Archive>Electrification

Updated Electrification Review makes the case

Roger Ford

On 11 February 1981, Norman Fowler, then Margaret Thatcher's Secretary of State for Transport, published a Report on main line electrification which, when re-read against the background of today's approach to transport policy, is breathtaking in its ambition, painstaking in its detail. And the case it made is even more compelling.

‘Review of main line electrification' had its genesis in a White Paper ‘The role of British Rail in public transport', published early in 1978. This included the decision to review the general case for more main line electrification in the light of the prospects for future energy supplies.

Bill Rodgers, the Labour Transport Secretary and BR Chairman Sir Peter Parker formally commissioned the Review in May 1978. It was to be carried out by a joint Department of Transport (DTp)/BR Steering Group.

Joint Chairmen of the Steering Group were BR's Vice Chairman Rail David Bowick and, for DTp, the Under Secretary Rail John Palmer. The latter appointment was significant.

Compared to John Palmer, all his successors I have known at DTp and now DfT have seemed middleweights. He was ever the scourge of woolly thinking and a relentless protector of the taxpayer against a feckless British Railways Board.

Rigorous

With John Palmer as co-chairman, and with representatives from the Treasury and the Department of Energy attending Steering Group meetings, the Review was going to be a rigorous examination of BR's aspirations. But by September 1979, when an Interim Report was published, the Government had changed.

Even so, Norman Fowler, the Conservative's new Minister of Transport, welcomed the report, saying ‘There are encouraging signs that a case for further railway electrification can be demonstrated'. Calling for further work he emphasised ‘It is very important that a project which involves such a very large commitment of resources should be thoroughly analysed'.

This thorough analysis occupied another 17 months. Remember that this was before the days of powerful desk top computers, which meant that the extensive modelling involved took time.

Ambitious

But the conclusions were ambitious and unambiguous. Michael Posner had replaced David Bowick in January 1980 and at the press briefing to mark publication of the final report he told us that the best chapter was the introductory submission signed by himself and John Palmer. Agreeing this had occupied his colleagues on the steering group ‘for several man years'.

And when you see what Mr Palmer had put his name to, this was not surprising. The main conclusion read ‘On the assumptions made a substantial programme of main line electrification would be financially worthwhile. All the larger electrification options examined show an internal real rate of return of 11%; the faster options give the highest net present values'.

And for those who peddle the tired old story of a railway in historic decline, the Steering Group had assumed that ‘these businesses (InterCity and Freight) will continue to win traffic and to flourish in a way that will justify substantial further investment - more than the Railways Board is investing in them at present'.

How much was ‘substantial'? For the largest, fastest programme taking 20 years, the net cost would have been around £3 billion at current prices. The overhead line would have stretched from Penzance to Aberdeen .

By the time the review was complete, Mr Fowler had been promoted to Secretary of State for Transport. The economy was in crisis with inflation worsening and the national debt rising. Not surprising, then, that his response this time was more cautious.

‘Several broad issues' presented in the report would need to be ‘considered'. Electrification would require ‘substantial extra investment' – primarily benefiting the commercial InterCity and Freight businesses. It would also ‘make cash-flow demands well into the 1990s'.

But having put the downside in the context of the current economic situation, Mr Fowler admitted that ‘in economic terms the report presents a favourable assessment'. He would have to ‘consider carefully, the prospects of BRB's commercial businesses and how the funds might be generated.

But the economy did recover, and with it BR's fortunes. Both the MML suburban and ECML electrification, went ahead. Subsequently, in the balkanised post-privatisation railway, there was no way to extract the benefits from investment in electric modern traction.

This magazine was the first to appreciate that diesel traction would dominate the franchised railway. And so it turned out, even to the extent of Intercity East Coast operator GNER leading a de-electrification campaign. Recently Network Rail Chief Executive John Armitt was reported as favouring a diesel rather than electric traction for Greater Western on the grounds of a quieter life.

Energy

But the world turns. The primary reason for the 1981 Review was concern over future energy supplies. As BR Chairman Sir Peter Parker put it at the time, the railway would be arguing for its share of North Sea oil revenues to fund an oil-substitution strategy based on electrification.

‘If the aim is to provide a railway which would give an economic transport system in the year 2000 and beyond, work to convert the nations main public bulk-transport system to electric power by that time must start now', said Sir Peter. Well here we are in 2006, North Sea oil and gas is running down and little has been achieved.

But, electrification is back on the agenda and the rest of this article will update the results of that 1981 Review to 2006-07 prices. Is there a case for a rolling programme of main line electrification in the 21 st Century?

Options

In final report the Steering Group had refined their proposals to three options numbered II, III and V. The base case – Option I – was the existing network including the various electrification schemes in hand or planned to have started by 1981.

Table 1

First full year of electric services assuming 1981 start of work

Table 1 shows the main lines to be electrified within each option, plus the timescale. Note that the commercial impact of the rate at which electrification was carried out had also been studied for options III and V. The suffixes ‘S' and ‘F' indicate slow and fast

Not shown in this table is electrification of a number routes for freight traffic. These included Doncaster-Grimsby, Didcot-Coventry and Manchester-Sheffield via the Hope Valley .

Modelling

This programme was the basis for four mathematical models. Separate passenger and freight traffic models determined diesel and electric train mileages which fed into a fleet assessment model. Among other things this calculated the traction and rolling stock build programme needed.

Then, the two traffic models and the fleet assessment model were combined in the costing model which calculated the annual costs and revenues for each option over the assessment period to 2013. The Net Present Values of these annual costs, discounted back to 1979 at 7%, formed the basis for the evaluation.

To get an idea of the complexity, BR at the time had over 300 passenger profit centres, each representing an identified service. The review consolidated these into 133 ‘working units' for each of which all traffic levels, revenues, operations and traction and rolling stock requirements were determined.

In projecting ridership forward assumptions were made on three factors: exogenous effects, journey time elasticity (Q) and fares elasticity (Table 2). The exogenous trend quantified the impact on ridership of the economy, assuming fares and quality of service remained unchanged. Experience of the past decade suggests that the upper range was pessimistic.

Table 2 Passenger Traffic and revenue assumptions

Journey time elasticity gives the effect of shorter journey times on ridership. The standard Intercity figure of -0.85 means that for a 1% reduction in journey time, ridership should increase by 0.85%.

Finally, fares elasticity, is similar to Q. An elasticity of 0.65 means that a 1% increase in fares would see ridership fall by 0.65%.

Fares policy in the model was based on an annual 1% a year, reflecting BR's business targets and improving quality of service. Following the introduction of electric services there would be a once and for all 6% increase spread over three years. A further phased 6% increase followed the replacement of locomotive hauled services with high speed trains.

Table 3

Passenger traffic levels in Option I

Table 3 shows the ridership predicted for the base option. I have shown only the lower and upper ranges. ‘Lower' combines lower exogenous growth, lower journey time elasticity and the upper fares elasticity. ‘Upper', the most optimistic combination. combines high growth, high Q and low fares elasticity.

Shown in italics are the real world figures for 1989/89 and 2004/05. Note how ‘other' – London & the Southeast and Regional services have outperformed in the ‘Upper' case, while InterCity has lagged.

Freight

If anything, modelling freight, which estimated train mileages and traction requirements for each option, was the more-complex exercise because individual traffic flows had to be analysed. The different performance of diesel and electric traction, as reflected in running times and train loads, also had to be modelled for each route.

Table 4 shows the forecast tonnages. these applied to all options, since it was assumed that the greater speed and reliability of electric traction would not generate more business.

Today these forecasts seem wildly optimistic. But remember that the ‘dash to gas' in power generation had yet to start, the Selby coalfield was still new and BR still thought that Speedlink would make wagonload freight viable.

Table 4

Freight traffic forecasts (millions of tonnes lifted per year)

Comparisons

All the options were compared against the Base case, essentially the 1979 railway projected forward to 2013. Table 5 shows the annual revenues and costs of the Option 1 railway, discounted at 7% to give a Net Present Value (NPV). All the tables in this article are at 2006-07 prices.

Table 5

NPV of Option I (base case)

Fig 1 (Phil has a better version)

Figure 1 compares the electrification options with Option I in terms of the changes in NPV. The stacked bars show the cost savings or increases in each category, the single grey bar the net improvement in NPV. Note that I have combined the changes in fuel and electricity charges into a single figure for ‘energy costs'.

This figure shows why even John Palmer was willing to put his name to that optimistic submission by the co-chairmen. It would take an unlikely combination of adverse factors', they wrote, ‘to undermine entirely the prospect that a programme of main line electrification would … earn a return of at least 7%. But the outcome could be better than 11% ‘if favourable chances combine'.

Testing

Looking 35 years ahead, the one thing about which the modellers could be certain was that Harold MacMillan's famous ‘events' could invalidate the assumptions on which this optimistic view of a future electrified railway was predicated.Thus, each of the options was subjected to a number of ‘sensitivity tests'.

In essence these asked the question ‘what if this assumption is too optimistic or pessimistic'? In the case of energy, for example, there where high, standard and low forecasts for future diesel oil and electricity price.

Fig 2 shows that the forecasters at the Department for Energy – looking ahead 20 years, were remarkably accurate. Yes, they missed the 1980 oil crisis, but in the long run, which is what counts, they were within 9% of the actual price of a barrel of oil. And while electricity costs fell more slowly than expected, by 2000 the price was midway between their low and medium forecasts.

Six combinations of oil and electricity prices were used for sensitivity tests. The biggest gain in NPV (£338 million) came from high oil prices combined with the standard electricity price forecast. At the other extreme, low oil prices and standard electricity prices reduced the NPV by £338 million.

Figure 1 does not show the current rising trend in electricity costs, but oil is now at $60/barrel. It is unlikely that it will fall back to the 2000 level. The sensitivity test based on high prices for both electricity and oil increased the NPV by around £45 million

Traffic

Another sensitivity test looked at the effect of the differing traffic forecasts already mentioned. Table 6 shows the result, and note that none of the tests included the lower freight estimate.

Table 6

* Option not modelled

Complicating hindsight is the fact that freight was expressed in tonnes lifted rather than tonne miles. Today, tonnage has fallen, but the changing market has seen tonne miles increase and tonne miles is what generates revenue.

Overall, one might extrapolate from Table 6, that upper passenger/standard freight would improve the NPV of Option V Fast by around £450 million.

Other sensitivity tests looked at the effects of:

• reducing off-peak service frequency
• traction and rolling stock maintenance costs
• Fixed equipment maintenance costs
• Staff costs
• Train crew costs
• traction and rolling stock and fixed works capital costs.

All these tests are summarised in Table 7. None of the tests generating a negative NPV is a show stopper.

Table 7

Notes

a) The NPV falls by a further £58-£77 million if reductions in service levels are assumed.

b) This is a rough estimate of the effect of assuming a lower trend in freight carryings

Electrification costs

Electrification costs in the Review were based on the St Pancras-Bedford Midland suburban scheme with an allowance for potential economies of scale. Fig 3 combines long term data from BR with costs from the Review itself and Network Rail .

There are clearly minor anomalies but the trend is clear. The BR figures represent the cost of electrification fixed equipment, less the power supply points from the grid.

I have used official Network Rail data to replicate this definition for the Crewe-Kidsgrove scheme. The resulting figure is clearly low, but not unrealistically so.

If structural work for clearances and signalling immunisation is included, there is good correlation with the Review figure of £220,000-280,000 per single track km. This spread recognises the differing costs of signalling immunisation.

ECML cost £215,000 per single track km, including civil engineering works and signalling immunisation. Currently Network Rail electrification projects are coming in at between £200,000-£260,000k per single track km for the OLE plus associated civils works.

While current electrification costs are within the range assumed in the Review, the hardware includes a number of design improvements. There is not space for a brief history of OLE, but the lesson of the ECML electrification was that long term maintenance costs were mortgaged for fast, cheap installation.

Thus Crewe-Kidsgrove uses copper, rather than aluminium, for headspan and catenary wires and portal structures are preferred to headspans wherever possible. If, on a short route like Crewe-Kidsgove, Network Rail can get a better quality OLE for around the same cost as the ECML, costs could be reduced significantly on a main line scheme through economies of scale.

Table 8

Overall, a quarter of a century on, the case made in that 1981 ‘Review of main line electrification' is, if anything, stronger than ever. What we lack is the conviction of railwaymen and civil servants which drove the original document.

In today's political environment, where political attention spans can be measured in months, proposing a £3 billion rolling programme of main line electrification would be naïve. But, with the Great Western Main Line, last upgraded 30 years ago, now due for modernisation later this decade, including electrification to Bristol and Swansea would add roughly £250 million to the cost.

It is time for the railway industry to revive the spirit of '81.

Return to Archive