                 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                                  REGION III
	1650 Arch Street
	Philadelphia, Pennsylvania  19103

DATE:	
 
SUBJECT:	Technical Support Document for the Revisions to the Attainment Plans for the Commonwealth of Virginia Portion of the Washington, DC-MD-VA 1990 1-Hour and 1997 8-Hour Ozone Nonattainment Areas and the Maintenance Plan for the Fredericksburg 1997 8-Hour Ozone Maintenance Area to Remove the Stage II Vapor Recovery Program

FROM:	Asrah Khadr, Environmental Engineer, EIT		/s/
 Office of Air Program Planning

TO:	File		

THRU: 		Cristina Fernandez, Associate Director		/s/
Office of Air Program Planning


   I. Introduction
	
On March 18, 2014, the Commonwealth of Virginia (Virginia) submitted revisions to the Virginia State Implementation Plan (SIP) through the Virginia Department of Environmental Quality (VADEQ) to amend the SIP-approved ozone attainment plans for the Virginia portion of the Washington, DC-MD-VA 1990 1-Hour and 1997 8-Hour Ozone National Ambient Air Quality Standard (NAAQS) Nonattainment Areas (Northern Virginia Areas) and the maintenance plan for the Fredericksburg 1997 8-Hour ouHour fgbfOzone NAAQS Maintenance Area (Fredericksburg Area) (collectively referred to as the Virginia Areas or Areas).  The revisions also address the impact on the attainment/maintenance plans due to the removal of Virginia's Stage II vapor recovery equipment from subject gasoline dispensing facilities in the subject areas.  The SIP revisions contain a demonstration that removal of the Stage II vapor recovery program will not interfere with attainment or maintenance of the NAAQS in accordance with section 110(l) of the Clean Air Act (CAA).  The analysis in these SIP revisions also contains a demonstration that meets the requirements of section 184(b)(2) of the CAA which requires areas in the Ozone Transport Region (OTR) to implement the Stage II vapor recovery program or a comparable measure. 


   II. Background 
   
Stage II Vapor Recovery and Onboard Vapor Recovery Systems

Stage II vapor recovery (hereafter referred to as Stage II) is a means of capturing gasoline vapors displaced during transfer of gasoline from the gasoline dispensing unit to the motor vehicle fuel tank during vehicle refueling at a gasoline dispensing facility (GDF).  Stage II vapor recovery involves use of special refueling nozzles and coaxial hoses for vapor collection at each gasoline pump at a subject GDF.  Gasoline vapors belong to a class of pollutants known as volatile organic compounds (VOC).  Gasoline is made up of a number of toxic compounds, and gasoline dispensing operations emit gasoline vapor that contains air toxics (also known as "hazardous air pollutants") that can cause cancer and other serious health effects.  These include benzene, 1,3-butadiene, formaldehyde, acrolein, napththalene, and polycyclic organic matter (POM), all of which are gas-phase hydrocarbons that exist in gasoline vapor (except POM, which exists in gas and particle phases).     

Stage II gasoline vapor recovery systems have been a required emission control measure in areas classified as serious, severe, and extreme for the ozone NAAQS.  With the amendment of the CAA in 1990, Stage II controls were required for moderate ozone areas, under CAA section 182(b)(3).  

However, under section 202(a)(6) of the CAA, 42 U.S.C. 7521(a)(6), the requirements of section 182(b)(3) would no longer apply in moderate ozone nonattainment areas after EPA promulgation of standards for onboard refueling vapor recovery (ORVR) as part of new motor vehicles' emission control systems, and after EPA determines that ORVR is in widespread use throughout the motor vehicle fleet.  Under implementation rules issued in 2002 for the 1997 8-Hour Ozone standard, EPA retained the Stage II-related requirements under section 182(b)(3) as they applied for the 1-Hour Ozone standard (see 40 CFR 51.900(f)(5) and 40 CFR 51.916(a)).

Beginning with model year 1998, ORVR equipment has been phased in for newly manufactured light duty passenger cars and trucks, and has been required on nearly all new vehicles since 2006.  ORVR is an emission control system contained in the vehicle that captures gasoline vapors during refueling.  The gas tank and fill pipe are designed to capture refueling-related vapors and contain them in an onboard, activated carbon canister.  When the vehicle is running and warmed up, the vapors are drawn out of the canister and transferred to the engine to be combusted as fuel.  The ORVR system is very efficient and is monitored by the vehicle's on-board diagnostic system.  Unlike Stage II systems, which are geographically limited to moderate and above nonattainment areas, ORVR is applicable to all new vehicles everywhere those vehicles operate.

 Over time, non-ORVR vehicles continue to be replaced by ORVR-equipped vehicles.  On May 16, 2012, EPA issued a final rule determining that ORVR technology was in widespread use throughout the U.S. vehicle fleet and waived the requirement for states to implement Stage II vapor recovery at GDFs in nonattainment areas classified as serious or above for the ozone NAAQS (77 FR 28772).  EPA determined that emission reductions from ORVR-equipped vehicles are essentially equal to and would soon surpass the emission reductions achieved by Stage II alone.  EPA determined that a state previously required to implement a Stage II vapor recovery program may take appropriate action to remove the measure from its SIP.  77 FR 28772.  EPA hereafter refers to this action as the "widespread use determination" or the "Stage II waiver."


III.  Stage II As A Control Measure in the Attainment/Maintenance Plans 

Stage II vapor recovery was originally required by the CAA in 1990 in two Virginia ozone nonattainment areas.  Virginia adopted and published its Stage II vapor recovery regulations as Rule 4-37 of VR 120-01, entitled "Emission Standards for Petroleum Liquid Storage and Transfer Operations" in the November 2, 1992 edition of the Virginia Register of Regulations (Vol 9, Issue 3), effective January 1, 1993.  Virginia's Stage II rules applied to the Northern Virginia and Richmond VOC Emission Control Areas, as defined under 9 VAC 5-40-5200(B)(2) and 9 VAC 5-20-206.  

Virginia submitted its Stage II regulation to EPA as a SIP revision on November 5, 1992.  EPA approved Virginia's Stage II SIP revision on June 23, 1994 (59 FR 32353).      


Stage II and the Virginia portion of the Washington, DC-MD-VA 1990 and 1997 Ozone NAAQS Nonattainment Areas Attainment Plans

Under the 1990 1-Hour Ozone NAAQS, EPA originally designated the OTR portion of Virginia as serious nonattainment as a part of the Washington, DC-MD-VA 1990 Ozone NAAQS Nonattainment Area.  The Virginia portion of the OTR include the following areas: Stafford County, Arlington County, Fairfax County, Loudon County, City of Alexandria, City of Falls Church, City of Manassas, and City of Manassas Park.  On December 19, 1997, the District of Columbia, Maryland and Virginia (the three states), submitted an attainment plan for the Washington, DC-MD-VA 1990 Ozone NAAQS Nonattainment Area.  On April 17, 2003 (68 FR 19106), EPA gave the attainment plan conditional approval.  On November 13, 2002 (67 FR 68805), EPA reclassified the area as severe nonattainment; to meet the requirements of this classification, a severe area attainment plan was submitted on February 24, 2004.  This attainment plan was approved on May 13, 2005 (70 FR 25688).  The 1990 ozone NAAQS was then revoked on June 15, 2005. 

On July 18, 1997 (62 FR 38856), EPA promulgated the 1997 Ozone NAAQS.  On April 30, 2004 (69 FR 23941), EPA designated the Washington, DC-MD-VA Area as moderate nonattainment.  The Virginia portion of this area includes Arlington County, Fairfax County, Loudon County and Prince William County as well as the City of Alexandria, City of Fairfax, City of Falls Church, City of Manassas, and City of Manassas Park.  On June 12, 2007, the three states submitted an attainment plan which was proposed for approval by EPA on March 20, 2013 (78 FR 17161).  On February 28, 2012 (77 FR 11739), EPA determined that the Washington, DC-MD-VA area was attaining the 1997 Ozone NAAQS and published a clean data determination as well as a determination that the area attained the 1997 Ozone NAAQS by its statutory attainment date. The attainment plan was approved by EPA on April 10, 2015 (80 FR 19206).    

Stage II and the Fredericksburg 1997 Ozone Maintenance Area Maintenance Plan

On April 30, 2004 (69 FR 23941), EPA designated Stafford County, Spotsylvania County, and the city of Fredericksburg as the Fredericksburg 1997 Ozone Nonattainment Area.  Since the air quality in the Fredericksburg 1997 Ozone Nonattainment Area met the 1997 ozone NAAQS, a redesignation request and maintenance plan were submitted by Virginia on September 28, 2005; EPA approved the redesignation request and maintenance plan on December 23, 2005 (70 FR 76165). 

Even though the 1990 1-Hour Ozone NAAQS was revoked on June 15, 2005, EPA's subsequent implementation rules for the 1997 8-Hour Ozone NAAQS retained the Stage II-related requirements under section 182(b)(3), but only as they applied to the Area for the Area's classification for the 1-Hour Ozone NAAQS at designation for the 8-Hour Ozone NAAQS (see 40 CFR 51.900(f)). Therefore the attainment plan for the Washington, DC-MD-VA 1997 Ozone NAAQS Nonattainment Area and the maintenance plan for the Fredericksburg 1997 Ozone NAAQS Nonattainment Area also contain Stage II as a control measure, in addition to the attainment plan for the Washington, DC-MD-VA 1990 Ozone NAAQS Nonattainment Area. 

At the time of the development of the Washington, DC-MD-VA 1990 ozone NAAQS attainment plan SIP, Stage II was still a mandatory measure for areas classified as moderate ozone nonattainment under section 182(b)(3) of the CAA.  Although EPA had adopted ORVR rules for light-duty vehicles in April 1994, ORVR vehicles were not introduced for sale until model year 1998, with nearly all new gasoline-powered vehicles being equipped with ORVR by model year 2006.  Under section 202(a)(6) of the CAA, Stage II requirements for areas classified as moderate "shall not apply" upon EPA's promulgation of ORVR rules, and determination of widespread use of ORVR throughout the motor vehicle fleet. Although after ORVR promulgation Stage II was no longer a mandatory measure under the CAA, Virginia had relied upon reductions from Stage II in its ozone plans.  Although Virginia was no longer obligated to retain Stage II as a mandatory measure under the CAA, the Commonwealth is required under section 110(l) of the CAA to demonstrate that a SIP revision to remove Stage II would not interfere with the applicable requirements to attain the NAAQS.        
   

IV.  EPA Evaluation of Virginia's SIP Revisions to Remove Stage II Vapor Recovery

A.  Requirements

The removal of Stage II gasoline vapor control programs from the SIP has to meet certain requirements.  Any SIP revision that may relax the SIP must meet the requirements of section 110(l) of the Clean Air Act.  Section 110(l) requires a demonstration proving that the SIP would not be relaxed if the measure in question is removed.  Additionally, and specific to the removal of Stage II, certain areas must meet the requirements of CAA sections 184(b)(2) and 193.  CAA section 184(b)(2) applies to areas that are in the Northeast Ozone Transport Region (OTR); these areas in the OTR are required to implement Stage II or another program with comparable emissions reductions referred to as the "comparable measures" requirement.  Additionally, CAA section 193, also referred to as the "general savings clause," requires areas trying to modify control measures approved prior to November 15, 1990 to prove that modification of the control measure "insures equivalent or greater emissions reductions." The Area did not have Stage II approved into its SIP prior to November 15, 1990, therefore CAA section 193 does not apply.  The Area is a part of the OTR, therefore CAA section 184(b)(2) does apply.  These SIP revisions demonstrate compliance with CAA sections 110(l) and 184(b)(2).  


B.  State Submittal

VADEQ submitted an analysis to satisfy the requirements of CAA sections 110(l) and 184(b)(2).  VADEQ used the EPA guidance document Guidance on Removing Stage II Gasoline Vapor Control Programs from State Implementation Plans and Assessing Comparable Measures in developing its analysis to meet the requirements of CAA sections 110(l) and 184(b)(2). 

The EPA guidance document, Guidance on Removing Stage II Gasoline Vapor Control Programs from State Implementation Plans and Assessing Comparable Measures provides direction to states on the calculations necessary for the analysis of the emissions impact of Stage II removal. To analyze the emissions impact of Stage II removal, the area wide impact on the VOC inventory must be calculated. As recommended by the guidance document, there are several methods that can be used to demonstrate compliance with the requirements of CAA section 110(l); the method to be used depends on how the removal of Stage II would affect the area in question.  The first method replaces Stage II with a new control measure that would offset any emissions increases that would result from the removal of Stage II.  The second method involves substituting Stage II with excess emissions reductions not accounted for in the SIP.  In the event that the removal of Stage II does cause an emissions increase, reductions in the VOC inventory since the previously approved SIP inventory can be used to offset any potential emissions increases resulting from the removal of Stage II. The third method does not involve substitution for any emissions increases in the event of an increase in VOC emissions from Stage II removal; this method allows the Area to provide a phase-out plan that would allow a minuscule amount of emissions increases that decreases rapidly as ORVR continues to increase in the vehicle fleet. This minuscule amount of emissions increases would be too small to interfere with attainment or progress towards attainment. This noninterference can be demonstrated in several ways.  The area can provide its most recent air quality data showing attainment of the NAAQS; if the NAAQS has yet to be attained, air quality projections showing attainment can be provided.  Additional air quality analyses showing that the removal of Stage II does not interfere with attainment or maintenance of the NAAQS will also be considered.  To meet the requirements of CAA section 184(b)(2) comparability must be demonstrated by the State.  In removing Stage II and trying to prove comparability, EPA believes that it is reasonable for states to demonstrate that the removal of Stage II has no impact on VOC emissions or has de minimis (minimal) impact on VOC emissions.  EPA has defined de minimis impact as any additional incremental control achieved by Stage II beyond ORVR that is less than 10 percent (%) of the are wide VOC inventory. 

To calculate the area wide VOC inventory impact, two important equations must be used.  Equation 1, presented below, provides the overall Stage II-ORVR increment.  This increment provides the percentage impact on the annual area-wide emissions control gain from Stage II as ORVR is phased into the vehicle fleet.  Equation 2, also presented below, calculates the emissions benefits of Stage II on the VOC inventory in tons per year (tpy).  Equations 3 and 4 provide the variables necessary to calculate the Stage II impact on the VOC inventory, these equations are also provided below. 


Equation 1: incrementi = (QSII)(1-QORVRi)(ηiuSII) - (QSIIva)(CFi) 

Incrementi:  The increment percentage impact on the annual area-wide emissions control gain from Stage II as ORVR is phased into the vehicle fleet 
QSII:  Fraction of gasoline throughput covered by Stage II 
QORVRi:  Fraction of annual gallons of highway motor gasoline dispensed to ORVR-equipped vehicles
ηiuSII:  In-use Stage II control efficiency 
QSIIva:  Fraction of gasoline throughput covered by Stage II 
CFi:  Compatibility factor


Equation 2:  Tonsi = (Incrementi)(GCi)(EF)

Tonsi:  VOC emissions benefit of Stage II (tons)
Incrementi:  The increment percentage impact on the refueling inventory from removing Stage II 
GCi: The projected gasoline consumption (gal)
EF:  The uncontrolled displacement refueling emissions factor (g/gal) 


Equation 3:  EF (g/gal) = exp[-1.2798 - 0.0049(ΔT) + 0.0203(Td) + 0.1315(RVP)]

EF:  The uncontrolled displacement refueling emissions factor (g/gal) 
ΔT:  Values for this variable are area specific and are located in Appendix A-2 of the EPA guidance document (degrees fahrenheit)
Td:   Values for this variable are area specific and are located in Appendix A-3 of the EPA guidance document (degrees fahrenheit)
RVP:  Reid Vapor Pressure of the gasoline used for the area (pounds per square inch (psi))


VA DEQ submitted an analysis quantifying the VOC emissions benefits of Stage II.  Equation 1 was utilized to yield the increments provided in Tables 1 and 2. The calculations in Table 1 assumed a Stage II control efficiency of 77% while the calculations in Table 2 assumed a Stage II control efficiency of 56%.  Negative increment values are indicative of when excess emissions from non ORVR compatible Stage II systems overwhelm the emissions benefits of implementing Stage II.  Table 1 presents calculated increment values which show that the Area will no longer see emissions benefits from Stage II after 2012.  Table 2 presents calculated increment values which show that the Area will no longer see emissions benefits from Stage II after 2011.  VADEQ used the more conservative Stage II efficiency estimate of 77% to determine the date when there will be no emissions benefits from Stage II; therefore, the Area will no longer see emissions benefits from Stage II after 2012. 


         Table 1.  Increment Calculations for 77% Stage II Efficiency
Year
Increment 
                                                                           2008
                                                                        0.13760
                                                                           2009
                                                                        0.10597
                                                                           2010
                                                                        0.07168
                                                                           2011
                                                                        0.04235
                                                                           2012
                                                                        0.01821
                                                                           2013
                                                                       -0.00128
                                                                           2014
                                                                       -0.01632
                                                                           2015
                                                                       -0.02785
                                                                           2016
                                                                       -0.03673
                                                                           2017
                                                                       -0.04335
                                                                           2018
                                                                       -0.04817
                                                                           2020
                                                                       -0.05428
                                                                           2030
                                                                       -0.06204


         Table 2.  Increment Calculations for 56% Stage II Efficiency
Year
Increment 
                                                                           2008
                                                                        0.08684
                                                                           2009
                                                                        0.06302
                                                                           2010
                                                                        0.03732
                                                                           2011
                                                                        0.01534
                                                                           2012
                                                                       -0.00275
                                                                           2013
                                                                       -0.01736
                                                                           2014
                                                                       -0.02862
                                                                           2015
                                                                       -0.03725
                                                                           2016
                                                                       -0.04390
                                                                           2017
                                                                       -0.04886
                                                                           2018
                                                                       -0.05245
                                                                           2020
                                                                       -0.05702
                                                                           2030
                                                                       -0.06279


The emissions benefits of Stage II were calculated using Equations 2 and 3.  Using the more conservative increments produced by using the 77% Stage II efficiency estimate, the emissions benefits were calculated and the values tabulated in Table 3, below. The calculations also prove that there will be no emissions benefits from Stage II implementation past 2012.  


    Table 3.  Stage II Emissions Reductions in the Area-Wide VOC Inventory
Year
Emissions Reductions (tpd VOC)
                                                                           2008
                                                                           0.58
                                                                           2009
                                                                           0.46
                                                                           2010
                                                                           0.31
                                                                           2011
                                                                           0.19
                                                                           2012
                                                                           0.08
                                                                           2013
                                                                          -0.01
                                                                           2014
                                                                          -0.07
                                                                           2015
                                                                          -0.13
                                                                           2016
                                                                          -0.17
                                                                           2017
                                                                          -0.20
                                                                           2018
                                                                          -0.22
                                                                           2020
                                                                          -0.24


In addition to increment and area wide VOC inventory calculations, VADEQ submitted additional analyses to support its removal of Stage II from the Virginia SIP.  VADEQ included an analysis which calculated the non-ORVR compatible Stage II benefit on a gasoline basis and a vehicle miles travelled (VMT) basis and compared the analysis to the Stage II benefit calculated with the use of the Motor Vehicle Emission Simulator (MOVES2010a).  This comparison is provided below in Table 4, and shows that incompatibility excess emissions, which are the emissions that arise from the interaction of a Stage II vapor recovery system (VRS) with ORVR, exceed Stage II benefits in the 2014 time frame.  


        Table 4.  Excess and Benefit VOC Emissions Calculations Results
Year
                       Excess Vent Emissions  (tpd VOC)
                          Stage II Benefit (tpd VOC)
                Stage II Benefit with respect to ORVR (tpd VOC)
 
Gasoline Basis
VMT Basis

Gasoline Basis
VMT Basis
                                                                           2008
                                                                         0.5768
                                                                         0.5790
                                                                          1.937
                                                                          1.360
                                                                          1.358
                                                                           2009
                                                                         0.6218
                                                                         0.6271
                                                                          1.616
                                                                          0.996
                                                                          0.991
                                                                           2010
                                                                         0.6669
                                                                         0.6696
                                                                          1.343
                                                                          0.676
                                                                          0.673
                                                                           2011
                                                                         0.7040
                                                                         0.7042
                                                                          1.309
                                                                          0.605
                                                                          0.605
                                                                           2012
                                                                         0.7496
                                                                         0.7479
                                                                          1.072
                                                                          0.323
                                                                          0.325
                                                                           2013
                                                                         0.7755
                                                                         0.7721
                                                                          0.885
                                                                          0.110
                                                                          0.113
                                                                           2014
                                                                         0.7933
                                                                         0.7883
                                                                          0.730
                                                                         -0.063
                                                                         -0.058
                                                                           2015
                                                                         0.8071
                                                                         0.8008
                                                                          0.605
                                                                         -0.202
                                                                         -0.196
                                                                           2016
                                                                         0.8189
                                                                         0.8115
                                                                          0.505
                                                                         -0.313
                                                                         -0.306
                                                                           2017
                                                                         0.8211
                                                                         0.8129
                                                                          0.434
                                                                         -0.388
                                                                         -0.379
                                                                           2018
                                                                         0.8204
                                                                         0.8113
                                                                          0.378
                                                                         -0.442
                                                                         -0.433
                                                                           2020
                                                                         0.8184
                                                                         0.8081
                                                                          0.307
                                                                         -0.511
                                                                         -0.501
                                                                           2030
                                                                         0.8055
                                                                         0.7926
                                                                          0.249
                                                                         -0.556
                                                                         -0.544


Additionally, VADEQ submitted work completed by the Georgia Institute of Technology (Georgia Tech) for the Association for Southeastern Integrated Planning project.  Georgia Tech completed emissions sensitivity analyses utilizing the Community Multiscale Air Quality (CMAQ) model, for the sensitivity of ozone concentrations with respect to NOx and VOC emissions.  Three sensitivity analyses were conducted to gauge the sensitivity of ozone formation with respect to a 30% decrease in anthropogenic VOC, ground level NOx and point source NOx in the applicable Virginia areas.  It was found that Ozone concentrations were most sensitive to decreases in point source NOx.  The sensitivity to point source NOx is 100 to 1,000 times greater than ozone concentration sensitivity to anthropogenic VOC. The results indicate that NOx has a much larger impact on Ozone concentrations than VOC.   

Through the CMAQ modeling utilized for the sensitivity analyses, the effect of Stage II on ozone concentrations was calculated at the monitor in the Virginia Areas.  Table 5, below, presents the Stage II benefit in terms of the change of ozone concentration.  The results presented show that Stage II has a very minuscule effect on ozone concentrations in the Virginia Areas.  


               Table 5.  Stage II Effect on Ozone Concentration
Area
Monitor
Stage II Benefit to Ozone Concentrations (parts per billion (ppb))
Virginia Areas
Arlington
 -1.72x10-4 to -2.87x10[-4]


C.  EPA Evaluation and Conclusion  

EPA thoroughly evaluated the submittal made by Virginia through VADEQ.  EPA's detailed review has led EPA to conclude that VADEQ properly addressed the recommended increment and VOC inventory calculations which prove when Stage II no longer provides a benefit for the Virginia Areas.  EPA agrees with VADEQ that withdrawal of the Stage II requirement is appropriate on January 1, 2014.  EPA also reviewed additional data provided by VADEQ showing ozone formation in the Virginia Areas is much more NOx dependent.  In conclusion, EPA agrees with Virginia's recommendation that the Stage II requirement be removed on January 1, 2014, since at that point in time, the Virginia Areas will no longer see any emissions benefits from the implementation of Stage II.  Also, at that point in time, the Virginia Areas will have met and exceeded the de minimus requirement for removing Stage II.  EPA is approving the removal of Stage II from the attainment and maintenance plans for the Virginia Areas because the submitted analysis demonstrates that Stage II removal will not interfere with the attainment or maintenance of the NAAQS in accordance with section 110(l) of the CAA and meets the requirements of section 184(b)(2) of the CAA.  
