Cap-and-Trade Programs and Innovation for Climate Safety

If sources can reduce emissions cheaply, they can then try to sell excess allowances at whatever price the market will bear; in a number of CTPs, they can also “bank” these allowances fo

Cap-and-Trade Programs and Innovation for Climate Safety

Margaret Taylor1

1Richard and Rhoda Goldman School of Public Policy

University of California, Berkeley

2607 Hearst Avenue

Berkeley, CA USA 94720-7320

tel: 011-510-642-1048

fax: 011-510-643-9657

Abstract [Needs revision]

Analysts generally agree that considerable technological innovation will be necessary to reduce greenhouse gas (GHG) emissions to “safe” levels while minimizing economic impacts Market failures related to both environmental pollution and innovation reduce the likelihood that the private sector will provide that innovation without public

intervention Meanwhile, cap-and-trade programs (CTPs) for GHG reductions are

rapidly becoming the world’s dominant climate policy instrument This paper assesses the innovation effects of the three most prominent CTPs in existence that have lengthy-enough operations for evaluation and strong similarities to climate CTPs It shows that ineach CTP, lower-than-expected pollution prices emerged that led to smaller-than-

expected markets for a wide range of emissions reduction technologies Further, in two

of the three CTPs, significant cancellations of technology orders already in process compounded the reduced market expectations for these technologies during CTP

operations In addition, the paper shows that dramatic declines occurred in patenting activity – the most widely used indicator of the levels of inventive effort involved in developing technologies for later sale – in all of the identified technologies when CTPs were operating, as compared to periods of time that were dominated by more traditional environmental regulation The paper concludes by raising concerns about whether CTPs will be able to induce the levels of pre-commercial inventive activity necessary to

achieve climate safety without careful policy design and complementary policy efforts

Classification Codes and Keywords

Keywords: Environmental Policy, Innovation, Emissions Trading, Climate Change

JEL codes: Q54, Q55, Q58

1 Introduction

Analysts generally agree that the process of reducing greenhouse gas (GHG) emissions to

“safe” levels, while minimizing economic impacts, will require considerable

technological innovation.1 Large portions of global GHGs are emitted by key sectors of the economy; for example, electric power (24% of global emissions), transportation (14%), industry (14%), and agriculture (14%), when combined, contribute 66% of global emissions [1] In comparison, “safe” GHG levels have been set at 50-80% below 1990 total emissions by 2050 in recent initiatives by the European Union, Canada, Japan, and

1 This paper will distinguish between innovation as a process and its component activities, defined here to

include the invention and commercialization of new products and processes, as well as the initial adoption then diffusion of new technologies throughout the economy.

California These are very ambitious goals, both in terms of the absolute GHG levels required and the speed at which those levels need to be reached, particularly when one considers the long operating life of many major individual emissions sources, such as power plants But the specter that even these ambitious targets will be inadequate to ensure climate safety is being raised by the latest findings of an accelerating growth rate

of atmospheric carbon dioxide (CO2), faster-than-predicted ice melts, and growth in China's CO2 emissions that is outpacing previous estimates [2; 3; 4]

Market failures related to both environmental pollution and innovation decrease the likelihood that the private sector will provide the necessary levels of “climate-safe” innovation without public intervention.2 A critical question, therefore, is which policy approaches will best serve to foster that innovation This question is largely unanswered

by empirical scholarship on environmental innovation, however, despite more than thirty years of renewable energy, energy efficiency, and environmental policy experience to draw on

In the meantime, climate policy is rapidly evolving, and cap-and-trade programs (CTPs) for GHG mitigation are becoming the world’s dominant climate policy instrument, with the European Union (EU), Australia, over half of both the U.S States and Canadian Provinces, and one Mexican State either operating or developing programs.3 The primaryeconomic case for the use of CTPs is one of static efficiency; previous CTPs have

demonstrated that the instrument is capable of facilitating pollution reductions to meet relatively short-run caps at low cost in cases in which there are available technological options But there is another important factor driving the emerging dominance of CTPs

in climate policy: the claim that CTPs are better than other policy instruments in

providing an “incentive for innovation” [e.g 8].4

2 Although the literature shows the primacy of the private sector as a source of

innovation [5], it also shows that the private sector under-invests in research and development (R&D) when compared to the societal returns of that R&D [e.g 6; 7] This is compounded in the case of technologies that either control or prevent

pollutant emissions (“environmental technologies”) by the fact that they maintain the

“public good” of a clean environment Public goods are typically characterized by weak market incentives for private investment and development Different

environmental technologies reveal different combinations of public and private value For example, a pollution control device for a power plant does not create an

economically valued good in and of itself unless the negative externality of pollution

is somehow internalized by the power plant Similarly, the market that alternative energy technologies satisfy is shaped by a combination of the privately valued and publicly valued characteristics of the energy they provide; such privately valued characteristics include cost, availability, and other performance attributes of energy, while their publicly valued characteristic is their impact on the environment.

3 In a CTP, policy-makers set a cap on emissions and then allocate emissions “allowances” to polluting sources that are equivalent, in sum, to the cap If sources can reduce emissions cheaply, they can then try to sell excess allowances at whatever price the market will bear; in a number of CTPs, they can also “bank” these allowances for later use.

4 Such an incentive is environmentally important, as noted above, but it is also politically salient, as it raises the possibility that acting to combat climate change could also provide innovative spillovers that are economically beneficial to a jurisdiction.

This “dynamic efficiency” claim stems from the conclusions of theoretical environmentaleconomics studies, dating back to [9], that compare and rank such instruments as taxes, subsidies, CTPs, and traditional environmental regulation regarding their incentives for innovation The majority of these studies in the 1970s-90s supported the dominance of CTPs above other policy instruments on dynamic efficiency grounds [e.g., 9; 10; 11; 12; 13] Another set of studies, however, that is for the most part more recent than these consensus studies, have portrayed a more ambiguous situation [e.g., 14; 15; 16; 17; 18; 19; 20] As reviewed in [21], the majority of more recent authors “support the view that grandfathered permits [the dominant allowance allocation approach in CTPs] provide lower incentives [for innovation] than emission taxes and also question the notion that market-based instruments, specifically emission trading, are generally superior to direct regulation.” Significantly, [15] states that there is “no unambiguous case for preferring any of these policy instruments,” because assumptions about such things as innovation costs, appropriability concerns, the shape of environmental benefit functions, and market structure are critical to the outcomes of the models.5

In light of the political impact of the earlier literature and what seems to be a dissolving consensus about its conclusions, a brief overview is in order The dominant modeling approach analyzes the incentives of a polluting firm facing a binary choice between “its existing technology and the possibility of one single (exogenously given) new

technology” [18] In light of this choice, studies typically consider firm incentives for

“innovation” in pollution control under different policy scenarios, where innovation is defined to represent both the invention and the commercialization of a new product or process Most studies consider diffusion as well, either as an assumption or as a variable;

as pointed out in [19], the assumption of complete diffusion of the new technology acrossall the firms in an industry [e.g., 10; 11] can be critical to modeling outcomes In most ofthe studies, innovation incentives are determined by accounting for “innovator” rewards and costs Following [11], rewards are attributed to three sources: (1) savings regarding the direct cost of abatement (examples are equipment expenses and operating costs); (2) savings related to transfer losses associated with abatement (i.e., payments made by the firm, such as emission taxes); and (3) gains related to payments made to the firm

(examples include emission subsidies and patent royalties) Costs are the funds necessary

to develop and implement the technology, which are termed “R&D expenditures” [18]

A number of concerns have been raised about the validity of the dominant modeling approach and its assumptions, some of which are highlighted here First, both [18] and[22] point out that the representation of so-called “command and control” regulation for comparative purposes is inadequate, given the greater use in environmental law of such performance standards as limits on emissions per unit output or input Second, several studies raise issues about the potential disincentives for innovation that may arise from the dynamic nature of pollution prices in a CTP [e.g., 15; 16; 19; 20] They point out thatallowance prices in a CTP are likely to drop when marginal abatement costs fall with technology adoption by a subset of early-mover polluting firms, thereby reducing the incentives for later firms to similarly adopt new technologies Third, some studies focus

on the modeling treatment of polluting firms under CTPs, which typically does not

5 [16] similarly finds that assumptions about perfect competition and information are also critical.

differentiate between sellers of allowances (i.e., polluters who emit less pollution than their allowance allocations) and buyers of allowances (i.e., polluters who emit more pollution than their allowance allocations) The argument here is that the net incentives for innovation may be ambiguous under a CTP because, although the instrument may incentivize sellers to make more changes to their production processes than other

instruments, it may also incentivize buyers to make fewer changes [17; 22].6 Fourth, a few studies and one review of the literature focus on the situation in which non-polluting third-party firms – rather than polluting firms – invent and commercialize new

technologies of relevance to the environmental focus of a CTP [11; 15; 18; 23] In this situation, the rewards for innovation stem from technology sales to polluting firms rather than from abatement cost savings or revenues from allowance sales, and therefore turn the traditional accounting in models on its head.7

What does the empirical literature say about the validity of these concerns?

Unfortunately, as reviews [e.g 5; 24] have pointed out, there is a dearth of empirical studies on CTPs and innovation The focus of the few studies that exist is on individual CTPs [e.g 25; 26; 27; 28; 29], rather than on bringing our collective experience operatingCTPs over the last two decades to bear on understanding the innovative conditions defined by these policy instruments more generally As such, the empirical literature has not been as useful to the policy debate as the theoretical literature

This paper aims to rectify this situation, to some extent The first part of the paper

focuses empirically on the question of whether the major innovation sources of new products and processes related to existing CTPs are typically the polluting firms that are allotted allowances, or non-polluting third parties As predicted in [23], the evidence supports the idea that “innovations in pollution control are often (if not mostly) supplied

by special outside suppliers.” This condition also appears to hold for five of the major technologies of relevance to GHG mitigation from the electric power sector.8 The secondpart of the paper considers the implications of the distinction between the innovators and the adopters of environmental technologies in an empirical treatment of the innovation dynamics under the three most prominent CTPs in existence that have lengthy-enough operations for evaluation and strong similarities to climate CTPs This part of the paper shows that in each CTP, lower-than-expected allowance prices emerged that led to

smaller-than-expected markets for a wide range of emissions reduction technologies Further, in two of the three CTPs, significant cancellations of technology orders already

in process compounded the reduced market expectations for these technologies during CTP operations In addition, dramatic declines occurred in patenting activity in all of the identified technologies when CTPs were operating, as compared to periods of time that

6 This is because a CTP can make inexpensive compliance options (like allowance purchases) available that would not otherwise have been possible under traditional environmental policy instruments.

7 [15] suggests that there may be one additional source of revenue in a CTP for a non-polluting firm, however The authors explain that “markets might offer the innovator an opportunity to capture more of the industry gains from emissions price changes by shorting permits By contracting to sell permits in the second period at pre-innovation prices, the innovator gains from the post-innovation price fall on those promised permits.” The authors do not, however, find it “credible that an innovator could short the entire permit market and capture all the industry gains.”

8 This sector is prominently featured in most climate CTP proposals.

were dominated by more traditional environmental regulation The paper concludes by raising concerns about whether CTPs will be able to induce the levels of pre-commercial inventive activity necessary to achieve climate safety without careful policy design and possibly complementary policy efforts.

2 Technology Innovators and Adopters

This part of the paper focuses empirically on answering the question of whether the majorinnovation sources of new products and processes related to CTPs are typically the polluting firms that are allotted allowances, or non-polluting third parties Although this question has received very little attention in the literature, so far, the answer to it is fundamental to any understanding of the competitive dynamics of innovation under CTPs

If the innovators are distinct from the polluters, there will be additional uncertainties introduced into the innovation process under a CTP system than under either emissions taxes or traditional environmental regulation Any innovator has to cope with R&D investment decisions that are long-term and have uncertain technical outcomes, of course.But because the innovator rewards to non-polluting third-party firms stem from

technology sales to polluting firms, it is easier for a third-party firm to predict total rewards under conditions of fixed emissions prices or fixed emissions quantities than it is under the changing allowance price situation that occurs under a CTP This is because the polluting firm “potential customers” of a new technology can choose a less

predictable array of options under a CTP, including allowance purchases either alone or

in combination with lower cost, less effective technologies that might not have been considered competitors to an innovator under a different policy regime If the third-party

is the innovator, its investments in R&D under a CTP will necessarily be based in large part on allowance price expectations, and the portfolio of technological pathways that these innovators choose to follow will probably need to be justified internally by

potential payoffs that incorporate premiums for allowance price uncertainty

As mentioned above, a few theoretical economic studies and one review of the literature focus on the situation in which non-polluting third-party firms – rather than polluting firms – invent and commercialize new technologies of relevance to the environmental focus of a CTP [11; 15; 18; 23] The prevalence of this situation is not really touched upon, however, except in [23]

Meanwhile, [30] empirically investigates the composition of R&D expenditures in the electric utility industry, which has been a major target of the three CTPs in existence withlong enough operations for evaluation and strong similarities to climate CTPs, as well as all proposed and operating climate CTPs Using data from the Federal Energy

Regulatory Commission (FERC) and Energy Information Administration (EIA), this study finds that “most of the environmental research in pollution abatement technologies was conducted by electric equipmentmanufacturers such as Babcock and Wilcox and not by utilities.”[30] Electric utilities, by contrast, “conducted very little pollution abatementresearch—rather the bulk of abatement expenditure was concentrated

on compliance issues and is thus not considered R&D.” In other words,according to the R&D expenditure data considered in [30], non-

polluting third-party firms are the primary sources of the innovations that are most relevant to resolving the pollution issues targeted by CTPs

But many R&D programs do not result in commercialized innovations

To get a better sense of innovative activity at the intermediary step between invention and commercialization, it is helpful to turn to patentdata Patents are required by law to publicly reveal the details of a completed inventionthat meets thresholds of novelty, usefulness, and non-obviousness Studies have shown that patenting activity parallels R&D expenditures, which are often difficult to find at a disaggregated enough level for research purposes, and can also be linked to events that occur outside the firm Surveys [31; 32; 33] demonstrate that 40–60% of the innovations detailed in patent applications are eventually used by firms This indicates that patents are probably best thought of as a well-accepted intermediary outcome of inventive

activity, one that is tied both to the input of R&D expenditures and to hopes of

commercialization See [34] for a review of the use of patent statistics as economic indicators, including some of their strengths and weaknesses

This section focuses on identifying the “innovators” – as opposed to the “adopters” – of environmental technologies of relevance to CTPs, as revealed in published datasets of patents in the U.S Patent and Trademark Office (USPTO) system Before turning to the patent results, it briefly provides background information on the main technology

strategies for combating the emissions of sulfur dioxide (SO2) and nitrogen oxides (NOx) – the pollutants that are the subject of the most prominent existing CTPs – as well as on five of the major technologies of relevance to GHG emissions from the electric utility sector Note that the appendix contains tables of the search terms used to put together thepatent data, as well as details on the coding approach used to identify innovators; the datafrom these searches were first published in [28; 35; 36; 37].9

1 Emissions Reduction Technologies for Existing CTPs

The emissions to the air of sulfur dioxide (SO2) and nitrogen oxides (NOx) are the

primary focus of the existing CTPs that are most analogous to climate CTPs As these CTPs operate in the U.S., the technical descriptions here focus on the U.S., although many of the relevant technologies were developed in an international environmental regulatory context [for more on the international context, see 28; 39; 40]

Several technology strategies can reduce sulfur dioxide (SO2) emissions, the primary U.S.source of which has been coal-fired power plants since 1960 [see 41; 42].10 First, power

9 Only one patent dataset included in this paper relies on a non-reviewed dataset This dataset, however, was used in [38].

10 SO 2 is of concern for several reasons It is an eye, nose, and throat irritant, which in the extreme case has contributed to such infamous air pollution incidents as the “killer smogs” in Donora, Pennsylvania, in 1948 and London, England, in 1952 [43; 44] SO 2 is also a significant secondary chemical component of the emerging public health issue of “ultra-fine” particles less than one micron in size, which can deposit deeply into the lungs and reside there up to several months [45] Finally, SO 2 emissions are a major contributor to

plants can burn naturally lower-sulfur coals, although there are tradeoffs between the energy per unit mass – the heat content, or “heat input” to generators – of these coals and their sulfur content The delivered price of coal and the design of U.S boilers for specificcoals have historically been constraints on the widespread use of this approach Other strategies that the electric power industry has pursued have included: (1) tall gas stacks that disperse emissions away from immediate areas but have acid rain tradeoffs; (2)

“intermittent controls,” or operational adjustments that are used to reduce emissions in response to atmospheric conditions; (3) pre-combustion reduction of sulfur from fuels;11

and (4) removal of SO2 from the post-combustion gas stream (otherwise known as “flue gas desulfurization” (FGD) systems or “scrubbing” technologies).12 Since the 1960s, the focus has shifted away from tall stacks and intermittent controls, toward pre-combustion and post-combustion treatment technologies

Electricity generation is also the largest stationary source of NOx emissions in the U.S., where it accounted for about 22% of overall emissions in 2002 [46].13 Of the various environmental problems associated with NOx, either singly or in conjunction with other pollutants, CTPs have focused on the role of NOx in helping to constitute tropospheric (ground-level) ozone (O3, commonly known as “smog”).14

NOx control strategies can generally be divided into two categories: (1) “combustion modification” processes that reduce the production of NOx emissions within the power plant; and (2) higher-cost and more effective “post-combustion” processes that decrease the NOx emitted by the power plant after combustion [see 28 for more information] Combustion modifications (also known as “primary measures”) for NOx control include burner optimization, air staging, flue gas recirculation, fuel staging, and low-NOx burners.They generally require relatively little capital investment, do not entail the use of

chemical additives or reagents, and have typical NOx reduction capabilities of 30-60% Post-combustion processes (or “flue gas treatment”), on the other hand, use reagents, either via selective non-catalytic reduction (SNCR) or via selective catalytic reduction (SCR) technologies, to reduce the NOx in the flue gas downstream of the power plant furnace Typical NOx reduction capabilities are 30-50% for SNCR and 70-90% for SCR

acidic deposition (or “acid rain”), with resulting damage to lakes, streams, plants, and forest growth

11 Pre-combustion technologies use physical, chemical or biological processes to “clean” coals; in

commercial operation, they typically remove less than 30% of the sulfur.

12 FGD systems contact a post-combustion gas stream with a base reagent (or “sorbent”) in an absorber in order to remove SO 2 Although there are several system types (see figure 4 in [27] for a full typology), the two main options are wet once-through processes that use limestone as the scrubbing reagent (about 72% of world capacity; forced oxidation systems make possible 95%+ SO 2 removal efficiencies) and the cheaper lime spray drying process (about 8% of world capacity, capable of 80-90% SO 2 removal efficiencies) Note that the costs of both wet and dry systems are higher in “retrofit” application to “existing” power plants, as opposed to “new” application to new power plants.

13 Transportation is the main U.S emissions source, accounting for about 54% of emissions in 2002 [46].

14 NO x is an eye, nose, and throat irritant, a key constituent of acid rain, a contributor to the greenhouse effect (through both the indirect radiative forcing of ground-level O 3 and the actions of one NO x , nitrous oxide (N 2 O)), and a significant secondary chemical component of ultra-fine particles In combination with sunlight and volatile organic compounds (VOCs), NO x is a key constituent of tropospheric ozone, one of the six criteria pollutants NAAQSs were established for under the 1970 CAA.

systems Post-combustion NOx control technologies are more costly than combustion modifications, but both are more expensive in “retrofit” application to “existing” power plants, as opposed to “new” application to new power plants.

2 Technology Strategies for GHG Reductions from the Electric Power Sector

There are two basic technology strategies that can be used to reduce the CO2 emissions – the most prominent (80%+) GHG implicated in climate change – from fossil fuel

combustion In the first strategy, the focus is on retaining the existing fossil-fuel-fired

combustion process (and upholding the investments that went into that process) while controlling emissions This can be done through pre-combustion and post-combustion interventions or reducing demand for generation In the context of electricity generation, considered to be 24% of global GHG emissions (mostly from CO2) [1], a pre-combustion intervention might involve fuel-switching from coal to natural gas, while a post-

combustion intervention might involve carbon capture and sequestration (CCS).15 Meanwhile, reducing the demand for electricity can be achieved through measures such

as encouraging greater efficiency in end-use devices or by meeting some of the demand for power using end-use devices powered by alternatives to fossil-fuel fired generation

An exemplar technology is domestic solar water heating (SWH), which was reportedly used in 2.5% of households worldwide by the end of 2004 [47].16

In the second main strategy, the focus is on a more significant shift away from

fossil-fuel-fired generation and to such generation alternatives as water, wind, sun, and nuclear

power Three exemplar technologies are photovoltaic (PV) cells, solar thermal electric generation (STE), and wind power (Wind) In the generation of electricity, the most prominent solar energy technologies use either the photoelectric effect, as in the case of photovoltaic (PV) cells, or convert solar radiation to heat that then generates power through such mechanical means as driving a Stirling engine, as in the case of solar

thermal electric (STE) power Wind power, on the other hand, converts the kinetic energy in wind into mechanical energy that is then converted to electricity

3 Patent Ownership in CTP-Relevant Technologies

Fig 1 differentiates the patent owners in the seven technologies described above, as clustered by type of technology and type of organization The pollution control patents cover SO2 (pre-combustion technology combined with post-combustion technologies),

NOx (combustion modification technologies combined with post-combustion

technologies), and CCS technologies Alternative generation patents cover Wind, STE, and PV technologies And energy conservation patents cover SWH technologies Fig 1 clearly shows that the most prevalent innovators in these pollution control, alternative

15 This latter, emerging technology, involves the separation and capture of CO 2 from the flue gas stream of electric power plants and other industrial processes, which then needs to be managed either by injection into geologic formations (e.g., deep saline reservoirs, depleted oil and gas wells, unmineable coal seams) or

in other repositories including (potentially) the world’s oceans Note that the capture aspect of CCS is analogous to FGD and SCR.

16 SWH raises the temperature of a circulating working fluid – sometimes potable water – by exposing it to solar radiation In most cases, SWH systems work as hybrid systems in conjunction with a supplemental natural gas-powered or electric heater.

generation, and energy conservation technologies are not the polluting firms that are the subjects of CTPs, but third-party non-polluting firms and individuals.17

Fig 1 Patents in CTP-relevant technologies, by type of technology, as broken down by the

percentages owned by various types of innovative organizations.

3 Existing CTPs and the Dynamics of Innovation

This part of the paper considers the implications of the distinction between the innovatorsand the adopters of environmental technologies in an empirical treatment of the

innovation dynamics under the three most prominent CTPs in existence that have

lengthy-enough operations for evaluation and strong similarities to climate CTPs Most

of the short history of CTPs is concentrated in the U.S., where a handful of programs have either substituted for or supplemented the pre-existing regulation of traditional air pollutants, most significantly sulfur dioxide (SO2) and nitrogen oxides (NOx) The three main CTPs in operation since the 1990s vary by pollutant and by governance level: “TitleIV” is a national CTP for SO2 emissions; the Ozone Transport Commission/NOx Budget Program (“OTC/NBP”) is a seasonal and regional CTP for NOx emissions; and the Regional Clean Air Incentives Market (“RECLAIM”) is a southern California CTP for both NOx and SO2 emissions (the NOx program is more prominent and is discussed here) These CTPs cover similar emissions sources: in Title IV and the OTC/NBP, coal-fired power plants are the primary emitters that need to adopt technologies and other

compliance strategies, while in RECLAIM, gas-fired power plants are an important, although not primary, emissions source

Table 1 provides an overview of these CTPs according to their major design elements The table also provides general information on observed market behavior during the operation of these CTPs The next section goes into this material in greater detail

Table 1

Overview of U.S CTPs

17 Patent citation analysis further shows that the patents held by non-CTP targets are at least as important as those of potential CTP targets See appendix for this data

Implementation of Cap

Treatment of Banking

Allowance Price Behavior

Market Depth Title IV Two phases Unlimited Lower than initially

expected, with one large price spike.

The initial firm reaction was autarkic (i.e to respond independently and perform only limited trading) [48] Today, Title IV is

considered liquid.

OTC/NBP Multi-phase Restricted Lower than initially

expected, with two large price spikes.

The initial firm reaction was autarkic [49] Today, the NBP is considered liquid.

RECLAIM Annual reduction None Lower than initially

expected, with one large price spike.

existing stationary sources and submit it for EPA approval SIPS were submitted in 1972,

and almost all called for continuous reduction of SO2 emissions, which in effect gave utilities the opportunity to use low sulfur fuels, pre-combustion treatment, or FGD

systems to comply with the standards, rather than tall stacks or intermittent controls

Meanwhile, major new sources (or significantly modified existing sources) of SO2 were

to be subject to New Source Performance Standards (NSPS) based on the agency’s determination of whether relevant SO2 control technologies were adequately

demonstrated for commercial use In the case of SO2 control, the EPA determined that scrubber technologies developed in Japan were demonstrated enough to provide the technology basis for standard-setting The 1971 NSPS set a maximum allowable

emission rate of 1.2 lbs of SO2/MBtu heat input (2.2 kg/Gcal), a rate that effectively required 0-85% SO2 removal, depending on coal properties This standard was

technologically flexible, as it could be met through the use of low sulfur fuels,

pre-combustion treatment, and FGD systems The 1979 NSPS for SO2, however, required a 70% reduction of potential SO2 emissions from generation based on low sulfur coal and a90% reduction of potential SO2 emissions from high sulfur coal This was not

technologically flexible, as it essentially required that any new power plant operate a dry

or wet scrubber, respectively, no matter the sulfur content of the fuel Note that existing sources were not subject to this requirement See [41]

More than 70 bills were unsuccessfully introduced in Congress to reduce SO2 emissions from power plants after the 1979 NSPS before the passage of the 1990 CAA, which introduced the national CTP for SO2 control in Title IV [51] One of the most important successes of Title IV was its ability to overcome the political logjam that had arisen on

SO2 emissions control in those years [40]

4.2 NO x

As in SO2 regulation, the U.S first began to regulate NOx-relevant emissions in the 1970 CAA, which identified NO2 and O3 as two criteria pollutants for which it set NAAQS The role of NOx in ozone formation was not recognized until the mid-1980s, however, so most of the pre-CTP policy experience with NOx emissions relates to reducing NO2

emissions For the NAAQS NOx-relevant pollutants, the existing vs new source

dichotomy and NSPS revision timeline applied as it did in the SO2 case, at least in the 1970s and 1980s

Unlike its stance in the SO2 NSPS development process, the EPA was less willing to accept foreign demonstration of control technologies in the NO2 NSPS-setting process, with the result a less stringent NSPS for NO2 than SO2.18 It took until 1998, after SCR, the dominant post-combustion control process, had been installed in almost 70 GWe of coal-fired capacity worldwide, that SCR was finally considered to be adequately

demonstrated enough for the U.S to allow it to serve as the NSPS technology basis [36]

In that year, the federal NSPS was revised for utility boilers, requiring reductions on the order of 80% or more from new and modified sources

Starting in the late 1970s, however, experts at the South Coast Air Quality Management District (SCAQMD) of California began to hold a different view on Japanese experience with SCR than the federal government, and found its high performance persuasive

enough to provide the basis of power plant rules despite its expense.19 Related rule efforts

in SCAQMD culminated in the passage in 1989 of Rule 1135, which set a stringent emissions limit for utility boilers that required approximately 90% reductions in NOx

emissions from gas-fired generating units by 1997 in Southern California This limit could only be met by a combination of combustion modification and post-combustion control technologies Shortly thereafter, SCAQMD established a CTP that replaced this rule

18 The EPA established NAAQS for criteria pollutants, required states to submit SIPs for controlling emissions from existing sources, and created NSPSs based on the agency’s determination of whether relevant control technologies were adequately demonstrated for commercial use In December, 1971, the NSPS for NO 2 was published, based, in part, on findings of an earlier report that noted that although many primary NO x controls had been proven commercially, mostly in California, SCR was a “speculative” control technique and should not be the technical basis of the NSPS ([52]) The 1971 NSPS set a limit of 0.7 pounds of NO 2 per million Btu (lbs/MBtu) heat input for coal-fired units and 0.2 lbs/MBtu for gas-fired units In 1979, the NSPS was revised, and the limits shifted to 0.5 lbs/MBtu (bituminous coal) and 0.6 lbs/MBtu (sub-bituminous coal), with the NSPS for gas-fired units unchanged This was not as strict a standard as was being met in Japan at the time; the successful contemporaneous Japanese SCR application

to a coal-fired plant was not considered to be on a large enough unit to be a “proven” technology for the NSPS If SCR had been accepted as an adequately demonstrated technology, the NO x emissions limit in

1979 could have been set as low as 0.034 lb/MBtu heat input (44 FR 33602) It took until 1998, after SCR had been installed worldwide in almost 70 GWe of coal-fired capacity (data from ([53]), that SCR was finally considered to be adequately demonstrated enough for the U.S to allow it to serve as the NSPS technology basis In that year, the federal NSPS was revised for utility boilers, requiring reductions on the order of 80% or more from new and modified sources.

19 Following the lead of SCAQMD, the state-wide California Air Resources Board (CARB) adopted Rule 1135.1 in 1980, requiring all utility units to reduce NO x emissions by 90% between 1988 and 1990; it was rescinded in March 1982, however, by Order of Superior Court Case No C 323997

Although there was considerable progress in achieving the NAAQS for SO2 and NO2 in the 1970s and 1980s, the U.S had less success in achieving the NAAQS for O3 The

1990 CAA, therefore, focused on this problem, in part by recognizing the role of NOx as

an important contributor to O3 formation It also addressed the role of NOx as a

contributor to acid rain by introducing a two-phase, rate-based emissions reduction program Continuing the tradition of distrusting the performance of the more expensive and effective post-combustion NOx control technologies employed internationally, the technical basis of the 1990 CAA was combustion modification techniques Finally, the

1990 CAA established an interstate organization – later known as the Ozone Transport Commission (OTC) – to recommend to the EPA Administrator the measures states in the OTC could take to attain the ozone NAAQS as a region.20 This led to a multi-state CTP for NOx control in the late 1990s in the OTC area

5 CTPs for SO 2 and NO x Emissions Control

to prevent owners of Table A units from meeting their emission reduction obligations simply by reducing generation from those particular units and increasing generation from other units Although these provisions were “much more heavily used than had been anticipated,” the Table A units accounted for at least 95% of the emission reductions in both 1995 and 1996 [54]

Phase II (2000-10) applies the maximum allowable emission rate established in the 1971 NSPS, in aggregate, to about 2,500 existing units, or all fossil-fueled power plants larger than 25 MWe The 2010 cap was set at 8.95 million annual tons (8.06 million annual tonnes) of SO2 A unit’s compliance is judged annually in a “truing-up” period when sources have to be able to demonstrate sufficient allowances to cover emissions

Penalties are based on a 1990 fine of $2,000 per ton ($2,197 per tonne) of SO2 above allowance levels, adjusted for inflation (e.g., $3,042 per excess ton, or $3,343 per excess tonne, by 2005) [55]

The main supply of allowances for each source is the annual allocation made by the EPA,based on the product of the phased emission rate and a baseline heat input [56] For Table A units in Phase I, the baseline heat input was generally the 1985-87 average For Phase II, the emissions rate was the lower of either the 1985 actual emissions rate or 1.2

20 The OTC (until by-laws adopted in 1991, it was the Northeast Ozone Transport Commission) consisted

of government leaders and environmental officials from the District of Columbia, the EPA, and twelve states: Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, and Virginia In addition to its policy recommending duty, it was charged with assessing the degree of interstate transport of ozone and its precursors in the northeast.

lb/MBtu (2.2 kg/Gcal), converted to tons There are also several additional supply streams First, there is a small annual allowance auction designed to help new entrants; between 1995 and 2002, this accounted for between 1.7% and 2.6% of the total amount

of allocated rights per year [57] Second, there is a small pool of opt-in allowances provided to units entering the program voluntarily (for example, eight units opted in during 2005 [55] Third, there is a complex series of “bonus” allowances [56; 58]

Banking in Title IV is unlimited, and 75% of the allowances generated in Phase I were banked for use in future compliance, regardless of phase, rather than traded [58] The bank generated in Phase I was so large that sources have been able to emit more than the aggregate allocated annual allowances throughout Phase II [55], as predicted in the late 1990s [59]

In 1994, the OTC (which consisted of government leaders and environmental officials from the District of Columbia, the EPA, and twelve states: Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania,Rhode Island, Vermont, and Virginia) agreed to a Memorandum of Understanding

(MOU) which established a three-phase program for reducing NOx emissions from large combustion sources Phase I, which began on May 1, 1995, applied year-round, region-wide emissions limits based on “reasonably available control technology” (RACT) standards for large stationary sources in O3 non-attainment areas; this amounted to roughly a 40% NOx reduction from 1990 levels

Phase II, which began May 1, 1999, and Phase III, which was supposed to begin on May

1, 2003, established a nine-state CTP during the “ozone season” of May through

September, with trading allowed year-long (Maine, Vermont, and Virginia did not join theOTC trading program) Coincidental with the start of Phase III, the EPA established another ozone season CTP, the “NOx Budget Trading Program,” which superseded the OTC Phase III but also involved additional non-OTC states; this CTP allowed the

affected states to meet the mandatory “NOx SIP Call” reductions that EPA issued in 1998 (New Hampshire is not subject to the requirements of the NOx SIP Call) Litigation, however, delayed its implementation for non-OTC states As a result, the non-OTC states

of Alabama, Illinois, Indiana, Kentucky, Michigan, North Carolina, Ohio, South Carolina,Tennessee, Virginia, and West Virginia began the first compliance period on May 31,

2004 As of the writing of [60], affected parts of Missouri were required to comply with the NOx SIP Call by May 1, 2007 EPA had stayed the requirements for Georgia pending determination on a petition to reconsider Georgia’s inclusion in the SIP Call The NBP will be superceded in 2009 by the Clean Air Interstate Rule

Banking is restricted in the OTC/NBP in order to minimize the potential for banked allowances to be used to exceed budgeted emissions in a given ozone season Emissions

in the OTC did not exceed allowances (although the allowance bank was large,

accounting for 20% of allowances after the first year), and so far, emissions in the NBP have only exceeded allowances in two years: 2003 and 2005 [61; 62] The NBP uses a system called “progressive flow control” to restrict banking In this system, once the allowance bank becomes larger than 10% of the emissions budget in a given year, if a

source wants to use banked allowances for compliance, only a portion of that source’s allowances can be redeemed on the basis of one allowance for each ton of emissions The rest are redeemed on the basis of two banked allowances for each ton of emissions The portion of banked allowances subject to the 2:1 requirement is set annually by the EPA, based on the amount by which the total bank exceeds the 10% threshold Flow control has applied in 2000-03 and 2005-07

To cope with the transition from the OTC to the NBP, all OTC allowances were officially retired, although the EPA created a small “compliance supplement pool” (CSP) of

allowances for the NBP that most OTC states apportioned in exchange for banked OTC allowances There were a few exceptions: no 1999 vintage allowances were eligible for the CSP; Pennsylvania additionally excluded 2000 vintage allowances; and Maryland apportioned allowances according to an emissions-based formula instead of according to banked allowances [63]

emissions from permitted stationary sources in the region (the equivalent of 17% of total

NOx emissions in the region) and 85% of SO2 emissions from permitted stationary

sources (31% of total SO2 emissions) RECLAIM Trading Credits (RTCs) were initially freely allocated on the basis of peak production rates that occurred between 1989 and

1992, prior to the recession that Southern California was experiencing when RECLAIM began This high emissions baseline meant that the cap did not require much of sources:

in 1994 and 1995, allowances exceeded emissions by 58% and 40%, respectively [50] RECLAIM differs from the bigger U.S CTPs in several ways other than high initial allowance allocations First, it involves an annually declining cap rather than multi-year phases Through 2003, the annual decline was about 8% for NOx and 7% for SO2, based

on 1994 levels Second, the penalties for emissions in excess of credits are not automatic,nor are they a set amount Non-complying facilities must surrender future RECLAIM Trading Credits (RTCs) to cover the excess emissions, and are also subject to significant civil penalties negotiated on a case-by-case basis (some of these fines have been

substantial, including a $17 million fine for one company with over 300 tons (273 tonnes)

of excess emissions) [50] Third, RECLAIM involves a wider range of sources than the other CTPs, including power generators, refineries, industrial sources, aerospace

companies, asphalt producers, chemical plants, and cement plants emitting more than 4 tons (3.64 tonnes) annually of each pollutant The program excludes certain “essential public services”, such as landfills, public transit, fire fighting facilities, etc Fourth, each RTC accounts for one pound (0.4 kg) of emissions, rather than one ton (0.91 tonnes) as inthe other two CTPs

Finally, there is no banking in RECLAIM, with each RTC expiring annually There is an opportunity for something akin to a very limited type of banking, however, due to the fact