Return to Archive -by date - by topic - Archive 2002.
Southern Region's power supply, optimised for English Electric traction equipment, will find Desiros and Electrostars heavy going
If you want an example of vertical integration in action consider the Southern Railway and its successors. The 750 Volt Direct Current traction power supply and the trains it supplies, developed in parallel on either side of the contact rail/shoegear interface.
From 1936 when English Electric was awarded the first of three 10 year exclusive contracts to supply traction equipment, the camshaft controller and EE507 traction motor equipped all commuter stock. The power supply system was sized and configured to match the current drawn by hundreds of 507s notching up and down.
Power from the grid is fed into the Southern's electrical network by the Regional Electricity Companies (REC) at around 30 grid substations.
Between the grid sub-station and the trackside, Southern Region operates its own mini-national grid from seven control rooms. This is one of the Region's strengths, because its allows power to be switched around within the system, giving the traction power supply considerable resilience in the event of outages or problems with the incoming supply.
This ‘grid' of high voltage cables, mostly at 33,000 volts (33kV), although there are sections supplied at higher and lower voltages, runs alongside the railway. At roughly 3 mile intervals each cable feeds track-side substations which transform the incoming AC to 750V and rectify it to provide the DC which is supplied to the third rail.
Each sub-station feeds contact rails on each side through Direct Current Circuit Breakers (DCCB). At the mid point between adjacency substations a Track Paralleling (TP) Hut containing a DDCB which connects the contact rails to smooth the load as trains pass along from sub-station to sub-station.
When designing a traction power supply system the electrical engineers ask the traction engineers, how many amps do their trains want and how does this current vary over the speed range?
In Figure 1, the blue line with the saw tooth peaks shows the current taken by a Class 4XX 4 car unit at the 675V design voltage at the pick up shoe. For techie readers, the saw teeth are where the camshaft progressively cuts out resistance and the two steep rises are where the motor fields are weakened before the traction motor current follow the characteristic series-motor curve.
Peak current for a four car is about 2,300A as weak field is taken. The power supply system has evolved to feed an eight car railway of Class 4xx stock.
When the Class 365s arrived with three phase drive, instead of the EE507 motor doing what comes naturally, we had microprocessor controlled drives where the electronics could make the motors display all sorts of exotic characteristics.
Networker coincided with the Channel Tunnel and on the ‘boat train' routes the Southern power supply was strengthened to allow Eurostar to draw 6,800A. It was also beefed up on the Networker routes in Kent to allow 10 car operation.
Even so, the power Eurostar and Networker could draw was restricted. In the case of Networker, this meant permanently inhibiting the high power setting. Anyway, as the economy turned down, privatisation loomed and Network Southeast was fragmented, the power supply strengthening programme for Networker sort of fizzled out, unfinished.
British Rail got a lot of electrification for its money because it built a utility railway. Even on the later extensions of the 750V DC network, current supply was limited. Bournemouth Weymouth was equipped for five car units.
But now we have the second generation three phase drives in Juniper (Class 458/SWT and 460/Gatwick Express), Electrostar (Class 375 Connex South Eastern/South Central) and Desiro (Class 350/SWT)
With the Class 460 fleet of only eight units running an in electrically ‘strong' area, Southern Region asked Alstom to stay within the current demand profile of the existing Gatwick Express DMLV/EDL combination. This was done by tweaking the characteristics of the Onix drive – something you can do with computer control.
Less said about the Class 458, the better. Suffice to say, no one knows what the fleet demand might be on the power supply.
Which brings us to the Class 375. Southern Region has had a long debate with maker Adtranz, on the characteristics of the MACS IKON three phase drive. It has four current settings, 1500, 1800, 2000 and 2250A. Currently it is limited to 1,500A.
On Fig 1 the red line shows the current demand of a four car Class 375 at this setting. The good news is that it doesn't have peaks and is less than the camshaft controlled 4xx four car between 30 and 90km/h. The bad news is that it is higher at low speed. Worse, the flat characteristic continues up to maximum speed when the current is twice the 4xx at 160km/h.
What does this mean for the power supply? At 100% of the nominal rating the output of a track side sub station is, typically, 3,300A. At 750Volts that equates to 2.5MW (think Deltic).
But electrical equipment can run at substantial overload for short periods. The limiting factor is the temperature build up at the higher ratings
A typical Southern sub station, for example, can operate at 150% of the nominal output (5000A) for two hours, which is handy in the peak. And if things get really dire you can draw 10,000A for 2min.
So there is short term capacity to handle lots of trains pulling away at once.
But, while the transformer/rectifiers can deliver extra amps, there are also the circuit breakers. There are still over 1000 DCCB dating back to the original Southern Railway electrifications which have a maximum continuous rating of 2 500A. More recent DCCBs can handle up to 4000A, but the upgrading work for the Channel Tunnel confirmed that high sustained currents are not good for this equipment.
While the Class 460 fleet could be managed, the large number of Class 375s is a different matter. So Southern Region has been working with Bombardier for the past year to evaluate the effect of the new traction equipment.
Figure 2 shows the comparison between a Class 4xx and a Class 375 in terms of the traction current in the feeder cable between the substation and the contact rail. What counts is not the maximum value but the area under the graph which represents the energy used.
So while on starting the Class 4xx draws more current than the Class 375, it is soon in weak field and the current drawn falls rapidly. But the Class 375 is almost a constant current machine.
Now current in cables and electrical equipment generates heat – the more current, the more heat. And here were we come to the advantages of a vertically integrated railway.
As a train passes a substation it draws current, and the electrical equipment and cabling heats up. This is not a problem because the system then cools down in the pauses between trains. Since power supplies and the Class 4xx have developed together over 50 years or so, heating and cooling are complementary and keep the sub-stations in thermal equilibrium (Figure 3.
But when a Class 375 comes by, it draws more current, so the temperature reached is higher. The equipment cools down at the same rate, but when the next Class 375 comes by it is still warmer than it was.
Then, the combination of still warm equipment and more current means that the substation reaches a higher peak temperature after the next train which means it is even warmer when the train after that arrives. Repeat this cycle and the substation and cabling continues to get hotter until protective thermal cut outs operate and shut off the current.
Reducing headways accelerates the temperature build up as figure 3 shows in exaggerated form. In the trade this is known as ‘thermal runaway' (Figure 4).
Clearly, there is a major problem here. Or, rather, two problems.
First, there are all those Class 4xx trains which have to be withdrawn by 31 December 2004 (and cup and coned by the end of 2002) by law. This depends on over 1000 Electrostar and Desiro EMU vehicles entering service smoothly over the next three years. If the power supply can't cope it could be embarrassing.
Longer term, the TOCs, not to mention the Strategic Rail Authority, have aspirations to run more and longer trains. Given that the new trains are already heavier than Mk1 stock, more, longer and heavier trains mean even more Amps to be supplied.
Fortunately, electrical circuits lend themselves to computer modelling and, in conjunction with the TOCs and their train suppliers, Southern has been modelling the traction power supply network. BR's OSLO simulation is being used for the DC system while DINIS, developed by the RECs is used for the HVAC network.
Early on in these studies another factor emerged. So far we have been worrying about Amps, but Volts are also important. A Class 375 is designed to operate with a constant power characteristic. But on a low voltage DC supply supplying high currents, the nominal voltage (675V at the train) can vary up and down depending on demand.
If the voltage increases above nominal, a ‘dumb' Class 4xx gets more power and accelerates more briskly. The smart Class 375, adjusts the drive and runs as normal.
But when the voltage falls below nominal, both trains lose performance, so the Class 375 loses on both the swings and the roundabouts. Bombardier is looking at ways of tweaking the traction package to get round this problem.
Whatever: the studies have shown that the present system will not be able to meet the power needs of either the short term like-for-like Mk 1 replacement stock or the long term aspirations in the replacement franchises.
Before Mk 1 stock can be replaced the AC and DC networks will have to be upgraded across the Region. Southern reckons that another six months work will be needed to define the scale of the up-grade and all that can be said at present is that it will cost many tens of millions of pounds. Since you don't need many 10s to make a hundred I expect the cost to be north of £100million.
Starting at the supply end, more grid sub-stations will be required, plus new transformers and new HVAC distribution cables. At the trackside new rectifiers and DCCBs will be required. Currently there are 630 sub-stations and track paralleling huts.
One way of beefing up the DC power supply, used on the Channel tunnel routes, is to install rectifiers in the TP huts which already incorporate DCCB. But them of course, you have to strengthen the HVAC system to feed the extra rectifiers .
Sorry that this is a bit vague. Southern Region knows it has a problem with the Class 375, but the impact of the Class 350 Desiro has yet to be determined, because Siemens has only just provided the performance characteristics needed for the train to be ‘run' in the OSLO model.
Given that Siemens are noted for a max power approach to EMU traction, my best guess is that the effect on the network will be slightly more demanding than the Class 375. Stagecoach intend to run Desiro at 2,000A
‘We are looking at a substantial upgrade programme, which is not yet defined, which has no funding attached to it and which is essential to the Mk1 replacement programme over the next three years, and yet it will take us more than three years to do even if we get the funding'.
Southern Region Director Michael Holden
So, a huge problem which has slipped in under everybody's radar. And, as ever, the key issue will not be producing the equipment, which probably could be procured in time, but finding the manpower resources and possessions to install it on the working railway.
But surely, it is not beyond the wit of Sven or Heinrich the demon programmers to get the software of Electrostars and Desiros to emulate the characteristics of the English Electric camshaft controlled equipment?
Well sort of. But these high tech modern trains are heavier than the Class 4xx and need power for things like air conditioning. Emulating the current characteristics of the older train would mean having less power available which, with the extra mass to accelerate, would mean that they might not be able to maintain sectional timings. This would wreck peak timetables.
Noting the SRA's record in managing Mk1 stock replacement, you will not be surprised to learn that when the results of these studies became known, Chief Executive Mike Grant wrote a characteristically helpful letter to Railtrack's Administrators.
Given the Mk 1 stock replacement deadline Mike found it ‘staggering' that Railtrack had not started to plan for upgrading the traction power supply. He was also surprised that Railtrack did not seek to raise the matter with the SRA until early summer 2001, given Railtrack's ‘detailed involvement' in the franchise replacement process for South Central and South West and the length of time since the new generation EMUs were ordered.
I intend to set up a Steering Group consisting of the train operators concerned together with Railtrack and the rolling stock manufacture4rs to initiate urgent action to resolve this issue
Chief Executive Strategic Rail Authority
Nor could Rail Regulator Tom Winsor keep out of the issue. Illustrating his conclusions on the complaint from Adtranz and Alstom on train acceptance he wrote ‘Connex South Eastern is experiencing delays in bringing its new class 375 electric trains, built by Bombardier, into service… because it only emerged late in the acceptance process that the power supply capacity of the ‘third rail' was not adequate to supply the additional power requirements of the new trains'.
Well, for once I am on Railtrack's side here. Adtranz couldn't tell Railtrack what the characteristics of the Class 375 were until the end of 2000. And modelling started as soon as the details were available. A year on and Siemens is only just starting to supply details of the Desiros.
So how, Mike and Tom, could Southern Region have planned the upgrade in advance? After all, it was not obvious that the new high tech trains were going to be power-hungry lard-butts. And Railtrack did brief all potential tenderers that DC power supplies would be a major issue.
In effect Railtrack has been told, ‘here are the new trains, provide whatever power they need'. According to one cheesed off Railtrack Informed Source, if the TOCs had supplied full details of the new trains' power supply requirements under the established vehicle change process (Tom Winsor please note) when orders were placed, Railtrack could have started modelling much earlier and identified the necessary upgrades and the timescale for implementation.
And of course, we are still only talking short term. It is no use implementing an expensive upgrade to accommodate Class 375s and 350s, if the SRA then replaces the Thameslink franchise on the basis of extra services, requiring even more amps (thanks for the Strategic Plan Sir Alastair).
In fact, ‘cheesed off' is an understatement. Those at the sharp end of ‘Railtrack in administration' who want to run a railway not a political slanging match are in no mood to take flak from idle bystanders.
As a chum remarked to me, ‘it is quite clear where the shortfall in planning and leadership has been'. And he is not exactly impressed by the idea of yet another steering group with mutually conflicting commercial objectives.
So there we are then. You thought Mk 1 replacement was as bad as it gets? No chance.