nteg rated H nvironmental Strategies Integrated Environmental Strategies Philippines Project Report Metropol itan Manila Focus on the Transport Sector Prepared by Manila Observatory, PHILIPPINES http://www.observatory.ph June 2005 ------- ------- Contents Executive Summary 1 1 Introduction 1 1.1 Background: Metropolitan Manila, Philippines 1 2 Objective of the project 2 3 Concept of the framework 2 4 Scoping Decision and Policies Considered 2 5 Methodology 3 5.1 Scenario Development 3 5.2 Air Pollutant Concentrations 6 5.2.1 The ISCLT3 Model 6 5.2.2 Emissions Inventory 6 5.2.3 Modelling Results 7 5.3 Health Effects Estimation 8 5.4 Economic Valuation 9 6 Results 10 6.1 Scenario Development 10 6.2 Air Dispersion Modelling Results 12 6.3 Health Impacts of Each Scenario 14 6.4 Economic Costs of the Health Damages Averted by Each Policy Scenario 16 6.5 Co-Benefits Results of Each Policy Scenario in terms of Emission Reduction 17 7 Conclusions and Recommendations 18 7.1 General Conclusions and Recommendations 18 7.2 Other Recommendations 19 ------- Project Report 1 Introduction 1.1 Background: Metropolitan Manila, Philippines 1.2 Objective of the Project 1.3 Organization of the Report 2 Conceptual Framework 3 Focus on the Transport Sector 3.1 Background 3.2 Scoping Decisions and Policies Considered 4 Methodology 4.1 Scenario Development 4.1.1 Criteria for Selection of Policy Measures 4.1.2 Analytical Framework 4.1.3 Estimation of Travel Demands 4.1.3.1 Study Area and Zoning System 4.1.3.2 Socio-Economic Characteristics 4.1.3.3 Transportation Network Characteristics 4.1.3.4 The Four-Step Model 4.1.3.5 Additional Procedures for Estimation of Transportation Demand 4.1.3.6 Calculation of Vehicle-Kilometers 4.1.4 Estimation of Vehicle Emissions 4.1.4.1 Emission Factors for Particulate Matter (PM) 4.1.4.2 Emission Factors for C02 4.1.5 Description of Scenarios: Policy Elements and Assumptions 4.1.5.1 "BAU" Scenario - Baseline or Business-as-usual 4.1.5.2 "MVIS" Scenario - Implementation of the Motor Vehicle Inspection System 4.1.5.3 "TDM" Scenario - Transportation Demand Management 4.1.5.4 "4STC" Scenario - Replacement of 2-Stroke with 4-Stroke Motorcycles for Tricycles 4.1.5.5 "BWMK" Scenario - Construction of Bikeways in Marikina 4.1.5.6 "BWMM" Scenario - Construction of Bikeways in Metro Manila 4.1.5.7 "Rail 2015" Scenario - Expansion of the Metropolitan Railway Network by 2015 4.1.5.8 "DPTBJ" Scenario - Diesel Particulate Trap (DPT) for Buses and Jeepneys 4.1.5.9 "DPTB" Scenario - Diesel Particulate Trap (DPT) for Buses 4.1.5.10 "CNGB" Scenario - Compressed Natural Gas (CNG) for Buses 4.1.5.11 "CMEJ" Scenario - Coco-methyl ester (CME) for Jeepneys 4.2 Air Pollutant Concentrations 4.2.1 The ISCLT3 Model 4.2.2 Inputs and Assumptions 4.2.3 Emissions Inventory 4.2.4 Modeling Results 4.3 Health Effects Estimations 4.3.1 Health Impact Assessment 4.3.2 PM10 As Pollutant Indicator for Health Impact ------- 4.3.3 Exposure Response Function 57 4.3.4 Health Outcome Variables 58 4.3.5 Health Outcome Variables and Corresponding Exposure Response Coefficients 60 4.3.6 Mortality Endpoint: Premature Mortality 61 4.3.7 Morbidity Studies 61 4.3.8 Other Health Variables Considered 63 4.3.9 Summary of Assumptions 63 4.4 Economic Valuation 64 4.4.1 Assumptions (PHMAP, 2003) 64 4.4.2 Benefits Transfer 65 5 Results and Discussions 68 5.1 Impact of Scenarios on Travel Demand 68 5.1.1 Total Travel Demand 68 5.1.1.1 Year 2005 69 5.1.1.2 Year 2010 70 5.1.1.3 Year 2015 71 5.2 PM and C02 Emissions 71 5.2.1 Total PM Emissions 72 5.2.1.1 Year 2005 72 5.2.1.2 Year 2010 74 5.2.1.3 Year 2015 76 5.2.1.4 Annual PM Emissions Reduction for Each Policy Scenario 77 5.2.2 Total C02 Emissions 80 5.2.2.1 Year 2005 81 5.2.2.2 Year 2010 83 5.2.2.3 Year 2015 84 5.3 Air Dispersion Modeling Results 85 5.4 Health Impact of Each Scenario 90 5.5 Economic Costs of the Health Damages Averted by Each Policy Scenario 94 5.6 Co-Benefits Results of Each Policy Scenario in terms of Emissions Reduction 98 6 Conclusions and Recommendations 102 6.1 General Conclusions and Recommendations 102 6.2 Policy Implications 103 6.3 Other Recommendations 104 References 107 ------- List of Figures and Tables Figure 1 Design and Analysis of Integrated Strategies 22 Figure 2 Towns and Cities of the Study Area 27 Figure 3 Traffic Analysis Zone (TAZ) 28 Figure 4 Road and Railway Network forthe Baseline Scenario (2005) 30 Figure 5 Transportation Demand Analysis: The Classical FourStep Model 31 Figure 6 VWidroseatthe Science Garden Station 49 Figure 7 Emissions (tons peryear)torthe 2005 baseline, 2015 business-as-usual, and 2015 combination of policies 51 Figure 8 Map ofstatbnary sources found in the 2002 DENR inventory 53 Figure 9 Calculated particulate concentrations in |jg/Ncmfor1he 2005 baseline, 2015 business-as-usual, and 2015 combination of policies 55 Figure 10 Projected Travel Demand perScenario: 2005,2010,2015 68 Figure 11 Projected Travel Demand: 2005 69 Figure 12 Projected Travel Demand: 2010 70 Figure 13 Projected Travel Demand: 2015 71 Figure 14 Total PM Emission perScenario: 2005,2020,2015 72 Figure 15 Projected PM Emissions: 2005 73 Figure 16 Projected PM Emissions: 2010 75 Figure 17 Projected PM Emissions: 2015 76 Figure 18 PM Emissions Reduction for All Policy Scenarios, 2005-2015 in tons/per. 80 Figure 19 TotalC02 Emission perScenario:2005,2010,2015 81 Figure 20 Projected C02 Emissions: 2005 82 Figure 21 Projected C02 Emissions: 2010 83 Figure 22 Projected C02 Emissions: 2015 84 Figure 23 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and Metro Manila arising from the business-as-usual (BAU) scenario 85 Figure 24 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and for Metro Manila arising from the implementation of the Motor Vehicle Inspection System (MVIS), the use of compressed natural gas (CNG) for buses and coco-methyl esters (CME)forjeepneys 86 Figure 25 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and for Metro Manila arising from the installation ofbikev^ys in Metro Manila and Marikina 87 Figure 26 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and for Metro Manila arising from the business-as-usual (BAU) scenario and the projected operation of the proposed railway in 2015 88 Figure 27 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and for Metro Manila arising from the installation of diesel particulate traps for buses only, and particulate traps for buses and jeepneys 88 Figure 28 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and for Metro Manila arising from the conversion of tricycles to four-stroke engines and the continued implementation of Traffic Demand Management 89 Figure 29 Mean annual PM concentrations (|jg/Ncm) per cily/municipality and for Metro Manila arising from a combination of all proposed measures excluding railways and tricycle engine conversion (top left) and the implementation of all proposed measures 90 Figure 30 Percent Contribution of the Cost of Each Health Effect to the Total Costs 96 ------- Table 1 Share of Present Travel Distance of Buses and Jeepneys 33 Table 2 Share of Private Modes 33 Table 3 PM Emission Factors 35 Table 4 C02 Emission Factors (g/km) 35 Table 5 Summary ofScenarios and Corresponding Assumptions 36 Table 6 Baseline Scenarios for2005,2010 and 2015 38 Table 7 Reduction in Vehicle Emission Factors-"l/M" Standards (l/M) 38 Table 8 Reduction in Vehicle Emission Factors -"Substantial Reduction" (STDS2) 39 Table 9 Reduction in Vehicle Emission Factors- "Substantial Reduction+CNG Buses" (STDS3) 39 Table 10 Calculation of Reduction of Traffic Due to TDM 40 Table 11 Modal Share of Bicycles in Marikina City 42 Table 12 Proposed Expansion of the Railway Nelwork (by 2015) 43 Table 13 DOE Alternative Transport Fuel Program-Cumulative Inventory of Alternative Fuel Vehicles 44 Table 14 Estimate ofShareofCNG Buses in 2005 and 2010 45 Tatte15 EstimateofShareofU1ilityVehiclesRunningonCMEin2005and2010 47 Table 16 Summary of PM emissions (in tons peryear) from mobile sources (2002 baseline case) 50 Table 17 PM emissions from stationary sources (2002) 52 Table 18 Assumed mass distribution and particle sizes used in the modeling 53 Table 19 Health Outcome Variable, Source of Data and Time Reference 59 Tatte20 Health Outcome and Relative Risks (Per 10|jg/m3 of PM10) 60 Tatte21 BenefitsTransfer UnitValuesfbrMortality and Morbidity Effects 66 Tatte22 Direct Medical Costs Year2002 (Philippine Pesos-1995 prices) 67 Table 23 Cost of Lost V\fork Days Year2002 (Philippine Pesos-1995 prices) 67 Table 24 PM Emissions tor Each Policy Scenario, 2005-2015 (1), in tons/day 77 Table 25 PM Emissions tor Each Policy Scenario, 2005-2015 (2), in tons/day 78 Table 26 PM Emissions Reduction tor Each Policy Scenario, 2005-2015(1) in tons/day 78 Table 27 PM Emissions Reduction tor Each Policy Scenario, 2005-2015 (2) in tons/day 79 Table 28 Numberof Cases averted by different policy scenarios in 2005 91 Tatte29 NumberofCases averted by Different Policy Scenarios in 2010 92 Tatte30 NumberofCasesAverted by Different Policy Scenarios in 2015 93 Table 31 Cumulative NumberofCasesAverted bythe policy scenarios from year of implementation to 2015 93 Table 32 Costs of the Health Damages per Policy Scenario per Year in Million Pesos, Philippines 94 Table 33 Cumulative Total Health Costs Averted of each Policy Scenario, in millions, 2003 96 Table 34 % reduction from BAU tor each year 98 ------- Project Team Fr. Daniel McNamara Dr. Ronald Subida Ms. Mary Anne M. Velas Dr. Flordeliza Andres Dr. Karl Vergel Dr. Emmanuel Anglo Fr. Roberto Yap Atty. Angela Consuelo Ibay With Support from: Mr. Collin Green Mr. Kevin Roseel Mr. Adam Chambers Ms. Katherine Sibold Dr. Luis Cifuentes Ms. Shalini Ramanathan USAID-Manila, Manila Observatory, NREL and US EPA staff ------- Integrated Environmental Strategies Philippines Executive Summary 1. Introduction 1.1 Background: Metropolitan Manila, Philippines Metropolitan Manila or Metro Manila (MM) is located on the southwestern coast of the island of Luzon around the mouth of Pasig River which drains into the Manila bay. The metropolis is composed of 17 cities and municipalities. Approximately 15% of the land area is occupied by industrial establishments. These factories are located in areas with good transportation links. The number of businesses and industries had increased considerably in the past decade. In 1997, there were about 26,500 industries in the whole Metro Manila, composed mostly of light and medium scale types such as manufacturing, food and pharmaceutical industries as well as a few refineries. These were concentrated in the north of the metropolis. In terms of the transport sector, there has been more than a four fold increase in the number of road vehicles in the past two decades in the Philippines, from less than a million in the late 1980's to almost 4.2 million in 2003. In Metropolitan Manila alone, the number of vehicles increased from about 600,000 in the early 1990's to approximately 1.4 million in 2003. This is about 33% of the total for the whole country. The increase is more pronounced among the diesel fuelled vehicles. The number of diesel fuelled vehicles in the whole country increased from about 331,000 in 1987 to 1.3 million in 2002, a more than three fold increase. Meanwhile, the number of gas fuelled vehicles increased from approximately 800,000 to 2.3 million in the same time period, a little more than a two fold increase. Fuel consumption naturally also increased considerably. In a recent review of the state of environmental health in the Philippines that reviewed available studies and data, four main problems were identified. Two of these main problems were related to ambient air pollution namely, dust- related diseases such as bronchitis, and lead poisoning particularly in Metropolitan Manila. Indoor air pollution was likewise considered. However, a recent study on indoor exposure to particulate matter, 10|im (PM10) and ------- nitrogen oxides (NOx) in Metropolitan Manila showed that there is no significant difference in the pollutant levels between inside and outside the homes. Therefore this particular aspect was deemed to be a consequential issue of the overall air quality problem. Realizing that more than 90% of the air pollution comes from mobile sources based on the 2003 emission inventory, the project decided to concentrate primarily on the transportation sector particularly in Metropolitan Manila. Metropolitan Manila, as described before, has more than one-third of all the vehicles in the country. It also has the highest density of population in the country and probably the highest levels of air pollution as well. 2. Objective of the Project This project quantified and assessed the public health benefits of different mitigation measures with special focus on transport issues, common to both controlling ambient air pollution and greenhouse gases emissions and made use of health and economic impact as parameters in evaluating the benefits of the mitigation measures. 3. Conceptual Framework To estimate the co-benefits of policies with regards both greenhouse gases (GHG) and local air pollution reduction, the following framework was pursued in this project. The analysis included policies which could include mitigation measures for both GHG and local air pollution. These policies are assessed as to their capability to reduce emissions of GHG and local air pollutants from both stationary and mobile sources. Scenarios are developed from these policies. From these assessments or scenarios, GHG emissions reduction are estimated, at the same time, the reduction of local air pollutants is also estimated. Together with the development of scenarios from the policies, a business-as-usual scenario is also set. Ambient concentrations of the air pollutants are eventually modelled. Changes in ambient concentrations of the air pollutants are calculated by comparing the different policy scenarios with the business-as-usual scenario. Health benefits are then estimated based on the differences in ambient concentrations and concentration response functions that associate changes in ambient pollutant levels with specific health impact endpoints. Economic values of the avoided health impacts (or benefits) are likewise computed. 4. Scoping Decisions and Policies Considered Based on a scoping meeting held early on in the project, several mitigation measures were considered for the scenario development. For this project's ------- purpose, these mitigation measures are called policy scenarios. Eight individual policies and three combinations of policies are assessed. Although other mitigation measures were suggested during the scoping meeting, these were not considered in this assessment due to either political reasons or data availability. The base year used is 2002 and projections to years 2005, 2010 and 2015 are made. The following are the policies considered and for which scenarios were developed. (i) Transportation Demand Management through license plate scheme (TDM) (ii) Construction of Rail-based Mass Transit System (iii) Construction of Bikeways (iv) Implementation of the Motor Vehicle Inspection System (MVIS) (v) Introduction of the Compressed Natural Gas buses (CNG) (vi) Introduction of Cocodiesel for diesel-fuelled vehicles particularly jeepneys (CME) (vii) Two stroke tricycles switching to four-stroke engines. (viii) Improvement of vehicles by the Use of Diesel Traps (ix) Combo 1 - combination of policies: all policies except railways and switching of two stroke to four stroke tricycles (x) Combo 2 - all policies except railways (xi) Combo 3 - all policies including railways 5. Methodology 5.1. Scenario Development: The policy measures analyzed were screened based on the following criteria: (a) feasibility of implementation; b) socio-economic and political acceptability, and c) availability of information. The scenarios with the corresponding policy measure and assumptions were summarized in Table 1. Table ES-1. Summary of Scenarios and Corresponding Assumptions Scenario Policy and Assumptions Baseline or BAD 2005: 2005 transportation demand + 2005 transport Business-As-Usual (BAD) network +I/M Standards BAD 2010: 2010 transportation demand + 2005 transport network + l/M Standards BAD 2015: 2015 transportation demand + 2005 transport network +primary and secondary road network in 2015 + l/M Standards ------- Scenario Policy and Assumptions Reduction of PM emission factors and the corresponding percentages of vehicle types with reduced emission factors MVIS2005: + STDS2 - I/M Implementation of the STDS2 scenario without tie I/M scenario reduction in PM emission factor by 60% percent of vehicles: cars=25%, jeepneys=100%, buses=30%, trucks=30% MVIS2010: + STDS2 Implementation of the STDS2 scenario on top of the I/M Scenario Reduction of PM emission factor by 60% after the 30% reduction of emission factor under the I/M scenario percent of vehicles: cars=25%, jeepneys=100%, buses=30%, trucks=30% MVIS2015: + STDS3 Implementation of the STDS3 scenario on top of the I/M Scenario Reduction of PM emission factor by 60% after the 30% reduction of emission factor under the I/M scenario percent of vehicles: cars=50%, jeepneys=100%, buses=100%CNG, trucks=40% Transportation Demand Vehicle-kilometers of private transport modes such as gas car, Management (TDM) gas jeepney/utility vehicle and diesel car/utility vehicle were reduced by 11.08% in all 98 traffic analysis zones Replacement of 2-Stroke The PM emission factor of tricycles was reduced to 1/5 of the with 4-Stroke Motorcycles emission factor of tricycles in the baseline scenario applied to for Tricycles (4STC) 100% of the tricycles in all zones Construction of Bikeways The rates of shift (1.5% in 2005 and 3.5% in 2015) from tricycle (BWMK and BWMM) to cycling modes were applied as reduction rates of the tricycle vehicle-kilometers of traffic analysis zones Marikina (BWMK): applied to zones 74 and 76 only Metro Manila (BWMM): applied to all 98 zones Expansion of the Expansion of the metropolitan railway network by 2015 by Metropolitan Railway approximately 164.1 kilometers of new MRT/LRT lines and 19.7 Network by 2015 kilometers of busways according to the MMUTIS Master Plan (Rail 2015) resulting to reduced road-based traffic demand Implementation of the Motor Vehicle Inspection System (MVIS) ------- Scenario Policy and Assumptions Diesel Particulate Trap (DPT) Installation of the diesel particulate trap is expected to reduce for Buses and Jeepneys the PM emission factor of buses and jeepneys by 30% (DPTBJ and DPTB) DPTB: reduction of PM emission factor of buses only Compressed Natural Gas Reduction of emission factor of buses by 86% if diesel is (CNG) for Buses (CNGB) replaced by CNG 2005 (Low: 0.88%/High: 1.76% applied to zones passed by C-5. EDSA and SLEX) 2010 (Low: 11.47%/High: 22.93% applied to zones passed by C-5. EDSA, SLEX and NLEX) Coco-methyl ester (CME) Reduction of emission factor of jeepneys by 86% if diesel is for Jeepneys (CMEJ) blended with CME 2005 (Low: 0.64%/High: 1.27% applied to all zones) 2010 (Low: 2.0%/High: 4.0% applied to all zones) Combo 1 Combination of all scenarios except railways and switching of two stroke to four stroke tricycles (2005) Combo 2 Combination of all scenarios except railways (2010) Combo 3 Combination of all scenarios (2015) The methodology used for the analysis of the impact of transport- and fuel- related measures on emissions is similar to the environmental analysis model developed by the Metro Manila Urban Transportation Integration Study or MMUTIS (JICA, 1999). In this study, the total emissions for various policies were calculated as: Emissions = f (travel distance, travel speed, emission factors) Travel distance in terms of vehicle-kilometers and travel speed in terms of kilometers per day by planning zone were estimated using the 4-step travel demand forecasting model using the JICA STRADA, a travel demand analysis software. The MMUTIS Study defined 171 planning zones where 94 zones are in Metro Manila and the rest are in the nearby provinces. For the purposes of the IES study, these zones were combined to form 98 traffic analysis zones wherein 94 traffic analysis zones were constructed for Metro Manila and four (4) other zones corresponding to the four (4) adjacent provinces. 5 ------- Vehicle-specific and speed range-specific emission factors for PM are used to estimate the total emission for each policy. The emission factors were derived from earlier studies such as the ADB VECP (1992), MMUTIS (1999) and JSPS Manila Project (2002). The share of travel distance of jeepneys and buses, and the share of gasoline and diesel-fed vehicles by mode were also estimated. 5.2. Air Pollutant Concentrations 5.2.1. The ISCLT3 Model The ground-level concentration of particulates in Manila was predicted using the Industrial Source Complex Long Term Model (ISCLT3), which was selected due to its capability to predict the long-term concentration of pollutants from many sources and of many types using minimal meteorological data. ISCLT3 was run to predict concentrations in a 100-m receptor grid covering the entire Metro Manila. The study models PM10 only, and assumes that finer particles such as PM25 are part of the PM10 load. The model does not include secondary particulate formation. No background levels were added to the model results owing to the lack of data, although the contribution of stationary sources was included. 5.2.2. Emissions Inventory PM emissions associated with the traffic generated by the model are summarized in Table ES-2. Diesel vehicles appear to account for bulk of the emissions; private vehicles, due to their numbers, contribute more than public vehicles. Emissions from public gasoline-driven vehicles are exclusively from two-stroke tricycles, which are often poorly maintained and overburdened. Table ES-2. Summary of PM emissions (in tons per year) from mobile sources (2002 baseline case) Fuel Private Public Total Gas 939 4,254 5,193 Diesel 7,392 3,823 11,215 Total 8,331 8,077 16,408 The traffic flow generated by the NCTS traffic model was converted into emissions using appropriate emission factors. These traffic emissions, which were assumed to be uniformly distributed in each traffic zone, were then assigned to approximately 60,000 area sources each 100 m by 100 m in size covering Metro Manila. A color-coded map of PM emissions for the baseline, best-case and worst-case scenarios are shown in Figure ES-1. 6 ------- From the leftmost map in Figure ES-1, vehicular emissions in Metro Manila can be seen to be highest at the center of the city where business, commercial and educational facilities are clustered. The series of zones with significant emissions that trace rough lines leading to this center indicate the major traffic routes. Traffic after population growth in 2015 does not appear to alter these routes, as shown in the business-as-usual (BAU) map at the center of the same figure, but the increase in emissions from all the zones is evident. The potential for reduction in 2015 under the most optimistic scenario is presented in the rightmost map, where emissions may be seen to fall to less than half of the 2002 levels. Figure ES-1. Emissions (tons per year) for the 2005 baseline (left), 2015 business-as-usual (middle), and 2015 combination of policies (right) 5.2.3. Modelling Results Modeling results are shown as isopleth maps of ambient PM concentrations in Figure ES-2 for the 2002 baseline, 2015 business-as-usual (worst-case scenario), and 2015 under a combination of air quality management policies (best-case scenario). Poor air quality may already be seen from the 2002 baseline where highest annual concentrations from mobile sources alone reach 105 micrograms per Normal cubic meter (pg/Ncm), well above the Philippine standard of 60 pg/Ncm (indicated as red in Figure ES-2). These levels are confirmed by the observations of the Manila Observatory at the Epifanio de los Santos Avenue, Metro Manila's main artery, where 24-hour concentrations are consistently higher than this value. Conditions get worse 7 ------- in 2015 under the business-as-usual case (center map in Figure ES-2), where population growth heightens the exceedances. The potential for improving air quality resulting from the implementation of several policies is shown in the rightmost map of Figure ES-2, which shows the best-case scenario. Annual PM concentrations fall to less than half of their 2002 levels, and all exceedances disappear. Clearly, the adoption of even just a few strategically selected policies can result in dramatic improvement in Metro Manila's air quality. Figure ES-2. Calculated particulate concentrations in |ig/Ncm for the 2005 baseline (left), 2015 business-as-usual (middle), and 2015 combination of policies (right) 5.3. Health Effects Estimation The main methodology used was the health risk assessment approach using epidemiological studies, based on the Krzyanowski proposed method of assessing the extent of exposure and effects of air pollution in a given population. The basic principle of risk estimation could be illustrated by the following equation: Attributable Number of Cases = Exposure-response coefficient X excess exposure level X exposed population X baseline mortality/morbidity rates The number of attributable cases for each policy scenario was calculated including that of the 'Business-as-usual' scenario (BAU) for the projected years. The attributable numbers of cases for the policy scenarios are then 8 ------- subtracted from the BAU scenarios for the respective projected years. These latter figures comprise the averted number of cases for each policy scenario. The exposed populations which cover the whole population of Metro manila for the 2005, 2010 and 2015 are projected based on the population growth rates predicted for those years. The predicted population growth rates consider both the birth and migration rates. With regards the baseline morbidity and mortality rates, these rates are assumed to be constant and similar to the rates in 2002 for 2005, 2010 and 2015 in this estimation. All data input and calculations of the estimates were made using the Analytica software. In this study, PM10 was used as an indicator of urban air quality and a proxy indicator for concurrent exposure to different pollutants. The exposure response function or exposure correlation coefficient which is a measure of the relationship between variables, indicates the expected change in a given health outcome per unit of change of pollutant (PM10, in this case). The exposure response coefficient values used here were derived from time series studies rather than the cohort studies. The following health outcome variables were considered for this assessment: mortality, asthma, acute and chronic bronchitis and cardiovascular and respiratory hospital admissions. 5. 4. Economic Valuation In order to conduct the economic valuation, the unit cost values to translate health impacts into economic values were needed. Several methods were used to estimate unit cost values: benefits transfer, direct cost of illness (medical costs), indirect cost of illness (lost work days). Benefits transfer method was used in calculations of cost of most of the health outcomes. Values used were largely based on the adjusted values found in a study by Orbeta and Rufo (2003). The values were readjusted to set them in 1995 prices computed using Philippines Consumer Price Index and using the 2002 U.S. Dollar - Philippine Peso exchange rate. To estimate the unit costs of health impacts of different transport scenarios for years after 2002 (e.g., 2005, 2010, 2015), present values of the unit costs were calculated using a discount rate of 12%. Direct cost of illness or medical costs were also used to estimate 'avoided medical cost'. Data from the Philippine Health Insurance system were used for this purpose. For indirect cost of illness, the lost income due to work loss days was utilized. Estimates of the number of work days lost for a specific illness was made by expert judgment of physicians. The income lost per day was assumed to be the minimum daily wage rate in year 2002 mandated by Philippine law (PhP 181.53 in 1995 prices). ------- 6. Results 6.1. Scenario Development Figure ES-3 shows the results of the estimation of travel demand for 2005. The total travel demand for the BAU scenario in 2005 was estimated at 74.5 million vehicle kilometers. million vehicle-km per day 80 ^ ^ f f f f / / ~ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ~ gas jeepney ~ gas car Scenarios Combi = MVIS+TDM+CNGBH+CMEJH+DPTBJ Figure ES-3. Projected Travel Demand : 2005 Figure ES-4 shows the results of the estimation of PM emissions in 2005. For the BAU scenario in 2005, the total PM emission in Metro Manila was estimated at 48.4 tons per day or about 17,670 tons per year. The results of the simulation indicate that the greatest reduction in PM emissions relative to BAU scenario may be achieved with the combination of scenarios, which includes the MVIS, TDM, CNGBH or CNG for buses- high case, CMEJH - high case, DPTBJ or diesel particulate trap for buses and jeepneys, and the BWMM scenario. However, the 4STC scenario alone indicates an SPM reduction of almost the same magnitude as the combination scenario, that is, about 11.6 tons per day or 31 %. Figure ES-5 shows the results of the estimation of PM emissions in 2010. The total emission for the BAU scenario in 2010 increased by 9.8 tons per day or ------- 20% compared to the BAU scenario in 2005. The increase in emission between the scenarios in 2005 and in 2010 is largely driven by the changes in assumptions that affect emission levels, for example, emission standards and increase in vehicles using alternative fuels. ¦ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ¦ gas jeepney ~ gas car B M C C C c T 4 B B D D C A V N N M M D S W W P P 0 U I G G E E M T M M T T M S B B J J C K M B B B L H L H J 1 s ce n a r i os 'emissions from exhaust and idle Combi = MVIS+TDM+CNGBH+CMEJI-H-DPTBJ Figure ES-4. Projected PM emissions: 2005 * 70 60 50 t o n 40 s / d a y 30 20 10 =i=i ~ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ~ gas jeepney ~ gas car c N G B H Scenarios c emissions from exhaust and idle Combi = MVIS+TDM+CNGBH+CMEJH+DPTBJ+BWMM+4STC Figure ES-5. Projected PM emissions: 2010 11 ------- 6.2. Air Dispersion Modelling Results Calculated ambient PM concentrations for the business-as-usual scenario were averaged for the entire municipality and are presented in Figure ES-6. Air pollutant levels are expected to rise due to project growth if no measures are taken to actively reduce emissions from the transport sector. The city of Manila appears to receive the brunt of both the concentrations of PM as well as the growth of its levels, particularly between 2005 and 2010. The municipalities of Valenzuela and Navotas, which host the most number of stationary sources, appear to have lower spatially averaged particulate levels and minimal escalation, although concentrations at specific areas can be very high. BAU NCR-All ¦ Kalookan Las Pinas _x_ * Mandaluyong Makati I Marikina Muntinlupa Navotas Rasay Raranaque X Rasig * Rateros Quezon City San Juan Taguig Valenzuela o -5- c\i (o m s s s s s s Figure ES-6. Mean annual PM concentrations (|ig/Ncm) per city/municipality and Metro Manila arising from the business-as-usual (BAU) scenario Figure ES-7 shows the mean concentrations in Metro Manila arising from all the scenarios considered. The Motor Vehicle Inspection System (MVIS) appears to be the most effective. Remarkably, after 2010 the MVIS can actually cause air pollution levels to start decreasing as a result of the phaseout of the most pollutive vehicles and their replacement by new and cleaner units. Also effective is the shift to four-stroke tricycle engines, which can reduce mean PM concentrations nearly as much as the MVIS. The impact of fuel shift from diesel to compressed natural gas (CNG) for public buses and the use of coco-methyl esters (CME) in jeepneys each give 12 ------- rise to a reduction in ambient levels by about 10 percent compared to BAU. These results show that a shift to cleaner fuel will not reverse the rising trend in PM levels unless a higher percentage of vehicles currently on the road is targeted for conversion. ¦ Bus in ess-as-Usual ¦ Traffic Demand Mgmt. CNG for Buses CME for Jeepneys - Bikeways ¦ Diesel PMTraps (Buses and Jeepneys) ¦ Diesel PMTraps (Buses) -Shiftto 4-Stroke TCs Motor Vehicle Inspection ¦ Com bination-1 Com bination-2 Figure ES-7. Mean annual PM concentrations (|ig/Ncm) in Metro Manila arising from the business-as-usual and other scenarios The impact of traffic demand management, or the banning of certain vehicles daily based on their plate numbers, results in improvement only as much as those from the shift in fuel. Improvement is also marginal with the installation of particulate traps for diesel buses. However, if jeepneys are also fitted with these devices, the reduction will be markedly larger. Virtually no change in PM levels is expected from the construction of bikeways. However, the new railway envisioned to be completed in 2015 will cause a significant decrease (Figure ES-8), even if not all municipalities in Metro Manila will benefit. 13 ------- 2015 BAU 2015 Railway ¦*NCR-All -¦ Kalookan Las Pinas Malabon -* Mandaluyong -Makati iManila Marikina Muntinlupa Navotas Pas ay Paranaque Pasig Pateros Quezon City San Juan Taguig Valenzuela Figure ES-8. Mean annual PM concentrations (|ig/Ncm) per city/municipality and for Metro Manila arising from the business-as-usual (BAU) scenario and the projected operation of the proposed railway in 2015. Applying the combination of measures described earlier to address Metro Manila's air quality is forecast to cause a dramatic improvement in PM levels. In 2005, a 25 percent reduction is immediately expected. But more important, the increase in pollution levels is minimized all the way to 2015 where by this average levels in Metro Manila fall to about 40 percent of baseline levels. This scenario, which is based on realistic estimates in the application of proposed air quality measures, reiterates an earlier prediction that PM levels in the capital can be controlled through concerted effort. 6.3. Health Impact of Each Scenario Table ES-3 shows part of the overall results of the estimation exercise. It summarizes the cumulative health impact of the most significant single policy scenarios and the combination 1 scenario if implemented in 2005 until 2015. Combination 1 does not include the replacement of 2-stroke tricycles to 4 - stroke and the railways policies. In this analysis, the implementation of combination 2 which includes all policies except the railways starts in 2010 until 2015 and combination 3 with all the policies is implemented only in 2015. Apart from the combination scenarios, the policies which could avert the most number of cases in all the years presented, are the conversion of tricycles from two stroke to four-stroke and the Maintenance of Vehicles and 14 ------- Inspection System. The Maintenance of Vehicles and Inspection System policy assumes that through an emission testing system, the quality of vehicles would be maintained and emissions will be decreased. This system depends entirely on the sustainability and consistency of implementation. Other policies also yielded results showing cases of health outcomes that could be averted, however, they were not as large as the two policies featured here. Table ES-3. Cumulaiwe Number of Cases Averted by ihe policy scenarios from year of implementation to 2015. (Note: Combo 1: from 2005-2015, Combo 2: from 2010-2015 and Combo 3: Only 2015.) Respiratory Cardiovascular Asthma Asthma Bronchitis Policy Natutal Hospital Hospital Attacfe Attacfe Episodes Chronic Scenario Mortally Admission Admissions <15 >15 <15 Bronchitis [VMS 1899 603 111 102354 13639 3495 27997 (1424-2376) (47-1154) (60-161) (63328-142988) (6710-20437) (1658-5323) (2683-53099) CNG 614 195 35 33117 4414 1130 9058 (461-768) (15-373) (19-52) (20491-46265) (2171-6613) (537-1725) (867-17181) CME 585 185 34 31509 4199 1077 8618 (439-730) (15-357) (1849) (1949544017) (2065-6292) (510-1640) (827-16343) RAIL- 173 55 10 9341 1245 319 2555 WAYS (130-217) (4-105) (5-15) (5780-13050) P12-1865) (151-486) (24494846) Diesel 1203 382 69 64859 8643 2215 17739 Traps (902-1505) (29-732) (37-101) (40131-90608) (4252-12951) (1051-3375) (1700-33646) Bike MM 51 17 1 2690 359 94 736 (37-61) (0-32) (0-3) (1665-3759) (175-538) (42-139) (73-1397) TC 2082 660 121 112156 14946 3829 30677 48troke (1562-2603) (52-1266) (65-175) (69394-156683) (7352-22393) (1817-5834) (2939-58182) TDM 860 275 49 46370 6177 1584 12681 (645-1076) (22-524) (27-71) (2869064778) (3039-9258) (750-2413) (1216-24054) Combol 2827 898 163 152325 20299 5200 41667 (2120-3535) (70-1719) (89-239) (94247-212800) (9986-30413) (2465-7924) (3993-79021) Combo2 2643 840 153 142376 18972 4861 38944 (1982-3303) (64-1605) (23-224) (88092-198900) (9335-28428) (2307-7408) (3733-73860) ComboS 590 188 34 31801 4238 1086 8698 (443-738) (15-359) (18-50) (1967644425) (2085-6349) (515-1654) (834-16497) Replacement of 2 stroke with 4-stroke tricycles 15 ------- The combinations of policies scenarios are still ideal and yield the most health gains. Moreover, this approach of estimating the health gains of individual policies also illustrates the importance of each policy. This would help decision and policy makers to appraise the merits of each policy especially in the event that due to budget constraints, only one or a few of the measures could be implemented. 6.4. Economic Costs of the Health Damages Averted by Each Policy Scenario From the health damages section, economic costs of each policy scenario are calculated. The percent contribution of the costs of each averted health effect to the total cost of individual policy scenario shows that the total costs of the different measures or policy scenarios are dominated by the costs of the averted deaths and the morbidity due to chronic bronchitis. The premature mortality account for about 50% of the total cost of the policy scenarios while chronic bronchitis account for about 46% of the total. This occurrence is expected since cost per case of these two health effects is also quite high. In addition, this is consistent with another IES project result, e.g. Santiago, Chile. In Table ES-4, the cost of cumulative health impact is shown. The policy scenarios with the largest costs averted are the combination scenarios, the MVIS and the replacement of 2-stroke to 4 stroke tricycles. In addition, it should be noted that, in spite of the low targets for the CNG and CME scenarios, the cumulative costs averted are still quite staggering at more than 3 billion pesos each. Despite the large figures seen here, these estimates remain as conservative, since they do not cover all the health damages that caused by particulate air pollution. Table ES-4. Cumulative Total Health Costs Averted of each Policy Scenario, in millions, 2003 Policy Scenario Cumulative Total Cost Averted, in millions MVIS CNG Buses CME Jeepneys Railway Diesel Traps Bikeways in Metro Manila Tricycle Replacement to 4-stroke TDM Combo 1 Combo 2 Combo 3 12,126 3,705 3,523 538 7.812 304 14,141 5,468 19,083 13,359 2.813 16 ------- 6.5. Co-Benefits Results of Each Policy Scenario in terms of Emissions Reduction An evaluation of the policy scenarios in the reduction of emissions of particulates and the GHG gas, C02, is presented here. Table ES-5 shows that the replacement of 2-stroke to 4 stroke tricycles and the diesel particulate traps do not have an impact in the reduction of C02 emissions but had significant reduction in particulate emissions. In all the other policy scenarios, the reduction in particulates emissions approximates the reduction in C02. Minimal impact is also seen with the CNG for buses and CME for jeepneys scenarios. Of the six single policies, the MVIS and the railways have the most impact in mitigating both the particulates and the C02. As in the health damages and economic impact, the best results for mitigating both particulates and C02 are seen with the combination of policies. This comparison would be of additional assistance to policy makers in determining which policy or combination of individual policies would have the most impact for both local air pollution and GHG emissions. Table ES-5. % reduction from BAU for each year Policy Scenarios 2005 2010 2015 SPM C02 SPM C02 SPM C02 MVIS 12.5 14.7 21.0 21.3 29.0 30.0 CNG for buses 0.6 0.04 8.6 0.6 9.8 3.6 CME for jeepneys 0.1 0.1 8.6 0.4 9.8 3.0 TDM 4.1 4.8 10.0 4.6 13.0 4.7 Replacement of 2-stroke tricycles 24.0 0.0 27.5 0.0 28.0 0.0 Bikeways 0.5 0.1 0.5 0.1 1.6 0.1 Diesel Particulate Trap for buses 2.0 0.0 7.7 0.0 11.4 0.0 Diesel Particulate Trap for jeeps/buses 7.0 0.0 32.7 0.0 34.0 0.0 Railway NC NC NC NC 18.2 13.0 Combination 1 27.0 31.6 NC NC NC NC Combination 2 NC NC 57.1 32.0 NC NC Combination 3 NC NC NC NC 69.0 53.0 17 ------- 7.Conclusions and Recommendations 7.1. General Conclusions and Recommendations The savings from the health impact projected in 2005, 2010 and 2015 in implementing the single or combination of policies could mean very substantial savings on the health budget. In 2005, the savings from implementing the policy with the least health impact to the combination of policies would range from 0.13% to about 16% of the health expenditures for that year. In 2010, savings go up to almost 3% for the least, to more than 19% of the health budget, for the combination. In 2015, it's from 0.21% to 13% of the health budget. The savings in the national health budget that could be incurred is only based on savings from Metro Manila. If these policies could be implemented in the secondary cities, the savings on the national health budget could even be more. The following are the specific conclusions of the study: 1. Based on the assumptions made in the scenario development, three single policies have the advantage of having more health and economic benefits. These are the implementation of the maintenance of vehicle and inspection system, switching from four-stroke to two stroke tricycles, and use of the metro railways. These three policies must be seriously considered by decision makers particularly the Department of Transportation and Communication and the Metro Manila Development Authority. 2. The use of CNG in buses and CME among jeepneys did not show very important benefits because the assumptions on the targets, which were based on the government plans, were too low for any significant impact. The Department of Energy must exert extra effort to increase its target with regards the CNG and CME buses and jeepneys to have a more meaningful impact on air pollution. 3. C02 emissions can also be considerably reduced with the policies proposed specially the MVIS and the TDM. However, at most benefits in terms of reduction of both PM and C02 can be seen with the MVIS and the railways policy scenarios. 4. The other single policy scenarios also contributed to the reduction of air pollution and resulted to some health and economic benefits. These single policy scenarios are the collective responsibility of the DOTC, DOE, MMDA and the DENR. 18 ------- 5. As expected, the combinations of policies resulted to the most health and economic benefits as well as reduction of C02. Hence, if at all possible, these combinations of policies must be implemented. 7.2. Other Recommendations 1. The cost effectiveness of the policy scenarios must be calculated to be able to have a more comprehensive evaluation of these policy scenarios 2. Stringent implementation of the Metropolitan Manila Air Quality Plan 3. Extension of this type of analysis to other sectors and sources of pollution, e.g. stationary sources or industrial air pollution 4. Implementation of similar type of assessment for other cities in the country, e.g. Cebu, Cagayan de Oro, Baguio and Davao. 5. Collection of more reliable data, e.g. meteorological data, and general Improvement of the models used. 19 ------- lES-Philippines Project Report \ Metropolitan Manila Introduction .1. Background: Metropolitan Manila, Philippines Metropolitan Manila or Metro Manila (MM) is located on the southwestern coast of the island of Luzon around the mouth of Pasig River which drains into the Manila bay. Thirty seven percent of its land area of 636 square kilometers are used for housing which includes single and multiple residential units and slum and squatter areas. In the year 2000, approximately 12.7% of the population of Metro Manila was estimated to be at or below the poverty line which was equivalent to $450 per annum per person. This percentage is equivalent to more than 211,000 families and has increased from about 8% in 1995. It is likewise estimated that 35% of the population live in slum settlements or squatter areas. The "income gap ratio" in Metro Manila is about 22.1%. This means that the income of those below the poverty line would have to be raised on average by 22.1 percent to reach the poverty threshold. This ratio has increased from 1995 that was about 18%. Poor housing areas with high density populations are usually located around industrial, commercial and tourist areas that provide formal and informal employment for the residents. The rest of the land area is used either for commercial and industrial or government services, i.e. buildings and offices, streets, etc. The more affluent residential areas are located in the southern portions of Quezon City, at Greenhills in San Juan, in the western area of Mandaluyong and Pasig, and around the business district of Makati. All of these affluent areas are situated around EDSA (Epifanio Delos Santos Avenue) which is the principal circumferential 12-lane highway running in a semicircle with a radius of seven kilometres. In the past two decades, more affluent neighbourhoods have been established in the south of Metro Manila, particularly in the cities of Muntinlupa and Las Pinas. Approximately 15% of the land area of Metro Manila is occupied by industrial establishments. These factories are located in areas with good transportation links like the port area, the districts of Paco and Pandacan in the city of Manila, and in some areas of Pasig and Mandaluyong. The number of businesses and industries had increased considerably in the past decade. In 1997, there were about 26,500 industries in the whole Metro Manila, composed mostly of light and medium scale types such as manufacturing, food and pharmaceutical industries as well as a few refineries. These were ------- concentrated in the north of the metropolis especially in the north of Quezon City and in the cities of Caloocan and Valenzuela. In a recent review of the state of environmental health in the Philippines that reviewed available studies and data, four main problems were identified. Two of these main problems were related to ambient air pollution namely, dust- related diseases such as bronchitis, and lead poisoning particularly in Metropolitan Manila. Indoor air pollution was likewise considered. However, a recent study on indoor exposure to particulate matter, 10mm (PM10) and nitrogen oxides (NOx) in Metropolitan Manila showed that there is no significant difference in the pollutant levels between inside and outside the homes. Therefore this particular aspect was deemed to be a consequential issue of the overall air quality problem. The other problems cited in the review were the persistence of diarrhea and skin diseases due to inadequacy of water supply and sanitation based on an epidemiological study which examined association between the services available and the diseases, and pesticide-use related morbidity based on acute poisoning data from the Poison Control Center and extent of usage of pesticides. As seen from this review, a combination of communicable and non-communicable diseases is attributed in part to environmental pollution. The review concluded that ambient air pollution especially in the Metropolitan Manila area is one of the most important environmental health problems in the Philippines. This important environmental health problem and its relevance to the mitigation of greenhouse gases is the topic of this project. 1.2. Objective of the Project This project quantified and assessed the public health benefits of different mitigation measures with special focus on transport issues, common to both controlling ambient air pollution and greenhouse gases emissions. This project made use of health and economic impact as parameters in evaluating the benefits of the mitigation measures. 1.3. Organization of the Report This project report is organized as follows: Section 2 briefly describes the conceptual framework which is the basis of the health and economic benefits calculations. Section 3 focuses on the transport sector in Metro Manila and the Philippines and the scoping decisions with regards the policies to be considered. Section 4 discusses the methodologies for each component of the model namely the scenario development, air pollutant concentrations modelling, health effects estimation and economic valuation. Section 5 gives the results including some discussion and finally section 6 contains the conclusions and recommendations. The references and additional annexes are at the end of the report. 21 ------- 2 Conceptual Framework To estimate the co-benefits of policies with regards both greenhouse gases (GHG) and local air pollution reduction, the following framework was pursued in this project. As can be seen in figure 1, the analysis included policies which could include mitigation measures for both GHG and local air pollution. These policies are assessed as to their capability to reduce emissions of GHG and local air pollutants from both stationary and mobile sources. Scenarios are developed from these policies. From these assessments or scenarios, GHG emissions and local air pollutants reduction are estimated. Together with the development of scenarios from the policies, a business-as-usual scenario is also set. Ambient concentrations of the air pollutants are eventually modelled. Changes in ambient concentrations of the air pollutants are calculated by comparing the different policy scenarios with the business-as-usual scenario. Health benefits are then estimated based on the differences in ambient concentrations and concentration response functions that associate changes in ambient pollutant levels with specific health impact endpoints. Economic values of the avoided health impacts (or benefits) are likewise computed. Ideally, as part of policy analysis, the economic value of health benefits must be compared with the cost of implementing the specific policies to demonstrate the cost effectiveness of the policies. However, in this project, it was not possible to do so because of time constraint primarily and the lack of some data that is needed for such a comparison. In addition, other types of social costs, e.g. environmental and aesthetic impact of air pollution, are needed to warrant a more credible comparison. This research is suggested as a necessary part of future follow-on work in this area. Figure 1. Design and Analysis of Integrated Strategies 22 ------- 3 Focus on the Transport Sector 3.1 Background There has been more than a four fold increase in the number of road vehicles in the past two decades in the Philippines, from less than a million in the late 1980's to almost 4.2 million in 2003. In Metropolitan Manila alone, the number of vehicles increased from about 600,000 in the early 1990's to approximately 1.4 million in 2003. This is about 33% of the total for the whole country. The increase is more pronounced among the diesel fuelled vehicles. The number of diesel fuelled vehicles in the whole country increased from about 331,000 in 1987 to 1.3 million in 2002, a more than three fold increase. Meanwhile, the number of gas fuelled vehicles increased from approximately 800,000 to 2.3 million in the same time period, a little more than a two fold increase. Fuel consumption naturally also increased considerably. For regular gasoline, the increase was from 2,971 barrels in 1983 to 5,555 barrels in 2000, and for unleaded gasoline, from 1,047 barrels in 1994 to 10,973 barrels in 2000. With diesel fuel, the increase was even more staggering, from 18,879 barrels in 1983 to more than 42,000 barrels in year 2000. In the busiest road networks, traffic peaks at 11,000-12,000 vehicles per hour and daily volumes could exceed 140,000 -150,000 vehicles. In many sections of EDSA, for example, 2.34 million passengers travel daily, of which 1.43 million travel by bus which are mostly diesel fuelled. Road vehicles contribute substantially to particulate matter loads. Changes in air quality had not been adequately and reliably documented before 1986. Since 1986, TSP pollution had shown no overall trend but the 24-hour average has consistently remained almost double the WHO air quality guideline (WHO 24-hour standard for TSP - 120 ug/m3). In the recent past five years, the ambient monitoring data in Metropolitan Manila revealed that PM10 had consistently been measured to be above the air quality guideline (PM10 National Annual Standard - 60 ug/m3). Personal monitoring measurements done in conjunction with an epidemiological study of jeepney and bus drivers and commuters in 1991 and schoolchildren in 1993 showed that levels of total suspended particles, sulfur dioxide, carbon monoxide, total oxidants and lead were way above the WHO air quality guidelines. 23 ------- 3.2 Scoping Decisions and Policies Considered Realizing that more than 90% of the air pollution comes from mobile sources based on the 2003 emission inventory the project decided to concentrate primarily on the transportation sector particularly in Metropolitan Manila. Metropolitan Manila, as described before, has more than one-third of all the vehicles in the country. It also has the highest density of population in the country and probably the highest levels of air pollution as well. Based on a scoping meeting1 held early on in the project, several mitigation measures were considered for the scenario development. For this project's purpose, these mitigation measures are called policy scenarios. Eight individual policies and three combinations of policies are assessed. Although, other mitigation measures were suggested during the scoping meeting, due to either political reasons or data availability, these were not considered in this assessment. The base year used is 2002 and projections to years 2005, 2010 and 2015 are made. The following are the policies considered and for which scenarios were developed. Full descriptions of the policies and the combination of policies can be seen in the scenario development section of the methodology. Transportation Demand Management through license plate scheme (TDM) Construction of Rail-based Mass Transit System Construction of Bikeways Implementation of the Motor Vehicle Inspection System (MVIS) Introduction of the Compressed Natural Gas buses (CNG) Introduction of Cocodiesel for diesel-fuelled vehicles particularly jeepneys (CME) Two stroke tricycles switching to four-stroke engines. Improvement of vehicles by the Use of Diesel Traps Combo 1 - combination of policies: all policies except railways and switching of two stroke to four stroke tricycles Combo 2 - all policies except railways Combo 3 - all policies including railways 1 Held 13-14 February 2004, the Scoping Meeting brought together stakeholders from the government, civil society, academe and private sectors. Specifically, the meeting provided a venue for the development of objectives, identification of alternative scenarios, forging agreements on geographic scope, timeframe and time steps for IES analysis, and discussion of technical issues such as on models and methodologies. ------- 4 Methodology 4.1 Scenario Development The study simulated scenarios of future transport and energy policy measures in Metro Manila and assessed their impact both in terms of mitigating ambient air pollution and emissions of greenhouse gases. The scenarios are framed around 2005, 2010, and 2015 future years in order to provide short, medium and long-term perspectives. Simulation results for the different scenarios were evaluated in terms of the total mass of emissions of PM10 as an indicator of ambient air pollution and of C02 as a greenhouse gas or indicator of global pollution. In Section 5.5, the study assessed the health and economic impact of air quality improvement as the ultimate benefits of these mitigation policy scenarios. This section summarizes the framework for developing and analyzing the scenarios, including the (a) criteria for selection of policy measures, (b) methodology for analysis of travel demand and emissions and (c) description of the policy scenarios and assumptions for estimating their impacts. 4.1.1 Criteria for Selection of Policy Measures Consistent with the IES objective to build awareness, support and in-country capacity for analysis and quantification of the local and global benefits of integrated energy and environmental policies and thereby promote their adoption and implementation, the measures to be analyzed were screened based on the following criteria: (a) Feasibility of implementation - measures must be consistent with the objectives, plans and programs of government agencies and non- governmental organizations that are involved in the implementation of air pollution and GHG mitigation programs. (b) Socio-economic and political acceptability - the social and economic costs as well as the political risks of implementing the measures must be identifiable although they may not necessarily be imputed into the models for air quality and health impact analysis. (c) Availability of information the measures must lend themselves to scenario modeling and quantitative analysis of their impact on reducing air pollution and GHG emissions. 25 ------- 4.1.2 Analytical Framework Policy measures for land transportation and energy use that are designed to improve air quality may achieve either one or more of the following objectives: (a) Reduce vehicle-kilometers- these are policies that can indirectly influence modal shift (i.e. increase switching to public transport) or even reduce travel demand. (b) Reduce fuel used per vehicle kilometer - these are policies that can improve the efficiency of vehicles. An example is the inspection and maintenance of vehicles. (c) Reduce emissions per unit of fuel used - these are policies that can clean the fuel inputs or exhaust coming from the vehicles. Examples are engine upgrades, use of alternative fuels, and emissions control technologies. Given the foregoing objectives, the analysis involves quantifying the impact of the policy measures on travel demand and vehicle emission per unit of fuel used. The methodology used for the analysis of the impact of transport- and fuel-related measures on emissions is similar to the environmental analysis model developed by the Metro Manila Urban Transportation Integration Study (MMUTIS, 1999). In this study, the total emissions for various policies were calculated as: Emissions = f (travel distance, travel speed, emission factors) The following sections describe the methodology for estimating the travel demand and vehicle emissions. 4.1.3. Estimation of Travel Demands Travel distance in terms of vehicle-kilometers and travel speed in terms of kilometers per hour by planning zone were estimated using the 4-step travel demand forecasting model discussed in Sec. 5.1.3 (b). The MMUTIS Study defined 171 planning zones where 94 zones are in Metro Manila and the rest are in the nearby provinces. Vehicle-specific and speed range-specific emission factors for PM are used to estimate the total emission for each policy. The emission factors were derived from earlier studies such as the MMUTIS (1999), ADB VECP (1992) and JSPS (2002). The share of travel distance of jeepneys and buses, and the share of gasoline and diesel-fed vehicles by mode were also estimated. 4.1.3.1. Study Area and Zoning System Although the study area is limited to Metro Manila, the modeling for the transportation demand covered Metro Manila and towns/cities of the adjoining provinces of Bulacan, Rizal, Cavite and Laguna (see Figure 2). The 1996 Metro Manila Urban Transportation Integration Study or MMUTIS 26 ------- _ PANDI j ¦A LAG T AS ) IAGONOV fMALOLOS J ^ PAOM0ONG * f' ' JHJL MCY.CAUArAN' OBAMDO^^SL- 5AN MATEO AKT1POLO NAVOTA! BAflAS iNGONO. MORON_G p^A^etWy j /Ef^TAGUIG' / tf>?C v J \ (JARANAOUE CAVITEGf .RUONA HOVEEET, /< ROSARI0 UAOONAOeeAYh GENERAL TfllAS , LA-JALA x > A DASMAHIMAS -¦ r - r! J k \-i- TflECE MARTiRES GSTY \ SILANG Figure 2. Towns and Cities of the Study Area (Japan International Cooperation Agency, 1999) established 265 traffic analysis zones for the 17 cities/towns of Metro Manila and 51 zones for the adjoining towns and cities. For the purposes of the IES study, these zones were combined to form 98 traffic analysis zones wherein 94 traffic analysis zones were constructed for Metro Manila and four (4) other zones corresponding to the four (4) adjacent provinces, as shown in the MMUTIS digital map in Figure 3. 27 ------- 4.1.3.2. Socio-Economic Characteristics Figure 3. Traffic Analysis Zone (TAZ) Transportation demand modeling for the present and the future years requires a database of socio-economic characteristics aggregated to the traffic analysis zones. The following socio-economic characteristics were considered in the modeling: population employment by residence employment by workplace school by residence school by school car ownership 28 ------- The above data exists for the 265 zones that were derived from the 1996 household interview survey (HIS) of the MMUTIS database. Data were subsequently aggregated for the 98 zones under the IES Study. The socio- economic data is also available in the years 2000 and 2010 for the 98 zones. The zonal socio-economic characteristics were calculated for the years 2002, 2010 and 2015 using a growth rate based on the data in 2000 and 2010 and applying the rate on the 1996 data from the HIS of MMUTIS database. The zonal car ownership was estimated for 2002 using the growth rate based on the 1996 and 2005 data. The zonal car ownership for 2010 and 2015 were estimated using the growth rate based on the 2005 and 2015 data. The future socio-economic, land use and transport network development (until 2015) scenario as stated in the MMUTIS Final Report is shown briefly in the "Assumptions on the Future Socio-Economic and Urban Land Use/Transport Network Development" in the Annex of this Report. 4.1.3.3. Transportation Network Characteristics Transportation demand modeling for the present and the future years also requires data on the transportation network that consists of roads and public transport lines. The public transport network mainly consists of railway lines and buses and jeepney routes. The study utilized the MMUTIS transportation network built in 1996 and updated to include the following roads and rail transit lines in 2005: a) Road Network Primary and Secondary Road Network Expressway Network North Luzon Expressway South Luzon Expressway Manila-Cavite Expressway Metro Manila Skyway (Makati-Bicutan) b) Rail-Based Mass Transit Lines LRT Line 1 (Monumento - Baclaran) MRT Line 3 (Monumento - Taft Avenue) LRT Line 2 (Santolan - Recto) 29 ------- The network consisting nodes and links is encoded in a digital map in JICA STRADA format, as displayed in Figure 4, This is the transportation network for the baseline scenario in 2005. Figure 4: Road and Railway Network for the Baseline Scenario (2005) 30 ------- 4.1.3.4. The Four-Step Model Base-Year Data Zone Networks Future Planning Data t Database TRIP GENERATION a © ¦N * b TRIP DISTRIBUTION MODAL SPLIT TRIP ASSIGNMENT Evaluation 3 o How many person trips? Where are they going? a, What mode are they using? What route will they take? Source: Teodoro (2003), Lecture 2:Transportation Infrastructure and Planning Concepts,UP-NCTS. Figure 5. Transportation Demand Analysis: The Classical Four Step Model The 4-Step JICA STRADA (JICA System for Travel Demand Analysis) Model utilized in the MMUTIS Study was used to estimate the transportation demand for the baseline scenarios in 2005, 2010 and 2015, and the railway scenario in 2015. The procedure for demand forecasting used by MMUTIS was adopted by this study (Figure 5). The zoning system, socio-economic characteristics and transportation network which came from the MMUTIS database, served as inputs to the transportation demand modeling procedure. Following is a description of the different parts of the 4-Step Model. Step 1: Trip Generation/Attraction The trip generation/attraction step calculates the number of person trips generated from and attracted to each traffic analysis zone. Model functions for trip generation and attraction are estimated with zonal socio-economic attributes calculated in Section 2 as explanatory variables. The trip generation model for Metro Manila is correlated with the variable "employment by workplace". The number of person trips generated from and attracted to each of the 98 zones by private and public modes is estimated. Step 2: Trip Distribution The trip distribution step estimates the number of person trips originating from a traffic analysis zone (origin zone) and ending in another traffic analysis zone 31 ------- (destination zone). Trips generated from and attracted to each zone are distributed among the zones generating a 98 by 98 origin-destination (O-D) matrix of person trips. In this step, two matrices (98 by 98) are generated consisting of private and public person trips. Step 3: Modal Split The modal split or mode choice step estimates how many of the person trips for each pair of origin and destination (O-D) zones will use private or public transport modes. However, the procedure for demand forecasting in this study has already segregated the public and private trips earlier in the trip generation step. Step 4: Traffic Assignment The traffic or trip assignment step identifies the exact routes that will be taken by each of the person trips. It involves assigning traffic to a road network or a transit network. The road and transit network of Metro Manila for the baseline scenario in 2005 is shown earlier in Figure 4. Traffic is assigned to available transit or roadway routes using a mathematical algorithm that determines the amount of traffic as a function of time, volume, capacity, or impedance factor. The three common methods are all-or-nothing, diversion and capacity restraint. The highway-type assignment for private and public modes is adopted as the model. The JICA STRADA outputs of the traffic assignment are the following: link vehicle traffic volumes per day, link average traffic speed per day, link volume-to-capacity ratios per day and link trip lengths per day. The traffic volumes were obtained by dividing the number of person trips assigned to each route by the average vehicle occupancy. The vehicle occupancy data were obtained from the MMUTIS. 4.1.3.5. Additional Procedures for Estimation of Transportation Demand a) Estimation of Transportation Demand for Tricycles The transportation demand for the local three-wheelers (a public transport mode in residential areas) called the "tricycles" was estimated separately in 2005, 2010 and 2015 and then the demand was added to the transportation demand which was earlier estimated for private and public trips. b) Application Program to Aggregate Transportation Demand From Links to Zone The main outputs of the transportation demand analysis software are daily link-level traffic flows and average traffic speeds. An application program was developed to aggregate the outputs from links to the traffic analysis zones. The program first identified which traffic analysis zone a specific link belongs. After identifying zone membership, aggregation was done for traffic volumes 32 ------- and average traffic speeds to produce zone-level traffic volumes and average speeds per day. 4.1.3.6. Calculation of Vehicle-Kilometers The daily vehicle-kilometers for the zones were calculated by multiplying traffic volumes by the length of the road links. The calculations yield the private and public vehicle-kilometers per day for the 98 zones of the study area. It is necessary to further classify the vehicle-kilometers in terms of vehicle types as input to the calculation of vehicle emissions. Table 1. Share of Present Travel Distance of Buses and Jeepneys Vehicle Trips Average Veh-Km Share of Trip Length Veh-Km Bus 57,000 13.0 741,000 31.5% Jeepney 460,000 3.5 1,610,000 68.5% Source: Japan International Cooperation Agency (1999) MMUTIS Technical Report No. 10; Traffic Environmental Study, Air and Noise Pollution in Metro Manila The composition of public transport vehicles was assumed to be uniform in all zones and the estimates were based on the share of the vehicle kilometers of each type, as shown in Table 1. The share in percent was multiplied by the total vehicle-kilometers of public trips to get the total vehicle-kilometers of each public transport vehicle type for each zone. Since the tricycle demand was estimated separately, it was not necessary to get its share of vehicle- kilometers of the public transport trips. Table 2. Share of Private Modes Cars (Gas) Utility Vehicles (Gas) Utility Vehicles (Diesel) Trucks (Diesel) 41.9% 21.7% 30.4% 6.0% Source: Land Transportation Office (LTO), Philippines For private vehicle trips, the percentages coming from the vehicle registration data of the Land Transportation Office of the Philippines in 2001 (Table 2) were multiplied by the private vehicle kilometers to get the share of vehicle- kilometers of each private vehicle type per zone. The share of vehicle- kilometers of road-based transport modes is assumed to be the same for 2005, 2010 and 2015. 33 ------- 4.1.4.Estimation of Vehicle Emissions (a) PM Emission The daily PM emissions for each traffic analysis zone were calculated as follows (Equation 1): 6 6 PM = x EF'exhaust. (v)+^Tsxd,x EFidlei (1) i=1 ;=1 where: PM = PM emissions per traffic analysis zone (g) di = travel distance of vehicle type /' (veh-km) per zone v = average travel speed per zone (km/h) EFexhaustj (v) = exhaust emission factor of vehicle type /' as a function of travel speed (g/veh-km) Ts = idle or stopping time (min/veh-km) per zone EFidlei = idle emission factor of vehicle type /' (g/min) The stopping or idle time per zone was obtained from the "Two-Fluid Model" developed for Metro Manila by MMUTIS (JICA, 1999). Using the output of travel demand estimation, which is the travel time (min/km) in each zone, the stopping time is calculated using the equation of the Two-Fluid Model as shown below (Equation 2): 1 fl T = T T w+i T w+i /q\ s m \£-) where: Ts = stopping time per unit distance (min/km) T = trip time per unit distance (min/km) 7~m = average minimum trip time per unit distance = 1.966 min/km for Metro Manila n = 1.889 for Metro Manila (b) C02 Emission The emission of C02 can be calculated based on the same principle, that is, emission of a vehicle type using a particular fuel is a function of distance traveled, the average travel speed per zone and the emission factor. The calculation can be done for the exhaust both at running and idle time, provided data on the applicable emission factor is available. Since there were no data available on emission factors for C02 per speed range and at idle time, the calculation of C02 emission in the study was done simply by multiplying the distance traveled by the emission factor. 34 ------- Table 3 shows the PM (particulate matter) emission factors for the six (6) vehicle-fuel types according to speed range. The study adopted the locally developed emission factors of the 1992 VECP Project (ADB, 1992) as the emission factors at 20 kph speed and patterned the variation after the speed- specific and vehicle-specific emission factors provided by the MMUTIS Technical Report No. 10 (JICA, 1999). Table 3. PM Emission Factors* Fuel Type Vehicle Type Idling -10 km/h 10 km/h 20 km/h- -20 km/h Gasoline Car 0.15 0.12 0.1 0.1 Jeepney 0.17 0.14 0.13 0.12 Tricycle (2-stroke) 2.05 2.01 2.00 2.02 Diesel Car 1.73 2.03 0.9 0.9 Jeepney 1.59 1.89 0.99 0.9 Bus 1.6 2.4 1.6 0.9 Unit g/min. g/km. g/km. g/km. 'Derived from the 1992 ADB VECP Project and 1996 MMUTIS Emission factors One of the outputs of the 4-step travel demand forecasting model estimated in JICA STRADA is the average speed of each zone. When the average speed is known, the emission factor for each vehicle-fuel type is taken from Table 3 as a function of speed. 4.1.4.2 Emission Factors for C02 Table 4 shows the C02 emission factors used for the different vehicle types analyzed. These emission factors were derived from an ADB study. Table 4. C02 Emission Factors (g/km) Fuel Type Vehicle Type Emission Factor Gasoline Cars 399 Utility/Jeepney 456 Motorcycles/T ricycles 186 Diesel Car 537 Utility/Jeepney 559 Buses 1249 Source: ADB, 2002 35 ------- 4.1.5. Description of Scenarios: Policy Elements and Assumptions The following paragraphs discuss the policy measures simulated for each scenario and the assumptions made in the estimation of transportation demand and vehicle emission for each scenario. The discussion includes a brief overview of government and private sector initiatives with respect to the different policies and where data is available, the technological and economic considerations. Table 5. Summary of Scenarios and Corresponding Assumptions Scenario Policy and Assumptions Baseline or BAD 2005: 2005 transportation demand + 2005 transport Business-As-Usual (BAD) network +I/M Standards BAD 2010: 2010 transportation demand + 2005 transport network + l/M Standards BAD 2015: 2015 transportation demand + 2005 transport network +primary and secondary road network in 2015 + l/M Standards Reduction of PM emission factors and the corresponding percentages of vehicle types with reduced emission factors MVIS2005: + STDS2 - l/M Implementation of the STDS2 scenario without tie l/M scenario reduction in PM emission factor by 60% percent of vehicles: cars=25%, jeepneys=100%, buses=30%, trucks=30% MVIS2010: + STDS2 Implementation of the STDS2 scenario on top of the l/M Scenario Reduction of PM emission factor by 60% after the 30% reduction of emission factor under the l/M scenario percent of vehicles: cars=25%, jeepneys=100%, buses=30%, trucks=30% MVIS2015: + STDS3 Implementation of the STDS3 scenario on top of the l/M Scenario Reduction of PM emission factor by 60% after the 30% reduction of emission factor under the l/M scenario percent of vehicles: cars=50%, jeepneys=100%, buses=100%CNG, trucks=40% Implementation of the Motor Vehicle Inspection System (MVIS) 36 ------- Scenario Policy and Assumptions Transportation Demand Vehicle-kilometers of private transport modes such as gas car, Management (TDM) gas jeepney/utility vehicle and diesel car/utility vehicle were reduced by 11.08% in all 98 traffic analysis zones Replacement of 2-Stroke The PM emission factor of tricycles was reduced to 1/5 of the with 4-Stroke Motorcycles emission factor of tricycles in the baseline scenario applied to for Tricycles (4STC) 100% of the tricycles in all zones Construction of Bikeways The rates of shift (1.5% in 2005 and 3.5% in 2015) from tricycle (BWMK and BWMM) to cycling modes were applied as reduction rates of the tricycle vehicle-kilometers of traffic analysis zones Marikina (BWMK): applied to zones 74 and 76 only Metro Manila (BWMM): applied to all 98 zones Expansion of the Expansion of the metropolitan railway network by 2015 by Metropolitan Railway approximately 164.1 kilometers of new MRT/LRT lines and 19.7 Network by 2015 kilometers of busways according to the MMUTIS Master Plan (Rail 2015) resulting to reduced road-based traffic demand Diesel Particulate Trap (DPT) Installation of the diesel particulate trap is expected to reduce for Buses and Jeepneys the PM emission factor of buses and jeepneys by 30% (DPTBJ and DPTB) DPTB: reduction of PM emission factor of buses only Compressed Natural Gas Reduction of emission factor of buses by 86% if diesel is (CNG) for Buses (CNGB) replaced by CNG 2005 (Low: 0.88%/High: 1.76% applied to zones passed by C-5. EDSA and SLEX) 2010 (Low: 11.47%/High: 22.93% applied to zones passed by C-5. EDSA, SLEX and NLEX) Coco-methyl ester (CME) Reduction of emission factor of jeepneys by 86% if diesel is for Jeepneys (CMEJ) blended with CME 2005 (Low: 0.64%/High: 1.27% applied to all zones) 2010 (Low: 2.0%/High: 4.0% applied to all zones) Combo 1 Combination of all scenarios except railways and switching of two stroke to four stroke tricycles (2005) Combo 2 Combination of all scenarios except railways (2010) Combo 3 Combination of all scenarios (2015) 37 ------- 4.1.5.1 "BAH" Scenario - Baseline or Business-as-usual a) Transportation Network Table 6. Baseline Scenarios for 2005, 2010 and 2015 Baseline or Transportation Transportation Business-as-usual Scenarios Demand Forecast Network and Measures BAU 2005 2005 transportation demand 2005 transport network BAU 2010 2010 transportation demand 2005 transport network BAU 2015 2015 transportation demand 2005 transport network + 2015 primary and secondary roads The BAU scenarios for 2005 and 2010 are based on the transportation network, that is, the roads and public transport lines, as outlined in Section 4.1.3.3, which is due for completion in 2005. The 2015 baseline scenario assumes that the primary and secondary road networks have expanded according the MMUTIS Master Plan for roads in 2015. b) Motor Vehicle Emission Standards Table 7. Reduction in Vehicle Emission Factors - "l/M" Standards (l/M) CO, HC - 40% reduction NOx, SPM - 30% reduction Percentage of Vehicles Meeting the "l/M Standards" Cars Jeepneys Buses Trucks 100% 100% 100% 100% Source: Asian Development Bank (1992) Vehicular Emission Control Planning (VECP) Project The BAU scenarios in 2005, 2010 and 2015 assume that the motor vehicle emission standards have already been initially implemented as reflected by the l/M Scenario of 2005 (Table 6) in the Vehicular Emission Control Planning (VECP) Project Report (Asian Development Bank, 1992). The PM emission factors in Table 7 were reduced by 30% for all vehicles. The implementation of the motor vehicle exhaust emission standards in the Clean Air Act started in January 2003. In Table 7, the category "Cars" include gasoline-fed cars, jeepneys/utility vehicles, diesel-fed cars/utility vehicles while the category "Jeepneys" means diesel-fed jeepneys, the category "Buses" means diesel-fed buses, and the category "Trucks" means diesel-fed trucks. 38 ------- 4.1.5.2. "MVIS" Scenario - Implementation of the Motor Vehicle Inspection System The implementation of the MVIS would involve stricter l&M (inspection and maintenance) scheme on motor vehicles will reduce the emission levels from the source significantly, since vehicles will be forced to be maintained properly to comply with emission standards of the Clean Air Act of 1999. This measure will follow the set of standards and compliance rates in STDS3 (Table 9) and the reduced vehicle emission factors and percentages of compliance will be inputs to the calculation of total emissions. Table 8. Reduction in Vehicle Emission Factors - 'Substantial Reduction" (STOS2)* CO, HC - 75% reduction NOx, SPM - 60% reduction Percentage of Vehicles Meeting the "Stringent Standards" Type Cars Jeepneys Buses T rucks Share 25% 100% 30% 30% * Source: Asian Development Bank (1992) Vehicular Emission Control Planning (VECP) Project Table 9. Reduction in Vehicle Emission Factors - 'Substantial Reduction+CNG Buses" (STDS3)* CO, HC - 75% reduction NOx, SPM - 60% reduction CNG Buses Percentage of Vehicles Meeting the "Stringent Standards" Type Cars Jeepneys Buses T rucks Share 50% 100% 100% CNG 40% * Source: Asian Development Bank (1992) Vehicular Emission Control Planning (VECP) Project The MVIS scenario assumes that the MVIS facilities provided for in the Implementing Rules and Regulation of the Clean Air Act of 1999 are established in Metro Manila. With the MVIS, it was assumed that certain percentages of the vehicle fleet are expected to comply with the emission standards. The corresponding reduction in the PM emission factors as applied to the shares of the vehicle fleet (Table 8 and Table 9) is assumed. The scenarios for the MVIS in 2005, 2010 and 2015 are discussed as MVIS2005, MVIS2010 and MVIS2015 below. 39 ------- MVIS2005: STDS2 - l/M In 2005, it was assumed that the STDS2 scenario in the VECP Project Report will be implemented but without the l/M scenario. This means a reduction in PM emission factor by 60% applying the percentages of compliance in the STDS2 scenario (Table 7) MVIS2010: l/M + STDS2 In 2010, the 60% reduction in emission factor under the STDS2 scenario was further applied after the 30% reduction of emission factor under the l/M scenario (Table 8) MVIS2015: STDS3 + l/M In 2015, the 60% reduction in emission factor under the STDS3 scenario (Table 9) was further applied after the 30% reduction of emission factor under the l/M scenario. The percentage of the vehicle fleet complying with such standards is higher than the STDS2 and all buses are expected to be running on compressed natural gas (CNG) fuel. 4.1.5.3. "TDM" Scenario - Transportation Demand Management Table 10. Calculation of Reduction of Traffic Due to TDM Alternative to Car Percentage of Car Users Percentage of Total Affected by TDM Scheme Use Public Transport 38.2% 8.74% Share A Ride 10.2% 2.34% Total 11.08% Source: Japan International Cooperation Agency (1999), MMUTIS: Metro Manila Urban Transportation Integration Study Final Report According to the MMUTIS (JICA, 1999), 22.9% of private car users were affected by the UWRP or Color Coding Scheme, the TDM scheme implemented in Metro Manila. Only those who used public transport and those who shared a ride as alternative modes of transport to private car were considered to affect the private vehicle traffic. Table 10 above shows the estimated reduction in traffic of private cars due to the TDM scheme. The reduction of 11.08% was applied to vehicle-kilometers of private transport modes such as gas car, gas jeepney/utility vehicle and diesel car/utility vehicle in all 98 traffic analysis zones. 40 ------- 4.1.5.4. "4STC" Scenario - Replacement of 2-Stroke with 4-Stroke Motorcycles for Tricycles With the support of the mayors of the different cities and municipalities in Metro Manila, the Metro Manila Development Authority (MMDA) has issued a circular prohibiting new franchises for tricycles using 2-stroke engines in Metro Manila. With this circular as well as other policy trends, it is expected that 4-stroke tricycles will increase its share in the current 120,000 registered tricycles in Metro Manila. Motorcycle manufacturers, however, claim that it is not actually the motorcycles which emit significant pollutants but the type of engine oil used (Philippine Star, 2003). Accordingly, motorcycles that use the so-called street oil commonly called takal produce the pollutant carbon monoxide and hydrocarbon (unburned gasoline/oil). It was argued that the processed two- stroke engine oil called "low smoke" produces lower harmful emissions and that the Clean Air Act does not ban or call for a phaseout of the 2-stroke technology as long as it meets emission standards. It was further argued that government should thus crack down on illegal oil refiners/manufacturers rather on the motorcycle manufacturers. Moreover, phasing out two-stroke motorcycles would affect some 12 million people who rely on income from their tricycles.2 With due regard to the political difficulty of implementing this policy, the 4STC scenario examine the impact of this policy assuming that all tricycles will be required to shift from using 2-stroke motorcycles to 4-stroke motorcycles. About 95-98% of tricycles in Metro Manila are claimed to be using 2-stroke motorcycles. Based on the study of Shah and Harshadeep (2001), the PMi0 emission factor of 4-stroke motorcycles is approximately 1/5 of the emission factor of 2-stroke tricycles and this ratio was applied to all the zones. 4.1.5.5. "BWiK" Scenario - Construction of Bikeways in Marikina The Government of Marikina City, one of the seven cities in Metro Manila, has started to construct exclusive bikeways in 2001. The master plan of the city will involve construction of bikeways on 66 kilometers of roads in the city. With this, it is assumed that there will be a significant increase in the use of bicycles for work trips. The BWMK scenario simulates the changes in modal share of bicycles shift from non-cycling modes in Marikina City for 2004 and 2014 as shown below in Table 11. Only the internal trips (trips within the city) were considered in the analysis. The shift in 2004 is assumed to take effect in 2005 while the shift in 2014 is assumed to take effect in 2015. 2 During the presentation of IES results at ADB, Secretary Leandro Mendoza also argued that the phaseout of 2-stroke motorcycles for tricycles is not politically feasible. ------- Table 11. Modal Share of Bicycles in Marikina City Year Internal Trips (Trips within the City) Percentage Shift Pedicab Pedicab Bicycle Walking Motorcycle T ri cycle 2004 1.00% 1.00% 1.85% 0.85% 0.50% 1.50% 2014 2.50% 2.50% 3.50% 3.0% 0.80% 3.50% Source: U.P. National Center for Transportation Studies Foundation, Inc. (2000), Marikina Bikeways Feasibility Study Final Report The rates of shift (1.5% in 2005 and 3.5% in 2015) from tricycles to cycling modes were applied as reduction rates of the tricycle vehicle-kilometers of traffic analysis zones 74 and 76, representing Marikina City. 4.1.5.6. "BWMNP Scenario - Construction of Bflceways in Metro Manila This scenario was simulated assuming that similar bikeways will be constructed in other parts of Metro Manila. The same rates of modal shift for Marikina were applied to reduce the tricycle vehicle-kilometers in all the 98 zones of Metro Manila. 4.1.5.7. "Rail 2015" Scenario - Expansion of the Metropolitan Railway Network by 2015 In Metro Manila, there are already 45.3 kilometers of LRT/MRT lines (Line 1, Line 2 and Line 3) as of April 2003 and there also exists a heavy rail line, which is approximately 30 kilometers (Philippine National Railways (PNR) Commuter Line), as shown in Table 12 below. In the 1996 Metro Manila Urban Transportation Integration Study (JICA, 1999), there is a master plan for the expansion of the railway network through the extension of existing LRT/MRT lines, construction of new LRT/MRT lines and busways, and upgrade of the PNR lines. By 2015, it is planned that there will be approximately 164.1 kilometers of new MRT/LRT lines and 19.7 kilometers of busways. With the expansion of the railway network, it is expected that more people will shift their transport mode from private vehicles, buses and jeepneys to rail due to a more convenient railway service. The model used for demand shift from private cars to public transit was based on the MMUTIS. The model calculated the probability of shifting as a function of difference in travel time (in minutes) and travel cost (in pesos) of the public mode and private mode of transport. This is expected to reduce the vehicle-kilometers of travel of private cars and road-based public transport. 42 ------- Table 12. Proposed Expansion of the Railway Network (by 2015) Railway Line Route Distance (km) LRT Line 6 South Extension of Line 1 (Baclaran-lmus) 30.0 MRT Line 3 Extension Extension from North Avenue (North Avenue-Navotas) South Extension (Taft Avenue-Reclamation) 10.0 2.0 MRT Line 2 (East, West Extension) West Extension (Recto-North Harbor) - MRT East Extension (Santolan-Masinag) - MRT East Extension (Masinag-Antipolo) - Busway 4.0 4.0 7.7 MRT Line 2 (South-East Extension) Fort Bonifacio-Taytay - MRT Taytay-Binangonan - Busway 19.8 12.0 MRT Line 4 Recto-Novaliches (via Quezon Avenue and Commonwealth Avenue) 22.8 North Rail Sta. Mesa-Meycauayan, Bulacan (using PNR line) 26.0 MCX Sta. Mesa-Sta. Rosa, Laguna (using PNR line) 45.5 Source: Japan International Cooperation Agency (1999), Metro Manila Urban Transportation Integration Study Final Report (MMUTIS) 4.1.5.8. "DPTBJ" Scenario - Diesel Particulate Trap (DPT) for Buses and Jeepneys The current technology of DPT can reduce particulate matter by 30% from diesel exhaust based on the experience in Hong Kong (Eco-Tek Holdings Limited, 2002). The scenario assumes that the installation of the diesel particulate trap will result in the reduction of PM emission factor of buses and jeepneys by 30% and applies such reduction rate to the baseline scenario of transportation demand in 2005. Diesel particulate trap oxidizers or diesel particulate filters seem to have no adverse effects on engine or vehicle performance (http://www.cleanairnet.org/caiasia/1412/article-37162.html). 4.1.5.9 "DPTB" Scenario - Diesel Particulate Trap (DPT) for Buses Another scenario was simulated for 2005 assuming the application of DPT for buses only. The same rate of reduction of PM emission factors of buses was assumed. 4.1.5.10 "CNGB" Scenario - Compressed Natural Gas (CNG) for Buses Many countries all over the world are promoting the use of cleaner and alternative fuels for transport to control urban air pollution while promoting 43 ------- energy diversity and reliability. In the Philippines, the road transport sector is almost entirely dependent on diesel and gasoline. Although there had been efforts to develop and promote the use of alternative fuels since the early 1980s, the interest has not been sustained due to fluctuations in oil prices. The passage of the Clean Air Act that sets emission limits for motor vehicles prompted the search for innovative technologies that will allow motor vehicles to meet the standards. The Act prescribes the maximum allowable emissions of CO, HC, NOx, SOx, and PM. Beginning 2003, compliance with emission standards has become a requirement for vehicle registration. Table 13. DOE Alternative Transport Fuel Program - Cumulative Inventory of Alternative Fuel Vehicles 2003 2004 2005 2006 2007 2008 2013 CNG 30 60 90 120 160 200 1500 LPG 70 100 200 300 400 500 1500 CME 800 1600 2500 3300 4100 5000 10000 Total 900 1760 2790 3720 4660 5700 13000 Source: DOE - unofficial targets as of July 2003 The Department of Energy (DOE) has launched an Alternative Transport Fuels program that focuses mainly on the promotion of alternative fuels that have been established to be technically feasible, namely, compressed natural gas (CNG), liquefied petroleum gas (LPG) and coco-methyl ester (CME). Table 13 shows the tentative targets of the DOE in the Philippine Energy Plan 2004-2003 for the cumulative inventory of vehicles running on these alternative fuels. The first initiative of the DOE to promote the use of natural gas for transport in cooperation with the Philippine National Oil Company (PNOC) involved the conversion of a diesel-engine Isuzu Hi-Lander vehicle to a 100 percent natural gas engine unit. This was followed by the conversion of three more vehicles, i.e., two buses and one Nissan Patrol utility vehicle. Jointly with PETRONAS Malaysia, PNOC is currently demonstrating six ENVIRO 2002 taxi units to promote factory-built or original equipment manufactured (OEM) natural gas vehicles or "NGVs". In March 2002, the first CNG refueling station was inaugurated at the San Antonio Gas Plant of PNOC Exploration Corporation (PNOC-EC) in Isabela province. On October 16, 2002, President Gloria Macapagal-Arroyo launched the Natural Gas Vehicle Program for Public Transport (NGVPPT), which involves a package of incentives to encourage public transport companies to use CNG as an alternative fuel. Such incentives include financial packages, preferential franchise for NGVs to newly opened routes, automatic issuance of Certificate of Compliance with Emission Standards for NGVs and fast-track issuance of Environmental Compliance Certificate (ECC) for NGV facilities and refueling 44 ------- stations. As the lead agency for the program, the DOE is working with various government agencies to operationalize these incentives. The DOE is also pursuing with the private sector various initiatives to promote the use of CNG. These initiatives include the conduct of a joint feasibility study with Shell Philippines Exploration B.V. (SPEX) and Pilipinas Shell Petroleum Corporation for a pilot project on the establishment of a "mother- daughter system" CNG refueling station, that is, a mother station in Batangas and a daughter station either in Pasay or Paranaque City, which is expected to be operational by 2004. For the long-term, the DOE is promoting the construction of CNG refilling stations in Metro Manila along a 40-km EDSA and Taft Avenue loop to be hooked up to the proposed 80-100 km natural gas pipeline from Batangas to Manila. This study analyzed scenarios of penetration of CNG-fueled vehicles based on the DOE's unofficial targets as of July 2003 as shown in Table 13 above. It was assumed that 90 CNG-fueled buses would be running on the major thoroughfares (EDSA, C-5 and South Luzon Expressway or SLEX) in 2005. The target of 1500 buses for 2013 was used for 2010. By this time, the buses are also assumed to reach the North Luzon Expressway. These targets will translate to only .88 % and 11.47% of the projected diesel bus fleet by 2005 and 2010, respectively. The study simulated a "high" scenario" by doubling the amount of these targets. The shares improved to 1.76% in 2005 and 22.93 in 2010 - still insignificant but it may be realistic considering that this is a new technology and uptake could be slow. Table 14. Estimate of Share of CNG Buses in 2005 and 2010 Percentage of Bus Fleet 2005 (C-5, EDSA, SLEX) 2010 (C-5, EDSA, SLEX, NLEX) Low Estimate 0.88% 11.47% High Estimate 1.76% 22.93% From a survey of literature, the emission factor for particulate matter would be reduced by 86% with the shift from diesel-fueled to CNG-fueled buses (Goyal and Sidhartha, 2003) while the C02 emission factor will be reduced by 45- 70% (ADB, 2002) 4.1.5.11 "CME J" Scenario - Coco-methyl ester (CME) for Jeepneys Efforts to develop CME as a diesel substitute started in the 1970s when the then ailing coconut industry required new market (SATMP, 2000). The project was shelved after the industry showed signs of recovery and world prices of 45 ------- petroleum plunged. Interest in CME was revived recently as the price of petroleum fluctuates while the price of copra remains depressed. In contrast with other alternative fuels like CNG and LPG, using CME does not require modifications on the engine, tanks, tubings, etc. CME is a chemical compound derived from the reaction of coconut oil and methanol. The fuel has been subjected to various technical evaluations by different government agencies. A significant finding is that although CME has a slightly lower heating value than diesel, it is no less efficient than diesel (SATMP, 2000). Moreover, CME burns more completely than diesel, thus producing less smoke in the engine's exhaust. It also has strong detergent properties that tend to eliminate grease and residues in the oil tanks. Yet the main advantage of CME over diesel is the absence of sulfur emission. At B6 or blend of 5% (95% diesel-5% CME), biodiesel derived from CME reduces smoke emission by 40-50%, net C02 emissions by 78%, CO by 20-40%, SOx by 80-99% and PM by 20-40%. Pure CME, however, was found to have adverse effects on the engine; hence, a CME-diesel blend is recommended. Sources at DOE also claim that pure CME would be very expensive at 40 to 50 pesos per liter compared to diesel at 15 pesos per liter. Car manufacturers have also raised a concern that CME destroys the engine although CME producers are disputing this. According to SATMP (2000), at 60:40 CME-diesel ratio, the diminution in engine's power output is considered tolerable at 6.45%. At 40:60 blend, the impact of CME on engine power is deemed negligible, i.e., unrecognizable by the vehicle operator. Also at this proportion, the sulfur content of the fuel mix is 0.05%, the prescribed sulfur content for fuel by the Clean Air Act. Advocates of CME are nonetheless pushing for 5:95 CME-diesel blend for at least two reasons. First, the dwindling local supply of coconut may not permit a higher proportion of CME in the mixture. Second, considering usual market skepticism over new product, experts reckon that it would be prudent marketing-wise to begin with incremental change over existing fuel. Data on vehicles using CME-diesel blend is not available but the DOE targets a total of 800 vehicles for 2003, to be increased to 500 and 1,500 by 2008 and 2013, respectively. To support these targets, the DOE is conducting workshops and tri-media campaigns to increase public awareness and advocate the use of CME. It also hopes to develop incentive packages for CME producers and users, including tax credits, reduced import tariffs and financing packages. Government agencies such as the Philippine Coconut Authority (PCA), the Department of Environment and Natural Resources (DENR) and the Department of Science and Technology have also initiated CME-related activities. A 20,000-liter storage tank was installed at the PCA while DENR- 46 ------- Quezon City has installed a 10,000-liter pump dispensing 5%-CME and 95-% diesel blend to cater to their service vehicles. SENBEL Fine Chemical Company, Inc., sole and exclusive manufacturer and dealer of bio-diesel grade fuel in the local market, has offered to supply NRDC/DENR with product called "Estrol Biodiesel (Coco Methyl Ester) fuel additive "when its use is proven to comply with the standard set by the CAA (RA 8479) - product to be blended with diesel or in pre-mix form. SENBEL got a one-year Provisional Registration for Diesel Fuel use only of its product from DOE on Oct. 23, 2002, pending promulgation of specifications for CME by the BPS. Table 15. Estimate of Share of Utility Vehicles Running on CME in 2005 and 2010 Percentage of Utility Vehicles 2005 2010 Low Estimate 0.64% 2.0% High Estimate 1.27% 4.0% As with CNG, the study simulated the DOE targets on the displacement of some of the diesel-fueled utility vehicles with those running on diesel-CME blend. Thus, 2,500 jeeepneys were assumed to be in place in 2005 and 10,000 in 2010. As in the CNG-fueled buses, these targets were doubled to simulate a "high" scenario. These jeepneys are assumed to be plying all the 98 zones in Metro Manila. As shown in Table 15, these jeepneys will only form a small part of the total fleet of utility vehicles. Based on data from DENR, blending diesel with at least 5% CME reduces PM emission up to 40% and C02 emission by 78%. 4.2. Air Pollutant Concentrations 4.2.1. The ISCLT3 Model The ground-level concentration of particulates in Manila was predicted using the Industrial Source Complex Model (ISC3). ISC3 is a Gaussian dispersion model often used for regulatory applications as recommended by the U.S. Environmental Protection Agency (2001). The model is however suited to the objectives of this study due its capability to model the dispersion of pollutants from many sources and of many types. 47 ------- ISC3 has two versions, a short-term (ISCST3) and a long-term (ISCLT3) model. ISCST3 calculates short-term concentrations for each hour of the year, and, by averaging the hourly values, can also yield daily, monthly, and annual concentrations. However, the ISCLT3 model was selected in this study for several reasons, not the least of which is the absence of hourly meteorological data required by ISCST3. ISCLT3 has the advantage of generating annual average concentrations far more quickly than ISCST3, a critical requirement given the number of sources and receptor points that are considered in this study. ISCLT3 estimates this annual value as the frequency-weighted average of the short-term concentrations arising from each possible meteorological condition. ISCLT3 generally results in slightly lower peak concentrations than ISCST3 owing to the added averaging it imparts on the calculated concentrations. ISCLT3 was run in order to predict concentrations in a receptor grid covering the entire Metro Manila. With receptor points placed 100 meters apart, this receptor grid had 256 points along the east-west direction and 491 points along the north-south direction. This provides high resolution in the concentration field needed to estimate average PM concentrations in each municipality of Metro Manila, but comes at the expense of computing time. 4.2.2. Inputs and Assumptions As mentioned earlier, the study suffered from the lack of meteorological data in Metro Manila. Despite the presence of three weather stations in the capital, data of suitable length and quality was not immediately available for this study. Instead, existing older data from the 1994 Urbair study (Shah and Nagpal 1994) in Metro Manila was used. Assuming that the same variability in weather conditions applies to the present, errors introduced by the use of older data are not expected to be serious. The available data consists of wind direction, wind speed and cloudiness readings at the Science Garden Station taken every 3 hours during the months of January, April and August. January represents conditions during the northeast monsoon, August the southwest monsoon, and April the transition period between these two seasons. Wind roses at the Science Garden station are shown in Figure 6. 48 ------- Figure 6. Windrose at the Science Garden Station From the wind speed, cloudiness and time of the reading, a stability class was generated using the so-called Pasquill criteria. The sequential data were then converted into the Stability Array (StAr) format required by ISCLT3. For mixing heights, a uniform value of 2000 m was used for unstable and neutral stability conditions, while unlimited mixing was assumed for the stable classes. No background levels were added to the model results owing to the lack of data, although as discussed later the contribution of stationary sources were included. The influence of this simplification is expected to be minor since the concern of the study is the change in the concentrations and their corresponding health impacts. The study also models PMi0only, and assumes that finer particles such as PM2.5 are part of the PM10 load. The model does not allow the inclusion of secondary particulate formation, which was neglected. ------- It should be reiterated that because the source emissions are specified as uniform for each traffic zone, the resulting concentrations are also uniform within each zone except at the boundaries between two zones. This simplification precludes a comparison of the model results against observed data in Metro Manila because the resulting model concentrations are essentially area-wide averages rather than receptor concentrations that can be compared to monitoring data. But since the output of the model required by the health component of this study is only the annual average PM concentration in each municipality and for the entire Metro Manila, the representation of air quality by traffic zone is enough to meet the objectives of the study. 4.2.3. Emissions Inventory PM emissions associated with the traffic generated by the model are summarized in Table 16. Diesel vehicles appear to account for bulk of the emissions; private vehicles, due to their numbers, contribute more than public vehicles. Emissions from public gasoline-driven vehicles are exclusively from two-stroke tricycles, which are often poorly maintained and overburdened. Table 16. Summary of PM emissions (in tons per year) from mobile sources (2002 baseline case). Fuel Private Public Total Gas 939 4,254 5,193 Diesel 7,392 3,823 11,215 Total 8,331 8,077 16,408 The traffic flow generated by the NCTS traffic model was converted into emissions using appropriate emission factors. These traffic emissions, which were assumed to be uniformly distributed in each traffic zone, were then assigned to approximately 60,000 area sources each 100 m by 100 m in size covering Metro Manila. A color-coded map of PM emissions for the baseline, best-case and worst-case scenarios are shown in Figure 7. From the leftmost map in Figure 7, vehicular emissions in Metro Manila can be seen to be highest at the center of the city where business, commercial and educational facilities are clustered. The series of zones with significant emissions that trace rough lines leading to this center indicate the major traffic routes. Traffic after population growth in 2015 does not appear to alter these routes, as shown in the business-as-usual (BAU) map at the center of the same figure, but the increase in emissions from all the zones is evident. The potential for reduction in 2015 under the most optimistic scenario is presented 50 ------- in the rightmost map, where emissions may be seen to fall to less than half of the 2002 levels. 2005 Baseline 2015 Combination 10000m 20000m 0m 10000m 20000 m 0m 10000m 20000m Figure 7: Emissions (tons per year) for the 2005 baseline (left), 2015 business-as-usual (middle), and 2015 combination of policies (right) In addition to the mobile source emissions, stationary sources were also included in the modeling in order to provide more accurate estimates of the ambient concentrations. Information on emissions from factories and power plants was obtained from a database maintained by the Environmental Management Bureau (EMB) of the Department of Environment and Natural Resources (DENR) of the Philippines. The database currently lists 680 sources operated by 226 industrial facilities, and includes estimated emissions of sulfur dioxide and particulate matter (PM) based on fuel usage. Table 17 summarizes the data and emissions from this inventory. 51 ------- Table 17. PM emissions from stationary sources (2002). Fuel Type Annual usage Emission factor PM emission (kg) Diesel 51,486 m3 240.2 g/m3 (a) 12,365 Bunker 192,630 m3 3697.2 g/m3 (a) 712,193 (3 % sulfur) Wood Waste 22,780 tons 17.3 kg/ton (b) 358,267 IPG 241 tons 54.0 g/m3 (c) 24 (442 m3) Petroleum coke 25 tons 25.0 kg/ton (d) 561 Total (kg) 1,083,410 Total (tons per year) 1,083.41 Notes: a. From AP-42 Table 1.3-1 distillate and No. 6 oil c. From AP-42 Table 1.5-1 b. From AP-42 Table 1.3-12 (default emission factors) d. From AP-42 Table 1.1-3 (as bituminous coal) Compared to emissions from mobile sources, stationary sources appear to constitute a small fraction of total PM emissions in Metro Manila. Since the focus of the study is on vehicular emissions, the contribution of stationary sources was assumed constant in all the scenarios. Figure 8 shows the location of these industrial facilities. Three clusters of sources are evident: one at the northwestern section, another at the center, and a third at the south. The most important sources in the inventory are two oil-fired power plants at the northwestern section of the city. In modeling the dispersion of PM from both stationary and mobile sources, ISCLT3 requires the specification of distribution by mass of the different particle sizes that make up the total particulate matter load. Information on both particle size distributions and mass fractions were obtained from appendices of the USEPA Compilation of Air Pollutant Emission Factors. The assumed particle size data used in the modeling are summarized in Table 18. A uniform density of 2.0 grams per cubic centimeter was used for all particles. 52 ------- A " *- \ KALOOKAN1 CITY 1 QUEZON CITY 3KAN thy MARIKINA CITY \ UU^- ¦MftWltA MANOAlJjYgfc . B2j T 1 V*' i rjfr- -. ¦A5JGCITY wr /? TICITYpAffeROS a PASAY CTTY W \ TAGUlG PAfeANAGiCfE «i "V j LAS Pf^AS CITY MUNTINLUPA city / Locations of stationary sources , - * ~ ¦ Stationary source Figure 8. Map of stationary sources found in the 2002 DENR inventory Table 18. Assumed mass distribution and particle sizes used in the modeling. Source and Fuel Type Mass distribution (percent) by particle size class Stationary 10 pm 6 [jm 2.5 pm 1 pm Bunker oil 30 25 22 23 Diesel 7 3 8 82 L.PG 0 0 0 100 Wood 73 11 10 6 Petroleum coke 30 25 22 23 Others (Furnace) 11 7 10 72 Mobile 10 (jm 5 |am 2 pni 1 (jrn Gasoline 4 5 7 84 Diesel 2 4 2 92 53 ------- 4.2.4. Modeling Results Modeling results are shown as isopleth maps of ambient PM concentrations in Figure 9 for the 2002 baseline, 2015 business-as-usual (worst-case scenario), and 2015 under a combination of air quality management policies (best-case scenario). Poor air quality may already be seen from the 2002 baseline where highest annual concentrations from mobile sources alone reach 105 micrograms per Normal cubic meter (|jg/Ncm), well above the Philippine standard of 60 |jg/Ncm (indicated as red in Figure 9). These levels are confirmed by the observations of the Manila Observatory at the Epifanio de los Santos Avenue, Metro Manila's main artery, where 24-hour concentrations are consistently higher than this value. Exceedances appear to be confined to specific zones in the center of the capital where traffic flow is heaviest. Conditions get worse in 2015 under the business-as-usual case (center map in Figure 9), where population growth heightens the exceedances. Due largely to the input from the traffic model, the exceedances are found in the same places as the baseline. The potential for improving air quality resulting from the implementation of several policies is shown in the rightmost map of Figure 9, the best-case scenario. Annual PM concentrations fall to less than half of their 2002 levels, and all exceedances disappear. Clearly, the adoption of even just a few strategically selected policies can result in dramatic improvement in Metro Manila's air quality. Additional results showing mean PM concentrations within each city and municipality arising from the various proposed scenarios are presented in a later section. 54 ------- Figure 9. Calculated particulate concentrations in |ig/Ncm for the 2005 baseline (left), 2015 business-as-usual (middle), and 2015 combination of policies (right) 4.3. Health Effects Estimations 4.3.1. Health Impact Assessment The main methodology used was the health risk assessment approach using epidemiological studies, based on the Krzyanowski proposed method of assessing the extent of exposure and effects of air pollution in a given population. This framework for assessment of impact was mainly based on the Covello-Merkhofer model with four stages: assessment of release of pollutant; assessment of exposure; assessment of the consequence; and the risk estimation. He emphasized that the use of epidemiological data is critical in assessing the consequences of exposure to the pollutants. The modei used in this study utilized epidemiological data. The basic principle of risk estimation could be illustrated by the following equation: Attributable Number of Cases = Exposure-response coefficient X excess exposure level X exposed population X baseline mortality/morbidity rates 55 ------- The number of attributable cases for each policy scenario was calculated including that of the 'Business-as-usual' scenario (BAU) for the projected years. The attributable numbers of cases for the policy scenarios are then subtracted from the BAU scenarios for the respective projected years. These latter figures comprise the averted number of cases for each policy scenario. The exposed populations which cover the whole population of Metro Manila for the 2005, 2010 and 2015 are projected based on the population growth rates predicted for those years. The predicted population growth rates consider both the birth and migration rates. With regards the baseline morbidity and mortality rates, these rates are assumed to be constant and similar to the rates in 2002 for 2005, 2010 and 2015 in this estimation. All data input and calculations of the estimates were made using the Analytica software. 4.3.2. PM 10 As Pollutant Indicator for Health Impact Various air pollutants have different health effects. A number of epidemiologic studies however point to Particulate Matter, 10|_im (PM10). The respirable fraction of total suspended particulates is the most important air pollutant with severe consequences for human health (D Schwela, WHO, 1996). High levels of PM (e.g. 500|jg/m3) are known to cause premature death. Recent studies in U.S., Europe, Asia, South America have found association of PM with death at much lower levels, with no evidence of a "threshold" or safe level. When efforts have been made to identify a threshold, little conclusive evidence has been found that one exists (Ostro, 1984). An analysis conducted by Cakmak et al. (1999) indicated that if threshold concentrations did exist, the current statistical models are sufficiently robust to be able to identify them. PM10 may serve as a surrogate measure for the complete mix of particles and gases that result from fuel combustion from automobile. Studies showing an association between PMi0 and the different health endpoints are quite robust. PM10 appears to be most consistently associated with health outcomes ranging from acute respiratory symptoms to premature mortality. Taken together, the studies provide strong evidence of a causal association between PM and health and provide a quantitative basis for which to generate a risk assessment. There is also little evidence, based on reviewing studies conducted at a wide range of air pollution concentrations, that the slope of the dose-response function diminishes significantly at lower concentrations (U.S. EPA, 1996). Most of the epidemiological studies have estimated linear or near-linear functions that suggest a continuum of effects down to the lowest PM levels observed in the study sample. For example, for mortality, most studies report a linear association between the relative risk in a population (percent increase in mortality) and the concentration of PMi0. 56 ------- Thus, in this study, PM10 was used as an indicator of urban air quality and a proxy indicator for concurrent exposure to different pollutants. In addition and a very important consideration as well, data of other pollutants (such as S02, NOx, CO, VOCs, 03) in Metro Manila were not regularly collected in the past three years. 4.3.3. Exposure Response Function The exposure response function or exposure correlation coefficient is a measure of the relationship between variables. This function indicates the expected change in a given health outcome per unit of change of pollutant (PM10, in this case). The exposure-response/relative risk is the increase in the probability of a given health effect associated with a given increase in exposure (usually 10|jg/m3) in epidemiological studies of PM10. Most exposure-response (ER) functions commonly used by other impact assessment studies were derived from meta-analytical assessment of various (international) studies selected from peer reviewed epidemiological literature (Kunzli, 1999). These ER functions serve as central estimates with 95% confidence limits which are applied to the baseline morbidity/mortality data. In this study, most of the exposure response coefficient values were derived from time series studies rather than cohort studies reviewed. Data collection also employed the time series studies of daily mortality. Time series studies correlate daily variation in air pollution with variation in daily mortality in a given city and measure primarily, the effects of acute exposure to air pollution (WHO, 2000). These studies capture the effects of short-term exposure to the probability of dying. People who are in a weakened physical state or who have a history of chronic obstructive pulmonary disease (COPD) or cardio- pulmonary problems are thought to be the most vulnerable. Time series studies give a reliable lower bound of effect in studies on mortality and good estimates with lower measurement error and potential confounding compared to cohort studies. The advantage in using time series studies is the fact that these studies did not have to control for a large number of confounding factors since the population characteristics (age distribution, smoking, occupational exposure to pollutants, income levels, health habits, nutritional status, access to medical care, among others) were basically unchanged. The only factors likely to vary with daily mortality and morbidity were environmental and meteorological conditions (temperature, humidity, precipitation), season, day of the week, presence of other pollutants. These factors were likely to be accounted for in the statistical analysis. On the other hand, the exposure-response functions from mortality cohort studies imply the effects of air pollution in the long term survival. Combining 57 ------- the acute mortality effects with of the chronic mortality effects in a single estimation is not compatible, as they are two totally different effects. The exposure response functions from the cohort mortality studies are better used in trying to consider the effects on the life expectancy of specific populations and coming up with the years of life lost due to chronic exposure. Hence, for this project, only the exposure response functions from acute studies are utilized. 4.3.4. Health Outcome Variables As regards to indicators for health impact of air pollution, measurements of mortality and morbidity in Metro Manila in 2002 were utilized. Mortality and morbidity statistics were used as measures of divergence from health in the affected population. Mortality, refers to deaths recorded in the study population of interest while morbidity refers to occurrence of illnesses recorded during the study period. Mortality rate (per 10000 population) is obtained by dividing the number of reported deaths that occurred in a particular time and place by the population at that same time and place multiplied by a factor of 10000 in this case. Mortality rate is an accurate indicator of the state of health of a given population as well as the effectiveness of the health care delivery system. Mortality rates are determined by a milieu of factors both genetic and environmental. Morbidity rates (per 10000 population) are obtained similarly as the mortality rates, by dividing the number of people who got sick of a certain disease in an area in a given period of time by the total population of the area then multiplied by a factor of 10000. This study included all 17 cities and municipalities for the Metropolitan Manila-wide analysis. The data used in this project were largely taken from the Public Health Monitoring of Air Pollution Study by the Department of Health, World health Organization and The Asian Development Bank, 2003. The following describe the data collection process of the said study. For each city/municipality, age- and sex-specific morbidity rates were calculated for each disease condition. The population distribution, age and sex distribution were obtained from the records of the National Statistics Office (NSO) for 2002. The number of reported illnesses (morbidity) was taken from the records of each city/municipality. Specific morbidity rates were computed by dividing the number of cases in a specified age and sex group by the corresponding size of this population sub-group. Health centers provide preventive and curative health services to specific areas in a locality. These communities are often defined as catchment areas. Morbidity data were obtained from 414 health centers in Metro Manila and were used to determine the age- and sex- specific morbidity rates. 58 ------- The numerator in the determination of the age- and sex-specific morbidity was obtained through the following procedures: a) Age- and sex-specific data from the identified catchment area were obtained from NSO. Available data were not disaggregated by age and sex. b) To identify the site-specific age and sex distribution, it was assumed that the age and sex distribution of the health center catchment population were similar with the distribution in the city/municipality they belonged. c) Consequently, the percentage in age and sex group for the city/municipality is multiplied by the catchment population size of the health center to obtain the age and sex distribution of the latter population. d) Age- and sex-specific morbidity rates for each center catchment area were then computed using these derived age and sex distributions. The following table summarizes the health variables definition, source of data and time reference. Table 19. Health Outcome Variable, Source of Data and Time Reference Population Characteristics Definition Level of Aggregation Source of Data Time Reference Mortality data Total and age- and sex-specific numbers of deaths/midyear population per 10000 Metropolitan and city level City death registration system 2002 Cause of Death Number of deaths per cause/midyear population per 10000. Specifically for cardiovascular, pulmonary diseases and natural causes (accidents not included - termed here as natural mortality) Metropolitan and city level City death registration system 2002 Morbidity data Number of illnesses per cause/midyear population for each age group per 10000. Diseases included - Asthma, Acute bronchitis, Chronic bronchitis City and Metropolitan level and health centers City health office and health centers 2002 Hospital Admissions Number of hospital admissions all through out Metro Manila per cause of hospitalization. Specifically for Cardio-pulmonary diseases and respiratory Metropolitan level Philippine Health Insurance System 2002 59 ------- 4.3.5. Health Outcome Variables and Corresponding Exposure Response Coefficients A comparison of health outcomes, exposure response coefficient for every 10|jg/m3 increase in PM10, in various time series studies in developed countries as well as in developing countries (Annexes 1 - 4) was made. The following table illustrates the health endpoints used and their corresponding exposure response coefficients or relative risks per 10|jg/m3 of PM10 used in this study. Table 20. Health Outcome and Relative Risks (Per lOjig/m'of PM1t) CAUSE CENTRAL ESTIMATE LOWER LIMIT 95% CI UPPER LIMIT 95%CI NOTES SOURCE Mortality (total, excluding deaths from accidents, homicides, suicides) 1.011 1.009 1.015 Adults age 30+ Bangkok study (Ostro, 1998) Respiratory Mortality 1.052 1.047 1.054 Bangkok study (Ostro, 1998) Cardiovascular Mortality 1.014 1.004 1.021 Bangkok study (Ostro, 1998) Respiratory Hospital Admission 1.0131 1.001 1.025 All ages European studies (pooled, 1998) Cardiovascular Hospital Admissions 1.0125 1.007 1.019 All ages European study (Kunzli, 1999) US study (Schwartz, 1997) Asthma Attacks 1.044 1.027 1.062 < 15 years old European (panel studies, 1990, 1995) US study (Whittemore and Korn,1980) Asthma Attacks 1.039 1.015 1.059 =/>15 years old European (panel studies, 1995, 1998) Bronchitis Episodes 1.306 1.135 1.502 < 15 years old US study (Dockery, 1986,1989) European (1992,1993) Chronic Bronchitis 1.098 1.003 1.1194 >25 years old European study (Kunzli, 1999) These exposure response coefficients should be interpreted as change in the relative risk for a specific illness per increase of 10|jg/m3 PM10 level in ambient air. For example, in the tableabove, for asthma among children below 15 years old, a 10|jg/m3 increase in the PM10 air pollution would increase the said disease incidence by 1.044 times from baseline in the affected population. 60 ------- 4.3.6. Mortality Endpoint: Premature Mortality Among all endpoints associated with air pollution, mortality is the most important one both from the health and economic perspective. In the present analysis, both short-term and long-term changes were included. The exposure response functions used were derived from the studies done in developing countries, particularly from the Bangkok study (Ostro et al, 1998). For total mortality (excluding deaths from accidents, suicides, or homicides), a relative risk of 1.011 (95% CI 1.009, 1.015) was used. Bangkok has similar characteristics with Metro Manila in terms of baseline health status (population sensitivity), access to health care, demographics (in terms of age structure), growth rates, socio-economic status (lifestyle, and housing characteristics), climate, temperature and humidity among others. ER coefficients for mortality were not derived from time series studies in developed or industrialized countries due to differences in factors mentioned. Also, levels of particulate are often 3-4 times higher in developing countries than in industrialized ones. In addition, in developing countries such as the Philippines, people die at younger ages and from different causes than in industrialized countries, implying that extrapolations of the impacts of air pollution on mortality may be especially misleading. 4.3.7. Morbidity Studies Dose response functions and associated relative risks have been developed from studies in Europe and the United States for several morbidity endpoints. For many endpoints, multiple studies were available. To combine the study results into one central estimate and associated confidence intervals, a meta- analytic approach similar to that used for mortality was used. Ultimately, both European and American studies were pooled to obtain the final estimate of relative risk. Thus in this study, ER values were derived from various European and US studies. Respiratory Hospital Admission There is extensive evidence to indicate that short-term changes in PM impact hospital admissions for respiratory disease. Studies are available from both Europe and the US covering various different respiratory endpoints, including all respiratory diseases. From European studies, statistical analysis provided heterogeneity and gave a general estimate of 1.0131 (95%CI 1.001,1.025). A US study by Scwartz (1995) gave a value of 1.0172. In our study, the ER value of 1.0131 was used for respiratory hospital admission. 61 ------- Cardiovascular Hospital Admission There is evidence from studies conducted in several cities of an association between daily changes in hospital admissions for cardiovascular disease. Following Kunzli (1999), from studies conducted in European cities a central estimate of 1.0125 was derived. A US study by Schwartz (1997) gave a similar value of 1.012. In this study, an ER function of 1.0125 (95%CI 1.007,1.019) was used for cardiovascular hospital admission. Asthma Attacks, <15 years old Asthma attacks were defined by wheeze or shortness of breath. There have been several studies conducted in Europe and US where many asthmatics are studied over time in panel studies. Three European panel studies on asthmatics generated a relative risk of 1.044. Evidence from the US was provided by Pope et al (1991) and Ostro et al (1995) and generated a risk estimate of 1.051. Another US study by Whittemore and Korn (1980) gave a relative risk of 1.044. In this present study, the value 1.044 (95%CI 1.027,1.062) was used for asthma occurrences among children, aged below 15 years old. Asthma Attacks, =/ >15 years old Three European panel studies are available for adults. A joint European estimate was 1.039. Two US studies are also available, by Pope et al (1991) and Ostro et al (1991) which gave a relative risk of 1.002. In this present study, the ER function used was 1.039 (95%CI 1.015,1.059) for asthma attacks among the population aged 15 years old and above. Bronchitis Episodes <15 years old Several studies indicate an association between annual exposure to PM and bronchitis episodes in children. Two US studies by Dockery et al (1989, 1986) and a European study in Switzerland (1992,1993), yielded a pooled risk estimate of 1.306, applied to the population of children below age 15. In this present study, an ER function of 1.306 (95%CI 1.135, 1.502) was used for bronchitis episodes below 15 years old. Chronic Bronchitis, > 25 years old In this study, the ER function used for chronic bronchitis among the population aged above 25 and above was from the three cities European study by Kunzli (1999) which is 1.098 (95%CI 1.003, 1.194). 62 ------- 4.3.8.0ther Health Variables Considered Aside from the health variables used in this study, other health outcomes potentially relevant for health impact assessment of air pollution are: Acute Outcomes - Restricted activity days- where an individual misses work, spends the day in bed or is otherwise forced to limit his normal activity - Emergency room visits for respiratory and cardiac problems - Lower respiratory illness in children - Cough days - Chest discomfort days - Work absenteeism - School days missed Chronic Disease Outcomes - Lung cancer Reproductive outcomes - Pregnancy complications (including fetal death) - Low birth weight - Pre-term delivery Lack of local data either from the health centers or hospitals on the above variables made us confine our study to the health variables mentioned in the previous section. In addition, data on "work absenteeism" comes from the Department of Labor and on school days missed from the Department of Education. The limited time frame of the study hampered us from coordinating with the agencies concerned. Furthermore, there are not enough international studies on which to base the exposure response coefficient on, and therefore unable to proceed with risk estimation. 4.3.9. Summary of Assumptions 1. Exposure response functions of the acute air pollution studies from other cities are applicable to the population of Metro Manila. 2. The present population growth rate would be consistent all throughout the projected years and include the in and out migration. 3. The baseline morbidity and mortality rates in year 2002 would remain the same all throughout the projected years. 4. Exposure parameters are taken from the scenarios developed in the previous two sections. 63 ------- 4.4. Economic Valuation In order to conduct the economic valuation we need to know the unit cost values to translate health impacts into economic values. Several methods were used to estimate unit cost values: benefits transfer, direct cost of illness (medical costs), indirect cost of illness (lost work days). 4.4.1. Assumptions ( PHMAP, 2003) Although the study was able to collect a large volume of data for analysis, certain specific assumptions were made to progress the analysis and in some cases due to absence or lack of supporting local and international data. The most important assumption is that air pollution contributes to adverse health outcomes in Metro Manila. This assumption is based on the multitude of epidemiological studies linking air pollution and health in many cities in the world. In addition, in two Manila studies of specific populations, namely jeepney and bus drivers and schoolchildren, such an association was likewise seen. However, in this assessment, all of Metro Manila population is assumed to be affected. The second most important assumption and related to the first one is that the magnitude of the effect of air pollution on health is similar to that seen in other cities of comparable features to Metro Manila. It is assumed that the exposure-response coefficients, derived from the Bangkok daily time series study on mortality and daily time series studies on morbidity from several developed countries, are applicable to the conditions in Metro Manila. The rationale for this assumption is that the effects seen in most of these studies, either from developed or developing countries are almost similar in magnitude especially with the mortality studies. The third assumption is the level of exposure. It is known that populations in a big city such as Metro Manila are not uniformly exposed. In this assessment, based on the modeled annual levels per city, it is assumed that in each city and for the whole metropolis, the level of exposure is the same for everyone. This assumption is unavoidable since individual exposures are difficult and impractical to obtain. The studies from where the coefficients were derived also had a similar assumption - that a single monitoring station could represent the exposure of everyone in the city. Corollary to this, it is likewise assumed that the mix of pollutants and how it behaves in the atmosphere are similar for Metro Manila and the other cities from where the coefficients were taken. A fourth assumption is concerned with the morbidity and mortality rates. These rates taken in 2002 are assumed to be constant for all the future years - 2005, 2010, 2015. It is quite difficult to project these to the future as there are so many factors which could affect these rates and historical data, from 64 ------- which projections could be based on, are not as reliable as it should be. Thus, this assumption was necessary for this exercise. Thus with such assumptions, the interpretation of the results of the health risk assessment as in any other health risk assessment studies must be treated with caution. Nevertheless, in the absence of local epidemiological studies and hard data, this type of study provides information as to the extent of the impact of air pollution on health. 4.4.2. Benefits Transfer The economic valuation of health effects in the Philippines has been hampered by the lack of sufficient and reliable data and the resources to conduct a primary study to estimate the benefits of the health policy in question. In the face of such resource and data constraints, analysts have relied on "benefits transfer" to assess the mortality and morbidity impacts of environmental scenarios. Benefits transfer is the shorthand term referring to the practice of taking and adapting value estimates from past research (e.g. values for a change in health conditions) and using them (hence the term "transferring") to assess the value of a similar, but separate scenario (e.g., a comparable health condition). One elementary procedure is to "borrow" a single estimate in one context and transferring it to another context. However, transferring estimates which are unadjusted is hazardous. Differences of values between two sites can arise from differences in the socio-economic characteristics of the relevant populations; the physical characteristics of the two sites; the proposed change in the provision between the sites of the good to be valued and the market conditions applying to the two sites. Estimates which are transferred need to be adjusted. The most widely used modification is the adjustment for differences in income. The feature that is changed between the two sites is income, perhaps because it is thought that this is the most important factor resulting in the estimates of economic value. In their study, "Air Quality Impacts of Increased Use of Indigenous Fuels for Power Generation in the Philippines," Orbeta and Rufo (2003)3 used benefits transfer to estimate the unit values for mortality and morbidity effects. They "borrowed" the values found in the Rowe, et al (1995)4 study of New York state environmental externalities costs and adjusted the original values for differences in income levels and exchange rates. Faced with the practical pressures to short-cut full scale procedures to estimate unit economic values, we used the adjusted values found in Orbeta and Rufo (2003). We readjusted these values to set them in 1995 prices computed using 3 Orbeta, E. M. and C. M. Rufo (2003), Air Quality Impacts of Increased Use of Indigenous Fuels for Power, Research Report No. 2003-RR3, Economy and Environment Program for Southeast Asia. 4 Rowe, R.D., C.M. Lang, L.G. Chrstnut, D.A. Latimer, D.A. Rae, S.M. Bernow and D.E. White (1995). New York State Environmental Externalities Cost Study (Volume I). ESEERCO, New York. 65 ------- Philippines Consumer Price Index and using the 2002 U.S. Dollar - Philippine Peso exchange rate. The readjusted values are presented in the table below. Table 21. Benefits Transfer: Unit Values for Mortality and Morbidity Effects Value per case (1995 Prices) Health Effects Philippine Pesos Year 2002 Low Central High Mortality 3,439,459 6,676,606 13,353,190 Morbidity Respiratory hospital admissions 14,166 28,331 42,486 Emergency room visits 543 1,075 1,607 Child bronchitis 282 543 825 Restricted activity day 76 141 217 Asthma attack day 22 65 109 Acute respiratory symptom day 11 22 33 Adult chronic bronchitis 254,915 424,881 679,796 To estimate the unit costs of health impacts of different transport scenarios for years after 2002 (e.g., 2005, 2010, 2015), present values of the unit costs presented in the table above were calculated using a discount rate of 12%.5 Direct Cost of Illness (Medical costs) Another way to derive unit values for morbidity is to estimate "avoided medical costs." Information was collected from records of claims made on Philhealth which is the legally required medical insurance scheme for salaried workers in the Philippines. Data was available on payments made by Philhealth for hospital room and board, drugs and medicines, medical tests (x-ray, laboratory) and professional fees. The data can considered as a conservative estimate of the medical costs for particular health effects because Philhealth payments do not cover the full costs paid by patients for their medical treatment. The table below presents the data computed from Philhealth records. 5 Rowe, R.D., C.M. Lang, L.G. Chrstnut, D.A. Latimer, D.A. Rae, S.M. Bernow and D.E. White (1995). New York State Environmental Externalities Cost Study (Volume I). ESEERCO, New York. 66 ------- Table 22. Direct Medical Costs Year 2002 (Philippine Pesos - 1995 prices) Medical Item Cardiovascular diseases Asthma Room and board 3,789 2,722 Drugs and medicine 5,390 4,830 X-ray, lab and others 9,968 4,096 Professional fees 2,012 1,567 Total 21,158 13,215 To estimate the unit costs of health impacts of different transport scenarios for years after 2002 (e.g., 2005, 2010, 2015), present values of the unit costs presented in the table above were calculated using a discount rate of 12%. Indirect Cost of Illness (Lost Work Days) Aside from medical costs, the costs of illnesses include also lost income due to lost work days. Estimates of the number of work days lost for a specific illness was made by expert judgment of physicians. The income lost per day was assumed to be the minimum daily wage rate in year 2002 mandated by Philippine law (PhP 181.53 in 1995 prices). The estimates for indirect cost of illness are presented in the table below. Table 23. Cost of Lost Work Days Year 2002 (Philippine Pesos-1995 prices) Disease Low Central High Asthma 545 908 1,271 Acute Bronchitis 545 726 908 Chronic Bronchitis 908 1,271 2,541 Respiratory Admissions 908 1,271 2,541 Cardiovascular Admissions 1,271 3,812 5,446 To estimate the unit costs of health impacts of different transport scenarios for years after 2002 (e.g., 2005, 2010, 2015), present values of the unit costs presented in the table above were calculated using a discount rate of 12%. 67 ------- 5 Results and Discussions 5.1 Impact of Scenarios on Travel Demand This section discusses the results of the simulation of different transport and fuel policy measures and their impact on travel demand. The results are presented and discussed in terms of the comparative impact of the different scenarios for the three base years studied, namely 2005, 2010 and 2015. 5.1.1 Total Travel Demand This section discusses the results of the estimation of total travel demand (expressed in vehicle-kilometers) for each of the three base years based on the methodology described in Sec. 4.1.3.6. The travel demand projection was also disaggregated in terms of the different vehicle types and the highlights of the projections for each of the three base years are discussed in Sections 5.1.3.1 to 5.1.3.3. As noted earlier, these projections will be used as the basis for the calculation of vehicle emissions. ~ BAU ¦ MVIS ~ CNGBL ~ CNGBH ¦ CMEJL ~ CMEJH ~ TDM ~ 4STC ¦ BWMK ¦ BWMM ~ DPTB ~ DPTBJ ¦ RAIL ¦ COMBI 2005 2010 2015 Figure 10. Projected Travel Demand per Scenario : 2005, 2010, 2015 Figure 10 shows the results of the estimation of total travel demand for 2005, 2010 and 2015 for each scenario. The total travel demand for the BAU scenario in 2005 is estimated at 73.5 million vehicle kilometers. For 2010, the travel demand for the BAU scenario increased by 11.1 % to 81.7 million vehicle-kilometers. For 2015, the projected demand for the BAU scenario is 68 ------- around 84.3 vehicle kilometers, about 14.5% higher than the BAU scenario for 2005 but only 3.1 % higher than the BAU scenario for 2010. Among the alternative scenarios, only the Traffic Demand Management (TDM), the Bikeways Construction for Marikina and Metro Manila (BWMK and BWMM), the Policy Combination (COMBI) and Railway expansion (RAIL) scenarios resulted in a reduction in travel demand relative to the BAU scenario. The travel demand for the other policy scenarios, that is the Motor Vehicle Inspection System (MVIS), the replacement of 2-stroke motorcycles with 4-stroke motorcycles for tricycles (4STC), the Low- and High- assumptions of penetration of Compressed Natural Gas (CNGBL and CNGBH), the Low- and High-assumptions of penetration of Coco-methyl ester blend for diesel-fueled Jeepneys (CMEJL and CMEJH) and the Diesel Particulate Trap for Buses and Jeepneys (DPTB and DPTBJ) are the same as in the BAU scenario. The specific trends in each of the base years are discussed below. 5.1.1.1 Year 2005 ~ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ~ gas jeepney ~ gas car ///// ^ Scenarios Combi = MVIS+TDM+C NGBH+CMEJH+ DPTBJ Figure 11. Projected Travel Demand: 2005 Figure 11 shows the results of the estimation of travel demand by vehicle type for 2005. Compared to the BAU scenario, the travel demand for the COMBI and TDM scenarios are lower by 7.9%. It should be noted that the COMBI scenario includes the TDM scenario, which has in fact caused the reduction in the total travel demand. The BWMK scenario results in the reduction of 1,702 vehicle-kilometers per day for tricycles, which translates to only .02% of the total vehicle kilometers for tricycles or a mere .002% of the 69 ------- total vehicle-kilometers for all vehicles. For the BWMM scenario, the reduction in vehicle-kilometers for tricycles increased to 128,523 vehicle-kilometers per day or about 1.5% of the total vehicle-kilometers for tricycles. However, as a percentage of the total vehicle-kilometers, such reduction is still insignificant at 0.18%. In terms of distribution, private vehicles account for a much larger share of the travel demand at about 72% of the total under the BAU and other scenarios which do not involve a reduction or change in travel demand as mentioned earlier, that is, the MVIS, CNGBL, CNGBH, CMEJL, CMEJH 4STC, the DPTB and the DPTBJ scenarios. In the TDM and COMBI scenarios, this share was slightly reduced to 69% due to the assumed reduction of private traffic volume by 11.08%. Among the different vehicle types, the gasoline-fuelled cars showed the highest share of the travel demand at about 25.2 %, followed by diesel car and diesel truck at 18.3 % and 15.1%, respectively, in the BAU scenario. The diesel bus had the lowest share of 5.2 %. In the TDM scenario, the shares of gasoline car, diesel car and gasoline jeepney changed slightly by 1.3%, 0.9% and 0.7%, respectively while the shares of the rest of the vehicle types increased slightly by one percentage point or less. As with the total travel demand, the shares of tricycle for the BWMK and BWMM scenarios also decreased only slightly relative to the BAU scenario. 5.1.1.2 Year 2010 ~ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ¦ gas jeepney ~ gas car f /<* ~ / / Scenarios Ccmbi = MVIS+TDM+CNGBH+CMEJU+DPTBJf B WM M+4STC Figure 12. Projected Travel Demand: 2010 Figure 12 shows the results of the estimation of travel demand for each of the scenarios for 2010. The total travel demand for the combination scenario was lower by 5% relative to the BAU because it includes the TDM scheme with the 70 ------- same rate of reduction (11.1%) in the vehicle-kilometers of private transport modes as in 2005 and also the BWMM scenario which assumes a 1.5% reduction of vehicle-kilometers for tricycles. The distribution of travel demand between public and private vehicles and the relative shares of the different vehicle types in 2010 are almost the same as in 2005. As noted in Sec. 4.1.3.6, the share of vehicle-kilometers of road-based transport modes is assumed as the same for 2005, 2010 and 2015. 5.1.1.3 2015 Figure 13 shows the results of the estimation of travel demand for 2015. The RAIL scenario which was not included in the previous base years (i.e., 2005 and 2010) shows a very significant impact with a reduction in travel demand of about 10.3 million vehicle-kilometers or about 12.3% of the total vehicle- kilometers for the BAU scenario in 2015. The combination scenario, which includes the TDM, BWMM and RAIL scenarios, results in a total reduction of about 12.5 million vehicle-kilometers or about 14.9% of the total travel demand for BAU 2015. Scenarios ¦ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ~ gas jeepney ~ gas car Cornbi = MVIS+TDM+CNGBH+CMEJH+DPTBJ+BWMM+4STC Figure 13. Projected Travel Demand: 2015 5.2 PM and C02 Emissions This section discusses the results of the estimation of PM and C02 emissions based on the methodology described in Sec. 4.1.4. As with the travel demand projections, the results are discussed in terms of the shares of the different 71 ------- vehicle-types under the various scenarios for the three base years considered, that is, 2005, 2010 and 2015. 5.2.1 Total PM Emissions Figure 14 shows the results of the estimation of PM emissions in 2005, 2010 and 2015. For the BAU scenario in 2005, the total PM emission in Metro Manila was estimated at 48.4 tons per day or about 17,673 tons per year. For 2010, the total emission in BAU scenario increased by 9.8 tons per day or 20%. The total emission for the BAU scenario in 2015 was estimated at 60.9 tons per day, which translates to an increase of 2.7 tons per day or 4.5% relative to the 2010 BAU scenario and an increase of 12.5 tons per day or 25.8% compared to the 2005 BAU scenario. As noted earlier, the share of vehicle-kilometers of road-based transport modes is assumed as the same for 2005, 2010 and 2015. The number of vehicle-kilometers for tricycle was also assumed constant throughout the three base years. Thus, the increase in emission between the scenarios in 2005 and in 2010 is largely driven by the changes in assumptions that affect emission levels, for example, emission standards and increase in vehicles using alternative fuels. 70.00' 60.00' 50.00' as "ft 40.00f a o ~ 30.00 £0.00-)- 10.00-1- 0.00 ~ BAU ¦ MVIS ~CNGBL ~CNGBH ¦CMEJL ~CMEJH ¦TDM ~4STC ¦ BWMK ¦ BWMM ~ DPTB ~ DPTBJ ¦ RAIL 2005 2010 2015 Figure 14. Total PM Emission per Scenario: 2005, 2020, 2015 5.2.1.1. Year 2005 Figure 15 shows the results of the estimation of PM emissions in 2005. While travel demand is dominated by private vehicles, the results of the analysis showed that PM emissions are divided almost equally between private and public vehicles, with 48% and 52 %, respectively, in the BAU scenario. ------- Among the vehicle types, gas tricycles account for the highest share of PM emissions of 29.8%, followed by diesel car, diesel truck and diesel jeepneys with 23.1%, 19.6% and 14.9% respectively. 60 50 40 >s KS 3 30 20 10-- Jill I I ¦ gas tricycle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ~ gas jeepney ~ gas car Scenarios Comtt = MVIS+TDM+CNGGH +CMEJH +DPTBJ+ BWMM 'emissions from exhaust and idle Figure 15. Projected PM Emissions: 2005* The results of the simulation indicate that the greatest reduction in PM emissions relative to BAU scenario may be achieved with the combination of scenarios or "COMBI", which includes almost all the individual policies (i.e., except the 'low' cases for CNG buses and CME jeepneys, the bikeways construction for Marikina and the diesel particulate traps for buses only). However, the 4STC scenario alone indicates an SPM reduction of almost the same magnitude as the combination scenario, that is, about 11.6 tons per day or 24%. Following are other highlights of the results of the estimation of PM emissions for 2005 in the various scenarios: In the 4STC scenario, the PM emission from tricycles was reduced by 80% relative to the BAU scenario. Furthermore, the share of PM emission from tricycles to the total emissions from all vehicles declined drastically from 30 to 8%. Such indication of the high impact of the 4STC scenario may be explained by the fact that gas-fuelled tricycles have the highest emission factor for Suspended Particulate Matter among the different vehicle types (see Table 3) and that these vehicles also account for a significant share (9%) in the total travel demand. As noted earlier, however, this scenario has a very ambitious target of 100% conversion of tricycles from using 2- stroke to 4-stroke motorcycle engines. 73 ------- The MVIS scenario showed a great potential for emission reduction because of the assumption of a significant reduction (60 %) in emission for all vehicle types as envisaged in the Clean Air Act and 100 % compliance of the jeepneys, which are among the most polluting. Thus, in the MVIS scenario, the emission of diesel jeepneys was significantly reduced by 3.1 tons per day relative to the BAU scenario and its share correspondingly declined from 15% to 10%. The scenarios involving diesel particulate traps (DPT) also show significant PM emission reduction, especially when applied to both the buses and jeepneys, wherein the reduction amounts to 3.2 tons or about 7% of the total emission in the BAU scenario. In the TDM scenario, the 11.1% reduction in the vehicle-kilometers from private vehicles, namely gasoline cars, gasoline jeepneys and diesel cars, translated to a relatively significant PM emission reduction of 1.8 tons per day. Because of the very small targets for the displacement of diesel buses by CNG-fuelled buses, the reduction in PM emissions from diesel buses under the low and high CNGB scenarios are quite insignificant at .011 tons per day or .3% and .021 tons per day or .6%, respectively, notwithstanding that the emission factor of CNG is 86% lower than that of diesel. Thus, the share of diesel buses in the PM emission also remained almost unchanged. With the very low rate of displacement of diesel jeepneys by those fuelled by CME-blended diesel, the reduction in PM emissions from diesel jeepneys under the high and low CMEJ scenarios are also quite insignificant at .06 tons per day or .01% and .08 tons per day or .07%, respectively. Thus, the share of diesel jeepneys in the total PM emission also barely changed. PM emission reduction is also insignificant for the BWMK scenario about .003 tons or 3,000 grams per day or a mere .006 % of the projected PM emission in Metro Manila. If the bikeways program is extended to all other areas in Metro Manila, that is, with the BWMM scenario, the reduction increases to .22 tons or a slightly higher but still insignificant .45% share. This trend could be attributed to the very low assumption of the increase in the rate of shift from tricycles to cycling modes of 1.5%. It should also be noted that Marikina only accounts for 2 out of the 98 traffic zones in Metro Manila. 5.2.1.2. Year 2010 Figure 16 shows the results of the estimation of PM emissions in 2010. As with the results for 2005, the combination scenario, which includes almost all the individual policies (i.e., except the 'low' cases for CNG buses and CME jeepneys, the bikeways construction for Marikina and the diesel 74 ------- particulate traps for buses only) shows the highest potential for reduction of PM emission at 25 tons per day or 57% relative to the BAU scenario. A number of single-policy scenarios, however, also indicate significant reductions particularly the DPTBJ with 19 tons per day or 32.7%, the MVIS with 12 tons per day or 21%, the 4STC with 16 tons per day or 27.5%. 70 60 50 > (C "O 40 "H O 30 20 10-H 0 I - l_l^l l_l I I I I 1 , I I I I I I 1 1 I ~gas tricycle ~diesel bus ¦diesel jeepney ~diesel truck ~diesel car ~gas jeepney ~gas car Scenarios Combi = MVIS+TDM+CNGBH +CMEJH+ DFTEJ+BWMM 'emissions from exhaust and idle Figure 16. Projected PM Emissions: 2010* Following are some of the other significant results of the estimation of PM emissions for 2010 in the various scenarios: In the MVIS scenario, the implementation of the inspection and maintenance (l/M) standards under the Clean Air Act, which assumes a 30% reduction in PM emissions, in addition to the STDS2 standards, which assumes a 60% reduction in PM emissions, resulted in a total reduction of 4.4 tons per day from the 2005 MVIS scenario. Relative to the BAU 2010 scenario, the reduction in total PM emission in the 2010 MVIS scenario is 11.6 tons per day or about 8%. The increase in the targets for the displacement of diesel-run buses by CNG-fuelled buses between 2005 and 2010 resulted in a total PM emission reduction of 5.4 tons per day for the low scenario and 5.2 tons per day for the high scenario. For the scenarios involving the displacement of diesel-run jeepneys by those fuelled with CME-blended diesel, that is, the CMEJ scenarios, the magnitude of reduction is almost comparable to those of the CNG scenarios with 5.49 tons per day in the low scenario and 5.42 tons per day in the high scenario. 75 ------- 5.2.1.3. Year 2015 ¦gas tricycle ~diesel bus ¦diesel jeepney ~diesel truck ~diesel car ¦gas jeepney ~ gas car Scenarios Combi = M VIS+T D M +C NG BH4-CMEJH+DPTBJ+BWM M +4STC+RAl L * emissions from exhaust and idle Figure 17. Projected PM Emissions: 2015* Figure 17 shows the results of the estimation of PM emissions in 2015. The combination scenario, which includes almost all the individual policies (i.e., except the 'low' cases for CNG buses and CME jeepneys and the bikeways construction for Marikina) shows the highest potential for reduction of PM emission at 41.8 tons per day or 69% relative to the BAU scenario. A number of single-policy scenarios, however, also indicate significant reductions particularly the DPTBJ with 20.7 tons per day or 34%, the MVIS with 17.5 tons per day or 29%, the 4STC with 17.1 tons per day or 28 % and the RAIL with 11.1 tons per day or 18%. Following are some of the other significant results of the estimation of PM emissions for 2015 in the various scenarios: The continued implementation of the policy to require diesel particulate traps for buses (DPTB), and that of the traffic demand management (TDM) policy albeit at the same rate of private traffic reduction as in 2005 and 2010 also show a significant impact. Likewise, although the number of buses fuelled with CNG and jeepneys fuelled with CME-blended diesel did not increase from the 2010 targets (and thus accounted for a much smaller percentage of the bus and jeepney 76 ------- stocks) the results of the simulation show that they still make a significant impact in the reduction of PM emissions relative to the BAU scenario. The BWMK and BWMM scenarios do not show a significant impact in terms of the overall reduction of PM emissions relative to the BAU scenario. This could be attributed to the very low assumption of the increase in the rate of shift from tricycles to cycling modes of 1.5% in 2005 to 3.5% in 2015. Such assumption resulted in a reduction of the PM emission from gas tricycles of 0.76 tons per day from the BAU scenario for the BWMM scenario and only 0.01 tons per day for the BWMK scenario. 5.2.1.4. Annual PM Emissions Reduction for Each Policy Scenario Table 24 and Table 25 show the magnitudes of the total daily PM emissions for each policy scenario annually from 2005 to 2015. The PM emissions in years 2006-2009 and 2011-2014 was estimated by linear interpolation of transportation demand in terms of vehicle-kilometers per day for private and public transportation. The transportation demand levels estimated in 2005, 2010 and 2015 were used in the interpolation. A policy introduced in 2005 or in 2010 is assumed to still have an effect in future years thereby contributing to emissions reduction in those years. Table 24. PM Emissions for Each Policy Scenario, 2005-2015 (1), in tons/day Year Baseline MVIS TDM 4STC BWMK BWMM Rail 2015 2005 48.419 42.966 46.661 36.867 48.414 48.203 2006 50.157 42.754 48.327 38.505 50.152 49.939 2007 51.970 44.981 50.064 40.215 51.964 51.749 2008 54.072 46.652 52.084 42.209 54.067 53.850 2009 56.118 48.272 54.044 44.146 56.113 55.894 2010 58.236 46.636 56.072 46.152 58.230 58.009 2011 58.738 47.039 56.542 46.627 58.733 58.511 2012 59.328 47.512 57.097 47.188 59.322 59.100 2013 59.823 47.910 57.558 47.655 59.817 59.595 2014 60.326 48.315 58.026 48.130 60.321 60.098 2015 60.910 44.205 58.573 48.684 60.897 60.154 49.807 77 ------- Table 25. PM Emissions for Each Policy Scenario, 2005-2015 (2), in tons/day Year DPTBJ DPTB CNGB CMEJ Combo 1 Combo i 2005 45.217 47.381 48.358 48.343 36.433 2006 46.703 49.037 50.091 50.075 37.473 2007 48.250 50.762 51.897 51.878 38.556 2008 50.054 52.757 53.976 53.956 39.837 2009 51.813 54.707 56.012 55.991 41.076 2010 53.637 56.728 57.762 58.004 42.158 30.255 2011 54.316 57.222 58.247 58.489 42.526 30.596 2012 54.673 57.797 58.825 59.071 42.955 30.998 2013 55.146 58.285 59.319 59.565 43.319 31.333 2014 55.627 58.781 59.822 60.068 43.688 31.675 2015 56.184 59.355 60.401 60.647 43.592 31.971 21.17885 Table 26 and Table 27 show the reduction in total daily PM emissions for each policy scenario annually from 2005 to 2015 based on the difference between the PM emission of the policy scenario and the PM emission of the baseline scenario for the year. Measures such as the motor vehicle inspection scheme and introduction of 4-stroke tricycles are the single policies that effected higher reduction in emissions from 11-16 tons/day in 2010 to around 12-16 tons/day in 2015. These are followed by the completion of the railway network where it had reduced the emissions by 11 tons/day in 2015 and by the installation of DPTs in buses and jeepneys that reduced emissions by 4.6 tons/day in 2010 and 4.7 tons/day in 2015. Figure 18 shows the corresponding graph of emissions reduction for all the policy scenarios from 2005 to 2015. Table 26. PM Emissions Reduction for Each Policy Scenario, 2005-2015 (1) in tons/day Year MVIS TDM 4STC BWMK BWMM Rail 2015 2005 -5.453 -1.759 -11.552 -0.005 -0.217 2006 -7.403 -1.830 -11.652 -0.005 -0.218 2007 -6.989 -1.906 -11.755 -0.005 -0.220 2008 -7.420 -1.989 -11.863 -0.005 -0.222 2009 -7.846 -2.074 -11.972 -0.005 -0.224 2010 -11.600 -2.164 -12.084 -0.005 -0.227 2011 -11.699 -2.197 -12.112 -0.005 -0.227 2012 -11.816 -2.230 -12.140 -0.006 -0.228 2013 -11.913 -2.265 -12.168 -0.006 -0.228 2014 -12.012 -2.300 -12.196 -0.006 -0.229 2015 -16.704 -2.337 -12.226 -0.013 -0.756 -11.103 78 ------- Table 27. PM Emissions Reduction for Each Policy Scenario, 2005-2015 (2) in tons/day Year DPTBJ DPTB CNGB CMEJ Combo 1 Combo 2 Combo 3 2005 -3.203 -1.039 -0.062 -0.077 -11.987 2006 -3.454 -1.120 -0.066 -0.082 -12.684 2007 -3.720 -1.208 -0.073 -0.091 -13.413 2008 -4.018 -1.316 -0.096 -0.116 -14.236 2009 -4.306 -1.412 -0.106 -0.127 -15.042 2010 -4.599 -1.507 -0.473 -0.232 -16.078 -27.980 2011 -4.422 -1.517 -0.491 -0.250 -16.213 -28.143 2012 -4.654 -1.531 -0.503 -0.257 -16.373 -28.330 2013 -4.677 -1.538 -0.504 -0.258 -16.504 -28.490 2014 -4.699 -1.545 -0.505 -0.259 -16.638 -28.651 2015 -4.726 -1.555 -0.508 -0.263 -17.318 -28.939 79 ------- year 0.000 -5.000 -10.000 -15.000 >> CO T3 o -20.000 c o '¦4' o "I -25.000 QL -30.000 -35.000 -40.000 -45.000 IT) co CO CD o T C\l CO C > o o o O T T T T T T O o o o O O o o o o o C\l C\l C\l C\l C\l C\l CN CN CN CN C\l 1- M - r-r r-r ¦ i¦ - ¦ <¦ 1¦ ¦ ¦ ¦ ¦ ¦ H 1 1 h MVIS -¦TDM 4STC BWMK BWMM ¦«- Rail 2015 h DPTBJ DPTB CNGB CMEJ -o Combo 1 Combo 2 Combo 3 Figure 18. PM Emissions Reduction for All Policy Scenarios, 2005-2015 in tons/day 5.2.2. Total C02 Emissions Figure 19 shows the results of the estimation of total C02 emissions for each scenario in 2005, 2010 and 2015. The total emission for Metro Manila in 2005 in the BAU scenario was estimated at 44,069 tons per day or 16,085 kilotons for the whole year. With due regard to the differences in the calculation methodology, and mainly for the purposes of benchmarking this figure, it is worthy to note that the 1994 inventory of greenhouse gas emissions (The Philippines' Initial National Communication on Climate Change, 1999) ------- recorded a nationwide C02 emission for the transport sector of 15,801 kilotons. 60,000 50,000 40,000 > re «- 30,000 a> o. (/) £ O 20,000 10,000 rrruxm Mil wl ~ BAU ¦ MVIS ~ CNGBL ~ CNGBH ¦ CMEJL ~ CMEJH ¦ TDM ~ 4STC ¦ BWMK ¦ BWMM ~ DPTB ~ DPTBJ ¦ RAIL 2005 2010 2015 Figure 19. Total C02 Emission per Scenario : 2005, 2010, 2015 The total C02 emission in the BAU scenario for 2010 was estimated at 49,771 tons per day, which is about 13% higher than the figure for the 2005 BAU scenario. The total C02 emission for the BAU scenario in 2015 was estimated at 51,345 tons per day, which translates to an increase of 1,574 tons per day or 3% relative to the 2010 BAU scenario and an increase of 7,276 tons per day or 43% compared to the 2005 BAU scenario. 5.2.2.1. Year 2005 Figure 20 shows the results of the estimation of C02 emissions in 2005. The combination scenario showed a total reduction of 15,736 tons per day or 31.6% relative to the BAU scenario. Among the single-policy scenarios, the MVIS scenarios showed the highest reduction in the total C02 emission relative to the BAU scenario of 6,459 tons per day or 14.7%, followed by the TDM scenario with a reduction of 2,116 tons per day or 4.8%. 81 ------- t o n s / d a y 50,000 45,000 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 Scenarios ¦ gas tricycle ~ diesel bus ¦ dieseljeepney ~ diesel truck ~ diesel car ¦ gasjeepney ~ gas car 1 emissions from exhaust only Combi = IVM&-TDIVH-CNGBI-H-CMEJH Figure 20. Projected C02 Emissions: 2005' Following are other significant results of the calculation of C02 emissions for 2005: Among the different vehicle types, the diesel truck had the highest share of C02 emissions at 31%, followed by the gasoline car and the diesel car at 16.8% and 16.4%, respectively. This trend is explained in part by the high emission factor of diesel truck (1,249 grams per km) despite its small share in the total vehicle-kilometers (15%) on the one hand, and on the other hand, the high share of gasoline cars in the travel demand (25%) despite its lower emission factor (399 g/km). Despite the significantly lower (45%) C02 emission factor of CNG-fuelled buses compared to diesel buses, the total reduction in the CNGB scenarios is only 9 tons per day or .02% for the low case and 18 tons per day or .04% for the high case. Again, this is due to the very low targets for the penetration of CNG-fuelled buses in the total bus fleet. Blending diesel with CME even more significantly reduces the emission factor by 78%. However, with the low targets for the displacement of diesel jeepneys by those fuelled by CME-blended diesel, the reduction in the CME scenarios were estimated at only 23 tons per day or .05% for the low scenario and 46 tons per day or .1% for the high scenario. For the bikeways construction, the impact is nil if implemented in Marikina only with a reduction of only 0.32 tons per day and still insignificant if 82 ------- spread throughout Metro Manila with a reduction of 24 tons per day or 0.08% of the total emission in the BAU scenario. 5.2.2.2. Year 2010 Figure 21 shows the results of the estimation of C02 emissions in 2010. Again, the combination scenario showed the greatest reduction in the total C02 emission relative to the 2010 BAU scenario at 15,736 tons per day or 32%. With the MVIS scenario alone, the reduction is a little bit lower than that of the COMBI scenario at 10,619 tons per day or 21.3%. The TDM scenario also showed a significant reduction of 2,296 tons per day or 4.6%. 60,000 50,000 t 40,000 o " 30,000 s / d 20,000 a y 10,000 j g' II '1 II II 1 [ i 1 J r i i ¦ 1 T " k u p - ¦ r ¦gas tricycle ~desel bus ¦desel jeepney ~desel truck ~desel car ¦gas jeepney ~gas car Scenarios * enissions from exhaust only Gorrti = IVMS^TIX/KH3H-IK3VEJI+DPTBJ^0l/\M/l Figure 21. Projected C02 Emissions: 2010* Following are other significant results of the calculation of C02 emissions for 2010: Compared to 2005, the higher targets for the penetration of CNG-fueled buses also increased the reduction in C02 emission due to this policy to 154 tons per day or .31 % relative to the BAU scenario for the low case (CNGBL) and 318 tons per day or 0.64% for the high case (CNGBH). Similarly for the scenarios with CME blended diesel for jeepneys, the reduction is also more significant with 85 tons per day or 0.17% for the low case (CMEJL) and 181 tons per day or 0.36% for the high case (CMEJH). 83 ------- For the bikeways construction, the result of the simulation is the same as in 2005 with the same assumptions on the rate of shift from tricycles to cycling modes and the constant growth in travel demand for tricycles. 5.2.2.3. Year 2015 Figure 22 shows the results of the estimation of C02 emissions in 2010. Again, the combination scenario showed the greatest reduction in C02 emission relative to the BAU scenario of about 27,000 tons per day or more than half (53%). With the MVIS scenario alone, the reduction in C02 emission is 15,170 tons per day or 30%. With the RAIL scenario, the reduction is 6,725 tons per day or 13%. The diesel particulate traps for buses and jeepneys (DPTBJ) also registered a significant reduction of 3,722 tons per day or 7.3% as well as the TDM scenario with a reduction of 2,396 tons per day or 4.7%. 60,000 ¦ gas trickle ~ diesel bus ¦ diesel jeepney ~ diesel truck ~ diesel car ¦ gas jeepney ~ gas car Scenarios 1 emissions from exhaust only Combi = IVMS^TDIVH-CMBBI-H-CMEJI-H-BVyVIIVH-RW Figure 22. Projected C02 Emissions: 2015 Following are other significant results of the calculation of C02 emissions for 2010: It is noteworthy that the greatest reduction (5,140 tons per day) in the MVIS scenario comes from diesel buses because of the associated assumption 84 ------- that all buses will be running on CNG. By contrast the CNG bus scenario for the low case (CNGBL) involved a displacement of only 9% of diesel buses and thus registered a corresponding reduction of only 1,693 tons per day or 3.3 % of the total emissions in the BAU scenario. For the high case (CNGBH) the assumed displacement was doubled to 18% but the corresponding reduction was still only 1,821 tons per day or 3.6 %. In the RAIL scenario, the greatest reduction comes from the private transport modes particularly from diesel truck (2,246 tons per day), gasoline car (1,197 tons per day) and diesel car (1,169 tons per day). 5.3. Air Dispersion Modeling Results Calculated ambient PM concentrations for each scenario were averaged by municipality and are presented in Figures 23 to 28. Each box in these figures show either how various policies compare with business-as-usual or how some policies compare with each other. Contributions of industrial emissions are included, but are assumed to remain the same, unaffected by population growth or the transport scenario. Figure 23 indicates how the air pollutant levels are expected to rise due to project growth if no measures are taken to actively reduce emissions from the transport sector. The city of Manila appears to receive the brunt of both the concentrations of PM as well as the growth of its levels, particularly between 2005 and 2010. By contrast, Valenzuela and Navotas appear to have lower particulate levels and minimal escalation. It should be noted, however, that while these two locations have, on average, lower local PM levels, they also host the most number of stationary sources can cause concentrations to be very high in small specific locations. ¦ NCR-All . Kalookari Las Pinas Malabon ¦ Mandaluyong . Makati ¦ Manila ¦ Marikina Muntinlupa Navotas Pasay Paranaque Pasig Pateros Quezon City San Juan ¦Taguig Valenzuela Figure 23. Mean annual PM concentrations (|ig/Ncm) per city/municipality and Metro Manila arising from the business-as-usual (BAU) scenario. Motor Vehicle Inspection System (MVIS) can improve air quality in all municipalities and therefore the whole Metro Manila (upper right of Figure 24). ------- From the present up to 2010 the levels under MVIS is about 20 percent of BAU levels. Remarkably, after 2010 the MVIS can actually cause air pollution levels to start decreasing as a result of the phaseout of the most pollutive vehicles and their replacement by new, cleaner units. The impact of fuel shift from diesel to compressed natural gas (CNG) for public buses and the use of coco-methyl esters (CME) in jeepneys are presented in separate graphs in Figure 24. Taken separately, the two measures each give rise to a reduction in ambient levels by about 10 percent compared to BAU. The degree of improvement is of course influenced by how dependent a municipality on the mode of transportation. For instance, Manila appears to benefit from both measures, but not Marikina. This finding implies a need for each local government to select its own strategy for cleaning up the air. However, the more important result is that a shift to cleaner fuel will not reverse the rising trend in PM levels unless a higher percentage of vehicles currently on the road is targeted for conversion. 30 MVIS 40 CNG for Buses 40 35 30 I 25 S 6> E 20 o 5 15 10 5 CME for Jeepneys ~ ~~ -*~«~~ ¦» NCR-All Navotas U Kalookan Pasay Las Pinas Paranaque Malabon Paslg w Mandaluyong Pateros » Makati Quezon City i Manila San Juan Marikina Taguig m Valenzuela Muntinlupa Figure 24. Mean annual PM concentrations (|ig/Ncm) per city/municipality and for Metro Manila arising from the implementation of the Motor Vehicle Inspection System (MVIS, top left), the use of compressed natural gas (CNG) for buses (top right) and coco-methyl esters (CME) for jeepneys. 86 ------- Bikeways have been put up in Marikina and are being proposed to be buiit in the whole Metro Manila, Unfortunately, the measure appears to result in very minimal improvement even in Marikina, as seen in Figure 25. The reason is that Marikina is inherently clean to begin with, thus leaving little room for improvement. Bikeways - Metro Manila Bikeways - Marikina E o 3 ¦ NCR-All . Kalookan Las Pinas Malabo ri ¦ Mandaluyong ¦ Makatl ¦ Manila . Marikina Muntinlupa Navotas Pasay Paranaque Pasig Pateros Quezon City San Juan Taguig Valenzuela Figure 25. Mean annual PM concentrations (pg/Ncm) per city/municipality and for Metro Manila arising from the installation of bikeways in Metro Manila and Marikina. For the rest of Metro Manila, the public is not expected to shift to bikes in significant numbers that could result in reduced vehicular demand. The impact of a new railway expected to be completed in 2015 shown in Figure 26 is a significant absolute decrease in PM levels for Metro Manila on average. However, not all municipalities will benefit. Levels in the city of Manila, for one, which will likely be served heavily by the railway, will exhibit a major improvement, but those in Las Pinas are still likely to rise with the population. 87 ------- NCR-All Navotas Kalookan Pasay Las Pinas Paranaque Malabon Pasig Mandaluyong Pateros Makati Quezon City Manila San Juan Marikina Taguig Muntinlupa Valenzuela Figure 26. Mean annual PM concentrations (|ig/Ncm) per city/municipality and for Metro Manila arising from the business-as-usual (BAU) scenario and the projected operation of the proposed railway in 2015. The installation of particulate traps for diesel engines reduces ambient levels by a slightly lesser degree, but the improvement over the BAU case is less than 1 [jg/Ncm (Figure 27). A clearer reduction may be seen if both buses and jeepneys are fitted with particulate traps since the traffic attributed to the latter are higher than those for buses. NCR-All m Kalookan Las Pinas Malabo n w Mandaluyong ~ Makati i Manila ¦ Mari kin a » Muntinlupa Navotas Pasay Paranaque Paslg Pateros Quezon City San Juan --Taguig Valenzuela Figure 27. Mean annual PM concentrations (pg/Ncm) per city/municipality and for Metro Manila arising from the installation of diesel particulate traps for buses only (left), and particulate traps for buses and jeepneys. 88 ------- Conversion to 4-stroke Tricycles 5 B Kalookan Las Pinas Malaban w Mandaluyong a Mahati l Manila m Marl kin a m Muntinlupa Navotas Pasay Paranaque Pasig Pateros Quezon City San Juan -- Taguig b Valenzuela Figure 28. Mean annual PM concentrations (|ig/Ncm) per city/municipality and for Metro Manila arising from the conversion of tricycles to four-stroke engines (left) and the continued implementation of Traffic Demand Management. The impact of other measures, such as the continued implementation of Traffic Demand Management (TDM), also known as the "color-coding" or vehicle reduction scheme, is also a 10 percent decrease for the whole Metro Manila (Figure 28). But the best improvement among the options may be expected from the replacement of two-stroke public tricycles with four-stroke units, a measure that is nearly as effective as the MVIS. Applying the combination of measures described earlier to address Metro Manila's air quality is forecast to cause a dramatic improvement in PM levels (Figure 29). In 2005, a 25 percent reduction is immediately expected. But more important, the increase in pollution levels is minimized all the way to 2015 where by this average levels in Metro Manila fall to about 40 percent of baseline levels. This scenario, which is based on realistic estimates in the application of proposed air quality measures, reiterates an earlier prediction that PM levels in the capital can be controlled through concerted effort. 89 ------- Combination - 1 Combination - 3 NCR-All Kalookan Las Pinas Malabo n *t Mandaluyong » Makati i Manila ¦ Marikina » Muntinlupa Navotas Pasay Paranaque Pasig Pateros Quezon City San Juan - Taguig Valenzuela Figure 29. Mean annual PM concentrations (|ig/Ncm) per city/municipality and for Metro Manila arising from a combination of all proposed measures excluding railways and tricycle engine conversion (top left) and the implementation of all proposed measures. 5.4. Health Impact of Each Scenario As reviewed in the methodology section, there are a number of health effects caused by different air pollutants. In this analysis of the health impact of air pollution particularly PM10, based on different policy scenarios, only 8 health outcomes were considered. Choice of such health outcomes depended on the availability of the health data and the exposure-response coefficients. Thus, in the interpretation of the following results, it is important to keep in mind that these are conservative estimates and does not reflect the total picture of the health impact of air pollution. However, this estimation exercise gives an idea of the extent and severity of the health damages that could be caused by the local air pollution and those that could be averted by different policy scenarios. 90 ------- Table 28 Number of Cases averted by different policy scenarios in 2005 Health Effect MVIS CNG CME Diesel Bike TC4 TDM Combo Traps Ways Stroke 1 Natural Mortality 88 1 1 46 2 132 25 166 (66-110) (1-2) (34-57) (2-3) (99-165) (19-31) (125-208) Respiratory Hospital 28 0 0 15 1 42 8 53 Admissions, All ages (2-54) (0-1) (0-1) (1-28) (0-1) (3-80) (1-15) (4-101) Cardiovascular Hospital 5 0 0 3 0 8 1 10 Admissions, All ages (3-8) (1-4) (4-11) (1-2) (5-14) Asthma Attacks, 4748 52 78 2465 130 7109 1349 8951 <15 y/o (2938-6633) (32-73) (48-109) (1525-3443) (80-181 )(4399-9931 )(835-1885)(5538-12505) Asthma Attacks, 633 7 10 328 17 947 180 1193 =/>15 y/o (311-948) (3-10) (5-16) (162-492) (9-26) (466-1419) (88-269) (587-1787) Bronchitis Episodes, 162 2 3 84 4 243 46 306 <15 (77-247) (1-3) (1-4) (40-128) (2-7) (115-370) (22-70) (145-466) Chronic Bronchitis, 1299 14 21 674 35 1945 369 2448 Adults (124-2463) (1-27) (20-40) (646-1279) (3-67) (186-3688) (35-700) (235-4644) Estimates of health damages that could be averted were calculated for eleven policy scenarios including three combinations of policies applied in different time periods. The following tables show the results of this estimation exercise. Each table reflects the year by which the respective policies would be implemented. Thus, each table shows 'what-if scenarios for each health outcome. Apart from the combination scenarios, the policies which could avert the most number of cases in all the years presented, are the conversion of tricycles from two stroke to four-stroke and the Maintenance of Vehicles and Inspection System. The Maintenance of Vehicles and Inspection System policy assumes that through an emission testing system, the quality of vehicles would be maintained and emissions will be decreased. This system depends entirely on the sustainability and consistency of implementation. In 2005, the Diesel traps and the Traffic Density Management policies could also contribute to reduction of cases but to a lesser degree as compared to the former two policies mentioned. The Traffic Density Management includes the numbers coding scheme primarily. In 2010, the CNG and CME policies contribute modestly to averting the number of cases due to air pollution, and in 2005, their would-be impact are negligible. The contributions of the CME and CNG policies are quite low mainly because the assumptions regarding 91 ------- their utilization in the respective scenarios are also quite low as compared to that of the policy on the conversion of the tricycles from 2 stroke to 4-stroke. The assumptions used for the CME and CNG policies, as mentioned in the scenario development section, are based on the government targets. Perhaps higher targets for the use of CNG and CME could be set in order to have maximum benefits. The least of effects are seen with the construction of bikeways. Table 29 Number of Cases averted by Different Policy Scenarios in 2010 Health Effect MVIS CNG CME TC4 Stroke Combo 2 Natural Mortality 172 (129-215) 68 (51-84) 64 (48-80) 341 (256-426) 416 (312-520) Respiratory Hospital Admissions, AH ages 55 (4-104) 21 (2-41) 20 (2-39) 108 (8-207) 132 (10-253) Cardiovascular Hospital Admissions, AH ages 10 (5-15) 4 (2-6) 4 (2-5) 20 (11-29) 24 (13-35) Asthma Attacks, <15 y/o 9245 (5720-12914) 3638 3446 18380 (2251-5082) (2132-4814) (11372-25676) 22400 (13859-31293) Asthma Attacks, =/>15 y/o 1232 (606-1846) 485 (238-726) 459 (226-688) 2449 (1205-3670) 2985 (1468-4472) Bronchitis Episodes, <15 316 (150-481) 124 (59-189) 118 (56-179) 628 (298-956) 765 (363-1165) Chronic Bronchitis, Adults 2529 (242-4796) 995 (95-1887) 943 (90-1788) 5027 (482-9535) 6127 (587-11620) In the 2015 scenario, the implementation of the utilization of the railways or the metro is a major player in averting the number of cases. Expectedly, the combination of policies scenarios in all the years show the most number of cases averted when implemented. The most number of cases are averted in 2015 with combo 3, wherein, apart from the implementation of the other measures, the railways or metro system are also supposed to be in full operation. In summary, the combinations of policies scenarios are still ideal and yield the most health gains. Moreover, this approach of estimating the health gains of individual policies also illustrates the importance of each policy. This would help decision and policy makers to appraise the merits of each policy especially in the event that due to budget constraints, only one or a few of the measures could be implemented. 92 ------- Table 30. Number of Cases Averted by Different Policy Scenarios in 2015 Health Effect MVIS Railways Bikeways Combo 3 Natural Mortality 274 (205-342) 173 (130-217) 10 (7-12) 590 (443-738) Respiratory Hospital Admissions, AH ages 87 (7-166) 55 (4-105) 3 (0-6) 188 (15-359) Cardiovascular Hospital Admissions, AH ages 16 (9-23) 10 (5-15) 1 (0-1) 34 (18-50) Asthma Attacks, <15 y/o 14733 (9116-20581) 9341 (5780-13050) 519 (321-725) 31801 (19676-44425) Asthma Attacks, =/>15 y/o 1963 (966-2942) 1245 (612-1865) 69 (34-104) 4238 (2085-6349) Bronchitis Episodes, <15 503 (239-767) 319 (151-486) 18 (8-27) 1086 (515-1654) Chronic Bronchitis, Adults 4030 (386-7643) 2555 (2449-4846) 1420 (136-269) 8698 (834-16497) Finally, Table 31 summarizes the cumulative health impact of the single policy scenarios and the combination 1 scenario if implemented in 2005 until 2015. Combination 1, as mentioned earlier, does not include the replacement of 2- stroke tricycles to 4 -stroke and the railways policies. In this analysis, the implementation of combination 2 which includes all policies except the railways starts in 2010 until 2015 and combination 3 with all the policies is implemented only in 2015. Table 31. Cumulative Number of Cases Averted byttie policy scenarios from year of implementation to 2015. (Note: Combo 1: from 2005-2015, Combo 2: torn 2010-2015 and Comb© 3: Only 2015.) Policy Scenario Natutal Mortally Respiratory Hospital Admission Cardiovascular Hosptal Admissions Asthma Attacfe <15 Asthma Attacfe >15 Bronchitis Episodes <15 Chronic Bronchitis MVIS 1899 603 111 102354 13639 3495 27997 (1424-2376) (47-1154) (60-161) (63328-142988) (6710-20437) (1658-5323) (2683-53099) CNG 614 195 35 33117 4414 1130 9058 (461-768) (15-373) (19-52) P349148265) (2171-6613) (537-1725) (867-17181) CME 585 185 34 31509 4199 1077 8618 (439-730) (15-357) (1849) (1949544017) (2065-6292) (510-1640) (827-16343) 93 ------- RAIL- WAYS 173 (130-217) (4-105) 10 (5-15) 9341 (5780-13050) 1245 (612-1865) 319 (151-486) 2555 (24494846) Diesel Traps 1203 (902-1505) 382 (29-732) 69 (37-101) 64859 (40131-90608) 8643 (4252-12951) 2215 (1051-3375) 17739 (1700-33646) Bike MM 51 (37-61) 17 (0-32) 1 (0-3) 2690 (1665-3759) 359 (175-538) 94 (42-139) 736 (73-1397) TC 48trake 2082 (1562-2603) 660 (52-1266) 121 (65-175) 112156 (69394-156683) 14946 (7352-22393) 3829 (1817-5834) 30677 (2939-58182) TDM 860 (645-1076) 275 (22-524) 49 (27-71) 46370 (2869064778) 6177 (3039-9258) 1584 (750-2413) 12581 (1216-24054) Combol 2827 (2120-3535) 898 (70-1719) 163 (89-239) 152325 (94247-212800) 20299 (9986-30413) 5200 (2465-7924) 41667 (3993-79021) Combo2 2643 (1982-3303) 840 (64-1605) 153 (23-224) 142376 (88092-198900) 18972 (9335-28428) 4861 (2307-7408) 38944 (3733-73860) ComboS 590 (443-738) 188 (15-359) 34 (18-50) 31801 (1967644425) 4238 (20856349) 1086 (515-1654) 8698 (834-16497) As expected, the cumulative health impact of the policy scenarios is considerably greater than any of the previously-defined single year scenarios. However, the policy scenarios which have the most health impact remain the same except for the railways and combination 3 policy scenarios. The railways and combination 3 policy scenarios are implemented only in a single year. The importance of the impact of these two policy scenarios are better expressed in the comparison of the results seen in Tables 28-30. These cumulative estimates give a more accurate representation of the health impact of the different policy scenarios. 5.5. Economic Costs of the Health Damages Averted by Each Policy Scenario From the health damages section, economic costs of each policy scenario are calculated. The methodology of such computations and its uncertainties are outlined in section 4.3. Figure 30 presents the percent contribution of the costs of each averted health effect to the total cost of individual policy scenario. It shows that the total costs of the different measures or policy scenarios are dominated by the costs of the averted deaths and the morbidity due to chronic bronchitis. The premature mortality account for about 50% of the total cost of the policy scenarios while chronic bronchitis account for about 46% of the total. This occurrence is expected since cost per case of these two 94 ------- health effects is also quite high. In addition, this is consistent with another IES project result, e.g. Santiago, Chile. Table 32 shows the sum of costs of the health damages that could be averted by each policy scenario in different years as well as the total amount for some policy scenarios for the 3 years. As evident in this costing, the total amount of doing a combination of the policies is the most advantageous of all. Table 32 Costs of the Health Damages per Policy Scenario per Year in Million Pesos, Philippines Policy Scenario 2005 2010 2015 MVIS 850 939 849 CNG Bus 9.29 370 NC CME Jeep 13.9 350 NC Railway NC NC 538 Diesel Trap 441 NC NC Bikeway 23.2 NC 29.9 TC4 Stroke 1,270 1,870 NC TDM 241 NC NC Combo 1 1,600 NC NC Combo 2 NC 2,280 NC Combo 3 NC NC 1,830 'present exchange rate 1 $=55.38 PhP NC : Not Calculated (Policies Were calculated independently per year. Cumulative calculations are in the next table These combinations of policies will save at least 1.6 billion pesos. This amount could even be bigger in the long term as shown in the 2010 and 2015. The major contributors to these costs as seen earlier in the health damages section are the conversion of tricycles from 2 stroke to 4 stroke, the MVIS and the Railways scenarios. The policy on the conversion of tricycles to 4-stroke alone could save up to 1.2 billion pesos in 2005 and 1.8 billion pesos in 2010. With the MVIS, the benefits could also be quite considerable at 850 to almost 940 million pesos. Implementation of the railways alone by 2015 would save up to more than 500 million pesos in health damages. As mentioned also in the health damages section, these are conservative estimates since these cost estimates do not cover all the health damages that are caused by particulate air pollution. 95 ------- In Table 33, the cost of cumulative health impact is shown. These figures are expectedly much bigger than those in Table 32. As in the description of the results in Table 32, however, the policy scenarios with the largest costs averted in Table 33 are the combination scenarios, the MVIS and the replacement of 2-stroke to 4 stroke tricycles. In addition, it should be noted that, in spite of the low targets for the CNG and CME scenarios, the cumulative costs averted are still quite staggering at more than 3 billion pesos each. Despite the large figures seen here, these estimates remain as conservative, since they do not cover all the health damages that caused by particulate air pollution. Table 33: Cumulative Total Health Costs Averted of each Policy Scenario, in millions, 2003 Policy Scenario Cumulative Total Cost Averted, in millions MVIS 12,126 CNG Buses 3,705 CME Jeepneys 3,523 Railway 538 Diesel Traps 7,812 Bikeways in Metro Manila 304 Tricycle Replacement to 4-stroke 14,141 TDM 5,468 Combo 1 19,083 Combo 2 13,359 Combo 3 2,813 ~ Natural Mortality NCR ~ Respiratory hospital admissions NCR ~ Cardiovascular hospital admissions ~ Asthma attacks <15 ¦ Asthma attacks adults ~ Bronchitis episodes <15 ¦ chronic Bronchitis Figure 30. Percent Contribution of the Cost of Each Health Effect to the Total Costs 96 ------- However, a caveat must be observed in interpreting these figures since there are a number of uncertainties. For example, due to the lack of local data, a benefits transfers method is used for some of the costs of the health effects. The benefits transfer method, as has been discussed, assumes that the figures from another study or country are transferable to Metro Manila. These foreign figures were adapted and 'localized' accordingly for it to be more utilizable in Metro Manila. These uncertainties are inevitable because of the nature of data needed for this type of estimation. Using benefits transfer is far from ideal. In this study, benefits transfer was used as a convenient quick method. We simply took average values of a Philippine study which had transferred values from a U.S. study. A better way which could be explored is to transfer benefit functions and not average values. Transferring benefit functions would be more accurate. The need to improve the quality of unit economic values underlines the importance of primary valuation studies to estimate willingness to pay (WTP) to avoid occurrence of specific illnesses and unit values for mortality. It would vastly improve the study if contingent valuation methods were used to estimate unit values for morbidity, i.e., WTP for avoiding certain diseases in Metro Manila. For unit values of mortality, hedonic wage analysis would be required to estimate the value of a statistical life in Metro Manila. The value of a statistical life is derived from an analysis of how much more workers are paid for riskier jobs. Different occupations involve different risks, and it is reasonable to expect that employers need to pay a wage premium to induce workers to undertake jobs involving higher risks. If we could measure this premium, we could obtain an estimate of the market value of small changes in fatality risk. In studies of wage-risk tradeoffs, the marginal wage with respect to fatality risk has been called the value of a statistical life. The value of a statistical life is the ratio of the increase in the wage rate to the increase in the risk of fatality that induced the increase in wages. More analysis could also be applied to the data available in Philhealth records for estimating medical costs. Comparison of average costs of different years could be considered to arrive at low, central and high estimates. The payments could be adjusted upwards to account for not only payments made by the insurance scheme but full medical costs. The indirect cost of illness (lost work days) could be improved with sufficient and reliable data on off-sick days due to specific illnesses. It would be interesting to check if such labor statistics were available. 97 ------- 5.6. Co-Benefits Results of Each Policy Scenario in terms of Emissions Reduction. An evaluation of the policy scenarios in the reduction of emissions of particulates and the greenhouse gas, C02, is presented here. Table 30 shows that the replacement of 2-stroke to 4-stroke tricycles and the diesel particulate traps do not have an impact in the reduction of C02 emissions but had significant reduction in particulate emissions. In all the other policy scenarios, the reduction in particulates emissions approximates the reduction in C02. Minimal impact is also seen with the CNG for buses and CME for jeepneys scenarios. Of the six single policies, the MVIS and the railways have the most impact in mitigating both the particulates and the C02. As in the health damages and economic impact, the best results for mitigating both particulates and C02 are seen with the combination of policies. This comparison would be of additional assistance to policy makers in determining which policy or combination of individual policies would have the most impact for both local air pollution and GHG emissions. Table 34 % reduction from BAU for each year Policy Scenarios 2005 2010 2015 SPM C02 SPM C02 SPM C02 MVIS 12.5 14.7 21.0 21.3 29.0 30.0 CNG for buses 0.6 0.04 8.6 0.6 9.8 3.6 CME for jeepneys 0.1 0.1 8.6 0.4 9.8 3.0 TDM 4.1 4.8 10.0 4.6 13.0 4.7 Replacement of 2-stroke tricycles 24.0 0.0 27.5 0.0 28.0 0.0 Bikeways 0.5 0.1 0.5 0.1 1.6 0.1 Diesel Particulate Trap for buses 2.0 0.0 7.7 0.0 11.4 0.0 Diesel Particulate Trap for jeeps/buses 7.0 0.0 32.7 0.0 34.0 0.0 Railway NC NC NC NC 18.2 13.0 Combination 1 27.0 31.6 NC NC NC NC Combination 2 NC NC 57.1 32.0 NC NC Combination 3 NC NC NC NC 69.0 53.0 NC : Not Calculated (The policies were not calculated independently per year) Under the Kyoto Protocol of the United Nations Framework Convention on Climate Change (UNFCCC), industrialized countries and economies in transition (Annex B countries) have legally binding emission reduction targets and collectively are to reduce their emissions by 5.2% below 1990 levels over the commitment period 2008-2012. The Protocol comes into force upon ratification by at least 55 parties whose C02 emissions represent at least 55% of the total from Annex I Parties. At the present, 150 countries have 98 ------- ratified the Protocol, with 61.6% of the emissions represented, despite the USA decision to pullout of the international agreement. Last year, the global arena banked on the ratification of Russia which represented 17.4% of the emissions. Upon Russia's ratification last 18 November 2004, it paved the way for the Protocol's entry into force last 16 February 2005. In complying with the Protocol, flexibility mechanisms have been set to assist developed countries achieve their C02 reduction targets. The Clean Development Mechanism (CDM) is the only mechanism by which developing countries such as the Philippines can participate in GHG mitigation under the Protocol. Developing countries benefit from the CDM projects which result in "certified emission reductions" (CERs) that developed countries can use to comply with their quantified emissions reduction commitments under the Protocol. Such CDM projects should likewise help in achieving sustainable development in the host developing country. Under a recent report and study, "State and Trends of the Carbon Market 2004" by Franck Lecocq, Development Economics Research Group, World Bank6 based on data and insights provided by Natsource LLC and PointCarbon, his findings included the following: The carbon market is growing steadily. A total of 64 million metric tonnes of carbon dioxide equivalent (tC02e) has been exchanged through projects from January to May 2004, nearly as much as during the whole year 2003 (78 million), which suggests that the market might double by the end of the year. The vast majority of this volume is from project-based transactions intended for compliance with the Kyoto Protocol. Beyond the market data reported in this document, there is a clear sense of momentum in the market, notably in Asia and Latin America. The approval of some methodologies by the CDM Executive Board clearly reinforces these dynamics. HFC23 destruction represent nearly a third of the volume supplied in 2003- 2004from only two project sites.7 Landfill gas to energy projects are the second largest suppliers of emission reductions, with 18%. This is a dramatic change compared to last year, which underlines the important potential of this technology, which has low mitigation costs, an approved and published methodology, and a large total supply per site. 6 The findings and opinions expressed in this paper are the sole responsibility of the author. They do not necessarily reflect the views of the World Bank, its executive directors, or the countries they represent; nor do they necessarily reflect the views of the World Bank Carbon Finance Business Team, or of any of the participants in the Carbon Funds managed by the World Bank. They do not necessarily represent the views and opinions of the two contributors: Natsource LLC and PointCarbon. 7 HFC23 is a by-product of the production of HFC22, which is used as a refrigerant and as a raw material for the production of fluorinated resins. HFC23 is a very potent greenhouse gas. The release of one ton of HFC23 in the atmosphere has the same long-term effect on climate change than the release of 11,700 times tonnes of C02. 99 ------- The demand remains heavily concentrated. Japanese companies now represent the single largest group of buyers in the carbon market, before the World Bank Carbon Finance Business and the Government of the Netherlands. These three groups of buyers account for nearly 90% of the demand in 2003-2004. Asia is now the largest supplier of emission reductions, followed by Latin America, developed economies, and Eastern Europe. Five countries (India, Brazil, Chile, Indonesia and Romania) represent two thirds of the supply in terms of volume. Increased concentration of CDM flows to a limited number of countries continue to leave Africa essentially bypassed, raising concerns about the long-term distribution of the benefits of the Clean Development Mechanism. The fact that the potential HFC23 destruction projects are heavily concentrated in a handful of countries reinforces this concern. Prices of project-based emission reductions in early 2004 have remained essentially stable compared with 2003, and depend strongly on the segment of the market, and on the structure of the transaction. They reflect the distribution of risks between buyer and seller. For example, assuming the risk that the ERs might ultimately not be registered under the Clean Development Mechanism or Joint Implementation commands a significant premium. End of 2003 and early 2004 have seen major regulatory breakthroughs that also contribute to the sense of momentum in the markets. Chief among them is the approval of the EU Emission Trading Schemes (EU-ETS) which, along with the "Linking Directive" connecting it to the world of project-based opportunities, creates a welcome and appropriate structure for managing and pricing carbon. The EU-ETS sends a strong signal to the markets and can be a driver of additional volume of carbon trades achieving climate mitigation in the future. At the time of writing, most EU Member States had published their National Allocation Plans for the scheme's pilot phase (2005-2007) which were less stringent than expected. This has apparently contributed, along with the approval of the Linking Directive, to a significant drop in the price of European allowances on the early market, although the early market for EU Allowances is still very thin and does not a priori reflect long term equilibrium between supply and demand. While these developments send the right signal to the markets, interested parties should note that, absent clarification of the validity of project-based emission reductions beyond 2012, the window of opportunity for initiating project-based transactions is rapidly closing. Given the long lead time between project preparation and the first "yield" of ERs, project developers have only a couple of years to act before carbon payments for only a few years cease making a meaningful contribution to project finance. 100 ------- Taking into consideration the results of the said report and applying them to the current study, projects that would revolve around MVIS and the railways or involving a combination of policies are promising prospects for CDM, as well as in mitigating both particulates and C02. The carbon market is growing and the Philippines is poised to have a share of supplier market. The CDM Executive Board has already set out the "Indicative Simplified Baseline And Monitoring Methodologies For Selected Small-Scale CDM Project Activity Categories" which includes low-greenhouse gas emitting vehicles. A project activity in this category shall both reduce anthropogenic emissions by sources and directly emit less than 15 kilotonnes of carbon dioxide equivalent annually. However, there are a number of challenges that face projects under the CDM. The first commitment period of 2008-2012 is fast approaching and negotiations in the international arena have not definitively tackled the prospect of a second commitment period. Hence, projects have a tendency to cover only the first period. Further, the high transaction costs for developing a project and getting the reductions verified and certified have had some deterring effect on project developers. Some have indicated that guidelines and processes for obtaining the national approval of the Designated National Authority are still unclear. However, the entry into force of the Protocol has indeed had an impact on the prices of per ton of carbon to be traded in the carbon market. Increasing activity in the carbon market has indicated that prices will stabilize and become robust, once more of the uncertainties are taken cared of. Yet, despite the Kyoto Protocol's entry into force, there are opinions that the aim of stabilizing levels of greenhouse gases in the atmosphere to avert dangerous climate change will not be realized because of the fact that the biggest emitter of greenhouse gases continues to remain on the sidelines. This matter will be left with the global community to tackle and debate in the years to come. 101 ------- Conclusions and Recommendations 6.1. General Conclusions and Recommendations The Philippine government's expenditure or budget on health every year is about 3.4% of the country's Gross National Product (GNP) (2000 World Development Report) and about 3.5 to 4% of the National Budget. In 2003, this amounted to about 9.281 billion pesos and will steadily increase in the years to come. It is forecasted to increase to about 9.956 billion pesos in year 2005, an increase of about 7.2% from 2003. If this trend continues, it could be as much as 11.853 billion pesos in 2010 and 14.094 billion pesos in 2015. The savings from the health impact projected in 2005, 2010 and 2015 in implementing the single or combination of policies could mean very substantial savings on the health budget. In 2005, the savings from implementing the policy with the least health impact to the combination of policies would range from 0.13% to about 16% of the health expenditures for that year. In 2010, savings go up to almost 3% for the least, to more than 19% of the health budget, for the combination. In 2015, it's from about 0.21% to 13% of the health budget. The savings in the national health budget that could be incurred is only based on savings from Metro Manila. If these policies could be implemented in the secondary cities, the savings on the national health budget could even be more. The amount to be saved may be used by the Health Department in other programs such as immunization, nutrition, provision of toilets and other preventive programs. The Education Department may also benefit by building more schools and classrooms or providing the much needed textbooks. The outcome of such transport policies in due time could be enormously beneficial for the whole country. However, as noted a few times in the report, there is a need to evaluate this health impact together with the cost of implementing each policy and combination thereof for a more comprehensive assessment as will be discussed further in the recommendations section. The following are the specific conclusions: 1. Based on the assumptions made in the scenario development, three single policies have the advantage of having more health and economic benefits. These are the implementation of the maintenance of vehicle and inspection system, switching from four-stroke to two stroke tricycles, and use of the metro railways. These three policies must be seriously considered by decision makers particularly the Department of Transportation and Communication (DOTC) and the Metro Manila Development Authority (MM DA). 102 ------- 2. The use of CNG in buses and CME among jeepneys did not show very important benefits because the assumptions on the targets, which were based on the government plans, were too low for any significant impact. The Department of Energy (DOE) must exert extra effort to increase its target with regards the CNG and CME buses and jeepneys to have a more meaningful impact on air pollution. 3. C02 emissions can also be considerably reduced with the policies proposed specially the MVIS and the TDM. However, at most benefits in terms of reduction of both PM and C02 can be seen with the MVIS and the railways policy scenarios. 4. The other single policy scenarios also contributed to the reduction of air pollution and resulted to some health and economic benefits. These single policy scenarios are the collective responsibility of the DOTC, DOE, MMDA and the DENR 5. As expected, the combinations of policies resulted to the most health and economic benefits as well as reduction of C02. Hence, if at all possible, these combinations of policies must be implemented. Again, several government agencies are responsible to implement the combination of policies. However, the DOTC must take lead as most of the policies are related to transport. 6.2. Policy Implications As earlier enumerated, the policy scenarios involved in this Study are the following: Motor Vehicle Inspection System (2005, 2010,2015) Conversion to 4-stroke tricycles (2005,2010) Railways (2015) Diesel Particulate Trap(2005) Compressed Natural Gas - Buses (2005,2010) CocoMethyl Ester - Jeepneys (2005,2010) Traffic Demand Management (2005) Bikeways (2005) Combo-1: MVIS+TDM+CNG+CME+DPT+Bikeways (2005) Combo-2: MVIS+TDM+CNG+CME+DPT+Bikeways+TC4stroke (2010) Combo-3:MVIS+TDM+CNG+CME+DPT+Bikeways+TC4stroke+Railways (2015) Flowing from the health impact and economic valuation, in order to reduce PM10 levels in Metro Manila by half, the best scenario to implement is a 103 ------- combination of policies. In particular, for the immediate term (2005), Combo-1 will produce the highest economic value with the most number of health cases averted. In the mid-term (2010), it is Combo-2 that will the best set of policies to be implemented. On the other hand, for the long-term (2015), the full implementation of the Motor Vehicle Inspection System (MVIS) will result in the most benefit. This is because the MVIS can offset the growth air pollutants such as PM10 since cleaner vehicles will then replace old, pollutive units. While it is also espoused that the use of cleaner fuels should likewise be pushed, the study showed that improvements resulting from use of CNG and CME will be minor unless the targeted number of public vehicles converted are increased. Based on our assumptions, the high estimate targeted is 3000 buses (2010) and 20000 jeepneys (2010). Of the scenarios enumerated, at present, the government is implementing a traffic demand scheme and components of Motor Vehicle Inspection System. However, there are still no clear-cut policies for a number of scenarios presented. For example, a policy of outright requiring the conversion of two- stroke tricycles to four-stroke has not been made by the government; although certain local government units have undertaken to require this in their jurisdiction. In Metro Manila, the issue is very tricky as the government found out last year when the tricycle sector conducted a rally, protesting the government's apparent singling-out of their sector as a key contributor to air pollution in the city. In response, the government engaged in dialogue with the sector to discuss ways to address their concerns. Looking at the short-term scenario for 2005, a number of policies have to be firmed up. In particular, in terms of the use of alternative fuels and additives, it is apparent that the market has to be made ready and friendly for business and the consuming public to favor their use over-and-above the more pollutive fuels. However, the infrastructure for the distribution of these cleaner fuels has yet to be set up. A clearer direction is needed so that the public can begin using these cleaner fuels. The use of Diesel Particulate Trap (DPT) for buses and jeepneys needs to become an established practice as well. Other physical infrastructure is likewise needed to complete the combination which will produce high benefits. Except for certain areas in Metro Manila, bikeways are more likely cannot be found in the metropolis. 6.3. Other Recommendations Apart from the recommendations based on the policies presented here, there a number of other recommendations which could improve the models used in this project and those that still could be done as a follow-up activity or research. One of these important recommendations is with regards the cost effectiveness of the policy scenarios. The study has estimated the economic costs of health damages from air pollution of each transport scenario. In a 104 ------- further study, the marginal benefit in terms of avoided health damages from abating a unit of air pollutant could be estimated. Compared to the baseline scenario, we will need to calculate the marginal benefit of pollution abatement that occurs in each transport scenario. Then, we could estimate pollution abatement costs of each scenario. Compared to the baseline scenario, we will need to estimate the marginal abatement cost of reducing a unit of pollutant in each transport scenario. We would then be able to check whether there is a positive net marginal benefit from pollution abatement in each scenario. We could also compare the net marginal benefits of different transport scenarios. Such a comparison would guide policymakers on which transport scenarios should be prioritized. At present, the only transport scenario for which total cost estimates are available is the railway scenario. Railway Scenario (2015) Assumptions: The following railways are assumed to be constructed/ in existence by 2015 as indicated in the MMUTIS plan (JICA, 1999) Railway Line Route Distance (km) LRT line 6 South extension of line 1 30 MRT line 3 extension North ave - Navotas; Taft - reclamation 12 MRT line 2 (east west extension) Recto - North harbor; Santolan - Masinag; Masinag - Antipolo 15.7 MRT line 4 Recto-Novaliches 22.8 North rail PNR line 26 MCX PNR line 45.5 Project Cost (US$ million) Fixed Cost Variable Cost Railway Line (Operating and Maintenance) TOTAL LRT Line 6 600 750 1,350 MRT Line 3 extension 306 261 567 MRT Line 2 East West Extension 351 182 533 MRT line 2 288 150 438 MRT Line 4 724 646 1,370 North Rail 589 649 1,238 MCX 554 996 1,550 TOTAL 3,412 3,634 7,046 105 ------- These total cost figures of the railway scenario need to be related to the air pollution abated under this transport scenario compared to the baseline scenario and marginal costs of abatement would have to be estimated. Likewise, the total costs of health damages under the railway scenario estimated in this study would have to be related to the air pollution abated by using railways compared to the baseline scenario. Marginal benefits of abatement under the railway scenario could then be estimated. We could then compare the marginal benefits (avoided health damages) with the marginal costs (costs of railways) of abating air pollution under the railways scenario. To evaluate the cost effectiveness of each transport scenario, similar economic analysis outlined above will need to be done for each scenario. The first necessary step in this direction is to calculate the total costs involved in implementing each transport scenario. Other recommendations: 1. Stringent implementation of the Metropolitan Manila Air Quality Plan 2. Extension of this type of analysis to other sectors and sources of pollution, e.g. stationary sources or industrial air pollution 3. Implementation of similar type of assessment for other cities in the country, e.g. Cebu, Cagayan de Oro, Baguio and Davao. 4. Collection of more reliable data, e.g. meteorological data, and general Improvement of the models used. 106 ------- References Anderson HR, et al. "Air Pollution and Daily Admissions for Chronic Obstructive Pulmonary Disease in 6 European Cities: Results from the APHEA Project." European Respiratory Journal 10: 1064-1071,1997. Asian Development Bank (2003) RETA 5937: Action Plans for Reducing Vehicle Emissions. http://www.adb.orgA/ehicle-Emissions/reta5937.asp Asian Development Bank (1992) Final Report for Vehicle Emission Control Planning in Metro Manila, July, ADB T.A. No. 1414-PHI Chestnut LG, BD Ostro, N.Vichit-Vadakan. "Transferability of Air Pollution Control: Health Benefits Estimates from the United States to Developing Countries: Evidence from the Bangkok Study". American Journal of Agricultural Economics 79(5).1630-1635,1997. Cifuentes, LA et al. "Assessing the Health Benefits of Urban Air Pollution Reductions Associated with Climate Change Mitigation (2000-2020): Santiago, Sao Paolo, Mexico City, and New York City". Environmental Health Perspective 109: supplement 3, 2001. Clean Development Mechanism Executive Board Department of Energy (2001), Feasibility Study on CNG Development for Public Utility Vehicles in Metro Manila Department of Energy (2002), Philippine Energy Plan 2003 - 2012 Dockery DW. et al. "An Association between Air Pollution and Mortality in 6 U.S. Cities" New England Journal of Medicine 329(24).1753-1759,1993. Dockery DW, J .Schwartz, and JD Spengler. "Air Pollution and Daily Mortality: Associations with Particulates and Acid Aerosols". Environmental Research, 59 ,362-373,1993. Eco-Tek Holdings, Limited (2002) Products and Services: Air Pollution, Eco- Trap (Diesel Particulate Trap). Environmental Management Bureau (1993) Air Pollution Emission Inventory for Metro Manila 1990, March Galassi I. et al. "Exposure to PM 10 in the Eight Major Italian Cities and Quantification of the Health Effects". ISEE 2000, New York, USA, August 19- 22,2000. Integrated Environmental Strategies. Co- Controls Benefits Analysis Project for Argentina, USEPA, 2000. 107 ------- Japan International Cooperation Agency (1999) Metro Manila Urban Transportation Integration Study: Final Report, May Japan International Cooperation Agency (1999) Metro Manila Urban Transportation Integration Study Technical Report No. 10: Traffic Environmental Study, Air and Noise Pollution in Metro Manila. March 1999. Japan International Cooperation Agency (1999) MMUTIS: Metro Manila Urban Transportation Integration Study Appendices. March 1999. Japan Society for the Promotion of Science (2002) Metro Manila: In Search of A Sustainable Future - Impact Analysis of Metropolitan Policies for Development and Environmental Conservation, edited by Tatsuo Ohmachi and Emerlinda R. Roman, JSPS Manila Project and the University of the Philippines Press, Philippines Kjellstrom T. "Preliminary Assessment of Human Health Indicators of Air Quality". Technical Paper no.55, Ministry of Environment, Auckland, New Zealand, June 1999. Kunzli N. et al. "Public Health Impact of Outdoor and Traffic-related Air Pollution: A European Assessment". Lancet 356:795-801,2000. Kyoto Protocol of the UNFCCC Land Transportation Office, Philippines (2002) Number of Motor Vehicles Registered, By Region, By Type of Motor Vehicle and Type of Fuel Used (unpublished). Marino, A. (2001) ASEAN-NZ Natural Gas Utilization for Transport Program Philippines Experience, the 6th ASEAN Science and Technology Week, Brunei Darussalam, 17-20 September McMichael AJ (1997) "Transport and Health: Assessing the Risks". In: Fletcher T and McMichael AJ eds. Health at the Crossroads. Transport Policy and Urban Health. New York, John Wiley & Sons. pp. 9-26. National Air Quality Action Plan (http://www.hangin.org/legal/aqap/index_AQAP.htm) National Statistics Office, Demographic Data, Manila, Philippines, 2000. Ostro B.et al. "Air Pollution and Mortality: Results from a Study of Santiago,Chile", Journal of Exposure Analysis and Environmental Epidemiology, 6(1 ),97-114, 1996. Partnership for Clean Air (PCA) Web Site http://www.hangin.org/ Philippine Council for Industry and Energy Research and Development and the U.P. National Center for Transportation Studies (2003) Standards Development for Local Motorcycle/Tricycle Sector 108 ------- Pope C. et al. "Daily Mortality and PM10 Pollution in Utah Valley", Archives of Environmental Health, 47 (3),211-217,1992. Schwartz,J. and D.Dockery, "Increased Mortality in Philadelphia Associated with Daily Air Pollution Concentrations", American Review of Respiratory Disease, 145,600-604,1992. Schwartz,J. "Air Pollution and Daily Mortality in Birmingham, Alabama, American" Journal of Epidemiology, 137 (10), 1136-1147, 1993. Schwartz J. "Air Pollution and Hospital Admissions for Cardiovascular Disease in Tucson", Epidemiology 8:371-377 ,1997. Schwartz J (1994) "Air pollution and Daily Mortality: a Review and Meta- analysis". Environmental Research, 64, 36-52, 1994. Schwela, D. "Exposure to Environmental Chemicals Relevant for Respiratory Hypersensitivity: Global Aspects". Toxicology Letters 86: 131-142, 1996. Shah, J. and Harshadeep, N. (2001) Urban Pollution from Two Stroke Engine Vehicles in Asia: Technical and Policy Options, Regional Workshop on Reduction of Emissions from 2- 3 Wheelers, September 5- 7, 2001, Hanoi, Vietnam. Manila [online]. Available from: http://www.adb.org/Documents/Events/2001/RETA5937/Hanoi/documents/20 _Shah.pdf [Accessed 1st December 2003], Shah, J. J., and Nagpal, T., 1994: Urban Air Quality Management Strategy in Asia (Urbair). World Bank Technical Paper No. 380. Washington D.C. Synerject LLC/Orbital Fluid Technologies and the Colorado State University (2002) Transfer of Clean 2-Stroke Engine Technology to Asian Markets, April, Concept Paper submitted to the USAID Asia Environmental Partnership The World Bank (1997) Urban Air Quality Management Strategy in Asia: Metro Manila Report (World Bank Technical Report No. 380), edited by Jitendrah J .Shah and Tanvi Nagpal, December, International Bank for Reconstruction and Development/The World Bank, Washington, D.C., United States Toman, Michael and Cazorla, Marina, The Clean Development Mechanism: A Primer, Resources for the Future, 29 September 1998, (http://www.weathervane.rff.org/features/feature048.html) United Nations Framework Convention on Climate Change (UNFCCC) U.P. National Center for Transportation Studies Foundation, Inc. (2000) Marikina City Bikeways Feasibility Study-Final Report U.S. Environmental Protection Agency, 2001. Guideline on Air Quality Models: Appendix W, Part 51, 40 CFR Ch. 1. U.S. Environmental Protection Agency, 1995. User's Guide for the Industrial Source Complex (ISC3) Dispersion Models, Volume 2. EPA Publication No. 109 ------- EPA-454/B-95-003b. Environmental Protection Agency, Research Triangle Park, NC. (NTIS Nos. PB PB 95-222758). U.S. Environmental Protection Agency Technology Transfer Network. Clearinghouse for Inventories & Emission Factors. Compilation of Air Pollutant Emission Factors, AP-42, Fifth Edition, Volume I: Stationary Point and Area Sources, http://www.epa.gov/ttn/chief/ap42/index.html (visited July 2003). Whittemore AS and EL Korn. "Asthma and Air Pollution in the Los Angeles Area", American Journal of Public Health 70: 687-696, 1980. World Health Organization. "Air Quality Guidelines for Europe". WHO Regional Publications, European series No. 23 Copenhagen, World Health Organization. World Health Organization. "Guidelines for Concentration and Exposure Response Measurement of Fine and Ultra Fine Particulate Matter for Use in Epidemiological Studies" WHO Regional Publications, European Commission, World Health Organization, 2002. World Health Organization. "Health Impact Assessment of Air Pollution in the Eight Major Italian Cities" WHO Regional Publications, Europe, 1989. World Health Organization and United Nations Environment Program, "Urban Air Pollution in Megacities of the World" Cambridge, MA, Blackwell Publishers, 1992. Xu, X. et al. ,"Air Pollution and Daily Mortality in Residential Areas of Beijing,China", Archives of Environmental Health, 49 (4),216-222, 1994. 110 ------- |