The Local Benefits of Global Air Pollution Control in Mexico City - Final Report of the Second Phase of the Integrated Environmental Strategies Program in Mexico ppt

The Local Benefits of Global Air Pollution Control in Mexico City Final Report of the Second Phase of the Integrated Environmental Strategies Program in Mexico by Galen McKinley, Miriam Zuk, Morten Hojer, Monserrat Avalos, Isabel González, Mauricio Hernández, Rodolfo Iniestra, Israel Laguna, Miguel Ángel Martínez, Patricia Osnaya, Luz Miriam Reynales, Raydel Valdés and Julia Martínez Instituto Nacional de Ecología, México Instituto Nacional de Salud Publica, México August 2003 Table of Contents I Executive Summary McKinley II Project Summary III Emission Reductions and Costs III.1 General Methodology III.2 Renovation of the taxi fleet III.3 Extension of the Metro III.4 Hybrid buses III.5 Measures to reduce leaks of Liquefied Petroleum Gas III.6 Co-generation McKinley and Zuk McKinley Hojer Osnaya McKinley McKinley Laguna IV Air quality modeling McKinley and Iniestra V Health impacts analysis Zuk with Avalos, Martínez, Hernández, González, Reynales and Valdés VI Valuation Zuk with Avalos, Martínez, Hernández, González, Reynales and Valdés VII Integration: The Co-Benefits model VIII Results McKinley IX Conclusions McKinley McKinley and Zuk Appendix A Air Quality Modeling McKinley and Iniestra Appendix B Capacity Building Zuk and McKinley Appendix C Basic User’s Guide for the Co-Benefits Model McKinley and Zuk Acknowledgements: We thank the U.S Environmental Protection Agency (EPA) and the U.S.-Mexico Foundation for Science (FUMEC) for their support of the project We appreciate the input of Dr Adrián Fernandez of INE We also thank Dr Jason West of the US EPA for his attention to the project ii Contact Information: Consultants to Instituto Nacional de Ecología: Galen McKinley Miriam Zuk Morten Hojer galen@alum.mit.edu mzuk@alum.mit.edu mhoejer@hotmail.com Instituto Nacional de Ecología: Julia Martínez Montserrat Avalos Isabel González Rodolfo Iniestra Miguel Ángel Martínez Israel Laguna Patricia Osnaya jmartine@ine.gob.mx moavalos@ine.gob.mx igmerino@ine.gob.mx riniestr@ine.gob.mx mmartine@ine.gob.mx ilaguna@ine.gob.mx posnaya@ine.gob.mx Instituto Nacional de Salud Publica: Mauricio Hernández Luz Miriam Reynales Raydel Valdes mhernan@correo.insp.mx lreyns@yahoo.com rayvs@insp.mx iii Chapter I Executive Summary From September 2002 to August 2003, the Second Phase of the Integrated Environmental Strategies Program in Mexico was undertaken at the Instituto Nacional de Ecología (INE; National Institute of Ecology) of Mexico In this report, activities and findings are summarized During this project, the following goals have been achieved: • • • • Estimate cost savings due to health improvements related to air pollution reductions occurring simultaneously with greenhouse gas (GHG) emissions reductions, Compare costs and benefits for the specific policy measures, Build capacity in the Mexican government for integrated, quantitative environmental and economic assessment, and Provide results and tools with relevance to emission control decision- making process in Mexico City We produce estimates of annualized reductions of emissions of local and global air pollutants and program costs for three transportation measures (taxi fleet renovation, metro expansion, and hybrid buses), one residential measure to reduce leaks of liquefied petroleum gas (LPG) from stoves, and one industrial measure for cogeneration for the periods 2003-2010 and 2003-2020 at several discount rates Using reduced-form air quality modeling techniques, the impacts of changed emissions on exposure are calculated Then using dose-response methodology, public health improvements due to reduced exposure are estimated Finally, various valuation metrics are applied to determine the monetized health benefits to society of the control measure We find that the measures considered in this study could reduce annualized exposure to particulate air pollution by 1% and to maximum daily ozone by 3%, and also reduce greenho use gas emissions by 2% (more than 300,000 tons C equivalent per year) for both the time periods We estimate that for both time horizons, over 4400 quality-adjusted lifeyears (QALYs) per year could be saved, with monetized public health benefits on the order of $200 million USD per year In contrast, total costs are under $70 million USD per year The mean cost per QALY is estimated to be under $40,000 for the measures Of the measures considered, transportation measures are most promising for simultaneous reductions of both local and global pollution in Mexico City This analysis has been integrated in to an user-friendly modeling tool using Analytica software The Co-Benefits Model has been made available to decision- makers and their staffs in Mexico City There is interest from these groups in applying the model to their work and in modifying it for use in other regions of Mexico, particularly the City of Toluca in the State of Mexico Capacity building has been a major part of this project A large group of INE staff have actively contributed to the research effort Regular meetings and training sessions have been held with members of the Metropolitan Environmental Commission (CAM) and other environmental agencies in the Mexico City These meetings have encouraged active participation in this project and aided the integration of this work with other air pollution control efforts in the region Chapter II Project Summary II.1 Introduction Due to complex socio-political, economic and geographical realities, Mexico City suffers from one of the worst air pollution problems in the world Greenhouse gas emissions from the City are also substantial In this study, we compare the costs and benefits of a set of politically- relevant air pollution control measures for the City and simultaneously consider the greenhouse gas emission impacts of these measures We find that with control measures, it would be possible to reduce annualized exposure to particulate air pollution by 1% and to peak ozone by 3%, and also to reduce greenhouse gas emissions by 2% (more than 300,000 tons C equivalent per year) for the time periods 2003-2010 and 2003-2020 We estimate that for both time horizons, over 4400 quality-adjusted life- years (QALYs) per year could be saved, with monetized public health benefits on the order of $200 million USD per year In contrast, total costs are under $70 million USD per year The mean cost per QALY is estimated to be under $40,000 for the measures We find that transportation measures are likely to be the most promising for simultaneous reductions of both local and global pollution in Mexico City II.2 Motivation With nearly 20 million inhabitants, 3.5 million vehicles, and 35,000 industries, Mexico City consumes more than 40 million liters of fuel each day It is also located in a closed basin with a mean altitude of 2240m The combination of these and other factors has led to a serious air quality problem In 2002, Mexico City air quality exceeded local standards for ozone (110 ppb for hour) on 80% of the days of the year Particulate 24- hour standards were exceeded on 5% of the days (SMA, 2002) Greenhouse gas (GHG) emissions from Mexico City are also significant In 1998, Mexico ranked as the 13th largest GHG producing nation Mexico City emits approximately 13% of the national total (Sheinbaum et al., 2000) Using a 3.3% annual growth rate (West et al., 2003) and a 1996 base year estimate of 45,585,000 tons of CO2 (Sheinbaum et al., 2000), we estimate that the annualized GHG emission of Mexico City for the period 2003-2010 and 2003-2020 will be 17 million tons of C equivalent per year and 20 million tons C equivalent per year, respectively As emissions of GHG and local air pollutants are often generated from the same sources, there may exist opportunities for their joint control In this study, we have developed a cost-benefit analysis framework to analyze the trade-offs between costs, public health benefits, and GHG emission reductions for a select set of control measures In an effort to disseminate the knowledge collected in this work, we have also created a reduced- form analysis tool for use by policy makers This study fits into an ongoing process of analysis and action regarding Mexico City air quality At present, Mexico City government is currently in the process of implementing its third air quality management plan The first plan, PICCA (Programa Integral para el Control de la Contaminación Atmosférica) was initiated in 1990 and had several major accomplishments, including the introduction of two way catalytic converters, the phase out of leaded gasoline, and establishment of vehicle emissions standards The second program, PROAIRE (Programa para Mejorar la Calidad del Aire en el Valle de México 1995-2000) achieved the introduction of MTBE, restrictions on the aromatic content of fuels and reduction of sulfur content in industrial fuel While significant improvements in ambient air quality have improved, levels remain dangerously high, therefore the government has recently initiated the third plan, PROAIRE 2002-2010, as an extension of previous plans PROAIRE 2002-2010 includes 89 control measures targeting emissions reductions from mobile, point and area sources, as well as proposing education and institutional strengthening measures to combat the air pollution that afflicts the city While some of these measures are slowly being implemented, little quantitative analysis has been done prior to designing this plan Decision makers are now faced with the difficulty in setting priorities when presented with a such a large range of control options Several studies are currently quantitatively analyzing these issue (Molina et al., 2002) A recent study by West et al (2003) aimed to analyze a large number of PROAIRE and climate change control measures to determine the least cost set of options for joint control This study builds on these works, by simplifying and integrating the analysis to provide real time answers to policy makers II.3 Methodology Emissions Reductions and Costs for Specific Control Measures We estimate the time profiles of local pollutant (PM10 , SO2 , CO, NOx , and HC) and global pollutant (CO2 , CH4 , and N2 O) emission reductions, and costs for control measures that address transportation, residential and industrial emission sources We estimate emissions reductions and costs for each year from 2003 to 2020 such that the different time-profiles of the programs’ costs and impacts can be studied These two time horizons were chosen to allow us to analyze the short term on the time frame of the plan itself, and a longer term analysis on the scale of the project implementation For incorporation into the cost – benefit analysis, results are annualized using several discount rates In this Project Summary, we present results using a 5% discount rate only Below, key aspects of the control measures analyzed in this study are outlined In Tables II.1 and II.2, the estimated emissions reductions and costs of these measures are presented Taxi fleet renovation • 80% of old taxis are replaced by 2010 • Fuel efficiency increases from 6.7 km/L to km/L • Tier I technology is assumed in 1999 and newer models • Changes in emissions of primary particulate matter are not estimated Metro expansion • 76 km of new construction by 2020 (5 km between 2003 and 2010, 71 km from 2011 to 2020) • Riders assumed to come from microbuses and combis • Recuperation value of capital is included, using a 30 year useful life Hybrid buses • 1029 hybrid buses are brought into circulation, replacing diesel buses, by 2006 • Emissions factors from detailed study for New York City (MJ Bradley and Associates, 2000) LPG leaks • • Stove maintenance is performed in million households to eliminate leaks This is a combination of measures that each address a specific part of LPG stove systems (TUV, 2000) Cogeneration • Installation of 160 MW of capacity by 2010 • Recuperation value of capital is included, using a 20 year useful life Table II.1 Annualized emissions reductions (tons / year) Control Measure PM 10 SO2 CO NOx HC CO2 CH4 N2 O Time horizon 2003-2010 Taxi Renovation 64 165,483 5,135 16,863 275,007 64 498 Metro Expansion 3,518 155 324 19,567 Hybrid Buses 73 14 566 -119 274 54,063 LPG Leaks 0 75 2,480 7,475 590,080 10 Cogeneration 0 Time horizon 2003-2020 Taxi Renovation 59 146,380 3,060 12,811 257,542 60 466 Metro Expansion 65 28,835 1,271 2,653 160,368 39 Hybrid Buses 82 16 635 -134 307 60,656 LPG Leaks 0 0 1,954 5,888 0 0 13 110 856,031 15 Cogeneration Table II.2 Annualized abatement costs (2003 million US$ / year) Control Measure Public Investment Taxi Renovation Metro Expansion Hybrid Buses LPG Leaks Cogeneration 16.10 5.37 54.33 1.31 Taxi Renovation Metro Expansion Hybrid Buses LPG Leaks Cogeneration 8.90 44.05 30.04 0.73 Private Investment Time Horizon 2003-2010 53.66 0 1.81 4.83 Time Horizon 2003-2020 29.67 0 1.00 7.33 Fuel, Operations, Maintenance Total Cost -61.16 -0.01 -9.10 -1.39 -4.33 8.59 5.37 45.24 1.74 0.49 -57.33 -0.02 -10.21 -0.84 -6.40 -18.76 44.03 19.84 0.89 0.92 Exposure Modeling For the estimation of the impacts of emission reduction on ambient concentrations and population exposures, we have developed a range of reduced-form modeling approaches Results from a source apportionment study are used to estimate changes in primary and secondary PM10 Ozone isopleths from Salcido et al (2001) are used to estimate peak O3 changes occurring with changes in hydrocarbon and NOx emissions In order to account for the spatial relationship of population and pollution concentrations, as well as to account for annual exposures, we use reduced form models to provide a reduction fraction (RF) of pollutant concentration (Cesar et al., 2002; USEPA, 1999) This reduction fraction is then multiplied by projected population-weighted concentrations for the appropriate time horizon These projected concentrations use as a baseline the mean 1995-1999 observed, population-weighted (1995 census) 24- hour mean PM10 (64.06 ug/m3 ) or O3 maximum concentration (0.114 ppm), from Cesar et al (2000) The projection to future population-weighted concentrations is achieved by a linear interpolation of mean concentration results from the Multiscale Climate Chemistry Model (MCCM) model for 1998 and 2010 based on the emissions inventory for 1998 and emissions inventory projection for 2010 of the CAM (PROAIRE, 2002; Salcido et al 2001) To estimate changes in PM10 concentrations, the chemical species in the observed particulate matter are attributed to primary pollutants based on chemical analyses of the composition of particulate matter in the MCMA (Chow et al 2002) Fractional changes in the emission inventories of primary pollutants can then be related to fractional reductions in particulate concentrations Results of chemical analyses of the composition of particulate matter from sampling sites during the IMADA campaign of March 1997 (Chow et al 2002) are averaged, with weighting based on the total mass of each sample In order to attribute organic carbon to its primary (combustion) and secondary (hydrocarbon) sources, observed organic carbon is disaggregated into its primary and secondary contributions Following Turpin et al (1991), we estimate the primary organic contribution to total organic carbon based on a fixed ratio to elemental carbon mass of 1.9, a mean value for the Los Angeles basin The mass of secondary organic carbon is then the difference of the total organic carbon mass and the mass of primary organic carbon Total primary particulate mass from combustion sources (25%) is the sum of primary organic and elemental carbon Secondary organic carbon mass (2%) is attributed to hydrocarbon emissions Additionally, the mass of particles associated with geological sources (45%) is attributed to primary PM10 emissions from geologic sources; the mass of particles associated with total particulate ammonium nitrate (7%) is attributed to NOx emissions; and the mass of particles associated ammonium sulfate (11%) is attributed to SO2 emissions The peak mean O3 reduction fraction (RO max) is estimated from the fractional reductions in hydrocarbon (RHC) and NOx (RNOx ) by: RO3 max = 0.5353*RNOx - 0.2082*(RNOx )2 + 0.1112*RHC This equation is derived from a series of runs of the MCCM for Mexico City (Salcido et al., 2001) where HC and NOx emissions are varied in equal proportion from all sources and O3 concentration changes were recorded The above equation results from a polynomial regression fit to the results of Salcido et al (2001) These reduced- form air quality modeling approaches are limited by the still large uncertainty about fundamental processes responsible for ozone and particulate formation in the Mexico City Valley Further, the approaches have uncertainty due to the lack of spatial and temporal resolution and imperfections in the modeling and measurement techniques on which the approaches are based An exact quantification of the uncertainty is beyond the scope of this analysis Based on the work of Cohen et al (2003) and comparisons made during this study, we make a conservative estimate of 30% uncertainty on primary particulate results, and 50% uncertainty on the secondary particulate and maximum ozone results In Table II.3, concentration change estimates based on Source Apportionment and the Ozone Isopleth methods are shown for each of the control measures Table II.3 Annual particulate and maximum ozone exposure changes (ìg/m3 ) Taxi Renovation Metro Expansion Hybrid Buses LPG Leaks Cogeneration Taxi Renovation Metro Expansion Hybrid Buses LPG Leaks Cogeneration Particulates (PM 10 ) Mean 95% CI Time Horizon 2003-2010 0.36 (0.17 : 0.58) 0.01 (0.01 : 0.02) 0.14 (0.06 : 0.23) 0.07 (0.02 : 0.28) (0 : 0) Time Horizon 2003-2020 0.24 (0.12 : 0.38) 0.12 (0.07 : 0.18) 0.15 (0.07 : 0.25) 0.06 (0.02 : 0.12) (0 : 0.01) 10 Maximum Daily O3 Mean 95% CI 5.13 0.14 -0.07 0.91 0.06 (1.59 : 9.97) (0.04 : 0.28) (-0.14 : -0.02) (0.14 : 1.76) (0.02 : 0.11) 3.02 1.07 -0.07 0.74 0.08 (0.94 : 5.87) (0.33 : 2.08) (-0.14 : -0.02) (0.23 : 1.43) (0.02 : 0.15) Table VIII.4 Health Benefit per ton of GHG Reduced, 2003-2020 Taxis Fleet Renovation Metro Expansion Hybrid Buses LPG Leaks Cogeneration Mean $824 $906 $2,187 $2,447 $8 95 % CI $349 : $1,677 $430 : $1,864 $815 : $5,323 $924 : $5,950 $3 : $18 VIII.5 References Cohen, J.T., J.K Hammitt, and J.I Levy (2003) Fuels for urban transit buses: A costeffectiveness analysis Environ Sci Technol 37 1477-1484 161 IX Conclusions and Future Work IX.1 Conclusions Taxi fleet renovation offers the most promising opportunity for the joint control of local and global pollution of the measures studied here Further, benefits might be found to be significantly larger than estimated here if changes in primary particulate matter emissions could be estimated The large potential benefits of this measure have already been recognized by decision- makers in Mexico City, and the implementation of this measure has begun as of 2002-2003 with public funding for the replacement of 3,000 taxis The LPG leak measure also provides benefits than are much larger than the total costs Emissions reductions and local benefits from this measure are small compared to the taxi fleet renovation, but investment costs are quite small, making implementation of the LPG leak measure relatively feasible from a decision- making standpoint Cogeneration provides more than 50% of the GHG benefits from this set of measures, but essentially no local benefit because it moves emissions of local pollutants into the valley, and health benefits from the reduced emissions at power plants located outside the valley are assumed negligibly small Were a similar study conducted at the national level, Cogeneration may turn out to be a promising joint local / global option because health benefits derived in populations living near to power plants could be considered This will depend, of course, on population exposure to emissions generated by electricity production across the country Metro Expansion has large local benefits, particularly for the long time horizon when the measure has been fully implemented However, investment costs for building more Metro are extremely high making its implementation unlikely Finally, the Hybrid Bus measure may have positive net benefits if the long time horizon is considered However, the analysis of this measure has large uncertainty because the emission factors used were derived for the altitude, driving conditio ns, and fuel mix of New York City, not for Mexico City Altitude has been shown (Yanowitz et al 2000) to significantly impact emissions behavior from heavy-duty vehicle technology, but these impacts have not been specifically calculated for the technologies under consideration here We recommend that a better understanding of emissions factors be obtained and also that the cost-effectiveness of other types of advanced technologies (e.g Cohen et al., 2003) also be considered in order to determine what would be the best advanced bus technology to introduce in Mexico City This work indicates that measures to improve the efficiency of transportation are key to joint local / global air pollution control in Mexico City The three measures in this category that are analyzed here all have monetized public health benefits that are larger than their costs when the appropriate time horizon is considered Global benefits, due to improved fuel efficiency, are also large In contrast, we find that traditional “no-regrets” electricity efficiency provide large GHG emission reductions, but not provide local benefits to 162 Mexico City because the majority of electricity is produced away from the valley in which Mexico City is located IX.2 Future Work Further work is needed to analyze more measures that cover a wider range of opportunities for joint local / global air pollution control Also very important is to quantify the air pollution improvements and cost savings that could be acquired were congestion reduced in MCMA Such an analysis may indicate that the benefits from transportation efficiency improvement are, in fact, much larger than estimates here Improved understanding of emission factors from new and old vehicles under Mexico City driving conditions is also greatly needed, and could significantly impact results Sensitivity analysis is also needed on the control options studied in this project The results presented here are, of course, dependent upon assumptions made about baselines, emission factors, implementation plans, etc Since we are attempting to predict the future, there is much uncertainty In order to address this uncertainty, sensitivity of results to these basic assumptions needed to be tested Results that are robust to the gamut of possible futures is the ultimate goal of this kind of analysis, making a complete sensitivity analysis a key next step Technical working groups among various agencies and institutions in Mexico City are needed in order to more precisely define control measures, and to improve the emissions factors used Working groups would be mutually beneficial to all parties involved, particularly given the limited resources available for this work in Mexico City, by facilitating interchange of the best-available information This project has evidenced in many ways the pressing need that decision-makers have for reliable rapid-assessment tools Reduced form air quality modeling techniques is one example; and the Co-Benefits Model that integrates this analysis is, of course, another The methods used here should be further studied and improved so that they can give ever- more reliable answers On the long term, maintenance and technical support must be continued so that the Model and the methodology upon which it is based can remain pertinent to the decision- making process In the near future, improved documentation and a more complete User’s Guide (ref Appendix C) is also needed 163 Appendix A: Air Quality Modeling A.1 Box Model Box models are the simplest of numerical models The region to be modeled is treated as a simple cell, or box, bounded by the ground on the bottom, the inversion base (or some other upper limit mixing) on the top, and the east-west and north-south boundaries on the sites The box may enclose an area on the order of several hundred square kilometers Primary pollutants are emitted into the box by the various sources located within the modeled region, undergoing uniform and instantaneous mixing The ventilation characteristics of the modeled region are represented by specification of characteristic wind speed and rate of rise of the upper boundary Fundamental to the box model concept is the assumption that pollutant concentration in a volume of air, a “box”, are spatially homogeneous and instantaneously mixed Under this assumption, pollutant concentrations can be described by a simple balance among the rates at which they are transported in and out of the air volume, their rates of emission from sources within the volume, the rate at which the volume expands or contracts, the rates at which pollutants flow out the top of the volume, and the rates at which pollutants flow out the top of the volume, and rates at which pollutants react chemically or decay Because of their formulation, box models can predict, at best, only the temporal variation of the average regional concentration for each pollutant species Consequently, they are capable of addressing only broad-scale regional questions The combined effects of local emission patterns and meteorological conditions generally give rise to significant spatial variations in pollutant concentrations So, clearly box models cannot be used to asses the effectiveness of emission control strategies that lead to spatially inhomogeneous emissions We have developed a box model for the MCMA that represents emissions, advection and dry deposition of primary PM10 The governing equation is: dPM  PM o − PM  PM o − PM  = u ⋅  + v ⋅  dt ∆x ∆y    ⋅  E PM  − vd h∆ t  + − 1  ∆x∆y∆z + ∆t ⋅ e      Equation A.1 Where u and v are mean zonal and meridional winds, respectively; Äx, Äy and Äz are the horizontal and vertical dimensions of the box; PMo is the concentration of PM on the boundaries of the box; PM is the concentration inside the box; E is the emission of primary particles; Ät is the residence time of a parcel of air in the box (=Äy/v); vd is the dry deposition velocity for particles of 0.1-10ìm ; and h is the height of the deposition layer Thus, the first two terms represent advection, the third represents emissions into the volume of the box, and the fourth represents dry deposition (Scire et al., 2000) 164 The steady-state (i.e time invariant) concentration in the box is found by setting  u v  E PM o ⋅   ∆x + ∆y  + ∆x∆y∆z    PM = v ⋅∆ t   u v   − dh  + − ⋅ e − 1  ∆x ∆y  ∆t       dPM =0: dt Equation A.2 We can solve this equation for the baseline emissions (E1) and for emissions under a given control scenario (E2), and then difference the results to arrive at the change in PM (ÄPM) concentration due to the emission change If ÄE = E2-E1: ∆E ∆x∆y∆z ∆PM = v ⋅ ∆t   u v   − dh  +  − ⋅ e − 1  ∆x ∆y  ∆t       Equation A.3 The result is an estimate of the change of concentration of primary particulates in the MCMA that results from the changes in emissions To estimate the reduction fraction of primary particulates using the box model, we simply divid e equation B3 by equation B2, to find: ∆E ∆PM ∆x∆y∆z RF = = PM  u v  ∆E PM o ⋅  +  ∆x ∆y  + ∆x∆y∆z    Equation A.4 It is important to say that following commentary from technical staff of the government agencies attending our regular meetings, we determined that the box model previously used in the study is particularly uncertain, and a less useful tool than Source Apportionment Thus, we have eliminated the box model as an explicit component of the analysis A.2 Marginal PM Method for Primary and Secondary PM By using a 3-dimensional tochemical model for Mexico City (MIT-CIT) to determine the sensitivity of 2o particulate precursors to changes in emissions of SO2 and NOx, and then an chemical equilibrium model to determine the sensitivity of 2o particulate formation to change in precur sor concentrations, West and San Martini (2001) estimate changes in secondary sulfate and nitrate particle formation with changes in SO2 and NOx emissions Using data from the La Merced monitoring station, during the IMADA campaign in March 1997, they find the following relationships: (dPM10 /dNOx) = 2.25e-5 (ug/m3 ) / (ton/y) (dPM10 /dSO2 ) = 3.36e-5 (ug/m3 ) / (ton/y) 165 We apply these relationships to the estimated emissions reductions from the control measures to estimate changes in secondary particle concentrations A.3 References for Appendix A: Air Quality Modeling Scrire, J.S., D.G Strimaitis, and R.J Yamartino (2000), “A User’s Guide for the CALPUFF Dispersion Model (version 5),” Earth Tech, Inc 521 pp West, J and I San Martini (2001) Report of the Fourth Workshop on Mexico City Air Quality, March 8-10, 2001, El Colegio de Mexico, Mexico MIT-Integrated Program on Urban, Regional and Global Air Pollution Report No 25, November 2001 166 Appendix B Capacity Building B.1 Introduction A key component of this co-benefits project has been in the building of capacity in the INE team and Mexican policy makers The main goals of the capacity building have been to: 1) Facilitate the continuation of the project by INE staff once this phase of project is completed 2) Introduce the analysis and the Co-Benefits model to Mexican policy makers, and to encourage their use of the model 3) Train individuals on Analytica software so that they are able to use this program to conduct integrated analyses for other projects Some of the main activities of the capacity building component have been regular meetings with the CAM, a final workshop, several short-courses and close collaboration with INE staff In this Appendix, we discuss some of our key accomplishments and lessons learned during the project B.2 Key Accomplishments Throughout the project we have held regular meetings with members of the Metropolitan Environmental Commission (CAM) and other environmental agencies in the MCMA, including the Secretariat of the Environment of the Government of the Federal District (SMA-GDF), the Secretariat of the Environment of the government of the State of Mexico (SEGEM), and the Directorate of Air Quality of the Federal Secretariat of the Environment and Natural Resources (SEMARNAT) These meetings have encouraged active participation in this project and to aid the integration of this work with other efforts in the MCMA, particularly the first revision of PROAIRE Our close collaboration with INE personnel has also been a fruitful one From the beginning of the project, INE researchers were encouraged to attend our regular presentations to the CAM and other government agencies From the beginning of 2003, we have also made efforts to bring multiple INE researchers into active participation on the Co-Benefits team Following is a list of the major capacity building activities undertaken: • On November 26, 2002 , we held a meeting with the CAM and other government representatives in which they explained their plans for the first 2-year revision of PROAIRE and we presented the goals for this project The discussion that followed considered how we can make this work useful to their PROAIRE revision There was much interest in this analysis from the CAM staff and several members from 167 the State of Mexico asked to be involved closely in the development of the CoBenefits model • On December 16, 2002, we met with Soledad Victoria from the State of Mexico to introduce her to the Analytica software and an initial version of the model • On February 12, 2003, we held a meeting with the CAM and other government representatives to illustrate the Analytica software and to introduce the developing Co-Benefits model Much valuable feedback was received about the level of detail appropria te for the model if is to be useful to decision- makers • On March 6, 2003, we had a productive meeting in which the estimation of costs and emissions were presented We focused on the accelerated retirement of taxis as an example • Our capacity building with a specific focus on INE researchers began with a shortcourse was led by Miriam Zuk on the estimation of health impact and valuation of these impacts (April 1, 2003), and on the Analytica model (April 3, 2003) INE investigators then began to study and work on specific exercises related to each of the modules • On April 22 and 24, 2003 intensive working meetings with the entire INE team were led by Dr Fernandez and Dr McKinley to discuss results, uncertainties, priorities for future work, and to begin the planning for the final workshop • On April 24, 2003, the air quality models were discussed with CAM and other governmental representatives We received valuable feedback during this meeting that led us to eliminate the box model from our final results • On April 30, 2003, the health and valuation modules were presented to CAM and other governmental representatives, and valuable feedback was received • On May 20, 2003, the final workshop was held Policy makers and technical staff were invited and m any key figures attended Consistent with their ever- increasing involvement with the project, INE investigators presented the bulk of the technical details of the project Interest in the project and the integrated analysis was high, and valuable comments regarding the work were provided by the audience • The depth and breadth of the comments during the final workshop led to an additional meeting for technical comments and discussion that occurred at INE on May 26, 2003 Representatives from the National University of Mexico Center for Atmospheric Science, the Mexican Institute of Petroleum (IMP), the SEGEM, and SEMARNAT attended to share their thoughts and for further discussion of the details of the analysis It is promising possibility that technical working groups, particularly between INE, IMP and SEGEM, will develop out of this meeting 168 • On June 4, 2003, we met with Dr Adrián Berrera of IMP to begin discussions about such collaboration These kinds of cross- institutional working groups would be highly beneficial to moving forward this type of integrated analysis in Mexico City We believe that INE is in a very good position to become a focal point for such effort • On June 16, 2003, we held a day- long course on the use of Analytica software and on the use of the Co-Benefits model The majority of time was spent doing modeling exercises using the software and the model Eight attendees were from the CAM, SEGEM, and GDF Another attendees from INE also took the course The feedback on the course was extremely positive, and there were multiple requests for an advanced course in the near future This course was a key step in the dissemination of results to the multiple government offices responsible for air quality in Mexico City, and therefore to improving decision-making on local and global pollution control B.3 Lessons Learned While capacity building has been a key component of the project since its initiation, we have learned many things over the course of 10 months and our understanding of the best ways to truly achieve capacity building has significantly improved At the start of the project, our focus was on designing the analysis and determining its scope Most time was spent in this development phase on detailed technical issues; few INE personnel participated in the planning phase Later, it was determined that if the analysis was to eventually be transferred completely to INE staff, more people needed to be involved As such we developed a large work group and had many fruitful meetings on the project as a whole and collaborations in the execution of specific parts of the analysis Through this process, we have learned that it is essential to involve key INE staff in the project planning and implementation phase from the beginning It is difficult to encourage participation and re-train staff every time new work groups are determined We recommend that in the future, a maximum of to staff be selected to work on the project and be asked by their superiors to dedicate a substantial portion of their time to the project Of this group, one leader would be assigned who understands the broad vision of the project, and is able to integrate the pieces Depending on the time available of the leader, either they could be responsible for maintaining and updating the model, or perhaps another staff member could be responsible for model maintenance Additionally, or members should be able to conduct and constantly improve the technical analysis in each of the modules: emissions and costs, air quality modeling, and health benefits assessment If tasks are not assigned and time is not dedicated, the gains achieved through this project may be lost once this phase is over We have also learned that in order for people to be truly involved with the project, they not only need to be present at meetings and understand the basics of the analysis, but they must be responsible for a part the project After key decisions are made, it is difficult to help people understand how the project evolved into its existing form Though we have made 169 much progress with increasing technical knowledge in the fields touched by this project, capacity building about developing and growing a project is still very much needed at INE Gathering an interdisciplinary team is difficult, only to be made more difficult when people have little time to devote or have minimal intellectual investment in a project We therefore recommend that for the continuation of this phase of the project and for future phases (or for other projects), it is very important to: 1) Designate an in-country leader who is responsible for both learning the technical details of the analysis and for further development and dissemination of the project 2) Give in-country participants tasks and responsibilities for the work to encourage their participation and learning 3) Keep in mind that the capacity building process must be started during the planning stage of the project If staff participate from the start of the project, their ownership and intellectua l interest will drive the capacity building process forward Capacity building in a top-down format is inherently ineffective 170 Appendix C Basic User’s Guide for the Co-Benefits Model C.1 Introduction This appendix is meant to provide the reader with basic instructions on how to access and use the Co-Benefits Model It indicates where to find the Model and the Analytica software required to run it on the worldwide web It assumes that the user has access only to the “Browser” version of Analytica software, which is free of charge This documentation is written based on version 5.7 of the Model This guide is written on the assumption that the user has read and fully understands the entirety of this Final Report in which the methodology implemented in the Model is discussed at length We strongly recommend all users to carefully read this Report before beginning to use the Model It is essential that the user remember that this Co-Benefits Model is designed only for use in Mexico City It is not applicable to other locations in its current form It is also key to remember that the emission reductions and costs required as inputs to the Model are: • Annualized emissions changes at discount rates of 0%, 3%, 5%, 7% for 2003-2010 and 2003-2020 • Annualized costs 0%, 3%, 5% and 7% for 2003-2010 and 2003-2020 The use of emission changes and / or costs calculated on other bases will cause erroneous results to be derived from the Model We also note that because our goal is to attract decision- makers in Mexico City to using the Model, the Model’s text is in Spanish C.2 Accessing the Co-Benefits Model The most recent version of the Model, as well as other information about this project, can be downloaded from: http://www.ine.gob.mx/dgicurg/cclimatico/benlg.html C.3 Accessing Analytica It is free to download the “Browser” version of Analytica from the website of Lumina Systems, Inc http://www.lumina.com With this version, the Model can be run, but it cannot be modified and input choices cannot be saved Input choices in the first window can be changed for each run, and new measures can be evaluated under the “Medida Nueva” module However, such changes cannot be saved and must be re-entered after the Model has been closed and re-opened 171 We suggest tha t before trying to use the Co-Benefits Model, the prospective user first familiarize herself with Analytica by following the Analytica tutorial guide, which can also be downloaded from Lumina, or by taking an Analytica training course C.4 The User’s Module of the Co-Benefits Model Upon opening the Model, the user will find the screen shown in Figure C.1 Figure C.1 The User’s Module of the Co-Benefits Model All options found in this first window are further defined and implemented under the “MODELO” module that appears in the upper right corner of the window All technical details required to understand these calculations are described in Chapters IV, V, VI, and VII of this report C.4.1 Input Options This section describes options that the user can change in order to customize their Model run All options are found on the left side of the first window of the Co-Benefits Model (Figure C.1) Input buttons are gray in color 172 Part 1: All Measures or Individual Ones (“Parte I: Todo o Individua l”) With this drop-down button, the user chooses if she will analyze all measures included in the Model (“Todo”) or only a single measure (“Individual”) If “Todo” is selected, results for all measures will be calculated simultaneously, and the Model will take longer to run Also, if “Todo” is selected, Parte II is deactivated by the Model and can be ignored by the user Part II: Control Measures (”Parte II: Medidas de Control”) If “Individual” is chosen in Part I, then this button is used to choose which measure to evaluate, or the option “Combination” can be selected If “Combination” is selected, the user also needs to use the “Tabla para Combinacion” to determine which measures will be summed together before the Model is run Combination (“Tabla para Combinacion”) Clicking on this button will bring the user to a table with all the control measures listed to the left and a series of 0’s and 1’s in the right column If the user wants to include a measure in her combination, she should alter the right column in order to have a “1” next to that measure Excluded control measures should have a “0” beside them Add a new measure (“Añadir su ‘Medida Nueva’”) If the user wants to define her own set of emission reductions and costs to be evaluated by the Co-Benefits Model, she should use this module Before doing so, there is likely significant work to be done (see Chapter III) in order to calculate emissions reductions and costs as: • Annualized emissions changes at discount rates of 0%, 3%, 5%, 7% for 2003-2010 and 2003-2020 • Annualized costs 0%, 3%, 5% and 7% for 2003-2010 and 2003-2020 Once emissions reductions and costs are calculated in this way, the user should double click on the “Añadir su ‘Medida Nueva’” module In part 1a, emission changes are entered into a table Pay careful attention that the measure identified is “Medida Nueva” and that the data are entered under the appropriate time horizon and discount rate Here, total changes in primary PM10 (PM2.5 is not currently active) should be entered (combustion + geological) In part 1b, changes in geological emissions of primary PM10 (PM2.5 is not currently active) should be entered The Model accounts for the fact that the total (combustion + geological) was entered in part 1a In part 2, costs are entered In each part, it is also possible to alter the values for the control measures inherent to the Model by entering the altered values into the “Medida Nueva” table 173 Time Horizon and Discount Rate (“Horizonte de tiempo” and “Tasa de Descuento”) With the first button, the user can choose to evaluate either the 2003-2010 time horizon or the 2003-2020 horizon With the second, the user opts for either not using discounting (0%) or to run the Model using a 3%, 5% or 7% discount rate Mortality Valuation (“Valoración de Mortalidad”) Here, the user chooses if mortality will be valued based on each case of mortality (”Casos”) or if each year of life lost will be valued (“Años”) Morbidity is always valued based on cases C.4.2 Output Options Result buttons are pink in color, and are located on the right side of the Model’s first window “Costos Acutal” returns annualized costs (US$ / yr) for the measure(s), time horizon, and discount rate chosen “Reducciones actual” gives the emission reductions (ton / yr) for the measure(s), time horizon, and discount rate chosen “Reduccion C equivalente” gives the change in emissions of greenhouse gases (tons Cequivalent / yr) for the measure(s), time horizon, and discount rate chosen “Cambio de concentración por medida” returns, for each measure and each contaminant, the change in population-weighted exposure (ug/m3) for the representative annualized year as estimated by the Model “Impactos totales” gives the number of avoided cases per year (if “Casos” is selected) or QALYs per year (if “Años” is selected) due to the selected control measures “QALYs totales” returns the total number of QALYs saved each year by the measure, independent of whether “Casos” or “Años” is selected in the input options section “Beneficios Monetarios” returns the monetary benefits, calculated with each of the three valuation metrics, for each avoided health impact (US$ / yr) “Escenarios de Beneficios” provides the total benefits for the high “Alta” and low “Baja” valuation scenarios (US$ / yr) “Costos / beneficios totales” provides Cost over Benefit ratios for the high “Alta” and low “Baja” scenarios If costs are negative, a negative ratio will be returned If benefits are negative, the result will be zero 174 “Costo / QALY” compares the cost of the control measure(s) to the QALYs it saves (US$ / QALY) “Beneficios Netos” returns the net benefits of the control measure(s), which is the Benefits minus Costs (US$ / year) “Beneficio / ton C eq.” compares the local health benefit and the GHG reduction (US$ / ton C eq.) “Precio de C equivalente” calculates the income per ton of GHG reduction (US$ / ton) that would be needed in order to make the sum of benefits and such “GHG Income” equivalent to the costs of the control measure If the result is zero, it indicates that either the benefits are already larger than costs without consideration of “GHG Income”, or that there is a net increase in GHG emissions due to the measure C.5 Citing Results If the results from the Model are used in a report or publication, it should be cited (with the appropriate year and version number) as: Co-Benefits Model for Mexico City, Version x.x, Instituto Nacional de Ecología, Mexico, 200y This report should also be cited: McKinley et al (2003) Final Report of the Mexico City Co-Benefits Project, Instituto Nacional de Ecología and the US Environmental Protection Agency Integrated Strategies Program, August 2003 If a modified version of the Model is used, or new control measures are entered into the existing Model, please cite this report as indicated above For the Model, please indicate the changes that have been made and by whom: Co-Benefits Model for Mexico City, Version x.x, Instituto Nacional de Ecología, Mexico, 200y Modified by (who) of (institution) in inputs for (costs and emissions of measure X, de dose-response Z, etc.) and in modules (A,B,C) in manner (D,E), etc 175 ... 2003, the Second Phase of the Integrated Environmental Strategies Program in Mexico was undertaken at the Instituto Nacional de Ecología (INE; National Institute of Ecology) of Mexico In this report, ... decision- makers and their staffs in Mexico City There is interest from these groups in applying the model to their work and in modifying it for use in other regions of Mexico, particularly the City. .. -6 3.37 -5 9.59 -5 5.95 -5 2.44 -4 8.79 -4 5.58 -4 2.53 -3 9.65 -3 7.16 84.39 64.82 45.53 27.20 20.18 -7 9.56 -7 5.29 -7 1.31 -6 7.28 -6 3.37 -5 9.59 -5 5.95 -5 2.44 -4 8.79 -4 5.58 -4 2.53 -3 9.65 -3 7.16 Table III.2.7