Martin Drake Power Plant SO NAAQS Compliance Modeling Report€¦ · · 2017-09-11Continuous...

Martin Drake Power Plant SO2 NAAQS Compliance Modeling Report

Colorado Springs Utilities Martin Drake Power Plant in Colorado Springs, CO Project Number: 60537165 August 2017

Martin Drake SO2 NAAQS Compliance Modeling Report, August 2017

i

AECOM

Quality information Prepared by Checked by Approved by

Olga Samani Air Quality Scientist Air Quality Services

Robert J. Paine Associate Vice President Air Quality Services

Thomas Damiana Air Quality Engineer/Meteorologist Air Quality Services

Prepared for: Colorado Springs Utilities Colorado Springs, CO

Prepared by: AECOM 250 Apollo Drive Chelmsford MA, 01824 USA www.aecom.com

Copyright © 2017 by AECOM


ii

AECOM

Table of Contents 1. Introduction ...................................................................................................................................... 1-1

1.1 Background ............................................................................................................................ 1-1 1.2 Document Organization ......................................................................................................... 1-1

2. Emission Source Inventory .............................................................................................................. 2-1 2.1 Description of Martin Drake Emission Sources ..................................................................... 2-1 2.2 Exhaust Parameters Sensitivity Analysis .............................................................................. 2-2 2.3 Nearby Background Sources ................................................................................................. 2-3

3. Site-Specific Meteorological Database ............................................................................................ 3-1 3.1 Overview of the Monitoring Program ..................................................................................... 3-1 3.2 Description of Monitoring Locations ...................................................................................... 3-1

4. Modeling Procedures ....................................................................................................................... 4-1 4.1 Dispersion Model Selection ................................................................................................... 4-1 4.2 Land Use Classification ......................................................................................................... 4-1 4.3 Good Engineering Practice (GEP) Analysis .......................................................................... 4-1 4.4 Meteorological Data Processing with AERMET .................................................................... 4-1 4.4.1 Surface Characteristics for Use in AERMET ......................................................................... 4-2 4.5 Receptor Grid ........................................................................................................................ 4-3 4.6 Review of Ambient Background Monitoring Data .................................................................. 4-4 4.7 Model Configurations and Options ........................................................................................ 4-4

5. Modeling Results and Conclusion ................................................................................................... 5-1 5.1 NAAQS Compliance Modeling Results ................................................................................. 5-1 5.2 Sensitivity of the Results to Operational Load ....................................................................... 5-1

List of Figures Figure 2-1: Location of Martin Drake ....................................................................................................... 2-4 Figure 2-2: Martin Drake Annual Emissions in 2006 – 2016 .................................................................... 2-5 Figure 3-1: Location of the Meteorological Tower and SODAR Locations Relative to the Martin Drake Facility ............................................................................................................................................... 3-2 Figure 4-1: Stacks and Buildings Used in the GEP and Building Downwash Analysis for Martin Drake 4-5 Figure 4-2: 10-meter On-site Wind Rose (11/7/2015 – 11/6/2016) .......................................................... 4-6 Figure 4-3: 60-meter On-site Wind Rose (11/7/2015 – 11/6/2016) .......................................................... 4-6 Figure 4-4: Location of the Meteorological Stations and Ambient SO2 Monitor for Background Concentrations ........................................................................................................................................... 4-7 Figure 4-5: Far-field Receptor Grid for Modeling ..................................................................................... 4-7 Figure 4-6: Near-field Receptor Grid for Modeling ................................................................................... 4-8 Figure 5-1: Far-Field Isopleths of 1-Hour SO2 Design Concentrations .................................................... 5-2 Figure 5-2: Near-Field Isopleths of 1-Hour SO2 Design Concentrations ................................................. 5-3

List of Tables Table 2-1: Martin Drake Stack Information Used in Modeling NAAQS Compliance .............................. 2-1 Table 2-2: Exhaust Parameters Sensitivity Analysis Input Data ............................................................. 2-3 Table 3-1: Meteorological Tower Location .............................................................................................. 3-1


iii

AECOM

Table 3-2: SODAR Location ................................................................................................................... 3-1 Table 5-1: AERMOD-Predicted Design 1-hour SO2 Concentration ........................................................ 5-1 Table 5-2: Sensitivity Analysis Results ................................................................................................... 5-1


1-1

AECOM

1. Introduction

1.1 Background The United States Environmental Protection Agency (EPA) promulgated a 1-hour National Ambient Air Quality Standard (NAAQS) for sulfur dioxide (SO2) in 2010. The 1-hour SO2 NAAQS has a level set at 75 ppb and the form of the standard is the average of the 99th percentile of the daily maximum 1-hour average monitored concentrations realized in each of three consecutive calendar years (the “design value,”).

The Martin Drake Power Plant (“Martin Drake”) is owned and operated by Colorado Springs Utilities (CSU). In their July 12, 2016 SO2 NAAQS area designations,1 EPA agreed with the State of Colorado that the area around Martin Drake Power Plant should be assigned an unclassifiable designation for SO2. During that process, Colorado Department of Public Health and Environment (CDPHE) asked that a 1-hour SO2 compliance demonstration be made using dispersion modeling driven by site-specific meteorological data. To fulfill that request, a meteorological monitoring program was initiated. Data covering the period October 18, 2015 through January 2017 was collected, submitted to and approved by CDPHE. While the data was being reviewed, a modeling protocol describing the technical approach to the compliance demonstration was drafted and approved by CDPHE. That modeling protocol was posted for public review and comment from June 21 through July 21, 2017. As a result of comments received, CDPHE requested minor changes to the procedures described in the protocol. These changes are referenced later in this document.

The purpose of this report is to provide an overview of the approved modeling procedure that was subject to public review and comment, and to provide modeling results documenting the SO2 NAAQS compliance demonstration.

1.2 Document Organization Section 2 of this report describes the emission sources of SO2 at the Martin Drake Power Plant, modeled emission rates and exhaust parameters, and discusses recent operational changes.

Section 3 describes the site-specific meteorological monitoring program.

Section 4 describes how the meteorological data was processed for AERMOD modeling, and dispersion model approaches that were used in this modeling application, including load analyses and emissions-dependent temperature and exit velocity values to be used in the modeling based upon a review of recent Continuous Emissions Monitoring system (CEMs) data. This section also describes the proposed source of regional monitoring data that represents impacts from sources that were not explicitly included in the modeling effort.

Section 5 discusses the modeling results and the results of the exhaust parameters sensitivity.

1 81 FR 45039, July 12, 2016, Technical discussion at https://www.epa.gov/sites/production/files/2016-03/documents/co-epa-tsd-r2.pdf

https://www.epa.gov/sites/production/files/2016-03/documents/co-epa-tsd-r2.pdf



2-1

AECOM

2. Emission Source Inventory

2.1 Description of Martin Drake Emission Sources Martin Drake is a coal-fired power plant located in Colorado Springs, CO (Figure 2-1). This facility, operated by CSU, consists of three boiler units, Units 5, 6, and 7. Unit 5 began operation in 1962, Unit 6 began operation in 1968, and Unit 7 began operation in 1974.2 Recently, the Martin Drake facility has undergone operational changes that have resulted in significantly reduced SO2 emissions. Effective December 31, 2016, Martin Drake’s Unit 5 was permanently retired from operations and excluded from the 1-hour SO2 compliance demonstration modeling described in this report. This retirement is federally enforceable under the Martin Drake Colorado Air Pollution Control Division (APCD) Construction Permit.3 Additionally, Martin Drake has completed the installation of two SO2 pollution control devices, one for Unit 6 and the second for Unit 7, in September 2016. Testing and optimization of these controls are currently ongoing and will continue throughout 2017. These pollution control devices satisfy the requirements of Best Available Retrofit Technology (BART) to improve air quality and assist Colorado in complying with the EPA’s Regional Haze Rule. As a result of the EPA Regional Haze Rule, a new rolling average 30-day permitted SO2 emission limit of 0.13 lb/MMBtu will be effective no later than December 31, 2017.4 Table 2-1 lists information on the stacks modeled for the NAAQS compliance analysis. Stack elevations, stack heights, and exit diameters are consistent with previously submitted Regional Haze modeling analyses5,6 and 2017 Colorado Air Pollution Emission Notices (APEN) data and were held constant for this modeling analysis.

Table 2-1: Martin Drake Stack Information Used in Modeling NAAQS Compliance

Stack Information Unit 6 Unit 7

Base Elevation 1,814 m 1,814 m

Actual Stack Height 61.0 m 76.2 m

SO2 Emission Rate 19.863 g/s 30.822 g/s

Exit Velocity (1) 12.79 m/s 15.99 m/s

Exit Temperature (1) 379.82 K 381.04 K

Exit Diameter 3.84 m 4.57 m

Boiler Rated Heat Input (2) 861 MMBtu/hr 1,336 MMBtu/hr

Electrical Power Rating, Gross Megawatts 85 MW 142 MW (1) The flue gas temperature and exit velocity were based on the median high load (greater than

65 MW for Unit 6 and greater than 113 MW for Unit 7) scrubbed normal operations using certified flue gas flow and temperature instrumentation installed in the stack. The period analyzed was from March 28 to August 20, 2017 for Unit 6; and from January 17 to August 20, 2017 for Unit 7.

(2) CDPHE. Operating Permit for Colorado Springs Utilities Martin Drake Power Plant. As described in the modeling protocol, short-term emissions were calculated using the Martin Drake 30-day allowable emissions limit and EPA’s Appendix D, Table 1. This table shows an average ratio of 99th percentile 30-day average SO2 emission values vs. 99th percentile 1-hour average SO2 emission values for sources with wet scrubbers, such as Martin Drake Power Plant, to be 0.71. Applying the 2 USEIA Form EIA-860, 6_2_EnviroEquip2015.xlsx. http://www.eia.gov/electricity/data/eia860/ 3 CDPHE, Permit 10EP402, Issuance 2. 4 CDPHE Draft Technical Support Document. https://www.epa.gov/sites/production/files/2016-03/documents/co-epa-tsd-r2.pdf 5 CDPHE, BART CALPUFF Class I Federal Area Individual Source Attribution Visibility Impairment Modeling Analysis for CSU Martin Drake Power Plant Units 5, 6, and 7 (2005): https://www.colorado.gov/airquality/tech_doc_repository.aspx?action=open&file=BARTCalpuff+Report-drake.pdf 6 CSU, CALPUFF Modeling Results for the Martin Drake Power Plant’s Synthetic Minor Permit Application (2007): http://www.colorado.gov/airquality/tech_doc_repository.aspx?action=open&file=CSU_Drake_BART_CALPUFF_Report-17Aug07.pdf

http://www.eia.gov/electricity/data/eia860/


https://www.colorado.gov/airquality/tech_doc_repository.aspx?action=open&file=BARTCalpuff+Report-drake.pdf

http://www.colorado.gov/airquality/tech_doc_repository.aspx?action=open&file=CSU_Drake_BART_CALPUFF_Report-17Aug07.pdf


2-2

AECOM

EPA-recommended 0.71 ratio to the 30-day limit leads to the 1-hour emission rates for Units 6 and 7 shown below.

• Martin Drake 30-day Allowable Emission Limit = 0.13 lb/MMBtu

• The design firing rate of Unit 6 is 861 MMBtu/hr; for Unit 7 it is 1,336 MMBtu/hr.

• For Unit 6, the resultant 30-day SO2 emission rate is 0.13 lb/MMBtu x 861 MMBtu/hr = 111.9 lb/hr.

• For Unit 7, the 30-day emission rate is 0.13 lb/MMBtu x 1,336 MMBtu/hr = 173.7 lb/hr.

Incorporating the ratio to account for 30-day to 1-hour fluctuations, the resultant 1-hour emission rate for Unit 6 becomes 111.9 lb/hr ÷ 0.71 = 157.6 lb/hr (19.863 g/s). For Unit 7, this becomes 173.7 lb/hr ÷ 0.71 = 244.6 lb/hr (30.822 g/s). These emission rates were used as input to the dispersion modeling to demonstrate compliance with the 1-hour SO2 NAAQS.

Due to recent operational changes previously mentioned (i.e., newly installed SO2 pollution controls currently undergoing testing), exit velocities (derived by volumetric flow rate) and exit temperatures used in previously submitted modeling analyses or cited in state APENs or national databases to this point are not considered representative of operations with the new pollution controls. Operation of the SO2 pollution controls modifies these two parameters. CSU’s CEMs data from each of the units provides (from 1/1/2017 to 8/2/2017) representative information on hourly emission rates, exit velocity (the protocol describes how the exit velocity was computed from CEMs volumetric flow data) and exit temperatures for each unit at Martin Drake. Based upon the newly available CEMs data, a median exit velocity and temperature, representative of high load scrubbed operations, was computed and used in the dispersion modeling. Impacts determined over a range of operational conditions are addressed in a sensitivity analysis, as described in Section 2.2 below.

For historical purposes, it is useful to compare recent trends in annual emissions with prior years. As shown in Figure 2-2, annual emissions at Martin Drake have an overall downward, or decreasing, trend over time. Emissions levels in 2014 were affected by the large fire experienced by Martin Drake which resulted in significant downtime. Emission levels in 2015 can be considered as typical prior to the installation of pollution controls. Emissions for 2016 correspond to the recent testing phase of the new SO2 pollution controls.

As described in the modeling protocol, historic and projected SO2 emissions from the only other SO2-emitting Martin Drake source, an emergency generator, are negligible and intermittent and thus, per EPA’s March 1, 2011 guidance7 has been excluded from the compliance modeling study.

2.2 Exhaust Parameters Sensitivity Analysis In addition to the design high capacity analysis described in Section 2.1, sensitivity modeling has been conducted per CDPHE’s request8 to evaluate the range of stack conditions during scrubbed operations. Only the hours when the scrubber was operational were considered for this analysis. Typically, scrubbing results in a cooler plume and less dispersion, which leads to a lower plume rise than that for the previously unscrubbed plant operations.

The sensitivity modeling consists of modeling the lower and upper bounds of the scrubbed operating conditions, conservatively paired with the permit-limited emissions at full capacity. In practice, the emissions associated with the low temperature and velocity case are two to three times lower than the high temperature and velocity case due to the ratio of the fuel flow rates. In particular, for the larger

7 EPA Additional Clarification Regarding Application of Appendix W Modeling Guidance for the 1-hour NO2 NAAQS, March 1, 2011. http://www.epa.gov/scram001/guidance/clarification/Additional_Clarifications_AppendixW_Hourly-NO2-NAAQS_FINAL_03-01-2011.pdf 8 Email from Lisa Devore (CDPHE) to Kevin Weiner (CSU) on August 11, 2017 with the subject “Address Partial Loads in the Final Modeling Report”.

http://www.epa.gov/scram001/guidance/clarification/Additional_Clarifications_AppendixW_Hourly-NO2-NAAQS_FINAL_03-01-2011.pdf

http://www.epa.gov/scram001/guidance/clarification/Additional_Clarifications_AppendixW_Hourly-NO2-NAAQS_FINAL_03-01-2011.pdf


2-3

AECOM

Unit 7, the emissions for the low temperature and velocity case would be expected to be less than one third of the emissions for the high temperature and velocity case.

To develop the lower and upper bounds to model, CEMs data for 2017 was reviewed to determine the lowest valid temperature and velocity values for each unit. The high temperature and velocity values were determined using the highest of the valid values recorded. Then each set of conditions was modeled with a fixed reference emission rate (157.6 lb/hr for Unit 6 and 244.6 lb/hr for Unit 7). Results of this analysis are presented in Section 5.2. The dataset used to develop the values is included in the digital modeling archive transmitted separately from this document.

Table 2-2: Exhaust Parameters Sensitivity Analysis Input Data

Stack Information Unit 6 Unit 7

Base Elevation 1,814 m 1,814 m

Actual Stack Height 61.0 m 76.2 m

Low Flow Exit Velocity (1) 7.63 m/s 6.09 m/s

Low Flow Exit Temperature (1) 351.34 K 328.99 K

High Flow Exit Velocity (2) 14.76 m/s 18.55 m/s

High Flow Exit Temperature (2) 410.59 K 408.95 K

Exit Diameter 3.84 m 4.57 m

Reference SO2 Emission Rate 157.6 lb/hr 244.6 (1) The low flow flue gas temperature and exit velocity are based on a lower bound of all scrubbed

operation conditions for that unit from March 28 to August 20, 2017 for Unit 6; and from January 17 to August 20, 2017 for Unit 7.

(2) The high flow gas temperature and exit velocity is based on the upper bound of all scrubbed operation conditions for that unit from March to August, 2017 for Unit 6; and from January to August 2017 for Unit 7.

2.3 Nearby Background Sources As described in detail in the modeling protocol, based on available reported data and with the Colorado Draft Technical Support Document,9 it was determined that there no nearby sources need to be included explicitly in the dispersion modeling.

9 CDPHE Draft Technical Support Document. https://www.epa.gov/sites/production/files/2016-03/documents/co-epa-tsd-r2.pdf



2-4

AECOM

Figure 2-1: Location of Martin Drake


2-5

AECOM

Figure 2-2: Martin Drake Annual Emissions in 2006 – 2016


3-1

AECOM

3. Site-Specific Meteorological Database

3.1 Overview of the Monitoring Program The Martin Drake meteorological monitoring program was developed and conducted to support dispersion modeling for this facility. The meteorological monitoring program began in October 2015 and continued through the end of January 2017. The program consists of one station with a 30-m instrumented meteorological monitoring tower, supplemented by a Sound Detection and Ranging (SODAR) system. The program elements are fully described in the modeling protocol.

3.2 Description of Monitoring Locations The coordinates, as well as the elevation for the meteorological tower and SODAR, are provided in Tables 3-1 and 3-2 and are shown in Figure 3-1. The meteorological tower and SODAR were cited according to EPA Prevention of Significant Deterioration guidance to minimize the impacts of nearby structures to the greatest extent possible. Additionally, the SODAR was positioned north of the meteorological tower and on the northern edge of the holding ponds in order to minimize noise and other physical obstacles to the sound detection and ranging operation. A comprehensive description of the monitoring program is available in the CDPHE-approved QAPP and periodic data reports submitted to CDPHE.

Table 3-1: Meteorological Tower Location

Data Meteorological Monitoring Station

Latitude N38°49’14.66’’

Longitude W104°49’58.09’’

UTM North (m; NAD83) 4296897.4

UTM East (m; NAD83) 514514.4

UTM Zone 13

Elevation (Feet-mean sea level [msl]) 5,980

Table 3-2: SODAR Location

Data SODAR Station

Latitude N38°49’20.06’’

Longitude W104°50’00.8’’

UTM North (m; NAD83) 4297063.7

UTM East (m; NAD83) 514448.7

UTM Zone 13

Elevation (Feet-mean sea level [msl]) 5,970


3-2

AECOM

Figure 3-1: Location of the Meteorological Tower and SODAR Locations Relative to the Martin Drake Facility


4-1

AECOM

4. Modeling Procedures 4.1 Dispersion Model Selection This modeling analysis used the most recent version of the AERMOD dispersion modeling system to evaluate air quality impacts from the Martin Drake Power Plant. The software versions listed below were used:

• AERMOD version 16216r

• AERMAP version 11103

• AERSURFACE version 13016

• AERMET version 16216

• BPIPPRM version 04274

4.2 Land Use Classification Both the Auer land use method and the population-based method were previously conducted as described in the protocol and resulted in an urban classification. As a result of these analyses, the AERMOD option “URBANOPT” with a 2015 population estimate of 456,56810 was used.

4.3 Good Engineering Practice (GEP) Analysis A GEP formula stack height analysis was performed for Martin Drake in accordance with the EPA guidance11 and modeling protocol using the latest version of the EPA’s Building Profile Input Program software (BPIP-PRIME version 04274). BPIP-PRIME also generates building downwash parameters for use in AERMOD and automatically accounts for building separations by merging the influence of buildings.12 The dominant building is approximately 46 m tall which becomes 48.56 m tall as adjusted for the base elevation difference between the boiler stacks and the building. The projected building width for this building for each unit is 42.94 m for Unit 6 and 41.60 m for Unit 7. Thus, Hg = 48.56 + (1.5 x 42.94) = 113 m for Unit 6 and Hg = 48.56 + (1.5 x 41.60) = 111 m for Unit 7.

The actual physical height of the Unit 6 stack is of 61 m, which is 52 m shorter than GEP stack height. For Unit 7, the actual physical height is 76.2 m, which is about 35 m shorter than GEP stack height. Because the actual stack heights of the main boiler units are lower than the GEP stack height, the actual stack heights were used in this modeling application. The locations and dimensions of the buildings/structures relative to the exhaust stacks are depicted in Figure 4-1.

4.4 Meteorological Data Processing with AERMET AERMET is a multi-stage processing package that ultimately produces two meteorological input files used by AERMOD, the SURFACE file and the PROFILE file:

• SURFACE: a file with boundary layer parameters, such as sensible heat flux, surface friction velocity, convective velocity scale, vertical potential temperature gradient in the 500-meter layer above the planetary boundary layer, and convective and mechanical mixing heights. The file also provides values of Monin-Obukhov length, surface roughness, albedo, Bowen ratio, wind speed, wind direction, temperature, and heights at which measurements were taken.

• PROFILE: a file containing multi-level meteorological data with wind speed, wind direction, and temperature. For this application involving site-specific data, the PROFILE file contains data for up to 15 levels: 2, 10, 30, 60, 90, 120, 150, 180, 210, 240, 270, 300, 330, 360, and 390-m.

10 US Census Bureau, 2017. QuickFacts: Colorado Springs City, CO. https://www.census.gov/quickfacts/ 11 EPA, 1985. Guideline for Determination of Good Engineering Practice Stack Height. http://www.epa.gov/scram001/guidance/guide/gep.pdf 12 EPA, 1995. User’s Guide to the BPIP. https://www3.epa.gov/scram001/userg/relat/bpipd.pdf

https://www.census.gov/quickfacts/

https://www3.epa.gov/scram001/userg/relat/bpipd.pdf


4-2

AECOM

Site-specific surface meteorological data from the 30-m height instrumented tower and SODAR operated on the Martin Drake property from October 2015 until January 31, 2017 was used as input to the dispersion modeling. The 12-month time period from November 7, 2015 through November 6, 2016 had the most complete set of measurements across all measured parameters and was selected for modeling as discussed in the modeling protocol. The final validated dataset was submitted to CDPHE on July 27, 2017 and approved in an August 1, 2017 letter13 issued by CDPHE. The following files were approved for use in the final modeling:

• CSU_Validated_Tower_Data_Rev01.xslx, dated 5/17/2017.

• CSU_Validated_Sodar_Date_rev_July_2017.xlsx, dated 7/28/2017.

The surface data (wind direction, wind speed, temperature, solar radiation, delta-T, and relative humidity) used included tower measurements from 2-m to 30-m, supplemented at higher levels with the SODAR data starting at the 60-m level and extending to 390-m. Though turbulence measurements were available, they were not used consistent with the AERMOD Implementation Guide,14 when using the urban option.

Per CDPHE request15, AECOM has processed the site-specific data together with concurrent National Weather Service (NWS) upper air data from the Denver International Airport by AERMET using the Bulk Richardson Number approach to determining stability. The locations of the surface and upper air meteorological stations used in this analysis are shown in Figure 4-4.

Following processing, the data completeness of the combined meteorological parameters by quarter meets EPA’s 90% data capture requirement; therefore, missing meteorological data were not substituted or filled in. Wind roses for the on-site meteorological data are shown in Figures 4-2 and 4-3 for 10-m and 60-m levels (closest to the winds at stack top), respectively.

4.4.1 Surface Characteristics for Use in AERMET

AERMET requires specification of site surface characteristics including surface roughness (zo), albedo (r), and Bowen ratio (Bo). While reviewing comments received during public review, CDPHE decided16 that the area around the Martin Drake Power Plant has not changed significantly since before the 1990’s, and that surface characteristics should be developed using AERSURFACE according to the guidance provided by EPA in the recently revised AERMOD Implementation Guide (AIG). This was done instead of using AERGIS which CDPHE prefers to use in areas that have experienced significant development recently. AECOM followed the EPA’s AIG guidance to determine the surface characteristics.

The revised AIG provides the following recommendations for determining the site characteristics:

1. The determination of the surface roughness length should be based on an inverse distance weighted geometric mean for a default upwind distance of 1 km relative to the measurement site. Surface roughness length may be varied by sector to account for variations in land cover near the measurement site; however, the sector widths should be no smaller than 30 degrees.

2. The determination of the Bowen ratio should be based on a simple un-weighted geometric mean (i.e., no direction or distance dependency) for a representative domain, with a default domain defined by a 10-km by 10-km region centered on the measurement site.

3. The determination of the albedo should be based on a simple un-weighted arithmetic mean (i.e., no direction or distance dependency) for the same representative domain as defined for Bowen ratio, with a default domain defined by a 10-km by 10-km region centered on the measurement site.

13 Nancy Chick (CDPHE) to Robert Iwanchuk (AECOM) letter on August 1, 2017. 14 EPA, December 2016. AERMOD Implementation Guide. https://www.epa.gov/ttn/scram/models/aermod/aermod_implementation_guide.pdf 15 Email from Emmett Malone (CDPHE) to Tom Damiana (AECOM) on August 17, 2017 with the subject “Martin Drake AERMET Processing Files”. 16 Email from Lisa Devore (CDPHE) to Tom Damiana and Olga Samani (AECOM) on August 21, 2017 with the subject “Martin Drake AERMET Processing Files”.

https://www.epa.gov/ttn/scram/models/aermod/aermod_implementation_guide.pdf


4-3

AECOM

The AIG recommends that the surface characteristics be determined based on digitized land cover data.

Accordingly, AERSURFACE was used to determine the site characteristics based on digitized land cover data following the AIG recommendations discussed above. AERSURFACE incorporates look up tables of representative surface characteristic values by land cover category and seasonal category. AERSURFACE was applied with the instructions provided in the AERSURFACE User’s Guide.

The current version of AERSURFACE uses digital land cover data from the 1992 National Land Cover Database (1992 NLCD) archives. The 1992 NLCD archive provides data at a spatial resolution of 30 meters based upon a 21-category classification scheme applied over the continental U.S. The AIG recommends that the surface characteristics be determined based on the land use surrounding the site where the surface meteorological data were collected.

As recommended in the AIG for surface roughness, the 1-km radius circular area centered at the tower site was divided into twelve sectors for the analysis; each sector generally has a unique mix of land uses.

In AERSURFACE, the various land cover categories are linked to a set of seasonal surface characteristics. As such, AERSURFACE requires specification of the seasonal category for each month of the year. The following five seasonal categories are supported by AERSURFACE, with the applicable months of the year which were specified for this site.

1. Midsummer with lush vegetation (June – August)

2. Autumn with un-harvested cropland (September – November)

3. Late autumn after frost and harvest, or winter with no snow (December, January – February, if applicable)

4. Winter with continuous snow on ground (December, January – February, if applicable)

5. Transitional spring with partial green coverage or short annuals (March – May)

For Bowen ratio, the land use values are linked to three categories of surface moisture corresponding to average, wet and, dry conditions. The surface moisture condition for the site may vary depending on the meteorological data period for which the surface characteristics are applied. AERSURFACE applies the surface moisture condition for the entire data period. Therefore, if the surface moisture condition varies significantly across the data period, then AERSURFACE can be applied multiple times to account for those variations. As recommended in AERSURFACE User’s Guide, the surface moisture condition for each month was determined using precipitation for a 30-year period of data, selecting “wet” conditions if precipitation is in the upper 30th-percentile, “dry” conditions if precipitation is in the lower 30th percentile, and “average” conditions if precipitation is in the middle 40th-percentile. The surface moisture condition was determined using precipitation from Colorado Springs Municipal Airport Station. Snow cover was determined using snow depth data from the NOAA Online Weather Data (NOWData) where months with 50% measurable snow depth or greater at the nearest NWS surface station, Colorado Springs Airport, are designated as “Winter with continuous snow on the ground”.

4.5 Receptor Grid Receptors were placed in a nested Cartesian grid centered on Martin Drake in areas of ambient air. The following spacing was used for receptor placement:

• 50 meters along the secured property boundary which in this case is the facility security fence,

• 100 meters out to a distance of 5 km,

• 250 meters between 5 and 7.5 km,

• 500 meters between 7.5 and 10 km, and

• 1,000 meters between 10 and 20 km.


4-4

AECOM

This receptor grid is more refined than that required in the CDPHE Modeling Guidance. No additional receptors were needed as the location of the peak concentrations were in the area of 100-m spaced receptors.

The current version of AERMAP was used to process terrain data to assign receptor elevations. The appropriate terrain files for 1/3 arc-second, or 10 m, USGS National Elevation Dataset (NED) data were obtained from the Multi-Resolution Land Characteristics Consortium.17

A far-field view of the receptor grid as well as a near-field view including the property boundary is shown in Figures 4-5 and 4-6, respectively.

4.6 Review of Ambient Background Monitoring Data Data collected at the Rocky Mountain Steel Mill (RMSM) facility’s Print Shop SO2 monitor located in Pueblo, CO (shown in Figure 4-4) and analyzed by CDPHE, was used to establish the background concentration used for the compliance demonstration. The CDPHE determined the background concentration of 12 ppb was based on the 3-year average (2013-2015) 1-hour SO2 99th percentile design concentration from the monitor.

4.7 Model Configurations and Options AERSURFACE, AERMET and AERMOD were run with default options using urban dispersion coefficients. As discussed in the protocol, the site-specific turbulence data was not used in this modeling application.

17 Multi-Resolution Land Characteristics Consortium (MRLC). http://www.mrlc.gov/viewerjs/

http://www.mrlc.gov/viewerjs/


4-5

AECOM

Figure 4-1: Stacks and Buildings Used in the GEP and Building Downwash Analysis for Martin Drake


4-6

AECOM

Figure 4-2: 10-meter On-site Wind Rose (11/7/2015 – 11/6/2016)

Figure 4-3: 60-meter On-site Wind Rose (11/7/2015 – 11/6/2016)


4-7

AECOM

Figure 4-4: Location of the Meteorological Stations and Ambient SO2 Monitor for Background Concentrations

Figure 4-5: Far-field Receptor Grid for Modeling


4-8

AECOM

Figure 4-6: Near-field Receptor Grid for Modeling


5-1

AECOM

5. Modeling Results and Conclusion 5.1 NAAQS Compliance Modeling Results Martin Drake Units 6 and 7 were modeled with a constant 1-hour emission rate developed to account for short-term fluctuations in operations. The emission rate was derived from the facility 30-day permitted SO2 emission limit using a ratio recommended by EPA as described in the protocol. The exhaust temperature and velocity are representative of high load scrubbed normal operations developed using CEMs data collected during 2017. The resulting modeled stack parameters are shown in Table 2-1. Using this input data and site-specific meteorological data, the maximum design concentration predicted over the modeling domain is listed in Table 5-1 and demonstrates compliance with the NAAQS by a large compliance margin (cumulative impacts are less than half of the NAAQS). Figure 5-1 and Figure 5-2 show the far-field and near-field concentration isopleths of the design concentrations used to demonstrate compliance with the NAAQS, respectively. Table 5-1: AERMOD-Predicted Design 1-hour SO2 Concentration

AERMOD-Predicted Design

Concentration (µg/m3)

Receptor Location UTM Zone 13, NAD83

(m)

Background Design

Concentration (µg/m3)

Total Design Concentration

(µg/m3)

NAAQS (µg/m3)

64.50

UTMe=515034.0, UTMn=4296954.5

(~1 km to the southeast of

the stacks)

31.44 95.94 196.5

This dispersion modeling of Martin Drake using AERMOD was conducted using EPA- and CDPHE-approved models, data and assumptions described in this report and approved modeling protocol. Modeling input and output files (including meteorological data processing, emissions, exhaust parameters, CEMs data used to develop stack parameters, and building information), are provided separately from this report in a model archive to CDPHE and EPA for review.

5.2 Sensitivity of the Results to Operational Load As detailed in Section 2.2, an analysis has been conducted to evaluate the sensitivity of the compliance demonstration (reference case) to a broad range of facility operations by modeling allowable emissions at both low velocity/low temperature and high velocity/high temperature stack exit cases for scrubbed operations. These cases bracket the plume dispersion conditions affecting facility impacts.

Both cases were modeled with AERMOD for the full year of site-specific meteorological data with identical reference case emission rates (the same as those used in the NAAQS compliance modeling discussed above). The sensitivity modeling results, shown in Table 5-2, indicate that with the same, maximum, emission rate as the reference case, the low velocity/low temperature case results still demonstrates compliance with the NAAQS. Therefore the results of this sensitivity analysis do not alter the conclusions of the compliance demonstration.

Table 5-2: Sensitivity Analysis Results

Case Facility-Only, AERMOD-Predicted Design Concentration (µg/m3)

Low Temperature/Velocity Case 122.64

Reference Case 64.50

High Temperature/Velocity Case 47.36


5-2

AECOM

Figure 5-1: Far-Field Isopleths of 1-Hour SO2 Design Concentrations


5-3

AECOM

Figure 5-2: Near-Field Isopleths of 1-Hour SO2 Design Concentrations

AECOM

Aecom.com

Technical Support Document

1-Hour SO2 National Ambient Air Quality Standard

Air Quality Redesignation Recommendation

for Martin Drake Power Plant

September 6, 2017

Colorado Department of Public Health and Environment Air Pollution Control Division

4300 Cherry Creek Drive South Denver, Colorado 80246

1 | P a g e

Table of Contents

Introduction .......................................................................................... 2

Designation and Boundary Recommendations ................................................. 2 EPA Guidance on SO2 Designations ................................................................ 2

EPA SO2 Designations Process ......................................................................... 3

Analysis of Source Specific Designations .......................................... 3

Colorado Springs Utilities – Martin Drake Power Plant ............................... 3

Air Quality Characterization: Ambient Monitoring or Dispersion Modeling Results . 3

SO2 Ambient Monitoring in El Paso County ...................................................................... 3

SO2 Dispersion Modeling Results ........................................................................................ 4

Emissions-Related Data ....................................................................................................... 9

Meteorology (Weather & Transport Patterns) .............................................................. 11

Geography and Topography (Mountain Ranges or Other Air Basin Boundaries) ...... 14

Jurisdictional Boundaries .................................................................................................. 14

Other Relevant Information ............................................................................................. 15

Conclusion ........................................................................................................................... 16

Figures

Figure 1: Cumulative Modeled concentrations (µg/m3) with background all domain ... 7 Figure 2: Cumulative modeled concentrations (µg/m3) with background around the Drake Power Plant ................................................................................. 8 Figure 3: Meteorological Stations and Background Concentration SO2 Monitor ......... 12 Figure 4: 10-meter On-site Wind Rose (November 7, 2015 - November 6, 2016) ...... 13 Figure 5: 60-meter On-site Wind Rose (November 7, 2015 - November 6, 2016) ...... 14 Figure 6: CDPHE Community Health Equity Map Example (Income/Poverty Rate) ..... 15 Figure 7: Martin Drake Power Plant Boundary Recommendation ......................... 17

Tables

Table 1: CDPHE Highway 24 Monitor Data (2013 - present) ................................. 4 Table 2: Summary of the Modeled Design Concentration .................................... 6 Table 3: Summary of the Modeled Design Concentration .................................... 9 Table 4: Martin Drake Power Plant SO2 Emissions .......................................... 10

2 | P a g e

Introduction

Designation and Boundary Recommendations This Technical Support Document (TSD) provides the basis for the source specific sulfur dioxide (SO2) National Ambient Air Quality Standard (NAAQS) air quality redesignation and boundary recommendation for Martin Drake Power Plant (Drake). EPA previously designated this facility unclassifiable as part of the “Round 2” process for the 2010 1-hour SO2 NAAQS as part of the Consent Decree requirements. On July 1, 2016, Colorado informed EPA that Drake would be subject to a federally enforceable emission limit. The Division used modeled results, data from the current ambient monitoring network, and permit limits to propose SO2 designation or redesignation recommendations for seven Colorado facilities as attainment/unclassifiable, which EPA has since agreed with. In a July 1, 2016 letter, Colorado opted to limit the SO2 emissions from the Martin Drake Power Plant to 2,000 tons per year or less. The Division did so on December 27, 2016 by revising the plant’s federally enforceable construction permit to prohibit future operation of Unit 5 and reduce emissions from Units 6 and 7. This action satisfied the Data Requirements Rule (DRR) and Consent Decree requirements affecting the Martin Drake Power Plant.

Although there are no remaining DRR or Consent Decree requirements applicable to the Martin Drake Power Plant, the Division committed to the Colorado Air Quality Control Commission in the fall of 2015 that it would further evaluate SO2 impacts in the vicinity of the plant. Modeling results submitted in August 2017 based on site-specific meteorological data collected from October 2015 through January 2017 demonstrate attainment of the 2010 1-hour SO2 NAAQS. This technical support document provides details of this modeling analysis, including the Division’s review of the results, along with other considerations that support redesignating the area as attainment/unclassifiable.

EPA Guidance on SO2 Designations The Division is providing this additional information in response to updated designations guidance issued by EPA through a March 20, 2015 memorandum which identifies factors that the EPA intends to evaluate in determining area boundaries under the 2010 SO2 NAAQS. Five factors outlined in this memorandum are:

1) Air quality characterization via ambient monitoring or dispersion modeling results; 2) Emissions-related data; 3) Meteorology; 4) Geography and topography; 5) Jurisdictional boundaries.

In this TSD, the Division considers each of these five factors along with other relevant information regarding each of the detailed area designations and boundary recommendations for the Martin Drake Power Plant. The Division requests EPA consider this additional information prior to finalizing this redesignation.

3 | P a g e

EPA SO2 Designations Process The DRR establishes an SO2 emissions applicability threshold of 2,000 tons per year (tpy) that identifies priority sources subject to a source specific SO2 designation process. States have three options under the DRR to characterize current air quality in areas with large SO2 sources (2,000 tpy or greater): (1) establish federally enforceable emission limits (under 2,000 tpy) by January 13, 2017; (2) conduct air quality modeling by January 13, 2017; or (3) begin operating an appropriate monitoring network by January 1, 2017. The EPA will promulgate Round 3 SO2 designations (by Court Order) no later than December 31, 2017. For sources that are monitored and any remaining undesignated areas, EPA will promulgate Round 4 SO2 designations (by Court Order) no later than December 31, 2020.

The Division and Commission have made designation recommendations for all of Colorado’s applicable SO2 sources under Round 2 and 3. EPA designated Round 2 facilities on July 2, 2016. On August 22, 2017, EPA sent Colorado a 120-day letter that stated its’ agreement with Colorado’s recommendations for all remaining undesignated areas in the state. There will be no Round 4 designations in Colorado.

Analysis of Source Specific Designations This section of the TSD analyzes each of the five factors set forth in EPA’s March 20, 2015 memorandum for the Martin Drake Power Plant. The redesignation and boundary recommendation is presented at the conclusion of the five-factor analysis.

Colorado Springs Utilities – Martin Drake Power Plant

Air Quality Characterization: Ambient Monitoring or Dispersion Modeling Results

SO2 Ambient Monitoring in El Paso County Currently, there are four SO2 monitoring locations in operation in Colorado, of which one is located in Colorado Springs at Highway 24 and 8th Street (AQS-ID: 08-041-0015). Absent representative meteorological data for modeling, the existing “Highway 24” carbon monoxide site was selected to place a SO2 monitor to meet the monitoring requirements for the 2010 standard for the Colorado Springs area. This location was expected to be a maximum concentration SO2 location for population exposure from both the Drake Power Plant emissions and from diesel truck traffic. Monitored concentrations are decreasing due primarily to a reduction in emissions from the Drake Power Plant. The Division regularly evaluates its monitoring network and may consider relocating this site or adding another site in the future, depending on the results of the final modeling analysis and site logistics. The primary SO2 National Ambient Air Quality Standard (75 ppb) is the 99th percentile of 1-hour daily maximum concentrations, averaged over three years. This means

4 | P a g e

that, depending on monitoring data completeness, there could be approximately three exceedances of the standard in each year of a three-year period before the monitoring data would be considered in violation of the NAAQS. Historical monitoring data for SO2 in Colorado Springs (and the rest of the state) never approached the level of any SO2 standard prior to the installation of the Highway 24 monitor in January 2013. There has been substantial SO2 monitoring in the Colorado Springs area, with up to ten monitors operating during different periods between 1988 and 2007. For 2017, the 1-hour daily maximum is 17 ppb (as of July). The Highway 24 monitor has not recorded enough exceedances to place the region out of attainment of the 2010 SO2 NAAQS. The last exceedance was in March 2015. The 99th percentile value through the end of 2016 remains below the 75 ppb 1-hour NAAQS, with the 2014-2016 design value at 52 ppb. Table 1 below details the concentrations at the Highway 24 monitor over the past three years. Table 1: CDPHE Highway 24 Monitor Data (2013 - present)

AQS Site ID

Site “Name” Address Year

1st Max 1-hour (ppb)

2nd Max 1-hour (ppb)

99th% 1-hour (ppb)

08-041-0015

Highway 24

690 W. Highway 24

2014 83 57 57

2015 87 70 53

2016 61 57 45

2014-2016

n/a n/a 52

SO2 Dispersion Modeling Results The Division coordinated with CSU to conduct SO2 modeling for the area surrounding Martin Drake using validated on-site meteorological data collected between October 2015 and January 2017. CSU and its consultant, AECOM, submitted a draft modeling protocol to the Division in March 2017. The Division and EPA each reviewed this draft protocol and requested revisions. CSU/AECOM revised the protocol as requested, termed “Martin Drake Power Plant SO2 Modeling Protocol: June 2017” (Attachment 1). The Division posted the draft modeling protocol and agency comments for a 30-day public comment period that ended on July 21, 2017. The Division received seven comments on the draft modeling protocol, posted all comments on the Division’s website, and responded to the comments on August 24, 2017 both via email and on the website. As a result of the public comments, at the request of the Division, CSU and AECOM evaluated the sensitivity of potential regional SO2 impacts when Martin Drake operates at various stack parameters (maximum and minimum temperature and velocity under scrubbed conditions). The Division believes that the modeling protocol is appropriate and consistent with Colorado and EPA rules and guidance.

CSU/AECOM submitted a final modeling report on August 29, 2017 demonstrating attainment of the 2010 1-hour SO2 NAAQS as the “Martin Drake Power Plant SO2

5 | P a g e

NAAQS Compliance Modeling Report: August 2017” (Attachment 2). Permit applicants and facilities routinely submit modeling protocols and analyses to the Division, which the Division then reviews in detail. The protocol adheres to the applicable rules, regulations, and principles in the “Colorado Modeling Guideline.”1 The Division completed a high-level review of the final modeling results to determine compliance with applicable requirements and did not note any major issues. Additionally, the Division concurs with the final modeling protocol. The Division will include a final comprehensive modeling report as an addendum to this TSD. This modeling analysis used the most recent version of the AERMOD dispersion modeling system to evaluate air quality impacts from the Martin Drake Power Plant. The software versions listed below were used: • AERMOD version 16216r • AERMAP version 11103 • AERSURFACE version 13016 • AERMET version 16216 • BPIPPRM version 04274

In both the Division’s review and the CSU/AECOM modeling analysis, ground-level

concentrations (with the background concentration added) were compared to the 1-

hour SO2 standard. Spatial contour plots of the modeled concentrations are included.

All input and output files related to the modeling were provided in the final report,

including all emissions information and spreadsheets, data associated with the

background concentration, building specifications, and other relevant information

used to support the assumptions in the modeling analysis.

1-hour SO2 NAAQS Assessment: Normal Operations

The modeled design concentration was calculated by AERMOD as 99th percentile of the

annual distribution of the daily maximum 1-hr SO2 concentration for the year of

meteorological data used. This concentration was calculated at the allowable

emission rate (adjusted per EPA guidance to appropriate 1-hour rates); actual

emissions are likely to be lower. This is equivalent to the highest fourth highest

modeled daily maximum 1-hr concentration and the result can be seen in Table 2

below.

1 https://www.colorado.gov/airquality/permits/guide.pdf

https://www.colorado.gov/airquality/permits/guide.pdf

6 | P a g e

Table 2: Summary of the Modeled Design Concentration

Ranked Value

Concentration (ppb)

Receptor Location (X, Y UTM Z13 NAD27)

Background Concentration (ppb)

Total Impact (ppb)

1-hr SO2 NAAQS (ppb)

H4H 24.62 515034.0 E 4296954.5 N

12.00 36.62 75

As shown, the total impact is below the corresponding 1-hr SO2 NAAQS. The location

of the highest impact is about one kilometer to the southeast of the Drake facility;

the overall distribution of concentration can be seen in Figure 1 and Figure 2. Please

refer to the “Martin Drake Power Plant SO2 NAAQS Compliance Modeling Report:

August 2017” and associated files submitted to the Division and EPA electronically on

August 31, 2017 for more detailed information.

7 | P a g e

Figure 1: Cumulative Modeled concentrations (µg/m3) with background all domain

8 | P a g e

Figure 2: Cumulative modeled concentrations (µg/m3) with background around the Drake Power Plant

9 | P a g e

1-hour SO2 NAAQS Assessment: Variable Operations

CSU/AECOM conducted sensitivity modeling analyses per the Division’s request to

evaluate the associated concentration impacts during varying stack conditions for

temperature and velocity. The sensitivity modeling consists of modeling the minimum

and maximum bounds for these two parameters during scrubbed operation conditions.

Both of these scenarios maintained emissions at full capacity as described in the

“Martin Drake Power Plant SO2 Modeling Protocol: June 2017.”

The modeled design concentration for each of these scenarios was calculated by

AERMOD as 99th percentile of the annual distribution of the daily maximum 1-hr SO2

concentration for the year of meteorological data used. This is equivalent to the

highest fourth highest modeled daily maximum 1-hr concentration and the result can

be seen in Table 2 below.

Table 3: Summary of the Modeled Design Concentration

Scenario Ranked

Value Concentration (ppb)

Background Concentration (ppb)

Total Impact (ppb)

1-hr SO2 NAAQS (ppb)

Low

Temperature/

Velocity Case

H4H 46.81 12.00 58.81 75

High

Temperature/

Velocity Case

H4H 18.08 12.00 30.08 75

As shown, the total impact is below the corresponding 1-hr SO2 NAAQS. The location

of the highest impact will be included in the addendum. Please refer to the “Martin

Drake Power Plant SO2 NAAQS Compliance Modeling Report: August 2017” and

associated files submitted to the Division and EPA electronically on August 31, 2017

for more detailed information.

Emissions-Related Data Martin Drake Power Plant – SO2 Emissions Analysis All three Martin Drake units have seen dramatic changes in their emission rates over the last 12-18 months. Martin Drake Unit 5 stopped operating in March 2016; the

10 | P a g e

shutdown of the unit became legally enforceable in a construction permit on December 31, 2016. SO2 control systems began operating on Drake Units 6 and 7 in May and September 2016, respectively. The SO2 controls are still undergoing operational and reliability testing. It is typical for large, custom-built emission control systems to undergo a lengthy period of testing and adjustment. Emission controls may operate intermittently during testing, causing emissions to fluctuate in ways that are not representative of normal operation. This fact does not indicate that the emission controls would be periodically shut down or bypassed during normal plant operations. CSU accepted a federally enforceable plant-wide SO2 emission limit of 1,995 tons per year per the DRR, which is below the 2,000 tons per year threshold. This new emission limit became effective on December 27, 2017 and will reduce actual SO2 emissions by approximately 2,000 tons per year in 2017 and future years when compared to the facility’s 2015 actual emissions. Drake Units 6 and 7 will be subject to mandatory 30-day rolling emission limits starting December 31, 2017 as required in Colorado Regional Haze State Implementation Plan (RH SIP) which will reduce actual SO2 emissions by approximately 2, . Emissions from the Drake Power Plant are shown in Table 4 below (from the EPA Air Markets Program Data system). Table 4: Martin Drake Power Plant SO2 Emissions

Year* Unit Number of

Months Reported

SO2 Annual Emissions

(tons/year)

SO2 Average 30-day Rolling Emission Rate

(lb/MMBtu)

2012

5

12 1,108 0.53

2013 12 982 0.49

2015 12 580 0.42

2016 12 6.5 0.25

2012

6

12 1,680 0.57

2013 12 1,595 0.51

2015 12 1,448 0.47

2016 12 724.1 0.31

2012

7

12 2,004 0.58

2013 12 2,004 0.53

2015 12 1,932 0.47

2016 12 891.8 0.22

2012

All

12 4,707 0.56

2013 12 4,580 0.51

2015 12 3,960 0.45

2016 12 1,622 0.26

2015** 6 & 7 12 3,379 0.47 *Drake Power Plant experienced a fire event that resulted in the units not operating for various periods of time. Unit 6 was down for nine weeks. Unit 5 was down for 16 weeks. Therefore, 2014 is not included in recent emission evaluations. **This row is included since Unit 5 operated at 70% of average operating hours (years 2006-2013) in the year 2015 to give perspective regarding the operation of Units 6 and 7.

11 | P a g e

SO2 Emissions in El Paso County The most current comprehensive emission inventory available from Division records for El Paso County indicates that approximately 99% of the 2013 SO2 emissions in the county are from point sources. The Nixon Power Plant, approximately 15 miles away from Drake, is the only other significant SO2 source in El Paso County, and contributed about 46% of point source SO2 emissions in 2013 while Drake contributed 53%. As discussed in the modeling section, air quality impacts from the Drake Power Plant have not previously been estimated due to the lack of available representative meteorological data. Prior analyses indicate that SO2 emission concentration gradients and potential hotspots from the Drake and Nixon Power Plants do not overlap because of geographical distance and terrain features that result in each plant existing in separate airsheds with regard to SO2.

Meteorology (Weather & Transport Patterns) As noted, the Division previously determined there were no representative meteorological datasets for the transport and dispersion conditions at the Drake Power Plant. EPA agreed with this determination in its February 16, 2016 letter and associated draft technical support document. As a result, CSU collected site-specific meteorological data from October 18, 2015 through January 31, 2017. The Division then conducted a comprehensive Quality Assurance process to validate the data, including reviewing the CSU contractor's audits on SODAR performance, checking project data recovery and validity, and looking at seasonal time period completeness and representativeness. We also reviewed the contractor's records concerning calibration of sensors for wind direction and speed, and the other sensors. The 12-month time period from November 7, 2015 through November 6, 2016 had the most complete set of measurements across all measured parameters and was used in the dispersion modeling analysis. The locations of the surface and upper air meteorological stations as well as the SO2 monitor used for the background concentration applied in the dispersion analysis are shown in Figure 3. Wind roses for the on-site meteorological data are shown in Figure 4 and Figure 5 for 10-meter and 60-meter levels (closest to the winds at stack top), respectively.

12 | P a g e

Figure 3: Meteorological Stations and Background Concentration SO2 Monitor

13 | P a g e

Figure 4: 10-meter On-site Wind Rose (November 7, 2015 - November 6, 2016)

14 | P a g e

Figure 5: 60-meter On-site Wind Rose (November 7, 2015 - November 6, 2016)

Geography and Topography (Mountain Ranges or Other Air Basin Boundaries) Fountain Creek and Monument Creek converge just to the west of the Drake Power Plant, which directly impact the plume and wind flows. There are two prominent terrain features that affect wind conditions at the Drake Power Plant: Pikes Peak approximately two miles to the west and the Palmer Divide approximately five miles to the north.

Jurisdictional Boundaries Clearly defined legal boundaries can be used to determine an appropriate geographic area regarding potential NAAQS impacts from the Martin Drake Power Plant. The city of Colorado Springs is the largest city in Colorado by land area, encompassing 194 square miles, and the second largest in population (445,830 as of 20142). Many scattered unincorporated county designated areas are enclosed within city limits. The Drake Power Plant is located at the western edge of downtown Colorado Springs. Therefore, it is appropriate to examine roadway intersections as major defining lines

2 All population statistics are from the United States Census Bureau.

15 | P a g e

for the designation boundary since city limits extend much farther east than the areas that monitoring, meteorological and topographical analyses suggest are being impacted. The Division also considers the city of Manitou Springs to be a clearly defined legal boundary that should be included in the designation area due to Fountain Creek airflows and potential impacts in the areas of the foothills.

Other Relevant Information The Division used CDPHE’s public-facing Community Health Equity Map (2011-2015 data)3 to examine census-tract level geographic disparities for selected social determinants of health, including income/poverty levels and race/ethnicity population percentages, and key health conditions and outcomes, including asthma-related hospitalization rates, heart disease mortality rates, and preventable conditions. An example of these maps (income/poverty rates) is shown below in Figure 6. Several areas in the southern area of the city clearly indicate above average poverty rates, increased minority populations as well as higher than state averages for asthma-related hospitalization rates and heart disease mortality rates. The Division also reviewed population density using this Map to incorporate appropriate potentially affected populations and assess population density in the mountains west of Manitou Springs and Colorado Springs. This information is important in considering boundary area recommendations for a source located in an urbanized area with varied population factors. Figure 6: CDPHE Community Health Equity Map Example (Income/Poverty Rate)

3 http://www.cohealthmaps.dphe.state.co.us/cdphe_community_health_equity_map/ -

http://www.cohealthmaps.dphe.state.co.us/cdphe_community_health_equity_map/

16 | P a g e

Conclusion Upon consideration of the five factors and other relevant information, CDPHE recommends redesignating the area around the Martin Drake Power Plant as an attainment area for purposes of the 2010 one-hour SO2 NAAQS. CDPHE recommends establishing a boundary around a portion of the city of Colorado Springs, including enclosed unincorporated county areas, bounded to the north by East Woodmen Road, North Academy Boulevard, and city limits, to the east by North/South Powers Boulevard, and to the south and west by city limits, with the addition of the ‘census designated place’ termed “Stratmoor” bounded by South Academy Boulevard remain as the boundary for the attainment/unclassifiable area redesignation recommendation as shown in Figure 74. This conclusion is based on the following information:

Monitoring and emissions-related data shows that this boundary incorporates the primary source (Martin Drake Power Plant). Emissions from the only other notable SO2 source in the vicinity are in a separate airshed;

Meteorological and topographical information indicate that potential impacts are contained within this boundary;

The mountains that generally begin just to the west of the cities of Colorado Springs and Manitou Springs constitute a geographical boundary limit;

Basing the area boundary on the guideline 10 km radius for a modeling domain is not appropriate because of the complex terrain and urban demographics;

Affected populations, including sensitive subpopulations, within the city of Colorado Springs, including those living in unincorporated enclosed county areas, and the city of Manitou Springs are located within this boundary. EPA has previously approved area boundaries for other Round 2 SO2 designation boundaries that align with roadways, rather than the outermost jurisdictional boundaries of a unit of local government5.

SO2 emissions from the Drake Power Plant are significantly reduced with the installation of controls on Units 6 and 7 and the shutdown of Unit 5.

CSU accepted a federally enforceable plant-wide SO2 emission limit of 1,995 tons per year, effective December 27, 2016. This limit will reduce actual SO2 emissions by approximately 2,800 tons annually in 2017 and future years.

The total modeled impact is below the corresponding 1-hr SO2 NAAQS.

4 City boundaries applicable as of April 2016. 5 North Dakota Mercer County (portion) SO2 Designation: road boundaries specified in Technical Support Document (February 16, 2016)

17 | P a g e

Figure 7: Martin Drake Power Plant Boundary Recommendation

Date post:	22-Apr-2018
Category:	Documents
Upload:	trancong
View:	217 times
Download:	3 times

Martin Drake Power Plant SO NAAQS Compliance Modeling Report€¦ · · 2017-09-11Continuous...

Documents