figure 11 Low-Carbon generation capacity growth

3.4.3. Biofuels Production

3.4.3.1. Liquids

The expansion of biofuels production is a critical strategy for reducing emissions as the economy transitions towards high levels of electrification. Even at the conclusion of the electrification transition, liquid biofuels play an important role in mitigating emissions in hard- to-electrify end-uses such as heavy industry and aviation. The United States already has a biofuels industry of significant size, but it primarily produces corn-derived ethanol, a relatively

© 2019 by Evolved Energy Research 41

high carbon form of biofuel over its lifecycle. As light-duty vehicle travel is electrified, the demand for liquid transportation fuels decreases, and this sector is reduced in importance. This analysis did not find cellulosic ethanol to be a critical strategy during the transition from gasoline to electricity due to the high cost of developing cellulosic refining and distribution, and the pace of electrification (the market-size for gasoline alternatives shrinks very quickly). This analysis also finds that the focus of biofuels should be on displacement of liquid fossil fuels, rather than gaseous fuels. This is due to: (a) natural gas has a lower cost per MMBtu than refined liquid fuels; and (2) natural gas CO2 emissions are lower than liquid fossil fuels on an energy basis. Liquid biofuels production is shown Figure 12. While this represents a rapid expansion of production capacity of up to 4 million barrels per day by 2040, it is still only a fraction of the current capacity of U.S. petroleum refineries. In cases where bio-energy carbon capture and storage (BECCS) is allowed, it dominates fuel production, which is understandable given the economic attractiveness of biorefineries for carbon capture, with concentrated CO2 streams and high utilization factors.

Carbon capture and storage (CCS) is distinct as a strategy from carbon capture and utilization (CCU) in this analysis. In CCU, the captured carbon is used in combination with electrically produced hydrogen to produce methane and other synthetic liquid or gaseous fuels that can be substituted for gasoline and diesel or natural gas, respectively. Bio-energy carbon capture and utilization (BECCU) is used either when the marginal cost of sequestration becomes high or in regions where there are substantial biomass resources but low sequestration potential, for example due to geographic unsuitability. In the No Tech NETS case, BECCS is primarily displaced by biofuels production without capture, but this case also has the highest BECCU production capacity.

© 2019 by Evolved Energy Research 42

Figure 12 Liquid biofuels production capacity in Million Barrels Per Calendar Day (MBCD) and as a % of current U.S. petroleum refining capacity

3.4.3.2. Gaseous Fuels

The analysis finds that the priority use for biomass feedstocks is liquid fuel production, but in some scenarios, there is also limited production of gaseous biofuels. The Low Land NETS case requires a displacement of almost all fossil fuels, including gas in the pipeline, and so we see deployment of biogas in this case (as well as significant amounts of fuels produced from electricity). This is shown in Figure 13 both in units of TBTU/Year, which is more appropriate for gaseous fuels, and also on the right-hand axis in units of million barrels per calendar day, for purposes of comparison to the scale of liquid biofuels.

© 2019 by Evolved Energy Research 43

Figure 13 Gaseous biofuels production capacity in TBTU/Year and as a % of current U.S. refining capacity

3.4.4. Electricity Storage

Electricity storage provides capacity to balance the electricity system during times of low renewable energy output. Battery storage is the lowest-cost capacity resource available to address system peaks of limited duration. For this reason, it is deployed on a significant scale even in the Baseline scenario which has no carbon constraints (Figure 14). We find that significant amounts of new electricity storage are needed in all 350 ppm-compatible scenarios starting in 2030, and this storage is deployed with an average duration of four to six hours. Without a significant technological breakthrough, however, the high cost of stored electricity

© 2019 by Evolved Energy Research 44

limits its value as a long-duration balancing resource (i.e. on scales from days to months of energy shortfalls from renewables). Thus, it operates primarily as a diurnal resource, using excess solar generation in the middle of the day on a consistent basis to avoid curtailment and to displace thermal generation off-peak (capacity and energy).

Figure 14 Energy storage capacity in gigawatts, gigawatt-hours, and average duration

3.4.5. Electricity Transmission

Many deep decarbonization analyses emphasize the importance of transmission to match the supply and demand for renewable electricity spatially across the country. Our findings are consistent with these studies in terms of the value of transmission as a resource. However, transmission has historically proven difficult to permit, site, and build in the U.S., especially in the case of large inter-regional lines. For this reason, in our analysis we have constrained new transmission construction to a doubling of currently existing capacity between regions. This is likely conservative, as some regional interties are quite small at present, not because of being technically or societally difficult but due to a lack of economic justification. However, this

© 2019 by Evolved Energy Research 45

admittedly arbitrary limit serves as a useful proxy for barriers to transmission construction, and at any rate is non-binding (does not constrain inter-regional flows) in almost every instance.

Limits on new transmission build may present less of a handicap in our analysis compared to some because our analysis employs other methods to transfer renewable energy between regions, namely through pipelines in the form of fuels produced from electricity (storage of such fuels within regions also provides a form of renewable energy storage). Still, we do see significant new interties between some regions, with almost all regions seeing some new economic transmission build by 2050, in all scenarios. Figure 15 shows all the regional intertie capacity built from 2020 to 2050. The largest such builds are between the CAMX region (California and northern Baja California), which requires imports of wind energy from NWPP (Northwest) and AZNM (Arizona and New Mexico) in all scenarios. There is also major development of transmission capacity between the RFC and SR (Mid-Atlantic and South- Atlantic) regions in the Low Land NETS scenario, related to the higher use of DAC and production of electric fuels in this scenario.

© 2019 by Evolved Energy Research 46

Figure 15 Incremental electric transmission capacity (gigawatts) by corridor

© 2019 by Evolved Energy Research 47

3.4.6. Hydrogen Electrolysis

The use of electricity to produce hydrogen from the electrolysis of water plays a key role in balancing the electricity system during periods of renewable energy overgeneration. The hydrogen produced is then used to create synthetic fuels that can be used in applications that are difficult to electrify. As illustrated in Figure 16, all pathways require more than 100 GW of electrolysis capacity, and the cases that require substantially more electric fuel production – Low Biomass, Low Electrification, Low Land NETS, and No Tech NETS – have up to 400 GW. This situation can be said to constitute a type of “hydrogen economy,” but not the type that has typically been discussed in the literature, in which hydrogen itself becomes a principal energy carrier for end uses. Many of the objections raised regarding that form of hydrogen economy center on the difficulty of developing a delivery infrastructure for this highly flammable fuel. Instead, in our 350 ppm scenarios electrically produced hydrogen is used as a feedstock in the production of renewable liquid and gaseous fuels that already have existing delivery mechanisms. Hydrogen can be combined with captured carbon dioxide to produce methane, the main component of natural gas, and further chemical synthesis using the Fischer-Tropsch process can produce synthetic liquid fuels comparable to (and interchangeable with) refined petroleum products, including diesel, gasoline, and jet fuel. Produced hydrogen can also be injected into a natural gas pipeline directly (limited to 7% by energy, which research has shown can be blended with fossil-based or synthetic natural gas without damaging end use equipment or delivery infrastructure). The hydrogen intended for pipeline injection is represented by the blue wedge in Figure 16. In sum, the “hydrogen economy” used in the 350 ppm scenarios avoids many of the infrastructure challenges typically associated with the use of hydrogen at large scale.

The production of electrolytic hydrogen and synthetic fuels provide the primary method of long-duration energy storage for a system with high penetrations of renewable generation. When peak electricity generation exceeds demand, the extra electricity is used to synthesize these fuels. These fuels can be used directly to meet demand for liquid and gaseous fuels and— to a limited extent— also be used to produce electricity at times of fallow renewable production. However, unlike previous hydrogen economy conceptions, in the 350 ppm scenarios the principal mechanism by which electric fuels balance the electricity system is not

© 2019 by Evolved Energy Research 48

round-trip electricity storage (production and storage, then burning in a power plant to produce more electricity), but instead by enabling the economically efficient over-building of renewable resources, in which curtailment (wasted energy) is minimized because the energy is used to produce fuels used elsewhere.

Figure 16 Capacity of hydrogen electrolysis for pipeline injection and synthetic fuels

3.4.7. Direct Air Capture

Direct air capture (DAC) is the removal of CO2 directly from ambient air. It has traditionally been imagined as a post-2050 technology, as 2oC scenarios called for achieving net-zero and net-negative emissions levels in the 2070 time frame. This analysis, however, demonstrates a role for DAC even before a net-zero economy is reached (Figure 17). In scenarios where there are insufficient biomass-based alternatives to replace fossil fuels (Low Electrification and Low

© 2019 by Evolved Energy Research 49

Biomass), DAC plays a role in accelerating the transition necessitated by the cumulative emissions cap, either by creating a carbon feedstock used for electric fuel production (DAC for utilization) or through geological carbon sequestration. We find that a heightened emphasis on the early commercialization of DAC is warranted due to its role as an accelerator of the overall transformation, as well as its obvious role as a technological backstop in the event of such contingencies as slower electrification or limited biomass deployment.

Figure 17 Direct air capture capacity for sequestration and utilization (MMT/Year)

Historically there has been reticence to treat DAC as a legitimate portfolio technology for achieving deep emissions reductions, not necessarily for reasons of technological maturity or acceptance but because of “moral hazard”: the not unwarranted concern that the presence of this technology could be used to justify continued unabated combustion of fossil fuels. Our analysis, however, shows that there is clearly a place for DAC in the rapid transition to low-

© 2019 by Evolved Energy Research 50

carbon energy systems, not as an alternative to decarbonization but as a complementary technology to hasten energy decarbonization and increase sequestration. Our analysis also shows that DAC pairs best economically with low-cost zero carbon resources such as wind and solar, because DAC (like hydrogen electrolysis) is a large industrial load that has high variable costs relative to fixed costs, and can therefore operate flexibly at less than full utilization, taking advantage of periods of renewable overgeneration. Alternative carbon capture scenarios in which grid electricity continues to be provided by fossil thermal generation do not offer the same economic opportunities.

© 2019 by Evolved Energy Research 51

4. Discussion

In addition to high-level summary results presented above, this discussion details additional features and components of low-carbon energy systems that can achieve 350 ppm trajectories in the U.S.

4.1. Four Pillars

Deep decarbonization analyses have relied on three primary strategies for achieving emissions targets: (1) electricity decarbonization, the reduction in the emissions intensity of electricity generation; (2) energy efficiency, the reduction in units of energy needed to provide energy service demands; and (3) electrification, the conversion of end-uses from fuel to electricity. These have been referred to as the “three pillars” and the use of these strategies to achieve deep decarbonization is a robust finding across many jurisdictions both domestically and internationally. Under our scenarios, which assume EIA projections for economic growth and increased consumption of “energy services”, achieving 350 ppm requires the inclusion of a fourth pillar, carbon capture, which includes the capture of otherwise emitted CO2 from power plants, industrial facilities, and biorefineries. It also includes the use of direct-air capture facilities to capture carbon from the atmosphere. Once captured, this CO2 can either be utilized in the production of synthesized electric fuels or it can be sequestered. Both strategies are used extensively in the scenarios analyzed here.

Figure 18 below shows the four pillars of decarbonization employed in the Base scenario. The emissions intensity of electricity has declined to less than 50 tonnes/GWh in 2050 from 350 tonnes/GWh in 2020, which is itself less than the current U.S. average of 424 tonnes/GWh in 2016 (U.S. Energy Information Administration 2018). One of the principal strategies employed in the early years is a change in the merit order dispatch (the prioritization used by system operators to determine the order in which electric generation is employed to meet demand) so that electricity generation from gas plants is prioritized over generation from coal plants. This

© 2019 by Evolved Energy Research 52

accounts for the difference between 2020 Electricity Decarbonization in Figure 18 and 2016 historical. Energy consumption per dollar GDP, one metric for energy efficiency, decreases substantially from 3.2 in 2020 to 1.2 in 2050. This is due to significant same-fuel energy efficiency, economic transition towards services, and direct electrification of end-uses, which contributes efficiency gains over fuel alternatives. Electricity, used either directly (e.g. in electric vehicle) or as an electrically produced fuel represents almost 60% of final energy demand by 2050. Carbon capture contributes 800 MMT of emissions reductions by 2050, either when directly sequestered or when utilized for making synthetic electric fuel.

Figure 18 Four pillars of deep decarbonization in the 350 – Base case

4.2. Regional Focus

Our current energy economy exhibits significant regional variation in terms of energy demand, energy supply, and overall energy costs. The future energy economy will exhibit these same regional variations and it is worth identifying geographic regions that may be at the center of the new energy economy. While all regions require significant investment in generation resources to decarbonize their sources of supply, some regions, due to particular resource endowments, will see additional investment as they become both the center of the fuel

© 2019 by Evolved Energy Research 53

production and direct air capture sectors. These regional dynamics are illustrated in Figure 19 for the Low Land NETS case, which requires the most significant infrastructure investments of all the cases and therefore shows the regional dynamics most clearly.

Figure 19 Regional infrastructure needs in the 350 – Low Land NETS case

© 2019 by Evolved Energy Research 54

Zero-carbon generation follows a predictable pattern of resource endowments, with regions that have access to the wind-belt extending from Texas up through Wyoming developing wind- heavy portfolios. The Southwest and Southeast have solar-heavy portfolios. The Northeast relies primarily on wind, both onshore and offshore. The Southeast and Midwest also rely on nuclear energy in this case.

Biofuels production will be concentrated in areas with significant biomass resources, primarily the Midwest and the Southeast. In addition, biofuels production is best located in areas with available saline aquifers in which captured CO2 can be stored. Electric fuels production will be determined by availability of CO2 as well as grid conditions that support low-cost electrolysis. These conditions include either low-cost solar or wind generation, which explains the high concentration of electrolysis in the desert southwest, as well as in the wind-belt. One additional consideration not modeled explicitly here is the availability of water, which may affect siting of electrolysis facilities as well. Direct air capture facilities will depend similarly on low-cost renewables as well as the availability of saline aquifers for sequestration or hydrogen for synthetic electric fuels production.

4.3. Electricity Balancing

Electricity balancing, which is the matching of electricity supply and demand at all time-scales is one of the principal technical and economic challenges of decarbonization. The systems modeled here have a large percentage of non-dispatchable generation resources. On important characteristic is that variable costs for these resources are low and curtailing production represents lost economic value. In many studies of low-carbon electricity systems, the principal resource used to balance these types of systems is electricity storage (batteries, pumped hydro, etc.). However, this is an incomplete toolkit, specifically when dealing with imbalances that can persist over days and weeks. This analysis expands the portfolio of options available to address the balancing challenge, employing solutions such as flexible electric fuel production, dual-fuel boilers systems (i.e. gas and electric), and direct air capture in addition to traditional solutions such as batteries, thermal generation, and transmission expansion. Figure 20 shows balancing behavior in the ERCOT (Texas) dispatch region in 2050 in the Low Land NETS case.

© 2019 by Evolved Energy Research 55

The range of daily system balancing operations are shown by the set of transparent lines, while the daily average behavior shown by the thicker opaque lines. It can be seen, in this scenario, the lion’s share of balancing needed is provided by direct air capture and electrolysis loads. Thermal generation is needed infrequently but must be maintained on the system for purposes of reliability. Storage exhibits a diurnal pattern, common across all resources, of increased load in the middle of the day responding to regular solar overgeneration conditions. Due to limited physical interties between ERCOT and other regions, transmission plays a relatively minor role here.

Figure 20 Electricity balancing from key technologies in ERCOT in 2050 in the 350 – Low Land NETS case

One can see the relative economics of building each type of capacity to balance load by looking at the average operations of each resource compared to the maximum operations in any single period. This average operation represents the utilization factor of each resource (this is referred

© 2019 by Evolved Energy Research 56

to as a “capacity factor” for generation resources). More expensive capacity such as direct air capture facilities operate at a much higher utilization factor than do cheaper forms of capacity such as electrolysis or batteries.

All these solutions contribute to addressing the balancing challenges posed by large amounts of non-dispatchable resources. Figure 21 shows the overall contribution by resource type, case, and year. How a resource contributes to electricity balancing is a function of its unique characteristics. Thermal generation and hydro contribute to balancing the system by generating during periods of some combination of low renewable output and high load; storage moves energy from overgeneration periods to hours where thermal generation would otherwise be needed; flexible fuel production and direct air capture balance the system by soaking up overgeneration and turning it either into electric fuels or sequestering carbon directly; finally, renewable curtailment balances the system by reducing overgeneration when there is no economic case for utilizing it.

The relative contributions are unique to each case and resource build, but there are commonalities. First, the scale of balancing needs in 2050 compared to 2020 is drastically different. That’s because the net-load signal that the system is trying to balance is significantly more volatile, as renewables make up a larger portion of generation.

In all cases, thermal generation provides most of the balancing through 2030 before the significant renewable penetration that ramps up post-2030. Flexible electric loads (Ex. fuel production and duel fuel boilers) play a role in all cases and become the dominant resource in cases where they are needed to displace fossil fuels. Storage plays a key role but not a solitary one, as its primary use is to operate diurnally and balance out solar overgeneration. Renewable curtailment is present in all cases.

© 2019 by Evolved Energy Research 57

Figure 21 Balancing contribution by resource (TWh)

Figure 22 shows the initial net-load signal for ERCOT across sample days in 2020, 2030, and 2050. In 2020, variation is primarily a result of the electricity demand shape. By 2030 and certainly by 2050, that load variation is swamped by variability in renewable output, with average daily swings in the net load of almost 100 GW.

© 2019 by Evolved Energy Research 58

Figure 22 Net Load in 2050 in ERCOT in the 350 – Low Land NETS case

Even with the magnitude of the different balancing solutions employed in the 350 ppm scenarios, there is in all cases still some level of renewable curtailment that is economically efficient. That is, during periods of significant or sustained overgeneration, the capacity that would be needed to be built to fully utilize that renewable energy generation is not economic. Economic curtailment exhibits a distinctly seasonal pattern, with much of it occurring in the spring and fall, shown in Figure 23. This is due to either generally high renewable production (wind, in particular, has a strongly seasonal shape), generally low-load conditions (i.e. no heating or air conditioning load), or a combination of the two. In regions with significant hydro resources, seasonal release requirements can contribute to spring overgeneration conditions as well. The baseline scenario has very low curtailment because without carbon constraints, the impetus to push renewable penetrations to levels that result in curtailment is diminished.

© 2019 by Evolved Energy Research 59

Figure 23. Renewable resource curtailment patterns in 2050

4.4. Sector Integration

The 350 ppm-compatible scenarios demonstrate the need for sectoral integration in deeply decarbonized economies. The lines between traditionally distinct sectors become blurred when decisions and their effect are so tightly linked across sectors. For example, the need for electric fuels to replace fossil liquid or gaseous fuels has a huge impact on renewable resource needs in the electricity sector, as well as on the need for supplementary balancing resources such as electric storage. Electric fuel production even competes with the need for transmission, as energy can instead be transferred between high renewable production zones either as gaseous or liquid fuels.

© 2019 by Evolved Energy Research 60

The demand-side transformation, especially rapid electrification in buildings, transportation, and industry, will also require sectorally integrated planning both to ensure that new generation resources are developed to meet the growing demand, and also to plan distribution system upgrades and charging infrastructure, and to leverage the ability of new electric loads (specifically, space heating, water heating, and vehicle charging) to operate flexibly. Figure 24 shows the rate of load growth in each of our cases, with rates exceeding 4% in some cases during the 2030 to 2040 timeframe.

Figure 24 Electric load growth by year (%)

Allocation of limited biomass resources is another area in which cross-sector integration is critical. Some jurisdictions have undertaken policies that emphasize 100% renewable electricity. The ambition of these types of targets is consistent with the challenge of deep emissions reductions targets, but 100% renewable or zero-carbon electricity can be regarded de facto as a

© 2019 by Evolved Energy Research 61

biofuels allocation policy—achieving a zero emissions target requires some portion of biomass to be burned in electric generators rather than used as liquid biofuels. Allocation of biomass towards liquids, however, might be lower cost and provide the same overall emissions reduction, illustrating the fungibility of emissions reductions between sectors.

4.5. Circular Carbon Economy

The circular carbon economy, or CCE, is a term for an energy economy that uses CO2 embodied in biomass feedstocks or through direct air capture to produce electric fuels. Given existing energy service delivery mechanisms, both fuel delivery and fuel consumption infrastructure, large portions of energy demand in 2050 is still met as it is today, with liquid and gaseous fuels. These fuels can no longer be fossil-based and so require drop-in, non-fossil-based alternatives.

These fuels begin as electrolyzed hydrogen before they are catalyzed with captured CO2. Critical sources of carbon for utilization in this analysis are biorefineries and direct air capture facilities. Biorefineries that are located in areas with limited sequestration potential are specifically good candidates as they can run at high utilization factors and have extremely concentrated sources of CO2 emissions for low-cost capture. DAC facilities with utilization are also employed to a lesser extent as seen in Figure 17. This is a critical strategy in the long-term, even before net-zero emissions economies have been achieved.

© 2019 by Evolved Energy Research 62

5. Conclusions

Based on the analyses described in this report, we conclude that achieving U.S. emissions consistent with 350 ppm globally is technically feasible. This result is robust against five key strategies not materializing at the scale expected in the base case – biomass deployment, electrification, new nuclear deployment, technological NETS deployment, and land NETS implementation. While feasible, achieving the outcomes modeled here requires ambitious early action in order to maintain reasonable trajectories towards mid-century. Without this ambitious early action, it will require the achievement of net-negative emissions energy economies before mid-century and then sustain them at these low-levels through the end of the century.

These scenarios are intended to answer the question of whether the U.S. and its anticipated growth in consumption of energy services can develop an energy system that is consistent with 350 ppm in the atmosphere and we conclude that it can. We do not assert the necessity of, nor model the effects of, behavioral changes and energy service demand reductions (i.e. lower VMTs, lower temperature setpoints, lower consumption of material goods) though all would contribute to lower system costs, lower material requirements, lower infrastructure needs, and could improve quality of life in ways not measured by this analysis. There are co-benefits aside from CO2 including improved air quality, energy price predictability, job creation and energy security that are not modeled here.

We observe large shifts in energy spending away from fossil fuels towards fixed infrastructure, both demand-side (electric vehicles, heat pumps, etc.) and supply-side (low-carbon generation, hydrogen electrolysis, electric storage, etc.). That said, the overall net costs of decarbonization found here are well within the range that a major industrial economy can manage, and indeed that the U.S. has managed historically. Based on this analysis, achieving 350 ppm-compatible pathways would maintain energy system costs within the low-range of historical values.

© 2019 by Evolved Energy Research 63

5.1. Key Actions by Decade

In conclusion, “Key Actions by Decade” below describes the sequence of actions needed to achieve a 350 ppm trajectory in the U.S. The list is by no means comprehensive, but it does highlight the most important physical transformations required and when each needs to occur. These actions make up a general blueprint for the U.S.—regional differences in resource endowment, existing infrastructure, and societal preferences will mean that not every step is universally relevant. In some cases, these actions need to build on one another, so that later actions are path dependent on earlier successes.

This and previous research have indicated that many pathways to decarbonize the energy system exist. The list below represents our current best understanding of how to achieve mid- century carbon targets at lowest cost while delivering the energy services projected in the 2017 AEO. Inherently this blueprint relies on projections of cost and performance that are unknowable. Despite this, a long-term blueprint is essential because of the long lifetimes of infrastructure in the energy system—making decisions that have long-term consequences using imperfect information is an enduring challenge. Uncertainty means an energy system plan is never static. Thus, we expect future work to revise this plan as decisions get made, technology improves, energy service projections change, and as our understanding of the climate science evolves.

From a policy perspective, this provides a list of the things that policy needs to accomplish, for example the deployment of large amounts of low carbon generation, rapid electrification of vehicles, buildings, and industry, and building extensive carbon capture, biofuel, hydrogen, and synthetic fuel synthesis capacity. Some of the policy challenges and opportunities that must be managed include: land use tradeoffs related to carbon storage in ecosystems and siting of low carbon generation and transmission; electricity market designs that maintain natural gas generation capacity for reliability while running it very infrequently; electricity market designs that reward demand side flexibility in high-renewables electricity system and encourage the development of complementary carbon capture and fuel synthesis industries; coordination of planning and policy across sectors that previously had little interaction but will require much

© 2019 by Evolved Energy Research 64

more in a low carbon future, such as transportation and electricity; coordination of planning and policy across jurisdictions, both vertically from local to state to federal levels, and horizontally across neighbors and trading partners at the same level; mobilizing investment for a rapid low carbon transition, while ensuring that new investments in long-lived infrastructure are made with full awareness of what they imply for long-term carbon commitment; and investing in ongoing modeling, analysis, and data collection that informs both public and private decision-making. These topics are discussed in more detail in Policy Implications of Deep Decarbonization in the United States (Williams et al. 2015).

2020s

• Begin electrification – Electrification of buildings, transportation, and industry is necessary for affordable decarbonization. The initial focus should be on making new buildings all electric and building markets to electrify vehicles of all types. The transportation electrification goal is not near-term carbon emissions reductions but instead transformation of an industry to eliminate carbon emissions in the long term as the carbon intensity of electricity drops. Replacing air conditioners or furnaces with heat pumps in existing buildings is also a priority, pushing a technology that has improved markedly in recent years to further maturation. Steps towards electrification also involve removing systemic bias preventing electrotechnology adoption that are often good intentioned around energy efficiency goals but self-defeating in the long term. Examples include providing incentives on high-efficiency gas furnaces but no such incentives on heat-pumps or policies that discourage electric utility load growth of any type.
• Switch from coal to gas in electricity system dispatch – Dispatching gas in preference to coal is one of the most impactful and cost-effective ways to curtail carbon emissions in the near-term. Natural gas has approximately half the carbon intensity of coal but costs only slightly more on an energy basis at time of writing and is generally burned more efficiently than coal. Coal to gas switching in dispatch is distinct from retiring all coal, which will happen more gradually due to considerations on reliability and speed at which replacement generation can be built. Natural gas plants also are better complementary generation in the medium-term as renewable generation is deployed.
• Build renewables and reinforce TX where possible – Due to their abundance and based on current cost projections, wind and solar will form the backbone of a future low carbon energy system. Meeting 2050 goals requires a truly enormous quantity of renewable deployment, which must accelerate. Complementarity between wind and solar profiles means both get built wherever possible, but regional specialization will

© 2019 by Evolved Energy Research 65

occur depending on resource quality. Offshore wind should be emphasized in places, like the Northeast, where this resource holds promise as a vital part of the electricity system long-term. Transmission that connects renewable resources to loads takes time to permit and build and thus planning must start early for this critical infrastructure.

• Allow gas build to replace retiring plants – Even in a future electricity system with 80%+ energy coming from renewables, difficult long-duration (seasonal) electricity balancing challenges mean that dispatchable thermal capacity that can be dispatched during fallow periods of renewable production will be a part of a low-cost energy system. Our modeling shows that an optimized pathway to deep decarbonization shows little change to gas capacity relative to today over the next 30 years but eventual retirement of all other fossil electricity generation.
• Start electricity market reforms to prepare for a changing load & resource mix – Future electricity systems must accommodate rapid load growth from electrification, increasingly flexible demand, and increasingly inflexible supply resources. Fossil generation in the future without carbon capture will operate for far fewer hours than today making capacity markets more and more attractive. In those capacity markets the need to distinguish resources that can offer capacity over long durations will become important. Future energy markets must also compensate balancing services, with full symmetry between supply and demand side balancing.
• Maintain nuclear – Nuclear is an important source of low-cost carbon free electricity and when possible to do safely, the lowest cost path to decarbonization involves maintaining these resources. Retiring nuclear to ‘make room’ for renewable resources is ultimately self-defeating. Reducing climate change should be the priority when weighed against nuclear accidents given relative risk and consequence except where specific circumstances dictate otherwise (E.x. reactors in active seismic zones). This is not an assertion of the safety of generation III nuclear but rather a recognition of the urgency of the latest climate science.
• Pilot new technologies that will be deployed at scale after 2030 – Among these are carbon capture of many varieties including direct air capture, carbon storage and utilization including creating drop-in replacement fuels through methanation, and generation IV nuclear technologies.
• No new infrastructure to transport fossil fuels – Consumption of every fossil fuel declines in a pathway to 350 ppm. Thus, new infrastructure to transport fossil fuels run a high risk of either becoming stranded or locking in a higher emission pathway. Some infrastructure built for a 20th century energy system is still useful in the 21st century such as natural gas storage and transmission pipelines and should be maintained.
• Start building carbon capture on industrial facilities – Carbon capture on industrial processes should be prioritized because many processes result in higher CO2 concentrations than post-combustion capture on electricity generation and operate at

© 2019 by Evolved Energy Research 66

higher utilization factors, reducing cost, and because some industrial processes offer no ready alternatives making this type of carbon capture a necessary long-term strategy.

2030s

• Large renewables push – The 2030s is when the bulk of new renewable generation is built. Renewable curtailment is a necessary if sometimes transient balancing solution while transmission is expanded, market rules with high variable generation mature, and other balancing solutions get built.
• Reach near 100% sales on key electric technologies – All new vehicle sales must become electric or zero carbon compatible, for example fuel cells or biodiesel for heavy equipment. Similar transitions must occur in buildings for heating and cooking equipment. In industry electric or dual-fuel equipment should be installed for process heating and steam production which can be called upon based on electric system conditions (i.e. they can utilize overgeneration).
• Start significant biofuel production in diesel & jet fuel – Diesel and jet fuel are two of the largest residual fuels after high electrification. Bio-fuels used as drop-in replacements for fossil are a major strategy for reducing emissions. In the 2030s both are beginning to be produced in significant quantities, often with carbon capture on the biorefineries.
• Large scale carbon capture on industrial facilities – This completes the carbon capture on industry begun in the 2020s. By the late 2030s the marginal carbon abatement cost exceeds the capture cost for most industrial processes making this a cost-effective measure to pursue. The main challenge becomes geographic mismatch between where industry is located and where CO2 is sequestered or used.
• Electrical energy storage for capacity – As fossil capacity retires, electric energy storage technologies are deployed at a modest scale for reliability and to assist with diurnal balancing between electricity supply and demand. The phrase ‘modest’ is used because energy storage technologies cannot cost effectively replace all types of other dispatchable generation without a major cost breakthrough in long duration storage. Just like in the 2020s, some new gas power plant capacity is needed. When the duration of need for dispatchable capacity is less than 8 hours, energy storage will most likely be the most cost-effective option, for anything longer than 8 hours, gas turbines are the cheapest option for the system.
• Fossil power plants with 100% capture – If competitive with renewables and nuclear, fossil power plants with pre-capture or oxy technologies should start to be deployed. It’s possible that CCS technologies in electricity are unable to compete with a combination of renewables and energy storage, in which case most carbon capture stays focused on industry and refining.

© 2019 by Evolved Energy Research 67

• Maintain nuclear – As in the previous decade, continue to maintain nuclear where safe to do so.

2040s

• Reach near 100% stock penetration on electric technologies – The key building heating and transportation technologies that approached 100% new technology adoption in the 2030s have lifetimes of 10-15 years; and therefore, stock shares of these technologies should approach 100% in the 2040s based on natural replacement.
• Deploy circular carbon economy – In the 2040s synthetic fuel production & direct air capture (DAC) become important strategies to further reduce emissions and to balance a system with high renewables. The degree to which each are needed is dependent on many factors including: how much sustainable biomass can be produced, how much electrification is achieved, how cheap and efficient can DAC become, how much annual sequestration potential is there and at what cost, and how cheap are renewables and competing balancing strategies?
• Maintain/grow renewables together with new flexible loads – As synthetic fuel industrial loads grow it gives a new tool for balancing a grid composed of large amounts of variable generation. This, in turn, allows for further increases in renewables at low cost. Distributed fuel production also avoids the need for some new transmission.
• Replace nuclear at the end of its lifetime – As generation three nuclear retires, it should be replaced with fourth generation nuclear technologies if possible. By the 2040s renewables make up most of all electricity generation. Because of high marginal balancing costs when installing further wind and solar, dispatchable zero-carbon technologies such a nuclear are highly competitive.
• Fully deploy biofuels including bio-energy with carbon capture – Biofuel production and deployment reaches its limit in the 2040s. Biofuels find only marginal application in electricity because of higher value uses in transport and industry. Those industrial applications that can also deploy carbon capture allow opportunities of negative life- cycle emissions. Carbon capture on biofuel refining becomes an important technology.

© 2019 by Evolved Energy Research 68

Bibliography

Ashe, Mary, Dan Chwastyk, Caroline de Monasterio, Mahima Gupta, and Mika Pegors. 2012. “2010 U.S. Lighting Market Characterization.” U.S. Department of Energy; Office of Energy Efficiency and Renewable Energy; Building Technologies Program; Solid-State Lighting Program. https://www1.eere.energy.gov/buildings/publications/pdfs/ssl/2010- lmc-final-jan-2012.pdf.

Boer, Eelco den, Sanne Aarnink, Florian Kleiner, and Johannes Pagenkopf. 2013. “Zero Emissions Trucks: An Overview of State-of-the-Art Technologies and Their Potential.” CE Delft. https://www.theicct.org/sites/default/files/publications/CE_Delft_4841_Zero_emission s_trucks_Def.pdf.

Brooker, Aaron, Jeffrey Gonder, Sean Lopp, and Jacob Ward. 2015. “ADOPT: A Historically Validated Light Duty Vehicle Consumer Choice Model.” In . https://doi.org/10.4271/2015-01-0974.

Capros, P., E. Dimopoulou, S. Evangelopoulou, T. Fotiou, M. Kannavou, P. Siskos, and G. Zazias. 2018. “Technology Pathways in Decarbonisation Scenarios.” Advanced System Studies for Energy Transition. https://ec.europa.eu/energy/sites/ener/files/documents/2018_06_27_technology_path ways_-_finalreportmain2.pdf.

Congress of the U.S., Office of Technology Assessment. 1991. Changing by Degrees: Steps to Reduce Greenhouse Gases. Congress of the U.S., Office of Technology Assessment.

Dentz, Jordan, David Podorson, and Kapil Varshney. 2014. “Mini-Split Heat Pumps Multifamily Retrofit Feasibility Study.” U.S. Department of Energy; Office of Energy Efficiency and Renewable Energy. https://www.nrel.gov/docs/fy14osti/61620.pdf.

Eurek, Kelly, Wesley Cole, David Bielen, Nate Blair, Stuart Cohen, Bethany Frew, Jonathan Ho, et al. 2017. “Regional Energy Deployment System (ReEDS) Model Documentation: Version 2016.” NREL/TP-6A20-67067. Golden, Colorado: National Renewable Energy Laboratory. https://www.nrel.gov/docs/fy17osti/67067.pdf.

Fargione, Joseph E., Steven Bassett, Timothy Boucher, Scott D. Bridgham, Richard T. Conant, Susan C. Cook-Patton, Peter W. Ellis, et al. 2018. “Natural Climate Solutions for the United States.” Science Advances 4 (11): eaat1869. https://doi.org/10.1126/sciadv.aat1869.

Fulton, Lew, and Marshall Miller. 2015. “Strategies for Transitioning to Low-Emissions Trucks.” Davis, California: UC Davis Institute of Transportation Studies. https://steps.ucdavis.edu/files/06-11-2015-STEPS-NCST-Low-carbon-Trucks-in-US-06- 10-2015.pdf.

© 2019 by Evolved Energy Research 69

G. del Alamo, J. Sandquist, B.J. Vreugdenhil, G. Aranda Almansa, and M. Carbo. 2015. “Implementation of Bio-CCS in Biofuels Production.” IEA Bioenergy Task 33 Special Project ISBN 978-1-910154-44-1. IEA Bioenergy. https://www.ieabioenergy.com/wp- content/uploads/2018/08/Implementation-of-bio-CCS-in-biofuels-production_final.pdf.

Griscom, Bronson W., Justin Adams, Peter W. Ellis, Richard A. Houghton, Guy Lomax, Daniela A. Miteva, William H. Schlesinger, et al. 2017. “Natural Climate Solutions.” Proceedings of the National Academy of Sciences 114 (44): 11645–50. https://doi.org/10.1073/pnas.1710465114.

Hansen, James et al. 2008. “Target Atmospheric CO2: Where Should Humanity Aim?” Open Atmos. Sci. J. vol. 2, pp. 217-231 arXiv:0804.1126v3

Hansen, James, Pushker Kharecha, Makiko Sato, Valerie Masson-Delmotte, Frank Ackerman, David J. Beerling, Paul J. Hearty, et al. 2013. “Assessing ‘Dangerous Climate Change’: Required Reduction of Carbon Emissions to Protect Young People, Future Generations and Nature.” PLOS ONE 8 (12): e81648. https://doi.org/10.1371/journal.pone.0081648.

Hansen, James, et al. “Ice melt, sea level rise and superstorms: evidence from paleoclimate data, climate modeling, and modern observations that 2 C global warming could be dangerous.” Atmospheric Chemistry and Physics16.6 (2016): 3761-3812.

Hansen, James, Makiko Sato, Pushker Kharecha, Karina Von Schuckmann, David J. Beerling, Junji Cao, Shaun Marcott, Valerie Masson-Delmotte, Michael J. Prather, and Eelco J. Rohling. 2017. “Young People’s Burden: Requirement of Negative CO2 Emissions.” Earth Syst. Dynam., 8, 577-616, doi:10.5194/esd-8-577-2017.

Intergovernmental Panel on Climate Change. 2018. “Global Warming of 1.5C.” IPCC Special Report. United Nations. https://www.ipcc.ch/pdf/special-reports/sr15/sr15_draft.pdf.

International Energy Agency GHG Programme. 2017. “Techno – Economic Evaluation of SMR Based Standalone (Merchant) Hydrogen Plant with CCS.” International Energy Agency. https://ieaghg.org/exco_docs/2017-02.pdf.

Jadun, Paige, Colin McMillan, Matteo Muratori, Laura Vimmerstedt, and Trieu Mai. 2017. “Electrification Futures Study: End-Use Electric Technology Cost and Performance Projections through 2050.” NREL/TP-6A20-70485. Golden, Colorado: National Renewable Energy Laboratory. https://www.nrel.gov/docs/fy18osti/70485.pdf.

Johnson, Timothy L., Joseph F. DeCarolis, Carol L. Shay, Daniel H. Loughlin, Cynthia L. Gage, and Samudra Vijay. 2006. “MARKAL Scenario Analyses of Technology Options for the Electric Sector: The Impact on Air Quality.” EPA/600/R-06/114. Washington D.C.: United States Environmental Protection Agency Office of Research and Development. https://nepis.epa.gov/Exe/ZyPDF.cgi/P10089YQ.PDF?Dockey=P10089YQ.PDF.

Keith, David W., Geoffrey Holmes, David St. Angelo, and Kenton Heidel. 2018. “A Process for Capturing CO2 from the Atmosphere.” Joule 2 (8): 1573–94. https://doi.org/10.1016/j.joule.2018.05.006.

© 2019 by Evolved Energy Research 70

Kuramochi, Takeshi, Andrea Ramírez, Wim Turkenburg, and André Faaij. 2012. “Comparative Assessment of CO2 Capture Technologies for Carbon-Intensive Industrial Processes.” Progress in Energy and Combustion Science 38 (1): 87–112. https://doi.org/10.1016/j.pecs.2011.05.001.

Langholtz, M. H., B. J. Stokes, and L. M. Eaton. 2016. “2016 Billion-Ton Report: Advancing Domestic Resources for a Thriving Bioeconomy.” DOE/EE-1440, ORNL/TM-2016/160, 1271651. https://doi.org/10.2172/1271651.

Lashof, Daniel, and Dennis A. Tirpak. 1990. “Policy Options for Stabilizing Global Climate.” U.S Environmental Protection Agency EPA–21P-2003.1. http://inis.iaea.org/Search/search.aspx?orig_q=RN:23016941.

Le Quéré, Corinne, Robbie M. Andrew, Pierre Friedlingstein, Stephen Sitch, Julia Pongratz, Andrew C. Manning, Jan Ivar Korsbakken, et al. 2018. “Global Carbon Budget 2017.” Earth System Science Data 10 (1): 405–48. https://doi.org/10.5194/essd-10-405-2018.

National Renewable Energy Laboratory. 2017. “Annual Technology Baseline.” 2017. https://atb.nrel.gov/electricity/2017/.

Navigant. 2014. “Updated Buildings Sector Appliance and Equipment Costs and Efficiencies.” U.S. Energy Information Administration. https://www.eia.gov/analysis/studies/buildings/equipcosts/archive/2015/pdf/full.pdf.

Risky Business Project. 2016. “From Risk To Return: Investing in a Clean Energy Economy.” https://riskybusiness.org/site/assets/uploads/sites/5/2016/10/RiskyBusiness_FromRisk ToReturn.pdf.

Rogelj, J., G. Luderer, R. C. Pietzcker, E. Kriegler, M. Schaeffer, V. Krey, and K. Riahi. 2015. “Energy System Transformations for Limiting End-of-Century Warming to below 1.5 °C.” Nature Climate Change 5 (June): 519–27. https://doi.org/Rogelj J <http://pure.iiasa.ac.at/view/iiasa/254.html>, Luderer G, Pietzcker RC, Kriegler E, Schaeffer M, Krey V <http://pure.iiasa.ac.at/view/iiasa/166.html>, & Riahi K <http://pure.iiasa.ac.at/view/iiasa/250.html> (2015). Energy system transformations for limiting end-of-century warming to below 1.5 °C. Nature Climate Change 5 (6): 519- 527. DOI:10.1038/nclimate2572 <https://doi.org/10.1038/nclimate2572>.

State of California. 2018. “Exxecutive Order B-55-18 to Achieve Carbon Neutrality.” https://www.gov.ca.gov/wp-content/uploads/2018/09/9.10.18-Executive-Order.pdf.

TA Engineering Inc. 2012. “TRUCK5.1: Heavy Vehicle Market Penetration Model Documentation.” National Petroleum Council. https://www.npc.org/reports/HDV/TRUCK5.1_Model_Documentation-final.pdf.

U.S. Department of Energy: National Energy Technology Laboratory. 2017. “FE/NETL CO2 Saline Storage Cost Model.” U.S. Department of Energy: National Energy Technology Laboratory. https://edx.netl.doe.gov/dataset/fe-netl-co2-saline-storage-cost-model- 2017/resource/154e5905-e84b-4f80-8e6e-89e0228f28c5.

© 2019 by Evolved Energy Research 71

U.S. Energy Information Administration. 2012. “Commercial Building Energy Consumption Survey.” Commercial Building Energy Consumption Survey. 2012. https://www.eia.gov/consumption/commercial/.

———. 2013. “Residential Energy Consumption Survey (RECS) 2009 Technical Documentation- Summary.” Washington D.C.: U.S. Energy Information Administration. https://www.eia.gov/consumption/residential/methodology/2009/pdf/techdoc- summary010413.pdf.

———. 2015. “Annual Energy Outlook 2015.” U.S. Energy Information Administration. https://www.eia.gov/outlooks/aeo/pdf/0383(2015).pdf.

———. 2017. “Annual Energy Outlook 2017.” U.S. Energy Information Administration. https://www.eia.gov/outlooks/aeo/pdf/0383(2017).pdf.

———. 2018. “State Electricity Profiles.” https://www.eia.gov/electricity/state/.

U.S. Environmental Protection Agency. 2018. “Emissions & Generation Resource Integrated Database (EGRID).” February 5, 2018. https://www.epa.gov/energy/emissions- generation-resource-integrated-database-egrid.

U.S. Global Change Research Program. 2017. “Fourth National Climate Assessment.” Climate Science Special Report. U.S. Global Change Research Program. https://science2017.globalchange.gov/.

Veron, J. E. N., et al. “The coral reef crisis: the critical importance of< 350 ppm CO2.” Marine pollution bulletin 58.10 (2009): 1428-1436.

Williams, James H., Andrew DeBenedictis, Rebecca Ghanadan, Amber Mahone, Jack Moore, William R. Morrow, Snuller Price, and Margaret S. Torn. 2012. “The Technology Path to Deep Greenhouse Gas Emissions Cuts by 2050: The Pivotal Role of Electricity.” Science 335 (6064): 53–59. https://doi.org/10.1126/science.1208365.

Williams, James H., Benjamin Haley, Fritz Kahrl, Jack Moore, Andrew D. Jones, Margaret S. Torn, Haewon McJeon. 2014. Pathways to deep decarbonization in the United States. Energy and Environmental Economics, Inc. and the Deep Decarbonization Pathways Project. Revision with technical supplement, Nov 16, 2015. http://usddpp.org/usddpp-reports/

Williams, James H., Benjamin Haley, and Ryan Jones. 2015. Policy Implications of Deep Decarbonization in the United States. Energy and Environmental Economics, Inc. and the Deep Decarbonization Pathways Project. http://usddpp.org/usddpp-reports/

© 2019 by Evolved Energy Research 72

Technical Supplement

Figure 25 350 – Low Biomass Emissions Reductions Breakdown

© 2019 by Evolved Energy Research 73

Figure 26 350 – Low Electrification Emissions Reductions Breakdown

© 2019 by Evolved Energy Research 74

Figure 27 350 – Low Land NETS Emissions Reductions Breakdown

© 2019 by Evolved Energy Research 75

Figure 28 350 – No New Nuclear Emissions Reductions Breakdown

© 2019 by Evolved Energy Research 76

Figure 29 350 – No Tech NETS Emissions Reductions Breakdown

© 2019 by Evolved Energy Research 77

Figure 30 Net Costs by Sector

© 2019 by Evolved Energy Research 78

Figure 31 Four Pillars

© 2019 by Evolved Energy Research 79

Figure 32 Final Energy Demand

© 2019 by Evolved Energy Research 80

Appendix

1. Scenario Descriptions

EnergyPATHWAYS scenarios consist of combinations of energy demand scenarios as well as emissions targets and other constraints applied to the entire energy economy. In this framework, we have three energy demand scenarios – Baseline, Base 350, and Low Electrification 350 – that are used for our seven energy economy scenarios. These relationships are described in the below table.

Table 2 Energy demand and energy economy scenarios

1.1 Energy Demand Scenario Descriptions

1.1.1 Baseline

This represents an assumption of stasis in terms of technology adoption. For example, gas storage water heaters in the residential sector are replaced with newer gas storage water heaters. These new technology vintages have changing parameters of cost and efficiency but

© 2019 by Evolved Energy Research 81

represent the same technology type and class (i.e. they use the same fuel and represent the same level of relative efficiency in the market).

1.1.2 Base 350

This scenario assumes rapid adoption of electrification technologies and high efficiency technologies where the end-use is already electric (i.e. refrigeration) or where complete electrification is infeasible. Adoption rates of these technologies accelerates through 2030, with the stock of these technologies lagging but making steady progress through 2050.

1.1.3 Low Electrification 350

This scenario assumes difficulty in inducing electrification of end-uses. Instead of adoption rates peaking by 2030, the adoption of these technologies is much slower, with peak adoption rates not being achieved until the 2050 timeframe. This slower rate of adoption leaves much more fuel combustion in intermediate years and also represents an incomplete electrification process by 2050, as much of the existing fuel combustion stock is still in service.

1.2 Demand-Side Mitigation Measures

1.2.1 Stock Rollover

The tables below show the stock shares (Table 3) and sales shares (Table 4) for three demand technology groups (Electrified Techs; HE Techs; Other Techs).13 The demand-side consists of over 380 technologies across all subsectors or end-uses, but we aggregate here for presentation purposes to show broader trends in our input values. The stock shares shown are determined by stock rollover assumptions specified in the measure for each technology as well as the lifetimes of the infrastructure and the methodology described in section 4.2.1.2.

13 Electrified Techs == Technologies that use electricity for end-uses where other fuels are competitors (i.e. water heating but not lighting); HE Techs == High efficiency technologies; Other Techs == Technologies not categorized as Electrified Techs or HE Techs.

© 2019 by Evolved Energy Research 82

Table 3 Stock shares

© 2019 by Evolved Energy Research 83

© 2019 by Evolved Energy Research 84

© 2019 by Evolved Energy Research 85

© 2019 by Evolved Energy Research 86

© 2019 by Evolved Energy Research 87

© 2019 by Evolved Energy Research 88

Table 4 Sales shares

© 2019 by Evolved Energy Research 89