Daily Rules, Proposed Rules, and Notices of the Federal Government


40 CFR Part 52

[EPA-R09-OAR-2012-0345, FRL-9727-1]

Approval and Promulgation of Implementation Plans; State of Hawaii; Regional Haze Federal Implementation Plan

AGENCY: Environmental Protection Agency (EPA).
ACTION: Final rule.
SUMMARY: The FIP establishes an emissions cap of 3,550 tons of sulfur dioxide (SO2) per year from three specific fuel oil-fired, electric utility boilers on the Island of Hawaii beginning in 2018. The Hawaii Electric Light Company (HELCO) can minimize impacts on the ratepayers by meeting the cap through the increased use of renewable energy and energy conservation. EPA finds that this control measure, in conjunction with other emissions control requirements that are already in place, will ensure that reasonable progress is made during this first planning period toward the national goal of no man-made visibility impairment by 2064 at Hawaii's two Class I areas.

EPA worked closely with the State of Hawaii in the development of this plan and the State has agreed to incorporate the control requirements into the relevant permits. The State has indicated that it intends to take full responsibility for the development of future Regional Haze plans.

DATES: This rule is effective on November 8, 2012.
ADDRESSES: EPA has established docket number EPA-R09-OAR-2012-0345 for this action. Generally, documents in the docket are available electronically athttp://www.regulations.govor in hard copy at EPA Region 9, 75 Hawthorne Street, San Francisco, California. Please note that while many of the documents in the docket are listed at,some information may not be specifically listed in the index to the docket and may be publicly available only at the hard copy location (e.g., copyrighted material, large maps, multi-volume reports or otherwise voluminous materials), and some may not be available at either location (e.g., confidential business information). To inspect the hard copy materials, please schedule an appointment during normal business hours with the contact listed directly below.
FOR FURTHER INFORMATION CONTACT: Gregory Nudd, Air Planning Office (AIR-2), U.S. Environmental Protection Agency Region 9, 415-947-4107,

Throughout this document, wherever “we,” “us,” or “our,” is used, we mean the United States Environmental Protection Agency (EPA).

Table of Contents I. Background and Purpose A. Definitions B. Overview II. EPA Responses to Comments A. EPA Responses to Written Comments 1. Baseline Visibility, Natural Visibility and Uniform Rate of Progress 2. Estimating Natural Visibility Conditions 3. Contribution Assessment According to IMPROVE Monitoring Data 4. Impact of Fugitive Dust on Visibility Impairment in Hawaii Class I Areas 5. Subject to BART Analysis 6. BART Determination for Kanoelehua Hill 7. NOXReasonable Progress Analysis for the State of Hawaii 8. Reasonable Progress Analysis for SO2Emissions on the Big Island 9. Point Source SO2Emissions on Maui 10. Reasonable Progress Analysis for SO2Emissions on Maui 11. Agricultural Burning on Maui 12. Integral Vista Issue and Reasonably Attributable Visibility Impairment 13. Comments on the Monitoring Strategy 14. Other Comments B. Comments From the Public Hearings III. Summary of EPA Actions IV. Statutory and Executive Order Reviews I. Background and Purpose A. Definitions

For purposes of this document, we are giving meaning to certain words or initials as follows:

1. The words or initialsActorCAAmean or refer to the Clean Air Act.

2. The initialsb extmean or refer to total light extinction.

3. The initialsBARTmean or refer to Best Available Retrofit Technology.

4. The termBig Islandrefers to the Island of Hawaii.

5. The termClass I arearefers to a mandatory Class I Federal area.1

1Although states and tribes may designate as Class I additional areas which they consider to have visibility as an important value, the requirements of the visibility program set forth in section 169A of the CAA apply only to “mandatory Class I Federal areas.”

6. The initialsDOHrefer to the Hawaii Department of Health.

7. The initialsdvmean or refer to deciview(s).

8. The initialsEGUmean or refer to Electric Generating Units.

9. The wordsEPA, we, usorourmean or refer to the United States Environmental Protection Agency.

10. The initialsFIPmean or refer to Federal Implementation Plan.

11. The initialsFLMsmean or refer to Federal Land Managers.

12. The wordsHawaiiandStatemean or refer to the State of Hawaii.

13. The initialsHECOmean or refer to the Hawaiian Electric Company.

14. The initialsHELCOmean or refer to the Hawaii Electric Light Company.

15. The initialsIMPROVEmean or refer to Interagency Monitoring of Protected Visual Environments monitoring network.

16. The initialsMECOmean or refer to Maui Electric Company.

17. The initialsMWmean or refer to megawatt(s).

18. The initialsNO Xmean or refer to nitrogen oxides.

19. The initialsNPmean or refer to National Park.

20. The initialsOCmean or refer to organic carbon.

21. The initialsPMmean or refer to particulate matter.

22. The initialsPM 2.5mean or refer to particulate matter with an aerodynamic diameter of less than or equal to 2.5 micrometers (fine particulate matter).

23. The initialsPM 10mean or refer to particulate matter with an aerodynamic diameter of less than or equal to 10 micrometers (coarse particulate matter).

24. The initialsPSDmean or refer to Prevention of Significant Deterioration.

25. The initialsRAVImean or refer to Reasonably Attributable Visibility Impairment.

26. The initialsRPmean or refer to Reasonable Progress.

27. The initialsRPGorRPGsmean or refer to Reasonable Progress Goal(s).

28. The initialsSIPmean or refer to State Implementation Plan.

29. The initialsSO 2mean or refer to sulfur dioxide.

30. The initialstpymean or refer to tons per year.

31. The initialsTSDmean or refer to Technical Support Document.

32. The initialsURPmean or refer to Uniform Rate of Progress.

33. The initialsVOCmean or refer to volatile organic compounds.

34. The initialsWRAPmean or refer to the Western Regional Air Partnership.


On May 29, 2012, the EPA proposed a FIP to address regional haze in the State of Hawaii. We proposed to determine that this FIP would meet the requirements of the CAA and EPA's rules concerning reasonable progress towards the national goal of preventing any future and remedying any existing man-made impairment of visibility in mandatory Class I areas.2 A detailed explanation of the requirements for regional haze plans and an explanation of EPA's Plan are provided in our Notice of Proposed Rulemaking and are not restated here.3

2Areas designated as mandatory Class I Federal areas consist of NPs exceeding 6000 acres, wilderness areas and national memorial parks exceeding 5000 acres, and all international parks that were in existence on August 7, 1977. 42 U.S.C. 7472(a). In accordance with section 169A of the CAA, EPA, in consultation with the Department of the Interior, promulgated a list of 156 areas where visibility is identified as an important value. 44 FR 69122 (November 30, 1979). The extent of a mandatory Class I area includes subsequent changes in boundaries, such as park expansions. 42 U.S.C. 7472(a). Although states and tribes may designate as Class I additional areas which they consider to have visibility as an important value, the requirements of the visibility program set forth in section 169A of the CAA apply only to “mandatory Class I Federal areas.” Each mandatory Class I Federal area is the responsibility of a “Federal Land Manager.” 42 U.S.C. 7602(i). When we use the term “Class I area” in this action, we mean a “mandatory Class I Federal area.”

3 See77 FR 31691 (May 29, 2012).

In our Notice of Proposed Rulemaking, we proposed to find that there was only one source in Hawaii that was subject to Best Available Retrofit Technology (BART) requirements, the Kanoelehua Hill Generating Station (Hill) on the Island of Hawaii (the Big Island). We also proposed to find that the current level of pollution control at Hill was consistent with BART and no additional controls would be required to meet the BART requirement. In addition, the EPA proposed to find that sufficient emissions reductions were expected on Maui to make reasonable progress during the first implementation period of 2001-2018. We also proposed to find that additional SO2reductions were required on the Big Island to ensure reasonable progress. We proposed that those reductions should be derived from controlling emissions on three oil-fired power plants on the Big Island: Hill, Puna and Shipman. The proposed control measure would cap the emissions of these three plants at 3,550 tons of SO2per year beginning in 2018. EPA received several comments during the public comment period on our proposal. We have provided summaries of and responses to significant comments below. Following consideration of all comments, EPA has decided to finalize the Hawaii Regional Haze FIP as proposed with one clarification regarding the compliance date for the emissions cap. We will work with the Hawaii Department of Health on developing future regional haze plans.

II. EPA Responses to Comments

EPA held two public hearings in Hawaii on May 31 and June 1, 2012 to accept oral testimony and written comments on the proposal. The first meeting was held at Maui College in Kahului on the Island of Maui. Twenty people provided oral comments and four provided written comments at this hearing. The second hearing was at the Waiakea High School in Hilo on the Big Island. Four people provided oral comments at this hearing and one provided written comments. Verbatim transcripts of the public hearings are available in the public docket for this rulemaking, Docket ID No. EPA-R09-OAR-2012-0345, which can be accessed through thewww.regulations.govWeb site.

We also received an additional 18 written comments through email, postal mail and the rulemaking docket. These comments are also available in the public docket for this rulemaking, Docket ID No. EPA-R09-OAR-2012-0345, which can be accessed through thewww.regulations.govWeb site.

A. EPA Responses to Written Comments

EPA received 18 written comments on the proposal. Commenting organizations include: Friends of Haleakala National Park (FHNP), Alexander and Baldwin, the parent company of Hawaii Commercial and Sugar (HC&S), Maui Electric Company (MECO), Hawaii Electric Light Company (HELCO), National Park Service (NPS), Maui Tomorrow Foundation (Maui Tomorrow), Law office of Marc Chytilo on behalf of Preserve Pepe'ekeo Health and Environment and private citizens (Chytilo), Robert W. Parsons on behalf of the Office of the Mayor of Maui (Parsons) and Earthjustice on behalf of the National Parks Conservation Association, Sierra Club, and Blue Planet Foundation (Earthjustice). Seven private citizens also submitted comments on the proposal.

1. Baseline Visibility, Natural Visibility and Uniform Rate of Progress

Comment:Four commenters (Earthjustice, HC&S, HELCO, and MECO) believe that EPA's proposed analysis contains a fundamental flaw in including the contribution of the Kilauea Volcano in baseline visibility conditions, but excluding it from natural visibility conditions. The commenters asserted that EPA must revise its analysis and the resulting uniform rate of progress (URP) in the final FIP.

Two of these commenters (HELCO, MECO) stated that EPA's exclusion of volcanic emissions from the determination of natural visibility conditions is arbitrary and capricious. Another of the commenters (Earthjustice) stated that EPA's methods for incorporating volcanic emissions into its analysis are internally inconsistent and arbitrary. These commenters asserted that while emissions from the volcano vary from year to year, there is no reasonable basis for EPA to completely exclude them from the estimate of natural conditions.

According to two of the commenters (HELCO, MECO), EPA has expressed the opinion that Kilauea could stop erupting at any time and that natural visibility conditions in 2064 might not include emissions from the volcano. In the view of the commenters, this does not justify EPA's use of the default conditions developed by the Western Regional Air Partnership (WRAP) for western states in the continental United States to determine requirements for Hawaii; rather, it displays a fundamental misunderstanding of Kilauea's emissions profile. Based on a report attached to the comments, the commenters asserted that significant SO2emissions would continue venting from the volcano even if it were to stop erupting immediately because, although SO2output is greatest during eruptive events, Kilauea emits SO2at all times, even during non-eruptive periods. The commenters contend that a substantial amount of data on Kilauea's emissions has been collected, and EPA should, at a minimum, use existing data to develop a “non-eruptive” emissions profile. The commenters stated that like particulate emissions from fire, SO2emissions from Kilauea are naturally occurring and would continue to occur in the absence of human activities. Accordingly, the commenters asserted that EPA cannot simply ignore emissions from Kilauea.

These commenters (HELCO, MECO) stated that by including emissions from Kilauea in baseline visibility conditions but excluding them from natural visibility conditions, EPA has created an “apples to oranges comparison” that artificially inflates the amount of manmade emissions reductions necessary in Hawaii. As a result, the commenters asserted, the proposed FIP would establish reasonable progress goals that would be impossible to achieve through the reduction ofanthropogenic emissions, which is inconsistent with EPA's own guidance. The commenters conclude that EPA must revise its analysis and the URP based on a proper evaluation of volcanic emissions and EPA's failure to appropriately evaluate volcanic emissions is arbitrary and capricious and must be addressed in the final FIP. However, the commenters recognized that EPA may opt not to revise its reasonable progress analysis in this way during this planning period. In that event, the commenters requested that EPA commit to addressing Kilauea's emissions in the next planning period because continuing to exclude these emissions that are the dominant cause of visibility impairment would create untenable results—increasingly expensive controls in successive planning periods that would not result in perceptible improvements in visibility.

Another commenter (Earthjustice) stated that the goal of the haze program is to eliminate visibility impairment “from manmade air pollution” [citing 42 U.S.C. 7491(a)(1)]. According to the commenter, failing to include the volcano in natural conditions distorts the analysis of the impacts from human sources and the corresponding BART controls and reasonable progress goals to achieve natural visibility conditions. The commenter asserted that based on this “skewed analysis,” EPA summarily eliminated any controls for NOXfor the BART analysis and reasonable progress goals. The commenter contended that EPA avoided evaluating the actual URP for anthropogenic SO2pollution; instead, rejecting a URP inflated by volcano impacts (which the commenter termed a “strawman of EPA's own making”), then proposing arbitrary progress goals of its own choosing. The commenter indicated that EPA's approach toward volcano conditions is unjustified and prevents the Agency from providing a rational and transparent justification for its pollution control determinations. According to the commenter, this approach also deprives the public of proper notice and opportunity to comment; in the commenter's view, EPA must rationally review and address impacts from human sources unskewed by volcano impacts and allow the public a meaningful opportunity to review and comment on such determinations.

The fourth commenter (HC&S) pointed out that under EPA's methodology, the URP incorporates reductions in visibility impairment that are sufficient to offset both the portion of baseline impairment that comes from anthropogenic emissions and the portion that is caused by the volcano. The commenter believes that to make a more accurate assessment of the reduction in emissions from anthropogenic sources necessary to achieve natural visibility conditions, emissions from Kilauea either need to be included in, or excluded from, both the estimate of baseline visibility conditions and the estimate of natural visibility conditions. The commenter recommended that EPA adopt the Hawaii DOH's proposed method to adjust the baseline visibility impairment to account for the impacts of volcano emissions as well as for the impacts of Asian dust. According to the commenter, under this approach, the URP target for 2018 would be 0.32 deciviews (dv), which is only slightly greater than what would be achieved through the proposed FIP.

Response:The central concern of these comments appears to be that the approach EPA used to determine the uniform rate of progress (URP), in particular how we considered volcanic emissions, led to inappropriate regulatory decisions in the proposal and/or may lead to inappropriate regulatory decisions in the future. EPA disagrees with this concern. The commenters mistakenly conclude that the URP sets a target or goal for the first planning period. In fact, the development of the URP is an analytical exercise that is intended to inform the setting of reasonable progress goals (RPGs) rather than a standard or presumptive target for the plan to meet.

In establishing RPGs, the states and EPA must “consider” both the URP and the emission reduction measures needed to achieve the URP.4 More specifically, EPA has recommended that states use the following approach in setting their RPGs:

4In addition, as noted in our proposal, CAA section 169A and the RHR at 40 CFR 51.308(d)(1)(i)(A) require consideration of the following four factors in determining “reasonable progress”: (1) The costs of compliance; (2) the time necessary for compliance; (3) the energy and non-air quality environmental impacts of compliance; and (4) the remaining useful life of any potentially affected sources. The weighing of these four factors is sometimes referred to as a “four-factor analysis” to distinguish it from the “five-factor analysis” for BART determinations. Comments concerning the URP and related issues are addressed in this section. Other comments on our RP analysis are addressed below.

1. Establish baseline and natural visibility conditions.

2. Determine the URP (i.e., a straight line between baseline visibility in 2000-2004 for the worst 20 percent days and projected natural conditions for the worst 20 percent days in 2064).

3. Identify and analyze the measures aimed at achieving the URP.

a. Identify the key pollutants and sources and/or source categories that are contributing to visibility impairment at each Class I area.

b. Identify the control measures and associated emission reductions that are expected to result from compliance with existing rules and other available measures for the sources and source categories that contribute significantly to visibility impairment.

c. Determine what additional control measures would be reasonable based on the statutory factors and other relevant factors for the sources and/or source categories identified.

d. Estimate through the use of air quality models the improvement in visibility that would result from implementation of the control measures found to be reasonable and compare this to the URP.

4. Establish an RPG.

In this case, the commenters' concerns relate primarily to how EPA performed step 1 of this analysis. Specifically, the commenters object to EPA's inclusion of volcanic emissions in the baseline and exclusion of volcanic emissions in our estimate of natural conditions. EPA acknowledges the commenters' concerns, but does not agree that our approach is arbitrary or unjustified in this case. Rather, we have followed the statutory and regulatory requirements for Reasonable Progress analyses, while also accounting for unique circumstances in Hawaii that severely limit the utility of the URP as an analytical tool for setting RPGs for the state's Class I areas.

Pursuant to 40 CFR 51.308(d)(2), baseline visibility conditions must be calculated using actual monitoring data from 2000-2004. Therefore, the baseline conditions for Class I areas in Hawaii necessarily include volcanic emissions. It is difficult to include volcanic emissions as part of natural background visibility in 2064 because of the extreme variability in volcanic emissions from year to year. In this case, a 2064 projection would be little better than a guess. Therefore, in estimating natural conditions for purposes of this first planning period, we have not attempted to forecast the future contribution of the volcano to natural background visibility. Even if we could quantitatively estimate “natural” volcanic emissions and air quality effects in 2064 with any accuracy, the URP would be of very limited value in setting RPGs for Hawaii.

As explained in EPA's Reasonable Progress Guidance, the URP is intended to serve as a gauge against which to measure the improvement in visibility conditions that is projected to resultfrom implementation of reasonable control measures during the first planning period which ends in 2018.5 However, the variability of volcanic emissions from Kilauea renders this type of analysis unhelpful for Hawaii's Class I areas. To understand why this is the case, it helps to look at Figure II.B-6 in EPA's technical support document (TSD). This figure shows the URP calculation for Hawaii Volcanoes National Park (NP). The points on the left side of the figure are the actual, measured visibility impairment at Hawaii Volcanoes for the past several years; these measurements vary by at least 13 dv, as compared to the difference between baseline conditions and natural conditions of 11.7 dv. This dramatic variation in visibility impairment on the worst 20 percent days is driven by the extreme variability of the volcanic emissions, which dominate visibility impairment on those days. Thus, the only way EPA could accurately estimate the improvement in visibility on the worst 20 percent days by 2018 is if we could accurately predict volcanic emissions on those days. In the absence of an accurate projection of volcanic emissions for 2018, there is no reasonable estimate of visibility conditions in 2018 to compare with the URP. Therefore, EPA has used a different method of gauging reasonable progress for this first planning period, as explained in Section F of the proposal, “Reasonable Progress Goals for Hawaii.”6

5See Section 1-7 of “Guidance for Tracking Progress Under the Regional Haze Rule”, Document No. EPA-R09-OAR-2012-0345-0003-B10.

677 FR 31707, May 29, 2012

However, given the dominance of volcanic emissions on the worst 20 percent days in Hawaii, it may be appropriate for future plans to focus on other days when the proportion of anthropogenic contribution to visibility impairment is larger. We expect that the State of Hawaii will develop future regional haze plans, consistent with the CAA and EPA's implementing regulations. We plan to work with the State of Hawaii on those future plans and we will consider different approaches to gauging reasonable progress, and different approaches to determining the URP.

2. Estimating Natural Visibility Conditions

Comment:One commenter (NPS) noted that emissions from the Kilauea Volcano vary from year to year, making it difficult to project future emissions levels or the specific contribution of these emissions to visibility impairment in 2018 or 2064. For clarity, the commenter recommended that EPA revise the conclusion in section III.B.1 of the preamble to the proposed FIP (77 FR 31699, May 29, 2012) to read “* * * in estimating natural conditions for purposes of this first planning period, we have not tried to forecast the future contribution of the volcano to natural background visibility” rather than stating an assumption that there will be no visibility impact from the volcano.

Response:The NPS is correct in saying that EPA did not attempt to forecast the future contribution of the volcano to natural background visibility. However, since the default natural conditions do not include volcanic emissions, we implicitly assumed that there would be no visibility impact from the volcano in our URP analysis. EPA does not consider this implicit assumption to be problematic because the URP analysis is not useful in the case of Hawaii due to the infeasibility of accurately accounting for volcanic emissions in the 2018 projections (see Section II.A.1. of this notice).

We would consider a refined estimate of natural conditions at these Class I areas if the State of Hawaii were to propose such a change as part of the next Regional Haze plan for Hawaii. Any such estimate would need to be consistent with our guidance on this subject.7

7See “Guidance for Estimating Natural Visibility Conditions under the Regional Haze Rule” Document No. EPA-R09-OAR-2012-0345-0003-B9.

3. Contribution Assessment According to IMPROVE Monitoring Data

Comment:One commenter (NPS) generally agreed with EPA's assessment of contributions to visibility impairment.

Response:EPA appreciates NPS' support of our contribution assessment, given their extensive expertise in this subject.

4. Impact of Fugitive Dust on Visibility Impairment in Hawaii Class I Areas

Comment:One commenter (Parsons) stated that EPA is incorrect in stating that there are no impacts or degradations in visibility to Haleakala NP as a result of fugitive dust. According to the commenter, EPA did not examine the impacts of particulate matter carried into the atmosphere from Maui's agricultural fields, which affects air quality on many days. The commenter asserted that Maui is subjected to strong trade winds on many days, and plantation practices of clearing and tilling hundreds of acres at a time means that tons of windborne topsoil are lost each year. The commenter believes that best management practices might help mitigate this loss, and preserve Maui's air quality, but plantations are exempt from the sort of requirements that would be applied to other land-altering activities, such as construction site grading. The commenter suggested that EPA may be able to work with the Hawaii Department of Agriculture and DOH to revise standards for dust control in order to protect the health and welfare of the community and near-shore coral reef ecosystems, and to help mitigate impacts that contribute to regional haze.

Two commenters similarly asserted that fugitive dust from the sugarcane fields affects the haze in Haleakala NP and is killing Maui's coral reefs. The commenter indicated that after harvest, the cane fields are left bare and the loose topsoil is picked up by the trade winds and carried across the island, coating everything in its path and eventually settling on and killing the coral reefs south of Maui.

Response:EPA disagrees with the commenter's assertion that we did not consider the impact of dust from agricultural activities when evaluating causes of haze at Hawaii's national parks. Dust from agricultural activities and other sources is measured at the IMPROVE monitors as coarse mass and soil. Section II.A.3. of the TSD discusses causes of haze at Haleakala NP. Section II.B.3 discusses the causes of haze at Hawaii Volcanoes NP. Both of these sections of the TSD address the contribution of coarse mass and soil to visibility impairment on the best and worst days. Coarse mass contributes about 9 percent to visibility impairment on the worst 20 percent days at Haleakala.8 The source of the coarse mass measured at the IMPROVE site is unclear. It could be dust from the low elevations transported up to the park, or it could be from nearby sources such as unpaved roads.

8See “Technical Support Document for the Proposed Action on the Federal Implementation Plan for the Regional Haze Program in the State of Hawaii, Air Division, U.S. EPA Region 9”, [hereinafter TSD] p. 12, Document No. EPA-R09-OAR-2012-0345-0003-A3.

EPA shares the commenters' concerns about the impact of dust emissions on public health, the loss of topsoil and possible impacts to water quality and marine life. However, in the context of this rulemaking, EPA does not consider it reasonable to require additional pollution control without clear evidence that the dust is causing or contributing to haze at the Class I area. Further analysis of the source of this coarsemass should be conducted as part of the reasonable progress review for the next planning period.

5. Subject to BART Analysis

Comment 1: Agreement with analysis to identify sources subject to BART.

Three commenters (HC&S, HELCO, and MECO) agreed with EPA's analysis to determine which sources should be subject to BART requirements. Their comments are summarized in the following paragraphs.

Two of the commenters (HELCO, MECO) noted that CAA section 169A(b)(2)(A) requires the relevant regulatory agency to review a state's BART eligible sources and determine whether they emit “any air pollutant which may reasonably be anticipated to cause or contribute to any impairment of visibility in [a Class I] area.” These commenters cited the BART Guidelines (70 FR 39109, July 6, 2005) to add that if a source does not meet this threshold, it may be exempt from further BART review. Based on these principles, the commenters believe that EPA's analysis of which sources in Hawaii should be subject to BART is sound and consistent with the BART Guidelines, and they urged EPA to retain it for the final FIP.

These two commenters stated that the BART Guidelines provide regulatory agencies with three options for making a “cause or contribute” finding and that EPA reasonably chose to use an “individual source attribution approach” and a threshold of 0.5 dv in this case. According to the commenters, the BART Guidelines explain that the appropriate contribution threshold depends on the number and proximity of sources affecting a Class I area, and a threshold lower than 0.5 dv is justified where there are a large number of BART-eligible sources within the state and in proximity to a Class I area. The commenters added that in Hawaii there are few BART-eligible sources and they are not concentrated near a single Class I area. On this basis, one of the commenters (MECO) explicitly expressed agreement that 0.5 dv is the appropriate threshold.

These two commenters went on to note that, consistent with the BART Guidelines, EPA applied the 0.5 dv contribution threshold to the results of computer modeling that was used to predict visibility impacts from each BART-eligible source in Hawaii, with the result that six of the eight BART-eligible sources fell below the 0.5 dv contribution threshold.9 The commenters agreed that EPA appropriately determined that only HELCO's Kanoelehua Hill Generating Station is subject to BART and exempted all other BART-eligible sources in Hawaii from further BART review. One of the commenters (HELCO) specifically stated that this analysis correctly excluded Hawaiian Electric Company's (HECO's) Waiau and Kahe facilities, and the other (MECO) stated that MECO's Kahului facility was appropriately excluded.

9Six of the eight BART-eligible sources had a less than 0.5 deciview impact and so were exempted from BART. One of the remaining facilities, Hu Honua Bioenergy is no longer permitted to burn fossil fuels and is therefore also exempt from BART. This leaves one facility in Hawaii as subject to BART, the Kaneolehua Hill facility. See 77 FR 31704, 31705.

The third commenter (HC&S) also agreed with the proposed threshold (0.5 dv) used to assess whether the impact of a single source contributes to visibility impairment at the Hawaiian Class I areas. This commenter pointed out that of the six sources in Hawaii exempted from BART because their modeled impact is below 0.5 dv, none has a modeled impact of as much as half of this threshold level. The commenter also stated that the combined impact from all six sources is 0.715 dv at Haleakala NP, which is lower than the level (1.0 dv) at which the BART Guidelines consider a single source to “cause” visibility impairment. The commenter added that while this combined impact is somewhat higher than the proposed 0.5 dv contribution threshold, the BART Guidelines explicitly caution that visibility effects of multiple sources should not be aggregated and their collective effects compared against the contribution threshold, because this would inappropriately create a “contribution to contribution” test. The commenter concluded that it is therefore reasonable to conclude that these six sources do not cause or contribute to visibility impairment at Haleakala NP. The commenter also asserted that since the combined visibility impact of these six sources at the Hawaii Volcanoes NP totals 0.35 dv, it is reasonable to conclude that these sources do not cause or contribute to visibility impairment at Hawaii Volcanoes NP. The commenter noted that setting the contribution threshold at 0.5 dv for sources subject to BART will capture those BART-eligible sources responsible for more than half of the visibility impacts at Haleakala NP and nearly 90 percent of the visibility impacts at Hawaii Volcanoes NP while still excluding other sources with very small impacts. The commenter believes that given the relatively small number of BART-eligible sources potentially impacting Class I areas in Hawaii and the magnitude of both the individual sources' impacts and the combined impact from all sources proposed to be exempt from BART, the proposed 0.5 dv threshold for determining whether a single source contributes to visibility impairment is wholly consistent with the BART Guidelines and is therefore an appropriate threshold for use in the Hawaii FIP.

Response 1:EPA agrees that the determination of sources subject to BART was conducted appropriately and in accordance with the applicable regulatory requirements and guidance. EPA also agrees that the 0.5 dv threshold is appropriate for Hawaii.

Comment 2: Disagreement with part of the analysis.

Two commenters (FHNP, Parsons) disagreed with some aspects of EPA's analysis to determine which sources should be subject to BART requirements.

One commenter (FHNP) suggested that the MECO Kahului and HECO Kahe facilities should not be exempted from BART analysis. The commenter noted that EPA used a modeled increase of 0.5 dv at the Haleakala IMPROVE monitoring site (HALE) to identify candidates for BART analysis, and that the measured concentrations of pollutants taken at the Haleakala Crater monitoring site (HACR) are approximately half of those measured at HALE. According to the commenter, it is expected that the point sources analyzed in this report would contribute similar densities of non-anthropologic elements at both the HALE and the HACR sites (with the exception of some smoke sources) and that, hence, a point source modeled to produce a 0.25 dv change at HALE would be expected to produce an approximate 0.5 dv change at HACR in the Class I area. The commenter pointed out that the MECO Kahului site and the HECO Kahe site were modeled to produce changes of 0.232 dv and 0.221 dv, respectively, at HALE. The commenter believes that extrapolating these contributions to the HACR site suggests that these sources are very close to contributing 0.5 dv in the Class I area. The commenter concluded that since actions recommended in the report are projected to produce less than the target rate of progress, the MECO Kahului and HECO Kahe sites should not be exempted from BART analysis.

The second commenter (Parsons) objected to EPA's omission of the Kahului facility from BART requirements. The commenter stated that air emissions from MECO's Maalaeaand Kahului facilities rank them as the fifth and seventh worst polluters in Hawaii. The commenter noted that EPA has not proposed additional pollution controls at either facility. According to the commenter, at the public hearing on May 31, 2012, EPA stated the belief that the Kahului facility will cease operations by 2018. The commenter asserted that this is conjectural and indicated that it would be prudent to apply more stringent pollution control standards to this facility, especially in the short term.

Response 2:We do not agree that the Kahului or Kahe facilities should be subject to BART.

As an initial matter, we would like to clarify that the modeling upon which we have based our subject-to-BART determinations did not use either the Haleakala (HALE) IMPROVE monitoring site or the Haleakala Crater (HACR) site as a receptor.10 Rather the subject-to-BART modeling predicted visibility impacts at gridded receptor locations spaced approximately one kilometer apart within the Class I area domain.11 Therefore, the modeled impacts cited by the commenter (i.e. 0.232 dv for Kahului and 0.221 dv for Kahe) represent the 8th highest delta-deciview values for the year modeled (2005) from all modeled receptors at Halekala National Park and do not reflect modeled impacts at either HALE or HACR.12

10See “Subject-to-Best Available Retrofit Technology (BART) Modeling for the State of Hawaii, Application of the CALPUFF Modeling System; Prepared for: Hawaii State Department of Health, Environmental Management Division Clean Air Branch by Alpine Geophysics, LLC (March 3, 2010)”, Document No. EPA-R09-OAR-2012-0345-0006-C3d.

11 Id.Section 3.2.4. These receptor locations were provided by the National Park Service and are available at

12The Hawaii BART/RP Supplemental Modeling Results Report does include modeling results at individual receptors placed at the location of the Haleakala (HALE) IMPROVE monitoring site and Haleakala Crater (HACR) site. See Hawaii BART/RP Supplemental Modeling Results, Alpine Geophysics (March 29, 2010), (Document No. EPA-R09-OAR-2012-0345-0011-Attachment 1) Table 12. The predicted visibility impacts from the Kahului and Kahe BART-eligible sources at the HALE and HACR receptors were similar to predicted visibility impacts at the NPS receptors, and were below the 0.5 deciview threshold for all receptors. Specifically, the modeled 98th percentile delta deciview impact from the BART-eligible units at Kahului was 0.227 at HALE and 0.247 at HACR.Id.Table 15. The modeled 98th percentile delta deciview impact from BART-eligible units at Kahe was 0.262 at HALE and 0.255 at HACR. Id.Table 15. Therefore, even if we had used HACR as a receptor for purposes of subject-to-BART modeling, neither the Kahe nor the Kahului facility would have been found to be subject-to-BART.

To the extent that the commenter is arguing that the subject-to-BART modeling should have used background conditions for HACR rather than HALE, we also disagree. Consistent with the BART Guidelines, the subject-to-BART modeling for Hawaii was performed against natural visibility conditions.13 Natural conditions have not yet been established for the HACR site. Therefore, EPA reasonably relied on the available information regarding natural conditions at the HALE site for purposes of conducting subject-to-BART monitoring.

13Specifically the modeling was performed against natural visibility baseline conditions for the best 20% of days. Use of either the best 20% days or average natural conditions is permissible under the BART Guidelines. See Memo from Joseph W. Paisie regarding Regional Haze Regulations and Guidelines for BART (July 19, 2006) (Document No. EPA-R09-OAR-2012-0345-0003-B15). However, use of the 20% best days is more conservative (i.e., it tends to increase the baseline impacts for a given source).

With respect to measuredpollutant concentrations, the commenter (FHNP) correctly notes that the values of the measured concentration of pollutants taken at the HACR site are smaller than those measured at the HALE site (i.e., the HACR site is “cleaner”). The commenter suggests that, therefore, a point source modeled to produce a 0.25 dv change at HALE would be expected to produce an approximate 0.5 dv change at HACR, and hence a 0.5 dv change in the Class 1 area. However, as noted by the commenter, doubling the source impact is a rough approximation of the effect of reducing the background light extinction by half. The effect of reducing the background extinction by half on the change in deciviews (delta dv) varies depending on the source extinction bext(source) and the background extinction, bext(bkg), but would be smaller than doubling the source impact. Therefore, the rough approximation proposed by the commenter (doubling the source impact) to estimate the potential change in visibility impacts from the facilities from using the new HACR site to calculate background light extinction would not be appropriate.14 So, even if natural conditions for the HACR site had been available, we do not expect that the impact of the sources would approximate 0.5 dv. EPA therefore disagrees that these sources should be subject-to-BART. Nonetheless, because Kahului and Kahe are both significant sources of pollution and include non-BART-eligible units as well as BART-eligible units, they may be appropriate candidates for controls as part of future Regional Haze plans.

14EPA does not believe doubling the source impact is appropriate. However, EPA notes that doubling the source impact of 0.23 deciviews and 0.22 deciviews would result in values below the 0.5 dv threshold.

We also disagree with the commenter that the URP is relevant to whether Kahului and Kahe are subject to BART. Under the Regional Haze Rule (RHR), the determination of which sources are subject-to-BART is a separate analysis from the calculation of the URP and the setting of RPGs.15 Moreover, as discussed in Section II.A.1 of this document, the URP is not a target and is particularly poorly suited for regulatory decisions in Hawaii.

15Compare 40 CFR 51.308(d)(1) and 51.308(e).

Regarding the comments from Parsons, EPA agrees that the electric power plants Maalaea and Kahului are relatively large sources of pollution. However, as noted above, the modeled 98th percentile visibility impact of the BART-eligible Kahului source is 0.23 dv, less than one-half of the 0.5 dv subject-to-BART threshold. Due to the age of its equipment, the Maalaea power plant does not have BART-eligible units and therefore is not subject-to-BART.

EPA disagrees that we represented at the hearing in Maui that the Kahului Power Plant would no longer be operating in 2018. That assertion is not supported by the transcript.16 Regardless, we did not base our decision on BART for the Kahului plant on future operation, but instead based it on the current emissions level for the facility.

16See transcript of Kahului hearing, Document No. EPA-R09-OAR-2012-0345-0022.

Comment 3: Puunene Mill.

One commenter (HC&S) stated that the small contribution to visibility impairment from the Puunene Mill warrants a determination that the facility should not be subject to BART. While conceding that it was reasonable to use maximum actual 24-hour emissions to model worst-case visibility impacts, the commenter indicated that typical visibility impacts from the Puunene Mill are likely to be lower than the modeled results. The commenter noted that even so, modeling results for both coal and bagasse firing showed that the impact of the facility was well below the 0.5 dv contribution threshold at both Haleakala NP and Hawaii Volcanoes NP, at both the maximum 24-hour 98th percentile impact and the highest modeled impact. According to the commenter, the highest modeled impact for the facility (i.e., during coal firing) was less than half the contribution threshold at Haleakala NP and less than 20 percent of the threshold at Hawaii Volcanoes NP. The commenter added that modeling of the combined impacts of both BART-eligible (Boiler 3) and Reasonable Progress-eligible (Boilers 1and 2) sources at the Puunene Mill demonstrated that the maximum 24-hour 98th percentile visibility impacts from the facility during both bagasse firing and coal firing scenarios are well below the 0.5 dv contribution threshold. The commenter believes that the modeling analysis clearly shows that even worst-case emissions from the Puunene Mill do not cause or contribute to visibility impairment at either Haleakala NP or Hawaii Volcanoes NP, and that additional controls are therefore not warranted.

In contrast, one commenter (Parsons) believes that with regard to the Hawaii Regional Haze FIP and for public health concerns, air emissions at HC&S's Puunene Mill should be subjected to BART determinations, Maximum Achievable Control Technology (MACT) Hammer standards and both continuous opacity monitoring systems (COMS) and continuous emissions monitoring systems (CEMS) guidelines. The commenter stated that the Puunene Mill is the second worst polluter in Hawaii with regard to air emissions. The commenter indicated that Boilers 1 and 2 at the facility predate the Act, and thus have been exempt from those standards for decades. The commenter also contended that EPA's “revised MACT Hammer provisions” have not been applied to these units because HC&S and sugar growers in Florida and Texas submitted a report replacing these emission limits with their own subcategory of bagasse-fired boilers. The commenter added that HC&S combusts 100,000 tons of coal annually without emission standards that apply to other coal-burning facilities. The commenter also stated that Boiler 3 at the facility has not been held to Federal standards required for COMS and CEMS or regulatory oversight by the Hawaii DOH.

Two other commenters also stated that EPA should include all of the HC&S smoke stacks in its review. In particular, the commenter asked that EPA review and closely monitor the electric power production emissions on the Puunene Mill.

Response 3:We reaffirm that the Puunene Mill is not subject-to-BART. In accordance with the BART Guidelines, the subject-to-BART modeling for Puunene Mill was performed using worst-case emissions from the Mill and best-case visibility (under natural conditions) at the parks.17 This analysis assumed the Mill was powered entirely by coal, its most polluting fuel, for 24 hours. That worst-case 24-hour emission rate was then compared to the clearest days at Haleakala NP and Hawaii Volcanoes NP. This comparison of very high emissions at the Puunene Mill to very clean conditions at the park was modeled for every weather condition during the 365 days of the year. The resulting visibility impact was less than the 0.5 dv threshold that would make the facility subject to BART.18 EPA reviewed this analysis and concurs with the results.

17See Section 2.4 of “Subject-to-Best Available Retrofit Technology (BART) Modeling for the State of Hawaii, Application of the CALPUFF Modeling System; Prepared for: Hawaii State Department of Health, Environmental Management Division Clean Air Branch by Alpine Geophysics, LLC (March 3, 2010)”, Document No. EPA-R09-OAR-2012-0345-0006-C3d.

18Id. Table 6.

The commenter indicates that he believes the Mill should be subject to Federal guidelines for CEMS and COMS. We are confident that the methods used to calculate worst-case emissions are appropriate and conservative. Therefore, the absence of continuous monitors does not weaken the analysis.

The commenter also indicates that he believes the Puunene Mill should be subject to MACT controls. The applicability of MACT is outside the scope of this rulemaking.

Comment 4: Hu Honua and Tradewinds should be subject to BART controls.

One commenter (Chytilo) objected to the exclusion of the Hu Honua Bioenergy Facility and the Tradewinds Veneer Mill and Cogeneration Facility's electric generating EGUs from the proposed FIP. The commenter argued that these facilities should be subjected to BART controls and emissions limits. The commenter stated that even though neither source was operating during the baseline period, the emissions from each source are significant and will interfere with progress towards national visibility objectives.

The commenter also asserted that EPA improperly exempted these sources from controls, reporting and reasonable further progress based on what the commenter believes is the irrelevant and incorrect belief that each source is entirely biofueled. The commenter stated that emissions controls will be less successful for these facilities because steady state operations are more difficult to achieve and the operators contemplate diurnal fuel source changes and other operational shifts daily. The commenter added that biofueled sources still cause visibility impairment, and alleged that the FIP rulemaking offers no explanation for why biofueled sources should be exempted from haze controls. The commenter indicated that the two facilities are permitted to burn wood waste which, according to the commenter, is a variable fuel that actually produces increased emissions and should be subject to enhanced controls. The commenter believes that haze objectives cannot be met if these sources are exempted.

The commenter made the following additional points related to the Hu Honua facility:

• The emissions calculations for the Hu Honua facility are questionable. The commenter expressed agreement with EPA's comments on the facility's Covered Source Permit (Hawaii's term for a title V permit), which the commenter characterized as saying that unrealistic emissions factors were used and actual plant emissions are likely to be considerably higher.

• The facility's permit allows the use of conventional diesel fuel during startup and off-peak periods, so any exemption for biofuels is not warranted.

• Sulfur oxides (SOX) emissions are not insignificant. The commenter asserted that SOXemissions from the facility “constitute nearly 93 percent of the * * * NAAQS,” and that EPA's rationale that these emissions may be ignored due to background volcanic emissions is misplaced. The commenter stated that the Pepe'ekeo area is only affected by volcanic emissions during certain wind conditions, and during other periods the facility's SOXand other visibility-impairing emissions will be significant and should not be exempted.

Response 4:The definition of BART-eligible facility may be found in 40 CFR 51.301. It provides a list of types of facilities that may be eligible for BART if they were built between 1962 and 1977. These types of facilities include “fossil-fuel fired steam electric plants of more than 250 million British thermal units per hour heat input.” When Hu Honua converted to biofuels, it was no longer “fossil-fuel fired” and therefore was no longer BART-eligible. The permit for this facility allows for it to be fueled by wood or biodiesel. Neither of these is a “fossil fuel.” Therefore, the facility is not eligible for BART. The Tradewinds Veneer cogeneration facility was not built between 1962 and 1977 and does not burn fossil fuels. Therefore, this facility is also not eligible for BART.

We note that the commenter's general concern about emissions from new large facilities possibly interfering with visibility goals is a common consideration in air quality planning and is not limited to these two facilities. Such emissions are regulated in large part under the CAA's Prevention ofSignificant Deterioration (PSD) permitting program, which applies to new major sources and major modifications at existing sources for pollutants where the area in which the source is located has been designated attainment or unclassifiable with one or more of the NAAQS. Among other requirements, PSD review requires an air quality analysis, using dispersion modeling, of ambient concentrations that would result from the applicant's proposed project. The PSD regulations provide special protection of Air Quality Related Values, including visibility, in Class I areas, including oversight by and coordination between the permitting authority and the Federal land managers (FLMs). The RH rule also requires reviews of plans every 5 years and complete new regional haze plans every 10 years. The 5-year update of this plan will include a verification that emission trends on the Islands are consistent with reasonable progress and an analysis of whether anthropogenic visibility impairment is decreasing The next full plan, required in 10 years, will include a new reasonable progress analysis that would take into account these and any other new sources of pollution.

6. BART Determination for Kanoelehua Hill

Comment 1: General comments on BART for Hill.

One commenter (Earthjustice) stated that EPA's proposal to exempt the Hill facility from any BART controls neither meets the requirements of the BART program, nor promotes necessary visibility improvements at Hawaii's Class I areas. The commenter pointed out that even though EPA has stated that this facility is by far the largest source of anthropogenic SO2emissions on the Big Island (citing 77 FR 31706), no BART control is required. The commenter believes that this result lacks a reasoned and lawful justification. The commenter asserted that EPA must require demonstrably cost-effective controls, rather than readily relieving polluters of these obligations. The specific arguments made by the commenter related to nitrogen oxides (NOX) and SO2are detailed in the following subsections.

In contrast, another commenter (HELCO) agreed that EPA appropriately determined that BART controls are not justified for the Hill facility. The commenter noted that EPA's analysis of the five statutory factors that must be considered in establishing BART was based largely on an analysis performed by a consultant for HELCO (“the Trinity BART report”19 ) which, according to the commenter, was consistent with the BART Guidelines even though the guidelines are not mandatory for a facility the size of the Hill facility. However, because the BART Guidelines were designed for power plants, the commenter believes they are both an apt and a conservative guide in this instance.

19See “BART Five-Factor Analysis Prepared for Hawaiian Electric Light Company, Trinity Consultants” (April 12, 2010), Document No. EPA-R09-OAR-2012-0345-0010-attachment 3, [hereinafter “Trinity BART Report”].

Response 1:EPA disagrees with Earthjustice's comments with respect to our BART determination for the Hill facility. The EPA BART determination is appropriate for Hill for reasons detailed below. And, it is important to note that we are requiring SO2reductions from Hill and two other plants on the Big Island in order to ensure reasonable progress toward eliminating anthropogenic visibility impairment at Hawaii Volcanoes NP. EPA agrees with HELCO's comment that the BART analysis conducted by Trinity for HELCO was consistent with the BART guidelines, although EPA does not agree with the company's cost estimates for lower sulfur fuels.

Comment 2: HELCO's comments on particulate matter (PM) and NO X.

One commenter (HELCO) agreed with EPA's determination that the Hill facility should not install NOXor PM BART. The commenter stated that the Trinity BART report evaluated the available control technologies and that EPA, based on that report, found that the controls considered for PM would not be cost effective and that the controls considered for NOXwould not provide a measurable visibility benefit at Haleakala NP or Hawaii Volcanoes NP. Given what the commenter characterized as the high cost of controls and low degree of improvement in visibility that might result from controls, the commenter supports EPA's determination that “no control for NOXand PM at the Hill Plant is consistent with BART” (citing 77 FR 31706).

Response 2:EPA agrees that the existing emission levels of PM and NOXfrom Hill are consistent with BART, given the unique conditions in Hawaii.

Comment 3: Earthjustice comments on NO X.

One commenter (Earthjustice) stated that EPA proposed no control as BART for NOXeven though HELCO admitted that the control option of low-NOXburners (LNB) is cost effective and proposed them as BART (citing the Trinity BART report, p. 5-11). According to the commenter, EPA reached this conclusion based on the rationale that “due to the overwhelming contribution of sulfate to visibility impairment at the nearby Hawaii Volcanoes Class I area, it is unlikely that reductions in NOXwould have a measurable impact on visibility at that area” (citing 77 FR 31705). However, as detailed in section II.A.1., the commenter believes that EPA inflated the impact of sulfate by including the natural contributions of the Kïlauea Volcano in baseline conditions but not in natural conditions. The commenter believes that this approach ignores the goal of the haze program of controlling anthropogenic visibility impairment. The commenter asserted that EPA cannot justify dismissing a pollution control that the utility already acknowledged as BART by burying it within the background impact of the volcano.

The commenter also stated that EPA summarily dismisses post-combustion controls such as selective catalytic combustion (SCR) because “they were not found to be cost effective” in the Trinity BART report (citing 77 FR 31706). However, according to the commenter, that report showed that the cost effectiveness of SCR for Hill falls within the range established in EPA and state BART determinations. The commenter quoted the Trinity BART report as including SCR costs of $2,600 and $2,200/ton for the units at the Hill facility, while EPA's proposal for BART at the Four Corners Power Plant considered cost estimates of $4,887 to $6,170/ton to be cost-effective (citing 75 FR 64227, October 19, 2010) and states have established thresholds for cost-effectiveness such as $7,300 (Oregon), $7,000 to $10,000 (Wisconsin), $5,946 to $7,398 (New Mexico), and $5,500 (New York).

The commenter believes that, in any event, EPA has no basis for eliminating BART controls without engaging in the statutorily mandated five-factor BART analysis. According to the commenter, EPA simply waived any analysis, and any pollution reduction benefit, based on speculation. The commenter alleged that proper inquiry would confirm, for example, that LNB would prove much more effective at controlling NOXthan the relatively high figures the utility cited.

Response 3:EPA disagrees with the commenter's assertion that a full five-factor analysis was not conducted for NOXcontrols at Hill. The Trinity BART report contains a complete five-factor analysis of NOXcontrols at Hill, which is consistent with EPA requirementsand guidance.20 We have relied in large part on that analysis in conducting our own five-factor BART analysis for NOXat Hill. As noted by Earthjustice, the Trinity BART Report found low-NOXburners to be cost-effective at Hill.21 The Trinity BART Report also estimated that installation of LNB would result in an improvement of 0.21 dv at Hawaii Volcanoes and 0.02 dv at Haleakala (based on the 98th percentile impacts).22 However, the Trinity BART Report noted this projection “does not reflect reality” because it “relies on an approach for establishing natural conditions that does not consider the local volcanic activity.”23 In other words, the actual visibility benefit of LNB will be significantly less than 0.21 dv, due to the impact of volcanic emissions. Taking this fact into account, EPA concluded that the costs of LNB were not justified by the visibility benefit that would actually result from installation of controls.

20See Trinity BART Report Chapter 5.

21See Trinity BART Report at