Interim Guide On Identifying Prioritising and Treating Hazardous Locations On Roads in Malaysia - JKR 20708-0022-95
September 2, 2022 | Author: Anonymous | Category: N/A
Short Description
Download Interim Guide On Identifying Prioritising and Treating Hazardous Locations On Roads in Malaysia - JKR 20708-002...
Description
JKR 20708-0022-95
Interim Guide On Identifing, Prioritising And
Treating Hazardous
Locations On Roads
In Malaysia
m m .0 7
m m .0 5
Roads Branch Public Works Department Malaysia Jalan Sultan Salahuddin 50582 Kuala Lumpur
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
FOREWORD Road accidents have now become a major concern in Malaysia. The latest accident figures from the Royal Malaysia Police show that the numbers have increased by 23% and 15% over the last two years alone, with 135,995 cases recorded for 1993. Malaysia's rapid economic growth may be partly par tly res respon ponsib sible le for for this this worry worrying ing trend trend as as the the count country ry is is curre currentl ntly y seein seeing g regis register tered ed vehi vehicle cless increase by over 7 % per year. In order to sustain high rates of economic growth, utilisation of the skills of our people is of paramount importance, and the nation can ill afford to waste such valuable resources in road accidents. Many would argue that to minimise human suffering in any way possible is sufficient justification in itself to devote greater efforts to reducing the road accident casualty toll. Realising this, the Government set a reduction target in 1991 to reduce road accident fatalities by 30 per cent by the year 2000 with 1989 chosen as the base year. To achieve this target, all relevant Government Departments and Agencies need to contribute to this effort. The approach to be taken follows the 3 E's concept: Engineering, Education and Enforcement. Jabatan Kerja Raya (JKR) being bei ng the the main main Engine Engineeri ering ng arm arm of of the the Gover Governme nment nt can can play play a major major rol rolee in thi thiss respe respect. ct. In line with this, Institut Kerja Raya Malaysia (IKRAM) has undertaken a road safety research progra pro gramme mme in coll collabo aborat ration ion with with the the Tran Transpo sport rt Rese Researc arch h Labor Laborato atory ry (TRL (TRL)) of the Unite Unite King Kingdom dom.. As part of this initiative, IKRAM is now able to provide a reference guide for use by all practising road engineers in Malaysia. I am very pleased to be able to introduce this document, entitled Interim Guide on Identifying, Prioritising and Treating Hazardous Locations on Roads in Malaysia and am grateful to the Overseas Development Administration, U.K. for their contribution to the funding of this project. This interim guide provides information specifically for Malaysia and sets out a standard methodology for analysing accident data to help identify the most hazardous locations, select appropriate remedial measures, and evaluate this action. It is my hope that this Guide will be useful to all who are involved in the road safety field, helping them to channel limited resources in a more efficient manner, and thereby ensuring that our road network is as safe as it can possibly be made.
(T (Tan an Sri Dato’ Ir Wan Wan A Rahman Yaacob) Yaacob) Director General of Public Works Works Jabatan 1995 Kerja Raya Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 1
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
ACKNOWLEDGEMENTS This Guide has been prepared within the Road Safety Group of the Institut Kerja Raya Malaysia (IKRAM) in association with the Transport Research Laboratory (TRL), United Kingdom. The author of the Guide is: Mr. Chris Baguley
-
TRL.
As its intended users are all road authority engineers in Malaysia, the Guide was reviewed at various stages of its production by the following representative Committee: Ir. Mohamed Shafii Mustafa IKRAM -Chairman Pn. Subiah Sulaiman IKRAM -Secretary Pn. Norliah Saidin Highway Planning Unit, Min.of Pub Works Ass. Prof. Radin U R Sohadi Universiti Pertanian Malaysia P/PPP Ru R uslan b. Khalid Polis Di Raja Malaysia PPP Ooi In Boo Polis Di Raja Malaysia Ir. Sabudin Mohd Salleh Dewan Bandaraya Kuala Lumpur En. Sanusi b. Ismail Jabatan Kerja Raya Daerah, Hulu Langat and representatives from Jabatan Perumahan & Kerajaan Tempatan; Majlis Keselamatan Jalan Raya; and Cawangan Jalan, Jabatan Kerja Raya. The author is indebted to the above committee members for their valuable contributions, and to En. Othman Hussin of IKRAM for his assistance in preparing some of the material. Gratitude is also extended to Doncaster Metropolitan Borough Council and Berkshire County Council/Babtie Group for permission to reproduce parts of their Road Safety Plans, and to the Royal Society for the Prevention of Accidents (UK) for various extracts from their Road Safety Engineering Manual. Finally, the author would also like to express his gratitude to the Director General of Public Works Malaysia for his permission to publish the Guide.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 2
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER I INTRODUCTION 1.1 PURPOSE AND SCOPE This Guide is intended for use by traffic engineers and road safety officers in the Public Works Department (JKR), Ministry of
country.
Transport (MoT), and all Local Authorities of Malaysia.
found to be effective in many countries of the world. It is hoped that this is achieved in as concise a manner as possible, and it is therefore recommended that for more in-depth discussions of the various topics the reader should make use of the References listed.
It has been estimated that about 350,000 people die in road accidents in developing countries each year. This represents 70 per cent of those killed on the roads throughout the world. In Malaysia there are about 4,500 fatalities and over 36,500 injured per year. This means that more than 1 person in every 450 of the country will suffer injury or death in a road accident each year. This situation is worsening, and traditionally the "three E's" have invariably been quoted when discussing ways of tackling the proble pro blem m namel namelyy Education, Enforcement and Engineering A fourth "E", Encouragement (by (by setting targets, support for initiatives, publicity material to promote positive attitudes, etc) has now also been added.
This Guide is designed to be an easytoassimilate outline of procedures that have been
Indeed the Guide has been based on several of these publications, particularly the Indonesian Accident Investigation Procedures Manual, the Institution of Highways and Transportation Transportation (UK) Guidelines for Accident Reduction and Prevention2, UK Department of Transport's Accident Investigation Manual3, and RoSPA Road Safety Engineering Manual4. 1:1.1 Summary of contents
Chapter I as an introduction, highlights the scale of the accident problem of Malaysia and introduces the types of strategies generally applied in many other countries to reduce accidents. The national accident reduction target is quoted and the need to manage road safety stressed.
Although the most effective approach for many road safety initiatives will be a combination of these elements, it is the
Chapter II is devoted to the accident database of Malaysia, its production, and the responsibilities of all those agencies
Engineering on which this focused. It isapproaches likely that some aspect of Guide is highway design, layout, state of road or traffic control is a contributory factor in most accident occurrence. It is well established that considerable safety benefits may result from the application of appropriate road engineering or traffic management measures at hazardous road locations. In order to reduce accidents effectively and help to achieve the nationally set reduction targets, it is essential that a systematic approach to the identification of hazards and selection of appropriate treatments be carr carried ied out at the the loca locall level level thr throug oughou houtt the the
contributing to its content.
Cawangan Jalan, Ibu Pejabat JKR, K.L
The third and subsequent chapters include a step-by-step approach (totalling 10 steps) to tackling the safety problems within a road authority's area. These steps are illustrated in the flow chart shown in Fig. 1.1. The three steps of Chapter III comprise the investigation process. A list of the worst blacks bla ckspot pot sit sites es need needss to be prod produce uced d first first fro from m the computerised database. Preliminary analysis is described and early initial site visits are also recommended. Page 3
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 4
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Chapter IV (Steps Chapter (Steps 4 and 5) deals with diagnosis of the problems, including the collection of sketch diagrams and other available data together with the likely need to carry out specific site studies. Detecting accident patterns (a skill to be acquired) is introduced.
points in the network. For example, even at a “high risk” junction in Malaysia having 42,000 vehicles (16-hour count) passing through it and where 53 accidents occurred in one year with 10 involving injury, the actual accident occurrence rate is still relatively low. These accidents occurred during about 18,300,000 vehicle
The two steps steps (6 and 7) of Chapter V are concerned with the selection of possible countermeasures and prioritising both these and the sites to be treated. Lists of the most common problems and treatments are included. A method of carrying out cost benefit estimates to assist in i n the decisionmaking process for the most effective measures is also described.
movements. That is, one damage-only accident every 425,000 movements and one injury accident every 1,830,000 movements.
Chapter VI discusses the implementation stage (Step 8), that is detailed design and installation, briefly, and emphasises the need for safety audits and for carrying out all
The precise moment when an accident will occur can never be predicted, and even if long periods such as a year are considered the numbers will fluctuate randomly about a longer term average. Statistical tests can show whether, during one particular ‘high’ year, a real change has occurred. The occurrence of accidents along the network tends to be less random as accidents are often clustered at so-called "hazardous
roadworks as safely as possible.
locations" or "blackspots".
The final two steps (9 and 10) in Chapter VII cover evaluation of the completed works. Some observational measurements for monitoring the sites are described in brief. Simple statistical techniques are outlined to estimate the size of the effect of the measures introduced. All necessary statistical tables are contained in the Appendices, as are examples of the Police POL27 accident report form and UK Road Safety Plans (see 1:5.1).
1:2 WHA T IS A AC ACCIDEN CIDENT? A fullWHAT definition of ROAD a road traffic accidentT? is: "a rare, random, multi factor event always preceded by a situation in which one or more road users have failed to cope with their environment, resulting in a collision on the public highway which should be recorded recor ded by the police".
Although, from the previous section, it may be concluded that road accidents are far too common in Malaysia (see Section 1:5), they are comparatively rare events at specific Cawangan Jalan, Ibu Pejabat JKR, K.L
There is rarely an accident situation in which only one "thing" or person is truly the sole cause of the accident: hence accidents are multi-factor events. There are three basic categories of factors:
Road user errors
Page 5
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Road and environment faults
Vehicle defects
It has been shown that road user errors are by far the most prevalent factor but often another factor(s) is present. For example, “adverse environment” implies a situation where a driver has had difficulty manoeuvring a vehicle safely; drivers are frequently provided with insufficient or unclear information with respect to signs and markings; sometimes poor design can cause a driver to have a misleading visual impression; and occasionally defective tyres and brakes can contribute significantly to an accident occurring. The chain of events leading up to an accident include people’s experiences on that day which can lead to stress or lack of adequate concentration on the driving task, and ultimately mean one or more persons failed to copethat with their environment. An assessment of the factors in the chain could indicate which road and environment factors may benefit from safety engineering remedial measures. Consider the example of a junction where skidding on a wet road surface has been recorded in several accidents by the police. Care in analysis is needed as an inexperienced investigator may immediately recommend that the skid resistance of road surfaces on the approaches to the junction need to be upgraded. Careful study of the site (during similar conditions), however, may reveal that tree branches weighed down by rainwater tend to obscure road signs and advance warning signs. The drivers thus failed to cope primarily because of obscured signs rather than the wet road surface itself. Treating the road surface may still be necessary but would be much less effective if the trees were not cut back or the signs not re-sited for improved visibility. visibility.
Cawangan Jalan, Ibu Pejabat JKR, K.L
1:3 APPROACHES TO IMPROVING ROAD SAFETY The main objective of improving road safety through road engineering and traffic management simply means measures taken primarily to avoid some accidents happening in the future, or at the very least, reducing the severity of future accidents. This may be achieved by following two distinct approaches: (i) (ii)
ACCIDENT ACCIDEN T PREVENTION PREVENTION ACCIDENT REDUCTION
ACCIDEN ACC IDENT T PREVENT PREVENTION ION
Involves the application of safety principles in the planning, design, upgrading and maintenance of roads. (See ref 5: The safety audit of highways). ACCIDEN ACC IDENT T REDUCTIO REDUCTION N
Involves the application of appropriate road engineering or traffic management schemes at hazardous locations on the existing road network. Such applications, particularly those of relatively low-cost, at known high accident locations or “blackspot” have yielded very high returns in many of the more industrialised countries of the world. Although accident prevention plays an extremely important role in maintaining a safe environment, this Guide concentrates on accident reduction and is aimed at engineers and technicians who have responsibility for safety on the existing road network. It is recommended that for more information about accident prevention the reader refer to ref.6: ‘Towards safer roads in developing countries - a guide for planners and engineers’.
1:4 THE FOUR BASIC STRATEGIES The four basic strategies for accident reduction through the use of countermeasures are:
Page 6
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Single sites/Blackspots the treatment of specific types of accident at
a single location [Eg. usually junctions, but could be areas 200-400m in diameter, 300-500m stretches of road] Mass action schemes
The application of a remedy to locations with a common accident problem. [Eg. skidding on wet road surface, head-on collisions, excessive speed approaching roundabouts] Route action plans
The application of remedies along a route with a high accident rate. Area-wide schemes
The application of various treatments over a wide area of town/city. town/city. [Eg. including traffic management and traffic calming (speed reducing devices) in areas bounded by links on a network, housing areas or l km squares having higher accidents than a preset level]. Blackspot treatment is likely to be the most Cawangan Jalan, Ibu Pejabat JKR, K.L
effective as amoving startingto point, theand roadstraightforward authority perhaps the other wider types of application as experience is built up. All these strategies rely on the availability of data which contains full information about accidents and their locations so that common features which have contributed to the accidents can be identified. Accident data and the use of collision and stick diagrams, key analysis tools for the traffic engineer, are discussed in Chapters II and III.
1:5 SAFE SAFETY TY MANAG MANAGEMENT EMENT One of the most effective initiatives in improving safety has been found to be the setting of realistic accident reduction targets around which all authorities can properly plan reduction programmes. In 1991 the Government of Malaysia set a national target:- to reduce the number of fatalities resulting from road accidents by 30 per cent by the year 2000. This was based on the year 1989 which means, in practise, a reduction from 3773 to 2641 deaths per year (fig. 1.3). Page 7
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
This was expressed the special target into (which needs to be targets) to Committee for Roadby Safety in 1990Cabinet in terms national disaggregated State and local of a fatality rate reduction from 7.12 down be reflected in safety initiatives for each to 3.14 deaths per 10,000 vehicles registered local area. to be achieved by the year 2000. The somewhat larger percentage reduction For example, Fig. 1.5, from the U.K.’s expressed in the form of fatality rate is due Institution of Highways & Transportation’s to the fact that this takes some account of (IHT) road safety guidelines, illustrates the the fairly substantial, steady increase in large number of policy initiatives which can vehicular traffic in Malaysia. The be input into each local area scheme, ie. the assumption was that of continuing linear integrated approach where agencies need to growth in the number of vehicles registered co-operate in agreeing targets and plans, and as that experienced during the past ten years. need to monitor the effects of the safety (see fig. 1.4). work. It can be seen from fig. 1.3 that there is a generally increasing trend in deaths which is particularly marked in more recent years. This disturbing feature makes it more important than ever for all authorities to work hard to achieve their particular target.
1:5.1 Road Safety Plans
Although aimed at fatalities, it is likely that the accident countermeasures employed will also help to reduce the levels of severely injured casualties.
Sample contents of some annual Plans published in the U.K. are included in Appendix A. Before producing such a document, the following will be required:
It is recommended that each highway authority produces an annual Road Safety Plan in which the local casualty reduction target is stated and a strategy for achieving the targets is developed.
Safety management should allow the Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 8
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Review of existing transportation policyhighways and
Investigation of accident trends for various road user groups in the authority’ss geographical area authority’
knowledge of the working structure of the authority (relationship between departments, committees external agencies concerned with safety).
The key to success lies in setting a series of achievable casualty reduction targets (short and long-term) that can be monitored, and being able to acquire the increased resources that will inevitably be required.
Summary proposals (including of major capitalplanned schemes, smaller remedial engineering work, safety audit, maintenance, costs, relationships with other agencies, safety publicity, traffic law enforcement)
Methods for monitoring and evaluation
Report of previous year's work and effect on accidents
The Plan should be a comprehensive document containing photpgraphs, graphs and figures and should be made available to the public.
The Plan should include:
Background to the road accident situation in the authority area (accident trends with respect to road user groups, road features etc)
Aims of the Plan (casualty reduction targets)
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 9
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
REFERENCES 1.
TRANSPORT AND ROAD RESEARCH LABORATORY LABORATORY.. Interim Interi m manual on accident investigation procedures and the development of low-cost engineering improvement schemes. TRRL, Pusat Litbang Jalan, Ministry of Public Works Works Indonesia, Jalan Raya Timur No. 264, Bandung - 1993.
2.
INSTITUTION INSTITUTI ON OF HIGHWA HIGHWAYS AND TRANSPORTA TRANSPORTATION. Highway safety guidelines: accident reduction and prevention. International edition. editi on. IHT, IHT, 3 Lygon Lygon Place, Ebury Street, Street, SWIW OJS, London - 1990.
3.
DEP DEPARTM ARTMEN ENT T OF TRAN TRANSP SPOR ORT T. Acci Accide dent nt Inv Inves estig tigati ation on Man Manua ual. l. Roy Royal al Soc Socie iety ty for the Prevention of Accidents, Cannon House, The Priory Queensway, B4 6BS, Birmingham - 1986.
4.
ROYAL ROYAL SOCIETY FOR THE PREVENTION OF ACCIDENTS. Road safety engineering manual. RoSPA, RoSPA, Cannon House, The Priory Queensway, Queensway, B4 6BS, Birmingham - 1992.
5.
INSTITUTION INSTITUTI ON OF HIGHWA HIGHWAYS AND TRANSPORTA TRANSPORTATION. Guidelines Guideli nes for: the safety audit of highways. IHT, IHT, 3 Lygon Place, Ebury Street, SWlW OJS, London - 1990.
6.
TRANSPORT & ROAD RESEARCH LABORATORY LABORATORY.. Towards safer roads in developing countries: a guide for planners and engineers. TRRL & Oversea’s Oversea’s Development Admin., Old Wokingham Wokingham Road, RG11 6AU, Crowthorne - 1991.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 10
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER II ACCIDENT DATA 2.0 INTRODUCTION An essential element of any accident reduction and prevention strategy is the collection and investigation of road accident data. Accident investigation procedures in the context of this Guide depend on the existence of a reliable database. This chapter outlines the database which is now being established in Malaysia with the microcomputer system that can be used for analysing this data.
2:1 PRODUCTION OF THE ACCIDENT DATABASE All accident data originates with the recording of details by the police either at the scene of an accident or as subsequently reported to them at the local police station by those involved. There will inevitably be a substantial number of road accidents that are not reported to the police at all. The accident details are first recorded in the police officer’s notebook and a simple record is made in the 24-hour incident book at the police station in which a unique reference number is assigned to the accident. If human injury has occurred and/or a prosecution is likely, an accident investigation file is opened, again having a reference number. This contains all documents associated with the accident, eg. witness statements, photographs, description and sketch diagram of the scene as found by the police reporting officer. This file is required in the law courts in the event of a prosecution. The traffic accident report form, POL27 (Pin. 1/91), is a pre-printed standard form and is the basis of all computerised data. It is also completed by the police accident reporting officer. officer. The form comprises 15 pages, though Cawangan Jalan, Ibu Pejabat JKR, K.L
several pages are repeated (for reasons given below): see copy of the form included in Appendix B. The first page contains some brief general instructions on filling in the form, though full instructions are contained in a separate booklet (Panduan Mengisi Borang POL271). Data entry begins on page 2 which includes information of a general nature, such as the police station, accident reference number, time and date of the accident, number of vehicles and casualties, road type and condition, road geometry, collision type, weather and lighting conditions, etc. This page also contains a small section for information on the closest kilometre post to the accident. This important page of the POL27 form is repeated twice (on pages 3 & 4) such that carbon paper can be inserted to produce copies for the Police District and also the local JKR office or local government department (see Section 2:3). The second main page (page no. 5) provides space for details of the vehicles involved in the accident including their type, model registration number, damage suffered, defects, and their movement prior to the collision. The details of the vehicle’s driver or rider are also included on this page. These include or herany age,driving sex, licence, (if any),his whether errors injury were made and whether they were wearing a seat belt or crash helmet. This page of the form is repeated for up to three vehicles involved in the accident with an additional carbon copy of each for the Police District. If more than three vehicles are involved then the reporting officer simply needs to attach additional pages, as necessary. The next new page (page 11) deals with any passengers or pedestrians involved in the accident, where again Page 11
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
simple details such as age, sex, severity of injury, are recorded. The remainder of this page contains estimates of the cost of damage to vehicles and/or property, and also codings for the location of the accident (see Section 2:2). Space for up to twelve injured passengers and six pedestrians is included on this page, where a carbon copy
recording the coding for both systems for each accident also serves as a cross-check on the accident location (it has been found elsewhere that errors are often made in the recording of grid coordinates).
is again provided for Police Headquarters.
i) ii) iii) iv) v)
The final page (page 13) includes a short space for the reporting officer to describe how the accident happened. Another box is provided for a sketch of the accident including the position of the vehicles prior to the collision, and the collision point in relation to the road layout. Separate space is provided for a location sketch where a simple map should be drawn showing clearly where the accident occurred on the road network.
Roads in Malaysia are one of five categories: Expressway (Toll) Federal State Urban Other (District Council or private)
2:2.1 Rural Areas
For rural roads (types i,ii,iii or v above), kilometre posts should be used as the network coding system. The Police reporting officer should estimate the position of the accident (to the nearest 100m) from the closest kilometre post. On Expressways this task is made easier as 100m posts have also
Two carbon copies of this last page are provided for both the Police District and the local JKR District Office or Local Government Department. The original/top copy of the form is sent to Police Headquarters in Kuala Lumpur. Lumpur.
been installed.
2:2 LOCATION CODING Location coding is a very important feature of the accident data as it provides the only way in which an engineer can obtain a true picture of where his safety problems exist. Two methods of location coding have been adopted in Malaysia and these are:
2:2.2 Urban Areas
1. 2.
Network coding Grid coordinate coding
The national grid coordinate system should be included on all maps used such that the position of the accident can be read off easily.
For urban areas (types iv or v above), a Node system needs to be devised where each major road junction is given a unique number (for that particular map). Sections of road between junctions are known as Links and can be uniquely by the node numbers on each side of defined the accident.
The function of the Network coding is to provide a means of examining easily and reliably a particular route or junction, or listing those with the worst accident records. The Grid coordinate coding provides an absolute location reference and means of plotting accident maps (essential for Geographic Information Systems).
Where an accident occurs off the main road network the location can be approximately defined by a unique Cell number. Cell numbers are simply areas containing smaller roads (eg. housing areas) which are normally bounded by the main road network, and should be assigned a different numerical sequence (eg. 900-999).
As well as facilitating both analysis feature,
Again, the national grid coordinates must
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 12
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
also be recorded by reading off values on the appropriate map.
2:3 RESPONSIBILITIES WITH RESPECT TO ACCIDENT DATA The production of the accident database for Malaysia is shown diagrammatically in Fig. 2.1. and the responsibilities of various authorities are outlined in the following sections. 2:3.1 Police
As mentioned above, when the local police station has completed a POL27 form according to their coding instructions1, the top copy is sent to the Police District Office, where it is checked, and then sent on to Police Headquarters at Bukit Aman, Kuala Lumpur. In practise, this tends to be done in batches, normally each week. At the same time pages 4 and 15 of the form are sent out to either the local JKR District Office or the
of each accident and complete the coding of:
Route number
Lowest Section number of km posts on either side of accident
Nearest 100m from the above post
Relevant map code and series
X-coordinate (easting)
Y-coordinate (northing)
Direction in which vehicle at fault was travelling
In order to provide this data the JKR Office will first need to acquire 1:25000 scale topographical maps of their area. Ideally with the use of a calibrated measuring wheel
Local Government Office, as appropriate.
attached to a car, all State and Federal roads in the area need to be driven along slowly and logged; ie. the position of kilometre The forms are further checked for posts and other landmarks in relation to completeness at Bukit Aman and then reference points already marked on the map processed (with other crime records) for (eg. road junctions) are noted as accurately entry onto the mainframe computer. At the as possible. On average, a feature or end of each month, the accident file is downloaded to a microcomputer diskette and permanent land-mark should be noted at least every 1/4km. the data-file converted into the Transport Research Laboratory's Microcomputer Accident Analysis Package2 (MAAP), which This information can then be transferred to the maps: an example of part of such a map has now been adopted as the standard is given in Fig. 2.2. If not already shown, accident analysis tool for Malaysia. one kilometre squares also need to be 2:3.2 JKR
The responsibility for precise location coding for each accident now lies with the various road authorities, as this is of primary importance to them. This is why copies of pages 2 and 13 (ie. pages 4 & 15 containing road number, kilometre post and collision sketches) of each accident are either delivered to or collected from the JKR District Office or Local Government Offices for completion of the relevant sections.
accurately drawn maps grid corresponding to the position of on the the national coordinates. With the additional landmark information and using the Police descriptions and location sketch on the POL27 form, it should now be possible for the JKR engineer to pinpoint accidents much more easily on the appropriate map, and thus fill in the aforementioned location data.
The appropriate officers at the JKR District Office are required to check location details Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 13
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 14
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
At the 1:25000 scale, just be possible to record the it X should and Y coordinates for each accident location to the nearest 10m (though the error may be ±25m). All completed forms should be sent to the Highway Planning Unit (HPU) of the Ministry of Works Works in Kuala Lumpur every month. 2:3.3 Local Government
For towns or cities it has been found to be more practical to use a node system. The Local Government Department are thus required to complete the following information on the POL27 forms sent to them:
Route number (where one exists)
Relevant map code and series
X-coordinate (easting)
Y-coordinate (northing)
Node number side of accident (if notonatother junction) Direction in which vehicle at fault was travelling
Again this requires the production of special maps, in this case producing a standard node system. For town/city maps a scale of 1:5000 (or at most 1:10000) is recommended. The Local Government Department should then assign a unique node numbering system preferably to all junctions in the' city. An example of part of such a node map is shown in Fig. 2.3. Again, if not already shown, squares corresponding to the national grid should be accurately drawn on the maps. (preferably 100m grid squares). This will enable X-Y coordinates to be noted easily to an accuracy of 10m. All completed forms should again be sent to HPU on a monthly basis
Node number of accident or nearest node (if not at junction) or cell
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 15
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 16
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
2:3.4 Ministry Of Public Works
On receipt of the completed parts of the forms (pages 4 and 15) the location information is further checked by HPU and entered with the accident identifying parameters onto computer. This will eventually be merged (once per month) with the corresponding accident records received from the Police (see Fig. 2.1). In theory, the database is now complete and can be sent out in the form of relevant MAAP data files for for use by the road authorities or analysis by other interested groups.
and safety improvements). For larger larger scale safety improvements it is recommended that central government provide a Special Road Safety Fund from which road authorities can apply for grants supplementary to their annual budget. The application will, of course, need to be justified for each scheme in terms of expected accident reductions which will contribute to achieving the local target.
2:4 RESOURCES REQUIRED In order to be successful in not only maintaining the accident database but achieving the local casualty reduction targets, resources will be required for both capital expenditure and staff time. Both the Institution of Highways and Transportation (IHT)3 and Local Authority Associations4 in the UK strongly recommend that a local road authority should establish a specific Accident Investigation Unit for this data maintenance, analysis and engineering side of accident reduction and prevention. The advantage of such a group is that it can dedicate its time to the task and not be diverted onto other traffic or highway matters. The IHT Guidelines suggest suggest a staffing staffing level of one engineer or technician for each 4001000 reported accidents per year, depending on whether the Unit can pass detailed design and implementation of schemes to another section. The staff must be trained as safety engineering is a specialised area of work. Adequate capital resources are also required in order to implement the extensive remedial work necessary to meet the local targets. It is therefore recommended that a set amount be specified in each annual budget of the road authority which is reserved solely for safety expenditure (maintaining the database Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 17
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
REFERENCES 1.
POLIS DI RAY RAYA MALAYSIA. MALAYSIA. POLIS 27 (Pindaan 1/91) - Panduan Mengisi Borang. Cawangan Trafik, Ibu Pejabat Bukit Aman, Kuala Lumpur - 1991.
2.
HILLS, B L, G J ELLIOTT & D CLARKE. Microcomputer Accident Analysis Package v5.0 (MAAPfive) User guide. Transport Research Laboratory, Overseas Centre, Crowthorne - 1994.
3.
INSTITUTION INSTITUTI ON OF HIGHWA HIGHWAYS AND TRANSPORTA TRANSPORTATION. Highway safety guidelines: accident reduction and prevention. International edition. editi on. IHT, IHT, 3 Lygon Lygon Place, Ebury Street, Street, SWIW OJS, London - 1990.
4.
LOCAL AUTHORITY ASSOCIATIONS. ASSOCIATIONS. Road safety code of good practice. C/o Hertfordshire County Council, Highways Dept, North Road, Hertford, SG14 2PY, 2PY, U.K. - 1989.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 18
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER III INVESTIGATION 3:0 INTRODUCTION The following chapters contain a stepby-step approach to tackling the problem of hazardous locations on a regular basis.
This Guide assumes that a microcomputer, the MAAP software (see 2:3.1), and relevant datasets are available to the investigator 3:1.1 Ranking blackspot sites
The main objective of this engineering safety work is to change the road environment in the most efficient manner (ie. within a specified budget) such that the maximum benefit in terms of accident savings is gained. This chapter is concerned with finding out where problem locations exist and the preliminary investigation required to try to determine the nature of the safety problems. Step 1: Identifying And Prioritising Sites
3:1. ACCIDENT ACCIDENT DA DAT TA SEARCH It is necessary to identify high accident sites in the network for which the road authority has responsibility. Ideally, Ideally, a period of 3 to 5 years of accident data should be reviewed. This is because accidents, even at very hazardous locations, are relatively rare events having a considerable random element, particularly in the time at which they occur. Statisticians tend to agree, therefore, that as a general rule, three years is really the minimum period needed to smooth out any abnormally large random fluctuations, to produce a reliable ranking of hazardous sites, and eventually to make evaluations of the treatments (ie. compare with a 3-year ‘after’ ‘after’ period). period). However, if such a long period is not yet available on the local computer database, rather than wait for this time to elapse, shorter periods can be investigated as long as caution is exercised over the conclusions made.
The first stage is to study the data in a logical manner to rank problem sites. It is important to note at this stage that the initial listing will need to be modified to produce one of ‘treatable’ sites. For example, consider Fig. 3.1 (a) and (b) summarising accident data types for two roundabout sites with similar accident numbers. In (a) there are a large number of similar accidents involving loss of control or skidding during the hours of darkness. This may well be treatable by improved skid resistant surface or drainage, improved signing and lighting. However, at site (b) there is no obvious dominant pattern; thus only site (a) can therefore be classed as a treatable site. It important to try to define a “reaction level” * , ie. the number of accidents or points above which the investigator takes some action. The reaction level is set based on the following three variables:
Number of accidents: a) all injury accidents b) severity points weighting c) all pedestrian injury accidents
Type of highway unit: a) kilometre length b) within 50m of junction c) links or mid-block accidents d) all roads in a defined area
Time period: a) 12-month periods of consecutive months, (not necessarily a calendar year) are the normal periods used.
*The Highway Planning Unit's current accident points weighting system is accidents involving fatality= 6 serious injury = 3 slight injury = 0.8 damage-only = 0.2 Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 19
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 20
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
An example of a reaction level criterion could be a blacksite definition of: 9 or more injury accidents, [or 15 points or more], within 50m of a junction, [or on a 200m road section], over the past 3 years. It is probably better to focus on injury accidents in setting a reaction level as these tend to be more reliably reported than damage-only ones. The following sections explain how sites can be ranked according to the four approaches to accident reduction mentioned in Chapter 1. 3:1.1.1 Route action sites
The simplest way of ranking sites, and the one currently recommended for use in Malaysia, is to list them in descending order of accident totals for either sections of road,
417 (includes 2 fatal accidents) 419 422 415 420 409 (includes 1 fatal accident). If the reaction level for this particular road authority was set at 9 injury accidents in a 100m section in 3 years (ie. as in Fig. 3.3), the first 7 locations are included in the above list (Table 3.1), though in a slightly different ranking order. order. Taking this a step further, if the severity of accidents is further taken into account by weighting factors (which are normally related to the average accident cost of each severity level), and damage-only accidents are also included (having a real cost), this results in the ranking shown in Table 3.1.
nodes, or grid referenced cells. Highway authorities elsewhere sometimes use accident rates but these necessitate traffic flow counts to be available at all points on the network, and also tend to give lower rankings to the high-flow, high-flow, high-accident where potentially more accidents could be saved.
It can be seen that in this particular case, which considers a small 27km length of road, the same seven sites appear at the top of the list irrespective of the ranking method used. However, the priority order varies somewhat, and is likely to change again when more indepth investigation is carried out to determine treatable sites.
To produce a list using MAAP, select the Location option from the main menu bar and select Kilometr Kilometree & Link/Node Link/Node analysis. analysis. If a particularrural road is being studied, select either the 1km or 100m analysis option, and
3:1.1. 3:1 .1.2 2 Single Single site sitess
then anyaccidents Conditions, such as, to include only set those involving personal injury accidents.
data entered onto computer. computer.
The priority listing for single sites in a town or city can be handled in much the same way as in the previous section once a node numbering system has been established and
For a particular town, town, MAAP can produce a list of the worst nodes or links (mid-block accidents between adjacent nodes) as in the example shown in Fig. 3.4.
An example of a kilometre plot of accidents for a 27km stretch of Federal Route 1 over a period of three years is shown in Fig. 3.2 To produce such a list, select the Location (using the Zoom option to specify lkm lengths). A list of the worst 100m sections of option from the main menu bar and select Kilometre & Link/Node analysis as above. this stretch is shown in Fig.3.3 (using the Now select Worst Worst node. Set any Conditions, Worst button which will prompt for the such as, to include only those accidents number of sites to include in the analysis). With reference to Fig. 3.2 it can be seen that involving personal injury accidents. Finally Select data files to be included in the the worst kilometres are: Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 21
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
analysis. Once a pass has been made through the data files, the user will be prompted for the number of nodes to include in the list. Again weighting factors can be applied to the different severity levels to obtain a cost-related ranking of sites.
It is often more difficult toaccident rank mass sites because a particular typeaction normally only represents a sub-set of the data at any one site. Thus smaller numbers are usually involved. However, However, the simplest approach is to try to assess the likely accident saving for each mass action plan and rank these, producing a list with the greatest potential for accident savings at the top of the list.
3:1.1.3 3:1.1 .3 Mass action sites
In order to determine sites for mass action treatment it is necessary to relate a selected type of accident feature to individual sites
3:1.1. 3:1 .1.4 4 Area Area wide wide action action
and to initially the latter according the numbers of rank accidents of the selected to factor.
scattered, andareas, so attention generally tends to turn to wider particularly urban residential areas. In urban areas in Malaysia, it is not uncommon for parts of towns to have well over 100 accidents (and sometimes over 50 injury accidents) per square kilometre per year. year.
Some examples of these would be as follows: Locations with the worst records of: Accidents on bends Right-turn accidents Overtaking accidents Nighttime accidents Pedestrians crossing road accidents Bicycle/motor-cycle accidents Cawangan Jalan, Ibu Pejabat JKR, K.L
As single blackspot sites are gradually treated accident occurrence can be rather
Again, ranking areas for treatment is not a simple matter and an assessment of the potential accident savings for each action plan should be made. Those yielding the best returns in terms of accident savings related to cost of implementation should be
Page 22
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 23
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 24
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 25
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
made. Those yielding the best returns in terms of accident savings related to cost of implementation should be placed at the top of the list. MAAP can also be used to produce counts of accidents in grid squares or irregularly shaped areas if a digitised or scanned map (to national grid coordinates) is available and accident locations have also been recorded by coordinates on the database (see Fig. 3.5 example). It is probably best to initially use MAAP MAAP to produce a grid square count and then consider, say, say, ten areas which could possibly each be treated as a package, that is, bounded by roads, railway lines, rivers or other geographic features. This type of ranking is normally carried out by focusing on the vulnerable road users. For example, total accidents involving some or all of the following groups could be plotted:
All motorcycle riders all pedal cyclists Child pedal cyclists (under 16 years) All pedestrian casualties Child pedestrians (under 16 years)
Cawangan Jalan, Ibu Pejabat JKR, K.L
STEP 2: Preliminary Accident Analysis
3:2. REFINING THE RANKING BY ST STA ATISTIC TISTICAL AL TECHNIQUES Before embarking on an in-depth investigation at any site, it is advisable to check that the site has higher numbers of accidents than might be expected, and that this difference is statistically significant. The following sections outline some simple statistical techniques which may be used. 3:2.1 Averages or "Norms"
It is important to know whether the level of accidents is higher than expected, for example, whether the number of skidding accidents at a site is worse than average. If a particular route is under consideration, this can be divided up into equal lengths (eg. kilometres) and the average number of accidents per section calculated. This is referred to as the arithmetic mean or norm. To determine whether particular sections warrant further investigation, the standard deviation (measure of the variability in the data) is normally calculated. The coefficient of variation Cv is a simple measure of how a set of data varies from its mean, with values of Cv > 1 regarded as very substantial deviation.
Page 26
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Those sites that have more accidents than the mean plus 1 standard deviation should be the first to be singled out for investigation.
It is used to calculate the probability of a particular frequency of accidents occurring in a year when the long term average is known.
Example
Example
Consider the example stretch of Federal Route 1 shown in Fig. 3.2 and take x as the
Let us assume the injury accident figures for a site are as follows:
frequency of injury accidents in three years.
1991 = 2 accidents 1992 = 0 accidents 1993 = 1 accident 1994 = 5 accidents If this site is selected on the basis of the last year, it is better to confirm that some change has happened at the site such that the next year will also be high, and not that the apparent increase has occurred by chance. Long term average = (2+0+1+5)/4 =2 Using the Poisson Probability (Single factor values) tables given in Appendix C, look for the high year value of 5 in the left hand column (k=5) and across to the column of λ(mean) =2. The value here is 0.0361 which means that the probability of 5 accidents occurring where the long term average is 2, is 0.0361 or 3.61 %.
Thus there is considerable variation between 1km sections along this road in their accident occurrence. Those sections with more than 12 accidents (ie. 6.22 + 6.62) are certainly worthy of further investigation, ie: Section:
417 419 422 419
However, the likelihood of 5 or more accidents occurring at the site should be quoted. To To do this simply add the probabilities of k=5, k=6, k=7, k=8 etc. That is: 0.0361 + 0.0120+ 0.0034+ 0.0009+ 0.0002 = 0.0526 Thus the probability of 5 or more accidents occurring due to random fluctuation is 5.26%
3:2.2 The Poisson Test
This test (for randomly occurring events) is commonly used to determine whether a recent increase in accidents at a site was due to random fluctuation only (and will return to previous levels). Cawangan Jalan, Ibu Pejabat JKR, K.L
ie. about a 1 in 20 chance that this is random, or a 94.74% (100-5.26) (100-5.26) chance that this is a real increase.
Page 27
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
FOR INTERNAL INTERNAL USE ONL ONLY Y
3:2.3 Chi Squared test
This test is normally used for two purposes:
To determine whether the number of accidents of a particular type is "significantly" higher than at similar sites
To check whether there has been a "significant" change in the number of accidents at a site after treatment has been carried out.
Example
A particular junction is suspected of having a poor skid resistant road surface, and has the following accident record: ‘Skidding’ accidents No skidding reported
=7 =5
For all other similar junctions along this road the accident record over the same period was: ‘Skidding’ accidents No skidding reported
= 37 = 178
We need to test whether the skidding accidents are significantly different from what might be expected. The following (2x2) table should be set up:
* N.B.
In the Chi Squared Distribution Table (Appendix D), looking along the line with one degree of freedom (v=1), the value just below the 9.81 calculated above is 6.64 which is the 0.01 “significance” level, ie. 1%. This means that the chance of getting 7 skidding accidents at a site with a total of 12 by chance is only 1 % (one in data is interrogated. 100 chance). Thus it seems fairly certain that there is some reason why the skidding accidents are occurring at this site 3:2.4 Interpretation of “significance”
The significance or confidence levels of results from the above statistical tests can be interpreted with the following practical meanings:
The above formula formula allo allows ws for Y Yates' ates' co correct rrection ion which overcome overcomess the inaccuracies inaccuracies which which could occur with the test when using whole numbers - as with accident frequencies. Also, note that the test becomes less reliable if any cell has a value less than S.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 28
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
it is generally agreed that only results significant at (or better than) the 5% level can be regarded as conclusive. 3:2.5 In-depth analysis - initial stage
Having now obtained a priority list of sites for investigation, it is advisable to produce a “working file” of accidents for each site. This can be done easily easily on MAAP and means that all subsequent analysis can be carried out of the working file, without having to search the whole database each time the data is interrogated. Use the Find Records option from the main menu bar and choose Find Selected Records. In this menu select Create Working file and then Set Conditions to extract all accidents for the site under investigation. The conditions to be set may be of the following types: i) Kilometre (Section) No. also set: Road Number Nearest 100m (if possible) ii) Node No. also set: Map code or State and District and PoliceStation No iii) X-coordinate range and Y-coordinate range The working files should include as many years data as available. Clear, meaningful names should be given to these working files so that they can be easily identified at a later date or by other users. A stick diagram diagram of each site site can now be produced using these files. This is produce a “working file” of accidents simply a way of displaying each accident record as a column of data. The purpose of the stick diagram is toassist the investigator to look quickly patterns of similar types of accident for which some appropriate engineering solution can be devised. The factors leading to accidents will be dealt with in more detail in Step 5, but at this Cawangan Jalan, Ibu Pejabat JKR, K.L
solution can be devised. The factors leading to accidents will be dealt with in more detail in Step 5, but at this stage the stick diagram gives investigator a “feel” for the types of accidents occurring and provides an indication of what to look for during the preliminary site visit. Select Stick from the main menu bar and set any conditions if necessary. It is possible to set up a number of different stick formats if required. An example of a stick diagram from MAAP is given in Fig. 3.6, where each accident is represented symbolically by a column (or stick) of key information. In this example, a T -junction (and one of the worst blackspots in Seremban), only injury accidents have been included; though it should be noted that there were also 74 damage-only accidents recorded in the 3 year period. It can be seen immediately from the stick diagram that the injury accidents all included the vulnerable road users: motorcyclists and/or pedestrians. Five of the six pedestrian casualties were in fact struck by a motorcycle. Most of the other motorcycle accidents (75 %) were side impacts or side swipes involving cars emerging from the side road and apparently not noticing or misjudging the motorcycle on the main road. Only two of the 13 injury accidents were in darkness, thus poor lighting is unlikely to be a particular problem at this site. The initial site visits should therefore concentrate on the turning manoeuvre problem particularly with motorcyclists, and also the pedestrian problem.
Page 29
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 30
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
STEP 3: Initial Initial Site Visit Visit
capital works programme for the area.
3:3 PRELIM PRELIMINAR INARY Y VISIT The site visit is a very important element of any accident investigation. The main purpose of the first site visit is to become familiar with the site and to ensure that available plans are up to date and detailed enough to identify specific features which may be contributing to accidents; for example, visibility sight lines, street furniture, buildings.
For operational purposes, the easy sites should be tackled first as they should provide:
The investigator should identify the manoeuvres indicated in the accident reports and try to visualise the accidents, particularly those with common characteristics. It might be necessary to make visits at different times of the day, or in dark and/or wet conditions, in accordance with the factors revealed in the stick diagram.
good return on money spent;
an immediate improvement in the accident record (- useful argument for allocation of funds for future yes);
an important psychological boost to staff to see successful results from implementation of schemes
It is often possible at this early stage to make a preliminary assessment of the likely causes of certain accident types. The use of photographs taken at driver/pedestrian eye height or an overall view can be an invaluable aid in the office or at presentations to committees. 3:3.1 “Easy” & “Hard” sites
It may now be possible to attempt to further rank sites even at this stage into whether they will be easy or hard to treat. This can be done more accurately later when costs and benefits are estimated. Easy sites are those where effective remedial measures can be readily identified and are of low-cost. Hard sites are those which do not provide a clear indication of appropriate treatment or where this is likely to be very costly. In the former case the site should be selected for further, more detailed investigation if it has high numbers of accidents. In the latter case it may be necessary to include the site in a Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 31
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER IV DIAGNOSIS 4.0: INTRODUCTION This chapter is concerned with collecting further data about the sites now selected for study, and using these to diagnose what are the common prime contributory factors that help explain how the road users involved in the actual collisions. This in-depth analysis of an accident site, area or group of road users is necessary in order to formulate an appropriate remedial measure. The following sections consider a single site analysis, the principles applying also to mass and route action approaches. STEP 4 Collection Of Further Data And Analysis 4:4.1 Collision Diagrams
Having produced a working computer file for each site, inspection of all this accident data and relevant police records is essential. It should be related to a plan of the area showing all on-site features relevant to the study period. It is suggested that attention should first be focused on injury accidents (the most reliably reported) unless these are very small in number.If accident pattems are not obvious from these accidents (discussed below),then it may be helpful to include the damageonly accidents where available. It is strongly advised that the POL27 sketch plms and accident description are retrieved by using the accident reference numbers for each accident.The values of data items which will uniquely define a particular accident record are:
State code District code Police Station no. Year
Cawangan Jalan, Ibu Pejabat JKR, K.L
Accident Reference no.
Copies of the sketch plans of accidents referenced by the above numbers from 1992 onwards can currently be obtained on application from: Accident Research Unit, Fakulti Kejuruteraan, Universiti Pertanian Malaysia, 43400 Serdang, Selangor Darul Ehsan Having obtained these and printed out the computerised record for each accident using MAAP, the next step is to produce collision diagrams at each site by drawing an approximate plan, preferably to about 1:500 scale showing the main site features (eg.kerb lines,street furniture, trees and buildings,and road markings. Details for this should have been noted/drawn during the initial site visit (Step3). On this plan, the POL27 sketch plans for each accident should be referred to in order to mark the positions of the accidents, and also the approach and intended departure paths of the vehicles immediately involved. An example of such a collision diagram for a crossroads where there were 10 injury accidents is shown in Fig.4.1. It is suggested that standard symbols be used for this as given in Appendix E The most important use of the collision diagram is toprovide a starting point for the classification of each accident into clusters. 4:4.2 Classiflcation Of Accident Types
As was stated in Chapter 1,accidents are generally multi-factor events and it is thus important not to try to assign a single cause to each accident during the initial examination of data.To do this Page 32
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
could mask underlying factors which can often be treated by simple low-cost remedial action.In practice an accident can be assigned to many underlying factors. For example, depending on circumstances, the basic single collision type shown in Fig.4.2 might be assigned to any of the following accident factors:
Approach visibility restricted Violation of mandatory sign Overshooting give way line Collision on restart from give way line Obscured give way sign Give way line worn away or conceaIed by uneven road surface Junction ahead not apparent from side road Excessive speed of main road traffIc Uneven lighting concealing main road vehicles
This list is not exhaustive, but illustrates the fact that a single collision type can be classified according to many factors or accident types. Some of these may suggest a suitable treatment whereas others may not Cawangan Jalan, Ibu Pejabat JKR, K.L
thus the aim should be to assign accidents to a class for which there is a remedial action. Unfortunately,, many of these underlying Unfortunately factors will not appear in the accident report or original police file. Let us consider again Fig.4.l where the accidents could be classified as: 2 double cross-overs (crossing both main road streams), a right turn, a left tum off (or nose-to-tail), and a pedestrian accident. At first sight no distinct accident pattern is revealed and thus no indication of my remedial action that would help. If left here no improvement to the junction would be made. However, after reading the written However, description of the POL27 form it was discovered that all 4 drivers pulling out of one minor road (fig 4.2) stopped first but collided with a main road vehicle on restart, because their “view to the right was obscured by street furniture or parked vehicle”. The rear-end collision in fig.4.1 occurred when the first vehicle braked for a third vehicle emerging from the side road but whose vision was masked by street
Page 33
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
furniture. Similarly the pedestrian stepped always necessary to verify and add data to out from behind the same street furniture. the computer grid. In Fig.4.5 additional Hence all 6 accidents could be assigned to information obtained by reading the text the class: “view to the right obstructed”, and description, looking at the sketch diagrams, thus the necessary remedial action is clear; and observations during the site visit, has ie. remove the obstructions. been incorporated by adding additional items and symbols (such as the “obscured vision” and “double cross over” taking into 4:4.3 Searching Searching For A Dominant Dominant Acciden Accidentt Pattern account main road direction). Consider the collision diagram in Fig.4.3 where again on first viewing there does not To help reveal common factors a useful appear to be any pattern which indicates a technique is to cut up the grid to produce treatment. There is a need to re-classify the individual sticks for each accident. These accidents to produce a dominant accident can be rearranges repeatedly on a new sheet type whichcould there be is at least one factorinwhich treated. Thecommon simplest way is to produce a stick diagram and, because every accident cluster is unique, using a standard stick format may be too restrictive. Such a stick diagram has been produced manually in Fig.4.4. Note that although other sticks can be produced and automatically sorted using MAAP,, the investigator MAAP i nvestigator is restricted to using only the computer computer coded items of POL27. Even if MAAP MAAP or other software software packages were used, manual checking is almost Cawangan Jalan, Ibu Pejabat JKR, K.L
of paper until a pattern is noticed. example of one rearrangement (byAn main road direction and collision type) is shown in Fig.4.5). Pattern recognition is a skill which improves with use. In the example it can be seen that the eastbound and westbound accidents reveal different characteristics. The westbound direction all possess “visibility obstructed by parked vehicles and trees” and the remedial action may involve new or enforced parking restrictions and tree lopping.
Page 34
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 35
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
On the eastbound approach, all accidents occurred on a wet road surface, the wet surface condition for the whole site being statistically no worse than the ‘norm’. The further data required in this case are skid resistances and any reasons why the eastbound approach may be wetter than westbound. In the subsequent site visit it was found that skid resistance was indeed considerably lower on this side due to reinstatement of the west side following extensive utility works. Also, lorries leaving a nearby plant were regularly depositing
junctions they will increase speed, and they may use a line of trees or telegraph poles to gauge the sharpness of a bend ahead and so judge how much to slow down.
water on the road on the eastern side. Furthermore the “Give way” line on the minor road southern approach was worn away and partly concealed due to a surface depression.
Unfortunately, the road engineer is unlikely Unfortunately, to have the opportunity to interview drivers involved in accidents at a problem site. But by relating the dominant factors from the police reports to his own site observations it is often possible to identify contributing defects in the road system.
4:4.4 Human Factors Need To Be Considered
Human factors are important to the road engineer as the roadside environment constantly presents visual cues to the driver as to the nature of the road ahead. Drivers tend to drive on expectancy in that when they see a wide, straight road ahead with no Cawangan Jalan, Ibu Pejabat JKR, K.L
Sometimes, however, the environment gives false visual cues. These are known as “perceptual traps” and are where some drivers are misled by the visual appearance of the road, commonly failing to recognise the presence of a give way junction ahead or a bend.
4:4.5 Example
A major-minor cross-roads had recently been improved on the major road arms by local widening and installation of clearly-marked right-turn bays. However,, concern was expressed over the However Page 36
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
the number of accidents which were still occurring. Only one year of accident data were available, and Fig.4.6 Fig.4.6 gives a MAAP stick diagram for this period which in this case includes all recorded accidents. It can be seen immediately that most (>90%) of the accidents were either right angle, side or side swipe impacts which implies collisions between one vehicle approaching the junction on the major road and another along a side arm.
right-angle collision. Ideally, the police accident descriptions should also be read (as should witness statements, if available) to determine whether any other factors could be added to the stick. It is clear that many drivers in these accidents were failing to give way to oncoming traffic at the stop line, and as no further information could be obtained from POL27, the essential site visit should now be carried out. STEP 5: Site Studies Studies And Analysis
4:5 RELEVANT RELEVANT DA DAT TA FROM SITE Most vehicles classed(DIR=7), as “at fault” werefour Before embarking on expensive new data travelling westbound though collection studies it is important to ensure of the sixteen accidents involved eastbound that all existing data about the site has been vehicles. There does not appear to be any obtained. Having studied this, together with pattern to the time t ime of day or day of week on the accident analysis above, it should then which the accidents occurred. Also there be possible to decide on studies which are were only four accidents during the hours of relevant to the actual safety problems at the darkness (the junction is lit) and only one site. accident on a wet road surface. Fig.4.7 shows the sorted stick diagram after viewing the POL27sketch diagrams to determine primarily which direction the main road vehicle was travel1ing prior to the Cawangan Jalan, Ibu Pejabat JKR, K.L
4:5.1 Simple Observation
It is possible that obvious difficult features of the site may have been observed at the initial site visit. However, with a more detailed knowledge of the types of accident Page 37
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
that have occurred, and also by driving through the site making the same manoeuvres, the investigator is now likely to be able to notice new features.
e) Would Would accidents be prevented by the prohibition of any single movement such as a right turn at a minor road?
Some of the most useful questions an investigator should ask are:
f) Could some of the traffic be diverted to other (safer) streets where problems are unlikely to be transferred?
a) Are accidents being caused by the physical condition of the road or adjacent property,, and cm the problem be property eliminated or corrected?
g) Are night time accidents out of proportion today time ones thus needing special night time protection, eg reflectorised signs, street lighting or traffic signals?
b) Is a ‘blind’ corner or restricted sight-line
h) Are there any particular times of day, year
at a junction responsible? If improvement is impossible, have steps been taken to warn drivers? c) Are the existing signs, signals and markings performing the job for which they were intended? Have conditions at the site changed since the devices were installed? Are replacements needed? Could the devices be causing accidents rather than preventing them? d) Is traffics properly channelled to minimise accident occurrence? Cawangan Jalan, Ibu Pejabat JKR, K.L
or weather condition when accidents are common? i) Do conditions indicate the need for additional levels of law enforcement? 4:5.2 Example
Continuing the previous cross-roads example, figs.4.8 and 4.9 show views of the approach to the junction fromeach minor road arm. The minor road is long and straight and relatively wide such that approach speeds of some drivers are probably quite high.
Page 38
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Despite the fact that there are stop signs, the visual cues to the driver provided by the kerb lines, line markings and light columns suggest that the road is continuous without a break, whereas it is, in fact, crossed by a major road. It is possible that even if a driver is a regular user of the road, the fact that he has been travelling a considerable distance in a fairly straight line could mean that he is not paying full attention and may thus fail to stop at the junction :there is then obviously the chance of a collision with a major road vehicle of the side impact type which can often result in injury.
Cawangan Jalan, Ibu Pejabat JKR, K.L
To improve this cross roads the engineer should concentrate on breaking up the long distance view, such that the illusion of a continuous road is removed. A relatively low-cost method of doing this would be to construct new offset traffic island in the centre of each minor arm carriageway with suitable chevron ghost islands and arrow signs. There is also likely to be a need to shave some area off the existing splitter islands to maintain adequate road width in the curved chicane created (see Fig.4.10). The carriageway's intersection by the major road should then be much more obvious to approaching drivers.
Page 39
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
As the junction carries relatively low volumes of traffic (including motorcycles), an alternative treatment could be the installation of aroundabout on which are mounted chevron boards opposite each approach arm. This would also have the additional benefit of slowing down traffic approaching on all arms of the junction (eg.see Fig.4.11).
Nevertheless, some of the following measurements or techniques would be worthwhile in certain circumstances to provide justification for decisions on remedial action. Table 4.1gives a list of the more common types of accident problems with studies that are likely to be appropriate. However, it must be noted studies will not be essential in every case. Further details of site studies can be obtained from ref. 1 4:5.3.1 Traffic Flow
A range of traffic data can be collected to assist with analysis, and this needs to be appropriate to the task in hand and comparable with the accident data, eg. same year or particular day of week. To help decide on the most appropriate provision for a particular manoeuvre (eg. right turners)at a junction, it is necessary to know the numbers of drivers normally making this manoeuvre. Comparing previous counts with more recent ones may reveal some changes in traffic pattern that could help to explain changes in the accident pattern. Although automatic axle counters can be used to measure straight road flows quite accurately,, manual counts will probably be accurately required to obtain turning manoeuvres at junctions. Although Although time-consuming to obtain, this method has the advantage of providing more accurate vehicle classifications (eg. for motorcycles, buses). 4:5.3 Other Other Observational Measurements
It is obviously desirable to have as much information about a site as possible when making decisions about how best to improve its safety. It is, however, recognised that additional observation studies may be difficult for some road authorities to carry out for reasons of cost and manpower. Cawangan Jalan, Ibu Pejabat JKR, K.L
If pedestrian accidents are a problem then the engineer will need to know how many pedestrians are crossing the road and where they do s. Guidance on carrying out vehicle flow counts is given in ref.2.
Page 40
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 41
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
4:5.3.2 Speed Measurements
4:5.3.3 Photography And Video.
Excessive speed is frequently reported as being a major contributory factor in accidents, and there can be no disputing the fact that safety margins are reduced and the likelihood of escaping injury in a collision is reduced with increasing speed.
The use of still photography or video taken from various positions (eye height, side or plan view) can be invaluab1e for presentations or measurements made in the office.
However, to provide evidence for a suspected speed problem at a particular site, speed measurements need to be taken. Road surface vehicle detectors linked to electronic timer counters provide accurate spot speed measurements if installed correctly, correctly, but hand-held radar guns have perhaps proved to be more popular a method due to their ease of use. However, care needs to be exercised in the use of radar guns to avoid secondary reflections from oncoming vehicles which can cause incorrect readings. For example, they are usually impractical for busy dual carriageways unless used from an overhead
Using colour video with a tenth second time display facility taken at various times of day, the movement of vehicles and pedestrians can be studied repeatedly in the office. Studying the road user behaviour in this way can sometimes provide valuable clues as to why accidents are occurring and whether there are any deficiencies in the site geometry which could be improved. If there are indications from the accident data of, say, a peak hour or wet weather effect then filming should be made in the same conditions. 4:5.3.4 Traffic Conflict Studies
bridge.
It is often difficult to establish the factors that lead to accidents from accident data accidents, and there can be no disputing alone due to incomplete or unreliable the fact that safety margins are reduced information. An additional measure which and the likelihood of escaping injury in can assist in the diagnosis of problems a collision is reduced with increasing involves the observation of conflicts or speed. “near-misses”. Conflicts are those events where there is a possibility of an accident, The siting of a radar meter is very but where a collision does not occur because important. It must be as inconspicuous as one or more of the parties involved takes possible so as not to affect drivers, choice of avoiding action. speed and away from bus stops, parked cars, junction mouths etc. The meter should be A conflict study is simply a formalised pointed as straight as possible along al ong the road method of observing the interaction of (in line with traffic movement):an error of 10 degrees either way will cause the meter to under-read by 1 1/2 %. A sample of at least 100 (preferably 200) should be taken comprising all freely -moving vehicles (or platoon leaders) to obtain a good estimate of the true mean and 85th percentile speeds. The standard deviation of the sample should be about one sixth of the mean. If it is much higher than themean (say, one quarter) or much lower (say, one tenth), then the measurements should be regarded with suspicion. Cawangan Jalan, Ibu Pejabat JKR, K.L
traffic at a location and recording the more hazardous events. There is obviously subjectivity involved in the identification of conflicts, and observers do need to be carefully trained to maintain conformity of results. Several slightly varying techniques have evolved in different parts of the world 3 and a relatively easy-to-learn and reliable technique has been developed over a number of years by TRL3,4. In a conflict study the numbers of conflicts are recorded and graded according to a scale of severity. This ranges from controlled Page 42
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Fig. 4.13 4.13 Collision diagram diagram (accidents (accidents overr 2-year ove 2-years) s) for juncti junction on in Seremban
slight braking to extreme emergency evasive action. The resulting data, usually expressed in the form of daily rates of particular types of conflict, should be used in conjunction with accident information to identify particular manoeuvres, road user groups, or site factors which contribute to a poor safety record. Conflict studies do have limitations to their use, and advice on the choice of sites, numbers of observers, length of study periods, etc .is given in ref 4. 4:5.4 Example Of5A Site Study Stu Fig.4.13 (from ref. ) shows thedycollision
diagram over a period of two years for an urban TT- junction between two one-way roads in Seremban. The stick diagram produced by MAAP is shown in fig.4.14 and this has been sorted according to collision type and severity of accident. It can be seen that the majority of collisions (33 out of a total 68) are side swipes (including all side impacts). Six of these
involved injury and all 6 injuries were suffered by motorcyclists. These collisions were chiefly between vehicles turning right out of the side road, Jalan Sheikh Ahmad, and merging with main through traffic. It is also likely that the 12 rear-end accidents were as a result of vehicles braking for these merging vehicles further upstream. The other main type of injury accident, comprising one fatal and four injury, that occurred at or near the junction were between pedestrians and motorcyclists. Pedestrians frequently do not notice the smaller visual area that a motorcyclist presents compared with a 44 wheeled vehicle. There does not appear to be any noticeable time of day or darkness effects in the accident pattern. Owing to various tune constraints, only a one day study could be carried out at this junction, but it was decided that this should include collecting data on traffic conflicts, approach speeds, vehicle manoeuvre now and pedestrian road crossing flows. The traffic flows are shown in Fig 4.15 and main conflict counts in Fig 4.16.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 43
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 44
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
It can be seen that the merge type of conflict with vehicles from Jln. Sheikh Ahmad was the most common due to relatively high merging flows. Drivers making this manoeuvre, particularly motorcyclist,were frequently observed relatively high merging flows. Drivers making this manoeuvre, particularly motorcyclist, were frequently observed as possible in order to use either the access road on the left (Jalan Khalsa see Fig.4.11) or adjoining petrol station. From Figs.4.15 & 4.16 it can also be seen that the frequency of pedestrians crossing the road is very high with a maximum of 890 in one hour. Conflicts with vehicles tend to occur mostly when business activity is high particularly in the morning.
The proposed accident countermeasures at this example site will be discussed in the next chapter.
Approach speeds of freely-moving vehicles were also measured using radar (see Fig.4.17) with a mean speed of 30km/h and 85th percentile of 37km/h, vehicle speeds were not considered to be excessive. Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 45
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 46
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
REFERENCES 1.
RADIN UMAR R S. Panduan Diagnosis dan Rawatan Kemalangan Jalan Raya. To be published by Dewan Bahasa dan Pustaka, Kuala Lumpur - 1995
2.
CHE MAT MAT BAHRI BAHRI BIN HJ KASRI. KASRI. Traffic Traffic Survey Survey and Studies. Studies. Institut Institut Kerja Raya Malaysia training reference: JLN/RA/201/1 IKRAM, Jalan Serdang, 43000 Kajang1989.
3.
ASMUSSEN, E. International Calibration Study of Traffic Conflicts Techniques. NATO NA TO ASI ASI Series F: Computer & System Sciences. Springer-Verlag, Springer-Verlag, Berlin, Heidelberg, New York, Tokyo - 1984
4.
TRANSPORT & ROAD RESEARCH LABORATORY LABORATORY.. The Traffic Conflict Conflic t Technique Guidelines. TRRL. Institution of Highways and Transportation, 3 Lygon Lygon Place, Ebury Street, SW1W SW1W 0JS, London - 1987
5.
BAGULEY, BAGULEY, CJ, & RADIN UR SOHADI. The improvement of accident data quality in Malaysia. In: Proceedings of First Malaysian Road Conference 1994. JKR, Roads Branch, Jalan Sultan Salahuddin, 50582 Kuala Lumpur - 1994
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 47
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER V SELECTION 5.0 INTRODUCTION This chapter discusses the steps of se1ecting a package of possible countermeasures for a site and of prioritising the potential treatments and sites. This is done by simply deciding on appropriate objectives of tile various safety strategies based on achieving satisfactory accident reductions which match or exceed the expenditure planned .
between RM250,000 and RM500,000. 5:6.1.2 Mass Action Objectives
STEP 5: Select Select Possible Possible Countermeasures 5:6.1 Objectives Of Countermeasure Countermeasure Scheme
For Malaysia, the precise objectives for the four accident reduction strategies outlined in Chapter I (Section l:4) will need to be decided based on local experience but those adopted in the UK are given below as a guide. The First Year Year Rate of Return (FYRR) is a measure of the net benefits in terms of accident reductions from the scheme expressed as a percentage of the total capital cost. This is defined fully later in this chapter in Step7. 5:6.1.1 Single Site Objectives
To achieve an accident reduction of at least 33%at treated sites. To obtain a significant FYRR To out not the exceeding remedial work at a costcarry per site a fixed maximum amount
An average FYRR of 50% for schemes should be achievable nationally. nationally. As time goes on schemes with a smaller FYRR may be worth considering provided that they meet the other two objectives. It is suggested that a maximum of RM25,000 be an appropriate level for the first application of accident remedial work. Thereafter Therea fter this this maximum maximum may be increas increased ed Cawangan Jalan, Ibu Pejabat JKR, K.L
To achieve an accident reduction of at least 15% at treated sites for each plan. To obtain a FYRR of not less than 40%. To carry out the remedial work at a cost per plan not exceeding a fixed maximum amount.
The maximum scheme cost is likely to depend on the type of measure used and the number of sites covered. 5:6.1.4 Area-Wide Objectives
To achieve an accident reduction of at least l0% within the area covered by the plan. To obtain a FYRR of 10% to 25%. To carry out the remedial work at minimum cost. The maximum sum will depend on the area size and inclusion of environmental enhancements.
5:6.2 Treatments
Having identified dominant accident types at a location or area under study, this will hopefully give an indication of an appropriate remedial measure (or package of measures if there is more than one accident group). It is desirable to consider a number of alternative proposals for each site. For every proposal it should be checked that: a) The measures are likely to decrease the type of accident at which it is aimed
Page 48
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
b) No further increase in other types of accident is likely to occur as a result of the selected measure. c) There are not likely to be any unacceptable effects on traffic movement or the environment. It should be stressed that safety at the site under study should not be the only consideration when choosing an appropriate countermeasure. The effect of that measure on the surrounding network should be estimated. For example, a self-enforcing speed reducing device like a series of road humps on a local collector road may have the effect of making a large proportion of drivers choose an alternative route along quieter residential streets. As well as being undesirable by residents of these streets, the safety will also most probably be worsened by the increased traffic now.
the sum of the individual percentages given in these tables. It is also recommended that reference5 be used as a source of ideas on many other treatments for typical blackspots, and that the reader keeps up-to-date with new techniques. 5:6.2.1 Example
Let us consider the same example T-junction discussed in detail at the end of the previous chapter (Section 4:5.4|& ref.6). It is clear that the vehicle sideswipe and pedestrian collisions problems should be where attention is focused in designing remedial action. In view of the former type of accident tending to involve motorcyclist, it was decided that the best policy would be to restrict the crossing manoeuvre and make all turning vehicles perform more of a merge type manoeuvre. This could be done by first narrowing down the two-lane flow along Jalan Sheikh Ahmad which currently joins The following tables (5.1 to 5.3) give a list of simple, chiefly low-cost, measures for the main road, Jalan Yam Tuan (see Fig.4.15) into a single lane. This would have general, urban and rural situations which the extra advantage of providing more area have been found to be effective. Where which can be utilised for motorcycle available, the average percentage reduction in accidents that has been achieved 1,2 is also parking. included. It should be noted, however, that the list is based largely on experience in the A solid delineator kerb kerb was suggested for 3,4 U.K. and Australia and should therefore this purpose as shown in the sketch in only be used as a guide or “ideas” list. It is Fig.5.1.This kerb is extended along Jalan likely that some of the measures will not be Yam Tuan so that the merge is actually applicable in Malaysia and the reduction in carried out further downstream where accidents will almost certainly be different. vehicles are travelling parallel to one It is thus very important that all remedial measures are properly monitored and evaluated, and results published or at least, centrally recorded so that a similar list based on actual Malaysian experience cm be built up.
If more than one group of accidents has been identified at a site, then the remedial work may consist of a package of measures with each one designed to reduce a particular accident group. It does not, of course, follow that the total effect of such a combination of measures at one site will be Cawangan Jalan, Ibu Pejabat JKR, K.L
another, and thus main road drivers have a clearer view of merging vehicles. This smoother merge should also help prevent queuing along Jalan Sheikh Ahmad despite its exit now being restricted to a single lane. The solid channelisation, which will reduce the width of the main road slightly, will need to be extended beyond Ja1an Kha1sa and the petrol station to prevent drivers making the immediate crossing manoeuvre to the left hand side (100 to150 vehicles per hour - see Fig.4.14).This relatively small proportion of the traffic will Page 49
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 50
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 51
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 52
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
however, now need to use a road on the left (slightly further downstream) to access these entries. Two proposals are shown in the concept sketches of Figs.5.l and 5.2 to deal with the pedestrian conflict problem. The first is to build a footbridge which would only be feasible between Jalan Sheikh Ahmad and Jalan Khalsa without affecting existing buildings. As there are considerable numbers of pedestrians at present crossing the road further upstream (Fig.4.14), it would probably also be necessary to install extensive lengths of pedestrian guard rail to channel pedestrians to the footbridge. This is obviously an expensive solution. The second option of installing a pedestrian refuge, as shown in Fig.5.2, is much cheaper. The refuge should make it easier Cawangan Jalan, Ibu Pejabat JKR, K.L
for pedestrians as they only have to make gap judgments about one stream of traffic at a time. Also, this refuge, together with the new chicane now already at this point, should also help to slow down traffic in this vicinity.. Large road studs along the vicinity boundary line of the hatched area and direction arrows on the refuge would also be required to help minimise the likelihood of co11isions with the new refuge. As well as being much cheaper this second option is preferred as it is like1y that many pedestrians would not choose to climb the footbridge. 5:6.3 Approaches To To Area-Wide Treatment
Where accidents are widely dispersed, often in urban areas, over several square kilometres rather than at obvious individual sites, an area-wide treatment may need to be considered Page 53
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
A road hierar hierarchy chy should should be be drawn drawn up by a road authority where residential streets, access roads, local distributors, district distributors and primary distributors are marked together with land-use (see Fig.5.3).
This can then be used as a base map on which to mark road accidents, vehicle and pedestrian flows. There is a wide variety of treatments that can be used in residential areas but good consu1tation with local residents is always strongly recommended. Traffic calming can generally be regarded as ways in which vehicle speeds can be reduced from an average of 50km/h down to 30km/h.
Cawangan Jalan, Ibu Pejabat JKR, K.L
The safety objective is to reduce both the number and severity of accidents, especially to vulnerable road users. This is usually done with self-enforcing speed reducing measures like chicanes, traffic throttles, road humps or speed tables (see Figs.5.4 to 5.9)
The main principle of traffic calming techniques is that they still permit motorised traffic to use the same route which they are unable to do with more restrictive measures like road closures, turning bans and one-way traffic. Although limited use can be made of these latter measures (eg.Fig.5.5), they are rarely popular with residents and can lead to accidents being transferred to other areas which become used as “rat-runs”.
Page 54
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
It should be remembered that solutions to a detailed study at a site may not exclusively involve civil engineering works. An integrated approach to improving road safety should really be adopted whereby road safety education and training may need to be combined with an engineering measure. For example, the introduction of a new signa1 controlled pedestrian crossing in a village may require a local publicity campaign to inform pedestrians and drivers how to use it together with training for children in the local school(s). STEP 7: Priortise Treatments Treatments & Sites 5:7.1 Estimating Accident Savings
The standard approach for the ranking of treatments is to carry out a cost-benefit analysis based on estimated benefits of the scheme and simply place these in priority order on the basis of the best returns. However, if there is currently little or no data on which to make an estimate of the likely effectiveness of a treatment, then perhaps the best way to proceed is to implement the lowest cost schemes first as these are likely to provide the greatest overall benefit. If the least cost scheme proves in practice to be ineffective then the a1ternative schemes in order of increasing cost should be tried. In most uses a pessimistic estimate can be assumed to t o be an average reduction in accidents of around 25-33%of a1l accidents. In these applications it is recommended that temporary materials be employed where possible for initial trials; for example, pre-cast concrete slabs tied together and pinned to the road surface to try a particular size and position of splitter island (see Figs.5.10 & 5.11). An economic assessment of projected schemes is important to ensure that the benefits likely will be greater than the cost of implementing and maintaining the scheme and that the best value for money is obtained. Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 55
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
There are two methods of economic assessment used for this purpose:
5:7.2 First Year Year Rate of Return (FYRR)
i) First Year Rate of Return(FYRR), and ii) Net Present Value(NPV)
This is simply the net monetary value of the accident (and any other) savings and drawbacks expected in the first year of the scheme, expressed as a percentage of tile total capital cost.
Both methods need the following basic information:
FYRR (%) = Benefits (1st year) x 100 Capital costs
a) The capital cost of the scheme. b) An estimate of al1 benefits (monetary value)expected to result. c) An estimate of all disbenefits.
where benefits = accidents savings + change in maintenance costs(+/-) + change in journey costs(+/-) Unfortunately, at present the only available accident costing used in Malaysia is that produced in 1985 (see Table 5.4 - from ESCAP7). However, applying national inflation figures However, since that time, figures for 1995 are estimated in Table 5.4; and using the recent numbers of each severity class of accidents in Malaysia, the average cost of an injury accident in 1995 is approximately RM33,000. It is likely that this figure is nevertheless an underestimate, and it is hoped that a study will be conducted soon to determine more up to-date and realistic values for Malaysia. Malaysia. A column for the most most recent national accident costings has thus been left in Table 5.4. Thus let us consider, as an example, a junction which had l2 injury accidents in 3 years, and nine of these involve side collisions with drivers overshooting the Stop line - these being the treatable group of accidents. If the target FYRR is 50%, then the maximum budget for the scheme may be calculated as: %FYRR = Annual Acc. Acc. Saving x 100 Scheme cost 50 = (9 x RM33,000 /3) x 100 Scheme cost
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 56
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Scheme cost = RM99,000 x 100 50 = RM198,000 That is, the scheme should not cost more than RM198,000 in order to achieve a 50% rate of return. A more detailed assessment may, may, however, however, be needed with schemes where traffic accidents and traffic levels are expected to change considerably from year to year. For example, a scheme with an 80% FYRR may not be worthwhile if subsequent road closures due to construction of a planned new road, say, restricts the benefits just one year. 5:7.3 Net Present Value (NPV)
This type of evaluation expresses (in a single lump slim) the difference between costs and benefits of a scheme which may occur over a period of several years. Unfortunately, it would be incorrect to simply assume that year 1 benefit can be summed to obtain the overall benefit over the life of the scheme. This is because society,, in general, prefers benefits which society occur sooner rather than later. Future benefits must therefore be adjusted , or “discounted” before being summed to obtain a “present value”. The current rate used by the Treasury for highway schemes is l1% which means that for each RM l of benefit occurring this year,
if this also accrues next year then this is valued at 11% 1ess, 1ess, ie.89 Sen. A further year's delay will reduce the benefit again by 11% of 89 Sen, ie.79 Sen, and soon. These figures can be summed over the life of the scheme to obtain the Present Value Value of Benefits (PVB). The overall economic worth of the scheme is then obtained by deducting the Present Value of Costs (PVC) {these may also have to be discounted if they are spread over more than one year}: NPV = PVB-PVC The scheme is only usually considered worthwhile if this figure is positive. 5:7.3.1 Example Example of NPV assessment.
Let us assume that the expected costs of a junction redesign will be initially RMl00,000 spread over 2 years with annual maintenance costs over the next 5 years (the life of the scheme) of RM8,000. The benefits are always difficult to estimate and will often require a simple educated guess. If in this case we assume that 4 injury accidents over the first two years(2 per year) will be saved, and this will reduce to 0.5 per year following that due to changes in traffic. This equates at present to RM66,000 for two years followed by RMl6,500 for the remaining 3 years. The Net Present Value Value is calculated in Table 5.5 to be RM6,865.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 57
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
In cases where the estimated benefits do not vary throughout the scheme, the calculation of NPV is simplified simplified by the the use of cumulated discount values and these are given for various discount percentages in Appendix F. For example, for a benefit of RM20,000 per annum over 5 years with reference to Appendix F, F, the net benefit at 11% discount rate would be:
all options are ranked in order of their NPV/PVC ratio [the highest ratio at the top of the list].
If funds are limited, those with the highest NPV/PVC ratios are preferable on economic grounds. Table 5.6 shows an example of a remedial works priority programme ranked in terms of the schemes, NPV/PVC ratio for a 5 year period.
RM20,000 x 3.57 = RM71,400 5:7.4 Priorities For Implementation
The economic criteria for scheme assessment using using the NPV approach are:
all schemes schemes where NPV NPV is positive positive are worthwhile in economic terms;
for a particular site, the most worthwhile option is that with the highest NPV;
Cawangan Jalan, Ibu Pejabat JKR, K.L
It can be seen that in this example the NPV/PVC ratio gives only a slightly different ranking of the sites to that using FYRR. Using this listing, a line can be drawn for a particu1ar budget: in this use RM350,000. The full 1ist of l0 sites could only be implemented if a budget of RM500,000 were allocated. If the authority is receiving local political or other pressures to treat a site which is Page 58
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
outside this list or below the cut-off level, then the table can be used to point out that resources should be concentrated at the sites where greater benefits are likely to occur. This is more 1ikely to yield the best contribution to the nation’s casualty reduction targets. In some uses a site may be at a location which is included within a major capital works programme such as a flyover or traffic signals. If the time at which these are scheduled for introduction is fairly close, it may be best to “do nothing” at this stage and incorporate necessary work within the major scheme. If, however, the scheme is unlikely to be carried out for 2 or 3 years, then short-term (perhaps lower-cost) measures will probab1y be justified .
current budget period to allow for these minor re-allocations of funds. In practice “easy” sites are normally best tackled first to yield cost effective results as quickly as possible. However, it is likely that the “harder” sites, which may require more staff resources to study extensively, will have high numbers of accidents. These sites should not thus be put on one side and forgotten about.
For this reason and others which might lead to “s1ippage” in timetables, it is always worth investigating more sites and preparing more schemes than can be carried out in the Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 59
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
REFERENCES 1.
ROYAL ROYAL SOCIETY FOR THE PREVENTION OF ACCIDENTS. Road safety engineering manual. RoSPA, RoSPA, Cannon House, The Priory Queensway, Queensway, B4 6BS, Birmingham - 1992.
2.
ACCIDENT REDUCTION 2000 GROUP. GROUP. Progress Report December 1993. Transportation 2PY, Hertford - 1994.Dept, Hertfordshire County Council, Goldings, SG14 2PY,
3.
NATIONAL NATIONAL ASSOCIATION ASSOCIATION OF AUSTRALIAN ST STA ATE ROAD AUTHORITIES. AUTHORITIES . Guide to traffic engineering practise: part 4-Road crashes. NAASRA, 2, Dind Street, PO Box 489, Milsons Point, NSW 2061-1988
4.
ANDREASSEN, ANDREASSEN, D C. Strategies for safety problems. Australian Road Research Board. Research Report ARR163. ARRB, 500 Burwood Highway, Vermont South, Victoria-1989
5.
TRANSPORT & ROAD RESEARCH LABORATORY LABORATORY.. Towards safer roads in developing countries: a guide for planners and engineers. TRRL & Oversea’s Oversea’s Development Admin., Old Wokingham Road, RG11 6AU, Crowthorne-1991
6.
RADIN UMAR RADIN SOHADI. Analisis Terperinci Kemalangan Jalan Raya: Projek Pilot Seremban, Shah Alam dan Petaling Jaya. JK3P, JK3P, Laporan Penyelidikan No. 4. Majlis Keselamatan Jalan Raya Malaysia, Wisma Semantan, Jalan Gelanggang, 50490 Kuala Lumpur-1993
7.
ECONOMIC AND SOCIAL COMMISSION COMMISSION FOR ASIA AND THE PACIFIC. Report on improvement of the traffic accident recording and analysis system in Malaysia. UN ST/ESCAP/478, Bangkok-1985
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 60
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER VI IMPLEMENTATION 6:0 INTRODUCTION Having selected an appropriate measure or package of measures to deal with the
the technical term for the systematic checking of safety aspects of new schemes carried out on the public road.
accident problems at a site area, the next stage is detailed design andorconstruction. It is not within the scope of this Guide to cover in depth the actual detailed design or physical implementation. This will generally be the responsibility of another appropriate department. However, it is essential that the team who made the scheme recommendation continue to be involved.
Although it is assumed that national standards will be followed in any design unless unusual local conditions dictate a departure (which needs special approval), a combination of elements perhaps close to their respective recommended minimum standard, may combine to create safety problems. Safety audit seeks to address such problems .
The road safety audit is discussed and general advice on the safety of construction work noted. STEP 8: Detailed Detailed Design Design And Installation
At present the UK Institute of Highway Engineers’ Engin eers’ Guidelines Guidelines1 are being applied in Malaysia but these should be amended to suit local conditions as experience is gained.
6:8.1 Detailed Design
6:8.1.2 Aims Of The The Safety Audit
As stated above, the next stage after selecting an appropriate remedial measure will usual1y be detailed design. This is likely to be carried out by a different unit to those investigating the problems, and is beyond the scope of this Guide. However, the design drawings will, of course, need to be based on the proposals/outline plans of the accident investigators and this same team should also remain actively involved with the designers. 6:8.1.1 Road Safety Audit
Road safety audit is a means of accident prevention rather than accident reduction (a change in philosophy to the previous chapters). It is the application of safety experience to ensure that future safety problems are not designed into new schemes. It is discussed here because ideally an audit should also be carried out on safety remedial work, both at the design stage and again immediately after the scheme implementation: that is, prior to opening to normal traffic. Safety audit is simply the Cawangan Jalan, Ibu Pejabat JKR, K.L
To ensure that all road schemes operate as safely as possible.
To ensure that preventable potential accident-generating elements are not present in a completed scheme, for example ,moving lamp columns to the back of the footway.
To ensure suitable accident-reducing elements are included in the scheme, for example, "anti-skid" surfacing on down hill approach to traffic signals, guard rail and chevron boards on unavoidably sharp bend, crash cushion before essential solid structure.
6:8.1.3 Organisation Of Of The Audit
Safety audit should be part of the overall safety management strategy for the road authority within its Road Safety Plan. An arbitration procedure should be agreed in use of differences in opinion. Page 61
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
In a road authority it is likely that safety audit will be carried out within the accident investigation/road safety unit, though preferably by more than one person. Before deciding on who should carry out the audit the following should be considered: considered:
The audit team should be independent of the design team. It should contain (and must certainly be led by) persons with safety engineering experience. A knowledge of design standards is important. Other specialists such as traffic signals and structural engineers may also need to be consulted depending on the scheme. The police may also be required, particularly in the t he latter stages of the audit, where special road users, requirements may need consideration.
is overlooked. However, the audit team should not rely solely on these lists. The auditors should imagine “walking” or “riding” the scheme, and should physically do this at the final audit stage to check, for example, that signs are of the correct type and in the right place, road markings and island are correctly placed and that there are no unforseen conflicts between the treatment and other existing site features. Although the audit team should discuss their findings with the design team, a formal report should always be produced. This should state the potential safety problems as precisely as possible and should include a recommendation or options for improvement. The recommendation should be in outline form only and it may be desirable to annotate copies of the original scheme drawings.
6:8.1.4 When To To Carry Out The Audit
The scheme should be monitored and feedback given to the design team.
Safety audits can be performed at the following stages:
6:8.2 Installation
Feasibility study
Completion of preliminary design
Completion of detailed design
Prior to opening to traffic
As stated above, it is beyond the scope of this Guide to include guidance on all engineering aspects of altering existing road geometry or installing countermeasure devices. This section is thus limited to general advice on maintaining safety during installation. 6:8.2.1 Safety At Roadworks
Other times on an informal basis
6:8.1.5 The Audit Task
Information such as plans, list of standards followed, departures made traffic and pedestrian counts, and accident records should be collected from the design team. It may be helpful to discuss the purpose behind the design of the scheme and it is essentia1to carry out a site visit at a1l stages of the audit. Appropriate check lists1 should be used to systematically ensure no safety problem
Accidents tend to occur at a higher rate at roadworks sites and involve more vehicles than on normally operating sections of the road network. A study of major roadwork sites in the UK 2 found that, despite the fact that the contraflow sites were genera1ly well signed and laid out, accidents still occurred 1.6 times more frequently than on non-roadworks sections, and the percentage of accidents involving 4 or more vehicles was 29% compared with only 8% without works. It issuspected that the ratio may be considerably higher than this in Malaysia, particularly at sites where advance warning signs are poor.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 62
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
their roadworks sites. Often with relatively short-term work due to the extra trouble required, warning signs are not set out sufficiently in advance of the works site or are two few in number: this is particularly hazardous where drivers vision of the site may be obscured by a bend or other traffic. The use of modern electrically-powered flashing arrow lights mounted on trailers tend to provide a more effective means of attracting drivers, attention. Attention should be paid to the use of adequate lengths of “safety zone” or “buffer space” (see Fig.6.4) which provides an escape area if drivers fail to notice the advance warning signs. Lengths of these zones are specified in ref.3. Also, tapers of traffic cones to close off a lane before the work area, or to move traffic into other contraflow lanes, are often not made long enough. The standards should again be followed to ensure that the traffic movement is gradual, and thus smoother and safer. 6:8.3 Implementation log
It is important to keep a record of the precise dates of the beginning and completion of major parts of the remedial work for all jobs. This is essential for the monitoring of the scheme (see following chapter). Similarly, details of all costs involved, including variation orders, must be kept. The actual costs often differ considerably from original estimates, and this record will facilitate a more reliable cost-benefitanalysis. It is very important, therefore, that countermeasure installations themselves are made as safe as possible. The road engineer must attempt to regularly enforce contractors to follow the standards laid down in Arahan Teknik 2C/85 3 for traffic control, temporary signs and work zones at Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 63
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 64
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
REFERENCES 1.
INST INSTIT ITUT UTIO ION N OF H HIG IGHW HWA AYS AND AND TRAN TRANSP SPOR ORT TATI TION ON.. Guid Guideli eline ness for for the safety audit of highways. IHT, IHT, 3 Lygon Place, Ebury Street, SWIW OJS, London 1990.
2.
MARL MARLOW OW M, and and R D COOM COOMBE BE.. A stud studyy of th thee ssaf afet ety y of of maj major or moto motorw rway ay roadworks in 1987. Research report RR223. Transport and Road Research Laboratory, Old Wokingham Road, RG11 6AU, Crowthorne - 1989
3.
JA JABA BAT TAN KER KERJA JA RAY RAYA. Man Manua uall on tra traff ffic ic con contr trol ol dev device ices, s, tem tempo pora rary ry sig signs ns and and work zones control . Arahan Teknik (Jalan) 2C/85. Cawangan Jalan, Ibu Pejabat JKR, Jalan Sultan Salahuddin, 50582 Kuala Lumpur - 1985.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 65
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER VII EVALUATION 7:0 INTRODUCTION Having introduced a countermeasure or package of measures it is important to
It is essential to carry out the monitoring effectively, not least to avoid the “bad publicity” which could occur if a road safety
establish thework effectiveness of the safety engineering carried out; first to check that nothing has gone wrong and that it is working as intended, and later to learn lessons which may influence future decisions on improvements.
scheme was seen to be actually causing accidents.
This chapter emphasises the need to monitor by observing the changes at the site in operation. To evaluate these it concentrates on accident changes and describes the simple statistical tests needed to obtain. STEP 9 Monito Monitoring ring 7:9.1 Initial Observations
The treated site should be observed immediately after completion of the construction and regular visits made in the following days, weeks or months until the team is satisfied that the scheme is operating in the way expected. It is strongly advised that my earlier behavioural measurements that were made during the investigation stage of Step5 (eg, traffic conflict counts, speed measurements, skid resistance) are now repeated as this will lend weight to any argument making further changes at the site or, for indeed, proving success. It can happen, for instance, that some feature of a scheme may produce an unforseen reaction in drivers which creates a potentially hazardous situation. Monitoring should highlight this problem at an early stage so that appropriate action can be taken quickly to remove this danger. At best it may be possible to alleviate this danger easily, for example, by a realignment of kerb lines to prevent a hazardous manoeuvre. At worst, it could lead to the complete withdrawal of a scheme and need to reassess alternative schemes. Cawangan Jalan, Ibu Pejabat JKR, K.L
Recording the results of the monitoring measures is also important to build up a database of types of treatment and the effects they produced to provide information for future safety engineering work. 7:9.2 Measures Used In Monitoring
For monitoring or measuring the effect of a safety improvement, the technique employed is usually by “before” and “after” analysis. The most important measure of success is, of course, whether the safety work has improved the accident situation at the site. This will always need to be assessed for a scheme and statistical methods for evaluation will be discussed in the next Step. A simple visual method that that has been used, though is perhaps more suitable for mass action plans rather than single sites, is that cumulative accident numbers (& types) are plotted together with their cumulative mean. Fig.7.1(from ref.1)is an example of this method of data presentation, and illustrates that the daytime running headlight campaign in Malaysia was apparently being effective in reducing those accidents related to daytime conspicuity (MSTOX = motorcycles moving straight or turning when other road users cross their path), whilst having no effect on night-time accidents. In the table of Fig.7.l the cumulative mean number of related accidents has been calculated. The cumulative mean is obtained by simply adding on the average monthly accident frequency over the before period (in this case 6 months) to each month after the first one. As long as the standard deviation is not large, the two comparable lines of Page 66
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
cumulative accidents and cumulative mean accidents lie close together during the before period, but after implementation of the scheme the cumulative mean line represents what would normally be expected if no action had been taken, and the amount they drift apart (marked * in Fig.7.1) represents the effect of the measure. However, to be sure that the random nature of accidents has been taken into account, it will normally be necessary to wait for several years for a valid result to be available. More immediate feedback is often necessary which is why the above method is suggested. Other behavioural data, as mentioned in 7:9.1, can also be collected to give indications that a scheme is working. It would, of course, be impractical to carry out detailed behavioural studies for all minor alterations, but studies may be particularly important for expensive schemes like areaCawangan Jalan, Ibu Pejabat JKR, K.L
wide or mass action treatments. It must be noted, however, that non-accident variables have the disadvantage that they do not give direct measures of the size of safety improvement. There are practically no variables for which the precise relation to accidents is known. This means that a measured reduction in mean speed, for example, cannot be translated into an estimate for the number of accidents saved: this is a considerable drawback. However, before carrying out a behavioural However, “after” study it is generally better to wait for a period of about 2 months after the scheme has been operating. This serves as a “settling in” period during which regular users get used to a new road feature and any learning effects have disappeared. Some of the factors that may need to be examined (see also Step5,Chap IV) are noted below: Page 67
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
7:9.2.1 Speed Of Traffic
7:9.2.5 Public Perception
If speed reduction is one of the objectives of the scheme then speeds should obviously be monitored. Similar and appropriate locations should be carefully chosen for the before and after studies preferably using automatic equipment. If radar guns are used then these need to be unobtrusive otherwise warning signals invariably given by drivers in the opposing direction will yield unreliable results.
Often one of the main reasons why an area-wide scheme has been implemented is due to campaigns by residents for something to be done. One of the most important parts of an area-wide scheme, therefore, is public consultation .Thus, an important monitoring measure is how the residents and other road users feel about the safety elements of the scheme after implementation. 7:9.2.6 Effects On Other Other Areas
The t-distribution can be used to compare whether my changes in the mean speeds in the two periods of measurement are statistically significant (see Appendix G).
It is important to examine whether the scheme has led to an increase in accidents, traffic speeds and volumes in adjacent areas. 7:9.3 Control Data
7:9.2.2 Traffic Conflicts
As mentioned earlier these are generally of use at junctions only. The “after” study should be carried out in the same conditions and for the same periods as the “before” study,, and preferably using the same study observers (to minimise subjectivity between individuals). The frequencies of occurrence of conflicts cm be analysed in the same way as the methods used for accidents, as outlined in the next Step. 7:9.2.3 Traffic Volumes
If the measure is expected to affect manoeuvres at a junction or drivers choice of route in any other way, then it is desirable to collect traffic flow data throughout the local network. It may also be necessary to expand this survey to provide origin and destination information so that estimates in through-traffic can be obtained to determine how this has been affected by the scheme. 7:9.2.4 Travel Times
In some cases monitoring may require an estimate of changes in travel time for residents and through-traffic. This will be important where traffic severance forms part of the scheme, and traffic is being re-routed.
In most of the above monitoring measures (and particularly accident changes) it is necessary to take into account other factors not affected by the treatment which might also influence that measure. Examples are: a change in speed limit on roads which include the site; national road safety campaigns; traffic management schemes which might affect volume of traffic. These changes may be compensated for by comparing the same “before” and “after” periods with accidents (or other measurements) at “control” sites which are untreated. Control data can be either by matched pairs or area controls. A matched pair control control site should be similar to the treated site in general characteristics and also geographically fairly close to it (but not close enough to be affected by my traffic diversion). This is so that the control will be subject to the same local variations which might affect safety (eg. weather, traffic flows, enforcement campaigns). Although the matched pair is the best statistical method to use, in practice it is very difficult to find other sites with the same problems which are left untreated purely to carry out statistical tests. Area controls which comprise anumber of sites are, therefore, much more frequently used.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 68
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
when choosing control sites:
they should be as similar as possible to treats sites;
that due to other sources. Some of the other factors that need to be considered are discussed below:7:10.1.1 Changes In The Environment
they should not be affected by the treatment;
there should be more than 10 times the number of accidents at the contorl sites.
For example, if the traffic signals at a site are modified then a control group of sites might be a11other signalised sizes in the town. But if there were only two other signalized junctions and these had lower flows and much fewer accidents as did other uncontrolled junctions, then it would be better to use all signalized junctions in the State.
This feature was mentioned in the last section of Step 9 whereby a change in the environment or driving habits can affect the accidents occurring at the study site. For example, a change in the nationa1 speed limit for the class of road at the site, or closure of a nearby junction to the site producing a marked change in traffic t raffic patterns. This feature can be taken into account by the use of control site data but for this to be valid it is important that these other sites experience exactly the same changes as the site under evaluation. 7:10.1.2 Random Fluctuation
As explained in Chapter I, the rare and random nature of road accidents can lead to quite large fluctuations in frequencies 7:10.1 The Effect On Accidents occurring at a site from year to year, even This step of the procedure focuses on evaluation of whether the treatment has been though there has been no change in the underlying accident rate. This extra successful in achieving its objective of variability makes the effect of the treatment reducing the number of accidents. This therefore requires comparison of the number more difficult to detect; but a test of statistical significance can be used to of accidents in the target group before the determine whether the observed change in treatment with the number after treatment accident frequency is likely to have occurred (with the assumption of a similar before by chance or not. pattern if nothing were done), and to study whether my other accident type has 7:10.1.3 Regression To The Mean increased. This effect complicates evaluations at high accident or blackspot sites in that This Guide does not attempt to delve deeply accidents at these sites tend to reduce even into the different statistical techniques, but when no treatment is applied. Even if a to suggest practical and simple ways in 3.year total is considered at the worst which schemes can be evaluated. The accident sites in m area, it is likely that the following sections generally refer to “a site” accident frequencies were at the high end of but the same techniques can be used for the naturally occurring random fluctuations, mass, route and area-wide action as long as and subsequent years will yield lower appropriate control groups are chosen. numbers. This is known as regression to the mean. The main problem when using accident data for evaluation (even assuming high recording accuracy) is to distinguish between a change due to t o the treatment and STEP 10 Evalua Evaluation tion
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 69
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
As an example consider Table Table 7.1 which gives the actual numbers of recorded accidents involving personal injury for 122
The effect does, however, tend to be diminished if longer periods of time are selected. For example, Abbess et a1.3, in a
nodes in the town of Seremban over a two year period. For sites with 5 or more accidents in year l there were overall fewer accidents in the following year. Conversely, Conversely, sites with 4 or less acc1dents have more accidents in year 2. If an accident countermeasure had been installed at the worst 9 sites at the end of year l then a highly significant reduction of 37% might be claimed after year 2, even though the measure had been completely ineffective (this same result would be obtained by doing nothing). An even higher false resu1t would be obtained if the other 113 sites were used
study in two counties of the UK calculated that regression-to-mean had the following effects at high accident sites(ie. more than 8 injury accidents per year), on average, on their accident rate:-
as a control group.
Due to the uncertainty andreliably complexity of allowing for this effect at any site it is suggested, therefore, that where the highest accident sites are chosen for treatment, then the above order of allowance should be made when calculating any estimate of the actual reduction in accidents the countermeasures have produced.
Possibly the most straightforward way of allowing for both the regression to mean effect and changes in the environment would be to use control sites chosen in exactly the same way as the treated sites, and identified as having similar problems, but left untreated. In practice, it is both difficult to find matched control sites and, if investigated, to justify not treating them. There has been much debate among statisticians over many years on this subject and the best way to deal with it (see refs. 2,3,4,5). Cawangan Jalan, Ibu Pejabat JKR, K.L
Period of accident data considered 1 year
Regression to mean change in annual accident rate 15 to 26%
2 year
7 to 15%
3 year
5 to 11%
7:10.1.4 Accident Migration
There is still some controversy over whether or not this effect exists but it has been reported by several researchers 6,7,8. It is simply that an increase in accidents tends to be observed at sites adjoining a successfully Page 70
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
treated site giving an apparent transfer or “migration” of accidents. It can be detected by comparing the accident frequencies in the surrounding area before and after implementation of treatments at sites in the area with a suitable control. It is unclear precisely why this effect occurs but is suspected that drivers are “compensating” for the improved safety at treated sites by being less cautious elsewhere. Again, there are no established techniques yet available to estimate this effect for a particular site. The first reported occurrence of this feature 6 found an overa1l increase in surrounding areas of about 9% and a later study8 of a larger number of sites estimated 0.2 accidents/site/year acci dents/site/year..
7:10.2 Before And After Periods
There are a number of points to taken into account when choosing periods to compare before and after the treatment was applied:
Whilst again the extent of this effect is extremely difficult to monitor, the engineer should be aware of the possibility of risk compensation when introducing countermeasures. For example, a scheme giving pedestrians more apparent priority using speed tables or raised pedestrian crossings (which give the impression of extensions to the footway) may lead the pedestrian into taking much less care in crossing the road. For further reading on this subject see references 9,10.
that at the control site. The period during which work was carried out should be omitted from the study. If this period was not recorded precisely, a longer period containing it should be omitted.
The before period should be long enough to provide a good statistical estimate of the true accident rate (so as to remove as far as possible random fluctuations).It should not, however,, include periods where the however site had different characteristics. Three years is widely regarded as a reasonable period to use.
The same applies to the after period which ideally should also be three years. However, results are often required much sooner than this. A one year after period can initially be used if there is no reason why this should bias the result (as long as the same period is used at the control sites). However, sensitivity is lost and the estimate of the countermeasure’ss success should be updated measure’ later when more data becomes available.
7:10.1.5 Risk Compensation
This is an even more controversial effect, though related to the previous section. The philosophy of “risk compensation” or “risk homeostasis theory” suggests that road users will change their risk-taking behaviour to compensate for any improvements in road safety. That is, road users tend to maintain a fixed level of accepted risk, so will take more risks when given greater accident protection, for example, if provided with seat belts or anti-lock brakes.
Before and after periods at the treated site should be identical to
7:10.3 Standard Tests On Accident Changes
In evaluating a treatment the answers to the following questions will usually be required:
Has the treatment been effective?
If so, how effective has it been?
It is assumed that the user of this Guide will Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 71
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
need to interpret accident data practically without necessarily understanding the underlying statistical theory. For this purpose it is sufficient to assume that the before and after accidents are drawn from a normal or Gaussian distribution. This means that we can use the Chi-square test to answer the first question above, ie. whether the changes at the site were statistically significant. However, let us first consider the size of that change by using the k test. 7:10.3.1 The k Test
It is possible that although accident levels reduced at a treated site in an “after” period, the general level of accidents is also reducing; the “real” reduction at the site due to the treatment thus being less than the actual numbers observed (ie. over estimating effectiveness). Conversely, Conversely, if the general level of accidents is increasing an underestimate of the treatment would be obtained. The k test can be used to show how the accident numbers at a site change relative to control data. For a given site or group of similarly treated sites, let:a b c d then
= before accidents at site = after accidents at site = before accidents at control = after accidents at control k = b/a d/c
If k < l then there has been a decrease in accidents relative so the control; if k = l then there has been no change change relative to the control; and if k > 1 then there has been an increase relative to the control. The percentage change at the site is given by
Example:
Let us assume that Table 7.2 gives the annual injury accident totals for a priority T-junction in a semi-urban area which had Stop on the minor originally, but wheresigns a roundabout wasroad installed three years ago. The control data used are accidents on all other priority junctions in the District over exactly the same 6 year period.
Therefore, as k < 1there has been a decrease in accidents relative to the controls of:
or, if any of the frequencies are zero then 1/2 should be added to each, ie: Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 72
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
7:10.3.2 The Chi-Squared Test
7:10.3.3 Group Of Sites With Same Treatment
It is important to answer whether the above change in accidents was indeed produced by the treatment or whether this occurred by chance. This test thus determines whether the changes are statistically significant. with reference to the above table chi-squared
For a number of sites, N, which have had the same treatment, the overall effect is a rather more complex calculation, ie. by solving the following equation for K over all the sites, ie. i = l to N. The other symbols
is calculated by the formula:
are as in previous equations.
For testing, the natural logarithm of a variable such as K is usually found to have a more symmetrical distribution (amenable to standard statistical treatments), and the standard error, σ, of loge K can be approximated to the following:
Example:
Now looking at the chi-squared distribution table (Appendix D) and the first line (one degree of freedom, v=1), the value for chisquare of 5.38 lies between 3.84 and 5.41. This corresponds to a value of significance level (on the column header line) between 0.05 and 0.02, which is normally quoted as greater than the lower level, ie. better than the 5% level of significance. This means that them is only a 5% like1ihood (or l in 20 chance) that the change in accidents is due so random fluctuation. Another way of stating this is that there is a 95% (100% - 5%) confidence that a real change in accidents has occurred at the junction.
The following ratio should then be calculated using loge of the value of K calculated above and its standard error from the previous equation:
±
and if this t), value outside theisrange 1.96 (Student's thenisthe change statistically significant at least the 5 per cent level. Now to test whether the changes at the treated sites are in fact producing the same effect on accident rates, calculate the following chi-squared value.
The 5% level or better is widely accepted as the level in which the remedial action has certainly worked, though the 10% level can be regarded as an indication of an effect.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 73
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
If this is significant with N-l degrees of freedom [refer to the (N-1)th] row in the table of Appendix D, where N is the number of treated sites], then unfortunately, the changes at the sizes are not producing the same effect. If non-significant, then it is likely that they are producing the same
to the treatment is negligible, the First Year Year Rate of Return (FYRR) is simply given by:
effect.
The above FYRR should be rounded so 90% to give m figure indication of the possible effect of using this treatment in the future.
7:10.4 Economic Evaluation
For every scheme the evaluation should include an indication of the benefits actually achieved in relation to cost. In the previous sections we have already seen how we can determine a best estimate of the size of the effect on accidents. Considering again the example in Fig.7.2, the estimate of the reduction was 72.2%. If the site was one of the worst blackspots in the District, then we ought to make some allowance for 7:10.1.3,let the regression-to-mean effect. From Section us assume this amounts to as much as 11%, such that our best estimate of the true reduction in accidents due to installation of the roundabout is 61.2% (72.2%-11%). (72.2%-11%). Since the original number of accidents at the site was 20, this represents a saving of 12.24 accidents over the study period. As the “before” period in this case was 3 years, the best estimate of savings is 4.08 accidents per year. It should be noted that only injury accidents have been considered here but if there had been reliable numbers of damage-only accidents which were also reduced, then a separate costing of these should perhaps also be carried out. Using the average injury accident cost of RM 33,000 used in Step7 (see Section 5:7.2), this accident saving amounts to RM134,640 per year. This figure is then compared to the costs of the treatment which totalled say, RM150,000. Assuming Assuming delay to traffic due Cawangan Jalan, Ibu Pejabat JKR, K.L
This is the same technique as outlined in Chapter 5 and the Net Present Value Value figures can also be calculated for completed schemes following the example given in that Chapter.. This would be particularly Chapter advisab1e if there are considerable new maintenance costs associated with the installed measure. It is only by evaluating and recording results in this way that a listing of implemented remedial measures and their effectiveness can be built up for the use of road authorities throughout the country. country. 7:10.5 Evaluating Overall Effectiveness
This chapter has concentrated on evaluating the effects of specific schemes. In addition there is a need for the regular strategy document mentioned in Chapter I to provide a summary of the overall achievements of road safety programmes. As background information in the strategy document it is normal to present and examine aggregate accident statistics over the State, District or Municipality, broken down in various ways by, for example, Class of road user, class of road. These aggregate figures can be useful not only in indicating general priorities but also in evaluating the effects of wide-scale safety campaigns, legislative and/or enforcement changes. However, as schemes are usua1ly localised, However, their effects are often difficult to detect among much larger accident totals. Hence in the strategy document or Road Safety Plan it will probably be better to give a summary Page 74
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
listing of the effectiveness of all the low-cost schemes (see Appendix A examples). This is more informative than a single overall figure as it displays the range
of safety efforts taking place and the relative success of the various methods used.
REFERENCES 1.
RADI RADIN N UMA UMAR R R S, S, G M MACK MACKA AY, B L HI HILL LLS. S. Prel Prelim imin inar ary y ana analy lysi siss of of motorcycle accidents:short-term impacts of the running headlights campaign and regulation in Malaysia.To be published in J.of Traffic Medicine-1995
2.
HAUER E,& P BYER. Bias by selection: the accumcy of an unbiased estimator. estimator. Accid-Anal.& Prevention-15, 5, pp323-328-1983.
3.
ABBESS C, D JARRETT, JARRETT, C C WRIGHT-Accidents WRIGHT-Accidents at blackspots: estimating the effectiveness of remedial treatment,with special reference to the ‘regression to mean’ effect. Traffic Traffic Engineering & Control, 22 22 10-1981.
4.
WRIGHT C C, C R ABBESS & D F JARRETT. JARRETT. Estimating the regression-to-mean effect associated road accident blackspot treatment:- towards approach. Accid. with Anal.& Prevention. 20, 3, pp199-214, 1986. a more realistic
5.
MOUNTAIN L,B FAWAZ & L SINENG. SINENG. The assessment assessme nt of changes in accident accide nt frequencies on link segments: a comparison of four methods.Traffic Engineering & Control, 33, 7 -1992.
6.
BOYLE, A J & C C WRIGHT. WRIGHT. Accident migration after remedial treatment at accident blackspots. Traffic Engineering & Control, 25, 5 -1984.
7.
PERSAUD, B.Migration of accident risk after remedial treatment at accident blackspots. Traffic Traffic Engineering & Control, 28,1-1987.
8.
MOUN MOUNT TAI AIN N L, & B FA FAWAZ. AZ. T The he eff effec ects ts of of engi engine neer erin ing g meas measur ures es on on safe safety ty at at adjacent sites. Traffic Traffic Engineering & Control, 33, 1-1992
9.
ADAM ADAMS S J. J. Ris Risk k and and fre freed edom om:th :thee reco record rd of roa road d saf safet ety y regu regula lati tion on.. T Tra rans nspo port rt Publishing Projects. Cardiff -1985.
10 10..
TRIM TRIMPO POP P R M, & G J S WILD WILDE E Chal Challe leng nges es to to acci accide dent nt pre preve vent ntio ion: n: the the iss issue ue of of risk compensation behaviour. STYX Publications, Postbus 2659, 9704 CR Groningen, The Netherlands -1994.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 75
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
CHAPTER VIII SUMMARY Summary Of Chapter Summary Chapter I - INTRODUCTION
Thesummarised main pointsascovered be follows:in this chapter can
Road accidents are a serious problem in Malaysia with more than one person in every 450 of the population suffering injury or death on the road each year. This Guide focuses on the Engineering aspects of improving safety.
The four basic strategies for reducing accidents are: -Single sites/blackspots -Mass action schemes -Route action plans -Area-wide schemes
Maintenance a reliable accident database is mof essential element of safety work since it constitutes the base measure used to: i) identify the nature and location of problems, and ii) to monitor the effects of remedial action taken.
The Guide also concentrates on the accident approach (althoughreduction the priciples applied in devising colinsermeasures also be adopted at the design stage in order to prevent accidents)
Summary Of Chapter Summary Chapter II - ACCIDENT DAT DATABASE
Accident Investigation Units should be set up in all road authorities with one full-time staff per 400-1000 accidents per year.
The data originates with the Police but the responsibilities of various authorities are summarised below:-
Balai Police
- Attend scene of accident and record details - Enter basic details in 24.hour incident report book - Open investigation file (for all injury and some damage-only accidents) - Complete POL27 form
The national reduction to targets shouldcasualty be disaggregated the local level, and a planned giving details of how they will be met, and reporting progress/success.
An annual Road Safety Plan should be produced by all road authorities to include a local accident analysis, statement of the targets, giving details of how they will be met, and reporting progress/success.
Output:
- Main copy of POL27 to HQ - Pages 4 & 15 copies to District JKR/Local Government Office - Investigation papers for court case Police HQ-Bukit Aman
- Receive POL27 forms, check, and enter onto mainframe computer - Download Download data to MAAP MAAP files Output:
- Send copy of MAAP files to HPU
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 76
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
- Single site listing by worst nodes and links, ie. descending order of accidents.
JKR District Office
- Produce appropriate scale maps marked with landmarks. - Check POL27 forms for accident location on all State & Federal roads - Fill in Section No, 100m distance, map code, X-Y coordinate coordinate coding. coding.
- Mass action sites can be ranked according to numbers of accidents of a selected factor (eg. night time accidents) - For Area wide action , residential areas need to be divided into approximately l km squares (though irregular shapes bounded by rail lines, roads, rivers etc. will ultimately be used).These areas are usually ranked by numbers of vulnerable road user accidents.
Output:
- Completed pages 4 & 15 of POL27 send to HPU Local Government Department
- Produce appropriate nodal scale maps. - Check POL27 forms for accident location on all Urban principal and minor roads. - Fill in Node, Link Nos, map code, X-Y coordinate coding.
STEP 2: Preliminary Accident Analysis Output:
- Completed pages 4 & 15 of POL27-send to HPU
Test sites in first listing to ensure high accident numbers have not occurred by chance.
Highway Planning Unit (MOW)
- Coordinate all location data - Merge this with data from Police HQ into MAAP computer computer files. files.
Produce stick diagrams to help look for common patterns of accident.
STEP 3: Initial Initial Site Visit Visit Output:
- Complete data set copy sent to Police HQ - Appropriate data files made available to all authorities.
Summary Of Chapter Summary Chapter III - INVESTIGATION
Step 1: Identifying And Prioritising Sites
Check plans are up to date. Visualise accidents on record to confirm manoeuvres and make preliminary judgement of causes. Photograph site. Classify sites if possible as ‘easy’ or ‘hard’.
Define a reaction level above which action should be taken. Search data ideally covering a period of 3 years using MAAP to produce initial ranking of sites. These can be:- Route action listing in descending order of accident totals ( or points) per Section Number per year.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 77
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Summary Of Chapter Summary Chapter IV - DIAGNOSIS
i) Supplement accident data, ii) Help diagnose diagnose problems, problems, and
STEP 4 Collection Of Further Data And Analysis
iii) Use in future evaluation of remedial work implemented.
Study accident data at the site in more detail including sketch diagrams, and produce collision diagrams.
Classify accidents into types.
Amend stick diagrams to include my further information and search for dominant accident patterns.
Gather any available data such as traffic flow, dates of road alterations.
Determine likely human factors any perceptual traps.
Summ Summary ary Of Chapter Chapter V - SELECTION
STEP 6 Select Select Posibble Posibble Countermeasures
A road hierarchy hierarchy should should be estab1ished.
Decide on economic objectives for the different scheme types (eg.50% FYRR at single sites to10-25% area-wide action).
Several treatment proposals(lists for a site should be considered of schemes shown to be effective in several countries are given), and each should:-
STEP 5: Site Studies Studies And Analysis Analysis
Make further site visit and look for likely features which may be contributing to accidents.
- aim to reduce the prevalent accident type(s);
Plan the following further studies, as necessary:
- not increase other types of accident at the site or in the surrounding area;
- Traffic flow manoeuvre counts. - Pedestrian road crossing flow in marked road lengths if relevant.
- not cause undesirable effects on other traffic movements or on the environment.
For area-wide residential schemes, aim to reduce speeds as opposed to restricting vehicle movements and maintain good consultation with locals and the emergency services.
Consider whether schemes need to incorporate road user training and media campaigns.
- Speed measurements on approach to junctions or bends indication of possible problems. - Take still photographs and/or video as a record for report, or use to study problem behaviour.. behaviour - Conduct traffic conflict study most useful at junctions to :
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 78
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
STEP 7: Priortise Treatments Treatments & Sites
Estimate costs and benefits of each treatment for which the following is required:
Monitoring should be carried out at different levels: for the whole area covered by the road authority; immediately after each individual scheme; and for longer term conclusions.
- capital costs - estimate of benefits - estimate of any disbenefits.
Select most cost-effective solution in terms of best First Year Year Rate of Return (FYRR) or best Net Present Value to Present Value Cost ratio (NPV/PVC). Only those schemes with FYRR > 50% should initially be considered.
Draw up list of sites in priority order of best NPV/PVC ratio, and decide on cut-off of sites to be
Variables other than accident frequency can be used to monitor the effectiveness, particularly in relation to the objectives of the particular countermeasure. A series of other factors are strongly recommended for area-wide schemes. It is important to monitor other surrounding areas which could be affected by the scheme and to identify as large a group of control sites as possible of similar nature but well away from the study sites.
treated within the budget. Summary Of Chapter Summary Chapter VI - IMPLEMENTATION
STEP 8: Detailed Detailed Design Design And Installation
Carry out a road safety audit at the design stage and immediately before opening to traffic.
Remedial action schemes should be evaluated so that knowledge can be gained about relative performances. This will assist decision making on efficient allocation of resources in the future.
Statistical tests should be used in before and after studies to compare accident changes at the treated sites with the control sites.
The investigator should make allowance for the known other factors that can affect the estimate of the effect of the measure on accidents (eg. regression-to-mean).
As a guide to overall effectiveness of a road authority’s road safety programme, a summary list of individual schemes, grouped in an appropriate manner, should be produced and included in the Road Safety Plan document.
Ensure adequate safety standards are followed at the safety improvement construction site.
STEP 10: Evalu Evaluation ation
Maintain a 1og of dates of the works and of actual costs.
Summary Of Chapter Summary Chapter VII - EVALUATION
STEP 9: Monito Monitoring ring
The importance of monitoring after scheme implementation is stressed .
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 79
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX A Exampl Exa mple e of Ro Road ad Safety Safety P Plan lans s This Appendix contains extracts from recent road safety strategy documents or Road Safety Plans of two highway authorities authorities in the United Kingdom, ie. Doncaster Metropolitan Borough Council and Berkshire County Council. They are included as examples only of the format anddocuments type of information which, it is recommended, be incorporated in similar published in Malaysia. Road Safety Plans are now produced annually by most road authorities in the United Kingdom regardless of the current success or failure of each in achieving its target. They serve as a means of ensuring that the road authority concentrates on managing its particular problems effectively, effectively, and provide the public and higher authorities with a valuable record of the efforts it is making on their behalf towards improving the safety of the road network. It should be noted that Berkshire County Council, having responsibility for a larger road network, have now appointed a commercial highways and planning consultants, the Babtie Group, to carry out the task of producing their Plan.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 80
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 81
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 82
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 83
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 84
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 85
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 86
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 87
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 88
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 89
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 90
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 91
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 92
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 93
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 94
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 95
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
2.0 PROGRESS TOWARDS TARGET Berkshire Progress
In 1989 when the County Council Council adopted the casualty reduction target, some some 4354 casualties had resulted from accidents on Berkshire’s roads. In order to achieve the target reduction of one third of the average 1981-85 figure by the year 2000, this figure must reduced to 2847. Progress towards the target is shown opposite. Whilst casualty numbers continue to decline it can be seen that an increased rate of reduction is now required if the objective is to be achieved. National Progress
For comparison purpose the national progress towards the year 2000 target is also shown opposite. It can be seen performance in Berkshire is better than the national average. It can be seen that both the Berkshire and National trends show the same effect: an early and relatively rapid reduction as the “easier” sites are tackled, followed by a levelling off as more difficult problems (often related to behaviour as much as to the road layout) have to be tackled.
WHILST CASUALTY NUMBERS NUMBERS CONTINUE TO DECL DECLINE, INE, MORE SIGNIFICANT REDUCTIONS ARE REQUIRED TO MEET THE YEAR 2000 TARGET
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 96
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
3.0 ACCIDENT TRENDS AND ANALYSIS An analysis of accident trends has shown that speeding particularly in urban areas continues to be a major factor in accident causation. The young and inexperienced road users feature predominantly in Berkshire's accident statistics, and the fact that two-thirds of all casualties resulting from accidents on Berkshire's roads are the drivers and passengers of cars is a cause for concern. It is at these key areas that resources must be targeted if the year 2000 targets is to be met. The Size Of The Problem
Throughout 1993 on roads in Berkshire there was a total of 2773 accidents involving personal injury. injury. These resulted in 3672 casualties of which: 33 347 292
were fatal were sseerious in injuries 3 were slight injuries
The total number of fatalities and serious injuries has declined every year since 1989 and now represents a 65% reduction over the 1981 - 85 average figure.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 97
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
The only specific road user group showing and increasing casualties trend is that of car occupants. Casualty levels in this group are currently 10% higher than the 1981-85 average. It is clear that this casualty group holds the key to achieving the casualty reduction target and the County Council's objective in the next few years will be to reduce casualties in this area.
THE KEY TO TO ACHIEVING ACHIEVING THE YEAR 2000 TARGET TARGET WILL BE REDUCE T TO O REDUCE CASUALITIES IN THE CAR OCCUPANT ROAD USER GROUP Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 98
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Accidents Locatios In Berkshire
As can be seen in the figures below, the majority of accidents in Berkshire take place on urban roads. (i.e. those roads subject to a speed limit of 40mph or less). It is evident that if the casualty reduction target is to be met then a significant proportion of engineering programmes and police enforcement activity must be focused on these areas .
THE COUNTY COUNTY COUNCIL COUNCIL WILL CONTINUE TO FOCUS FOCUS ROAD SAFETY RESOU RESOURCES RCES TOWARDS TOW ARDS ACCIDENT CONCENTRATIONS, CONCENTRATIONS, PARTICULARLY PARTICULARLY THOSE IN URBAN AREAS
Vulnerable Road User Groups In Berkshire Berkshire
The distribution of road traffic casualties by age and road user class shows that inexperienced users of each mode are more likely to be involved in accidents. This is particularly evident with the young car driver since accident records show that over one-third of all accidents involve drivers in the 17 to 24 age range.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 99
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
As discussed in later sections, Education Training and Publicity initiatives are directed principally at these inexperienced and hence vulnerable groups. Speed
Speeding by drivers and riders is the most common contributory factor to deaths and injuries on Berkshire's roads. A study by the Transport Transport Research Laboratory (TRL) (TRL) has indicated that between 22 and 32 percent of the accidents studied had excessive speed as a contributory factor. In terms of the 1993 injury road accident figures for Berkshire, this would indicate that speeding had been involved in over 800 injury road accidents. As discussed in later sections, the theme of excessive speed and road safety will play a major part in the County Council's engineering and education programmes and in the enforcement activity of Thames Valley Valley Police.
AN ANALYSIS BY TRL INDICATES INDICATES THAT A REDUCTION IN AVERAGE AVERAG E SPEED EVERYWHERE OF THE ORDER OF 1 KPH COULD SAVE SAVE 5 PERCENT PERCE NT OF ALL INJUR INJURY YA ACCID CCIDENTS ENTS
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 100
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
4.0 ENGINEERING ACTION The great majority of schemes in the County Council's highway works programme will result in road safety improvements. Indeed the Council's scheme selection procedures are heavily biased towards tackling sites with a persistent injury accident problem. In recognising the importance of reducing the level of accidents on the County's roads the Government allows Supplementary Credit Approval (SCA) to be claimed against engineering projects which can demonstrate potential injury accident savings. Hence it is in the County's interest to promote such schemes above those to which no potential injury accident savings can be ascribed. Recent engineering road safety initiatives are listed below:
TRAFFIC CALMING has continued to be heavily promoted within Berkshire with numerous projects including 20mph zones, chicanes, road narrowings, road humps and mini roundabouts being installed around the County.
In recognition of the accidents caused through excessive speed by drivers and riders in urban areas, Berkshire County Council quickly responded to the regulations issued in 1992 by installing the first 22 fixed speed camera sites in Berkshire. This work will continue in 1994/95 with a further 50 sites planned and complements red light enforcement cameras already installed.
The first non-trunk road variable speed limit scheme in the country was introduced in Slough.
The first Berkshire Toucan crossing is being progressed in Newbury.
Berkshire’ss 3rd Puffin crossing is being progressed in Thatcham. Berkshire’
Appendix A lists local safety engineering schemes planned for 1994/95 together with comment on their estimated effect on injury accidents over a three year period. As can be seen considerable emphasis will continue to be placed on tackling the problems of speeding in urban area with many speed camera and traffic calming schemes programmed at identified measures high risk accident schemes are seen as one theexpected prime accident reduction currentlysites. beingSuch implemented in Berkshire andofare to make a major contribution towards changing attitudes to inappropriate speeds. The County Council carries out a continuous programme of investigation of accident patterns at High Risk Sites. Potential schemes arising from this process are presented to Members for funding consideration The potential for reducing accidents will continue to be taken into account when deciding on the priorities for traffic management works promoted by the Area Highway Sections. Appendix B lists all High Risk sites in Berkshire with comment on action already taken or proposed.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 101
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
A minimum of 100 sites will be investigated in depth depth in 1994/95 either as individual sites or as part of an area wide or route approach. In addition to this, accident investigation will continue to feature predominantly in the County Council's continuing programme of Area Traffic and Transport Studies. The County Council continues to monitor its success in accident remedial work to ensure a cost effective approach to the work. Indeed, this feedback is essential to maintain levels of expertise in this rapidly developing field. Appendix C gives before and after information and clearly demonstrates the cost effectiveness of this work. The County Council continues to improve road safety through its close control of development proposals and liaison with the local Planning Authorities. Every opportunity will be taken to resolve existing road safety problems or prevent others arising as a result of new development. Developer contributions will also be actively sought to enable positive accident prevention and remedial work to be pursued at identified problem sites.
CONSIDERABLE EMPHASIS EMPHASIS WILL CONTINUE TO TO BE PLACED ON TACKLING THE PROBLEM OF SPEEDING IN VRBANAREAS AT IDENTIFIED HIGH RISKACCIDENT RISKACCIDENT SITE S. S.
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 102
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 103
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 104
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 105
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX B POL27 Accident Accident Report Form
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 106
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 107
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 108
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 109
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 110
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 111
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 112
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX C Poisson Probabilities (Single Factor Values)
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 113
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX D Chi-squared Table
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 114
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX E Standard Symbol For Collision Diagrams
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 115
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX F Discount Factor Tables
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 116
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
APPENDIX G Comparison Compa rison Of Mean Speed Measu Measurr emen ements ts and and T-Distribution Table COMPARISON OF MEAN SPEED
Having found the value of t we look at
MEASUREMENTS To determine whether the mean speed of one set of speed measurements is significantly different from another (ie.between a “before” and “after” study), it is necessary to determine the means and standard deviation of the difference in means.
b- 2) degrees the table over with of (n ta+n of freedom. If page the value exceeds that for the 5% level (the t =0.05column) we can be 95% conndent that the true mean speed has changed.
Let b1,b2..................bnb the before speed readings and a1,a2,..............ana be the after speed readings we then ca1culate the equations below:
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 117
FOR INTERNAL INTERNAL USE ONL ONLY Y
Interim Guide On Identifying, Priortising And Treating Hazardous Locations On Roads In Malaysia
Cawangan Jalan, Ibu Pejabat JKR, K.L
Page 118
View more...
Comments