![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081107114350im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Final Report: Investigation of Indicators for Ozone-NOx-VOC Sensitivity
EPA Grant Number: R826765Title: Investigation of Indicators for Ozone-NOx-VOC Sensitivity
Investigators: Sillman, Sanford
Institution: University of Michigan
EPA Project Officer: Shapiro, Paul
Project Period: October 1, 1998 through September 30, 2001
Project Amount: $306,407
RFA: Air Pollution Chemistry and Physics (1998)
Research Category: Engineering and Environmental Chemistry
Description:
Objective:This research project seeks to investigate the use of observation-based indicators for identifying the sensitivity of ozone to nitrogen oxides (NOx) and volatile organic compounds (VOCs). The sensitivity of ozone to NOx and VOCs is a critical issue for pollution-control policy, and has repeatedly been the subject of controversy. O3-NOx-VOC sensitivity is usually determined based on predictions from three-dimensional (3-D) Eulerian models for dynamics and photochemistry. It is difficult to test whether these predictions are correct for real-world conditions. The uncertainty concerning the accuracy of these ozone-precursor predictions has spurred interest in "observation-based methods," which seek to derive information about ozone formation from ambient measurements, rather than just from models.
In previous work, it was found that model NOx-VOC predictions correlate with the values of certain "indicator ratios," involving total reactive nitrogen (NOy) and peroxides. Measured values of these ratios might be used directly to determine whether ozone is primarily sensitive to NOx or VOC. Measured values of these species also can be used to evaluate whether model predictions for the impact of reduced NOx and VOC emissions on ozone are accurate. The same secondary species also have been useful for evaluating other features of the ozone formation process, such as the rate of ozone production per NOx ("ozone production efficiency"). This research project seeks to investigate the behavior of the secondary species that have been identified as possible indicators for ozone-NOx-VOC sensitivity, with special attention to the following issues:
· Do the species ratios proposed as indicators for NOx-VOC sensitivity show consistent behavior for all environments, or does their behavior show significant variation from location to location?
· Can it be shown that models correctly represent ambient concentrations and formation rates of secondary species in a way that is independent of the question of local NOx-VOC sensitivity? Use of secondary species ratios as NOx-VOC indicators is valid only if it can be shown that models predict ambient levels correctly.
· Does the behavior of secondary species in models critically depend on the photochemical mechanism used in the model calculations?
A second goal of this research project was to use secondary species to investigate the process of ozone formation in large power plants. Large power plants are responsible for 30-50 percent of NOx emissions in the eastern half of the United States, and they are believed to contribute significantly to the ozone formation process. However, there is considerable uncertainty about the amount of ozone resulting from these power plant emissions. Other studies, suggested that the rate of ozone production per NOx emitted was much lower for NOx emitted from large power plants than for NOx emitted from other sources. This research project used photochemical models to investigate these results, with special attention to the following issues:
· Are results from photochemical models consistent with recent observations of the rate of ozone formation in power plant plumes?
· Does the low rate of ozone production per NOx in large power plants, based on measurements, also appear in model calculations?
· Does the low rate of ozone production per NOx found in recent measurements appear to be a general feature of large power plants? Alternatively, do models predict a higher rate of ozone production per NOx for power plants under different circumstances (e.g., greater downwind distances, multiday transport, etc.)?
Summary/Accomplishments (Outputs/Outcomes):Two broad topics that we researched were: (1) the possible use of measured secondary species as "indicators" for the sensitivity of ozone to NOx versus VOC; and (2) the rate of ozone production per NOx in large power plants. Results are summarized separately. A summary of relevance to the U.S. Environmental Protection Agency (EPA) programmatic goals also is given.
Secondary Species as NOx-VOC Indicators. Ambient ratios such as O3/NOy, O3/NOz, and O3/HNO3 show different values depending on whether O3 is sensitive to NOx or VOC. Because of this property, the ratios are potentially useful for evaluating whether models predict O3-NOx-VOC sensitivity correctly. However, the behavior is more complicated than suggested by a simple "rule of thumb," linking ratio values to NOx-sensitive or VOC-sensitive conditions. Use of the ratios as a simple "rule of thumb" can lead to misleading results.
A composite of results from 3-D Eulerian models for different cities shows a distinct pattern of correlation for O3 versus NOy, O3 versus NOz, and O3 versus HNO3, which holds for all NOx-sensitive environments. A different correlation pattern is found for locations in which O3 is primarily sensitive to VOC.
These correlation patterns provide a more accurate basis for interpreting measurements than the simpler "indicator ratios" originally proposed. They also are advantageous because they allow for some quality assurance. Measured correlation patterns that are at variance with model predictions would be identified as errors (either in measurements, models, or basic theoretical understanding) and deemed ineligible for interpretation.
Measured correlation patterns for O3 versus the sum of NOx reaction products (NOz), at a number of locations, including Nashville, Houston, Los Angeles, several rural sites in the eastern United States, and Paris, France, are broadly comparable to the range of correlation patterns found in models. NOz includes all components of NOy other than NOx. These patterns show a variation between apparent NOx-sensitive locations and apparent VOC-sensitive locations.
In contrast to O3 versus HNO3, model correlations between O3 and organic nitrates show no difference for NOx-sensitive versus VOC-sensitive environments. This feature can be used in the future to test the accuracy of model representations of O3-NOx-VOC sensitivity. If models represent O3-NOx-VOC chemistry correctly, then the measured correlation for O3 versus organic nitrates should show no difference between NOx-sensitive and VOC-sensitive locations.
The values of ratios such as O3/NOy, etc., associated with NOx-sensitive and VOC-sensitive conditions tends to be different in relatively unpolluted air (NOy is equal to the sum of NOx, HNO3, organic nitrate, and particulate nitrate [NO3-], etc).
The ratio, H2O2/HNO3, and other ratios involving H2O2 are strongly correlated with O3-NOx-VOC in models, and provide a more reliable basis for evaluating model NOx-VOC predictions than other measurements.
Ozone Production in Power Plant Plumes. The rate of ozone production is much smaller on a per-NOx basis for plumes from large power plants than for NOx emissions from other sources. However, the rate of ozone production per NOx is much higher in models for power plant plumes that have been transported overnight. Ozone production per NOx in these transported power plant plumes are comparable to ozone production per NOx from other emission sources.
The highest ozone production per NOx is associated with power plant emissions during the late afternoon or evening, followed by dispersion overnight, and production of ozone the next day. Peak ozone concentrations in these plumes is much lower than in plumes from power plants emitted during the early morning, but the total amount of ozone produced is greater.
Models for ozone formation in power plants can accurately predict rates of ozone production and loss of NOx, but only if the initial amount of NOx in the plume is correct. An error in the initial amount of NOx in the plume would escalate into a major error, because overestimated NOx causes an underestimate in the NOx removal rate.
Model predictions for ozone production from power plants is critically dependent on the assumed rate of horizontal dispersion of the power plant plume as it travels overnight.
Relevance to EPA Programmatic Goals. Sensitivity of ozone to its two main precursors, NOx and VOC, is one of the major uncertainties for pollution-control policy. This issue also is important for particulates and other pollutants, because the formation of these species is linked to the ozone-formation process.
Results of the research project have potential practical applications for developing and evaluating State Implementation Plans (SIPs) for ozone. The approach investigated here is one of the "observation-based methods" (OBMs) recommended by the North American Research Strategy for Tropospheric Ozone (NARSTO) Assessment of Tropospheric Ozone Production, for possible use in developing SIPs.
Results of this research project provide a basis for interpreting measurements from the EPA's Photochemical Assessment Monitoring Stations (PAMs) network. The PAMs network currently includes limited measurement of NOy, and it was recently proposed that the PAMs network extend its coverage of NOy. Results of this research project are important for interpreting these measurements.
The impact of large power plants on ozone formation in the eastern United States, especially during events with multiday transport, has been a major policy uncertainty. This research project provides information that addresses this issue.
Journal Articles on this Report: 6 Displayed | Download in RIS Format
| Other project views: | All 16 publications | 6 publications in selected types | All 6 journal articles |
| Type | Citation | ||
|---|---|---|---|
|
|
Liu HH, Molz FJ. Comment on . Water Resources Research. 1998;34(11):3207-3208. |
R826765 (Final) R826171 (1998) R826171 (Final) |
not available |
|
|
Sillman S. On the erroneous use of receptor modeling to diagnose O3-NOx-hydrocarbon sensitivity. Atmospheric Environment 1999;33(14):2289-2291. |
R826765 (Final) |
not available |
|
|
Sillman S. Ozone production efficiency and loss of NOx in power plant plumes: photochemical model and interpretation of measurements in Tennessee. Journal of Geophysical Research 2000;105(D7):9189-9202. |
R826765 (1999) R826765 (Final) |
not available |
|
|
Sillman S, Odman MT, Russell AG. Comment on "On the indicator-based approach to assess ozone sensitivities and emissions features." by C-H. Lu and J. S. Chang. Journal of Geophysical Research 2001;106(D18):20941-20944. |
R826765 (Final) |
not available |
|
|
Sillman S. Comment on "The impact of an 8 hr air quality standard on ROG and NOx controls in Southern California," by Chock et al. Atmospheric Environment 2001;35(19):3367-3369. |
R826765 (Final) |
not available |
|
|
Sillman S, He D. Some theoretical results concerning O3-NOx-VOC chemistry and NOx-VOC indicators. Journal of Geophysical Research - Atmospheres 2002;107(D22):art. no. 4659. |
R826765 (Final) |
not available |
tropospheric, oxidants, nitrogen oxides, organics, volatile organic compounds, VOCs, nitrates, acid rain, hydrogen peroxide, point sources. , Toxics, Air, RFA, Engineering, Chemistry, & Physics, VOCs, tropospheric ozone, risk assessment, ambient emissions, nitrous oxide, ozone, bias, power plants, Nitric oxide, EPA Model 3 Systems, measurement methods , air quality standards, air modeling, indicator ratios, modeling predictions
Relevant Websites:
http://www.cgenv.com/Narsto/critrev.papers.html 
http://www1.psi.ch/~loop/gerzensee/prot_f.html
Progress and Final Reports:
1999 Progress Report
Original Abstract