To: David Peter, Environmental Engineer, U.S. EPA Region 7, Air Permitting Branch
Subject: Justification for the SO2 Emission Limit and Compliance Mechanism for MPW
From: Peter Zayudis & Shawn Corbin, Air Construction Permit Section, Air Quality Bureau, Iowa Department of Natural Resources
Date: 10/08/2015

Introduction: Muscatine Power and Water (MPW) is a municipal utility that operates fossil fuel fired boilers for power generation (SIC code 4911).  MPW operates three existing coal fired boilers, Boiler 7 (UNIT 7), Boiler 8 (UNIT 8), Boiler 9 (UNIT 9) and ancillary equipment to support power generation.  Boiler 7 is a traveling grate stoker boiler with a maximum heat input of 289 MMBtu per hour with no sulfur dioxide (SO2) controls.  Boiler 8 is a cyclone boiler with a maximum heat input of 890 MMBtu per hour with no SO2 controls.  Boiler 9 is tangentially fired boiler with maximum heat input of 1,556 MMBtu per hour with scrubbers (CE93 and CE94) that control SO2 emissions.  
MPW is located within the portion of portion of Muscatine County that is classified as a nonattainment area for the 2010 1-hr SO2 NAAQS.  As part of the SO2 attainment plan development, the 2009 and 2010 operating years for MPW were evaluated to determine if attainment may be achieved without additional control measures already installed at the facility.  Based on this analysis, the Air Quality Bureau of the Iowa Department of Natural Resources (DNR)  determined that additional controls (actual SO2 emissions reductions) were not necessary to achieve predicted attainment of the SO2 NAAQS.  However, reductions in potential SO2 emissions on Boilers 7, 8 and 9 were necessary to demonstrate predicted attainment with the NAAQS.  
On April 23, 2014, EPA released non-binding guidance titled Guidance for 1-hr SO2 Non-attainment Area SIP Submissions to assist states and regions in the development of 1-hr SO2 attainment submittals.  Within this guidance EPA discusses various methodologies that may be employed to establish 1-hr emission limits and other mechanisms that may be used to ensure attainment of the NAAQS.  The information presented below discusses how the DNR utilized this guidance in the development of SO2 emission limits and a compliance mechanism for MPW to demonstrate predicted attainment of the 1-hr SO2 NAAQS.  

Critical Value Development:  As specified within the previously mentioned guidance, the first step is to conduct dispersion modeling to determine the critical emission values or the maximum 1-hr values that demonstrated predicted attainment with the NAAQs.  MPW conducted dispersion modeling to determine critical values for Boilers 7, 8, and 9.
As part of this demonstration, MPW also requested to retain operational flexibility for Boilers 7, 8, and 9 while still maintaining the SO2 NAAQS.  Table 1 below illustrates the critical values determined for each proposed boiler operating scenario.

Table 1: SO2 Critical Values for Each Boiler Operating Scenario
Source
                        Emission Rate Scenarios (lb/hr)

                                      All
                                    U9 Off
                                    U8 Off
                                    U7 Off
                                    U7 Only
                                    U8 Only
                                    U9 Only
Boiler 7 (U7)
                                      250
                                      275
                                      700
                                      ---
                                      800
                                      ---
                                      ---
Boiler 8 (U8)
                                     1,000
                                     1,050
                                      ---
                                     1,350
                                      ---
                                     1,525
                                      ---
Boiler 9 (U9)
                                      120
                                      ---
                                      250
                                      250
                                      ---
                                      ---
                                     1,500

In the April 23, 2014 guidance EPA stated, "...that it may be possible in specific cases for states to develop control strategies that account for variability in 1-hour emission rates through emission limits with averaging times that are longer than 1-hour, using averaging times as long as 30-days, but still provided for attainment of the 2010 SO2 NAAQS."
As part of MPW's application submittal, MPW requested a 21-day averaging period for each operating scenario specified in Table 1.  To support MPW's request for a longer averaging period, MPW provided operational and hourly SO2 data for each boiler from 2010 to 2014.  

Emission Limit and Compliance Mechanism Development:  Based on MPW's request for a longer averaging period in their permit limit, the DNR reviewed EPA's guidance to determine the criteria that should be used to establish a longer averaging period for the 1-hr standard and the mechanisms that should be utilized to establish a longer averaging period. 
EPA states that it is their "...general expectation that, if periods of hourly emissions above the critical emission value are a rare occurrence at a source, particularly if the magnitude of the emissions is not substantially higher than the critical emissions value, these periods would be unlikely to have a significant impact on air quality..." since they would be unlikely to occur repeatedly at the times when meteorology is conducive to high ambient concentrations of SO2.  
Further, EPA's guidance states "EPA believes that making this option available to states could reflect an appropriate balance between providing strong assurance that NAAQs will be attained and maintained, while still acknowledging the necessary variability in source operations and the impairment to source operations that would occur under what could be cases [of] an unnecessarily restrictive approach to constraining the variability."  
Based on this guidance the DNR conducted a review of MPW's Boilers 7, 8 and 9 operation for 2010 to 2014 to determine variability in SO2 emissions due to such factors as boiler start-up, shutdown, malfunction, control equipment malfunction, and coal sulfur content fluctuations.  Based on this review, the DNR determined the following: 

 The 99[th] percentile of the hourly emission values are significantly lower than the critical emission values when Boilers 7, 8, and 9 are operating simultaneously.  The 99[th] percentile of the hourly emission rate for Boiler 7 is 178.1 lbs/hr, Boiler 8 is 782.8 lbs/hr, and Boiler 9 is 78.5 lbs/hr, which are all much lower than the critical values specified in Table 1.
 The maximum 24-hour rolling average emission rate for Boiler 7, during the past 5 years, is 222.5 lb/hr.  In comparison, the lowest 1-hour modeled critical emission value for Boiler 7 is 250 lb/hr when all three boilers are operating at the same time.  The 24-hour average emission rate for Boiler 7 and Boiler 8 never exceeded the lowest 1-hour modeled critical emission value during the past five years. 
 During that same time span, Boiler 9 had 11 occurrences when the 24-hour average emission rate exceeded the lowest 1-hour modeled critical emission value for that boiler, which equates to approximately 99.4% of the time the 24-hour average emission rate was below the lowest 1-hour modeled critical emission value for Unit 9.  
 For Boiler 9, each occurrence of the 24-hour average emission rate that exceeded the lowest 1-hour modeled critical emission value for Boiler 9 is related to startup, shutdown, or malfunction, or SSM, of the SO2 control equipment, which is included in the permitted 21-day rolling average emission limit.  For example, on May 3, 2012, Boiler 9 had 6 hours of uncontrolled emission rates that led to one of the aforementioned 24-hour average emission rate exceedances of the lowest 1-hour modeled critical emission value.  
 The DNR also reviewed how many days between 2010 and 2014 that MPW acted under these 7 operating scenarios as specified in Table 1, specifically Boilers 7, 8 and 9 operating alone to determine the likelihood of these scenarios occurring for a 21-day consecutive period.  Based on this review, MPW had zero occurrences of operating Boilers 7, 8 or 9 by themselves for 21-consecutive days.  For example, Boiler 7 operating without Boilers 8 and 9 operating occurred for less than 1 day consecutively over the 5 year period.  The most frequent combination (except for all three boilers operating at the same time) was Boilers 8 and 9 operating at the same time and Boiler 7 not in operation.  This occurred for a maximum of 7 consecutive days over the over 5 year period.  
 From 2010 to 2014, Boilers 7, 8 and 9 operated at the same time for approximately 89% of the total boiler firing time.  Therefore, a vast majority of the time all three boilers will be in operation at the same time.  

After completing review of MPW's operation of Boilers 7, 8 and 9 for 2010 to 2014, the DNR developed the appropriate adjustment to allow for a longer averaging period as compared to critical values specified in Table 1 while balancing "the necessary variability in source operations and the impairment to source operations that would occur under what could be cases [of] an unnecessarily restrictive approach to constraining the variability"
According to EPA's guidance, any emission limits based on averaging periods longer than 1-hour should be designed to have comparable stringency to a 1-hour average limit at the critical emission value.  The longer term average emission limit would allow occasions when emissions exceed the 1-hour critical emission value; the use of the lower limit requires most values be lower than required for a 1-hour limit at the critical emission value.
The DNR developed a 1-hr SO2 emission limit based on a 21-day rolling average for Boilers 7, 8 and 9.  To develop this limit, the DNR used the boiler specific ratio of the 99[th] percentile values for the 1 hour and 21-day values and the most stringent critical value for each boiler as specified in Table 1 (all boilers operating scenario).  These boiler specific ratios are detailed in Table 2 below:

Table 2:  Boiler Variability Ratios as Compared to 21-day Averaging Period
Source
                              Variability Ratios


Boiler 7
                                     0.71
Boiler 8
                                     0.90
Boiler 9
                                     0.63

Based on the ratios in Table 2, the 21-day rolling average emission limit development is detailed below:
 Overall 21-day Roling Average = U7 (1hr Critical Value) x U7 Ratio (21-day) + U8 (1hr Critical Value) x U8 Ratio (21-day) + U9 (1hr Critical Value) x U9 Ratio (21-day) 
 Overall 21-day Roling Average = (250 lb/hr)(0.71) + (1000 lb/hr)(0.90) + (120 lb/hr)(0.63) = 1153 lb/hr 

To address the scenarios requested to allow for 21-day averaging period for Boilers 7, 8, and 9 operating separately in any combination,. the DNR determined that establishing a 21-day limit for each boiler operating separately based on the ratio specified in Table 2 was not appropriate.  This is due to the fact that the 21-day emission limit (1153 lbs/hr) already accounts for each boiler's variability. An additional correction for boiler variability based on these ratios would unnecessarily penalize MPW operationally by essentially double counting boiler variability .  
As specified in the guidance, "EPA does not preclude states from using other approaches to determine appropriate longer term average limits; however, EPA recommends all analyses begin with determination of the critical emission values."
Therefore, the DNR used operational data from 2010 to 2014 for Boilers 7, 8 and 9 operating independently as the basis to develop (as detailed above) the appropriate adjustment factor to address the scenario when Boilers 7, 8, and 9 are operating independently and allow a longer averaging period.   The adjustment factor to allow for a longer averaging period to address the scenario when Boilers 7, 8, and 9 are operating independently is detailed below:
Boiler 7:
 U7 Fraction = Modeled total emission rate, lb/hr (All scenario)/Modeled U7 only emission rate, lb/hr
 U7 Fraction = 1370/800 = 1.71 
Boiler 8:
 U8 Fraction = Modeled total emission rate, lb/hr (All scenario)/Modeled U8 only emission rate, lb/hr
 U8 Fraction = 1370/1525 = 0.898
Boiler 9:
 U9 Fraction = Modeled total emission rate, lb/hr (All scenario)/Modeled U9 only emission rate, lb/hr
 U9 Fraction = 1370/1500 = 0.913

The "fraction" factor is the sum of the critical emission values as specified in Table 1 for the "all" scenario (all boilers operating at the same time) divided by the critical emission value for each boiler operating independently (U7 Only, U8 Only, U9 Only).  The fraction factor developed accounts for the different impacts of each boiler on ambient concentrations due to each boiler's unique dispersion characteristics. Based on the adjustment for a 21-day averaging period for all scenarios as specified in Table 1, the flowing compliance formula is derived to ensure attainment of the 1-hour SO2 NAAQs:

 1153 lb/hr (21-day rolling average, For Boilers 7, 8 and 9 combined)
 = 1.713(Boiler 7, 24 hr average lb/hr) + 0.898(Boiler 8, 24 hr average lb/hr) + 0.913(Boiler 9, 24 hr average lb/hr) 

The sum of each boiler operating scenario emission rate cannot exceed the calculated 21-day rolling average for all the three boilers combined, 1153 pounds per hour.  In addition, under no circumstances (i.e., boiler emission scenarios) can the monitored emission rate on a 21-day rolling average exceed the 1-hour critical value for each boiler under any applicable operating scenario.  For example, if MPW operates only Boiler 7 for 21 consecutive days, the maximum 21-day rolling average emission rate allowed would be approximately 673 lb/hr, as compared to the critical value for this scenario of 800 lb/hr. (i.e., Unit 7 alone).  The proposed compliance formula considers MPW operation flexibility while protecting the critical value and establishing appropriate adjustments for boiler variability.  The proposed mechanism provides the maximum operating flexibility for MPW and eases the recordkeeping burden while remaining protective of the NAAQS
Conclusion: The proposed SO2 limit and compliance mechanism provide flexibility to accommodate operational flexibility and is comparable in stringency to the 1-hour emission limit while balancing "the necessary variability in source operations and the impairment to source operations that would occur under what could be cases [of] an unnecessarily restrictive approach to constraining the variability."  This justification demonstrates  that the SO2 emission limit, compliance mechanism, and averaging period are protective of the 1hr SO2 NAAQS under any MPW boiler operating scenario.
