Click to get your own widget

Wednesday, October 10, 2007


Some time ago I made an FOI inquiry to the Scottish Executive as to how the government's costing of a tunnel under the Forth of £4673 million had been arrived at when the Norwegians had been able to build hundreds of kilometres of tunnels at between £3.5 & £11 million a kilometre.

I have been sent a table which comes out as follows:

1 Cost of actually building the tunnel £1,023 million
2 Road Network £ 355 million
3 "Optimism Bias" £ 456 "
4 10% extra contingency fee £183 "

5 subtotal £2,018 "

6 Transport Scotland budget £100 "
7 VAT at 17.5% £353 "

8 more subtotal £2,471 "

9 allowance for 7.5% inflation £1,655 "
10 capital charges £545 "

11 Total £4,673 "

Neat. It proves that the civil service can make the figures jump through hoops to provide any answer they want.

Note that the actual cost of the tunnel is acknowledged to be 1 billion & the tunnel & roadbuilding come to only 29.5% of the total bill. This is, at least by comparison, relatively close to the figure of £250 suggested by Mr Roy Pedersen or the £500 million suggested by FTAG & various engineers on online discussions in the Scotsman. I have not been given more detailed figures of how the actual building is costed but if it has the same generous allowances for paperwork costs as these calculations I can well see that a more demanding calculation could come in at the lower independent figures.

Looking at these individually

(2) Why should building roads to the tunnel mouth cost this improbable figure? In fact I was given a more detailed breakdown of this & since almost all the expenditure on roads comes in the 2014-2017 period while most of the cost incurred on building the actual tunnel comes in 2011-2013 the 7.5% inflation indexing means that the road proportion is even higher. Iincluding this inflationary factor, the roads & TS budget come to half as much as the tunnel itself. This means that, all the other costs being expenses proportional to the building costs, building roads will have a total cost of £1600 million. Assuming that a bridge will also require roads leading up to it this means that a flat £1600 million will also apply to roads to it. Since the bridge is calculated at £2,500 million this means that the bridge plus all the extra factors will cost £1 billiion or the actual building of the bridge must come to a mere £295 million. Isn't Scottish engineering wonderful! This magnificent feat would only be possible, of course, if the civil servants were calculating the cost of a tunnel on exactly the same basis as the bridge - but of course we must assume that is what they are doing since anything else would be fraud.

(3) I am assured that "contingency costs called Optimism Bias" are "required by HM government Treasury Guidance" though I have not ben told whether the Treasury insists they be 1/3rd of the actual building ocsts or indeed whether they must be separate from the 10% contingency (4) more common in the industry - note that because the Bias increases costs by 1/3rd this in turn increases the contingency to 13% of the actual doing things cost.

(5) I am not sure that Transport Scotland's oversight of this scheme really justifies their £100 million budget.

(7) VAT is perfectly fair though since it is merely a bookkeeping exercise between government departments it would be reasonable to suggest that the Treasury forego part of it. No wealth is actually created or lost in charging VAT & it would be a pity if a government project were to be cancelled because the government cannot afford to pay the government their own extra taxes on it.

(9) The government are indexing all their costs on the assumption of a 7.5% inflation rate. They must know something we & the Treasury, don't.The Treasury says 2%. An interesting effect of this is that while the government are saying work will not start on this till 2011/12 if it were to start immediately we would save 3 years of 7.5% inflation(compounded to 24.2% or 19.5% of the final figure) - this would, in the government's theory, cut the cost of a tunnel by £910.

(10) I have no idea what "additional capital charges" are or why they should be introduced. or indeed why they should cost £545 million. The more detailed year by year breakdown shows that 2/3rd of this additional capital is only introduced from 2014 when most of the expense of the actual tunneling has been done.

These figures have all been officially calculated on "Quarter 4 2006" prices which suggests to me that it was all fixed up before the SNP, who came out for a tunnel during the election campaign, got there.
I believe these figures bear no real relationship at all to what a tunnel could actually cost. By comparison I have pointed out before that the Norwegians have been able to build tunnels at as little as £3.5 million a kilometre & that the Danes & Germans are colaborating on a 20 kilometre bridge costed at £3.7 - & take 12 years. If they are making the same assumption about inflation as our own government prices will have doubled by then meaning that the current cost of their proposed bridge would be £1.8 billion Perhaps the Danes & Germans, unlike the Scottish Office, don't expect 7.5% inflation.

What we should do, instead of having this fraudulent comparison is to throw an open competion for a fixed price contract for a crossing. Let the world's civil engineering companies make open bids, our government guarnateeing only to pay for land & government legal & regulatory costs & see what bids are made.

I have shown before how the cost of the original Forth Road Bridge was £15 million & how at real world inflation rates this would equate to £360 million today not the 8 times greater figure we see being foisted on us. Until such contracts are awarded openly & open to the world we are going to see such massive artificial increases continue.

My own highest estimate of a tunnel cost, based on the highest Norwegian priice of £11 million would be £11 million X 3 (assuming a 3 kilometre tunnel to achieve depth) X 2 (2 lanes) X 2 (assuming extra large motorway sized tunnels) X 2 (allowing high fixed start up costs buying equipment for this one project not currently available) x 121% (2% inflation over 10 years) = £319 million. If it was decided to use the same equipment to make a series of tunnels as per my Scottish Tunnels Project the last doubling would disappear. One suggestion that has been made is to convert Rosyth dockyard to manufacture sections for a tunnel laid across the seabed - obviously if this were done only for the Forth Tunnel all the costs would accrue to it whereas if it was used to create such tunnels to the Scottish Isles etc this cost would be spread very wildely. If the basic cost was £3.5 milliion, like the basic Norwegian price a Forth Tunnelt could cost as little as £26 million without start up costs.

I will be forwarding this article to interested parties. It may be that somebody in Transport Scotland or in Scottish politics or business will be able to dispute these figures or verify the government's ones - we shall see.


You make some interesting points and I think clarity in the process is something that is vital if we are all to have the comfort of knowing that we are getting value for money. Not being a cost consultant I couldn't answer all your points. However, I would pick up on a few as follows:

(2) The main reason the roads element for the tunnel option is so much is that to the south of the river a bored tunnel would be required to pass underneath the Hopetoun estate and to the north of the river the depth of cutting into rock will be significant (by crude calculations perhaps as much as 40m) most of it into rock. While there will be road costs for the bridge option they would not be nearly as much as the lengths of connection are far less (<1km on each side as a minimum)

(3): I understand that this optimism bias is a requirement of these major government funded projects. Not sure how the figure is arrived but perhaps a name 'pessimism bias' would be more appropriate. This would apply to the tunnel and bridge options.

(4): 10% extra contingency. I actually feel this is quite light for such a major project, particularly for the tunnel option as the ground risk is significant. I would have expected a contingency of between 15% and 20% at this stage of the project perhaps pared down to 10% by the time the design is developed to tender stage

(6): TS budget will include all the designer's fees, specialist testing, legal departments for getting the project through parliament and then tying up the contract as well as a host of other costs. Although a large amount, this works out as a relatively small percentage of the overall cost of the project so may well be reasonable. It would be interesting to see how this compared to other major infrastructure projects which have been delivered.

(7): Agree with you on the VAT issue. This is such a vital part of our infrastructure it should be VAT free but who knows how government departments work.

(9): Inflation at 7.5% in the construction industry is actually quite low. There is an awful lot of work out there at the moment and not enough people to build everything. The London Olympics / Crossrail and the Thames Gateway projects in particular are going to soak up a lot of the UK workforce over the next 10 - 15 years. Contractors can virtually pick and chooose their contracts and current inflation in construction is easily running at this rate and in some sectors perhaps even higher.

(10): Not sure what these capital charges would be but it may be things like land purchase, dare I say toll booths??, and various capital expenditure to get the crossing running but which haven't been included in the tunnel costs. Just speculation.

Bear in mind that 3,4,6,7 and 9 would apply proportionately the same for the tunnel or the bridge so in terms of a comparative study they make no difference. Depending on what is in (10) it may be higher for one option over the other, I'm not sure.

The SNP councillors in Fife came out in favour of an tunnel in July 2005, almost 2 years before the TS reports were concluded. I can't quite see how they arrived at this conclusion without any expert reports to base their opinions on.

Q4 2006 is used as a base line for costs as it is the most recent point at which construction costs for actual projects which are on going have been confirmed so it provides a comparative baseline. I don't think it suggest that the route was fixed up before hand.

Norwegian tunnels are generally very different in form. They are bored through rock on dry land rather than in an excavated trench in a river and they have been doing them for so long that they probably have the process down to a fine art.

A fixed price contract would certainly seem to protect the taxpayer's interests. However, what would happen would be that if we asked contractors to take on board the risk by requesting such a contract, the contractors would price the risk associated with this project (higher for a tunnel) accordingly and the tender returns could well be much higher than budgeted. The government may have to hold on to some of the risk (ie delays due to unforeseen ground conditions) simply to get an affordable bid and to encourage the industry to compete for it.

It is difficult to compare the cost of the original bridge to now just by applying inflation. Health and safety legislation and environmental legislation in particular have increase through the roof in the last 40 years, with a knock on effect on construction prices. Also inflation in the construction industry over the past few years in particular has been running higher than general inflation levels. mind you so has everything (bills, council tax, petrol, the list is endless.

not sure your comparison to the Norwegian tunnels stack up directly as the tunnels are a different form here with different gorund conditions, different links into the existing infrastructure and different market forces/legislative requirements.Also the tunnel here would be an immersed tube for 4km rather than 3km and there would be a bored tunnel for an additional 3km on the south side as well as a deep cutting on the north side for at least 2km.

What you are absolutely right about is the need for clarity in the costing involved in this project. It doesn't help anyone if we are just given a bottom line figure and a 'trust us' pat on the head. We do need to understand what we are spending our money on. We all remember too well the cost issues at the bottom end of the Royal Mile.

If you want to see an alternative view to ForthTAGs have a look at some interesting points there for consideration.
Thanks Alasdair for the detailed answer. I will make a few short points.

The number of points a tunnel can start from is greater than a bridge since bridges are constrained pretty much to the shortest crossing & does not have to be a straight line as bridges are so I doubt if it is impossible to come up with a site 7 access route at least as convenient as for the bridge.

The Norwegians have built tunnels through all sorts of ground costing between £3.5 & £11 million per km. My estimate was based on their highest priced roads. See (11 kroner to the £ on the link). Your point about the Norwegians costs going down because of their long production run is right - this is why I have been saying that we should be building lots of tunnels.

I do think we should compare the costs of the original bridge with the current one (also with the cost of the Denmark/Germany bridge). I think you may well be right that the 8 fold above inflation rise is largely due to government regulation (council tax & indeed most things governmentbill us for, is also clearly rising far faster than inflation for no clearly defined technical reason). In which case this would be a very good argument for changing our regulatory structure. I do not know but I suspect the health & safety record in German & Danish construction is not appreciably worse than ours.

Technologically projects which are at the cutting edge tend to rise below inflation over time as the technology improves (electronics is obviously an extreme example) so that an 8 fold increase from non-technolgical factors may be an underestimate. Certainly it seems the rest of the world can manage these projects far more cheaply than us.
Looking elsewhere I have found that the Sydney Cross City Tunnel cost 680 million Aus$ (£300 million) for a 2.1 km structure. This seems closely in line with my extimate that a Forth tunnel should cost well under 1/10th what we are told.

Even at that price it lost money.
Apart from Alastair, who is the engineer out of the 4 members of StayonForth, a group basicly opposed to FTAG & its tunnel proposal, nobody in an official or party positiion has responded.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

British Blogs.