Manual for Calculating Greenhouse Gas Benefits of Global Environment Facility Transportation Projects Prepared by the Institute for Transportation and Development Policy For the Scientific and Technical Advisory Panel of the Global Environment Facility Scientific and Technical Advisory Panel The Scientific and Technical Advisory Panel, administered by UNEP, advises the Global Environment Facility Continual Updates This Manual and the TEEEMP Models are continually updated as more accurate information is contributed to this process. To ensure you are working with the most current files, go online to the following internet address and download the appropriate files for your project. http://www.unep.org/stap/calculatingghgbenefits This document was prepared by the: Institute for Transportation and Development Policy 127 W. 26th Street, 10th Floor New York, NY 10001 USA Project team: Walter Hook, ITDP Michael Replogle, ITDP Colin K. Hughes, ITDP With support from: Clean Air Initiative for Asian Cities Cambridge Systematics, Inc. Edited by Michael A Kinder Layout by Bill Pragluski The Transport Emissions Evaluation Models for Projects (TEEMP) are excel-based models for estimating GHG impacts of transport projects. The TEEMP tools were developed by the Clean Air Initiative for Asian Cities (CAI-Asia) and the Institute for Transportation and Development Policy ( ITDP) for evaluating the emissions impacts of ADB’s transport projects and were modified and extended for GEF projects by ITDP, CAI-Asia and Cambridge Systematics, Inc for the GEF-Scientific and Technical Advisory Panel (STAP). Models were further refined through their application to projects of the ADB and the World Bank. Manual for Calculating Greenhouse Gas Benefits of Global Environment Facility Transportation Projects Prepared by the Institute for Transportation and Development Policy For the Scientific and Technical Advisory Panel of the Global Environment Facility Scientific and Technical Advisory Panel The Scientific and Technical Advisory Panel, administered by UNEP, advises the Global Environment Facility Table of Contents List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv I. Introduction, Concepts, and Definitions . . . . . . . . . . . . . . . . . . . . . 1 GEF: Partnering for Global Environmental Benefit . . . . . . . . . . . . . . . . . . . . . . . . . 1 Why this Manual? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 What Distinguishes the GEF Methodology from other Models for CO2 Accounting? . . . . . . 2 Principal Attributes of the GEF Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 What Is “Direct” GHG Impact in Transportation Sector Projects? . . . . . . . . . . . . . . . . . 3 What Is “Direct Post-Project” GHG Impact of Transportation Sector Projects? . . . . . . . . . 3 What Are “Indirect” GHG Emission Savings of Transportation Sector Projects? . . . . . . . . . 4 What Are Local Co-Benefits and Why Are They Important to Global Benefit? . . . . . . . . . . 5 II. Overview for Applying GEF Tools and Methodologies . . . . . . . . . . 6 TEEMPs: The Core of the GEF Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Broad Assumptions in Applying the GEF Methodology . . . . . . . . . . . . . . . . . . . 7 Required Data for GEF Methodologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Sequence of the GEF Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Lifetime of the Investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Baseline Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Emission Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Calculating Direct Emission Impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Calculating Direct Secondary Impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Calculating Direct Post-project Emission Reduction Effects . . . . . . . . . . . . . . . . . . . 13 Calculating Indirect Impacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Calculating Indirect Impacts—Bottom-up Approach . . . . . . . . . . . . . . . . . . . . . . . 16 Calculating Indirect Impacts—Top-down Approach . . . . . . . . . . . . . . . . . . . . . . . 18 Calculating the Local Co-Benefit of Transportation Projects . . . . . . . . . . . . . . . . . . . 19 III.  Step-by-Step Guide toEstimating The Direct Impacts of Transportation Efficiency (Vehicle, Fuel, Network Efficiency) Projects . . . . . . . . . 20 Before Proceeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Data Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Eco-Driving TEEMP Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Baselines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Calculating Direct Emissions Impact of Transportation Efficiency Projects . . . . . . . . . . . 22 Calculating Indirect GHG Impact in Transportation Efficiency Projects . . . . . . . . . . . . . 23 IV.  Step-by-Step Guide to Estimating Direct Impacts of Rapid Transit and Railway Projects . . . . . . . . . . . . . . . . . . 24 Before Proceeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 BRT TEEMP Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Shortcut Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Full Scenario Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Data Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Projecting Ridership on the New System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Calculating Peak Hour Ridership using Boarding and Alighting or Frequency and Occupancy . . 27 Calculating CO2e Emissions Using the TEEMP Detailed Model . . . . . . . . . . . . . . 27 ii Calculating Greenhouse Gas Benefits Estimating the CO2e Impact of the Project over the No-Project Baseline Scenario . . . . . . 29 GHG Impact of Shifting Passengers to Newer, More Fuel Efficient Buses . . . . . . . . . 30 Impact on Mixed Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 CO2e Generated in the Production of Vehicles . . . . . . . . . . . . . . . . . . . . . . . 30 Construction Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Indirect Effects: Impact of Land Use Changes . . . . . . . . . . . . . . . . . . . . . . . . 30 Special Notes for Calculating Indirect Impacts: Dissemination of Mass Transit Best Practice . 31 Summarizing Total CO2e Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 V.  Step-by-Step Guide to Non-Motorized Transportation Projects (Bicycle & Pedestrian) . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Before Proceeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Estimating Direct GHG Impact for Bike-Sharing Systems with TEEMP Model . . . . . . . . . . 32 Estimating Direct GHG Impact for Pedestrian Improvement Projects with TEEMP Model . . . 32 Data Requirements for Walkability Model Project Scenario: . . . . . . . . . . . . . . . . 33 Estimating Direct GHG Impact for Bikeways Improvement with TEEMP Model . . . . . . . . . 33 Data Requirements of Bikeways TEEMP model . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Sketch Analysis (Short cut) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Full Model (Detailed) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Developing a Baseline for NMT Projects Without TEEMP Model . . . . . . . . . . . . . . . . 35 VI.  Step-by-step Guide for Travel Demand Management Projects . . . . . 36 Before Proceeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Data Requirements: Baseline Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Calculating Direct GHG Impact for Commuter Strategies, Parking Pricing, Pay-As-You-Drive Insurance using TEEMP Modules . . . . . . . . . . . . . . . 37 Commuter Strategies (Employer-based Strategies) . . . . . . . . . . . . . . . . . . . . . . . . 37 1) Employer Support Programs (Transport Support) . . . . . . . . . . . . . . . . . . . 37 2) Telework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3) Compressed Work Week . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4) Commute Strategies (Rideshare/Transit Subsidies) . . . . . . . . . . . . . . . . . . . . 39 Parking, Pricing, and Company Car Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Company Cars: Employer-Provided Vehicles . . . . . . . . . . . . . . . . . . . . . . . . . 40 Parking Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Parking Density (Availability) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Pay-As-You-Drive (PAYD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Calculating Direct Emission Reductions for Other TDM Projects . . . . . . . . . . . . . . . . . 42 VII. Step-By-Step Guide For Comprehensive Regional Transport Initiatives 43 Before Proceeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Describing the Baseline and the GEF Impact Case . . . . . . . . . . . . . . . . . . . . . . . . 43 Calculating Direct Emission Reductions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 VIII. Appendices: TEEMP Model Data Defaults & Sources . . . . . . . . . . 45 Appendix 1: Data Required and Defaults Provided for Eco-Driving Module . . . . . . . . . . . . . . . 45 Appendix 2: Data Required and Defaults Provided for Employer-Based Commuter TDM Strategies . 46 Appendix 3: Data Required and Defaults Provided for PAYD . . . . . . . . . . . . . . . . . . . . . . . 50 Appendix 4: Data Required and Defaults Provided for Employer-Based Commuter TDM Strategies . 51 Appendix 5: Default Values for Various TEEMP Models . . . . . . . . . . . . . . . . . . . . . . . . . . 53 IX. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Calculating Greenhouse Gas Benefits iii List of Abbreviations ASIF Activity-Structure-Intensity-Fuel BU Bottom-up BRT Bus Rapid Transit CDM Clean Development Mechanism CF Causality Factor CO2 Carbon dioxide CO2 eq Carbon dioxide equivalent in global warming potential DP Direct Project DPP Direct post-project GEF Global Environment Facility GHG Greenhouse Gas GWP Global Warming Potential IPCC Intergovernmental Panel on Climate Change kt or ktonnes kilo-tonnes or 103 metric tonnes M&E Monitoring and evaluation MRT Mass Rapid Transit PAD Project Document PM Particulate Matter PPG GEF Project Preparation Grant PIF GEF Project Information Form RF Replication Factor SOV Single Occupancy Vehicle TF Turnover Factor TAR IPCC Third Assessment Report TD Top-down TDM Transportation Demand Management TEEMP Transportation Emissions Evaluation Model for Projects t or tonnes 103 kg or one metric tonne UNFCCC UN Framework Convention on Climate Change VKT Vehicle Kilometers Traveled iv Calculating Greenhouse Gas Benefits I. Introduction, Concepts, and Definitions GEF: Partnering for Global the lessons learned from experience to tailor these methodologies expressly for transportation projects. Environmental Benefit The GEF models are designed to develop ex- The primary purpose of the Global Environment ante estimations of the GHG impacts of transport Facility (GEF) is to generate global environmental interventions (projects) as accurately as possible, benefit. The essential path for achieving this goal is without requiring data so exacting that it discourages the financial support of projects whose completion investment in the sector. The methodology provides delivers substantial, measurable reductions in uniformity in the calculations and assumptions used greenhouse gases (GHG). The more projects that can to estimate the GHG impact over a very diverse array be brought to fruition, the greater is the fulfillment of potential projects. These include projects that: of our purpose, and the more profound is the positive impact on the environment. This effort is a • Improve the efficiency of transportation vehicles collaboration between the GEF and those applicants and fuels; proposing projects designed to yield these benefits. • Improve public and non-motorized transportation This Manual is designed to assist proponents in modes; shaping their projects accurately and responsibly, and presenting them for consideration in consistent, • Price and manage transport systems more quantifiable terms. The GEF is committed not only efficiently; to supporting the national and regional goals of • Train drivers in eco-driving; each group, but to extending, as far as possible, the results of these projects so that they contribute to the • Package multiple strategies as comprehensive, reduction of greenhouse gases (GHG) on a global integrated implementation packages. scale. The purpose of the methodologies, however, goes We welcome you into this process, and encourage beyond mere impact estimation: they are designed you to use this Manual—and all of GEF’s resources—to to encourage high quality project design, increase compose your project as an asset to your community consistency and maintain objectivity in impact and the world. estimation. In addition to environmental benefit, transportation Why this Manual? projects also produce significant “local co-benefits” that, in many cases, could be the primary justification Every GEF project requires an assessment of the for the host country to pursue the project. Therefore, greenhouse gas (GHG) emissions (in CO2 equivalence) this document also seeks to articulate the related that the projects are expected to reduce. In 2008, co-benefits appropriate to the unique nature of GEF the GEF developed a manual detailing specific projects. While co-benefits do not directly create methodologies for calculating the GHG impacts global benefit, they increase the engagement and of energy efficiency, renewable energy, and clean investment of local stakeholders in project success and energy technology projects. This new Manual they increase the replication potential of projects— provides the first methodology designed specifically both of which do result in increased global benefit. for projects in the transportation sector. It follows the For this reason, GEF project applicants are asked to general framework, terminology, and principles of consider co-benefits in all proposals, and guidance those earlier GEF modules. More importantly, it uses for doing so is included in this Manual. Introduction, Concepts, and Definitions 1 What Distinguishes the GEF projects include additional elements such as establishing financing mechanisms that leverage Methodology from other Models local private sector financing; capacity building for CO2 Accounting? and technical assistance; and the development and implementation of government policies supporting Most of the methodologies used to measure the climate-friendly investments. These elements do GHG impacts of projects focus on the emissions not have direct GHG impacts, yet are necessary for savings from a specific investment. The GEF model effectively avoiding emissions in the long run. So, they considers these impacts from multiple perspectives. are calculated separately within GEF methodologies Additionally, GEF projects differ in other ways such as “indirect” impacts. as funding schedules, project activities, and strategic market development. Because of these distinct Compared to the CDM (and some other models), attributes, a different technique for calculating end GEF projects are intentionally and necessarily riskier. results must be applied. Their outcomes are less certain, and are subject to greater variation in the degree of uncertainty both For comparison, consider that projects under the between and within projects. Yet, because they are Clean Development Mechanism (CDM) of the Kyoto less rigorous and data-intensive, GEF projects are Protocol must specify the technical characteristics more accessible to project hosts who command fewer of the hardware, location, ownership, and operating data resources. GEF projects are also more flexible to hours, in order to accurately calculate the amount of accommodate a more diverse array of project types. emissions reductions produced from an investment. So, while it is essential to be able to estimate the GHG Those methodologies are well designed for impact with reasonable confidence, there are other assessing GHG impacts of CDM projects, and are critical purposes addressed in this process. The GEF continually reviewed by the relevant bodies of the methodologies are designed specifically to address United Nations Framework Convention on Climate all of these intersecting factors. Change Convention (UNFCCC). They can also serve as helpful tools to analyze the results of GEF projects, but, by themselves, they do not satisfy the Principal Attributes of the GEF model. GEF Methodology Also, with the CDM, proponents receive the funding for CO2 emissions reductions only upon delivery of An adequate methodology to assess the effects of a Certified Emission Reductions analysis based on GEF investment in transport projects must account observed results after the project is implemented. for the direct mitigation impact of both GEF and Because the financing is directly tied to the GHG co-financing investments. It must also estimate the impact measurement, precision is highly important. indirect impacts which come from the replication a project inspires in other places, and the market GEF financing, on the other hand, happens before expansion which results from these investments. project implementation. Thus, the GEF applicant Since the estimates for direct and indirect impacts are must create a projection of the expected impact fundamentally different in their accuracy and degree of a project in an early phase of planning, when of certainty, the methodology must report separately advanced data is not available and the future on direct and indirect impacts. impact is more difficult to forecast. In many of the developing countries where the GEF operates, To quantify and inter-relate these factors, GEF has transportation data is often incomplete, unreliable, constructed a set of formulas with extensive default or all-together non-existent. The GEF methodologies factors. Identified as the Transportation must recognize these realities, and cannot be overly Emissions Evaluation Model for Projects data-intensive. (TEEMP), they are the core estimating tools for GEF projects and account explicitly for the factors Also, GEF funding is not revoked if reduction targets noted above. TEEMPs are explained extensively in all are not attained or certified. “Success” has many Sections of this Manual. measurements, and GEF weighs multiple factors in It is important to note that no single, general- assessing the value of a project. In addition, a GEF purpose methodology can be used to quantify GHG method for GHG estimates must take into account emission reduction effects for GEF projects. Further, the investments that can happen after the actual GEF a methodology that results in only one aggregate intervention. number for the portfolio does not provide meaningful Yet another difference lies in the types of project and comparable values for GHG abatement costs activities supported by the GEF. Many proposed (US$/tons) because of the following: 2 Introduction, Concepts, and Definitions a. The GHG emission reductions are achieved by 3. Amount of transport activity, integrating many different strategies in GEF projects. 4. Mode of transport chosen, and b. The weights of these strategies vary greatly 5. Amount of capacity/occupancy used. among different projects. Direct emission reductions in any of these five c. In the interest of sustainability and replicability, categories are calculated by assessing the expected the GEF-sponsored part of the project often change in GHG emissions that would be attributable focuses on interventions that have long-term cost- to the GEF (and co-financed) investments. These reduction effects (e.g., through capacity building reductions are projected for, and totaled over, the or enabling environments), but by themselves do respective lifetime of the investments both during and not have impacts on GHG emissions. post implementation. (These concepts are thoroughly discussed in Section II of this Manual.) Intersecting these influences is the system of categorization used to organize areas of results for a All CO2 savings resulting from investments made GEF project. A GEF project can yield results in three within the boundaries of a project will be counted general areas: toward a project’s direct effects. The boundaries of a project are defined by the logframe (a commonly- 1. Direct CO2 emission reductions achieved by used project management matrix used to track project investments that are directly part of the results of activities and outcomes), either using GEF resources or the projects; the resources articulated by co-financiers, and tracked 2. Direct post-project emission reductions achieved through monitoring and evaluation [M&E] systems. through those investments that are supported by The GEF methodology also includes what will be GEF-sponsored revolving financial mechanisms referred to in this Manual as “direct secondary still active after the project’s conclusion; impacts,” often referred to by transport and 3. A range of Indirect impacts achieved through environmental planners as “indirect” effects. These market facilitation and development. include such items as GHG impacts that come from changes in land use or vehicle ownership, which Clearly, no single formula can be applied unilaterally in turn resulted from a GEF investment. (These are to calculate these divergent impacts. For that reason, detailed in Section II) the GEF methodology estimates direct and indirect impact figures separately, and applies numerical values for uncertainties that are appropriate to each What Is “Direct Post-Project” scenario. In each instance, conservative assumptions are used to account for uncertainties, including GHG Impact of Transportation the influence of the GEF intervention itself and the Sector Projects? possibility of shifting baselines. Although it is rare in transportation projects, the GEF The three areas of potential impacts articulated above does allow the establishment of financial mechanisms are examined briefly in the following sub-sections, that could continue to operate after the project ends. and are thoroughly detailed throughout this Manual. These mechanisms may include such tools as partial credit guarantee facilities, risk mitigation facilities, or revolving funds. Such ongoing mechanisms may What Is “Direct” GHG Impact in facilitate investments that yield GHG reductions. Transportation Sector Projects? However, because these impacts occur or continue beyond the timeframe of scheduled project In the GEF methodology, there are five categories of monitoring—and the fund continues to recycle the transport sector1that GEF projects can influence itself—they are considered separately as “direct post- to reduce GHG emissions: project impacts.” These impacts can be estimated by 1. Vehicle fuel efficiency, using the same methodology as the direct impacts. The formulas used here are the same as those used 2. Greenhouse gas intensity of the fuel used, in calculating direct emission reductions. However, the nature of projecting direct post-project emissions 1 Salon, Deborah. An Initial View on Methodologies for Emission dictates that conservative assumptions be used with Baseline, 2001. Schipper, Lee, Celine Marie-Lilliu, and Roger Gorham, June 2000, Flexing the Link between Transport and reference to leakage rates and financial instruments’ Greenhouse Gas Emissions, 2000. effectiveness Introduction, Concepts, and Definitions 3 To date, only one GEF transportation project, has used attributes that increase the potential for its replication. a revolving fund or credit guarantee facility. That project These can include the quality of the project design, the educated mechanics in Pakistan on improving engine amount of co-benefits a project achieves, and activities efficiency with tune-ups, and provided facilities used for designed specifically to encourage replication,. those tune-up services. As loans to set up the project were paid back, the funds were scheduled to be re- Indirect impacts are measured using two different approaches, referred to as “Bottom-up” and “Top- cycled to fund more training and facilities, continuing down.” Each provides a different range of potential until the fund was depleted due to leakage. indirect impacts. This approach has succeeded in other GEF initiatives The “Bottom-up” approach provides the lower, and in non-GEF transportation investments. Similar more conservative extent in the range of possible revolving funds might be considered when proposing indirect impacts. It estimates the likely effectiveness GEF transportation projects. Examples could include of a project’s potential power to inspire and catalyze the development of private sector parking management similar projects. To arrive at this figure, the direct and concessions linked to urban improvement districts; or direct post-project impacts of a project (calculated the development of road user charging; and smart separately) are simply multiplied by the number of traffic management systems linked to performance times that a successful investment under the project contracts for corridor operations and management. is likely to be replicated after the original project’s Credit guarantee facilities could be used to help secure activities have ended. “Bottom up” requires an expert low-cost private financing for development of GEF judgment on the degree to which a project is likely to projects, cutting the risk premium attached to bonds replicate within its sphere of influence. supporting private or public project financing. (In the The “Top-down” approach is generally used to find United States, the Transportation Infrastructure Finance the highest extent in the range of potential indirect and Innovation Act - TIFIA - provides Federal credit as- impacts. It estimates the combined technical and sistance in the form of direct loans, loan guarantees, economic market potential for the project type within and standby lines of credit to finance surface transpor- the 10 years after the project’s lifetime. Using the tation projects of national and regional significance.) maximum realizable market size further implies that Capitalization of such loan guarantee programs might there would be no baseline changes over considerable be done on a national or regional basis to leverage periods of time, and that all emission reductions in substantial additional short-term investment capacity that sector or market can be attributed entirely to the by expanding access to credit markets. This could ac- GEF intervention. celerate the timetable of investments in such measures as BRT, non-motorized transportation network im- Clearly, both of these assumptions are unlikely to provements, high quality vehicle registration and traf- hold in reality. Therefore, the assessment contains fic management systems, or freight system efficiency a correction factor variable, the “GEF causality improvements. (The process of estimating direct post- factor,” that expresses the degree to which the GEF project impacts is thoroughly covered in Section II.) intervention can take credit for these improvements. This causality factor is used to calibrate the “Top- down” estimate for the indirect benefits. What Are “Indirect” GHG For some types of transport projects, such as bus or rail, Emission Savings of there is currently enough historical data to support the Transportation Sector Projects? estimation of a replication rate based on documented experience with previous projects. Accepted replication All GEF projects strive to catalyze replication of rates based on historical observations may be used in- successful projects by emphasizing capacity building, stead of creating a range of indirect impacts using the promotion of project activities, the removal of market two methods described above—Bottom-up and Top- barriers, and development of innovative approaches. down. The summaries of other types of transport sec- The GHG emission reductions that result from tor projects—both GEF and non-GEF projects—should replication are referred to as “indirect” GHG impacts. also be tracked so that the documented dissemination They are counted separately from direct impacts rates can be used to inform future projects. because they occur outside the project logframe. Because the level of uncertainty and accuracy is To estimate these indirect impacts, one must rely heavily different from those of direct or direct post-project upon informed assumptions and expert judgment. savings, it is not appropriate to consolidate the two The potential of a project’s replicability springs not types of savings. Projects should be conservative in only from its market potential, but also from project projecting the size of the affected geographic area 4 Introduction, Concepts, and Definitions or market when calculating likely indirect impacts. local co-benefits in the areas of public health, travel The majority of projects should not go beyond the time, and economic growth. In many cases, these co- regional or country area, although in some cases a benefits are the primary justification—and motivation— wider sphere of influence can be permitted. for the host country to pursue the project. The greater the co-benefit to the local stakeholder, the greater is their interest in implementing the project successfully. What Are Local Co-Benefits Similarly, projects with high local co-benefits are also and Why Are They Important more likely to be replicated in other cities/regions. For these reasons it is advantageous to account to Global Benefit? for co-benefits, as they are essential ingredients in While the main objective of GEF investments is to transforming local investment into global impacts. generate global environmental benefit, the very nature The GEF methodology is designed to weigh local co- of transportation projects also produces significant benefits in assessing a project. Table 1: Three Types of GHG Emission Reductions in GEF Projects Evaluation Tool Direct Direct post-project Indirect Definition of Reduction Project activities and Investments supported by Project components that Type: investments whose outputs mechanisms (e.g., revolving encourage replication such and secondary impacts funds) that continue as study tours, capacity are tracked in the project’s operating after the end of building, public promotion, logframe the project etc. Logframe level Has a corresponding activity Not corresponding to a Outcome/impact on level or investment with an output specific logframe level of global environmental that is tracked in the logframe objective Quantification method Use of GEF TEEMP models Based on assumptions of Based on the replication with default values (or functioning post-project rate o the project using provision of additional data) mechanisms Bottom-up or Top-down methods Quality of assessment Highest level of certainty and Reasonable level of Lower levels of accuracy accuracy for minimal data accuracy, medium level of and certainty inputs (lower than the CDM) certainty Introduction, Concepts, and Definitions 5 II. Overview for Applying GEF Tools and Methodologies TEEMPs: The Core of TEEMP Spreadsheet Model Cell Color-Coding the GEF Methodology The process of calculating GHG reductions from GEF Green Cells Required User input projects has several steps. The complexity depends on the number and type of project components Default Value, which can be involved. As discussed in Section I, these can include Red Cells replaced with local data, if available Direct GHG emission reductions, Direct Post-Project reductions and Indirect reductions. Since there are Blue Cells Output: GHG Impact many different ways to achieve GHG reductions in (User does not modify) the transport sector, there is no “one size fits all” Internal Calculation Cells Yellow/Orange Cells methodology that can effectively evaluate their (User does not modify) impact. To confidently project the GHG reductions for a GEF The TEEMPs can provide an ex-ante estimation of the project, specific methodologies have been developed direct GHG impact of a project in a consistent way for common types of transport projects. At the heart with very little local data. This is possible because of these methodologies are a series of models the formulas use conservative default values. These (Excel-format formulas) called the Transportation values are based on research, observed results from Emissions Evaluation Model for Projects (TEEMP). similar projects, and expert opinion. However, when The methodologies are derived from international local data is available, it can easily be inputted into the experience and best practices, and are kept as simple models to provide a more accurate—and potentially as possible. larger—estimation of the direct GHG impact. TEEMP models streamline the process of estimating TEEMP Release 1.0 was developed with support from emissions impacts for transportation projects in five the Asian Development Bank (ADB). It was used to categories: estimate the carbon footprint of ADB’s transportation I. Transportation Efficiency Projects (Clean Vehicles/ projects between 2000-2009 and to evaluate various Fuels) strategies that might reduce transport CO2 emissions. TEEMP Release 1.1 has been expanded and enhanced II. Public Transportation Projects (Bus/Rail) with support from the United Nations Environment III. Non-Motorized Transportation Projects Program and Climate Works Foundation for GEF. IV. Transportation Demand Management Projects Version 1.1 addresses more types of interventions V. Comprehensive Regional Transport Initiatives and transport management strategies. Currently, TEEMP models exist for bike-sharing, In 2010-2011 the TEEMP models are being more fully bike-ways, bus rapid transit (BRT), expressways validated and enhanced, by applying them to various alternatives, mass rapid transit (MRT), pedestrian projects for which data is available, and/or is being facility improvements, railway alternatives, as well collected. Through these refinements, many region- as several different transport demand management specific default values have been updated and made (TDM) programs. Each of the models has a “Basic more accurate. This process of enhancing and updating Guide” and “Home” worksheet tab which explain the models is continuous. To ensure you are working how to get started using the model. When using with the current values and models, we strongly these spreadsheet models the cells are color-coded advise you download the most recent formulas at the according to the following scheme: http://www.unep.org/stap/calculatingghgbenefits. 6 Overview for Applying GEF Tools and Methodologies Sequence of the GEF Methodology Even though there is a vast variability in the types of GEF projects, there is a consistent sequence that is followed in calculating CO2 emission reductions for a GEF application: 1. Establish a baseline: Calculate the estimated baseline emissions of the scenario without a GEF intervention. The baseline emissions estimation will be compared against the estimated GHG emissions reduction achieved by the GEF project. When using TEEMP models to find direct impact, no separate baseline need be established in this step because TEEMP models automatically calculate a baseline by using a market-shed analysis approach. Instead, the user should be sure to input all dependable local transport data that is available into the TEEMP model. If dependable local data is unavailable, default values are provided. 2. Calculate the direct emissions impact for the GEF scenario. This includes all GEF and co-financing investments that are tracked in the logframe during the project’s implementation. The difference between this GEF project scenario emissions and the baseline emissions equals the direct emission impact of the project. If TEEMP models are used, this figure is the model‘s main output. 3. Estimate the direct post-project emission reductions, if any are expected. Direct post-project impacts occur beyond the supervised timetable of the project. They result when a financial mechanism, established as part of a project, remains in place and keeps providing support for GHG-reducing investments beyond the lifetime of the project. 4. Calculate the indirect emission reductions. These are reductions that occur from replication and market expansion outside of the logframe or in the post-project period which have a “causal” link to the GEF intervention. If it is appropriate for the situation, use both the Bottom-up and the Top-down methodologies to create a range of potential impacts. In some cases, only the Bottom-up method will make sense. For certain types of transportation interventions, accepted (default) replication rates based on observed impacts can be used. Each of these steps will be discussed in more detail in the following sub-headings. Figure 1 (next page) contains a flowchart illustrating this process. Broad Assumptions in Applying e. As a general rule when applying this methodology, the GEF Methodology the project proponent should err on the side of transparency, and generally be cautious and The data and assumptions necessary for the GHG conservative when making assumptions on GHG emissions reduction assessment will vary by the type emission reductions. of transportation sector intervention. However, some general rules are important in all steps of the GHG emission reductions assessment for the GEF: Required Data for a. All GHG impacts are converted to metric tons of CO2 equivalence. GEF Methodologies b. The CO2 reductions reported are cumulative For each GEF project, the proponent is required reductions, calculated for the lifetimes of the to provide extensive data in the following broad investments. No GEF projects may claim impacts categories: for more than 20 years. (a) Lifetime of the Investment c. There is no discounting for future GHG emission reductions. (b) Baseline Scenarios d. Whether or not the TEEMP models are used, all (c) Emission Factors GEF impact estimations should incorporate as much local measured data as possible. When none These three categories are the “input channels” for is available, applicants can rely on the conservative the TEEMPs. default values provided in the TEEMP. The default values are based on research and past experience A detailed explanation of these three categories agreed upon by experts. follows. Overview for Applying GEF Tools and Methodologies 7  teps for Data Collection and Development of Baselines, Impact Estimations, and Figure 1: S Calibration over GEF Transport Project Lifetime GEF Default Local Values & Transport Transport Sector Data Emissions Evaluation Models for Projects Ex-Ante (No-Project) Baseline Established Transport Public Non-Motorized Transport Comprehensive Efficiency Transport Transport Demand Transport Strategy Methodology Methodology Methodology Management Methodology Methodology Project Impact Reported and Data Lifetime Direct Impact from Project Used to Calibrate GEF Default Estimate Sector Values Direct Is & GHG there a post- Reduction Post-Project project financial Effects mechanism? Rates Yes No Direct Post-Project Impacts Estimate Bottom-Up: Indirect Project Top-Down: Impacts Estimate Total Replication via TD and/or BU Factor Potential with approach Causality Add in any direct secondary effects and apply a causality factor. Range of Indirect Project Impacts (based on replication) 8 Overview for Applying GEF Tools and Methodologies Lifetime of the Investment developed by combining local traffic and travel counts/surveys and the default values from the A critical parameter that must be determined is the TEEMP models for fuel cycle and emissions lifetime of the investment (project). This lifetime is factors. impacted by the various technologies, investment conditions, and assumptions associated with each In cases where local travel activity data is weak, its project. Since these vary widely from one project to acceptance is subject to GEF approval and could the next, applicants should use sound judgment is possibly be disallowed. So, a strong effort must assigning values for each of these interwoven factors. be made to collect valid local data in the project The GEF methodology specifies preapproved default preparation phase. A potential source of funding values for the lifetimes of the relevant technologies, to support this task can come from applying for a and proponents are encouraged to utilize these GEF Project Preparation Grant (PPG) in the initial default values. The calculation of these values for Project Concept (PIF) document. each type of project is discussed in detail in the later Sections of this Manual. Each Section addresses a The GEF has separate guidelines for Incremental specific project type and discusses how the TEEMPs Cost Analysis. These guidelines relate to are used to support the project design. the incremental costs incurred through developmental activities of national governments Baseline Scenarios and implementing agencies in caring for the environment. Whatever methodology is applied, it is imperative that a dynamic, “no-project” baseline scenario be c) Include impacts for other major, planned developed. The baseline scenario should incorporate transport sector interventions that are not GEF- analyses of the sector’s current static conditions as funded but are within the impact area of a proposed well as growth trends of transport behavior, different GEF-funded transport initiative. If, for example, technologies, mode shares, carbon-intensity of fuels, a new ring road or major roadway expansion is fuel economy of vehicles, etc. This measurement being implemented in or around the impact zone must forecast emission values for the specific market of a proposed GEF project, the impacts of these that would occur without the GEF or co-financing should be included in the baseline analysis. The intervention over the period of the intended project. GEF TEEMP includes sketch models that can be TEEMP models are constructed to generate the used to evaluate these impacts. baseline so that it overlaps the GEF alternative scenario (the GEF investment). d) GEF projects should incentivize the development of plans for gathering observation-based data When developing the Baselne Scenario, GEF at all points of the project. More accurate data can applications should follow the guidance below: be used to strengthen the baseline developed in the project application phase. It better informs a) A dynamic baseline forecasts the emissions planning and regulation, helps secure wider inventory of the affected market in a “business- funding, and is valuable in monitoring and as-usual” scenario. The baseline ignores any evaluating the project. Better data can help refine contribution that would be made by a GEF or co- the TEEMP models, and, later, makes a successful financing project. (If TEEMP models are used for project easier to replicate. For these reasons, all the ex-ante direct impact estimation—discussed projects should design tools for monitoring and below—a separate baseline need not be created evaluation, and for the systematic collection of because the TEEMP automatically calculates a data that relates to the GEF project. Collection “no-project” dynamic baseline in its market-shed tools could include traffic counts, household analysis of the GHG impact from a GEF project.) surveys, GPS vehicle and personal activity b) Baselines should contain a description of the monitoring, local fuel and emissions testing, etc. market’s likely development and transportation The GEF also encourages the use of enhanced activities as they would evolve without investments modeling methodologies, when possible, that can from the GEF or co-financing. The baseline should co-relate fluctuations in transport demand with also include all non-GEF interventions that would changes in travel time and cost of different modes, be introduced to the sector by the implementing and have some capacity to estimate longer-term agency. Proponents should describe the impacts on land development patterns.  characteristics of the transportation sector, the emission factors, the markets to be transformed, e) Baselines must include all transportation modes and the lifetime of the investments. In absence affected by the project within the project of good local data, the ex-ante baseline will be area. Thus projects that shift travel loads among Overview for Applying GEF Tools and Methodologies 9 multiple modes will need to establish baselines activity survey data is encouraged, where these are which include multiple modes over the entire area. available and deemed to be adequately calibrated to Others may only need basic data about a small observed local conditions. group of vehicles to establish a reliable baseline for estimating the eventual impact of the project. The default emissions factors used in all TEEMP Projects that combine multiple interventions models are illustrated in the table below. will need to establish baselines for each type of For many GEF projects, the principal GHG emission intervention for which they claim direct impacts. focus will be on CO2, which is closely tied to fuel use. In general, significant benefit will be realized by However, there are several other contributing factors combining multiple strategies into an integrated to GHG emissions and, where possible, applicants are approach. encouraged to include them. Emission Factors Table 2 reproduces the Intergovernmental Panel For the baseline technologies, as well as for the on Climate Change (IPCC) figures, which should be technologies to be deployed under the GEF Alternative used for all purposes in GEF projects where non-CO2 Scenario, the proposal needs to contain the expected gases are considered. Typically, the 100-year figures emissions factors, i.e., how many kilograms of CO2e are used. are going to be emitted for each vehicle-kilometer Not included in this table is black carbon, formed of travel (VKT) by mode and vehicle type. This value through the incomplete combustion of fuels. Black is derived using either the default emission factors carbon is a potent climate forcing agent emitted in provided by the GEF TEEMP or more accurate locally- the transport sector. Its effects on global warming measured data. Emission factors will vary considerably are considered to be second only to CO2. Mitigating based on vehicle fleet composition, vehicle speed black carbon may be one of the most effective means and operating conditions, and vehicle occupancy, of controlling climate change. with additional variation based on temperature, fuel characteristics, and other factors. Use of emission This manual does not incorporate emissions of black factor models, such as COPERT, in conjunction with carbon in its methodologies because, at the time regional travel models and local travel and vehicle of publication, the UNFCC has not yet assigned a Table 1: Default Emission Factors for GEF TEEMP Models Fuel Efficiency CO2 emissions factor Average CO2 Speed Fuel Type CO2 emissions per vkt @ 50 km per liter of fuel efac by veh type Vehicle Type % Split km/liter kg CO2/liter kg CO2/km kg CO2/km km/hour Petrol Diesel   Petrol Diesel Petrol Diesel Petrol Diesel All Fuels Cars 22 95% 5% 100% 9 11 2.75424 2.94348 0.306026667 0.267589091 0.304105 2-Wheeler 22 100%   100% 60 0 2.75424 2.94348 0.045904   0.045904 3-Wheeler 22 100%   100% 22 24 2.75424 2.94348 0.125192727   0.125193 Taxi 22 30% 70% 100% 8 11 2.75424 2.94348 0.34428 0.267589091 0.290596 Bus 22   100% 100% 1.8 2.2 2.75424 2.94348 1.530133333 1.337945455 1.337945 Jeepney/RTV 22   100% 100% 6 7 2.75424 2.94348 0.45904 0.420497143 0.420497 Walking 4           Cycling 12           LRT             10 Overview for Applying GEF Tools and Methodologies GWP for black carbon. Even so, projects are urged Almost all GEF projects combine different categories to account for black carbon in their calculations when of investments that yield reductions in different ways. reliable data is developed. Shifting from fossil fuels Tangible investments in infrastructure or planning to other fuel sources and adopting newer engine yield direct emissions impacts. technology and emissions standards are all methods Other investments intervene through less tangible of reducing black carbon. project components such as education, capacity- building, and/or public outreach. The impacts from these investments accrue beyond the lifetime of the Calculating Direct project, following only indirectly from the project Emission Impacts activities. When this is the case, these reductions should be calculated separately. The TEEMP models Many advanced approaches can produce GHG provide methods for calculating both direct and indirect reductions in the transportation sector. These include impacts. These approaches are discussed thoroughly the development of voluntary carbon funds, voluntary in this Section and throughout this Manual. markets for certified emission reductions, obligatory markets for carbon emissions, and the methodological The most clear-cut criterion to decide whether progress in the Clean Development Mechanism. All of investments should be counted toward direct or indirect emission reductions is whether the investment these mechanisms target emission reductions resulting is included in the log frame of the GEF project, from specific investment projects. In GEF projects this and whether it is monitored as part of the project’s target measurement is referred to as “direct emission success indicators. Even so, in many cases, a project reductions,” or “direct GHG impacts.” component’s impact is included in the project’s log Several methodologies have been published to frame but there is no reliable way to quantify its impact on emissions. In this case, no impact should analyze the direct emission reduction effects of CDM be recorded. Normally, direct impacts should only be projects. These methodologies tend to be more recorded for investments with known and quantifiable rigorous and data-intensive than the TEEMP models. impacts, such as infrastructure, policy, and planning. However, they can be applied to calculate direct emission reductions for GEF projects in place of using TEEMP models incorporate baseline calculation in the TEEMP. their “market-shed” approach to calculating GHG Table 2: Global Warming Potential of Other Greenhouse Gases1 Global Warming Potential Time Horizon Gases Lifetime (years) 20 years 100 years 500 years Methane (CH4) 12 72 25 7.6 Nitrous Oxide (N20) 114 289 298 153 HFC-23 (hydro fluorocarbon) 270 12,00 14,800 12,200 HFC-134a (hydro fluorocarbon) 14 3830 1430 435 Sulfur Hexafluoride 3200 16,300 22,800 32,600 *IPCC AR3 figures in parenthesis where different from AR4 values. 2007 IPCC Fourth Assessment Report (AR4), Chapter 2: Changes in Atmospheric Constituents and in Radiative Forcing. http://ipcc- 1 wg1.ucar.edu/wg1/Report/AR4WG1_Print_Ch02.pdf) Overview for Applying GEF Tools and Methodologies 11 impact. If a project’s impact cannot be calculated Calculating Direct using a TEEMP model, the general equation below should be followed. It is derived from international Secondary Impacts best practices and based on the “ASIF” model. Another type of direct impact—referred to collectively All investments responsible for direct effects are as “direct secondary impacts”—may also accrue from evaluated in terms of the energy or fuel saved over secondary effects of GEF and co-financer investments. the lifetime of the respective investments. Different These include GHG impacts from supportive policy technologies have different assumed lifetimes. reforms, fuel standards, motorization rates, and land The saved fuel or energy is then multiplied by the use changes that are catalyzed by GEF and co-financer marginal CO2 intensity of the energy supply. The investments. An example of a direct secondary impact formula is: would be when there is an intensification of land uses as a result of a GEF-financed transit project (BRT), that in turn further reduces private auto trips within the BRT corridor. CO2 direct = E * c = e * l * c; with Direct impacts from these secondary effects can be CO2 direct = direct GHG emission calculated using the same methodologies used to savings of successful project calculate direct impacts. However, a GEF causality implementation in CO2 eq, in factor should always be applied to reflect the degree of influence the project provided in creating the GHG tonnes impact. For instance, in the BRT project example E = cumulative fuel or energy above, the project may not have been the only factor saved or substituted, e.g., volume/ contributing to the intensification of land use. A supportive zoning reform may have occurred within mass of fuel used (or MWh if the timeframe that the project was implemented, electric); E = Σl e and, thus, also become an additional inducement for c = CO2 intensity of fuel/energy the intensification of land use. Therefore, the GEF project by itself cannot claim full credit for the GHG e = annual fuel/energy replaced, impact of the land use intensification. Instead a GEF e.g., in volume/mass of fuel used causality factor—expressed as a percentage—should (or MWh if electric) be applied in proportion to the degree of influence generated by the GEF project. l = average useful lifetime of The general guidelines for applying the GEF causality equipment in years factor are: i. Level 5 = “The GEF contribution is critical and The lifetime of the infrastructure determines nothing would have happened in the baseline,” the duration over which the GHG savings may GEF causality = 100 percent occur. Regardless of when they occur, savings are ii. Level 4 = “The GEF contribution is dominant, but represented as totals at the completion of the project. some of this reduction can be attributed to the That means that the impact of all investments that are baseline,” GEF causality = 80 percent made during the project is the same, irrespective of whether they are realized in year one or five of project iii. Level 3 = “The GEF contribution is substantial, implementation. However, they must be introduced but modest indirect emission reductions can be during the project’s supervised operations to count as attributed to the baseline,” GEF causality = 60 “direct” GHG emission reductions. percent Because of the structure of GEF projects (and a iv. Level 2 = “The GEF contribution is modest, and conservative interpretation of the GEF co-financing substantial indirect emission reductions can be rules), investments are counted toward this sum attributed to the baseline,” GEF causality = 40 regardless of whether they are financed by GEF percent support or by co-financing. The decisive criterion for the question of whether to include or exclude v. Level 1 = “The GEF contribution is weak, and most an investment is whether it is included in the M&E indirect emission reductions can be attributed to framework proposed in the logframe. the baseline,” GEF causality = 20 percent 12 Overview for Applying GEF Tools and Methodologies The chart in Figure 3 summarizes the process and keeps looking for new investments. Depending on variables in calculating direct GHG emission reductions the leakage rate, facilities of this type can lead to a for transport projects. multiple of the original direct investment, which in turn can lead to a multiple of the associated emission savings long after the project itself has ended. (An Calculating Direct Post-project example of a successful project in Pakistan, illustrating Emission Reduction Effects this dynamic, was given in Section I of this Manual.) In some cases, GEF projects implement a GEF- These “direct post-project” emissions are calculated supported financing mechanism that will continue by extrapolating from the direct effects achieved to support direct investments after the supervision during project implementation. Clearly, some period of the project. An example is a revolving fund assumptions are needed. For a revolving fund, for up-front financing of bus rapid transit, parking for example, the rates of reflow and leakage will management, and urban improvements, which is determine how many investments can be financed then refinanced from user fees, loan repayments. after the supervised implementation period. A There might also be a partial credit guarantee facility “turnover factor” (tf) is defined as the number of that could be fully exposed at the end of the project, times the post-project investments will be larger than but then reduces its credit risk exposure and thus the direct investments. Figure 3: Flowchart for Calculating Direct GHG Emission Reductions For Transport Projects Does the activity in the project logframe no direct reductions include tangible No installations? Yes x + = Sum of avg. Average Secondary annual GHG useful Direct Effects Lifetime Direct Reductions reduction lifetime of (totaled over a (for one activity) from project investment 20 yr. max lifetime) activity in transport sector (years) Sum of all activity- level reductions Emissions factor (incl. upstream Avg Lifecycles: emissions) (Defaults) per Total Direct Impact mode/fuel Vehicles & affected Equipment – 10 years Infrastructure – 20 years ∆ in fuel use for each mode/fuel affected Overview for Applying GEF Tools and Methodologies 13 The general formula for calculating direct post-project In this equation, the turnover factor “tf” is equal to GHG reductions is: the number of times that the whole fund volume is expected to be invested and reinvested after the project. The first turnover will usually happen within CO2 DPP = CO2 direct * tf; with the project’s supervised implementation period, and thus count toward the direct emission reduction. CO2 DPP = emissions saved with Subsequent turnovers would be counted as direct- investments after the project, post-project emissions impact. supported by post-project financial By their very nature, the estimates for direct post- mechanisms project effects carry a slightly higher degree of uncertainty than the direct GHG project outputs. CO2 direct = direct emissions But since they clearly can impact GHG emission savings to the degree that they are reductions, they need to be accounted for in a GEF supported through the mechanism project. To provide this measurement, direct post- that causes the post-project impacts project effects should be reported separately from the direct emission impacts (described above). Direct tf = turnover factor, determined post-project effects are actually a form of indirect for each investment based on emission reductions (covered below) But they can be assumptions on the fund leakage assessed with a higher degree of certainty and so are and financial situation in the calculated as a distinct category. project country Figure 4 illustrates how to calculate direct post-project GHG impacts. Figure 4: Direct Post-Project GHG Emission Reductions Calculation Leakage Reflow Years Lifetime Direct fund will rate rate Project Impacts operate after after during after project project project supervision (d) close (n) close (k) close (r) Based on r= 1-k Because the Depends on expert fund begins fund design, Lifetime Direct assessment before project local financial Post-Project of local close, there market. Impact context are direct Leakage and = [d*(1-r(n+1))/k)]-d reductions reflow rates from the initial will affect this investments. variable, These although reductions generally a are also well-managed entered here. fund will last 5-10 years. Lifetime Direct Post-Project Impact 14 Overview for Applying GEF Tools and Methodologies Calculating Indirect Impacts the risk of exaggerated project expectations, one should use conservative estimates when using either For many projects, direct GHG emission reduction methodology. impacts tell only half the story. GEF projects that catalyze replication of sustainable transport projects Market and replication potential for a project is not in multiple cities or regions, or remove barriers and the only factor to drive indirect impacts. Three other bring sustainable transport technologies to a wider factors must be considered in the expert analysis of a market, can—indirectly—accrue large GHG reduction project’s indirect impact: impacts. These impacts—referred to as Indirect 1. Project activities which facilitate replication; Emission Impacts—could potentially be larger than the direct impacts, and must be assessed within a 2. The creation of attractive local co-benefits from GEF project. project activities; and During project design, proponents must estimate 3. The quality of a project and its potential to be long-term (indirect) impacts of the interventions, successful. and must include the data and assumptions used to estimate this impact. This is sometimes a difficult These activities, detailed in later Sections of this exercise. Essentially, the proponent is projecting the Manual, increase a project’s replication factor in the likelihood of a project’s repetition after the original Bottom-up method and may increase a project’s project is complete. The variables are considerable, causality factor in the Top-down method. and the initiative for project replication may not be in Some assumptions must be made to calculate indirect the hands of the proponent of the original project. impacts: Thus, it is not practical to use a straight line formula a. A standard project influence period for GEF effects to estimate indirect emission impacts. Instead, has been assumed to be 10 years. This means that complimentary techniques are used to create portions a typical project will exert some influence on local of a broad vision for future possibilities. The results market development for about 10 years. Thus, of these approaches are then merged to compose investments that happen within 10 years after the as responsible a picture as possible of a project’s project—that were not projected in the baseline— potential replication. can be counted toward indirect impacts. The The two techniques used for these calculations are GHG reductions of each subsequent investment called “Top-down,” and “Bottom-up.” Top-down are summed over their respective lifetimes for presents the most optimistic estimate of potential a cumulative measurement. Depending on the replication. Bottom-up presents the most conservative lifetimes of these investments, the influence estimate. period might be shorter than 10 years. The Top-down methodology uses the size of the b. When applying either the Bottom-up method or entire national/regional market as a starting point, the Top-down method, inserted data and values applying given assumptions for costs and benefits should be conservative and limited to a realistic of the technology. For instance, a GEF bus-related scope. investment may be designed to impact a city-wide bus fleet. But the potential market for replication could be c. If a project envisions a second phase or tranche the bus fleet of the entire nation or region. Clearly, this at a later stage, and the GEF contribution to this results in the most optimistic assessment – full market second phase is not yet approved by Council, penetration—and thus it is the upper-most limit for the GHG reductions achieved during the second the range of potential GEF project impacts. phase are counted as indirect effects. Alternatively, using the Bottom-up methodology, one d. Most transport sector GEF projects should limit makes a conservative estimation of the number of the tabulation of indirect impacts to impacts within times the project is likely to multiply in the long run, the same region or country as the project. In some resulting in a lower limit of the range of the potential cases, innovative transportation projects have indirect impact. influence beyond their own country’s borders. For example, small nations with only a single large Whenever appropriate, both methodologies city and no potential to replicate a large-scale should be used in a complementary manner. This transport project within their national borders may is described in more detail below. Expert opinion is still play a catalytic role in the immediate region. required to determine the Top-down market potential This is especially true in regions of smaller, closely and the Bottom-up replication factor. To minimize connected countries with strong cultural and Overview for Applying GEF Tools and Methodologies 15 commercial links. Countries within such regions as This includes publication of results, public outreach, Central America or Southeast Asia could argue to educational outreach, capacity building, support for accrue indirect impacts beyond a country’s borders study tours and exchanges, etc. but within its sphere of influence. Examples of internationally catalytic projects are well-known: Figure 5 illustrates how to calculate the indirect GHG congestion pricing in Singapore, BRT in Curitiba, impacts of GEF projects using both approaches. and bicycle-sharing in Paris. Details on each approach are covered discussed in the next pages. e. To maintain integrity across the different segments of a project, double counting issues for indirect impacts need to be addressed and managed. Calculating Indirect Impacts— Some reality checks can be used to test the final results. Bottom-up Approach For example, the Bottom-up indirect calculation The Bottom-up approach for calculating indirect GHG should exceed the sum of the direct and direct reductions generally provides the lower extent in the post-project results. On the other hand, it should be range of possible indirect impacts from a project. smaller than the Top-down total market potential of It starts with the direct impacts of the investments the technology. under a project, and multiples that number by a factor The potential for replication and indirect impact representing the number of times the project is likely should also be linked to the funding and quality of to be replicated in other places/markets. For example, project components which encourage replication. a bus rapid transit project developed through a GEF Figure 5: Flowchart for Indirect GHG Emission Reductions Bottom-up Choose the Total Bottom-Up or Top Down Direct x Replication Factor Effects Approach Top-Down The number of times a Total project is likely to replicate, Technical considering market Potential potential, project activities, qualities, & co-benefits based on expert opinion. Total Economic Determined by Potential expert assessment Causality P10=Best-case replication x Factor – Top-down replication for lifetime influence (0.2 – 1.0) Range of Indirect Reductions 16 Overview for Applying GEF Tools and Methodologies project might save 200,000 tons of CO2 over the In the BRT example above, the replication factor lifetime of the infrastructure. Judging from the local would be 5, and the resulting indirect savings conditions, one could assume that within 10 years calculated by the Bottom-up methodology would after the project ends, five more cities in the country be 1 million tons. will adopt BRT systems with similar levels of GHG reduction. Mathematically, the direct GHG emission To date, there is no empirical assessment of the reductions are then multiplied by the assumed factor replication factors for the GEF portfolio, partly of replication (five) to find the Bottom-up indirect because the portfolio is not mature enough for reduction. systematic observation, and partly because no post- project evaluations are taking place. Therefore, for the The Bottom-up replication factor should be time being, the replication factors should be explicitly determined by an expert and based on four factors: determined in the project proposal for each project. When assessing these replication factors, two major a. Market Potential: a conservative estimate of its aspects should be taken into account: real potential for places and markets where it is likely to replicate (a) The first is the expected probability of replication, which is mostly related to the question of b. Project Quality: high-quality, full-featured proj- whether a particular transportation intervention ects are more likely to succeed, and successful is profitable or politically desirable and for that projects are more likely to replicate. reason offers some incentives to the local public or private stakeholders for replication. c. Project Activities Designed to Encourage Replication: study tours, capacity building, (b) The second is the question of how this likelihood technical assistance, public promotion, publication compares to the amount of investment already and dissemination of project information and taking place directly under the project. results all help to promote and facilitate project replication. In the absence of empirical assessments, generalized replication factors can be employed in the assessment, d. Local Co-Benefits: When a project has strong relating to the design and activities of the project. local co-benefits in addition to global benefit, it becomes more attractive to other places and Developing these replication factors on the basis of markets and thus more likely to replicate. experiences collected within GEF projects and from similar projects outside the GEF is underway but far The formula for estimating indirect impacts with the from concluded. What is clear is that for a project to be bottom up approach is: widely replicated, it needs to be a ‘high-quality, full- featured’ project that is politically popular in the host city with sufficient status and visibility to impress other CO2 indirect BU = CO2 direct* RF; with cities. The parameters of a ‘high-quality, full-featured’ project are defined in the project-specific sections of CO2 indirect BU = emissions saved this document as necessary. Secondly, promotional with investments after the project, and capacity building project components such as as estimated using the Bottom-up public outreach, study tours, policy guidance, and approach, in tons of CO2 eq technical training all also drive replication. RF = replication factor, i.e., how The potential for replication and indirect impact should also be linked to the funding and quality of project often will the project’s investments components (noted earlier) that encourage replication. be repeated during the 10 years These activities increase a project’s replication factor after project implementation, in the Bottom-up method and increase a project’s determined by expert and reflects causality factor in the Top-down method. the degree to which the project emphasizes activities which In the next Sections of this Manual, guidance is encourage replication provided in calculating indirect emission impacts for all major categories of GEF transportation projects. CO2 direct = estimate for direct However, in cases where the guidance may not be and direct post-project emission precise, each project should decide on a replication reductions, in tons of CO2 eq factor based on the knowledge of the local market. Keep in mind that the assessment should be conservative. Overview for Applying GEF Tools and Methodologies 17 Some reality checks: really be attributed to the GEF intervention, and how much would have occurred in the business-as-usual a) The replication (Bottom-up) should always be scenario. smaller than the overall market potential (Top- down), and; The calculation of indirect impacts should also account for the degree to which projects budget funding b) A comparison with the direct and direct post- for specific program components that promote project impacts should lend itself to a reasonable the project. The aggressiveness of a GEF project’s explanation. promotion affects its causality of replication. In most GEF climate change interventions, estimates Calculating Indirect Impacts— for full economic potential are created in the project Top-down Approach development phase. Many technologies that reduce greenhouse gases are already widely available and The underlying assumption of the top down the trend in longer term production costs are widely approach is that each investment has the potential known. So broader dissemination trends are easier to to economically impact 100% of the market being targeted by the initiative. This assumes the effective estimate with some degree of reasonableness. These removal of barriers to sustainable transportation estimates should be given greater credence than initiatives through capacity building and the promotion projects for newer technologies where performance of the initiative. Therefore, the starting point for and future production costs are difficult to determine. the Top-down Approach is forecasting the whole Such projects rely on expert estimates that are still economic potential for GHG abatement of a given unknown and/or difficult to verify independently. (The application in the project’s host country or sphere of relatively disappointing results of previous GEF efforts influence. This assumption has sweeping implications. involving hydrogen fuel cell vehicle development It asserts that, if all barriers to market implementation serve as a cautionary lesson in this regard.) are removed, market forces would move to exploit the full economic potential offered by the impacted In addition, the identification of specific GEF causality market. Following this paradigm, in the case of a in the dissemination of the technology needs to be public transit intervention, full economic potential carefully documented. Because market forces or would be the maximum provision/demand for public government policies might generate some of these transit within the region/country/sphere of influence of achievements at a later point in time even without a the project—all buses in the country, for instance. GEF intervention (baseline shifts), this figure is then multiplied by an assumed GEF causality factor, to be As you can see, the Top-down indirect impact assigned by an expert in the field, which indicates to calculation generally presents the high extent of the what degree the GEF intervention can claim causality range of potential indirect impacts. It starts by assessing for the reduction. the maximum possible market that could be leveraged using the specific transportation infrastructure initiated For the GEF causality factor, five levels of GEF impact in the GEF project. This assessment assumes 100% and causality have been assumed: impact within the project’s entire host country or sphere of influence. This is determined, rather simply, a. Level 5 = “The GEF contribution is critical and by determining the number of cities or regions that nothing would have happened in the baseline,” could support such infrastructure, technical capacity, GEF causality = 100 percent and typical investment rates in the country that can be expected under post-project circumstances. If it seems b. Level 4 = “The GEF contribution is dominant, but technically unfeasible to achieve 100% impact within some of this reduction can be attributed to the 10 years of the project’s completion, the total amount baseline,” GEF causality = 80 percent of potential additional project locations should then be corrected downward. c. Level 3 = “The GEF contribution is substantial, but modest indirect emission reductions can be The Top-down calculation must also adjust the 10-year attributed to the baseline,” GEF causality = 60 potential by accounting for “baseline shift.” Baseline percent shift is that part of the potential that would have been progressively achieved by the market even without a d. Level 2 = “The GEF contribution is modest, and GEF intervention. To make this adjustment, the GEF substantial indirect emission reductions can be causality factor is used. The GEF causality factor attributed to the baseline,” GEF causality = 40 describes how much of the buildup of capacity can percent 18 Overview for Applying GEF Tools and Methodologies e. Level 1 = “The GEF contribution is weak, and most Calculating the Local Co-Benefit of indirect emission reductions can be attributed to the baseline,” GEF causality = 20 percent Transportation Projects Wherever possible, local benefits that would be a While the GEF causality factor is useful and can deliver direct result of project impacts should be quantified consistent results, GEF causality factors should rely and included in the Project Document. In this manual, on situation-specific justifications and be estimated these are referred to as Local Co-Benefits. As noted in conservatively. If, in the future, the methodology shifts the above methodologies, the presence of significant to a different method of setting the baseline, the GEF local co-benefits in a project increases its likelihood causality factor could be simplified. of achieving success and the replication factor that The formula for calculating indirect impacts with the determines its indirect impact. Co-benefits include, Top-down approach is: but are not limited to: a. Travel time savings CO2 indirect TD = P10 * CF; with b. Expanded travel options and opportunities c. Job growth CO2 indirect TD = GHG emission savings in tonnes of CO2 eq d. Technical capacity building as assessed by the Top-down e. Economic development methodology f. Income growth P10 = technical and economic g. Additional employment potential GHG savings with the h. Air pollution reductions respective application within 10 years after the project (not i. Increases in physical activity that improve public health including direct and direct post- project impacts) j. User cost savings CF = GEF causality factor Wherever possible, the TEEMP models calculate savings in particulate matter linked to respiratory illness and safety issues like traffic fatalities. These are detailed in the specific methodologies in the following sections of this Manual. Any and all verifiable co- benefits which result from transport projects should be detailed in GEF project documents, whether calculated by TEEMP models or via another methodology. Overview for Applying GEF Tools and Methodologies 19 Step-by-Step Guide to III.  Estimating The Direct Impacts of Transportation Efficiency (Vehicle, Fuel, Network Efficiency) Projects Before Proceeding It is essential that the proponent read Section I (Introduction, Concepts and Introductions) and Section II (Overview for Applying GEF Tools and Methodologies) before moving forward. The core critical concepts, terminologies and foundations are detailed in those sections and are not repeated here. Unless the proponent is already quite familiar with GEF methodologies through prior experience, it is doubtful this current Section can be successfully navigated without first reading Sections I and II. In this Section you will be working with the following TEEMP model: EcoDriving_TEEMP.xlsx Introduction changes in demand due to the rebound effect—the change in the amount of fuel consumed due to the Transportation efficiency projects reduce the GHG increase in travel resulting from the reduced time- emitted per vehicle kilometer traveled by reducing cost of travel. The methodology does not account for the GHG intensity of: changes in transportation activity levels—such as motor 1. The vehicle operation, vehicle travel demand, trip length, or modal share. 2. The fuel, or Special care must be taken to evaluate whether there 3. The transportation network. will be any associated changes in service, speed, or pricing related to an efficiency project, as these are Transportation efficiency projects generally focus on likely to impact transportation activity. If such changes supply-side approaches to making existing transport are anticipated, then the proponent must also use the services, infrastructure, and behavior less GHG- Step-by-Step Guide to Public Transit Projects (Section intensive, rather than changing transport modes, IV) to calculate the GHG impact across the modes demand, or behavior. Examples of past GEF projects that will be affected. This choice has to be made early which would fall under this category include clean in the development of the project. The only projects vehicle projects that replaced diesel buses with fuel of this type that would not result in changes in travel cell buses, clean fuels projects, market development behavior would be technology projects that have no for electric plug-in two-wheelers, training programs impact to users in the cost or performance of the that taught mechanics how to improve fuel efficiency transportation mode. via engine tune-ups, and transportation network efficiency projects which may include coordinated Measuring net gains in GHG reductions from signal timing and enhanced real-time transit transportation efficiency projects is a complex process. dispatching and operations management. As efficiency factors are improved, GHG emissions are reduced. Yet, the benefit to the public, predictably, This methodology does not require the use of a TEEMP entices more travelers into the transportation system. model. It should be used to find the reduction in GHG This “give and take” dynamic must be quantified in a emissions in cases where an existing vehicle, mode, or GEF project. The TEEMP model is designed to bring network will be replaced or reconfigured to be more order to this process. GHG-efficient. This methodology can be used in conjunction with the BRT model, for instance, if more For example, an area-wide traffic signal system efficient buses are introduced. It accounts for simple coordination that boosts average network travel 20 Estimating The Direct Impacts of Transportation Efficiency Projects speeds by 10 percent is likely to induce a several to operate vehicles more fuel efficiently, a TEEMP percent increase in traffic as travelers find that the spreadsheet model is available to streamline the generalized cost (in time and money) of travel is calculation of direct GHG impacts by the program. lower, spurring travelers to drive more vehicle- The eco-driving model examines the effectiveness kilometers. Estimating complex interactions may be of implementing eco-driving training programs for a challenging analytic exercise, even with good travel passenger and truck drivers. It also examines the data and models. Where data and models are lacking, effectiveness of adding on-board display tools to the evaluation will have to rely on ad-hoc sketch provide real-time feedback on fuel efficiency to analysis while encouraging collection of better data drivers. and development of better models. The model uses effectiveness rates based on a study The GEF Alternative Scenario in some cases will that included U.S., European, and developing world simply identify the acceleration of emission reductions results. It has been documented that when reinforcing that would have happened anyway in the baseline lessons or tools are not applied, the effectiveness of scenario. For example, reduced emission intensities the training declines after the first year. For that reason, that would be reached in 10 years under a baseline the model includes a 66% reduction in effectiveness scenario could be reached in four years under a GEF in Year 2 for drivers who do not have on-board display Alternative Scenario. This has to be included in the tools. The user must specify the percentage of the GHG analysis, as the difference in the emission paths population reached by training programs, as well as of the two scenarios gives the cumulative emission the degree of penetration of on-board display tools. reduction of the GEF intervention. Keep in mind that to be consistent with past estimates and reduce The TEEMP model requires the user to select the the number of assumptions necessary, cumulative type of program offered, and the number of people emission reductions for GEF projects are calculated expected to be involved. It also allows the user to over the lifetime of the investment. input VKT, vehicle mode share data, and emissions factor data, if local data is available. Table A-1 in the appendix illustrates all data required and default Data Requirements values provided for Eco-Driving TEEMP Model. The formulas used to calculate the impact of Types of Programs transportation efficiency projects are designed to The user has the option of measure the amount of fuel saved by the project. The selecting from two categories basic data requirements include an estimation of the of training programs, with three amount of each type of fuel to be used (or saved) in levels of intensity for each: each scenario (baseline and GEF alternative). This estimation will be based on changes in the fuel/ 1. The “Structured Training consumption of the vehicles and/or networks. In order Program” refers to training to calculate this, fuel economy and VKT must be courses, generally targeted known. It may also be necessary to know passenger at relatively small groups of kilometers traveled, vehicle speeds, price/speed drivers. The levels of intensity sensitivity of travel demand, and other data, depending in the model, from least to on the project type. If the project focuses on vehicle or most effective, include: fuel technology, the specific emissions factors for the Basic Structured Training Program – Classroom a.  vehicle models affected by the project must be known program, in which participants are instructed and should be calibrated for local conditions. This is on ecodriving techniques via presentations, true for both the Baseline estimation and the GEF lectures, or videos. investment. General defaults used in other scenarios are not sufficient for this type of project. Hands-on Training Program – A classroom b.  program augmented with hands-on driving training, in actual vehicles or a simulator. Eco-Driving TEEMP Model Intensive Training Program with Benefits – A c.  For projects which program with the characteristics of a hands-on include eco-driving training program that also includes an incentive and/or implementation structure to reward drivers who implement the of on-board display course in practice. For instance, a commercial components, which fleet might provide bonuses to drivers who use instruct drivers how less fuel on the job. Estimating The Direct Impacts of Transportation Efficiency Projects 21 2. “General Marketing and engines are becoming more efficient over time. Program” refers to a mass- In cases where a technology already shows an upward market campaign, designed trend in usage, and the GEF projects will accelerate to reach a wide audience, but this trend, the baseline shift needs to be accounted not including a formal training for and described in the baseline scenario. program. This could include, from least to most effective: Clearly, this is a complex area of estimation. However, the process of completing the TEEMP models is Basic Outreach Program a.  designed to bring clarity to these items. with Information Brochures—A marketing campaign in which brochures or other materials with information on eco-driving techniques are Calculating Direct Emissions Impact distributed to the public. of Transportation Efficiency Projects Interactive b.  Marketing Program with In transportation efficiency projects, the direct Multimedia—A marketing campaign which emission reductions can be calculated in two steps: also includes interactive multimedia to engage the audience to a greater level than brochures mproving Vehicle Efficiency – For transpor- 1. I and static materials. tation network efficiency projects and projects which improve average vehicle fuel economy, Interactive (c)  Marketing Program with step one in calculating the direct emission Feedback—A marketing campaign that involves reduction is to multiply the projected fuel some degree of personal interaction with savings by the corresponding emissions factor marketers or trainers to reinforce the messages and summing this for each fuel affected. From and provide individualized information. this sum is subtracted the rebound effect, which is the estimated additional fuel consumed by traffic generated by the lower fuel cost of travel Baselines by this mode: Projects that intend to introduce standards or new technology for specific vehicles or sectors—such as CO2 direct = Σ Fx,y,z (Fx * cFx) – (Fr* cFx), taxis, private cars, or buses—can focus on the local market baseline for technology, and the developmental where trajectory the baseline would likely take in the market without GEF intervention. Typically, this baseline CO2 direct = sum of direct GHG trajectory already contains some planned initiatives emission savings from reducing use that would yield GHG reductions without a GEF of fuels x,y, z due to successful project intervention. This is what is referred to as “baseline implementation of project, in tonnes of shift.” In forecasting GHG emission reductions, the CO2 eq. effect of baseline shift must be accounted for as much as is reasonably possible. It cannot be assumed Fx =  amount of fuel x saved by the that the energy use and GHG emissions in a market intervention, and cumulated would remain the same in the baseline throughout over the lifetime of the respective the implementation of the project. investments, fuel savings are to be Whether baseline shift is an issue depends on the corrected by the “baseline shift,” situation in the country in question. In some cases, i.e., the amount of fuel savings the GEF project supports a technology that is not that would have happened due to currently available or used in the country. In that case, improved technology anyway, even baseline shift does not need to be accounted for, without a GEF intervention. except through the GEF causality factor in the indirect Top-down methodology. However, if a clean vehicle c = CO2 emission factor for fuel x program replaces a bus fleet with an average age of 10 years, the baseline must assume that the buses the rebound effect, or the amount Fr =  would have been replaced over time regardless, (most of fuel consumed by the increase likely at a rate to maintain this 10yr average age) and in travel resulting from the that the replacement buses would be more efficient reduced fuel cost of travel than those running in the base year because buses 22 Estimating The Direct Impacts of Transportation Efficiency Projects Evidence suggests that in the developed world, In cases where energy from the electric grid is involved the rebound effect related to improvements in fuel (e.g. the case of electric vehicles), the energy per vkt economy standards is relatively modest. However, (in watts) should be multiplied by an emissions factor no similar assessment has been made for emerging calibrated for the local power mix or the next power economies where the price sensitivity of demand plant to come on line. is generally much higher. So, further research is needed to develop some reasonable expectations As a default, the CO2 emission factor for additional with regard to projected rebound effects. power from a power grid should be for the marginal factor. “Marginal” refers to the emission factor for the mproving Fuel Efficiency – For projects which 2. I additional energy demanded (not the average of all involve the substitution for one vehicle fuel type the energy produced). In exceptional cases where with a different fuel that is less carbon-intensive grid electricity is being saved or supplied at peak (e.g. substituting hybrid or fuel cell buses for times, the emission factor can be an average emission diesel buses), or changes in the carbon-intensity factor. For example, if grid electricity is being saved, of the same type of fuel, the cumulative carbon the formula uses the overall average emission factor emissions must be calculated for both the of the local power sector, as opposed to the emissions baseline and the intervention scenarios: attributable to the next power plant to come on line. All emissions reductions are aggregated across all affected markets, modes, etc. for the expected useful CO2 direct = (Fintervention * cFintervention) – economic lifetime in years. If annual savings vary, sum (Fbaseline*cFbaseline) - Fr, with them for all years of useful lifetime. These economic lifetimes might be different for various vehicle types CO2 direct = direct GHG emission savings and intervention types. of successful project implementation in tonnes of CO2 eq., F= c  umulative fuel used, in appropriate Calculating Indirect GHG Impact in metric, cumulated over the lifetime of Transportation Efficiency Projects the respective investment The general guidance for calculating indirect impacts  O2 emission factor for Fuel F. This c= C provided in section II should be used for transportation emission factor should not only efficiency projects. Refer to page 22—Bottom-up include direct carbon content of approach. fuel, but also account for upstream greenhouse gas emissions connected Sometimes a project may include bringing to the with the extraction, production, and market a vehicle technology utilizing fuel with lower CO2 emissions that also lowers the price of fuel but distribution process for the fuels. increases vehicle procurement costs. In such cases, These factors will vary from locale it is known that—at least in the freight sector—the to locale, depending on fuel type, savings in fuel needs to recoup the increased vehicle refining source, distance from refining cost within 18 months or else the product will not sell.2 source, fuel raw material source1 This also assumes that fuel prices are sufficiently stable and raw material type. The GEF to project the economic value of these fuel savings. methodology provides a 14% default This dynamic is important to note when calculating factor which should be adjusted based the indirect impact of a project. on local data, where available.  ebound effect, or the amount of Fr = r fuel consumed or saved by the additional or reduced travel induced by the higher or lower cost of using the new fuel Synfuels, like gasoline produced from oil shales or coal, have 1 much higher GHG emissions than conventional crude oil derived fuels. As the crude oil is extracted from ever more challenging and higher cost sources, its associated GHG intensity is likely to rise as well. Thus sourcing of fuels should be accounted for in any analysis. comments by Cummings Engine Representative at MIT, 2008 2 Estimating The Direct Impacts of Transportation Efficiency Projects 23 Step-by-Step Guide to Estimating IV.  Direct Impacts of Rapid Transit and Railway Projects Before Proceeding It is essential that the proponent read Section I (Introduction, Concepts and Introductions) and Section II (Overview for Applying GEF Tools and Methodologies) before moving forward. The core critical concepts, terminologies and foundations are detailed in those sections and are not repeated here. Unless the proponent is already quite familiar with GEF methodologies through prior experience, it is doubtful this current Section can be successfully navigated without first reading Sections I and II. In this Section you will be working with the following TEEMP models: BRT.xls • BRT_MAC.xls • TEEMP-Railway.xlsx • TEEMP-MRT.xlsx • TEEMP-Roads.xlsx • Metro.xls Introduction more complex methods for estimating the emissions impact from BRT projects and the modal shift and There are a very wide array of Bus Rapid Transit (BRT) other changes they can spur in urban transportation and Mass Rapid Transit (MRT) systems worldwide, systems. The MRT TEEMP model enables users with widely varying performance metrics. GEF has to consider the energy characteristics of electric funded many BRT projects but generally avoided generation used to power electrified trains. funding metro and rail projects. Although these can The Railways TEEMP model also enables users to also reduce transport CO2 emissions, they tend to evaluate the impact of shifting a portion of freight have higher costs and longer delivery times. Freight from trucks to rail due to new or modernized railway mode-shifting from truck to rail and logistics efficiency lines and services, also considering energy sources initiatives also offer potential to curb transport CO2 and construction emissions. emissions, but have not yet been funded by GEF. Transit Projects generally create direct GHG impacts In the interest of facilitating consideration of a variety in five main ways: of GHG reducing strategies, this chapter provides references to CO2 analysis tools for BRT, MRT, and a. Induced modal shift resulting from new or improved Railways, while focusing its discussion primarily on the transit service. application of the BRT TEEMP tool. These tools and b. Total transit vehicle kilometers are reduced by methods could be extended to handle other forms of reorganized routes. mode shifting and logistics improvement in the freight sector and other aspects of system modernization and c. Fuel efficiency is increased due to improved transit operational enhancement in public transport. vehicle speed and operations. Due to the size, scale, and variability in BRT and MRT d. New or improved transit vehicles yield lower projects, creating an ex-ante estimation of their direct emissions per passenger-km due to more efficient impacts can be a very complicated, data-intensive vehicles and/or higher passenger capacities than the exercise. TEEMP models have been developed to vehicles from which the passengers were drawn. streamline this process for projects in the early planning e. The new system could impact land use changes by stages. The models increase consistency of methods stimulating higher density development around the and assumptions, without requiring high levels of data. system which in turn shortens future trip distances, The BRT TEEMP model offers both simplistic and reduces auto-mobility, induces modal shifts, and 24 Estimating The Direct Impacts of Mass Rapid Transit Projects (BRT & Rail) slows the conversion of land to urban usage on Data Requirements the periphery (Lane use changes are calculated as secondary direct impacts). The calculations used to find the GHG impact of mass transportation projects are based on existing bus These potential benefits have to be weighed against ridership in the corridor, the quality of the transit system construction emissions and any special emissions design, and operation variables (which determine caused by traffic impacts of the construction of the speed and shift from other modes). The basic data public transit system, which can be significant and are requirements include mode share, ridership, length accounted for in the TEEMP models. of routes, frequency, passenger trip length, as well as bus capacity, engine type, fuel and average speeds currently found in the corridor. Planning information BRT TEEMP Model regarding the length, route, capacity, and features Since GEF projects require a GHG estimation of the proposed transit project is also required. The before a project is implemented, and in some default values used by the model can be found in cases before detailed planning has begun, Annex Table 5 (A-5). the BRT TEEMP model has two modes of impact The model requires the following basic data about estimation that can be selected by the user: existing bus services on the planned mass transit • Short Cut BRT Method corridor, including: a. the total round trip length of each route, (input • Full Scenario Method onto the ‘BRT Operations’ worksheet) b. the km or percentage of the route that overlaps the project corridor, (BRT Operations) c. the peak hour frequency (BRT Operations) and average observed occupancy on the section of The user’s choice will depend on the amount of local the corridor most heavily utilized by buses OR data available. Users can click on which method they total boarding and alighting counts for each bus would like to use when opening the BRT TEEMP model route serving the corridor (BRT Operations) excel spreadsheet (“Model Choice” worksheet). d. the bus engine types (% of pre-Euro, Euro II, Depending on which method the user selects, they Euro III), entered onto ‘Tech%’ Worksheet’. will be guided through the model. e. the bus fuel type (petrol, diesel, CNG, LPG, hybrid, Shortcut Method etc) entered into the ‘Fuel Type’ worksheet. The Shortcut Method is a sketch analysis mode f. the buses capacity, (BRT Operations). which works as a very simple calculator. It multiplies g. average speeds, entered on the ‘Speed’ the proposed BRT corridor length times the average worksheet. certified emissions reductions from several previously implemented projects. This provides an order of h. average passenger trip length entered on the magnitude estimate for potential GHG reduction. The worksheet Trip. Shortcut Method is a very low-confidence estimate that may be appropriate at an early stage of planning, such as a PIF for a GEF project to scope or plan a BRT system. Projecting Ridership Ultimately, a more detailed estimate must be provided on the New System using the Full Scenario Method outlined below. The model measures the changes in emissions brought Full Scenario Method about by the introduction of a new mass transit system by first identifying the likely number of future riders The Full Scenario Method allows the user to input on the new system and making certain reasonable local and project-specific data for all data fields and assumptions about how they would have made the produce a higher-confidence GHG impact estimate trip if the new system never been built. It allows the of the project. While some data-points are required user to assume that, without the intervention, historical (green cells) for the Full Scenario Method, many other trends towards ongoing modal shift will continue data-points have default values (red cells) which can to occur. The benefits accrue because the potential be used if dependable local data is not available. passengers are presumed to generate far fewer CO2e These defaults are conservative, encouraging the emissions using the new system than they would have collection of local data. by using their previous mode. Estimating Direct Impacts of Rapid Transit and Railway Projects 25 Thus, the model first requires the user to generate an Some projects define a ‘direct service’ BRT system, estimate of the projected number of passengers the meaning that some bus routes will operate in mixed new system will serve. The project proponent has two traffic, enter the trunk BRT infrastructure, and then options for generating this ridership estimate: leave the BRT infrastructure. As such, the passengers • Input specific measures obtained from local using the system are likely to be a much greater surveys, or share of total bus passengers currently using the BRT corridor than would normally be the case for a closed • Use the default values provided in the TEEMP ‘BRT’ or MRT system. model. Average Speeds Ideally the ridership estimate should be based on a Average speeds can generally be collected by simply detailed operational plan which is then run as a scenario in an acceptable traffic model designed to handle this measuring them during the peak hour using GPS. type of demand analysis. The development of a clear Average Passenger Length operational plan, and the creation of a transit model for the system, are the best indicators of good The average passenger length is more difficult to project planning and greatly increase the likelihood collect, and using standard values from a household of project success. Where a full operational plan has survey or spot survey is acceptable. Alternatively, a been developed and a demand estimate made using an default value of 6 can be used. accepted traffic model (we recommend Emme III, Visum, Frequency and Occupancy Counts or TransCad, with other applications requiring review) no discount on projected demand should be applied. The peak hour frequency and average peak hour occupancy for each route can easily be collected by The following data contributes to the calculation of taking surveys at the ‘critical link—the most crowded ridership on the new system: part of the road. One approach is to simply count each bus and minibus servicing each route at the peak Price Change hour, and then estimating their average occupancy as If the new Mass Transportation System will introduce a percentage (25%, 50%, 75%, etc). This percentage is a change in the price of the mode, sketch modeling then multiplied by the bus type’s estimated capacity. that shows the elasticity of transit demand—induced by the price change—should be employed. These Another method utilizes video recordings (20 minutes estimates are calculated as part of the development of in length) captured periodically during the day. The the new system’s operational plan. However, the GEF video can be reviewed later in slow motion to arrive recognizes that the preparation of the new system’s at an estimated capacity. operational plan is frequently included as one of the more important functions of the eventual project. So Boarding and Alighting Counts the preparation of a detailed plan should not be made Boarding and alighting surveys, while more labor a precondition for receiving GEF funding. intensive than frequency and occupancy counts, are tremendously valuable to the project design. It is also recognized that, while inadvisable, the final At stations along the existing routes—or on board operational plan is frequently not decided until weeks the vehicles—riders are given the opportunity to before project launch. So, the model provides a simple complete a short survey that indicates basic details methodology for estimating future ridership that the such as the origin and end of their trip, the frequency applicant is required to follow in the planning phase. of their travel on this route and other simple details. The methodology recommended here is only reliable Paper or electronic systems can be used directly on plus or minus about 20%. For this reason, it is recom- board, and in some cases kiosks may be effective. mended that demand estimates be discounted 20% if this method is used rather than a traffic model. Station by station boarding and alighting counts can be aggregated and used to estimate the potential Existing Routes in the Project Corridor boarding and alighting numbers on the new system. When recording the existing bus and minibus routes on This is critical in designing the stations to avoid the planned new transit corridor, using data collected saturation. from departments of transportation is notoriously unreliable. So, it is recommended that the project Using the collected data, the project proponent should proponent collect this information directly by observing then clarify which of the existing routes identified which bus routes are actually using the corridor and are planned to be cut and replaced by new routes then using a GPS to record the coordinates of each bus in the BRT or MRT system, and which will continue route and bus stop that overlaps the planned corridor. operation outside the BRT or MRT system. If the 26 Estimating Direct Impacts of Rapid Transit and Railway Projects specific methodology to be used by the operational SF’ worksheet identifies all the different elements of planning team has been selected, then that a high quality mass transit system that are likely to methodology should be used. If the methodology is affect ridership and attributes to them a value totaling not determined, then a simple assumption should be up to a maximum score of 100%. Each point counts made that routes with greater than 50% of their length for a percentage of ridership bonus. overlapping the proposed corridor will be replaced This ridership bonus percentage is then multiplied by the new BRT or MRT system. This assumption is by the baseline ridership (BRT Operations) times the used only to determine a baseline demand for bus discount factor for unreliability fixed at 0.8, times the transport estimate. It will yield a low level of demand system type scaling factor ridership bonus to give the because it is not known whether the system will be estimated peak hour ridership. ‘open’ or ‘closed’, or have feeder routes. The new system’s passengers will not be all of the current bus If the other methodology was used—Frequency and passengers using the corridor but a subset of them, Occupancy counts—and maximum load on the critical because in any MRT or BRT project, some bus routes link was applied (total buses times average occupancy), are likely to be cut or scaled back, while other bus then deriving the total peak hour passenger demand routes remain to compete with the new system for requires also multiplying this number times the certain trips not well served by the new system. renovation rate (2.5 in this case). The frequency and occupancy counts will yield an To increase the ridership figure to a daily figure, estimated maximum passenger volume on the critical the peak hour estimate should be multiplied by a link estimate, whereas the boarding and alighting data default value of 10 unless full day bus occupancy and will give an estimated total of passengers using the frequency counts have been done to give a more link. The relationship between these two is called the accurate multiplier. If a more accurate multiplier is ‘renovation rate.’ The renovation rate is the number known, then this multiplier can be used if the data of times the bus turns over all of its seats in a single backing up this multiplier is submitted. route. If both types of data are available, then the To convert the daily estimated baseline demand for renovation rate can be calculated. If not, a renovation bus transport to an annual baseline demand for bus rate of 2 can be used as a default value. transport, one of two techniques can be used. • If occupancy and frequency counts have been Calculating Peak Hour Ridership conducted on a weekday, a weekend, and a using Boarding and Alighting or holiday, then the average daily demand for a weekday can be multiplied times the total Frequency and Occupancy weekdays in the year, the weekend demand by On the ‘BRT operations’ worksheet, peak hour the total weekends in the year, and the holiday boardings in the corridor are calculated initially by demand by the total holidays in the year. adding up all the boarding and alighting passengers • Alternatively, the daily demand can simply be (one boarding and one alighting equals one passenger) multiplied by 310. on all the lines which the operational plan determined are likely to use the BRT corridor. This initial number is At this point, a reasonable total annual trips on the then multiplied by a specific default value to allow for new BRT system has been estimated. This is the total the appropriate degree of uncertainty. number of trips impacted, and thus forms the basis of the CO2e calculation. Not all of the passengers on the new BRT or MRT system will come from the existing bus and minibus Calculating CO2e Emissions Using system, but most of them will. The question is how the TEEMP Detailed Model many additional passengers are likely to be attracted from other modes, such as cars, motorcycles, etc. Now that the baseline demand for bus transport has been estimated, the TEEMP model can calculate the The TEEMP model calculates the likely additional estimated bus kilometers of the new system, if the demand resulting from modal shift by multiplying majority of system details are known. If not, the sketch the directly impacted bus and minibus passengers function is used and default numbers are used. times a multiplier that is tied to ‘System Type SF’ (for ‘Scaling Factors’) worksheet. Based on empirical The user needs to first input the total length of the evidence collected from various BRT systems around BRT trunk corridor in both directions. Then, the user the world, an assumption is used to determine the must input the capacity of the planned buses or MRT maximum number of passengers that might be vehicles that the new system plans to use. Average attracted from alternative modes. The ‘System Type bus capacity is usually easily known. If not, a simple Estimating Direct Impacts of Rapid Transit and Railway Projects 27 formula can be used. Bus capacity is simply the length v. Operational control system to reduce bus bunching of the bus in meters less 3 meters for the driver and w. Extensive feeder bus services integrated into BRT engine, times ten, which is 10 passengers per meter of bus length (less 3 meters for the engine and driver). x. Integrated fare collection with other public A standard 12 meter bus thus yields a capacity of transport 90 ( [12 - 3] x 10 = 90) y. Peak-period pricing The TEEMP model then imports the projected average z. Performance based contracting for operators speed. If the system is an MRT fully segregated from aa. Passenger information at stops, surface traffic, then the design speed can be input. If headway > 5 min., info on vehicles the system is a BRT or LRT operating on an existing bb. Quality branding of Vehicles & stations road, the ‘System Type SF’ worksheet speed calculator cc. Brochures/schedules should be used. These components affect system speed and service On the ‘Operations’ worksheet, all of the BRT system quality. If a BRT has all of the above components design components that impact system speed have (100% of component points) it should achieve an been included. Each of these components carries a average operating speed of 30 mph.. A score of 90% score that is weighted by its ability to induce speed of points would yield a speed of 27 kph, etc. increases and modal change. If the characteristic is absent, a zero value is given. (These parameters can This projected average speed for the new BRT system be refined using local data). is then multiplied by the percentage of the total system that is operating inside the trunk corridor. The The characteristics are these: existing average bus speed is used for that portion of a. Dedicated right of way in central verge, w/barrier the BRT system that is operating outside of the trunk b. Station separated from junction by min of BRT infrastructure. From this the average speed for 70 meters the entire system is calculated. This average system speed is then imported back into the other relevant c. Passing lanes at station stops, if pphpd >6000 sections of the BRT Operations worksheet. d. Unique/attractively designed shelter At this point, the model can calculate the average fleet e. Weather protection at stations size needed, and the average peak hour bus kilometers f. Illumination operated. The only additional piece of information g. Security personnel at stations needed is the multiplier for converting the peak hour bus kilometers to daily bus kilometers. Because buses h. Stations =>3.5 m wide tend to operate with a lower occupancy off peak, i. Multiple docking bays w/ space to pass, the multiplier used for the bus kilometers should be pphpd <6000 higher than for the daily demand multiplier. If no local j. 3 or more doors information is known, a multiplier of 14 can be used. For an annual figure, the same multiplier used for the k. Boarding platform level with bus floor bus ridership can be used. Now, the model can give l. Safe & attractive pedestrian access system and estimated totals for daily and annual bus kilometers corridor environment for the new system. m. Bicycle parking at stations The ‘BRT Operation’ worksheet then takes this total n. Bike stations/bike rentals/public bikes at stations projected bus kilometer data and derives the CO2e o. Compliant w/ Access International BRT estimate of the new system from that. The model Accessibility guidelines provides some default values for the fuel efficiency of different vehicle types using different fuels at 50 p. Bike paths leading to stations kph. These are listed on the worksheet ‘Fuel Eff @ 50’ q. Service offered throughout day and should be used by the applicant unless a clear r. High frequency service < 5 min. avg. justification can be given for using different default s. Off-vehicle fare collection values. The model also provides default values for how fuel efficiency will vary for the same vehicle t. On bus camera enforcement of ROW type and the same fuel type at different speeds. It u. Turning restrictions across > 60% of intersections multiplies this fuel efficiency by a scaling factor linked (high volume) or bus priority at junctions to a projected speed (+ or – the fuel efficiency at (low volume) 50kmph depending on the actual speed). 28 Estimating Direct Impacts of Rapid Transit and Railway Projects The model multiplies the planned total bus kilometers household survey conducted in the city, if per day and per year for the bus type used by the new available. A preferable methodology is to create a system times the fuel consumption per kilometer at cordon around the planned BRT corridor. At each the estimated speed of the new system (drawn from road with any significant traffic volume entering the speed calculator mentioned above). This then is and leaving the planned mass transit corridor, simply multiplied by CO2e per liter of fuel, which is traffic counts and occupancy surveys should be taken from standard CO2e emissions factors supplied conducted in both directions during the morning by the model. and evening peak if possible. The total number of vehicle trips passing through the cordon are The model now knows how many passengers are likely then simply added up. The passenger trips are to ride the new system, how many new transit vehicle then multiplied by the average occupancy per kilometers the new system will create, and how much vehicle type to derive the total trips per mode. CO2e this will generate. The total CO2e that will be This method will give a reasonably site specific generated by the new transit system then appears in modal split without requiring too much work. the ‘BRT Operations’ worksheet. This figure is then car- ried forward to the ‘GEF CO2’ summary worksheet. b. Average occupancy data can usually be taken from the same traffic counts used to conduct the modal split counts, but secondary source materials on Estimating the CO2e Impact of average vehicle occupancy can be used. the Project over the No-Project c. Average speed data can be collected by riding up and down the planned corridor in a car or taxi Baseline Scenario and measuring the speeds during the peak and The model now estimates the amount of CO2e that is off peak hour using a GPS. Alternatively, average likely to be removed from the existing baseline traffic speed data can be used from secondary sources. system and projected future baseline traffic system. d. Average trip length can be taken from the latest The passengers on the new system will be diverted household survey, user surveys in the corridor, or from their existing trips where they currently use some secondary sources can be used. alternative mode and alternative vehicle. If the modes e&f. In terms of the engine type and fuel type of and vehicles these passengers were using before the bus and the non-bus modes, it is ideal to have new system was introduced generated more CO2 detailed records of vehicle type and fuel type than these same passengers generate using the new from the vehicle registry. If this is not available, system, then CO2 is reduced. whatever secondary sources are available can Before the CO2e impact of this change can be be used. Fuels and vehicle experts should be calculated, however, we need some additional consulted to derive better methodologies for information, which includes the following: estimating these when no secondary source materials are available. a. Existing modal split (share of trips made by each mode) data for the base year and projected modal The estimated CO2e benefit from drawing passengers split for 10 and 20 years after the base year needs to the new transit system from other modes is calculated to be input into the ‘Mode Share’ worksheet. on the ‘modal shift’ spreadsheet. The ‘modal shift’ worksheet imports the total ridership for the base year b. Average occupancy for all the non-bus modes from the ‘BRT Operation’ worksheet. In some projects, needs to be input into the ‘Occupancy’ there is only one type of bus involved. If, however, worksheet. these original passengers are drawn from multiple bus c. Average speed of all the non bus modes needs to types, then the passengers drawn from these modes be input into the ‘Speed’ worksheet. should be the same ones that were used to calculate d. The Average trip length for all modes including the demand on the new system, and should be drawn buses needs to be input into the ‘Trip Length’ in the same proportion. For example, if minibuses worksheet. constitute 60% of baseline public transit trips, and 12 meter buses constitute 40%, then the total new system e. Engine type of the non-bus modes needs to be passengers should be drawn from where they came input into the ‘Tech%’ worksheet. from, 60% to minibuses and 40% from buses. f. Fuel type of the remaining modes needs to be input into the ‘Fuel Type’ worksheet. Any additional passengers now need to be drawn from different modes. Since we do not know exactly which Some guidelines on each item follow. modes they would be drawn from, we just assume a. Modal split data can be taken from the last that they are drawn from private vehicles in rough Estimating Direct Impacts of Rapid Transit and Railway Projects 29 proportion to the preponderance of those modes in of the benefits were derived from energy efficiency the general traffic. On the ‘Mode Share’ worksheet, the improvements from reduced congestion in the mixed modal split of private modes is entered. The ‘mode shift’ traffic lanes. worksheet derives the trips drawn from private cars by simply subtracting the total bus trips from total trips, CO2e Generated in the Production of Vehicles and then multiplying the remainder by the proportion of private vehicle trips accounted for by private cars. Some CO2e will also be generated in the production of the new transit vehicles. However, CO2e will be These trips are then assigned to vehicles using the abated if modal shift results in fewer private vehicles average vehicle occupancy figures collected above consumed. No impact of this type was measured as it and recorded on the ‘Occupancy’ worksheet. These did not seem to add any additional useful information vehicles are then multiplied by average trip distances for the selection of better projects. Besides, the CO2 from the ‘trip length’ worksheet to yield the total from the production may have already been considered vehicle kilometers removed by vehicle type. Each if the vehicles were manufactured in another jurisdiction vehicle type was assigned an average speed from that is already under a carbon restraining regime. the ‘Speed’ worksheet. These vehicle kilometers per vehicle type at a specific average speed are then assigned an estimated fuel consumption based on Construction Emissions the fuel efficiency factors included in the model. An BRT construction emissions account for the emissions estimated CO2e emission per litre of fuel consumed generated during material production and construction is also based on factors included in the model. Each of infrastructure such as additional lanes, stations etc. of the withdrawn trips by mode—and their associated In general for BRT projects they are not that significant CO2e reduction—are then added up. Results appear in terms of total CO2e impact, but ideally they should in the ‘modal shift’ worksheet. be included. MRT construction includes highly energy intensive processes consuming vastly more construction GHG Impact of Shifting Passengers to Newer, materials, and neglecting construction-related CO2e More Fuel Efficient Buses production will fundamentally change the CO2e profile of the project. The TEEMP model has the capacity to capture a shift from dirtier and less fuel efficient buses to cleaner and The model includes some averages of typical tons of more fuel efficient buses. The engine type recorded by cement, steel and bitumen that are used per kilometer the user in ‘Tech%’, and the fuel type recorded in ‘Fuel in constructing some MRT and BRT projects. It relates Type’ affect the fuel efficiency of the vehicles used. In emissions to their production based on default values some cases, the vehicles are assumed to be the same taken from available literature. Ideally, the project vehicles as would otherwise operate in mixed traffic. proponent should collect data specific to the project, but the model allows for the use of default values which are simply multiplied by the length of the planned system. Impact on Mixed Traffic The worksheet ‘construct’ contains some baseline A BRT project will also have significant impacts on values of how much CO2e is likely to be generated mixed traffic. Currently the model does not account per ton kilometer of bitumen, cement and steel. If the for these impacts, though they could be highly total tons used is known for the project, these can be significant and complex. included here. If they are not known, then the user A BRT system could increase or decrease mixed traffic can use the default values included in the model. congestion depending on design and circumstances. This figure is then multiplied by the total proposed Some of the scaling factors included in the ‘System kilometers of trunk mass transit infrastructure on the Type SF’ worksheet are good indicators of mixed ‘construction’ worksheet. This yields the estimated traffic impacts, but there is not enough information project CO2e generated by the construction. This in the sketch planning tools to make this predictable is then reflected in the first year of the ‘GEF CO2’ in any reasonable way. For this reason, we agree worksheet. It is assumed that there will be no additional that the sketch model should not include any effort construction related emissions after year one. to measure these impacts without more complete demand modeling to back it up. Indirect Effects: Impact of Land Use Changes However, it should be recognized that where these BRT projects have been shown to spur land-use impacts can be captured, they can be sizeable, more intensification along their corridors1 which has a than tripling the CO2e savings benefit. In the Mexico City analysis done by Rogers (using Tranus), about 1/3 Cervero, Robert, Kang, Chang Deok: Bus Rapid Transit Impacts 1 30 Estimating Direct Impacts of Rapid Transit and Railway Projects resultant effect on decreasing VKT2. The TEEMP model both in additional corridors in Curitiba and in other accounts for land use changes by merely multiplying Brazilian cities. Its replication effect, however, stalled the total emissions savings from operations times after that. This is a multiplier of 4.4. a land use factor multiplier supplied by UNEP GEF based on very limited empirical evidence. To avoid • Quito was then built, also a high-quality, full- distortions all applicants are required to use the featured system for its day, and its Phase I inspired same multiplier. additional BRT km about 3.3 times the original Phase I kilometers, mainly in Quito and Guayaquil. The multiplier currently uses a 1.45 multiplier after 10 years and a 1.9 multiplier after 20 years. This figure • TransMilenio was then built, and it had by far will be gradually improved as the database from the biggest replication impact, with 17 times the original Phase Ia length being more or less directly which it is derived is improved. However, given the attributable to its inspiration, not only in Bogota unpredictability of land use impacts, for GEF purposes and other cities in Colombia, but all around this should be retained as a fixed multiplier to avoid the world. This was in part due to much more distorting outcomes. The land use factor might also aggressive international promotional efforts which be scaled by the same scoring process used in the clearly had an impact on hastening the replication, ‘system type SF’ worksheet to greater incentivize and in part due to a superior, second generation good design practice. technology. Alternatively, the land use impact could be dropped The way the dissemination factor works in the model all together. There is certain to be some land use is as follows. The average replication multiplier is impact of a good transit system, and a reasonable 8.4 for a BRT with all the recommended features. assessment of that impact is useful. However, at this If on the ‘System type SF’ worksheet, the system time too little is known about land use impacts to make receives a score of 80 out of 100 points or more, this a significant element for the GEF in determining then it receives a dissemination multiplier of 8.4 * the quality of project submissions. .01 times the total number of points. If the system receives a score lower than 80 points then it receives Special Notes for Calculating zero dissemination multiplier points. This roughly simulates the degree to which only systems of a very Indirect Impacts: Dissemination of high standard have proven to have any significant Mass Transit Best Practice dissemination impact. Good practice engenders replication and good prac- tice elsewhere. The world class TransMilenio BRT sys- Summarizing Total CO2e Results tem, for example, has inspired other cities to follow suit in spectacular fashion. Ignoring this increase in On the ‘GEF CO2e’ spreadsheet, the results are replication potential for high-quality, full-featured BRT summarized in the CO2 Emissions Savings table. systems would be to miss one of the most important Direct impacts are listed separately from indirect roles played by the GEF. Therefore, as part of this impacts. The direct impacts are calculated by taking model, a ‘Dissemination Rate’ worksheet has been the total emissions generated by the new system, and created tallying total global kilometers of BRT systems the emissions related to construction, and subtracting and linking them to the specific systems that inspired them from the emissions reduced from pulling trips them. This spreadsheet should be used to calculate off other modes. the Bottom-up, or lower extent, of the range of the The dissemination and land use multiplier is applied project’s indirect impacts. The general approach out- to the direct operational benefit as described in the lined in Section II should still be used to calculate the previous section. higher extent of the range of indirect impacts. The CO2 Emissions Savings table also includes a To illustrate the impacts of good practices on shortcut method of calculating CO2e benefits. successive projects: This method requires only the baseline estimate of • Curitiba Phase I was roughly 46.2 kilometers. This passenger ridership in the new system. This figure high-quality, full-featured system inspired the is then multiplied by the average CO2e benefit per passenger of all existing empirical data on BRT construction of 206.28 km of new BRT systems systems. Currently this data set is very limited. But on Land Uses and Land Values in Seoul, Korea, 2009 as the methodology for collecting these estimates is 2 Reid Ewing, Keith Bartholomew, Steve Winkelman, Don Chen: “Growing Cooler: The Evidence on Urban Development and standardized—and more data points are collected— Climate Change,” 2008. the reliability of this approach should improve. Estimating Direct Impacts of Rapid Transit and Railway Projects 31 Step-by-Step Guide V.  to Non-Motorized Transportation Projects (Bicycle & Pedestrian) Before Proceeding It is essential that the proponent read Section I (Introduction, Concepts and Introductions) and Section II (Overview for Applying GEF Tools and Methodologies) before moving forward. The core critical concepts, terminologies and foundations are detailed in those sections and are not repeated here. Unless the proponent is already quite familiar with GEF methodologies through prior experience, it is doubtful this current Section can be successfully navigated without first reading Sections I and II. In this Section you will be working with the following TEEMP models: Bikeways_TEEMP.xls • Bikeway_TEEMP-MAC.xls • Bike-sharing_TEEMP.xls • Pedestrian_Improvement_TEEMP.xls Introduction Estimating Direct GHG Impact Similar to transit projects, non-motorized transportation for Pedestrian Improvement (NMT) projects seek to induce modal shift away from Projects with TEEMP Model more GHG-intensive modes and toward bicycling and For projects that make walking, which are GHG-neutral. TEEMP models are an urban environment provided for use with Bikeways, Bike-sharing, and more walkable—be it Pedestrian Improvement. These models will guide by expanded sidewalks, the user through the steps necessary to estimate the block density, improved direct GHG impact of such projects. To estimate the crossings, or otherwise direct GHG impact without the TEEMP model, use improving pedestrian the no-project baseline scenario to compare against facilities—the Pedestrian project scenarios to find impacts. Improvement TEEMP model can be employed to estimate GHG impact. The calculation is done in two stages: Estimating Direct GHG Impact for Bike-Sharing Systems with • Calculating the No Improvement Scenario TEEMP Model (Baseline) Bicycle-sharing systems make a large number of bicycles available for public use. For projects that incorporate the development of a bicycle sharing system, the Bike-Sharing TEEMP model can be used to estimate the direct emissions impact. This simple spreadsheet model requires the user to input details • Calculating the Improvement Scenario (GEF about the scale of the system and the types of trips Alternative Scenario) avoided through modal shift, and then calculates the GHG impact of the system. 32 Non-Motorized Transportation Projects (Bicycle & Pedestrian) In the No Improvement Scenario, the user estimates total motorized trips at the starting year. A maximum the number of walking trips as a % of total trips. cap of 50% walking mode share is applied. The model assumes a decrease over time due to deteriorating pedestrian facilities coupled with The third option (c) works similarly as the second increased motorized traffic. The user has 2 options to option but the change in the walking mode share is define this decrease: calculated using the change in the walkability score. Pre-defined indicators are used in determining the a. The TEEMP model generates annual mode share walkability score. changes using the values inputted by the user. b. The mode share changes of the other travel modes Data Requirements (bus, auto, etc.) are automatically generated by for Walkability allocating the trips shifted away from walking Model Project to the motorized modes. The allocation of the Scenario: shifted trips is calculated on the basis of the size a. Streets with protected walkway with width portion of each motorized mode in relation to the adequate to accomodate pedestrian volume and total motorized trips at the starting year of the which are kept barrier free (including parked cars project. A minimum capping limit of 10% walking & hawkers) with non obstructing furniture. trip mode share is also applied as a check. b. Adequately safe crossing facilities (crossing lights, The shifted walking trips are segregated in the crosswalk striping, raised crossings, or accessible calculations. The number of shifted walking trips are grade separated as needed depending on traffic multiplied by the lengths of the walking trips. The volume) with active traffic calming. non-shifted trips are multiplied by the respective trip lengths as defined in the input (basic) sheet. These c. Streets with lighting. two sets are added to get the adjusted trips (total km d. Blocks/streets with shade/trees. traveled). The adjusted trips are multiplied with the respective emission factors (defaults) to arrive at the e. Block Size discount factor. final emissions for the No Improvement Scenario. f. Land Use Heterogeneity discount factor. For the Improvement Scenario, the same quantification concept is applied, but after the project. This Scenario The emissions in the No Improvement Scenario minus assumes that the number of walking trips as a % of the the emissions in the Improvement Scenario would total trips will increase over time. The increase can be give the emissions savings generated by the project. defined in three ways: a. Direct input of the mode shares at the final year of the project life Estimating Direct GHG Impact b. Assume an annual increase in the percentage of for Bikeways Improvement with walking trips out of the total trips TEEMP Model c. Determine the change in walking trip mode share by estimating the increase in the walkability score For projects that in the project areas incorporate the development of The model will generate the mode shares through bicycle lanes or time depending on the method chosen by the user. paths, the Bikeways TEEMP model can In the first option (a), the model generates annual % be used to estimate changes in the mode shares using the values inputted GHG impact. The Bikeways model has two modes of by the user. (This is similar to the No Improvement estimation: Scenario’s option a). 1. A shortcut “shortcut” Sketch Analysis Method For the second option (b), the changes in the mode to be used to generate an ‘order of magnitude’ shares of the other travel modes (except for biking, estimation of the impact if there is little local data, which is assumed to be constant throughout time) are and generated by distributing the trips shifted away from motorized modes to walking. The allocation of the 2. A “detailed” Full Model to be used to calculate mode origin of the shifted trips is dependent on the the impact if there is a high level of local data and % of the different motorized modes in terms of the project design details available. Non-Motorized Transportation Projects (Bicycle & Pedestrian) 33 The user can determine which method to pursue on Rio de Janeiro and Bogota. Thus, the proponent is the first worksheet of the model. The model will guide still able to estimate the impacts of such variables as the user from there. mode share, modal shift, trip lengths, etc. It is assumed that roughly, 1 km of bikeways would Data Requirements of attract 2173 trips. If narrow bike lanes are constructed with width less than 2m, the trips are scaled down Bikeways TEEMP model by 50%. Average trip length suggested as default by The Data Requirements for calculating GHG impacts the model is 6 km, and a 90% shift from public and for Bikeways projects vary depending on which Model intermediate public transport modes is assumed. The the user selects: user can vary the shifts to quantify the impacts using local data. Sketch Analysis Inputs – Width and length of Bikelanes, Average Bike trip length (6 km assumed as default) Future refinements of the bikeway model may incorporate other factors that are likely to be significant. Detailed Model for different scenarios – BAU – base These may include the population and employment year, BAU – Horizon year, With Project – Horizon Year density of areas served by the bikeway; whether the bikeway connects to larger networks; the degree to a. Average mode speeds - Cars, Two Wheelers, Three which the area served by the bikeway is pedestrian Wheelers, Taxi, Bus, Jeepney/RTV’s, Walking and and bicycle friendly or being made so as part of Cycling the initiative; the topography of the area; and other b. Vehicle Emission Standards for modes project elements that may provide added legitimacy and support for cycling in the area, such as car-free c. Fuel Type (Gasoline and Diesel) days, bike parking, and promotion programs. d. Mode share of modes - Cars, Two Wheelers, Three Wheelers, Taxi, Bus, Jeepney/RTV’s, Walking , Full Model (Detailed) Cycling and LRT Using the data e. Average Trip Length - Cars, Two Wheelers, Three supplied by the user Wheelers, Taxi, Bus, Jeepney/RTV’s, Walking and and ASIF logic, the Cycling Full Model tries to f. Average Occupancy capture emissions for both the Baseline and the GEF Alternative Scenario. g. Fuel Consumption at 50 km speed (kmpl) The emissions savings are highly dependent on h. Quantity of Cement, Steel and Bitumen/km the modal shift achieved, Trip Lengths and “stream speeds”. Using the base 50 kph speed emission i. Emission factors for Cement, Steel and Bitumen/ factors and estimated traffic speeds, the model first Ton (production) calibrates the emission factors and then processes the CO2 emissions. Air pollutants – PM and NOx—are For quantifying the emissions generated at the quantified in similar manners. construction stage, the projected quantities of cement, bitumen and steel are requested. The The emissions savings are assumed to be linear and emissions generated during the energy consumption thus using the base value and horizon year savings, for the production of these materials are calculated total emissions during the project lifetime are as construction emissions and are included in project quantified by the model. The emissions are increased analysis. This procedure may result in conservative by default 14% to include the “well-to-tank” upstream estimates because emissions generated due to GHG emissions typical for motor fuels. The outputs material movement, construction machinery usage, include: traffic diversion etc. are not included. • Total emissions from scenarios, Sketch Analysis • Total savings over lifetime, and (Short cut) • Tons/km/year savings due to bike lane In cases where no local construction. data is available, the Sketch Analysis is useful Figure 6 shows the structure of the TEEMP bikeway tool. Default values are model. drawn from case studies of successful projects in 34 Non-Motorized Transportation Projects (Bicycle & Pedestrian) Figure 6: Structure of Bikeways Emissions Impact Model Input Data Sketch Analysis Detailed Model Construction Business With Data As Usual Project Base Horizon Horizon Year Year Year CO2, PM NOX - Emissions Developing a Baseline for NMT a. Estimate the growth trend of travel in this impact area using corresponding historical transportation Projects Without TEEMP Model data if available or by applying a growth factor A “quick but reasonable” baseline must be established based on related trends such as regional trends, in the application phase of a GEF Project. If the TEEMP land use forecasts, etc. for a no-project scenario. model is not used, the baseline can be developed using Take into account capacity limits. a combination of any existing local data. The general b. The baseline inventory should be calculated sequence in establishing a baseline is as follows: over the lifetime of the project. Annual a. Define a project impact area. The project impact emissions inventories are then summed to find area includes any area where there will be a traffic the cumulative emissions for the no-project impact from the project. It will most likely loosely scenario over the lifetime of the proposed project follow the corridor of the project, including the for comparison: corridor itself and any and all competing corridors. b. Estimate the number of trips, average trip distance, and vehicle occupancy occurring within CGHGNP = Σ n YGHGn, with the project impact area for all modes during peak and non-peak hours. Data for this estimation can be CGHGNP = Cumulative GHG derived from local traffic counts and travel surveys. Emissions for no-project scenario over lifetime of proposed project c. Apply emissions factors and vehicle fleet (years 1-n) data. If local vehicle fleet and emissions factors (including upstream emissions) are not available, AGHGn = Annual GHG Inventory use the GEF Transportation Default Values. for all years of project lifetime d. Enter values into the following formula: Pick Y or A in the formula. YGHG1 = Σ Tx,yz…[Tx *cx], with YGHG1 = Yearly GHG Inventory for year 1; Tx= yearly VKT for mode/vehicle X; cx = Emission factor of mode/vehicle X Non-Motorized Transportation Projects (Bicycle & Pedestrian) 35 Step-by-step Guide for Travel VI.  Demand Management Projects Before Proceeding It is essential that the proponent read Section I (Introduction, Concepts and Introductions) and Section II (Overview for Applying GEF Tools and Methodologies) before moving forward. The core critical concepts, terminologies and foundations are detailed in those sections and are not repeated here. Unless the proponent is already quite familiar with GEF methodologies through prior experience, it is doubtful this current Section can be successfully navigated without first reading Sections I and II. In this Section you will be working with the following TEEMP models: Pricing_TEEMP.xlsx • Commuter_Strategies_TEEMP.xlsx • PAYD_TEEMP.xlsx Introduction a baseline must be created which quantifies the emissions for the area and modes that are affected by Travel Demand Management (TDM) Projects include the proposed project. This can be developed through an array of strategies that use modal shift to either a combination of any existing local data. reduce the demand for transportation or encourage more efficient consumption of transportation resources Once the baseline is known, the degree to which the through modal shift. Strategies include: project will reduce that baseline transport activity and GHGs must be estimated to find the direct • Various transport pricing schemes, emission reduction. • Integrated transport and land use planning, The general sequence in establishing a baseline is as follows: • Parking management, a. Define a project impact area: The project impact • Car-sharing programs, or area includes any area where there will be a traffic • Encouraging telecommuting, among others. impact from the project. b. Estimate the number of trips, average trip distance, and vehicle occupancy occurring within Data Requirements: the project impact area for all modes during peak Baseline Calculations and non-peak hours. Data for this estimation can be derived from local traffic counts and travel For all GEF projects, a quick but accurate baseline surveys. must be established in the application phase. If a TEEMP model is employed, the user can disregard c. Apply emissions factors and vehicle fleet the steps described below because the TEEMP will data: If local vehicle fleet and emissions factors automatically calculate the baseline. (including upstream emissions) are not available, use the GEF Transportation Default Values and To estimate the emissions impact without a TEEMP, apply them to this formula: a no-project (no GEF or co-financing investment) 36 Travel Demand Management Projects values derived from empirical experience. These tools can be customized with locally-available data where YGHG1 = Σ Tx,yz…[Tx *cx], with it provides a better basis for analysis, and they will be enhanced over time by better documentation of YGHG1 = Yearly GHG Inventory for global empirical experience. year 1; These models are provided within three Excel files, Tx= yearly VKT for mode/vehicle X; corresponding to the main model type: cx = Emission factor of mode/vehicle X 1. Commuter Strategies 2. Pricing 3. Pay As You Drive (PAYD) This note applies in two places in this Section and in The following sections describe each model and other Sections. The “Y” and the “A” in these formulas worksheet. represent distinct and different values. Making these accurate requires someone with an intimate Commuter Strategies understanding of the purpose of the formulas. (Employer-based Strategies) a. Estimate the growth trend of travel in this impact There are four TDM strategies area using corresponding historical transportation covered in the ‘Commuter data if available or by applying a growth factor Strategies’ TEEMP. Each is a based on related trends such as regional trends, separate worksheet (tab) within land use forecasts, etc. for a no-project scenario. the same Excel file: Take into account capacity limits. 1. Employer support programs, b. The baseline inventory should be calculated over the lifetime of the project. Annual 2. Telework, emissions inventories are then summed to find 3. Compressed work week, the cumulative emissions for the no-project 4. Commute Strategies (Rideshare/transit subsidies) scenario over the lifetime of the proposed project for comparison, using this formula: Table A-2 in the appendix illustrates all the data required for and defaults provided for the Employer- Based Commute Strategies TEEMP module. The worksheets for these four areas are described more CGHGNP = Σ n YGHGn, with fully in the following sub-headings CGHGNP = Cumulative GHG 1) Employer Support Programs Emissions for no-project scenario (Transport Support) over lifetime of proposed project The Employer Support Model (years 1-n) (tab) examines the effect of employer support programs AGHGn = Annual GHG Inventory which encourage employees for all years of project lifetime to utilize alternative modes. These may include provision of an on-site transportation coordinator, ride-matching, transit information, and other Calculating Direct GHG Impact actions aside from time and for Commuter Strategies, cost incentives. Parking Pricing, Pay-As-You-Drive For a regional analysis, necessary inputs include both the existing and the alternative scenario participation Insurance using TEEMP Modules rates (percent of employers participating) by program The TEEMP models developed for various commuter level for each mode. Program levels of “1” through strategies, parking pricing, pay-as-you-drive insurance. “4” indicate varying levels of effort for the programs. These tools are all based on simple elasticity analyses Alternatively, levels of support and participation rates applied to a market share framework, using default can be defined for office employment compared to Travel Demand Management Projects 37 non-office employment. This approach is borrowed 2) Telework from the US EPA Commuter Model, developed in part by Cambridge Systematics. The Telework Model (tab) examines the effect of Examples of the four levels, which can be customized employers implementing to match local support program conditions are: teleworking policies. The user can apply multiple a. Level 1 = Employer provision of baseline assumptions, including the information activities (transit fare and route share of employers where information, rideshare matching, etc.) teleworking would be a feasible practice and the share of b. Level 2 = Level 1 plus employer assisted carpool/ employees at these employers vanpool matching, work hours flexibility, bike that would telework. parking and shower facilities. The TEEMP model calculates VKT reduction from c. Level 3 = Level 2 plus preferential carpool parking, teleworking based on share of jobs amenable to vanpool development and operating assistance, teleworking and rates of participation, with a 25% transit pass sales, secure bike parking. rebound effect offset to account for additional non- d. Level 4 = Level 3 plus additional financial and commute trips taken by teleworkers on their non- commute days. Specific values used in the analysis technical support, guaranteed ride home, include: promotional activities. • Existing rates of telework participation, VKT reduction impacts of support activities act as a multiplier to the effectiveness of employer incentive • Average days per week teleworking, programs. Thus the results of employer financial incentives are multiplied by the VKT reduction • Total working days per year, effectiveness of support programs. Effectiveness • Average round trip commute length, estimates are based on a matrix evaluation of a full range of before and after participation rates in the • Rebound effect offset percentage (workers make Commuter model. trips from home on telework days that were previously chained with the work trip) and The critical inputs impacting employer support program strategy results are: • SOV commute mode share. a. Total employment, All of these values can be user specified based off local characteristics. b. Baseline and Alternative Scenario participation rates and The critical inputs for scenario testing are: c. Baseline and Alternative Scenario support a. Existing and scenario rate of telework participation, levels. and In many US cities there are trip reduction ordinances, b. Existing and scenario average days of telework or Transportation Management Associations which per week. support implementation of support programs. The successes of the incentive programs are tied directly Surrounding conditions play a key role in determining to employee exposure, knowledge and ease of use of both the potential for teleworking and its actual the programs that are available. utilization. Despite a high level of interest for teleworking—expressed by most employees, as well In developing world urban locations, the potential as by a growing number of employers—the share of level of deployment for employer support programs regular teleworkers is still relatively low (5 percent or is predominantly tied to the size of the employer, less) in most countries. This is due to a combination industry type and scale (i.e. local, national, of factors, including technological and economic continental or international). The inputs of Baseline barriers, legal and administrative barriers (such as lack and Alternative Scenario participation rates should of permission to telework from the company or lack of take into account these types of characteristics of all approval from the superior), and the perceived need the employers in the region, city, or district within for physical presence and face-to-face interaction in a which the strategy is tested. number of jobs. 38 Travel Demand Management Projects There are a number of research reports on potential ridesharing and transit. These programs work best ranges of teleworking participation worldwide that to encourage commuters to switch from driving can serve as good references on potential scenarios alone to carpooling or transit in dense employment for testing. Alternatively, given a specific employment districts where alternative modes to driving alone are profile for a region and a knowledge of the existing available, traffic congestion is a significant challenge policy and technology environment, regional specific and parking is at a premium. inputs can be entered. The model differentiates among the response expected for offices located in low density suburbs, 3) Compressed Work Week activity centers, and CBDs. The model calculates the percentage reduction in VKT, based on the rate of The Compressed Work Week employer participation, for each $1(USD) increase in Model (tab) examines the daily subsidy provided. This estimate is based on a effect of shifting workers Victoria Transportation Policy Institute (VTPI) study to shorter workweeks while of expected vehicle trip reduction in response to maintaining the same total different subsidies in the U.S. work hours per week or two week period (such as 4-day The critical inputs impacting employer financial workweeks or 9 days of work incentives strategy results are per two weeks). a. Total employment, base and scenario participation Average VKT reduced per rates and week per worker who formerly drove is based on average daily round trip commute length. The b. Base and scenario subsidy levels. reduction is offset by a rebound effect (estimated at 25% per US research) similar to the approach for In many US cities there are trip reduction ordinances, telework. or Transportation Management Associations, which support education and implementation of incentive or The critical inputs are: subsidy programs. The successes of these programs are tied to these management or regulatory practices. a. Existing rates of participation b. Scenario rates of participation The trip reductions used in this approach are estimated by VTPI, and include the impact of these Other local commuting characteristics such as external factors. In developing world urban locations, total employment, SOV commute mode share and this regulatory and supportive institutional framework average round trip commute length are required for may not exist, suggesting that the effectiveness of the calculation. The VMT reduction per week varies subsidy programs may be less than the US example. depending on a 4/40 (e.g., a 5-day 40-hour workweek In addition, competitive supporting infrastructure compressed into 4 days) as opposed to a 9/80 such as public transit, or programs such as regional compressed schedule (i.e. the average weekly impact ridesharing databases, may not exist, meaning that of a 9/80 is 50% of a 4/40). the effectiveness rates could, on average, be less in The same caveats at play for the teleworking strategy these regions. The consideration of user inputs and are relevant for compressed work weeks, except that understanding of results from the model should take barriers with regard to technology or costs are not these factors into account. an issue. Therefore this strategy is highly reliant on employer policy and willingness to offer flexible work schedules to employees. In addition, users should be Parking, Pricing, and sure to modify the constants for number of work days Company Car Programs per week and year depending on local practice (such as whether work weeks are customarily 6 days instead The Pricing TEEMP contains of 5). worksheets (tabs) for calculating the direct GHG 4) Commute Strategies impact of TDM programs that (Rideshare/Transit Subsidies) focus on: a. Parking, The Commute Strategies model (tab) examines the effect of providing new incentives (subsidies), b. Pricing, and or increasing existing incentives to commuters for c. Use of company cars. Travel Demand Management Projects 39 Table A-4 in the appendix illustrates all the data uses a base parking pricing required—and defaults provided—for the Pricing elasticity of -0.15 (from TEEMP module. Shenzhen, China), which indicates a decrease of 1.5% The worksheets for these three areas are described of VMT as a response to 10% more fully in the following sub-headings increase in parking price in the CBD. For parking converted Company Cars: Employer-Provided Vehicles from free to paid parking, the model uses a base -0.2 elasticity (e.g. 20% reduction in trips that made The Company Cars Model use of the free parking spots that are now priced) (tab) examines the effects based on VTPI data. The model allows the user of reducing or eliminating to implement separate policies for 4-wheel versus subsidies associated with the 2-wheel vehicles. provision of company cars. Outside the U.S., company To account for local characteristics, scoring factors cars are a benefit often within the model are factored into the calculation of provided to employees. strategy effectiveness through a lookup process that Since employees do not pay the costs of owning or modifies the base parking price elasticity upwards or operating the cars, they have little or no incentive to downwards based on the combination of three unique reduce costs by limiting driving. This model examines region factors: two policy options: a. City Size – Characterized either as “Large” a. Eliminating the company car. The impacts will (generally the top tier metropolitan regions include some measure of increase in private auto with an international presence) or “Small” (the travel. second tier, rapidly developing cities serving as subnational or regional economic generators). b. Keeping the company car but eliminating free May also follow official national classification fuel for non-business travel. schemes such as India’s Compensatory City The model uses elasticities for response to elimination Allowance. of company cars and elimination of the free fuel benefit b. Parking Location – Characterized by “Urban Core” to calculate VKT reduced. This analysis is based on (Central Business District), “Near Core” (other a United Kingdom Revenue and Customs evaluation regional employment centers), and “Suburb” report on company car tax reform. Since households (regional activity or town centers). in the U.K. are more likely to have a private car available to replace the company car than households c. Transit Level of Service – Characterized as “High” in the developing world, this analysis would likely be (presence of a large Metro system), “Medium” a conservative estimate of GHG reductions possible (small or no Metro system, may have BRT or other from this strategy. high capacity bus transit), and “Low” (local and regional bus services only). Since the basis for application of the company car model is a study of company car tax reform in the UK, In general, larger city size, more densely developed factors like vehicle ownership and level of employees location and more robust transit system results in a who are accorded provision of company cars is likely to more sensitive price elasticity. be different from cities in developing countries. Due to scant data on company car mileage in the developing The worksheet is set-up to accommodate the impacts world, data from UK and Canada has been used to of parking pricing for three income groups – low, determine the relationship between regular passenger medium, and high (representing the lowest third car mileage and company car mileage. This is heavily of households in income, the middle third, and dependent on policies guiding usage of company the highest third). This stratification is designed to cars for private use, and hence is very policy specific. capture the differing levels of price sensitivity of these Other policies regarding provision of free fuel for non- groups—lower income travelers are more sensitive to business travel might vary from case to case. price increases, and thus will respond with greater VKT reduction. Where different stratifications will be more useful, the user may aggregate finer income groups Parking Pricing or divide into groups relative to median income or The Parking Pricing Model (tab) examines the effects actual income data through acceptable means of data of increasing parking fees in urban areas. This analysis collection. 40 Travel Demand Management Projects This worksheet does not account for the possible b. Transit Availability – Characterized as “High” effect of drivers switching from four-wheelers to (presence of a large Metro system), “Medium” two-wheelers to mitigate the effect of parking price (small or no Metro system, may have BRT or other increases. In the U.S., it is generally assumed that the high capacity bus transit), and “Low” (local and trips are shifted to non-auto modes (transit, bike/ped), regional bus services only). or are not taken. In the developing world, however, since parking for two-wheelers is on the order of 25- In general, larger city size and more robust transit 50% of the cost of parking a car, it is possible that system results in greater sensitivity to parking some car trips would be converted to 2-wheeler trips availability. (as an inexpensive way to deal with the increased cost One of the primary design factors that determines of parking). the number of parking spaces per unit floor space available is the peak demand for parking utilization. Finally, the effectiveness of parking pricing policies The Parking Density model bases its parking space depends heavily on the degree to which parking availability on this peak demand. Since such parking laws are enforced in an urban area. If drivers can park design factors vary across different countries, the user on the sidewalk with impunity, they are not likely to should be sensitive to the basis of determination of pay for parking. The input “share of total parking parking availability and demand. Availability of parking affected by fee increase” can be used as a proxy for by employees can be used as a proxy to availability the tolerance to illegal parking, with the relative share by floor space or area, given the ability to adequately of parking affected decreasing in urban areas where estimate the number of employees working in the illegal parking is not effectively controlled. targeted area. Parking Density (Availability) Pay-As-You-Drive (PAYD) The Parking Density Model (tab) examines effect of The PAYD TEEMP examines reducing the number of the effects of turning the fixed parking spots available in the costs of auto insurance into a Central Business District (CBD) per-mile (variable) cost. per square foot of office space, or per employee. Thus it can The PAYD Model’s result be implemented as a parking is calculated by adding the cost of PAYD insurance spot reduction, or a parking to the per-mile cost of driving, and using the price spot freeze if the amount of elasticity of VKT to calculate a reduction in VMT. This office space and employment in the CBD is growing. price elasticity of VKT is stratified by income level, to account for drivers’ increasing price sensitivity Our model uses a North American elasticity for as incomes decline. The user should input driver the effect of parking availability on number of trips participation by income category, with the income taken, obtained from a Canadian study which is cited categories determined by regional or national income in Transit Cooperative Research Program (TCRP) distribution: 95—Parking Management and Supply Traveler Response to Transportation System Changes. a. Low income represents the lowest third of households To account for local characteristics, scoring factors within the model are factored into the calculation of b. Medium income represents the middle third of strategy effectiveness through a lookup process that households modifies the parking availability elasticity upwards or c. High income represents the highest third of downwards based on the combination of two unique households regional factors: The analysis uses a U.S.-based price-elasticity of VKT, a. City Size – Characterized either as “Large” but with default values of insurance and the price of (generally the top tier metropolitan regions driving based on developing world data. The user with an international presence) or “Small” (the can specify different values for insurance costs and second tier, rapidly developing cities serving as elasticities if known. sub-national or regional economic generators). May also follow official national classification The user can also specify different shares of the schemes such as India’s Compensatory City population that participates in PAYD, depending on Allowance. whether the policy is voluntary or mandatory. For a Travel Demand Management Projects 41 voluntary system (e.g., a system in which drivers have The formula for calculating the direct emissions for a choice of purchasing other types of auto insurance other TDM projects is: instead of PAYD insurance), we again use a U.S.-based default value of 30% participation. For a mandatory system, it is assumed that drivers seeking insurance can only purchase PAYD. Table A-3 in the appendix CO2 direct year 1 = Σ Tx,y,z [(1- R x )* Tx illustrates all the data required for and defaults *cx], with provided for the PAYD TEEMP module. Tx= VKT for mode/vehicle X The cost of driving (expressed as the total cost per mile, including fuel, maintenance, and vehicle R x= Reduction factor for travel depreciation or capital costs) and the cost of insurance activity of mode/vehicle X due to vary considerably around the world, and users should TDM program in year 1 input local data wherever possible. Also, it is important to note that given the relatively low rates of insurance cx = Emission factor of mode/vehicle among the driving population in the developing X world, insurance reform (mandating more consistent levels of insurance) might need to be part of a PAYD policy. PAYD policies are most effective when purchasing The reductions for each year of the projects’ life should insurance is mandatory in order to own a vehicle. then be summed together to find the cumulative If insurance is optional or not widespread, PAYD emissions reduction of the TDM project: effectiveness will be reduced since higher mileage drivers may choose not to purchase insurance at all to avoid the added cost per mile. CGHGP = Σ 1-n CO2 direct year n, with CGHGP = Cumulative GHG Calculating Direct Emission Emissions for the project scenario Reductions for Other TDM over lifetime of proposed project Projects (years 1-n) TDM projects generally increase transportation GHG CO2 direct year n = CO2 direct efficiency by reducing the demand for—or distribution impact for each year of project of—transportation activity. The impact of the TDM lifetime project on the transportation sector must employ appropriate project-specific methods – surveys or the use of local models, etc. – to reliably estimate the effect of the TDM strategy on the transportation sector in question. An emissions factor is then applied to the changes in transportation activity and the direct GHG impact of the project is known. 42 Travel Demand Management Projects VII. Step-By-Step Guide For Comprehensive Regional Transport Initiatives Before Proceeding It is essential that the proponent read Section I (Introduction, Concepts and Introductions) and Section II (Overview for Applying GEF Tools and Methodologies) before moving forward. The core critical concepts, terminologies and foundations are detailed in those sections and are not repeated here. Unless the proponent is already quite familiar with GEF methodologies through prior experience, it is doubtful this current Section can be successfully navigated without first reading Sections I and II. In this Section you will be working with the following TEEMP model: Expressway_TEEMP.xls Describing the Baseline and the The data requirements for this baseline include: GEF Impact Case a. Two to three modal splits are desired: a recent modal split and 1-2 modal splits that pre-date the Comprehensive Regional Transport Initiatives involve most recent modal split data by approximately five the coordination of multiple strategies – at least three years and/or ten years respectively. The historical from different transportation sub-sectors –that have data is used to track trends in the transportation mutually reinforcing impacts and are implemented sector and project the future growth of emissions in concert to reduce the GHG intensity of a regional in the no-project (no GEF or co-financing transport sector. This approach focuses on strategies investment) baseline. that are complimentary and synergistic, allowing the impact of each project component to leverage greater b. Average trip distances by mode (if possible also impacts of accompanying components. by trip purpose) are also desired to accompany all modal splits unless specific distances can be A simple example of this approach could be a taken from modeled project specific traffic system strategy combining increased residential density impacts (if available). along new BRT corridors with new parking pricing program in the central business district. Each of these c. The mix of the vehicle fleet by vehicle type is also components can reduce transport sector emissions, desired. but when implemented in concert they leverage the efficacy of each other. For this very reason, the d. If freight transport is targeted in the comprehensive comprehensive approach is considered to be highly strategy, then data should be provided for all effective, although quantifying the impacts becomes relevant freight modes in the baseline. more complicated. This data is best derived from recent household Comprehensive Regional Transport Initiatives are best origin-destination surveys if possible. If these are not evaluated by the assemblage of a comprehensive available, use vehicle and vehicle occupancy counts ex-ante baseline including historical trends for the around, within, and across cordons. Once assembled, region. This baseline should be submitted in two the data can be combined with appropriate emissions forms: one baseline, which includes all walking factors to create a transport sector emissions inventory and bicycling modes, and other non-motorized for the region, based on a simplified ASIF philosophy of transportation (NMT) modes, and another that quantifying for transport sector emissions, which relies excludes all NMTs. on per kilometer emissions factors for various vehicles: Comprehensive Regional Transport Initiatives 43 The second step is to sum the direct lifetime emissions impacts for each project components with AGHG1 = Σ Tx,yz…[Tx *cx], with direct impacts. AGHG1 = Annual GHG Inventory The third step, unique to comprehensive regional for year 1 transportation initiatives, is to apply a leveraging factor to the total lifetime emissions reductions for all the Tx= yearly VKT for mode/vehicle X components, which recognizes the enhanced efficacy of a comprehensive and synergistic approach. The cx = Emission factor of mode/ leveraging factor should be determined by an expert vehicle X and justified within the text of the Project Document. The following guide breaks down the extents of leveraging factor awards for comprehensiveness of The baseline inventory should be calculated over strategies. the lifetime of the project, requiring at least two Leveraging Factor data points (project start and finish) to interpolate the annual emissions over the lifetime of the project. Minimum - 10% Low Leveraging Factor – project These annual emissions inventories are then summed components will have mutually to find the cumulative emissions for the no-project reinforcing synergistic effects on one another, but countervailing scenario over the lifetime of the proposed project for actions will undermine this effect. comparison: Maximum - 30% High Leveraging Factor – Project Components will have highly significant mutually reinforcing CGHG = Σ 1-n AGHG1-n, with synergistic effects on another without interference from other CGHG = Cumulative GHG countervailing policies or actions. Emissions for no-project scenario over lifetime of proposed project (years 1-n) A high leveraging factor shall be used only if, in the applicable metropolitan area, there are no planned AGHG1 = Annual GHG Inventory or currently underway major transport sector for year 1 investments or policies that might undermine the synergistic impact of the comprehensive approach that is being proposed. Transport investments and policies that should be considered as undermining a comprehensive approach include: Calculating Direct Emission Reductions a. Motorway expansion or flyover development for private vehicles expected to increase lane-km of The previous sections of this Manual discuss how motorways or flyovers in the impact area or region to find the direct emissions reduction for the vast by more than 5% in the next decade. majority of transportation efficiency interventions that would be funded by the GEF. Comprehensive b. Any planned increase in direct or indirect motor regional transportation initiatives are likely to have fuel subsidies or tax reductions. components—such as a public transportation or NMT strategies—for which direct emission c. Any planned increase in parking requirements or reduction estimation methodologies are found in this subsidies for new parking developments. document. d. Any new restrictions to limit non-motorized The first step of calculating the direct emissions vehicle travel in the region or impact area. reduction for a comprehensive regional transportation A low leveraging factor shall be used if these conditions initiative is to calculate the direct emissions impact are not met but an otherwise comprehensive strategy of each component of the initiative separately is being advanced. according to the methodology outlined in this guide. 44 Comprehensive Regional Transport Initiatives VIII. Appendices: TEEMP Model Data Defaults & Sources Appendix 1: Data Required and Defaults Provided for Eco-Driving Module Telework Model Acceptable Means Data Point Default Value Source Remarks of Collection VKT by mode Cars 85,000,000 Dummy Value Input 2W 22,000,000 Dummy Value Input 3W 13,000,000 Dummy Value Input Taxi 19,000,000 Dummy Value Input Highway Statistics Bus 3,600,000 Dummy Value Input Data, Regional Studies, Surveys with Jeepney/RTV 7,700,000 Dummy Value Input inventory information Walk — Dummy Value Input from Vehicle Registrations Cycle — Dummy Value Input LRT — Dummy Value Input Medium Freight Truck 10,000,000 Dummy Value Input Heavy Freight Truck 40,000,000 Dummy Value Input Passenger Percent of population Penetration Rates from reached by Ecodriving 10% Dummy Value Input Studies training programs Percent of population with Penetration Rates from 0% Dummy Value Input on-board display tools Studies Freight Percent of population Penetration Rates from reached by Ecodriving 10% Dummy Value Input Studies training programs Percent of population with Penetration Rates from 0% Dummy Value Input on-board display tools Studies Ecodriving Training: SUTP Review Structured Training Program Nature of Ecodriving Training Types of Ecodriving In-vogue training of Ecodriving OR General Marketing (Choose one) Training Programs programs Training Program Programs Basic Structured Training Depending upon the Program nature of ecodriving Hands-on Training Program training program, Intensive Training Program the program details with Benefits OR provide different levels Literature review of Program Type in Detail under Basic Outreach Program of participation and ecodriving training corresponding Training with Information Brochures involvement resulting programs. Interactive Marketing in varying levels of Program with Multimedia results when it comes to Interactive Marketing implementation rates and Program with Feedback fuel reduction rates. Appendices: TEEMP Model Data Defaults & Sources 45 Telework Model Acceptable Means Data Point Default Value Source Remarks of Collection International Transport Forum Leipzig 2008 Transport and Energy: The Challenge of Climate Change Research Findings 50 country Individual Fuel Use Reduction and Indonesian Study Case Studies or members include Based on Scoring Factors from training Eco Driving: Saving Fuel Surveys some Asian Around Countries the World Clean Fleet Management Toolkit Training 3 March 2009 Michigan Department Percent of population of Environmental reached that implements Based on Scoring Factors Quality (DEQ) - based lessons learned on European examples (Netherlands and Sweden) Percent of that population Michigan DEQ - based that continues to implement 33% on European examples ecodriving in Year 2 (Netherlands and Sweden) On-board display tools: International Transport 50 country Forum Leipzig 2008 Individual Fuel Use Reduction Case Studies or members include 5% Transport and Energy: from on-board display tools Surveys some Asian The Challenge of Climate Countries Change Research Findings Percent of population Michigan DEQ - based reached that implements 50% on European examples lessons learned (Netherlands and Sweden) Appendix 2: Data Required and Defaults Provided for Employer-Based Commuter TDM Strategies Telework Model Default Acceptable Means Data Point Source Remarks Value of Collection “From workplace to anyplace assessing the opportunities to Base rate of telework reduce greenhouse 0.062 participation gas emissions with virtual meetings and telecommuting”, WWF Report “From workplace to anyplace assessing the opportunities to Average days per week reduce greenhouse 1.14 (telework) - Base Year gas emissions with virtual meetings and telecommuting”, WWF Report 46 Appendices: TEEMP Model Data Defaults & Sources Telework Model Default Acceptable Means Data Point Source Remarks Value of Collection Average round trip commute Millennium Cities User can pick details at City/Region 10 length (Km) Database (IATP) from the database “From workplace to anyplace assessing the opportunities to Scenario rate of telework reduce greenhouse 0.084 participation gas emissions with virtual meetings and telecommuting”, WWF Report “From workplace to anyplace assessing the opportunities to Average days per week reduce greenhouse 1.29 (telework) - Scenario Year gas emissions with virtual meetings and telecommuting”, WWF Report “From workplace to anyplace assessing Assumptions on this rebound effect the opportunities to are based on the review of over 30 reduce greenhouse studies undertaken by Steven Rebound effect offset 0.25 gas emissions with Sorrell UKERC (2007). (http://www. virtual meetings and ukerc.ac.uk/support/tiki-index. telecommuting”, php?page=ReboundEffect) WWF Report Employment Department Data/ Total employment 10000 Government Agency/ Census Data SOV Vehicle Type Mode Split Millennium Cities Cars 0.6 Database Supply Indicators Millennium Cities 2-Wheeler 0.2 Database Supply Indicators Only used if the three wheelers are a significant share of traffic - Example India 3-Wheeler 0.2 Dummy Input (Collection and use based on Individual City Studies/vehicle registration data) Knowledge Intensive Sectors’ share of Total Share of employment suitable 0.4 Dummy Input Employment (highest for telework participation rate of all sectors) Appendices: TEEMP Model Data Defaults & Sources 47 Compressed Work Week Model Default Acceptable Means Data Point Source Remarks Value of Collection Existing rate of compressed US Data - US EPA 0.1 No Developing Country Data work week participation Commuter Model US Data - US EPA Split to 4/40 0.08 No Developing Country Data Commuter Model US Data - US EPA Split to 9/80 0.02 No Developing Country Data Commuter Model Average round trip commute Millennium Cities User can pick details at City/Region 10 length (km) Database (IATP) from the database Scenario rate of compressed US Data - US EPA 0.2 No Developing Country Data work week participation Commuter Model US Data - US EPA Split to 4/40 0.16 No Developing Country Data Commuter Model US Data - US EPA Split to 9/80 0.04 No Developing Country Data Commuter Model Commute Strategies Model Default Acceptable Means Data Point Source Remarks Value of Collection Employment/Census Total Employment 10000 Dummy Input Data, Other Govt Sources Employment/Census Share office 0.7 Dummy Input Data, Other Govt Sources Employment/Census Share non-office 0.3 Dummy Input Data, Other Govt Sources Split of total employment by area type Low Density Suburb 0.2 Dummy Input Regional Studies Activity Center 0.3 Dummy Input Regional Studies Regional CBD 0.5 Dummy Input Regional Studies Base Employer Participation 0.1 Dummy Input Regional Studies Rate - Office Base Employer Participation 0 Dummy Input Regional Studies Rate - Non Office Base Financial Incentives Daily Transit/Rideshare 0 Dummy Input Regional Studies Subsidy (in USD) SOV Vehicle Type Mode Split Millennium Cities Database Supply Cars 0.6 Indicators (Dummy Input) Millennium Cities Database Supply 2-Wheeler 0.2 Indicators (Dummy Input) 48 Appendices: TEEMP Model Data Defaults & Sources Commute Strategies Model Default Acceptable Means Data Point Source Remarks Value of Collection Only used if the three wheelers are a significant share of traffic - Example India 3-Wheeler 0.2 Dummy Input (Collection and use based on Individual City Studies/vehicle registration data) Millennium Cities Travel Demand Model, Average roundtrip commute 10 Database Mobility Highway Mobility length (Km) Indicators Statistics, Sutveys 2000 US Census default being used (average occupancy for 2-4 person Average HOV occupancy 2.25 carpools) - varies by vehicle occupancy data for individual countries Employer Support Model Default Acceptable Means Data Point Source Remarks Value of Collection Employment/Census Total Employment 10000 Dummy Input Data, Other Govt Sources Employment/Census Based on USEPA Commuter Model Share office 0.5 Dummy Input Data, Other Govt Methodology Sources Employment/Census Share non-office 0.5 Dummy Input Data, Other Govt Sources Millennium Cities Travel Demand Model, Database -Supply SOV commute mode share 0.75 Surveys/Inventory Indicators (Dummy Statistics Value - US example) Total employee use of employer operated commute 0 Dummy Input shuttles Millennium Cities Travel Demand Model, Average roundtrip commute 10 Database Mobility Highway Mobility length (Km) Indicators Statistics, Sutveys 2000 US Census default being used (average occupancy for 2-4 person Average HOV occupancy 2.25 carpools) - varies by vehicle occupancy data for individual countries Appendices: TEEMP Model Data Defaults & Sources 49 Appendix 3: Data Required and Defaults Provided for PAYD Acceptable Means of Data Point Default Value Source Remarks Collection VKT by mode Cars 85,000,000 Dummy Value Input 2W 22,000,000 Dummy Value Input 3W 13,000,000 Dummy Value Input Highway Statistics Taxi 19,000,000 Dummy Value Input Data, Regional Studies, Bus 3,600,000 Dummy Value Input Surveys with inventory information from Vehicle Jeepney/RTV 7,700,000 Dummy Value Input Registrations Walk — Dummy Value Input Cycle — Dummy Value Input LRT — Dummy Value Input Designing a New Automobile Insurance Pricing System in China- Actuarial and Social Data from Insurance Percent of Drivers who Chinese and Indian 30-50% Considerations Daqing Corporations, Regulation are insured Studies Huang and J. Tim Agencies Query AND India Road Transportation Efficiency Study, World Bank, 2005 Percent of policies that are PAYD (rate of 30% Bordoff and Noel participation by insured drivers) Impacts of Policy Instruments to Reduce Census Data, Price elasticity of VKT/ Congestion and Discuss short/long term Pegged to Income Government Income VMT Emissions from Urban elasticities Data Sources etc. Transportation The Case of São Paulo, Brazil Transport Cost data also Cost per km of driving, Informal survey of India available in Millennium $0.145 Regional Studies without insurance experience Cities Database (Data from India = 0.13) Data from Insurance Informal survey of India Insurance cost per km $0.005 Corporations, Regulation experience Agencies PAYD Driver Low Income Aggregation of Income Participation by Income Medium Income Income stratification quintiles, other fine Category High Income classifications. 50 Appendices: TEEMP Model Data Defaults & Sources Appendix 4: Data Required and Defaults Provided for Employer-Based Commuter TDM Strategies Parking Pricing Model Default Acceptable Means Data Point Source Remarks Value of Collection Average trip length Millennium Cities Database (IATP) Mobility 10.0 (Km) Indicators, average trip distance The demand for road-based passenger mobility in India: 1950-2030 and relevance Average occupancy - 2 for developing and developed countries Indian/Malaysian 1.5/1.3 wheelers and Vehicle Occupancy in Malaysia Sources According To Land Use and Trip Purpose - Easts Conference The demand for road-based passenger mobility in India: 1950-2030 and relevance Average occupancy - for developing and developed countries Indian/Malaysian 3.2/1.6 private cars and Vehicle Occupancy in Malaysia Sources According To Land Use and Trip Purpose - Easts Conference Total daily vehicle trips 25,000 Dummy Input Values OR:   Total regional daily Country Highway   Dummy Input Values VKT/VMT Statistics Data or Travel OR:   Demand Model or Estimation of VMT Total Person Vehicle based on average trip   Trips length and population Low Distribution of Income Dummy Input Values. Percentage travelers Medium for Impacted Travelers in Each Income group High Scoring Factors Small City City Size TCRP 95 Chapter 13 Large City Urban Core Parking Location Near Core TCRP 95 Chapter 13 Suburb Low Level of Transit Service Medium TCRP 95 Chapter 13 High Impacts of Policy Instruments to Reduce Elasticity to Travel Cost Pegged to Congestion and Emissions from Urban by Income Income Transportation The Case of São Paulo, Brazil AND TCRP 95 Chapter 13 % of total public parking 60% Dummy Input Values on-street (with charge) Got Parking Spaces by Parking Purpose from Paul Barter (Singa- % of total public parking 15% Dummy Input Values Parking Studies, Sur- pore), followed up for off-street (with charge) veys. aggregate data. ADB Study for 12 Asian Cit- ies about to come out % of total public parking soon. 25% Dummy Input Values free of cost Appendices: TEEMP Model Data Defaults & Sources 51 Parking Pricing Model Default Acceptable Means Data Point Source Remarks Value of Collection Vehicular Mode split Millennium Cities Database Supply Cars 60% Indicators (Dummy Input Values) Millennium Cities Database Supply 2-Wheeler 20% Indicators (Dummy Input Values) Millennium Cities Database Supply Vary largely by city, can 3-Wheeler 20% provide some mode Indicators (Dummy Input Values) split by City data points Private vehicle mode Millennium Cities Database Supply to the user for familiar- 40% share (all trips) Indicators (Dummy Input Values) ization Millennium Cities Database Supply Split to 2 wheelers 40% Indicators (Dummy Input Values) Millennium Cities Database Supply Split to private cars 60% Indicators (Dummy Input Values) Parking Density Model Default Acceptable Means Data Point Source Remarks Value of Collection City Department of Transportation/ Parking Enforcement/ Revenue Collection # of off-street spaces   Dummy Value Input Office statistics, Other in CBD Govt/Commerce Associations Data, Parking Studies/ Surveys Department of Office/Commercial Commerce, Town   Dummy Value Input Space (sq ft) Planning, Commerce Studies/Surveys Vehicular Mode split Millennium Cities Database Supply Cars 60% Indicators (Dummy Input Values) Millennium Cities Database Supply 2-Wheeler 20% Indicators (Dummy Input Values) Millennium Cities Database Supply 3-Wheeler 20% Indicators (Dummy Input Values) Millennium Cities Database Mobility Travel demand models, Average trip length 10.0 Indicators - Trip Length by Regions of the surveys, Insurance/ (Km) World Govt agency statistics Scoring Factors Small City City Size TCRP 95 Chapter 13 Large City Low Level of Transit Service Medium TCRP 95 Chapter 13 High Impacts of Policy Instruments to Reduce Congestion and Emissions from Urban Elasticity to Travel Cost Transportation The Case of São Paulo, Pegged to by Combination of Brazil AND TCRP 95 Chapter 13 and Income Scoring Factors Parking Density Elasticity TCRP 95 travelers response to parking strategies Chapter 18 52 Appendices: TEEMP Model Data Defaults & Sources Company Cars Model Default Acceptable Means Data Point Source Remarks Value of Collection Canadian and UK     study Average daily commute Millennium Cities Database (IATP) Mobility 10.0 trip length (round trip) Indicators, average commute trip distance Drive Green: Company Car Tax Shift - Average daily business Analysis of Proposed Changes in Tax 15 trip length Treatment for Company Cars in Canada (Company Car Tax Shift) Travel Elasticity to HM Revenue and Customs “Report on the Reduction in Company 0.004 Evaluation of Company Car Tax Reform - Cars Stage 2” - March 2006 Travel Elasticity to HM Revenue and Customs “Report on the Reduction in Company 0.0003 Evaluation of Company Car Tax Reform - Cars with Free Fuel Stage 2” - March 2006 Benefit Dummy Value Input (Data from HM Company Car Studies - Base total company Revenue and Customs “Report on the 1,000,000 Vehicle Registrations as cars Evaluation of Company Car Tax Reform - Company Cars Stage 2” - March 2006) Share of company cars 12% Dummy Value Input Surveys/Studies with free fuel benefit General rule of thumb Drive Green: Company Car Tax Shift - is 50% more than Annual mileage for Analysis of Proposed Changes in Tax 19,500 regular trip - translates company car Treatment for Company Cars in Canada to about the same (Company Car Tax Shift) annually Per Capita Trip Rate Default Values Appendix 5: Default Values (in Number of Trips) for Various TEEMP Models Per Capita Region Source Total Trips Trip Rate The total trips by motorized and non-motorized Latin America 1.71 UITP-MCD transport modes refer to cumulative daily one-way trips between an origin and destination. Based on Africa 1.60 UITP-MCD the economic growth, city planning and transport network, the total number of trip varies among zones, India 1.13 MOUD cities and regions. In case the user does not have any indication of total number trips in the study area,1 per China 2.58 GEF capita trip rates can be multiplied by the population data from the zone/city/region to estimate the Other Asia 2.21 UITP-MCD total number of trips. Per capita trip rate values are available from the International Association of Public Transport’s Mobility in Cities Database (UITP-MCD)2. Trip Mode Share This would allow the user to compute emissions at The trip mode share indicates the distribution of the sketch level. trips in the study area with different modes of trans- port. The trip mode share is one of the indicators for measuring sustainable transport. Trip mode share is an integral parameter for calculating emissions from any urban transport project as it helps in converting person trips to vehicular trips when combined with average occupancy. If trip mode share data is not available, the following default values (expressed in %) are proposed 1 Can refer to zone, city, region. based on literature survey from different countries: 2 See http://www.uitp.org/publications/Mobility-in-Cities- Database.cfm Appendices: TEEMP Model Data Defaults & Sources 53 Default Trip Mode Share (%) Two Description Walk Cycle Car IPT Bus Metro wheeler Average of 30 cities, Ministry of Urban India 31 11 21 16 5 16 - Development GEF and other sources (Average of 16 China 32 26 6 11 5 19 1 cities) Latin America UITP-MCD 25 36 40 Africa World Bank (average of 14 cities) 37 4 12 12 8 27 - Average Trip Length It is the average distance travelled during a trip i.e. one system has implications on the average trip length of way between an origin and destination. This is generally the study area. The data on average trip length allows estimated as the ratio of total passenger- kilometers to the analyst to link the trip characteristics with vehicle the total number of trips and by using origin and desti- emission factors to determine emissions. The following nation (O-D) surveys and often represented in km. The default values can be used for sketch analysis in case size, structure, economic growth, density and transport the average trip length data is not available. Default Values for Average Trip Length (kilometers) Two   Walk Cycle Car IPT Bus Metro Source wheeler Asia 1.1 3.5 6.7 9.9 7.3 10.5 10.0 various - GEF, UITP-MCD, others Africa - - - 12.39   13.1 13.1 UITP-MCD Latin America - - - 13.79   11.8 11.8 UITP-MCD Average Occupancy The average occupancy is calculated person- emissions per person trips. Average occupancy can kilometers per vehicle  –kilometers or simply as the be easy calculated using field occupancy surveys. In number of people traveling divided by the number case no data is available, following default values can of vehicles. Higher the occupancy rates, the lesser the be used: Average Occupancy Two Public Region   Walk Cycle Car IPT wheeler transport Asia UITP-MCD and others 1.00 1.01 1.26 2.38 41.34 1.92 Latin America UITP-MCD and others     2 2 26.47   Africa UITP-MCD         36.3   54 Appendices: TEEMP Model Data Defaults & Sources Emission Factors “Two methodological alternatives are pro- posed for the fuel consumption data (in order Emission factors are generally derived from of preference):Alternative 1: Measurement of dynamometer-based drive cycle tests to simulate fuel consumption data using a representative typical driving conditions and traffic speeds. They are sample for the respective category and fuel generally represented in grams per kilometer traveled type. Factors such as the specific urban driving or one of its derivatives. Fleet-based emission factors conditions (drive-cycle, average speed etc), often used in sector calculations depend on “driving vehicle maintenance and geographical con- behavior” (how do we drive), “fleet characteristics” ditions (altitude, road gradients etc) are thus (what vehicles we drive), “infrastructure” and included. The sample must be large enough geographical conditions (where we drive). It is to to be representative … and Alternative 2: Use be noted that “no two vehicles will have the same of fixed values based on the national or in- emission factor profile, even if they are nominally ternational literature. The literature data can identical models, produced on the same day on the either be based on measurements of similar same production line.”3However, in order to simplify vehicles  in comparable surroundings (e.g. the calculations, the analyst needs to tailor the from comparable cities of other countries) or emission factors to fit “best possible local conditions may include identifying the vehicle age and and the fleet”. These “tailoring” are often done using technology of average vehicles circulating in local studies on various models. the project region and then matching this with In other words, by using an on-road mobile source the most appropriate IPCC values. The most emissions model like the International Vehicle important proxy to identify vehicle technolo- Emissions (IVE) Model with local data on vehicle gies is the average age of vehicles used in the technology distributions, power-based driving factors, area of influence of the project….” vehicle soak distributions, and meteorological factors, In the present TEEMP models, a detailed set of one can tailor the model to suit the local conditions. emission factors based on IVE has not been suggested This would give the best accuracy for computing due to the time and data availability.6 Instead as an emission factors. For example, IVE Model has over alternative option, it is recommended that analyst base emission rates for over 1300 vehicles4 to capture use city-specific studies and national/city surveys to the different fleet characteristics and thus allow better generate the emission factors for the TEEMP models. representation. In order to capture the impact of speed, following In case, the data is not available for the analyst to use default index values have been proposed taking models such as IVE, one can use national averages, insights from COPERT and other studies.7 Many local averages or use fuel consumption data reported studies have suggested that vehicle travelling near via surveys etc. It is to be noted that the approved 50 kmph have best efficiency. Thus 50kmph was kept CDM baseline methodology AM0031 “Baseline as the basis to compute the effect on efficiency and Methodology for Bus Rapid Transit Projects”5 suggests calibrate the emission factor. the following alternatives: 3 DIESEL study- PCD Bangkok, http://www.cleanairnet.org/ caiasia/1412/article-48845.html 6 Corrective factors need to applied to the base emission rates in 4 Different combinations of vehicle types, fuel, weight, air/fuel order to adjust them to local conditions. control, exhaust emission controls and age. 7Copert-3, CORINAIR, green transport, diesel, updated road 5 See http://cdm.unfccc.int/UserManagement/FileStorage/ user cost study of India and trl emission factors for 2009 for CDMWF_AM_IK6BL2878HZ4NHV86V65CBJ2Y1ZBDI department of transportation, UK. Appendices: TEEMP Model Data Defaults & Sources 55 Speed and Emission factors Index (assuming 0 at 50 kmph)8   CO2 PM NOx SPEED 2W 3W Cars LCV Bus HCV Car LGV Bus HGV Car LGV Bus HGV 15 -70 -70 -61 -69 -61 -61 -43 -30 -21 -60 -43 -35 -56 -44 20 -43 -43 -34 -38 -51 -51 -26 -18 -16 -55 -32 -23 -46 -36 25 -26 -26 -20 -22 -39 -39 -18 -10 -12 -45 -23 -14 -37 -28 30 -21 -21 -12 -18 -23 -23 -11 -4 -9 -35 -16 -8 -29 -22 35 -7 -7 -5 -6 -15 -15 -6 -1 -7 -25 -10 -3 -21 -15 40 -4 -4 -3 -3 -9 -9 -3 1 -4 -16 -5 -1 -14 -10 45 -1 -1 0 0 -3 -3 -1 1 -2 -7 -2 0 -7 -4 50 0 0 0 0 0 0 0 0 0 0 0 0 0 0 55 0 0 -1 -1 2 2 0 -2 2 6 1 -2 6 6 60 -2 -2 -3 -4 5 5 -1 -4 3 10 1 -4 13 9 65 -4 -4 -6 -7 5 5 -3 -8 3 12 1 -7 13 9 70 -8 -8 -9 -12 6 6 -6 -11 3 12 -1 -11 13 9 75 -12 -12 -13 -16 0 0 -9 -15 1 12 -3 -15 10 7 80 -18 -18 -18 -23 -4 -4 -13 -19 -1 10 -5 -19 7 4 85 -23 -23 -24 -29 -7 -7 -17 -23 -5 7 -9 -24 4 1 90 -30 -30 -30 -37 -12 -12 -22 -28 -8 4 -12 -28 1 -2 95 -37 -37 -36 -45 -16 -16 -27 -32 -8 -14 -16 -33     100 -37 -37 -36 -45 -16 -16 -32 -36 -8 -16 -20 -38     8 % decrease in fuel efficiency assuming fuel efficiency at 50kmph as 0, - value is indicative The TEEMP model allows users to quantify the air level emission factors for local projects. As a first pollutants PM and NOx using the emission factors. approximation, several studies in Asia were collated The analyst is encouraged to look for national to capture a set of default vales for Asian fleet. Fuel Consumption and Emission Factors for Different Vehicles in Asia Fuel Consumption  CO2 PM (g/ CO2 g/ Vehicle distribution NOx g/Km KMPL L/100KM (kg/L) Km) VKM Two Stroke 1.8 2.416 0.057 0.050   24.170 MC-two P Four Stroke 1.8 2.416 0.015 0.540   24.820 NO data   2.416 0.03 0.34   24.56 Two Stroke 3.5 2.416 0.045 0.200   62.410 MC- P Four Stroke 3.5 2.416 0.015 0.530   73.800 three NO data   2.416 0.03 0.4   69.24 Pre Euro 8 2.416 0.008 0.950 12.5 193.28 Euro I 8 2.416 0.000 0.200 12.5 193.28 P Euro 2 8 2.416 0.000 0.090 12.5 193.28 Euro 3 and Above 8 2.416 0.000 0.080 12.5 193.28 NO data 8 2.416 0.004 0.518 12.5 193.28 PC Pre Euro 7 2.582 0.145 0.450 14.3 180.74 Euro I 7 2.582 0.060 0.490 14.3 180.74 D Euro 2 7 2.582 0.015 0.280 14.3 180.74 Euro 3 and Above 7 2.582 0.050 0.250 14.3 180.74 NO data 7 2.582 0.087 0.359 14.3 180.74 56 Appendices: TEEMP Model Data Defaults & Sources Fuel Consumption  CO2 PM (g/ CO2 g/ Vehicle distribution NOx g/Km KMPL L/100KM (kg/L) Km) VKM Pre Euro 10 2.416 0.008 0.950 10.0 241.6 Euro I 10 2.416 0.000 0.200 10.0 241.6 P Euro 2 10 2.416 0.000 0.090 10.0 241.6 Euro 3 and Above 10 2.416 0.000 0.080 10.0 241.6 NO data 10 2.416 0.004 0.518 10.0 241.6 LCV Pre Euro 8 2.582 0.655 1.710 12.5 206.56 Euro I 8 2.582 0.475 1.600 12.5 206.56 D Euro II 8 2.582 0.100 0.820 12.5 206.56 Euro III and Above 8 2.582 0.050 0.250 12.5 206.56 NO data 8 2.582 0.3675 1.151 12.5 206.56 Pre Euro 28 2.582 1.213 6.240 3.6 722.96 Euro I 28 2.582 0.610 6.660 3.6 722.96 BUS D Euro II 28 2.582 0.150 6.240 3.6 722.96 Euro III and Above 28 2.582 0.100 5.930 3.6 722.96 NO data 28 2.582 0.6715 6.178 3.6 722.96 Pre Euro 30 2.582 1.294 6.450 3.3 774.6 Euro I 30 2.582 0.601 7.620 3.3 774.6 HCV D Euro II 30 2.582 0.366 6.450 3.3 774.6 Euro III and Above 30 2.582 0.100 5.860 3.3 774.6 NO data 30 2.582 0.7768 6.332 3.3 774.6 For the references of the above emission factors please see the endnote. Appendices: TEEMP Model Data Defaults & Sources 57 Construction Emissions Emissions quantification from transport projects should the methodology adopted. In absence of any data, in ideally consider construction emissions. The quantum order to have ballpark estimates, default values have of construction emissions varies depending upon the been proposed for per km construction based on quantity and type of construction materials used and materials used (cement, steel and bitumen). Construction Emission Factors 1 km of tons of Description Source infrastructure CO2 Assuming material quantity - Cement -737.8 tons/km, Asphalt - 403.5 tons/km and Steel - 143.2 tons/km. A multiplier of 1.75 has Considering only been proposed for actual construction works based on Kwangho BRTS the quantity of steel, 1900 Park, et. al. (2003),. Estimates from Mexico BRTS ( Lee at al.) and cement and asphalt. Transmilenio ( monitoring report) have indicated 3475 and 1390 tons . Considering only Assuming material quantity - Cement -15.5 tons/ Bikeways the quantity of steel, 20 km, Asphalt - 40 tons/km and Steel - 1 tons/km for cement and asphalt. constructing 1km of 2.5 m wide bikeway Bangalore metro calculations using quantity of materials 2 lines for 80% used - steel and cement. Research from japan as MRTS elevated and 20% 15600 summarized in TEEMP model indicates a range between underground 7119 to 19487 tons of CO2 Assuming a track requires 570 tons of concrete and 117 tons of steel, 350 tons of CO2 is generated during Considering only material production. Scotland Transport depatment the quantity of steel Railways 875 recommends 500 tons of CO2 per track based on material and concrete for production.A multiplier of 1.75 has been proposed for single track actual construction works based on Kwangho Park, et. al. (2003) for Road works An analysis based on the quantity of construction materials used – cement, steel and bitumen indicates that the approximate emissions of a two lane to four lane Considering only improved highway is approximate 1100 tons/km. When the quantity of steel, Roads 2100 all the quantities are considered including the emissions cement and asphalt generated by machinery, the emissions range from for a four lane road 2100 to 2400 tons/km for high-speed roads (four-lanes) based on traffic, topography and type of improvements suggested. 58 Appendices: TEEMP Model Data Defaults & Sources Mode shift from different modes to a bike share program The development of bike sharing scheme would sharing schemes are proposed. The majority of the attract new riders from different modes. Actual surveys riders using bike sharing schemes come from public can determine the extent of transition from different transport modes. The analysis of 51 schemes in modes. In case the analyst does not have any insights Europe by the “Optimising Bike Sharing in European on the magnitude of transition, the following default Cities” study9 indicates that nearly 25% and 9.3% of values derived from the evaluation of different bike trips have been shifted from walking and cycling. Mode Shifts towards Bike Sharing Schemes Around the World Mode shift from Default Hangzhou Shanghai Beijing Paris Barcelona Lyon London (%) Values Pedestrian 16 26 23 20 26 37 21 22 Bus 51 40 48 65 51 50 34 46 Taxi 4 4 3 5       4 Car 4 4 5 4 8 10 7 6 E-Bike/ Motorcycle 4 5 3 4 Private Bicycle 8 14 8     4 6 10 Others/No Trip 13 7 10     2 23 10 Source: Various studies 9 http://www.obisproject.com Appendices: TEEMP Model Data Defaults & Sources 59 IX. Acknowledgements The authors wish to thank especially the Scientific Vera Lucia Vicentini and Maria Cordiero of the Inter- and Technical Advisory Panel (STAP) of the Global American Development Bank, Sam Zimmerman, Environment Facility (GEF) and the Partnership for Shomik Raj Mehndiratta, Sameer Akbar, and Holly Sustainable Low-Carbon Transport (SLoCaT) who Krambeck of the World Bank, Pai Madhav of WRI/ dedicated many hours to reviewing and commenting EMBARQ, Li Yuwei of the United Nations’s ESCAP, Faris on various versions of this document and the related Khader of the United Nations Development Program, models. We especially want to thank Cornie Huizinga, Francois Cunot and Lew Fulton of the International co-convenor of SloCaT, who synthesized over 500 Energy Agency, and Elisa Dumitescau of UNEP. These comments on the first draft into an effective summary individuals contributed through participation in a one- and John Rogers, of the World Bank, and Axel day workshop of the STAP in Manila in October 2009 Friedrich, former head of transport for the German or through other interactions with the project team. environment agency, who contributed throughout the project, and especially to the July 2010 peer review This work would not have been possible without the of the TEEMP models and pointed us to new sources support of the Clean Air Initiative for Asian Cities, of data. especially Sudhir Gota, who has been a primary developer for the TEEMP models, along with Alvin We are particularly grateful to Lev Neretin, of STAP Mejia, Bert Fabian, and Sophie Punte. Secretariat, who oversaw our work and helped facilitate our interactions with many other stakeholders The authors appreciate the support of the project team and to Osamu Mizuno, from the GEF Secretariat, who at Cambridge Systematics, Inc., who participated in provided constructive ideas and facilitated progress at developing an effective peer review process and in many stages of this work. We especially appreciated devising the TDM and ecodriving models, especially the early and invaluable support for the development Robert Hyman, Christopher Porter, Suseel Indrakanti, of these tools from Narendra Singru, of the Asian Joanne Potter, and David Jackson. Development Bank, who brought both confidence and vision to this work and helped finance related We are grateful for the generous support of the efforts to develop and apply the TEEMP tools to ADB Climate Works Foundation, which enabled ITDP to projects. contribute to this methodology development an initiative far beyond what would have been possible Other STAP members who we want to recognize with UNEP support alone. and thank include Lee Schipper, from the University of California Berkeley and Stanford, Holger Dalkman Many other individuals also contributed towards this of the Transport Research Laboratory, Jaime Leather effort. We are grateful for their support. Of course, the and Shared Saxena of the Asian Development Bank, final result is the responsibility of the authors. 60 Acknowledgements www.unep.org/stap 62 Introduction, Concepts, and Definitions