GWR Class 800

[Mag_seven]

It always takes time for new stock to "bed in" so to speak. Reliability should level off at a high level eventually.

 

[cactustwirly]

But what figure are you using for fleet size? That isn't yet a fixed number as deliveries of 802s is still in progress, and the last 800 was only delivered a couple of weeks ago... Like I said you're taking the published figures - which I agree don't paint the IETs in a great light for reliability - and reading in something that isn't there.

As @TractiveEffort states several units are managing in excess of 900 mile diagrams without issue; and there is (now) becoming a surplus of 80x stock while we wait for the New TT to be introduced - units that are sat spare on NP or SG are going to be contributing zero miles to those run; while still be counted towards fleet size.

Fleet miles run isn't a sensible reliability metric.

Interesting there are spare 80x units, as a few Bristol diagrams were worked by HSTs this week, when they could be covered by 800s.
I'm guessing GWR is getting the most out of their HSTs, as the leading costs are much less than the 80xs

 

[JN114]

Interesting there are spare 80x units, as a few Bristol diagrams were worked by HSTs this week, when they could be covered by 800s.
I'm guessing GWR is getting the most out of their HSTs, as the leading costs are much less than the 80xs

There’s far more IETs welsh side of Severn Tunnel block than are needed for day to day service; as just one suggestion why HSTs seem to be more prevelant in England at present.

 

[jimm]

The published data include the total miles run by the fleet in any four week period. Also included is the number of units in the fleet in that four week period and the number of TINs.

Dividing the total miles by the number of TINs gives the MTIN. If I do this I obtain the same value for the MTIN as Modern Railways reports.

So total fleet miles divided by fleet size divided by 28 gives the miles each unit is doing per day. Unless somebody is feeding MR with porkies the arithmetic is correct and this yields a figure which is 200 miles less per day than an HST.

Do your sums make any allowance for the Oxford and Cotswold Line Class 180 and Turbo diagrams that IETs have also taken over in the past 12 months?

Even if it does, a set on a diagram with a couple of London-Worcester and back trips, plus a couple of London-Oxford and back trips, is clocking up 730-odd miles in the course of a single day, never mind the 910 miles that a Penzance-Paddington-Penzance diagram covers.

It always takes time for new stock to "bed in" so to speak. Reliability should level off at a high level eventually.

Though looks like there will be a level of GW fleet-specific problems until the antique 1990s ATP technology is removed from IETs, as Modern Railways says faults with this system account for 28 per cent of all technical incidents and 20 per cent of delay minutes.

 

[CC 72100]

They are indeed. Averaging fleet miles of a fleet that is still in delivery is going to give you some really dud figures.

The last HST workings on GWR are due to finish with effect from the May timetable. If the cl.802 delivery programme continues to progress as it was doing before Xmas, that may come earlier, driver training at Penzance permitting.

The last SX diagrams to finish on the Friday before the T/T change date are a PZ-LA, a LA-PZ and possibly a PZ-PZ, if that hasn’t gone earlier. Both of the first 2 diagrams I quoted involve lunchtime Pullman dining.

Would I be correct in my assumption that those diagrams therefore are)

1) 1W003 (PZ-LA)

2) IW001 or IW006 (LA-PZ)

3) IW002 or 1W007 (PZ-PZ)

(Numbers from this post - Remaining GWR HST Diagrams )

 

[coppercapped]

But what figure are you using for fleet size? That isn't yet a fixed number as deliveries of 802s is still in progress, and the last 800 was only delivered a couple of weeks ago... Like I said you're taking the published figures - which I agree don't paint the IETs in a great light for reliability - and reading in something that isn't there.

As @TractiveEffort states several units are managing in excess of 900 mile diagrams without issue; and there is (now) becoming a surplus of 80x stock while we wait for the New TT to be introduced - units that are sat spare on NP or SG are going to be contributing zero miles to those run; while still be counted towards fleet size.

Fleet miles run isn't a sensible reliability metric.

The data published in Modern Railways show the Class 800 fleet size being constant at 44 units during periods 6, 7 and 8 of the current financial year and the Class 802 fleet size being constant at 10 units.

If these are not the correct numbers of units in service at any given time then either somebody is not reporting the numbers correctly or there are errors occurring in the treatment of these data. If this is happening then it calls into question the whole basis of the railways' rolling stock reliability data.

I never suggested that fleet miles run is a reliability metric — but it is a performance metric.

Now that some services are operated with two (5 coach) sets twice the unit-miles are needed for the same number of train-miles. For example a train will run (roundly) 120 train-miles between London and Bristol; if this is an HST or a 9 coach IET this will be reported as 120 fleet miles, but for two 5 coach IETs this should be reported as 240 fleet miles in order to give consistent fleet reliability statistics.

It is possible to work out the total number of train-miles which are needed to fulfil the published timetable. In the case of fGW/GWR's long distance, high speed services and fixed length HSTs then the number of miles each unit has to run in a day to meet the timetable requirements can be determined; over the last few years of HST operation with a fleet of some 50-odd HSTs this worked out to be around 750 miles per unit per day. The increasing use of double units which can split en-route will obviously affect the number of miles run per day needed to meet the timetable which, as you said, will be intensified in the near future. This makes a direct comparison more difficult but I suspect that in future the miles per day demanded of the IET fleet will be higher than that required of the HST fleet

I am not surprised to learn that some IETs are covering more than 900 miles per diagram - but that is what I would expect with newer trains. But these miles are captured in the fleet mileage per period so feed into the average miles run per unit per month figure.

Do your sums make any allowance for the Oxford and Cotswold Line Class 180 and Turbo diagrams that IETs have also taken over in the past 12 months?

Even if it does, a set on a diagram with a couple of London-Worcester and back trips, plus a couple of London-Oxford and back trips, is clocking up 730-odd miles in the course of a single day, never mind the 910 miles that a Penzance-Paddington-Penzance diagram covers.
Though looks like there will be a level of GW fleet-specific problems until the antique 1990s ATP technology is removed from IETs, as Modern Railways says faults with this system account for 28 per cent of all technical incidents and 20 per cent of delay minutes.

Why do people have difficulty with averages? The data published lists the total miles run by each fleet — it makes no difference at all whether the miles run are to Hereford, Penzance or anywhere else.

I also read the MR article mentioning the ATP issues. To my mind it raises two questions:

• is it the ATP itself that is unreliable or is it the interface to the IET's train management system?
• do the ATP issues also have a similar effect on the reliability of the HST fleet? (fGW/GWR's HSTs have always returned worse figures than HSTs used on other lines which don't have ATP).

 

[jimm]

Why do people have difficulty with averages? The data published lists the total miles run by each fleet — it makes no difference at all whether the miles run are to Hereford, Penzance or anywhere else.

I have no difficulty with averages but it is pretty clear from points made above that your calculation is based on figures that do not seem to make much sense, such as the number of sets supposedly in service remaining constant over an extended period when extra IETs were entering service at regular intervals. Whatever it is that you are trying to prove in the first place.

For the average mileage to be as low as you claim, there would have to be a lot of diagrams covering pitifully few miles each day if lots of the others cover 700, 800 or 900 miles.

That 730-mile diagram I noted above would be an entirely typical one for any IET, HST or anything else allocated to Oxford and Cotswold services on any given day.

There are likely to be some IET diagrams that will see sets work into Paddington on a morning peak train, visit the depot for maintenance in the middle of the day, then head back out for the afternoon peak which won't cover many miles, but that is no different from the kind of thing that was happening with some HSTs in the past.

 

[FGW_DID]

Shortest / Longest daily working Diagrams (miles): HST v IET

HST: 242 (evening work only) / 1044
800/0: 436 / 1025
800/3: 255 (8 hr layover between the peaks) / 869
802: 776 / 990

 

[Charlie Smythe]

Why do people have difficulty with averages? The data published lists the total miles run by each fleet — it makes no difference at all whether the miles run are to Hereford, Penzance or anywhere else.

We don't have a problem with averages its that your figures don't add up right. For you very low mileage figure to be right, there would have to be a seriously large amount of sets doing almost 0 miles day. Sets doing almost 0 miles a day seems very unlikely to me.

 

[FGW_DID]

Sets doing almost 0 miles a day seems very unlikely to me.

Not every unit will be in use every day, there are always (or should be ) more units than working diagrams.

As an example, there are 45 Class 387 and 40 working diagrams therefore 5 units a day, will be doing 0 miles.

 

[FGW_DID]

Average daily working diagram mileage HST & IET:

HST: 750
800/0: 799
800/3: 643
802: 940

 

[Charlie Smythe]

I appreciate that, however, @coppercapped has stated that the IET's are doing at least 200 miles less per day than the HST's, which to me seems unlikely given that a lot of diagrams have been replaced like for like. Even though there will be sets sat all day, there would have to be a lot for them to bring down the average that much.

 

[coppercapped]

I have no difficulty with averages but it is pretty clear from points made above that your calculation is based on figures that do not seem to make much sense, such as the number of sets supposedly in service remaining constant over an extended period when extra IETs were entering service at regular intervals. Whatever it is that you are trying to prove in the first place.

We don't have a problem with averages its that your figures don't add up right. For you very low mileage figure to be right, there would have to be a seriously large amount of sets doing almost 0 miles day. Sets doing almost 0 miles a day seems very unlikely to me.

These are NOT MY FIGURES!

Modern Railways has been publishing tables of reported fleet sizes, fleet miles run and the number of TINs for several types of new train over the past year. The derivative values, as also published by Modern Railways, are Miles per TIN (MTIN) and the Moving Annual Average of the MTIN time series. The magazine states that the data originate from the Rail Delivery Group which collates the performance of individual fleets for Fleet Challenge within the National Performance Task Force.

Doing simple arithmetic on the published numbers gives the results I posted. On the face of it the miles run per unit per day are a long way below that which even 40 year old HSTs achieve and would be insufficient to meet the demands of the timetable if only IETs were available. Believing these data to be accurate I could not see how it was possible that the GWR would be an all IET fleet by 'Spring 2019'. I thought that such statements had close parallels to Crossrail which suddenly slipped 18 months or so within four months of the scheduled completion date and so thought it was worth posting the result of my arithmetic. I considered the required leap in miles per day figures to be too great to be reached in five or six months as the values had flatlined for the previous three periods.

If anybody doubts the accuracy of the published numbers then this should be taken up with the organisation which collates the data because it implies irregularities in one or all of the reporting of the number of trains in service, the miles they have covered and possibly the number of TINs attributable to the trains. Any incorrect values would affect the published reliability statistics.

The next question is: do the private, internal statistics agree or disagree with the published ones?

If the railway uses questionable statistics for planning and management purposes then I am not surprised that performance is not where it should be.

 

[coppercapped]

I appreciate that, however, @coppercapped has stated that the IET's are doing at least 200 miles less per day than the HST's, which to me seems unlikely given that a lot of diagrams have been replaced like for like. Even though there will be sets sat all day, there would have to be a lot for them to bring down the average that much.

The figures that FGW_DID posted are DIAGRAM (that is, scheduled) miles, the figures I posted are based on the published numbers and give ACTUAL run miles. There is no discrepancy between the two sets of numbers.
But possibly lots of questions!

 

[FGW_DID]

Any chance of posting the actual figures for the MTIN per fleet please so we know what numbers we’re dealing with?

 

[Charlie Smythe]

The figures that FGW_DID posted are DIAGRAM (that is, scheduled) miles, the figures I posted are based on the published numbers and give ACTUAL run miles. There is no discrepancy between the two sets of numbers.
But possibly lots of questions!

Could you post a link so we can look for our selves?

 

[Clarence Yard]

As the person who has drawn up every version of the LTP cascade plan, I can categorically state that each SX HST diagram has been replaced by a 1 x 9 car or 2 x 5 car IET diagrams. The former class 180 diagrams were replaced by 4 x 5 car. The 3 Bedwyns Turbo diagrams will eventually be replaced by 3 x 5 car.

In addition, there have been LTP driver training diagrams, currently 2 x 800 5 car and 3 x 802 5 car.

I am away from my desk for a few days but that 44+10 fleet total for P6-P8 looks very wrong. I don’t recall either the deliveries or the cascade plan stopping for a full 3 periods!

 

[coppercapped]

Any chance of posting the actual figures for the MTIN per fleet please so we know what numbers we’re dealing with?

Could you post a link so we can look for our selves?

I am loath to quote the complete figures out of the magazine, if you want them I suggest that you buy it.

However, in addition to the values I posted upthread the published MTIN values for the 800 and 802 fleets are


The best units on the railway at the moment are the Class 350 at 61,000 MTIN. Class 142 Pacers come in at around 9,000 MTIN.

 

[cactustwirly]

I am loath to quote the complete figures out of the magazine, if you want them I suggest that you buy it.

However, in addition to the values I posted upthread the published MTIN values for the 800 and 802 fleets are
View attachment 57331

The best units on the railway at the moment are the Class 350 at 61,000 MTIN. Class 142 Pacers come in at around 9,000 MTIN.

And what are the MTIN figures for the GWR HSTs over the same period?

Will be interesting to see what the FLIRT MTIN figures are like!

 

[SamYeager]

And what are the MTIN figures for the GWR HSTs over the same period?

Will be interesting to see what the FLIRT MTIN figures are like!

Modern Railways (MR) is only publishing the figures for new trains (345(FLU),345(RLU),700,707,800, 802). IIRC MR is only counting TINS due to a fault with the train. I've seen some mention of DfT being the source of these figures but a quick search on gov.uk didn't turn up anything.

 

[43096]

Modern Railways (MR) is only publishing the figures for new trains (345(FLU),345(RLU),700,707,800, 802). IIRC MR is only counting TINS due to a fault with the train. I've seen some mention of DfT being the source of these figures but a quick search on gov.uk didn't turn up anything.

You need to read the January Modern Railway’s, then!

 

[jimm]

These are NOT MY FIGURES!

Modern Railways has been publishing tables of reported fleet sizes, fleet miles run and the number of TINs for several types of new train over the past year. The derivative values, as also published by Modern Railways, are Miles per TIN (MTIN) and the Moving Annual Average of the MTIN time series. The magazine states that the data originate from the Rail Delivery Group which collates the performance of individual fleets for Fleet Challenge within the National Performance Task Force.

Doing simple arithmetic on the published numbers gives the results I posted. On the face of it the miles run per unit per day are a long way below that which even 40 year old HSTs achieve and would be insufficient to meet the demands of the timetable if only IETs were available. Believing these data to be accurate I could not see how it was possible that the GWR would be an all IET fleet by 'Spring 2019'. I thought that such statements had close parallels to Crossrail which suddenly slipped 18 months or so within four months of the scheduled completion date and so thought it was worth posting the result of my arithmetic. I considered the required leap in miles per day figures to be too great to be reached in five or six months as the values had flatlined for the previous three periods.

If anybody doubts the accuracy of the published numbers then this should be taken up with the organisation which collates the data because it implies irregularities in one or all of the reporting of the number of trains in service, the miles they have covered and possibly the number of TINs attributable to the trains. Any incorrect values would affect the published reliability statistics.

The next question is: do the private, internal statistics agree or disagree with the published ones?

If the railway uses questionable statistics for planning and management purposes then I am not surprised that performance is not where it should be.

If your initial post had been framed as a question, such as 'do these figures I have come up with on the basis of the tables in Modern Railways make sense?', rather than making statements on the basis of your sums such as

So the best of the IETs are running some 200 miles per day fewer than the HSTs they are replacing

or

It is quite obvious that the diagrams are NOT being replaced like for like.

and not questioned when the full GWR IET fleet would be in passenger service, perhaps you might have got a different reaction.

Never mind then digging your heels in and going 'published data' over and over when several people pointed out that the said data did not seem to stack up, so perhaps the conclusions you drew from your sums based on that data might not stack up either.

I've no idea what Crossrail is supposed to have to do with entry into service of GWR IETs - that delay is mainly about infrastructure, getting the central London stations finished off and the railway through the tunnels commissioned. Though looking at Class 345 reliability so far, it will give some useful breathing space to try to sort them out as well.

In his piece about the Golden Spanners in the current Modern Railways, Roger Ford also looks at the apparent dramatic decline in MTIN of SWR dmus - but this seems to be down to how SWR records technical incidents compared with the way that SWT used to do it, which suggests all the figures need to be treated more as a broad guide to the overall situation, not absolutely precise details.

 

[E_Reeves]

Has anybody else noticed a clunking sound as an IET leaves a station?

Sounds like the doors locking or moving but surely it can't be?

 

[coppercapped]

If your initial post had been framed as a question, such as 'do these figures I have come up with on the basis of the tables in Modern Railways make sense?',

SNIP

I've no idea what Crossrail is supposed to have to do with entry into service of GWR IETs - that delay is mainly about infrastructure, getting the central London stations finished off and the railway through the tunnels commissioned. Though looking at Class 345 reliability so far, it will give some useful breathing space to try to sort them out as well.

In his piece about the Golden Spanners in the current Modern Railways, Roger Ford also looks at the apparent dramatic decline in MTIN of SWR dmus - but this seems to be down to how SWR records technical incidents compared with the way that SWT used to do it, which suggests all the figures need to be treated more as a broad guide to the overall situation, not absolutely precise details.

Oh please! I did not pose a question because Modern Railways has been publishing train reliability data for several years and promotes the Golden Spanners industry awards every November which are based on these figures. The accuracy of these numbers has never before been called into question - so why should I doubt them?

On the other hand the industry has form in not completing work to the published schedules - this includes the introduction of new and rebuilt rolling stock as well as infrastructure work. It is entirely appropriate to take some statements with a pinch of salt.

I repeated my statements because it seems that my original post had not been completely understood although I tried to make it as clear as possible. As it is now clear from Clarence Yard's and JN114's posts that the numbers published in Modern Railways do not reflect the situation on the ground, and have not done so for several months, but show the trains as performing worse than they are in practice, I would hope that GWR and Hitachi will now ensure that the correct data is made available. Even the DfT might be interested to show that it has not bought a lemon!

 

[J-2739]

Has anybody else noticed a clunking sound as an IET leaves a station?

Sounds like the doors locking or moving but surely it can't be?

Isn't it the mechanism of the door that pressure seals them?

 

[JN114]

Isn't it the mechanism of the door that pressure seals them?

Or something similar, yes. Certainly nothing to be concerned about. My technical knowledge of IETs isn’t amazing; @TractiveEffort might have a more detailed explanation of what’s involved mechanically.

 

[spark001uk]

Has anybody else noticed a clunking sound as an IET leaves a station?

Sounds like the doors locking or moving but surely it can't be?

It is moving, slightly. You can observe it if you look carefully. It's a set of cam locks pressing the door against its seal.

 

[spark001uk]

Have the 80x been cleared for Castle Cary - Yeovil - Exeter?
Reason I ask is there was an incident at Taunton yesterday, and looking on Twitter there was one doing a Penzance, that passengers were initially told may be diverted, but was instead held at CC for almost 1.5hrs, before taking the normal route through Taunton.

 

[Mag_seven]

I've noticed that all the chairs shake quite violently when the switching to/from diesel mode.

 

[PHILIPE]

Have the 80x been cleared for Castle Cary - Yeovil - Exeter?
Reason I ask is there was an incident at Taunton yesterday, and looking on Twitter there was one doing a Penzance, that passengers were initially told may be diverted, but was instead held at CC for almost 1.5hrs, before taking the normal route through Taunton.

There have been gauging runs but a problem exists between Exeter Central and St.Davids. Haven't heard anything about it being resolved yet.