Krish Vijayaraghavan,

1
Modeling of Atmospheric Nitrogen Deposition to
the Escambia Bay and Watershed

• Krish Vijayaraghavan,
• Rochelle Balmori, Shu-Yun Chen,
• Prakash Karamchandani and Christian Seigneur
• AER, San Ramon, CA
• Justin T. Walters and John J. Jansen
• Southern Company, Birmingham, AL
• Eladio M. Knipping
• EPRI, Palo Alto, CA
• CMAS Conference, Oct 1-4, 2007
• Chapel Hill, NC

2
Overview

• Objective
• Estimate impact of NOx and SO2 controls at the
  Crist power plant on nitrogen deposition in
  Escambia Bay and its watershed in
  Florida/Southern Alabama
• Tools
• Three versions of CMAQ v. 4.5.1

3
Escambia Bay and Watershed
Escambia Bay Watershed
4
Air Quality Models

• CMAQ
• CMAQ v. 4.5.1 with SOA modifications by VISTAS
• CBM-IV for gas-phase chemistry
• AERO4 aerosol module
• Heterogeneous nitrate formation in the PM phase
  only
• Includes sea salt emissions but does not account
  for coarse nitrate formation due to sea-salt/HNO3
  interactions
• CMAQ-MADRID
• Based on CMAQ 4.5.1 and also utilizes CBM-IV
• Aerosols Model of Aerosol Dynamics, Reaction,
  Ionization, and Dissolution
• Heterogeneous nitrate formation in aqueous and PM
  phase
• Comprehensive sea-salt/HNO3 chemistry (fine and
  coarse size ranges)
• CMAQ-MADRID-APT
• Builds upon CMAQ-MADRID
• Advance Plume Treatment (APT) for 40 power
  plants, including Plant Crist

5
Plume Chemistry Dispersion Relevance to
Nitrogen Chemistry
3
2
Early Plume Dispersion
1
Long-range Plume Dispersion
Mid-range Plume Dispersion
NO/NO2/O3 chemistry
Reduced VOC/NOx/O3 chemistry Slow PM formation
from OH and NO3/N2O5 chemistry
Negligible PM formation
Full VOC/NOx/O3 chemistry PM and O3 formation
(NO3-, SO4)
6
ALGA Modeling Domain

• 2002 reference year
• 12 km horizontal grid
  resolution
• 19 layers up to 15 km altitude
• 40 power plants, including Plant Crist, with APT
• Inputs from VISTAS/GEPD

7
Emission Reductions at Plant Crist

• Anticipated emission reductions due to
  installation of FGD and SCR/SNCR

8
Nitrogen Species
Coarse mode not present for CMAQ
9
Spatial Distribution of Total Nitrogen Deposition

• Change in annual dry wet deposition flux due to
  controls on Plant Crist

Maximum reduction in deposition flux
0.68 kg/ha
0.85 kg/ha
0.42 kg/ha
APT Less oxidation of NOx to HNO3 gt Less dry
deposition near the plant
10
Dry Deposition of Nitrogen over the Escambia Bay
Watershed (tpy)

• Highest contributor Gas I-NOz (mostly HNO3)

11
Ammonia Dis-benefit due to SO2 Controls

• SO2 controls gt Less PM ammonium sulfate gt More
  gas NH3
• Dry deposition velocity of gas NH3 gt PM NH4
• Increase in NH3 deposition gtgt Decrease in PM NH4
  deposition
• Outcome
• Planned controls result in an increase in the NHx
  component of nitrogen deposition.
• Caveat
• Downward revision of the NH3 dry deposition
  velocities will decrease the dis-benefit.
• Difference between models
• APT has less dis-benefit because of less sulfate
  formation in the plume.

12
Sea-salt Dis-benefit due to NOx/SO2 Controls

• MADRID and APT account for coarse NaNO3 formation
  from sea-salt/HNO3
• SO2 controls gt Less fine and coarse sodium
  sulfate
• Less coarse sodium sulfate gt More coarse sodium
  nitrate
• Outcome
• Planned controls result in a small increase in
  coarse nitrate deposition.
• Caveats
• There is enough HNO3 so coarse nitrate formation
  is not affected by NOx controls.
• Dis-benefit will be lower if more fine NaCl in
  sea-salt emissions
• CMAQ will also exhibit sea-salt dis-benefit if it
  accounts for coarse NaNO3.
• Difference between models
• APT has less dis-benefit than gridded models
  because of less sulfate/nitrate formation.

13
Wet Deposition of Nitrogen over the Escambia Bay
Watershed (tpy)

• Models can not distinguish between gas and PM wet
  deposition
• Largest contributors I-NOz followed by NHx

14
Conclusions

• Three versions of CMAQ were used to estimate the
  decrease in atmospheric nitrogen deposition in
  Escambia Bay/watershed due to NOx and SO2
  emissions controls at the nearby Crist power
  plant.
• Differences in results between the three models
  are due to differences in model formulation and
  configuration
• CMAQ-MADRID has more comprehensive heterogeneous
  nitrate and coarse sea-salt nitrate chemistry
  than CMAQ.
• CMAQ-MADRID-APT includes plume-in-grid treatment
  of large point sources (here, 40 large power
  plants including Crist).

15
Conclusions

• Gaseous inorganic NOz (mostly HNO3) has the
  largest contribution (60) to nitrogen dry
  deposition in all three models.
• Inorganic NOz (gaseous particulate) is the
  largest contributor (52) to wet deposition with
  a slightly lower contribution from NHx.
• NOx emission controls result in reductions in
  nitrogen deposition but SO2 controls result in an
  increase in nitrogen deposition due to an
  ammonia dis-benefit. NH3 dry deposition
  velocities in CMAQ need to be investigated
  further.

16
Conclusions

• Over Escambia Bay and its watershed, CMAQ, MADRID
  and APT predict
• total N deposition reductions of 91, 100, and 106
  tons/yr, respectively.
• maximum reductions in gridded deposition fluxes
  of 0.68, 0.85, and 0.42 kg/ha/yr, respectively.
• APT simulates less dry deposition of HNO3 and PM
  sulfate near Plant Crist than CMAQ and MADRID due
  to its correct treatment of plume dispersion and
  chemistry. It is important to use a
  plume-in-grid treatment of emissions from large
  elevated point sources so that nitrogen
  deposition can be correctly simulated.
• Air quality modeling results were subsequently
  used in a watershed modeling study to estimate
  net nitrogen loading to Escambia Bay.