Transport Watch UK Focusing on UK's Traffic & Traffic Systems

Facts sheet No 12: Rail to road, costs of conversion and rates of return

Updated August 2012
Price base updated
Links to archive and spread sheet added


This facts sheet deals with the costs of conversion and touches briefly on conversion strategy.

The data and analysis shows that (a) the first year rates of return from railway conversion would be overwhelming and (b) paving the entire national rail network would cost perhaps £19 billion at 2010 prices, a fraction of the £63 billion - £88 billion forthe  rail modernisation programme.


A contract let by the then Department of the Environment (Contact DG 466/3) culminated in the Hall Smith report "Better Use of Railways". The second edition was published in 1976 (ISBN 0 7049 0349 0). It attracted vitriolic criticism, which was rebutted in the companion volume, Comments and Rejoinders, (ISBN 0 7049 0473 x). There the critics appear, for the most part and at the politest, in a very poor light. In any event they never responded.

The main report examined the potential for converting six railway lines to motor roads in and around London. The conclusion was that conversion would yield first year rates of return in the range 18% to 530% (except in one case where the return was infinite because the value of scrap including surplus land exceeded the cost of the conversion). Those very high rates of return are hardly surprising in view of our analyses, which show rail 3 to 4 times as expensive as road while being used to a fraction of its potential, if paved.

The main report may be updated by factoring the time and accidents savings according to the ratio of today's unit values to those used in the study and by factoring the construction costs by the Road Construction Prices index. That analysis showed benefits would rise in real terms but that costs would fall, so strengthening the conclusions.

Comments and Rejoinders provides a useful cost summary. It compares the all-in study costs of conversion with 5 actual schemes where all prices are per sq. metre and are at the price base 1973 base. The data is as follows:

Actual Conversions: Edinburgh £10.95, Southport: £3.73, Radnor CC 1970: £2.49, Radnor CC 1969: £2.94, Radnor CC 1968: £3.35 (Motorway Construction standards would add about £3 per sq. metre to these costs. In comparison study estimates, including all ancillary works, had the range £8.21 to £19.40 per sq. metre. That implies that the study estimates were, if anything, too high

Converting the study estimates to 2010 equivalents using the Road Construction Tender Price Index, RCTPI, provides £(67-160) per sq. metre and the mid range value of £113.5. see data on conversion costs here where there are two pages, the first dealing with historic costs, the second providing the Hall smith comparisons.

In comparison the Highway Agency puts the cost of a new six-lane motorway complete with hard shoulders and central reserve at £29 million per mile.  That includes VAT at 17.5% and optimism bias at 45% (a factor induced by the gross underestimates for railways e.g.. the West Coast Main Line Modernisation Programme). Stripping out the VAT and optimism bias and setting the overall width to 9 lanes each 3.65 metres wide yields a cost of £250 per sq m.  Of that 21.5% is earthworks and 46% structures and 12% land, Earthworks on converted lines would be minimal and new structures relatively few. Little or no new land would be required. Removing those items provides a cost of £92 per sq metre.

Against that background we here a apply £120 per sq m to Network Rail’s 32,000 km of track and assume a generous 5-metre lane width.  That yields a total of £19 billion, a cost that is dwarfed by the rail modernization programme.  It had the range £50 billion to £70 billion at 2000 prices, which, after applying the GDP deflator, converts to £63 billion to £88 billion at 2010 prices.


As to conversion strategy, first remember that they can lay a mile of road surfacing in a day.

The strategy would be to stockpile materials and plant at 5 to 10 mile intervals for a particular route. Then on Death of Rail Day the tracks would off and replaced by a road surface sufficient for buses in a matter of weeks, at least to the edges of town and city. During that period special traffic orders would be in place to enable buses, previously procured etc. to operate reasonably in urban areas.


If paved the rights of way would be used 3 to 10 times as intensively as are the railways. All London's crushed peak hour (surface) rail commuters would have seats in express coaches at a fraction the cost of the train. Those coaches would use only one fifth to one seventh of the capacity available. Hence, many thousands of lorries and other vehicles would be able to transfer from the unsuitable city streets that they currently clog. (See facts sheets 1, 7, 8, and 9). Additionally many thousands of hectares of derelict railway land would become open to development and hence intensely valuable.


© Transport Watch UK 2011 | Webdesign by 1PCS