So ... what does this tell us?

Something went amiss in the priority of passenger trains over freight trains.

* An intermediate signal on the TransWilts would significantly increase capacity at time like this where there's more traffic around that scheduled because of late running.

* A loop at Chippenham station (where there's already a platform face available) would significantly reduce delays to main line trains by giving somewhere to get freights and / or TransWilts passenger trains out of the way.

While on the other hand the 18.32 from Westbury to Swindon ran on time. I noted this as I passed it between Bradford Junction and Trowbridge at 18.39 on the delayed (in theory) connecting 16.50 out of Gloucester

A single evening of laterunning freight trains is an indicator, but insufficient evidence to form a quanitative evaluation of what improvements should be made to let traffic and service levels grow between Swindon, Westbury and beyond.

I was struck on Sunday as to just how effetively the South Wales main line was being diverted from Swindon via Kemle and Gloucester, with an unimaginable previous efficiency as trains were passing at speed on the newly-redoubled section.   Even though trains were turing up with various delays on them, that delay was not being bounced back from Westbound to Eastbound by any remaining significant lengths of single track.  There is a comparison to be made in the medium and long term between various options on the TransWilts, and the ultimate solution may well be the "Kemble" one - i.e. redoubling the line all the way, or nearly all the way.   I think that the junction near Stonehouse is still single for a short distance, and I suspect that costwise a similar thing could be required / sensible at the river Avon bridge at Staverton.

The question, though, is wider that just the TransWilts on this story.

Re-instatement of a loop at Chippenham. Railway property still allows for this on the South side, Vision Chippenham is looking to it, Network Rail should be making passive provision in resignalling and electrification to allow for it - i.e. not plonking signals or masts in what would be the trackbed.  This will allow:
* Sidetracking of trains to / from the TransWilts
* A turnback for the Bristol Metro service rather than Batheaston
* Overtaking by London to Bristol express trains of the Oxford to Bristol Regional service
And the Bristol Metro and Oxford to Bristol regional services are both providing a train that can stop at Corsham.

Other possible options / improvements short of full re-doubling include
* An Intermediate signal on the TransWilts to let trains follow at shorter headway
* Re-engineering Thingley and Bradford Junctions back to double junctions with sufficient double track clear of the main lines to hold a long freight train
* Provision of an intermediate passing loop - either fairly short or dynamic

Electrification also come in somewhere here ;-)

Finally on track / operational improvements, I need to add in the request to re-instate the triangle at Bradford Junctions.  That would be superbly useful during next summer's diversions of Chippenham - Bath services to avoid Box tunnel, and I understand that one of the reasons that the curve was taken out in 1990 - the lack of power for power operated points at the junction - would no longer be an issue.

I would love to see it redoubled.  I think the loop at Chippenham is needed for quite other reasons too, and I like the idea of trains being able to wait off the main line for / from the single line without stopping a train going the other way.   I'm personally les convinced about an intermediate loop, simply because of the wide variety of trains and the tendency for them to be a few minutes late - loops are fine on single ended services where 2 passenger trains pass each other every hour, but I'm less convinced for th TransWilts at current robustness levels and think that signal and section end solutions may be better.  I'm not an expert on this and need to learn more to make a properly educated comment.


Yes,  the northbound freight waited at Westbury until it had gone and followed it after a decent interval.  The 18:32 gets to Chippenham just a few minutes before the 18:52 from Swindon heads back the other way - or that's how it's supposed to be - but the pair of late freights got in the way tonight; an effective round trip over the single line of 36 minutes.   An intermediate signal would have cut that to 20 minutes (saving 8 minutes headway between the 2 northbound and 8 minutes between the 2 southbound)  and the loop at Chippenham would have allowed the 18:00 from Paddington to overtake the Southampton train had the Paddington been on time.   Total delay time to TransWilts passenger trains would have been reduced from around 60 minutes to around 20, and the Bristol express, already 18 late, may only have suffered another couple of minutes if that due to conflict movements at Thingley.

It's a train of stone empties - not available to passengers

There were delays earlier, also.

The 1514 ex-Swindon was delayed between Trowbridge and Westbury (it left Trowbridge on time, contrary to what Realtimetrains reports). The return working, the 1615 from Westbury consequently departed late.

The cause isn't clear - there were two late-running freights off the Berks&Hants, then the Cardiff-Portsmouth was held outside Westbury - the Swindon was behind that.

Does anybody know if this was just congestion (freight standing fould of the junction?) or a failure of some kind?

That was one of the first things the advocates of privatisation and the fragmentation of the railways came up with. I remember hearing a Miniser say if a freight company wanted a path, in the middle of the evening peak, from the North London Line at Stratford to Barking, which involves crossing all 4 GE lines,they could have it!

Do freight trains pay the same level of track access charges as passenger trains? In the case of pasenger trains I understand that the access charge varies according to the weight and length of trains as the bigger the train, the more wear and tear is made to the track. If so a typical freight would be many times heavier and longer than a passenger train and would, I would have thought, paid much higher track access charges than, say , a 2-car sprinter.