RandstadRail: Increase in Public Transport Quality by Controlling Operations

Transcription

1 RandstadRail: Increase in Public Transport Quality by Controlling Operations Niels van Oort HTM, Urban public transport, Research and development P.O. Box 28503, 2502 KM The Hague, The Netherlands, Rob van Nes Delft University of Technology, Transport and Planning Section, P.O. Box 5048, 2600 GA Delft, The Netherlands, Abstract RandstadRail will be a new regional public transport system between The Hague, Zoetermeer and Rotterdam. RandstadRail will be a high level public transport system: short travel times and high frequencies. During peak hours the intensity on some trajectories will be about 24 vehicles an hour. RandstadRail will share some tracks with other tramlines as well. Dealing with these high frequencies, a system for controlling operations is necessary: because of offering travellers a high quality product, according to waiting times and chance of getting a seat, as well as making optimal use of the restricted capacity. To prevent congestion the operation of RandstadRail should be as much as possible according to the timetable. To reach this, HTM designed a three step controlling philosophy. The most important aspect is distribution in deviation of the timetable. First step is preventing this deviation to occur: the infrastructure will be as much as possible exclusive right of way and at intersections RandstadRail will get priority over the other traffic. RandstadRail will stop at every stop and will never leave before schedule time. Second step in the philosophy is dealing with the deviation by planning extra time in the schedule at stops, trajectories and terminals. Small deviations can get solved in this way. Final step to get vehicles back on schedule will be done by the traffic control centre: they have a total overview of all vehicles and can take measures like slowing down vehicles nearby a delayed vehicle. In case of big disturbances derouting and shortening of lines is possible. Keywords Reliability, Operations control, light rail, punctuality, delay management 1

2 1 Introduction In the south of the Dutch Randstad-area, the need for space to live and work is growing rapidly. The area between the cities of The Hague and Rotterdam urbanises fast. New residence areas are built of already in use. All these developments lead to an increasing demand of transport. RandstadRail is a new regional public transport system with high quality standards: high frequent, fast, comfortable and reliable. RandstadRail is a connection of tram-, metro- and heavy rail lines. RandstadRail will be completed in 2007 and it will consist of two main networks (illustrated in figure 1)[3]: Zoetermeerlijn. The tram lines 3 and 6 in The Hague will be connected to the former heavy rail line in Zoetermeer (called Zoetermeerlijn ). HTM, the public transport company of The Hague will operate on this line; Hofpleinlijn. The secondary, former, heavy rail line between The Hague and Rotterdam (called Hofpleinlijn ) will be connected to the metro network in Rotterdam. The public transport company of Rotterdam (i.e. RET) will operate on this line. Figure 1: RandstadRail network 2

3 Figure 2: HTM RandstadRail network This paper only focuses on the Zoetermeerlijn: the connection of the tramlines of The Hague with the former heavy rail line in Zoetermeer. Figure 2 shows this network. The network exists of the following lines: Line 3 (turquoise line): The Hague Loosduinen- Central Station - Zoetermeer Centrum West and vv; Line 3 Short (blue line): Only operates during peak hours on a part of line 3; Line 6 (yellow line): The Hague De Uithof- Central Station - Zoetermeer Oosterheem en vv; Line 6 Short (green line) Only operates during peak hours on a part of line 6; RandstadRail 3 and 6 will partially share tracks with HTM tram lines and the RET RandstadRail line (purple line in figure 2). 2 Why controlling? Nowadays bus and tram lines in The Hague are not controlled in any way: The driver knows the departure time at the first stop and the arrival time of the last stop. During the trip some deviations will occur: this results in a distribution of driving times [4]. The deviations of the tram lines 3 and 6 during rush hours are illustrated in figure 3. This figure shows a distribution in deviation of 3 minutes at the first stop and approximately 10 minutes at a stop in the city centre of The Hague. 3

4 Figure 3: Punctuality on tram line 3 and 6 at the first stop (u) and halfway the route (l) All 4 RandstadRail lines will operate 6 times an hour in both directions, during peak hours, which results in headways of 2,5 minutes on the shared part. Because of these high frequencies, deviations as showed in figure 3 are unacceptable. This research deals with the minimization of these deviations to reach schedule adherence or regularity. This is necessary because of two main goals: To offer high service quality In case of public transport service with high regularity and a uniform arrival of passengers at a stop the chance of having a seat is maximized. Regular arrivals of vehicles prevent a poor distribution of passengers over the vehicles: a mix of empty vehicles and overloaded vehicles. Besides maximizing the chance of having a seat, the waiting times are minimized in case of high punctuality and regularity [5]. 4

5 To prevent congestion On some parts of the RR lines, signalling is applied. On these parts, capacity is limited, compared to the non-signalled parts. On the busiest parts of the line, the total frequency of all lines (tram and RR) is approximately 40 vehicles an hour in both directions. The capacity of the infrastructure is sufficient, if regular services are assumed. If substantial irregularity occurs, vehicles will be delayed, because of congestion. Bunching of vehicles should be prevented. To achieve this, attention needs to be paid to prevent deviations next to curing afterwards. RandstadRail must become a regional public transport system with high reliability. To achieve this, the transport authority also demands quality standards. Table 1 shows the threshold values, related to reliability, that are demanded. Without extra measures these values are not achievable. Table 1: Threshold values of reliability Maximum percentage Trips ahead of schedule 0 % Delay greater than 2 min 5% Delay greater than 4 min 1% 3 Controlling philosophy To prevent large deviations as well as to decrease the impact of the effect of deviations, HTM designed a new, three step controlling philosophy. Figure 4 illustrates these steps, with their main actor per phase. The three steps are described below in more detail. Disturbances A. Preventing B. Coping C. Adjusting Planner Planner/Driver Dispatcher/driver Figure 4: Controlling philosophy 5

6 3.1 Step A: Preventing The first step is preventing the deviation to occur. This is the most important step: An ounce of prevention is worth a pound of cure. There are different ways to prevent deviations to occur. For RandstadRail the next changes will be applied: Improvements of infrastructure An important cause for deviations is infrastructure. RandstadRail exist of two different parts: the former tramway and the former heavy railway. The latter is 100% own right of way without any crossing of other traffic. Whereas the former tram part has crossings and shared use of lane. Before the start of RandstadRail a lot infrastructure is improved: the shared use with other traffic is decreased, several crossings are closed and priority at intersections is improved. This won t prevent all deviations, but it will decrease significantly. Besides shared use of lanes with cars, RandstadRail shares some tracks with other public transport too. At intersections RandstadRail will have priority over other tram lines as well. Punctuality of the vehicle will be shown to the driver in the cabin To prevent deviations to occur, the driver gets real time information of his actual schedule adherence. Figure 5 shows the display which is in the cabin of the vehicle. In the right the delay is showed in quarter of minutes, supported by a background colour, depending on whether the vehicle is too early, on time or too late. This information will help the driver to depart on time at the first stop and prevent him to depart ahead on schedule during the trip. If the vehicle is delayed the driver can try to speed up by using the slack time in the schedule (e.g. using a shorter dwell time). Figure 5: Information display in cabin of RandstadRail vehicle 6

7 Departing on time at the first stop At the moment, schedule adherence at the first stop is not sufficient (like figure 3 illustrates). The schedule adherence differs per endpoint and time of day. To achieve a high punctuality, which is demanded for RandstadRail large delays at the start of a trip is unacceptable. One of the main causes of high unreliability is poor punctuality at the start point. The following aspects are of influence on departing on time and will be important for RandstadRail: The vehicle will be ready to start on time at the first stop. To achieve this it is necessary to ensure a high level of punctuality and enough slack layover time; The driver will be ready to start on time at the first time. If a driver is not present at some point before the scheduled departure time the central dispatching room is warned by the control system. They decide how to solve this (e.g. contact the driver, use a spare driver); The departure time of the next trip will be shown to the driver in the vehicle as well as in the waiting area for the drivers; It is important the driving time is achievable. If it s not, the driver will adjust his departure time to reach the endpoint on time. Driver should trust the timetable. Excellent training on this is unbearable; The central dispatchers will be warned by the control system if the delay of a vehicle on the first stop exceeds a certain threshold; The schedule adherence will be continuously monitored and published to the management, dispatchers and especially drivers. New dwell process Besides infrastructure the dwell process is a significant cause of deviations as well. These deviations can be prevented as well. Important factors are: RandstadRail vehicles stop at every stop, like a train system. Travellers do not need to use a stop-at-demand button. Using constant dwell times in the schedule, deviations will be decreased. Figure 6 shows the total dwell time of trips of tram line 3. This figure illustrates the deviations in total dwell time per trip. These deviations are mainly caused because of skipping stops or non-constant dwell times. Introducing stopping at every stop will decrease the deviation greatly. 7

8 Figure 6: Deviation in total dwell time per trip (tram line 3; peak hour) RandstadRail will be operated with new state of the art vehicles. The vehicle and platform are on an equal level and broader doors enable a smoother way of boarding and lightening. Factors that are of great influence on dwell times, like trolleys, will have less effect. In this way, the new characteristics of the vehicles will eliminate a significant cause of delay. A significant change with RandstadRail will be that vehicles are not allowed to leave ahead of schedule at any stop. Driving ahead of schedule creates deviations and could lead to bunching of vehicles. This will result in a lesser quality level of service: long average waiting times and a high probability of overcrowded vehicles. To prevent driver to depart early a display in his cabin will show the deviation of the timetable, such as showed in figure 5. When a vehicle arrives too early at the stop it has to extend its dwell time. This extra dwell time will be short though, because of a tight schedule and a no early departure at the last stop. At the moment bus and tram drivers sell tickets and give information to travellers. In RandstadRail vehicles these tasks are automated: tickets can be sold at machines and an information button in the vehicle connects the travellers to the customer information centre of HTM. The driver is seated in a closed cabin and no contact is possible with travellers, like in a train. This change will decrease delays as well New planning process RandstadRail will be public transport system with characteristics between a tram and a train. Besides new infrastructure and a new way of operation this requests another way of planning as well. Normally, tram lines are planned calculating the total trip time. This calculation will be used to estimate the departure time of each stop. To achieve a better accuracy, the RandstadRail schedule will exist of driving times from stop to stop, like in heavy railway planning. This way, the schedule will be more accurate and driving ahead of schedule is easier to prevent. Because of the high punctuality standards, an accurate as well as an up to date schedule is very important. To improve this schedule, a feedback loop from to operation to schedule design is necessary. In tramway planning, this is an often used loop, in contrary to heavy railway scheduling. 8

9 Figure 7: Distribution of driving time during the trip (black=current; white= after improving infrastructure and vehicles; grey=after improvement of departing at first stop and no more early delays) Figure 7 roughly shows the expected effect of all these changes. The black lines are the maximum current deviations. Improvements of the infrastructure and the vehicle will reduce the increase in deviations (white line). Improving the punctuality at the first stop and preventing vehicles to depart early will decrease deviations as well (grey line). 3.2 Step B: Coping with deviations The second step in the philosophy is dealing with the deviation by planning extra time in the schedule at stops, trajectories and terminals. Small deviations can be solved in this way. Carey [1] and Israeli [2] deal with this topic as well. Adding extra time in the timetable will enable late vehicles to catch up. This extra time is a trade off of operational speed and reliability. The bigger the slack time is, the bigger reliability is, but the lower the average operational speed of the vehicles is. The slack time can be added to different parts of the trip time: driving time, dwell time and layover time: Slack time in driving time Driving time is planned from stop to stop. In this way, more accurate schedules can be designed. A small amount of slack time is added to the driving time. This amount depends on the characteristics of the track and the expected deviations. Especially tracks in the city need some slack time, because of crowded junctions and shared tracks with other traffic. Slack time in dwell time It is also possible to plan some extra time per stop. This slack time depends on the expected distribution of dwell time and the effects of vehicles waiting longer than necessary at a stop. Because some tracks and stops are shared with other lines 9

10 blocking of other vehicles is possible. This should be prevented, especially at tracks where high capacity is needed. To use the extra dwell time efficiently, it is recommended to dwell longer at stops with a high number of boardings and alightenings and relatively little passengers travelling over the stop. Because of demands of the transit authority, slack time in dwell won t be that much: the maximum dwell time is set to 20 seconds, with an exception to crowded stops where this value can exceeded with 50%. Slack time in layover time Like stated earlier, extra time will be added to the layover time to achieve a high punctuality of departing vehicles. 3.3 Step C: Adjusting Sometimes, even after step A and B, deviations occur. In that case, the final step, adjusting, should be applied. In Wilson [6] several operational measures are described as well. Adjusting the operations of RR will be done by the dispatchers in the central dispatch room. They have a total overview of all vehicles and their punctuality. Software tools are available to adjust operations and inform drivers as well as travellers. 10

11 Figure 8: Overview of all vehicles on a line and their schedule adherence The first goal is to guard punctuality. Dispatchers are warned by the system, when punctuality is about to exceed certain thresholds, because of, for instance: Broken vehicles; Late drivers; Early departures; Late departures. 11

12 Early vehicles are requested to wait. Different measures can be applied, for instance: Shortening a trip; Adding a vehicle; Skipping stops to catch up; Change order of vehicles (of different lines) at junctions; These measures are always a trade off between the travel time of travellers in the vehicle and travellers at the stops. The secondary goal is to achieve regularity: If it is not possible to restore schedule adherence (e.g. because disturbances are too large) the dispatcher can apply regularity control. This principle is explained in figure 9. By slowing down some vehicles, regularity can be improved, which leads to shorter waiting times and a better distribution of passengers over successive vehicles. Figure 9: Principle of regularity control A: Vehicles drive with equal headways B: One vehicle gets delayed: the headway in front increases, behind it decreases C: By slowing down successive and preceding vehicles the regularity will partially be restored 12

13 4 Conclusion RandstadRail is a new public transport line that connects two tram lines with a former heavy rail line. RandstadRail will be a high frequent system sharing its track with trams and metros. To offer high quality service, especially in the field of punctuality and regularity, and to make efficient use of the infrastructure it is decided to apply a new control strategy. The main elements of the control philosophy are preventing, coping and adjusting. The punctuality of the vehicle is shown to the driver, so he can adjust his driving style. On top of that, all vehicles with positions and punctuality are shown in the central dispatching room. The dispatchers use a system, supporting them in adjusting operations, if necessary. RandstadRail has a high percentage of exclusive right of way and priority at traffic lights. The vehicles got low floors and broad doors, which reduces deviations in dwell time. All these new characteristics and measures enable RandstadRail to operate in high frequencies with high reliability. 5 Acknowledgements This research is performed in cooperation with HTM, the public transport company in The Hague and Delft University of Technology, Faculty of Civil Engineering and Geosciences, department of Transport & Planning. This research is supported by the Transport Research Centre Delft. 6 References [1] Carey M., Optimizing scheduled times, allowing for behavioural response, Transportation Research B, vol. 32, No. 5, pp , 1998 [2] Israeli Y., A. Ceder, Public transportation assignment with passenger strategies for overlapping route choice, Lesort J.B., Transportation and Traffic Theory, Elsevier Science, Amsterdam, 1996 [3] Oort, N van, Post, MR, Randstadrail: Kwaliteitssprong in operationele kwaliteit door exploitatiebeheersing, Duurzame mobiliteit: hot or not? Bundeling van bijdragen aan het colloquium, pp , Rotterdam: CVS, 2005 (in Dutch) [4] Oort N. van, R van Nes, Reliability of urban public transport and strategic and tactical planning, a first analysis, TRAIL conference 2006, Rotterdam, 2006 [5] Oort N. van, R van Nes, Service regularity analysis for urban transit network design, 83rd Annual Meeting of the Transportation Research Board, Washington DC, pp. 1-24, 2004 [6] Wilson, et al. (1992), Improving service on the MBTA green line through better operations control, Transportation Research Record 1361, TRB, National Research Council, Washington, D.C., pp ,