Supporting Online Materials
Can Radiative Forcing Be Limited to 2.6Wm-2 Without Negative Emissions From Bioenergy AND CO2 Capture and Storage? In Climatic Change
Authors: James Edmonds,[1],[2] Patrick Luckow,[3] Katherine Calvin2, Marshall Wise2, Jim Dooley2, Page Kyle2, Sonny Kim2, Pralit Patel2,and Leon Clarke2
Limitations and Future Research Directions
Integrated assessment models, like all models, are simplified representations of real world phenomena. But, they are not the real world itself. Simplifications are intended to allow a clearer focus on phenomena of interest, these simplifications are also limiting. The potential virtue of integrated assessment models grows out of the fact that they couple a diverse set of phenomena to shed light on the implications of interactions among the major elements of the human and biogeophysical Earth system. While many interesting results flow from the numerical experiments that we report here, it is important to bear in mind that, like any model results, they depend on numerous explicit and implicit assumptions ranging from model structure and parameterization to driving assumptions such as technology performance, future economic activity levels, future population, and the future policy environment. Several of these areas stand out as potentially fruitful areas for future research.
We have undertaken this analysis under the assumption that the climate remains unchanged. This is clearly an unrealistic assumption. The question that needs to be tested by future research is the degree to which this assumption might affect the results reported here. By experimental design we are exploring a world in which RF is limited. That in turn will limit the degree of climate change feedback, particularly when coupled with feedbacks from changing CO2 concentrations. But, fully coupled experiments, including albedo and other direct land-use-land-cover effects, need to be undertaken.
Assumptions about economic activity and demographics determine the scale of human activities and hence energy and land-use systems. We explore only one set of underlying assumptions. Other socioeconomic pathways are clearly possible and experiments to test the robustness of results in the face of these assumptions are also needed.
Similarly, this analysis is predicated on one set of future technology performance assumptions. Other possible technology development paths are also possible. This analysis explores the relative contribution of some technologies but clearly not all technologies. Potentially one of the more interesting technology assumptions left unexplored was the rate of improvement in crop yields. Land-use patterns and land-use change emissions are affected by the rate of change of crop yields and exploring the effect of varying crop-yield improvement assumptions is a potentially fruitful direction for future research. Similarly, given the importance of transportation technology in a world with deep emissions reduction, would these results be different in a world with advanced hydrogen or electric vehicle technologies available?
The specific set of policy assumptions that we employ are another potentially rich vein of future research. We have assumed that institutions exist that reward negative emissions at the same carbon price as emissions are penalized—everywhere in the system including land-use systems as well as energy and industrial systems. We assumed that terrestrial carbon emissions could be priced at the same rate as fossil fuel and industrial emissions. Implications of imperfect policy institutions need to be considered. Given the importance of land-use change emissions in limiting RF to low levels, research to explore alternative land-use policies is clearly a high near-term priority. One interesting question that needs investigation is, would BECCS increase or decrease the cost of emissions mitigation in a world that continued to value terrestrial carbon emissions at a price of zero? What if the REDD were the world’s terrestrial emissions mitigation policy? What if bioenergy were treated as if it were a fossil fuel rather than a renewable energy form?
Implementation of the Delayed Accession Policy Scenarios
SOM Table 1: Assumed pattern of regional accession to emissions mitigation
Regions Joining Mitigation Regime / Year when Region First Takes on Emissions Mitigation Obligation / Year when Region Ends Phase-in Period and Faces Common Global Emissions PriceWestern Europe, Eastern Europe, Japan / 2020 / NA
Australia/NZ, Canada, China, Korea, USA / 2030 / 2040
India, L. America, Other South & East Asia / 2050 / 2070
Africa, FSU, Middle East / 2070 / 2095
A price is applied to carbon emissions as soon as a region begins emissions mitigation activities. It is assumed that each region phases in the carbon price over a period of years after which the path follows the reference international price Table 1[4].
Geographic Distribution of Carbon Reservoirs in GCAM Scenarios
SOM Figure 1: Regional Distribution of GCAM CO2 Storage Capacity
Energy System Implications of Technology Availability
The energy system varies dramatically both as a consequence of technology availability and the policy environment employed to affect RF limits. The global energy system is substantially larger in the reference scenario than in any of the control scenarios. Limiting RF to 2.6Wm-2 reduces aggregate end-use energy demand, increases reliance on a decarbonized power generation[5]. The relative roles of bioenergy and CCS depend importantly on the technology environment in which the 2.6Wm-2 limit is affected. When bioenergy is assumed to be available for emissions mitigation, bioenergy production expands dramatically relative to the reference scenario. Deployment under mitigation is more than double reference scenario deployment[6]. When both bioenergy and CCS are assumed available, bioenergy is used extensively in combination with CCS but deployed relatively evenly between use in power generation and as a liquid fuel feedstock in refining. When CCS was unavailable, bioenergy was used predominantly as a liquid fuel feedstock.
Decarbonization of end use energy use leads to expanded use of electricity both in relative and absolute terms except under the T5 (LowTech) assumptions. Decarbonization of the power sector takes different forms reflecting both the cost assumptions in GCAM’s technology set and their assumed availability. When all technologies are assumed available (T1), then there is something of a balanced expansion in the use of CCS with fossil fuels and bioenergy, expanded use of nuclear power, and expansion of other renewable energy forms. Except in the T5 (LowTech) scenario, nuclear expands to fill the base-load power requirements when either CCS or bioenergy is assumed unavailable. In the T5 (LowTech) scenario decarbonization is affected entirely by non-biomass renewable energy forms.
SOM Figure 2: End-use Energy by Fuel (Oil, Gas, Coal, Biomass, Electricity, & H2) and Sector (Residential, Industrial, & Transport)Reference
T1 (Ref) x Idealized
/ T1 (Ref) x Delayed
T2 (No Bio) x Idealized
/ T2 (No Bio) x Delayed
T3 (No CCS) x Idealized
/ T3 (No CCS) x Delayed
T4 (No Bio & No CCS) x Idealized
/ T4 (No Bio & No CCS) x Delayed
T5 (LowTech) x Idealized
/ T5 (LowTech) x Delayed
SOM Figure 3: Electric Power Generation Technology Mix
Reference
T1 (Ref) x Idealized
/ T1 (Ref) x Delayed
T2 (No Bio) x Idealized
/ T2 (No Bio) x Delayed
T3 (No CCS) x Idealized
/ T3 (No CCS) x Delayed
T4 (No Bio & No CCS) x Idealized
/ T4 (No Bio & No CCS) x Delayed
T5 (LowTech) x Idealized
/ T5 (LowTech) x Delayed
SOM Figure 4: Bioenergy Utilization (EJ per year)
Reference
T1 (Ref) x Idealized
/ T1 (Ref) x Delayed
T3 (No CCS) x Idealized
/ T3 (No CCS) x Delayed
The marginal cost of BECCS is set by the carbon price. That is, like all technologies, BECCS is deployed up to the point at which the marginal cost of mitigating another ton of carbon with BECCS is equal to the price of carbon. Since the price of bioenergy is driven by the carbon price, the bioenergy price rises in all of the control scenarios in which bioenergy is available, see SOM Figure 5 below, whereas it is relatively stable in the reference scenario. Part of the increase in the price of bioenergy is an expansion effect. There is more land allocated to bioenergy than in the reference scenarios. But a larger part of the story is the escalation of the land rental rate to reflect the asset value of carbon in terrestrial systems.
SOM Figure 5: Global Bioenergy Prices
Terrestrial Responses to Climate Policy
As discussed in the paper, GCAM includes energy, economy and terrestrial system interactions. The effect of introducing a carbon price dramatically changes land-use in GCAM. One feature of the scenarios that we examine here is the effect of a carbon price on land use. Consider for example the T1(Ref)xIdealizedscenario. The carbon tax in the first time policy period of the T1(Ref)xIdealized scenario is $60/tC. The imposition of this price on carbon dramatically changes the economics of land use and dramatically changes the allocation of land in the model.
At a price of $60/tC, the asset value of the terrestrial carbon stock, which is approximately 2000 billion tons, immediately shifts from zero to approximately $120 trillion. Every ton of carbon stored is worth $60, which is the same price that is charged for each ton emitted. This in turn changes the expected profitability of every activity undertaken by landowners. Carbon management becomes an important economic consideration. The initial impact changes relative food prices and therefore shifts dietary composition. The shift away from high-carbon-intensity elements such as cattle, frees up land both because less pasture is needed, but also because less land is tied up growing crops to feed animals. In addition, carbon emissions/sequestration effects of land use directly affect profitability. This in turn affects relative profitability of every land-use,shifting land use toward options that are less carbon-intensive. Land owners shift toward holding a larger stock of carbon. Unmanaged forest land holdings are rewarded and increase dramatically. Carbon prices, which are plotted in Figure 2 in the paper, for the 10 policy scenarios are given in the Table below:
SOM Table 2: CO2 Prices for Ten Policy Scenarios
The monetization of the carbon stocks in the terrestrial biosphere leads to dramatic shifts in land use. Between 2015 (year in which policy is initiated) and 2020 in the T1(Ref)xIdealized scenario, is approximately 10Mkm2 (1Bha) change use to forest systems predominantly, but not exclusively in the tropics. By the end of the century approximately 18Mkm2 (1.8Bha) have changed use relative to 2015 shifting toward unmanaged forest systems. Total land-use change between 2015 and 2095 is approximately 14% of terrestrial land area tracked by GCAM (128.8 x 10^6 km2).
GCAM uses a dynamic terrestrial carbon cycle model and treats each land use and land-use change explicitly. For example it uses a forest-growth model which assumes an “S” shaped forest growth curve. GCAM does NOT assume constant carbon uptake for land-use change. See for example the carbon density function for above-ground biomass for Tropical Unmanaged Forest in AEZ6 of Latin America shown below. Mature age there is 40years, with a potential carbon density of 19.7KgC/m2. This is a Chapman-Richards equation, which is a pretty standard function to model forest growth. See for example, Pienaar, L. V., and K. J. Turnbull. 1973. “The Chapman-Richards Generalization of Von Bertalanffy’s Growth Model for Basal Area Growth and Yield in Even - Aged Stands”. Forest Science, Volume 19, Number 1, 1 March 1973, pp. 2-22(21)
SOM Figure 6: Above-ground carbon dynamics in a tropical forest in GCAM land region AEZ6 located in Latin America
In the GCAM T1(Ref)xIdealized scenario carbon uptake is approximately 4.1PgC/y (15 PgCO2-e/y) in 2025. Cumulative uptake over the course of the century is ~160PgC (between 2020 and 2095). This is smaller than the ~300 PgC cumulative sequestration that are implied by a continuation of the linear assumptions cited in the comment. Between 2015 and 2095 the GCAM T1(Ref)xIdealized scenario stores ~160 PgC(584 PgCO2-e)on ~18Mkm2.
SOM Figure 7: Land Use by Major Category, 2020 to 2095, Across Scenarios
SOM Figure 8: CO2 Capture from Fossil Fuels, Cement Manufacture, and Bioenergy, 2020 to 2095
SOM Figure 9: Primary Energy Consumption by Energy Form (Primary Energy Equivalent, Fixed Values)Reference
T1 (Ref) x Idealized
/ T1 (Ref) x Delayed
T2 (No Bio) x Idealized
/ T2 (No Bio) x Delayed
T3 (No CCS) x Idealized
/ T3 (No CCS) x Delayed
T4 (No Bio & No CCS) x Idealized
/ T4 (No Bio & No CCS) x Delayed
T5 (LowTech) x Idealized
/ T5 (LowTech) x Delayed
SOM Figure 10: Refined Liquid Fuel Supply by Source
Reference
T1 (Ref) x Idealized
/ T1 (Ref) x Delayed
T2 (No Bio) x Idealized
/ T2 (No Bio) x Delayed
T3 (No CCS) x Idealized
/ T3 (No CCS) x Delayed
T4 (No Bio & No CCS) x Idealized
/ T4 (No Bio & No CCS) x Delayed
T5 (LowTech) x Idealized
/ T5 (LowTech) x Delayed
1
[1] Corresponding author, email: ; telephone: +1 301.314.6749; fax: +1 301.314.6719.
[2] Joint Global Change Research Institute (JGCRI) of the Pacific Northwest National Laboratory, 5825 University Research Court, Suite 3500, College Park, MD, 20740.
[3] Synapse Energy Economics, Inc., 485 Massachusetts Avenue, Suite 2, Cambridge, MA 02139
[4] Edmonds et al. (2008) discusses the potential problem of carbon price shock to participants joining a mitigation coalition already in progress.
[5] See Supporting Online Material Figures 2 and 3.
[6] See Supporting Online Material Figure 4.