United States Office of Research and EPA/600/R-18/094
Environmental Protection Development April 2018
Agency. Washington DC 20460
&EPA
Summary Report of the
Special Reactive Nitrogen
(Nr) Inter-Comparison Study:
Ammonia CASTNET CSN
Study (ACCS)
-------
Summary Report of the Special Reactive
Nitrogen (Nr) Inter-Comparison Study:
Ammonia CASTNET CSN Study (ACCS)
Ralph E. Baumgardner, Jr.
Exposure Methods and Measurement Division
National Exposure Research Laboratory
Office of Research and Development
Research Triangle Park, NC
Christopher M. Rogers
Amec Foster Wheeler Environment & Infrastructure, Inc.
Jacksonville, FL
Melissa A. Puchalski
Clean Air Markets Division
Office of Atmospheric Programs
Office of Air and Radiation
Washington, DC
Nealson Watkins
Air Quality Assessment Division
Office of Air Quality Planning and Standards
Office of Air and Radiation
Research Triangle Park, NC
Thomas F. Lavery
Environmental Consultant
Cranston, Rl
Kevin P. Mishoe
Amec Foster Wheeler Environment & Infrastructure, Inc.
Newberry, FL
Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect
official Agency policy. Mention of trade names and commercial products does not constitute
endorsement or recommendation for use.
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Office of Research and Development
Research Triangle Park, NC 27711
-------
Table of Contents
I.0 Executive Summary 1
2.0 Introduction 4
2.1 Background 4
2.2 Purpose of Study 4
2.3 Supporting Agencies 5
3.0 Equipment and Site Specifications/Selection 7
3.1 Sampling Equipment Considered/Evaluated 7
3.2 Sample Site Selection 8
3.3 Monitoring Site Modifications 11
3.4 Equipment Modifications 11
3.5 Calibration Procedures 11
3.6 Sample Preparation and Shipping 12
4.0 Test Site Results 13
4.1 Sampling Period 13
5.0 Network Sampling 20
5.1 Sampling Schedule and Laboratory Analyses 20
5.2 Data Collection and Handling 20
5.3 Data reporting 21
6.0 Quality Assurance/Quality Control 22
7.0 Results of NH3 Concentration Measurements 23
8.0 Evaluation of Filter Pack S02, Total N03, S024, and NH+4 Concentration Measurements 26
8.1 S02 Concentrations 26
8.2 Total N03 Concentrations 26
8.3 Particulate S024 Concentrations 31
8.4 Particulate NH4 Concentrations 31
9.0 Evaluation of S024, N03, and NH+4 Concentrations Measured on Filter Pack, ADS and CSN
SuperSASS Ion Module 37
9.1 Particulate S04 Concentrations 37
9.2 Particulate N03 Concentrations 39
9.3 Particulate NH4 Concentrations 40
10.0 Dry Deposition of Atmospheric Nitrogen 44
II.0 References 51
List of Appendices
Appendix A Monitoring Site and Sampler Modifications
Appendix B Sample/Sampler Preparation Procedures
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
List of Figures
Figure 1. Map of CASTNET/AMoN Sites Selected
Figure 2. Legend for Figures 3 through 8
Figure 3. S024 Results from Test Study
Figure 4. NH+4 Results from Test Study
Figure 5. NH3 Results from Test Study
Figure 6. S02 Results from Test Study
Figure 7. NHX Results from Test Study
Figure 8. N03 Results from Test Study
Figure 9. Sampling Schedule
Figure 10. Example Chain-of-Custody Form
Figure 11. Comparison between AMoN and ADS NH3 Concentrations at CTH110 / NY67
Figure 12. Comparison between AMoN and ADS NH3 Concentrations at PAL190 / TX43
Figure 13. Comparison between AMoN, ADS, and SuperSASS MPPD NH3 Concentrations
at ARE128 / PA00
Figure 14. Comparison between AMoN, ADS, and SuperSASS MPPD NH3 Concentrations
at CHE185 / OK99
Figure 15. Comparison between AMoN, ADS, and SuperSASS MPPD NH3 Concentrations
at ROM206 / C088
Figure 16. Comparison between CASTNET Filter Pack and ADS S02 Concentrations at
Three CASTNET Sites
Figure 17. Regression Analysis of CASTNET Filter Pack and ADS S02 Concentrations at
Three CASTNET Sites
Figure 18. Comparison between CASTNET Filter Pack and ADS Total N03 Concentrations
at Three CASTNET Sites
Figure 19. Regression Analysis of CASTNET Filter Pack and ADS Total N03
Concentrations at Three CASTNET Sites
Figure 20. Comparison between CASTNET Filter Pack, CSN SuperSASS Ion Module, and
ADS S04 Concentrations at Three CASTNET Sites
Figure 21. Regression Analysis of CASTNET Filter Pack and ADS S024 Concentrations at
Three CASTNET Sites
Figure 22. Comparison between CASTNET Filter Pack, CSN SuperSASS Ion Module, and
ADS NH+4 Concentrations at Three CASTNET Sites
Figure 23. Regression Analysis of CASTNET Filter Pack and ADS NH+4 Concentrations at
Three CASTNET Sites
Figure 24. Regression Analysis of CASTNET Filter Pack and CSN SuperSASS S04
Concentrations at Three CASTNET Sites
Figure 25. Regression Analysis of ADS and CSN SuperSASS S024 Concentrations at Three
CASTNET Sites
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
List of Figures (continued)
Figure 26.
Figure 27.
Figure 28.
Figure 29.
Figure 30.
Figure 31.
Figure 32.
Figure 33.
Figure 34.
List of Tables
Regression Analysis of CASTNET Filter Pack and CSN SuperSASS N03
Concentrations at Three CASTNET Sites
Regression Analysis of ADS and CSN SuperSASS N03 Concentrations at Three
CASTNET Sites
Regression Analysis of CASTNET Filter Pack and CSN SuperSASS NH+4
Concentrations at Three CASTNET Sites
Regression Analysis of ADS and CSN SuperSASS NH+4 Concentrations at Three
CASTNET Sites
Dry Deposition of Nitrogen for ARE128, PA
Dry Deposition of Nitrogen for CHE185, OK
Dry Deposition of Nitrogen for ROM206, CO
Dry Deposition of Nitrogen for CTH110, NY
Dry Deposition of Nitrogen for PAL190, TX
Table 1. Summary of Regression Analyses of Concentrations Measured with Filter Packs
and ADS
Table 2. CASTNET Sites Selected for Study
Table 3. Summary of Regression Analyses of Concentrations Measured with Filter Packs
and ADS by Site
Table 4. Summary of Regression Analyses of Concentrations Measured with CSN
SuperSASS Ion Modules, CASTNET Filter Packs, and ADS
Table 5. Summary of Measured Nitrogen Concentrations and Modeled Dry Deposition
Fluxes Calculated using ADS Concentrations
-------
1.0 Executive Summary
The primary purpose of the Ammonia CASTNET CSN Study (ACCS) was to conduct reactive
nitrogen measurement inter-comparison research at five Clean Air Status and Trends Network
(CASTNET) sites for one year. The objectives of the ACCS were to assess the precision,
accuracy, and bias of passive ammonia (NH3) samplers (Radiello), standard 3-stage CASTNET
filter packs, Met One SuperSASS NH3 denuders, and Met One SuperSASS ion canisters.
SuperSASS instruments are utilized by the Chemical Speciation Network (CSN). The
instruments were operated over one-week sampling periods, and single and duplicate annular
denuder systems (ADS) were used as the reference method. The study also evaluated sulfur
and oxidized nitrogen components measured using the various instrument systems.
Five study sites were selected based on location, land use type, predicted or known nearby
ammonia emission sources, site operator capability, and co-location with the National
Atmospheric Deposition Program (NADP) Ammonia Monitoring Network (AMoN). The sites
included Arendtsville, PA (ARE128); Cherokee Nation, OK (CHE185); Rocky Mountain National
Park, CO (ROM206); Connecticut Hill, NY (CTH110); and Palo Duro State Park, TX (PAL190).
Prior to the start of actual field sampling, three test phases were completed at the Amec Foster
Wheeler Gainesville, FL facility between April and June 2010. Two major modifications were
made to the method design for the ACCS because of the test phase results. A Teflon filter was
added to the ADS for the collection of particles, and 4-stage experimental filter packs were
excluded during the ACCS field campaign due to an inability to collect both gaseous sulfur
dioxide and NH3 using the same filter pack.
ACCS field sampling started at the end of August 2010 and continued through September 2011.
A principal goal of the field studies was to use measured data to quantify the performance of
AMoN passive NH3 samplers, CASTNET filter packs, SuperSASS mini-parallel plate denuders
(MPPD) in terms of measuring reactive nitrogen, and CSN SuperSASS ion modules to measure
particulate cations.
AMoN passive samplers performed very well at all five sites compared with the ADS reference
concentrations. Evaluation of the AMoN samplers was reported by Puchalski etal. (2015), who
performed detailed comparisons between AMoN and ADS measurements. They determined
that AMoN samplers were comparable to ADS with an overall mean relative percent difference
(MRPD) of -9 percent for the study.
SuperSASS MPPD samples consistently collected less NH3 than the reference ADS denuders
and results worsened over the course of the field sampling. This outcome was especially
evident at the high-elevation site at ROM206 where the concentrations sampled by the MPPD
system were less than half of the ADS values.
1
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
The CASTNET filter pack data compared well with ADS-measured concentrations for all
CASTNET analytes. Table 1 summarizes linear regression results for four filter pack parameters
compared with ADS measurements. Good correlation (R2 equal to approximately 0.9) was
exhibited for all four parameters with the best results for sulfur dioxide (S02) and the worst for
particulate ammonium (NH+4). Gaseous S02 data showed a slope (0.97) near unity with a small
y-intercept. The three results for particulate pollutants indicate the filter pack sampled higher
concentrations than the ADS, perhaps because the filter pack did not utilize a device to exclude
larger particles (Lavery eta!., 2009).
Table 1. Summary of Regression Analyses of Concentrations Measured with Filter
Packs and ADS
Parameter
n
ads_avg
FP
Slope
y-intercept
R2
S02
56
1.253
1.253
0.973
0.033
0.977
Total N03
54
1.444
1.800
1.108
0.201
0.891
so24
56
1.494
1.734
1.138
0.035
0.963
nh;
92
0.561
0.702
1.270
-0.002
0.907
Note: ads_avg is average of concentration measurements from dual ADS.
FP = filter pack
The results for sulfate (S024) were good for ARE128 with a regression line slope of 1.0349.
Higher filter pack S024 concentrations were measured at CHE185, again likely because of no
particle size selection device on the filter pack and the fact that CHE185 is more influenced by
fugitive dust than ARE128 and ROM206. S024 concentrations sampled at ROM206 were not as
comparable because concentration values were much lower, i.e., less than 20 percent of the
concentrations measured at ARE128 and CHE185.
Particulate pollutant measurements made by the CSN SuperSASS ion modules were consistent
with both ADS and filter pack measurements. The best comparisons were between SuperSASS
S04 data and both filter pack and ADS S04 data with correlation coefficients of 0.98. Good
results were also obtained between SuperSASS nitrate (N03) and both filter pack and ADS data
with correlation coefficients of 0.88 and 0.99, respectively. The coefficients for NH4 were 0.82
for filter pack data and 0.81 for ADS data.
CASTNET was designed to estimate dry deposition of sulfur and nitrogen pollutants across the
United States. Estimates of dry deposition of nitrogen, however, have been incomplete because
measurements of all species of nitrogen have not been available for the dry deposition
calculations. This study helps remedy that problem by providing data on NH3 and volatized NH4
concentrations. Estimates of dry deposition of nitrogen, including fluxes of NH3 and NHX, were
made for the five monitoring sites for the 20-week sampling period. The results demonstrate that
NH3 and nitric acid (HN03) were the principal constituents of dry deposition of nitrogen. The dry
nitrogen deposition estimates varied from a low of 0.029 kilograms per hectare per year
2
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
(kg ha"1yr1) at ROM206 to a high of 0.186 kg ha"1yr1 at PAL190. The contribution of NH3 to dry
deposition of nitrogen was greater than 70 percent at all five sites and greater than 80 percent
at PAL190 and CHE185, the two sites with the highest rates of deposition of nitrogen.
3
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
2.0 Introduction
2.1 Background
CASTNET is a long-term environmental monitoring program that measures changes in ambient
air quality and assesses atmospheric deposition over broad geographic regions of the United
States. Operating since 1987, CASTNET has evolved into a robust national monitoring program
that currently consists of 96 monitoring stations nationwide. The Environmental Protection
Agency (EPA) operates a majority of the CASTNET monitoring stations. In cooperation with
EPA, the National Park Service (NPS) currently operates 26 of the 96 stations and the Bureau
of Land Management (BLM) operates five sites in Wyoming. View the EPA CASTNET website,
https://www.epa.gov/castnet, for more information on CASTNET. NADP's AMoN operates NH3
samplers at 103 sites with 68 of the AMoN sites at or near CASTNET locations. CSN measures
fine particulate matter (PM2.5) and its chemical species at 151 sites throughout the United
States.
2.2 Purpose of Study
The primary purpose of the ACCS was to determine the comparability of several different types
of reactive nitrogen (Nr) measurements at five CASTNET sites for one year. Nr includes all
forms of nitrogen that are biologically, photochemically, and radiatively active. Compounds of
nitrogen that are reactive include nitrous oxide (N20), N03, nitrite (N02), NH3, and NH+4. The
current CASTNET 3-stage filter pack captures particulate NH+4 and particulate N03 on the first
(Teflon) filter. There is a known loss of particulate NH+4 and particulate N03 due to the
volatilization of ammonium nitrate (NH4N03) during the typical one week sampling period when
the filter pack is exposed on the sampling tower (Lavery et al., 2009). In addition to the bias
introduced by volatilization, an error is also introduced in the partitioning between particulate
N03 and gaseous nitric acid (HN03), which is collected on the second (nylon) filter
The goals of the ACCS were to assess the precision, accuracy, and bias of passive ammonia
samplers (Radiello), standard CASTNET filter packs, CASTNET filter packs with an additional
fourth stage filter impregnated with phosphorus acid (H3P03) to collect atmospheric NH3 and any
volatilized NH4 (an NHx-filter), Met One SuperSASS NH3 denuders, and Met One SuperSASS
ion canisters with duplicate ADS as the reference method.
Study sites were selected based on location, land use type, predicted or known nearby
ammonia emission sources, site operator capability, and co-location with AMoN sites, which
measure NH3 concentrations at 2-week intervals. The current AMoN protocol was used for the
passive Radiello samplers operated in ACCS. View the NADP website, http://nadp.slh.wisc.edu,
for more information on AMoN. The modified (4-stage) CASTNET filter pack was designed to
run for two 1-week sampling periods at the same locations as the 3-stage filter pack. The
SuperSASS NH3 denuder, SuperSASS ion canisters and ADS (EPA Modified Compendium
Method I.O. 4.2) were run for two 1-week periods every six weeks to save laboratory and
4
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
equipment costs. The passive samplers were analyzed by the NADP Central Analytical
Laboratory (CAL) and reported in the AMoN database, which is accessible via the AMoN web
page. The SuperSASS NH3 denuders and annular denuders were shipped to and analyzed by
the Amec Foster Wheeler laboratory. The SuperSASS ion canisters were shipped to the RTI
International (RTI) laboratory for analysis as part of CSN. All SuperSASS systems and the
reference ADS were operated in duplicate. Radiello samplers were deployed in triplicate every
six weeks at the five study sites.
2.3 Supporting Agencies
Funding for the work assignments that encompassed ACCS involved multiple EPA offices. The
partnering offices and their roles in accomplishing the various statements of work (SOW) were:
EPA Office of Air and Radiation (OAR)/Clean Air Markets Division (CAMD) CASTNET
IV Contract (Contract No. EP W 09 028), Task Order 0005:
This task order provided support for Amec Foster Wheeler-related field operations and
activities necessary to purchase the required equipment such as ADS, filter packs, and
mass flow controllers; perform testing of all systems in Amec Foster Wheeler's Gainesville,
FL office prior to deployment; prepare necessary training materials, including preparation
and revision of standard operating procedures (SOPs); and train technicians and site
operators on the SOPs.
EPA Office of Research and Development (ORD)/National Exposure Research
Laboratory (NERL) CASTNET IV Contract (Contract No. EP W 09 028), Task Order
0008:
This task order provided support for laboratory analysis costs for the Amec Foster Wheeler
laboratory covering 4-stage filter packs, ADS, and SuperSASS NH3 denuders. It also
supported the data analysis and reporting effort at Amec Foster Wheeler.
EPA/OAR/CAMD Support of AMoN (through NADP):
Analysis of the passive NH3 samples was supported as part of regular AMoN network
operation. For some sites, sponsorship of the NH3 laboratory analyses was provided by the
NADP site sponsor.
5
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
EPA Office of Air Quality Planning and Standards (OAQPS)/Air Quality Assessment
Division (AQAD):
OAQPS loaned to Amec Foster Wheeler 3 Met One SuperSASS speciation samplers,
16 SuperSASS denuders, and 8 Teflon-coated cyclones for use during the study period.
Equipment was returned to OAQPS following completion of the study. OAQPS also
supported (via the CSN contract) the analysis costs for the ion canisters that were run at the
sites operating the SuperSASS system.
6
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
3.0 Equipment and Site Specifications/Selection
3.1 Sampling Equipment Considered/Evaluated
Six different samplers were originally designated for inclusion in the ACCS. All measurements
were to be made at approximately 10 meters, which is the sampling height used for the
standard CASTNET 3-stage filter pack. Each of the proposed sample methods is described in
the following sections.
1. 3-Stage CASTNET filter pack
A 3-stage CASTNET filter pack was deployed for each sampling period as part of the routine
operations of CASTNET.
2. 4-Stage CASTNET style filter pack
In addition to the 3-stage CASTNET filter pack, a 4-stage CASTNET-style filter pack with the
final stage housing a H3P03-impregnated filter to collect NH3 and any volatilized particulate NH4
that escaped from the Teflon filter located in the first stage was deployed. The four stages were:
First stage: A Teflon filter that was analyzed using the standard CASTNET protocol for
particulate S024 and N03 by ion chromatography (IC) and particulate NH+4 by automated
colorimetry (AC).
Second stage: A nylon filter analyzed for gaseous S02 and HN03 by IC.
Third stage: Dual potassium carbonate (K2C03)-impregnated cellulose fiber filters that
were analyzed using the standard CASTNET protocol for S02 by IC.
Fourth stage: The added H3P03-impregnated filters were designed to be analyzed by AC
using deionized water (DIW) as the extraction fluid for NHX (particulate NH+4 lost from the
Teflon filter plus gaseous NH3 trapped on the nylon filter by other species collected on
the filter).
3. Dual ADS
The third sampler was a URG (URG Corporation, 2018) 2-denuder, multi-channel ADS that
featured:
A sodium carbonate (Na2C03)-coated annular denuder for collecting gaseous HN03 and
S02 that was analyzed by IC using DIW as the extraction fluid.
A H3P03-coated annular denuder for collecting gaseous NH3, which was designed to be
analyzed by AC using DIW as the extraction fluid.
A 2-stage CASTNET-style filter pack consisting of:
o A nylon filter for collecting S024 and N03, which was analyzed by IC.
7
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
o A H3P03-impregnated filter for collecting any extra NHX either from denuder capture
inefficiency or loss of NH+4 through volatilization off the nylon filter (or both), which
was analyzed by AC using DIW as the extraction fluid.
4. Met One SuperSASS mini-parallel plate denuder (MPPD)
A Met One SuperSASS NH3 denuder, which consisted of:
A H3P03-coated MPPD for collecting gaseous NH3, which was analyzed by AC using a
DlW/methanol mixture as the extraction fluid.
A NH3 canister for the Met One SuperSASS, which was developed by Met One and
tested by researchers at Colorado State University. Amec Foster Wheeler personnel
were trained on the sampling and analysis techniques at the Colorado State
Atmospheric Science laboratory in July 2009.
A H3P03-impregnated filter for collecting particulate NH+4 and any extra NH3 not captured
by the MPPD that was analyzed by AC using DIW as the extraction fluid.
A Teflon PM2.5 cyclone was used to provide particle size selection for the SuperSASS
canister.
5. Single denuder ADS
A single modified ADS that featured:
A H3P03-coated annular denuder for collecting gaseous NH3, which was analyzed by AC
using a DlW/methanol mixture as the extraction fluid.
A 2-stage CASTNET-style filter pack consisting of:
A nylon filter for collecting S024 and N03, which was analyzed by IC.
A H3P03-impregnated filter for collecting any extra NHX either from denuder capture
inefficiency or loss of NH+4 through volatilization off the nylon filter, which was
analyzed by AC using DIW as the extraction fluid.
6. SuperSASS ion module
A SuperSASS ion canister was run on the SuperSASS sampler for nine sampling periods. The
ion module was prepared, shipped, and analyzed according to CSN protocols by RTI.
3.2 Sample Site Selection
1. Test site - Gainesville, FL
Before deployment to any of the selected CASTNET sites, a test site was established at the
Gainesville, FL Amec Foster Wheeler location. The site's data collection designation was
MEC099. All sample types were installed at this site to review operational procedures and to
evaluate the precision of the Amec Foster Wheeler laboratory. Data collection from the test
8
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
study, which consisted of four consecutive one-week sample periods, were processed and
reviewed by Amec Foster Wheeler and EPA. The following sample types were run during each
of the four sample weeks beginning on April 5, 2010:
3-stage standard CASTNET filter pack,
4-stage CASTNET-style filter pack,
Dual denuder ADS with 2-stage filter pack (in triplicate),
Single denuder ADS with 1-stage filter pack (in triplicate), and
SuperSASS-style NH3 denuder (in triplicate).
2. Primary Sites
Table 2 shows the primary sites within CASTNET that were selected for this study along with
the types of samplers originally expected to be deployed at each site. Figure 1 shows the
locations of the sites along with the AMoN site identification.
Table 2. CASTNET Sites Selected for Study
Site ID
CASTNET/AMoN
Sampler Configuration
CTH110, NY/NY67
PAL190, TX/TX43
AMoN passive sampler
3-stage CASTNET filter pack
Single denuder ("short") NHX only ADS
ARE128, PA/PA00
CHE185, OK/OK99
ROM206, CO/CO88*
AMoN passive sampler
3-stage CASTNET filter pack
Dual denuder ("long") S02, HN03, and NHX ADS
SuperSASS MPPD for NHx
SuperSASS CSN ion module
Note: * Co-located with ROM406, the NPS-sponsored CASTNET site of the ROM406/206 pair. Data for
ROM406 are included in several figures for informational purposes, only. The ROM406 data are
not included in any summary calculations.
9
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 1. Map of CASTNET/AMoN Sites Selected
3. SuperSASS sites
At three of the sites selected for the ACCS, a SuperSASS was deployed. At each of these sites,
the following sample types were used:
AMoN passive NH3 sampler
3-stage standard CASTNET filter pack,
Duplicate dual denuder ADS with 2-stage filter pack,
Duplicate SuperSASS-style NH3 denuders, and
SuperSASS ion module.
In order to support the samplers needed to perform these measurements, a second tilt-down
tower was installed at the three SuperSASS sites. These sites are listed in Table 2.
4. Non-SuperSASS sites
At the other two sites selected for the ACCS, a SuperSASS was not deployed. At those sites
(also listed in Table 2), the following samplers were used:
AMoN passive NH3 sampler
3-stage standard CASTNET filter pack, and
Duplicate single denuder ADS with 2-stage filter pack.
10
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
3.3 Monitoring Site Modifications
Several modifications to the sampling sites were required depending upon which sampler
configuration was deployed. Details of all site modifications are provided in Appendix A. The
adjustments included:
Modification of the pot head to accommodate the 4-stage filter pack, two ADS systems,
and flow tubing sufficient to sample at 10 meters; and
Installation of a new base, tower, mast, cabling, and flow tubing to support the
SuperSASS equipment.
3.4 Equipment Modifications
Modifications to several of the samplers were required. Details on the modifications to the
equipment are summarized in Appendix A. The modifications to the equipment included:
Addition of a fourth stage to the standard CASTNET filter pack to add an H3P03-
impregnated cellulose filter. This required modification of the top support ring to the
filter pack.
Modification of the 2-stage ADS filter pack. This also required modification of the top
support ring to the filter pack.
3.5 Calibration Procedures
1. Mass flow controller calibration
Following installation of the ADS, the mass flow controllers were verified according to CASTNET
SOPs and calibrated if necessary. The target flows for the ADS were 3.0 liters per minute (Ipm).
The target flow for the filter pack was 1.5 Ipm, except at ROM206, CO, which maintained a flow
of 3.0 Ipm. Results were recorded on electronic calibration forms.
2. SuperSASS MPPD and canister flow system calibration
Following installation, the SuperSASS operation was verified as follows:
Date/time
The SuperSASS system date and time were verified using the site data logger and
adjusted, if necessary, using the SuperSASS control box.
Mass flow controllers
A BIOS DryCal Volumetric Flow Meter was connected to each SuperSASS cyclone inlet
using a short piece of flexible tubing. With the pump on, the SuperSASS flow rate was
verified to within 5 percent of the reference value. The target value set by the
SuperSASS control software was approximately 6.7 Ipm. If the measurement was
outside this 5 percent range, the reference value was entered in the SuperSASS control
box, and the reading was calibrated to this reference value. Results were recorded on
electronic calibration forms.
Filter temperature sensors
11
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
The sampling canisters were removed, and a reference thermometer was placed in the
inlet opening of each channel. If the filter temperature reading was greater than
ฑ 2ฐCelsius (C) outside the reference value, the reference value was entered in the
SuperSASS control box and the filter temperature reading was calibrated to this value.
Results were recorded on electronic calibration forms.
Ambient temperature sensor
The ambient temperature sensor was calibrated at two points. The probe was removed
from the shield and placed in an ice bath with the reference temperature sensor. If the
ambient temperature probe reading was greater than ฑ 2ฐC outside the reference value,
the reference value was entered in the top reference window on the SuperSASS control
box, and the temperature probe was calibrated to this value. The water bath was heated
to approximately 30ฐC, and the ambient temperature probe was compared to the
reference value. If the ambient temperature probe reading was greater than ฑ 2ฐC
outside the reference value, the reference value was entered in the second reference
window on the SuperSASS control box, and the temperature probe was calibrated to this
value. Results were recorded on electronic calibration forms.
Pump box pressure sensor
The SuperSASS ambient pressure sensor was verified to within 10 millimeters (mm)
mercury (Hg) of a reference barometer. If the readings were outside this range, a syringe
was connected to the reference barometer and the SuperSASS pressure test point. The
syringe plunger was adjusted to simulate a low test pressure of approximately
600 mmHg and a high test pressure of approximately 800 mmHg. The reference
barometer readings for these points were entered in the SuperSASS control box, and
the ambient pressure sensor was calibrated to these values. Results were recorded on
electronic calibration forms.
Leak check
With the SuperSASS control box operating in test mode, small caps were placed on the
sampling head inlet. A successful leak check was indicated by flows of no more than
0.1 Ipm.
3.6 Sample Preparation and Shipping
Sample preparation and shipping procedures are described in Appendix A. The procedures
included the following tasks.
Cellulose filters were impregnated with H3P03 for use in the 4-stage CASTNET filter
packs
H3P03-coated denuders were washed, coated, and dried.
Na2C03-coated denuders were washed, coated, and dried.
SuperSASS MPPD were washed, coated, and dried.
SuperSASS canisters were assembled.
12
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
4.0 Test Site Results
Prior to the start of actual field sampling, three test phases were performed at the Gainesville,
FL Amec Foster Wheeler facility. All sample types that were analyzed by the Amec Foster
Wheeler laboratory were included in the tests. These included:
4-stage CASTNET filter pack,
Single denuder ADS,
Dual denuder ADS, and
SuperSASS MPPD.
In addition to these samplers, a standard CASTNET 3-stage filter pack was also analyzed.
4.1 Sampling Period
The test phases at the Gainesville test site were conducted between April and June 2010. Each
sampler was operated on a 1-week time scale.
1. Sampling set up
Four different sampler types were considered for inclusion in the ACCS and tested during the
test phases in Gainesville, FL. The standard CASTNET 3-stage filter pack was also used during
the test phases in order to have CASTNET-type results available for the comparisons. All
measurements were made at approximately 10 meters, which is the sampling height used for
the standard CASTNET filter pack.
All associated flow systems were calibrated prior to the beginning of the test study. All
laboratory and calibration procedures are detailed in the CASTNET Quality Assurance Project
Plan (QAPP; Amec Foster Wheeler, 2011).
2. Results
Phase 1, which was conducted during April 2010, compared all the proposed sampling
methods. After reviewing the results, a second test period (Phase 2) was designed to evaluate
the 4-stage filter packs. No NH3 was measured on the H3P03-impregnated cellulose filter.
Therefore, Phase 2 is excluded in the figures that follow. Phase 3 tested different ADS
configurations, including the addition of a Teflon filter to the ADS filter pack for both single and
dual configurations; the switch from a H3P03-impregnated filter to a citric acid (C6H807)-
impregnated filter in the CASTNET-style 4-stage filter pack; and a CASTNET-style filter pack
with three stages where the third filter type was a H3P03-impregnated filter instead of a K2C03-
impregnated filter.
Figures 3 through 8 show the results from Phase 1 and Phase 3. Figure 2 provides a legend for
the six figures. The legend uses "SHORT" to correspond to a single denuder ADS and "LONG"
to denote a dual denuder ADS.
13
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 2. Legend for Figures 3 through 8
Legend
Phase
Comment
ฆ 3-stage
1 and 3
Standard CASTNET filter pack
4-stage (nyl)
1
4-stage CASTNET filter pack - nylon filter extracted using DIW,
phosphorus acid filter in 4th spot
ฆ 3-Stage (TNPa) 1
3
3-stage CASTNETfilter pack - potassium carbonate filter
replaced with phosphorus acid filter in 3rd spot
ฆ 3-Stage (TNPa) 2
3
3-stage CASTNETfilter pack - potassium carbonate filter
replaced with phosphorus acid filter in 3rd spot
ฆ 4-stage 1
3
4-stage CASTNET filter pack - nylon filter extracted using
standard extraction fluid, citric acid filter in 4th spot
ฆ 4-stage 2
3
4-stage CASTNET filter pack - nylon filter extracted using
standard extraction fluid, citric acid filter in 4th spot
ADS-NYL L0NG1
1
2-denuder ADS, nylon filter used for particle collection
ฆ ADS-NYL L0NG2
1
2-denuder ADS, nylon filter used for particle collection
ADS-NYL L0NG3
1
2-denuder ADS, nylon filter used for particle collection
ADS-NYL SH0RT1
1
1-denuder ADS, NH3 and NH4 only, nylon filter used for particle
collection
ADS-NYL SH0RT2
1
1-denuder ADS, NH3 and NH4 only, nylon filter used for particle
collection
ADS-NYL SH0RT3
1
1-denuder ADS, NH3 and NH4 only, nylon filter used for particle
collection
ฆSASS1
1
SASS MPPDwith phosphorus acid backup filter for NHx
ฆSASS2
1
SASS MPPDwith phosphorus acid backup filter for NHx
ฆSASS3
1
SASS MPPDwith phosphorus acid backup filter for NHx
ฆ ADS-TEF L0NG1
3
2-denuder ADS, Teflon filter used for particle collection
ฆ ADS-TEF L0NG2
3
2-denuder ADS, Teflon filter used for particle collection
ADS-NYL L0NG1
CNETSTA
3
2-denuder ADS, nylon filter used for particle collection, nylon
filter extracted using standard extraction fluid
ADS-NYL L0NG2
CNETSTA
3
2-denuder ADS, nylon filter used for particle collection, nylon
filter extracted using standard extraction fluid
14
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figures 3 and 4 compare results for particulate S024 and NH+4 concentrations in micrograms per
cubic meter (|jg/m3), respectively. Results from Phase 1 were not satisfactory. For example, the
S024 concentrations measured on the three denuders were about 80 percent of the
concentrations collected on the nylon filters. Similarly, the NH+4 concentrations from the
denuders were about half the nylon filter values based on the Phase 1 data on the left side of
Figures 3 and 4. For Phase 3, a Teflon filter was added to the ADS filter pack to improve S024
and NH4 collection efficiency. The results improved significantly as shown on the right side of
Figures 3 and 4.
Figure 3. S04 Results from Test Study
4.5
04/05/10 04/12/10 04/19/10 04/26/10 06/10/10 06/17/10 06/24/10
15
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 4. NH4 Results from Test Study
2.2
2.0
1.8
^1-6
CO
o) 1 -4
ง1.2
Phase 1 I Phase 3
CD
i
C
<1)
ฃ
o
O
1.0
0.8
0.6
0.4
0.2
0.0
ill
1
04/05/10 04/12/10 04/19/10 04/26/10 06/10/10 06/17/10 06/24/10
Figure 5 shows NH3 results from Phases 1 and 3. The poor performance of the CASTNET-style
4-stage filter pack is evident. The results for other samplers compared reasonably well.
Figure 5. NH3 Results from Test Study
2.2
2.0
1.8
^1-6
CO
o) 1 -4
J1-2
TO
Phase 1 | Phase 3
c
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 6. S02 Results from Test Study
1.4
04/05/10 04/12/10 04/19/10 04/26/10 06/10/10 06/17/10 06/24/10
Figure 7 shows measured NHX concentrations, which for this study were defined as NH+4 + NH3.
Again, the poor performance of the 4-stage CASTNET-style filter pack is evident in Phase 1
data. The concentrations measured by the standard CASTNET 3-stage filter pack were also
lower, but these results were expected because the 3-stage filter pack excluded sampling media
to collect NH3. The NHX concentration results for the two filter packs during Phase 1 were
similar, further illustrating the inefficiency of the 4-stage filter pack.
Figure 8 shows particulate N03 concentrations from the test studies. The influence of larger N03
particles [diameter greater than 2.5 micrometers (jjm)] is evident during the first testing period
because N03 concentrations measured by the open-faced CASTNET filter packs were much
higher.
17
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 7. NHX Results from Test Study
2.2
2.0
1.8
1.6
F
1.4
CD
zL
c
1.2
o
05
1.0
c
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
3. Lessons learned/modifications for actual sampling
Two problems were encountered during Phase 1. First, the initial ADS design featured a nylon
filter that proved to be inadequate for the collection of particles. Comparing S024 concentrations
from the standard filter pack and the ADS (Figure 3) showed S024 concentrations were
consistently lower as measured by the ADS even though concentrations from the two sampling
systems should have been similar. EPA found similar problems during the verification study for
the Monitor for AeRosols and Gases in Ambient Air (MARGA) automated gaseous and
particulate sampler that EPA was testing (Rumsey et ai, 2014). Consequently, for Phase 3, a
Teflon filter was added to the ADS for the collection of particles, and results improved.
The second problem involved the 4-stage filter pack. No NH3 was collected by the H3P03-
impregnated filter despite expectations to the contrary. In high humidity environments, use of a
4-stage filter pack is not viable because NH3 reacts with the S02 collected by a hydrated K2C03-
impregnated filter prior to encountering the H3P03-impregnated filter. In the third test study, a
3-stage filter pack with the K2C03 filter replaced with a H3P03 filter was used to successfully
collect NH3. Also, a 4-stage filter pack using C6H807 instead of H3P03 was deployed, but the
same results were obtained (no NH3 was collected) as in Phase 1.
Because of these results, 4-stage filter packs were not included during the ACCS field-sampling.
Additional testing of other sampling configurations was considered, but budgetary constraints
and the availability of a viable passive collection device (from AMoN) eliminated the impetus to
perform more tests.
Following the successful outcome of Phase 3 testing, ACCS field sampling started at the end of
August 2010 and continued through September 2011. Results from the field sampling and data
comparisons are presented in Sections 7.0 through 9.0. The results from the field study were
used to evaluate the performance of the AMoN samplers, standard CASTNET filter pack,
SuperSASS MPPD, and SuperSASS ion module. Concentrations measured by the ADS were
considered reference values.
19
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
5.0 Network Sampling
5.1 Sampling Schedule and Laboratory Analyses
Sampling began in August 2010 and was completed in September 2011. In total, there were 10
2-week sampling periods. The deployment schedule (Figure 9) was designed to permit sampling
during all four seasons. The samples that were analyzed by the Amec Foster Wheeler
laboratory were collected using:
Single denuder ADS,
Dual denuder ADS, and
SuperSASS MPPD
CASTNET standard filter pack
AMoN passive samples were analyzed by the NADP CAL and SuperSASS ion modules were
analyzed by RTI. The ADS samples were considered the reference method samples for the
study analyses. In addition, a regular CASTNET filter pack was collected and analyzed using
normal CASTNET sampling protocols at each site.
Figure 9. Sampling Schedule
Note: Puchalski etal., 2015
5.2 Data Collection and Handling
The output of each mass flow controller was collected and stored on a Campbell Scientific
CR3000 data logger. Voltage measurements were made by the data logger, and calibration
factors were applied to convert the output to standard Ipm. Flow data were aggregated to
5-minute and hourly averages. Data were collected remotely using Campbell Scientific's
LoggerNet polling software.
For SuperSASS-related data, the actual volumetric flow data were measured for each channel
along with filter temperature, ambient temperature, and ambient pressure. All measurements
were aggregated to 5-minute averages. The SuperSASS control box was connected to the
serial port of a RavenX cellular modem. Data were polled remotely using SuperSASSComm AQ
software developed by Met One Instruments. Serial communications data were transmitted via
the internet to provide a connection to the SuperSASS control box.
20
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
5.3 Data reporting
In general, data from the test study and the ACCS were processed according to existing
CASTNET protocols as documented in the CASTNET QAPP. Data produced by the Amec
Foster Wheeler laboratory were transferred from the laboratory information management
system (LIMS) to the Amec Foster Wheeler Data Management Center (DMC) by the submittal
of electronic data deliverables (EDDs), which are DBF files. Data were imported at the DMC into
SQL Server and/or Oracle.
Hourly flow data were imported into SQL Server and/or Oracle. Atmospheric concentrations
were calculated by dividing the laboratory data (in total micrograms) by the flow volume (in cubic
meters). Hourly flow data were reviewed on a daily basis during sampling periods. Hourly flow
data were validated using the results of the semiannual audits conducted by field calibrators.
Laboratory data were validated by reviewing atmospheric concentrations and laboratory quality
control (QC) data. Each batch was peer reviewed prior to being finalized as summarized in
Section 6.0.
Validated and reviewed data were transferred from the DMC to EPA via established data
submittal protocols, which involve a direct database-to-database data transfer via copy
command.
21
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
6.0 Quality Assurance/Quality Control
Quality assurance (QA)/QC procedures utilized for the study followed those described in the
CASTNET QAPP Revision 7.0 with slight modifications for SuperSASS sampling as described
previously. All measurements were expected to meet a precision criterion of a relative percent
difference (RPD) of 20 percent. Mass flow control flow rates were audited and calibrated using
the standard CASTNET protocols. Results were documented by field technicians using standard
CASTNET electronic calibration forms. The data recorded on these forms were reviewed for
compliance after each site visit.
Audit procedures for SuperSASS measurements consisted of flow procedure compliance audits,
as described previously, along with review of data recorded on Study Report Forms and site
narrative logbook pages. Additionally, at least one method audit was performed to verify
compliance with packing, unpacking, and extraction procedures during both the Amec Foster
Wheeler test period and the study period. QC sample results were reviewed for each batch and
included those for reference samples, drift checks, analytical and co-located sampling precision,
and method and laboratory blanks. One field blank for each sample type was analyzed for each
site during the one-year study period. A complete analytical data package audit was performed
for data associated with each measurement system at mid-study and study end.
Figure 10 provides an example of a completed chain-of-custody (COC) form that was used
during ACCS. The example form is for a single denuder ADS.
Figure 10. Example Chain-of-Custody Form
CASTNET
AMMONIA STUDY SAMPLE REPORT FORM
o. poM 20C&
Jk &
CHAIN OF CUSTODY
WEEK 25 2011
LAB1D#: 1125004-21 to 23 Site:ROM206-ADSl
Scheduled On Date: 06/21/2011
SHIPMENT OPENED BY:
AA
PACKED BY: 1996
DATE:
ฃ-31-11
LAB USE ONLY:
LAB TECH RECEIPT-
IVC
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
7.0 Results of NH3 Concentration Measurements
Figures 11 through 15 show results for the five sites at which NH3 concentrations were
measured by AMoN passive samplers. In these figures, the NADP/NTN site designator is
provided adjacent to the CASTNET site designator. Figures 11 and 12 also present NH3
concentrations measured by the ADS. Figures 13 through 15 illustrate concentrations measured
by passive samplers, ADS, and SuperSASS MPPD. See Table 2 for the sampler configurations
at the five sites. At ROM206/CO88, AMoN samples did not start until May 2011, and only three
AMoN samples were included in the comparison. Statistical analyses of the ADS and AMoN
NH3 measurements were provided by Puchalski et al. (2015). In general, the results shown in
Figures 11 through 15 demonstrate that the AMoN passive sampler performed very well at all
five sites compared with the ADS reference concentrations. Puchalski etal. (2015) determined
an overall MRPD of -9 percent over all sites, which included a range of NH3 concentrations plus
diverse monitoring elevations and climate regimes.
SuperSASS MPPD samples consistently collected less NH3 than the reference ADS (Figures 13
through 15), and results worsened over the course of the field sampling. This outcome was
especially evident at the high-elevation site, ROM206 (Figure 15), where the concentrations
sampled by the MPPD systems were less than half the ADS values.
Figure 11. Comparison between AMoN and ADS NH3 Concentrations at CTH110 / NY67
1.2
1.0 -
NY67 AMoN
i ADS H3P03 Denuderl
ADS H3P03 Denuder2
23
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 12. Comparison between AMoN and ADS NH3 Concentrations at PAL190 / TX43
7
6
TX43 AMoN
i ADS H3P03 Denuderl
ADS H3P03 Denuder2
CD
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 14. Comparison between AMoN, ADS, and SuperSASS MPPD NH3 Concentrations
at CHE185 / OK99
4 0 1 ฆ OK99 AMoN
3.5- ฆ ADS H3P03 Denuderl
ADS H3P03 Denuder2
^3 0 " ฆ SuperSASS MPPD1
5 - 1 SuperSASS MPPD2
ElllUUili
OOOt T
^c\jco^LooSo^c\ico
CO
CO O T- CM CO LO CD a>
Figure 15. Comparison between AMoN, ADS, and SuperSASS MPPD NH3 Concentrations
at ROM206 / C088
25
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
8.0 Evaluation of Filter Pack S02, Total N03, S024, and NH4 Concentration
Measurements
Figures 16 through 23 show concentrations of S02, total N03, S024, and NH+4 from the three
sites with "long" ADS (ARE128, CHE185, and ROM206; see Table 2 for sampling
configurations). Each figure provides a separate plot with results for each of the three study
sites. The time series plots show weekly concentrations measured by CASTNET filter packs,
ADS, and SuperSASS ion modules as appropriate for the specific analytes.
Regression statistics for all sites and four filter pack pollutants are summarized in Table 3. The
table illustrates overall good results and also shows the variability of the comparisons among
the sites.
8.1 S02 Concentrations
In Figure 16, S02 concentrations are presented in three time series, one for each site. The
concentrations were measured by the standard CASTNET filter pack and the first denuder
(coated with Na2C03) on the two ADS. The figures show good comparison among the three
measurement systems. Figure 17 provides a scatter diagram based on S02 concentrations
measured by the filter packs at all three sites combined versus S02 concentrations averaged for
the two denuders at the three sites. There was a positive relationship between the average
denuder and filter pack measurements with a regression line:
y = 0.973x + 0.033.
The correlation coefficient for the two sets of measurements is 0.977.
8.2 Total N03 Concentrations
Time series of total N03 concentrations (HN03 + N03) are presented in Figure 18 for the three
sites. The total N03 concentrations were measured using the filter pack and the first denuder
plus the Teflon and nylon filters at the back end of the two ADS (ADS1 and ADS2). Total N03
measurements were higher for the CASTNET filter pack compared with the ADS/filter pack
systems because CASTNET filter packs had an open face while the other two measurement
devices had a 2.5 |jm size cut. The open face allows coarse-sized N03 particles to be collected
by the filter pack.
A scatter diagram based on total N03 concentrations measured by the filter packs at all three
sites combined versus total N03 concentrations averaged for the two denuders at the three sites
is provided in Figure 19. The regression line for the average ADS/filter pack concentrations and
filter pack measurements is:
y = 1.108x + 0.201.
26
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
The correlation coefficient for the two sets of measurements is 0.891.
Figure 16. Comparison between CASTNET Filter Pack and ADS S02 Concentrations at
Three CASTNET Sites
a) ARE128, PA
12
05
O
O
10 -
o o o o o o
CO
N CM O) CO O ^
CJ} \ \
00 o o
C\J CO
ARE128 CASTNET FP
ARE128 ADS NA2C03 DENUDER1
ARE128 ADS NA2CQ3 DENUDER2
1111 ฆ 111 ll I ฆ
LOC\JCT>Loor^^-coc\JO>coo
^CsjCsj^^^CNjCNjSocO^^!
c\] c\] CO LD LD CD CD CD CD
b) CHE185, OK
3.5
^ N (N O) CO O
5 55 - - (N 5 ^
CO O O t- T
LO C\J CT> LO O h-
^ CNj gNJ
CN CM CO
^ ^ CNj gNJ
in in cd cd
CO CM O) CO O
CO CO
_ CM
ct> (j>
27
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 16. Comparison between CASTNET Filter Pack and ADS S02 Concentrations at
Three CASTNET Sites (Continued)
c) ROM406/206, CO
1.0
^ N (N O) CO O
5 55 - - (N 5 ^
CO O O T
LO CM O) LO O N
^ CM ฃM
(N CM CO
^ ^ CNj ฃM
in in (D (D
CO CM CD CO O
00 00
_ CM
ct> (j>
Figure 17. Regression Analysis of CASTNET Filter Pack and ADS S02 Concentrations at
Three CASTNET Sites
w
y = 0.9734X + 0.033
~2 - n Q7Rซ;
.r''
jL
./
J*
0123456789
ADS Concentration (jig/m3)
28
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 18. Comparison between CASTNET Filter Pack and ADS Total N03 Concentrations
at Three CASTNET Sites
a) ARE128, PA
7
ARE 128 CASTNET FP
ARE 128 ADS1
ARE 128 ADS2
CO
CO
CNJ CT> CO O ^f
CT>
C\J CO
LO C\J CT> LO O h-
O O
^ CNJ CNJ
cm CM CO
^ ^ CM CM
LO LO CD CD
CO CM O) CO
CO CO
^ CNJ
^ ^ ฃ! ฃ2 ^
CO O O t- T
LO CM O) LO
r- csJ csJ ^
CM CM CO
or^^-cocMO>coo
rrCNCNJowrM
LO LO CD CD
CD
29
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 18. Comparison between CASTNET Filter Pack and ADS Total N03 Concentrations
at Three CASTNET Sites (Continued)
c) ROM406/206, CO
1.4
CO
CO
N (N O) (O O ^
^ CM CO ^
CD
O O
LO CM CD LO
^ ^
CM CM CO
Or^T-COCMCT>COO
^^ฃ!ฃ!SoSo^ฃ!
10 in 0 CD CD CD
Figure 19. Regression Analysis of CASTNET Filter Pack and ADS Total N03
Concentrations at Three CASTNET Sites
ฆ L
II
1075x + 0.200;
5 .
hr = u.oyi4
ฅ *
/-:
/
012345678
ADS Concentration (jig/m3)
30
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
8.3 Particulate SO* Concentrations
Time series of measured particulate S024 concentrations are presented in Figure 20 in three
graphs for the three sites. The S024 was measured on the CASTNET filter pack, Teflon filters on
the back end of the two ADS (FP1 and FP2), and the SuperSASS ion modules used in CSN.
The results from the four samplers are reasonably comparable with higher filter pack S04
concentrations measured at CHE185 during the spring, again likely because of no device on the
filter pack for the removal of particles greater than 2.5 |jm in mean diameter. CHE185 is
influenced by fugitive dust more than ARE128 and ROM206, as indicated by measured weekly
concentrations of cations/earth metals (Amec Foster Wheeler, 2011). For example, calcium
(Ca2+) concentrations measured at CHE185 were consistently higher than the other two sites
during the period of ACCS. S024 concentrations sampled at ARE128 were highest of the three
sites during the summer. Throughout the study, concentrations at ROM206 were less than 20
percent of the concentrations measured at the other two sites.
Figure 21 shows a scatter diagram based on S04 concentrations measured by the filter packs
at all three sites combined versus S04 concentrations averaged for the two ADS (FP1 and FP2)
at the three sites. A regression line calculated between the average ADS and filter pack
measurements is:
y = 1.138X+ 0.035.
The correlation coefficient for the two sets of measurements is 0.963.
8.4 Particulate NH4 Concentrations
Comparisons of NH4 concentrations measured on the filter pack, Teflon and nylon filters on the
two ADS (FP1 and FP2), and the SuperSASS ion modules are presented in Figure 22. The filter
pack and the ion modules measured higher NH4 concentrations than the ADS, especially at
ARE128 and CHE185.
A scatter diagram based on NH4 concentrations measured by the filter packs at all five sites
combined versus NH4 concentrations averaged for the denuders at the five sites (Table 3) is
provided in Figure 23. The regression line for the ADS/filter pack and filter pack measurements
is:
y = 1.270x- 0.002.
The correlation coefficient for the two sets of measurements is 0.907.
Figure 20. Comparison between CASTNET Filter Pack, CSN SuperSASS Ion Module, and
ADS SO4 Concentrations at Three CASTNET Sites
a) ARE128, PA
31
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
6 -
ARE 128 CASTNET FP
ARE128 ADS FP1
ARE128 ADS FP2
ARE 128 SuperSASS
o o o o o o
CO
N CM O) CO O ^
CO o o
CM CO
LOC\JCT>Loor^^-coc\JO>coo
CM CM CO LO LO CD CD
b) CHE185, OK
5
o o o o o o
C\J CT> CO O
co
CD
CM CO
o o
LO (M O) LO O
^ ^ CNj ^ ^
CM CM CO
Qsl Qsi
LO LO CD CD
CO CM CD CO O
CO CO
_ CM
cd (j>
32
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 20. Comparison between CASTNET Filter Pack, CSN SuperSASS Ion Module, and
ADS SO4 Concentrations at Three CASTNET Sites (Continued)
c) ROM406/206, CO
15 1 1 ROM406 CASTNET FP
ฆ ROM206 CASTNET FP
ฆ ROM206 ADS FP1
^ 1 ROM206 ADS FP2
o>1 0 " ฆ ROM206 SuperSASS
0
0
0
0
O
O
T
T
T
t
t
T
T
T
t
t
t
T
t
t
C\J
CD
CO
O
LO
Cvl
CD
UO
0
CO
CNJ
CD
CO
O
CO
CT>
^
Cvl
CO
^
^
Cvl
Cvl
^
^
c\]
Cvl
CO
CO
^
Cvl
CO
0
O
ฆซ
ฆซ
Cvl
csi
CO
LO
LO
CD
CD
CD
CD
Figure 21. Regression Analysis of CASTNET Filter Pack and ADS S04 Concentrations at
Three CASTNET Sites
y = 1.1383X + 0.03E
R2 = 0.9632
t
.V
1
.1
#*
r
1
.ฆi
12 3 4
ADS Concentration (|ig/m3)
33
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 22. Comparison between CASTNET Filter Pack, CSN SuperSASS Ion Module, and
ADS NH4 Concentrations at Three CASTNET Sites
a) ARE128, PA
3
C\J CT> CO O
CO O O T- T-
LOCMCT>LOOI^T-CO
CM CM CO
LO LO CD CD
T
T
t
t
CM
CD
CO
O
CO
CO
CM
CD
(7)
b) CHE185, OK
3
N CM O) (O O
C? 3> ^ ^ ฃ! ฃ2 ^
CO O O T- T-
LOCMCDLOOI^T-COCMO>COO
CM CM CO lO lO CD CD CD CD
34
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 22. Comparison between CASTNET Filter Pack, CSN SuperSASS Ion Module, and
ADS NH4 Concentrations at Three CASTNET Sites (Continued)
c) ROM406/206, CO
0.6
CO
CO
CT>
CM CT> CO O
v- v- C\l CO
00^^
LO CM CD LO
^ ^ ^ ^
CM CM CO
Or^T-COCMCT>COO
^^^CMSoSo^^!
10 in (D CD CD CD
Figure 23. Regression Analysis of CASTNET Filter Pack and ADS NH4 Concentrations at
Five CASTNET Sites
y
= 1,2696x - 0.0023
R2 = 0.9065
v*'
/
/
ฆฆ
>
. / ""
&
0.0 0.5 1.0 1.5 2.0 2.5
ADS Concentration (jag/m3)
3.0
35
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Table 3. Summary of Regression Analyses of Concentrations Measured with Filter
Packs and ADS by Site
n
S02 (mean)
Linear Regression (ads_avg vs. FP)
adsl
ads2
ads_avg
FP
Slope
y-intercept
R2
All Sites
56
1.271
1.236
1.253
1.253
0.9734
0.0330
0.9765
ARE128
17
2.694
2.657
2.675
2.565
1.0318
-0.1958
0.9626
CHE185
19
1.152
1.085
1.119
1.177
0.9138
0.1542
0.9483
ROM206
20
0.175
0.170
0.173
0.211
1.0018
0.0381
0.8667
n
Total N03 (mean)
Linear Regression (ads_avg vs. FP)
adsl
ads2
ads_avg
FP
Slope
y-intercept
R2
All Sites
54
1.428
1.460
1.444
1.800
1.1075
0.2005
0.8914
ARE128
17
2.267
2.455
2.361
2.717
1.2047
0.1271
0.7125
CHE185
19
1.696
1.622
1.659
2.196
1.0453
0.4622
0.9094
ROM206
18
0.352
0.351
0.352
0.515
1.2234
0.0850
0.8879
n
SO4 (mean)
Linear Regression (ads_avg vs. FP)
adsl
ads2
ads_avg
FP
Slope
y-intercept
R2
All Sites
56
1.508
1.480
1.494
1.734
1.1383
0.0350
0.9632
ARE128
17
2.267
2.309
2.288
2.622
1.0349
0.2542
0.8907
CHE185
19
2.023
1.909
1.966
2.313
1.1793
-0.0059
0.9189
ROM206
20
0.374
0.367
0.370
0.429
1.1283
0.0132
0.9745
n
NH4 (mean)
Linear Regression (ads_avg vs. FP)
adsl
ads2
ads_avg
FP
Slope
y-intercept
R2
All Sites
92
0.568
0.555
0.561
0.702
1.2696
-0.0023
0.9065
ARE128
17
0.911
0.907
0.909
1.239
1.3076
0.0507
0.5811
CHE185
19
0.753
0.700
0.727
0.917
1.2401
0.0159
0.9595
CTH110
18
0.696
0.688
0.692
0.798
1.1453
0.0056
0.9594
PAL190
19
0.381
0.385
0.383
0.432
0.8627
0.1352
0.4549
ROM206
19
0.141
0.139
0.140
0.187
1.1273
0.0290
0.8799
Note: ads_avg = average of concentration measurements from dual ADS
FP = filter pack
36
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
9.0 Evaluation of S024, N03, and NH4 Concentrations Measured on Filter Pack,
ADS, and CSN SuperSASS Ion Module
Figures 24 through 29 show regression analyses for particulate S024, N03, and NH4 measured
by the CASTNET filter pack, duplicate ADS, and CSN SuperSASS ion modules.
Regression statistics for the three sites and measured (filter pack, ADS, and SuperSASS)
particulate pollutant data are summarized in Table 4. The table illustrates generally good results
with some variability in the comparisons among the sites.
9.1 Particulate SO4 Concentrations
Particulate S024 was measured on the CASTNET filter pack, Teflon filters on the back end of the
two ADS, and the CSN SuperSASS ion modules. Figure 20 shows S024 data from the four
samplers were reasonably comparable with higher filter pack S024 concentrations, again likely
because of no device on the filter pack for the removal of larger particles greater than 2.5 |jm in
mean diameter.
Figure 24 provides a scatter diagram based on S024 concentrations measured by the filter packs
at all three sites combined versus S04 concentrations measured by the SuperSASS ion
modules. There was a positive relationship between the average SuperSASS and filter pack
measurements with a regression line:
y = 0.897x + 0.073.
The correlation coefficient for the two sets of measurements is 0.978.
37
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 24. Regression Analysis of CASTNET Filter Pack and CSN SuperSASS S04
Concentrations at Three CASTNET Sites
o
II
8967x + 0.0728
F
V = 0.9779
. y
%
i.-
X
12 3 4
Filter Pack Concentration (ng/m3)
Figure 25. Regression Analysis of ADS and CSN SuperSASS S04 Concentrations at
Three CASTNET Sites
y = 1,0428x + 0.0728
R2 = 0.9829 ....
4
>
*
.ซ
jr
r*
9'
/
12 3 4
ADS Concentration (ng/m3)
38
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 25 provides a scatter diagram based on S04 concentrations measured by the ADS at all
three sites combined versus S024 concentrations measured by the SuperSASS ion modules.
The data produced a regression line:
y = 1.043X+ 0.073.
The correlation coefficient for the two sets of measurements is 0.983.
9.2 Particulate N03 Concentrations
Particulate N03 concentration regression analyses are shown in Figures 26 and 27. A scatter
diagram based on N03 concentrations measured by the filter packs at all three sites combined
versus the SuperSASS data is provided in Figure 26. The regression line is:
y = 0.857x+ 0.105.
The correlation coefficient for the two sets of measurements is 0.883.
Figure 26. Regression Analysis of CASTNET Filter Pack and CSN SuperSASS N03
Concentrations at Three CASTNET Sites
7
CO
E
O)
35
c
o
CO
~ 4
3
2
1
0
CD
O
c
o
O
CO
CO
<
CO
CD
Q.
3
CO
= U.ซS/4X+ U.1U01
R2 = 0.8832 ,ฆฆฆฆ'
H
9
1 2 3 4 5 6
Filter Pack Concentration (ng/m3)
Figure 27 provides a scatter diagram based on N03 concentrations measured by the ADS at all
three sites combined versus N03 concentrations measured by the SuperSASS ion modules.
The data produced a regression line:
39
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
y = 1.037X+ 0.032.
The correlation coefficient for the two sets of measurements is 0.989.
Figure 27. Regression Analysis of ADS and CSN SuperSASS N03 Concentrations at
Three CASTNET Sites
7
:o
E
CD
O
S3
CO
CO
<
w 2
CD
Q.
3
CO
1
0
0 1 2 3 4 5 6 7
ADS Concentration (ng/m3)
9.3 Particulate NH4 Concentrations
Figures 28 and 29 depict regression analyses for NH+4 concentrations. A scatter diagram based
on NH 4 concentrations measured by the filter packs at all three sites combined versus the
SuperSASS data is provided in Figure 28. The regression line for the average SuperSASS and
filter pack measurements is:
y = 0.992x + 0.029.
The correlation coefficient for the two sets of measurements is 0.823.
= 1.0366X+ 0.0318 ,
R2 = 0.9889
4
/
40
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 28. Regression Analysis of CASTNET Filter Pack and CSN SuperSASS NH4
Concentrations at Three CASTNET Sites
3.0
O)
zL
i20
-t'
03
-t'
o 1.5
c
o
O
CO
% 1.0
CO
CD
Q.
w 0.5
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Filter Pack Concentration (ng/m3)
Figure 29 provides a scatter diagram based on NH4 concentrations measured by the ADS at all
three sites combined versus NH4 concentrations measured by the SuperSASS ion modules.
The data produced a regression line:
y = 1.329X+ 0.004.
The correlation coefficient for the two sets of measurements is 0.809.
y = 0.9917x + 0.0288 *
R2 = 0.8233
4
1"
*
V
41
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 29. Regression Analysis of ADS and CSN SuperSASS NH4 Concentrations at
Three CASTNET Sites
0,5 1 1.5 2
ADS Concentration (|.ig/m3)
2.5
42
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Table 4. Summary of Regression Analyses of Concentrations Measured with CSN
SuperSASS Ion Modules, CASTNET Filter Packs, and ADS
so24
n
Concentration (|jg/m3)
Linear Regression (FP vs. SuperSASS)
FP
SuperSASS
Slope
y-intercept
R2
All Sites
48
1.793
1.681
0.8967
0.0728
0.9779
ARE128
16
2.649
2.460
0.8273
0.2682
0.8761
CHE185
15
2.437
2.277
0.8865
0.1170
0.9907
ROM206
17
0.420
0.421
0.9447
0.0247
0.9720
n
Concentration (|jg/m3)
Linear Regression (ads_avg vs. SuperSASS)
ADS
SuperSASS
Slope
y-intercept
R2
All Sites
47
1.527
1.665
1.0428
0.0728
0.9829
ARE128
15
2.323
2.462
0.9628
0.2257
0.9902
CHE185
15
2.049
2.277
1.0512
0.1227
0.9323
ROM206
17
0.363
0.421
1.0752
0.0307
0.9838
N03
n
Concentration (|jg/m3)
Linear Regression (FP vs. SuperSASS)
FP
SuperSASS
Slope
y-intercept
R2
All Sites
48
0.991
0.954
0.8574
0.1051
0.8832
ARE128
16
1.493
1.486
0.8308
0.2461
0.8862
CHE185
15
1.415
1.259
0.8460
0.0623
0.8340
ROM206
17
0.143
0.185
1.0137
0.0395
0.5172
n
Concentration (|jg/m3)
Linear Regression (ads_avg vs. SuperSASS)
ADS
SuperSASS
Slope
y-intercept
R2
All Sites
45
0.886
0.950
1.0366
0.0318
0.9889
ARE128
15
1.360
1.392
1.0497
-0.0351
0.9771
CHE185
15
1.144
1.259
1.0393
0.0703
0.9930
ROM206
15
0.154
0.199
0.9945
0.0457
0.9463
NH+4
n
Concentration (|jg/m3)
Linear Regression (FP vs. SuperSASS)
FP
SuperSASS
Slope
y-intercept
R2
All Sites
48
0.803
0.825
0.9917
0.0288
0.8233
ARE128
16
1.289
1.215
0.6395
0.3902
0.3748
CHE185
15
1.000
1.152
1.2246
-0.0730
0.9066
ROM206
17
0.171
0.170
0.6930
0.0517
0.6230
n
Concentration (|jg/m3)
Linear Regression (ads_avg vs. SuperSASS)
ADS
SuperSASS
Slope
y-intercept
R2
All Sites
46
0.613
0.819
1.3291
0.0042
0.8090
ARE128
15
0.943
1.171
0.8935
0.3285
0.2416
CHE185
15
0.792
1.152
1.4971
-0.0335
0.8528
ROM206
15
0.136
0.173
0.784
0.0703
0.5651
Note: ads_avg = average of concentration measurements from dual ADS
FP = filter pack
43
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
10.0 Dry Deposition of Atmospheric Nitrogen
CASTNET was designed to estimate dry deposition of sulfur and nitrogen pollutants across the
United States by combining measured pollutant concentrations and modeled values of
deposition velocity (Vd). The Multi-Layer Model (MLM) is used to calculate hourly Vd for each
monitoring site based on meteorological measurements and information on vegetative cover
within 1 km of each site. Estimates of dry deposition of nitrogen, however, are incomplete
because measurements of all species of nitrogen have not been available for the dry deposition
calculations. This study helps remedy that problem by providing data on NH3 and volatized NH+4
concentrations.
Estimates of dry deposition of nitrogen, including fluxes of NH3 and NHX, were made for the five
monitoring sites for the 20-week sampling period. Vd values for NH3 were taken from 2002
simulations using the Community Multiscale Air Quality (CMAQ) modeling system (EPA,
2018b). The average NH3 Vd values (cm/sec) for the five monitoring sites were calculated as
follows:
ARE128
1.455
CHE185
1.778
CTH110
1.877
PAL190
0.967
ROM206
1.287
The MLM was used to calculate Vd values for the other pollutants for the 20-week period in
2011. If Vd values for the weeks in 2011 were not available, the average of the weekly values for
those weeks in 2009 and/or 2010 was used. The calculated dry fluxes are presented in
Figures 30 through 34. For the three sites (ARE128, CHE185, and ROM206) with dual
denuders, two illustrations are given in the figures. The first was calculated based on ADS
measurements of HN03 and N03 concentrations, and the second was calculated based on
CASTNET filter pack measurements of those two parameters.
Figure 30 shows estimates of dry deposition of nitrogen for ARE128. The nitrogen deposition
fluxes illustrated by the two circles are comparable. The two principal contributors to dry
deposition were NH3 and HN03. The deposition estimate of nitrogen based on the ADS data is
0.136 kg ha-1 yr1, and the estimate based on ADS/CASTNET filter pack data is 0.145 kg ha-1 yr1.
The results for CHE185 are given in Figure 31. Again, NH3 and HN03 were the principal
contributors to dry deposition of nitrogen, although the percentage contribution from NH3 was
higher for this site. The deposition estimate of nitrogen based on the ADS data is 0.150 kg ha-1
yr1 and is 0.154 kg ha-1 yr1 based on ADS/CASTNET filter pack data.
44
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 32 provides the deposition data for ROM206. The ADS and CASTNET estimates are
comparable with values of 0.029 kg ha-1 yr1 and 0.032 kg ha-1 yr1, respectively, but are about
three to five times lower than fluxes for ARE128 and CHE185. Also, the CASTNET estimate of
the contribution of HN03 measured on the nylon filter at ROM206 is about 60 percent higher
than the flux calculated from the ADS HN03 concentration.
Nitrogen fluxes for CTH110 and PAL190 are given in Figures 33 and 34. Total nitrogen fluxes
for the two sites were 0.061 and 0.186 kg ha-1 yr1, respectively. Table 5 summarizes measured
concentrations of ambient nitrogen pollutants and dry nitrogen deposition simulations for the five
sites. The results demonstrate that NH3 and HN03 were the principal constituents of dry
atmospheric nitrogen fluxes.
45
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 30. Dry Deposition of Nitrogen for ARE128, PA
a) Calculated using ADS Concentrations
0.14 kg ha1 yr1
5% 2% 2%
19%
ฆ HN03, NA2C03 DENUDER
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
N03, TEF+NYL
72%
b) Calculated using CASTNET Concentrations
0.14 kg ha1 yr1
5% 2%2%
^ 1 23%
ฆ HN03, CASTNET
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
i N03, CASTNET
69%
Note: For the three sites (ARE128, CHE185, and ROM206) with dual denuders, two illustrations are given in Figures 30-32.
The first illustration was based on calculations using ADS measurements of HN03 and N03 concentrations and the
second was based on calculations using CASTNET filter pack measurements of those two parameters.
46
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 31. Dry Deposition of Nitrogen for CHE185, OK
a) Calculated using ADS Concentrations
0.15 kg ha1 yr1
3% 2% 1 % 11%
ฆ HN03, NA2C03 DENUDER
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
N03, TEF+NYL
84%
b) Calculated using CASTNET Concentrations
0.15 kg ha1 yr1
3% 2% 1 % 130/0
ฆ HN03, CASTNET
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
i N03, CASTNET
81%
47
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 32. Dry Deposition of Nitrogen for ROM206, CO
a) Calculated using ADS Concentrations
0.03 kg ha1 yr1
4% 3ฐ^0/o^^14%
ฆ HN03, NA2C03 DENUDER
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
N03, TEF+NYL
77%
b) Calculated using CASTNET Concentrations
0.03 kg ha1 yr1
4% 3%2%
21%
ฆ HN03, CASTNET
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
ฆ N03, CASTNET
71%
48
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure 33. Dry Deposition of Nitrogen for CTH110, NY
0.06 kg ha1 yr1
5% 2% 1%
ฆ HN03, CASTNET
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
i N03, CASTNET
Figure 34. Dry Deposition of Nitrogen for PAL190, TX
0.19 kg ha1 yr1
2o/01%1% 10%
ฆ HN03, CASTNET
ฆ NH3, H3P03 DENUDER
NH4, TEFLON
ฆ NHX AS NH4, H3P03 FILTER
ฆ N03, CASTNET
86%
49
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Table 5. Summary of Measured Nitrogen Concentrations and Modeled Dry
Deposition Fluxes Calculated using ADS Concentrations
ARE128
HN03, Na2C03
nh3, h3po3
NH+4,
NHx as NH+4,
N03,
TOTAL
DENUDER
DENUDER
TEFLON
h3po3 FILTER
TEF+NYL
cone
0.116
0.119
0.009
0.003
0.012
flux
0.026
0.098
0.007
0.002
0.003
0.136
pet of total
19%
72%
5%
2%
2%
CHE185
HNO3, Na2C03
nh3, h3po3
NH+4,
NHx as NH+4,
N03,
TOTAL
DENUDER
DENUDER
TEFLON
h3po3 FILTER
TEF+NYL
cone
0.074
0.152
0.006
0.003
0.007
flux
0.016
0.125
0.004
0.002
0.002
0.150
pet of total
11%
84%
3%
2%
1%
ROM206
HNO3, Na2C03
nh3, h3po3
NH+4,
NHx as NH+4,
N03,
TOTAL
DENUDER
DENUDER
TEFLON
h3po3 FILTER
TEF+NYL
cone
0.018
0.027
0.002
0.001
0.002
flux
0.004
0.022
0.001
0.001
0.000
0.029
pet of total
14%
77%
4%
3%
2%
CTH110
HNO3, Na2C03
nh3, h3po3
NH+4,
NHx as NH+4,
N03,
TOTAL
DENUDER
DENUDER
TEFLON
h3po3 FILTER
TEF+NYL
cone
0.058
0.052
0.004
0.001
0.004
flux
0.013
0.043
0.003
0.001
0.001
0.061
pet of total
21%
71%
5%
2%
1%
PAL190
HNO3, Na2C03
nh3, h3po3
NH+4,
NHx as NH+4,
N03,
TOTAL
DENUDER
DENUDER
TEFLON
h3po3 FILTER
TEF+NYL
cone
0.082
0.194
0.005
0.003
0.010
flux
0.018
0.159
0.004
0.003
0.002
0.186
pet of total
10%
86%
2%
1%
1%
All Sites
HNO3
nh3, h3po3
DENUDER
NH+4,
TEFLON
NHx as NH+4,
h3po3 FILTER
N03,
TEF+NYL
TOTAL
min cone
0.018
0.027
0.002
0.001
0.002
min flux
0.004
0.022
0.001
0.001
0.000
0.029
max cone
0.116
0.194
0.009
0.003
0.012
max flux
0.026
0.159
0.007
0.003
0.003
0.186
50
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
11.0 References
Amec Foster Wheeler Environment and Infrastructure, Inc. (Amec Foster Wheeler). 2011. Clean
Air Status and Trends Network (CASTNET) Quality Assurance Project Plan Revision
7.0. Prepared for the U.S. Environmental Protection Agency (EPA), Office of Air and
Radiation, Clean Air Markets Division, Washington, D.C., Contract No. EP-W-09-028,
Gainesville, FL. https://java.epa.gov/castnet/documents.do.
Lavery, T. F., C. M. Rogers, R. E. Baumgardner Jr., K. P. Mishoe. 2009. Intercomparison of
CASTNET N03 and HN03 measurements with data from other monitoring programs. J.
Air& Waste Manage. Assoc., 59, 214-226.
National Atmospheric Deposition Program. 2018. Ammonia Monitoring Network.
http://nadp.slh.wisc.edu (accessed January 2018).
Puchalski, M. A., C. M. Rogers, R. E. Baumgardner Jr., K. P. Mishoe, G. Price, M. J. Smith, N.
Watkins, C. M. Lehmann. 2015. A Statistical Comparison of Active and Passive
Ammonia Measurements Collected at Clean Air Status and Trends Network (CASTNET)
Sites. Environ. Sci.: Processes Impacts, 2015, 17, 358.
Rumsey, C., K. A. Cowens, J. T. Walker, T. J. Kelly, K. Mishoe, C. M. Rogers, R. Proost, G. M.
Beachley, G. Lear, T. Frelink, and R. P. Otjes. 2014. An Assessment of the Performance
of the Monitor for AeRosols and Gases in Ambient Air (MARGA): a Semi-Continuous
Method for Soluble Compounds. Atmos. Chem. Phys., 14, 5639-5658, 2014
U.S. Environmental Protection Agency (EPA). 2018a. Clean Air Status and Trends Network
Home Page, https://www.epa.gov/castnet (accessed January 2018).
U.S. Environmental Protection Agency (EPA). 2018b. Estimating Atmospheric Deposition with
the Community Multiscale Air Quality Model (CMAQ),
https://www.epa.gov/cmaq/estimating-atmospheric-deposition-cmaq (accessed
January 2018).
URG Corporation. 2018. Annular Denuder System.
http://www.urgcorp.com/index.php/systems/manual-sampling-systems/annular-denuder-
system (accessed January 2018).
51
-------
Appendix A
Monitoring Site and Sampler Modifications
Appendix A
-------
MONITORING SITE MODIFICATIONS
Pot Head
The existing air sampling enclosure ("pot head") was modified to accommodate an additional
4-stage filter pack and two ADS. Samples were taken at 10 meters. Each sampler was required
to have individual tubing connected to a mass flow controller inside the shelter. The two ADS
shared a common flow pump. The 4-stage filter pack shared the flow pump that serves the
existing CASTNET 3-stage filter pack.
1. Modification to the existing pot head
At each site, a black color-keyed quick disconnect was installed on the pot head plate for the
4-stage filter pack. Yellow and orange color-keyed quick disconnects were installed on the pot
head plate for the ADS running in duplicate to differentiate between the two systems. The full
ADS filter packs used orange filter pack clamps, and the NH3-only ADS filter packs used black
filter pack clamps to differentiate between the two ADS configurations. The ADS tubes were
braced for stabilization.
2. Flow tubing
At each site, 3/8 inch Teflon tubing was run from each sampler to a wall-mounted, glass
knockout bottle inside the shelter. The knockout bottle was used as a water trap. Each sampling
train had an individual knockout bottle. Teflon tubing (1/4 inch) was run from the knockout bottle
to an Apex mass flow controller. Teflon tubing (1/4 inch) was run from the flow controllers to the
suction side of a Thomas Pump & Machinery (Thomas) pump. The Thomas pump, which is
used for the existing 3 stage filter pack flow, was fitted with a "T" connector. This pump provided
flow for the existing 3 stage filter pack and the new 4-stage filter pack. An additional Thomas
pump was installed at each site to provide flow for the two ADS.
SuperSASS Tower
At each of the three SuperSASS sites, a new tilt-down 10-meter flow tower was installed to
mount the SuperSASS sampling head. The cables and tubing from the SuperSASS pump box
were extended to allow for 10 meter sampling. Sample flow control was provided by mass flow
controllers in the pump box installed at the base of the SuperSASS tower. Data were collected
by a control unit inside the shelter.
1. SuperSASS tower installation
At each SuperSASS site, the base of the new flow tower was installed approximately one week
before the tower was erected, to allow the concrete sufficient time to cure. The SuperSASS
tower was secured to the shelter at one site, and guy wires were installed for stabilization on
freestanding towers at two sites.
2. Mast fabrication for SuperSASS Head
The mast of the new SuperSASS flow tower was modified to the correct size for the sampling
head. Alignment holes were drilled to orient the top and bottom portions of the sampling head. A
holding hook was installed to hold the shield in place during canister changes.
A-1
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
3. SuperSASS cabling and flow tubing
Factory extension cables were installed for the SuperSASS canister temperature sensors,
ambient temperature probe, and control box. Each cable extension was approximately 40 feet
long. Teflon tubing (1/4 inch) was installed to extend the flow tubing for canisters. Each canister
had its own extended tubing. Each tubing extension was 30 feet,
SAMPLER MODIFICATIONS
1. 4-stage CASTNET style filter pack
A fourth stage was added to the existing CASTNET 3-stage filter pack configuration to contain
the H3PO3- impregnated cellulose filter. In the current CASTNET configuration, a support ring
without grid is placed on top (upstream in Figure A-1) of the Teflon filter. In order for the inlet
and the clamp to seal with the fourth stage added, the top support ring required a simple
modification of removing the lip with a razor knife. Figure A-2 shows the required modification.
Figure A-1. CASTNET 3-Stage Filter Pack
Shipping Cap
(removed during sampling)
Teflon Nylon Two
Filter Filter
Direction of Air Flow
Two
Cellulose
Filters
nCH
h
Quick Disconnect
I
Teflon Spacers
Figure A-2. Modified Support Ring (Left) versus Non-modified (Right)
2. Dual Denuder ADS filter pack
2
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure A-3 provides a photograph of the dual denuder ADS with filter pack. A modification was
made to the ADS filter pack inlet by removing part of the lip with a razor knife (Figure A-4). The
top support ring (without grid) was modified as discussed for the 4-stage filter pack. The two
modifications were necessary in order for the inlet and the clamp to tighten completely.
Figure A-3. Dual Denuder ADS with Filter Pack
WKHi - .a
J
i 'ฆ
ฎ j
ฆ
-ป
Figure A-4. Filter Pack Inlet (Left) versus Modified Inlet (Right)
-------
Appendix B
Sample/Sampler Preparation Procedures
Appendix B
-------
PROCEDURE FOR IMPREGNATING CELLULOSE FILTERS WITH PHOSPHOROUS ACID
1. Gather two boxes of Whatman brand #41 47mm ashless cellulose filters, gloves, clean
forceps, a 1-liter wide-mouth bottle, and 3 percent and ~5 percent H3P03 solutions.
2. Place two boxes of filters (50 per box) into a 1-liter wide-mouth polypropylene bottle and
cover with 3 percent H3P03-impregnating solution and seal.
3. Sonicate for 30 minutes. Drain impregnating solution into an appropriate waste receptacle.
4. Cover the lab bench for cleanliness with aluminum foil, and wipe with a towel soaked with a
-5 percent H3P03 solution.
5. Tear another piece of foil and fold to fit inside the modified nitrogen desiccator. Wipe the foil
with a freshly soaked towel of a ~5 percent H3P03 solution. Clean gloved fingertips and
forceps with a -5 percent H3P03 solution.
6. Arrange a flat layer of the impregnated filters on the cleaned foil inside the modified nitrogen
desiccator (Figure B-1). Cover promptly, and place the modified nitrogen desiccator in hood
with exhaust turned on.
7. Attach the nitrogen gas source to the modified desiccator with flow control.
8. Turn on nitrogen gas flow to ~5 Ipm. Close hood sash completely, and check for dryness in
six hours; continue the nitrogen purge overnight (if necessary) until the filters are visibly dry.
9. After placing in a sealed and marked plastic resealable bag, refrigerate the filters.
Contamination during storage can be reduced by including a ~5 percent H3P03-soaked
paper towel in a plastic resealable bag and placing the sealed bag of impregnated filters
inside.
10. The impregnated filters should be acceptance tested after each impregnating procedure. A
pair of laboratory blanks should be pulled for extraction with each extraction of exposed
filters.
Figure B-1. Modified Nitrogen Desiccator
B-1
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
PROCEDURES FOR URG PHOSPHOROUS ACID-COATED DENUDER
Procedure for Washing URG H3P03 Denuder
This procedure is performed in the washroom where all required components are available.
1. Rinse each cap with DIW, filling, swishing, and dumping three times.
2. Using a properly labeled squeeze bottle, rinse the denuder with methanol (CH3OH) over the
labeled denuder waste container.
3. Rinse the denuder, running DIW through the channels for -10 seconds on each side. Rinse
the outside threads of both ends.
4. Put the denuder into a large washtub and fill with DIW.
5. Soak for -30 minutes.
6. Repeat steps 4 and 5 two times, for a total of three soaking periods.
7. Shake water out of the denuder and caps and allow to dry in the drying oven in an aluminum
foil lined heat resistant tray while covered with aluminum foil for about two hours at 60ฐC.
Procedure for Coating and Drying URG H3P03 Denuders
This procedure is performed in the washroom where all required components are available.
1. Pipette 10 milliliters (ml_) of 1 percent H3P03 coating solution into the denuder with bottom
cap attached. The bottom of the denuder is the end in which the inner quartz tube is
recessed about 1/8 inch from the outer edge of the aluminum denuder tube.
2. Secure top cap of the denuder and invert 20 times. The top of the denuder is the end into
which the inner quartz tube is recessed about one inch from the outer end of the aluminum
denuder tube.
3. Remove the top cap and pour excess coating solution into the labeled waste container.
4. Screw the denuder securely into the URG drying manifold (Figure B-2) so the direction of
drying flow is downward; this will allow gravity to remove excess solution from the denuder.
5. Dry for approximately 20 minutes with a 5 Ipm nitrogen purge. The "honeycomb" quartz
inside the URG denuder will cloud when dry.
6. Remove the denuders from the drying manifold, and install both caps; refrigerate until
needed. A blank red label should be attached near the top end to signify H3P03 coating. The
red laboratory identification (ID) label will be attached below it.
7. Two laboratory blanks should be pulled for extraction from each coating session and/or each
batch of coating solution made.
B-2
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure B-2. Denuder Secured to Drying Denuder Manifold
PROCEDURES FOR URG SODIUM CARBONATE-COATED DENUDER
Procedure for Washing URG Sodium Carbonate Denuder
This procedure is performed in the washroom where all required components are available.
1. Rinse each cap with DIW, filling, swishing, and dumping three times.
2. Using a squeeze bottle, rinse the denuder with CH3OH over the appropriate waste
container.
3. Rinse the denuder, running DIW through the channels for -10 seconds on each end. Rinse
the outside threads as well.
4. Put the denuder into a large washtub and fill with DIW.
5. Allow to soak for -30 minutes.
6. Repeat steps 3 through 5 two times, for a total of three soaking periods.
7. Shake water out of the denuder and caps and dry in the drying oven in an aluminum foil
lined heat resistant tray covered with aluminum foil for about two hours at 60 C.
Procedure for Coating and Drying URG Sodium Carbonate Denuder
This procedure is performed in the washroom where all required components are available.
1. Pipette 10 ml_ of 1 percent Na2C03 coating solution into the denuder with bottom cap
attached. The bottom of the denuder is the end in which the inner quartz tube is recessed
about 1/8 inch from the outer of the aluminum denuder tube.
B-3
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
2. Secure top cap of the denuder and invert 20 times. The top of the denuder is the end in
which the inner quartz tube is recessed about one inch from the outer end of the aluminum
denuder tube.
3. Remove the top cap and pour excess coating solution into the labeled waste container.
4. Remove the bottom cap, and place the denuder and caps in desiccators to dry overnight.
The desiccant should be fresh.
5. Remove the denuders from the drying manifold and install both caps. Refrigerate until
needed. A blank green label should be attached near the top end to signify Na2C03 coating.
The green Laboratory ID label will be attached below it.
6. Laboratory blank(s) should be pulled for extraction from each coating session and/or each
batch of coating solution made.
PROCEDURES FOR SUPERSASS MPPD PREPARATION
Procedure for Washing SuperSASS MPPD:
This procedure is performed in the washroom where all required components are available.
1. DiSuperSASSemble the denuder, retaining O-rings in their grooves.
2. Rinse each cap with DIW, filling, swishing, and dumping three times.
3. Using a squeeze bottle rinse the denuder with CH3OH over the labeled waste container.
4. Rinse the denuder, running DIW through the channels for -10 seconds on each side. Rinse
the outside threads as well.
5. Put the denuder into the larger top cap, and fill with DIW; fill bottom cap with DIW.
6. Set aside and allow soaking for -30 minutes.
7. Repeat steps 4 through 6 two times, for a total of three soaking periods.
8. Shake water off of the denuder and caps and allow to dry in the drying oven in an aluminum
foil lined heat resistant tray covered with aluminum foil for about two hours at 60ฐC.
B-4
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure B-3. SuperSASS MPPD
Procedure for Coating and Drying SuperSASS MPPD:
This procedure is performed in the washroom where all required components are available.
1. Screw denuder firmly into the bottom cap, ensuring that the O-ring is correctly positioned.
2. Pipette 6 mL of 1 percent H3P03-coating solution into the denuder attached to the bottom
cap; this should completely fill all sampling channels (Figure B-4).
3. Secure the top cap loosely over the denuder; let the denuder soak for 10 minutes.
4. Remove the top cap and unscrew the denuder from the bottom cap. Tap gently the denuder
against the bottom cap to drain as much coating solution as possible.
5. Screw the denuder securely into the drying manifold and hang the manifold so the direction
of drying flow is downward (Figure B-5). This will allow gravity to remove excess solution
from the denuder.
6. Dry for approximately 30 minutes with 5 liters per minute nitrogen purge. This should be
performed in a hood without the hood exhaust. The hood sash must be completely closed.
The glass plates of the MPPD will appear cloudy when dry.
7. Remove the denuders from the drying manifold and install directly into both caps. Seal in a
resealable plastic bag, and refrigerate until needed.
8. Two laboratory blanks should be pulled for each extraction, each coating session, and/or
each batch of coating solution made.
B-5
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure B-4. Location for Pipetting Coating Solution and Extraction Fluid in MPPD
The coating solution and the
extraction fiuids pipette here: j.
Figure B-5. MPPD Attached to Drying Manifold
Procedure for SuperSASS Canister Assembly
This procedure is performed in the washroom where all required components are available.
1. Cover the laboratory bench for cleanliness with aluminum foil. Wipe with a fresh towel
soaked with a ~5 percent H3P03 solution. Clean gloved fingertips and forceps with the same
towel.
2. Gather all clean components for assembly of the SuperSASS Canister as seen in
Figure B-6.
B-6
-------
Summary Report of the Special Reactive Nitrogen (Nr) inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
3. Use clean forceps to place a H3P03-impregnated filter in the filter holder, and place filter
holder in the base of canister as seen in Figure B-7.
4. Place filter spacer atop filter holder with O-ring facing up. Follow with the H3P03 coated
MPPD denuder with O-ring facing up and lastly, with the cone-shaped spacer as seen in
Figure B-8.
5. The stainless steel canister is installed over the assembly after firmly pressing all parts
together.
6. The three stainless steel screws are installed and tightened firmly.
7. When ready for sampling, the dust plug is removed and the cyclone is attached as seen in
Figure B-9.
B-7
-------
Summary Report of the Special Reactive Nitrogen (Nr) Inter-
Comparison Study: Ammonia CASTNET CSN Study (ACCS)
Figure B-6. Components Ready for Assembly Figure B-7. Base of Canister
with Filter Holder
Figure B-8. Cone-Shaped Spacer on Top of MPPD Figure B-9. Assembled Canister
B-8
------- |