CONTENTS EXECUTIVE SUMMARY 2 1. INTRODUCTION 8 2. ROAD PRICING IN CONCEPT AND PRACTICE TECHNOLOGICAL OPTIONS FOR ERP SYSTEM 14

Transcription

1 CONTENTS EXECUTIVE SUMMARY 2 1. INTRODUCTION 8 Background of the Study 8 Study Approach 9 2. ROAD PRICING IN CONCEPT AND PRACTICE 10 Principe Underying Road Pricing 10 Economic Theory for Road Pricing 10 Wordwide Deveopments in Road Pricing 11 ERP System in Singapore 11 Piot Scheme on ERP in Hong Kong TECHNOLOGICAL OPTIONS FOR ERP SYSTEM 14 Desktop Review 14 Fied Evauation 15 Performance of Tria Systems 15 Technoogica Option for Hong Kong ALTERNATIVE TRAFFIC MANAGEMENT MEASURES 17 Measures to Improve the Transport System 17 Suppy Management 17 Demand Management 17 Aternative Measures to ERP NEED FOR ERP 18 Traffic Forecast 18 Need for ERP POSSIBLE ERP SYSTEM FOR HONG KONG 20 Conceptua Design of ERP Scheme 20 Compementary Measures 22 Impacts of Potentia ERP Schemes 23 Impementation Strategy 23 Impementation Programme ERP & ITS INTEGRATION BENEFITS OF ERP PUBLIC CONSULTATION 26

2 2 EXECUTIVE SUMMARY Background i. In 1994, Government set up a Working Party to examine measures to tacke traffic congestion in Hong Kong. The Working Party proposed to continue to pursue the baanced transport strategy of buiding new infrastructure and managing vehice use and ownership. A package of both ong term and short term measures was put forward for consideration. One of the measures considered was the subject of this Study, Eectronic Road Pricing (ERP), in view of technoogy advancement and its effectiveness in managing road use. ii. Unti the economic downturn in 1998, the vehice feet was increasing at a fast rate. Despite that, average traffic speeds in the urban areas have been maintained at about 20 km/h over the past twenty years. This has been achieved by a combination of ownership restraint measures such as First Registration Tax (FRT) and Annua Licence Fee (ALF), an efficient pubic transport system, road network improvements and effective traffic management schemes. Considering the high GDP per capita, these restraint measures had been successfu. In Hong Kong there are ess than 50 private cars vehices per 1,000 popuation, compared to more than 100 in Singapore and Seou, more than 200 in Jakarta and over 300 in Bangkok. Ony about 11 per cent of daiy person trips are made by private cars vehices whie 89 per cent use pubic transport. Nevertheess, it was considered that the adoption of the "user-pays" principe woud offer a more efficient, equitabe and fexibe means of managing the road space particuary in congested areas during busy hours. Study Approach iii. iv. The Feasibiity Study on ERP (the Study) was commissioned in March 1997, with the objective of examining the practicabiity of impementing an ERP system in Hong Kong and assessing the need for such a system to meet transport objectives. The Study has evauated ERP's essentia components, cost effectiveness and consequences, giving due consideration to transport panning, systems technoogy and pubic acceptance. The Study reviewed wordwide deveopment of road pricing, assessed future traffic conditions under a range of scenarios, considered aternatives to ERP, undertook fied evauation of the technoogy options, considered environmenta impacts, considered ega and procedura issues, investigated system design issues, expored integration with Inteigent Transport Systems (ITS) appications and prepared an outine for pubic consutation. ERP in Concept and in Practice v. In a congested environment, each additiona trip causes a other simiar trips to be deayed. This deay to other users has a cost which is referred to as the "externa" cost, and is much greater than the cost of the trip to individua users. The economic theory underying road pricing is to bring these costs into consideration and, as a resut, aocates road space more efficienty. Those prepared to pay obtain a higher quaity of service, and the absence of those not prepared to pay benefits a remaining road users incuding pubic transport passengers. vi. There are schemes being operated in the United States of America and Singapore which have proven that the concept is workabe. At present, ERP is being considered in the Netherands, United Kingdom and Japan.

3 Technoogica Options vii. viii. ix. Various technoogica options were assessed by the Study under a set of criteria incuding capita and maintenance costs, system accuracy and reiabiity, imitations, security against fraud, integration with existing to systems and smart cards, anonymity, etc. Two options, namey the Dedicated Short-Range Communications (DSRC) System and the Vehice Positioning System (VPS), were seected for fied evauation. The DSRC system is based on an interchange of information between roadside readers and in-vehice units (IVU) using ow power microwave communication. This is simiar to the operation of the existing Autoto system at various to faciities in Hong Kong. The VPS effects charges by an IVU based on the ocation of the vehice using the sateitebased Goba Positioning System (GPS). No roadside equipment is required at charge points but vioation enforcement stations are required at strategic ocations. A wireess data communication network is provided between the vehices and the contro centre for transaction data transmissions, database updating and enforcement verification. Fied Evauation x. Fied trias were conducted at both off-street and on-street test sites for a period of two months in ate 1998 under a range of conditions to determine the robustness and reiabiity of the systems. The resuts proved that both DSRC and VPS technoogies coud be adopted for an ERP system in Hong Kong. Whie both technoogies achieved a high degree of accuracy for its charging function, the enforcement accuracy of both systems coud be further enhanced by a vehicespecific coding system. xi. xii. DSRC technoogy is more mature and avaiabe for impementation in the near future. However, it requires roadside equipment at every charging point. This makes DSRC ess fexibe and adaptabe to changes in charging zones due to dynamic traffic conditions. It aso presents instaation difficuties because of the existing underground utiities making it progressivey more expensive with the introduction of each change. In addition, DSRC can ony communicate with the vehice when it is directy under the charging station. This imits its adaptabiity to ITS and Transport Information System needs. VPS technoogy, on the other hand, offers greater fexibiity and adaptabiity. It provides vehice ocation information and wide area communications to the driver. It may aso better fit future ITS and Transport Information System needs. Athough the present cost of VPS is higher than DSRC, the projected cost of VPS is dropping at a rate faster than that of DSRC. Hence, taking into account various aspects such as adaptabiity, fexibiity and better integration with ITS, VPS technoogy offers the best-baanced choice for ERP in the onger term. Aternative Traffic Management Measures xiii. Measures that coud improve the functioning of the transport system may be categorised into suppy management and demand management. Suppy management are measures that increase the suppy of road or rai capacity, add or improve pubic transport services. Demand management are measures that imit the number or usage of vehices using the road network. There are two aternative approaches to managing traffic demand. The first one is ownership restraint by reguation and fisca measures such as quota systems, parking poicies, FRT and ALF. The second aternative measure is usage management that imposes restraints on the use of vehices such as an "odd-even" number 3

4 4 xiv. pate system, restrictions on access or charging for usage. However, as the causes of traffic congestion differ from district to district, there is no simpe soution that can be appied universay. ERP and the aternative traffic management measures are not mutuay excusive. ERP can reinforce the effects of both ownership and usage restraints. It aso provides the financia incentive for motorists to switch to pubic transport. The diversion and mode switch make more efficient use of the existing road network by reducing the number of vehices on the roadway thereby deaying or eiminating the need to buid more infrastructure. Need for ERP xv. xvi. The existing poicy is that growth in the private vehice feet shoud not significanty exceed 3% per year. If the growth of the private vehice feet size is not greater than 3% per year, the traffic conditions wi be simiar to those experienced over the past two decades. As a resut, additiona restraint measures on the main east-west corridor on Hong Kong Isand may not be warranted on traffic management grounds before 2006, athough some ocaised traffic congestion may occur at the busy hours. Whether ERP wi be required after 2006 depends on pubic acceptabiity of the forecast traffic conditions in terms of traffic speeds, contro on the growth of the vehice feet, improvements to transport infrastructure and pubic transport services. The need for ERP wi aso diminish after 2010, as the competion of Centra-Wan Chai Bypass wi reieve the traffic congestion aong the main east-west corridor. For Kowoon, additiona restraint measures wi not be required at east unti 2011, the panning horizon of the current study. It is worth noting that the projected traffic speed on the main corridors on Hong Kong Isand in 2006 and 2011 are higher than those of present day Tokyo, centra London and New York. Tokyo and London are currenty considering road user charging systems to hep aeviate traffic congestion during peak periods. Possibe ERP System for Hong Kong xvii. xviii. xix. xx. The Study has ooked into a possibe ERP system for Hong Kong covering the foowing aspects. Charging Method - A cordon-based charging scheme is preferred in comparison with distance-based, time-based and congestionbased for its simpicity in both operation and enforcement, and technoogy readiness. Directiona charging, i.e. charging on entering in the morning and exiting in the afternoon, is preferred to bi-directiona charge whereby charging wi be imposed on both directions during the busy hours. Charging Zone - It shoud ony cover areas that are perceived to be the most congested and we served by pubic transport to provide opportunities for road users to change mode of trave. A singe zone encompassing Centra, Wan Chai and Causeway Bay is preferred to mutipe charging zones which wi add compexity to the scheme. Charging Period - Charging periods are considered in terms of their overa economic benefits according to traffic voume at different times of the day. The anaysis showed that peak hours charges shoud appy from 8:00 am to 9:00 am and from 5:30 pm to 7:00 pm and a sighty ower charge for the inter-peak hours. To reduce the probem of "bunching", shouder periods at ower charge rates for 30 minutes before the morning and after the evening peaks are proposed. No charge is envisaged for the overnight period from 7:30 pm to 7:30 am or on Sundays and Pubic Hoidays.

5 xxi. xxii. Charging Rate - The required eve of charge is determined by the target traffic speed to be achieved. Based on the historica average traffic speed, a target speed of 20 km/h was adopted for the mode tests. Charges under the test target speed range from $8 to $31 depending on the period of the day and the traffic demand growth scenarios. Exemption - ERP provides the fexibiity to differentiate charges by vehice type and to exempt a particuar vehice type from charge. However, exemptions other than for emergency vehices must be thought through carefuy because a trips contribute to congestion. Any exemption to be granted has to be evauated against the basic principes of equity, efficiency and pubic acceptabiity. Impact of Possibe ERP Scheme xxiii. xxiv. It was estimated that 40% of car trips in the morning peak may be diverted to pubic transport and 10% may change time of trave. The remaining 50% may stay-and-pay the road user charges but wi benefit from higher trave speeds and ess congestion. Whie the ERP may bring about some improvement in the air quaity in the charging zone, the environmenta conditions of other areas show some deterioration due to the overa redistribution of traffic in reaction to the ERP charging zone. Compementary Measures xxv. ERP wi divert car and taxi trips to pubic transport, producing a sma percentage increase in overa demand for pubic transport. Compementary measures such as providing park and ride faciities, improving pubic transport interchanges, providing feeder services and improving pubic transport information system shoud be considered. Impementation Strategy xxvi. ERP offers a more precise too to tacke traffic congestion in specific areas. Singapore experience indicates that a revenue neutra scheme woud increase pubic acceptance. However, it shoud be noted that Singapore maintained a quota on the vehice feet. The impementation strategy may, for instance, incude a reduction in FRT/ALF and use of the ERP revenue towards a traffic management and information centre and ITS appications to provide rea-time information about existing traffic conditions to road users and suggest modes of trave for pubic transport. Impementation Programme xxvii. Singapore took about six years to impement the ERP system. Given that the DSRC technoogy has since become more mature, the ead time for the instaation of a DSRC system in Hong Kong is expected to take about five years. The ead-time for VPS woud ikey be onger because there is no off-theshef product avaiabe in the market and the instaation of the in-vehice units may be more compicated. An estimated six years ead time woud be required. ERP & ITS Integration xxviii. The Study identified potentia areas that both DSRC and VPS technoogies coud be empoyed with other ITS appications in Hong Kong, most of which were demonstrated during the Fied Evauation. If ERP was not required to be impemented in the near future, the ERP technoogy coud sti be empoyed in feet management, taxi dispatch, non-stop toing, traveer information and integration with other ITS appications. These appications woud have a positive economic benefit and provide for greater operationa efficiencies that woud benefit motorists and 5

6 6 pubic transport aike. The overa effect of integrating ERP with ITS is synergy. Together, ERP and ITS woud provide a more robust, effective and cost efficient transport system than either coud deiver separatey. Benefits of ERP xxix. xxx. xxxi. The transport operation benefits reate to reduced vehicuar traffic, particuary during congested periods, reduced trave times, increased pubic transport use and higher car and taxi occupancy. The economic benefits derive from reduced traffic deays accrued to a road users and to the community at arge. The estimated net economic benefit resuting from journey time savings and ower vehice operating cost is about $2 biion/year. On the other hand, the estimated cost for the proposed ERP scheme is $1 biion (incuding the cost of in-vehice units for existing vehices) with an annua recurrent cost of $200 miion. ERP is forecast to generate annua gross revenue of $0.4 to 1.3 biion. If it is decided to adopt a revenue neutra scheme, this revenue can be poughed back for transport infrastructure investment. The environmenta benefits invove reduced vehice emissions and reduced exposure to traffic noise inside the charging zone. However, the environmenta conditions of other areas may witness some deterioration due to the overa redistribution of traffic. ERP therefore can assist in the overa improvement to the environment, but is not the soe soution to the compex issue of air quaity and noise abatement. If ERP is to be used to meet environmenta objectives, further studies wi need to be conducted to define the parameters, benchmarks and reated impementation detais. Pubic Consutation xxxii. Impementation of ERP wi not be possibe without a genera pubic consensus on the objectives and principes of the proposa. A we-panned and executed pubic consutation programme to demonstrate how ERP works and to aow pubic input into the deveopment of system objectives can promote awareness of the ERP system as a measure to reieve traffic congestion and deveop understanding in the community for its possibe introduction. The pubic shoud be consuted specificay on the acceptabe traffic condition in terms of traffic speeds, size of the charging zone, the charging eve, the type of vehices which may be charged and the use of the ERP revenue. Findings and Recommendations xxxiii. Major findings of the Study Both DSRC and VPS technoogies can be adopted for a possibe ERP system in Hong Kong. DSRC is more mature and ready for impementation in the near future with a ead time of about 5 years. VPS offers the best-baanced choice in the onger term because of its adaptabiity, fexibiity and better integration with ITS and the ead time is about 6 years. ERP technoogy coud be integrated with or empoyed in other ITS appications such as feet management, muti-ane toing and traveer information. ERP coud generate transport operation, economic and environmenta benefits. ERP and other aternative traffic management measures are not mutuay excusive.

7 ERP coud be used to support the funding of new transport initiatives and an integrated transport poicy. Whie ERP can be so designed to achieve environmenta objectives, a separate study wi be required to define the parameters, benchmarks and impementation detais. xxxiv. Key recommendations of the Study Drastic restraint measures are not warranted on traffic management grounds before 2006 if the growth of the private vehice feet is no more than 3% per year. Cosey monitor the GDP growth, vehice feet size and infrastructure programme. Activey monitor the deveopment of ERP technoogies and ERP projects in Singapore, USA, UK, the Netherands and Japan. Initiate a pubic consutation programme to promote pubic understanding of the traffic congestion probem and how a possibe ERP system wi work in Hong Kong. 7 xxxv. Specific issues for pubic consutation: Whether ERP is needed after 2006 in consideration of acceptabe traffic speeds, vehice feet size and expansion and improvement of transport infrastructures and pubic transport services? If needed, whether ERP shoud be impemented as a revenue neutra scheme. If needed, what shoud be the charging method, charging zone, charging period, charging rate and exemptions.

8 1. INTRODUCTION Background of the Study The economic boom of the mid-1990s and the associated growth in transport demand threatened a rapid worsening of traffic congestion. Consequenty, Government set up a Working Party in 1994 to examine measures to tacke traffic congestion in Hong Kong. One of the measures considered Eectronic Road Pricing (ERP) was the subject of this study Working Party Report 1.2 The Feasibiity Study of ERP was commissioned in March 1997, with the objective of examining the practicabiity of impementing an ERP system in Hong Kong and assessing the need for such a system to meet transport objectives. The Study has evauated ERP s essentia components, cost effectiveness and consequences, giving due consideration to transport panning, systems technoogy and pubic acceptance. 1.3 Unti the recession-induced sowdown in 1998, the number of vehices on the roads at times was increasing at a rate faster than that we fee comfortabe with. Nevertheess, average traffic speeds in urban areas were maintained at about 20 km/h over the past twenty years. This was achieved by a combination of ownership restraint measures incuding increase of First Registration Tax Growth of Vehice Feet in ast 20 years (FRT) and Annua Licence Fee (ALF), an efficient pubic transport system, road network improvements and effective traffic management schemes. 1.4 Recognising that the Hong Kong road network cannot accommodate an unrestrained number of vehices, the growth of the vehice feet has been controed in the past by means of occasiona increases in FRT and ALF. These ownership restraint measures together with reativey high operating costs due to fue taxes and parking fees have imited the growth of vehices to provide Hong Kong with one of the owest rates of cars per capita in the region. In Hong Kong there are ess than 50 private cars per 1000 popuation, compared to more than 100 in Singapore and Seou, 200 in Jakarta and 300 in Bangkok. Considering the high gross domestic product (GDP) per capita, these restraint measures were successfu as approximatey 11 per cent of daiy person trips are made by private vehices whie 89 per cent use pubic transport. Distribution of Persona Trips

9 1.5 The Study evauated means to improve the mobiity of peope and goods, based on the premise that the "user-pays" principe woud offer a more efficient, equitabe and fexibe means of managing the road space. In addition to evauating road pricing, aternatives to ERP were aso considered to compare the effectiveness of such measures. To accompish the project objectives, the Study aso reviewed compementary measures that coud be introduced in conjunction with road pricing. assessed future road network operating conditions under a range of future scenarios to determine the requirements for road pricing; considered the aternatives to ERP as approaches to manage or accommodate traffic growth; undertook fied evauation of the technoogy options; examined the effect of aternative ERP strategies under a range of economic and trave demand growth scenarios; Study Approach considered how ERP might be used to achieve environmenta objectives; 1.6 The Study seeks to provide the definitive answer concerning the feasibiity of introducing ERP in Hong Kong. The Study deveoped a structured framework that provides a sound basis for decision-making and pubic acceptance. To this end, the Study: reviewed wordwide deveopment of the concept and appication of road pricing; considered ega and procedura issues surrounding the impementation of ERP; investigated system design issues incuding an assessment of the technoogy, resource impications of impementing and operating ERP; expored the reationship between ERP and ITS for possibe cost savings and efficiency; prepared an outine of options for pubic consutation. 9 INCEPTION REPORT TRANSPORT MODELLING FORMULATION OF TECHNOLOGY OPTIONS NEED FOR AND ALTERNATIVES TO ERP INITIAL EVALUATION OF TECHNOLOGY OPTIONS PREFERRED ERP STRATEGY FIELD EVALUATION OF TECHNOLOGY OPTIONS CONCEPTUAL DESIGN OF ERP SYSTEM

10 2. ROAD PRICING IN CONCEPT AND PRACTICE Principe Underying Road Pricing but ceary in order for road pricing to be impemented charges must be subject to a genera consensus amongst users concerning acceptabe eves. It is quite possibe that as the benefits of congestion reief become apparent, the "toerated" eve of congestion wi become ower Charging for the use of infrastructure is not a new concept. Toed tunnes, bridges and highways have been in operation for centuries. Generay, tos are coected to cover the cost of provision of the infrastructure or for funding new infrastructures. In Hong Kong tos are evied on most road tunnes as we as in the Tsing Ma Contro Area. Autoto System 2.2 Unike to road systems, ERP charges vehices for usage of the road at times when its capacity fas we short of demand. It manages demand and, by varying charges to refect the degree of congestion, smoothes out peak fows. Road users are encouraged to choose whether, when and where to trave, depending on the importance of the trip, and the aternative routes, trave times or transport modes avaiabe. They may decide to change mode, trave time, destination or not to undertake the trip. Those who choose to pay and trave as before benefit from an improved quaity of service and more reiabe trave times, due to the absence of those who choose not to pay. 2.3 The price of road space can be set to produce any desired eve of service on the network, Economic Theory for Road Pricing 2.4 On congested roadways, each additiona trip causes a other simiar trips to be deayed. This is because each vehice reduces the speed of other road users. This deay to other users has a cost which is referred to as the "externa" cost, and is much greater than the cost of the trip to the individua user. The economic theory underying road pricing is that it is inefficient for these externa costs not to be refected in deciding whether to make the trip or not. Road pricing brings these costs into consideration and, as a resut, aocates road space more efficienty. Those prepared to pay obtain a higher quaity of service and the absence of those not prepared to pay benefits a remaining road users. 2.5 Motorists incur trave time and out-of-pocket costs for fue, tos and other vehice operating costs associated with each trip they make. Athough these costs may be substantia in congested conditions, a trip may sti be made if the vaue derived from the trip outweighs these costs. However, if the vaue is ess than the sum of the individua costs and the externa cost imposed on others, then road users as a whoe suffer. 2.6 By charging for the use of a congested road network, the motorist woud ony use the road if the vaue of the trip exceeded the sum of the road pricing charges and the other trip costs. In this manner, trip-making (for which the cost to society as a whoe exceeds the benefits) can be deterred. Congestion costs aso incude other externa costs, such as air poution, noise and accident costs. Hence,

11 it can be argued that the price to be charged shoud refect the overa socia cost of congestion, environmenta damage and accidents. Wordwide Deveopments in Road Pricing 2.7 There are operationa congestion pricing schemes on existing to faciities in the USA incuding the SR91 and I-15 Congestion Pricing Projects in Caifornia, Lee County Expressway in Forida and Hardy To Road in Texas. Unique in approach is the Caifornia SR91 project, in which new anes were added in the median. These new anes are toed in accordance with congestion or traffic eves on the free parae anes. As congestion of the free portion increases, the congestion charge for crossing over into the to anes aso increases. 2.8 In other countries, road user charging has been introduced to generate a new source of funding for buiding additiona infrastructure. In each case, traffic is charged on existing road space to generate a transport fund. This transport fund is used to buid new transport infrastructure and services. Traffic congestion reief is a secondary benefit of road pricing. Exampes incude the to rings in Bergen, Oso and Trondheim in Norway. South Africa, Tasmania and New Zeaand are panning a simiar approach to Norway's. In South Africa's Gauteng Province, the motivating factor is aso the ack of funding for transport and the attempt to make transport sef-sustaining. 2.9 ERP can aso be used as part of a sustainabe and integrated transport poicy. It acts to suppress demand whie other components of the poicy designate baanced and use and improve infrastructure suppy and pubic transport services. Utimatey, it aims to baance these objectives whie improving the environment. ERP's roe is to support these initiatives by freeing vauabe road space and thereby reducing the needs for road infrastructure and minimising environmenta impacts. It pays a roe in supporting pubic transport by diverting car and taxi users to bus, tram, and rai services. Exampes incuded the proposed Rekening Rijden Scheme in the Netherands, proposed Road User Charging scheme in London, Bristo, Leeds, Edinburgh and other cities in the UK, and road pricing studies in Dubin, Ireand and Tokyo, Japan. ERP Test in the Netherands ERP System in Singapore 2.10 In September 1998 Singapore became the first city in the word to impement an ERP system. Prior to the impementation of ERP, Singapore had a manua Area Licensing Scheme (ALS) in operation since The most congested parts of the city of about 720 ha was designated as the Restricted Zone (RZ) and an imaginary cordon was demarcated on the approach roads eading to it by 33 overhead gantries. This area, which is about 1.2% of the tota area of the state of Singapore, is served by a good arteria road network and has a ring road skirting it. To enter this RZ during the period of 7.30am to 7.00pm on weekdays and 7.30am to 2.00pm on Saturdays, vehices (other than emergency vehices such as fire engines, ambuances and poice cars) had to purchase and dispay an area icence on their windscreens or hande bars (for motorcyces) Area icences came in monthy and daiy forms and were coded by coour and shape 11

12 for easy identification by enforcement personne stationed at the entry points at the overhead gantries on the approach roads. The area icences had to be bought in advance from post offices, petro stations, area icence saes booths (ocated on the approach roads) or convenience stores prior to the entry. They coud not be bought at the entry points, as in the case of a to road. pass under these charging points on the three expressways during the hours of operation The ALS for the Restricted Zone (RZ) and the manua RPS for the expressways had been successfu in keeping traffic congestion within manageabe eves on the Singapore road network. The ERP System was impemented on 1 September 1998 to repace the two manua systems. The ERP system has the foowing advantages over the manua system in these areas: Eectronic Road Pricing in Singapore The area icences had different prices for different categories of vehices and aso for peak and off-peak usage Enforcement was carried out at the entry point. Poice stationed at the entry points during the hours of operation of ALS visuay checked whether the vehices dispayed a vaid area icence. Vioators were not stopped, but the detais of the vehices were noted and they were sent a notice to pay a fine for entering the RZ without a vaid icence. There was no poicing within the RZ. Vehices were free to move around or eave the RZ. Using the same area icence vehices coud make mutipe trips daiy into the RZ In the 1990's, foowing the success of the Area Licensing Scheme, a simiar manua pricing system caed the Road Pricing System (RPS) was introduced progressivey aong congested sections on three major expressways to operate on weekdays from 7.30am to 9.30am. Vehices had to purchase and dispay specia road pricing icences to Area Licence Scheme in Singapore it is easier to assign different rates to different vehices it is easier to change rates it is easier to change the hours of operation it is reativey easier to incude more areas under pricing or to remove areas from pricing 2.15 The Singapore ERP System empoys dedicated short range communication (DSRC) technoogy using microwave wireess communication for interchange of information between overhead readers and the in-vehice units (IVU). Gantries were instaed at a charging points for supporting the overhead charging and enforcement equipment. Each vehice passing through is detected and cassified by the system, and the appropriate charge is sent to the IVU through the DSRC ink. The IVU then deducts the amount from a smart card inserted in the

13 IVU. If the charge is made successfuy, there is no need to record the vehice's identification or its ocation. If there is insufficient baance on the smart card or intent to evade payment, the enforcement cameras wi capture an image of the vehice's number pate for enforcement As compementary measures, rebates on annua icence fees were offered, an estimated 15,000 to 20,000 additiona Certificates of Entitement (CoE is a document required for purchase of a vehice) was made avaiabe over a three year period, and park and ride faciities at MRT stations and bus interchanges have been introduced After the change of ALS to ERP, there are significant reductions in the number of vehices entering the RZ during the morning peak hours and for the rest of the day. The daiy reduction for the ERP period after one year of operation as recorded in August 1999 is 15%, whie the reduction for the morning peak traffic period of 7.30am to 9.30am is about 16% ERP rates in Singapore are reviewed at quartery intervas. Changes are triggered by the prevaiing traffic speeds aong the roads within the RZ and the expressways during each haf-hour interva. The ERP rates wi be adjusted if the prevaiing traffic speeds fa outside the optima ranges of 20 to 30 km/h and 45 to 65 km/h in RZ and the expressway respectivey. When the average prevaiing speeds on the affected section of the expressway or seected roads in the RZ was beow 45 km/h or 20 km/h respectivey, the ERP rates wi be raised. Simiary, when speeds exceeded the higher vaue of 65 km/h or 30 km/h, ERP rates wi be reduced. These speed figures for rate adjustments are made pubic and such adjustments of ERP rates have found favour with the motorists. In essence, the motorists' behaviour decides the ERP rates. Since the start of the scheme, ERP rates for the RZ have been revised twice within the first year and the charges on Saturdays discontinued. Piot Scheme on ERP in Hong Kong 2.19 Hong Kong conducted a Piot Scheme on ERP from Juy 1983 to March It demonstrated that ERP was technicay feasibe and coud produce economic benefits. The then-proposed technoogy tested over a six-month tria period was a form of automatic vehice identification, in which each vehice had an eectronic number pate, the size of a videocassette tape, mounted underneath. This device was queried when it passed over a power oop embedded in the road, reaying the vehice's identification number to a roadside computer. Cosed circuit teevision identified vehices whose automatic number pates were fauty or tampered with. There were eighteen to sites around the Centra Business District (CBD) on Hong Kong Isand and 2,500 vehices were fitted for the tria. A centra accounts computer produced a biing for each vehice For a number of reasons, ERP was not impemented and the Study was sheved. One of the factors was the improved traffic conditions resuting from the triping of ALF and the doubing of FRT and fue tax in There was an economic downturn at the time of the Study that raised concerns over the addition of another perceived tax on the road user. Another infuence was the opening of the Isand Line which provided better pubic transport and improved trave times. The combination of these factors provided a genera sense of improved transport operating conditions and undermined the need for further restraint measures. In addition, there was strong objection against the possibe invasion of privacy resuting in a ow eve of pubic acceptabiity. 13

14 3. TECHNOLOGICAL OPTIONS FOR ERP SYSTEM 3.4 Desktop Review " 3.1 A desktop review of wordwide road pricing and reated experience together with survey on ERP components currenty avaiabe in the market or under deveopment were conducted to identify technoogica options that may be suitabe for impementing ERP in Hong Kong. 3.2 The various technoogica options were then assessed under a set of 20 criteria incuding capita and maintenance cost, system accuracy and reiabiity, imitation, security against fraud, integration with existing to system and smart cards, anonymity, etc. Two options, namey Dedicated Short Range Communication (DSRC) System and Vehice Positioning System (VPS), were seected for fied evauation. Reader Gantries or stee frames attached to fyovers or other overhead structures are instaed at the charge sites for supporting the roadside equipment. Each vehice passing under the gantry is detected and cassified by the system, and the appropriate charge is sent to the IVU through the DSRC ink. The IVU then deducts the amount from a smart card or transmits an account number for debiting the charge to a centray-hed account, if the motorist opts for it. If the charge is made successfuy using a smart card, there is no need to record the vehice's identification or its ocation. The privacy of the individua is thus protected. DSRC In-Vehice Unit with Smartcard 3.5 Camera If there is insufficient baance on the smart card or intent to evade payment, the vioation enforcement system (VES) instaed at the charge site captured an image of the vehice's icense pate for enforcement. Vehice Detector Roadside Cabinet The Dedicated Short-Range Communications System 3.3 The DSRC system is based on an interchange of information between roadside readers and in-vehice units (IVU), using 5.8 GHz microwave communication between the IVU and overhead readers over short distances and ow power emissions. This is simiar to the operationa characteristics of the existing AutoTo system operating at various to faciities in Hong Kong. VPS In-Vehice Unit with User Interface 3.6 The VPS effects charges by an IVU based on the ocation of the vehice using the sateitebased Goba Positioning System (GPS). Positiona accuracy is improved using Differentia GPS (DGPS), dead reckoning through a soid-state compass and odometer

15 connection, and map-matching for parts of the network where GPS reception is impossibe, such as tunnes. 3.7 Data defining the charging zones, charging periods and charge rates are stored in the IVU. When the IVU determines that it has entered a charging zone, it deducts the appropriate charge from the smart card without the need to communicate with any roadside equipment. If the smart card is charged successfuy, there is no need to record the vehice's identity or ocation, thus protecting privacy. Centra accounting is specified as an aternative, in which case the identity of the vehice together with accumuated charges are periodicay transmitted back to the contro centre for processing. Fied Evauation 3.10 Fied trias were set up at two off-street tracks at Kai Tak and five fixed on-street sites in Wan Chai. The trias took pace over a two-month period with 75 equipped vehices. The test programmes were designed to assess the suitabiity of each system as the basic technoogy to impement ERP in Hong Kong. Tests incuded a range of conditions designed to provide information about the performance of the system in adverse conditions and to determine their robustness and reiabiity. 15 GPS Sateites Vehice cassification sensor Smart Card Vehice testing under Gantries at Kai Tak Performance of Tria Systems Portabe VES The Vehice Positioning System 3.8 No roadside equipment is required at charge points. A wireess data communication network is provided between the vehices and the contro centre for transaction data transmissions, database updating and enforcement verification. 3.9 Enforcement for VPS is carried out at a few fixed enforcement sites and suppemented by portabe and mobie equipment. The enforcement equipment consists of a Vehice Detection and Cassification (VDAC) subsystem and a Vioation Enforcement System (VES). Images of the rear of vehices identified as potentia vioators are transmitted to the contro centre The performance of an ERP system is measured by its abiity to charge vehices correcty and to detect vioation vehices. In the fied evauation, an overa weighted average accuracy of 99.24% and 99.25% was achieved for transactions under DSRC and under VPS systems respectivey. With finetuning and system adjustments, suppementary testing suggested that 99.99% accuracy was achievabe with both technoogies Both DSRC and VPS systems use a vehice detection and cassification sub-system (VDAC) and vehice enforcement sub-system (VES) for enforcement. The camera wi be triggered to take an image of the vehice icence pate if a proper transaction cannot be verified. The accuracy of the enforcement system for DSRC was 70%. The accuracy of the VPS enforcement system tested in the fied

16 16 evauation at ony 24% was far from satisfactory and required substantia fine-tuning of the system. It is expected that the accuracy of the VPS enforcement system wi improve with better system integration and technoogica advancement in the coming years The accuracy of both enforcement systems shoud be abe to be improved to over 90% by incuding a vehice specific coding system, in which a unique seria number is assigned to each in-vehice unit. If a vaid transaction cannot be made, the roadside computer wi read the seria number of a vioation vehice and extract its identity from the database kept by the operator The fied evauations demonstrated that technoogy exists to impement an ERP system whie sti protecting the privacy of the individua as defined in the Persona Data (Privacy) Ordinance (see ). However, if a higher accuracy rate (above 70%) of the DSRC or VPS enforcement system is considered necessary, a vehice-specific coding system is required. The coding system can hep to identify the vioation vehice but may aso expose the identity of the motorists. Consideration wi need to be given as to how privacy issues shoud be baanced against enforcement accuracy. Technoogica Option for Hong Kong 3.15 Fied evauations proved that both DSRC and VPS technoogies coud be adopted for an ERP system in Hong Kong. DSRC technoogy is more mature and avaiabe for impementing ERP in the near future. It has been used successfuy for the ERP systems in Singapore and free-fow, non-stop to coection systems in the United States, Canada and Austraia. However, it requires overhead and/or roadside equipment at every charging point. This makes DSRC ess fexibe and adaptabe to changes in the charging zone. It aso presents programme issues for instaation because of the prevaence of utiities aong sections of the existing road network. The dynamics of traffic and the need to adjust and/or expand the charging zone over time aso make it progressivey more expensive with the introduction of each change. In addition, DSRC can ony communicate with the vehice when it is directy under the communications antenna. This imits its adaptabiity to Inteigent Transport Systems (ITS) and Transport Information System needs VPS, on the other hand, offers greater fexibiity and adaptabiity. It provides vehice ocation information and wide area communications to the driver at a ower systems cost. It aso better fits with future ITS and Transport Information System needs. However, whie such a system has been tested in Germany, New Zeaand and Tasmania, there is as yet no working VPS based ERP system. Furthermore, the present cost of VPS is higher than DSRC, but the projected cost of VPS is dropping at a rate faster than that of DSRC. It is estimated that VPS wi be about 2% more expensive than DSRC in the immediate future. Hence, taking into account various aspects such as adaptabiity, fexibiity and better integration with ITS, VPS-based technoogy offers the best-baanced choice for ERP in the onger term. It is recommended that the VPS technoogy be activey monitored over the coming years Locay, Goba Positioning Systems (GPS - the core system upon which the VPS system is buit) are finding more appications. The CityBus Company has used GPS ocation system for their bus feet management for Route No. 11. Both the Fire Services Department and Poice are instaing their feet management and dispatch systems using GPS. Athough their appications are different from the VPS based ERP system, their experience shoud aso provide vauabe insights into the use of positiona ocation technoogy over the next severa years.

17 4. ALTERNATIVE TRAFFIC MANAGEMENT MEASURES Measures to Improve the Transport System 4.1 Measures which woud improve the functioning of the transport system can be divided into two categories: suppy management; and demand management. Suppy Management 4.2 Suppy management incudes measures that increase the suppy of road or rai capacity, either through new construction or widening, or through techniques designed to maximise or enhance the capacity of existing infrastructure. The foowing measures have been considered: road infrastructure - unikey to provide significanty increased capacity due to cost, environmenta impact and socia concerns particuary in buit-up areas. pubic transport provision - spare road capacity as a resut of shift from private to pubic transport is ikey to be taken up by suppressed demand quicky. traffic management measures - probaby scope for improvement at oca spots adding itte overa system-wide capacity. Demand Management 4.3 There are two aternative approaches to managing traffic demand. ownership restraint addresses the congestion probem at source by discouraging or imiting the ownership of vehices by - reguatory measures such as quota systems and parking poicies (avaiabiity, contros and charges) - fisca measures such as First Registration Tax (FRT) and Annua Licence Fee (ALF). Both are effective measures. ALF in particuar causes car owners to re-evauate the need for the vehice reguary. usage management imposes restraints on the use of vehices by - operating permits such as the "oddeven" number pate system. - restrictions on access such as pedestrianisation scheme. - pubic transport priority such as bus ony anes. - high occupancy vehice priority to encourage car pooing. 4.4 Ownership or usage reguation may be suppemented by introducing incentives to encourage more efficient use of vehices and the transport system in genera. Such poicies, cassified as behavioura incentives, incude: park and ride faciities at rai stations and major pubic transport interchanges; inteigent transport systems with a variety of measures to provide better information to the commuters; mode change incentives such as vanpooing subsidised by the empoyer; fexi-time or staggered working hours; and tee-commuting. 4.5 Finay, fisca contros may be empoyed to suppement other poicies by imiting vehice use through pricing. Fisca contros incude: fue tax; and road tos at tunnes and bridges. 17

18 18 Aternative Measures to ERP 4.6 As can be seen above, different aternatives to ERP are avaiabe to tacke forecast traffic congestion. Fisca measures such as FRT and ALF have been used to contro vehice ownership successfuy in the past whie traffic management measures such as bus priority schemes and park-and-ride schemes have been impemented to reieve oca traffic congestion. The increased use of ITS wi enhance the efficiency of the road network and thus better manage the avaiabe suppy and smooth traffic fows without buiding more roads. As the causes of traffic congestion differ from district to district, there is no simpe soution that can be appied universay. 4.7 Raising FRT and ALF are rather bunt but very effective measures in improving traffic conditions. Increasing them to high eves woud have a significant impact on traffic voume at the busiest times and ocations. Yet, it undermines peope's desire to own a car for eisure trip and use it in areas which are ess congested or at times when the traffic is ess busy, e.g. on Sundays and Pubic Hoidays. 4.8 The adoption of the "user-pays" principe woud offer an efficient and fexibe means of managing the road space. ERP appies the "user-pays" principe for managing road traffic demand. It compements the continued "suppy-side" oriented infrastructure programme by managing demand to baance traffic voumes with avaiabe capacity. It aows a pricing structure to be appied that is time and congestion sensitive. Charging on a time and distance basis reates the charge to the frequency and duration of use. Other aternatives tend to be ess precise in managing traffic congestion where it occurs. 4.9 ERP and the aternative traffic management measures are not mutuay excusive. ERP can operate with existing poicies, in conjunction with the other aternatives or as the primary measure. It reinforces the effects of ownership restraints such as ALF, FRT, usage restraints such as parking poicies and charges, and moda priority measures such as pubic transport priorities, and park-and-ride. ERP, ike FRT and ALF, provides the financia incentive for motorists to switch to pubic transport. It aso provides the incentive for drivers to use a park-and-ride faciity rather than incurring ERP charges. The diversion and mode switch make more efficient use of the existing road network by reducing the number of vehices using the roadway. This effect can deay or eiminate the need to buid more infrastructure. 5. NEED FOR ERP Traffic Forecast 5.1 An ERP Traffic Mode (ETM) was deveoped to estabish traffic forecast and assess the impact of ERP on traffic demand and performance of the road network. ETM is compatibe to the mode deveoped in the Third Comprehensive Transport Study (CTS- 3) except more detaied zoning and road network are incuded to better represent traffic characteristics in the urban areas. 5.2 The Study adopted a range of possibe growth scenarios in forecasting future traffic demand. Each scenario comprised different assumptions on socio-economic factors incuding Gross Domestic Product (GDP), popuation and empoyment, private and goods vehice feet sizes and infrastructure provision. 5.3 The Study identified that depending on the growth of the private vehice feet the average traffic speed in Centra, Western, Mid-Leves, Wan Chai and Causeway Bay

19 woud be 17.8 to 19.7 km/h in 2006 and 20.6 to 23.9 km/h in 2011 as a resut of infrastructure projects expected to be competed by then. The average traffic speed in Kowoon woud be 23.4 to 25.9 km/h in 2006 and 23.0 to 25.3 km/h in In comparison, the average traffic speeds in the urban area have been maintained at about 20 km/h over the past twenty years. Need for ERP 5.4 The existing poicy is that growth in the private vehice feet shoud not significanty exceed 3% per year. If the growth of the private vehice feet size is controed to no more than 3% per year, the forecast trave speeds wi be towards the upper end of the average speed ranges noted in 5.3 above. The traffic condition wi be simiar to that experienced over the past two decades. As a resut, additiona restraint measures may not be warranted on traffic management grounds before 2006 athough some ocaised traffic congestion may occur at the busy hours. Whether ERP wi be required after 2006 depends on pubic acceptabiity of the forecast traffic conditions in terms of traffic speeds, contro on the growth of the vehice feet, improvements to transport infrastructure and pubic transport services. The need for ERP wi diminish by 2010, with the competion of Centra-Wan Chai Bypass that wi reieve the traffic congestion aong the main east-west corridor. 5.5 The community needs to decide whether those projected speeds are acceptabe. If growth in the private vehice feet size is not greater than 3%, the forecast traffic speeds in Hong Kong in 2006 and 2011 are higher than the existing traffic speeds of 18.5 km/h in Tokyo, 16.0 km/h in centra London and 11.0 km/h in New York. However, the traffic speeds in Tokyo and London has aready raised the attention of their respective Government who pans to consut their citizens on a package of measures to reieve congestion, incuding a possibe ERP system. 5.6 There is therefore a need to consut the pubic on toerabe traffic speeds and cosey monitor the GDP growth and vehice feet size as we as the infrastructure programme. 5.7 Furthermore, essons can aso be earnt from the deveopment of road pricing projects currenty under consideration in the Netherands, Japan, the United Kingdom and the United States. 19 Traffic Conditions in London

20 20 6. POSSIBLE ERP SYSTEM FOR HONG KONG Conceptua Design of ERP Scheme 6.1 The Study has ooked into a possibe ERP system for Hong Kong. The definition of an ERP scheme invoves specifying the foowing parameters: type of charging system geographic extent of the charged area number of charging zones time periods to be charged vehice types to be charged charge eves by period and vehice type compementary measures to be introduced. 6.2 The Study has considered different types of charging methods, incuding Cordon-based: vehices are charged each time they pass an operationa charging point. Distance-based: charges are reated to the distance traveed within the area during the charging period. Time-based: charges are reated to the time spent within the charged area: either the time the vehice is in operating mode, or the tota time (incuding inactive time). Congestion-based: charges are reated to the time spent in the charged area whie traveing beow a set (congested) speed for onger than a set time 6.3 Time-based and congestion-based charging are not preferred because they may encourage dangerous driving, as motorists seek to minimise the time spent in the charged area. Secondy, the charge for the same daiy trip may vary and cause confusion on the cost of the trip. A cordon-based charging scheme is recommended for its simpicity. Athough both technoogica options can be adopted for distance-based charging, proper enforcement may be difficut and DSRC wi aso require additiona connection to the odometer. Directiona charging, i.e. charging woud ony be imposed on motorists entering the charging zone in the morning and eaving the zone in the afternoon, is preferred to bi-directiona charging whereby charges woud be imposed on both directions during the peak periods. 6.4 A practica and pubicy acceptabe ERP scheme shoud ony cover areas that are perceived to be the most congested. The charged area shoud aso be we served by pubic transport to provide opportunities for road users to change mode. Different options for the charging area have been studied. A typica charging zone encompassing Centra, Wan Chai and Causeway Bay is shown overeaf. The zone incudes wings to discourage the use of Mid- Leves as a bypass to the charged area. Both the Eastern Harbour Crossing and Western Harbour Crossing are excuded from the charging zone and act as aternative routes for North-South movement. 6.5 The use of mutipe charge zones woud aow ower rates to be charged on each cordon, thereby reducing the equity probem associated with high charges for short trips across a singe zone boundary. Secondy, it woud intercept trips whoy within the charging area of a singe zone system, which woud otherwise be free of charge. The main disadvantage is that it adds compexity to the scheme, thereby reducing pubic acceptabiity. 6.6 The Study evauated aternative charging periods in terms of their overa economic benefits. Typicay, there is a steep rise in traffic voume towards the morning peak. The same traffic voume remains throughout

21 Typica Charging Zone