truth is life - Clearly you have no knowledge of geotechnics, one of the first rules of which is "You will have no idea on what the ground actually is."
There is therefore no point being numerically accurate as the numbers are guess work and your making assumptions on ground properties based on tiny samples. Uncertainty tends to scare academics, especially when you ask them to produce something that will actually happen rather than just an abstract model with no useful application.
Ciryandor - When the client says "Much of our current asset base is ~150 years old, still in critical use and survives with bugger all maintenance, give us the same for the new stuff." I have yet to hear a decent argument why they're wrong.
If you ever got your nose out of that spreadsheet and looked at the reality of the matter you would know that something designed to need maintenance is in effect being designed to fail. When that thing is a) underground and b) critical infrastructure the consequences of failure are unpredictable but usually severe and almost always very expensive.
Big risks and a lifetime of operational costs just to save minimal up front capital costs sound a bad deal to me, but then I can't produce a contrived spreadsheet that assumes the world is simple and that unpredictable failure events can be reduced to a number.
As many before you, you under-estimate British computing. The hardware is easy, take the code-breaking Colossus or the work at Manchester Uni, while on the industry side don't forget the world's first commercial computer was British (LEO I) and was just the first of many. And that was with an essentially broke country that was suffering under a Labour government. Imagine what they could have achieved with money and a competent govt. not nationalising everything and taxing them till they bled?