Deployment, Design, and Commercialization of Carbon-Negative Energy Systems

Home , Daniel Kammen, Panicum virgatum

Daniel Lucio Sanchez

A dissertation submitted in partial satisfaction of the

requirements of the degree of

Doctor of Philosophy

Energy and Resources

in the

Graduate Division

of the

University of California, Berkeley

Committee in charge:

Professor Daniel M. Kammen, Chair Professor Duncan S. Callaway Professor Solomon M. Hsiang

Fall 2015

Abstract

Deployment, Design, and Commercialization of Carbon-Negative Energy Systems

Daniel Lucio Sanchez

Doctor of Philosophy in Energy and Resources

University of California, Berkeley

Professor Daniel M. Kammen, Chair

Climate change mitigation requires gigaton-scale carbon dioxide removal technologies, yet few examples exist beyond niche markets. This dissertation informs large-scale implementation of bioenergy with carbon capture and sequestration (BECCS), a carbon- negative energy technology. It builds on existing literature with a novel focus on deployment, design, commercialization, and communication of BECCS.

BECCS, combined with aggressive renewable deployment and fossil emission reductions, can enable a carbon-negative power system in Western North America by 2050, with up to 145% emissions reduction from 1990 levels. BECCS complements other sources of renewable energy, and can be deployed in a manner consistent with regional policies and design considerations. The amount of biomass resource available limits the level of fossil CO2 emissions that can still satisfy carbon emissions caps. Offsets produced by BECCS are more valuable to the power system than the electricity it provides. Implied costs of carbon for BECCS are relatively low (~$75/ton CO2 at scale) for a capital-intensive technology.

Optimal scales for BECCS are an order of magnitude larger than proposed scales found in existing literature. Deviations from optimal scaled size have little effect on overall systems costs – suggesting that other factors, including regulatory, political, or logistical considerations, may ultimately have a greater influence on plant size than the techno- economic factors considered.

The flexibility of thermochemical conversion enables a viable transition pathway for firms, utilities and governments to achieve net-negative CO2 emissions in production of electricity and fuels given increasingly stringent climate policy. Primary research, development (R&D), and deployment needs are in large-scale biomass logistics, gasification, gas cleaning, and geological CO2 storage. R&D programs, subsidies, and policy that recognize co- conversion processes can support this pathway to commercialization. Here, firms can embrace a gradual transition pathway to deep decarbonization, limiting economic dislocation and increasing transfer of knowledge between the fossil and renewable sectors.

Global cumulative capital investment needs for BECCS through 2050 are over $1.9 trillion (2015$, 4% real interest rate) for scenarios likely to limit global warming to 2 °C. This scenario envisions deployment of as much as 24 GW/yr of BECCS by 2040 in the electricity sector. To achieve theses rates of deployment within 15-20 years, governments and firms must commit to research, development, and deployment on an unprecedented scale.

Three primary issues complicate emissions accounting for BECCS: cross-sector CO2 accounting, regrowth, and timing. Switchgrass integration decreases lifecycle greenhouse gas impacts of co-conversion systems with CCS, across a wide range of land-use change scenarios. Risks at commercial scale include adverse effects on food security, land conservation, social equity, and biodiversity, as well as competition for water resources. This dissertation argues for an iterative risk management approach to BECCS sustainability, with standards being updated as more knowledge is gained through deployment. Sustainability impacts and public opposition to BECCS may be reduced with transparent measurement and communication.

Commercial-scale deployment is dependent on the coordination of a wide range of actors, many with different incentives and worldviews. Despite this problem, this dissertation challenges governments, industry incumbents, and emerging players to research, support, and deploy BECCS.

2 Table of Contents

List of figures ...... iii List of tables ...... v Acknowledgements ...... vi Chapter I. Background and motivation ...... 1 1. Negative emissions and climate change mitigation ...... 1 1.1. Reduce, geoengineer, or remove: three approaches to climate change mitigation ...... 1 1.2. The scale of negative emissions in climate change mitigation ...... 4 1.3. Negative emissions technologies and techniques ...... 7 1.4. Permanence of negative emissions technologies ...... 8 2. Bioenergy with carbon capture and sequestration ...... 11 2.1. Technology description ...... 12 2.2. Process engineering and techno-economic assessment ...... 13 2.3. Contribution to long-term climate change mitigation ...... 17 2.4. Commercial deployment ...... 18 2.5. Research needs for implementation ...... 19 3. Contributions ...... 21 Chapter II. Deployment of carbon-negative energy systems ...... 23 1. Preface ...... 23 2. Excerpt from Biomass enables the transition to a carbon-negative power system across western North America (Sanchez et al., 2015a) ...... 23 2.1. Main text ...... 23 2.2. Materials and methods ...... 30 2.2.1. Biomass technologies ...... 30 2.2.2. Biomass supply ...... 31 2.2.3. Biomass cofiring and modeled scenarios ...... 31 2.2.4. CCS reservoirs and transportation ...... 31 2.2.5. Scenario development ...... 32 2.3. Additional methods and information ...... 32 2.3.1. Overview of the SWITCH model ...... 32 2.3.2. Materials and methods ...... 33 2.3.3. Additional results from core scenarios ...... 40 2.3.4. Additional biomass cofiring results ...... 48 2.3.5. Sensitivity results for carbon-negative power systems ...... 49 2.3.6. Comparison of net energy of bioelectricity and biofuels ...... 52 3. Excerpt from Emissions accounting for biomass energy with CCS (Sanchez et al., 2015b) ...... 53 4. Conclusions ...... 55 Chapter III. Communication of carbon-negative energy systems ...... 57 1. Preface ...... 57 2. Excerpt from Removing harmful greenhouse gases from the air using energy from plants (Sanchez and Kammen 2015) ...... 58 2.1. Main text ...... 59 2.2. Glossary ...... 65 i

3. Conclusions ...... 66 Chapter IV. Design of carbon-negative energy systems ...... 67 1. Preface ...... 67 2. Excerpt from Optimal Scale of Bioenergy with Carbon Capture and Storage (BECCS) Facilities (Sanchez and Callaway) ...... 67 2.1. Introduction ...... 67 2.2. Materials and methods ...... 69 2.2.1. Problem statement ...... 69 2.2.2. Multiple facility case ...... 69 2.2.3. Single facility case ...... 71 2.2.4. Data sources and parameters ...... 71 2.2.5. Solution method ...... 75 2.3. Results and discussion ...... 75 2.3.1. Single facility case ...... 75 2.3.2. Multiple facility case ...... 80 2.4. Implications ...... 81 3. Conclusions ...... 82 Chapter V. Commercialization of carbon-negative energy systems ...... 83 1. Preface ...... 83 2. Excerpt from A commercialization strategy for carbon-negative energy (Sanchez and Kammen, In Press) ...... 83 2.1. Introduction ...... 83 2.2. Flexibility as a virtue ...... 84 2.3. Research and policy needs ...... 86 2.4. Uncertainty clouds deployment at scale ...... 88 2.5. An agenda for transition ...... 89 3. Conclusions ...... 89 Chapter VI. Conclusions and implications for policy ...... 91 1. Net-negative emissions energy systems ...... 91 2. Value of BECCS in climate change mitigation ...... 91 3. Incentives for scale ...... 92 4. Flexibility and commercialization ...... 92 5. Communicating BECCS ...... 92 6. Sustainability limitations ...... 93 7. Scale of the transition ...... 94 8. Conclusion ...... 94 Chapter VII. References ...... 96

List of figures

Figure 1. Schematic of climate change mitigation techniques...... 2 Figure 2. Framework for climate policy design ...... 5 Figure 3. Historical and projected CO2 emissions through 2100, based on IPCC 2014a ...... 6 Figure 4. Use of globally aggregated annual BECCS in Pg C yr−1 for each of four RCPs scenarios ...... 7 Figure 5. Pathways for carbon dioxide removal ...... 8 Figure 6. Representation of BECCS process ...... 13 Figure 7. Routes to biomass with CO2 capture ...... 14 Figure 8. Process diagram for biomass IGCC-CCS facility designed for maximum capture of CO2 ...... 15 Figure 9. Process diagram for the coal/biomass to Fischer-Tropsch liquid (FTL) synthetic fuels CO2 capture and storage process ...... 16 Figure 10. Carbon dioxide emission pathways until 2100 and the extent of net negative emissions and bioenergy with carbon capture and storage (BECCS) in 2100 ...... 18 Figure 11. The four components of consistent negative emissions narratives ...... 20 Figure 12. Continuum of research tasks to inform large-scale implementation of BECCS ...... 22 Figure 13. Supply Curve of Available Solid Biomass post-2030 ...... 25 Figure 14. (a) Generation (105 GWh, gross), and cost of electricity (2013$/MWh) in 2050. (b) Yearly carbon emissions (MtCO2/yr) in 2050...... 27 Figure 15. Hourly dispatch in 2050 in the -145% case ...... 28 Figure 16. Carbon caps defined for -86% (86% reduction), -105% (105% reduction), -120% (120% reduction), and -145% (145% reduction) cases through 2050...... 39 Figure 17. Cost of electricity in each period for core scenarios ...... 41 Figure 18. Carbon costs (2013$/tCO2) in each period across core scenarios ...... 42 Figure 19. Storage installed (GW) in 2050 for each core scenario ...... 43 Figure 20. Installed biomass, coal, and gas CCS capacity in each load area in 2050 in the (a) -86% case and (b) -145% case...... 45 Figure 21. Cost of electricity (2013$/MWh) in 2050...... 46 Figure 22. Yearly biomass consumption (PJ) by period in core scenarios...... 47 Figure 23. Histogram of biomass prices in 2050 in -145% case in modeled load areas ...... 47 Figure 24. Hourly dispatch in 2050 in the -86% No Biomass case ...... 48 Figure 25. Biomass cofiring installation in each period for the -86% case and -145% case, including both retrofits of existing plants and co-installation on new plants ...... 49 Figure 26. Average generation (GW) in 2050 in sensitivity scenarios (-120%)...... 50 Figure 27. Yearly carbon emissions (MtCO2/yr) in 2050 under sensitivity scenarios (-120%)...... 51 Figure 28. An artist’s depiction of bioenergy with carbon capture and sequestration (BECCS)...... 59 Figure 29. The world’s carbon dioxide emissions for “Business as Usual” and “Climate Change Mitigation” scenarios...... 61 Figure 30. Diagram of (a) a bioenergy power plant, and (b) a bioenergy with carbon capture and sequestration (BECCS) power plant ...... 62 Figure 31. A supply curve for biomass ...... 64 Figure 32. (a) Electricity produced (average generation in GW, an amount of energy over 1 year), and (b) CO2 emissions (MtCO2/year, an amount of CO2 emitted every year) in 2050 for an electricity system in Western North America with negative emissions ...... 65 Figure 33. County supply limit (105 bone dry ton biomass), and potential facility locations in Illinois for base case scenario ...... 72 Figure 34. (a) Normalized biomass consumption (values between 0 and 1), and (b) biomass market clearing price ($/BDT) for Illinois counties for base case in Decatur (scale = 2713 MW)...... 77

iii

Figure 35. Difference in cost (% difference from optimal objective) and implied cost of carbon ($/tCO2) for different fixed scale sizes in Decatur (base case) ...... 78 Figure 36. Sensitivity of base case to biomass availability, transportation costs, capital cost scaling parameters, location, and technology type ...... 79 Figure 37. Scaled size results among potential facility locations at different scaling exponents, α, for a scaled size goal of 2500 MW...... 80 Figure 38. Flow diagrams for carbon capture and storage processes ...... 85 Figure 39. Lifecycle greenhouse gas carbon intensity for IGCC-CCS facilities ...... 88

List of tables

Table 1. Overview of general differences between carbon dioxide removal and geoengineering proposals ...... 4 Table 2. Biomass technology and biomass cofiring-enabled coal technology in the SWITCH model 34 Table 3. Specifications for selected biomass technologies, in 2013$ ...... 34 Table 4. CO2 emissions factors for various biomass, coal, and gas technologies with and without CCS ...... 35 Table 5. Biomass Supply in the SWITCH model for years 2030 and beyond ...... 37 Table 6. Variables for biomass cofiring dispatch, capacity, and flexibility constraints in the SWITCH model...... 37 Table 7. Key parameters for modeled scenarios...... 40 Table 8. Comparison of bioelectricity production options to cellulosic ethanol ...... 53 Table 9. (a) Objective function and (b) constraints for total cost problem (multiple facility case) ... 71

Acknowledgements

Dissertations are collective efforts. First, I would like to thank the outstanding advisors I have had at the University of California-Berkeley. My chair, Daniel Kammen, has been a strong promoter of my career and interests. Duncan Callaway has provided a wealth of academic support. Others, such as Hanna Breetz, Michael O’Hare, Solomon Hsiang, and Timothy Lipman have provided academic and personal support throughout my graduate studies. Other advisors during summer experiences, such as Florian Kraxner at the International Institute for Applied Systems Analysis, Cheryl Martin at the Advanced Research Projects Agency-Energy, and Nika Rogers at the California Public Utilities Commission have provided invaluable career direction.

Second, I owe much to my colleagues at Berkeley, especially in the Renewable and Appropriate Energy Laboratory and Energy and Resources Group. I extend special thanks to the SWITCH modeling team, including Josiah Johnston, Ana Mileva, James Henry Nelson, Juan Pablo Carvallo, Patricia Hidalgo Gonzalez, Anne-Perrine Avrin, and Diego Barido.

Third, I would like to thank my friends at Berkeley, many of whom have also served as academic and personal advisors. These include Noah Deich, Danny Cullenward, John Romankiewicz, James Matthew Lucas, and Rebecca Hernandez. I owe additional thanks to many colleagues from the Berkeley Energy and Resources Collaborative.

Finally, I thank my family, Christine, Israel, and David, for their support during my graduate studies.

Chapter I. Background and motivation

Climate change, a change in the statistical distribution of weather patterns over an extended period of time, is expected to contribute to sea level rise, changing precipitation patterns, ocean acidification, species extinction, and more frequent extreme weather events. Recent climate change, also known as anthropogenic global warming, describes the increase in average temperature since the industrial revolution, which is primarily due to human influence. Greenhouse gases (GHGs) absorb and remit terrestrial radiation emitted by Earth’s surface, causing an increase in global average temperatures (IPCC 2014a).

The principal greenhouse gas is carbon dioxide (CO2), which is released as by-product of burning fossil fuels (such as oil, gas, and coal), land use changes, and industrial processes. Other greenhouse gases include water vapor (H2O), nitrous oxide (N2O), methane (CH4), ozone (O3), hydrofluorocarbons (HFCs), and perfluorocarbons (PFCs). Reduction of CO2 emissions, along with other greenhouse gases, is an important strategy to reduce future climate change. Interventions by humans to reduce or reverse climate change are known as climate change mitigation.

1. Negative emissions and climate change mitigation

This Section will briefly describe three methods for climate change mitigation: reduction of CO2 emissions, solar radiation management, and CO2 removal (also know as carbon dioxide removal [CDR], or negative emissions). It will focus specifically on negative emissions, including the scale of necessary negative emissions for climate change mitigation, and negative emissions technologies. 1.1. Reduce, geoengineer, or remove: three approaches to climate change mitigation

As discussed previously, CO2 and other greenhouse gases trap heat in earth’s atmosphere, causing climate change (Figure 1 (a)). In general, there are three approaches to climate change mitigation. These approaches are represented in simplified form in Figure 1. The approaches described in this Section are meant to be broadly representative of options for mitigation rather than comprehensive. They are loosely based on distinctions introduced by the U.S. National Academy of Sciences (NAS) in 2015 in two reports (Climate Intervention: Reflecting Sunlight to Cool Earth 2015; Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015).

First, humans can choose to reduce emissions of greenhouse gases. Reducing emissions can proceed by a variety of mechanisms: these include, but are not limited to, reducing demand for carbon-intensive services (including provision of energy), substitution of lower-carbon processes for higher-emitting processes, or capture and destruction or sequestration of GHGs prior to atmospheric release. In the electricity sector, these actions are generally characterized as energy efficiency, electrification of end uses, and substitution of lower- 1 carbon energy sources (such as renewable energy or fossil fuel systems with carbon capture and sequestration [CCS]) in place of existing fossil fuel power plants (J. H. Nelson 2013; J. H. Williams et al. 2012). A schematic for emissions reduction is shown in Figure 1 (b). Reducing GHG emissions is the subject of a wide range of existing efforts, including academic research, activism, government policy, and small-scale action.

Figure 1. Schematic of climate change mitigation techniques. (a) Climate change is caused in part by CO2 emissions from energy infrastructure. Responses include (b) reduction of CO2 emissions, (c) geoengineering via solar radiation management, and (d) carbon dioxide removal.

Second, humans can choose to manage incoming solar radiation to reduce earth’s temperature. This approach is commonly called geoengineering, or more specifically albedo modification, to describe techniques that seek to enhance the reflectivity of the planet to reduce global temperature (Climate Intervention: Reflecting Sunlight to Cool Earth 2015). Albedo modification techniques mask the effects of greenhouse warming; they do not reduce greenhouse gas concentrations. According to the NAS:

Modeling studies indicate that significant cooling, equivalent in amplitude to the warming produced by doubling the CO2 concentration in the atmosphere, can be produced by the introduction of tens of millions of tons of aerosol-forming gases into the stratosphere… Modeling results also suggest that the benefits and risks will not be uniformly around the globe… [Costs] are small enough that decisions are

likely to be based primarily on considerations of potential benefits and risks, and not primarily on the basis of direct cost.

The risks and benefits of solar geoengineering, or solar radiation management (SRM), depend on assumptions about its implementation. Implementation may introduce new global risks, such as changes in precipitation and stratospheric ozone depletion (Keith and MacMartin 2015). A schematic for solar radiation management is shown in Figure 1 (c). Additional observations about geoengineering proposals are shown in Table 1.

Finally, humans can remove carbon dioxide from the atmosphere. The NAS defines CDR as: “intentional efforts to remove carbon dioxide from the atmosphere.” They also discuss several candidate CDR technologies, including: • land management strategies • accelerated weathering • ocean iron fertilization • bioenergy with carbon capture and sequestration [BECCS] • direct air capture and sequestration Because CDR techniques remove atmospheric CO2, these approaches are often characterized as ‘negative-emissions’ technologies. Barriers to CDR approaches include slow implementation, limited capacity, policy considerations, and high costs of presently available technologies (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015). A schematic for CDR is shown in Figure 1 (d). In contrast to the NAS, the IPCC includes CDR in their definition of mitigation (IPCC 2014a).

Removal and geoengineering are very different processes for climate change mitigation. CDR proposals are similar to CO2 reduction proposals, in that they address the root cause of climate change, are relatively low risk, and are both slower and more expensive than geoengineering. Table 1 summarizes the primary differences between CDR and geoengineering. The remainder of this Chapter will focus on CDR approaches.

Carbon dioxide removal proposals… Geoengineering (albedo modification) proposals… …address the cause of human-induced ... do not address cause of human-induced climate change (high atmospheric GHG climate change (high atmospheric GHG concentrations). concentrations).

... do not introduce novel global risks. ... introduce novel global risks.

... are currently expensive (or comparable ... are inexpensive to deploy (relative to to the cost of emission reduction). cost of emissions reduction).

... may produce only modest climate effects ... can produce substantial climate effects within decades. within years.

... raise fewer and less difficult issues with ... raise difficult issues with respect to

3 respect to global governance. global governance.

... will be judged largely on questions ... will be judged largely on questions related to cost. related to risk.

... may be implemented incrementally with ... could be implemented suddenly, with limited effects as society becomes more large-scale impacts before enough research serious about reducing GHG concentrations is available to understand the risks relative or slowing their growth. to inaction.

... require cooperation by major carbon ... could be done unilaterally. emitters to have a significant effect.

... for likely future emissions scenarios, if ... for likely future emissions scenarios, if abruptly terminated would have limited abruptly terminated would produce consequences significant consequences.

Table 1. Overview of general differences between carbon dioxide removal and geoengineering proposals. The risks, effects, costs, scale, and consequences of CDR are similar to GHG reduction techniques. Adapted from (Climate Intervention: Reflecting Sunlight to Cool Earth 2015).

1.2. The scale of negative emissions in climate change mitigation

Prior academic work has studied the potential scale of negative emissions necessary for climate change mitigation. Much work on global-scale emissions limitations is based on avoiding ‘dangerous anthropogenic interference’ (DAI) with our climate system, as defined in 1992 by the United National Framework Convention of Climate Change. Policymakers seeking to limit impacts of climate change must choose policies and goals for limiting GHG concentrations (Figure 2). This normative choice is complicated by multiple kinds of uncertainty in physical and social systems. As argued by Mann:

To properly define DAI, one must take into account issues that are not only scientific, but, as I have argued elsewhere, economic, political, and even ethical in nature. Defining DAI begs the question, for example, ‘‘Dangerous to whom?’’ It amounts to the tacit adoption of some level of risk, risk that will not be shared equally among all nations and people (Mann 2009).

Despite uncertainties, international diplomacy has focused on stopping global warming at 2 °C above pre-industrial levels, with more than 100 countries adopting such a goal. Meinshausen et al. attempt to capture the uncertainty in the climate system to quantify the likelihood that certain levels of cumulative emissions in 2050 will lead to mean global temperature change below 2°C. They find that limiting cumulative CO2 emissions over the years 2000–50 to 1,000 GtCO2 yields a 25% probability of warming exceeding 2°C —while a limit of 1,440 GtCO2 yields a 50% probability—given a representative estimate of the distribution of climate system properties (Meinshausen et al. 2009).

Temperature Radiative GHG Impacts Change Forcing concentration

Emissions budgets?

What we care What we about (maybe) can control

Figure 2. Framework for climate policy design. Policymakers seeking to limit impacts of climate change must choose policies and goals for limiting GHG concentrations. This normative choice is complicated by multiple kinds of uncertainty in physical and social systems. The construct presented in this figure, an emissions budget, represents a total amount of CO2 that may be emitted into the atmosphere over a given period of time (Meinshausen et al. 2009).

Working Group III of the Intergovernmental Panel on Climate Change (IPCC) developed representative concentration pathways (RCPs) describe four possible climate futures, all of which are considered possible depending on how much greenhouse gases are emitted in the future. The four RCPs, RCP2.6, RCP4.5, RCP6, and RCP8.5, are named after a possible range of radiative forcing values in the year 2100 relative to pre-industrial values (+2.6, +4.5, +6.0, and +8.5 W/m2, respectively) (IPCC 2014a).

RCPs are given additional technical detail based on evaluation in global energy-economy models, also known as integrated assessment models (IAMs) (Riahi et al. 2011; Vuuren et al. 2011). Global CO2 emissions through 2100 for some of these RCPs are shown in Figure 3. RCP8.5, defined in Figure 3 as ‘Business as Usual’, is a scenario of comparatively high greenhouse gas emissions, combining assumptions about high population and relatively slow income growth with modest rates of technological change and energy intensity improvements. This leads, in the long term, to high energy demand and GHG emissions. RCP2.6, defined as ‘Climate Change Mitigation’ in Figure 3, was defined to explore the possibility of keeping the global mean temperature increase below 2°C.

20 Climate Change Mitigation Mitigation without Negative Emissions 15 Business as Usual Emissions 2 (PgC / yr) / (PgC 10 Emissions Overshoot Net-negative

CO emissions 5

-5 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Year Figure 3. Historical and projected CO2 emissions through 2100, based on (IPCC 2014a). Business as Usual scenario corresponds to RCP8.5, while Climate Change Mitigation corresponds to RCP2.6. Mitigation without Negative Emissions is found by subtracting the amount of negative emissions from BECCS from the Climate Change Mitigation Scenario. Negative global emissions by 2100 entail “emissions overshoot,” where global CO2 emissions exceed a global carbon budget for a short period of time, only to be removed later. Negative emissions derived from (Kato and Yamagata 2014).

RCPs also include detail about the evolution of energy systems. RCP2.6 requires cumulative emissions of greenhouse gases from 2010 to 2100 to be reduced by 70% compared to a baseline scenario, requiring substantial changes in energy use and emissions of non-CO2 gases. This includes substantial deployment of negative emissions technologies, represented as BECCS and afforestation (Vuuren et al. 2011). BECCS deployment in RCP2.6 is responsible for as much as 3 PgC/yr (PgC is equivalent to GtC) in carbon dioxide removal by the year 2100. Current global emissions are roughly 10 GtC/yr. Global CO2 emissions without negative emissions are shown in Figure 3 as ‘Mitigation without Negative Emissions’. Negative global emissions by 2100 entail “emissions overshoot,” where global CO2 emissions exceed a global carbon limitation for a short period of time. Put another way, CDR strategies remove as much emissions as three ‘stabilization wedges,’ as proposed by Pacala and Socolow, by 2100 in RCP2.6 (Pacala and Socolow 2004).

Kato and Yamagata report global negative emissions from BECCS deployment for all four RCPs, as shown in Figure 4 (Kato and Yamagata 2014). The amount of cumulative BECCS during 2005–2100 is 161.7 Pg C, 40.4 Pg C, 4.5 Pg C, and 0 Pg C for RCP2.6, RCP4.5, RCP6.0 and RCP8.5, respectively. To place this in perspective, BECCS deployment in RCP2.6 is roughly 24 GW/yr in new installations by the year 2040, if installed as electricity infrastructure. Limiting temperature increases below 2°C—for instance, 1.5°C—requires even larger amounts of BECCS (Rogelj et al. 2015).

Figure 4. Use of globally aggregated annual BECCS in PgC yr−1 for each of four RCPs scenarios. The amount of cumulative BECCS during 2005–2100 is 161.7 PgC, 40.4 PgC, 4.5 PgC, and 0 PgC for RCP2.6, RCP4.5, RCP6.0 and RCP8.5, respectively. From (Kato and Yamagata 2014). 1.3. Negative emissions technologies and techniques

There are several techniques for carbon dioxide removal. To discuss these techniques, it is useful to distinguish between biological approaches and chemical approaches. Biological approaches enhance or manipulate natural sinks for CO2 to store more carbon. Chemical approaches, in contrast, apply chemical processes to capture and reliably sequester CO2 (Figure 5). These distinctions are not absolute: for example, biochar production relies of thermal treatment to create pyrogenic carbon for soil sequestration, while BECCS relies on sustainable biomass production before energy production from combustion.

Biological approaches include land management strategies, afforestation, and biochar. These approaches attempt to enhance the quantity and permanence of carbon storage above or below ground. Afforestation, the establishment of a forest or stand of trees in an area where there was no forest, increases above ground biomass stock. In contrast, biochar or other land management techniques aim to sequester more carbon in agricultural or grassland soils, and to increase the permanence of such sequestration. In general, biological approaches to CDR are considered cheaper than chemical approaches, but less permanent, as discussed in Section 1.4 of this Chapter (IPCC 2007). Priorities for future research include basic science to understand the permanence and magnitude of CDR from different techniques.

Chemical approaches include enhanced weathering, bioenergy with carbon capture and sequestration (BECCS), and direct air capture (DAC) and sequestration. Enhanced weathering uses mined minerals, such as silicates, to remove CO2 from the atmosphere. DAC and BECCS use absorption or adsorption to capture dilute or concentrated CO2, then compress, transport, and reliably sequester that CO2. The net energy of the process can be 7 positive or negative, depending on technique: DAC has negative net energy, while BECCS produces energy. Concentrated CO2 can also be used in carbon utilization schemes. While likely more permanent than biological approaches, chemical CDR approaches are generally more costly. The permanence of geological sequestration is discussed in greater detail in Section 1.4 of this Chapter.

Figure 5. Pathways for carbon dioxide removal. Pathways can be divided into biological approaches (shown in teal) or chemical approaches (shown in grey). Removal proceeds in three phases: capture, transformation, and sequestration. Figure from the Center for Carbon Removal (“What Is Carbon Removal?” 2015).

Cost and supply estimates for carbon removal approaches are fairly uncertain as most carbon removal technologies have not entered large-scale commercial deployment. In addition, more systems-level analysis of carbon removal approaches is needed to assess global scale potential. For example, Integrated Assessment Models lack key CDR techniques, relying primarily on BECCS and afforestation (Vuuren et al. 2011). According to the NAS, cost and lack of technical maturity are factors limiting the deployment of CDR strategies (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015). 1.4. Permanence of negative emissions technologies

The permanence of carbon removal technologies is also uncertain. In general, the long- term reliability of biological sequestration is more ambiguous than chemical sequestration. 8

In forests, it is unclear how a changing climate will affect sequestration: impacts depend both on the rate of climate change and the spatial scale of disturbance. Sequestration may also be reversed by accidental or intentional release of carbon by fire or future harvesting. On agricultural lands, no-till practices must be maintained without interruption to be used effectively as a form of long-term carbon sequestration. CDR from improved agronomy practices, such as cover cropping, likely levels off after roughly a decade. In contrast, biochar is a relatively stable form of organic carbon, and has fewer questions about permanence. However, its ultimate impact on emissions is not well understood, particularly should application to soils involve mixing and further oxidation of soil organic carbon (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015).

Chemical carbon dioxide removal approaches are generally much more permanent. The discussion and conclusions in the remainder of this Section are based on literature review and syntheses from three reports, the 2012 Global Energy Assessment, the 2014 NAS report on Carbon Dioxide Removal and Reliable Sequestration, and the 2005 IPCC Special Report on carbon dioxide capture and storage (Benson et al. 2012; Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015; Metz et al. 2005). Both BECCS and DAC ultimately rely on CO2 storage in geological formations.

Three lines of evidence indicate that well-selected and well-managed geological formations will have very high CO2 retention rates. First, hydrocarbon and CO2 reservoirs show that buoyant fluids can be trapped underground for millions of years. Natural analogues in the Bravo Dome of the United States, for example, demonstrate that geologic CO2 storage is possible for millions of years (Sathaye et al. 2014). Second, multiple trapping processes contribute to long-term retention and increase security of storage over time. Third, existing CO2 injection projects in diverse geologies have exhibited a high degree of containment. At least five commercial-scale CCS projects are operational today, with over 35 million tons of CO2 captured and stored between 1996 and 2012. The US Department of Energy’s Clean Coal Research, Development, and Demonstration Programs have sequestered over 11 million tons of CO2 as of 2015 (Office of Fossil Energy 2015).

There are two primary trapping mechanisms in geological formations: structural and stratigraphic. Structural trapping contains CO2 beneath a dome-shaped seal while stratigraphic trapping confines CO2 through changes in porosity and permeability. In addition, three secondary trapping mechanisms provide additional security: capillary, solubility, and mineral trapping. Capillary trapping traps residual gases inside the reservoir while solubility trapping involves the dissolution of CO2 within saline aquifers. Finally, mineral trapping involves the formation of stable carbonate minerals. These secondary trapping mechanisms continue to increase storage security as time goes on. They have been the subject of significant scientific research over the past decade, with hundreds of relevant publications (Benson and Cole 2008). The range of trapping contributions from each of these processes is highly site-specific.

Nevertheless, there are still risks in commercial-scale CO2 sequestration. Risks of CO2 storage are typically separated into two categories: 1) risks associated with the release of CO2 into the atmosphere (limiting effectiveness), and 2) health, safety, and environmental 9 risks associated with the local impacts of the storage operations and potential leakage (Benson et al. 2012). Because of operational and technological similarities, the probability and consequences of risks from geological storage are generally assumed to be similar to those of existing activities such as oil and gas production, natural gas storage, and acid gas injection (Metz et al. 2005). Here, risks are managed on a routine basis through a combination of operational controls, management oversight, monitoring, maintenance, regulatory oversight, and insurance. Similar practices are necessary to manage the risks of geological storage.

Allowable CO2 leakage rates from reservoirs consistent with climate change mitigation depend on the overall volume of CO2 sequestered, but are generally between .01% and .1%/yr (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015). This is well above achieved and envisioned leakage rates (Hepple and Benson 2004).

Benson et al. describe the conceptual process to limit health, safety, and environmental risks over the lifetime of a typical geological storage project, as illustrated in Figure 6:

Performance specifications, or acceptable risks, will be set by regulatory authorities. Projects will be designed to conform to regulatory requirements – or even lower depending [on] design specifications. Primary risk management tools include site characterization and selection, identification and assessment of abandoned wells (including potential remediation), and storage engineering to ensure CO2 containment and management of injection pressures. Actual risks will change over time, with growing risks during the early stages of the project, as CO2 is first injected into the storage reservoir. Eventually, information gained from the combination of performance modeling and acquisition of monitoring data will provide assurance that the project is conforming to the design specifications – or remediation measures will be taken to address unforeseen risks. After injection stops, the pressure in the storage reservoir will begin to decrease, lessening the risk of CO2 leakage or brine migration. Over time… secondary trapping mechanisms will further reduce the risks of health, safety and environmental impacts. (Benson et al. 2012)

Figure 6. Conceptual schematic illustrating the anticipated magnitude of health, safety and environmental risks over the lifetime of a typical geological storage project. From (Benson et al. 2012).

The technology for storing CO2 in deep underground formations is adapted from oil and gas exploration and production technology, and is relatively mature. Technologies to drill and monitor wells, methods to characterize sites, and models to predict where CO2 moves when it is pumped underground are all available, though more research could improve their performance. Every storage project is likely to use a combination of monitoring techniques that will track migration of the CO2 plume, detect leakage out of the storage reservoir, monitor injection rates and pressure, and detect seismic activity. Technology for monitoring geologic storage of CO2 is available from a variety of other applications. Large- scale storage projects in saline aquifers are needed to further refine site characterization and selection, capacity assessment, risk management, and monitoring.

To summarize scientific understanding of the permanence of geological CO2 sequestration, the Global Energy Assessment offers the following statement:

Observations from commercial storage projects, engineered and natural analogues as well as theoretical considerations, models, and laboratory and field experiments suggest that appropriately selected and managed geological storage reservoirs are very likely to retain nearly all the injected CO2 for very long times, more than long enough to provide benefits for the intended purpose of CCS (Benson et al. 2012).

2. Bioenergy with carbon capture and sequestration

This Section describes one leading negative emissions technology, bioenergy with carbon capture and sequestration (BECCS). In particular, it will focus on a description of the technology, and summarize existing research, existing commercial deployments, and research needs. Research for BECCS has focused on process engineering, techno-economic assessment, and its contribution to long-term climate change mitigation. BECCS is unique among CDR approaches is that produces energy products like electricity or fuels, while removing CO2 (Keith and Rhodes 2002). As such, it is a carbon-negative energy system. 2.1. Technology description

Implementation of BECCS involves several different systems: upstream biomass cultivation, facilities for energy conversion, carbon capture and sequestration systems, and downstream energy markets. A general representation of BECCS is shown in Figure 7. Aside from carbon products (CO2 and, in some cases, solid or liquid carbon byproducts), BECCS technologies produce a variety of energy products including heat, electricity, liquid fuels, hydrogen, or biomethane. Canadell and Schulze describe BECCS as:

A mitigation technology that combines biomass use with geological carbon capture and storage to produce negative CO2 emissions. Atmospheric CO2 is effectively captured by plants during photosynthesis and fixed into biomass, which is then used for bioenergy production. The CO2 emitted during this process is subsequently captured and sequestered in geological reservoirs, most commonly saline aquifers. This chain of processes, including the substitution of fossil fuel use with the bioenergy produced and the subsequent crop regrowth or replanting, creates a net carbon sink, or net negative emissions, in addition to a valuable source of energy. Biomass can be combusted, fermented, digested or gasified leading to different energy products such as heat, electricity, methane and synthetic biofuels. (Canadell and Schulze 2014)

Figure 7. Representation of BECCS process from (Canadell and Schulze 2014).

2.2. Process engineering and techno-economic assessment

Most existing research on BECCS focuses on engineering and economics at the energy facility level. This subsection summarizes this existing research, focusing on three primary pathways: integrated-gasification combined cycle with CCS for electricity (IGCC-CCS), polygeneration units producing electricity and fuels via the fischer-tropsch process with CCS (FT-CCS), and CO2 capture from fermentation of sugars to transportation fuels. Notably, these existing analyses focus on facility design or optimization, with a relatively simple representation of the greater energy-economy or environmental constraints on facility design.

As mentioned previously, there are several routes to bioenergy production with CO2 capture. Rhodes and Keith provide a high-level view of these processes in Figure 8. In general, they include biological processing with capture of CO2 by-products to produce liquid fuels, biomass gasification with shift and CO2 separation to produce hydrogen (or other products), and biomass combustion to produce electricity with CCS—either by oxyfuel (combustion in pure oxygen) or post-combustion capture (PCC). These basic routes can be combined or integrated (Rhodes and Keith 2005). For instance, polygeneration facilities produce both electricity and fuels (via FT and IGCC, for instance), while lignin that is a byproduct of ethanol fermentation can produce electricity with or without CCS. Rhodes and Keith do not include anaerobic digestion processes that produce CH4 and CO2.

Importantly, these routes produce gaseous byproducts with different concentrations of CO2, and energy products with different carbon intensities. For example, liquid fuel production processes, such as fermentation or gasification/gas-to-liquids, produce a relatively pure CO2 stream (Xu, Isom, and Hanna 2010; Liu et al. 2011). Hydrogen and electricity products from biomass produce no downstream biogenic CO2 emissions, unlike

13 liquid hydrocarbons. These factors affect BECCS system design and economics under carbon emissions limitations or carbon prices.

Figure 8. Routes to biomass with CO2 capture. The routes to biomass energy products with CCS include (from top to bottom) biological processing with capture of CO2 by-products to produce liquid fuels, biomass gasification with shift and CO2 separation to produce hydrogen, and biomass combustion to produce electricity with CCS—either by oxyfuel or PCC routes. These basic routes can be combined or integrated. From (Rhodes and Keith 2005). Figure 9 presents an IGCC-CCS facility for carbon-negative electricity production, from the National Energy Technology Laboratory (NETL) (Mike Matuszewski, James Black, and Eric Lewis 2013). Upon delivery, biomass is dried and fed into a high-temperature, oxygen- blown, entrained flow gasifier. Here, biomass and steam undergo a transformation into several gaseous products, notably a mixture of CO and H2 known as syngas, as well as CO2 and H2O. Following cleaning for impurities, this product gas is shifted in a water-gas shift reactor, producing H2 and CO2:

�� + �!� ��! + �!

Following shift, the shifted gas undergoes acid-gas removal in a dual-stage selexol unit. Here, CO2 and H2S separate from H2. H2S undergoes further processing to elemental sulfur (with CO2 recycle), and CO2 is compressed for sequestration. H2 is diluted with N2 from an air separation unit and combusted in a combined-cycle facility to produce electricity. As mentioned previously, carbon-negative electricity can also be produced using alternative system configurations, including combustion with post-combustion capture (PCC), or oxycombustion, followed by carbon sequestration (Mike Matuszewski, Eric Lewis, and Steve Herron 2013).

Figure 9. Process diagram for biomass IGCC-CCS facility designed for maximum capture of CO2. Explicit provision is made for CO2 recycle from the sulfur removal process. From (Mike Matuszewski, James Black, and Eric Lewis 2013).

NETL’s BECCS facility capacity (IGCC-CCS) is 262 MW, based on a maximum biomass supply constraint of 5,189 dry tons switchgrass per day, which is an explicit design consideration. Levelized cost of electricity (LCOE), which is internally consistent with several other coal and biomass facilities discussed in their study, is $257/MWh. The plant captures 90% of gross CO2 emissions, with net efficiency (HHV) of 26.6%.

Thermochemical conversion of solid fuels can produce a wide-range of energy products, including multiple outputs from a single facility, known as polygeneration. Figure 10 shows a process diagram for coal/biomass to Fischer-Tropsch liquid (FTL) synthetic fuels production, as well as electricity co-production (R. H. Williams et al. 2011). To contrast with the IGCC-CCS facility in Figure 9, we present a process diagram for polygeneration with altered inputs and outputs: 1) biomass is co-converted with coal, a fossil input, and 2) fuels production is prioritized.

Following separate coal and biomass gasification in entrained flow and fluidized bed gasifiers, respectively, shift occurs to form a product gas with H2/CO ratio of 1:1, which is beneficial for fuels production. Following acid gas removal (CO2 and H2S), the H2 and CO is sent to a FT synthesis reactor, which is then refined into finished gasoline and diesel 15

blendstocks. Residual syngas and hydrocarbons are subjected to another, optional, CO2 removal, before hydrogen is combusted in a combined-cycle facility to produce electricity. CO2 is compressed for sequestration.

Figure 10. Process diagram for the coal/biomass to Fischer-Tropsch liquid (FTL) synthetic fuels CO2 capture and storage process. Assumptions: separate O2-blown gasifiers for bituminous coal and for switchgrass; FT liquids obtained via slurry-phase synthesis with iron catalyst; H2/CO = 1.0 for syngas entering the synthesis reactor; on-site refining to upgrade crude FT liquid products to finished diesel and gasoline; CO2 and H2S separated from shifted syngas in an acid gas removal (AGR) unit using Rectisol; H2S recovered in AGR unit is reduced to elemental S in a Claus/SCOT plant; CO2 is compressed to 150 bar for transport to storage; residual H2-rich syngas is mixed with N2 from the air separation unit (for NOx control) and burned in the combustor of a gas turbine combined cycle (GTCC). From (R. H. Williams et al. 2011)

Polygeneration units with CCS have been the subject of a wide-range of techno-economic analysis (Liu et al. 2015; Larson et al. 2010; Chen, Adams, and Barton 2011). In general, results show that co-production of electricity and fuels holds economic advantages over single product facilities. This is due to the several factors, including economies of scale in fossil energy conversion, the inherently low cost of CO2 capture, and the thermodynamic advantages of coproduction (R. H. Williams et al. 2011). Such facilities can advance dual climate and energy security goals (Rhodes and Keith 2009).

One final BECCS process considered here is CO2 capture from fermentation of sugars to transportation fuels. CO2 is a byproduct of fermentation, and is available in concentrations of ~99% in flue gas (Xu, Isom, and Hanna 2010). Availability of concentrated CO2 reduces capture costs, and eases retrofit of existing facilities:

!"#$% �!�!"�! 2�!�!�� + 2��!

Existing deployments of BECCS, including those in the United States, use this configuration. This high-concentration source of CO2 is a good candidate for merchant markets for

capture and purification. The ethanol industry produced ~50 MtCO2 in 2008, with half of production in the United States. This amount is two orders of magnitude lower than U.S. emissions, which has been over 5000 MtCO2 in recent years. One gallon of ethanol produces 6.29 lbs of CO2 available for capture. 2.3. Contribution to long-term climate change mitigation

BECCS’ role in climate change mitigation has been studied primarily through integrated assessment models (IAMs), which model long-term evolution of the energy-economy at the global level, typically through 2100. As IAMs rely on BECCS in their representation of CDR technologies (along with afforestation), the discussion in Section 1.2 also informs the discussion to follow.

Rose et al. undertook a model intercomparison to understand the role bioenergy in climate management, specifically focusing on BECCS. Using 15 IAMs, they found:

Bioenergy is found to be valuable to many models, with significant implications for mitigation and macroeconomic costs of climate policies. The availability of bioenergy, in particular biomass with carbon dioxide capture and storage (BECCS), notably affects the cost-effective global emissions trajectory for climate management by accommodating prolonged near-term use of fossil fuels, but with potential implications for climate outcomes…Model scenarios project, by 2050, bioenergy growth of 1 to 10 % per annum reaching 1 to 35 % of global primary energy, and by 2100, bioenergy becoming 10 to 50 % of global primary energy. (Rose et al. 2013)

The ability of long-term models to incorporate realistic constraints on bioenergy supply, including sustainability constraints, is a matter of disagreement (Slade, Bauen, and Gross 2014). On this, Rose notes that less than half of the models explicitly consider land-use. Of those that do not, some implicitly consider land-use via bioenergy cost and/or sustainability constraints on biomass supply, such as allowable land-use.

Fuss et al. focus on BECCS deployment, rather than all forms of bioenergy, using results from the IPCC fifth assessment report (AR5) database (Fuss et al. 2014). Results are shown in Figure 11. They find that most IPCC scenarios that keep climate warming below 2 °C above pre-industrial levels use BECCS and many require net-negative emissions (that is, BECCS exceeding fossil fuel emissions) in 2100. Based on these results, Fuss et al. argue that the world is “Betting on Negative Emissions” for energy and climate mitigation strategies.

Figure 11. Carbon dioxide emission pathways until 2100 and the extent of net negative emissions and bioenergy with carbon capture and storage (BECCS) in 2100. a) Historical emissions from fossil fuel combustion and industry are compared with the IPCC fifth assessment report Working Group 3 emissions scenarios (pale colors) and to the four representative concentration pathways (RCPs) used to project climate change in the IPCC Working Group 1 contribution (dark colors). b) The emission scenarios have been grouped into five climate categories measured in ppm CO2 equivalent (CO2eq) in 2100 from all components and linked to the most relevant RCP. Most scenarios that keep climate warming below 2 °C above pre- industrial levels use BECCS and many require net negative emissions (that is, BECCS exceeding fossil fuel emissions) in 2100. From (Fuss et al. 2014).

2.4. Commercial deployment

Despite the large role of BECCS in global mitigation scenarios, there are relatively few deployments of BECCS technology at scale. This Section will highlight existing commercial deployments, based on academic literature and reports to the International Energy Agency (Arasto 2015). While not comprehensive, it is intended to highlight the diversity of BECCS implementation globally.

The largest BECCS project to date, which concluded in 2014, is in Decatur, IL. A partnership between Midwest Geological Sequestration Consortium, Archer Daniels Midland Company and Schlumberger Carbon Services facilitated the capture and geological storage of 1 MtCO2 over three years of operation (1000 tCO2/day). CO2 from the fermentation process at a corn ethanol production site was captured by dehydration and compressed before injection into the Mount Simon Sandstone formation, a saline aquifer, at 2km depth. The main objectives with the project were to validate capacity, injectivity and containment of the saline aquifer. Substantial funding was provided by the US Department of Energy Office of Fossil Energy (Finley 2014; Gollakota and McDonald 2012).

CCS from ethanol production facilities has also employed CO2 reuse, both for utilization and enhanced oil recovery (EOR) in the United States. Notably, three injections for EOR have sourced CO2 from facilities in Kansas, in Russell, Liberal, and Garden City. Kansas has ideal

18 geography for BECCS, producing both corn ethanol and conventional oil (Biorecro 2010). In the Netherlands, a grain ethanol plant operated by Abengoa Bioenergy captures and transports CO2 for fertilization of a nearby greenhouse facility (Arasto 2015).

BECCS has also been implemented or contemplated in several forms in the electricity sector. The Buggenum IGCC project in the Netherlands has studied both biomass and CCS integration (Damen et al. 2011). The White Rose Carbon Capture and Storage project is a planned oxycombustion plant that co-fires coal and biomass in the United Kingdom. The total output from the power plant is estimated to be 448 MW: with 90% CO2 capture efficiency, an estimated 2 MtCO2 will be captured and stored annually in North Sea for permanent geological storage (“About White Rose | White Rose CCS” 2015). Finally, Norway is exploring CO2 capture from the Klemetsrun waste incineration plant, a combined heat and power plant is Oslo.

Finally, fischer-tropsch conversion facilities are nearing commercial deployment. Notably, Total’s BioTFuel project is focused on commercial deployment of torrefaction, entrained- flow biomass co-gasification, and FT synthesis (Liu et al. 2015). The project, located in France, aims to produce 200,000 tons of biodiesel from 1 million tons biomass each year (~2700 tpd) by 2020. The project will not have a CCS system.

2.5. Research needs for implementation

To date, research on BECCS has proceeded primarily along two lines: process engineering / techno-economic analysis, and modeling of long-term mitigation portfolios with negative emissions. This Section attempts to answer this question: what new modeling and analysis is necessary to inform responsible, incremental implementation of BECCS, towards a global contribution to mitigation efforts?

Fuss et al. note the following high-level uncertainties, with emphasis on impacts of and requirements for large-scale implementation (Figure 12):

Four major uncertainties need to be resolved: (1) the physical constraints on BECCS, including sustainability of large-scale deployment relative to other land and biomass needs, such as food security and biodiversity conservation, and the presence of safe, long-term storage capacity for carbon; (2) the response of natural land and ocean carbon sinks to negative emissions; (3) the costs and financing of an untested technology; and (4) socio-institutional barriers, such as public acceptance of new technologies and the related deployment policies. (Fuss et al. 2014)

Figure 12. The four components of consistent negative emissions narratives. These inform key uncertainties and research needs. From (Fuss et al. 2014). I focus on a smaller subset of questions that are important to near-term installation of BECCS. While global-scale physical constraints and global ecosystem responses are important for understanding the impacts of gigaton-scale deployment of BECCS, they are likely better informed by measurement than models, and by iterative frameworks for risk management. Youngs and Somerville argue, in the case of cellulosic biofuels:

The academic literature… abounds in hypothetical scenarios about possible negative effects of continued biofuel expansion. The abundance of such concerns highlights the importance of implementation, by both importing and producing nations, of standards based on verifiable sustainability criteria and good governance… A disturbing trend in the treatment of model predictions as equivalent to knowledge or data based on actual measurement. Models are important tools, but they are often built on a partial state of knowledge and reflect assumptions and simplifications… Prediction will be tested by empirical studies that should ultimately settle the matter… It is therefore important that the scientific community remain clear about the relative power of measurements versus model predictions. (Youngs and Somerville 2014)

This is also echoed by the statement by the NAS in Table 1, that CDR “may be implemented incrementally with limited effects as society becomes more serious about reducing GHG concentrations or slowing their growth” (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015).

I instead focus on research needs that inform technology diffusion. BECCS will begin with small-scale regional implementations, driven by utilities, firms, and project developers. Results and measurement from local installations can inform global scale up. With this focus in mind, I identify four critical research needs for implementation:

1) Regional deployment: Existing engineering analyses focus on facility design or optimization, with a relatively simple representation of the energy economy,

including biomass availability. Alternatively, BECCS deployment has historically been evaluated through global IAMs, which lack necessary spatial and temporal detail to inform the operation of realistic energy systems. Finally, the evolution of energy systems and climate policy are influenced by regional considerations, which IAMs also lack (Victor and Keohane 2010; Hughes 1993). An evaluation of regional deployment of BECCS at high resolution is necessary to inform near- term efforts to build facilities, integrate them into existing energy and power systems, respect regional design considerations, and complement existing energy and climate policy.

2) Practical design: Existing BECCS process engineering studies often lack consideration of physical, or environmental, constraints on facility design, including the spatial availability of biomass. In particular, system scale is an important design consideration, as bioenergy facilities likely exhibit both economies-of-scale and diseconomies-of-scale (Bomberg, Sanchez, and Lipman 2014). Evaluation of optimal scale in a spatially-explicit context informs the engineering, logistics, and systems planning for building BECCS facilities.

3) Commercialization strategy: BECCS is very valuable to long-term climate change mitigation (Rose et al. 2013). Yet, there are few commercial deployments of BECCS outside of niche markets, creating uncertainty about commercialization pathways. This uncertainty is exacerbated by the absence of a strong policy framework, such as high carbon prices and research coordination. Strategies for commercial deployment can inform opportunities for governments, industry incumbents, and emerging players to research and support BECCS technologies.

4) Communication: Existing research has identified a key problem for BECCS as cultural, lacking in a ‘community of support’, awareness and credibility amongst its own key stakeholders and the wider public (Dowd, Rodriguez, and Jeanneret 2015). There is low public familiarity with BECCS, as well as some public resistance to bioenergy and CCS technology. Simple, accessible, and balanced introductions to BECCS can build awareness and credibility in the wider public and among central stakeholders.

3. Contributions

This dissertation attempts to fill critical gaps on the continuum of research tasks to inform large-scale implementation of BECCS (Figure 13). Research needs for implementation span from techno-economic analysis and global energy-economy modeling, to studies of deployment, design, commercialization, and communication. I attempt to focus my research on actionable information to inform key stakeholders responsible for deployment: utilities, firms, project developers, corporate entities, and governments. These entities are key actors in the ‘Energy Technology Innovation System’ responsible for commercializing new technologies (K. S. Gallagher et al. 2012).

In particular, this dissertation focuses on the deployment, design, commercialization, and communication of bioenergy with carbon capture and sequestration (BECCS). As discussed in Section 2.5 of this Chapter, it attempts to answer the question “what modeling and analysis is necessary to inform responsible, incremental implementation of BECCS, towards a global contribution to mitigation efforts?” It focuses on analysis that informs technology diffusion. As such, it is a bottom-up analysis, rather than top-down assessment. Put another way, it is applied research, practically using science for a specific, outcome-driven, purpose. Near-term implementation of BECCS is essential to understanding the costs, risks, and contributions of this technology to global scale mitigation.

Techno-Economic Deployment Design Commercialization Communication Global Energy- Analysis (TEA) Economy Modeling

30 Options to decrease carbon intensity of products 1 - Increase ratio of biomass / coal inputs 25 Sulfur 2 - Increase shift of syngas in WGS reactor Coal 3 3 - Recycle CO from sulfur removal to AGR 2 20 CO SuSullffuurr Climate Change Mitigation Air 2 Steam 2 Recycle RRemoval Mitigation without Negative Emissions Electricity 15 Business as Usual Integrated Gasification Air O2 Water-Gas Acid Gas Emissions 2 Combined (PgC / yr) 10 Separation Gasification Shift Removal Unit Cycle (WGS) (AGR) CO (I GCC) 5 H2 O CO +H2 O->CO2 +H2

N2 Biomass 0 1 Bypass CO2 (for compressionand sequestration) -5 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Year Research Well-established Incomplete Incomplete Incomplete Incomplete Well-established Status: Figure 13. Continuum of research tasks to inform large-scale implementation of BECCS. Research needs span from techno-economic analysis and global energy-economy modeling, to studies of deployment, design, commercialization, and communication. This dissertation focuses on incomplete research tasks. Characterization of research status by the author.

Commercial-scale deployment is dependent on the coordination of a wide range of actors, many with different incentives and worldviews. Despite this challenge, this dissertation attempts advance knowledge to inform responsible, incremental implementation of BECCS.

Chapter II. Deployment of carbon-negative energy systems

1. Preface

This Chapter explores the economic and deployment implications for BECCS at the regional level. We focus on the electricity system of western North America, known as the Western Electricity Coordinating Council (WECC). WECC has rich renewable energy resources, such as wind and solar, but relatively limited bioenergy resources. In particular, we study BECCS deployment under aggressive (pre-2050) time frames and carbon emissions limitations, with rich technology representation and physical constraints. We employ the SWITCH model for analysis, which leverages a unique combination of spatial and temporal detail to design realistic power systems that meet policy goals and carbon emission reduction targets at minimal cost (J. Nelson et al. 2012). We also discuss emissions accounting issues relating to BECCS, which is important in examining economy-wide decarbonization.

Studying regional deployment of BECCS remedies at least two shortfalls in existing knowledge. First, existing engineering analyses focus on facility design or optimization, with a relatively simple representation of the energy economy. As a result, systems planners and operators have relatively little understanding of BECCS integration in existing, or future, energy systems. Second, BECCS deployment has historically been evaluated through global IAMs, which lack necessary spatial and temporal resolution to inform the operation of realistic energy systems. Augmenting these shortfalls is the importance of regional planning and analysis: the evolution of energy systems and climate policy are influenced by regional considerations, which IAMs also lack (Victor and Keohane 2010; Hughes 1993). Similarly, biomass availability and cost has distinct regional characteristics (Downing et al. 2011).

More broadly, an evaluation of regional deployment of BECCS at high resolution is necessary to inform near-term efforts to build facilities, integrate them into existing energy and power systems, respect regional design considerations, and complement existing energy and climate policy.

2. Excerpt from Biomass enables the transition to a carbon-negative power system across western North America (Sanchez et al., 2015a)1

2.1. Main text

1 First published in Nature Climate Change 5, 230–234, 2015 by Nature Publishing Group (NPG). NPG does not require authors of original (primary) research papers to assign copyright of their published contributions. Authors grant NPG an exclusive license to publish, in return for which they can reuse their papers in their future printed work without first requiring permission from the publisher of the journal. For more information, see http://www.nature.com/reprints/permission-requests.html

Sustainable biomass can play a transformative role in the transition to a decarbonized economy, with potential applications in electricity, heat, chemicals, and transportation fuels (Demirbaş 2003; Farrell and Gopal 2008; Liu et al. 2011). Deploying bioenergy with carbon capture and sequestration (BECCS) results in a net reduction in atmospheric carbon. BECCS may be one of the few cost-effective carbon-negative opportunities available should anthropogenic climate change be worse than anticipated or emissions reductions in other sectors prove particularly difficult (Hansen et al. 2008; Read and Lermit 2005). Previous work, primarily using Integrated Assessment Models (IAMs), has identified the critical role of BECCS in long-term (pre- or post-2100 timeframes) climate change mitigation, but has not investigated the role of BECCS in power systems in detail, or in aggressive timeframes (Milne and Field 2014; IPCC 2014b), even though commercial-scale facilities are starting to be deployed in the transportation sector (Gollakota and McDonald 2012). Here, we explore the economic and deployment implications for BECCS in the electricity system of Western North America under aggressive (pre-2050) timeframes and carbon emissions limitations, with rich technology representation and physical constraints. We show that BECCS, combined with aggressive renewable deployment and fossil emission reductions, can enable a carbon-negative power system in Western North America by 2050 with up to 145% emissions reduction from 1990 levels. In most scenarios, the offsets produced by BECCS are found to be more valuable to the power system than the electricity it provides. Advanced biomass power generation employs similar system design to advanced coal technology, enabling a transition strategy to low-carbon energy.

An assessment of BECCS deployment as part of a suite of low-carbon technologies is a critical research need (Benson 2014). Such an analysis requires detailed spatial and temporal assessment of distributed biomass supply, electricity demand, deployment of intermittent renewables, and electricity dispatch capabilities. We employ the SWITCH optimization model for long-term strategic planning of the electric system (Wei et al. 2013; Mileva et al. 2013). SWITCH leverages a unique combination of spatial and temporal detail to design realistic power systems that meet policy goals and carbon emission reduction targets at minimal cost (J. Nelson et al. 2012). The version of the SWITCH model used here encompasses the region of the Western Electricity Coordinating Council (WECC), which includes the Western United States, two Canadian provinces, and a small portion of Mexico. WECC contains high quality wind and solar resources, but relatively limited bioenergy resources: the Eastern United States, for example, has a larger absolute resource (Downing et al. 2011). Existing studies of low-carbon transitions in Western North America have generally reserved biomass for biofuels production, rather than for electricity (Wei et al. 2013; J. Long et al. 2011).

Western North America contains biomass resources from forestry, wastes, agricultural residues, and dedicated energy crops, though supply is limited by land and sustainability practices (Figure 14) (Downing et al. 2011). In total, we identify 1.9x109 MMBtu (2000 PJ) of economically recoverable bioenergy available annually from solid biomass by the year 2030, sufficient for ~7-9% of modeled demand for electricity in 2050. Our estimates for availability in California are smaller than other studies, which tend to focus on ‘technical potential’ rather than ‘economically recoverable’ resources (J. Long et al. 2011; J. H. Williams et al. 2012). While barriers to biomass recovery exist even for economically 24 recoverable resources, we choose these resources as a reasonable approximation of biomass potential. We model solid biomass fuel costs as a piecewise linear supply curve disaggregated for 50 regions across Western North America. Biomass supply from dedicated energy crops represents only 7% of the total supply, so direct land use impacts from the biomass feedstocks used in this study would be minimal. Dedicated feedstocks, such as switchgrass and pulpwood, tend to have higher prices than wastes and residues.

Feedstock Type! 15 Forest Residue [R] Mill Residue + Pulpwood [R + D] Waste [W]! Forest Thinning [R] Municipal Solid Waste (MSW) [W] Residue [R] ! Agricultural Residues (Canada Only) [R] Dedicated Feedstock [D]! Switchgrass [D] Corn Stover [R] 10 Wheat Straw [R] Pulpwood! Orchard and Vineyard Waste [W] Switchgrass! 0 100 200 300 400 500 Supply (PJ) Agricultural Residues! 5 Orchard and Vineyard Waste! Corn Stover! Price (2013$/MMBtu) Wheat Straw! Mill Residue! MSW! Forest Residue!

0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Quantity of Biomass Available (PJ) Figure 14. Supply Curve of Available Solid Biomass post-2030. Biomass can provide up to 2000 PJ/yr of energy in 2030 for the electricity system, from a number of waste and dedicated sources. Labels on supply curve represent the principal price region of a given biomass source. Feedstocks are classified as wastes (‘W’), residues (‘R’), or dedicated feedstocks (‘D’). Dedicated feedstocks tend to be the most expensive.

The implications of BECCS for the economics and carbon emissions of regional power systems through 2050 have not been previously investigated in detail. To address this gap, we explore scenarios for the electricity sector that are consistent with economy-wide decarbonization, but vary the allocation of biomass across sectors of the economy (Table 7, SI Text). We explore scenarios with WECC-wide power sector CO2 emissions reductions from 1990 levels by 2050 ranging from 105% to 145%, which previous work has found would be consistent with economy-wide goals should biomass be used for electricity (Max Wei et al. 2013). Our case without biopower mandates an 86% reduction in CO2 emissions from 1990 levels by 2050 (-86% No Biomass). We vary this scenario by disallowing CCS technologies (-86% No CCS No Biomass) and allowing biomass (-86%). To understand biomass deployment in carbon-neutral and carbon-negative power systems, we mandate a 105% reduction (-105%), 120% reduction (-120%), and 145% reduction (-145%) in CO2 emissions by 2050. These scenarios require aggressive R&D on CCS and BECCS over the coming decades. We continue operation of some existing nuclear plants, but do not allow new nuclear power. We do not conduct a complete economy-wide assessment of CO2 emissions across WECC or optimal biomass allocation among sectors.

Without biomass technologies (-86% No Biomass), the resource mix is reliant on other renewable energy technologies including wind, solar, hydro and geothermal for 86% of total electricity generated in 2050 (Figure 15 (a)). Low carbon power systems employ gas technologies (with and without CCS), storage, and transmission to compensate for 25 renewable intermittency. Coal (with and without CCS) plays little to no role in energy generation because of its relatively high level of CO2 emissions (Figure 15 (b)). While CCS technology reduces CO2 emissions from coal, coal CCS still has higher emissions than gas CCS (Table 4, SI Text). Without CCS technologies (-86% No CCS No Biomass), the resource mix is even more reliant on renewable energy, up to 94% in 2050.

(a) -86% Carbon Cap Scenarios Carbon-negative Scenarios 20

15 Wind Solar GWh/yr) 5 Gas Gas CCS Coal CCS 10 Biomass CCS Hydro Geothermal Nuclear Generation in 2050 in (x10 Generation 5 Other

0 -86% No -86% No CCS -86% -105% -120% -145% Biomass No Biomass

Power Cost (2013$) $184/MWh $219/MWh $134/MWh $138/MWh $160/MWh $188/MWh

(b)! 200!

150!

100! ! /yr) 2 Coal! 50! Coal CCS! CCGT CCS! CCGT! 0! -86% No -86% No -86%! -105%! -120%! -145%! Gas Combustion Turbine! Biomass! CCS No Compressed Air Energy Storage! Biomass! -50! Biomass Coﬁring!

Emissions in 2050 in (MTCO Emissions Biomass CCS! 2 CO -100!

-150!

-86% Carbon Cap Scenarios! Carbon-negative Scenarios! -200! Figure 15. (a) Generation (105 GWh, gross), and cost of electricity (2013$/MWh) in 2050. Fossil fuel use is phased out as the power system becomes carbon-negative, transitioning from coal CCS and gas, to gas combined with CCS. ‘Other’ includes generation from coal, biogas, and bioliquid inputs. Total generation exceeds system load because of transmission, distribution, and storage losses as well as curtailment of generation on resources. (b) Yearly carbon emissions (MtCO2/yr) in 2050. Biomass CCS and biomass cofiring CCS on coal CCS plants provide negative CO2 emissions. As emissions limits are reduced, fossil CO2 emissions shift from coal and CCGT to CCGT with CCS. BECCS can sequester ~165 MtCO2/yr. Biomass CCS technologies enable a power system more reliant on baseload and fossil technologies in 2050 at moderate power sector emission caps (between -86% and -105%). In the -86% case, coal CCS, biomass cofiring, and BECCS cumulatively provide 20% of electricity generated, enabling lower-cost gas resources to generate 22% of electricity while still meeting CO2 emission constraints. 43 GW of coal and biomass technologies are installed throughout western North America in 2050 (Figure 21 (a), SI Text). Due to the dispersed nature of the fuel resource, biomass deployment is distributed across the WECC. In the context of the electric power sector, if the cap on carbon emissions is held constant, the introduction of bioenergy for BECCS reduces power system costs, carbon abatement costs, and the need for electrical energy storage for intermittent renewable energy (Figure 15 (a) and SI Text).

As the carbon cap becomes more stringent between the -105% and -145% case, we see CO2 emissions from combined-cycle gas turbine (CCGT) technology shrink before being

27 captured via CCS, as well as increased renewable generation from wind and solar (Figure 15 (a) and (b)). Coal CCS and biomass cofiring CCS play a significant role in the -105% case (~13% of average 2050 electricity generated), a smaller role in the -120% case (2%), and no role (0%) in the resource mix under the -145% case given the severity of the CO2 emissions constraint (Figure 21 (b), SI Text). Gas turbines are installed across all scenarios to provide flexibility, dispatchability, and system reserves.

Our -145% scenario demonstrates a power system that generates almost all electricity from renewable resources, representing how the power sector might be configured if climate change is severe, or emission reductions in non-electricity sectors are more expensive than the electricity sector. In our -145% case, biomass CCS plants provide carbon-negative baseload power in 2050, resulting in overall emissions of -135 MtCO2/yr in the WECC (Figure 15 (a) and (b)). Generation, electricity costs, (Figure 15 (a)) and dispatch (Figure 16 and Figure 25, SI Text) are similar between the -86% No Biomass and - 145% cases, with the exception of BECCS technology deployment. Low-carbon scenarios without BECCS and carbon-negative scenarios with BECCS ultimately result in qualitatively similar deployment of gas and renewable generation.

! 300! Jan! Feb! Mar! Apr! May! Jun! Jul! Aug! Sep! Oct! Nov! Dec! 250!

200!

150!

100!

50!

-50! 2! 10!18! 3! 11!19!2! 10!18! 3! 11!19!2! 10!18! 3! 11!19!2! 10!18! 1! 9! 17! 2! 10!18! 1! 9! 17! 2! 10!18! 1! 9! 17! 2! 10!18! 1! 9! 17! 2! 10!18! 1! 9! 17! 2! 10!18! 1! 9! 17! 2! 10!18! 1! 9! 17! 2! 10!18! 3! 11!19!2! 10!18! 3! 11!19! Generation in 2050, -145%in Generation Case (GW) Hourly Dispatch over Simulated Year! Nuclear! Geothermal! Other Biomass! Solid Biomass! Gas CCS! Gas! Storage (discharging)! Hydro (non-pumped)! Solar! Wind! Storage (charging)! System Load! Figure 16. Hourly dispatch in 2050 in the -145% case. Power system dispatch is shown in sampled hours from two days (peak and median day) each month between January-December. With the exception of BECCS, dispatch is similar between the -86% No Biomass and -145% case (Figure 25, SI Text). Low-carbon scenarios without BECCS and carbon-negative scenarios with BECCS ultimately result in similar deployment of gas and renewable generation. “Other Biomass” includes both liquid and gaseous biomass supplies. In all cases where biomass is allowed, the power system employs between 90-98% of all biomass supply available in 2050, regardless of the extent of CO2 reduction or availability of low-carbon flexible assets. This indicates that biomass systems are cost-effective in the context of low carbon power systems in Western North America, especially due to negative CO2 emissions from BECCS. Given the very small amount of net CO2 emitting infrastructure in the -145% case, we do not expect that that emissions could fall well below a 145% reduction with projected levels of biomass availability. While technology cost, lifecycle CO2 emissions, and performance assumptions in carbon-negative power systems alter the

28 relative deployment of coal CCS and intermittent renewables, they have little effect on biomass deployment (SI Text).

We find that the value of BECCS lies primarily in the sequestration of carbon from biomass, rather than electricity production. This result reconfirms previous results found using IAMs (Klein et al. 2013). To illustrate this point, we explore cases in which BECCS plants capture CO2 emissions but do not produce electricity. The average cost of electricity when BECCS is used exclusively for carbon sequestration is only slightly higher (~6%) than when BECCS provides both sequestration and electricity (Figure 22, SI Text). Carbon sequestration from biomass, regardless of the technology employed or capital cost, could be a key driver of climate change mitigation pathways in the 2050 timeframe.

Our analysis has several implications for CO2 reduction, technology development, and biomass allocation. Negative emissions from BECCS can offset CO2 emissions from fossil energy across the economy. The amount of biomass resource available limits the level of fossil CO2 emissions that can still satisfy carbon emissions caps. Efforts to expand biomass supply can increase demand for water, land, and fertilizer, or other ecosystem impacts (Solomon 2010; Abbasi and Abbasi 2010). Given the level of projected biomass availability in WECC, it would appear that there is little room for coal CCS technology to play a role in an energy system consistent with economy-wide emissions reductions goals. Gas CCS, however, can contribute moderately to economy-wide decarbonization due to its operational flexibility (Figure 21, SI Text).

Our analysis suggests that installation of up to 10 GW of BECCS capacity between 2030 and 2040, with additional capacity additions thereafter, could be a key part of meeting stringent climate goals in the WECC. Such a goal would require a concentrated effort in finance, site selection, biomass sourcing, geological characterization, permitting, site- specific environmental impact assessments and community consultation. Biomass harvesting, drying, and transportation present logistical challenges to rapid deployment. However, we find necessary capacity deployment rates for BECCS to be smaller than that for other intermittent renewables or gas.

Advanced biomass power generation technology employs similar system design to advanced coal technology, including CCS and integrated gasification combined cycle (IGCC) systems (Corti and Lombardi 2004). Such systems boast higher efficiency and more easily capture CO2 emissions than conventional steam turbines; these characteristics become even more desirable in light of biomass’ lower energy density, higher feedstock cost, and distributed geographic nature. Research needs include systems integration and technology advancement in gasification, air separation, gas cleaning, shift catalysis, and gas turbines that operate on H2-rich syngas (Farrell and Gopal 2008; Maxson et al. 2011). Moving forward, the fossil fuel industry could embrace higher and more efficient levels of biomass utilization combined with CCS technology development as a transition strategy to low- carbon energy. BECCS could enable some of the world’s largest carbon emitting entities to instead become some of the world’s largest carbon sequestering entities.

Biomass could enable CO2 reduction not only in the electricity sector, but also the transportation and industrial sectors for fuels, heat, and chemicals. We estimate that cellulosic biofuel production from available biomass in WECC can reduce emissions by 75 MtCO2/yr by displacing gasoline, based on literature conversion efficiency and near-term carbon intensity values (Farrell et al. 2006). In contrast, if biomass is made available to the power sector, BECCS can sequester 165 MtCO2/yr and also displace fossil electricity. At the conversion efficiencies assumed in this study, bioelectricity contains 28-45% of the net energy of candidate cellulosic ethanol conversion pathways, but can provide as much as 41% more transportation miles because of the high efficiency of battery electric drive vehicles (Table 8, SI Text) (Campbell, Lobell, and Field 2009; Farrell et al. 2006).

Our analysis indicates that while valuable to the power sector, carbon sequestration from biomass may be more cost-effective in other sectors. We find BECCS technology deployment at abatement costs as low as $74/tCO2 in the -86% case, with more stringent emission caps incurring higher abatement costs. Such costs are slightly higher than afforestation schemes (~$5 – 40/tCO2), biochar projects in North America ($30 – 40/tCO2), and cellulosic biofuel production ($35/tCO2), but are far lower than projected abatement costs for direct air capture of CO2, which has been assessed as high as $1000/tCO2 (IPCC 2007; Pratt and Moran 2010; Lutsey and Sperling 2009; House et al. 2011). Should carbon sequestration be more effective via alternative abatement methods, the electric power sector would find it economical to purchase those offsets. A roadmap of economy-wide biomass policy focused on CO2 reduction should account for both the technical potential and economic costs of biomass deployment across sectors. Increasing efficiency, reducing costs, and commercializing carbon-negative biomass technologies could make such a roadmap possible. 2.2. Materials and methods

Model description and additional methods are presented in detail in the SI Text.

2.2.1. Biomass technologies

SWITCH inputs include technology cost profiles, construction timeframes, outage rates, generation flexibility, retrofit ability, heat rate, and cycling penalties for a broad range of existing and new conventional and renewable energy generation technologies. Technical performance metrics and evolution of capital and operations and maintenance costs are drawn primarily from Black and Veatch (“Cost and Performance Data for Power Generation Technologies” 2012). We assume that future biomass plants will use IGCC technology, while existing plants use steam turbines (Table 2 and Table 3, SI Text). CCS technologies are modeled with a default capture efficiency of 85%, and are available for installation on biomass IGCC, coal, and natural gas technologies after 2025. We do not explicitly model criteria pollutants, which may require additional control technology to be installed on coal and biopower technologies.

While Black and Veatch estimates capital and operating costs for biomass IGCC plants, their dataset does not include similar values for BECCS plants. As assumptions between cost

30 datasets can differ substantially, we choose to estimate cost and efficiency parameters for BECCS plants from other similar plant types. We derive the capital cost of CCS equipment, the efficiency penalty of performing CCS, and increase in non-fuel variable operations and maintenance costs for BECCS from coal IGCC and coal IGCC CCS systems. Our BECCS capital cost estimates are within 5% of those by the National Energy Technology Laboratory (NETL) for biomass IGCC-CCS facilities (Mike Matuszewski, James Black, and Eric Lewis 2013). Increasing the capital cost of BECCS would likely not lower deployment due to the high value of carbon sequestration. As a large amount of the biomass resource is already deployed in our scenarios, lowering the capital cost would also be unlikely to affect deployment. 2.2.2. Biomass supply

Fuel costs for solid biomass are input into the SWITCH model as a piecewise linear supply curve for each load area. This piecewise linear supply curve is adjusted to include producer surplus from the solid biomass cost supply curve in order to represent market equilibrium of biomass prices in the electric power sector. As no single data source is exhaustive in the types of biomass considered, solid biomass feedstock recovery costs and corresponding energy availability at each cost level originate from a variety of sources (Table 5, SI Text). We consider two scenarios for biomass life-cycle assessment (LCA): 1) carbon-neutrality, as feedstocks are primarily wastes or low-input crops grown on marginal lands, and 2) a sensitivity scenario with solid biomass penalized at 10% of its biogenic carbon content. In the carbon-neutral cases, we assume that direct emissions from harvesting and transport— a small source of emissions—will be minimized as the entire economy is decarbonized (Rhodes and Keith 2005). The sensitivity case represents increased emissions such as those from transportation, fertilizer, or soil organic carbon (SOC) from residue collection, which recent empirical work suggests may be larger than previously thought (Liska et al. 2014). 2.2.3. Biomass cofiring and modeled scenarios

Cofiring is allowed up to 15% of total output from a single coal plant. When cofiring is installed on a plant with CCS technology, we assume that the heat rate increases by the same percentage when sequestering carbon as does coal IGCC relative to coal IGCC CCS.

2.2.4. CCS reservoirs and transportation

Large-scale deployment of CCS pipelines would require pipeline networks from CO2 sources to CO2 sinks. We require CCS generators that are not near a CO2 sink to build longer pipelines, thereby incurring extra capital cost. If a load area does not does not contain an adequate CO2 sink within its boundaries, a pipeline between the largest substation in that load area and the nearest CO2 sink is built. We derive pipeline costs from existing literature. CCS plants must send all of their CO2 output to their closest reservoir.

2.2.5. Scenario development

All scenarios enforce a carbon cap and existing Renewable Portfolio Standard (RPS) laws. We disallow new nuclear generation. Electricity demand profiles include extensive energy efficiency, electric heating, and electric vehicle penetration consistent with economy-wide decarbonization. We sample hourly demand for each of 50 areas within WECC for six hours of each of 12 representative days in the decades 2020–2050. Investment decisions are made in four periods between 2016-2055; these periods are 2016-2025 (“2020”), 2026- 2035 (“2030”), 2036-2045 (“2040”) and 2046-2055 (“2050”). In each modeled hour, demand must be met by the optimization, as well as capacity and operational reserve margin constraints to ensure system reliability.

2.3. Additional methods and information 2.3.1. Overview of the SWITCH model

SWITCH is a capacity planning and dispatch model of the electric power sector. In this study, SWITCH is used to model the entire geographic extent of the Western Electricity Coordinating Council (WECC). The model is a linear program whose objective function is to minimize the cost of delivering electricity from present day until 2050 with generation, transmission, and storage subject to policy, carbon emission, resource availability, reliability, and generator output constraints. SWITCH is well suited to explore the optimal deployment of a low-carbon WECC power system as it models a large geographic region in detail at a high temporal resolution. The WECC-wide version of SWITCH used in this study is maintained and was developed by Ph.D. students James Nelson, Daniel Sanchez, Ana Mileva, and Josiah Johnston and Daniel Kammen in Professor Kammen’s Renewable and Appropriate Energy Laboratory (RAEL). James Nelson is now with the Union of Concerned Scientists.

Many capacity expansion models of the electricity grid encounter difficulties with the spatially and temporally complex nature of intermittent resources relative to conventional generators. To address these issues, SWITCH uses time-synchronized hourly load and renewable generation profiles in a capacity expansion model. SWITCH determines the contribution of baseload, dispatchable and intermittent generation options alongside storage and transmission capacity on a least-cost basis, subject to policy constraints, while ensuring that projected electricity load is met reliably. The model concurrently optimizes investment in, and dispatch of, power system infrastructure, an approach that allows for proper valuation of intermittent renewable capacity at varying levels of intermittent resource penetration. SWITCH has perfect foresight and assumes a single global decision maker.

This version of the SWITCH model encompasses the synchronous region of the Western Electricity Coordinating Council (WECC). WECC includes 11 western U.S. states, Northern Baja Mexico, and the Canadian provinces of British Columbia and Alberta. WECC provides an ideal case to examine system dynamics in a complex, interconnected region with significant greenhouse gas (GHG) emissions and many low-carbon generation resources,

32 including biomass. In the version of SWITCH used in this study, WECC is divided into 50 ‘load areas,’ within which power is generated and stored, and between which power is transmitted.

In the model, four ‘investment periods,’ each ten years in length, span the time between the present day and 2050. The first of these investment periods represents 2016-2025 and the last represents 2046-2055. At the start of each investment period, SWITCH chooses which generation, storage and transmission projects to build. All investment periods are optimized simultaneously, so projects installation decisions in earlier investment periods affect decisions made in later periods, and vice versa. SWITCH operates existing power system infrastructure and can build new generation, transmission and storage capacity in order to meet load cost-effectively. Each optimization is given the option to build over 7500 generation projects, 200 storage projects, and 100 transmission projects in each investment period.

SWITCH makes power system investment and dispatch decisions simultaneously, thereby evaluating the present and future value of infrastructure investments within the context of their hourly value to the electric power system. We represent six different categories of flexibility with respect to dispatch: baseload, flexible baseload, intermittent, dispatchable, hydroelectric, and storage. Baseload generators (biogas, geothermal, nuclear, cogenerators) are operated at the same level of output in every study hour. Flexible baseload generators (coal, biomass) can vary their output on a daily basis between a minimum and maximum power output. Intermittent generators (solar, wind) produce power corresponding to their hourly capacity factor in each study hour. Dispatchable generators (non-cogeneration natural gas and hydroelectric) can vary their level of energy output as a function of installed capacity. Dispatchable generators can also adjust how much capacity to keep in both spinning and non-spinning reserve within each study hour. Hydroelectric generators can vary hourly output subject to average historic generation and minimum flow requirements. Storage projects are subject to daily energy balances, as well as minimum and maximum dispatch constraints based on installed capacity.

2.3.2. Materials and methods

This Section will document model development and model enhancements that enable a detailed examination of the role of biomass in WECC. First, we document modeled biomass and coal technologies. Second, we describe our construction of a spatially explicit biomass supply curve for the WECC. Third, we describe model implementation of biomass cofiring retrofits at existing and future coal plants. Fourth, we describe our modeling of CCS pipelines and reservoirs. Finally, we describe scenario development for this report.

Biomass technologies SWITCH inputs include technology cost profiles, construction timeframes, outage rates, generation flexibility, retrofit ability, heat rate, and cycling penalties for a broad range of existing and new conventional and renewable energy generation technologies. Technical performance metrics and evolution of capital and operations and maintenance (O&M) costs are drawn primarily from Black and Veatch’s Cost and Performance Data for Power 33

Generation Technologies (“Cost and Performance Data for Power Generation Technologies” 2012). Bioelectricity technologies include cofiring biomass with coal in new and existing coal plants, dedicated biomass steam turbines, biomass integrated gasification combined cycle (IGCC) systems, and bioenergy systems equipped with carbon capture and sequestration technologies (BECCS). We assume that future biomass plants will use IGCC technology, while existing plants use steam turbines. Existing biomass plants are available for cogeneration and cogeneration-CCS retrofits. Available cogeneration capacity is calculated using on-site demand for heat. Biogas internal combustion engine cost and performance is sourced from the Electric Power Research Institute (Table 2 and Table 3) (McGowin 2007). CCS technologies are modeled with a default gross capture efficiency of 85%. We do not model flexible biomass generation technologies that provide hourly load following or sub-hourly reserves.

Base Allow CCS Allow Cogeneration Existing Technology Retrofit Technology Biomass IGCC Y, in limited N/A N scenarios Biogas internal Y, with Y, where available Y combustion cogeneration engine (limited scenarios) Biomass steam Y, with Y, where available Y turbine cogeneration Coal steam Y Y, where available Y turbine Coal IGCC Y N/A N Table 2. Biomass technology and biomass cofiring-enabled coal technology in the SWITCH model.

Technology Capital Variable Fixed Heat Rate Heat Rate Construction Cost O&M O&M (non- (CCS) time (2013$ ($/MWh) ($/ CCS) (MMBtu/ /kW) kW- (MMBtu/ kWh) yr) kWh) Biomass 1120 ** 22.6 10 13.44 1 year Cofiring Biomass 3562 13.95 88.3 12.5 N/A 2 years IGCC 5 Biomass 5971 20.13 100. N/A 16.32 2 years IGCC-CCS 72 Table 3. Specifications for selected biomass technologies, in 2013$. Biomass cofiring assumes the Variable O&M costs of its parent technology.

While Black and Veatch estimates capital and operating costs for biomass IGCC plants, the dataset does not include similar values for BECCS plants. As assumptions between cost

34 datasets can differ substantially, we choose to estimate cost and efficiency parameters for BECCS plants from other similar plant types. To estimate the capital cost of CCS equipment, we assume that the capital and fixed costs for adding a CCS system to a biomass solid IGCC plant are the same (in $/W of capacity) as for coal IGCC relative to coal IGCC CCS. To estimate the efficiency penalty of performing CCS – input energy is necessary to capture and sequester carbon – we assume that the heat rate of a biomass solid IGCC plant increases by the same percentage when sequestering carbon as does coal IGCC relative to coal IGCC CCS. To estimate the increase in non-fuel variable operations and maintenance costs incurred by operating a CCS system on a biomass solid IGCC plant, we add a variable cost for sequestering carbon of $6.2/MWh to the biomass solid IGCC variable cost, which was calculated using the percentage heat rate increase due to carbon sequestration of both coal and biomass IGCC plants.

CO2 emissions factors assumed in this study for coal, biomass, and gas technologies with and without CCS are shown in Table 4.

Technology Emissions Factor (metric tons CO2/MWh) Coal Steam Turbine 0.86 Coal IGCC 0.76 Gas Combustion 0.55 Turbine CCGT 0.36 Coal Steam Turbine 0.17 CCS Coal IGCC CCS 0.15 CCGT CCS 0.08 Biomass Cofiring 0 Biomass IGCC 0 Biomass Cofiring CCS -1.08 Biomass IGCC CCS -1.31

Table 4. CO2 emissions factors for various biomass, coal, and gas technologies with and without CCS. Lifecycle emissions for solid biomass are omitted.

Biomass supply We construct spatially disaggregated biomass supply curves as an input to the SWITCH model. Fuel costs for solid biomass are input as a piecewise linear supply curve for each load area. This supply curve is adjusted to include producer surplus from the solid biomass cost supply curve in order to represent market equilibrium of biomass prices in the electric power sector. We do not allow transportation of biomass between load areas.

As no single data source is exhaustive in the types of biomass considered, solid biomass feedstock recovery costs and corresponding energy availability at each cost level originate from a variety of sources (University of Tennessee 2007; De La Torre Ugarte and Ray 2000; N. C. Parker 2011; Milbrandt and (U.S.) 2006; Kumarappan, Joshi, and MacLean 2009) 35

(Table 5). This table does not represent the technical potential of recoverable solid biomass – instead it depicts the economically recoverable quantity of biomass solid feedstock. The definition of ‘economically recoverable’ is dependent on each dataset, but the maximum cost is generally less than or equal to $100 per bone dry ton (BDT) of biomass. While the energy content per BDT of biomass varies by feedstock, a factor of 15 MMBtu/BDT can be used for rough conversion between BDT and MMBtu. Feedstock prices range are as low as $0.3/MMBtu, with a median price across WECC of $2.9/MMBtu. 90% of such biomass is available for prices under $100/BDT (in $2013, approx. $6.70/MMBtu).

Biomass California Rest of WECC Carbon Flux Sources Feedstock Availability Availability Potential Type [1012 Btu/yr] [1012 Btu/yr] (MtC/yr)

Corn Stover 19.1 82.3 3.4 (University of Tennessee 2007) (De La Torre Ugarte and Ray 2000) Forest Residue 41.3 408.8 14.2 (Milbrandt and (U.S.) 2006) (Kumarappan, Joshi, and MacLean 2009) Forest 72.3 211.0 8.9 (N. C. Parker Thinning 2011) Mill Residue + 39.5 341.0 12.0 (Kumarappan, Joshi, and Pulpwood MacLean 2009) (N. C. Parker 2011) (Milbrandt and (U.S.) 2006) Municipal Solid 81.4 117.1 5.7 (N. C. Parker 2011) Waste (MSW) (Milbrandt and (U.S.) 2006) Orchard and 66.1 10.5 2.4 (N. C. Parker

Vineyard 2011) Waste Switchgrass 0 123.7 4.0 (University of Tennessee 2007) (De La Torre Ugarte and Ray 2000) (Kumarappan, Joshi, and MacLean 2009) Wheat Straw 8.1 70.0 2.7 (University of Tennessee 2007) (De La Torre Ugarte and Ray 2000) Agricultural 0 183.2 6.1 (Kumarappan, Joshi, and Residues MacLean 2009)

(Canada Data Only) Total 327.8 1547.6 59.4 Table 5. Biomass Supply in the SWITCH model for years 2030 and beyond. We assume elemental carbon content of 47.5% in all biomass.

We classify biomass supply as wastes (Municipal Solid Waste [MSW], Orchard and Vineyard Waste), residues (Forest Residue, Mill Residue, Forest Thinning, Agricultural Residues, Corn Stover, Wheat Straw), or dedicated feedstocks (Switchgrass, Pulpwood). Wastes include trimmings, wood, cardboard or paper that are diverted from loads destined for landfills. Residues are typically left on the ground to return to the soil, which may have importance in soil carbon cycles. Dedicated feedstocks are energy crops or current commodities that may require additional land to be cultivated or harvested.

Biomass cofiring and modeled scenarios Biomass cofiring is available as a retrofit option for existing coal plants, and is available both as a retrofit and a co-installation option for new coal plants. Biomass cofiring costs and performance characteristics were again drawn from Black and Veatch’s Cost and Performance Data for Power Generation Technologies. Explicitly defined parameters are shown in Table 3. Other parameters, including forced and scheduled outage rates, variable O&M costs, dispatch flexibility, and thermal cycling penalties, are assumed to be the same as the parent technology. When cofiring is installed on a plant with CCS technology, we assume that the heat rate increases by the same percentage when sequestering carbon as does coal IGCC relative to coal IGCC CCS. Cofiring is allowed up to 15% of total output from a single plant.

For flexible baseload cofiring plants, we include explicit constraints on plant-level dispatch, capacity, and minimum loading.

∗ ∗ �!,! ≤ �!,!× �! ∗ �!,! ≤ .15 × �!,! ∗ �!,! + �!,! ≤ �!,!× �! ∗ �!,! + �!,! ≥ �!,!× �!

D = dispatch from parent plant C = capacity of parent plant

D* = dispatch from cofiring C* = cofiring capacity g = availability of plant, including m = minimum loading outages p = project t = technology i = investment period d = day Table 6. Variables for biomass cofiring dispatch, capacity, and flexibility constraints in the SWITCH model.

CCS reservoirs and transportation

Large-scale deployment of CCS pipelines would require pipeline networks from CO2 sources to CO2 sinks. We require CCS generators that are not near a CO2 sink to build longer pipelines, thereby incurring extra capital cost. If a load area does not does not contain an adequate CO2 sink within its boundaries, a pipeline between the largest substation in that load area and the nearest CO2 sink is built (National Energy Technology Laboratory, n.d.). We assume a pipeline cost of $90/km/ktCO2*yr based on costs for large pipelines found in existing literature (Middleton and Bielicki 2009). CCS plants must send all of their CO2 output to the closest reservoir.

Scenario development For consistency, all scenarios in this study are given identical system conditions and constraints except for those explicitly varied in each scenario. All scenarios include extensive electrification of end-use fuel demand and energy efficiency, as these have been highlighted as fundamental to economy-wide decarbonization (Wei et al. 2013). We disallow new nuclear generation in all scenarios. Demand is treated exogenously and is not price responsive. For our -86% carbon cap scenarios, “-86%” and “-86% No Biomass,” and “-86% No CCS No Biomass,” we reduce WECC-wide power sector emissions by 86% from 1990 levels by 2050 while enforcing existing Renewable Portfolio Standard (RPS) laws. As the electricity sector is likely easier to decarbonize than other sectors, we choose an 86% target because it represents a more aggressive target for the electricity sector relative to an economy-wide 80% reduction target (Wei et al. 2013). We sample hourly demand for each of 50 areas within WECC for six hours of each of 12 representative days in the decades 2020–2050. Investment decisions are made in four periods between 2016-2055; these periods are 2016-2025 (“2020”), 2026-2035 (“2030”), 2036-2045 (“2040”) and 2046-2055 (“2050”). In each modeled hour, demand must be met by the optimization, as well as capacity and operational reserve margin constraints to ensure system reliability. While we allow CCS technologies in the -86% case and -86% No Biomass case, our -86% No CCS No Biomass case excludes such technologies. Like our -86% No Biomass case, our -86% No CCS No Biomass case does not allow biomass.

Modeling very strict carbon caps (including carbon-negative constraints) allows us to understand the role of bioenergy in low-carbon energy systems. Our seven cases, including “-105%,” “-120%,” “-145%,” and four -120% sensitivity scenarios, allow new biomass (including with CCS), ultimately reducing emissions by 105%, 120% or 145% from 1990 levels (Figure 17). Previous research studying economy-wide decarbonization in California found that a 120% reduction in the electricity sector by 2050 was consistent with economy-wide goals, if biomass were to be used for electricity. We believe that a reduction between 105% and 145% is consistent with economy-wide goals for the WECC. As such, these scenarios are meant to simulate intensive development of biomass technology in the power sector. Such development would limit biomass availability in the transportation sector or other sectors.

Figure 17. Carbon caps defined for -86% (86% reduction), -105% (105% reduction), -120% (120% reduction), and -145% (145% reduction) cases through 2050.

To understand deployment of biomass technologies in carbon-negative scenarios, we explore sensitivities to our -120% case through four additional scenarios. These include 1) flexible loads from electric vehicles and buildings (-120% Demand Response), 2) solar costs matching those of the United States Department of Energy’s SunShot Initiative (- 120% Sunshot Solar), 3) a 75% carbon capture rate with CCS technologies (-120% Low CO2 Capture), 4) bioenergy lifecycle analysis (LCA) scenario with solid biomass penalized at 10% of its biogenic carbon content (-120% Biomass LCA). The -120% Demand Response case includes the ability to shift loads within each day of the optimization without cost or efficiency penalty. In the -120% Sunshot Solar case, solar technologies achieve targeted cost reductions set by the U.S. Department of Energy program by 2020 and then remain at these cost levels through 2050 (Mileva et al. 2013). We decrease CO2 capture rates from 85% to 75% in our -120% Low CO2 Capture case (Table 7). Our 10% penalty in the LCA scenario is meant to be conservative, as biomass production, harvest, transport and processing generally consume less than 5% of biomass energy (Rhodes and Keith 2005). Direct emissions from bioenergy may include fertilizer, land use change, or fossil fuel use in harvesting, transport, or conversion. Indirect emissions may include indirect land use change (iLUC) or other market-mediated effects. We expect indirect emissions to be minimal, as feedstocks are primarily wastes or low-input crops grown on marginal lands. In the carbon-neutral cases, we assume that direct emissions from harvesting and transport—a small source of emissions—will be minimized as the entire economy is decarbonized. 39

Scenario 2050 Carbon Cap Core or New Technology (% Emissions Sensitivity Biomass Sensitivities relative to 1990) Scenario for Electricity -86% No -86% Core No No new biomass for Biomass electricity -86% No CCS -86% Sensitivity No No CCS technology No Biomass -86% -86% Core Yes New biomass for electricity -105% -105% Core Yes Same as -86% -120% -120% Core Yes Same as -86% -120% -120% Sensitivity Yes Availability of Demand demand response Response in buildings and electric vehicles for load shifting -120% -120% Sensitivity Yes Solar technology Sunshot Solar cost targets set forth by U.S. Department of Energy -120% Low -120% Sensitivity Yes 75% capture of CO2 CO2 Capture from CCS plants -120% -120% Sensitivity Yes Solid biomass Biomass LCA assigned 10% of biogenic carbon content to account for lifecycle emissions -145% -145% Core Yes Same as -86% Table 7. Key parameters for modeled scenarios. 2.3.3. Additional results from core scenarios

Outputs from SWITCH include hourly dispatch, emissions, costs, fuel consumption, and generation capacity decisions for all modeled technologies. This Section will summarize relevant results for five core scenarios: -86%, -86% No Biomass, -105%, -120%, and - 145%.

Carbon caps and technology availability are the primary drivers of cost of electricity for our results (Figure 18). Costs remain relatively constant from present day through 2035, but generally increase after 2035, as carbon emissions are increasingly limited. After 2035, allowing biomass technologies reduces power costs, from $185/MWh in the -86% No

Biomass case to $134/MWh in 2050 (2013$) in the -86% case. Here, carbon-negative biomass CCS plants allow lower-cost fossil infrastructure to provide energy while still satisfying emissions constraints. The cost of electricity is similar between the -145% case and -86% No Biomass in all periods.

Figure 18. Cost of electricity in each period for core scenarios (2013$/MWh). Biomass technologies reduce total power system costs. Similar trends emerge when examining the carbon cost—an output from SWITCH—across scenarios (Figure 19). Carbon costs increase as emissions are limited, especially after 2035. Biomass technologies reduce the carbon cost between the -86% No Biomass case ($1108/tCO2 in 2050) and -86% case ($121/tCO2). Carbon costs are similar between the - 145% and -86% No Biomass case in 2050. Note that the cost of carbon calculated here represents a small portion of overall system costs due to the extremely stringent carbon caps enforced in this study.

Figure 19. Carbon costs (2013$/tCO2) in each period across core scenarios. Note that the cost of carbon calculated here represents a small portion of overall system costs due to the extremely stringent carbon caps enforced in this study.

System flexibility, emissions, and costs are all drivers of storage installation. In 2050, between 20 and 43 GW of storage is installed (Figure 20). Biomass technologies reduce storage installation, from 41 GW in the -86% No Biomass case to 20 GW in 2050 in the - 86% case. Here, biomass CCS, coal CCS plants, and flexible gas plants limit renewable energy adoption and provide balancing and reserves to the system. In the -105% case, a similar amount of coal CCS is installed relative to the -86% case, while gas is displaced by renewable energy generation from solar. Gas and Compressed Air Energy Storage (CAES) provide flexibility in the -120% case as renewable energy displaces other fossil emitting infrastructure such as coal CCS. However, storage capacity again decreases in the -145% case as the system installs CCGT CCS technology and less CAES for lower-emission flexibility.

Figure 20. Storage installed (GW) in 2050 for each core scenario. The -145% case replaces compressed air energy storage with battery storage in order to achieve lower CO2 emissions. As carbon emissions limitations become increasingly severe, fossil emissions shift from coal CCS to gas CCS. In the -86% case, coal CCS, biomass cofiring, and biomass CCS cumulatively provide 20% of energy generation, enabling lower-cost non-CCS gas resources to generate an additional 22% of generation while still meeting CO2 emission constraints. Other renewable energy such as wind and solar generate 33% of electricity. 43 GW of coal and biomass technologies are installed throughout western North America in 2050 (Figure 21 (a)). In the -145% case, virtually no coal CCS is built. Instead, 47 GW of biomass CCS and gas CCS are installed (Figure 21 (b)). In addition to providing relatively small amounts of electricity generation (10% of the 2050 total), gas combustion turbines without CCS and combined cycle gas turbines (CCGTs) with CCS contribute to balancing as well as planning and operating reserves (Figure 16).

(a)

(b)

Figure 21. (a) Installed biomass, coal, and gas CCS capacity in each load area in 2050 in the -86% case. Biomass and coal generation provide 43 GW of capacity. Biomass CCS (green), coal CCS (grey), coal (black), biomass cofiring (blue), or gas CCS (red) technology is installed in every simulated load area in the WECC. Small amounts of coal operating in California consist of existing cogeneration facilities, many with CCS retrofits. Pie chart area corresponds to the amount of installed capacity in each load area. (b) Installed biomass, coal, and gas CCS capacity in each load area in 2050 in the -145% case. As emissions limitations become more severe, fossil energy technology shifts from coal CCS to gas CCS.

We find that the value of BECCS lies primarily in the sequestration of carbon from biomass, rather than electricity production. To illustrate this point, we explore cases in which BECCS plants capture CO2 emissions but do not produce electricity. The average cost of electricity when BECCS is used exclusively for carbon sequestration is only slightly higher (~6%) than when BECCS provides both sequestration and electricity, while exclusion of biomass from the power sector increases costs by 37% (-86% No Biomass) relative to the -86% case (Figure 22). Removal of CCS technologies increases costs by 63% if biomass is also excluded (-86% No CCS No Biomass). Allowing bioelectricity without CCS (-86% No CCS) decreases costs, but has a smaller effect on both costs and renewable energy deployment than BECCS (-86% case).

Figure 22. Cost of electricity (2013$/MWh) in 2050. Availability of BECCS and sequestration technologies highly reduces the cost of electricity at 86% carbon emissions reduction. Power systems costs when BECCS is used only for carbon sequestration are 6% higher relative to when BECCS provides both sequestration and electricity, suggesting that the primary value of BECCS is from carbon sequestration. Biomass is responsible for between 135 (-86% case) and 165 MtCO2/yr (-145% case) of carbon sequestration. Biomass consumption reaches its highest points in the lowest emission scenarios (Figure 23). Consumption peaks at 1960 PJ/yr in 2050 for the -145% case, which is 98% of available biomass input into the SWITCH model. In all cases where biomass is allowed (- 86%, -105%, -120%, -145%) the power system employs between 90-98% of all biomass available as supply in 2050. Existing, rather than new, dedicated biomass plants consume small amounts of biomass in the -86% No Biomass case. Biomass prices show relatively little variation across load areas in 2050 in the -145% case (Figure 24). While prices reach as high as ~$11.20/MMBtu, they show a central tendency between $2-4/MMBtu.

Figure 23. Yearly biomass consumption (PJ) by period in core scenarios. The vast majority of available biomass is consumed in 2050 for scenarios that allow biomass.

Number of Load Areas (N=50) 10

0 0 2 4 6 8 10 12 Biomass Price in 2050 (2013$/MMBtu) Figure 24. Histogram of biomass prices in 2050 in -145% case in modeled load areas.

Hourly dispatch in 2050 in the -86% No Biomass case depends on hydropower, storage, and gas technologies for balancing. Existing nuclear, geothermal, and biogas provide baseload power (Figure 25). In the -145% case, baseload solid biomass displaces gas technologies to meet CO2 emissions constraints (Figure 16). Dispatch is similar between the -86% No Biomass and -145% cases, with the exception of BECCS technology deployment. Low-carbon scenarios without BECCS and carbon-negative scenarios with BECCS ultimately result in qualitatively similar deployment of gas and renewable generation.

! 300! Jan! Feb! Mar! Apr! May! Jun! Jul! Aug! Sep! Oct! Nov! Dec! 250!

200!

150!

100!

50!

Nuclear! Geothermal! Other Biomass! Coal CCS! Coal! Gas CCS! Gas! Storage (discharging)! Hydro (non-pumped)! Solar! Wind! Storage (charging)! Generation in 2050, -86%in Biomass No Generation Case (GW) System Load! Figure 25. Hourly dispatch in 2050 in the -86% No Biomass case. Power system dispatch is shown in sampled hours from two days (peak and median day) for each month of the year. “Other Biomass” includes both liquid and gaseous biomass supplies.

2.3.4. Additional biomass cofiring results

The deployment of biomass cofiring technologies depends on the cap on carbon emissions. In all cases, cofiring is employed as a near- to mid-term (through 2030) mitigation technique via retrofit of existing coal plants. Under moderate emissions caps (less than or equal to -105%) cofiring with CCS is installed on new coal CCS plants (Figure 26), but we do not observe this behavior in the -120% and -145% cases. While biomass cofiring provides short-term carbon emissions reductions at existing coal plants, under strict emissions caps the emissions from coal CCS are too large to support cofiring.

Figure 26. Biomass cofiring installation in each period for the -86% case and -145% case, including both retrofits of existing plants and co-installation on new plants. We simulate biomass cofiring retrofits in detail, and are able to distinguish between retrofits of existing coal plants and cofiring co-installation on new coal plants. In our -86% case, retrofits of existing plants comprise the majority of biomass capacity installation in early years, while co-installation on new plants drives deployment in later years (Figure 26). Co-installation occurs at the same time as plant installation, rather than retrofit of new plants. In total, 2800 MW of cofiring capacity exists in 2050 in the -86% case. In the -145% case, we observe greater levels of retrofits in early periods, but the system retires all coal and cofiring capacity by 2050 to satisfy stringent emissions limits. 2.3.5. Sensitivity results for carbon-negative power systems

To further understand biomass cofiring and BECCS deployment in low-carbon power systems, we simulate five scenarios with 120% reduction by 2050 (-120%, -120% Demand Response, -120% Sunshot Solar, -120% Low CO2 efficiency, and -120% Biomass LCA). This Section will summarize results from the four additional scenarios based on the -120% case (Figure 27 and Figure 28). While technology cost, lifecycle CO2 emissions, and performance assumptions in carbon-negative power systems alter the relative deployment of coal CCS and intermittent renewables, they have little effect on biomass deployment.

Figure 27. Average generation (GW) in 2050 in sensitivity scenarios (-120%).

150!

100! ! 50! /yr) 2 Coal! Coal CCS! 0! CCGT CCS! -120%! -120% -120% -120% Low -120% Demand Sunshot CO2 Capture! Biomass CCGT! Response! Solar! LCA! Gas Combustion Turbine! -50! Compressed Air Energy Storage! Biomass Coﬁring!

Emissions in 2050 in (MTCO Emissions Biomass CCS! 2

CO -100!

-150!

-200! Figure 28. Yearly carbon emissions (MtCO2/yr) in 2050 under sensitivity scenarios (-120%). Demand response technologies allow greater penetration of solar technologies, while reducing generation from wind and storage requirements (-120% Demand Response). Flexible loads are shifted to be coincident with daytime-peaking solar energy technologies, increasing solar generation while decreasing the need for electrical energy storage. While similar amounts of biopower are installed, more biomass is used for biomass cofiring with coal CCS instead of BECCS. CO2 emissions primarily occur from CCGTs and gas turbines.

Lower levels of CO2 capture from CCS plants (-120% Low CO2 capture) reduce the amount of coal CCS and biomass cofiring in 2050, but do not reduce the amount of BECCS capacity (~7% of 2050 generation). As BECCS sequesters less CO2 than at base technology assumptions, renewable energy generation increases to satisfy stringent carbon limitations. If higher levels of CO2 capture from coal are not available, coal will not be able to achieve the level of carbon reduction necessary to contribute to a low-carbon power system.

Accounting for potential lifecycle emissions from biomass (-120% Biomass LCA) reduces total carbon sequestration from BECCS, but does not reduce installed BECCS capacity. Instead, larger amounts of gas CCS, solar, and wind displace gas and coal CCS technologies, which have higher CO2 emissions factors. Non-BECCS renewable energy deployment increases compensate for fewer negative emissions (from 77% to 82% of 2050 generation). Accounting for lifecycle emissions from bioenergy, at levels assumed here, has a relatively small effect on generation portfolios and BECCS deployment. 51

Low solar technology costs in the -120% Sunshot Solar case increase storage and solar energy deployment. Solar energy provides 29% of generation, complimented by 66 GW of storage capacity. Emissions from CAES increase, while coal CCS and gas turbine emissions decrease.

2.3.6. Comparison of net energy of bioelectricity and biofuels

We also calculate the net energy of bioelectricity and biofuels based on our heat rate values and commonly accepted values for candidate cellulosic ethanol conversion pathways (Campbell, Lobell, and Field 2009; Farrell et al. 2006). Values for agricultural, transport, and biofuel conversion process energy are based on the Energy and Resources Group Biofuels Analysis Meta-Model (EBAMM), a lifecycle assessment of ethanol conversion pathways. As in EBAMM, we assume a solid biomass feedstock energy density of 18 MJ/kg for our comparison, which is slightly higher than the assumptions used in our supply estimates. Depending on generation technology, bioelectricity contains 28-45% of the net energy of cellulosic ethanol. However, bioelectricity provides as much as 41% more transportation miles than cellulosic ethanol because of the efficiency of battery electric drive vehicles over internal combustion engines (90% instead of 29%) (Table 8). As reported by Campbell, “Co-product credits in EBAMM favor the ethanol pathway by accounting for ethanol co-products but not potential bioelectricity co-products, including steam for heat and fly ash for cement.”

Technology Net Energy of % Net Efficiency Transportati % Increase in Fuel or Energy of of Primary on Motor transportation Electricity Cellulosic Conversion Efficiency distance over (MJ/ton Ethanol Step Cellulosic Ethanol biomass)

Biomass 3945 45 .34 .90 41 Cofiring Biomass 3156 36 .27 .90 13 IGCC Biomass 2935 34 .25 .90 5 Cofiring with CCS Biomass 2417 28 .21 .90 -14 IGCC with CCS Cellulosic 8676 -- .45 .29 -- Ethanol (with co- product credit)

Table 8. Comparison of bioelectricity production options to cellulosic ethanol. While cellulosic ethanol provides more net energy than bioelectricity, bioelectricity generally provides greater transportation distance due to more efficient drivetrain technology.

We also use the carbon intensity values in EBAMM for gasoline and cellulosic biofuels to estimate the amount of carbon reduction possible should biomass be used for biofuels instead of electricity. By displacing gasoline with cellulosic biofuels, 2000 PJ of annually available biomass can provide 75 MtCO2/yr of carbon reduction. Specifically, we assume a carbon intensity of 11 gCO2e/MJ for cellulosic biofuels and 94 gCO2e/MJ for gasoline. In contrast, if biomass is made available to the power sector, BECCS can sequester 165 MtCO2/yr and also displace fossil electricity. EBAMM includes direct emissions from biomass fertilization, harvesting, transport, and conversion, but excludes emissions from soil organic carbon (SOC), as well as direct and indirect land use changes. While this study does not directly address the question of biomass allocation between competing uses outside of the electricity sector, it suggests that biomass can play a large role in low-carbon power systems.

3. Excerpt from Emissions accounting for biomass energy with CCS (Sanchez et al., 2015b)

To the Editors (Gilbert and Sovacool 2015) Sanchez et al. provide a viable technological roadmap for using biomass energy with carbon capture and storage (BECCS) in the western United States (Sanchez et al. 2015a). However, they oversimplify emissions accounting by assuming a zero or negative carbon emissions factor. Accounting for total lifecycle emissions is perhaps the greatest challenge in deploying biomass (in solid, gaseous, or liquid form) to reduce carbon emissions (Timothy Searchinger et al. 2008; Tilman et al. 2009).

When utilized to generate electricity, emissions sinks and sources for biomass occur in two different sectors. As plants grow, they take up CO2 and store it. When combusted, the stored CO2 is released and contributes to emissions. Accordingly, counting the emissions factor for biomass electricity generation as zero, or negative in the case of BECCS, causes double-counting of emissions on a cross-sectoral basis (Tim Searchinger and Heimlich 2015). These accounting challenges persist when developing national or international carbon control regimes (T. D. Searchinger et al. 2009). Further, in a carbon-constrained world, both biomass producers and electricity generators will have competing claims concerning monetization of their low-carbon attributes.

Sectoral accounting is further complicated by the timing of emissions in the biomass electricity lifecycle. Power generation releases CO2 that was previously sequestered — and an implicit assumption made by Sanchez et. al. is that harvested biomass provides room for re-growth and sequestering of released emissions. This assumption, however, raises two problems.

First, if regrowth does not occur, net emissions will increase, even if CCS confines the majority of emissions. Measurement and verification are needed to ensure biomass is regrown and net negative emissions actually occur.

Second, the rate of CO2 uptake from biomass fuel sources varies considerably. Trees — the dominant source of utility-scale biomass fuel today — grow over decades with different CO2 uptake rates at different ages and across species. Ricke and Caldeira recently found that the climate impact of CO2 emissions could occur in as few as 10 years (Ricke and Caldeira 2014). The CO2 released by uncontrolled biomass burning can thus contribute to short-term radiative forcing before CO2 is sequestered by regrowth.

The concerns we raise suggest that additional, nuanced, and refined research is needed to improve our understanding of carbon flows in BECCS, develop efficacious legal regimes for CO2 emissions reduction ownership, and design successful monitoring regimes for biomass regrowth. Only then can the future role of bioenergy and BECCS be more fully contextualized and appreciated.

Reply (Sanchez et al. 2015b) Our Letter (Sanchez et al. 2015a) assesses the impact on regional carbon emissions if biomass energy is used to replace fossil fuels in the electricity system, and carbon capture and storage (CCS) is used to sequester most of the emissions associated with electricity production. Our bioenergy assessment prioritizes, but does not rely solely on, bioenergy production from wastes and residues, which do not compete with food crops and would otherwise be burned or left to decompose, releasing their carbon into the atmosphere as CO2. Most of these feed stocks minimize the impact of biomass regrowth and uptake rates. For feed stocks such as forest residues that may take many years to grow back, there will be some amount of short-term radiative forcing that was not accounted for in our analysis.

As Gilbert and Sovacool suggest (Gilbert and Sovacool 2015), it is important not to count the same CO2 emissions reductions in two separate sectors when quantifying economy- wide emissions. Our analysis avoids this accounting error by using a simplified methodology, ascribing all changes in atmospheric CO2 — from plant growth to combustion in a bioenergy and CCS (BECCS) plant — to the electricity sector. Should BECCS be adopted widely, it will be important to allocate emissions credits among all relevant actors across sectors.

Several forms of complementary analyses inform roadmaps for sustainable bioenergy production. In addition to the bottom-up engineering-economic analysis performed in our Letter (Sanchez et al. 2015a), our team at the Renewable and Appropriate Energy Laboratory at UC Berkeley, USA, and others, have engaged in commodity-chain theoretical bioenergy analysis, producing quantitative indirect land-use change estimates (Hertel et al. 2010), and evaluations of previous efforts through meta-analysis (Plevin and Kammen 2013). Based on this work, we agree with Gilbert and Sovacool (Gilbert and Sovacool 2015) that monitoring and verification should be a critical part of any long-term strategy for mitigating climate change. Of particular concern is whether the cultivation or extraction of 54 biomass for energy will degrade or enhance the ecological productivity and related carbon flows of the land (Liska et al. 2014).

Each step of energy extraction, preparation, combustion, and disposal demands a rigorous assessment of carbon impacts. This statement applies not only to bioenergy, but also carbon capture technologies. In addition to oft-cited concerns about sustainable bioenergy production, risks of CO2 leakage from long-term geologic sequestration raises additional uncertainties about BECCS and other CCS strategies (Benson et al. 2012). However, the choice of counterfactual is critical to any bioenergy analysis, including assumptions of population, future diet, and crop productivity (Slade, Bauen, and Gross 2014). Recent research shows that biofuel production can provide emissions benefits over non-bioenergy land-use decisions, including forest recovery on marginal land (Evans et al. 2015). Geologic storage of carbon through CCS can proceed for decades and potentially millennia if properly managed, which may be more stable than other carbon sequestration options from biomass. The emissions benefits of BECCS — encompassing displaced fossil-fuel CO2 emissions from energy production and geologic CO2 sequestration — may improve the desirability of biomass production for bioenergy over other land-use decisions, but more research is needed to directly compare it with other sequestration strategies. Moving forward, supportive policy should incentivize land-use decisions that are beneficial for the climate (Lemoine et al. 2010).

4. Conclusions

In this Chapter, we examine regional deployment of BECCS, focusing on the WECC. We also discuss emissions accounting for BECCS. Three primary results emerge:

1. Net-negative emissions electricity systems: BECCS, combined with aggressive renewable deployment and fossil emission reductions, can enable a carbon-negative power system in Western North America by 2050, with up to 145% emissions reduction from 1990 levels. This result is robust to realistic physical constraints, represented by the SWITCH model, which IAMs often lack. The amount of biomass resource available limits the level of fossil CO2 emissions that can still satisfy carbon emissions caps.

2. The value of BECCS in mitigation: Offsets produced by BECCS are more valuable to the power system than the electricity it provides. This confirms results from IAMs: BECCS is very valuable in the context of decarbonization, and can offset CO2 emissions from fossil energy across the economy. The availability of BECCS and a high supply of lignocellulosic biomass for bioenergy are projected to be critical requirements for achieving stringent mitigation targets and containing the associated costs (Bauer 2015).

3. Sustainability and life-cycle accounting: Three primary issues complicate emissions accounting for bioenergy with CCS: cross-sector CO2 accounting, regrowth, and timing. Our bioenergy assessment prioritizes bioenergy production

from wastes and residues, which would otherwise be burned or left to decompose, releasing their carbon into the atmosphere as CO2. The emissions benefits of BECCS — encompassing displaced fossil-fuel CO2 emissions from energy production and geologic CO2 sequestration — may improve the desirability of biomass production for bioenergy over other land-use decisions, but more research is needed to directly compare it with other sequestration strategies.

This work informs near-term efforts to build facilities, integrate them into existing energy and power systems, respect regional design considerations, and complement existing energy and climate policy. It also informs decarbonization pathways for other regions considering implementation of negative emissions technologies.

However, there are additional research needs to inform regional deployment. First, additional analysis can include a more detailed representation of BECCS technologies, including different production pathways, CO2 capture amounts, and potential economies- and diseconomies-of-scale. This could help prioritize systems design and R&D efforts. Second, future research can examine BECCS deployment across energy sectors, including electricity, transportation, heating, and industry. Biomass allocation, with and without CCS technology, is a contested and unresolved issue (Mullins et al. 2014). Finally, future modeling efforts can enhance representation of the land economy, as well as carbon cycles. This can help ensure that BECCS deployment does not interfere with sustainability, biodiversity, or food security goals.

Chapter III. Communication of carbon-negative energy systems

1. Preface

This Chapter discusses communication of BECCS. In particular, it documents a specific instance of BECCS-focused communication, the development of a journal article introducing BECCS to a middle-school audience. The journal of publication, Frontiers for Young Minds, is a non-profit scientific journal for which young people serve not only as the target audience, but also as critical participants in the review of manuscripts written by expert researchers. The end result is a journal of freely available scientific articles that are written by leading scientists and shaped for younger audiences by the input of their own peers (“Frontiers for Young Minds” 2015). Other journals such as Earth Science Journal for Kids have similar missions.

This article, ‘Removing harmful greenhouse gases from the air using energy from plants’ (Sanchez and Kammen 2015), resulted following the February 2015 publication of ‘Biomass enables the transition to a carbon- negative power system across western North America’ (Sanchez et al. 2015a). Prof. Berend Smit, an Associate Editor of Frontiers for Young Minds who researches CCS at the University of California-Berkeley, reached out the Authors to develop a “New Discovery” article explaining the content and importance of a recent discovery in language that can be understood by younger audiences. These articles are based on an academic article and written by researchers involved in the original publication.

The article introduces several concepts to its audience. First, it describes climate change and climate change mitigation. Then, it introduces the concept of carbon dioxide removal (CDR) as one form of mitigation. It describes BECCS technology, and concludes by describing their “New Discovery,” carbon-negative power systems. Along the way, it introduces the concept of an electricity system, and explains how carbon-negative power systems are very different than existing power systems. The authors strove to explain these concepts in a simple language as possible, and to develop resulting figures that can be used by the entire community to introduce BECCS.

More broadly, this article fills two critical needs, which are not discussed in the text of the article itself. First, it builds a ‘community of support’ for BECCS. Existing research has identified a key problem for BECCS as cultural, lacking in a community of support, awareness and credibility amongst its own key stakeholders and the wider public (Dowd, Rodriguez, and Jeanneret 2015). This community of support comprises of two sub-groups: elites, and the broader public.

Environmental education can play a role in closing the achievement gap, the observed, persistent disparity of educational measures between the performance of groups of students, especially groups defined by socioeconomic status, race/ethnicity and gender. Specifically, the environmental can serve as an “integrating context” for learning 57

(Lieberman and Hoody 1998). Recent academic science standards, such as Next Generation Science Standards (NGSS), have included the environment formally in science curriculum (National Academy of Sciences 2012). Contributory action from the broader public is essential for achieving just and effective environmental outcomes (Shallcross and Robinson 2008). Thus, tailored introductions to environmental technologies like BECCS can improve environmental education outcomes and contributory action.

Elites are a second group involved in communities of support. In the policy process, communities of elites can participate in ‘Advocacy Coalitions’ to affect change (Nowlin 2011). More broadly, elites are key stakeholders in many organizations involved in Energy Technology Innovation System (K. S. Gallagher et al. 2012). Support, credibility, and awareness among elites will be critical for commercializing carbon-negative energy.

Second, straightforward description can increase public familiarity with and decrease hostility towards BECCS. CCS technologies are controversial: Wong-Parodi et al. summarize attitudes towards CCS as negative, similar in scale and risk to nuclear energy, or seen as pointless when other alternatives exist (Wong-Parodi et al. 2011). Bioenergy suffers from similarly negative attitudes, particularly in Europe (Halder et al. 2010).

Experience on several continents has shown that local resistance to CO2 storage projects can lead to cancellation of planned CCS projects. Inhabitants of the areas around geological storage sites often have concerns about the safety and effectiveness of CCS. Early engagement of communities in project design and site selection, along with credible communication, can help ease resistance. Environmental organizations sometimes see CCS as a distraction from a sustainable energy future, despite future prospects for BECCS (Benson et al. 2012).

Trust is an important concept when discussing the transformation of energy systems (Rayner 2010). Experiences with technologies such as genetically modified foods suggest that if public acceptance of a technology is lacking, large-scale global deployment is limited (Kalaitzandonakes, Marks, and Vickner 2005). Likewise, public perception and support will be critical if CCS and BECCS technology are to achieve their potential for climate change mitigation.

Given these needs, simple, accessible, and balanced introductions to BECCS can help to build awareness and credibility in the wider public.

2. Excerpt from Removing harmful greenhouse gases from the air using energy from plants (Sanchez and Kammen 2015)2

2 © 2015 Sanchez and Kammen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

2.1. Main text

Figure 29. An artist’s depiction of bioenergy with carbon capture and sequestration (BECCS).

We know that the Earth’s atmosphere acts like a greenhouse: rays from the sun enter, heating the air and the ground. Instead of leaving the greenhouse as they came in, this heat is trapped by the walls and ceiling, making the greenhouse warmer than the air outside. This is just like a car on a hot day: light passes freely through the window (more than 90% gets through), but after being absorbed by the seats and upholstery, this energy is re- released as heat, which the windows reflect back into the interior. Hence, the heat can enter but cannot escape and the car will get warmer. Earth has its own greenhouse – an atmosphere, a blanket of air that traps heat through a process known as the “greenhouse effect.” The greenhouse effect is a good thing – without an atmosphere the earth would be about 33°C (60°F) colder and most of the water on the planet would be ice! The gases that heat the earth are known as greenhouse gases, the most well known of which is carbon dioxide (chemically written as CO2).

However, too much of CO2 in our air can be a bad thing. Humans are responsible for producing large amounts of CO2. Every day we use fossil fuels (such as coal, oil, or natural gas, which we find deep underground in solid, liquid, or gas forms) in our cars or power plants, or we cut down forests. All these activities combined have caused the amount of CO2 in our atmosphere to increase to levels not seen on earth in 55 million years. This increase in greenhouse gases heats the earth. The increase in temperature causes climate change, which is a change in average worldwide or regional weather patterns. Scientists project that climate change will cause a rise in sea level, more intense heat waves, extreme weather, species extinction, and other negative impacts on our world.

Luckily, there are several steps that we can take to reduce the impacts of future climate change. Scientists generally divide these helpful actions (known as “climate change

mitigation”) into three categories: reducing CO2 and other harmful greenhouse gas emissions (the release of these gases into the atmosphere), reducing the amount of sunlight that reaches the earth’s surface, or removing CO2 from the atmosphere. Each of these actions has different costs, benefits, and risks. It is important for governments and policymakers to understand how large of a part each of these solutions can play in society’s response to climate change.

Let us focus on carbon dioxide removal (CDR) from the atmosphere. How much CO2 removal might we need to do in order to fight climate change? One estimate comes from the Intergovernmental Panel on Climate Change (IPCC), a group of scientists from around the world working together to answer these questions. The IPCC report contained a strict plan for limiting temperature increases so that climate change could be slowed. This IPCC plan includes both ways to decrease our CO2 output, as well as ways to remove carbon from the atmosphere using CDR technologies. In fact, this plan requires CDR technologies to remove an amount of CO2 equal to 25% of the amount produced worldwide, by the year 2100. The rest of the CO2 reductions would have to come from other actions, including new energy sources that do not produce CO2, such as wind and solar power. Figure 30 shows what CO2 emissions might look in the future if we act to reduce climate change (“climate change mitigation”) or if we choose not to reduce emissions (“business as usual”).

Figure 30. The world’s carbon dioxide emissions for “Business as Usual” and “Climate Change Mitigation” scenarios. The figure shows how CO2 emissions (in PgC/year, an amount of CO2 emitted every year) evolve in the future. In the “Business as Usual” scenario, we assume that the world continues to use fossil fuels the same way as we have done in the past, whereas in a “Climate Change Mitigation” scenario, we assume that governments take actions to reduce CO2 emissions. Adapted from (IPCC 2014a).

There are several ways to perform CDR. Each of these ways has different costs and would remove a different amount of CO2 from the atmosphere. One idea is called direct air capture of CO2, an expensive procedure that “scrubs,” or removes, carbon dioxide directly from the air we breathe. Other ways of removing CO2 from the atmosphere include storing more carbon in trees, soils, or the ocean. One new technology for removing CO2 is known as bioenergy with carbon capture and sequestration (BECCS). BECCS removes CO2 from the atmosphere while producing valuable energy products, such as electricity or fuels. What is more, BECCS can also replace dirtier forms of energy production from fossil fuels. Because of these characteristics, scientists believe that BECCS could play a large role in reducing future climate change.

How does BECCS work, and why does it remove carbon dioxide from the atmosphere? Let us look at how this new technology removes harmful greenhouse gases from the air using energy from plants.

Humans have been producing energy from biomass – material derived from living things, such as plants, grasses, or trees – for their entire existence. We call this energy from biomass bioenergy. For example, early humans produced heat from plants using fire, just like you do when you build a campfire. Today, we use biomass to produce heat, electricity, or transportation fuels around the world. Biomass can be a carbon-neutral energy source, which means that it does not increase the amount of CO2 in the atmosphere. This is because

the CO2 released when producing energy is the same as the CO2 taken from the air to grow the biomass (Figure 31 (a)). This CO2 is taken from the air during plant growth. However, biomass is only carbon-neutral if grown sustainably – that is, if the biomass regrows after it is harvested and does not cause emissions from land (such as cutting down forests). Like wind and solar energy, bioenergy can be an important way to reduce the amount of CO2 created when we produce energy.

Figure 31. Diagram of (a) a bioenergy power plant, and (b) a bioenergy with carbon capture and sequestration (BECCS) power plant. Plants remove CO2 from the atmosphere while growing. If we harvest the biomass and burn it in a power plant we can generate electricity. With BECCS,

we do not release all the CO2 into the atmosphere, but we separate the CO2 (capture) and store it underground (sequestration).

Carbon capture and sequestration (CCS) is a CO2 management technique that captures CO2 from a source (such as a power plant), transports it to a storage site, and stores (sequesters) it underground (Smit et al. 2014). As a result, CO2 emissions that would normally enter the atmosphere are instead sequestered atmosphere, CCS reduces future climate change. Most scientists imagine using CCS in existing or future power plants that burn fossil fuels (such as coal, oil, and natural gas), to reduce CO2 emissions from these plants by as much as 90%. Although scientists and engineers have built the technology necessary for CCS, there are still not many power plants using this technology. In the future, we will need CCS in our electricity system, which makes electricity and delivers it to our homes. The electricity system is responsible for a large amount of the world’s CO2 emissions.

Bioenergy with carbon capture and sequestration uses both bioenergy and CCS to produce energy and remove CO2 from the atmosphere (Figure 31 (b)). Instead of releasing CO2 during energy production, as traditional bioenergy technologies do, CO2 is instead captured and stored using CCS technologies. The result is electricity or fuels that remove, rather than release, CO2. BECCS is different from other CDR technologies in that it produces energy, instead of only removing CO2 from the air. As a result, it could also replace power plants operating on fossil fuels, such as coal, oil, or natural gas.

Energy and climate scientists, like those at the IPCC, are just beginning to understand the role that BECCS can play in reducing climate change. Computer models used by the IPCC predict that BECCS and other CDR technologies will be very useful, but they are still missing several very important details (IPCC 2014a). For example, we would like to know how much biomass is available for energy production, and where it is located. This is important because it tells us how much energy we might be able to create. We also need to understand where we might store carbon dioxide underground. These locations for underground sequestration might not be in the same place as the biomass we need! Finally, we must understand the systems where BECCS might be used, like the electricity system that delivers electricity to our homes. Understanding and making sense of all of this information is a job for engineers, geologists, and other scientists across the world.

To get a better understanding of how BECCS might be used, we used computer models to explore using BECCS in the electricity system of Western North America, which includes parts of the United States, Canada, and Mexico (Sanchez et al. 2015a). This model collects important information about where biomass is located and possible locations for underground sequestration. The model also contains important details about how engineers can plan the electricity system in the future. All of the information in the model requires large supercomputers, which are computers with a very high level of computational capacity. These supercomputers can be 50 million times faster than your laptop! Using supercomputers can help us to understand BECCS better than we can use IPCC models.

The first step in understanding how BECCS might be used is to understand how much biomass is available, and where it is located. Biomass contains less energy than the same amount of fossil fuels, and is more spread out around the earth than the concentrated fuels from coal mines, or oil and gas fields. Western North America contains biomass from forests, agriculture, and the wastes we produce, like wood from when we build or tear down buildings. This supply of biomass is limited by its ability to grow back after being used, the steps we take to ensure that we do not degrade our land, water, or air, and the amount of available land to grow plants. Other scientists interested in energy from biomass have built models to estimate biomass availability. We use these existing estimates to create a biomass “supply curve” (Figure 32). This supply curve shows the price and quantity of available biomass. The amount of biomass available for energy production using BECCS could fill about 10% of the electricity demand of Western North America in 2050.

Figure 32. A supply curve for biomass. This figure shows the quantity of biomass available, and its cost. We classify biomass as wastes (what we throw away), residues (wastes left on our farms or in our forests), or dedicated feedstocks (plants that we grow explicitly for bioenergy). Adapted from (Sanchez et al. 2015a).

Using the data that we can get from our models, we explore the possibility of using BECCS to reduce climate change caused by the electricity system. At the same time, we look at other technologies, such as coal with CCS, and renewable energy technologies (meaning sources that do not “run out,” such as wind and solar). We find that BECCS, combined with a large amount of renewable energy from wind and solar power, and far less fossil fuel CO2 emissions, could create an electricity system in Western North America by 2050 that takes more carbon dioxide out of the air than it puts in. These carbon dioxide removing power systems of the future would rely on renewable resources, including biomass, for as much as 88% of the electricity produced in 2050 (Figure 33). The largest amount of electricity generation would come from wind energy. The remaining electricity would be provided by fossil fuel power plants, especially those using natural gas and CCS.

Figure 33. (a) Electricity produced (average generation in GW, an amount of energy over 1 year), and (b) CO2 emissions (MtCO2/year, an amount of CO2 emitted every year) in 2050 for an electricity system in Western North America with negative emissions. BECCS allows this system to achieve negative emissions.

What is more, since BECCS removes CO2 from the atmosphere, it might allow some CO2 emissions from fossil fuels to continue in the future. Since reducing emissions from our cars or factories can be expensive, BECCS may reduce the cost of slowing climate change. However, the amount of CDR is limited by how much biomass is available.

Despite the promise of BECCS, there are still several risks and uncertainties that scientists need to understand in the future. One key issue is the sustainability of biomass needed for BECCS. Efforts to grow more plants or increase biomass availability can impact our environment in negative ways. Planting more plants can increase demand for water, land, and fertilizer. For example, getting enough plants to meet our supply estimate could require a piece of land the size of the state of West Virginia! Another risk is that underground storage of CO2 may not last forever. Carbon dioxide sequestered from CCS could leak into the atmosphere, causing climate change.

The biggest uncertainty with BECCS, however, is that we have little experience building and operating these systems. Who will want to build these facilities? How large will we want to build them? How do we make sure that they make money? How much will they cost? And how can governments support this technology? In the future, we will need to make sure that BECCS technology works efficiently, does not cost too much, and is easy to build and use. By doing this, tomorrow’s scientists and engineers can make this technology a reality.

2.2. Glossary

Climate change: a change in average worldwide or regional weather patterns.

Climate change mitigation: actions that reduce the impacts of future climate change.

Carbon dioxide removal: actions that remove CO2 from the atmosphere, which will reduce future climate change.

Intergovernmental Panel on Climate Change: a group of scientists from around the world that assesses information to understand the risks climate change.

Bioenergy with Carbon Capture and Sequestration: a technology that removes CO2 from the atmosphere while producing energy products like electricity or fuels.

Bioenergy: the production of energy from biomass. This energy is derived from living things, such as plants, grasses, or trees, and can be carbon-neutral energy.

Carbon-neutral energy: energy that does not produce any CO2 emissions during its production or consumption.

Carbon capture and sequestration: a carbon dioxide management technique that captures CO2 from a source, transports it to a storage site, and stores it underground (sequesters it). Sequestering CO2 underground prevents it from entering earth’s atmosphere.

Electricity system: a network that supplies, transmits, and uses electricity.

3. Conclusions

This Chapter documents the process and importance of the development of a journal article introducing BECCS to a middle-school audience. Transparent, simple descriptions of BECCS can build awareness and credibility in the wider public and among central stakeholders. This process occurs in two ways. First, it builds a ‘community of support’ for BECCS. Second, it increases public familiarity with, and decreases hostility towards, BECCS.

We conclude simply, noting that exposition can describe the following topics to a middle- school audience: climate change, mitigation of climate change, carbon dioxide removal, BECCS, and the electricity system. Along the way, it explains how carbon-negative power systems are very different than their current configuration. The authors strove to explain these concepts in a simple language as possible, and to develop resulting figures that can be used by the entire community to introduce BECCS.

Chapter IV. Design of carbon-negative energy systems

1. Preface

This Chapter explores practical design issues for BECCS. Specifically, it focuses on system scale, identifying optimal scale using spatially explicit infrastructure optimization for the State of Illinois. Bioenergy facilities, including BECCS facilities, likely exhibit economies of scale in capital costs, but diseconomies of scale in biomass transportation and supply costs (Wright and Brown 2007). Previous high-resolution modeling of BECCS deployment, including the experimental procedure in Chapter II, has treated BECCS capital costs as linear, despite the likelihood of non-linear costs (Sanchez et al. 2015a).

Scale is an important design consideration for planners, project developers, and policymakers. Project developers can maximize their revenues—or minimize costs—by building projects near optimal scale. For policymakers, policies specifying project size (such as minimum or maximum scale for subsidies) may sacrifice important economic efficiencies and ratepayer benefits should they be improperly designed. Scale also has implications for sustainability, as the size of facilities will affect the nature and location of lifecycle impacts (Bomberg, Sanchez, and Lipman 2014).

This effort is novel in several respects. First, we focus on BECCS, an emerging bioenergy technology for climate change mitigation, which has not been subject to detailed examination of practical design issues. Second, we develop a spatially-explicit model for optimal scale, which leverages publicly available biomass supply and transportation databases. Existing process engineering studies often lack consideration of environmental constraints on facility design, including the spatial availability of biomass. Diseconomies of scale are often considered using heuristics, such as constraints on biomass delivery rates (Larson et al. 2010; Mike Matuszewski, James Black, and Eric Lewis 2013).

More broadly, this work informs the engineering, logistics, and systems planning for building BECCS facilities. It has application to entrepreneurs, power plant designers, and power system planners involved in the implementation of first-of-a-kind BECCS plants.

2. Excerpt from Optimal Scale of Bioenergy with Carbon Capture and Storage (BECCS) Facilities (Sanchez and Callaway)

2.1. Introduction

The urgency of climate change has led to societal pressure not only for technologies that reduce CO2 emissions, but also those that can reduce the net amount of CO2 in the atmosphere (Benson 2014; Hansen et al. 2008; Fuss et al. 2014). These technologies, collectively referred to as Carbon Dioxide Removal (CDR) options, include direct air capture of CO2, biochar, afforestation, soil carbon sequestration, ocean fertilization, and bioenergy with carbon capture and sequestration (BECCS) (Lenton 2010). Deploying BECCS results in a net reduction in atmospheric carbon, and may be an important 67 technology for dealing with abrupt climate change (Read and Lermit 2005). Currently, BECCS is being deployed at commercial scale in ethanol production facilities (Gollakota and McDonald 2012). Previous work, primarily using Integrated Assessment Models (IAMs), has identified the critical role of BECCS in long-term climate change mitigation, particularly should carbon-negative technologies be required (Milne and Field 2014; IPCC 2014b). Additionally, recent work exploring BECCS deployment low-carbon power systems indicates that BECCS could be a key technology for aggressive decarbonization in pre-2050 timeframes (Sanchez et al. 2015a; Christopher Yang et al. 2014). This work, however, has not focused on practical design issues for BECCS facilities, including systems scale.

Bioenergy facilities, including BECCS facilities, likely exhibit economies of scale in capital costs, but diseconomies of scale in biomass transportation and supply costs (Wright and Brown 2007). The intuition behind capital cost economies of scale is that the capacity of different plant components (e.g., boilers) is a function of volume, while the cost of these components is a function of the material involved, which will scale at a rate closer to the component surface area. For a perfect sphere, volume grows as a cubic function of radius while surface area grows quadratically, so facility capital costs exhibit economies of scale. In contrast, transport costs exhibit diseconomies of scale because as plant capacity increases, feedstock must be hauled from longer distances. Profit-maximizing or cost- minimizing producers must also buy more expensive sources of biomass to satisfy plant demand. Probing these tradeoffs results in an optimal scale for BECCS facilities, which minimize average costs for a single facility, or total costs for a portfolio of facilities. Previous high-resolution modeling of BECCS deployment has treated BECCS capital costs as linear, despite the likelihood of non-linear costs (Sanchez et al. 2015a).

Two key factors affecting optimal scale are biomass availability and cost. Estimates of lignocellulosic feedstock supply for bioenergy production are influenced by assumptions about available land, diet, population, and yield increases (Slade, Bauen, and Gross 2014). The spatial distribution and cost of bioenergy resources have been estimated via a variety of methods, resulting in detailed inventories in the continental United States (Downing et al. 2011). These publicly available biomass inventories can promote transparency and consistency in biomass and bioenergy analysis.

Biomass is less energy-dense and more spatially distributed than fossil fuels or other forms of renewable energy such as solar and wind, increasing the importance of management and logistics in cost-effective bioenergy supply. Optimization via linear programming, non- linear programming, and mixed-integer linear programing can help design bioenergy supply chains subject to constraints on supply and sustainability (N. Parker 2012). In this paper we use large spatial datasets to characterize the drivers of optimal sizing, both for a single facility and multiple facilities. This framework can be applied to a broad range of bioenergy technologies, including BECCS, and to diverse geographic areas.

We then apply this framework to optimally size BECCS facilities in Illinois, leveraging county-level biomass supply data, detailed road transportation networks, existing technology cost estimates, and previous geologic characterizations for long-term CO2 storage. Illinois contains relatively plentiful low-carbon biomass resources from corn 68

stover and other crop residues, as well as excellent geologic sequestration potential for CO2 in the Illinois Basin (Larson et al. 2010). For example, the Illinois Industrial Carbon Capture and Storage Project in Decatur, IL is one of the first commercial applications of BECCS in the world (Hnottavange-Telleen, Krapac, and Vivalda 2009). Several other commercial- scale CCS projects, such as the FutureGen Coal CCS facility in Meredosia, IL, are being planned around the state (Zitney et al. 2006).

Our study is novel in several respects. First, we focus on BECCS, an emerging bioenergy technology for climate change mitigation, which has not been subject to detailed examination of practical design issues. Second, we develop a spatially-explicit model for optimal scale, which leverages publicly available biomass supply and transportation databases. Relatively few studies have studied economies-of-scale in a spatially-explicit context. Those that have rely on self-generated, rather than publicly available, biomass availability and cost estimates, and have studied conventional, rather than emerging, biomass facilities (Sultana, Kumar, and Harfield 2010; Sultana and Kumar 2012).

We focus on BECCS systems for electricity production. Biomass can be converted to electricity via two methods: 1) direct combustion, and 2) gasification (Farrell and Gopal 2008). Different conversion methods have varying degrees of permanence, CO2 fixation efficiency, and technical potential, with important implications for climate change mitigation (Strand and Benford 2009). BECCS for electricity can proceed via post- combustion capture, pre-combustion capture, or oxycombustion processes (Rhodes and Keith 2005). BECCS for fuel or chemicals production can occur on biochemical or thermochemical conversion processes (R. H. Williams et al. 2011).

Using Illinois as a case study, we find optimal scale for several BECCS technologies, biomass availability scenarios, locations, capital cost scaling parameters, and transportation cost scenarios. We characterize the sensitivity of cost to scale. We also find the optimal scales of multiple facilities across Illinois. Our results highlight the importance of economies of scale for BECCS system design, with implications for entrepreneurs, power plant designers, policymakers, and power system planners. 2.2. Materials and methods

2.2.1. Problem statement

2.2.2. Multiple facility case

Our first formulation of the optimization problem minimizes the net present value (NPV) of the total cost ($) of operating a set of BECCS facilities over the entire project lifetime (Table 9 (a)). Costs include capital, variable operations and maintenance (O+M), fixed O+M, biomass purchase, fixed transportation, and variable transportation costs. Our problem is constrained by county-level biomass availability, and a constraint ensuring that sufficient biomass is delivered to meet expected electricity demand at the facility (Table 9 (b)). We also enforce a constraint on the minimum size of the sum of the portfolio of BECCS

69 facilities; this constraint can be interpreted as a total size goal, or portfolio standard, for a set of facilities throughout the study area.

This optimization problem chooses the size (S) of a set of potential facilities at several different locations, as well as biomass supplied from each county to each location (A), that minimizes total costs, subject to constraints. Decision variables include the size, S, of each facility, and amount of biomass supplied, A, from each county to each facility to satisfy demand. Capital costs are determined using a scaling parameter, α ≤ 1, which represents economies of scale. Biomass price, P, is a function of biomass supplied (A). Our sets include counties with available biomass, c, and a set of facility locations, l.

(a) Objective function: minimize total cost ($2014) Capital costs incurred for installing capacity at location l are ∝ calculated as the ratio of generator size in S to the base size, S , �! l 0 �! ∗ �!"# raised to a scaling exponent, α, and multiplied by the base capital �! Capital ! cost of generation, S0 * Ccap. Costs are annualized over the plant lifetime.

Fixed operation and maintenance costs paid for plant in location l + �! ∗ �!"#$% !!! are calculated as the total generation capacity of the plant, Fixed Fixed O&M ! multiplied by the recurring fixed costs Cfixed o+m. Generation

The variable costs paid for operating plant in location l are + � ∗ � ∗ � ! !"# !!! !" calculated from the power output in MWh, Sl * top, multiplied by ! the variable costs associated with the generator, Cvar o+m. Variable

The cost of biomass feedstock in each county c is calculated as the

product of the amount of biomass sourced from each county, Ac, and the county-level price for biomass, Pc. Price is represented as + �! ∗ �! � a one- or two-stage exponential function of biomass sourced in !

Feedstock each county (Eqs. 1 and 2). Sufficient biomass must be sourced to satisfy demand for input energy at each location.

Fixed biomass transportation costs are calculated as the product of

+ �!,! ∗ �!"#$% n the total amount of biomass transported from each county c to Biomass

Fixed Fixed !,! each location l, Ac,l, and the fixed transportation cost Tfixed. Transportatio Variable biomass transportation costs are calculated as the product of the amount of biomass transported from each county c to each + �!,! ∗ �!,! !"# location l, A , and the variable transportation cost T . Variable !,! c,l c,l var Variable Variable

Transport transportation costs are found using road transportation distances. (b) Constraints

�!,! ≤ �!,!"# Biomass availability. The amount of biomass !

delivered from each county c to each location l, Ac,l, must be less than or equal to the maximum amount of biomass available in each county, Ac,max.

�!,! ∗ 1 − �� ≥ Demand for biomass. The amount of biomass ! delivered from each county c to each location l, �!" ∗ ℎ�� ∗ �� Ac,l, must be greater than or equal to the amount � ∗ ! �� of biomass demanded at each location. Biomass demand is derived from the scaled size, Sl, time in operation, heat rate, and energy content of biomass.

Scaled size goal. The sum of scaled sizes at all �! ≥ �!"# locations, S , must be greater than or equal to a ! l specified scaled size goal, Smin.

Table 9. (a) Objective function and (b) constraints for total cost problem (multiple facility case). Objective for average cost problem can be found by dividing by the scaled size, Sl. Constraints for average cost problem are identical, except there is no scaled size goal.

2.2.3. Single facility case

Our second optimization problem minimizes the NPV of the average cost ($/W) of operating a BECCS facility at a given location over the entire project lifetime. For a fixed capacity factor, minimizing capacity and energy are equivalent. This “average cost” minimization problem determines the optimal scale of a single BECCS facility. This framework is similar to prior studies of optimal scale for bioenergy facilities that are not spatially explicit (Bomberg, Sanchez, and Lipman 2014; Jenkins 1997; Wright and Brown 2007). Constraints are identical to the multiple facility case, except no scaled size goal is enforced. We restrict our set of facility locations, l, to a single facility. Decision variables and parameters are identical between the two optimization problems.

2.2.4. Data sources and parameters

We leverage advances in public spatial databases to obtain model inputs, focusing on the State of Illinois. County-level biomass price and availability has been documented for several lignocellulosic feedstocks in the continental United States as part of the 2011 Billion Ton Study Update (BTS) (Downing et al. 2011). The BTS represents economically recoverable biomass resources from agricultural residues, wastes, forest residues, and dedicated energy crops. BTS data includes price and quantity of available biomass by feedstock type at different price points, which can form the basis of county-level biomass supply curves. We characterize biomass availability for 102 Illinois counties, disregarding

71 available biomass supply in neighboring states. Biomass in neighboring states might be reserved for other facilities or applications.

We develop three biomass availability scenarios: a base supply case, an expanded supply case, and a high price case. Our base supply case includes biomass from agricultural residues and waste streams, including corn stover, wheat straw, unused primary mill residues, and urban wood wastes. Total available supply for our base case is sufficient for 3112 MW in capacity (electricity output, throughout this paper) under assumed energy content, losses, and heat rates for an integrated gasification combined cycle facility with carbon capture and sequestration (IGCC-CCS). Maximum county-level availability is shown in Figure 34. Our base supply case does not account for biomass competition from lignocellulosic biorefineries, or other other uses of available biomass (Mullins et al. 2014).

Potential

Locations

Supply (100,000 tons) 0.0 - 0.5 0.50 - 1.0 1.00 - 2.0 2.00 - 3.5 3.50 - 6.0 Figure 34. County supply limit (105 bone dry ton biomass), and potential facility locations in Illinois for base case scenario. Our expanded supply case also includes woody biomass from forests, annual energy crops, perennial grasses, and logging residues. These feedstocks, especially energy crops used on primary agricultural land, may entail higher sustainability risks (Downing et al. 2011). Our high price case, used for sensitivity analysis, doubles the price of available biomass from the base supply case. This case may represent increased competition for biomass, or higher costs for biomass production, harvesting, and collection.

To ensure differentiability of our objective function, we fit county-level biomass availability using a single-term (Eq. 1) or two-term (Eq. 2) exponential model. We employ MATLAB Curve Fitting Toolbox, and chose the exponential form for each county that maximizes the coefficient of determination (r2). Biomass price is well-represented by this functional form; most counties have r2 ≥ .85. We assume that all biomass consumed in each county sells for a uniform price determined by the marginal cost of biomass. This is consistent with a horizontal integration structure, where plant operators purchase biomass from producers in each county.

!!! �!(�) = �!� (Eq. 1)

!!! !!! �!(�) = �!� + �!� (Eq. 2)

We choose candidate locations for BECCS facilities in identical locations to existing or planned CO2 injection locations in Illinois (Figure 34). Much of the state of Illinois is located on the Illinois Basin, a geological basin with large potential for long-term CO2 storage, which also underlies western Indiana and western Kentucky. For the (single site) average cost optimization problem and most of our sensitivity analyses, we use Decatur, IL. Decatur is the location of the Illinois Industrial Carbon Capture and Storage Project, an existing BECCS facility producing ethanol, and is close to the geographic center of the State (Hnottavange-Telleen, Krapac, and Vivalda 2009). This project is storing biogenic CO2, a byproduct of industrial fermentation, in the Mt. Simon Formation, a deep saline aquifer.

For the total cost optimization problem (which allows for more than one site) we use four additional facility locations: Meredosia, Taylorville, Loudon, and Tanquary. Meredosia, IL is the site of the proposed FutureGen Coal oxycombustion CCS facility located in Morgan County (Zitney et al. 2006). Taylorville was the location of a proposed IGCC-CCS facility in Christian County, with CO2 storage in the Mt. Simon Formation (“Taylorville Energy Center” 2014) . Loudon Field is located Fayette County, has been tested for enhanced oil recovery (EOR) potential by the Midwest Geological Sequestration Consortium (MGSC) (Robert Finley 2012). Tanquary Farms, located in Wabash County in Southwestern, IL, has been tested for enhanced coal bed methane recovery via CO2 injection by the MGSC. MGSC has also overseen injection for the Decatur Illinois Industrial Carbon Capture and Storage Project. Projects in these locations should be able to leverage geologic characterization and monitoring planned for these injections, while minimizing transportation distance for CO2.

We consider three scenarios for transportation costs: a base, low-cost, and high-cost scenario. Fixed and variable biomass transportation costs are drawn from existing literature (Marrison and Larson 1995; Searcy et al. 2007; Wright and Brown 2007). We employ the Google Maps API to find optimal transportation distances between the geographic center of each of Illinois’ 102 counties and our project facility locations, minimizing transport time. These distances are used to calculate variable transportation costs in our objective function. 73

We identify base costs, heat rates, and base sizes for several BECCS technology configurations, including pre-combustion capture (IGCC-CCS), oxycombustion, and post- combustion capture (Mike Matuszewski, Eric Lewis, and Steve Herron 2013; Mike Matuszewski, James Black, and Eric Lewis 2013; Rhodes and Keith 2005). Our base scenario, upon which sensitivity analysis is performed, assumes IGCC-CCS technology with a capital cost of 5,795 $/kW (2014$) and base size of 262.46 MW. These costs, along for those for oxycombustion and post-combustion capture, and derived from the National Energy Technology Laboratory (NETL). We also consider a low-cost IGCC scenario, with a capital cost of 2,089 $/kW and base size of 123 MW, based on estimates from Carnegie Mellon University. We assume that heat rate does not vary with scale.

We vary the capital cost scaling exponent parametrically across values found in existing literature. Chemical industry scaling exponents demonstrate a central tendency around .6, which we treat as our lower bound in study (Bomberg, Sanchez, and Lipman 2014). Such an estimate, however, may not hold for BECCS facilities. Gallagher et al., for example, reports a capital cost scaling factor of 0.86 for the U.S. dry mill ethanol industry, while others report values between .7-.9 (Leboreiro and Hilaly 2011; P. W. Gallagher, Brubaker, and Shapouri 2005). Scaling exponents for coal-fired power plants and nuclear plants in the United States have been found to be as high as .93-.94 (Fisher Jr, Paik, and Schriver 1986; Jenkins 1997). We set our base capital cost scaling exponent as .8, and vary this parameter between .6 and .95 for single site sensitivity analysis.

Several other parameters are fixed for the purposes of this study. We assume a plant life of 20 years, a discount rate of .1, and set 2014 as the base year for our study. All plants operate with a capacity factor of .85, a typical value for baseload plants. Carbon prices may need to rise to sufficient levels to justify baseload operation for BECCS plants. We assume a mean biomass energy content of 18 MMBtu/BDT (Farrell et al. 2006). We set biomass storage losses at 5%, based on existing literature (Sanderson, Egg, and Wiselogel 1997).

We calculate implied carbon abatement cost and levelized cost of electricity (LCOE) for our facilities. LCOE is calculated over the entire plant lifetime assuming no revenues from carbon prices. For our base abatement costs, we define the carbon abatement cost as the necessary carbon price ($/tCO2) to compensate for the difference between current average marginal prices in central Illinois, and plant LCOE (Eq. 3).3 We assume that BECCS displaces electricity from a modern coal steam turbine with emissions intensity of .86 tCO2/MWh. BECCS facilities are compensated both from electricity displacement (which we reason would come from an increased marginal electricity price) and geologic sequestration of CO2.

3 Data obtained from Midcontinent Independent Storage Operator (MISO), at: https://www.misoenergy.org/Library/MarketReports/Pages/MarketReports.aspx. For our implied cost of carbon calculations, we use the mean locational marginal price (LMP) between July and September, 2014, in Central Illinois, $29.74/MWh.

�� =

!"#$% !"#$!!"#$%&# !"#$%&"' !"#$% (Eq. 3) !"#$%& !"#$"%&#'!"#$%&'() !"#$%&'$'%( ! !"#$%& !"#$%&!"#$%&'$ !"#$%&'!"#$

We also consider several sensitivity cases for our implied carbon abatement costs. First, we calculate long-run abatement costs from displacement of electricity from coal IGCC-CCS facilities in Illinois. Costs and performance for these facilities are estimated by NETL (Mike Matuszewski, James Black, and Eric Lewis 2013). Second, we evaluate carbon abatement costs for displacement of combined cycle gas turbines (CCGTs), with emissions intensity of .36 tCO2/MWh, at current average marginal prices. Finally, we calculate abatement costs assuming non-zero lifecycle emissions from biomass. Here, we assign a lifecycle emissions penalty from biomass at 10% of its biogenic carbon content, as in (Sanchez et al. 2015a). 2.2.5. Solution method

This problem is non-convex. The objective function includes a number of terms that are linear in facility size but also a concave term (capital cost of capacity scales to the power α < 1) and a summation of convex terms (biomass feedstock costs increase exponentially with quantity). In general, it is not possible to ensure global optima from non-convex problems (Boyd and Vandenberghe 2009). To identify optima, we apply gradient-based methods for inequality-constrained, differentiable, non-linear programs from starting points across the feasible region (Byrd, Nocedal, and Waltz 2006). Starting points include a range of facility sizes, between 0 and 3100 MW, and county-level feedstock allocations sufficient to meet demand. By comparing several different local optima, we are able to approach global optimality. Novel algorithms for non-convex optimization are currently under research, and can be applied to this problem in the future (Vaz and Vicente 2009). We use the NEOS server, a free internet-based service for solving numerical optimization problems, to find solutions (Czyzyk, Mesnier, and Moré 1998; Dolan 2001; Gropp and Moré 1997). Our problem is written in AMPL, an algebraic modeling language for optimization (Fourer, Gay, and Kernighan 1993).

2.3. Results and discussion 2.3.1. Single facility case

Optimal scales are an order of magnitude larger than proposed scales found existing literature. We identify an optimal scale of 2713 MW (electricity) for the single IGCC-CCS facility case (Decatur, IL), for our base inputs. The facility produces carbon-negative electricity at a price of $142/MWh, with an implied carbon abatement cost of $59/tCO2. This scale is about ten times larger than the proposed base facility size reported by NETL (Mike Matuszewski, James Black, and Eric Lewis 2013). This suggests that BECCS facilities could be larger than previously assumed.

This BECCS facility consumes 87% of available biomass in the base case from all counties in the State. Highest levels of normalized consumption (the ratio of consumed biomass to 75 available biomass) occur near the facility location in Decatur (Figure 35 (a)). Normalized consumption has considerable range, between 40-100% of available biomass. Consumption tends to be lowest in counties farthest from the facility, where transportation costs are highest. We observe similar trends in county-level biomass market prices (Figure 35 (b)). The highest biomass prices occur immediately adjacent to the plant, to its Southeast. Biomass prices tend to be higher in counties near the facility. This is likely due to higher amounts of biomass consumption in those areas, and lower variable transportation costs from these counties, which lowers average costs. Biomass market prices range from $19- 87/BDT across the state. Lowest prices occur in the northeast of the state.

(a)

Decatur

Consumption (Normalized)

0.39 - 0.6 0.60 - 0.7 0.75 - 0.8 0.85 - 0.9 0.95 - 1.0

(b)

Decatur

Biomass Price ($/BDT)

18 - 40 40 - 55 55 - 65 65 - 75 75 - 90

Figure 35. (a) Normalized biomass consumption (values between 0 and 1), and (b) biomass market clearing price ($/BDT) for Illinois counties for base case in Decatur (scale = 2713 MW).

Deviations from optimal scaled size have little effect on overall systems costs. To test the sensitivity of costs to scale, we fix scale at several values and allow the average cost optimization to determine biomass allocation (Figure 36). While costs at optimal scale are roughly 24% lower than base costs, reducing LCOE from $177/MWh to $143/MWh, costs are relatively insensitive to scale at larger sizes. For example, a plant that is 60% smaller than optimal is only 5% more expensive than optimal. This suggests that non-techno- economic considerations – including investor availability, regulatory institutions, political processes, and public relations issues may ultimately have a strong impact on plant size. Power plant designers may instead wish to focus on practical limitations rather than optimal scale (Bomberg, Sanchez, and Lipman 2014).

60 110 Difference in Cost (%) 100 50 Cost, Base Size 90 Cost, Optimal Size 80 40 Implied Carbon Abatement Cost 70 60 30 50 20 $177/MWh

Difference in Cost (%) 10 $143/MWh

0 Implied Carbon Abatement Cost ($/tCO2) 0 500 1000 1500 2000 2500 3000 3500 Scaled Size (MW) Figure 36. Difference in cost (% difference from optimal objective) and implied cost of carbon ($/tCO2) for different fixed scale sizes in Decatur (base case). Optimal scale is denoted by a star, while base scale is denoted by a circle.

Biomass supply, scaling exponents, and technology costs are large drivers of optimal scale, while facility location and transportation costs are less important (Figure 37). Increasing biomass availability (expanded supply case) increases optimal scale to nearly 3500 MW (29% increase), while high biomass prices (high price case) lower scale to 2317 MW (15% decrease). Scale is similarly sensitive to scaling exponents and technology types. Scale exhibits a non-linear response to changes in scaling parameter; increasing the parameter to .95 from .8 decreases scale to 1150 MW (58% decrease), while decreasing the parameter to .6 leads to scale of 2900 MW (7% increase). All other facility types considered in this study—oxycombustion, post-combustion capture, and low-cost IGCC—have lower optimal scale than our base scenario. The low-cost IGCC facility, which is able to capture fewer economies of scale due to its lower cost, has the smallest scale, at 1245 MW. Varying transportation costs from the base scenario affect scale, between 2260-2848 MW, but have a smaller effect than other parameters. Finally, scale is relatively insensitive to the location of the facility.

Scaled Size: Decatur, IL (MW) 1000 1500 2000 2500 3000 3500

High Price Expanded Supply

Biomass Availability

Transport Costs High Costs Low Costs

Scaling Parameter .95 .6

Location Taylorville, IL

Low Cost IGCC Steam Turbine

Technology Type

Upside Downside Figure 37. Sensitivity of base case to biomass availability, transportation costs, capital cost scaling parameters, location, and technology type. Across sensitivity scenarios, LCOE ranges between $88-176/MWh, while implied carbon abatement cost ranges between $44-76/tCO2. Alternative base facility assumptions, including oxycombustion or low-cost IGCC, tend to have the lower carbon abatement costs than our base IGCC-CCS case. Scaling exponents also have a large affect of carbon abatement costs over the range of values tested in this analysis. Implied carbon abatement costs depend on assumptions about electricity displacement.

Long-run abatement costs from displacement of coal IGCC-CCS electricity (LCOE = $122.5/MWh, emissions intensity = .19 tCO2/MWh), is only $16/tCO2 at optimal scale. However, NETL employs different design considerations and operating cost estimates. For example, NETL identifies minimum biomass costs as $85/BDT, while we identify biomass prices between $19-87/BDT in our base case (Mike Matuszewski, James Black, and Eric Lewis 2013). In contrast, the implied cost of carbon for their base-case BECCS facility (262 MW, $257/MWh) is $108/tCO2, while ours is $43/tCO2.

Implied carbon abatement costs are also dependent on assumptions of carbon emissions intensity. Non-zero emissions from biomass, as a result of direct and indirect emissions from biomass cultivation, increase abatement costs from coal displacement from $59 to $65/tCO2 at optimal scale, assuming a 10% biogenic emissions penalty. Assuming that 79

BECCS displaces a CCGT instead of coal increases abatement costs to $80/tCO2. Future evolution of the electricity sector will have a large effect on abatement costs from BECCS.

Transportation logistics may present an additional barrier to scale. At our assumed heat rate, a BECCS facility at optimal scale would require nearly 40,000 dry tons/day of biomass. Assuming a truckload of 15 dry tons, this would require the delivery and unloading of nearly 110 trucks each hour. In contrast, NETL limits biomass delivery to 5,000 dry tons/day, citing traffic congestion and other barriers to large-scale biomass logistics (Mike Matuszewski, James Black, and Eric Lewis 2013). At this 5,000 ton limit, facility scale in our analysis would be limited to 354 MW. In contrast, advanced transportation logistics, including high-volume biomass railcars, would reduce railcar deliveries to 23 each hour (“Drax Unveils High-Volume Biomass Wagon” 2015). Pelletization and torrefaction, which we do not explicitly consider here, may improve transportation logistics for large-scale BECCS facilities.

2.3.2. Multiple facility case

When choosing between multiple facility locations, economies of scale lead to a centralized BECCS infrastructure. That is, across a range of scaled size goals and scaling parameters, the optimal solution is still a single facility. Here, the total cost minimization chooses to build facilities in Decatur equal to the scaled size goal. We did not observe solutions with more than one facility until we increased the capital cost scaling parameter, α, to greater than .96 (Figure 38). This result indicates that economies of scale for BECCS facilities have a larger effect than transportation or biomass costs in the context of promoting a decentralized BECCS infrastructure. Policymakers considering mandates for BECCS or other carbon dioxide removal (CDR) technologies should consider this result in policy design, especially as lower-cost CDR facilities benefit ratepayers. Similarly, entrepreneurs or power plant designers should consider the benefits of integration and scale when deploying BECCS.

Size (MW) 0.99 2500

0.98 1500 α 500 0.97

250

100 .96

Decatur Meredosia Taylorville Loudon Tanquary

Figure 38. Scaled size results among potential facility locations at different scaling exponents, α, for a scaled size goal of 2500 MW. A single facility is built in all cases where α ≤ .96.

One model enhancement, which we do not attempt here, is to allow for variable, rather than constant, cost scaling, as in Jenkins (Jenkins 1997). Under variable cost scaling, a

80 scaling parameter might increase asymptotically towards 1 with increasing capacity, to reflect increasing constraints or risks with scale. Jenkins, in applying this framework to conventional biopower facilities, finds that variable cost scaling decreases optimal scale 76%, from 1252 MW to 305 MW, under his reference cost and biomass availability assumptions. Given uncertainty about scaling, those implementing BECCS facilities may want to proceed with caution during industry scale up.

Our analysis may underestimate other benefits of decentralized BECCS infrastructure. These include decreased risk, ease of coordination, and impacts on the broader power system. Our results are consistent with other studies of optimal scale for bioenergy facilities, which tend to find optimal scales an order of magnitude larger than existing facilities (Bomberg, Sanchez, and Lipman 2014; Wright and Brown 2007).

2.4. Implications

This work has application to entrepreneurs, power plant designers, and power system planners interested in building BECCS facilities. In general, we find that economies of scale support centralized BECCS infrastructure, both when considering a single facility and a portfolio of facilities. This result derives from the fact that, at optimal scales, transportation cost and location have a small influence on costs. However if biomass supply is constrained due to competition with other uses, such as cellulosic biorefineries, optimal scales will decrease and levelized costs will increase.

Illinois, while an ideal case study for this exercise, contains many characteristics that will lead to larger optimal scales than in other locations. Corn stover is relatively plentiful in the state, and available at lower cost than dedicated feedstocks and some forest products that can be used for BECCS (Downing et al. 2011). Illinois roads have low tortuosity, leading to low transportation distances and variable transportation costs, an important diseconomy of scale. Illinois also has plentiful geologic storage capacity, lessening CO2 transportation costs. While we find that optimal scale is insensitive to location in Illinois, this may not be the case in areas with limited road infrastructure, or where biomass supply and geologic storage are not co-located.

Those wishing to implement policy to promote BECCS infrastructure, such as a portfolio standard, should consider large optimal scales in policy design. Mandates for distributed BECCS infrastructure, which may mimic mandates for distributed solar energy or feed-in- tariffs for small bioenergy facilities in the State of California, may sacrifice important economic efficiencies and ratepayer benefits (California Public Utilities Commission 2005; Rubio 2012). However, there are declining economic returns to increased scale (Figure 36). Additionally, we find relatively little variation in implied cost of carbon between different facility scales. This is due both to relatively small variation in LCOE, and the large carbon benefit of BECCS over other low-carbon generation options from negative emissions.

Though we find that biomass supply, technology cost and cost scaling do have a strong effect on optimal size, we also find that levelized cost and the implied carbon abatement 81 cost are relatively insensitive to deviations from the scaled size. This suggests that regulatory and political factors, transportation logistics, issues of public perception, and financing may ultimately drive what size facilities are built – rather than the techno- economic factors we consider here.

3. Conclusions

This Chapter explores system scale for BECCS, a practical design issue, using spatially explicit infrastructure optimization for the State of Illinois. We find that there are large incentives for scale. Optimal scales for BECCS are an order of magnitude larger than proposed scales found in existing literature. When choosing between multiple facility locations, economies of scale support a centralized BECCS infrastructure. Deviations from optimal scaled size have little effect on overall systems costs – suggesting that other factors, including regulatory, political, or logistical considerations, may ultimately have a greater influence on plant size than the techno-economic factors we consider.

This result indicates that economies of scale for BECCS facilities have a larger effect than transportation or biomass costs in the context of promoting a decentralized BECCS infrastructure. Policymakers considering mandates for BECCS or other carbon dioxide removal (CDR) technologies should consider this result in policy design, especially as lower-cost CDR facilities benefit ratepayers. Similarly, entrepreneurs or power plant designers should consider the benefits of integration and scale when deploying BECCS. Finally, this has implications for sustainability of BECCS, as scale affects the magnitude and location of lifecycle impacts.

Chapter V. Commercialization of carbon-negative energy systems

1. Preface

This Chapter considers commercialization pathways for BECCS. We also consider sustainability risks at scale. Commercialization is dependent on the coordination of a wide range of actors, many with different incentives and worldviews (K. S. Gallagher et al. 2012). Nevertheless, we draw on existing literature to propose a commercialization roadmap for BECCS. Our analysis is complicated by the relative lack of deployments at scale.

Numerous kinds of research inform commercialization of risky, capital-intensive technology. These include market research, assessment of technological readiness, techno- economic analysis, assessment of deployment, and analysis of existing or future supportive governmental policy. Market research and assessments of deployment are clear about the potential role for BECCS. Integrated Assesment Models indicate BECCS is very valuable to long-term climate change mitigation (Rose et al. 2013). Other analyses of deployment and design, such as those carried out in Chapter II and Chapter IV, indicate that near-term deployments could also be cost-effective given supportive climate policy (Sanchez et al. 2015a; Liu et al. 2011).

Our analysis focuses on the role of flexibility, enabled by thermochemical co-conversion of biomass and fossil fuels, in commercialization. In addition, we catalog R&D, policy, and funding needs for BECCS. Strategies for commercial deployment can inform opportunities for governments, industry incumbents, and emerging players to research and support BECCS technologies.

2. Excerpt from A commercialization strategy for carbon-negative energy (Sanchez and Kammen, In Press)4

2.1. Introduction

The Intergovernmental Panel on Climate Change envisions the need for large-scale deployment of net-negative CO2 emissions technologies by mid-century to meet stringent climate mitigation goals and yield a net drawdown of atmospheric carbon. These carbon dioxide removal technologies complement low- or zero-carbon energy technologies (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015; Sanchez et al. 2015a). Industrial-scale sequestration of carbon dioxide (CO2) from bioenergy

4 First published in Nature Energy 1, 2016 by Nature Publishing Group (NPG). NPG does not require authors of original (primary) research papers to assign copyright of their published contributions. Authors grant NPG an exclusive license to publish, in return for which they can reuse their papers in their future printed work without first requiring permission from the publisher of the journal. For more information, see http://www.nature.com/reprints/permission-requests.html

83 production, a process known as bioenergy with carbon capture and sequestration (BECCS), can produce fuels, chemicals, and electricity while removing atmospheric CO2. Yet there are few commercial deployments of BECCS outside of niche markets, creating uncertainty about commercialization pathways and sustainability impacts at scale (Fuss et al. 2014). This uncertainty is exacerbated by the absence of a strong policy framework, such as high carbon prices and research coordination. Here, we propose a strategy for the potential commercial deployment of BECCS via thermochemical co-conversion of biomass and fossil fuels, particularly coal, challenging governments, industry incumbents, and emerging players to research and support these technologies.

While biochemical conversion is a proven first market for BECCS, this trajectory alone is unlikely to drive commercialization of BECCS at the gigawatt scale. The early development of BECCS has been focused on biochemical facilities converting sugars to ethanol, a transportation fuel, for use in enhanced oil recovery and large-scale industrial sequestration demonstration (Finley 2014). Yet, biochemical conversion pathways are limited by the market size of fuel products, scale, sensitivity to biomass inputs, and throughput. For example, alcohol fuels from biochemical conversion processes face compatibility issues with existing transportation infrastructure, while high lignin content inhibits biochemical conversion. Reaching gigaton-scale carbon sequestration via this pathway will require at least four-fold higher ethanol production than current levels. While proposals for large-scale bioenergy deployment focus on conversion of lignocellulosic feedstocks to liquid fuels, BECCS is also valuable to the electricity sector (Sanchez et al. 2015a; Carroll and Somerville 2009). 2.2. Flexibility as a virtue

In contrast to biochemical conversion, thermochemical conversion of coal and biomass enables large-scale production of fuels and electricity with a wide range of carbon intensities, process efficiencies and process scales (Figure 39). We focus on two representative thermochemical pathways: electricity production via integrated gasification combined cycle with CCS (IGCC-CCS), and long-chain hydrocarbon fuels production via gasification and the Fischer-Tropsch process with CCS (FT-CCS). FT and combined cycle systems can be combined for polygeneration-CCS systems that produce both electricity and fuels. The energy and capital penalties of adding CCS are comparatively small for these processes, and, for FT, can reduce downstream equipment size requirements, further reducing capital costs (Liu et al. 2011). Addition of biomass into coal gasification increases the H2/CO ratio of syngas, which is beneficial for fuels production.

(a) Sulfur Coal 3 CO SulfurSulfur Air 2 Steam 2 Recycle RRemoval Electricity Integrated Gasification Air O2 Water-Gas Acid Gas Separation Gasification Combined Shift Removal Cycle Unit (WGS) (AGR) (I GCC)

H2 O CO + H 2 O->CO2 +H2

N2 Biomass 1 Bypass CO2 (for compression and sequestration)

(b) 2 Sulfur FuelsFuels CO2 Coal Water-Gas Recycle (Gasoline,asoline, DieDiesel)s Gasfication Shift SulfurSulfur (WGS) Removal

CO + H 2 O->CO2 +H2 Fischer- Fischer- Air Acid Gas Tropsch Tropsch Separation Bypass Removal Combined Synthesis Refining Unit (AGR) Cycle Power ElectricityElect O2 Recycle Island Steamteam Biomass Tar Cra cki ng 3 GasificationG and Filtering CO2 (forfor compressioncompres and ssequestration)equestratio Autothermal Water-Gas COO2 1 Reformer Shift Removal

Figure 39. Flow diagrams for carbon capture and storage processes. Simplified flow diagram for (a) IGCC-CCS and (b) polygeneration-CCS processes for production of electricity and fuels from coal and biomass, based on (Mike Matuszewski, James Black, and Eric Lewis 2013) and (Liu et al. 2011). Green boxes and dashed green lines indicate options to decrease the carbon intensity of resulting fuel or electricity products. Dashed lines indicate optional process enhancements. Options to decrease the carbon intensity of products in IGCC-CCS systems include 1) increasing the ratio of biomass to coal inputs, 2) increasing shift of syngas in the water-gas shift reactor, and 3) recycling CO2 from the sulfur removal process to the acid gas removal system. For polygeneration-CCS, options include 1) increasing the ratio of biomass to coal inputs, 2) recycling CO2 from the sulfur removal process to the acid gas removal system, and 3) autothermal reforming and shift prior to electricity production.

Other promising thermochemical pathways may complement IGCC-CCS and FT-CCS. For example, torrefaction, a mild thermal pretreatment, improves biomass suitability for gasification (Boerrigter and Drift 2004). Integration of reformed natural gas with syngas from coal and biomass presents further opportunities for electricity or fuels production, though this has been studied less extensively (Floudas, Elia, and Baliban 2012). Likewise, synthetic gasoline production via the methanol-to-gasoline process is a less widely implemented alternative to FT synthesis (Liu et al. 2015). Alternatively, all of the carbon in biomass feedstock could be converted to liquid fuels using external hydrogen and low- carbon electricity inputs from nuclear, renewable energy, or hydropower (Agrawal et al. 2007).

Thermochemical co-conversion creates flexibility for producers to balance product cost and carbon reduction goals. Incremental carbon reduction from these systems is cheaper

85 when leveraging both CCS and biomass in concert (Liu et al. 2011; Mike Matuszewski, James Black, and Eric Lewis 2013). IGCC-CCS producers can adjust biomass, water-gas shift, and CCS integration to produce low-carbon or carbon-negative electricity, with lower carbon-intensity systems having smaller scale and higher costs in the absence of supportive climate policy. Similarly, polygeneration producers can adjust biomass, CCS, or autothermal reformer integration (Figure 39). Polygeneration systems with flexible ratios of fuel inputs or product outputs can increase profitability for producers, though at additional cost (Floudas, Elia, and Baliban 2012). In addition to technical advantages, the flexibility of co-conversion also holds advantage for existing industries, supply chains, and workforces. Here, firms can embrace a gradual transition pathway to deep decarbonization, limiting economic dislocation and increasing transfer of knowledge between the fossil and renewable sectors.

The scale of both biomass and co-utilization systems holds unique advantages. In the United States, dedicated biomass plants average half the efficiency of coal plants (Milne and Field 2014). Co-utilization systems can leverage economies of scale associated with coal inputs to increase efficiency and decrease unit costs, while lessening feedstock variability issues associated with biomass-only systems. Co-utilization also requires less biomass per unit product than biomass-only systems, which further extends the impact of scarce sustainable biomass resources. Minimum capital expenditures for co-conversion systems are smaller than those for envisioned coal-to-liquid facilities, easing project finance (Liu et al. 2011). In practice, many deployed biomass systems are several-fold smaller than coal- only systems, enabling greater experimentation at lower cost (Trancik 2006).

This strategy is likely to work best in developed economies, such as the United States. The U.S., for example, has relatively advanced bioenergy, hydrocarbon production, and CCS sectors. Its advanced engineering, construction, and financial industries are capable of commercializing new, risky, capital-intensive technologies. Other regions, such as Scandinavia and British Columbia, have considerable expertise in thermochemical conversion of biomass. Following technology development, other developing countries, particularly rapidly expanding economies dependent on coal, could deploy BECCS and avoid lock-in with conventional fossil fuel infrastructure. International development finance could help promote BECCS in these locations. In contrast, certain regions with large biomass resources and advanced alcohol fuel industries, such as Brazil, may rely on biochemical conversion for BECCS in the future. Other developing countries could embrace biochar production for carbon dioxide removal (Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration 2015). Spatial optimization can account for biomass supply limitations, transportation logistics, and geologic CO2 storage capacity, and help balance economies- and diseconomies-of-scale inherent in bioenergy production, including gasification (Sanchez et al. 2015a; Floudas, Elia, and Baliban 2012).

2.3. Research and policy needs

While most system components for fossil CCS systems are technologically mature, there are very few commercial deployments at scale. We expect near-term deployment of co- conversion systems to demonstrate technical feasibility and reduce investment risks. For 86 example, the Buggenum IGCC project in the Netherlands has studied both biomass and CCS integration, while Total’s BioTFuel project is focused on commercial deployment of torrefaction, entrained-flow biomass co-gasification, and FT synthesis (Liu et al. 2015; Damen et al. 2011). However, IGCC-CCS systems under construction in the United States have faced cost overruns, construction delays, and regulatory uncertainty. Alternative electricity system CCS configurations, including post-combustion capture or oxycombustion systems, may face lower commercial hurdles while still maintaining the flexibility of IGCC-CCS.

Aside from systems integration, primarily technical barriers are in large-scale biomass logistics, gasification, and gas cleaning. Ultimately, commercial-scale BECCS will require high-temperature, oxygen- or steam-blown, pressurized, entrained flow gasification of multiple biomass streams (Boerrigter and Drift 2004). While fluidized-bed gasifiers are most commonly used for biomass gasification, coal-based entrained-flow gasifiers are more commercially advanced and exist at larger scale (Liu et al. 2015). In addition to the logistical challenges in feedstock management of both coal and biomass, multiple revenue streams from polygeneration systems complicate business models for producers. Business models may be informed by existing multi-output bioenergy systems, like sugarcane biorefineries or pulp and paper mills, which produce electricity and heat in addition to their primary product.

In contrast, large-scale CO2 storage faces deployment barriers. Delivery of identified, accessible, and permitted CO2 storage requires upfront investment of monetary and human resources (Scott et al. 2015). This is a particular challenge in regions without a developed hydrocarbon exploration and production industry, which lack regulation, finance, knowledge, public trust, and skills for geological sequestration. Natural analogues demonstrate the security of long-term CO2 storage, but social acceptance remains a barrier in many regions (Sathaye et al. 2014).

Existing policies and R&D programs largely ignore the synergies enabled by co-conversion. At the U.S Department of Energy, the Office of Fossil Energy supports CCS demonstration and clean coal efforts, but does not focus on standalone biomass gasification. Similarly, the Biomass Energy Technologies Office focuses on feedstock logistics and biomass conversion, but does not support upstream fossil energy integration or CCS. Given their distinct expertise, future Department of Energy deployment programs should take the form of a crosscutting initiative. U.S. biofuel policy, such as the Renewable Fuels Standard or California’s Low-Carbon Fuel Standard, does not recognize co-conversion or the benefits of CCS integration. In the European Union, current EU-ETS policy does not recognize the potential for carbon dioxide removal from BECCS (Lomax et al. 2015). This can be remedied by revising accounting principles to include biogenic carbon storage. In addition to carbon pricing, policymakers can de-risk BECCS technologies by providing credit subsidies to first-of-a-kind plants, or tax credits for geologic CO2 sequestration. Like conventional CCS technologies, BECCS deployment may require a stable investment climate, such as price guarantees, in absence of a sufficiently high carbon price (Scott et al. 2013). Public private partnerships may be an attractive vehicle to reduce risks and increase financing. Performance standards, rather than quantity mandates, can better recognize the 87 performance of multiple process configurations to produce low-carbon or carbon-negative electricity or fuels.

We estimate cumulative capital investment needs for BECCS through 2050 to be over $1.9 trillion (2015$, 4% real interest rate) under stringent climate change mitigation scenarios, based on existing cost estimates and deployment values from Intergovernmental Panel on Climate Change Representative Concentration Pathway 2.6, which is likely to limit global warming to 2 °C (Sanchez et al. 2015a; Kato and Yamagata 2014). This stabilization scenario envisions deployment of as much as 24 GW/yr of BECCS by 2040, if installed as IGCC-CCS, at costs similar to current global cleantech investment rates. Regional allocation of carbon removal, a subset of all emissions reduction, is not yet established. To achieve this rate of deployment within 15-20 years, governments and firms must commit to RD&D on an unprecedented scale.

2.4. Uncertainty clouds deployment at scale

Key uncertainties around large-scale BECCS deployment are not limited to commercialization pathways; rather, they include physical constraints on biomass cultivation or CO2 storage, as well as social barriers, including public acceptance of new technologies and conceptions of renewable and fossil energy, which co-conversion systems confound (Fuss et al. 2014). Lifecycle greenhouse gas (GHG) impacts of coal and biomass systems at scale are uncertain, due in large part to variation in estimates of direct and indirect land use change (LUC) resulting from biomass cultivation (Figure 40) (Argonne National Laboratory 2014). Nevertheless, we find that switchgrass (a dedicated feedstock) integration decreases lifecycle GHG impacts of IGCC-CCS systems, across a wide range of LUC scenarios. Polygeneration systems may not be carbon-negative due to tailpipe CO2 emissions (Liu et al. 2011). Assumptions about counterfactuals and timing of emissions further complicate analysis of lifecycle impacts.

200 Base LUC e/MWh) 2 0 Best-case Worst-case −200 Risks at Scale: −400 • Food security • Land conservation −600 • Biodiversity • Social equity −800 • Water resources

−1000 Lifecycle GHG Emissions (kgCO 0 10 20 30 40 50 60 70 80 90 100 Biomass (% wt) Figure 40. Lifecycle greenhouse gas carbon intensity for IGCC-CCS facilities. Base (black line), best-case (green dashed line), and worst-case (red dashed line) lifecycle GHG performance intensity for coal and biomass IGCC-CCS facilities. Shaded region represents the range of carbon intensity. Biomass is assumed to be switchgrass (Panicum virgatum) grown in the United States. 88

The IGCC-CCS facility is assumed to capture 90% of gross emissions (Mike Matuszewski, James Black, and Eric Lewis 2013). Assessment accounts for emissions in feedstock production, as well as modeled direct and indirect land use change (LUC) scenarios evaluated in GREET 2014 (Argonne National Laboratory 2014), and assumes biomass regrowth. LUC scenarios contain differing assumptions about soil carbon sequestration (best-case has deeper soil depth), crop yields (best-cast includes 1% yield increase each year), carbon in harvested wood products (best- case includes credits for carbon sequestered in products), and emissions from land conversion (worst-case includes more forest conversion for cropland). Performance intensity scenarios were modeled in GREET’s Carbon Calculator for Land Use Change from Biofuels Production over a thirty year production period. Flexible generation systems can alter their levels of coal and biomass to achieve different lifecycle impacts.

Other risks in large-scale biomass cultivation include adverse effects on food security, land conservation, social equity, and biodiversity, as well as competition for water resources. Fuel and electricity from biomass may also face competition from heating and manufacturing demands. While modeling can help reduce uncertainty, actual measurement will be necessary to generate knowledge around the impacts of BECCS deployment (Youngs and Somerville 2014). An emerging biomass industry can embrace sustainability standards to help ensure environmental and social benefits. We envision an iterative approach to sustainability, with standards being updated as more knowledge is gained through deployment. The Intergovernmental Panel on Climate Change and others have embraced iterative risk management as a framework for managing uncertainty and new information. 2.5. An agenda for transition

Despite sustainability risks, this commercialization strategy presents a pathway where energy suppliers, manufacturers and governments could transition from laggards to leaders in climate change mitigation efforts. Using the flexibility of thermochemical co- conversion, these entities could meet increasingly stringent climate policy, ultimately deploying commercial-scale BECCS facilities and transitioning away from fossil fuels. This transition strategy holds advantages for existing industries, supply chains, and workers, while sharing knowledge between the fossil and renewable energy sectors. Future plants would be able to avoid technology lock-in or unfavorable economics after policy changes by installing systems that can embrace biomass and CCS. All stakeholders in the energy technology innovation system can work in concert to support this end, employing stable policy support, targeted and crosscutting research and development, iterative risk management, and commercial deployment.

3. Conclusions

This Chapter proposes a commercialization pathway for BECCS. Specifically, we find that the flexibility of thermochemical conversion of biomass and fossil energy to electricity and fuels, when coupled with carbon capture and storage (CCS), provides a viable strategy for commercializing carbon-negative energy. Three primary results emerge:

1. Flexibility and commercialization: The flexibility of thermochemical conversion enables a viable transition pathway for firms, utilities and governments to achieve net-negative CO2 emissions in production of electricity and fuels given increasingly stringent climate policy. Primary research, development, and deployment (RD&D) needs are in large-scale biomass logistics, gasification, gas cleaning, and geological CO2 storage. R&D programs, subsidies, and policy that recognize co-conversion processes can support this pathway to commercialization. Here, firms can embrace a gradual transition pathway to deep decarbonization, limiting economic dislocation and increasing transfer of knowledge between the fossil and renewable sectors.

2. The scale of the transition: We estimate cumulative capital investment needs for BECCS through 2050 to be over $1.9 trillion (2015$, 4% real interest rate) under stringent climate change mitigation scenarios. This stabilization scenario envisions deployment of as much as 24 GW/yr of BECCS by 2040, if installed as IGCC-CCS. This work challenges governments, industry incumbents, and emerging players to research and support BECCS and co-conversion technologies. All stakeholders in the energy technology innovation system can work in concert to support this end, employing stable policy support, targeted and crosscutting research and development, iterative risk management, and commercial deployment.

3. Sustainability limitations: Risks at commercial scale include adverse effects on food security, land conservation, social equity, and biodiversity, as well as competition for water resources. Scale, for which we find strong incentive, also influences sustainability risks. We find that switchgrass integration decreases lifecycle GHG impacts of IGCC-CCS systems, across a wide range of land-use change (LUC) scenarios. While modeling can help reduce uncertainty, actual measurement will be necessary to generate knowledge around the impacts of BECCS deployment.

This work informs R&D, investment, policy, and coordination to support BECCS commercialization. As such, it is the culmination of this dissertation’s efforts to inform responsible, incremental implementation of BECCS, towards a global contribution to mitigation efforts.

Chapter VI. Conclusions and implications for policy

This dissertation attempts to fill critical gaps on the continuum of research tasks to inform large-scale implementation of bioenergy with carbon capture and sequestration (BECCS), centering on deployment, design, commercialization, and communication (Figure 13). It focuses on actionable information to inform key stakeholders responsible for deployment: utilities, firms, project developers, corporate entities, and governments. It builds on existing literature on BECCS, which largely centers on techno-economic analysis and global energy-economy modeling.

In particular, this dissertation is motivated by the question “what modeling and analysis is necessary to inform responsible, incremental implementation of BECCS, towards a global contribution to mitigation efforts?” The research presented here is applied, using science for a specific, outcome-driven, purpose. Thus, the analysis informs technology diffusion, and is bottom-up, rather than top-down. The following themes and results emerge:

1. Net-negative emissions energy systems

BECCS, combined with aggressive renewable deployment and fossil emission reductions, can enable a carbon-negative power system in Western North America by 2050, with up to 145% emissions reduction from 1990 levels. BECCS complements other sources of renewable energy, and can be deployed in a manner consistent with regional policies and design considerations. This result is found by using a power systems planning model with rich technology representation to examine the role of BECCS in power systems in detail, and in aggressive timeframes. Put another way, this result is robust to realistic physical constraints, represented by the SWITCH model, which IAMs often lack. The amount of biomass resource available limits the level of fossil CO2 emissions that can still satisfy carbon emissions caps.

Studies of regional deployment inform near-term efforts to build facilities, integrate them into existing energy and power systems, respect regional design considerations, and complement existing energy and climate policy. They also inform decarbonization pathways for other regions considering integration of negative emissions technologies.

2. Value of BECCS in climate change mitigation

Offsets produced by BECCS are more valuable to the power system than the electricity it provides. This result, found through a high-resolution power systems planning model for the Western Electricity Coordinating Council, confirms results from IAMs: BECCS is very valuable in the context of deep decarbonization, and can offset CO2 emissions from fossil energy across the economy. The availability of BECCS and a high supply of lignocellulosic biomass for bioenergy are projected to be critical requirements for achieving stringent mitigation targets and containing the associated costs (Bauer 2015).

Implied costs of carbon for BECCS are relatively low for a capital-intensive technology. In the WECC, we observe BECCS technology deployment at abatement costs as low as $74/tCO2 (86% emissions reduction by 2050), with more stringent emission caps incurring higher abatement costs. In Illinois, implied carbon abatement cost ranges between $44- 76/tCO2 at optimal scale when displacing coal steam turbines. Implied carbon abatement costs are dependent on assumptions degree of mitigation, displaced energy sources, and carbon emissions intensity. These costs, which are relatively low in the context of deep carbonization, are due to the fact that BECCS can displace carbon-intensive electricity while sequestering biogenic CO2 (Keith and Rhodes 2002).

3. Incentives for scale

We find large incentives for scale. Optimal scales for BECCS are an order of magnitude larger than proposed scales found in existing literature. When choosing between multiple facility locations, economies of scale support a centralized BECCS infrastructure. Deviations from optimal scaled size have little effect on overall systems costs – suggesting that other factors, including regulatory, political, or logistical considerations, may ultimately have a greater influence on plant size than the techno-economic factors we consider.

4. Flexibility and commercialization

Climate change mitigation requires gigawatt-scale carbon dioxide removal technologies, yet few examples exist beyond niche markets. The flexibility of thermochemical conversion enables a viable transition pathway for firms, utilities and governments to achieve net- negative CO2 emissions in production of electricity and fuels given increasingly stringent climate policy. Primary research, development and deployment (RD&D) needs are in large- scale biomass logistics, gasification, gas cleaning, and geological CO2 storage. R&D programs, subsidies, and policy that recognize co-conversion processes can support this pathway to commercialization. Here, firms can embrace a gradual transition pathway to deep decarbonization, limiting economic dislocation and increasing transfer of knowledge between the fossil and renewable sectors.

5. Communicating BECCS

We develop a journal article introducing BECCS to a middle-school audience, involving students in the review process. We find that simple exposition can describe the following topics to a middle-school audience: climate change, mitigation of climate change, carbon dioxide removal, BECCS, and electricity systems. During development, the authors strove to explain these concepts in as simple language as possible, and to develop figures that can be used by the entire community to introduce BECCS. The end result is a freely available scientific article written by scientists and shaped for younger audiences by the input of their own peers.

Transparent, simple descriptions of BECCS can build awareness and credibility in the wider public and among central stakeholders. First, it builds a ‘community of support’ for BECCS. Existing research has identified a key problem for BECCS as cultural, lacking in a community of support, awareness and credibility amongst its own key stakeholders and the wider public (Dowd, Rodriguez, and Jeanneret 2015). These communities of action are essential for achieving just and effective environmental outcomes, or advocating to affect change (Shallcross and Robinson 2008; Nowlin 2011).

Second, straightforward description can increase public familiarity with, and decrease hostility towards, BECCS. CCS and bioenergy technologies are controversial (Wong-Parodi et al. 2011; Halder et al. 2010). Trust is an important concept when discussing the transformation of energy systems, particularly for novel technologies like CCS (Rayner 2010). As such, simple, accessible, and balanced introductions to BECCS can build awareness and credibility in the wider public.

6. Sustainability limitations

Sustainability impacts, and sustainability limitations, are a pervasive theme of this dissertation. In Chapter II and Chapter IV, we prioritize waste and residues for lignocellulosic biomass feedstock, rather that dedicated energy crops or food-based supply. Nevertheless, substantial uncertainty remains about impacts at scale.

One form of uncertainty around sustainability impacts of BECCS is the lifecycle GHG emissions of energy products. Three primary issues complicate emissions accounting for bioenergy with CCS: cross-sector CO2 accounting, regrowth, and timing. As mentioned previously, our assessments prioritize production from wastes and residues, which would otherwise be burned or left to decompose, releasing their carbon into the atmosphere as CO2. The emissions benefits of BECCS — encompassing displaced fossil-fuel CO2 emissions from energy production and geological CO2 sequestration — may improve the desirability of biomass production for bioenergy over other land-use decisions, but more research is needed to directly compare it with other sequestration strategies. Nevertheless, we find that switchgrass integration decreases lifecycle GHG impacts of IGCC- CCS systems, across a wide range of land-use change (LUC) scenarios (Figure 40).

Risks at commercial scale include adverse effects on food security, land conservation, social equity, and biodiversity, as well as competition for water resources. Scale, for which we

93 find strong incentive, also influences sustainability risks. While modeling can help reduce uncertainty, actual measurement will be necessary to generate knowledge around the impacts of BECCS deployment. This dissertation argues for an iterative risk management approach to BECCS sustainability, with standards being updated as more knowledge is gained through deployment.

7. Scale of the transition

We study transitions on two scales: global, and regional. In the WECC, our analysis suggests that installation of up to 10 GW of BECCS capacity between 2030 and 2040, with additional capacity additions thereafter, could be a key part of meeting stringent climate goals. Despite this scale, we find necessary capacity deployment rates for BECCS to be smaller than those for other intermittent renewables or gas. Globally, we estimate cumulative capital investment needs for BECCS through 2050 to be over $1.9 trillion (2015$, 4% real interest rate) under stringent climate change mitigation scenarios. This stabilization scenario envisions deployment of as much as 24 GW/yr of BECCS by 2040, if installed as IGCC-CCS.

To achieve theses rates of deployment within 15-20 years, governments and firms must commit to research, development, and deployment (RD&D) on an unprecedented scale. All stakeholders in the energy technology innovation system can work in concert to support this end, employing stable policy support, targeted and crosscutting research and development, iterative risk management, and commercial deployment. Coordination of funding and research agendas is critical, both intra- and inter- nationally. For deployment, this a goal would require a concentrated effort in finance, site selection, biomass sourcing, geological characterization, permitting, site-specific environmental impact assessments and community consultation. This dissertation challenges governments, industry incumbents, and emerging players to research and support BECCS and co-conversion technologies.

8. Conclusion

Taken in sum, this dissertation makes several advances to inform responsible, incremental implementation of BECCS. It builds on existing literature about BECCS, which largely centers on techno-economic analysis and global energy-economy modeling.

Its findings on deployment, design, and commercialization present a strong conclusion to utilities, firms, project developers, corporate entities, and governments considering BECCS. Specifically, we find that BECCS is valuable in the context of climate change mitigation, complements renewable energy, enables additional decarbonization pathways, has large incentives for scale, and has a viable commercialization strategy leveraging co-conversion with fossil fuels. Sustainability impacts and public opposition to BECCS, two oft-cited concerns, may be reduced with transparent measurement and communication.

However, commercial-scale deployment is dependent on the coordination of a wide range of actors, many with different incentives and worldviews. Sustained coordination and effort will be necessary to prevent BECCS from being merely an academic exercise.

Chapter VII. References

Abbasi, Tasneem, and S.A. Abbasi. 2010. “Biomass Energy and the Environmental Impacts Associated with Its Production and Utilization.” Renewable and Sustainable Energy Reviews 14 (3): 919–37. doi:10.1016/j.rser.2009.11.006. “About White Rose | White Rose CCS.” 2015. Accessed October 6. http://www.whiteroseccs.co.uk/about-white-rose. Agrawal, Rakesh, Navneet R. Singh, Fabio H. Ribeiro, and W. Nicholas Delgass. 2007. “Sustainable Fuel for the Transportation Sector.” Proceedings of the National Academy of Sciences 104 (12): 4828–33. Arasto, Antti. 2015. “Special Project: Bio-CCS and Bio-CCUS in Climate Change Mitigation and Extended Use of Biomass Raw Material.” Proposal for 2016-2017 ExCo76. Berlin, Germany: International Energy Agency. Argonne National Laboratory. 2014. Argonne GREET Model: The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (version 1.2.0.11346). Argonne, IL: Argonne National Laboratory. http://greet.es.anl.gov/. Bauer, Nico. 2015. “Power Systems: Carbon Negative at the Regional Level.” Nature Climate Change 5 (3): 196–97. doi:10.1038/nclimate2548. Benson, Sally M. 2014. “Negative-Emissions Insurance.” Science 344 (6191): 1431–1431. doi:10.1126/science.1257423. Benson, Sally M., Kamel Bennaceur, Peter Cook, John Davison, Heleen de Coninck, Karim Farhat, Andrea Ramirez, et al. 2012. “Chapter 13 - Carbon Capture and Storage.” In Global Energy Assessment - Toward a Sustainable Future, 993–1068. Cambridge University Press, Cambridge, UK and New York, NY, USA and the International Institute for Applied Systems Analysis, Laxenburg, Austria. www.globalenergyassessment.org. Benson, Sally M., and David R. Cole. 2008. “CO2 Sequestration in Deep Sedimentary Formations.” Elements 4 (5): 325–31. Biorecro, A. B. 2010. “Global Status of BECCS Projects 2010.” Canberra, Australia: Global CCS Institute. Boerrigter, H., and A. van der Drift. 2004. “Biosyngas: Description of R&D Trajectory Necessary to Reach Large-Scale Implementation of Renewable Syngas from Biomass.” ECN-C--04-112. Energy research Centre of the Netherlands (ECN). https://www.ecn.nl/publications/E/2004/ECN-C--04-112. Bomberg, Matthew, Daniel L. Sanchez, and Timothy E. Lipman. 2014. “Optimizing Fermentation Process Miscanthus-to-Ethanol Biorefinery Scale under Uncertain Conditions.” Environmental Research Letters 9 (6): 064018. doi:10.1088/1748- 9326/9/6/064018. Boyd, Stephen, and Lieven Vandenberghe. 2009. Convex Optimization. Cambridge university press. http://books.google.com/books?hl=en&lr=&id=IUZdAAAAQBAJ&oi=fnd&pg=PR11 &dq=boyd+convex+optimization&ots=HMEBif1MJl&sig=wR_-06Kd4kzfS8EkB- EaVmJLJa8.

Byrd, Richard H., Jorge Nocedal, and Richard A. Waltz. 2006. “KNITRO: An Integrated Package for Nonlinear Optimization.” In Large-Scale Nonlinear Optimization, 35–59. Springer. http://link.springer.com/chapter/10.1007/0-387-30065-1_4. California Public Utilities Commission. 2005. California Solar Initiative. http://www.cpuc.ca.gov/puc/energy/solar/aboutsolar.htm. Campbell, J. E., D. B. Lobell, and C. B. Field. 2009. “Greater Transportation Energy and GHG Offsets from Bioelectricity Than Ethanol.” Science 324 (5930): 1055–57. doi:10.1126/science.1168885. Canadell, Josep G., and E. Detlef Schulze. 2014. “Global Potential of Biospheric Carbon Management for Climate Mitigation.” Nature Communications 5 (November): 5282. doi:10.1038/ncomms6282. Carroll, Andrew, and Chris Somerville. 2009. “Cellulosic Biofuels.” Annual Review of Plant Biology 60 (1): 165–82. doi:10.1146/annurev.arplant.043008.092125. Chen, Yang, Thomas A. Adams, and Paul I. Barton. 2011. “Optimal Design and Operation of Flexible Energy Polygeneration Systems.” Industrial & Engineering Chemistry Research 50 (8): 4553–66. doi:10.1021/ie1021267. Christopher Yang, Sonia Yeh, Kalai Ramea, Saleh Zakerinia, David L. McCollum, David S. Bunch, and Joan M. Ogden. 2014. “Modeling Optimal Transition Pathways to a Low Carbon Economy in California: California TIMES (CA-TIMES) Model.” UCD-ITS-RR- 14-04. Institute of Transportation Studies, University of California, Davis. Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration. 2015. Accessed March 6. http://www.nap.edu/catalog/18805/climate-intervention-carbon- dioxide-removal-and-reliable-sequestration. Climate Intervention: Reflecting Sunlight to Cool Earth. 2015. Accessed October 5. http://www.nap.edu/catalog/18988/climate-intervention-reflecting-sunlight-to- cool-earth. Corti, Andrea, and Lidia Lombardi. 2004. “Biomass Integrated Gasification Combined Cycle with Reduced CO2 Emissions: Performance Analysis and Life Cycle Assessment (LCA).” Energy 29 (12–15): 2109–24. doi:10.1016/j.energy.2004.03.015. “Cost and Performance Data for Power Generation Technologies.” 2012. Black & Veatch. Czyzyk, Joseph, Michael P. Mesnier, and Jorge J. Moré. 1998. “The NEOS Server.” Computing in Science and Engineering 5 (3): 68–75. Damen, Kay, Radoslaw Gnutek, Joost Kaptein, Nawin Ryan Nannan, Bernardo Oyarzun, Carsten Trapp, Piero Colonna, Eric van Dijk, Joachim Gross, and André Bardow. 2011. “Developments in the Pre-Combustion CO2 Capture Pilot Plant at the Buggenum IGCC.” Energy Procedia, 10th International Conference on Greenhouse Gas Control Technologies, 4: 1214–21. doi:10.1016/j.egypro.2011.01.176. De La Torre Ugarte, Daniel G, and Daryll E Ray. 2000. “Biomass and Bioenergy Applications of the POLYSYS Modeling Framework.” Biomass and Bioenergy 18 (4): 291–308. doi:10.1016/S0961-9534(99)00095-1. Demirbaş, Ayhan. 2003. “Sustainable Cofiring of Biomass with Coal.” Energy Conversion and Management 44 (9): 1465–79. doi:10.1016/S0196-8904(02)00144-9. Dolan, Elizabeth D. 2001. “NEOS Server 4.0 Administrative Guide.” arXiv Preprint cs/0107034. http://arxiv.org/abs/cs/0107034. Dowd, Anne-Maree, Michelle Rodriguez, and Talia Jeanneret. 2015. “Social Science Insights for the BioCCS Industry.” Energies 8 (5): 4024–42. doi:10.3390/en8054024. 97

Downing, Mark, Laurence M. Eaton, Robin Lambert Graham, Matthew H. Langholtz, Robert D. Perlack, Anthony F. Turhollow Jr, Bryce Stokes, and Craig C. Brandt. 2011. “U.S. Billion-Ton Update: Biomass Supply for a Bioenergy and Bioproducts Industry.” ORNL/TM-2011/224. Oak Ridge National Laboratory (ORNL). http://www.osti.gov/scitech/biblio/1023318. “Drax Unveils High-Volume Biomass Wagon.” 2015. Railway Gazette. Accessed September 8. http://www.railwaygazette.com/news/traction-rolling-stock/single- view/view/drax-unveils-high-volume-biomass-wagon.html. Evans, Samuel, Benjamin S. Ramage, Tara L. DiRocco, and Matthew D. Potts. 2015. “Greenhouse Gas Mitigation on Marginal Land: A Quantitative Review of the Relative Benefits of Forest Recovery vs. Biofuel Production.” Environmental Science & Technology. http://pubs.acs.org/doi/abs/10.1021/es502374f. Farrell, Alexander E., and Anand R. Gopal. 2008. “Bioenergy Research Needs for Heat, Electricity, and Liquid Fuels.” MRS Bulletin 33 (04): 373–80. doi:10.1557/mrs2008.76. Farrell, Alexander E., Richard J. Plevin, Brian T. Turner, Andrew D. Jones, Michael O’Hare, and Daniel M. Kammen. 2006. “Ethanol Can Contribute to Energy and Environmental Goals.” Science 311 (5760): 506–8. doi:10.1126/science.1121416. Finley, Robert J. 2014. “An Overview of the Illinois Basin – Decatur Project.” Greenhouse Gases: Science and Technology 4 (5): 571–79. doi:10.1002/ghg.1433. Fisher Jr, C. F., S. Paik, and W. R. Schriver. 1986. “Power Plant Economy of Scale and Cost Trends: Further Analyses and Review of Empirical Studies.” Tennessee Univ., Knoxville (USA). Construction Resources Analysis. http://inis.iaea.org/search/search.aspx?orig_q=RN:18019951. Floudas, Christodoulos A., Josephine A. Elia, and Richard C. Baliban. 2012. “Hybrid and Single Feedstock Energy Processes for Liquid Transportation Fuels: A Critical Review.” Computers & Chemical Engineering 41 (June): 24–51. doi:10.1016/j.compchemeng.2012.02.008. Fourer, Robert, D. Gay, and B. Kernighan. 1993. Ampl. Boyd & Fraser. http://isites.harvard.edu/fs/docs/icb.topic621109.files/AMPLFourer_at_Chicago.pd f. “Frontiers for Young Minds.” 2015. Frontiers for Young Minds. Accessed October 12. http://kids.frontiersin.org/. Fuss, Sabine, Josep G. Canadell, Glen P. Peters, Massimo Tavoni, Robbie M. Andrew, Philippe Ciais, Robert B. Jackson, et al. 2014. “Betting on Negative Emissions.” Nature Climate Change 4 (10): 850–53. doi:10.1038/nclimate2392. Gallagher, Kelly Sims, Arnulf Grübler, Laura Kuhl, Gregory Nemet, and Charlie Wilson. 2012. “The Energy Technology Innovation System.” Annual Review of Environment and Resources 37 (1): 137–62. doi:10.1146/annurev-environ-060311-133915. Gallagher, Paul W., Heather Brubaker, and Hosein Shapouri. 2005. “Plant Size: Capital Cost Relationships in the Dry Mill Ethanol Industry.” Biomass and Bioenergy 28 (6): 565– 71. doi:10.1016/j.biombioe.2005.01.001. Gilbert, Alexander, and Benjamin K. Sovacool. 2015. “Emissions Accounting for Biomass Energy with CCS.” Nature Climate Change 5 (6): 495–96. doi:10.1038/nclimate2633.

Gollakota, Sai, and Scott McDonald. 2012. “CO2 Capture from Ethanol Production and Storage into the Mt Simon Sandstone.” Greenhouse Gases: Science and Technology 2 (5): 346–51. doi:10.1002/ghg.1305. Gropp, William, and Jorge Moré. 1997. “Optimization Environments and the NEOS Server.” Approximation Theory and Optimization, 167–82. Halder, Pradipta, Janne Pietarinen, Sari Havu-Nuutinen, and Paavo Pelkonen. 2010. “Young Citizens’ Knowledge and Perceptions of Bioenergy and Future Policy Implications.” Energy Policy, The Role of Trust in Managing Uncertainties in the Transition to a Sustainable Energy Economy, Special Section with Regular Papers, 38 (6): 3058–66. doi:10.1016/j.enpol.2010.01.046. Hansen, J., M. Sato, P. Kharecha, D. Beerling, V. Masson-Delmotte, M. Pagani, M. Raymo, D. L. Royer, and J. C. Zachos. 2008. “Target Atmospheric CO2: Where Should Humanity Aim?” Open Atmos. Sci. J. 2: 217–31. doi:10.2174/1874282300802010217. Hepple, R. P., and S. M. Benson. 2004. “Geologic Storage of Carbon Dioxide as a Climate Change Mitigation Strategy: Performance Requirements and the Implications of Surface Seepage.” Environmental Geology 47 (4): 576–85. doi:10.1007/s00254-004- 1181-2. Hertel, Thomas W., Alla A. Golub, Andrew D. Jones, Michael O’Hare, Richard J. Plevin, and Daniel M. Kammen. 2010. “Effects of US Maize Ethanol on Global Land Use and Greenhouse Gas Emissions: Estimating Market-Mediated Responses.” BioScience 60 (3): 223–31. Hnottavange-Telleen, Ken, Ivan Krapac, and Claudia Vivalda. 2009. “Illinois Basin-Decatur Project: Initial Risk-Assessment Results and Framework for Evaluating Site Performance.” Energy Procedia 1 (1): 2431–38. House, Kurt Zenz, Antonio C. Baclig, Manya Ranjan, Ernst A. van Nierop, Jennifer Wilcox, and Howard J. Herzog. 2011. “Economic and Energetic Analysis of Capturing CO2 from Ambient Air.” Proceedings of the National Academy of Sciences 108 (51): 20428–33. doi:10.1073/pnas.1012253108. Hughes, Thomas Parke. 1993. Networks of Power: Electrification in Western Society, 1880- 1930. JHU Press. https://books.google.com/books?hl=en&lr=&id=g07Q9M4agp4C&oi=fnd&pg=PR9& dq=networks+of+power&ots=BCzD-BllHQ&sig=gS-z6YSAkqUYgbOkuqqTchgNH9M. IPCC. 2007. “Chapter 9: Forestry.” In Climate Change 2007: Mitigation of Climate Change. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 2007, edited by Burt Metz, Ogunlade Davidson, Peter Bosch, Rutu Dave, and Leo Meyer, 541–84. IPCC Fourth Assessment Report. New York: Cambridge University Press. http://www.ipcc.ch/pdf/assessment-report/ar4/wg3/ar4-wg3-chapter9.pdf. ———. 2014a. “IPCC WGIII Fifth Assessment Report - Mitigation of Climate Change 2014.” April 13. http://mitigation2014.org/. ———. 2014b. “IPCC WGIII Fifth Assessment Report - Mitigation of Climate Change 2014 - Section 11.13 - Bioenergy: Climate Effects, Mitigation Options, Potential and Sustainability Implications.” April 13. http://mitigation2014.org/. Jenkins, B.M. 1997. “A Comment on the Optimal Sizing of a Biomass Utilization Facility under Constant and Variable Cost Scaling.” Biomass and Bioenergy 13 (1–2): 1–9. doi:10.1016/S0961-9534(97)00085-8. 99

J. Long, John, Mariam, Greenblatt, Jeffrey, Wei, Max, Yang, Christopher, Richter, Burton, Hannegan, Bryan, and Youngs, Heather. 2011. “California’s Energy Future-The View to 2050.” California Council on Science and Technology. http://www.ccst.us/publications/2011/2011energy.php. Kalaitzandonakes, Nicholas, Leonie A. Marks, and Steven S. Vickner. 2005. “Sentiments and Acts towards Genetically Modified Foods.” International Journal of Biotechnology 7 (1-3): 161–77. Kato, Etsushi, and Yoshiki Yamagata. 2014. “BECCS Capability of Dedicated Bioenergy Crops under a Future Land-Use Scenario Targeting Net Negative Carbon Emissions.” Earth’s Future 2 (9): 2014EF000249. doi:10.1002/2014EF000249. Keith, David W., and Douglas G. MacMartin. 2015. “A Temporary, Moderate and Responsive Scenario for Solar Geoengineering.” Nature Climate Change 5 (3): 201–6. doi:10.1038/nclimate2493. Keith, David W., and James S. Rhodes. 2002. “Bury, Burn or Both: A Two-for-One Deal on Biomass Carbon and Energy.” Climatic Change 54 (3): 375–77. doi:10.1023/A:1016187420442. Klein, David, Gunnar Luderer, Elmar Kriegler, Jessica Strefler, Nico Bauer, Marian Leimbach, Alexander Popp, et al. 2013. “The Value of Bioenergy in Low Stabilization Scenarios: An Assessment Using REMIND-MAgPIE.” Climatic Change, 1–14. Kumarappan, Subbu, Satish Joshi, and Heather L. MacLean. 2009. “Biomass Supply for Biofuel Production: Estimates for the United States and Canada.” SSRN Scholarly Paper ID 1420944. Rochester, NY: Social Science Research Network. http://papers.ssrn.com/abstract=1420944. Larson, Eric D., Giulia Fiorese, Guangjian Liu, Robert H. Williams, Thomas G. Kreutz, and Stefano Consonni. 2010. “Co-Production of Decarbonized Synfuels and Electricity from Coal + Biomass with CO2 Capture and Storage: An Illinois Case Study.” Energy & Environmental Science 3 (1): 28–42. doi:10.1039/B911529C. Leboreiro, Jose, and Ahmad K. Hilaly. 2011. “Biomass Transportation Model and Optimum Plant Size for the Production of Ethanol.” Bioresource Technology 102 (3): 2712–23. doi:10.1016/j.biortech.2010.10.144. Lemoine, Derek M., Richard J. Plevin, Avery S. Cohn, Andrew D. Jones, Adam R. Brandt, Sintana E. Vergara, and Daniel M. Kammen. 2010. “The Climate Impacts of Bioenergy Systems Depend on Market and Regulatory Policy Contexts.” Environmental Science & Technology 44 (19): 7347–50. doi:10.1021/es100418p. Lenton, Timothy M. 2010. “The Potential for Land-Based Biological CO2 Removal to Lower Future Atmospheric CO2 Concentration.” Carbon Management 1 (1): 145–60. doi:10.4155/cmt.10.12. Lieberman, Gerald A., and Linda L. Hoody. 1998. “Closing the Achievement Gap: Using the Environment as an Integrating Context for Learning. Executive Summary.” http://eric.ed.gov/?id=ED428942. Liska, Adam J., Haishun Yang, Maribeth Milner, Steve Goddard, Humberto Blanco-Canqui, Matthew P. Pelton, Xiao X. Fang, Haitao Zhu, and Andrew E. Suyker. 2014. “Biofuels from Crop Residue Can Reduce Soil Carbon and Increase CO2 Emissions.” Nature Climate Change 4 (5): 398–401. doi:10.1038/nclimate2187.

100

Liu, Guangjian, Eric D. Larson, Robert H. Williams, and Xiangbo Guo. 2015. “Gasoline from Coal And/or Biomass with CO2 Capture and Storage. 2. Economic Analysis and Strategic Context.” Energy & Fuels 29 (3): 1845–59. doi:10.1021/ef502668n. Liu, Guangjian, Eric D. Larson, Robert H. Williams, Thomas G. Kreutz, and Xiangbo Guo. 2011. “Making Fischer−Tropsch Fuels and Electricity from Coal and Biomass: Performance and Cost Analysis.” Energy & Fuels 25 (1): 415–37. doi:10.1021/ef101184e. Lomax, Guy, Mark Workman, Timothy Lenton, and Nilay Shah. 2015. “Reframing the Policy Approach to Greenhouse Gas Removal Technologies.” Energy Policy 78 (March): 125–36. doi:10.1016/j.enpol.2014.10.002. Lutsey, Nicholas, and Daniel Sperling. 2009. “Greenhouse Gas Mitigation Supply Curve for the United States for Transport versus Other Sectors.” Transportation Research Part D: Transport and Environment 14 (3): 222–29. doi:10.1016/j.trd.2008.12.002. Mann, Michael E. 2009. “Defining Dangerous Anthropogenic Interference.” Proceedings of the National Academy of Sciences 106 (11): 4065–66. doi:10.1073/pnas.0901303106. Marrison, Christopher I., and Eric D. Larson. 1995. “Cost versus Scale for Advanced Plantation-Based Biomass Energy Systems in the USA and Brazil.” National Renewable Energy Lab., Golden, CO (United States). http://www.osti.gov/scitech/biblio/115349. Maxson, A., N. Holt, D. Thimsen, J. Wheeldon, and E. Worrell. 2011. “Advanced Coal Power Systems with CO2 Capture: EPRI’s CoalFleet for Tomorrow® Vision—2011 Update: A Summary of Technology Status and Research, Development, and Demonstrations.” Technical Report 1023468. Palo Alto, CA: Electric Power Research Institute. Max Wei, Jeffrey Greenblatt, Sally Donovan, James Nelson, Ana Mileva, Josiah Johnston, and Daniel Kammen. 2013. “Scenarios for Meeting California’s 2050 Climate Goals: California’s Carbon Challenge Phase II: Volume I: Non-Electricity Sectors and Overall Scenario Results.” 500-10-047. Lawrence Berkeley National Laboratory. http://eetd.lbl.gov/sites/all/files/ca-2050-climate-goals.pdf. McGowin, C. 2007. “Renewable Energy Technical Assessment Guide. TAG-RE:2007.” Palo Alto, CA: Electric Power Research Institute. Meinshausen, Malte, Nicolai Meinshausen, William Hare, Sarah CB Raper, Katja Frieler, Reto Knutti, David J. Frame, and Myles R. Allen. 2009. “Greenhouse-Gas Emission Targets for Limiting Global Warming to 2 C.” Nature 458 (7242): 1158–62. Metz, Bert, Ogunlade Davidson, H. C. De Coninck, Manuela Loos, and L. A. Meyer. 2005. “IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change.” IPCC, Cambridge University Press: Cambridge, United Kingdom and New York, USA 4. Middleton, Richard S., and Jeffrey M. Bielicki. 2009. “A Scalable Infrastructure Model for Carbon Capture and Storage: SimCCS.” Energy Policy 37 (3): 1052–60. doi:10.1016/j.enpol.2008.09.049. Mike Matuszewski, Eric Lewis, and Steve Herron. 2013. “Greenhouse Gas Reductions in the Power Industry Using Domestic Coal and Biomass - Volume 2: PC Plants.” NETL/DOE-2012/1547. National Energy Technology Laboratory.

101

Mike Matuszewski, James Black, and Eric Lewis. 2013. “Greenhouse Gas Reductions in the Power Industry Using Domestic Coal and Biomass - Volume 1: IGCC.” NETL/DOE- 2012/1546. National Energy Technology Laboratory. Milbrandt, Anelia, and National Renewable Energy Laboratory (U.S.). 2006. A Geographic Perspective on the Current Biomass Resource Availability in the United States. National Renewable Energy Laboratory. Mileva, Ana, James H. Nelson, Josiah Johnston, and Daniel M. Kammen. 2013. “SunShot Solar Power Reduces Costs and Uncertainty in Future Low-Carbon Electricity Systems.” Environmental Science & Technology 47 (16): 9053–60. doi:10.1021/es401898f. Milne, Jennifer L., and Christopher B. Field. 2014. “Assessment Report from the GCEP Workshop on Energy Supply with Negative Carbon Emissions.” Accessed March 10. http://www.stanford.edu/group/gcep/pdfs/rfpp/Report%20from%20GCEP%20W orkshop%20on%20Energy%20Supply%20with%20Negative%20Emissions.pdf. Mullins, Kimberley A., Aranya Venkatesh, Amy L. Nagengast, and Matt Kocoloski. 2014. “Regional Allocation of Biomass to U.S. Energy Demands under a Portfolio of Policy Scenarios.” Environmental Science & Technology 48 (5): 2561–68. doi:10.1021/es405222w. National Academy of Sciences. 2012. “Next Generation Science Standards. APPENDIX J - Science, Technology, Society and the Environment.” National Energy Technology Laboratory. n.d. “National Carbon Sequestration Database and Geographic Information System.” United States Department of Energy. Nelson, James H. 2013. “California’s Carbon Challenge: Scenarios for Achieving 80% Emissions Reduction in 2050.” Ph.D., United States -- California: University of California, Berkeley. Nelson, James, Josiah Johnston, Ana Mileva, Matthias Fripp, Ian Hoffman, Autumn Petros- Good, Christian Blanco, and Daniel M. Kammen. 2012. “High-Resolution Modeling of the Western North American Power System Demonstrates Low-Cost and Low- Carbon Futures.” Energy Policy 43 (April): 436–47. doi:10.1016/j.enpol.2012.01.031. Nowlin, Matthew C. 2011. “Theories of the Policy Process: State of the Research and Emerging Trends.” Policy Studies Journal 39 (April): 41–60. doi:10.1111/j.1541- 0072.2010.00389_4.x. Office of Fossil Energy. 2015. “11,202,720 Metrics Tons of CO2 Injected as of October 14, 2015.” Accessed November 25. http://energy.gov/fe/11202720-metric-tons-co2- injected-october-14-2015. Pacala, Stephen, and Robert Socolow. 2004. “Stabilization Wedges: Solving the Climate Problem for the next 50 Years with Current Technologies.” Science 305 (5686): 968– 72. Parker, Nathan. 2012. “Spatially Explicit Projection of Biofuel Supply for Meeting Renewable Fuel Standard.” Transportation Research Record: Journal of the Transportation Research Board 2287 (-1): 72–79. doi:10.3141/2287-09. Parker, Nathan Chandler. 2011. “Modeling Future Biofuel Supply Chains Using Spatially Explicit Infrastructure Optimization.” http://trid.trb.org/view.aspx?id=1134666. Plevin, R. J., and D. M. Kammen. 2013. Indirect Land Use and Greenhouse Gas Impacts of Biofuels. Encyclopedia of Biodiversity . Waltham. Academic Press.

102

Pratt, Kimberley, and Dominic Moran. 2010. “Evaluating the Cost-Effectiveness of Global Biochar Mitigation Potential.” Biomass and Bioenergy 34 (8): 1149–58. doi:10.1016/j.biombioe.2010.03.004. Rayner, Steve. 2010. “Trust and the Transformation of Energy Systems.” Energy Policy, The Role of Trust in Managing Uncertainties in the Transition to a Sustainable Energy Economy, Special Section with Regular Papers, 38 (6): 2617–23. doi:10.1016/j.enpol.2009.05.035. Read, Peter, and Jonathan Lermit. 2005. “Bio-Energy with Carbon Storage (BECS): A Sequential Decision Approach to the Threat of Abrupt Climate Change.” Energy 30 (14): 2654–71. doi:10.1016/j.energy.2004.07.003. Rhodes, James S., and David W. Keith. 2005. “Engineering Economic Analysis of Biomass IGCC with Carbon Capture and Storage.” Biomass and Bioenergy 29 (6): 440–50. doi:10.1016/j.biombioe.2005.06.007. ———. 2009. “Biomass Co-Utilization with Unconventional Fossil Fuels to Advance Energy Security and Climate Policy.” http://legacy.bipartisanpolicy.org/sites/default/files/Biomass%20co- utilization%20for%20energy%20and%20climate%20security%20- %20synthesis%20report.pdf. Riahi, Keywan, Shilpa Rao, Volker Krey, Cheolhung Cho, Vadim Chirkov, Guenther Fischer, Georg Kindermann, Nebojsa Nakicenovic, and Peter Rafaj. 2011. “RCP 8.5—A Scenario of Comparatively High Greenhouse Gas Emissions.” Climatic Change 109 (1-2): 33–57. doi:10.1007/s10584-011-0149-y. Ricke, Katharine L, and Ken Caldeira. 2014. “Maximum Warming Occurs about One Decade after a Carbon Dioxide Emission.” Environmental Research Letters 9 (12): 124002. doi:10.1088/1748-9326/9/12/124002. Robert Finley. 2012. “Demonstration of CO2-Enhanced Oil Recovery Potential in the Illinois Basin.” DOE Report DE-FC26-05NT42588. Midwest Geological Sequestration Consortium. http://www.sequestration.org/resources/publish/MGSC%20Final%20Phase%20II %20Report.pdf. Rogelj, Joeri, Gunnar Luderer, Robert C. Pietzcker, Elmar Kriegler, Michiel Schaeffer, Volker Krey, and Keywan Riahi. 2015. “Energy System Transformations for Limiting End- of-Century Warming to below 1.5 °C.” Nature Climate Change 5 (6): 519–27. doi:10.1038/nclimate2572. Rose, Steven K., Elmar Kriegler, Ruben Bibas, Katherine Calvin, Alexander Popp, Detlef P. van Vuuren, and John Weyant. 2013. “Bioenergy in Energy Transformation and Climate Management.” Climatic Change 123 (3-4): 477–93. doi:10.1007/s10584- 013-0965-3. Rubio. 2012. Energy: Renewable Bioenergy Projects. SB-1122. https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201120120SB1 122. Sanchez, Daniel L., and Daniel M. Kammen. 2015. “Removing Harmful Greenhouse Gases from the Air Using Energy from Plants.” Frontiers for Young Minds 3 (September). doi:10.3389/frym.2015.00014. Sanchez, Daniel L., James H. Nelson, Josiah Johnston, Ana Mileva, and Daniel M. Kammen. 2015a. “Biomass Enables the Transition to a Carbon-Negative Power System across 103

Western North America.” Nature Climate Change 5 (3): 230–34. doi:10.1038/nclimate2488. ———. 2015b. “Reply to ‘Emissions Accounting for Biomass Energy with CCS.’” Nature Climate Change 5 (6): 496–496. doi:10.1038/nclimate2634. Sanderson, Matt A., Richard P. Egg, and Arthur E. Wiselogel. 1997. “Biomass Losses during Harvest and Storage of Switchgrass.” Biomass and Bioenergy 12 (2): 107–14. Sathaye, Kiran J., Marc A. Hesse, Martin Cassidy, and Daniel F. Stockli. 2014. “Constraints on the Magnitude and Rate of CO2 Dissolution at Bravo Dome Natural Gas Field.” Proceedings of the National Academy of Sciences 111 (43): 15332–37. doi:10.1073/pnas.1406076111. Scott, Vivian, Stuart Gilfillan, Nils Markusson, Hannah Chalmers, and R. Stuart Haszeldine. 2013. “Last Chance for Carbon Capture and Storage.” Nature Climate Change 3 (2): 105–11. doi:10.1038/nclimate1695. Scott, Vivian, R. Stuart Haszeldine, Simon F. B. Tett, and Andreas Oschlies. 2015. “Fossil Fuels in a Trillion Tonne World.” Nature Climate Change 5 (5): 419–23. doi:10.1038/nclimate2578. Searchinger, Tim, and Ralph Heimlich. 2015. “Avoiding Bioenergy Competition for Food Crops and Land.” Working Paper. Creating a Sustainable Food Future. World Resources Institute. http://www.wri.org/publication/avoiding-bioenergy- competition-food-crops-and-land. Searchinger, Timothy D., Steven P. Hamburg, Jerry Melillo, William Chameides, Petr Havlik, Daniel M. Kammen, Gene E. Likens, et al. 2009. “Fixing a Critical Climate Accounting Error.” Science 326 (5952): 527–28. doi:10.1126/science.1178797. Searchinger, Timothy, Ralph Heimlich, R. A. Houghton, Fengxia Dong, Amani Elobeid, Jacinto Fabiosa, Simla Tokgoz, Dermot Hayes, and Tun-Hsiang Yu. 2008. “Use of U.S. Croplands for Biofuels Increases Greenhouse Gases Through Emissions from Land- Use Change.” Science 319 (5867): 1238–40. doi:10.1126/science.1151861. Searcy, Erin, Peter Flynn, Emad Ghafoori, and Amit Kumar. 2007. “The Relative Cost of Biomass Energy Transport.” Applied Biochemistry and Biotechnology 137-140 (1- 12): 639–52. doi:10.1007/s12010-007-9085-8. Shallcross, Tony, and John Robinson. 2008. Sustainability Education, Whole School Approaches, and Communities of Action. Springer. http://link.springer.com/content/pdf/10.1007/978-1-4020-6416-6.pdf#page=319. Slade, Raphael, Ausilio Bauen, and Robert Gross. 2014. “Global Bioenergy Resources.” Nature Climate Change 4 (2): 99–105. doi:10.1038/nclimate2097. Smit, Berend, Jeffrey R. Reimer, Curtis M. Oldenburg, and Ian C. Bourg. 2014. Introduction to Carbon Capture and Sequestration. EPFL-BOOK-200456. World Scientific. http://www.worldscientific.com/sda/1043/intro-carbon-capture- sequestration.pdf. Solomon, Barry D. 2010. “Biofuels and Sustainability.” Annals of the New York Academy of Sciences 1185 (1): 119–34. doi:10.1111/j.1749-6632.2009.05279.x. Strand, Stuart E., and Gregory Benford. 2009. “Ocean Sequestration of Crop Residue Carbon: Recycling Fossil Fuel Carbon Back to Deep Sediments.” Environmental Science & Technology 43 (4): 1000–1007. doi:10.1021/es8015556. Sultana, Arifa, and Amit Kumar. 2012. “Optimal Siting and Size of Bioenergy Facilities Using Geographic Information System.” Applied Energy 94: 192–201. 104

Sultana, Arifa, Amit Kumar, and Don Harfield. 2010. “Development of Agri-Pellet Production Cost and Optimum Size.” Bioresource Technology 101 (14): 5609–21. doi:10.1016/j.biortech.2010.02.011. “Taylorville Energy Center.” 2014. Accessed October 23. http://www.globalccsinstitute.com/projects/12566. Tilman, David, Robert Socolow, Jonathan A. Foley, Jason Hill, Eric Larson, Lee Lynd, Stephen Pacala, et al. 2009. “Beneficial Biofuels—The Food, Energy, and Environment Trilemma.” Science 325 (5938): 270–71. doi:10.1126/science.1177970. Trancik, J. E. 2006. “Scale and Innovation in the Energy Sector: A Focus on Photovoltaics and Nuclear Fission.” Environmental Research Letters 1 (1): 014009. doi:10.1088/1748-9326/1/1/014009. University of Tennessee. 2007. “Agricultural Policy Analysis Center: POLYSYS - Policy Analysis System.” http://www.agpolicy.org/polysys.html. Vaz, A. Ismael F., and Luís N. Vicente. 2009. “PSwarm: A Hybrid Solver for Linearly Constrained Global Derivative-Free Optimization.” Optimization Methods & Software 24 (4-5): 669–85. Victor, David G., and Robert O. Keohane. 2010. “The Regime Complex for Climate Change.” SSRN Scholarly Paper ID 1643813. Rochester, NY: Social Science Research Network. http://papers.ssrn.com/abstract=1643813. Vuuren, Detlef P. van, Elke Stehfest, Michel G. J. den Elzen, Tom Kram, Jasper van Vliet, Sebastiaan Deetman, Morna Isaac, et al. 2011. “RCP2.6: Exploring the Possibility to Keep Global Mean Temperature Increase below 2°C.” Climatic Change 109 (1-2): 95– 116. doi:10.1007/s10584-011-0152-3. Wei, Max, James H Nelson, Jeffery B Greenblatt, Ana Mileva, Josiah Johnston, Michael Ting, Christopher Yang, Chris Jones, James E McMahon, and Daniel M Kammen. 2013. “Deep Carbon Reductions in California Require Electrification and Integration across Economic Sectors.” Environmental Research Letters 8 (1): 014038. doi:10.1088/1748-9326/8/1/014038. “What Is Carbon Removal?” 2015. The Center for Carbon Removal. Accessed October 4. http://www.centerforcarbonremoval.org/what-is-carbon-removal/. 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. doi:10.1126/science.1208365. Williams, Robert H., Guangjian Liu, Thomas G. Kreutz, and Eric D. Larson. 2011. “Coal and Biomass to Fuels and Power.” Annual Review of Chemical and Biomolecular Engineering 2 (1): 529–53. doi:10.1146/annurev-chembioeng-061010-114126. Wong-Parodi, Gabrielle, Hadi Dowlatabadi, Tim McDaniels, and Isha Ray. 2011. “Influencing Attitudes toward Carbon Capture and Sequestration: A Social Marketing Approach.” Environmental Science & Technology 45 (16): 6743–51. Wright, Mark, and Robert C. Brown. 2007. “Establishing the Optimal Sizes of Different Kinds of Biorefineries.” Biofuels, Bioproducts and Biorefining 1 (3): 191–200. doi:10.1002/bbb.25. Xu, Yixiang, Loren Isom, and Milford A. Hanna. 2010. “Adding Value to Carbon Dioxide from Ethanol Fermentations.” Bioresource Technology 101 (10): 3311–19. doi:10.1016/j.biortech.2010.01.006. 105

Youngs, Heather, and Chris Somerville. 2014. “Best Practices for Biofuels.” Science 344 (6188): 1092–96. Zitney, S. E., M. O. Osawe, L. Collins, E. Ferguson, D. G. Sloan, W. A. Fiveland, and J. M. Madsen. 2006. “Advanced Process Co-Simulation of the FutureGen Power Plant.” In Proc. of the 31st International Technical Conference on Coal Utilization & Fuel Systems, 21–25.

106