Whatever happened to the 1991 aspiration of a 95 minutes train journey between Belfast and Dublin? #20yearrule

Northern Ireland Railways and Iarnrod Eíreann set up a Joint Study Group in 1988 to consider the optimum investment policy for the cross border rail network. One option was chosen out of the eight examined and files (ENV/34/1/8 and ENV/34/1/9) [selective scans] released under the 30/20 Year Rule show how Coopers & Lybrand Deloitte won a tender in 1990 to write a more detailed report about that option.

It was envisaged that nine diesel trains a day would run at 145 kph with a reduction to the non-stop journey time to 1 hour 35 minutes.

More than 25 years later, the journey time is never less than 2 hours 5 minutes, though the current timetable makes four stops along the way between Belfast Central and Dublin Connolly (calling at Portadown, Newry, Dundalk and Drogheda).

Like so many bundles of government papers released under the 30/20 Year Rule, the final chapter of the story is missing.

The final version of the consultants’ report which was published at the end of July 1991 was not included in either of the two files of papers. There’s an empty brown envelope at the front of the ENV/34/1/9 file labelled ‘Final Report’ that probably once contained a copy that an official forgot to put back!

However, the question remains how it was envisaged in 1991 that a north-south 95 minute train service was possible when journey times are now half an hour or more longer?

Local train traffic around Dublin has greatly increased since 1991 is causes some contention on the shared track. All services stop at four halts between Belfast and Dublin, adding ten or more minutes to the journey.

I asked Translink why the 1990/1991 aspiration of a faster journey time had not been realised. A spokesperson said:

“The level of investment available for the Enterprise service during the 1990s enabled sections of the cross border line to be upgraded to speeds of 145km/hr. Most of these are south of the border where the topography offers more opportunity to attain higher speeds (i.e. longer sections of straight line).

“The twists and turns that the railway takes north of the border allows only more modest opportunities for full 145km/hr running on sections closer to Belfast.

“This, along with busy urban and suburban services into Belfast and Dublin, has required that the original journey time aspirations for the Enterprise be tempered to achieve the best overall balance of service capacity and reliability for rail customers as a whole.”

, , , , ,

  • james

    Perhaps the costs outweighed the benefits?

  • ted hagan

    Come on, let’s not be too naive.
    Coopers & Lybrand Deloitte are commissioned to write a report (for sweeties?) and they come up with the ‘correct’ answers. Astonishing!

  • Brendan Heading


    There are a few pressing issues preventing the timetabled service described.

    The obvious one is that this is the non-stop time. The two railway companies decided that the services should stop at the major stations; it means more passenger journeys can be taken with one train.

    Only four new Enterprise trainsets were purchased, rather than the seven envisaged above. Further, after the order was placed, the railway companies decided to “borrow” the fourth set to make the other three sets longer, to take more passengers.

    On top of this, there are a number of speed restrictions that have recently become necessary due to track deterioration at various spots between Lisburn and Newry. And further, the economic growth experienced in the south led to increased use of the DART. The need for the Enterprise to slow down behind DARTs between Malahide and Connolly adds several minutes to the journey time.

  • Richard Gadsden

    It’s one of those routes where you really want to do a big intervention, rather than a small one. You could probably straighten out a few curves, do some more maintenance, and save a few minutes, but you ideally want to switch to electric to improve acceleration (so the stops cost less time) and to be able to raise the maximum speed. But DART uses a low-speed electrification system (and you obviously can’t have two electrification systems on the same track) so you’d really want to bypass the whole DART segment.

    There isn’t another good approach route into Connolly station, so the obvious approach to this would be to build at Heuston station, rerouting traffic so the train could cross the Liffey bridge, and then go into a new tunnel in Phoenix Park (at the same location as the current Phoenix Park tunnel portal), then run the new tunnel up to a new station at Dublin Airport and then a new direct route to Drogheda.

    You could then electrify the existing line all the way from Drogheda to Belfast, straightening as appropriate along the route. This should give you about sixty minutes Dublin-Belfast with a stop at the airport as well as the existing ones.

    It would be very tempting at that point to run a line from Portadown to Omagh, and then alongside the Strule/Mourne/Foyle through Strabane to Derry. That could be built at 250 km/h or so, allowing for a much faster direct Derry-Belfast train and also moving the intercity traffic off the Belfast-Coleraine line, which would allow for a single stopping pattern and a more frequent service – as well as obviously permitting a direct Dublin-Derry.

    Ninety minutes Belfast-Dublin is an awkward ambition – it’s achievable without major work (but a lot of minor work), but would involve messing up a lot of DART journeys, and all the money spent would be wasted if you wanted to do anything better than 90.

  • Richard Gadsden

    90 would need to give priority to the Enterprise over the DARTs, ie the DART just not running for at least half an hour to clear the tracks for the Enterprise. Or you’d have to four-track a section to let the Enterprise overtake – which would cost the same sort of money as a new Dublin-Drogheda route with far less benefit.

    One simple rule is never to add tracks to an existing route, unless they used to be there in the first place, or it was built with them designed in to be added later – otherwise you will either have to close the existing route for construction, or take four or five times as long as a new route. Building a completely new route is almost always cheaper and has more benefit (e.g. you can serve new intermediate stops, such as Dublin Airport in my proposal).

  • lizmcneill

    The old right of way for the Derry train was all sold off when it was Beeching’d, wasn’t it?

  • Richard Gadsden

    You’d want a new one anyway – you want sustained 200-250 km/h all the way. The saving on land acquisition would disappear just in the extra rolling stock you need on a slower line.

  • hgreen

    Now you can work on the train is there any need for a 90min journey time?

  • William Kinmont

    How much time will customs checks add to the overall journey time.

  • AndyB

    I think it was always four trains – three in operation, one for spare, each doing three return journeys a day. The ninth journey was lost somewhere along the way so that we only have eight each way – one set does three return journeys, and the other two do 2.5 each so that locomotives can be changed in Dublin as required and all passenger carriages get two nights out of three being serviced in Belfast.

  • Brian O’Neill

    I know nothing about it but in the future could electric trains not be battery powered like electric cars?

  • AndyB

    Each station stop takes up to five minutes, including deceleration, station dwell time, and acceleration. Portadown and Drogheda are low impact stops due to the severe curvature through the former Portadown junction (still having the same radius as it would have had when the lines to Armagh and Dungannon were in place) and Drogheda station, both of which and the Boyne Bridge being restricted to 15mph.

    One of the key phrases is “selective realignments”, which at the time would have included the curves at Poyntzpass and Scarva.

    In the end, there was precisely one realignment, through Dundalk Central signal cabin, leaving the curves at Portadown, Scarva, Poyntzpass, and Drogheda in place.

    For some years after 1996, the 08:00 and 18:10 (as were) ex Belfast and 19:00 ex Dublin ran as expresses with one or no stops, but there was a lot of political pressure to add the extra stops – and in the end, Translink had to face down the opposition in Lisburn to withdrawing the station stops there.

    One of the key problems with the project is that Knockmore-Moira was omitted because it had been relaid in the 1980s. A horrendously short-sighted move, and the files may well show whether it fell by the wayside at the same time as Scarva and Poyntzpass (ie because the Department wouldn’t provide the extra money) or earlier, because that section remained limited to 70mph and only now is work being carried out to relay that section and increase the permitted line speed – and the line is showing its age.

  • Oggins

    Andy, you obviously know your stuff!

    Do you see any development in rail? Will we get a faster more modern service?

  • Oggins

    I believe so Hugh. I travel to Dublin 2/3 a week. I drive as to get a train and then taxi or bus to my locations would still take fair longer the car. If we had a service that got me there in 1.30 with another 30 minutes in a Dublin commute, I would definitely think hard in parking the car at home.

  • AndyB

    Only if the ratepayers are willing to pay for it! There’s nothing in it commercially, so the private sector won’t do it.

  • Oggins

    Would there ever or could there be a point in which it became commercially viable?

    Or would that require bigger strategy itself on public transport within Belfast and Dublin?

  • mickfealty

    In my experience, the slowest part of the journey is from Malahide in when it becomes funereal. Given the infrequency of the service, I’ve no idea why that should be.

  • mickfealty

    Interesting Richard. I’m guessing we’ll the costs will outweigh the benefits argument, even if (as you would surely have to) you included similar investments in the Cork line?

  • mickfealty

    Yep. This could be a Brexit ask (if we had any politicians with the nerve/nouse to do it). A bigger investment like Richard suggests would actually be transformational in terms north south economy.

  • Oggins

    Also trains that leave you in Dublin at acceptable work times!

  • AndyB

    It would need an Irish Rail strategy that didn’t involve quite so intensive a DART service – or the politically brave strategy of demolishing enough houses to quadruple the track between Fairview and Malahide so that the Enterprise and outer suburban services could cruise into Connolly. The Irish politicians are even worse than ours at understanding public transport, so the chances of that…!

  • ted hagan

    The Dart service begins and ends at Malahide; I suspect this is the reason.

  • AndyB

    @mickfealty:disqus precisely the Dart, which runs at 10 minute frequency and stops everywhere. There are gaps for the Enterprise, but they don’t even rise to the level of inadequate.

  • AndyB

    It’s probably more economical to put up catenary for continuous supply due to the amount of power needed to haul a train – the 201 class diesels develop 3200 hp, and 25kV locomotives in GB and elsewhere develop an awful lot more, and I think the benchmark tractive effort for mixed traffic is around the 200kN mark. Doesn’t mean it can’t be tried on batteries, obviously, but I wouldn’t want to bet on success.

  • AndyB

    Benson’d, technically!

  • AndyB

    Yes and no. The advantage of building additional tracks (as was done out of Heuston) is that you would have to carry out an awful lot less demolition, and they can be built on one or both sides independently of the existing lines – for example, I would want to put fast lines on the Down (Belfast-bound) side so they don’t have to cross Howth Junction.

    Apart from a few closures to replace bridges and to slew track onto new alignments (assuming that you couldn’t uniformly build the new line all on one side) it wouldn’t actually be that disruptive to normal traffic. The Heuston project didn’t take that long for a fairly similar distance.

    I also suspect that it would be difficult to get planning and compulsory purchase orders for a new route. It would be bad enough getting them to widen the existing formation, and there are a lot of vested interests around.

  • Abucs

    Could be worse. Just came back from the Socialist Republic of Sri Lanka where they are still using the 100 year old British built system. The Ella to Colombo train ride (130 miles) took 10 1/2 hours.

  • Brendan Heading


    This is all utterly pie in the sky and unjustifiable for an island with 8 million people living on it.

    Electrification of the Belfast-Dublin line would be extremely expensive. To get an idea of this, look at the project to electrify the line from London-Cardiff, costing just under £3bn. That’s roughly a 150 mile route. The Belfast-Dublin railway route is about 120 miles. You’re in the £2bn ballpark.

    Before you spend your £2bn you have to, as you noted, straighten out some of the curves so that trains can travel at the higher speeds. Tunnels or overpasses would have to be constructed to eliminate the level crossings between Lisburn and Newry (such as at Lisburn, Lurgan, and a few other sites). You need to replace the rolling stock.The Virgin Pendolino 9-car trains cost £22m for a set. You’d need several of these just to run the Enterprise and nothing else. The route would almost certainly have to be resignalled to allow for higher speed running with four-aspect signals along the entire route. Again this would require significant expense.

    After you’ve spent all this vast sum of money, all you’ve done is put Belfast 45 minutes closer to Dublin. You’ve also annoyed the residents of Cork and the route between there and Dublin who’ve seen their line relegated to second class status.

    Most of this is unnecessary; a lot can be achieved by spending comparatively little. Adding passing loops between Connolly and Malahide would allow signallers to move DARTs out of the way of the Enterprise. Adding more sets would permit an hourly service, and would permit a few non-stop services which should be able to make the journey within 1hr45min. You can, if you wish, connect Dublin Airport by running a spur from the area around Portmarnock directly over the undeveloped fields there.

    Upgrading the track to support 125mph running, which does not require electrification, would get you to the point where an express train could complete the journey within 1hr15min. Once you’re able to do that, what’s the point in spending £2bn on electrification ?

    On the second part – opening up the railway line to Derry via Omagh and Strabane is ridiculous and unjustifiable in any sense. So, again, let’s look at the costs of reopening an abandoned railway. The irish government recently reinstated the Clonsilla-Dunboyne section of the old Navan railway line at a cost of €108m, excluding rolling stock. That’s just under 5 miles, so around €21.6/mile.

    It’s around about 77 miles from Portadown to Derry via Omagh and Strabane. In UK money that’s about £1.5bn.

    There is no way you can justify spending £1.5bn, before you’ve even operated a single service, to connect rural communities with a train that will stop at every hole in the hedge along a 77 mile route. It isn’t going to happen.

  • Brendan Heading

    Connecting towns with a few tens of thousands people with a railway service that is faster than any other train in the UK or Ireland outside of High Speed 1 is thoroughly away with the fairies.

  • Brendan Heading

    Not in the short term, batteries can’t store energy at the required density.

    Ironically, Ireland pioneered the battery train in and around the WW2 period.

  • Brendan Heading

    ie the DART just not running for at least half an hour to clear the tracks for the Enterprise.

    I don’t know where you get this. It takes approximately ten minutes for the Enterprise to reach Malahide from Connolly on a good day.

    I’m not a civil engineer. Quadrupling would be the ideal (Irish Rail did this on the Heuston-Kildare route) but very difficult as the line is crowded by residential property all the way along. Passing loops, on the other hand, seem eminently doable in several places. Converting some of the DART stations to have an island platform in one or the other direction would allow signallers to shift DARTs out of the way when the Enterprise is approaching.

  • AndyB

    Yes, but it takes the DART at least 25 minutes to clear the section (15 minutes for a Howth train). That’s 15 or 25 minutes at least of no DARTs before the Enterprise can leave, and a further five minutes before the next DART can leave.

    Beyond the range of the DART, it was possible to loop trains at Skerries and Mosney. Skerries loop was goods only, and while Mosney loop was passed for passengers, it is now a bay only – one of the turnouts was removed because Irish Rail couldn’t financially justify its replacement.

    Ideally you’d have something like the Cork line where the platforms in the Up direction at a couple of stations are on loops so that non-stop trains can overtake, but it’s all a question of land take. Not insurmountable, just difficult.

    Incidentally, while the northbound Enterprise is not known for its speed on approach from Lisburn, it’s not bad, and compared to Malahide-Dublin and, to a lesser extent, Dublin-Malahide, it’s flying.

  • AndyB

    Drogheda station platforms were realigned slightly, come to think of it, but it was fairly minor and didn’t affect the speed restriction.

  • Andy – This is what I hoped – but did not expect – Translink to say when I asked them about the journey time!

  • AndyB

    Even at the heights of the Save our Railways campaign, there was only so much that Translink could say and do to bite the hand that feeds them!

    There is a lot of politics in this, and part of it is not winding up Irish Rail too much, not least because in reality Irish Rail set the timetable due to the limited slots available into Connolly. Your article about the 6.45 Enterprise last year did all of us a great service, and eventually that plan of Irish Rail’s was abandoned.
    Trains to Belfast tend to run better, because there is so much less congestion. The intensity of the service means that local trains between Portadown and Belfast have to leave more or less on time to be able to continue on time for the day, but once you clear Malahide the running tends to be a lot nicer, and of course local trains can be looped at Portadown and Lisburn.

  • Reader

    AndyB: Each station stop takes up to five minutes, including deceleration, station dwell time, and acceleration.
    On the Belfast to Bangor line the difference between the Express and the Stopper is 8 minutes=6 stops.
    Are the Armagh and Louth folks a bit slow on their feet?

  • Brendan Heading

    I remember them removing the middle road through Drogheda. I’ve a few pictures somewhere of it being done.

    I didn’t know Dundalk had been realigned, as it’s more or less a straight line through there ?

  • Richard Gadsden

    They would. My point is mostly that there is a limit on what is achievable without a big intervention, and it’s a pretty small limit.

    Obviously, you could get some more rolling stock, run the trains more frequently, and you could do maintenance work to improve reliability … but there isn’t that much you can do on speed cheaply

  • Richard Gadsden

    Yes, but leaving a train stationary for hours to recharge is very expensive, so you would need to have most of the route electric anyway.

  • AndyB

    Several reasons.

    Intercity trains typically have one or two minute station stops where local trains are scheduled for 30 seconds, simply due to the number of people boarding and alighting and the amount of their baggage.

    Loco-hauled trains take longer to accelerate in particular, and potentially also stop. They also reach full line speed between stops – something that, for example, on the Bangor line only happens between Holywood and Sydenham for stopping trains and is only 70mph against 90mph on the Dublin line.

  • AndyB

    There is a bit of a turn at the junction end of the station – enough for non-stop trains to slow considerably. Platform 4 was removed and the Down platform realigned to allow a smooth run through the site of Dundalk Central cabin (indeed, the reason why it was closed so quickly!)

    I can’t remember if the alignment of Platform 3 was changed, but both main lines were designed for 80mph running. Not that anything other than special trains continue straight through without stopping any more…

  • Gavin Crowley

    If at some point in the future the new Navan line is finished, and justified on its own merits, then a link to the Dublin-Belfast line (perhaps using portions of the other lines out of Navan) would provide an out-sized passing loop. That approach to Dublin is not without its own problems of course.
    There’s a 4km gap at Lucan between the Kildare line and the Maynooth line. Heuston could be the terminus instead?
    It’s nice to dream.