So Sir Isaac Newton was right all along
FACT REPLACING FICTION
In motor racing the saying goes ‘when the flag drops the b*llsh*t stops’. And that applies for trains entering service. As the 2017 Golden Spanners rolling stock reliability review elsewhere in this issue shows, the cold numbers may not necessarily align with claimed failure rates. And now, courtesy of the Railway Performance Society, whose members have been out and about riding Great Western Railway’s first Class 800 bi-mode services, we can compare the actual performance of a bi-mode under diesel power against InterCity 125 schedules.
In one of the longest campaigns in ‘Informed Sources’ history I have been patiently explaining why the Department for Transport’s belief that distributed traction compensates for a shortage of installed power is delusional. This delusion was central to the original Inter-city Express Programme specification and despite me reducing the fallacy to simple O-Level (I think you mean GCSE old chap – Ed) physics formulae, it has refused to go away. Last year, while trying Virgin Trains East Coast’s ‘Azuma’ simulator, I mentioned to the trainers that the bi-mode might struggle on the climbs in Scotland. ‘No, it’s got distributed traction’, came the reply. So before looking at performance, time for a quick sense check. Two Class 800 bi-modes together weigh 494 tonnes and with the six engines at the 750hp (560kW) reliability rating have 4,500hp. A 2+8 IC125 weighs 408 tonnes and also has 4,500hp.
So which is going to have the better acceleration? At the height of the IEP heresy I asked my brother, a retired Professor of Mechanical Engineering, to run a simulation of IEP vs IC125 based purely on published acceleration curves. This showed the IEP, with its weird DfT-specified speed : tractive effort curve, romping away, only to be overtaken by the IC125 within just over two miles. GWR recently added fuel to the flames with a video showing a Class 800 out-dragging an IC125 from a standing start. What they didn’t show was the IC125 catching up seven miles later running at 115mph versus the Class 800’s 100mph.
But hang on, didn’t the Prof predict two miles? Well, yes, but he assumed what Cecil J would have called a ‘sparkling start’ by the IC125 driver while, out of habit, most IC125 station starts are still made in notch 1 or 2 to avoid the exhaust roar from the Paxman Valenta diesel engines (long replaced) disturbing passengers on the platform. But with a Notch 5 start, say from a signal stop, the IC125 is much sprightlier, catching up with the Class 800 within 2.5 miles with an 11mph advantage over the bi-mode’s 76mph. Nice one, bro.
So much for the distributed traction fallacy. Not surprisingly balancing speed is also lower, although analysis and modelling of their logs to date by Railway Performance Society members has so far failed to find any evidence of the Class 800’s MTU1600 engines being uprated from the 750hp (560kW) ‘IEP contractual reliability rating’, to the 940hp (750kW) commercial rating. Nor does analysis of performance with an engine out indicate that the automatic uprating of the other engines in the set to the commercial rating is happening. These are, of course, early days for the Class 800 fleet, with the first sets being bedded in. I would not be surprised if the engine management software was locked at the Reliability rating.
Mind you, according to ‘Informed Sources’, GWR is under pressure from DfT not to ‘unmuzzle’ the engines. As part of the £300 million deal with IEP owners Agility Trains West, the engine reliability regime with service contract is suspended for uprated power units. So if an engine fails, Hitachi is held harmless for any performance regime penalties during the rest of the diagram.
Both the RPS logs and observations during East Coast main line running reveal an interesting characteristic. Even with gravity assistance on a down grade, a Class 800 under diesel power can’t quite reach 125mph. A recent southbound run with a nine-car bi-mode could not get to 100mph on the climb from Stoke to Grantham and down Stoke Bank eventually managed to reach an indicated 123mph for a brief period before Peterborough. This reminded me of recent correspondence from an engineering chum who pointed out that the roof of Class 800 ‘is like a ploughed field’. The pocketed sliding doors, compared with the flush fitting doors of the Mk 3 coach, can’t help the aerodynamic drag either. But I seem to recall that the original IEP specification included improved aerodynamics.
While we’re on alternative facts, at the beginning of November I was invited to give evidence to the Parliamentary Welsh Affairs Committee on the cancellation of electrification between Cardiff and Swansea. My submission was a simple restatement of the advantages of electric traction and why bi-modes were a retrograde step. Hitachi also put in a written submission and, quite reasonably, put the counter argument that bi-modes are the future. However, some of the supporting claims made me pause. According to the Hitachi submission ‘A hybrid Class 800 is lighter than existing inter-city trains on the Great Western route’. Well, a 2+8 IC125 weighs 408 tonnes. A nine-car Class 800 bi-mode weighs 431 tonnes and the pair of five-car bi-modes replacing IC125 weigh 494 tonnes. Now, this may have been a reference to weight per seat, in which case a nine-car bi-mode is a bit better than a 40 year-old train and a five-car bi-mode a bit worse. So a draw.
But the more interesting, confusing and eventually enlightening claim read: ‘The wear per mile caused by a new train is 56% less than an HST (Class 800 £0.177 per mile compared to £0.4 for a HST)’. I checked with Hitachi and, as I suspected, it was quoting Network Rail’s Variable Track Access Charges, more properly Variable Usage Charges (VUC), per seat mile. Now I went through the issue of comparative VUC in excruciating detail back in several columns in 2011, concluding with an analysis of a DfT claim that IEP track access charges per seat would be 38% less than IC125. Using contemporary VUC rates for IC125 and provisional rates for IEP I could not get a 38% reduction. But after some intensive detective work (I think you mean ‘obsessive’ – Ed) I traced it to an Hitachi presentation to the 2011 National Rail Conference. This compared the nine-car bi-mode (25p per seat mile) and the nine-car electric (21p) with a 2+8 IC125 quoted at 40p. The IC125 figure was clearly wrong by a factor of two at 2011 VUC rates, but this did give a 38% saving for the bi-mode.
Time to update the comparison at current VUC prices. But when I downloaded the latest (November 2017) price list, the Class 800 was not there. So I asked Network Rail to explain the absence of trains already running around the network and after some to and fro with the press office my persistence resulted in chapter and verse from the VUC engine room.
Normally multiple-units have separate VUC for motored and trailer vehicles. The five-car Class 800/0 and 800/2 are straightforward with two driving trailer vehicles and three motor vehicles. However, the Class 800/1 and 800/3 nine-car units add two intermediate trailer vehicles to the two driving trailers and five motored vehicles.
Now the original IEP design had conventional H-frame bogies for the motor vehicles but inside-frame bogies for the driving and intermediate trailer cars. Inside frame bogies not only weigh less, the shorter axles reduce the unsprung mass that is a factor in track wear. Both of these advantages help reduce the suspension’s primary yaw stiffness, which is also good news for the track wear.
However, in the final design, Hitachi replaced the in-board bearing bogies on the driving trailer vehicles with conventional H-frame bogies. The driving cars each carry a pantograph and transformer and weigh about the same as a motored car. Switching to the H-frame increased the mass and primary yaw stiffness. But the in-board bearing bogies were retained on the intermediate trailer vehicles.
Now the VUC levied on vehicles reflects the wear and tear they cause to the infrastructure, which depends on speed, axle load, unsprung mass and primary yaw stiffness. So, to recapitulate, the five-cars have two different types of vehicle: motor and trailer vehicles, both with conventional bogies. But the nine-cars have three types: motor vehicles, plus trailer vehicles with two radically different bogie designs.
This mix of bogie designs combined with the different proportion of trailer vehicles in the two lengths causes a problem when estimating an appropriate single weighted average VUC rate for the trailer vehicle based on the typical train set formation. If the VUC for the trailer vehicles was based on an average for the conventional and in-board bearing cars then the ‘800/0’ and ‘800/2’ units would be undercharged for wear and tear per mile. Contrarily, if the VTAC for the trailer vehicles was based on the H-frame bogie parameters then the ‘800/1’ and ‘800/3’ units would be overcharged which, Network Rail points out, would remove the incentive for operators to procure more ‘track friendly’ vehicles.
Clearly Network Rail needs to ensure that the IEPs pay a VUC which reflects track wear. But this is likely to require replacing the current calculation of ‘average’ rates for motor and trailer cars with different rates for the different variants of motor and trailer vehicles.
This seems obvious but it would establish a new precedent for setting charges in CP5, which ORR would have to be content with. Resolving this new precedent is why the VUC rate for the Class 800s has taken longer than normal to approve. So when the VTAC price list is next updated I will recalculate the comparison of the Class 800 with IC125 and, for a bit of fun, even a spanner-winning Greater Anglia loco-hauled set.
TABLE 4: COMPARATIVE PERFORMANCE
Thanks to informed sources everywhere for your support and encouragement during the past year, without which this column would have to be renamed. I hope you enjoy a happy and peaceful Christmas and come back refreshed and ready for what promises to be a turbulent new year.