Impacts of Technology Uncertainty on Energy Use, Emission and Abatement Cost in USA: Simulation results from Environment Canada's Integrated Assessment Model.

• Author: Zhu, Yunfa

1. INTRODUCTION

   Historically energy use and economic activity level have been tightly coupled, and energy use is a vital and indispensable ingredient of economic growth (Toman and Jemelkova 2002; Stern and Cleveland 2004; Guttormsen 2009). However, energy production, energy transformation and energy use, in particular the combustion of fossil-fuels results in energy-related greenhouse gas (GHG) emissions, which accounts for the majority of anthropogenic GHG emissions. For example, in 2009, the total GHG emissions excluding land use change in the U.S. are some 6608 MTCO2e, of which 87% are energy related. (1) The United States Energy Information Administration (EIA) forecasts that in the absence of new policies, fossil fuel use will still dominate primary energy use in the U.S. in 2035 (US EIA 2011a and 2011b).

   The scientific evidence confirms that increasing anthropogenic GHG emissions is an important contributor to global warming (Farley 2008; IPCC 2001,2007). Actions by all large emitters in the developed and developing countries are necessary for meaningful global GHG reductions. In the absence of policy change, "the overall costs and risks of climate change will be equivalent to losing at least 5% of global GDP each year, now and forever and if a wider range of risks and impact is taken into account the estimates of damage could rise to 20% of GDP or more" (Stern 2006). The future path of emissions growth and the abatement costs under climate policy would heavily depend on the status of end-use technology and clean technologies as these can heavily influence the way energy is produced and used.

   However, there is a high degree of uncertainty around the evolution of the future energy system. To explore how different factors might influence the evolution of the energy system and GHG emissions, we focus on the possible future development status of five key technologies: (1) end use technology, (2) CCS technologies, (3) nuclear energy, (4) wind & solar energy, and (5) biomass electric energy. The possible technology scenarios analyzed in this paper were identified in a model comparison exercise undertaken by the Energy Modeling Forum (EMF 24 U.S. Scenarios: Final version, 2012). To assess the role of uncertainties, this paper considers two extreme scenarios for each technology category; optimistic or high status and pessimistic or low status. Technology parameter values are applied in simulation exercises employing Environment Canada's Integrated Assessment Model (EC-IAM). To understand the role of technology in energy development and consequent emissions pathways, the model is calibrated to several baseline scenarios based on differing mixes of technology development and then policy simulations are performed for each baseline projection.

   The rest of the paper is organized as follows: Section 2 provides a brief overview of the EC-IAM model. Section 3 presents the simulation results and Section 4 discusses the main findings and conclusions.

2. OVERVIEW OF EC-IAM

   Environment Canada's Integrated Assessment Model (EC-IAM) (2) is based on the structure of the Model for Evaluating the Regional and Global Effects (MERGE) (Manne 1976; Manne and Richels, 1992; Manne et al, 1995; US Climate Change and Science Program, 2007). Modifications specific to EC-IAM include the explicit representation of Canada as a model region with specific extensions to represent oil sands reserves that are central to the evolution of Canada's oil producing sector and electricity generation mix reflecting Canadian endowments (NEB 2011). EC-IAM is an intertemporal multi-regional global computable general equilibrium (CGE) model suitable for analyzing regional and global effects of climate policies. It integrates an economy-energy model consisting of a top-down macroeconomic submodel and a bottom-up energy supply submodel with an aggregate climate submodel into an integrated model system to quantify alternative ways to assess climate policies.

   2.1 Macroeconomic submodel

   In each region production is aggregated to a single macro sector with a nested constant elasticity of substitution (CES) function transforming the price responsive inputs comprising capital, labour, electric energy and non-electric energy into a numeraire good. The representative agent's instantaneous utility function in each region is a CES function of consumption of macro good and the passenger transportation. Economic decision in the model is described by Ramsey-Solow paradigm. The representative agent in each region chooses intertemporal consumption, level of various modes of passenger transportation, saving and investment to maximize total discounted utility subject to an intertemporal budget constraint. Investment forms next period's new capital. Regions are linked through international flows treating the tradable goods as internationally homogeneous goods. Production, input demand and consumption and passenger transportation demand and instantaneous utility as well are all vintaged as "putty-clay" formulation. Population, labour and automatic energy efficiency (AEEI) improvement index are exogenously specified based on best available information. In "putty-clay" formulation, old vintages equal the survival part of last period depending on the depreciation rate.

   The production of new vintage output ([YN.sub.rt]) at period t in regions r is given by a CES function as follows:

   [YN.sub.rt] = [A.sub.rt] [[[[theta].sub.rt][([KN.sup.[alpha].sub.rt][LN.sup.1-[alpha].sub.rt]).sup.[rho]] + (1 - [[theta].sub.rt])([[EN.sup.[beta].sub.rt][NN.sup.1-[beta].sub.rt]).sup.[rho]]].sup.1/[rho]] (1)

   Where [KN.sub.rt], [LN.sub.rt], [EN.sub.rt] and [NN.sub.rt] are respectively the inputs of new vintaged capital, labour, electric and non-electric energy at period t in region r, and [A.sub.rt] is the reference production efficiency index.

   The new vintaged instantaneous utility ([UN.sub.rt]) of representative agent at period t in region r is given by

   [UN.sub.rt] = [UREF.sub.rt][[[[alpha].sub.rt][CN.sup.[delta].sub.rt] + (1 - [[alpha].sub.rt])[TN.sup.[delta].sub.rt]].sup.1/[delta]] (2)

   Where [CN.sub.rt] and [TN.sub.rt] are new vintaged consumption and passenger transportation at period t in region r, and [UREF.sub.rt] is the reference utility index.

   The budget constraint for region r in period t implies that total macro production must satisfy the competing claims on resources including consumption ([C.sub.rt]), investment ([I.sub.rt]), energy costs ([EC.sub.rt]), transportation costs ([TC.sub.rt]), non-CO2 abatement costs ([AC.sub.r]) and net exports of the composite numeraire good ([NTXY.sub.r]). (3)

   [Y.sub.rt] = [C.sub.rt] + [EC.sub.rt] + [TC.sub.rt] + [AC.sub.rt] + [NTXY.sub.r], (3)

   The energy cost is determined by energy supply technologies described later. Passenger transportation services are provided by vehicles distinguished by 5 alternative technologies: (1) internal combustion engine, (2) plug-in hybrid electric, (3) full electric, (4) compressed natural gas, and (5) backstop (e.g. H2) vehicle.

   There are a limited number of goods that are tradable; macro good, oil, gas and emission permit. Heckscher-Ohlin paradigm is assumed to govern the international trade. This implies that all tradables are homogeneous rather than the region-specific heterogeneous goods usually represented in by Armington (1969) specification. For each tradable good i and each period t, there is a balance-of-trade constraint, i.e., at a global level, net exports or imports for all regions must sum up to zero.

   [summation over (r)][NTX.sub.rt] = 0 (4)

   For an optimization, the regional discounted utilities are weighted by Negishi weights (4). Thus, the objective function is a Negishi weighted global welfare (NWGW),

   NWGW = [summation over (t)][summation over (r)][NWT.sub.r][UDF.sub.rt]log([U.sub.rt]) (5)

   Where [NWT.sub.r] is the Negishi welfare weight and are updated iteratively according to the weights of regional consumption in the global consumption, [UDF.sub.rt] is the utility discount factor.

   The model is solved using sequential optimization of global discounted utility by iteratively updating Negishi weights (Rutherford 1999; Negishi 1972). It can be operated either in "cost-effectiveness" mode or in "cost-benefit" mode depending on the damage value of climate change is taken into account or not. Given the focus of the paper simulations in this paper are performed using the "cost-effective" mode. (5)

   2.2 Energy submodel

   The energy submodel consists of bottom-up representations of various energy supply and transformation technologies along with supply constraints for electric and non-electric energies based on Energy Technology Assessment (ETA) model (Manne 1976; Manne et al. 1995). Electric and non-electric energy supply in this submodel meet all energy demand in the macroeconomic submodel and incur energy cost from exploration, extraction and conversion.

   Levelized costs are used to describe all electric (vintaged and non vintaged) and nonelectric energy technologies whose advancement is assumed to be exogenous. The various electric energy technologies including fossil fuel and clean and/or renewables are shown in Table 1. The choices of these technologies are endogenously determined by the cost-minimization actions of agents with the climate policies taken into consideration. Extracted coal and gas can be used either for generating electric energy or directly used by the industry or transportation. However crude oil needs to be refined before it is used for electricity generation or by industry or transportation. Beside refined oil, there are two other liquid fuel supply technologies; biomass-based liquids and coal-based synthetic liquids as is shown in Table 1. Beside solid, gaseous and liquid fuels, a backstop technology such as H2 is also introduced to provide non-electric energy service to the industry and transportation, as is shown in Table 1.

   In addition to the bottom-up cost configurations of...