Experts Conferred on Improving Emissions Inventory

Contributed by Gerardo Mejía, Francisco Obregon and Robert Slott

The Integrated Program on Urban, Regional and Global Air Pollution organized the “Workshop on Mexico Emissions Inventory” at the Hotel Royal Pedregal, México D.F., México on February 25 and 26, 2003. The purpose of this workshop was to bring experts from both the United States and from Mexico to discuss the current state of knowledge in emissions inventories and to help improve the development of the Mexico emission inventory. Better defining the emission inventory in the Mexico City metropolitan area (MCMA) is of particular importance due to the large population exposed to air pollutants that often exceed the national standards. The workshop was sponsored by the North American Commission for Environmental Cooperation, the Mexican Ministry of Environment and Natural Resources (SEMARNAT) and the Mexican Metropolitan Environmental Commission (CAM).

Uncontrolled air pollution in our modern age can lead to serious adverse health effects, threaten the viability of ecosystems and contribute significantly to global climate change. Studies have shown that increased morbidity and mortality accompany air pollution that exceeds air quality standards. Pollutants may be directly emitted from air pollution sources and other pollutants may be formed in the atmosphere as a result of reactions often catalyzed by sunlight. Establishing the appropriate control strategy portfolio requires understanding how secondary air pollutants are formed, especially ozone, and particulate matter.

Reducing air pollution requires controlling emission sources. An accurate emission inventory is needed to help design and monitor the emission control program by defining the important emission sources and watching air pollution trends. Emission inventories can be approached by measuring source emissions (bottom-up) or monitoring pollutants in the atmosphere using ambient concentrations, source receptor analysis, or other methods (top-down). Air pollutant concentrations vary in time (time of day, day of the week, season of the year) and space. Because many of the sources are difficult to measure and monitoring cannot take place everywhere, every emission inventory should be considered as a best estimate with an associated uncertainty. Methods of estimating the accuracy and uncertainty of emission inventories are important.

The workshop was divided into six sessions:

1. Emission Inventory Applications, Uncertainties, and Data Requirements
2. Mexico Emissions Inventory
3. Innovative Techniques for Measuring Emissions
4. Bottom-up Methods
5. Top-Down Methods
6. Wrap up Session

In the introduction to the workshop, Mario Molina, MIT, gave an overview of the research and education activities of the Integrated Program’s Mexico City Case Study (hereafter referred as Mexico City Project). A review of the 1998 Mexico City Emission Inventory is included in a book published in 2002 (Air Quality in the Mexico Megacity: An Integrated Assessment, edited by L.T. Molina and M.J. Molina). The Project used emissions information from the 1997 IMADA measurement campaign to calculate the ambient ozone concentration as a function of time and location. The ambient ozone measurements did not agree with the air modeling calculation unless the hydrocarbon concentrations were multiplied by a factor of two to three. The Project conducted an exploratory field study in the MCMA to gather additional emissions measurements in February 2002 in preparation for an intensive field measurement campaign in the MCMA in April 2003. More than two dozens institutions from the US, Europe and Mexico, involving more than a hundred investigators will participate in this campaign. The 2003 effort adds a significant number of new state-of-the-art instruments, some on the Aerodyne Inc. mobile lab and some on a supersite installed at CENICA (Centro Nacional de Capacitación Ambiental), as well as in the boundary sites of the MCMA. We expect to generate lots of information that will enable us to further elucidate the atmospheric chemistry of the Mexico City Valley and reduce uncertainties in emission inventories. The insight obtained will be crucial for protecting human health and ecosystem viability in Mexico City as well as other megacities around the world.

Another important activity of the Project is the integrated scenarios analysis. While we cannot predict the future, we can estimate what might happen if Mexico City continues to grow and if different severity of control strategies are implemented. One of the unique features of the Project is the close collaboration with the Mexican government officials, so that our research findings can provide the scientific foundation for the decision making. Communicating to the public on how their actions can improve air quality is also part of the outreach program that is an important component of the Project.

Session 1: Emission Inventory Applications, Uncertainties, and Data Requirements

Bill Kuykendahl, US Environmental Protection Agency (EPA), described the emission inventory and air modeling requirements for states in the US that did not meet the National Ambient Air Quality Standards (NAAQS) to demonstrate that their control strategies were indeed working to reduce air pollution. The public in the US has access to these studies. The rules were established in the Clean Air Act. However, the Clean Air Act failed to account for the importance of pollutant transport from one state to another. Transport is especially a problem for long-lived pollutants and secondary pollutants such as ozone and particulate matter. Regional control strategies are now also important in the US. Jim Wilson, Pechan, described the shift from state to regional emission inventories. In the US power plant emissions are closely followed; hourly emissions are reported to the EPA. J. David Mobley, EPA, described the new NARSTO assessment on particulate matter to be released in March 2003. Robert Harley, University of California at Berkeley, and David Allen, University of Texas at Austin, each talked about the criteria for emission inventories that will be used as sources of information for air quality modeling. For air quality modeling emissions must be specified spatially and temporally. Spatial coordinates are a few square kilometers in area in a three-dimensional grid. Times are specified as hours of the day, days of the week, and seasons of the year. Particulate matter inventories need to account for size and chemical composition of the pollutant. Based on studies in Houston, TX, extreme values of point source hydrocarbon emissions from chemical plants and refineries can have a large effect on the ozone concentrations. Intermittent high hydrocarbon emission events caused an underestimation of the point source hydrocarbon emission inventory by factors of five to ten. This illustrates the critical role that field measurements play in the development and performance evaluation of emissions inventory.

Session 2: Mexico Emissions Inventory

Adrian Fernandez, INE/SEMARNAT, described the Mexico National Emission Inventory Project, which started in 1996 with the support of US EPA and the Western Governors Association (WGA), later with help from the North American Commission for Environmental Cooperation (NACEC). The national emission inventory is starting in the border area for the cities of Aguascalientes, Mexicali and Tijuana, but will cover the entire country. Richard Halvey, WGA, described the decision making process in forming an emission inventory and the importance of the public having access to the information. Gathering emissions inventory information in Mexico presents unique challenges: jurisdictional issues, confidentiality laws, reporting requirements, data availability must all be considered during the planning and implementation portions of any project.

Session 3: Innovative Techniques for Measuring Emissions

Charles Kolb, Aerodyne Research Inc., talked about new opportunities for emission information by using new measurement tools. He illustrated this with some on-road formaldehyde data obtained in 2002 in real time in both Boston and Mexico City. The much higher formaldehyde concentrations and their association with carbon dioxide indicated much of the increase in formaldehyde in Mexico City could be due to primary emissions from automobiles. Brain Lamb, Washington State University, described the Urban Tower Flux Experiments to be established near CENICA and the instruments that will be used in the MCMA-2003 field campaign. Gene Tierney, US EPA, mentioned that EPA is promoting the development and commercialization of Portable Emissions Measurement Systems (PEMS) because they offer many benefits over traditional laboratory measurement techniques.

Session 4: Bottom-up Methods

Rodrigo Perrusquía, Gobierno del Distrito Federal, reviewed the recommendations given by the MIT Mexico City Project to the 1998 emissions inventory. He said that the three most important recommendations for mobile sources are knowing better the number of vehicles, the distribution by model-year and the emission factors used. He discussed the basis and uncertainties in the vehicle counts. Sandeep Kishan, ERG, described the adoption of the US EPA MOBILE6 to conditions in Mexico. Matt Barth, University of California at Riverside, discussed the history and recent advances in mobile source emission modeling. He described in some details the Comprehensive Modal Emissions Model (CMEM) and the International Vehicle Emission (IVE) Model developed at UC Riverside. The later has been successfully developed and applied to locations outside the U.S., including Chile, Kenya, India, and China. Gene Tierney pointed out that EPA’s upcoming mobile source emissions model, MOVES, will be based much more on on-road vehicle emissions and will have the capability to make a life cycle analysis. Robert Harley commented to gain confidence in the estimates of emissions we need to make multiple independent estimates. His research activities use a highway tunnel as a laboratory to measure changes in vehicle emissions over a number of years. Vehicle emissions are estimated from emissions per fuel used multiplied by the amount of fuel sold in the area. The results from 1994 to 2001 showed an emission decrease of about 50% for NOx, 62% for CO and to 67% for VOC even though the amount of fuel used was increasing. The vehicle emission reduction is caused by better vehicle emission control equipment, especially better catalytic converters.

Sergio Sánchez, SEMARNAT, said all industries in Mexico are required to provide data related to general information of the industry, processes, raw material and inputs, air, water and soil emissions and pollutant transfers. Information about proprietary processes is confidential similar to US regulations. There is a list of 104 substances and the criteria pollutants are included as well as greenhouse gases. Jesús Contreras, SEMARNAT, gave an introduction to the stationary emissions inventory efforts in previous years and the air quality programs. The last EI is the 1998 estimate. Following the recommendation from Mario Molina’s team at MIT, several techniques were used in refining the 1998 EI: source monitoring, emission factors, materials balance, and extrapolation. A large effort is being made to reduce uncertainties in the estimation of emissions. The focus is on NOx, HC, PM10 and PM2.5 emissions. Jim Wilson talked about similarities between data registered in the US and Mexico and about differences in confidentiality of information. He discussed uncertainty analysis results found by the National Academy of Science (NAS). Proper characterization of uncertainty is essential; most analyses underestimate uncertainties, thus giving decision makers a false sense of confidence.

Rocío Reyes and Cuitlahuac Cruz, Secretaría de Ecología del Estado de México, reviewed the biogenic emission inventory for the Valley of Mexico. These talks included a description of the data requirements to be used in the Biogenic Emissions Inventory System (BEIS) model developed by EPA. They included climatology, humid and dry seasons, and land use, and emissions of isoprene, terpenes, VOC, and NOx. Luis Gerardo Ruiz Suárez, Centro de Ciencias de la Atmósfera, UNAM, discussed limitations in the use of the PC BEIS and in the use of the data for modeling. There is also no uncertainty analysis or validation with other methodologies. Thomas Pierce, National Oceanic and Atmospheric Administration (NOAA), described the evolution and advances of the BEIS and its application to air quality simulation modeling within the United States. Biogenic emissions are estimated to contribute more than 50% of VOC emissions in the US. However, depending on the methodology used, large differences in isoprene emissions may be estimated. He recommended ways to improve the estimation of biogenic emissions.

Jorge Sarmiento, Gobierno del Distrito Federal commented on the air quality problems and the efforts to estimate emissions inventories in the MCMA. The goal is to have a new estimate every 2 years. Mobile and area sources were considered the most important. The PM10 estimate is known to be low because it does not include emissions from either paved or unpaved roads. Many sources were not included in the 1998 EI due to lack of data and uncertainty about their reliability, especially in the informal sector (e.g., street vendors).

Session 5: Top-Down Methods

Rafael Ramos, Gobierno del Distrito Federal, described the current Mexico City Air Quality Monitoring Network. The cost of the monitoring operation is in the order of 2 million dollars per year. David Allen talked about the use of the results of an air quality study to assess the accuracy of the EI using aircraft data and its application to areas of Texas. Results of ozone formation and HC reactivity in Houston showed that VOC emissions are underestimated. Interaction with stakeholder and community leaders has increased the confidence on policy-making and the value of scientific partnerships. Eric Fujita, Desert Research Institute, talked about how tunnel studies have shown that gasoline and heavy-duty diesel vehicle emissions have been greatly reduced over time except for particle emissions, which have shown only a small decrease. He discussed how emissions could be traced to their source using chemical composition.

Session 6: Wrap up Session

As part of the project’s outreach efforts, this session was transmitted via the Mexican Education Network (EDUSAT) to other cities in Mexico. Sergio Sánchez gave an introduction to the session and explained the importance of reaching other audiences. Mario Molina emphasized the importance of the emission inventory and how it fits under the study being developed for the MCMA. A very precise emission inventory is expensive to produce and update, so only the degree of precision that is necessary should be used. However, different approaches to the inventory, using independent data, should be made so that the uncertainty can be estimated. These include measurements, fuel consumption, levels of activities, etc. The wrap up session included questions addressed to a panel of speakers and other workshop participants by Mario Molina and follow-up questions and comments from the audience. Topics discussed include what kind of training is needed to compile and use emissions inventories; what can Mexico learn from the international experience; how should Mexico decide which emissions control strategies should be put in place and how do we know whether the control strategies are working; and finally what can average person do to reduce emissions and hence to help improve the air quality. Luisa Molina closed the workshop by thanking the speakers, the participants and the sponsors for a very productive workshop.

Gerardo Mejía is a Professor and Francisco Obregon is a graduate student at Tecnológico de Monterrey (ITESM). Robert Slott is a Visiting Engineer with the Integrated Program on Urban, Regional, and Global Air Pollution.