Is Abundant Natural Gas a Bridge to a Low-carbon Future or a Dead-end?

• Author: Gillingham, Kenneth
• Position: Report

"[...] natural gas is the Rorschach test of energy policy. [...] it can be either an essential tool for meeting the challenge of climate change or another dirty fossil fuel that will speed the planet down the path to calamitous warming." --New York Times, December 22, 2014 1. INTRODUCTION

In recent decades, there has been a dramatic boom in natural gas production in the United States, spurred by the hydraulic fracturing ("fracking") technological innovation, which opened up vast shale formations to economic recovery of natural gas (Joskow, 2013; Huntington, 2017). This boom has led to low natural gas prices, fuel switching to natural gas for electricity generation (Newell and Raimi, 2014; Holladay and LaRiviere, 2017; EIA 2018). Fuel switching away from coal (and to a lesser extent oil), can reduce emissions from electricity generation due to lower carbon content of the fuel (Brown et al., 2010; McLeod et al., 2014). Indeed, this fuel-switching is responsible for a sizable fraction of the emissions decline in the United States over the past decade (CEA, 2017). However, many have argued that while a large-scale transition to natural gas may reduce emissions in the short-run, it may actually increase emissions in the long-run by leading to lock-in of a low-cost emitting technology (Shearer et al., 2014; McJeon et al., 2014).

This paper investigates the degree to which natural gas can serve as a "bridge" to a low carbon future by helping or hindering reductions in C[O.sub.2] emissions in the coming decades. Our study uses a large-scale energy-economic model of the United States--the National Energy Modeling System (Yale-NEMS)--to assess the extent to which abundant natural gas availability reduces or increases long-run emissions, and welfare, and the effectiveness in reducing emissions relative to a moderate climate policy. (1) While no model is perfect, the NEMS platform is widely accepted as the "gold standard" for prospective energy market analysis (Winebrake and Sakva, 2006). We find that a scenario of truly abundant natural gas does reduce local air pollution and greenhouse gas emissions, while at the same time providing a large welfare benefit. However, the reduction in emissions is modest relative to what a carbon price (e.g., from a carbon tax) set linearly rising to roughly $46 per ton C[O.sub.2] in 2040 could achieve. Furthermore, as we approach 2050, emissions under the abundant natural gas scenario are even slightly higher than the reference case due largely to less deployment of renewables. We further examine heterogeneous welfare effects across regions, illustrating that while all regions see welfare gains from abundant natural gas (regardless of whether there is a carbon policy), there is substantial heterogeneity in the welfare effects. Our welfare estimates are subject to the caveat that we cannot quantify all of the potential local impacts of natural gas production and distribution, but illustrative calculations suggest that even if these are included, the welfare benefits would remain large.

This is not the first paper to tackle the policy-relevant question of whether abundant shale gas increases or decreases net emissions. Indeed, there are several modeling analyses of this question that take different approaches (e.g., Brown et al., 2010; McJeon et al., 2014; Holladay and LaRiviere, 2017; Linn and Muehlenbachs, 2018; Johnsen et al., 2018). Our paper is distinctive by focusing on scenarios with abundant natural gas based on optimistic--but realistic--recent regional estimates of U.S. natural gas resources, consistent with an upper bound natural gas production case that may come about due to richer resources than expected in the reference case (but within uncertainty bounds) and/or an opening up of federal or state lands to natural gas production. Further, we make several new contributions to the literature.

First, we develop a concise economic theory framework that illustrates how adding natural gas could either increase or decrease emissions and elucidates the theoretical links between the outcomes from the computational model, which distinguishes this work from previous modeling papers on natural gas (e.g., Brown and Krupnick, 2010; Brown et al., 2010; Shearer et al., 2014). The static framework we present also clearly lays out the welfare effects of supply increase in natural gas, identifying the economic forces driving welfare. This short theory section crystallizes the basic insights that we then explore using Yale-NEMS, akin to how the theory framework of Gerarden et al. (2016) laid the groundwork for an analogous energy modeling exercise.

Second, we are not aware of any other study that uses the same modeling platform to tackle our research question from both economic and environmental perspectives. Yale-NEMS is an ideal tool for our analysis because of its granularity and comprehensiveness. It is a detailed regional equilibrium model that covers all major U.S. energy markets and end-use demand sectors. It also enables us to quantitatively analyze the impacts of fundamental and policy alterations on markets, welfare, and the environment relative to the baseline projections based on a comprehensive coverage of existing policy and technology status. This detail has led numerous authors to use the NEMS platform for the analysis of changes in the energy system (e.g., Nogee et al., 2007; Goulder, 2010; Brown et al., 2010; Auffhammer and Sanstad, 2011; Mignone et al., 2017).

Third, we are the first to calculate the welfare effects of abundant natural gas both with and without a carbon policy, including monetized C[O.sub.2] and air pollutant impacts. The welfare impacts of the natural gas boom have been of growing interest to economists. Bartik et al. (2016) estimate the local welfare impacts of fracking and conclude that fracking results in substantial oil and gas industry development and growth in the local economy. Hausman and Kellogg (2015) estimate the retrospective consumer and producer surplus effects associated with the recent natural gas supply boom using their own estimated natural gas supply and demand elasticities to find positive gains for consumers but losses for producers. Linn and Muehlenbachs (2018) investigate electricity generation and emissions over the past several years when natural gas prices have been low and econometrically estimate the relationship between fuel switching and electricity prices across regions. Our work differs in quantifying a broader set of welfare effects capturing how multiple inter-related energy markets and sectors respond to changes in market fundamentals and policy, and our modeling includes more detail on emissions than in previous research.

Finally, we examine the winners and losers of an abundant natural gas scenario, accounting for both direct welfare and environmental effects from greenhouse gas emissions and air pollution emissions based on recent air quality modeling work (Muller and Mendelsohn, 2009; Muller et al., 2011; IWG, 2016). We show that the regional impacts from abundant natural gas are quite similar regardless of whether there is a carbon policy or not. In both cases, the region around Texas and the West Coast region benefit the most. But, as mentioned already, nearly all regions benefit from abundant natural gas.

The paper is organized as follows. Section 2 provides background on the U.S. natural gas market. Section 3 presents the stylized economic model. Section 4 introduces Yale-NEMS and our scenario designs. Section 5 presents our primary quantitative simulation results, while section 6 examines the welfare implications. Section 7 concludes.

2. BRIEF BACKGROUND ON NATURAL GAS IN THE UNITED STATES

   Between 2001 and 2016 (the most recent estimate available from EIA), total natural gas production in the United States increased 35% to 27 Tcf (Figure 1). This increase can be attributed to a boom in shale gas production, which exceeded 53% of total production in 2016. Most of this new production is from six major shale plays: Barnett, Eagle Ford, Fayetteville, Haynesville, Marcellus, and Woodford spread in Texas and Pennsylvania (Joskow, 2013; Hausman and Kellogg, 2015; Cooper et al., 2016). As is also seen in Figure 1, the increase in production has been accompanied by a decrease in natural gas prices, with the Henry Hub trading price dropping by 75% between 2008 and 2016, reaching $2.5 per MMBtu (in 2016$). (2)

   Lower natural gas prices provide benefits to consumers and firms, with positive direct economic impacts (Mason et al., 2015). As natural gas has a lower carbon intensity than coal and emits fewer air pollutants when combusted, a switch from coal to natural gas for electricity generation can also have direct environmental benefits. Indeed, total U.S. greenhouse gas emissions have been dropping, and this has been determined to be partly attributable to fuel switching from coal to natural gas (CEA, 2017). In April 2015, for the first time ever in the U.S., natural gas replaced coal as the largest fuel source of power generation. (3) At the same time, the increased domestic natural gas production has also changed the U.S. from a natural gas importer to a natural gas exporter (EIA 2017a). As of April 2016, U.S. sent its first liquified natural gas (LNG) cargo to Europe from the Sabine Pass LNG terminal. (4) As production continues growing, LNG exports from the U.S. are expected to play a pivotal role in the international energy market (IEA 2016).

   This paper explores an upper bound estimate of the quantity of low-cost abundant natural gas--a world where some of the most optimistic assumptions on the extension of the 'shale gas revolution' come true. The next section develops a simple theory framework for understanding what abundant low-cost natural gas can mean for emissions and welfare.

3. A THEORETICAL MODEL

   To operationalize a world with substantially more abundant natural gas, a natural starting place is to...