At long last we have Transport Scotland's 25-year plan to electrify Scotland's railway. The wait has been worth it. By 2035 the lines from the Central Belt to Inverness and Aberdeen (including the Fife Circle), the GSWR line to Kilmarnock and Gretna, and the Borders Railway will all have been electrified. Three more discrete lines - Inverness to Aberdeen, Inverness to Tain and Ayr to Girvan are to be electrified thereafter. The Far North Line, the Kyle Line, the West Highland Lines to Fort William, Mallaig and Oban, and the Girvan to Stranraer line will remain unelectrified - no great surprise.

It's immediately clear from the Plan that Network Rail has been intimately involved in its preparation - a fact that could not have been said about electrification plans drawn up by the DfT for the Great Western. Indeed even in the Minister's Introduction (often a rather bland part of any government publication) we see these words. "This plan builds on...the recent devolution - at my request - of NR's planning and development functions to Scotland." This is surely the first step in bringing total control, including legal accountability, north of the Border.

TS is clear that where electrification will not happen there are only two other technologies: battery and hydrogen. Much is made of "discontinuous electrification" throughout the Plan, implying that an imaginative use of electrification is at least under consideration. It's suggested that "[battery-operated trains are] effective for lower-intensity services of up to around 55 miles". This is fascinating in the context of the FNL. Electrification as far as Tain is to be carried out soon after 2035. From Tain to Thurso is 110 miles. Were a battery-powered train fully charged on leaving Tain it would be able to reach Brora (45 miles) without running dangerously low. Were the FNL electrified from Brora to Helmsdale, and were the battery able fully to recharge over those 11 miles, the train could then run to Thurso (assuming a Halkirk chord), 54 miles away. Tight, but if battery technology improves in the intervening 15 years, entirely realistic. Is this what is meant by "discontinuous electrification"? Other routes have similar possibilities, allowing battery bi-modes for the whole of Scotland. Maintenance and other costs will be reduced if only two power sources are involved. It may be, of course, that hydrogen-powered trains offer a better outcome, but there is a problem of the size of the necessary hydrogen tanks for longer distances. A technical box advises that a hydrogen train is significantly less efficient than a battery train, which would seem to suggest that the ongoing running costs would, other things being equal, favour the battery option (see below).

In CP4 and CP5 TS committed to electrifying 100 stk a year. In other words 50 km of a double-track railway or 100 km of single track. This commitment was enough to carry out all the electrification now complete in the Central Belt, but it is only in this document that we find the commitment - as distinct from the intention - to continue. "We aim to electrify, on average, 130 stk per year in order to achieve our 2035 target."

It is clearly the intention to deliver the passenger benefits conferred by the elimination of diesel traction as swiftly as possible. Bi-mode diesel and electric trains will operate as an interim measure between now and 2035 - quite what these will be is not yet clear. There are many DMU classes which could be temporarily adapted to carry us though the next 15 years by the addition of an electrically-powered unit. It might be a locomotive or it might be carriages with a pantograph. The simplest is the former, and has been used for many years with, for example, Class 390s being loco-hauled by diesel to Holyhead.

The Plan notes that Class 156s (which operate on the WHL) and Class 158s (FNL and Kyle) have end of life dates "between 2025 and 2035", though which will last as long as the next 15 years is not spelt out. The following paragraph, somewhat dispiritingly, concerns itself with refurbished stock as well as new stock: it would be encouraging to think that those long straggly bits which will not see electric trains will see spanking new trains whose motive power is equally spanking new.

TS has managed to avoid the biggest pitfall of all, and one into which the DfT (and its predecessor departments) almost seems to have made its home: doing it all in fits and starts. FNE 80 carried a stark illustration of this. In contrast "we must have a rolling programme that provides a constant, sustainable design and delivery work-bank, both for NR and for contractors ... In addition to economies of scale, that will provide the supply chain with the confidence to invest in research, resource and talent thus supporting continuous efficiency improvement ..."

Here are a few more direct quotes:

• "Our aim is for a continuous electrified line of route from Edinburgh through Fife to Dundee and Aberdeen";
• "Based on fleet expiry dates, and rolling stock interworking, the FNL, WHL and Kyle Line are considered appropriate for the early introduction of alternative traction technology as a permanent solution" (emphasis added);
• "By 2035 we plan to have fully electrified routes from the Central Belt to both Aberdeen and Inverness East";
• "Though our plans are to electrify the entire route [Inverness East to Aberdeen] ..." the Plan talks of a lower priority for this than the other work to be completed by 2035. However there is an expectation "that the route, with an extension to Tain, will be fully electrified in the years shortly after 2035".

The Plan, to use a familiar expression, is the settled will of TS, of NR, of Scottish Ministers and by implication of ScotRail. FoFNL, had it had a magic wand to wave benignly over Scotland's railway, could not realistically have asked for more ...

Doubtless the Lentran Loop and other enhancements will appear in a different document soon.

This was all Pandora wanted to say on the subject until his attention was drawn to an article in Railway Gazette which puts much more flesh on the technical box mentioned above. This reviews a German paper which concludes that "hydrogen-powered multiple-units are more expensive [than battery-powered EMUs] and less efficient, being 35% more expensive to buy, operate and maintain". Hydrogen must be generated by electrolysis with an efficiency of less than 80%; it then has to be converted back into electricity for traction with efficiency less than 70%. Thus the energy costs "significantly exceed those of a train taking power from overhead wires or batteries" [emphasis added]. The review goes on to point out that the present hydrogen-cell technology would require these to be replaced seven times during the life of the vehicle, and while the life of a cell would be expected to be increased with technical developments, that's still a lot of replacement.

Pandora is drawn more and more to the idea of discrete sections of overhead electrification as suggested above. The great merit of hydrogen, we have been led to believe, is that a train can go a lot further without recharging or refuelling. But if that comes at a daily running and maintenance cost maybe one third higher than a battery train then the initial capital cost of doing a bit more electrifying here and there may be a small price to pay.