Many thanks Waverley47 for that history lesson and reality-check. I fear this is straying beyond the scope of this thread, but I'll ask the £350m question (i.e. £294m construction + ~£55m preparation) - at the time you did the analysis, what figure for the total cost was in play? My recollection is that at each major re-announcement of the project, another figure in the tens of millions got added, meaning that the BCR would have been even poorer if the final figure had been known.
Also, although as you say the spec was in important ways stripped down, it's fair to say there were some embellishments such as bowing to local pressure and putting Stow station (my local station) on its awkward original site closer to the village and the lengthy extensions to the platforms at Newtongrange, Galashiels and Tweedbank. Also, there were expensive road improvements linked to the railway which I think were charged to the project, especially at Falahill. Can you put a figure on any of these?
A final tuppence worth from me as a layman, is that given the extensive civil engineering that was required to create a robust trackbed for a single track, the idea of it now being doubled tracked feels like almost another rebuild.
I am afraid I cant put figures on any of the individual aspects of the project, but I can give some perspective. I still have copies of most of the documents sitting in my office, but equally there's a lot of sensitive data on them, and sharing some of the things would probably get me in trouble with someone, but I'm not sure whom.
I can say though that there.was a lot of goodwill, and a lot of costs were written off. For example, and I've talked about this before, but the bridge under the bypass was originally supposed to be paid for by the council, because the bypass came after the railway.
We originally wanted to route it straight down the old route, but the council had earmarked part of the route for a new waste from energy plant. So we agreed to route around the site, and pay for the bridge on a different alignment, but the council payed for Shawfair station.
Similarly, Scottish Borders Council agreed to waive all the costs associated with road closures in Galashiels if we agreed to raise the line up and over Glenfield Road instead of severing it, and if we moved the line over to save space for the footpath over the Redbridge Viaduct and onwards to Tweedbank.
Stow station came very late in the day, but that was an issue with funding in the original plan. Being told we had to provide the station after years of assuming we didn't have to meant we could go back to parliament and say "you've told us to provide this, we haven't costed for it, so that's an extra several million pounds that needs to be added to the bill."
There were bits that were stripped out, the wholesale remodelling of Portobello Jn being one, although that was basically a fait accompli; we all knew that at some point it would have to be done, and that removing it and kicking the can down the road would please the treasury, but that sooner or later we'd be back needing to do it again for a lot more money. Similarly, wiring the thing from the start was taken out, but that meant that the Whifflet line had to be wired, and extra diesel units kept on lease to work the thing. That particular decision was actually a net increase in cost to ScotGov, but the calculations of what we could kick the can down the road as OPEX versus what we could spend as CAPEX were brutal at points.
There were a couple of single track sections that didn't have to be, but by making them single track we saved ourselves the costs of rebuilding the formation for doubling,. particularly over old viaducts with very limited clearance. The Glenesk viaduct, while originally double track, was fundamentally unsuitable; we took some of the local councillors and MSPs out with plywood track panels to show them just how little clearance there was.
We'd cut the cost down by turning it into a single track line with loops at some stations, and saved a lot of money. The single track iteration came out to save about £160m versus full doubling, and the treasury were very happy. However, throughout the project, as computer modelling of timetables and train performance got better, the whole thing started to get more expensive as the locations and lengths of the loops changed.
As Bald Rick says, the original sum, the version 1.0 calculation as submitted to the Scottish Parliament, had a very convenient BCR of 1.01, and a cost somewhere in the region of about £140m. Unfortunately, I can't remember the exact value. But even that BCR was using the most recent set of economic and census data, to maximise potential use, but used construction cost estimates from the Larkhall line, which had been gathered before the financial crash, to keep costs as low as possible and fudge the numbers. That first BCR didn't even include ongoing costs for running the service and lighting the stations.
The financial crash itself was a nightmare for us, because it became apparent in early 2010 that the economy of Edinburgh was actually recovering quite well. In fact, it was recovering faster than the inflation in construction costs, because a lot of companies had suddenly lost work and were sending in desperately low estimates. So we had this problem where inflation as a whole was dragging the costs up faster than tax receipts, and so the treasury wanted it submitted as soon as possible, but the longer we waited, the better the BCR actually became.
It shows that BCR figures are not to be taken literally, given the passenger success of the line. You can't predict human behaviour, it doesn't fit into neat computer calculations. I doubt if any lines would have reopened, if BCR was taken as gospel, set in stone.
How B is calculated/other has always fascinated me.
The BCR magic at the end of the day is just a number, it's just maps and graphs and estimates. I've written about it before, but we did a lot of dividing into what we called static, semi-static and fluid captures; estimates of how much each person within a reasonable distance of a station could be expected to use the line. Speaking with the councils and asking them to give us maps of where they'd build 100, 500, 1000, 5000 new homes in their constituencies and modelling that demand. We had all the data from Lothian Buses, vehicle counters along all the main roads into the city from Midlothian to try and get accurate numbers of commuters
We did it incredibly thoroughly, we had a team clear the trackbed of vegetation and we walked the entire route over a week in June 2010 with NR and engineers and councillors, with cameras and surveying equipment. Every single bridge was looked at, access points were drawn up, negotiations with landowners had been going for months before the bill was published. It was an enormous job, which is where some of the costs came from; the project had paid the full time salaries of over fifty people for three years before a single spade hit the ground.
A BCR is basically black magic at the end of the day; it's trying to boil down two decades worth of hypothetical change in the future and put a monetary value on that. But you can't put a monetary value on the psychology of having a railway link into Edinburgh, you can't model working from home, you can't model political changes and which parties might or might now want to build houses on bits of land, you can't model the massive benefit towards public attitude and the positivity towards rail that came with the line.
As I said previously, this wasn't a proper BCR, because it was the first one of its kind, and we did a lot of things wrong. I'm sure doing it now we could come up with a more accurate BCR, but it would never be perfect.
