New approach wanted for traffic and public transport analysis
Page 62
Page 63
If you've noticed an error in this article please click here to report it so we can fix it.
PAPER SUMMARIZED BY DEREK MOSES
A NEW APPROACH to the analysis of shortterm public transport requirements, starting from first principles, completely ignoring any preconceived ideas on routeing, and employing the latest techniques for undertaking a traffic survey, was suggested at the annual conference of the Municipal Passenger Transport Association in Edinburgh yesterday.
Such a survey would be a valuable contribution towards the solution of practical operational problems, it was claimed. The suggestion was made by Mr. P. K. McIlroy and Mr. A. C. Dick, both of Freeman, Fox, Wilbur Smith and Associates, in a jointly prepared paper entitled -The role of public transport in the analysis of traffic studies".
Although much ground already familiar to public transport executives and officers, not to mention regular readers of Commercial Motor, was covered, new approaches to the problem were, in fact, discussed. The paper also included a description of the results of a transport study in East Central Scotland, employing a new computer program entitled -Transitnet".
Introducing their paper, the authors referred to the main objectives which should be part of any study of future urban transport planning and described some factors which should be taken into consideration. For exam ple, a particular road or bus route might affect travel patterns considerable distances away. Thus individual roads and routes needed to be considered not in isolation, but as part of a much wider transport system covering at least the whole of a town or conurbation.
The general planning process fell into three parts: (a) The design of alternative systems.
(b) A forecast of the usage of each system.
(c) The choice of the best system.
Design in this context included the physical design of road and public transport systems, the routes they would use, the frequency of public transport services, and policies for controlling the system and charging for its use.
Forecasting traffic use
Forecasting the usage of the system meant predicting how many people would go between each pair of traffic zones in the system by which mode, by which route, at what time of day, and for what journey purpose. (A traffic zone was defined by the authors as small zones into which the study area was divided such that, without significant loss of accuracy, it could be assumed that any journey to or from a zone started or finished at a single point within the zone known as the -zone centroid".)
In its present stage of development the forecasting process was based on the use of a mathematical model which had three main requirements: (a) Data on land use and transport networks for the design year; (b) Information on typical weekday travel habits, usually found by examining current travel patterns in the study area in relation to the existing distribution of population and employment; and (c) The suite of computer programs.
(A mathematical model is a series of mathematical formulae expressing the actions and reactions of relevant factors in such a manner that realistic travel patterns can be synthetized or estimated from given assumptions about these factors.) The experience gained during the last 10 years had given the transport planner a range of mathematical techniques and computer programs for carrying out most of the calculations needed in the forecasting process. In the early studies, the authors remarked, public transport was scarcely considered at all; in the latest studies it played a very important part, where it was represented by such programs as 'Transitnet", which took into account bus and train schedules and speeds and passenger walking and waiting time.
The input to this program was a description of public transport routes and an expression of public transport demand; the output included the number of trips in an average weekday on each link of the network, the total number of boardings, and the total passenger-time and passenger-miles on the system.
Public transport networks differed fundamentally from road networks by comprising a system of fixed services following given routes, generally at fixed intervals. This was in contrast to private road transport which generally had a free choice of route and time for the journey. Thus a different technique was required for the assignment of public transport.
Operation explained
At this point the authors explained the operation of Transitnet. This used a network built up by the description of individual public transport services and a special technique was used to determine the shortest path between any pair of zone centroids in the network. The criterion for deciding on the path was normally that of minimum time, which included both link travelling times and waiting times at points of boarding, together with walking times at each end of the journey and at transfer points.
If required, additional artificial time penalties could be added any time a new service was boarded in order to simulate preference for taking a slightly longer path to avoid a change. Transitnet then permitted the public transport demand to be allocated to the network producing: (a) Network link loadings in each direction.
(b) Total number of passengers carried by the system per day.
(c) Total passenger hours spent changing, travelling and waiting.
(d) Number of passengers boarding at each station or bus stop and total number of passenger hours waiting at each node.
le) Total passenger miles on the system in a day.
If) Total number of passenger hours spent walking.
Each of these factors played a part in the evaluation of a system in comparison with others tested.
With information such as that derived from the assignment process, calculations could be made of the cost /benefit returns for whole systems and parts of systems, allowing comparison to be made between one scheme and another—this could lead to recommendations on locations of new routes and services and on priorities for implementation.
Scottish study
The model described was fairly conventional and had been used in several studies of large urban areas and regions. It had been used in the Study of East Central Scotland, an area of over 1,000 square miles centred upon Edinburgh, in which the great bulk of public transport was at present by bus.
The present-day system was composed of successive amalgamations of a number of individual companies running on traditional routes, with the biggest concentration in the central part of Edinburgh. Forecast for future travel demand in the region (which assumed that measures would be taken to ensure that the degree of congestion was no worse 20 years from now than it was today) showed that while the proportion of all trips made by public transport would be reduced considerably, the total number of trips made by public transport would remain roughly constant. This was a generalized conclusion covering the whole region, and in the case of Edinburgh it was thought likely that under these conditions the usage of public transport would continue to drop slightly.
Flexibility
A number of modern developments in pub lic transport had been considered for their potential use in East Central Scotland, but none of them appeared to have the great advantage of a bus system with its inherent flexibility of response to demand.
Transitnet had been used to estimate the effectiveness of two contrasting future bus systems. The first was merely an adaption of the present-day system in which the existing routes were extended into new developments such as New Towns, but no completely new routes had been created. This was known as the 1986A system.
A rather different 1986B system had been built up from first principles, ignoring administrative boundaries and looking only at demand patterns. Major demand centres were linked by fast direct bus routes (or rail routes if these already existed); intermediate demand centres were linked to the city by less direct routes; small townships and villages were linked to larger communities by separate local systems operating as frequent circular services. When this system had been defined it was seen that many parts of the 1986A system had developed in a similar manner, but the 1986B system appeared to rationalize a number of anomalies and duplications.
The authors observed that the results of the tests on the contrasting systems were interesting. Not surprisingly, it was found that the slightly extended present-day system would cope with the forecast demand in general. Also predictably, it had been found that the 1986B system would give, in general, better service to passengers, but in order to make it work it required more buses and therefore cost more. However, the difference in performance between the two systems was marginal, and the case for a radical reorganization was certainly not made out by the results of the study.
Restraining traffic
The study described applied to a situation in which it was assumed to be possible to build sufficient road capacity to carry all the potential traffic. However, if the volume of traffic was such that it was not possible, for physical, financial or environmental reasons, to create a system to accommodate the demand, then quite a new problem arose. Depending upon what practical policies were envisaged for restricting traffic, different analytical techniques were required in the study. There were three basic policies which could be used, either alone or in combination, to restrain road traffic.
(a) To allow congestion to occur, making road travel increasingly unpleasant and time-consuming.
lb) To control parking in a manner designed to inhibit the use of cars for certain purposes at certain times of the day.
(c) To introduce direct cost penalties such as tolls or other road-pricing devices. Whichever policy or combination of policies was adopted, the effect was likely to be a reduction in the number of drivers attempting to make private car trips, because some would seek alternative, less difficult destinations by car, some of them would transfer to public transport and some of them abandon their trips altogether.
Allowing traffic to control itself by congestion might work quite well, providing that the congestion was only of modest extent and that the resulting speeds were above that corresponding to maximum capacity. Otherwise, some form of restraint was required and this was commonly provided by the limitation of parking space. In the past, the lack of parking facilities in city centres arose simply because of their size and cost. However, it must be accepted that in the future, limitation of parking facilities would be implemented as a positive method for controlling traffic, and this, indeed, was planned for Edinburgh.
An elaborate model had, therefore, been devised to determine how the number of parking spaces in each zone should be chosen so as to make the fullest possible use of the road system within whatever constraints were specified in the use of roads. These principles were applied in the London Transportation Study and were proposed for the detailed study of the City of Edinburgh which had just commenced.
Bus priorities represented
From a public transport point of view, the interesting thing was that this allowed such features as bus lanes and priorities to be represented in the models and their constraining effect on the remainder of the traffic to be taken into account in such a way that it could be said that public transport, parking and highway policies were interwoven.
It would seem, said the authors, that while most public transport operators were interested in long-term forecasting, they were more vitally concerned with short-term improvements designed to stem the slight but steady flow of travellers away from public transport. This would also be a major concern of the new Passenger Transport Authorities.
Bus services in most major urban areas had grown up with the development of the area, and as towns and cities had grown radially, so the bus routes had followed. They had seldom been designed as a complete system for an area, and there was now a requirement for operational research techniques to treat the complete network of services as one system, and to attempt to minimize some functions of the system such as running cost for any particular level of usage.
Certain other constraints had to be accepted, such as a limit on the total number of buses available to run the new services, and the basic financial object of the operator. This latter requirement could be defined as the necessity for the bus system as a whole to break even, make a profit, or run at a subsidized loss.
The authors concluded by remarking that the approach they had outlined had not yet been developed into a working program, and their object in expressing it at that stage was to obtain the reaction of a group of experts in the practical operation of public transport to the questions: (a) Was there a need for scientific analysis of present-day bus operation in relation to demand as an aid to decision on reorganization of routeing?
(b) Was there a better approach to the problem?
(c) What type of criteria should be applied in route selection? Demand, suitability of road and general cost criteria had been mentioned. Were there others?