Spectacular derailment in Western Australia.

[ben_issacs]

Folks, 

In the morning of the 5th of Nov., a BHP (Broken Hill Proprietary) loaded iron ore train of 268 hoppers and four diesel locos. was running from their Mt Newman mine  to Port Headland on the W.A. coast, a distance of about 300 km or so.  

For some reason the driver (one man op.) stopped the train midway along and got out  to investigate.

Inexplicably, the train started to move off, and it ran un-attended for 93 km., which it covered in about 50 minutes, when it was intentionally derailed.

What a mess!!

Speed at derailment point was calculated to be about 140 km/hr, so about 80 - 85 mph.

It looked like perhaps fifty cars at the rear were still upright, the front quarter or so was just a mass of twisted metal, unrecognisable , the remainder were considerably damaged.

One loco, probably the mid-train one was upside down minus its bogies, no doubt the lead locos were in a similar state

No one was injured.

From such photos that I've seen, I reckon that this would be the most costly railway accident to have happened in Australia.

The line is now back in operation.

To maintain the flow of iron ore to its port for export, BHP has to somehow make up for the loss of wagons and locos, plus the loss of the iron ore load, but this might have been salvagable to some extent. .

The official report on the cause of this accident won't be available util about mid next year, it will make very interesting reading, I'll bet!

Regards, 

Bill, 

Melbourne.

 

 

 

[katoftw]

got a link?

[ben_issacs]

Katoftw

Haven't found one, but inexpert in this field.

I've seen a video, taken out the side window of a car moving alongside the line, about a minute in length. 

Doesn't show the full length of the train, but shows enough to get a good idea of the wreckage.

Try Western Australia iron ore train derailment on Youtube.

I wonder if someone at the derailment site got a video, this would really be something to see!

Regards, 

Bill, 

Melbourne.

 

[cteno4]

 

 

[katoftw]

that looks expensive.

[Davo Dentetsu]

We are rather crap at doing railways here, it seems

[marknewton]

The reason the train stopped was because an ECP brake cable had parted. That gives you an immediate penalty brake application. The driver had no say in the matter. When he left the loco to investigate he didn’t apply the automatic/train brake, and after an hour had elapsed the ECP brake released, as it’s designed to do. When that happened the train ran away. There’s nothing “inexplicable” about it.

 

https://www.onrsr.com.au/__data/assets/pdf_file/0020/22475/Safety-Alert-RSA-2018-002-ECP-Braking.pdf

 

Cheers,

 

Mark.

Edited November 26, 2018 by marknewton

[NXCALE]

I thought BHP had robot trains like Rio Tinto.

One accident that could promote autonomous trains in that industry. 

 

[marknewton]

No, BHP still run trains with a human driver onboard.

 

Mark.

[ben_issacs]

Mark, 

If the ECP brake went off after an hour, that means the driver was out of the cab for this time.

Possible, of course..

Even with the length of the train, surely he could have done the check in less than an hour.

4.30 a.m., dark,  perhaps that slowed his walk  back.

No doubt we'll find out in due course.

Regards, 

Bill, 

Melbourne

 

[marknewton]

23 minutes ago, ben_issacs said:

Mark, 

If the ECP brake went off after an hour, that means the driver was out of the cab for this time.

Possible, of course.

 

Thats what happened, according to my contact at ONRSR. They’ve been in discussion about this with every operator who runs ECP stock, as well as the railways where those operators are running, including mine. I have no reason to doubt the veracity of the information we’ve received.

 

All the best,

 

Mark.

[Socimi]

One-man operation with no deadman? That's quite the recipe for a derailment.

 

Also, don't they have any kind of ATS (atleast in passing loops)? 

Edited November 26, 2018 by Socimi

[chadbag]

This is how changes get made.   After big accidents and the post mortem, laws are passed or new systems developed to help prevent future occurrences 

[marknewton]

The locos have task linked vigilance control, which is a more sophisticated system than simple deadman control. But the VC is suppressed if there is at least 200kpa in the loco brake cylinders.

 

BHP does use a form of automatic train protection, but it relies on the ECP brake being cut in and functional. The line this incident occurred on is double track. 

 

Mark.

[kvp]

1 hour ago, marknewton said:

The locos have task linked vigilance control, which is a more sophisticated system than simple deadman control. But the VC is suppressed if there is at least 200kpa in the loco brake cylinders.

BHP does use a form of automatic train protection, but it relies on the ECP brake being cut in and functional. The line this incident occurred on is double track. 

Mark.

Doesn't the railway make it mandatory for the driver to dump all air from the trainline before exiting the cab? This, together with spring force storage parking brakes makes it hard for a train to run away. My guess is that the ECP line break didn't open the dump valves automatically, which would have prevented it from releasing automatically. (at least on older european copies of westinghouse controllers, if the trainline is dumped, the brake handle has to be moved all the way to dump then back to full release for the release to work) I don't know anything about the ECP system on these trains though, espeicially the vigilance system, that in many countries usually just acts like any emergency brake valves along the train. (first you have to reset them manually, then reset the active brake controller too to be able to hold any air in the brake line)

[ben_issacs]

Folks,

Do the locos still have their independent brakes, if so, and these were on, and not affected by the ECP system, would they have been sufficient to hold the train when the ECP system shut down?

I suppose that if one knew the gradient on which the train stopped, it would be possible to calculate whether the gravitational pull on the 268 un-braked cars would overcome the holding power of the three (four?) locos with their brakes on.

Regards, 

Bill, 

Melbourne.

 

[marknewton]

20 hours ago, kvp said:

Doesn't the railway make it mandatory for the driver to dump all air from the trainline before exiting the cab? This, together with spring force storage parking brakes makes it hard for a train to run away. My guess is that the ECP line break didn't open the dump valves automatically, which would have prevented it from releasing automatically.

 

I don’t know what procedures BHP have in place for an incident like this, but I’ll make an educated guess and suggest that neither the driver or the train controllers fully understood this particular failure mode. Because if the driver had made an emergency application of the conventional auto air brake the ECP brake wouldn’t have released after 60 minutes. 

 

Neither the SD70s or the ore wagons have spring parking brakes, by the way. They’re applied and released manually.

 

When the ECP cable parted it caused an emergency application of the ECP brakes only. The auto air brake pipe remains charged.

 

Cheers,

 

Mark.

[marknewton]

12 hours ago, ben_issacs said:

Do the locos still have their independent brakes, if so, and these were on, and not affected by the ECP system, would they have been sufficient to hold the train when the ECP system shut down?

I suppose that if one knew the gradient on which the train stopped, it would be possible to calculate whether the gravitational pull on the 268 un-braked cars would overcome the holding power of the three (four?) locos with their brakes on.

 

The locos have an independent/straight air brake. I not know whether the driver applied it before leaving the cab or not.

 

But the train was standing on a 1/66 falling grade, and weighed 43620 tons. The four locos on the train couldn’t hold that sort of tonnage even if the independent was fully applied.

 

Cheers,

 

Mark.

[kvp]

19 minutes ago, marknewton said:

When the ECP cable parted it caused an emergency application of the ECP brakes only. The auto air brake pipe remains charged.

Thanks for the explanation! This failure mode however seems to a be a design fault as imho every emergency application should dump the brake pipe, just to be sure that the train stays put.

 

The ECP variants i've seen (Knorr-s) control the brake pipe directly, just open it on each car that supports it. (this also allows braking pneumatic brake only cars that only have passthrough cables or connected on the tail end or ones with faulty ECP equipment, but otherwise still working pneumatic ones)

[ben_issacs]

Mark,

Thanks for that info about total train weight and the grade on which it was standing.

Useful for some calculations that I want to do.

Are these trains permanently coupled, if so how is the train air line and the ECP cables passed from wagon to wagon?

How would one determine which car had a faulty ECP connection?

Regards, 

Bill, 

Melbourne.

 

[ben_issacs]

Mark,

If, on that train weight, we knock off the 620 tons for the four locos, we then get just on 160 tons per wagon.

130 tons for the load and 30 tons for the wagon, is that reasonable?

Regards, 

Bill.

Melbourne.

 

 

[Ochanomizu]

On 11/26/2018 at 9:55 PM, marknewton said:

The reason the train stopped was because an ECP brake cable had parted. That gives you an immediate penalty brake application. The driver had no say in the matter. When he left the loco to investigate he didn’t apply the automatic/train brake, and after an hour had elapsed the ECP brake released, as it’s designed to do. When that happened the train ran away. There’s nothing “inexplicable” about it.

 

https://www.onrsr.com.au/__data/assets/pdf_file/0020/22475/Safety-Alert-RSA-2018-002-ECP-Braking.pdf

 

Cheers,

 

Mark.

Hello,

 

Perhaps the ECP brake can be upgraded to require manual release in future?

[ben_issacs]

Folks, 

Just done a few calculations on this ore train runaway.

Using Mark's train weight and grade figures. on the 1:60 grade with a train of 43,620 tons, the component of the gravitational force parallel to the track , the 1:60 grade, is 727 tons.

This is the force tending to pull the train down the grade.

I've assumed that the four locos had a total weight of 600 tons.

Using the Static Co-efficient of Friction for steel wheel on steel rails of 0.78, it requires a force of 468 tons to move the locos against their brakes, less than the 727 ton force calculated above, so as Mark wrote, the independent loco brakes could not hold the train if all the train brakes went off.

Once the train starts to move the Co-efficent of Friction becomes a Sliding Co-efficent, which is much less than the Static one, so the train keeps on going.

With the loco wheels locked, they must have developed considerable flats at the place of derailment.

Regards, 

Bill, 

Melbourne.

 

 

[marknewton]

On 11/27/2018 at 11:13 PM, kvp said:

Thanks for the explanation! This failure mode however seems to a be a design fault as imho every emergency application should dump the brake pipe, just to be sure that the train stays put.

 

Yes, I agree. And i think that’s why it caught this driver and the train controllers unawares, as it’s a failure mode that is counterintuitive.

 

This is the AAR standard for ECP brake systems, which is what we’re using in Australia. It makes interesting reading.

 

http://www.spoornet.co.za/Website/tender_pdf/AAR Spec S-4200.pdf

 

Cheers,

 

Mark.

Edited November 29, 2018 by marknewton

[marknewton]

On 11/28/2018 at 11:16 AM, ben_issacs said:

Are these trains permanently coupled, if so how is the train air line and the ECP cables passed from wagon to wagon?

How would one determine which car had a faulty ECP connection?

 

Bill, the cars are not permanently coupled. They have knuckle couplers, one fixed and one rotary. The brake pipe hoses are connected in the normal way, and the ECP cables are slung from the brake pipe hoses. 

 

 

The ECP system incorporates what they like to call a “man-machine interface”. You and I would call it a data display screen. It shows the status of the ECP system and each vehicle in the train, so the position of a cable separating would be indicated by the MMI. This video shows a similar setup on an SSR coalie hauled by CEY class locos.

 

 

All the best, 

 

Mark.