Reactivity-Based VOC Control for Solvent Products: More Efficient Ozone Reduction Strategies

Transcription

1 Environ. Sci. Technol. 2006, 40, Downloaded via on December 17, 2018 at 14:05:16 (UTC). See for options on how to legitimately share published articles. Reactivity-Based VOC Control for Solvent Products: More Efficient Ozone Reduction Strategies ROBERT J. AVERY* PCA Services, Inc., 2704 Trail Wood Drive, Durham, North Carolina It has been almost 40 years since scientists and regulators began to publicly recognize that volatile organic compounds (VOCs) differ in their potential to form ozone. Since 1977, our understanding of the mechanisms by which VOCs contribute to ozone formation has grown substantially. An assessment of the science indicates that we now have sufficient understanding to develop and implement a more efficient approach to VOC control policy that will promote smarter, more cost-effective VOC controls. Furthermore, the U.S. Environmental Protection Agency s recently published Interim Guidance on Control of Volatile Organic Compounds in Ozone State Implementation Plans encourages States to consider recent scientific information on the photochemical reactivity of volatile organic compounds (VOC) in the development of State implementation plans (SIPs). As has been demonstrated by the California Air Resources Board s Aerosol Coatings Rule and the recent experience in Houston addressing high-reactivity VOCs, reactivitybased regulations may be more effective and efficient than mass-based rules in many applications. It is time for regulators, industry, and other stakeholders to work together to accelerate efforts to bring about a new paradigm in VOC control, in which the focus is on the ozoneforming potential of VOC emissions, rather than the mass of VOCs emitted. Introduction It has been almost 40 years since scientists and regulators began to publicly recognize that volatile organic compounds (VOCs) differ in their potential to form ozone. Even though modern atmospheric chemistry was in its infancy, the 1960s and 1970s saw several significant developments in VOC policy that acknowledged that the rates and routes of VOC photooxidation could differ dramatically among VOCs. The Los Angeles Air Basin s Rule 66 was a landmark air pollution regulation that first recognized the differential contributions of VOCs to ozone formation by allowing exemptions from control measures for many VOC species while singling out a few for special attention based on their relative potential for ozone formation (1). In 1971, following passage of the Clean Air Act a year earlier, the U.S. Environmental Protection Agency s (EPA) initial guidance to states for state implementation plans (SIPs) for ozone attainment included a statement that substitution of one compound for another might be useful where it would result in a clearly evident decrease in reactivity and thus tend to reduce photochemical oxidant formation (2). In its 1977 Recommended Policy on * Corresponding author phone: (423) ; fax: (919) ; rjavery@chartertn.net. Control of Volatile Organic Compounds, the EPA listed 12 organic compounds that were considered to have negligible contribution to ozone formation, and four of the 12 were exempted from regulation as VOCs (3). The underlying premise behind these recommendations was an acknowledgment that the atmospheric chemistry of VOCs could have a decided influence on the key reactions leading to the formation of tropospheric ozone (4). Since 1977, our understanding of the mechanisms by which VOCs contribute to ozone formation has grown substantially. The only significant change to the EPA s VOC policy since 1977 has been the addition of several more compounds to the list of exempt VOCs that are considered to have a negligible contribution to ozone formation (5). The U. S. federal regulatory approach continues to be a binary approach to VOCs: either a compound is considered to be a contributor to ozone formation or it is not. This approach can lead to a situation in which two low-reactivity compounds that have very similar ozone formation potentials (e.g., with reactivity just below and just above the line between exempt and nonexempt VOCs) are treated in completely different ways: one is regulated as a VOC and the other is not. Conversely, two VOCs with very different ozone formation potentials (one with reactivity just above the exemption cutoff and another with much higher reactivity) can be treated exactly the same. In other words, current regulations do not value or encourage shifts from higher to lower-reactivity compounds. An assessment of the science indicates that we now have sufficient understanding to develop and implement a more efficient approach to VOC control policy that will promote smarter, more cost-effective VOC controls (6-8). Furthermore, the EPA s recently published Interim Guidance on Control of Volatile Organic Compounds in Ozone State Implementation Plans (9) encourages States to consider recent scientific information on the photochemical reactivity of volatile organic compounds (VOC) in the development of State implementation plans (SIPs). As a final introductory note, we point out that this article is generally limited to a discussion of regulating solvent VOCs used in commercial and consumer products (CCPs) to limit the formation of ozone in the troposphere as directed by Section 183(e) of the Clean Air Act and related state regulations. It does not directly address the issue of VOCs that are hazardous air pollutants, which are regulated under Section 112 of the CAA as well as a variety of state regulations. Interested readers are also referred to a discussion of the EPA s Coatings and Composites Coordinated Rule Development (CCCR) program, an effort to coordinate regulations stemming from Clean Air Act Sections 112 and 183(e) (10). This article also does not directly address VOCs that may have the potential to form secondary organic aerosols (SOA). Preliminary evidence suggests that the relatively volatile VOCs used in most consumer and commercial products (having less than about six carbon atoms) do not participate appreciably in SOA formation (11). More research is needed to clarify the issues and identify potential candidates for regulation as SOA precursors. Photochemical Reactivity: What Is It? VOC reactivity in this article refers to a measure of the potential of a given compound to contribute to ozone accumulation in the troposphere (lower atmosphere) (12). Ground level ozone, a major component of what is commonly referred to as photochemical smog, forms in a complex series of reactions that is initiated when nitrogen dioxide /es060296u CCC: $ American Chemical Society VOL. 40, NO. 16, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY Published on Web 07/07/2006

2 (NO 2) is photolyzed by sunlight to form nitrogen oxide (NO) and atomic oxygen (O), and the latter then combines with molecular oxygen (O 2) to form ozone (O 3). In the absence of any VOCs, this ozone is consumed in a reaction that reoxidizes NO to NO 2, which starts the cycle once again. VOCs participate in this reaction scheme by providing an alternative pathway for the conversion of NO to NO 2 through free-radical intermediates formed when VOCs react with hydroxyl radicals in the atmosphere. The net result of these secondary reactions involving VOCs is an indirect increase in ozone formation brought about by the elimination of the reaction pathway that leads to ozone destruction (i.e., the NO to NO 2 conversion). Another often-used term is incremental reactivity, which can be thought of as the sensitivity of ozone formation to changes in emissions of a given VOC for a particular set of ambient conditions. The most common way of expressing VOC reactivity is in terms of the change in the amount of ozone in the atmosphere as a result of an incremental change in the amount of the VOC compound in question, frequently expressed in units of grams of ozone generated per gram of VOC (g O 3/g VOC) (13). Because of the complex chemistry involved, two VOCs may have profound differences in their contribution to ozone formation. Under a given set of conditions, a gram of solvent A might contribute less than 0.1 g of ozone, whereas a gram of solvent B might contribute more than 10 g, a 100-fold difference. Put another way, to reduce the amount of ozone in the air by 1000 kg under the same set of conditions, emissions of VOC A could be reduced by kg, or VOC B by only 100 kg: See Table 1 for specific reactivity values. VOC NO 2 + Sunlight f NO + O O + O 2 f O 3 O 3 + NO f NO 2 + O 2 VOCs + Sunlight f Radicals NO + Radicals f NO 2 max ozone formed (kg ozone/kg VOC) emissions reduction needed to reduce ozone by 1000 kg solvent A kg solvent B kg Note that we are not suggesting that solvent A can necessarily be used as a substitute for solvent B. They may have very different properties, and they may not be interchangeable in any formulation scenario. Nor are we suggesting that we should ban the use of higher-reactivity solvents altogether; many high-reactivity solvents are powerful solvents that are useful in some applications where most other solvents simply are not satisfactory. Our point is that to control the amount of ozone in the air, it makes sense to focus on controlling the VOCs that have the greatest potential to form ozone. It is the subtleties and nuances of how to go about focusing on those higher-reactivity compounds that originally confounded regulators and contributed to delaying development of a more efficient VOC policy. However, we are now at a point where available science can be effectively applied to achieve more efficient strategies for reducing tropospheric ozone. VOC reactivity is not a static function of a compound, but varies with the ambient conditions. Temperature, amount of sunlight available, and the concentration of NO x as well as the concentration and identity of other VOC compounds in the atmosphere can have a profound effect on the potential TABLE 1. Examples of MIR and k OH Values (21) compound MIR (g O 3/g VOC) k OH (10 13 ) (cm 3 /mol-s) methane methyl acetate tert-butyl acetate ethane acetone ethyl acetate methanol n-butyl acetate n-butane methyl ethyl ketone ethanol butoxyethanol n-butyl alcohol toluene o-xylene ethylene isoprene propylene of a given compound to react to form ozone. Meteorology affects all of these variables. In temperate latitudes ozone episodes, which generally do not happen on cool or cloudy days, tend to be concentrated in the summer months. For these regions, wintertime VOC emissions have little impact on ozone formation. NO x, another essential ingredient in ozone formation, tends to have concentrated emission sources in urban and suburban areas where there are numerous mobile sources as well as combustion industries, and thus ozone formation is generally lower in rural or regional areas (14). In fact, if the ratio of VOC/NO x levels is high, as occurs in many rural areas of the United States, changes in VOC emissions have little or no effect on ozone formation, a situation known as NO x-limited. Regulating VOC emissions in NO x-limited areas is ineffective at reducing ozone concentrations (15). Another factor contributing to NO x-limited conditions is that some areas that have significant NO x levels also have substantial biogenic (natural) VOCs released from trees and other green plants (16). Because these biogenic emissions tend to be highly reactive and are concentrated in the summer months, they may be sufficient to react all available NO x, again leading to a situation in which limiting anthropogenic (man-made) VOC emissions would not improve ozone levels. It is worth noting that the EPA has recently embarked on a massive NO x-reduction strategy, and as that strategy is implemented, reducing NO x levels in many parts of the country, even more areas are destined to become NO x-limited over the next several years, making control of anthropogenic VOCs less and less important in the overall ozone control strategy (17). Thus, there are dramatic differences from place to place in the response of ozone concentrations to changes in anthropogenic emissions of VOCs. In highly urbanized areas that have a lot of both anthropogenic VOC and NO x emissions and low-to-moderate biogenic VOC emissions, regulating anthropogenic VOC emissions may be a necessary component of the ozone-control strategy. In rural areas, especially those that are not downwind of urban centers, controlling VOC emissions may have a negligible impact on ozone levels. What about Metrics: Which Reactivity Measurement Should Be Used? It may be useful to spend a few minutes reviewing the basics of VOC reactivity metrics. Because most atmospheric reactions involving VOCs begin with a reaction of the VOC with hydroxyl radicals, early calculations of VOC reactivity were based on the hydroxyl radical reaction rate constant, k OH, for the VOC of interest ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 16, 2006

3 While the k OH method is useful for determining the initial rate of reaction for VOCs in the atmosphere, the ultimate impact of a VOC on atmospheric ozone is not limited to the initial reaction. The products of that initial reaction are available to react again, and those secondary reaction products react yet again, and so forth, until nonreactive products are produced, terminating the reaction series. Thus, while k OH may be useful to obtain an estimate of the relative reactivity of a compound, a more robust method is needed to determine the overall effect of a VOC on atmospheric ozone, which depends, as we saw in the previous section, on many variables including atmospheric conditions and the presence of other reactants (18). One way to directly measure the change in ozone resulting from a VOC emission is to add a known amount of a VOC to an environmental chamber containing a simulated atmosphere and measure the resulting change in ozone concentration in the chamber. Unfortunately, while chamber studies can be carefully controlled and monitored, even today s most sophisticated chambers cannot simulate the complexities of the troposphere with a sufficient degree of accuracy to make those studies directly meaningful. However, it is possible to use environmental chambers to verify and validate chemical mechanisms that can then be input into complex computer airshed models, and that is the basis for the various incremental reactivity scales that have been developed over the past two decades. The most common scale in use today, the maximum incremental reactivity, or MIR, scale, is a measure of the average ozone yield of VOCs derived by adjusting the NO x emissions in 39 base case urban scenarios to yield the highest incremental reactivity of the base ROG mixture, where ROG is reactive organic gas (13). Because MIR conditions are a reasonable approximation to the worst-case conditions during ozone episodes in a variety of urban surroundings, the California Air Resources Board (ARB) adopted MIR as its metric of choice first for its lowemission vehicles and clean fuels regulations and later in its aerosol coatings regulations (19, 20). Table 1 gives examples of MIR and k OH values for a range of solvents and a few other common VOCs. For comparison, the MIR of the base ROG is 3.71 g O 3/g VOC. MIR conditions do not reasonably approximate the conditions in most areas of the United States, and that has been a significant area of concern that has contributed to preventing the EPA from moving more quickly to implement nationwide reactivity-based regulations. This issue of metrics is one of the key issues addressed by the EPA- and NARSTOsponsored Reactivity Research Work Group (RRWG). The RRWG went through an exhaustive process of listing metrics considered to have potential for use on a national scale and evaluating them using a set of selection criteria including effective range, consistency, and variability over different areas of coverage. As a result of this work, the RRWG focused on a small number of metrics and addressed technical questions that had been raised (22). [The Reactivity Research Work Group (RRWG) was formed in 1998 as a result of a summit meeting of interested scientists, regulators, and industry representatives arranged by the USEPA and NARSTO (formerly the North American Research Strategy for Tropospheric Ozone) earlier that year. The stated Mission of the RRWG (condensed) is to provide an improved scientific basis for reactivity-related regulatory policies... by bringing together all parties actively interested in sponsoring, planning, performing or assessing policy-relevant scientific research on the reactivities of organic compounds emitted to ambient air, as related to the formation of ozone, PM2.5, and regional haze... for the purposes of coordinating such research and defining potential applications, while continuously involving key policy makers. ] The RRWG asked a panel of leading atmospheric scientists to address the issue of whether different scales would result in substantially different reactivity rankings among VOCs. They found that, although some scales may be more compressed than others, there is little variation in the rankings of VOCs among the commonly used reactivity scales. In general, a given VOC tends to remain in the same position relative to other VOCs on any of the reactivity scales (23, 24). Although the RRWG process continues, with a few remaining technical questions to be answered, it appears that the issue of metrics is no longer a significant stumbling block to the use of reactivity, even over such a widespread and diverse area as the United States. Advantages of Reactivity-Based VOC Controls Both regulators and the regulated community, which includes the consumers and producers of regulated products, have a common goal: a regulatory system that achieves the desired environmental improvement in the most efficient and effective way possible. In areas where anthropogenic VOCs are a significant factor in tropospheric ozone formation, using reactivity-based controls creates a more direct correlation between VOC emissions and ozone formation. Mass-based limits seek to control the amount of VOC emitted (directly or indirectly) but do not consider the ozone-forming potential of those emissions. This sometimes can lead to situations in which the ozone-forming potential of products meeting more stringent VOC standards has actually been higher than that for the products they replaced. Especially in recent years, as mass-based limits have become increasingly stringent, we have seen more and more cases where further reductions in VOC content have actually resulted in higher ozone-forming potential. This has occurred because as the VOC mass content of formulated products is driven downward, the formulator has fewer options in formulating. VOCs that are less effective at achieving the desired effect, be it dissolving paint resin or emulsifying dirt, must be replaced with VOCs that are more effective, regardless of their ozone-forming potential, and some of the VOCs that offer the most efficiency to the formulators also have a higher potential for ozone formation. Thus, while reductions in VOC emissions may in most situations be helpful, it is quite possible to substitute a lower mass of a higher-reactivity compound in place of a greater mass of a lower-reactivity compound, resulting in an overall VOC mass reduction but at the same time increasing the ozone-forming potential significantly. The point to be made is that a reduction in the mass of VOCs may or may not reduce the ozone formed from emissions of that product, but reducing the ozone-forming potential (relative reactivity) of the product will always result in reduced ozone formation under conditions where man-made VOCs are having an effect on ozone levels. VOC controls under the binary regulatory approach value reductions in VOC content and substitutions of exempt for nonexempt, but not shifts from high reactivity to medium or from medium reactivity to low. Therefore, there are no incentives for formulators to consider those possibilities, and as a result, many opportunities may be missed. California s ARB recognized this phenomenon in its justification of its reactivity-based aerosol coatings regulation (25). After the ARB had proposed a new round of VOC mass reductions for aerosol coatings, producers of aerosol coatings pointed out that in order to meet the new standards, the formulators were being forced to increase the percentage of higher-reactivity aromatic solvents in their formulations, and the ozone-forming potential of the resulting products was higher than before. A similar situation occurred when the South Coast Air Quality Management District (SCAQMD) proposed to reduce its coatings limits. To achieve the proposed lower VOC levels for varnish and sanding sealers, VOL. 40, NO. 16, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

4 many formulators found that they had to switch resin systems, and the new resins required a change to higher-reactivity solvents, again increasing the overall ozone-forming potential of the final product. One producer wrote to SCAQMD: We find that, to the extent that lowering the VOC content limit for Varnish from 350 g/l to 275 g/l causes a shift from conventional solvent borne varnishes to alternative waterborne clear wood finishes, a decrease of 21% in VOC content is accompanied by an increase of 153% in VOC reactivity. In terms of relative ozone formation impacts of VOC emitted, substituting waterborne clear finishes for conventional solventborne varnishes will almost double the amount of ozone formed (26). Some have suggested that regulators should consider simply banning the use of VOCs that have relatively high ozone-forming potential. While that might initially seem like a reasonable idea, it fails to take into account that many of those higher-reactivity VOCs have other properties (e.g., solvency or emulsifying power) that make them highly effective for use in some formulations, achieving the same level of performance with a lower ozone formation potential. One of the real strengths of a reactivity-based program is that it allows appropriate tradeoffs to be made, so that effective formulated products can continue to be available to the consumer, while at the same time overall ozone formation potential is minimized. How Has VOC Reactivity Been Used Thus Far? In addition to the regulations previously discussed, regulators have already used the concept of reactivity in a variety of innovative ways, including control of mobile sources and industrial emissions. In this section, we will examine several of those approaches and discuss their potential usefulness and relevance to a future reactivity-based policy. The EPA s Standard Binary Approach. The most basic use of reactivity is found in the binary or 2-bin approach implicit in the EPA s current VOC policy, in which compounds are considered to be VOCs unless they are on the list of exempt compounds. In the 1977 policy, the EPA named four negligibly reactive compounds on the basis of experimental work that indicated that those compounds were unlikely to contribute to violations of the National Ambient Air Quality Standard for ozone, and these compounds were exempt from regulation as VOCs (3). These were (i) methane, (ii) ethane, (iii) 1,1,1-trichloroethane (methyl chloroform), and (iv) trichlorotrifluoroethane (Freon 113). Because ethane was determined to be the most reactive of the original list of negligibly reactive compounds, it became the de facto benchmark by which other compounds were measured; those that could be shown to be equally or less reactive than ethane were considered negligibly reactive, and those that were more reactive than ethane were not. Compounds may be added to the exempt list in 40CFR51.100(s) if it can be demonstrated that their reactivity is equal to or less than that of ethane (18, 27). There are several significant problems inherent in this system. First, because most of the compounds on the exempt list are halogenated compounds with limited application, this system does little in practice to encourage substitution of lowreactivity compounds for higher-reactivity compounds. Further, the so-called bright line between exempt and nonexempt VOCs can result in scarce scientific and economic resources being diverted into determining whether a few compounds are slightly above or below the line, an exercise with little overall value to atmospheric quality. This brightline approach also causes very different treatment of compounds that are virtually identical in their ozonegenerating potential. That is, a substantial number of commercially useful compounds have very low potential to create ozone (only slightly greater than ethane), but instead of having their use encouraged, they are treated exactly the same as if they were highly reactive. Further, compounds with quite different reactivity (very reactive versus just above ethane) are treated as if they were the same, providing no incentive for switching from higher-reactivity to lowerreactivity compounds. Using Reactivity To Identify Categories for Regulation: Section 183(e). In Section 183(e) of the Clean Air Act Amendments of 1990, Congress recognized the significance of VOC reactivity by directing the EPA to develop a control strategy for VOC emissions from CCPs, taking into account the reactivities of such emissions. Specifically, the EPA was to (i) determine the potential of VOC emissions from CCPs to contribute to ozone levels that violate the National Ambient Air Quality Standards for ozone; (ii) identify the highly reactive species of such VOC emissions; and (iii) list those CCPs that account for at least 80% of the VOC emissions on a reactivityadjusted basis. Once it had identified the CCPs that account for >80% of VOC emissions on a reactivity-adjusted basis, the EPA was further directed to divide them into four groups and regulate one group every two years. At this writing, the EPA has promulgated national regulations for three categories: Architectural and Industrial Maintenance Coatings, Consumer Products, and Automobile Refinish Coatings (28). Interested readers are referred to an excellent discussion by Dimitriades of the scientific issues surrounding this seemingly straightforward mandate (18). Alternative and Reformulated Fuels for Mobile Sources. As early as 1991, the ARB approved reactivity adjustment factors (RAFs) as a way to compare emissions from lowemission vehicles (LEVs) using reformulated gasoline or other alternative fuels. According to an ARB staff report, RAFs are based on a comparison of the ozone reactivity of an alternative fuel or reformulated gasoline low-emission vehicle to the ozone reactivity of a comparable conventional gasoline lowemission vehicle. The comparison of the reactivities of the two classes of vehicles is accomplished through the application of a maximum incremental reactivity (MIR) scale which identifies MIR values for the over 140 individual hydrocarbon species that can be found in vehicle exhaust (29, 30). Control of Highly Reactive VOCs in the Houston/ Galveston Area. In the past few years, the Texas Commission on Environmental Quality (TCEQ) has taken a unique approach to the use of reactivity principles to address ozone in the Houston/Galveston/Brazoria (HGB) area. After the intensive TexAQS 2000 study found that the Houston/ Galveston area had frequent large, previously unidentified episodic and/or fugitive emissions of high-reactivity compounds, TCEQ in December of 2004 adopted the Highly Reactive Volatile Organic Compound Emissions Cap and Trade (HECT) program. The HECT program established a mandatory annual highly reactive volatile organic compound (HRVOC) emissions cap on all sites located in the Houston/ Galveston nonattainment area that have the potential to emit more than 10 tpy of HRVOC (31). The program applied to four industrial source categories: flares, vents, cooling towers, and fugitive emissions. HRVOCs were defined as the following: ethene, propene, 1,3-butadiene, and butenes (c-2- butene, t-2-butene, isobutene, and 1-butene). Although preliminary anecdotal indications have been positive, a follow-up air quality study, TexAQS II, planned for , will provide data on the effectiveness of this approach. Some of the issues TCEQ wants to study in more detail are the following (32): (1) Have the concentrations of HRVOC emissions decreased in Houston from 2000 to 2005/ 2006? (2) Has the frequency of high concentration events decreased in that time frame? (3) Can very high concentrations of lower-reactivity compounds cause transient high ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 16, 2006

5 ozone? If so, how often does this occur? (4) Do more VOCs need to be controlled to reach attainment of the eight-hour ozone standard? If so, which ones? (5) Are the reactivity metrics commonly used (e.g., MIR) appropriate for conditions found in industrial plumes in Houston? The answers to these questions will help determine whether the findings in the HGB area are generalizable to other urban and/or highly industrialized areas. ARB s Aerosol Coatings Regulation. Probably the most innovative and useful example of a reactivity-based regulatory scheme to date is the ARB s Regulation for Reducing the Ozone Formed from Aerosol Coating Product Emissions (Aerosol Coatings Rule). This regulation was developed and implemented in response to industry concerns that proposed mass-based limit reductions were not practically achievable. Industry had found that the more stringent mass-based limits could be met only by substituting high-solvency and highreactivity aromatic hydrocarbon compounds for other lowersolvency and lower-reactivity compounds. Their calculations indicated that on a reactivity-adjusted basis, the reduced mass-based limits would actually increase the ozone-forming potential of the products. ARB saw the logic in those concerns and set out to develop a reactivity-based rule that would yield the same ozone reductions as the mass-based rule it was replacing. Key elements of the development of this pioneering rule included (20, 33) (i) using the MIR scale, (ii) including compounds that were considered exempt in mass-based regulations, (iii) calculating upper-limit MIR values for compounds that did not have a known MIR value, (iv) calculating group MIR values for hydrocarbon solvent mixtures, (v) addressing uncertainty in the MIR scale, and (vi) calculating equal air quality benefit reactivity limits. We have watched this rule closely over the intervening years, and although it might be difficult or impossible to provide scientific evidence that the resulting ozone created is the same as or less than that from the original mass-based rule, it appears to have worked very well in terms of regulatory workability. Initial concerns from ARB about enforceability and from industry about frequent changes requiring numerous reformulations have not proven to be significant issues. One of the last major concerns with the program was answered when, on September 13, 2005, the EPA published in the Federal Register a final rule that approved the new aerosol coatings products regulation as part of the California State Implementation Plan, approved the use of ARB s Tables of Maximum Incremental Reactivity, and revised the EPA definition of VOC to allow compounds that the EPA had previously considered exempt to count toward a product s reactivity-based VOC limit for the purpose of California s aerosol coatings regulation (34). This action answered several questions about how the EPA might deal with new rules of this type and gave a significant boost to the concept of reactivity-based VOC limits. In developing this first reactivity-based regulation for VOCcontaining products, the ARB did much of the front-end work for other regulatory bodies that might be considering new reactivity-based regulations. Establishing Reactivity Values for Complex Hydrocarbon Solvents. In the previous section, we mentioned that the ARB had to develop a new system that would include reactivity values for hydrocarbon solvent mixtures. Many hydrocarbon solvents (e.g., mineral spirits and Stoddard solvent) are complex combinations of two major classes of hydrocarbon compounds: aliphatic hydrocarbons (paraffins and cycloparaffins) and aromatic hydrocarbons. These complex hydrocarbon solvents frequently contain hundreds of closely related isomers, most at very low concentrations (<1%). This compositional complexity initially precluded the establishment of MIR values based on individual chemicals. To address this issue, ARB worked with the hydrocarbon solvent manufacturers to develop a series of bins for which a MIR value is assigned to each class/type of hydrocarbon solvent. The bin MIR values consider the impact that each type of hydrocarbon chemistry (e.g., n-paraffin, isoparaffin, cycloparaffin, and aromatic hydrocarbons) has on the overall reactivity of the solvent, although in order to be practical to implement, they are defined using basic product specifications such as boiling range and aliphatic/aromatic content. There are 20 bins for aliphatic hydrocarbon solvents and four bins for aromatic hydrocarbon solvents. The complete list of the bins is located in Section of the California Consumer Products Regulations (33). Examples of common hydrocarbon bin reactivities are: hydrocarbon (H/C) bin and definition bin 11 H/C, low aromatics mineral spirits (B.P F, <2% aromatics) bin 6 H/C, VM&P Naphtha (B.P F, <2% aromatics) bin 22 H/C, Aromatic 100 (B.P F, >98% aromatics) This rule has demonstrated that available science is sufficient to effectively incorporate reactivity into policy, even for complex mixtures of solvents. Note that when ARB adopted amendments to the tables of MIR values in 2003, no changes to the MIR values for the 24 different hydrocarbon solvents contained in Section 94701, Title 17, CCR were deemed necessary (35). Summary. In summary, reactivity is an appropriate option for federal and state agencies to consider when developing VOC control strategies. As has been demonstrated by ARB s Aerosol Coatings Rule and the recent experience in Houston addressing HRVOCs, reactivity-based regulations may be more effective and efficient than mass-based rules in many applications. It is time for regulators, industry, and other stakeholders to work together to accelerate efforts to bring about a new paradigm in VOC control, in which the focus is on the ozone-forming potential of VOC emissions, rather than the mass of VOCs emitted. Acknowledgments The author acknowledges and thanks Dr. Rosemary Zaleski of Exxon Mobil Corporation and Dr. David Morgott of Eastman Kodak Company for their insightful comments on the technical issues addressed in this article; Andrew Jaques of the American Chemistry Council for his help on the section on complex hydrocarbon solvent bins; and the American Chemistry Council s Solvents Council for financial support that enabled the writing of this article. Literature Cited MIR, g O 3/g VOC (1) Krenz, W. B.; Dickinson, J. E.; Chass, R. L. An appraisal of Rule 66 of the Los Angeles County Air Pollution Control District. J. Air Pollut. Control Assoc. 1968, 18, (2) U.S. EPA. Requirements for preparation, adoption, and submittal of implementation plans. Fed. Regist. 1971, 36, (3) U.S. EPA. Recommended policy on control of volatile organic compounds. Fed. Regist. 1977, 42, (4) Dimitriades, B.; Joshi, S. B. Application of reactivity criteria in oxidant-related emission control in the USA; EPA 600/ l b; NTIS PB In Proceedings on the International Conference on Photochemical Oxidant Pollution and Its Control; Office of Research and Development Environmental Sciences Research Laboratory: Research Triangle Park, NC, 1977; Vol. II, pp (5) U.S. EPA. Revision to definition of volatile organic compounds - Exclusion of tert-butyl acetate. Fed. Regist. 2004, 69, (6) RRWG Policy Team. VOC Reactivity Policy White Paper, (accessed Apr 2006). VOL. 40, NO. 16, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

6 (7) RRWG Science Team. VOC Reactivity Science Assessment, (accessed Apr 2006). (8) McBride, S.; Oravetz, M.; Russell, A. Cost-benefit and uncertainty issues in using organic reactivity to regulate urban ozone. Environ. Sci. Technol. 1997, 31, 238A-244A. (9) U.S. EPA. Interim guidance on control of volatile organic compounds in ozone state implementation plans. Fed. Regist. 2005, 70, (10) U.S. EPA. Coatings and composites coordinated rule development (CCCR). (accessed Apr 2006). (11) Griffin, R.; Cocker, D., III; Seinfield, J. Incremental aerosol reactivity: Application to aromatic and biogenic hydrocarbons. Environ. Sci. Technol. 1999, 33, (12) Bergin, M. S.; Russell A. G.; Carter, W. P. L.; Croes, B. E.; Seinfeld, J. H. VOC Reactivity and Urban Ozone Control. In Encyclopedia of Environmental Analysis and Remediation; Meyers, R. A., Ed.; Wiley & Sons: New York, 1998; pp (13) Carter, W. P. L. Development of ozone reactivity scales for volatile organic compounds. J. Air Waste Manage. Assoc. 1994, 44, (14) Sillman, S. The relation between ozone, NOx and hydrocarbons in urban and polluted rural environments. Atmos. Environ. 1999, 33, (15) Blanchard, C. E. Report on Spatial Mapping of VOC and NOx Limitation of Ozone Formation in Six Areas. In Proceedings of the 94th Annual Meeting of the A&WMA, Orlando, FL, June, 2001; Paper No. 215, Session No. AB-2c. (16) Pierce, T.; Geron, C.; Bender, L.; Dennis, R.; Tonnesen, G.; Guenther, A. Influence of isoprene emissions on regional ozone modeling. J. Geophys. Res. 1998, 103, (17) National Research Council. Summary of VOC, NOx, and Ozone Observations. In Rethinking The Ozone Problem In Urban and Regional Air Pollution; National Research Council, Committee on Tropospheric Ozone Formation and Measurement, National Academy Press: Washington, DC, 1991; pp (18) Dimitriades, B. Scientific basis for the VOC reactivity issues raised by Section 183(e) of the Clean Air Act Amendments of J. Air Waste Manage. Assoc. 1996, 46, (19) California Air Resources Board. Proposed Regulations for Low- Emission Vehicles and Clean Fuels: Staff Report and Technical Support Document; California Environmental Protection Agency Air Resources Board: Sacramento, CA, (20) California Air Resources Board. Initial Statement of Reasons for the Proposed Amendments to the Regulation for Reducing Volatile Organic Compound Emissions from Aerosol Coating Products; California Environmental Protection Agency Air Resources Board: Sacramento, CA, 2000; Chapter IV, pp (21) Carter, W. P. L. The SAPRC-99 chemical mechanism and updated VOC reactivity scales; Report to California Air Resources Board Contracts No and Air Pollution Research Center and College of Engineering, Center for Environmental Research and Technology: University of California, Riverside, CA, (22) Hakami, A.; Arhami, M.; Russell, A. Further Analysis of VOC Reactivity Metrics and Scales; Final Report to U.S. EPA, Contract No. 4D-5751-NAEX, July (23) Carter, W. P. L.; Tonnesen, G.; Yarwood, G. Investigation of VOC Reactivity Effects Using Existing Regional Air Quality Models; Report to Reactivity Research Work Group; Air Pollution Research Center and College of Engineering, Center for Environmental Research and Technology: University of California, Riverside, CA, (24) Carter, W. P. L. Calculation of Reactivity Scales Using Regional Models; Report (draft) to Reactivity Research Work Group; Air Pollution Research Center and College of Engineering, Center for Environmental Research and Technology: University of California, Riverside, CA, (25) California Air Resources Board. Updated Informative Digest, Adoption of Amendments to the Regulation for Reducing Volatile Organic Compound Emissions from Aerosol Coating Products and Tables of Maximum Incremental Reactivity (MIR) Values, and Adoption of Amendments to ARB Test Method 310, Determination of Volatile Organic Compounds in Consumer Products ; Stationary Source and Monitoring and Laboratory Divisions, California Environmental Protection Agency Air Resources Board: Sacramento, CA, (26) Wendoll, R. Draft Environmental Assessment for Proposed Amended Rule Architectural Coatings - Comment Letter to South Coast Air Quality Management District, Dunn-Edwards Corporation, (27) U.S. EPA. Definition of Volatile Organic Compound, 40 CFR (s), (28) U.S. EPA. Fact Sheet: Final Air Regulation for Architectural Coatings; OAR Policy and Guidance, Office of Air and Radiation: Research Triangle Park, NC, (29) California Air Resources Board. Staff Report: Initial Statement of Rulemaking: Proposed Amendments to Low-Emission Vehicle Regulations; Mobile Source Division, California Environmental Protection Agency Air Resources Board: Sacramento, CA, (30) U.S. EPA. Reformulated Gasoline - Basic Information; Office of Air and Radiation, Office of Transportation and Air Quality: Washington, DC, (31) Stoeckenius, T.; Russell, J. Survey of HRVOC Regulations; Report Houston Advanced Research Center Project No. H EE-UT-TI( ); Environ International Corp.: Novato, CA, (32) Estes, M. VOC Reactivity Before, During and After TexAQS A Brief Discussion; TexAQS II Science Meeting, Austin, TX, (33) California Air Resources Board. Regulation for Reducing the Ozone Formed From Aerosol Coating Product Emissions Article 3. In The California Consumer Products Regulations; California Environmental Protection Agency Air Resources Board: Sacramento, CA, 2005; pp (34) U.S. EPA. Revisions to the California state implementation plan and revision to the definition of volatile organic compounds (VOC): Removal of VOC exemptions for California s aerosol coating products reactivity-based regulation. Fed. Regist. 2005, 70, (35) California Air Resources Board. Adoption of Amendments to the Tables of Maximum Incremental Reactivity (MIR) Values; California Environmental Protection Agency Air Resources Board: Sacramento, CA, Received for review February 10, Revised manuscript received April 25, Accepted May 24, ES060296U ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 16, 2006