|
|
|
|
4.0 SAN FRANCISCO NATIONAL UPA EVALUATION PLANThis chapter presents the San Francisco UPA Evaluation Plan. This material is presented in major subsections. The first of these sections, 4.1, Evaluation Analyses, discusses the potential benefits, costs, and impacts of the UPA projects; the Evaluation team's planned approach to measuring those effects; the kinds of data needed to perform this work; and the planned analytic approach. The second Section, 4.2, Preliminary Evaluation Test Plans, summarizes in somewhat more detail data sources and analysis methods. Once this evaluation plan has been finalized, the full detail on data collection and analyses will be presented through a set of separate test plan documents. 4.1 Evaluation AnalysesThe proposed approach to ten evaluation analyses is presented in this section. For the San Francisco UPA, two of the twelve analyses identified in the NEF—transit and safety—were determined to be unnecessary for the national evaluation. Thus, the ten San Francisco analyses address the following areas:
For each of these analyses, key hypotheses and questions to be addressed are presented. The hypotheses describe the results that the UPA projects are expected to produce, including benefits such as throughput improvements, congestion reduction, expanded traveler choices, improved mobility, and related outcomes. In a few cases, unwanted side-effects of the UPA investments are hypothesized. For each hypothesis and question, measures of effectiveness (MOE s) are presented. These are measurable aspects of the San Francisco deployment effects that relate to the evaluation hypotheses and questions. Each analysis discussion includes a table which summarizes the hypotheses/questions being asked, relevant MOE s, and the data required to compute those MOE s. Accompanying text discusses key aspects of the planned analytic approach and related matters. 4.1.1 Congestion AnalysisThe purpose of the congestion analysis is to determine what the combined impact of all the individual UPA projects were on congestion. Specifically, the congestion analysis is designed to answer the following question:
The Congestion Analysis utilizes the basic principles for monitoring freeway performance discussed in NCHRP's Guide to Effective Freeway Performance Measures.7 Specifically, the Congestion Analysis will attempt to quantify the following as a result of deploying the tolling, technology, transit, and telecommuting strategies in the San Francisco area:
Because parking pricing is expected to change every four to six weeks, the performance measures will be computed after each major parking price change in a parking management zone. Performance measures will be computed for each parking management zone and the overall impact will be computed by summing the effects across each parking management zone. Table 4-1 shows the hypotheses, measures of effectiveness and data that will be used to conduct the congestion analysis. In this case, the data that will be used to conduct analysis will be collected primarily through automated data collection equipment deployed specifically for this project and through use of transit vehicles as surrogates when data for other modes aren't available. These data requirements are discussed more fully in the test plans for traffic system and transit system data in Section 4.2.
4.1.2 Pricing AnalysisThe pricing analysis focuses on the affect of new parking management approaches and technology to manage San Francisco's parking supply and demand in ways that reduce the number and duration of vehicle trips, congestion, and double parking. There are seven parking pilot areas and three parking control areas as part of the SFpark project. The pilots include approximately 6,000 on-street metered parking spaces (about 25 percent of the city's total) and 12,250 parking spaces in 15 of 21 SFMTA-managed parking garages, as well as one SFMTA-managed parking lot. Table 4-2 presents the hypotheses/questions, measures of effectiveness, and data for the pricing analysis. SFMTA's demand-based pricing of parking is expected to result in:
Measures of effectiveness to test hypotheses will assess changes in parking supply, vehicles entering/exiting garages and parking on-street by time of day, modal split, parking turnover, parking search time, double parking, length of parking session, transit vehicle travel times (including schedule and headway adherence), and reported changes in travel behavior. The data for developing these MOE s will come from a variety of sources. The SFpark technologies will measure parking duration, turnover, and price. Tax data will be used to assess activity in parking garages. Observational data on search time, disabled placard use,8 double parking and motorcycle parking will be collected. The reported impact of the UPA strategies on travel behavior will be collected in a survey of visitors and shoppers. Muni bus system data will be used for assessing public transit improvements. 4.1.3 Telecommuting/TDM AnalysisThe telecommuting/TDM element of the San Francisco UPA will be of a supportive nature to the primary activities of the UPA, namely the SFpark and 511 enhancements. The telecommuting/ TDM activities will be conducted by the City of San Francisco DOE under the direction of the SFCTA. Three distinct activities are planned: promotion of SFpark at DOE outreach events; promotion of 511 enhancements at outreach events; and co-location of a bike-sharing station at an SFpark facility (e.g., parking structure). The primary objective of these activities is to inform downtown workers about the UPA initiatives and how to get additional information. By so doing, workers, as commuters and downtown travelers, can better use the parking, bike-sharing and information resources available to them. The bike-sharing component is contingent on pending grant activities and the timing of implementation of the city's bike-share system. As shown in Table 4-3, the three hypotheses focus on the impact of these outreach activities on awareness of the UPA activities and their influence on mode shift decisions.
The basic approach to analyzing telecommuting/TDM supportive efforts will be to document increases in the amount and type of information disseminated at DOE outreach events (to include SFpark and 511 enhancements information) and infer the potential impact of this information on mode shift. Using existing metrics collected by DOE as well as tracking changes in rideshare registration rates, the evaluation can infer the influence of these activities on mode shifts, albeit these shifts cannot be directly measured with available data. In terms of assessing the influence of this outreach on awareness of the SFpark initiative, additional information will be collected as part of the visitor/shopper survey. This will serve as a means to corroborate the findings from the event as to the proportion of visitors and parkers who heard about SFpark via the events and who changed their travel behavior as a result. 4.1.4 Technology AnalysisTechnologies, including intelligent transportation systems, underlie many of the UPA strategies being deployed in San Francisco. However, the technology analysis is not intended to be an assessment of the performance of the technologies, per se. Rather, the technology analysis is intended to quantify the degree to which those projects identified in the San Francisco UPA as “technology” projects contributed to the overall reduction in congestion and improved transportation system performance. As such, the technology analysis of the San Francisco UPA National Evaluation is structured to answer the following three evaluation questions:
Parking Technology. Table 4-4 summarizes the hypotheses, measures of effectiveness, and data that will be used in the analysis of how deploying advanced parking management technology improves the SFMTA's ability to better manage parking in the target parking management zones. In this portion of the technology analysis, parking usage, enforcement, and technology performance measures will be used to provide a basic understanding of how the technology was used by SFMTA in the SFpark zones. Interviews with SFMTA staff will then be conducted to assess if and how agency operations and efficiencies were improved as a result of deploying the parking sensor technology. Usage and performance statistics will be collected monthly for each parking management zone while interviews with the agency personnel will occur toward the end to the evaluation period after agency personnel have accumulated significant experience with the technology.
Parking Information Dissemination Technology. A number of information dissemination technologies will be deployed as part of the San Francisco UPA Deployment, including the following:
The purpose of these information dissemination technologies is to provide pricing and availability information related to the on-street and garage parking, and to facilitate way finding to SFpark parking management zones. The DMSs will also provide traffic information during incident conditions. The national evaluation will collect performance and usage statistics that will show how travelers used the different information dissemination technologies available to them in the deployment area. Usage statistics will be aggregated for each parking management zone. These performance measures will be tracked over time and correlated with the current pricing structure to determine how changes in parking pricing impacted travelers' decisions. Travelers surveys will be used to collect information on how travelers used the different information dissemination technologies and how that use impacted their decisions on where and when to park. Table 4-5 shows the hypothesis, measures of effectiveness and data associated with the evaluation of the parking information dissemination technology being deployed as part of the San Francisco UPA Deployment.
Electronic Payment Technology. Several electronic payment technologies will be installed as part of the San Francisco UPA deployment. The local partners are implementing a regional electronic payment card (TransLink®) that local travelers can use to pay for transit fares and on a pilot basis to pay for garage parking. In addition, SFMTA's new parking meters will allow travelers to pay their parking fee through an electronic payment system. It is envisioned that this approach will facilitate parking turnover and improve customer satisfaction in these parking management zones. As a result of the San Francisco UPA deployment, travelers in the downtown San Francisco areas can use electronic payment technologies in the following manner:
Performance measures will be monitored throughout the duration of the post-deployment period to examine how the use of these electronic payment systems varied over time. Travelers will also be surveyed during the post-deployment period to examine how having the ability to use these different forms of electronic payment influenced their mode and parking location decisions. Table 4-6 shows the hypotheses, measures of effectiveness and data that will be used in this portion of the Technology Analysis.
4.1.5 Equity AnalysisThis analysis will examine potential equity impacts of the SFpark pricing project. Experience with pricing projects throughout the country indicates that perceptions of fairness, or equity, may be a factor in the acceptance of proposed pricing projects, especially on how pricing impacts are distributed among minorities or lower income populations. Equity may also be a concern in the spatial distribution of services and infrastructure. Equity issues are important to assess because the impacts – both positive and negative – may contribute to public opinion and the potential success or failure of pricing projects. As presented in Table 4-7, equity will be examined in four ways. First, the direct social effects from the San Francisco UPA projects on various user groups will be examined. These social effects may include parking fees paid, travel-time savings, and adaptation costs. The second hypothesis addresses the spatial distribution of aggregate out-of-pocket and inconvenience costs, and travel time and mobility benefits. Third, possible differential environmental impacts on certain socio-economic groups will be examined. This question addresses possible environmental justice issues. Finally, the reinvestment of revenues from parking pricing and how this reinvestment impacts user groups will be examined. Data addressing these questions will come from a variety of sources as illustrated in Table 4-7.
4.1.6 Environmental AnalysisThe environmental analysis will address air quality and energy impacts. The analysis approach outlined in the National Evaluation Framework, which measures changes in traffic volumes in terms of VMT and then applies emission factors for each pollutant, is consistent with the approaches used in other pricing pilot projects, namely those in London and Stockholm and on FHWA Value Pricing pilot projects on I-15 (San Diego), I-394 (Minneapolis) and SR 91 (Orange County, California). One overall goal of SFpark is to reduce automobile impedance of transit vehicles, thus increasing transit speeds and travel time reliability. Decreased impedance of transit vehicles could improve overall traffic flow and speeds for all vehicles thereby attracting more non-transit vehicles and increasing non-transit VMT. Thus, one unintended impact of SFpark could be increased VMT due to induced travel. This will be very difficult to measure, as it would require measurement of net changes in VMT for all of downtown. For this reason, the evaluation will not attempt to measure air quality impacts from potential induced demand. Table 4-8 summarizes the environmental analysis approach. The first air quality hypothesis focuses on observed or estimated reductions in vehicle miles of travel (VMT) resulting from three possible outcomes of SFpark: 1) mode shift to transit, 2) reduced VMT from less search for parking, and 3) reduced idling from fewer cars double or illegally parked. The second hypothesis focuses on perceptions of the public and stakeholders as to the overall environmental impacts of the projects. The third hypothesis involves the potential for energy saving as estimated using VMT reductions.
The air quality impacts will be analyzed using two primary sources of “observed” data: parking search time data (from search time surveys) and transit ridership data (from transit operators) used to estimate VMT changes. Changes in the amount of idling due to double parking will be documented, but will not be sufficient for emissions analysis. Emission factors for the San Francisco region, provided by the California Air Resources Board and approved by MTC, will be utilized to estimate the emission reductions associated with reduced miles of travel attributable to SFpark. The Air Resources Board will use the EMFAC 2007 model for the San Francisco region to derive the emission factors (emissions reduced per mile). (FHWA has recommended that another alternative be considered for San Francisco and other UPA/CRD sites, and that is the use of EPA's new MOVES [Motor Vehicle Emissions Simulator] model. Use of MOVES would require additional data collection, including driving cycle data. The evaluation team will explore this alternative with FHWA and include the final approach in the detailed test plans.) VMT reductions from mode shift may use different emission factors than reductions from less cruising for parking, given the difference in average speeds for each kind of travel. Average fuel consumption per mile for the San Francisco Bay Area will be used to estimate energy savings. 4.1.7 Goods Movement AnalysisCommercial vehicle operators (CVOs), making deliveries to and parking in SFpark zones, are expected to experience benefits related to improved travel and parking conditions. Frequently there are no available parking spaces for CVOs and often loading and freight zones are occupied, and as a result CVOs double park, which can reduce street capacity by as much as 40 percent and contribute significantly to congestion. Table 4-9 presents the goods movement analysis approach for the San Francisco UPA. CVO travel and parking conditions are expected to improve based on four hypothesized effects of the SFpark program: 1) CVO double parking is expected to decrease; 2) CVO fines are also expected to decrease; 3) parking availability in response to pricing, including loading and freight zones, will increase in the SFpark areas; and 4) travel times will decrease in the SFpark areas for CVOs and other vehicles. These hypotheses will be tested by measuring changes in double parking, double parking fines, parking availability, and travel times before and after the implementation of the program in the SFpark areas relative to control areas. The data to test these measures will be obtained from double parking surveys, records of double parking violation, as well as parking supply and traffic data from sensors.
4.1.8 Business Impacts AnalysisThis element will examine the impact of the San Francisco UPA SFpark project on retail and similar businesses. The SFpark project is expected to improve parking availability and travel time for customers accessing businesses in the SFpark areas. However, it is also possible that the added parking costs of the program may discourage some customers from frequenting SFpark businesses. Table 4-10 highlights the approach for analyzing the impacts of the SFpark project on retail and similar businesses that rely on customers accessing their location. The first hypothesis is that sales will increase in the SFpark areas. Sales tax receipts are available in San Francisco at a geographic scale that is fine enough to assess the impact of SFpark compared to control areas.
The second hypothesis is that overall travel to access retail and similar businesses will increase in the SFpark areas. Parking supply and activity data from sensors will be used to test for significant changes in the number of parking events before and after the implementation of the SFpark program in the treatment and control areas. In addition, data from a survey of visitors and shoppers will be examined for changes in the number, destination, mode choice, and reasons for change in travel to the SFpark areas relative to the control areas. 4.1.9 Non-Technical Success Factors AnalysisThis analysis will collect lessons learned about non-technical success factors from the San Francisco UPA. These non-technical success factors include outreach, political and community support, and the institutional arrangements used to manage and guide implementation of the San Francisco UPA projects. Information on the non-technical success factors is of benefit to the U.S. DOT, state departments of transportation, MPOs, and local communities interested in planning and deploying similar projects. Table 4-11 presents the questions, measures of effectiveness and data sources associated with the analysis of the non-technical success factors. The first hypothesis/question focuses on understanding how a wide range of variables influence the success of the San Francisco UPA project deployments. The variables have been grouped into five major categories: (1) people, (2) process, (3) structures, (4) media, and (5) competencies. The categorization scheme emerged from the Hubert H. Humphrey Institute of Public Affairs' recent study of the Minnesota UPA process leading up to that site's UPA award by U.S. DOT.10 As indicated in Table 4-11 this analysis relies heavily on information provided by the San Francisco UPA partners. Input from the San Francisco UPA partners will be collected using the formal mechanisms shown in Table 4-11, which includes rounds of interviews followed by a group workshop addressing the non-technical success factors. Additionally, information will be gleaned informally through observation and interaction with the San Francisco UPA partners over the course of the demonstration, as well as an examination of formal partnership documents, outreach material, and media coverage. The second question guiding this analysis focuses on public opinion regarding the San Francisco UPA project. Does the public view the UPA projects as effective and appropriate ways to reduce congestion? Public opinion data, if available, and information from the stakeholder interviews will be used.
4.1.10 Cost Benefit AnalysisThe purpose of the cost benefit analysis (CBA) is to quantify and monetize the potential costs and benefits that may be incurred from implementing the San Francisco UPA projects. The net benefit from the UPA projects, which is the difference between the total benefits and the total costs, will indicate the potential returns from the public investment. The cost benefit analysis plays an important role in determining the feasibility of transportation projects because the results from the analysis are easily understood and acknowledged. The cost benefit analysis will be performed using a 10-year time frame (the 10 years following implementation of the San Francisco UPA projects). Within this evaluation time frame, the cost benefit analysis will estimate and compare annual benefits and costs between two scenarios—before and after implementation of the San Francisco UPA projects. Since the UPA projects focus on reducing congestion in the San Francisco downtown area, the expected benefits include travel-time savings, vehicle operating cost savings, increases in travel time reliability, and increase in business activities. On the cost side, the capital costs of the UPA projects will be included, as will operating and maintenance costs, and replacement and reinvestment costs for technology components, such as new facilities for charging parking prices. For communities, the potential benefits include reduction in emissions. The cost benefit analysis for the San Francisco UPA projects depends on several types of data. These data sources include the future traffic forecasts from the regional travel demand model, the data collected from surveys, and the project investment or the expenditures of the local government agencies. To examine the impacts of certain parameters on the net benefits calculated in the cost benefits analysis, a sensitivity analysis will be conducted. Vehicle operating cost savings, for instance, are one of the major benefits that will be experienced by drivers and freight transportation. The calculation of the vehicle operating cost savings depends on fuel price, which has been volatile in recent years. Because forecasting the future movement of fuel price is beyond the scope of the San Francisco UPA evaluation, a sensitivity analysis will be utilized to examine the impacts of fuel price on vehicle operating cost savings and the net benefit generated from the cost benefits analysis. Table 4-12 summarizes the key hypothesis/question that will be addressed by the cost benefit analysis and the main data components that will be calculated in the analysis.
4.2 Preliminary Evaluation Test PlansIndividual test plans will be developed and conducted to collect and analyze the data needed to assess the hypothesis in the 10 evaluation analyses presented in Section 4.1. The 10 test plans for the San Francisco UPA are:
Table 4-13 illustrates the relationship among the 10 test plans and the 10 evaluation analyses. The use of data from the various test plans in assessing the evaluation analyses – both as major input and as supporting input – is highlighted. Table 4-14 presents the more specific data needed for each of the 10 evaluation analyses that will be included in the test plans. Figure 4-1 summarizes the schedule for data collection. The remainder of this section summarizes the key elements of each of the 10 test plans. Preliminary information on the data sources, data availability, data analysis, and the data collection schedule and responsibilities is presented. The more detailed test plans will be developed as the next step in the evaluation process.
 — Major Input  — Supporting Input
Figure 4-1. San Francisco UPA Projects Deployment and Evaluation Data Collection Schedule 4.2.1 Traffic System Data Test PlanData Sources The following sources of traffic system data are recommended for use in the national evaluation. At the time of this writing, the national evaluation team and the local partners were investigating potential methods for collecting the recommended data. The specific methods will determine which of the recommended data sources end up being available, and these will be defined in the detailed test plan at a later date. Vehicle Travel Time Data. SFMTA is investigating use of parking sensors installed as a network of traffic sensors. These sensors would be deployed to provide a saturation of volume counts and potentially speed measurements throughout the different parking management zones. While these sensors are not likely to provide direct measurements of travel times, the national evaluation team is working with SFMTA to determine other potential surrogate measures for travel time, which can be used to assess congestion reduction impacts. The parking sensor vendor has tested the sensor for a variety of non-parking data collection applications, which need to be investigated further for use in the national evaluation. Transit Probe Travel Time Studies. Muni AVL-equipped buses (described in Section 4.2.3) are one potential source of travel time data in the study area. Probe travel time studies will need to be performed in the pre- and post-deployment conditions and potentially after significant pricing changes in each parking management zone. Traffic Volume Counts. As mentioned previously, SFMTA is considering deploying a network of sensors that will allow traffic volume information to be obtained from the network. These sensors are expected to be the same type of sensors used to determine vehicle occupancy in parking spaces. Information from this sensor network can be used to compute changes in vehicle throughput in the corridor. Traffic counts should be performed pre-and post-deployment, and after each significant rate change in each parking management zone. Vehicle Occupancy Counts. Data on the average number of occupants per vehicle class are needed for the congestion analysis. The anticipated types of vehicle classes from which these data are needed include passenger vehicles, HOV vehicles (carpools and vanpools) and transit. One of the stated objectives of SFpark is to promote mode shift (especially to transit), and, therefore, it will be particularly important to gather information on vehicle occupancy rates to determine if UPA improvements achieved this objective. Vehicle occupancy levels are critical in computing passenger throughput at both the facility and corridor levels. It is highly recommended that pre- and post-deployment sampling of vehicle occupancy rates be conducted as part of the evaluation. The approach would be to collect occupancy rates on selected routes. For transit occupancy the automatic passenger counter data can be sampled to calculate average transit occupancy for those routes. For other types of vehicles a protocol for field observation should be developed to count the number of occupants in each vehicle. Data Availability Much of the traffic system data is envisioned to be included in SFMTA's data warehouse, which is being constructed as part of the UPA deployment. While the bulk of the data that will be included in the data warehouse is parking sensor data, the warehouse is envisioned also to retain in-road sensor data and manually collected data.11 Data from the special studies recommended for the national evaluation could also be housed in the data warehouse, including the vehicle occupancy counts, traffic volume counts, and probe vehicle studies. SFMTA currently envisions that data will be added to the data warehouse on either a real-time basis or through nightly batch processes. SFMTA plans on granting the national evaluation team access to the data warehouse. It is anticipated that the evaluation team will access the data warehouse and extract the pertinent traffic system data on a monthly basis. Data Analysis Traffic system data will be used in a number of analyses including the congestion analysis, the technology analysis, the pricing analysis, and others. These data will be used to conduct before-and-after analyses to gauge the extent to which system performance was enhanced as a result of the UPA deployments. Examples of the measures that use traffic system data include the following:
Data Collection Schedule and Responsibility SFMTA will be responsible for securing the traffic system data and making it available to the Battelle team. Traffic sensor data should be archived on a continuous basis and incorporated into the SFMTA data warehouse system on a regular basis. SFMTA will be responsible for any special traffic data collection, which need to be collected on a routine basis – at least every quarter and/or after major parking price changes. However, what specific data and how SFMTA will collect the data will be specified in the detailed test plan. The Battelle team will be responsible for the analysis of the traffic data. Prior to using any of the traffic system data in the analysis, the Battelle team, with the assistance of SFMTA, will inspect the data so as not to include any suspect or obviously invalid data which could bias the results of the analyses. Pre- and post-deployment data from control sites will be compared to data from deployment sites to ensure data trends are similar. Table 4-15 shows the schedule for baseline and post-deployment data collection of the traffic system data.
4.2.2 Parking Data Test PlanData Sources The parking data test plan supports six evaluation analysis areas: pricing, goods movement, technology, environmental, equity, and business impacts. The data will come from four primary sources, including field observational survey data, data generated by parking system technology, parking operations and enforcement data, and tax data for garages. SFMTA's data warehouse will be the repository for the data for this test plan. Field Observational Survey Data. Field observation will be used to collect the types of data related to parking patterns in SFpark and control zones that are not available from the installed parking technology. The surveys will be collected by personnel stationed in the field and collecting the data using preset protocols to ensure data reliability. Pre-deployment and post-deployment surveys will be conducted to collect the following data:
Parking Technology Data. Technology installed as part of the SFpark program will provide automated data on usage and pricing. Parking sensors installed in spaces on-street and at one SFMTA parking lot will detect the presence of a parked vehicle, which will be used to measure parking availability and session durations. Parking payment data will come from parking meters for on-street and parking lot payment and from garage payment systems. The parking payment and transaction data include the following:
Parking Operations and Enforcement Data. The national evaluation team will also review any logs maintained by parking operations and enforcement personnel that can be used to quantify how agency operational procedures change as a direct result of implementing new parking pricing strategies. Examples of the types of data expected to be obtained through these logs include the following:
Parking Tax Data for Garages. Since privately run garages are also part of the parking supply, parking tax data are an available data source for monitoring usage at those locations. A 25 percent tax is assessed (with approximately a three to six month delay in tax data availability) for both SFMTA and non-SFMTA facilities. The locations of non-SFMTA garages (off-street) will be identified in a manual census of garages available to the public conducted by SFMTA. Parking tax data for those locations should result in useful tax data for non-SFMTA facilities. These tax data will be used to approximate changes in supply, activity, and price during the pilot. Such data for non-SFMTA garages will likely be aggregated for small geographic areas. Data Availability The data identified in this test plan are part of SFMTA's planned data collection process. The data will be available to the national evaluation team either from the SFpark data warehouse or other means. Data Analysis The data will be used to develop measures of effectiveness for several analyses. Examples of measures of effectiveness to be derived using data collected through this Parking Data Test Plan include:
Data Collection Schedule and Responsibilities Table 4-16 highlights the proposed data collection schedule for the parking data test plan. Baseline collection for the parking technology data should begin as soon as the sensors are installed in the winter of 2010. All other baseline data collection should be conducted prior to the start of variable pricing, which is now scheduled for April 2010 in the initial SFpark zones. Data collection would continue through the spring of 2011 to provide at least one full year of post-deployment data for the last SFpark zone to go operational.
Battelle team members will work with SFMTA and DOT personnel to establish the data collection and analysis protocol. SFMTA will be responsible for all data collection, with the exception of MTC providing TransLink® data. The Battelle team will analyze the data and report the findings. 4.2.3 Transit System Data Test PlanData Sources Muni APC System Logs. As part of their Transit Effectiveness Project (TEP), SFMTA installed automatic passenger counter (APC) devices on approximately 25 percent of their bus transit fleet. This system uses infrared sensor technologies to measure passenger boarding and alighting activity at each transit stop and to generate passenger loading information. Passenger loading and boarding information will be used by the national evaluation team to determine if changes in parking pricing caused a shift in travel demand to transit. Data by route, bus stop and time of day and week are needed. Muni AVL System Logs. Each APC-equipped bus also contains global positioning system (GPS ) technologies as part of the automatic vehicle location (AVL) system. The GPS technologies can potentially be used to derive transit running times on corridors where the intelligent parking system has been deployed. GPS technology records the position of the transit vehicle when the doors are opened and closed for passenger boarding. Using this information, the evaluation team will compute the average running time between bus stops and the average corridor running time (the actual travel time minus the transit dwell time) as a surrogate for vehicle travel time. Unfortunately, the Muni AVL transit logs may not shed light on the source of delays between transit stops (i.e., whether transit delays were caused by parking maneuvers or signalized intersections, etc.). NextBus (or NextMuni) System Logs. Another potential source of transit travel time information is the NextBus system installed on SFMTA Muni transit vehicles, known as NextMuni. NextBus technology uses GPS positioning, coupled with computer modeling, to track transit vehicles on their route. The software uses the actual position of the vehicle, their intended stops, and typical traffic patterns to estimate (within a minute) the expected arrival time of buses at each stop. Anticipated bus arrival times at each stop location are then made available to transit riders via the NextBus website, electronic signs located at stops, and through mobile communication devices. It is unclear at this time whether or not this is a viable source of transit travel time information. To be a viable source of transit information, the evaluation team would need access to position information collected by vehicles as they traversed through the study corridors. It is unclear at this time what (if any) vehicle position information is retained and by whom. San Francisco Bay Area Rapid Transit (BART) Ridership Data. The BART system automatically collects boardings and alightings of passengers using BART at the turnstiles in each station. Data for BART's stations in San Francisco will be collected by SFMTA and included in the data warehouse. Data Availability Pre-deployment and post-deployment transit data from Muni APCs and BART boardings/ alightings at San Francisco stations will be used in the analysis. Currently, approximately 25 percent of Muni buses are equipped with APCs. To facilitate comparisons, the evaluation team will focus on strategic roadways that include major transit lines that travel through the SFpark zones. Thus, transit data will be collected from strategic roadways within the deployment zones, such as Fillmore, Mission, Chestnut, Union, Market, Van Ness, and North Point. There is a possibility that the APC coverage on the routes of interest to the national evaluation will not be adequate, in which case the NextBus data may be a viable alternative in some cases. Data Analysis Transit travel times and passenger loading information from transit routes that travel through the SFpark pilot area will be used to assess the impacts of parking pricing on both mode shift and transit service. Standard statistical procedures will be used to compare changes in transit ridership and transit running times in the strategic corridors after each parking price change. SFMTA expects that parking rates are likely to change every four to eight weeks throughout the "after" period in each of the parking management zones. This pricing change will need to be correlated with transit vehicle performance to gauge the impacts of parking pricing strategies. Data Collection Schedule and Responsibilities SFMTA will be responsible for making the transit data available to the evaluation team; they will be available through the data warehouse. The Battelle team is responsible for analyzing the data and reporting the findings. Table 4-17 presents the transit system data collection schedule.
4.2.4 Telecommuting/TDM Data Test PlanThe telecommuting/TDM data test plan will be used primarily for the telecommuting/TDM analysis. It also supports the congestion, environmental, equity, business impacts, and cost benefit analyses. SFCTA and the DOE are still working out the final plans for including SFpark and 511 enhancements in their outreach events. As a result, the final telecommuting/TDM test plan will be developed in cooperation with SFCTA and DOE, based on the final outreach plan. Activities related to bike-sharing will be added if this element is added to the telecommuting/TDM element of the UPA. (This activity is pending separate funding approval and discussion between MTA and SFCTA.) Therefore the test plan description provided here focuses on evaluating the outreach element of the telecommuting/TDM activities of the UPA for San Francisco. Data Sources The primary source of data for the telecommuting/TDM analysis will be records of outreach events held by the DOE and surveys of travelers. The evaluation will also track trends in ridesharing registration statistics to infer whether the additional information provided at outreach events has an impact on rideshare registration levels. Finally, questions in the visitor/shopper survey can document the recall of outreach events as a source of information on SFpark. Records of Outreach Events. DOE plans to conduct outreach events in downtown San Francisco for commuters in order to provide information on alternatives to driving alone to work. These events are held in public places with significant foot traffic, at employment sites, or in conjunction with other events (e.g., Earth Day, Carfree Day). These events are planned on a 3-month rolling basis, and DOE or SFCTA will provide to the national evaluation team a listing of planned events over the course of the evaluation period. The events will provide a venue for distribution of information on SFpark and 511, and DOE will track the number of brochures distributed. Rideshare Registration Data. Ridesharing registration trends, as maintained by MTC as part of regional ridematching activities, will be tracked to assess potential changes in registration levels that might be influenced by the additional information provided at outreach events. Surveys of Travelers. The visitor/shopper survey (described in Section 4.2.6) that SFMTA will conduct will be used as a corroborative source of data. The visitor/shopper survey should collect information about respondents' awareness of the parking changes and the source of this information prior to their visit. One category for this source of information should be a DOE/employer outreach event. Data Availability It is anticipated that the data needed to assess the impacts of the telecommuting/TDM outreach information about UPA activities will come from data collection activities already planned by SFMTA and from regular tracking data available from DOE and/or SFCTA. DOE would also need to begin recording the number of SFpark and 511 brochures distributed at each event. Rideshare registration trend data will be available from MTC and is also a regular reporting activity that DOE tracks as part of their standard set of metrics of their outreach activities. The visitor/shopper survey will be undertaken by SFMTA. To support the telecommuting/TDM analysis, the survey should include a question about how the respondent learned of the variable parking prices, with one of the response categories being the DOE outreach events. Data Analysis The Battelle team will perform the analysis of telecommuting/TDM outreach effectiveness in informing commuters about parking changes and parking information. Data provided by DOE outreach events and survey data from SFMTA will be the basis of the analysis. The analysis will be largely qualitative or descriptive in terms of the amount of information distributed and role of outreach in providing information to commuters on UPA activities. To the extent possible, the analysis will attempt to infer the influence of this information on parking or mode behavior, based on rideshare registration data and SFMTA surveys. Data Collection Schedule and Responsibilities The DOE will be responsible for providing records of the events they conduct and the metrics they track including rideshare registration statistics. SFMTA will provide the visitor/shopper survey data. All the data collection will be post-deployment following the start of operation of SFpark pricing in April 2010 through the spring of 2011. The Battelle team is responsible for coordinating with SFMTA, DOE, and SFCTA, providing technical assistance to local partners to incorporate UPA-specific survey questions as appropriate, analyzing the data, and reporting the findings. 4.2.5 Traveler Information System Data Test PlanData Sources SFpark Website Use and Operations Logs. As part of the SFpark deployment, the SFMTA is planning on implementing a self-maintained website that provides up-to-the-minute parking availability and pricing information from all the different parking management zones. Travelers can access the website and/receive text messages about parking availability from their mobile devices. The national evaluation team anticipates that logs will be generated that can be used to monitor the number and duration of requests for parking information from SFMTA's website and from their text messaging system. 511 Logs and Usage Reports. SFMTA will feed real-time parking availability and pricing information to the Bay Area 511 system operated by the MTC. The national evaluation team will use operator logs and usage reports produced by MTC to monitor changes in parking availability and pricing information and usage of that information by travelers. Data Availability The SFpark data warehouse is expected to house data on SFpark website usage and operations logs which the national evaluation team would access on a monthly basis. Other data, such as the 511.org website and 511 phone service and operations logs, are expected to be provided to the evaluation team at the end of each month. Data Analysis The traveler information system data will be used in the technology analysis to track post-deployment patterns of operations of the systems by the SFMTA and MTC and the access to the parking information by travelers. Examples of the measures that use the traveler information system data include the following:
Schedule and Responsibility The data that will be used in this test plan will come from manual or automated logs of the different systems (i.e., the website usage logs, operator logs, etc.) as well as SFMTA's data warehouse. SFMTA plans to include in the data warehouse information about requests for parking information. MTC will be responsible for providing information related to requests for parking information associated with SFpark. It is expected that the national evaluation team would collect information from these agencies monthly. Table 4-18 presents the data collection schedule.
4.2.6 Surveys and Interviews Test PlanData Sources and Availability Surveys, interviews and workshops are critical for obtaining information needed to assess the influence of the San Francisco UPA projects on changes in travel behavior and perceptions. Possible behavior changes include shifting travel modes, frequency of trips and parking in San Francisco, and changing time-of-travel. While traffic counts and bus ridership data are important, the only way to ascertain if people have changed their travel mode or made other changes as a result of the UPA projects (as opposed to other factors) is to ask them. Surveys, interviews and workshops also provide information about individuals' perceptions of different strategies and projects, the ease or difficulty of using technologies and services, and concerns about equity. This test plan outlines the survey, interview and workshop-related UPA evaluation data needs. Planning and conducting special surveys can be costly and so the national evaluation team has, aided by the San Francisco partners, inventoried existing data sets and planned surveys for possible use in the UPA evaluation. The recommended approach includes identification of existing and planned local partner data and data collection that may be used in the UPA evaluation. It also identifies the additional UPA-specific surveys, interviews and workshops needed to fully evaluate the San Francisco UPA deployment. Table 4-19 presents the information needed from various populations and summarizes the recommended approach. A total of 5 population groups and the associated information needed for the evaluation are identified.
The sections that follow briefly discuss each survey, interview, or workshop to be used, first presenting the existing or planned local partner data to be utilized and then identifying the UPA-specific method that is recommended. Details on questions and survey protocols (recruitment, sampling method, etc.) will be presented in the full test plan documents and will include consultation with the local partners. Use of San Francisco Partners' Relevant Existing and Planned Surveys Mobility, Access and Pricing Study (MAPS ) Retail Survey. As part of its assessment of pricing to manage congestion, SFCTA was aware of potential concern among the business community that pricing may negatively impact downtown merchants. To collect data on existing travel and spending patterns, SFCTA conducted a random physical intercept study in the winter of 2007 and spring of 2008 with 1,390 visitors to shopping areas. The objective was to understand the behavior of patrons to two shopping areas in San Francisco and two areas outside the city. The survey averaged three minutes and asked shoppers about trip purpose, frequency and mode of travel to the shopping area, what is liked most and least about the area, spending levels at the area, and socio-economic data. This survey may provide some useful comparative data for the baseline analysis to the extent the surveyed areas overlap SFpark zones and control zones. Mobility, Access and Pricing Study (MAPS ) Public Opinion Survey. Also part of the assessment of congestion pricing in San Francisco, this survey collected information on commute habits, parking fees and subsidies, and opinions on congestion. A random telephone survey of 600 households in six Bay Area counties was conducted in the fall of 2007. This survey may provide useful baseline data on public opinion about congestion pricing. Doyle Drive Tolling Survey. This survey was conducted in 2008 by SFCTA in conjunction with plans for tolling Doyle Drive that were part of the original UPA proposal. Some of the opinion questions may be useful as baseline data for opinions about pricing as a travel demand management tool. 511 Transit 2009 Focus Groups. MTC conducted four focus groups in July 2009 to gather post-launch feedback from users of the 511 Transit website. The main purpose was to gather feedback about the website's specific features, but the focus groups afforded an opportunity to gain some initial feedback on MTC's concept for a multimodal trip planner. Parking information and cost of parking were some of the most frequently requested types of information by focus group participants. 511 Annual Phone and Website Satisfaction Studies. MTC has conducted annual 511 user satisfaction surveys on both their phone system and website since 2004. Both the phone and website samples are self-selecting based on a phone prompt or banner ad soliciting survey participation. The surveys in 2008 yielded 1,000 website users and 1,500 callers. In the past surveys no questions about parking were asked and a search of comments from the web survey respondents did not reveal any mention of parking. However, MTC is considering addition of a question related to parking in the 2009 survey now being planned. If so, the findings may provide useful comparative data. Needed Surveys and Interviews Visitor/Shopper Survey (Baseline & Post Deployment). These surveys will provide details on travel and parking behavior in response to parking pricing and other travel options, the awareness and use of real-time parking information disseminated through various means, the use of TransLink® integrated payment system, as well as perception of the impact and value of the UPA project for addressing congestion issues. Surveys will reveal the perceived personal advantages and disadvantages of the UPA strategies to the traveler, such as improved parking availability, travel time reliability, and the perceptions of the broader societal implications (e.g., equity, safety, and environment). Collecting information on travel behavior, including changes in travel patterns (e.g., frequency of travel to San Francisco or mode) and the reason for the changes is essential for differentiating the impact of the UPA from the influence of various exogenous factors and understanding traveler responses to specific UPA strategies. The surveys will focus on persons in the parking management and control zones of the SFpark pilot areas. Several options for conducting a survey of visitors and shoppers were considered by the national evaluation team and SFMTA, including cross-sectional and panel studies. Other methodological options pertain to the method of recruiting participants (e.g., on-street intercept and telephone sampling) and conducting the survey (e.g., in-person interview, telephone interview, and on-line). For cost-considerations, the recommended approach is a cross-sectional on street intercept survey and follow-up telephone survey conducted before and after the variable pricing is implemented in selected SFpark zones. A brief 3-minute on-street intercept can be used to solicit respondents and collect contact information for the more in-depth telephone interview of approximately 15 minutes. TransLink® Cardholders Survey (Post-Deployment). This survey would use e-mail addresses in MTC's database of TransLink® cardholders to solicit them to participate in an on-line survey. MTC has previously had success in conducting surveys of cardholders about topics related to TransLink® and have used commercially available on-line survey tools such as SurveyMonkey.com. Questions would ask about the cardholder's awareness of and usage of the card for parking in the SFMTA garages and their perception of the benefits of the parking payment feature. Directly surveying cardholders would result in a higher likelihood of reaching people who have used the card for parking than would a survey aimed at a general cross-section such as the visitor/shopper survey. Interviews with SFpark Operators (Post-Deployment). These interviews will collect information from staff of SFMTA who is responsible for parking operations and management of the SFpark system. The interviews will gather information about experience in using the technology associated with SFpark in terms of ease of using the system and advantages relative to the previously used technology. The interviews will also focus on the impact of SFpark on efficiency of operations of the on-street and off-street parking such as revenue collection, enforcement, and other types of improvements. Partnership Agency Representatives and Other Key Stakeholders (Baseline & Post- Deployment). Members of the national evaluation team will conduct one-on-one interviews with representatives of organizations that play an important role in planning, deploying and/or operating the UPA projects. This will include those organizations instrumental in the institutional, technical or public outreach aspects of the UPA projects. As the full test plan is developed the national evaluation team will work with the local partners to further specify interviewees. Two rounds of interviews will be conducted, one each near the end of the baseline and post-deployment periods. Each round of interviews will include a group workshop to discuss lessons learned. Data Analysis A variety of data analysis techniques will be used to analyze the wide range of survey and interview data, with techniques varying according to the type of data and the intended use of the resulting measures of effectiveness in the various evaluation analyses. In the case of interviews, key points from each interview will be compiled, summarized and discussed, and areas of agreement, disagreement and recurring themes cutting across multiple interviews will also be identified. Survey analysis will begin with checking the data for anomalies, outliers, or other data peculiarities and to prepare the data, including applying any necessary weighting to adjust for selection bias, unequal response rates in various strata, etc. Descriptive statistics will be prepared to characterize outcomes of interest such as the percentage of respondents reporting that they experienced less parking search time or found parking information on 511 useful in making travel decisions, as well as potential predictor variables such as trip purpose. Data Collection Schedule and Responsibilities The San Francisco local partners will be responsible for collecting the data in this test plan with the exception of interviews with the partnership agencies and other key stakeholders, which will be conducted by the national evaluation team. SFMTA will procure survey contractor services for the visitor/shopper survey. The national evaluation team will, through the full Surveys and Interviews Test Plan document to be developed, provide the local partners specific guidance and recommendations on the key aspects of the survey methodology, including specific information to be collected. Baseline surveys should be conducted shortly before the SFpark pilot areas go into operation in April 2010. Since the SFpark zones will be phased in over several weeks or months, it will be advantageous to focus the visitor/shopper survey on parking zones that go into operation later to maximize the time available for survey design and pretesting by SFMTA's survey consultant. Post-deployment surveying should occur after all the parking-related UPA projects are operational. The TransLink® integrated payment system is expected to be operational in December 2010 and the initial SFpark zones will have been in operation for one year in March 2011. Thus, the post-deployment visitor/shopper survey and the TransLink® cardholder survey should occur in the spring of 2011. Other post-deployment interviews will also occur during this period. 4.2.7 Environmental Data Test PlanData Sources The environmental data test plan will be used primarily in the environmental analysis, but it will also support the equity and cost benefit analyses. The environmental analysis for the national UPA evaluation will be based on assessing the impacts of the various projects on VMT rather than direct monitoring of air pollutants. Data will come from other test plans (including the traffic, parking, transit, telecommuting/TDM, and survey and interview data test plans), and include the following data:
Parking search time data collection is described in Section 4.2.2 the Parking Data Test Plan. This will involve a before and after survey in the pilot and control zones. Changes in idling will be derived from the disabled placard/double parking survey and will describe the incidence of idling, but not the duration, due to limitations with the data to be collected. SFMTA will be responsible for this data collection activity and providing VMT reduction estimates. VMT reduction from mode shift to transit will come from both the transit data described in the transit data test plan (to get changes in ridership) and from the visitor/shopper surveys to assess prior mode (hence mode shift). These surveys will be implemented toward the end of the deployment period or during post-deployment. The approach will be to apply the appropriate emission rates to VMT estimates for each pilot area based on observed changes in search time and idling. Emission factors will be obtained from the California Air Resources Board (ARB) and/or the MTC, who regularly generate these factors for other analyses, including conformity. Additionally, emission factors and fuel efficiency factors will be used to estimate air quality and energy impacts from VMT reductions. Data Availability The data for the air quality analysis will be obtained from the parking, transit, and survey test plans. The key data from the parking test plan will be changes in parking search time, vehicle speeds and incidence of vehicles idling. Changes in search time will be converted to miles of travel reduced. No additional data collection is anticipated for the environmental analysis. Emission factors will be obtained from ARB and/or MTC. Data Analysis The air quality analysis will assess emission impacts based on before and after estimates of VMT changes due to changes in parking search behavior. Reduction in specific pollutants will be estimated, including volatile organic compounds (VOC), nitrogen oxide (NOx), particulate matter (PM) and carbon dioxide (CO2). Additionally, mode shift data will also be assessed in order to estimate concomitant VMT reduction, which will be converted to emission reduction using emission factors. Fuel consumption factors applied to VMT will be used to assess energy impacts. The three basic sources of potential air quality impacts are:
Emission reductions from reduced search time will be possible using data from parking search time surveys. Emission analysis tied to idling and transit ridership increases require data that may not be available in a form convertible to VMT. If this is the case, the potential contribution of reduced idling and increased transit use will be discussed, but not necessarily converted to emissions reductions. Options for energy modeling will be examined in the development of the detailed test plan, but at a minimum will involve the application of average fuel consumption factors to VMT reductions. Data Collection Schedule and Responsibilities SFMTA is responsible for collecting the data identified in this test plan. SFMTA will begin collecting baseline data in the winter of 2010, with post-deployment data collection beginning in the spring of 2010 and running through winter of 2011. The Battelle team will be responsible for analyzing the data to produce aggregated VMT reduction estimates to assess the air quality impacts by applying the appropriate emission factors. 4.2.8 Content Analysis Test PlanData Sources The content analysis test plan focuses on collecting and analyzing information on San Francisco UPA outreach activities, media coverage, and reactions from the public, policy makers, and other groups. The information collected and analyzed in the content analysis test plan will be used primarily in the non-technical success factors analysis. Two primary data sources will be used in this test plan. The first data source is on-line search engines Google Alerts and Vocus. Information from the San Francisco UPA agency partners represents the second data source. Google Alerts and Vocus. Google Alerts is a free on-line search engine that tracks news articles, web-based information, blogs, video, and other media information based on search terms. Members of the Battelle team have signed up with Google Alerts and have entered key terms based on each of the UPA sites. Examples of key terms for the San Francisco UPA projects include SFpark, parking pricing, and San Francisco UPA. Vocus is a private company providing a subscription service that monitors media coverage based on key search words. Through team member TTI's subscription to Vocus, the Battelle team will collect information about media coverage of the SF UPA projects. San Francisco UPA Partnership Agency Information. Press releases and outreach, public education, and marketing materials issued by the San Francisco UPA agencies represent the second source of information for the content analysis test plan. Staff from the San Francisco partnering agencies will include Battelle team members on the distribution list for these efforts. Members of the Battelle team will monitor these activities and will document press releases and other outreach activities. To the extent the information is available, members of the Battelle team will also obtain information from the agencies on letters, e-mails, and telephone calls received about the UPA projects. Data Availability The availability of most agency data is assumed to be good in that the local partners will be maintaining archives of media coverage. It is anticipated that post-deployment information will be available, with the possible exception of extensive tracking of letters, e-mails, and telephone calls received by the partnership agencies. Google Alert data are already being collected by the Battelle team and Vocus data will be available soon. Both on-line sources are assumed to be available over the course of the evaluation. Data Analysis The information obtained in this test plan will be used in the lessons learned analysis and will support other analyses. The following questions provide examples of how the qualitative information obtained in the test plan will be applied in the evaluation.
Members of the Battelle team will document the results of the Google Alerts and Vocus on-line search tools and information obtained from the partnership agencies. Table 4-20 illustrates how the information will be tracked, categorized, and analyzed.
Data Collection and Responsibilities The San Francisco partners are responsible for providing their data for the content analysis to the Battelle team. Members of the Battelle team have already begun data collection activities related to this test plan. Battelle team researchers have registered with Google Alert and Vocus. Members of the Battelle team will continue to monitor these on-line resources over the course of the pre- and post-deployment periods. The Battelle team will be responsible for analyzing the content analysis data and reporting the findings. As Table 4-21 highlights, partial pre-deployment information is available from the two sources used in this test plan.
4.2.9 Cost Benefit Analysis Test PlanData Sources The cost benefit analysis test plan will be used in the cost benefit analysis. The cost benefit analysis test plan will use several sources of data. One source is the detailed costs associated with the UPA projects to be provided by the partner agencies. A second source of data is forecasts of travel from the region's transportation model. A third source is tax data for businesses in San Francisco. The fourth data source is data collected through other test plans. Cost Data from Participating Agencies. Cost data will be obtained from SFMTA, MTC, and SFCTA related to their UPA projects. Data include the capital costs associated with various projects, the operating and maintenance costs, and the replacement and re-investment costs. The following examples are some but not all of the cost categories needed for this test plan.
Travel Demand Forecasting Model Data. The SFCTA's travel demand forecasting model (SF-CHAMP) will be used to generate 10-year forecasts of travel patterns in San Francisco resulting from the UPA strategies. The model is equipped to handle parking pricing changes by trip purpose, but at the present time, its ability to handle variable parking pricing (for example, by time of day or by parking duration) at the level needed to represent SFpark is limited. Depending on availability of data, the San Francisco Model may be improved to capture this capability by the time the data are needed for post deployment evaluation. The San Francisco model was developed in 2002 using an activity-based framework to improve the simulation of travelers' responses to transportation policies relative to the traditional four-step approach.12 In an activity-based framework, individuals' activities drive their travel. Travel is modeled as a tour, which is a closed chain of trips that begins and ends at one location (e.g., home or work). The model includes six types of tours: Work Tours, Grade School Tours, High School Tours, College Tours, Other Tours, and Work-based Tours. There were several surveys used for model development, including the 1990 and 2000 Bay Area Travel Survey, census data, and other sources. A local stated preference survey was used to develop the model's pricing sensitivity. Microsimulation is used to forecast the travel of each person in the population. The synthetic population was generated from the U.S. Census Public Use Microdata Sample, San Francisco County population and employment data, and other demographic data developed by the Metropolitan Transportation Commission. The current population is about 800,000. The model includes 2,475 travel analysis zones (TAZs), 981 of which are in San Francisco and which are essentially blocks that conform well to the SFpark zones. The network used in the model is highly detailed: it includes every street, road, transit line, and transit stop. Currently, the supply and availability of parking is represented in the vehicle availability model through a parking availability index and in the mode choice models through an average parking cost for work trips (eight hours) and average parking cost for other trips (one hour). The model is currently being used to forecast up to 2035. The model can generate VMT by facility type and speed by neighborhood for emissions analysis. Commercial vehicle travel is represented very crudely in the model. The structure of the San Francisco model facilitates its representation of pricing policies because it can be adjusted to track daily payments and wide variations in individual response to pricing. In 2008, several enhancements were made to the model to further improve its representation of cordon and congestion pricing: 1) model expansion to nine Bay Area counties; 2) representation of toll payment in sub-models; 3) use of continuous value-of-time distributions using the results of a new stated preference survey; and 4) other structural improvements to better represent travel time and cost in the model and traveler response to change. The San Francisco model does have a parking cost sub-model, but its representation of parking supply and cost is too general to precisely replicate the UPA parking pricing scenarios at the present time. For example, the model uses average parking costs and a generalized measure of parking availability, and it does not differentiate on-street parking from off-street parking. SFCTA would like to update the model's parking supply data with the data that SFMTA is collecting for SFpark. In sum, although the model is very advanced relative to the state-of-the-practice in the U.S., its ability to precisely represent the UPA parking pricing policy is limited. SFCTA plans to make further improvements to the model's capabilities for handling variable parking pricing, which may be available in time for the post-deployment analysis of SFpark. Sales Tax Receipts. Quarterly sales tax revenue data will be provided by the Controller's Office for the City and County of San Francisco. This data is available by economic category, economic segments, and business code for the SFpark areas and the control areas. This data will also be accessible from the data warehouse. Table 4-22 lists the economic categories and segments available.
Other San Francisco UPA Test Plans. Another important source of data for the cost benefit analysis is other test plans. The data from each test plan will be used to compare the scenarios before and after the UPA projects are implemented. The following are examples of the data from other test plans that will be used in the cost benefit analysis:
Data Availability It is anticipated that agency cost data will be available from SFMTA, MTC, and SFCTA records. SFCTA has agreed to run the SF-CHAMP model to produce forecasts for the national evaluation, but the planned improvements to the parking pricing submodel will need to occur to produce the quality of the forecasts needed for the evaluation. Sales tax data are available as public records summarized by geographic areas of the city from Comptroller's Office. Other needed data for the cost benefit analysis will be obtained from other test plans. Data Analysis As noted previously, SFCTA's regional travel forecast model will be used to estimate the benefits related to congestion reduction resulting from the UPA projects. The cost benefit analysis will be performed using a 10-year time frame. This time frame includes the first year after implementation of the San Francisco UPA projects and again nine years into the future, for a total 10-year period after implementation of the projects. Within this evaluation time frame, the cost benefit analysis will estimate and compare annual benefits and costs between two scenarios—before implementation of the San Francisco UPA projects and after implementation of the San Francisco UPA projects. Data Collection Schedule and Responsibilities The cost benefit analysis will be initiated prior to deployment of the San Francisco UPA projects. The analysis will be completed after all the UPA projects are in operation. The local partners will be responsible for providing the cost information relevant to the UPA projects that each agency is deploying. SFCTA will run the regional travel model to provide forecasts based on collaboration with the Battelle team about the inputs and the outputs needed. Members of the Battelle team will perform the cost benefit analysis based on data from the various sources. It is anticipated that the Battelle team will work with the SFCTA staff running the regional travel model to perform test runs during the post-deployment operation phase of SFpark and the other UPA projects. Test runs will help identify any modeling issues that need to be resolved before the final modeling is performed. Once the modeling data and all the other data from this test plan are available following the post-deployment data collection, the cost benefit analysis will be performed. 4.2.10 Exogenous Factors Test PlanData Sources The exogenous factors test plan will be used to monitor elements unrelated to the SFpark UPA project that may influence travel in San Francisco and changes in travel modes, routes, and parking location. The data obtained in the exogenous factors test plan supports the SFpark project primarily. As outlined in this section, elements included in the test plan are changes in fuel prices; employment factors (e.g., unemployment rates); transit strike tracking (Bay Area Rapid Transit (BART) District); construction; and changes in overall parking supply, availability, and pricing. Energy Information Administration Gasoline Prices. The Energy Information Administration of the U.S. DOE monitors gasoline prices by the nation, selected states, and regions, including San Francisco. Historical data on weekly retail gasoline prices for various grades have been available on-line since 2000. Data will be monitored over the course of the San Francisco UPA evaluation at http://tonto.eia.doe.gov/dnav/pet/pet_pri_gnd_dcus_nus_w.htm and tracked in the SFMTA data warehouse. Employment Factors. Unemployment rates in the San Francisco Bay Area will be tracked, along with possible furlough activity in SFpark pilot and control areas and maintained in the SFMTA data warehouse. General System Impacts. General system impacts will be monitored throughout the pilot project, such as strikes or changes to the BART District and other transit providers and major incidents that significantly disrupt traffic. Such activity will be tracked throughout the pilot to assess any impacts on travel behavior that could result from a transit strike(s) or other system changes. These data will be recorded in the SFMTA data warehouse. Large-Scale Construction Events. In downtown areas, traffic patterns can be significantly impacted by both on-roadway and off-roadway construction activities. Off-roadway construction sites frequently remove on-street parking to accommodate construction equipment and/or pedestrian movements. Construction activities can sometimes infringe and impede traffic flow in travel lanes. For evaluation purposes, it will be important for the national evaluation team to know when and where major construction is occurring in the various parking management zones so that unusual or atypical changes in traffic patterns can be observed. SFMTA will track large-scale construction events for control and pilot areas. These events will be noted by duration and types and be available through the SFMTA data warehouse. Changes in Parking Supply. Changes in parking supply will be derived from a manual census of publically available garage data and parking tax data for those locations. A 25 percent tax is assessed (with approximately a three to six month delay in reporting) for both SFMTA and non-SFMTA facilities. It is hoped that the manual census, currently being conducted by SFMTA, will result in better parking tax data for non-SFMTA facilities. These parking tax data will be used to approximate changes in supply/activity and price during the pilot. Such data for non-SFMTA garages will likely be aggregated for smaller geographic areas. These data will be available through the SFMTA data warehouse. Control Areas. There are three SFpark control areas that will be used to compare changes due to parking pricing in the seven pilot areas. The three control areas are Inner Richmond, West Portal, and Union Street. Data Availability As noted previously, historical, pre-deployment, and post-deployment data are available for gasoline prices and unemployment rates. Historical and pre-deployment data on other exogenous factors are limited, but post-deployment data will be available on all of the elements in the test plan. The SFpark data warehouse will serve as a repository for the data elements in this test plan. Data Analysis The exogenous factors included in this test plan will be used as comparison checks in the analysis of SFpark primarily. Information on the exogenous factors will assist in identifying elements that may influence and explain changes in parking availability, travel patterns, traffic conditions, and modal changes that are not due to the UPA strategies by themselves. Data Collection Schedule and Responsibilities Table 4-23 presents the anticipated data collection schedule for the exogenous factors test plan. As noted, historical data and pre-deployment data are available for some factors, while post-deployment data are available for all factors. The responsibility for collecting data will reside with the local partners.
7
Margiotta, Richard A. et al., "Guide to Effective Freeway Performance Measurement: Final Report and
Guidebook," NCHRP Project 3-68, August 2006.
8
Margiotta, Richard A. et al., "Guide to Effective Freeway Performance Measurement: Final Report and
Guidebook," NCHRP Project 3-68, August 2006.
9 The deployment of the DMS has been delayed to December 2011, placing them a year behind the other UPA projects. Rather than delay evaluation of the rest of the projects, the decision was made not to include them in the
national evaluation.
10 John M. Bryson, Barbara C. Crosby, Melissa M. Stone, J Clare Mortensen (2008). "Collaboration in Fighting Traffic Congestion: A Study of Minnesota's Urban Partnership Agreement." Report no. CTS 08-25, University of Minnesota ITS Institute. December. http://www.its.umn.edu/Publications/ResearchReports/reportdetail.html?id=1714
11
SFpark Data Warehousing Plan. SFMTA. January 6, 2009.
12
SFCTA. San Francisco Travel Demand Forecasting Model Development, Executive Summary. October 1, 2002;
and Freedman, Joel and Charlton, Billy. Activity-Based Travel Models for Road Pricing. April 14, 2008.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
 |
US DOT Home | FHWA Home | Operations Home | Privacy Policy | |
