There are a number of factors which will permit the recommended rapid-transit system to give a performance not yet attained elsewhere. Of these, the most important is that the civil engineering works are planned so that long, wide cars can move freely over the system, without speed restrictions imposed by sub-standard curvature. The method of supplying electricity to trains is also free of the restrictions so frequently imposed by the necessity to inter-work with an existing system. The equipment necessary for operating the system will be built at a time when several important new techniques have been tried out elsewhere and can be recommended with confidence here. The most important of these are the automation of signalling and train speed control, the substitution of static (solid state) for many moving devices, and the elimination of friction in the mechanical parts of the trains by the use of electric braking, rubber components and pneumatic springing on the cars.
The equipment recommended is the most suitable for performing the duties required. These may be summarised as providing a reliable, punctual, fast and frequent service with the highest capacity and efficiency compatible with reasonable first cost and operating cost.
CAPACITY
The operating facilities and equipment required for a rapid-transit railway are primarily governed by the maximum volume of traffic to be carried over any section of the line during the busiest hours of the day. But variations in traffic density between different periods of the day, and between different sections of the line, also have an important bearing on the design of the train services and on some of the facilities and equipment.
The traffic volumes expected on the recommended system are abnormally high. For example, on their most heavily-loaded sections the Tsuen Wan and Kwun Tong Lines are both expected to carry about four times as many passengers a day as the heaviest sections of the London Underground; and the Island Line three times as many.
The distribution of this unique volume of daily traffic over the hours of the day is fortunately expected to be much less uneven than on any similar existing rapid-transit system. Thus the maximum effective capacity required in one direction in the peak hour is much lower in proportion to the total daily volumes than elsewhere. Even so, the estimated maximum hourly loads are very high. By 1986 the peak services will be required to move in one direction in one hour about 45,000 passengers on the Tsuen Wan Line, almost the same number on the Kwun Tong Line, and over 35,000 on the Island Line. The corresponding figure for the Sha Tin Line is about 25,000, but a large growth of traffic after 1986 is expected with the further expansion of Sha Tin. All these figures are high by any standards. Indeed, the movement of over 45,000 passengers on one track in one hour appears to have been attained so far only on one or two lines in New York and Tokyo, and then only under conditions which are generally regarded as involving excessive overcrowding. Nevertheless, it is quite possible on a new system specially designed and equipped for the purpose, to carry at least 45,000 passengers in the maximum peak hour over a single line of track, in crowded but tolerable conditions, with some margin for further expansion.
The theoretical capacity of a rapid-transit system depends on the maximum practical loading of an individual car. The total theoretical hourly capacity of the line is then the assumed car capacity multiplied (a) by the number of cars per train and (b) by the number of trains per hour which can be regularly worked.
The maximum practical loading of an individual car has been the subject of some operational research. The best available evidence seems to show that after all seats have been occupied, passengers will not or cannot pack themselves into the standing accommodation more tightly than about one person per two square feet of space between longitudinal seats and about one person per one-and-a-half square feet of space between doorways. At this density of loading the conditions are such that no passenger is readily able to move his position (or make his way to the doors to alight) without pushing or elbowing his way between other passengers and without some co-operation on their part.
File:MTS Fig72.pngFigure 72 — Theoretical Rapid-transit CapacityFigure 72 shows the theoretical single-track capacities for various combinations of size of cars, cars per train and trains per hour. It shows for example that 40 six-car trains or 30 eight-car trains per hour will provide a theoretical capacity of 48,000 passengers per hour, if each car has a theoretical capacity of 200 passengers. But the practical or effective capacity is quite a different thing. Full utilisation of all the available space is far from attainable and the theoretical capacity has to be heavily discounted for several reasons.
First, in crowded conditions and with necessarily limited station stop time, a given load of passengers is seldom in practice evenly distributed over all parts of a car or all cars in a train. This is partly because passengers accidentally obstruct each other, either while moving or by seeking to take up positions inside the car near the doors. It is also partly because, even if the waiting passengers on a platform are well distributed over its full length, as they should be, there is no way of ensuring that the passengers alighting from the train will be equally well spread. The numbers alighting and wishing to board at each door do not match, so congestion occurs in some parts of the train while spare space is wasted in others. The inequalities of loading thus developed may be either corrected or made worse at subsequent stations. Much can and should be done by car design, good station planning, crowd control and so on to mitigate these inefficiencies, but they cannot be eliminated. A large margin of theoretical capacity in excess of actual demand remains essential for these reasons, though it cannot be measured closely except after actual experience on any given system or line.
Secondly, turning from theoretical and practical capacities per train to track capacity per hour, it has to be remembered that the full peak-hour traffic will not be evenly spread throughout the hour. There is sure to be a "peak within the peak", not necessarily occurring at the same time on all parts of a line or at exactly the same times each day. Again the extent of such fluctuations in the flow of traffic cannot be closely predicted for a future system. On existing systems it varies from city to city, and even from line to line and year to year. But in the light of experience elsewhere it can be assumed for planning purposes that at least one third of the full peak-hour traffic will pass in the busiest 15 minutes. In other words if 45,000 passengers per hour are to be carried, the train service maintained through the peak period must be at least sufficient to carry traffic at the rate of 60,000 per hour. Otherwise delays both to passengers and trains are bound to occur during the critical 10 or 20 minutes when the capacity is overtaxed. Any delays to trains will of course carry over and the planned capacity scheduled over the full hour will not be realised.
| Considering the need for a high capacity as described above, the system has been designed to provide for at least 30 trains per hour and 8 cars per train. On the essential assumption that the lines will be built with the generous curvature recommended, the trains will consist of cars having an overall width of 10 feet 6 inches and an overall length of 73 feet 6 inches, except for the end cars which are 76 feet 3 inches long to provide space for the train operator. These cars will be larger than any so far built. Each car has five pairs of double doors 4 feet 6 inches wide, and contains approximately 50 seats arranged longitudinally in four groups on each side between the five doors in a manner designed to cause the minimum obstruction to the movement of passengers into, out of and within the cars. The cars will be connected by vestibules approximately to their full width, permitting free passenger movement from end to end of the train. This is an unusual feature in rapid-transit cars, though it was in use on the Lancashire and Yorkshire Railway before 1914 and the newer cars in Tokyo have it. Although no records of comparative research on the subject have been found, this feature should contribute to the attainment of high practical car capacity.
}}
The proposed car dimensions are shown in Figure 73. Each train has a driving position at each end, and may be made up of four, six or eight cars, (but not in odd numbers) to suit the growth of traffic over the years.
The theoretical passenger capacity of each car would be 370, but careful streamlining of all the interior arrangements and the features described in paragraph 11 may make higher loadings possible, and if they also have the effect of facilitating faster passenger movement through the doors than has been experienced with existing car designs, the practical capacity should also be increased, creating a margin for the growth of load after the design year. The availability of such a margin does not, however, solely depend upon these possibilities of higher effective capacity per train. There is also in reserve the possibility of increasing the number of trains per hour by 10 to 15 per cent, perhaps even up to 36 per hour. The regular operation of such a close service will be greatly helped by the independence of each of the four lines and the consequent freedom from the difficulties of interworking different services through junctions. Any increase above 32 trains per hour might however overtax the practical capacity of certain two-track terminals recommended later and involve some additional expenditure for enlarging them to three tracks.
PEAK AND OFF-PEAK SERVICES
In order to estimate for each stage the amount of rolling stock required and the daily car mileage, on which so many of the costs depend, it was necessary to prepare tentative timetables for each line. The primary need in planning the train services is to ensure that the capacity at peak periods is adequate to carry the estimated maximum volume of traffic on any part of the line. Capacity must, of course, also be sufficient for the reduced volume of traffic at other times, but in the design of off-peak services it is seldom the dominating factor. The aim then is rather to provide a frequent enough service to attract most or all of the potential traffic. In any case, the degree of crowding that may become acceptable in peak hours is generally unacceptable in less frequent off-peak services, when the public can see for themselves that more service and capacity could be provided and comfort increased.
It is wise to provide rather more capacity in relation to demand in the earlier years than will be necessary, or even perhaps possible, at later stages. This will help to popularise the system. It will give time for the public to become more experienced in boarding, alighting, moving within the cars and generally co-operating with each other and the staff—experience which is of great value to the smooth and efficient operation of the trains at stations, and takes time to acquire. It will also reduce the risk of complaint about the inadequacy of the service and congestion, which might well arise if "excessive" crowding were encountered at the outset, when only six-car trains are being worked in stations obviously designed for eight-car trains.
In planning the peak-hour services for the early stages, there will be a choice, within limits, between longer and less frequent trains or shorter and more frequent trains. The same capacity can for example be provided by 16 six-car trains as by 24 four-car trains. The former is appreciably cheaper, requiring one-third fewer train operators. The longer train is also better able to absorb an unexpected surge of traffic. On the other hand, the more frequent service of short trains would reduce the average waiting time of passengers at stations by nearly 40 seconds. For purposes of estimating costs, the services have been planned for the early stages on. each line on the most economical basis to provide the capacity required, subject to the reasonable judgment that the main peak-hour services should not operate at intervals wider than four minutes (15 trains per hour) gradually closing to intervals of two minutes (30 trains per hour) or thereabouts, as traffic develops through the years.
Similarly, it has been assumed, on commercial grounds, that for the mid-day services between the morning and evening peak hours, the intervals should not be wider than six minutes (10 trains per hour) at the first stage of each line, with frequencies again increasing as traffic grows.
The only cases in which the peak and off-peak intervals would be slightly wider than four and six minutes are east of Kwun Tong on the Kwun Tong Line, and west of Western Market and east of North Point on the Island Line. On all these three sections the relatively lighter traffic can be adequately carried by half the main line service. The plan accordingly provides for alternate trains to be reversed at Kwun Tong, Western Market and North Point.
On some systems it is the practice to shorten the trains between the peak periods, either dividing them into two or detaching a few cars and running them separately into sidings, mainly to save electric current costs. But the operation of uncoupling and later re-coupling trains in service is complicated and expensive in staff, though sometimes justified where off-peak traffic is very light. In Hong Kong, however, with its good all-day traffic, it is likely to prove more economic to operate the same length of train on any given line throughout the day. This has significant advantages in rolling stock design and cost. Proportionately fewer of the more expensive cars with driving ends are required and these cars, which interfere with the freedom of passenger movement and waste passenger space, never appear in the middle of trains.
Having settled the desired length and frequency of trains, both peak and off-peak, for each line at each stage, estimates of their capital and operating cost are derived from specific timetables worked out in sufficient detail to reveal the amount of rolling stock required and the maximum current demand by reference to the peak hour services, and the total car mileage by reference to the whole days' operation. The number of trains is arrived at by dividing the interval into the sum of (a) the running time, including station stops, for the round trip and (b) the terminal or layover time at each end. The running time is separately calculated between each pair of stations in each direction, and is governed by the proposed performance of the rolling stock (acceleration and deceleration rates and maximum speed) related to the distance and taking into account gradients, curves and all other factors imposing restrictions on speed.
STATION STOP TIME
Obviously station stops should be as short as possible, consistent with serving their purpose, to minimise total journey times. Moreover, the quicker the journey the lower the cost of the service, because it can be worked by fewer trains and staff. But on an intensively used system where trains have to be run at 2½-minute intervals or less, a further reason for firm control of station stop time is to prevent delay to the next train. With a 2-minute headway of very long trains, station stops over, say, 40 or 45 seconds may well begin to cause such delays—setting up a vicious circle of increasing congestion on platforms, still longer station stops and further delay to trains behind. Experience shows that the longer the trains—and long trains must be used for heavy traffic—the more difficult the control of station stop time becomes. To overcome this, three features are essential. First, the car layout must permit very quick boarding and alighting. Secondly, every car must suit all passengers—there must be no separate cars for first and second class, or smokers and non-smokers (and the case for allowing no smoking on other grounds is substantial). Thirdly, the station entrances and exits at street level and the passages, stairs and escalators to and from the platforms, must be arranged so as to attract an even distribution of incoming traffic over the whole length of the platforms, and facilitate quick clearance of the platforms when heavy traffic alights. Provided designs embodying these features are adopted, and the best use is made of detailed crowd-control techniques, maximum station stop times should normally be well within the limits consistent with efficient operation of the services envisaged. In the tentative timetables prepared for estimating purposes, average station stop times of 30 seconds in the peak hours, and slightly less at other times, have been assumed.
TERMINAL TIME
File:MTS Fig74.pngFigure 74 Proposed Terminating Track SegmentsAt the end of a line, or at any intermediate station at which part of a service is reversed, time has to be allowed for the staff to carry out certain minimum terminal duties associated with safety and the reversal of the train. If everything runs properly it is quite feasible, even with the longest trains recommended, to unload, reload and reverse 30 trains per hour in a simple 2-track terminal, with a scissors-crossover on the approach side, as diagrammatically shown in Figure 74. But such an arrangement allows practically no margin for recovering any time lost. If a train is delayed for any reason, most commonly by extended station stops, and arrives at the terminal late, it will also have to leave late on its next journey. To avoid this it is essential to provide somewhere a margin for correcting any such late running, and on a rapid-transit system this can only be done by scheduling the trains to have a "layover" of a few minutes at either or both terminals where they reverse. If trains are scheduled at 2-minute intervals and are to be allowed more than about 2½ minutes between arriving at a terminal and departing on their next journey, clearly a 2-track terminal is insufficient and a 3-track terminal must be provided. In normal circumstances the best track and platform arrangement for this purpose is of the type also shown in Figure 74. Such a layout permits the scheduling of trains on a 2-minute headway to have about 4½ minutes "recovery time" before they are due to leave. This is of great value in ensuring a high standard of adherence to the regular intervals scheduled, well balanced loadings as between successive trains, and limited and consistent station stop times.
To give the maximum flexibility in train operation, a case could be made for providing 3-track terminals at both ends of most of the lines. On the other hand, the capital cost of such a layout, especially if constructed underground, is much higher than that of a 2-track terminal. Moreover the latter is more convenient to passengers because all trains then start from one side or the other of the same island platform and direction signs become much simpler. The general conclusion has been reached after weighing all these factors that a 3-track terminal should be provided at one end only of each line and a 2-track terminal at the other, thus allowing some "recovery time" to be built into the schedule not at the end of each trip but once in each round trip. Provision has accordingly been made in the estimates for the following terminal arrangements:
On the Kwun Tong Line — Three tracks at Western Market; two tracks at Ma Yau Tong, and also at Central and Choi Hung when they are temporary terminals in Stages 1 and 2, and the headway will not be closer than three minutes.
On the Tsuen Wan Line — Three tracks at Admiralty; two tracks at Tsuen Wan and also at Lai Chi Kok in Stages 2 and 3.
On the Sha Tin Line — Three tracks at Tsim Sha Tsui; two tracks at Wo Liu Hang and also at Tsz Wan Shan in Stage 5, when the headway will not be closer than four minutes.
On the Island Line — Conditions differ because the full service is not required at either end and half the trains will be reversed at Western Market and North Point. This permits 2-track terminals both at Kennedy and at Chai Wan, but calls for special layouts at Western Market and North Point.
The reversal of close interval services, for example on 2-minute headways, in the 2-track terminals does not allow sufficient time for the train operator to change ends and take out the same train. Arrangements have therefore to be made for each train opera tor to "step back" one train and take out the train following the one which he brought in. This involves employment of an additional train operator, but makes possible the capital cost savings of a much simpler station.
Just as the timetable has to provide for the smooth conversion of the frequency of service between the two peak periods, and also allow gradually for the slightly faster running times at off-peak periods due to shorter station stops, so the service has to be gradually built up from the start of traffic in the early morning and thinned out in the evening. The volume of service has been varied in the different main periods of the day in accordance with the known variations which occur on existing public transport, except that it has been assumed that the proportion of the total days' traffic travelling in the peak hour will tend to rise over the years. Table 69 summarises the hourly variations assumed for the design year. For purposes of estimating car mileage and operating costs, it was decided that first trains would start from both ends of each line at approximately 6.00 a.m. and last trains would finish at both ends at 1.00 a.m. On grounds of cost no extension of the traffic day beyond these limits should be entertained. While first and last trains are likely to be well patronised whenever they run, the loadings before 7.00 a.m. and after midnight will be light, and revenue small. Train running costs can be reduced by working a thin service at these times, but station staffing costs continue till stations are finally closed; and any further shortening of the five-hour period when the lines are available for track and signal inspection at night would sharply increase maintenance costs.
In the first and last half-hour of the traffic day a service of trains at about 10-minute intervals should be sufficient. This means that on all lines only three or four trains will need to be stabled overnight at the ends of each line, away from the running maintenance depot, ready to start up the service in the mornings. In the completed system, sidings are provided for this purpose on the Kwun Tong Line beyond Western Market and at Ma Yau Tong; on the Tsuen Wan Line beyond Admiralty in a position where they could subsequently become part of the running tracks to Aberdeen, and also at Tsuen Wan itself; on the Island Line between Belcher and Kennedy; and on the Sha Tin Line at Hung Hom. In Stages 1 and 2 there would also be temporary sidings, which would form part of the extension to Western Market in Stage 3, for three trains beyond Central.
These same sidings, while primarily required for stabling a minimum number of trains away from the depots overnight, also serve the purpose of stabling a few trains, to save unnecessary mileage, between the morning and evening peak periods, and can be useful in emergency for taking a defective train out of service until it can be worked back to its depot.
The train services tentatively planned on the bases outlined in the foregoing paragraphs have shown that the rolling stock needed at each stage, including spares to allow for maintenance, and the corresponding annual car mileage to be operated would be as set out in Table 70 which also shows the total manpower requirements.
METHOD OF ELECTRIFICATION
Electricity may be distributed to trains either via collector shoes running on a conductor rail, known as the third rail system, or via pantographs to collect current from an overhead contact wire. Either method requires substations in which 3-phase alternating current at high voltage is converted into the form required by the railway. Their number and type depend upon the method and voltage used for distribution to the trains.
Both methods were considered and the overhead contact system was rejected for several reasons of which the more important are:
The cost of all underground sections would be increased due to the space needed between the tunnel roof and the car roof for the overhead equipment. In contrast, the third rail can be accommodated without additional height or width.
On those parts of the lines where overhead construction is recommended, special precautions would be needed to ensure that the catenary system and its supporting gantries were strong enough to withstand typhoon winds. There would also be risk of damage by objects flying in the wind or dropped from adjacent buildings.
There would be some loss of amenity on overhead sections due to the presence of the gantries and overhead wires.
The third rail system does not suffer from any of these objections. Moreover, it can be protected against the risk of flooding. Flood protection on the low lying sections is essential for the railway as a whole in any case.
In recent years high voltage alternating current at industrial frequency has been used for main line and some suburban railways associated with main line railways, but this is for reasons which are not applicable to rapid-transit systems. Direct current is therefore recommended, the running rails being used for return current.
Most third-rail systems use direct current at about 600 volts, but the new Bay Area Rapid Transit District, San Francisco, has recently decided to use 1,000 volts. The Manchester-Bury electrification has given satisfactory performance for over 50 years with a third rail at 1,200 volts. To minimise the overall cost of substations, distribution system and train equipment, a 1,500 volt third rail system is recommended. This voltage has been used for many years on main line and suburban systems using the overhead contact system.
POWER SUPPLY
A reliable supply of electricity is essential for satisfactory operation. Preliminary discussions have been held with the two electricity supply companies. Both are building new power stations and reinforcing their power transmission systems to ensure reliable supplies from their existing and new stations.
Either company will be able to supply the whole of the power required at each stage of the project. The peak demand in the design year will be less than 10 per cent of the present total maximum demand, and less than 5 per cent of the total capacity of the generators when the plant now under construction is completed. Both companies will be able to offer satisfactory guarantees that supply will be continuously available at or adjacent to the substations. Both agree that these substations should be designed, equipped and owned by the rapid-transit authority to ensure proper co-ordination of design and capacity with the demand of the trains and other electrical apparatus such as signalling equipment, pumps, fans, escalators and lighting. The choice of voltage at which the supplies may be taken, namely 33 KV or 11 KV, can only be made in the final design when the relative costs and reliability have been agreed with the supply companies. The estimates are based on taking supplies at the more expensive 33 KV source, and a material saving may be achievable by using 11 KV.
Material advantages in security of supply will be attained by dividing the load between the two companies. In particular, arrangements should be made so that the cross-harbour section can be supplied from both, and the estimates allow for this. Elsewhere the substations would be supplied from the nearest source, those on the Island from the Hong Kong Electric Company and those on the mainland from the China Light and Power Company.
For loads of the size in question, aggregating some 300 million units per annum in the design year, it should be possible to negotiate cheaper tariffs than those used for the estimates. This should at least reflect the presumptive fall in generating and other costs as the new and more efficient power stations come into use.
With the adoption of 1500 volts for supplying the trains instead of the more usual 600, less than half the number of substations will be needed. There will be three substations on the Island and seven on the mainland. They will house all necessary transformers, rectifiers and associated switchgear for giving supply to the trains, to the signal power system and to the medium voltage system supplying the tunnel ventilating fans, pumps, lifts, escalators, lighting and other station, yard and office equipment. All the switchgear would be under remote control from the central control room, and the substations would not be manned.
DESIGN AND PERFORMANCE OF TRAINS
Car weight is an important factor in train performance. An average speed of about 20 miles per hour inclusive of station stop time is desirable. Cars of the type recommended and capable of this speed can be built within a wide range of weights depending on the extent to which high tensile steels and light alloys are used. Final evaluation of the optimum weight can only be made in the detailed design stage when firm prices are available for cars of equal strength and comfort in the many variations technically permissible. Enquiries should therefore call for alternative prices for alternative weights. Then, when the cost of electricity will also have been established precisely, it will be possible to take account of all the economic factors involved in choice of weight—initial cost, maintenance costs for differing external and internal finishes, cost of electrical power equipment, and savings possible by adopting varying degrees of light weight construction. At this stage also, it will be necessary to decide the optimum speed and rates of acceleration and braking by close analysis of the final track profiles throughout the system.
Nevertheless, fair values of these variables, adequate for all purposes of this study, can be chosen in the light of knowledge and experience of existing systems. The estimates are based on cars in which all axles are driving axles, each powered by electric motors of approximately 1 10 H.P. The weight of the car, about 40 tons exclusive of passengers, is neither the lightest nor the heaviest that might be built. The cost allows for a measure of air conditioning, and for dynamic electric braking.
Dynamic electric braking not only avoids frequent renewal of brake blocks, but also allows the use of non-ferrous blocks under the best possible conditions, as the final brake is only used to bring the train finally to rest at stations and for emergency stops. A further advantage of this system of braking is that accumulation of iron dust on the tracks is avoided and maximum cleanliness is thus achieved. The system of control will ensure that the rate of change of acceleration and braking is kept low to ensure passenger comfort in spite of the relatively high acceleration and braking values necessary to maintain schedule speed on the short runs between the closely spaced stations.
The acceleration and the speed of the train both depend upon the amount of current allowed to flow into the motors. This is regulated by controlling the way the motors are connected (whether in series or in parallel) with one another, and the amount of resistance in circuit between them and the conductor rail. The electrical control equipment of the train, including the compressors supplying air for actuating the brakes, doors and traction motor current switches, is grouped in pairs of cars to give better distribution of weight and greater accessibility for inspection and maintenance.
The normal rate of acceleration and braking has been taken as 2.5 miles per hour per second but this value will need reconsidering during final design. Like the car weight, it is neither the highest nor the lowest that could be used.
The final decisions about the precise form of drive from motor to axle, the arrangement of the bogies and the springing of the vehicles should be taken on the basis of relative cost and performance as measured by firm estimates and guarantees and in the light of the behaviour of equipment in other parts of the world using the diverse techniques now available. The springs should undoubtedly be fitted with devices to maintain uniform rates of acceleration and braking with varying passenger load by controlling the current supply to the motors. It will also be for settlement at this stage to what extent static (solid state) control devices should be used to control train speed and provide energy for car lighting and battery charging; whether the doors should be air or electrically operated and whether the means of actuation of the final brake should be electro-pneumatic, mechanical or electric.
Suitable insulation should be provided in the sides, roofs and floors of cars to reduce noise to a comfortable level. The windows should be double glazed for the same purpose. These measures of sound insulation will also assist in keeping the cars cool in hot weather and reduce air-conditioning costs.
TRACK
It is recommended that British Railways' standard 109 pound per yard rail, or the International Railway Union (UIC) 54 Kg per metre (108.9 lb per yd) rail, which is almost identical, be used mounted on hardwood or concrete sleepers in a substantial bed of granite ballast. This relatively heavy rail for a rapid-transit system is fully justified, having regard to the length and loaded weight of the cars, the frequency of the trains and the high acceleration and braking rates. Rails should be welded together in long lengths without joints and the estimates allow for this. The absence of frequent rail-joints not only reduces the cost of track maintenance, but also reduces the cost of maintaining the mechanical and electrical components of the train equipment, by eliminating vibration and jolting. It correspondingly improves passenger comfort, and above all reduces noise both in and around the trains. Some reduction in first cost and perhaps in maintenance costs might be obtained by fastening the rails on rubber-like pads direct onto the structure or tunnel floor, but there is substantial evidence to show that quieter running is achieved by mounting them on sleepers in ballast. However, in the final design this matter will warrant reconsideration in the light of further experience with new types of construction.
The conductor rail, of about the same weight per yard, but of softer steel of high electrical conductivity, would be mounted on porcelain insulators, at suitable intervals, fixed to the sleepers at one side of the track. The estimates allow for guarding it from accidental contact on open sections of route and at points where railway staff will need to have access while the rail is "alive". It will also be guarded in the stations where it will be on the side of the track remote from the platform edge, or between tracks where side platforms are proposed.
SIGNALLING
The estimates provide for a signalling system of the highest quality essential for the safe and punctual movement of the trains. Every component would be built so as to "fail safe" by immediate application of emergency brakes designed to bring a train to rest if, by human or other mistake or failure, it attempts to move wrongly into a section of track.
Several techniques have now been developed and are already in use which, by associating the signalling devices with corresponding equipment on the train, permit economies in train staff (single manning of all trains), and reduce electricity consumption by ensuring accurate control of its rate of use. They are known as automatic train control systems. By the time firm specifications have to be written there will be a sound basis on which to select the most suitable system for Hong Kong. The present estimates are based on the system to be used by London Transport on the new Victoria Line, which is a development of one that has been in regular public service on the London Transport Hainault Line for over three years.
The essential principle of all such systems is to divide the track into sections and provide methods of detecting and controlling the speed of each train at all times. It can then be assured that each train runs at the optimum speed needed to perform its schedule unless the signalling apparatus demands a reduction of speed. It does this if it detects unsafe overtaking of the train ahead and, of course, at the approach to every station. The controlling devices located on the track regulate the supply of current to the motors and the degree of brake application.
In starting out from a station on a clear section of track, the current is automatically controlled to a level dependent on the passenger load by devices associated with the springing. The train will accelerate smoothly to the maximum speed or until it reaches the pre-determined position on the track where current is cut off and the train allowed to coast without taking power. At the correct distance from the next station the dynamic electric braking is brought into use, its strength being regulated to ensure an accurate stop at the station after application of the final brake which takes over as the electrical brake fades away with loss of speed.
The locations for the devices on the track to perform these functions are pre-determined, taking into account the distance and gradients between stations. The desired speed, the demand for electricity and the rate of wear of brake shoes and wheel treads are all considered. Signal indications are transferred to the driving cab, thus eliminating the cost of track-side signals except those few which are needed for the operation of service vehicles.
There are several ways of transmitting the necessary information from track to train. In the system recommended, four codes are used and the train cannot move unless it is receiving a code from the track. The various codes control the amount of power and braking necessary to maintain the schedule and stop at the appropriate location in stations. The emergency brake is applied when the maximum speed corresponding to the code being received is exceeded. The train operator can take over control subject to suitable restrictions as to the speed at which he may drive, and to the automatic application of the emergency brake if this is exceeded.
Several other ways of achieving these objects are being used, and some of them still under development show promise of being equally satisfactory and cheaper. Some reduction in the cost of automatic control may thus be possible before a final decision has to be taken.
The power requirements for signalling are supplied by medium voltage cables laid in concrete ducts on one side of the track. Most of the apparatus required for the operation of the devices mentioned above can be housed in cubicles at the ends of the station platforms, and the remainder in cubicles between tracks where cut and cover is adopted and in specially-shaped housings on bored tunnel sections of line. The telecommunication cables and apparatus are similarly accommodated.
TELECOMMUNICATIONS AND CENTRAL CONTROL
The system of telecommunication recommended would employ a combination of wired and wireless equipment. The objective is to ensure that at all times those responsible for the operation of the railway are in touch with every individual having local responsibilities for sections of the railway, and that wherever necessary all individuals are in touch with one another.
Although high-frequency short-wave techniques are increasingly effective and widely used, the time has not yet come when they can be relied upon entirely to supersede telephone networks: the two techniques are complementary to one another. Three railway telephone circuits have been provided for—one for track operations, one for control of power supply and one for general purposes. Suitable connections would be made to the public telephone service. Wireless communication will be provided to enable the men in charge of individual trains to speak to the control room and to broadcast throughout their trains. Public address systems will also be provided for station masters. At the stations with the heaviest traffic, closed-circuit television in association with station broadcasting will assist in control of passenger movement.
The telecommunication network would radiate from a central control room located at the Kowloon Bay maintenance depot with local auxiliary circuits as necessary. Control and indication desks would be provided to show the tracks and power supply network symbolically. The position of each train in relation to the track circuits and stations, and the state of each electrical circuit breaker, would be shown. The apparatus will, with suitable safeguards as to procedure, allow the control staff to isolate any section of the network and take over the movement of any train. However, in the ordinary course of events, train movements will proceed automatically on their pre-arranged routes controlled by the punched tape-computer technique which is an integral part of the automatic train control system.
Statistical information necessary for evaluating and improving service can be obtained directly from recording apparatus installed in the control room. Teleprinters will be used for all instructions relating to the safety of the public and of railway staff.
ANCILLARY EQUIPMENT
A large variety of ancillary equipment is necessary for the efficient operation of a rapid-transit railway. Although allowance has been made in the cost estimates for all such items, only the more important are described here.
Ventilation — It will be necessary to ventilate the underground sections of the system to ensure reasonable comfort for the passengers and prevent a gradual increase in temperature. The estimates therefore include electrically-driven fans to force air into the tunnels near the mid-points between stations. The ducts and shafts are included in the civil engineering works. The location of the fan rooms, which can often be associated with the drainage pumps described below, can vary from place to place depending on the availability of sites. In all cases air must be drawn in through shuttered openings well above ground and flood level. These need not be large and can be incorporated in buildings adjacent to the line of route. The estimates allow for all fans to be capable of reversal seasonally to exhaust air from the tunnel, although detailed investigations may show that this is not necessary in all cases.
Pumps and Flood Prevention — Although the underground structure will be reasonably watertight, some water will seep in and means must be provided for removing it. In general, water will be removed by self-priming pumps located at the low points of the track profile. Semi-watertight steel doors will also be provided at all the entrances to underground stations located in areas where flooding is likely to occur.
Escalators, Travelators and Lifts — The escalators in a rapid-transit system carry much greater passenger loads than those installed in department stores. Also, to avoid interference with passenger movements, it must be possible to carry out maintenance procedures without closing down. Thus, the design standards of the recommended escalators are higher than these of any currently operating in Hong Kong. The same high standards are required for lifts and travelators. In all, 133 escalators, with heights ranging from 12 feet to 60 feet, six lifts, and seven pairs of travelators in passageways will be required.
Station Lighting — In general, fluorescent lighting has been assumed for the stations. On the platforms the recommended level of illumination is 30 foot-candles but 40 foot-candles in the mezzanines. Certain facilities such as ticket-vending machines and ticket control barriers should be highlighted. For these and for all passageways and stairs giving access to the station, a level of illumination of 60-foot candles is recommended.
Fare Collection and Ticket Machines — With a graduated or mileage fare system, the cost of fare collection and of ensuring that the correct fare is paid for every journey is considerably higher than with a simple flat-fare system, and the methods to be adopted for selling and checking tickets are therefore an important matter. Graduated fare systems on rapid-transit undertakings handling very heavy flows of traffic are unusual in American and European Continental cities; most of the relevant experience has been accumulated in Britain, mainly by London Transport. In London, the passenger buys a ticket at the start of his journey which shows prominently the station of origin and the fare value, and is available to any destination covered by that fare. This device enormously reduces the number of different tickets in use and facilitates high speed mechanised ticket issue. The tickets may be purchased at ticket booth windows or by inserting the appropriate coin or coins in an automatic ticket machine. In either case the ticket is dispensed by machine. Varying numbers of automatic machines, according to the density of traffic at different stations, can be provided in the ticket halls for all or most of the fare values for which there is a large demand at the station concerned. The more modern types of these machines are equipped to give change, and some recent designs to give change for paper money as well as coin. All these machines, whether operated by the booking clerk or by the passenger, print the tickets at the time of issue on blank rolls of paper or flexible card. This eliminates the cost of providing for the security and audit of large stocks of preprinted tickets of high value. All the machines automatically record the number and value of the tickets issued thereby saving much clerical labour.
The separate processes of checking and collecting tickets in rapid-transit conditions with graduated fares have not been either mechanised or particularly efficient. They require a very large staff of ticket collectors but, even so, significant losses of revenue due to underpayment almost certainly occur.
It might have been necessary to recommend the adoption of these methods for want of anything better, were it not for the fact that a dramatic and most promising revolution in automatic fare collection (AFC) and ticket-issuing devices is in prospect, and is being actively developed by London Transport and other transit authorities in both America and Europe. This novel system is based on the use of electronic control of entrance and exit gates at stations by the use of magnetically-coded tickets. London Transport plans to use this system on the Victoria Line. Somewhat similar systems are already operating on a small scale on the new underground railway in Milan and on a heavily-used commuter line of the Central Illinois Railroad in Chicago. The Bay Area Rapid Transit District (BARTD) is also, it is understood, adopting a form of automatic fare collection on its new system. It is expected that the high cost of the electronic equipment, automatic barriers and associated automatic ticket-issuing machines, will be more than offset by savings in booking office and ticket-collecting staff and by more efficient fare collection.
Since firm prices for this new equipment are not known, estimates of equipment and staff have been based on existing London Transport methods. However, the new automatic methods, further improved and developed, are likely to have been well proved in two or three years' time and it is therefore recommended that the choice for Hong Kong be made in, say, 1970, in the light of the experience gained by BARTD, London Transport and other rapid-transit undertakings, of the operational efficiency and economics of these promising new methods.
Research is also proceeding into the possibilities of developing sophisticated electronic equipment of similar kinds which could be used on buses and might even make feasible the efficient collection of graduated fares on double-deck buses with one-man operation; but the solution of the technical and design problems involved is probably some years away.
MAINTENANCE AND INSPECTION ORGANISATION
Adequate facilities for maintenance and inspection of all components of a rapid-transit system are vital to its efficient operation. These facilities fall into two categories, namely, those for storage, cleaning, and short-term inspection of trains and associated equipment; and those for less frequent long-term inspection and overhaul of all equipment.
Facilities of the first kind should be provided on each of the four lines. Overhaul facilities need only be provided at one place, and it is recommended that the site for this main depot, as well as the short-term inspection sheds and storage yards for the trains normally working on the Kwun Tong Line, be at Kowloon Bay. This central workshop, stores, training, control and administration building to serve the whole system should be constructed and brought into use well before commencement of service in 1974, and gradually augmented in the later stages. For maximum operating flexibility, and for security against trains being locked in the depot on account of points failure, derailment, etc., rail connections to the main line are provided at both ends of the depot. The central stores organisation should be located here with substores at the inspection sheds on the individual lines. Road access with a car park and a vehicle servicing depot should also be incorporated.
A four-storey building is planned to house all these activities. Although the site is near Kai Tak airport it is fortunately free of height restriction; and in any case a very high building is neither necessary nor desirable. It was formerly customary to provide cranes and headroom for lifting one car over another in a main workshop. Most of the air space in such a building is very infrequently used. Modern hydraulic lifting jack techniques and track-level wheel-turning equipment avoid this waste of material and space. Overhead runways capable of light lifts over parts of the shop, and ample provision of fork lift trucks, provide all the lifting facilities required. It is recommended that the shops be constructed in this fashion.
The workshops, inspection sheds and stores will be at track level; the control room, training school and workshops for light current apparatus associated with telecommunication, automatic train control and ticket machine equipment will be on the next floor; engineering, accounting and statistical offices on the next floor, and management offices on the top floor. Each section should have access to the appropriate section of the messing facilities, which can then be efficiently organised for the particular needs of each grade of personnel. The estimates allow for full air conditioning of the main block and for partial air conditioning of the workshops, stores, inspection sheds and other covered accommodation.
The short-term inspection sheds, which should be adjacent to the storage yards on each of the other lines, can be of very simple character and should be ready for use six months before the lines come into service. They include covered accommodation for a full length train above a sunken floor so as to give convenient access to all under-car equipment, and gangways for access to the cars themselves for inspection and cleaning, and for attention to interior control equipment. The shed on the Tsuen Wan Line will have an additional track fitted with equipment for restoring wheel profiles of the cars on this and the Island Line. Covered stabling accommodation, adjacent to the inspection sheds, for trains not required for service between the morning and evening traffic peak, will be provided on all lines. If such trains were stored in the open they would become very hot in summer. Limited substores, under control of the shed foreman, to hold small components such as brake shoes, motor brushes, etc., should be provided.
The cost by stages of equipping the rapid-transit system is given in Table 71. The estimates are based on current prices of British built equipment delivered in Hong Kong. Although manufacturers in the Colony will be able to make some components, all the more important items, for example, rails, trains, electrical equipment for substations, signalling and control apparatus, will have to be imported. For all installation and erection work, including track-laying, allowance has been made for the prevailing cost of labour in Hong Kong.
The cost of engineering services has been assessed on the basis that for the early stages consulting engineers will be retained to design, specify and control all contracts; and supervise the testing and setting to work of all components. This allowance has been tapered off from six per cent for Stage 1 to one per cent for Stage 6 on the assumption that the permanent engineering officers of the rapid-transit organisation will undertake most of these executive duties in the later stages with only general advice from consulting engineers. Provision is, however, made for the retention of consultants for inspection and testing of equipment during manufacture abroad in all the stages.
The amount shown for preliminary expenses includes provision for the training of local engineers to assist the consulting engineers in the early stages and later to take over responsibility for design and construction. It also includes the training of staff of all categories to operate the railway before the beginning of public service. (Training expenses from 1974 onwards are included in the annual operating costs.) Although for some categories of staff it will be essential to send men abroad for this training, it will also be necessary to augment the facilities for technical training in the Colony. These facilities exist in embryo in the universities and technical colleges. Because it takes as long to train men for the technical duties involved in operating and maintaining a railway as it does to design and produce the equipment, this matter will demand detailed investigation and action immediately a decision is reached to proceed with the work.
The amount (10 per cent) included in item 8 of Table 71 for contingencies for major equipment reflects the fact that, provided the rapid-transit system is built to perform the functions described herein, there is more certainty as to its cost than in the case of the construction works described in Chapter 9. Twenty per cent has, however, been included for contingencies for minor works that have not been so closely investigated.
Where alternative methods have been indicated, it may be assumed that if the capital cost exceeded that allowed in the estimates, there would be equivalent or greater savings in operating expenses. Table 72 shows approximately the amount of capital that will be needed in each year for the whole of the works described in this chapter.
The costs of operating the rapid-transit system at each stage are given in Table 73. The approximate total numbers of staff, including management, required at each stage, are shown in Table 70. The estimates of cost for staff of all grades are based on current rates, taking into account employee benefits and allowances customary at the present time, and the differentials that would have to be paid to expatriate specialists with the appropriate experience. The cost of continued training of recruits for all purposes and training for promotion and refresher courses is included in these estimates. The cost of preliminary training prior to beginning of operation is treated as a capital expense.
The amount allowed for depreciation is calculated on the assumption that a sum appropriate to the prospective life of equipment of different kinds, which will eventually require complete renewal, will be set aside annually to accumulate at compound interest at five per cent so as to provide an amount for its renewal, equal to its cost, at the end of its life.
