A Clean Energy Standard Analysis with the US-REGEN Model.

• Author: Blanford, Geoffrey J.

1. INTRODUCTION

   Since the defeat in the U.S. Senate in 2010 of the American Clean Energy and Security Act, which included an economy-wide cap on carbon emissions and had been approved by the House of Representatives, the focus of U.S. climate policy has shifted to more regulatory and sectoral approaches. An alternative that appears to be gaining political traction is a Clean Energy Standard (CES) applied to the electric sector. In the 2011 State of the Union Address, President Obama called for 80% of United States electricity to be generated from clean sources by 2035 (1,2) a call he reiterated in the 2012 address. Later that year, a CES bill was introduced by the Senate Energy and Natural Resources Committee Chairman Jeff Bingaman, although its prospects for passage remain unclear at the time of writing.

   Under a CES policy, a portfolio constraint is enforced on electric generation in which a certain percentage of consumed energy must be generated by qualified sources. The percentage would increase over time, and the definition of qualified sources would be broader than in other similar policies, such as a renewable portfolio standard, possibly including natural gas, nuclear, and coal with carbon capture and storage (CCS). Additionally, qualified sources could be weighted, for example to reflect or approximate the carbon emissions intensity of various generation technologies. (3) The implementation would likely be market-based with a clean energy certificate (or fraction thereof for weights less than one) awarded to qualified generators and a compliance obligation on the part of load-serving entities to acquire certificates equal to the target percentage of delivered energy. A key distinction between a CES and a market-based emissions policy is that no implicit or explicit public sector transfer takes place, e.g. in the form of permit allocation or tax revenue. There is also the potential for large inter-regional transfers as a result of the policy given the geographic diversity of renewable resources.

   The idea of a CES is relatively recent. The earliest mention of clean energy credits appears to be in the work of Michel and Nielsen (Electricity Journal, 2008). They proposed trading in CO2 reduction credits for electricity generation, arguing that this approach has advantages in terms of administrative and distributional efficiency over the cap and trade alternative. Their paper was descriptive in nature with no analysis to measure the claimed efficiencies. Indeed, to our knowledge, no analytical modeling of the Clean Energy Standard had been conducted prior to its mention in the 2011 State of the Union address. A proposal by Aldy (2011), following the President's endorsement of the approach, argued that a CES "represents a simple, transparent, more cost-effective, and more effective alternative to greenhouse gas regulatory authority under the Clean Air Act and the patchwork of state renewable and alternative energy portfolio standards." [p48] At the state level, Coffman et al. (Energy Policy, 2011) draw some comparisons between a CES and an RPS in the specific case of Hawaii, using a dynamic optimization model of Hawaii's electric sector. They find that a CES policy where the weights on technology are determined by GHG emissions can reduce by up to 90% the cost of lowering emissions by an amount equivalent to Hawaii's current RPS scheme, due to the greater range of abatement options offered by the CES.

   In response to growing interest in the CES, a number of modeling studies at the national level have emerged, including Paul et al. (2012), Rausch and Mowers (2012), and Mignone et al. (2012). These papers all agree on the broad consequences of a CES to the electric sector, namely, a shift away from fossil fuels to nuclear and wind, and revenue transfers from fossil heavy states to those states with large renewable resources. Mignone et al. and Rausch and Mowers use models with macroeconomic modules, and thus can also show the impact on GDP and welfare. Mignone et al. find the impact on welfare to between $287B and $355B (cumulative NPV in 2009 $) through 2035; Rausch and Mowers using the USREP integrated model additionally find the policy hits welfare harder in the lower income brackets. Finally the Energy Information Administration has conducted an analysis of the Bingaman proposal using the NEMS model (EIA, 2012). In their reference case, they concluded there would be little impact on electricity prices until 2020, and that nuclear generation would dominate new generation as the standard tightened. There was a modest impact on GDP of less than 0.1% by 2050. Their case found lower interregional transfers and low use of Alternative Compliance Payments, due to the dominance of nuclear, and due to restrictions on credits from legacy generation (discussed further below).

   The Energy Modeling Forum (EMF) 24 study, in which multiple energy-economy models were asked to run a set of coordinated scenarios for future U.S. climate policy, included a CES alongside both economy-wide emissions caps and more specific regulatory approaches. In this paper, we follow the EMF 24 design using the US-REGEN model to explore the economic and environmental outcomes of a CES in relation to other options, with an emphasis on the role of technology. We find that alternative assumptions about technology can completely change the optimal compliance strategy and economic impact at the national and especially the regional level. The unique features of the model's design are applied to yield insights into the nature of a CES approach and how it differs from previously studied market-based carbon policies.

2. MODEL

   Overview

   Our analysis employs the US-REGEN model, an inter-temporal optimization model of the US economy through 2050 that combines a detailed dispatch and capacity expansion model of the electric sector with a dynamic computable general equilibrium (CGE) model of the rest of the economy. The model emphasizes details in the energy production sectors and different end-uses. Both the electric and CGE models are disaggregated into 15 state-based regions, and the two components are solved iteratively to convergence. The model has been developed over the last two years at the Electric Power Research Institute (EPRI). Because the model has not previously been presented in the literature, we take some time here to describe its key features. Further detail can be found in the US-REGEN Model Documentation (EPRI, 2013).

   The electric sector component is formulated as a linear process model with a bottom-up representation of power generation capacity and dispatch across a range of intra-annual load segments. In each time step, the model makes decisions about existing capacity (carry forward, retrofit, or retire) and investments in new capacity both for generation and inter-region transmission, as well as dispatch decisions for installed capacity in each load segment. A discount rate of 5% is applied. Individual existing generators in each region are aggregated into larger capacity blocks based on similar operating characteristics. The block is dispatched as a single unit, but the age profile of the underlying individual units is preserved. Several unique features of the electric sector make the explicit treatment of capacity vs. dispatch essential to accurately model decision-making and the impact of new policies. First, the "shape" or hourly profile of end-use demand and variable resource availability is crucial for appropriately characterizing the operational patterns and profitability or value of different types of capacity. Second, these patterns and hence the value of generating assets are also dependent on the mix of installed capacity in a region (and in neighboring regions). Third, capital investments in generating capacity tend to be long-lived, creating a strong link between dispatch and investment decisions across time periods.

   The CGE component of the model is formulated in the classical Arrow-Debreu general equilibrium framework, which describes the supply of factor inputs (labor, capital, and resources) owned by households to the producing sectors of the economy, and the supply of goods and services from these sectors back to households. US-REGEN has been designed with particular detail in the energy sectors and energy flows throughout the economy, with a high level of aggregation elsewhere. Non-energy production is described by an industrial sector, a commercial services sector, and a transportation sector. Household consumption is described by a residential sector with a single representative household. For each sectoral activity, a constant elasticity of substitution production function defines how inputs can be translated into outputs, including the structure of substitution opportunities. There is inter-regional domestic trade in industrial goods and commercial services, and foreign trade in all commodities. Capital stocks accumulate as a function of endogenous investment. Each region's welfare, representing its households' utility, is a function of residential consumption over time.

   The representation of the operational details of the electric sector allows a high-fidelity treatment of the trade-offs among candidate technologies under a policy scenario such as a CES, while the integration with the CGE representation of the economy allows a comprehensive analysis of feedbacks between sectors and overall cost effectiveness. The following subsections provide additional detail into a few specific elements of the model formulation. For the electric sector, we describe the treatment of intermittent renewable resources and the distinction between cost-of-service regulation and competitive electricity markets at the regional level. For the macro model, we describe the unique formulation of end-use specific energy-using capital inputs. Additional details, including regional definitions, cost and performance assumptions...