Date : 201 6 / 11 / 2 2 – Issue A BART P LATFORM G ATES R ETROFIT A Feasibility Study Prepared by: Kwan Henmi Arc Transit Proje Kive Lea In Professional Allian BART Platform Screen Door Feasibility Study December 2016 Draft FINAL 12-5-2016 Prepared by: Kwan Henmi Architecture & Planning + HNTB Corp. BART Platform Screen Door Feasibility Study Prepared for BART by: Kwan Henmi Architecture & Planning Inc. with KivettConsult Transit Projects Strategies LLC Lea + Elliott #6M6092 WP-2 06-30-2017 BART Platform Screen Door Feasibility Study June 30, 2017 BART Contract #6M6092 WP-2 Prepared by Kwan Henmi Architecture/Planning, Inc. with KivettConsult Transit Project Strategies Lea + Elliott BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 3 Table of Contents Executive Summary....................................................................................................................................... 4 I. Introduction .......................................................................................................................................... 6 II. Background ........................................................................................................................................... 7 III. Analysis ............................................................................................................................................. 9 A. Understanding of Problems, Issues, and Solutions........................................................................... 9 B. Field Observations .......................................................................................................................... 11 C. Risk Assessment .............................................................................................................................. 13 D. Design Concept Development ........................................................................................................ 14 1. Train Control Interfaces – Vehicles and PSDs ............................................................................. 14 2. Facilities....................................................................................................................................... 17 3. Emergency Protocol .................................................................................................................... 18 E. Experience in Other Transit Agencies in the US.............................................................................. 19 1. NYCTA’s 2 nd Avenue Subway Line, New York City ....................................................................... 19 2. WMATA, Washington DC Region ................................................................................................ 20 3. Regional Connector, Los Angeles MTA ....................................................................................... 20 4. BART, San Francisco Region ........................................................................................................ 20 5. Honolulu Rapid Transit Project (HRTP): ...................................................................................... 21 F. State of the Art in Overseas Systems .............................................................................................. 22 IV. Evaluation........................................................................................................................................... 23 A. Design Criteria:................................................................................................................................ 23 B. Capital Costs:................................................................................................................................... 23 V. Recommendations / Next Steps ......................................................................................................... 24 VI. Primary References ......................................................................................................................... 26 VII. Appendices ...................................................................................................................................... 27 A. Appendix A: Problems/Basis for Developing Solutions .................................................................. 27 B. Appendix B: Risk Assessment Matrix ............................................................................................. 31 C. Appendix C: Design Criteria/Synopsis of Findings.......................................................................... 36 D. Appendix D: Embarcadero Schematic Plans for PSD Installation .................................................. 38 BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 4 Executive Summary Context BART has experienced a substantial increase in ridership over the past three to five years, primarily because of dramatic increases in employment in downtown San Francisco as well as the inner East Bay and San Mateo County. The ridership growth has resulted in train and station crowding in the core of the system. Within stations, there is a particular concern with congested platforms, with greatest concern at the Embarcadero and Montgomery Stations. Recognizing that any significant increase in basic station capacity at these underground locations requires a substantial capital investment, BART requested a feasibility study be undertaken to document the issues, benefits, and challenges associated with installing platform screen doors (PSDs) to provide a physical barrier between the trainways and the platforms at congested underground stations. This feasibility study was undertaken by the Consulting Group, composed of Kwan Henmi, KivettConsult, Transit Project Strategies, and Lea + Elliott. The Group also reached out to potential PSD suppliers for input. The product of the assignment is “BART Platform Screen Door Feasibility Study,” produced by the Consulting Group. Key Findings • PSDs can be effective in addressing station overcrowding and potential safety issues in the heavily-used BART subway stations. • The PSD opening, closing, and locking can be synchronized with train door opening and closing, and can accommodate BART’s existing 2-door fleet as well as the next generation 3-door fleet. • Train berthing accuracy for existing fleet must be increased for PSDs to be successful; this is achievable subject to testing and validation by BART operations. Next Steps The study’s findings strongly suggest that core BART stations would be well served by the installation of PSDs. We highly recommend that a test be undertaken at one of the congested underground stations, with one platform of the 12 th Street Station in downtown Oakland the recommended site. Proposed next steps: 1. Station Infrastructure Evaluation a. Verification of the structural integrity of the existing BART subway station platform edges to support the doors and the screen walls b. Impacts on the station and tunnel ventilation systems, as evaluated using computer simulation models, with any necessary retrofits identified 2. Engineering Analysis a. In-depth Engineering and Systems Feasibility Report BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 5 b. Car and Platform Screen Door Equipment Interface Simulation Tests; c. Confirmation that stopping accuracy can be held to an acceptable plus/minus tolerance to the PSD center point d. Installation Engineering and Constructability Report Both the Station Infrastructure Evaluation and the Engineering Analysis would need to support the findings of the current study before a pilot project would be recommended for full definition, funding, and advancement. Upon completion of the two studies, if results are favorable, we urge BART to begin to deal with the continuing issue of platform crowding with a pilot PSD installation at a congested underground station such as 12 th St. Station. Data, including human behavior information, would be collected from the pilot location for review and evaluation to determine if PSDs could be valuable elsewhere in the system. BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 6 I. Introduction This feasibility study was undertaken because of BART’s ongoing concerns about overcrowding and potentially unsafe conditions for their riders at several of the downtown San Francisco subway stations. The study addresses the physical requirements to install the PSDs, along with any train control modifications to the current and future BART fleet for simultaneous opening of the rail car doors and the PSDs. The benefits to be derived from use of the PSDs at the core stations for crowd control and safety will be documented. Rough order of magnitude capital costs for installing the PSDs on a subway station, and any modifications to the train control system, and the existing fleet for full synchronization will be identified. We reviewed the experience of other US transit agencies that have explored the use of train screens, along with retrofit installations overseas that were undertaken while maintaining ongoing station operations. A synopsis of the benefits, issues, and potential next steps has also been provided. BART authorized a special study group with expertise and experience in PSDs to undertake this study. The Consulting Group consists of the following: • Kwan Henmi Architecture/Planning, Inc., Mona Tamari, Project Manager/Architect • Hanan Kivett, KivettConsult-Transit Station Architect/Planner • P. Takis Salpeas, Transit Project Strategies-Rail Transit Consultant • John Kennedy, Lea & Elliott-APM Systems Specialist BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 7 II. Background The barriers referred to throughout this Study as Platform Screen Doors (PSDs) operate in a manner similar to elevator doors and are commonly used in fully automated driverless airport people mover systems around the world. BART’s Oakland Airport Connector along with San Francisco International Airport’s landside people mover system are local examples of platform screen doors. The PSDs provide benefits related to crowd control, platform capacity enhancements and passenger information systems, while virtually eliminating the risk of riders falling into the trainway from the platform. PSDs have been installed in existing and new subway stations in Paris and are widely used in major Asian transit systems both for new stations as well as retrofits to existing stations, primarily to achieve the aforementioned safety benefits. Asian transit systems have also installed them to reduce air conditioning levels, with the full height platform doors substantially reducing the areas requiring cooling. The barriers also reduce the transference to the platforms of heat from the train operations, and from the tunnel air. In addition to BART, four transit agencies in the US, namely Washington Metropolitan Area Transit Authority (WMATA), New York City Transit (NYCTA), Honolulu Area Rapid Transit (HART), and LA MTA, have studied installation of train screen doors primarily for safety and crowd control purposes. To date, only HART, a fully automated system in Honolulu, is proceeding with installations. Examples of the three types of PSDs being used in transit systems overseas today follow. BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 8 Full Height PSDs: Tokyo Japan Partial Height PSDs: Hong Kong Low Height PSDs: Shanghai China BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 9 III. Analysis A. Understanding of Problems, Issues, and Solutions In order to provide a framework for undertaking the technical work associated with ascertaining the feasibility of using PSDs, the Consulting Group conducted a series of focused meetings with key BART managers and staff in Oakland on September 26 th and 27 th , 2016. These meetings provided a forum to identify benefits, issues, and potential solutions that would have to be addressed during the course of the Study. The meetings were organized into the following categories: A. General B. Vehicles C. Systems D. Operations E. Stations F. Safety Appendix A: Problems/Basis for Developing Solutions summarizes the feedback collected during the 2-day working sessions in Oakland. This led to the Study Methodology and Timeline that formed the basis of this Study and follow-on tasks, with a suggested timeframe for moving forward with a fully operational pilot demonstration program. During the course of advancing this initial task, our Group realized that there was a lack of any direct experience in the actual installation and use of PSDs in the United States. In order to provide real-world experience and data for installing PSDs in existing rail transit stations, we also conferred with potential PSD suppliers. BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 10 BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 11 B. Field Observations The consulting team toured three stations (Embarcadero, Montgomery and 19th St Oakland) to understand the current characteristics regarding passenger movements and behaviors, station layouts and location of vertical circulation elements. As- built drawings were obtained for these 3 stations to be used in understanding the physical dimensions, the structural characteristics of the platform, the location of signing, lighting and communications lines, etc. Photos were taken of key features to provide a reference for follow on work associated with the development of design concepts for the PSDs. While Embarcadero is a station most in need in of PSDs due to its heavy usage, the team initially believed 19th Street station would be suitable for a demonstration program because of its reduced passenger flow. However, after further research, we recommend 12th Street Station; 19th Street has timed transfers which would be disrupted should single tracking occur on the upper level for any reason. The 12th Street Station platform features are the same as the 19th Street Station, and should provide a suitable environment for a pilot of PSDs before implementing in high-traffic stations like Embarcadero. Embarcadero Station-Rush Hour Congestion Source: BART from AECOM Study 2016 BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 12 Embarcadero Station-Normal Activity Levels BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 13 C. Risk Assessment We prepared a risk matrix to document major technical issues that would have to be resolved through follow on studies focusing on train control, current operating procedures and the physical components required to install PSDs. This Matrix is organized by Fixed Facility or System Elements and identifies potential risks, possible mitigations and means and methods of providing solutions as part of follow-on engineering studies if there is a determination by BART to advance the installation of PSDs in major downtown subway stations in San Francisco and Oakland. The 4 Square Risk Matrix below summaries the major findings as a result of the initial assessment. See Appendix B: Risk Matrix for the full tabulation. Ongoing Issues Short - Term Acceptable Risks • Impacts on: Station & tunnel ventilation systems; platform structural integrity; power, etc. • Requirement to install Radio Frequency ID Tags to confirm number of cars in consist and door quantities (2 vs.3) • Stopping distance accuracy for fleet • Synchronizing BART and PSD door open and close signals • Synchronizing BART and PSD door closed and locked signals. • BART follow-on Tasks to Implement PSD Demo Program • Approvals for Pilot Demonstration Program • Demo Program duration and acceptance/rejection criteria • Available expertise to carry out Pilot Program • Follow-on program for permanent station installations if Demo Program is successful • Demo Station location selection: Passenger volumes sufficient for data collection; post PSD Demo; opportunities to minimize short-term constructability impacts on train operations/mitigations when single tracking • Funding for Engineering, Capital and O & M costs to support Pilot Demonstration Program • Procurement Method • PSD Demo Program feasible • BART Management support • BART can be first transit agency in the United States to undertake actual testing of the PSDs • PSDs are well proven transit system station appliances in several leading high-volume rail systems worldwide of high reliability/punctuality Potential Threats Good Trends BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 14 D. Design Concept Development The key technical areas where physical as well as systems changes will be necessary for the installation and operation of PSDs are outlined below. 1. Train Control Interfaces – Vehicles and PSDs a. Train Control Interfaces System installations that provide for platform screen doors (PSD) must be designed to allow for the interface with the vehicle and the train control system installed on the wayside and at the platform itself. Such interfaces are necessary to address the requirements for the platform doors to operate in concert with the vehicle doors in a safe and efficient manner. These requirements include the following, which will be discussed in detail below: • Precise (within an allowable tolerance) stopping alignment of the vehicle door and the platform door or gate; • Coordinated door opening and recycling; • PSD open or closed status signal; • Signal indicating all doors closed and locked. Dedicated space on the platform that doesn’t interfere with passenger movements will also be necessary for platform screen door controller equipment. b. Precision Stopping Passengers must be able to board or de-board trains of any length freely without encountering an obstruction to the vehicle door opening. The vehicle doors and the PSD automatic sliding doors must align to a certain tolerance when the train stops in a station. The tolerance describes the distance the train can stop short or long of the platform door opening and still allow for the passenger boarding or de-boarding movements. On Automated People Movers systems, such as that installed on the Oakland Airport Connector (OAC), this tolerance is restricted to plus or minus six (6) inches from the exact center of the station door opening. The wayside train control system (generally defined as Automatic Train Operations or ATO) and on-board train controller interface provides the precise alignment at the station. The system must also account for possible vehicle or wayside failure that causes stops that are longer or shorter than the normally specified +/- 6 inches. The maximum allowable misalignment would still provide passengers at least thirty-two and one-half (32.5) inches to evacuate a vehicle. To provide adequate exit width even in case of a severe BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 15 misalignment or emergency, however, the PSD system must incorporate emergency panels in addition to the automatic sliding doors. With the inclusion of the emergency panels, the PSD system now functions as a complete barrier wall system (as does the OAC), which incorporates all equipment necessary to allow for access or exit from a train in normal or emergency operations. As part of the design of the PSD barrier wall system, analysis must be conducted illustrating each potential stopping position at the platform and demonstrating that the minimum required egress opening width is achieved. For BART, this analysis would examine both precision stopping as well as the function of the emergency panels. c. Coordinated Door Opening, Closing, and Recycling The design of the PSD system must be such that the platform sliding doors will not automatically open – under any circumstance – unless there is a train in the station and it is aligned, to the specifications detailed above, with the sliding doors of the barrier wall system. Generally, the sequence of events that must occur to open the station doors is as follows: i. The Station Automatic Train Operations (ATO) signals the train as to the distance remaining before it is to precisely stop in the station; ii. The train’s onboard computer analyzes the Station ATO signal and begins to brake the train such that the precise door to door alignment is achieved; iii. Once the train is in the station and stopped – meaning that the train is at zero speed for an established period and the propulsion is disabled – the train signals the Station ATO through on-board antennas that it is safe to open the station doors. iv. The station doors are commanded to unlock and the train and the station doors open simultaneously; v. After opening, the PSD system sends a signal to the Station ATO that its doors are now open. This is necessary so that the control system knows that the doors are functioning as commanded; vi. After a specified dwell time, the Station ATO commands the station doors to close while simultaneously signaling the train to close its doors; vii. The doors now begin to close. If the doors are obstructed for any reason, the station doors will automatically open to what is called a recycle mode, while sending a signal to the Station ATO controller to command the vehicle doors to recycle open. BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 16 viii. If the obstruction clears, then the Station ATO will command both the train and station doors to close; ix. The station doors close and lock themselves and subsequently send a signal to the wayside Automatic Train Protection (ATP) – which is designed as a fail-safe – that it is safe to allow for the train to depart; and x. If all train doors are closed and locked, as verified by the trains on-board ATP, then the train can depart the station. This sequence will take approximately 5 seconds, not including dwell time. It should be noted that if the train is not aligned properly, the station doors will not automatically open. This is an absolute safety function. d. PSD Status – Doors Open or Closed or Locked PSD systems are design to prevent intrusions into the active guideway area. Thus, if a PSD automatic sliding door or an emergency exit panel within the barrier wall opens for any reason, then the wayside ATP commands the train control system to place a stop on the system to prevent the train from entering the station platform area and possibly striking a person. PSD automatic sliding doors are equipped with devices that lock the doors in a fail-safe manner. As described above, the lock can only be removed if there is train in station that is precisely aligned. In the event of an emergency, the automatic sliding door is also equipped with an emergency release accessible from the station and guideway side of the door. The release on the guideway side of the door is for passenger evacuation from inside the train, while the release on the platform side of the door is for emergency personnel to enter the train. The PSD emergency panels can only be opened from the guideway side and are equipped with panic or other type of emergency handles that allow a passenger to evacuate from a misaligned train. e. Mixed Fleet and Multi-Car Compatibility The PSDs will need to collect information from the train to determine if it is a newer type of car with three doors or an older car with two doors. Additionally, the operating system would need to understand how many cars are associated with the train. The operating system is capable of opening only the doors needed depending on which car type comes into the station. BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 17 2. Facilities a. Architectural Considerations Key physical design considerations for installing PSDs at a BART station include: • Spacing of screen doors and emergency exit doors to accommodate the mixed fleet: 2 doors on the existing rail car fleet and 3 doors on the new fleet • Material selection and finishes of PSDs to be compatible with the existing station aesthetics • Impact on existing platform materials, door opening positions, safety warning strips • Impact on existing overhead lighting, signing, PA systems, CCTV, conduit chases, etc. • Potential for perforated panels to provide some through ventilation from trainway • Structural attachment to platform; i.e., materials and assembly methods • Dimensions from existing platform edge to accommodate required gap between the rail car and the inside face of the PSDs. • Potential for incorporation of lighting, train information, PA system, and advertising (as an option) A set of schematic plans illustrating the location and positioning of the 5’-7” high doors and 5’-9” high screen walls has been developed for the Embarcadero Station. See Appendix C: Embarcadero Station Existing Conditions and Proposed PSD Layouts b. Structural Considerations There are two main considerations when it comes to how the PSDs will perform structurally: the dead weight and the weight when people push against the doors. The platform is also cantilevered at the edge, allowing for a continuous refuge area underneath that must be preserved. The dead weight of the load is not a significant impact, but the sideways force of people pushing against the doors will likely be the key factor in determining if the platform edge requires structural reinforcement through the use of vertical struts from the track invert to the underside of the platform edge. Spacing and sizing of the struts will have to be minimal to avoid any impact on the use of the refuge areas under the platform. A redesign and reconstruction of the platform edge may be required depending upon the condition of the aging concrete. It is recommended that a focused study be undertaken to verify the structural capability of the existing platforms and what reinforcements may be necessary. BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 18 c. Grounding A possible shock hazard is created between the PSD and vehicle doors. This is due to the current traction power system being an isolated (non- grounded except at the substations) Direct Current (DC) system and the PSD which is powered by an Alternating Current (AC) grounded system. To avoid stray currents affecting passengers, it will be necessary to insulate the facade from the station and connect it to the rail to ensure it is at the same potential as the car body d. Station Ventilation Part-height PSDs are recommended to avoid impact on the existing station and tunnel ventilation systems. Use of 5’-9” tall screen doors appears most suitable, high enough to discourage people jumping over but low enough to minimize impacts on the station and tunnel ventilation. As part of verifying these impacts, it is recommended that a focused technical study be undertaken use computation fluid dynamic (CFD) models. Supplemental ventilation devices may be necessary at selected stations to assure the effectiveness of these life-safety systems. e. Train Operator Visibility Today, the BART train operators stick their heads out from the car to determine if it is suitable to close the doors. The installation of a PSD will prevent train operators from doing so. Alternate methods of determining whether it’s clear to close the doors must be analyzed. Special mirrors or CCTV cameras and screens mounted on the end of the train room walls may be effective. f. Power Availability – California Public Utilities Commission (CPUC) One unit of the PSD takes 300 watts to operate, which is relatively low. Should an emergency occur and power should fail, there are manual ways of opening the PSDs from the trackside. It is possible to also manually open the PSDs from the platform side with a key. 3. Emergency Protocol Protocols would have to be developed for BART Central Control to manage these types of emergencies: • Power failure • Egress when train is misaligned • Egress when train is in the tunnel • Screen door failing to open or close • Passenger or object caught in PSD BART Platform Screen Door Feasibility Study 06-30-17 Kwan Henmi Architecture/Planning | KivettConsult | Transit Project Strategies | Lea + Elliott | Page 19 These protocols should be simple, easy for passengers to understand, and follow common sense. E. Experience in Other Transit Agencies in the US Members of our Consulting Group have been actively engaged in assessing the feasibility of installing PSDs in other US transit systems where consideration has been given to addressing platform crowding, safety and prevention of users entering the active trainway. New York City Transit, WMATA and LA MTA have all examined the potential for use of these devices. 1. NYCTA’s 2 nd Avenue Subway Line, New York City NYCTA through their AE design team (DMJM-Harris/Arup) advanced their efforts through Conceptual Engineering in the 2005-2007 timeframe and produced an extensive technical report entitled, “Feasibility of Installing Platform Edge Doors at 96 th , 86 th and 72 nd Street Stations” issued in 2007. The technical/engineering aspects of providing full height PSDs in the 3 new stations were feasible, and benefits related to safety, temperature control, and emergency ventilation systems (which would be designed at the outset to take full advantage of smoke control) were acknowledged. The large variations in mixed fleet door locations, lack of precision stopping and necessary changes to train control were considered too challenging to implement at the outset. Our current understanding is that provisions were made for future installation. The Phase 1 stations are now operational. Phase 2 is adding 3 more stations; now in Preliminary Design. New York City Transit- PSD Design Concept for New 2nd Avenue Subway Line Source: DMJM - Harris 2007