River Avon at Kelston, near Bath

[rodent279]

Yes, it is a railway related question. Looking at the Bristol-Bath section of the GWML, on NLS maps, I am a bit puzzled about the course of the River Avon. It appears to have been diverted at some point in the fairly recent past, say post about 1970. The map on the left is the OS 1:25000 1937-61 overlay, on the right is Bing Satellite. The river is clearly following a slightly different course, between the bridge under the former MR Mangotsfield-Bath GP line and where it comes parallel to the GWML.

Can anyone give any further info as to why and when this was done?

(I have an idea why, from school geography lessons, but that was a long time ago, and I paid more attention to the trains outside on the WCML than to what was going on in class!).

 

 

[young37215]

There was quite a bit of work done to reduce the impact of the regular flooding in the area culminating in the 1968 big flood. I dont know specifically but it seems reasonable to assume that the river would have been a focal point and the change has straightened out the river.  

[Bomag]

The river river (sorry Afon Avon) is a navigation at this point. There is a marina and wear to the west and a lock east. The bank is managed for mooring. Significant sections between Bath and Bristol have been altered.

[justin]

Another possibility, and it is only supposition, is that the outside of the bend is where bank erosion takes place and the river was very close to the railway embankment on the bend on the right, so it would remove the risk of undermining the embankment over time by realigining the river.

 

The article linked below doesn't mention any specific work east of Bath as part of the post 1960 flood alleviation programme

 

https://historyofbath.org/images/BathHistory/Vol%2007%20-%2009.Buchanan%20-%20The%20Floods%20of%20Bath.pdf

[rodent279]

1 hour ago, justin said:

Another possibility, and it is only supposition, is that the outside of the bend is where bank erosion takes place and the river was very close to the railway embankment on the bend on the right, so it would remove the risk of undermining the embankment over time by realigining the river.

 

The article linked below doesn't mention any specific work east of Bath as part of the post 1960 flood alleviation programme

 

https://historyofbath.org/images/BathHistory/Vol 07 - 09.Buchanan - The Floods of Bath.pdf

My thoughts exactly. One of the few things I picked up in Geography at school was how rivers erode banks more rapidly on the outside of the bend. Looking at the old maps, the bend is as you say very near to the railway, and I guess over time, it would start undercutting it, eventually causing subsidence.

Whether that's actually why it was diverted is another matter, but it's certainly plausible.

 

[Edwin_m]

3 hours ago, rodent279 said:

My thoughts exactly. One of the few things I picked up in Geography at school was how rivers erode banks more rapidly on the outside of the bend. Looking at the old maps, the bend is as you say very near to the railway, and I guess over time, it would start undercutting it, eventually causing subsidence.

Whether that's actually why it was diverted is another matter, but it's certainly plausible.

 

It wouldn't surprise me to find that Brunel had diverted the river for a short distance (near where it says "Bristol 8" on the left hand map), to avoid having to cross it twice, thereby introducing the bend in question.  

 

If you look at the modern OS map the boundary that follows the river in the 1968 map is still on the old course.  

[Nearholmer]

My thought exactly - that bit looks suspiciously straight, and if the flow is L-R, the would have a natural tendency to want to continue straight, cutting-off the small meander. The entire section that has been smoothed out would have a tendency to flood when in spate, as well at to under-cut the bank.

 

[Wickham Green too]

No, flow is right to left ........... that's downhill  towards Bristol and its Channel !

[rodent279]

Would diverting the river be less hassle/expense than building two railway bridges across it?

[Gilbert]

The GWR mainline is south of the river from St Annes in Bristol to Bath Spa station. The bridges are on the former Midland line to Bath Green Park. I'm very familiar with that stretch - I assumed the river was even more windy before the GWR was built but apparently not.

https://www.saltfordenvironmentgroup.org.uk/history/history014.html#1883

 

I know people who were on the river in the 60s and 70s so will try to find out more

This old image actually shows erosion on the embankment.

 

 

Bizarrely the foot/tow path still follows the old route!

 

[Gilbert]

3 hours ago, Edwin_m said:

It wouldn't surprise me to find that Brunel had diverted the river for a short distance (near where it says "Bristol 8" on the left hand map), to avoid having to cross it twice, thereby introducing the bend in question.  

 

If you look at the modern OS map the boundary that follows the river in the 1968 map is still on the old course.  

This map suggest it may have been so.

But this 1817 map implies the river was straight pre-Brunel...

[Nearholmer]

1 hour ago, Gilbert said:

This old image actually shows erosion on the embankment.

 

Or, is that a retaining wall, built there to accommodate the meander?

 

There is a short part of the WCML, just North of Linslade Tunnel, where the formation was widened in I think the 1850/60s, which has a hefty retaining wall to accommodate a meander. In that case the meander accommodated is in the canal, which was forced into that position by a meander in the River Ouzel. [Much to my surprise, I've just found a photo on-line of the very wall, taken way back in LNWR days. I can't quite be sure, but I think while the WCML was still three tracks

https://jonlinney.co.uk/scans/postcards/pc02.html ]

 

Even if it is a wall, it would likely be susceptible to scouring and undercutting of its foundations by a river, in a way that it wouldn't by a canal.

 

 

Edited April 23, 2021 by Nearholmer

[Wickham Green too]

52 minutes ago, Gilbert said:

Bizarrely the foot/tow path still follows the old route!

Your hoss would need a very long tow rope !

[Edwin_m]

1 hour ago, Gilbert said:

The GWR mainline is south of the river from St Annes in Bristol to Bath Spa station. The bridges are on the former Midland line to Bath Green Park. I'm very familiar with that stretch - I assumed the river was even more windy before the GWR was built but apparently not.

https://www.saltfordenvironmentgroup.org.uk/history/history014.html#1883

 

I know people who were on the river in the 60s and 70s so will try to find out more

This old image actually shows erosion on the embankment.

 

 

Bizarrely the foot/tow path still follows the old route!

 

I hope while they're at it they re-route the path at C so people following it don't have to swim the river twice...

[rodent279]

45 minutes ago, Nearholmer said:

 

Or, is that a retaining wall, built there to accommodate the meander?

 

There is a short part of the WCML, just North of Linslade Tunnel, where the formation was widened in I think the 1850/60s, which has a hefty retaining wall to accommodate a meander. In that case the meander accommodated is in the canal, which was forced into that position by a meander in the River Ouzel. [Much to my surprise, I've just found a photo on-line of the very wall, taken way back in LNWR days. I can't quite be sure, but I think while the WCML was still three tracks

https://jonlinney.co.uk/scans/postcards/pc02.html ]

 

Even if it is a wall, it would likely be susceptible to scouring and undercutting of its foundations by a river, in a way that it wouldn't by a canal.

 

 

What is now the Up Slow at Linslade Tunnel was added in 1857, with a single bore put in next to the original 1837 double bore. The current Down Fast was added in 1876, at the same time as the rest of what is now the WCML was quadrupled as far as Roade. A second single bore, even more restrictive than the first, was added on their other side of the double, forming what is now the Down Fast.

 

 

I'm not sure whether the retaining wall for the railway above the canal was built in the 1830's when the railway was built, and the trackbed widened to the west, and track slewed over, when the first additional line was built, or whether the retaining wall dates from when that additional line went in.

 

[Nearholmer]

20 minutes ago, rodent279 said:

or whether the retaining wall dates from when that additional line went in.

 

The latter, I'm 99% certain.

 

The original alignment is clearly defined by the original double bore, unless it had a dog-leg in it, and the new track on the eastern side must have been accommodated by a significant widening of the embankment, sufficient to allow the spread between old and new tracks to allow the new track to get to the new single bore. That significant widening would have put the toe of the bank in the canal, hence the retaining wall. 

 

I quite often cycle to Leighton Buzzard (nearest model railway shop and steam railway!) along the tow path, and that retaining wall always impresses me - a very dour bit of typically LNWR masonry.

 

back to Bath ......

[Gilbert]

9 hours ago, Nearholmer said:

 

Or, is that a retaining wall, built there to accommodate the meander?

 

Even if it is a wall, it would likely be susceptible to scouring and undercutting of its foundations by a river, in a way that it wouldn't by a canal.

 

 

The river is subject to major changes in flow and levels so I suspect it would be a difficult wall to maintain..

Edited April 24, 2021 by Gilbert

[rodent279]

8 hours ago, Nearholmer said:

 

The latter, I'm 99% certain.

 

The original alignment is clearly defined by the original double bore, unless it had a dog-leg in it, and the new track on the eastern side must have been accommodated by a significant widening of the embankment, sufficient to allow the spread between old and new tracks to allow the new track to get to the new single bore. That significant widening would have put the toe of the bank in the canal, hence the retaining wall. 

 

I quite often cycle to Leighton Buzzard (nearest model railway shop and steam railway!) along the tow path, and that retaining wall always impresses me - a very dour bit of typically LNWR masonry.

 

back to Bath ......

Does make you wonder why they chose to add the first line in that way. Why not widen the formation to the west, build the first single bore to the west of the double, then split the double bore between the new line (today's Up Slow) and there original down line? Then the retaining wall would not need to be built, and when the second line was put in, either a second single bore could be added next to the first, or the first single bore could be widened.

[Nearholmer]

I think it might have to do with the length of tunnel. Because of the way the hill comes down and tapers in width, I think the eastern bore must be slightly shorter. There might also have been a good balance between cut and fill doing it on that side, whereas on the west it would have created “all cut and no fill” I think.

 

The hills there are a bit interesting geologically too, I believe, because of the mix of sand, sandstone, and clay, with spring-lines at multiple levels.

 

One of the very early surveys for the route put it c150 yards to the east, running along the valley with the river and canal, no tunnels at all, but I assume that was abandoned because there is a pinch-point in the valley, which would have led to a set of curves a bit like Berkhampsted.

 

Maybe this needs its own thread!