Telephone: LTK Engineering Services Facsimile: 530-795-4403 27717 County Road 92F 530-795-4403 Winters, CA 95694

[email protected]

MEMORANDUM TO: Brent Efford, KiwiTram Wellington, New Zealand

FROM: Tom Matoff, Director, Transportation Planning LTK Engineering Services

DATE: 3 December 2013

SUBJECT: Johnsonville Line LRT Concept

This memorandum summarizes a rough conceptualization of conversion of the Johnsonville Line from suburban railway operation to a concept directly serving Wellington’s Central Business District (CBD). It was prepared as a partial response to statements made in published Spine Study documents, also quoted in the press, referencing high infrastructure costs associated with a “northern” extension of Light Rail Transit as the basis for eliminating from consideration the option which, on the face of it, would seem to be Wellington’s single most promising opportunity for a demonstration light rail line. The Spine Study’s technical conclusions appear to be summarized on page 5 of the report entitled “Option Evaluation Results, Technical Note” (16 June 2013). The following statements, among others, are made there, under Section 3.2 “Short List Corridors”:

For connections to the north 3.2.1

- Converting the Johnsonville line to Light Rail Transit as a sub-option, has its own costs and challenges:

• There would be significant costs for conversion (tunnel widening, platform lowering, additional passing loops) • Double the number of Light Rail Vehicles would be needed to replace the capacity of the 4 car Matangi holding 490 passengers. This would increase the number of vehicles required and necessitate additional passing loops or double tracks. • There would be significant disruption to services for a long period during construction/conversion.

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 2

For these reasons, extending the public spine to the north was not pursued further in this study. The proposal is to provide on-going train and bus service improvements.

I could not find, in the limited time available, any simulation data, operations calculations or cost estimates in the published documents to support these statements. I assume some documentation must exist somewhere for the consultant to have stated these conclusions so categorically, and perhaps if this documentation were made available, it would be possible to evaluate more closely the validity of the authors’ arguments in the context of this particular application.

My own brief analysis, and I emphasize this has been made in the absence of an actual site visit, leads me to question the validity of these three statements. I have other concerns as well, but absent sufficient time to delve into the minutiae of the demand forecasts and underlying network and service assumptions, I will only note that the report’s observation that…”The average distance for rail users to their final destination from the Wellington Rail Station is short (0.9 km), with the majority of these trips by walking or cycling (90% in the AM peak)…” may not be a justification for eliminating a CBD light rail extension of the Johnsonville Line from further consideration so much as it might be a reflection of the inadequacy of a service that terminates short of activity centers.

The following discussion deals with the main aspects of a conversion concept, and demonstrates that a reasonable basis exists to reach conclusions contrary to those cited above from the Spine Study.

1. Track and Station Layout Assumptions

The current track layout is assumed as shown in the segment of the line north of Wellington Station in the schematic sketch in Attachment 1. This is a single track line with passing loops, and significant physical constraints (grades, tunnels). Station platform lengths are assumed to be long enough to accommodate three two-car Matangi sets, or (3)(42)=126 meters, although you advise that not all platforms may actually meet this standard. Passing loops are assumed to be of sufficient length for three Matangi sets. Overall distance from Wellington Station to Johnsonville Station on the present alignment is 10.49 km.

The critical issue of passing loops on a single-track line is based on the understanding that they are located as follows, with inter-loop distances as indicated:

● at Wadestown, 3.09 km from Wellington Station;

● at Ngaio Station, 2.33 km from Wadestown Loop, and a total of 5.42 km from Wellington Station; and

● at Khandallah Station, 2.47 km from Ngaio Station Loop and a total of 7.89 km from Wellington Station.

Another critical consideration is the length of single track beyond the last loop, Khandallah, to the terminal at Johnsonville. This important distance is 2.60 km. As Johnsonville is presently a single-track terminal, trains travelling north from Khandallah Station must be able to make it to TO: Brent Efford RE: Johnsonville Line LRT Concept Page 3

Johnsonville, take their terminal time, change ends, and return to Khandallah Station within one headway. As we shall see, this appears to pose no problem.

The conceptual extension alignment through the CBD, and possible station locations, are assumed as shown in Attachment 2. The detailed alignment from Wellington Station into Lambton Quay, with a useful pocket track, is assumed as shown in Attachment 3. The distance from Wellington Station to Courtney Place via this alignment is understood to be 2.45 km. There are 7 intermediate stations in the central city. The entire alignment south of Wellington Station has double track, and therefore no passing loops are required in this segment. The timetable and rotation of trains are determined by the schedule requirements over the layout of track and passing loops north of Wellington Station.

Per the concept schematic you provided, a pocket track is assumed just south of Wellington Station in the transition area between the station and Lambton Quay. (In addition, if Wellington Station were to serve as the northern terminus of a CBD shuttle or a line from the south, the track layout at Wellington Station should enable northbound runs to be brought into the southbound station track to exchange passengers and change ends, without having to run through the length of the pocket track. This consideration, however, can be left to the future.) At the southern terminal, Courtney Place, I assume a simple facing point crossover, with a tail track sufficient to store one train beyond the facing turnout if necessary.

To accommodate level boarding, nominal platform height would be approximately 400mm above top of rail. At and north of Wellington Station, this would entail either reconstruction and lowering of platform surfaces, or, perhaps more cheaply, raising the track. Many platforms have significant curvature, and designers would eventually have to deal with this.

2. Current Running Times and Service Practices

Running Time

Current scheduled end-to-end running times on the Johnsonville Line are typically 21minutes, northbound and southbound. This reflects performance characteristics of current rolling stock, civil characteristics of the line (grades, curves), single-track segments, dwell time and train control efficiency (signal clearance times, switch and derail function times, driver/operator reaction times). Intermediate station-to-station and inter-loop distances and running times, and the cumulative distance from Wellington Station, are illustrated in the following table.

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 4

Station Distance from Distance Running time Running time Wellington Station between between Between (km) adjacent adjacent stations adjacent stations (km) (minutes: sb/nb) loops or terminals (minutes: sb/nb) Johnsonville 10.49 Johnsonville- Station Khandallah 1.16 3/2 Loop: 5/5 Raroa Station 9.33 1.44 2/3 Khandallah 7.89 Station and Loop 0.55 2/2 Khandallah Box Hill Station 7.34 Loop – Ngaio 0.45 1/1 Loop: 6/7 Simla Crescent 6.89 Station 0.88 1 /2 Awarua Street 6.01 Station 0.59 2/2 Ngaio Station 5.42 and Loop 0.71 3/2 Ngaio Loop- Crofton Downs 4.71 Wadestown Station Loop: 5/4 1.62 2/2 (est.) Wadestown 3.09 Loop 3.09 5/5 (est.) Wadestown Wellington ------Loop – Station Wlngtn Station: 5/5

Running time between loops is the critical factor in determining potential service levels, as discussed below.

Current Service Levels

Weekday base period, weekend and holiday service is provided on a 30-minute clock headway, requiring two trainsets. Ample terminal and schedule recovery time is provided with this arrangement – 9 minutes at each end, more than enough to change ends and operating cabs after a trip of 21 minutes. Base period meets take place at Ngaio Station and Loop. TO: Brent Efford RE: Johnsonville Line LRT Concept Page 5

The running times shown in the public timetable are used to reach the conclusions below, but it is not entirely clear that they are accurate. The on-board video of operation with the earlier English Electric equipment, now replaced, showed a delay to the southbound train of more than 2 minutes at Ngaio, with a narration suggesting this was habitual practice. It was not clear whether the delay was due to early arrival of the southbound at Ngaio, delayed arrival of the northbound at the loop, or somewhat sluggish functioning of the signal and train control system, but if this is a continuing and routine problem it should be addressed. It may be that the 2- minute offset of Johnsonville and Wellington departures, intended to reflect the difference in Johnsonville-Ngaio and Wellington Station-Ngaio running times (9 minutes versus 11) should be evaluated, and if necessary adjusted.

The current Johnsonville Line service pattern is based on a 30-minute base bi-directional headway, augmented by additional service in peak periods. It is asserted that the service delivered achieves a 13-minute headway, but while it is literally accurate that some headways do narrow to 13 minutes, it is somewhat misleading to think of the service in this way. In actual practice, the peak departures and arrivals at Wellington station never exceed an average of 4 per hour, or 15 minutes, and then for only one hour and in only one direction, as shown in the following table (taken from November 2011 public timetable):

Hourly Interval Departures from Average Arrivals at Average One Wellington One Hour Wellington Hour Headway Station Headway Station 5-6 AM 0 ------0 ------6-7 AM 4 15 min 1 60 min 7-8 AM 3 20 min 3 20 min 8-9 AM 2 30 min 4 15 min 9-10 AM 2 30 min 2 30 min

BASE PERIOD 2 30 min 2 30 min

3-4 PM 4 15 min 2 30 min 4-5 PM 3 20 min 3 20 min 5-6PM 3 20 min 4 15 min 6-7 PM 2 30 min 3 20 min 7-8 PM 2 30 min 2 30 min

While this service pattern is not atypical for an old-fashioned concept of a traditional suburban railway service, it does not reflect contemporary regional trunk transport standards for an important urban area.

3. Current Rolling Stock

The Johnsonville Line is currently operated with new two-car Matangi EMU trainsets. It is my understanding that the sets are arranged as married pairs, with one car a powered control motor, TO: Brent Efford RE: Johnsonville Line LRT Concept Page 6 and the other an unpowered control trailer, and furnished with cabs on both ends of the pair. Single two-car sets provide base service, with a second set added to lengthen trains to four cars during peak periods. Trainsets are understood to operate trailer-first southbound. Two two-car trainsets are required for 30-minute base service. During peak periods, train lengths are understood to be doubled to two-car units, with three consists, or 6 Matangi trainsets, being required

The cars are partially high floor at 1100mm in the motor car, and partially low floor at 730mm in the trailer (some sources say 680mm), but since many of the station platforms are curved, it appears that platform gaps prevent the achievement of full level boarding for wheelchairs, buggies and shopping carts. An on-board ramp is provided for wheelchair access where needed. I was unable to determine whether manipulation of the on-board ramp results in any kind of schedule delay, and if so, whether that delay cascades to other trains due to the inherent nature of bi-directional working on a single track. I could not find a specification sheet for the Matangi equipment, but gleaning data from several sources, I believe the following to be reasonably correct:

• Nominal capacity of a two-car set is 147 seated, 130 standing – total 277.

• Car weight is 76,900kg; car width is 2.73m.

• The Matangi sets have AC propulsion. Nominal traction power line potential is 1500VDC. Power output is 680kwh.

The introduction of the Matangi sets on the Johnsonville line was accompanied by some reconstruction of platforms, the signal system, and other infrastructure.

4. An Alternative Future for the Johnsonville Line

Given many examples worldwide of redevelopment of traditional “commuter rail” into true “regional rail” services by routing trains diametrically through Central Business Districts, it seems likely that the market share in Greater Wellington could be put on a strong growth trajectory by operating of most of the regional rail service through the central city, and eventually beyond. For many people this would provide direct access to all major central city destinations and trip generators within the 900m catchment area criterion of station platforms sited in the Options Evaluation report. In addition, this would provide single-transfer connections with central city bus and trolleybus lines making it easy to reach many other destinations outside Central Wellington. By increasing the range of trip origin/destination pairs that can easily be made by public transport, a significant impediment to making public transport more useful in Wellington – or any metropolitan area for that matter - could be removed. The termination of “commuter” rail lines on the edge of the CBD, with a transfer required just to complete a basic journey to the center of town, and a service orientation to peak hours, is an antiquated concept based on historic precedents that are no longer valid. Public transport must compete for the public’s business, and to do so must be arranged to make the kinds of trips characteristic of the modern city, with dispersed trip patterns and non-traditional travel times, easy to make by bus and train. Underground construction to bring trains into the central city involves very high investment costs, prohibitively so where travel density is not high enough to TO: Brent Efford RE: Johnsonville Line LRT Concept Page 7 generate the benefits required to justify investment costs. The -train concept, basically converting suburban railway lines shared with mainline railway traffic into light rail lines and running them at grade through the central area, offers most of the benefits of rapid transit at lower cost, and can be made to work at passenger volumes as high as 10,000 passengers per peak hour per direction (pphd), as in Calgary, Alberta. It does not appear that the Wellington system, and certainly not the Johnsonville Line, would have a peak passenger density exceeding this threshold.

The Johnsonville Line itself would be a somewhat less challenging starter line than other liens in the Greater Wellington network, as it is a stand-alone service, and would not significantly display the shared-trackage feature characteristic of the tram-train concept. It would not share trackage with the mainline railway lines except for storage sidings and access to the maintenance depot just north of Wellington Station. Absent common operation with mainline trains, there might also be reduced concern for signal system upgrades. Consequently the Johnsonville Line could represent a simpler and cheaper laboratory whose experiences could inform future decisions if further network development were to be undertaken. Still, in making technical decisions about Johnsonville Line conversion, it would be desirable to give consideration to the implications of ultimate network requirements, since service to Kapiti, Hutt Valley points and Wairarapa, would seem to represent a significant opportunity for eventual upgrading into a regional tram-train concept. For this reason, in resolving technological issues of a Johnsonville Line light rail conversion, consideration should also be given to the design needs of an ultimate system.

To the greatest degree possible, the LRVs should be compatible with existing rail standards: 1067 mm gauge, New Zealand Railways wheel profile, ability to use 1500 VDC electrification on the mainline (but, possibly, trolley bus voltage in the central city to simplify overhead construction), platform heights, vehicle width, and so forth. In general, at conversion, wherever a significant design commitment is made, the design choice should allow for future tram-train compatibility with any current light rail development.

a. Assumed light rail vehicles ( some examples of many):

Various international car builders can provide a vehicle that could be operated on the surface through Central Wellington, then on to the Johnsonville Line. I am not offering any endorsement here or pointing out any favorites, but only illustrating a few vehicles in the available range that show what might be available for a combined street and railway application:

• “Citadis” (Attachment 4): multiple-section 100% low floor vehicle

• Bombardier “Flexity” (Attachment 5): Double-articulated partial low floor vehicle

• Stadler RBS Triplets (Attachment 6): Partial low floor three-car trainset

Here is an incomplete summary table comparing some of the chief characteristics of these vehicles, and the Matangi sets (from readily available information):

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 8

Vehicle Length/ Width Floor % Seats Standing Total Max Empty Makeup height- Low (nom) (nom) Cap. Power weight “level” Floor door opening (ATOR) Bombardier 37m/ 2.65m 400mm 50% 96 147 233 960kwh 55.4t “Flexity” 3modules Alstom 42m/ 2.65m 370mm 100% 98 153 251 900kwh 77t “Citadis” 4modules (42m) Alstom 52m/ 2.4m 370mm 100% 116 176 292 “Citadis” 5modules ? ? (52m) Stadler 60m/ 2.65m 400mm 50%? 154 192 346 1400kwh 77t RBS triplets Matangi 43m/ 2.73m 730mm? 50% 140 130 270? 680kwh? 76.0t set 2 cars

The data shown here are from specification sheets I was able to obtain from colleagues and internet sources, and I have not confirmed the information, or my understanding of it. There are also other carbuilders (Siemens and CAF are obvious examples, not to mention the Japanese builders) that could build a 1067mm gauge car for Wellington (the Stadler set is meter gauge, for example, but easily could be re-gauged). One critical technical feature would be minimum commercial turning radius, which would be critical in Central Wellington and which is not specified in the sources I came up with. Therefore I would not really draw any conclusions from this, and I am not qualified to make any such recommendation, save one – that there is a reasonable range of low-floor equipment available with power, weight and length characteristics in the range of the Matangi equipment that could be used as examples in a train performance simulation of the extended Johnsonville Line. Specific evaluation of technically sound alternatives is a job for qualified specialists.

b. Surface Running Time Wellington Station to Courtney Place

I strongly emphasize that this is all subject to confirming analysis using a complete and sophisticated operational modeling tool that takes vehicle operating characteristics and details of track alignment and train control system efficiency and signaling into account.

For purposes of looking at Tram-Train operation of a through line to Courtney Place, I am allowing 15 minutes for this 2.45km segment. I believe this is an overly generous guess. By way of comparison, Sacramento Regional Transit, for which I was formerly responsible, schedules light rail trains of more than 100 meters in length (four Siemens U2A cars) a comparable distance between the railroad station and 16th Street, 2.7 km, in 10 minutes outbound, 12 minutes inbound (2 minutes timetable “slack”) with six intermediate stations. Absent a simulation, I believe this 15 minutes assumption is a reasonable but conservative approximation.

c. Minimum Alternative Service Level Example

As noted above, it is the inter-loop running times that determine the minimum operable headway. TO: Brent Efford RE: Johnsonville Line LRT Concept Page 9

These running times as shown in the table above are pretty rough – I have based them on the published Johnsonville Line timetable, and from your video, plus some guesswork. Under the present suburban railway operation, an average 15-minute headway provided in one direction, presumably taken as the more important market in a given peak, is always accompanied by a 20 or 30-minute average headway in the opposite direction. The memory clock headway breaks down, and integration with local bus routes, were that to be attempted, would be difficult.

The ideal is a consistent, reliable clock headway in both directions. The maximum acceptable headway for a modern regional rail service can be taken to be 15 minutes. (It is apparent that the present railway service does not achieve this – and possible cannot, given constraints). Clearly, shorter clock headways that repeat themselves every hour, and therefore lend themselves to intermodal service integration (12 minutes, 10 minutes, 6 minutes etc.) are desirable, but given the civil constraints, the current level of study, and concerns expressed regarding initial capital costs for LRT in Wellington, it seems practical to ask, first of all, whether or not a 15-minute headway can be sustained on the Johnsonville Line without, as has been asserted elsewhere, major new investment in tunnels or passing loops (or double tracking).

When operating a bi-directional service with a uniform headway on a single track line, the critical factor at work is the basic physics of the situation – trains heading in opposite directions will “meet” each other every half-headway. If the headway is 30 minutes, passing loops must be located 15 minutes apart. If the headway is 15 minutes, the loops must be 7½ minutes running time from one another. The passing loops on the Johnsonville line reflect legacy locations, and are hemmed in by difficult topography and tunnels; they are not ideally located for a15-minute headway. However, though they are not ideally situated, I believe they can be made to work where they are. To do this, trains must either go faster, in order to cover a greater distance in the allotted 7 1/2 minutes, or else in places they can go more slowly than the current schedule for the Matangi sets.

Faster equipment may be possible, but in order to avoid conclusions based on overly optimistic assumptions, given the lack of any simulation data, I am assuming here that in order to avoid the significant additional capital investment which would arise from tunnel widening or other major civil work, the Johnsonville trains could go a little more slowly in order for a couple of critical inter-loop segments to be covered in the right amount of time. (There are remedies for this, and I speculate on them below). There is a tradeoff – the end-to-end one-way running time from Wellington Station to Johnsonville might be lengthened from 21 scheduled minutes to 28, but this would be offset by direct operation through the core of Central Wellington, with a time savings of perhaps ten minutes in comparison with a transfer to buses for the “last kilometer”. There would also be an offset benefit in the form of reliable 15-minute service in both directions (I suggest all day), and reduced perceived waiting time in comparison with lengthier average waits for trains that currently run less frequently almost all day long.

The following timetable fragment, using part of the morning peak purely as an example, shows what a through light rail service might look like, based on these assumptions.

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 10

Southbound (Read Down) STATION Northbound (Read Up) F E D C B A G F <- Trainset A B C D E F G A -> 748 733 718 703 648 633 618 603 Johnsonville 627 642 657 712 727 742 757 812 751 736 721 706 651 636 621 606 Raroa 625 640 655 710 725 740 755 810 753 738 723 708 653 638 623 608 Khandallah 623 638 653 708 723 738 753 808 754 739 724 709 654 639 624 609 Box Hill 621 636 651 706 721 736 751 806 756 741 726 711 656 641 626 611 Simla 619 634 649 704 719 734 749 804 Crescent 758 743 728 713 658 643 628 613 Awarua 617 632 647 702 717 732 747 802 Street 800 745 730 715 700 645 630 615 Ngaio 615 630 645 700 715 730 745 800 803 748 733 718 703 648 633 618 Crofton 612 627 642 657 712 727 742 757 Downs 807 752 737 722 707 652 637 622 (Wadestown 607 622 637 652 707 722 737 752 Loop) 814 759 744 729 714 659 644 629 Wellington 600 615 630 645 700 715 730 745 Station 816 801 746 731 716 701 646 631 Bowen 558 613 628 643 658 713 728 743 Street 818 803 748 733 718 703 648 633 Kirkaldie’s 556 611 626 641 656 711 726 741 820 805 750 735 720 705 650 635 Hunter 554 609 624 639 654 709 724 739 Street 822 807 752 737 722 707 652 637 Civic 552 607 622 637 652 707 722 737 Centre 824 809 754 739 724 709 654 639 Manners 550 605 620 635 650 705 720 735 Street 826 811 756 741 726 711 656 641 Cuba Street 548 603 618 633 648 703 718 733 829 814 759 744 729 714 659 644 Courtney 545 600 615 630 645 700 715 730 Place F E D C B A G F <- Trainset A B C D E F G A ->

This is a conservative timetable. It provides 6-minute schedule recovery at Johnsonville, with no double-tracking required at that terminal, and 16 minutes at Courtney place. It provides for meets all day at all current passing loops on the line north of Wellington Station, and it assumes a fairly slow operation through Central Wellington, with 16 minutes layover at the south end of the line – more than a headway. It makes no unusual demands on the performance of the light rail trains. Even with conservative performance assumptions, it appears that the service could be operated with only seven trains (trainsets A through G in the example above).

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 11

d. Capacity

It appears that capacity adequate to current actual demand could be provided by light rail trains. Current ridership figures demonstrate that ridership is well within world light rail standards. The tables below show representative on-board peak load point counts in passengers per peak hour direction (pphd) from last April and May. These would be mid-autumn counts that can be taken as typical of the loads the Johnsonville Line is called upon to accommodate:

MORNING SB TOWARD WELLINGTON STATION (2013) Hour Tues 30 April Mon 6 May Thurs 9 May Fri 10 May 600-700 79 81 128 118 700-800 483 762 833 808 800-900 519 429 481 374 900-1000 92 49 153 79

AFTERNOON/EVENING NB FROM WELLINGTON STATION (2013) Hour Mon 6 May Weds 8 May Thurs 9 May Fri 10 May 1500-1600 283 270 193 310 1600-1700 178 277 253 209 1700-1800 667 670 688 562 1800-1900 210 423 246 217

With the 15-minute headway suggested, delivered capacity would be in the range of 900-1200 pphd, even without operation. It appears that seven single-car light rail trains with could conceivably provide service superior to that now provided by three two-unit Matangi trains, totaling 6 units or 12 cars.

Providing for anomalies, or should the service through Central Wellington generate additional ridership, longer, two-unit light rail trains could be provided for certain “peak-of-the-peak” schedules, though this would mean procurement of a fleet increment of one unit for each such train. There are several ways to operate occasional longer trains without having to lengthen CBD platforms, but one obvious way would be to run two units northbound in the peak direction two minutes apart, combining them at Wellington Station; southbound in the morning, one or two two-car trains could be operated into Wellington Station, then split into two separate movements through Central Wellington. This can be accomplished in little more than a normal station dwell time by a well-trained light rail operations staff. Other strategies are possible. Platforms of sufficient length would be needed north of Wellington Station, but it is understood that most, if not all platforms have already been lengthened.

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 12

5. Conclusions and Recommendations

My comments are entirely preliminary, and are based on the data you have given me, and materials available from colleagues in the field and from open public sources. I think it is clear that light rail service is feasible on the Johnsonville Line, and I see no reason why it could not be routed on the surface through Central Wellington if there were the political will to do so. That is as far as I can go with this. To reach a firm and definitive conclusion, a comprehensive feasibility study would have to be performed to accurately work out the details. Such a study would be as necessary in order to correctly rule out the conversion of the Johnsonville Line to light rail, as it would be to find it entirely feasible. Though I may easily have overlooked it, it does not appear that such a study has been conducted as part of the Spine Study.

Obviously there would be significant capital investments required for light rail vehicles, the CBD trackage, stations and electrification, proof-of-payment fare collection equipment (I assume), maintenance depot equipment, and adjustments to platform heights at, and north of Wellington Station. The signal system would also have to be evaluated. These incremental investment costs, and differences in operating and maintenance costs between railway and light rail practice, would have to be summed up and weighed against the potential benefits of a more useful service. It does not appear that that has been done.

To summarize, my main points are these:

1. Pending confirmation by a complete and adequate feasibility study, it appears that LRT could serve the Johnsonville Line with a surface extension through Central Wellington.

2. It also appears that a 15-minute bi-directional headway could probably be operated with seven trains in operation between Johnsonville, Wellington Station and Courtney Place, without any need for new passing loops or doubled tunnels.

3. The Central Wellington track construction could be undertaken without any disruption to current Johnsonville service. Platform height or track level adjustments could probably be made on a phased basis with little or no disruption. No significant disruption is likely.

4. A complete and detailed demand analysis on a train-by-train basis should be undertaken to assess capacity needs and capacity delivery capabilities of specific light rail vehicle options.

5. There are several candidate vehicles from world carbuilders that might have the characteristics necessary to provide this service. If this is to be explored, then a thorough performance simulation should be performed that incorporates the performance characteristics of various candidate vehicles, operating under a variety of assumptions (speed, passenger load, signal system etc.) over the known civil characteristics of the Johnsonville Line.

6. An adequate operational simulation of the line should also be undertaken to confirm this “sketch” conclusion, and to answer other obvious questions:

TO: Brent Efford RE: Johnsonville Line LRT Concept Page 13

• Can faster operation by readily available technology, and a modest extension of double tracking north of Wellington Station, speed up the service and permit the 15-minuite headway to be operated by only 6 trains? Faster service and lower operating costs are an obvious benefit.

•The analysis above is based on a 15-minute headway, which is good service for a suburban line, but may be inadequate for local circulation in an urban center. Should/can additional shuttle trains be operated between Wellington Station and Courtney Place to provide additional short-headway circulation in Central Wellington and provide additional capacity for passengers transferring to and from other regional rail lines at Wellington Station? Would more elaborate terminal trackwork be necessary at both Wellington Station and Courtney Place? Would a second terminal track be required at Johnsonville?

•Can/should the operation of the train control system be improved? Would there be a benefit to investing in a more efficient signal system?

As part of the demand analysis, a network study should be undertaken to determine whether some buses entering Central Wellington from the north could be redeveloped into local connecting services with a Johnsonville light rail line, reducing overall public transport operating costs, and providing more comprehensive network mobility in Wellington’s northern suburbs.

0.0 km A B

0.4 km C

0.7 km D

E 1.0 km F

1.5 km G

1.85 km

H

2.45 km

I LRT tracks to depot, Johnsonville, other lines (tram-train scenario).

Light rail connection to WRS Option 1: remove spiral pedestrian ramp. Potential air rights development at Platforms accommodate 2 coupled 45m 1st oor level, over LRT platforms, Draft 1, Drawn BLE, 21/4/13. integrated with walkway. Escalators replace ramp.

Tracks on west side of the median in Lambton Quay. citadis dualis

Accessibility Modularity . stOM

AL Reliability Comfort Customization tyling. S esign& D and their frames are registered trademarks of trademarks registered are and their frames he information given in this document is indicative only and may be changed without warning. given in this document he information T DUALIS

ITADIS asso - Pagécran - asso - Pagécran S - C. A logo, C stOM OM T AL ansport / r T , the stOM AL OM ALST 2010 - stOM

AL A lstom Transport 48, rue Albert Dhalenne 93482 Saint-Ouen Cedex France Telephone: +33 (0)1 57 06 90 00 Combining the best /eng/06.10/FR - © pso Facto, Paris – Photo credits: Paris – Photo credits: pso Facto, I www.alstom.com/transport of trains and tramways ITADISDUALIS ns /C A esign and creation: esign and creation: he other brands mentioned, whether registered or not, are the property of their respective owners. of their respective the property or not, are mentioned, whether registered he other brands tR T D citadis dualis

Optimized access for all Passenger comfort Operating tools Modular Design With its full low floor and wide gangways, Citadis Dualis offers many options dedicated Citadis Dualis also offers anti-vandalism Citadis Dualis may be configured to the needs Citadis Dualis is the first tram-train vehicle certified to passenger comfort, depending on the profile and passenger-counting options to optimize public of each operator. according to STI PRM* standards. of the line served. Among the possibilities are toilets, transit operations for maximum vehicle availability. Urban or peri-urban configuration, number of doors more comfortable seats, window shades per side, interior spaces... the design and layout of your and umbrella racks... Citadis Dualis are modular to fit the type of itinerary. Citadis Dualis makes on-board comfort a priority.

* Excluding the toilet option 4 Degrees of freedom Capacities

Length Electric power supply 42 Meters and 52 Meters Dual voltage 750 Vdc / 25 KVac 50 Hz Or dual voltage 750 Vdc / 1500 Vdc Width Version Version Version Version 2.4 Meters and 2.65 Meters 2.65 M 2.65 m 2.40 m 2.40 m 4 cars 4 cars 4 cars 5 cars Interior Fittings 4 doors 5 doors 4 doors 6 doors

l Length: 42 m Seats + Tip-up seats CE2 18 + 2 18 + 2 16 + 2 16 + 2 l Width: 2.4 m C2 20 l 4 doors CE2B 32 32 32 40 l 93 seated places C1 24 + 2 20 22 + 3 20 l 141 standing places CE1 18 + 2 18 + 2 16 + 2 16 + 2

Total 92 + 6 88 + 4 86 + 7 112 + 4

l Length: 42 m l Width: 2.65 m Passenger in Wheelchair space 2 2 2 2 l 4 doors Bicycle rack 1 0 1 0 l 98 seated places Luggage rack 1 0 1 0 l 153 standing places

Total passengers 251 251 234 292 4 persons/m2 2 330 332 307 382 6 persons/m 410 414 381 472 l Length: 42 m 8 persons/m2 l Width: 2.65 m l 5 doors l 92 seated places l 159 standing places

CE1 C1 CE2B C2 CE2

l Length: 51.5 m l Width: 2.4 m l 6 doors l 116 seated places l 176 standing places

2.4 meters 2.65 meters Technical Specifications Options

Type Articulated – 4 to 5 car bodies Length (in m.) 42 42 42 52 4 car bodies: 42 m Length 5 car bodies : 52m Width (in m.) 2,65 2,65 2,4 2,4 Width 2.65 m (4 car bodies) or 2.40 m (4 or 5 car bodies) Height 3.5 m Number of cars 4 4 4 5 ns Floor height (4 pers/m² + seats folded, new wheels) o ti Number of doors per side 4 5 4 6 • Access (gap-filler) 370 mm • Corridor central part 405 mm op • Corridor above bogies 537 mm Luggage space l l l l

Multiple Unit Up to 3 units (MU3) with 4 car bodies Design Bicycle space l l l l

Structure • Compression 600 kN Retractable steps l l l l • Passive safety complies with standards EN12663 and EN15227 Maximum number of doors per side 5 5 5 7 Performance • Maximum speed 100 km/h • Maximum acceleration from start up 1,09 m/s2 from 0 to 40 km /h (base version) Toilets l

Fixed steps for the doors l l l l Braking Types Electric, Electrohydraulic and Electromagnetic • 2.8 m/s speed ≤ 70 km/h • Maximum Emergency braking deceleration 2.5 m/s speed > 70 km/h Supply and installation of passenger-counting system l l l l

ns l l Weight o Lengthwise umbrella rack • Tare weight ≈77 t (4- car vehicule) ti op

• Maximum load per axle (6 persons/m2 + seats extended ) 11.5 t

l Exterior livery l l l l

Traction • Type IGBT, 3-phase permanent magnet synchronous motor Trash receptacles l l l l • Continuous Motor power Motor 6 x 150 kW ntractua

Co Passenger window shade l l l l Redundant MVB network / WTB Train Control / Command Cabled commands for safety functions More comfortable seats l (cannot be combined with anti-vandalism option) Dual voltage 750 Vdc / 25 kVac 50 Hz Electric power supply Dual voltage 750 Vdc / 1500 Vdc Anti-vandalism l (cannot be combined with more comfortable seat option) 400 Vac /50 Hz Auxiliary power 24 Vdc Minimum curve radius 25 m

Capacity in EL4 (4 pas/m²) Fixed seats / Total • 4-car vehicle 86-92 / 234-251 • 5-car vehicle 112 / 292

Corridor width 600 mm (2.65 m version) Passenger access width Double sliding door / 1,300 mm passage Platform access Fill-gap

Passenger Information • Base Interior and exterior LED display screens + audio • Option TFT screens Coromf t options

Toilet Option* Lengthwise umbrella Rack Option The toilet option* is offered aboard the CITADIS DUALIS in a suburban configuration (2 m65 in width) in order to improve passenger CITADIS DUALIS can also be equipped with lengthwise overhead racks. These allow passengers to put small objects, umbrellas, comfort for trips that exceed 30 minutes. This is a unique option in a tramway-type LRV. Two variants are possible for this option: jackets... above their seats and also improve passenger comfort in freeing up space. l Factory-fitted toilets, l Preparation for toilet integration at a later time (for example in the case of a line extension) * Option non-certified STI PRM Trash Receptacle Option Trash receptacles can be installed as an option aboard the CITADIS DUALIS. They facilitate maintaining a high level of on-board cleanliness, improving both the trip’s ambiance and the cleaning work.

Passenger Window Shade Option Passengers are able to adjust sunlight to what’s comfortable.

More comfortable Seat Option (cannot be combined with the Anti-Vandalism Option) Improved seating and an added headrest offer passengers greater travel comfort. Technical options

Fixed Steps Option Individualized Exterior Livery Removal of the retractable steps (standard edition) in order to reduce passenger exchange times at stations. To facilitate its integration in the regional landscape, the livery of the CITADIS DUALIS can be individualized This option requires an appropriate infrastructure. with the colors of the region.

Anti-Vandalism Option CITADIS DUALIS’s anti-vandalism resistance can be reinforced with an interior and exterior film protection on the car.

Bicycle Space / Luggage Space (Option depends on the body car’ design)

Passenger Counting Option All vehicle accesses are equipped with a system that makes it possible to note passengers getting on and off. The number of passengers aboard is thus known at all times. This is an essential element for tram-train operations. Urban Transport

Bombardier FLEXITY Link Saarbrücken, Germany

The FLEXITY* Link tram-train concept was developed by Bombardier for the operator GmbH. By using dual-system technology the vehicles can operate on the inner city rail network as well as on the heavy rail infrastructure of the surrounding countryside. Meeting both the BoStrab and EBO regulations, the 28 FLEXITY Link vehicles can be put into service on all German railway networks.

Since 1997, passengers using the FLEXITY Link vehicles can travel from the outskirts of the city to its downtown district without changing the means of transport. Level entrances, air-conditioned passenger areas, special communication features and attractively designed interiors ensure an enjoyable travel experience.

FLEXITY Bombardier FLEXITY Link 3,900

2 650 13,250 10,600 13,250 37,000

1,300 1,300 1, 300 1,300

GENERAL DATA TECHNICAL CHARACTERISTICS Contract award 1995 • Nominal current supply: bi-system 750 Vdc & 15 kV 16 2/3 Hz Type of vehicle Bombardier FLEXITY Link • Energy recuperation Model bi-directional vehicle • Low voltage: 24 Vdc Owner Stadtbahn Saar GmbH • 120 kW motor power Quantity 28 • Eight 3-phase asynchronous motors Train consist 3 modules • Air-cooled motor • Four powered bogies DIMENSIONS AND WEIGHT Rubber/metal primary suspension Length of vehicle 37 m Coil spring secondary suspension Length over coupler faces 37.9 m 8 sanders Height 3.9 m Flange lubrication device Width 2.65 m Anti slip, anti skid system Floor height above TOR • Electrical service brake: regenerating motor - low-floor entrance 400 mm • Mechanical service brake: disc brakes - low-floor area 400 mm • Parking brake: disc brakes - above powered bogie 600 mm • Magnetic brake: 8 x 60 kN Percentage of low-floor area 50% Doors PERFORMANCE AND CAPACITY Electric double-sliding doors 4 per side Maximum speed 100 km/h - door clearance height 2,000 mm Maximum acceleration (2/3 load) 1.1 m/s2 - door clearance width 1,300 mm Deceleration Electric single-sliding door near cab 1 per side - maximum deceleration (service brake) (3/3 load) 1.6 m/s2 - door clearance height 2,000 mm - emergency braking rate 2.8 m/s2 - door clearance width 650 mm Maximum gradient 80‰ Aisle width 610 mm Seated passengers 96 Wheel diameter (new/worn) 660 / 580 mm Standing passengers (4 pass/m2) 147 Gauge 1,435 mm Bicycle, pram and wheelchair locations 2 Minimum horizontal curve radius 25 m Minimum vertical curve radius, crest 500 m Minimum vertical curve radius, sag 500 m Car weight (empty) 55.4 t Car weight (loaded) (4 persons/m2) 72.8 t Maximum axle load 108 kN Buffer load 600 kN

Bombardier Transportation Donaufelder Straße 73-79, A-1211 Vienna, Austria Telephone (43-1) 25 110-760

Australia • Austria • Belgium • Brazil • Canada • China • Czech Republic • Denmark • France • Germany • Hungary India • Italy • Mexico • Norway • Poland • Portugal • Spain • Sweden • Switzerland • United Kingdom • USA

www.bombardier.com

* Trademarks of Bombardier Inc. or its subsidiaries. © 2005, Bombardier Inc. or its subsidiaries. All rights reserved.Austria / 0791 lrv / 01-2005 EN Recycled paper / Printed in Low-fl oor railcar for Regionalverkehr Bern–Solothurn (RBS), Switzerland

RBS has purchased 6 three-part low-fl oor express railcars from Stadler

Bussnang AG. These trains are fully air-conditioned and have a design that facilitates access to all compartments throughout their entire 60-metre length. This accessibility improves the distribution of passengers within the train and make them feel safer. Passengers are provided with 157 seats in total, of which 18 are found in 1st class, plus 289 standing spaces. The trains Stadler Altenrhein AG Park Altenrhein für Industrie und Gewerbe are designed to reach a maximum speed of 120 kilometres per hour. The CH-9423 Altenrhein, Switzerland Phone +41 (0)71 858 41 41 new Stadler multiple-unit low-fl oor railcars are in service between Solo- Fax +41 (0)71 858 41 42 [email protected] thurn and Bern since 2009. This means a better service for passengers, as A Company of Stadler Rail Group they benefi t from more comfortable trains and, little by little, from an Ernst Stadler-Strasse 1 CH-9565 Bussnang, Switzerland increased number of trains running during peak travel periods. The railcars Phone +41 (0)71 626 21 20 Fax +41 (0)71 626 21 28 have been specially tailored to the needs of RBS and its passengers. [email protected]

www.stadlerrail.com Technical features Vehicle data

• Light passenger compartments thanks to large windows Customer Regionalverkehr Bern–Solothurn RBS, Switzerland • Clear, open interior design Lines operated Solothurn–Bern • Air-conditioned passenger and driver compartments Gauge 1000 mm • Modern passenger information system Centenary supply voltage 1250 V DC • Low-fl oor railcar as lightweight aluminium construction Axle arrangement 2’2’+Bo’Bo’+2’2’ • Six entrance doors on each side Number of vehicles 6 • Large entrance platforms Service start-up 2009 st • All passenger compartments accessible with wide gangway Seating capacity 1 class 18 Seating capacity 2nd class 136 • Air-suspended drive and trailer bogies Folding seats 3 • Ergonomically designed driver compartment Standing spaces (6 pers./m2) 289 • Vehicle control system with train bus and diagnostics computer Floor height Low-fl oor in entrance 400 mm High-fl oor 1000 mm Door width 1400 mm Longitudinal strength 600 kN Overall length 60 000 mm Vehicle width 2650 mm Vehicle height 3945 mm Tara weight 77 t Bogie wheelbase Power bogie 1900 mm Trailer bogie 1800 mm Power wheel diameter (new) 770 mm Trailer wheel diameter (new) 770 mm Max. output at wheel 1400 kW Starting tractive power 120 kN (up to 42 km/h) Maximum speed 120 km/h

SRBS0709e