Hot weather warning 25th July 2019

[whole diary]

[Clan Line]

Daily Telegraph this morning:

[Timmer]

I dare say after yesterday's disruption to trains heading North out of London there will be a few more working from home requests going in.

[a-driver]

I am a bit more forgiving of snow/ice induced chaos*, because a warming climate suggests that the problem will become steadily less serious as the years go past.

Heat related problems are much more important in my view, because without significant mitigation action, they will get worse and worse.
And whilst OHLE in the GWR^▸ (Great Western Railway) area looks re-assuringly robust, this wont help if the wires expand and sag in hot weather. As more of the GWR network is eventually electrified we have large scale OHLE failures to look forward to.
Survival of the nice solid supports is little consolation if the wires attached thereto have fallen down.

Early on in the sorry IET^▸ (Intercity Express Train) saga, I expressed doubts as to the reliability of the units in hot weather. A supporter of the new DMUs^▸ (Diesel Multiple Unit) felt that I was unduly negative over this and pointed out that reliable operation at, IIRC^▸ (if I recall/remember/read correctly) 32 degrees was an "essential requirement"
I speculated as to what happens at 40 degrees, this was said to needless speculation as it "wont get that hot" Well today we found out what happens at nearly 40 degrees (met office) and almost certainly well over 40 degrees at engine height, in a cutting, and with another unit alongside.
Over a dozen short formations DESPITE numerous cancellations that SHOULD have resulted in plenty of spare units.

Anyone would think the IETs were the only trains to have any heat related problems today!  In reality they performed reasonably well, given recent issues with engine reliability, and suffered the same displacement issues due to speed restrictions as anything else on the GWR network - which it seems held up better than any other long distance route out of London on the second hottest day ever recorded.  I see a HST^▸ (High Speed Train) had to be evacuated near Peterborough.

Another textbook case of confirmation bias from Broadgage. 

Former colleagues tell me they performed far from reasonably well, backed up by a lot of their personal Facebook posts.

Call them cynical, but apparently they tell me there were no diesel only restrictions, is it a coincidence that happens for the first time on the hottest days of the year?!

[IndustryInsider]

There were no diesel only restrictions for two days last week and none today either.

[Adrian]

Our new OHLE is ugly, but very resilient!

In the case of drop weights, you can measure the drop, count the turns, and work out the temperature range for a given wire run length. For a Tensorex, it's not so obvious, but it still can only cope with a certain expansion length. Presumably that's enough for what was foreseen by the designers about ten years ago, plus a safety margin. The only other question is whether they were all set up right - I think that can be checked visually, though I don't have the details.

I did a quick calculation.  A 1000 metre solid copper cable would extend by 85cm if heated from -10 to +40 degC.  OK, so catenary is made of twisted strands, and is some sort of alloy, but I guess the extension would be of that order.  But then, the catenary is clamped at regular intervals - surely it must be able to slide through these to allow for expansion and contraction?

I also wondered whether the catenary and contact wire will stretch a bit over their lifetime - they're under considerable tension, after all.  Do the tensioners and weights need adjustment from time to time, and could that be part of the problem if they hit their end stops in very hot weather?   

[dviner]

Actually... how did the GWR^▸ (Great Western Railway) overhead wiring cope with the heat?

[stuving]

Actually... how did the GWR^▸ (Great Western Railway) overhead wiring cope with the heat?

Because it was designed to!

Specifically (and in answer to Adrian):

expansion coefficient of catenary wire: 18 ppm/K   (CuAg conductor wire is 17, so less demading)
maximum wire length (half run length)  750 m     (wire runs usually fixed at mid point)
Tensorex C+ travel                              1m   (there are ones with shorter travel, presumably for short runs)

therefore temperature range allowed for   74 K  which should cover (say) -20^o to +50^o

The Tensorexes (Tensorices?) have a helpful dial on the side to tell you where they are in their travel. There is a standard maintenance task to measure the wire temperature and check the dial. However, it isn't required unless something else has happened to mechanically upset the wire - so evidently stretching in service isn't an issue (i.e. too small to worry about).

[eightonedee]

A few observations from the sharp end (commuting on Thames Valley/North Downs) over the last two days-

1 - It does not seem to be the IETs^▸ (Intercity Express Train) that have failed but Turbos - two missing at peak evening period yesterday and today on the North Downs , with train problems blamed on Journey Check (OK - pinch of salt time)

2 - We knew in the morning that from 12 noon that a reduced speed service would be operating, so shouldn't GWR^▸ (Great Western Railway) have an emergency fall back timetable they can implement and use on all information systems?

3 - Transport Focus are clearly best prepared - there was already a "how was it for you" questionnaire about hot weather problems in my Inbox when I got home this evening!

[stuving]

I did a quick calculation.  A 1000 metre solid copper cable would extend by 85cm if heated from -10 to +40 degC.  OK, so catenary is made of twisted strands, and is some sort of alloy, but I guess the extension would be of that order.  But then, the catenary is clamped at regular intervals - surely it must be able to slide through these to allow for expansion and contraction?

Following on from the wire expansion range of 74^oC = 1m, setup is based on a central temperature of 20^oC so should be -17^o to +57^o. In fact that's also stated as -18/+56 or even -20/+60; I guess no-one's really too worried about the couple of degrees at the extremes ... yet.

As to support and registration, the principles are:

Most runs have tensioners at both ends, and are anchored or at least restrained in the middle to define where the wires are along-track at that point. Elsewhere the wires define their own positions along-track depending on temperature, and the supports mustn't resist that.

The catenary wire is held vertically and horizontally at each support, and drapes in a catenary shape between them defined by the load (its weight and the contact wire etc.) and tension. Some systems do have pulleys or slides, but Series 1 uses (almost always) cantilever arms that swing horizontally. That provides support at a fixed height, and laterally at position that is fixed (can't be pushed sideways) but moves as the arm swings (due to wire expansion). Arms are 1 to 1.8 m long, and the along-track movement is +/- 0.5 m, so this swing could be +/- 30^o! There may be a rule that the longer arms are used near the ends of a run.

The conductor wire hangs from the catenary wire on droppers, which holds it level but allows a pantograph to lift it against a reasonably constant force due to the tension and weight of the conductor wire. Its horizontal position is defined by registration arms, which are clamped to it but free to move vertically and along-track. They pivot at a point on the cantilever arm even further out (usually) than the catenary clamp, but as the two wires expand by almost the same amount the conductor wire stays below the catenary wire.

[IndustryInsider]

So, the all time temperature record wasn't beaten after all.

Global warming? Pah!

Then again, perhaps it was? 

https://www.bbc.co.uk/news/amp/uk-49132575

[stuving]

Following on again from the design temperature range of Series 1 being 74 K, I was wondering what it was in earlier OLE^▸ (Overhead Line Equipment, more often "OHLE"). I don't have technical data on that, but I did find this in BR^▸ (British Rail(ways)).12034/16 (from Railways Archive), which was a summary of current practice in about 1988 (it isn't dated):

"...the equipment is automatically tensioned, usually by weights and pulleys, in order to maintain constant tension within +/-1.5% in the catenary and contact wire over a temperature range of 56^oC. For this temperature range, the maximum tension length into which the equipment can be divided is just under 2 km."

Assuming that does mean 56K, not 0-56^oC, it doesn't look a big enough number even for the observed climate leading up to 1988, when account is taken of direct sun heating of the wires. How big that effect is, for example compared to rails, is another thing I'v never seen information about. What I can say, however, is that if anyone (like your boss) suggests going out to paint the wires white - look very caefully at the PPE you are provided with.

[Adrian]

Assuming that does mean 56K, not 0-56^oC, it doesn't look a big enough number even for the observed climate leading up to 1988, when account is taken of direct sun heating of the wires. How big that effect is, for example compared to rails, is another thing I'v never seen information about. What I can say, however, is that if anyone (like your boss) suggests going out to paint the wires white - look very caefully at the PPE you are provided with.

More of a breeze a few metres above the ground, and not much effect of heat radiated from the ground - I reckon the wires would be significantly cooler than the rails on a sunny day.  Could try using chrome plating the wire to keep it even cooler?

Thanks for all the technical details to satisfy my curiosity - all very interesting!

[stuving]

More of a breeze a few metres above the ground, and not much effect of heat radiated from the ground - I reckon the wires would be significantly cooler than the rails on a sunny day.  Could try using chrome plating the wire to keep it even cooler?

The extra wind is certainly going to help remove local hot spots. But the ground radiation effect is the same (nearly) at any height - the greater distance being exactly compensated by the greater area of ground in any solid angle. And the ground is always there, of course, filling all the directions that are "down" rather than "up", plus those blocked by things.

What I think matters more for rails is ground reflection, for example from a light-coloured cutting side wall. There may be some reflection from shiny (well a bit bright anyway) new ballast too. That may be a bit less at OLE^▸ (Overhead Line Equipment, more often "OHLE") height.

There is another mitigation for the wires, that the length depends on the average temperature over the whole 700 m. For a rail, the equivalent length is far shorter - less than 200 m, at least in the short term. On the other hand, there is a minimum length of rail that can buckle.

[ellendune]

The temperature record has now been confirmed.

[broadgage]

Presumably the overhead wires are also heated a bit by resistance losses caused by the current flowing. This would be very variable according to amount of traffic, not a constant.

Regardless as to the details, it is well known that the wires expand in the heat, and means to compensate for this are readily available.
So what went wrong ? Was the temperature greater than the maximum planned for ? If so that sounds like a basic error in design or installation. The temperatures though reaching a new record were still only about 1 degree higher than those previously achieved. I would expect infrastructure to function correctly up to at least 5 degrees higher than the previous record.

Or were the multiple failures not due to simple expansion of the wires, but something else related to the heat.