Is VLR a good fit for Stoke?
Disclaimer: Betteridge’s Law of Headlines applies to this article.
Talk has been made recently of whether the Coventry Monorail, no, Gadgetbahn wait, Very Light Rail (VLR) system currently in the works could be a good fit for Stoke-on-Trent. Indeed, a preliminary study was even included in the appendix of Stoke’s latest Transport Planning Document.
To be clear, I don’t want to wail too hard on the new sense of energy in Stoke - the drive to get steel rails in the ground and invest in transport infrastructure. This is undoubtedly good. Props to the councillors for even thinking about this stuff. I however have a number of reservations, both with VLR as a product and more specifically in its application to Stoke.
Let’s get something out the way - VLR is a small, battery-powered tram/streetcar, as defined by Gareth Dennis’ categorisation flowchart. It is a proprietary, single-vendor system with no options for things like tram-train operations. It has some innovative features (mostly to do with the track-form rather than the vehicle), but it’s not a new concept, much less a new mode. It’s a battery-powered tram. Don’t let anyone tell you different.
More akin to a historic streetcar than a modern tram, VLR is slower than conventional trams (40mph vs. 60-80mph) and offers the capacity of a small bus with none of the route flexibility. The trade-off is that it claims to be cheaper than conventional trams. It does this through a “lightweight” battery-powered vehicle, which reduces ground pressure and allows the innovative track-form to be laid directly over utilities. This reduces build cost by minimising civils and utility works at build-time. Battery power eliminates the overhead lines that would normally supply electricity to the vehicle.
This is fine. But we must consider it in that light (a small streetcar with none of the flexibility of buses), and not get carried away by marketing hype just because someone else has something new and shiny. Is a small battery-powered tram the right fit for Stoke?
Performance
Acceleration
Urban transit revolves around acceleration (as well as zero-emissions). On a dense urban route with many stops, a tram will spend as much time accelerating/decelerating as it does cruising at top-speed. It may also spend time at lights and junctions when running on-street rather than in a dedicated tramway. Consequently, the best way to minimise journey times is through brisk acceleration. Top speed is almost irrelevant because the vehicle spends very little time there.
This is why electric traction is favoured over diesel. To give a rail example, small diesel multiples (e.g. Class 171) will manage an acceleration of ~0.5m/s2 . A comparable electric multiple for urban patterns such as the Class 777 or Class 399 will manage 1-1.3m/s2 thanks to the superior torque characteristics of electric motors. Bringing your 0-60 time from 2 minutes down to 50seconds provides a profound improvement to journey times, as well as timetable reliability - timetables will not assume full performance, and having a bit of acceleration in reserve gives the driver the ability to “put their foot down” if they are delayed and make up time.
Fundamentally, it is desirable for acceleration to be limited only by passenger comfort, not by the inherent performance of the vehicle. This is easy when a tram is powered by a 750V overhead line, which can provide many Watts. In the case of VLR, acceleration is likely to be restricted by the maximum discharge rate of the battery. Although performance figures are not available, with a top speed of just 40mph, it is going to be relatively slow.
There is also a question of VLR’s ability to handle hills - Coventry is relatively flat compared with Stoke (in the foothills of the Peak District). Will the acceleration profile die off when faced with an uphill run? What of battery range? The advertised range is 20km - perfectly adequate in theory, but how is that range affected by terrain?
Top Speed
Having said that top speed doesn’t matter, it is important to note that in some cases it does. Coventry is a nuclear city with a defined centre, and the intention for VLR is to run on-street in mixed traffic (i.e. with cars, bikes and vans). But in a low density, poly-centric city like Stoke-on-Trent there may exist longer drags between calling points. Old rights of way are also available to provide dedicated tramways, where trams could safely achieve 50+mph (which is less common for on-street running). Indeed it is generally preferable to have dedicated tramways which are not affected by congestion and traffic - particularly that suffered by Stoke-on-Trent when the M6 has an incident. There is no point investing millions putting steel rails down if the VLR pods get stuck in car traffic. This is not a problem Coventry faces as it is bypassed effectively by motorways - inward traffic is typically visiting Coventry and may be modal-shifted to VLR. But Stoke experiences huge through-traffic on the A50 to the East Midlands. VLR or Trams need to offer an off-road, traffic-resistant travel option for locals when the roads snarl up.
Capacity
There seems to be some confusion over whether the VLR pod can carry 40, 50 or 70 passengers. In any case, this is much smaller than a conventional tram which can carry 200-300passengers, and comparable with a bus. During peak periods, VLR will require extremely high frequency service (every 3-5 minutes, compared with 10-minute service of the West Midlands Metro) in order to get a sensible throughput of passengers-per-hour.
If we intend to drive modal shift, then VLR needs to provide more capacity than the buses it replaces on high-intensity routes.
Cost
Urban VLR is pitched as a cheap way of deploying tram-like system, using its shallow track-form to reduce civils, eliminating overhead wires and using small, cheap vehicles.
This however seems too good to be true, and on closer inspection it feels there is a bit of salesmanship going on here. There seems at times to be a sly attempt to reduce “sticker shock”, but at the cost of loading on operating costs down the line, which may well outstrip the initial “savings” and affect the long-term viability of the system. We’re trading reduced CapEx for a lifetime of increased OpEx.
Track maintenance - If your plan for not moving utilities involves having a team on standby to lift sections of track out so that gas/electric/water/sewerage/telecoms can access their cables, then this is an expense the VLR is bearing forever in return for saving the one-off cost of moving utilities during the build. We’re trading one-off Capital Expenditure for heightened Operational Expenditure for the life of the VLR (which should be >100years).
Fleet size - Small, 40pax vehicles will require a larger fleet to meet peak demand (measured in passengers per hour, not trams per hour). Developing a cheap vehicle is of limited benefit if you have to buy 2-3 times more of them.
Fleet maintenance - Having bought 2-3x more vehicles, maintenance on the VLR vehicles will need to be cheap - 30-50% of the per-vehicle cost compared to a conventional tram in order to keep the overall fleet maintenance cost the same. Vehicles of this nature are expected to serve for at least 30years, but many run longer. Munich’s “Class A” U-Bahn stock was built in the 1970s. It is likely these vehicles will require at least two lifetime battery replacements. It is unclear whether the lifetime costs (TCO - Total Cost of Ownership) will be any cheaper than conventional trams.
Staffing - Use of a small vehicle with 40-50pax inherently means the driver-passenger ratio is 1:40 instead of 1:200. This makes the system very staff intensive, both in terms of drivers and conductors/revenue officers (I assume on such a small vehicle that revenue-operators would randomly inspect and float between vehicles, rather than having a staff member riding on every vehicle as with the West Midlands Metro).
Mitigations to criticism
The developers of VLR claim that their system will eventually be able to operate autonomously, which allows them to run frequent services with short headways (as often as every 2-3 minutes) without excessive staffing. In the interim, “virtual coupling” or platooning will allow one driver to lead two or three pods during peak hours. This will reduce staffing overheads.
In the first instance, platooning doesn’t exist. If it did, it would leave various questions like “what if the middle of 3 cars has a battery failure” (or just a failure of the wireless control channel). In a mechanically-coupled train, it could be put into neutral and towed by the rest of the train. With virtual coupling, it’s now just blocking the line.
The developers have offered no suggestion as to how they would safely achieve Unattended Operation (GoA4) without infrastructure elements like platform-edge doors to manage the platform-tram interface. Autonomy is best considered to be vapourware and marketing until it is actually demonstrated successfully. My money would bet that autonomy is eventually abandoned and they end up connecting 3 pods together with miniaturised Jacobs bogies to get a single vehicle that can accommodate 170+ people with one driver - i.e. a tram.
VLR in Stoke
Stoke’s Plan
Stoke council have tentatively proposed a VLR network comprising three lines - North, South and Central. These would feed into a mixed bus/VLR park-and-ride network. Their plan is shown below (from the Strategy Document)
Several things jump out.
The Southern loop appears to use the A34 (Newcastle/Stone Road). The A34 is heavily congested, and clogs up terribly when the M6 has an issue and people come off heading for J14 at Stafford. Now it is possible they intend to provide a dedicated tramway (possibly at the expense of un-dualling the A34). But if there is any suggestion of on-street running then this is madness - VLR/Tram should be resistant to those congestion events. They need dedicated tramways. Because although modal shift will reduce the number of local cars on those routes, will do nothing to reduce the tens of thousands of vehicles that clog the region during motorway closures.
Newcastle-Under-Lyme is basically disconnected. A patient travelling to RSU Hospital from the east of the city (Tunstall, Hanley) would be dragged down to Trentham on the South (Pink) line before travelling up. Why on earth is there not simple east-west link from Newcastle to the main station (also of value to students and Keele visitors, as well as NUL residents wanting to access national rail services)?
Why doesn’t the northern line spur down to Longport Station and into Newcastle through Wolstanton? I appreciate this involves the participation of NUL Council and falls outside SoTCC’s direct control. But an “aspirational” map with some dotted lines of “what would be best for the residents” would be nice though - something to push NUL council with to get Stoke and Keele better connected.
The VLR doesn’t meet half of the Park-and-Ride stations (Longport Station, Sideway).
The P&R stations are relatively central. Having battled as far as Sideway or Etruria, there will be little incentive to transition to VLR or buses. For access to Hanley maybe - since the council intends to increase the price of parking and discourage visiting by car. But this doesn’t help commuters avoid traffic from incidents on the M6 or A500. Having spent a fair amount of time parked on the A50 due to an M6 incident, I’d need the P&R site to be more like Blythe Bridge or Hanchurch/Junction 15 - and then ride in from the actual edge of town - by the time I got to Longton I was either at work or most of the way there! This is actually one of the mistakes committed by Beeching - he thought closing branch lines would simply see people drive to their nearest station and then ride the train. But actually once they were in their car (with the expense that entails) it was easier and cheaper to just drive the whole way.
No discussion is made of intermediate stations for the Stoke-Leek line. Presumably it is not the intention that shoppers travelling to Hanley or students to Burslem Campus should go down to the station and then back up on the VLR. Interchanges at (say) Bucknall Road are vital. Although a conventional tram from Hanley using the Leek line in a tram-train configuration would offer significantly improved service flexibility and frequency.
Conclusions
I’m afraid I do not believe VLR to be a good fit for Stoke. Coventry is a nuclear city 6 miles across, but with a much smaller core. A small hub-and-spoke VLR system offering high-frequency services could fit well. By contrast, although Stoke is not that much bigger, the city has no core - it is poly-centric with many disparate destinations - the six towns, Newcastle Under Lyme, as well as a need to reach out to Keele University, Blythe Bridge and up to Leek.
In some respects, this makes VLR very attractive - small, cheap lines running between cores, rather than a handful of big spines running through the city. However, Stoke needs its backbone services. RSU Hospital for instance is not central - it needs to be as easily accessible from Wetley Rocks as it is from Newcastle Under Lyme - not 90minutes and three changes away on the VLR!
Some of the backbone work will be handled with rail. But here is the big “but” - VLR cannot interoperate with rail. VLR in Stoke means two disparate systems - local rail and VLR. There is no prospect of merging the two as is possible with tram-train operations (as I have proposed to Leek). We cannot take the best of both worlds, or leverage existing rail corridors to get VLR vehicles from one place to the next as we could with trams.
Further, many of the advantages of VLR’s innovative track-form have limited value in Stoke. The track can achieve a 15metre bend radius - very impressive given that conventional trams usually stop at 28metres. But Stoke is low-density and would rarely use that sort of tight radius. Use of existing ways such as the former Potteries Loop Line would provide an off-street broad-radius alignment, suitable for faster running than could be achieved by VLR or on-street trams.
Worse yet, in the announcement of the Transport consultation, Councillor Dan Jellyman states:
“The VLR network is a key part to this strategy. I believe this should have been done 15 years ago when Manchester and Birmingham were doing it, so I am very keen not to waste any more time.”
Councillor Jellyman has seemingly misunderstood the difference between VLR and conventional trams. Manchester and Birmingham do not have VLR Networks - they have conventional tram networks. Whilst neither has joined Sheffield in running any tram-train segments yet, the potential exists for them to do so - because the tram system can support tram-train rolling stock. If Stoke were to commit to VLR, it would lock itself out of tram-train operations. This would be impossible to retrofit later - heavier conventional trams could not operate on the VLR track without largely rebuilding it. And there’s the rub. If you’re going to go to all this effort and get in the position where you’re talking about hundred-million-pound figures with Midlands Connect and the Treasury… then just build a tram.
I hope VLR works really well for Coventry. I just don’t want to see Stoke run down a technological dead end because someone else has a new-shiny monorail that promises the world. VLR is a small, slow, low-capacity tram. Maybe that’s right for some places. I just think it might handcuff Stoke down the line, given the city’s polycentric layout and the need to reach more distant destinations like Leek and Keele University. And if it were just the vehicles, I’d be more willing to try - but by doing half-a-job on the track and precluding heavier (proven) vehicles from Stadler or CAF if the VLR cars don’t work out… that’s a problem.
I’ve also seen it suggested that VLR might fit parts of Stoke and we could build proper trams elsewhere. This is of course possible, but then you saddle yourself with additional rolling-stock types - there’s a non-trivial maintenance and driver-training overhead for running two small, mutually-exclusive fleets instead of one larger one (and you’re stuck with that cost forever).
So it concerns me deeply if Stoke’s councillors are looking at Manchester and Birmingham (as they should) and concluding that Stoke needs trams (which it does) but concluding that they should procure VLR because it’s “cheap”. If they end up buying into a proprietary system with limited scope for expansion and interoperability for the sake of avoiding “sticker shock” then they may be selling the people of Stoke short. VLR may leave them committed to a high-maintenance, high OpEx platform which doesn’t achieve what they need it to. And that would be a very sad thing to happen.
For more discussion on Urban VLR, see Gareth Dennis’ VLR Q&A.