UNTIL now, the debate on project seems to have the Dublin Light Rail revolved around three issues the fact that when it was cut back to two routes rather than three, the one that was dropped was the line to Ballymun the temporary congestion that will be caused as streets are dug up to lay the track and the threatened demolition of a number of houses.
There are, however, even more fundamental issues that need to be addressed. First, will it be physically possible for a street level LRT to carry the estimated traffic volume on the proposed routes? Second, what will be the permanent impact (as distinct from the temporary impact during construction) of a street level LRT upon traffic in the city?
Third, should the determining factor in the choice of the operating system and route network be its net economic/social benefits, or should the sole consideration be the financial contribution sought and secured for it from the EU, plus the amount of Exchequer funding that was allocated to the project at a time some years ago when the issues involved had not been fully explored? In this article I shall address the capacity issue.
For any route system the choice of transport mode must depend upon the traffic to be carried. In the case of commuter services, the determining factor is the volume of demand in the peak morning hour for morning city bound traffic is more concentrated than evening outbound traffic.
The amount of traffic that a given system can carry depends upon three factors first, the maximum frequency that can be operated safely and, in the case of on street vehicles, without causing a snarl up of other public and private traffic with which it is competing for road space second, the theoretical capacity of the vehicle and, third, the proportion of that capacity that can in practice be filled under peak hour conditions.
THE LUAS leaflet issued by CIE proposes LRT vehicles with a theoretical capacity of 200 passengers (60 seated and 140 standing) and a peak interval of six minutes between services. Rather curiously, CIE calculates the theoretical capacity of such a frequency at 2,400 per hour although a six minute interval would in fact entail only 10 vehicles operating per hour, yielding a maximum hourly capacity of 2,000.
No provision has, however, been made for the fact that no transport system achieves 100 per cent utilisation of theoretical capacity even in the peak, especially on vehicles where most of the theoretical capacity consists of standing room. Except in Japan, where staff are employed at stations to push passengers physically into overcrowded trains, commuters are simply not prepared to tolerate the kind of density that would be involved in achieving the theoretical capacity they prefer to travel at something like half the theoretical standing room density and, accordingly, wait for the next vehicle.
Thus, on the DART service, 10 trains, each double sets of four carriages with a total theoretical capacity of 1,000, operate from Bray to the city in the morning peak hour (07.30-08.30). However, a census carried out by Iarnrod Eireann last November showed that these 10 trains carried only 4,860 passengers on the heaviest loaded sector, from Sandymount to Lansdowne Road just under 49 per cent of the theoretical capacity.
In the other direction, nine trains carried 4,770 passengers from Killester to Connolly Station just over 55 per cent of theoretical capacity and the most crowded train of all achieved only 82.5 per cent of theoretical capacity. Because an occupancy rate of 55 per cent in the peak hour is regarded by Suburban Rail as an unrealistically high figure, one of their four reserve sets is now being added to a peak hour train to ease this pressure.
It is possible, however, that an LRT service might do better than "this. Peak occupancy in Vienna is 70 per cent and Sheffield has given me an estimated peak capacity figure of 65 per cent. If, despite the DART experience of Irish passengers low toleration of overcrowding, we assume that LUAS could achieve a 70 per cent loading in the morning rush hour, that means that the peak frequency proposed by CIE would carry 1,400 passengers.
HOW does this compare with the estimated demand? On a service starting, as is currently proposed, just south of Dundrum, the morning peak hour in the year 2001 from Ranelagh inward was estimated some years ago by the Dublin Transport Initiative at 1,871. But this seems certain to be an under estimate, for as The Irish Times reported some weeks ago, a traffic count on a road into the city has recently recorded traffic already running at the level that the Dublin Transportation Office had been predicting for the year 2001. Moreover, the development of housing in the hinterland of the Dundrum LUAS seems to have been running, and to be likely to continue to run, well ahead of earlier expectations.
Against that background, it would clearly be most unwise not to provide for a morning peak flow significantly in excess of 2,000 in the opening year of the service rising to a flow in excess of 2,500 some years later. And that would require from the outset a four minute peak frequency. A minimum of 36 vehicles would for this purpose which is well in excess of the figure of 25 and 30 vehicles mentioned at different points in the DTI analysis. This has implications for the economics of the on street LRT proposal.
But there is another crucial factor that has not hitherto been high lighted. The DTI projections show that if and when the service is extended to Cabinteely, as was originally proposed, the morning traffic will be increased by over 50 per cent and the revenue on this route by about 80 per cent.
The DTI estimates show that this would bring the morning peak hour traffic flow up to 3,500 by the yearn 2011 which, given the way Dublin traffic already appears to be running ahead of these estimates, could well mean a flow in excess of 4,000.
To carry that kind of traffic at a 70 per cent occupancy rate on an on street LRT system would require a peak frequency of 30 plus vehicles per hour which would mean that every cross street would be blocked to other traffic at average intervals of less than one minute. In practice, allowing the necessary time to halt traffic to give priority to the LRT vehicles, all traffic including buses from Harcourt Street to O'Connell Street and across to Heuston Station and on to Inchicore would as a result be almost permanently jammed at peak hours.
This is why an unsegregated on street LRT operating without under passes at junctions cannot carry more than 2,500-3,250 passengers per peak hour. Peak flows of over 3,000 an hour are attainable in continental cities with vehicles operating on street on wide boulevards, but the lower figure is more realistic for a city with narrow centre city streets like the Nassau Street College Green bottleneck and the generally narrow streets running north from the quays between O'Connell Street and Heuston Station.
Under such congested conditions, an LRT cannot operate at intervals of less than 3.5 minutes which leaves an average interval of less than one minute 45 seconds between vehicles in the two directions. Such a minimum interval implies a maximum of 17 vehicles per hour, with an effective peak hour capacity of less than 2,500 passengers an hour.
It seems to me, therefore, that on capacity grounds alone an on street LRT system for Dublin is a grossly inadequate, and indeed counter productive concept in terms of traffic management.
