Technical Assistance Consultant’s Report

Project Number: 49063-001 June 2019

Islamic Republic of Pakistan: Enabling Economic Corridors through Sustainable Transport Sector Development

Prepared by NTU International Aalborg, Denmark

For Ministry of Communications

This consultant’s report does not necessarily reflect the views of ADB or the Government concerned, and ADB and the Government cannot be held liable for its contents.

The preparation of these Guidelines was assisted by a technical assistance grant financed by the UK Department for International Development and administered by the Asian Development Bank.

Published by: Ministry of Communications, Islamabad June 2019 Photo credits: Asian Development Bank, Pakistan www.communication.gov.pk

Guidelines for Road Safety Engineering|Part I

ABBREVIATIONS AND ACRONYMS

AASHTO American Association of State Highway and Transportation Officials AADT Average Annual Daily Traffic ADB Asian Development Bank AfDB African Development Bank AJK Azad Jammu and Kashmir CAREC Central Asia Regional Economic Cooperation CDA Capital Development Authority CMF Crash Modification Factor EC European Commission ECSP Engineering Consultancy Services Punjab EN European Norm EU European Union ESCAP Economic and Social Commission for Asia and the Pacific FYRR First Year Rate of Return GB Gilgit-Baltistan GDP Gross Domestic Product GoP Government of Pakistan GPS Global Positioning System iRAP International Road Assessment Programme ITP Islamabad Traffic Police KP Khyber Pakhtunkhwa MASH Manual for Assessing Safety Hardware MoC Ministry of Communications NHA National Highway Authority NH&MP National Highway and Motorway Police NPV Net Present Value NTRC National Transport Research Centre NUST National University of Sciences and Technology OECD Organisation for Economic Co-operation and Development PAK Pakistan PKHA Pakhtunkhwa Highways Authority PIARC World Road Association RSA Road Safety Audit RSI Road Safety Inspection

i Guidelines for Road Safety Engineering|Part I

RSIA Road Safety Impact Assessment TA Technical Assistance UN United Nations US United States WHO World Health Organisation

ii Guidelines for Road Safety Engineering|Part I

TABLE OF CONTENTS

Preamble ...... 1 Introduction ...... 3 1 Reference road safety framework ...... 6 1.1 Pakistani road network ...... 6 1.2 The current road safety situation ...... 7 2 Review of existing standards and guidelines ...... 16 2.1 Road design standards currently used ...... 16 2.2 New geometric design standards ...... 16 2.3 Asian Highway design standards ...... 17 2.4 CAREC Road Safety Engineering Manuals ...... 18 2.5 NHA Road Safety Audit guidelines ...... 19 3 Sustainable safety principles for road design ...... 22 3.1 Safe System principles ...... 22 3.2 Sustainable Safety principles ...... 23 3.3 Categorization of roads and network design ...... 35 3.4 Speed management ...... 40 4 Road infrastructure safety management ...... 52 4.1 Road safety strategies ...... 53 4.2 Road safety impact assessment ...... 54 4.3 Road safety audits ...... 57 4.4 Road safety inspections ...... 60 4.5 Treatment of crash locations ...... 65 References ...... 79 Annex – Road design standards’ benchmark ...... 81

iii

Guidelines for Road I Safety Engineering|Part I

Preamble

The objective of the TA-8990 PAK: Enabling Economic Corridors through Sustainable Transport Sector - Road Safety Component is to assist the Government of Pakistan (GoP) to establish sustainable national road safety institutions and to build the structures, processes and capacity required at all levels to achieve long term reductions in road traffic deaths. The project structure is aligned with the UN Road Safety Pillars and underpinned by the Safe System Approach. Globally, Pillar 2 ‘Safe Roads’ is a key focus area. Currently road safety, particularly the safety of vulnerable road users, such as motorcycle riders, pedestrians, bicyclists and occupants of 3-wheeled vehicles is not sufficiently prioritised in road design, construction, improvement and maintenance. These Road Safety Engineering Guidelines aim to provide a suitable document for practitioners both being engineers in agencies with a responsibility for road construction and network management and private consultants. The objective is therefore to provide a practical tool that will be used by managers and engineers responsible for designing, improving, and maintaining all classes of road at all government levels in Pakistan. The development of Guidelines content and format is being coordinated with two concurrent NHA projects: development of Pakistan Road Design Guidelines and the introduction of Pakistan International Road Assessment Program to ensure that the three initiatives are integrated. The overall objective is to support Federal, Provincial and Territory road agencies to deliver safer roads throughout Pakistan. The Guidelines consist of two parts as follows: Part I - General recommendations for safer roads Methodological document in which, after a brief overview of the main road safety issues affecting Pakistan, a systemic route is defined for solving problems in a cost-effective perspective. Reference is made to international best practices such as, for example, the European Directive 2008/96/EC, or road safety engineering manuals published in the countries historically most reputable in the field of road safety (e.g. UK, Netherland, Ireland, Australia, etc.). Part II - Catalogue of countermeasures for typical road safety issues Practical design guide consisting of a sample of forms describing typical cases of infrastructural deficiencies and possible countermeasures. The catalogue gives brief information, including pictorial representations, of well-known design errors in a readily understood way, will suggest a range of methods to overcome these and will give

1 Guidelines for Road Safety Engineering|Part I

an indication of the countermeasure costs and benefits to facilitate prioritisation of the work. The catalogue can be used both as a proactive safety tool to ensure the design faults do not arise in the first place, or a reactive safety tool to assist in designing cost-effective countermeasures where problems already exist on the road network.

2 Guidelines for Road I Safety Engineering|Part I

Introduction

In Pakistan, WHO statistics (WHO, 2018) indicate that more than 27,000 people were killed in road collisions in 20161. According to the international statistics2, we can estimate that the death of about one third of them is – at least partially – related to the road environment. In fact, although reliable crash data are still not available in Pakistan, there is information about many collisions where the road alignment, equipment or traffic management were one of the main factors Consequently, it is proven that a good and ‘safety-oriented’ design may reduce the probability and severity of crashes, and measurable safety gains have been achieved in all countries where this approach has been implemented. On the other hand, if this new approach to design is not part of a management system that allows the entire cycle from planning to the construction and maintenance of the infrastructure to be effectively controlled, there is a risk of providing a sterile tool that is not supported by the institutional and programming framework. Currently, Pakistan does not have a shared system for managing road infrastructure safety. Each activity related to road infrastructure security is mainly due to occasional initiatives of a few individual organisations (e.g. NHA, Punjab Road Safety Authority, etc.), without them being part of a broader national design. This is also in addition to a situation where the limited adherence to a specific set of road design guidelines is still an issue. This document, underpinned by the principles of the Safe System approach, is intended to respond to these needs. After a general description of the problem, it describes some key processes, already widely tested in other countries, which, if implemented, can constitute the architecture on which to implant a new (safety-oriented) approach to design. The Part I of the Guidelines consists of four main sections: 1. Reference road safety framework 2. Review of existing standards and guidelines 3. Sustainable safety principles for road design 4. Road infrastructure safety management

1 To make a comparison, in the same year, Pakistan lost 1,803 precious lives in terrorism and insurgency incidences (source: South Asia Terrorism Portal), events that always evoke a dramatic societal response; unfortunately, the same response was not there for the road carnage. 2 Cf. Treat et al. (1979) - Tri-level study of the causes of traffic accidents: Final report. US DoT NHTSA Report DOT HS-805-099, or AASHTO (2010) - Highway Safety Manual. 1st Edn. Washington, DC. 3

Guidelines for Road I Safety Engineering|Part I

5 Redacted

Guidelines for Road I Safety Engineering|Part I

Table 1 – Estimated length of roads in Provinces (km; 2016)

Category Punjab Sindh KP Balochistan GB & AJK Total

Low type 30,901 23,415 12,320 8,460 453 75,549

High type 76,817 58,209 30,625 21,030 1,126 187,807

Total 107,718 81,624 42,945 29,490 1,579 263,356 (source: NTRC)

1.2 The current road safety situation In the following paragraphs the main road safety issues observed across the Pakistani road network are summarized. This assessment does not intend to be exhaustive but is rather a brief overview of the most emblematic topics of a rather critical situation. 1.2.1 Highway hierarchy With the exception of few cases (e.g. Islamabad urban area), the road categories are not easily recognizable. It results in a situation where road users hardly understand which type of behaviour is expected in a specific road section, thus leading to speeding or other dangerous practices. In addition, if the function of a road link is not well defined or understood, a very dangerous mix of traffic categories can Figure 2 - Karachi city centre: disorganized mix be observed. The less homogeneous of traffic functions traffic is, the more dangerous conflicts are likely: differences in speed and mass between road users using the same link or junction at the same time should be reduced to a minimum. On the contrary, in Pakistan, especially on interurban highways or main urban arterials, often we can observe very different vehicles using the same link at the same time: heavy trucks and buses together with motorcycles and rickshaws, fast cars with donkey carts or pedestrians.

7 Guidelines for Road Safety Engineering|Part I

1.2.2 Road alignment The road alignment of the main road network, usually designed according to AASHTO standards, is generally satisfactory although there is a tendency to utilise long straights interspersed with fairly tight radii. This is now known to generate an unacceptably high number of crashes. A mixture of large and medium radii to give a more curvilinear alignment has been shown to reduce the crash rate. Unfortunately, we cannot say the same Figure 3 - Poor alignment of the road to Margalla Hills, Islamabad for most of the local roads, i.e. secondary, tertiary and access roads, that in general seem to be just laid following the terrain, without significant earthworks. This obviously leads to very dangerous alignments, especially in rolling environments. It is not rare to observe very sharp bends, steep gradients and sharp crest vertical curves, thus leading to very poor sight distances. Poor harmonisation of the geometric elements can be also observed. The alignment of some rural provincial roads is composed of long straights, sections of very large radii and, very sudden, very tight bends (without proper signage to alert drivers). As regards the motorways and other trunk roads, some major departures from standards have been observed, i.e. sharp bends without super-elevation, steep gradients and sharp crest vertical curves. These departures, which are justified because of the increase of construction costs in rolling or mountainous environments, lead to unexpected and dangerous situations such as: - Poor sight distances; - Lack of proper transition or termination of the vertical alignment; - Sudden bends, poorly signed, requiring sudden drastic reduction in speed; - Presence of heavy trucks travelling at a very low speed (less than 30 km/h), that on downhill sections, because of their Figure 4 - Poor alignment on the M2 motorway (Kallar Kahar section): sharp bends and steep poorly maintained brakes and state of gradients overloading, contribute to many serious crashes as a result of loss of control.

8 Guidelines for Road I Safety Engineering|Part I

1.2.3 Junctions The vast majority of non-signalised junctions identified both in urban and rural locations throughout Pakistan have no clearly stated priority and are devoid of both road markings and signing which could have been used to convey any instructions or advice to drivers. The practice of traffic management at intersections does not seem to be practised to any great extent and traffic

is invariably left to drivers to sort out how Figure 5 - Urban junction in Islamabad with no to negotiate the intersecting lines of markings or signs travel without actually hitting any other vehicles. Even large complex intersections in large urban areas are without any form of traffic management or control and drivers are left to find their way across many conflicting lines of traffic as best they can. As a result, even during relatively quiet off-peak periods there are massive queues at major junctions because of the need to proceed cautiously and uncertainty shown by many drivers. Roundabouts in Pakistan are not very common. Some examples have been observed in schemes recently implemented in urban or sub-urban areas. Unfortunately, in many cases, the junction layout contains very poor geometry, i.e. small central islands, wide circulating roadway and, above all, minimal or no deflection of trajectories. It can result in poor capacity (and therefore long queues), dangerous conflicts and insufficient speed reduction. As regards motorway interchanges, it has been observed that most of them are not provided with the fully required length of acceleration and deceleration lanes and tapers. In addition, these do not include a weaving section where the entering vehicles can attain the operating speed of the motorway lane and merge into the flow in safety. In case of heavy traffic volume, it could be difficult to carry out this manoeuvre in safety.

9 Guidelines for Road Safety Engineering|Part I

1.2.4 U-turns The need to provide a safe and efficient facility to accommodate U-turn movements is essential for many divided highways in Pakistan. In some cases, especially in urban areas, they are also used to avoid right-turn movements at junctions3. Unfortunately, in many cases the median width neither allows the complete manoeuvre within the median (vehicles are forced to cross the opposite Figure 6 – U-turns along N5 highway with no carriageway and merge in the traffic deceleration/acceleration lanes stream in the travel lane) nor to host safe deceleration and acceleration lanes. In some cases, in order to provide enough room to U-turning long trucks, the cross-section is narrowed, and one or more lanes are devoted to this manoeuvre. It is clear that such narrowing produces a dangerous bottleneck and provides further potential for crashes to occur. These narrowings of carriageway width are usually not signed in advance so that vehicle manoeuvres are invariably undertaken at the last moment. 1.2.5 Cross-section The road cross-section will inevitably have quite a bearing on the relative safety of road users inasmuch that the carriageway type and width, the presence or not of a hard shoulder and whether vehicle restraints are provided are factors which will permit or regulate appropriate vehicle speed. Generally, throughout Pakistan the more strategic roads are well equipped with 2- or 3-lane carriageways in each Figure 7 - Unsafe vehicle restraint system along direction, separated by a median and N5 highway with a hard shoulder. However, even these strategic roads have deficient safety aspects, the main concern being the lack of median crash barrier to prevent the occurrence of cross-over collisions and the lack of crash barrier to prevent errant vehicles from leaving the carriageway and proceeding out-of-control down an embankment or into a structure with obvious results of increased severity of casualties.

3 This is typified by one of the preferred solutions adopted by the Islamabad Traffic Police of closing the problematic junction and the provision of U-turns either side 10 Guidelines for Road I Safety Engineering|Part I

The provision not just of crash barrier but the correct installation of the most appropriate type of barrier is equally important. Incorrect installation will negate any effect that the crash is intended to have with regard to restraining errant vehicles. When installed, the crash barrier must be the correct working distance away from the object it is intended to protect from impact or from the point at which a vehicle would proceed down an embankment, i.e. the back of verge. Again, the type of crash barrier is important. Different types of barriers have different restraint capacity. Accordingly, for each type of road and traffic mix a specific type of barrier should be used. At the moment, on the contrary, we can say that the same type of barrier is used everywhere. 1.2.6 Traffic signs Traffic sign design is a very extensive subject in terms of the legibility, conspicuity, frequency of use, siting and location. Irrespective of the standards currently in use for signing, the general comment with regard to traffic signing in Pakistan in both rural and urban situations is that it is considerably lacking in all these criteria. Traffic signs are not noticeable as a means of communication and are often Figure 8 - Inconsistent signs mostly poorly designed, poorly fabricated and erected, badly located and mostly quite inappropriate for the purpose for which they are intended. Second only to road markings, traffic signs are the next most effective method of guiding, warning, informing and directing traffic. However, they are only effective if basic guidelines are followed. The same criteria should apply to warning and regulatory signs, but they are often not according to any imposed standard. An additional feature that is seriously lacking from the current provision of signing in Pakistan is compliance with the material specification. In many cases signs are made up locally to no set standards. It results in a lack of retroreflectivity – so that signs cannot be discerned during night-time – and use of inappropriate substrates and/or supports that can be hazardous in case of collisions. 1.2.7 Road markings With the exception of the motorway network, pavement markings are often worn, not retroreflective or even absent. In these conditions it is very difficult to have a precise spatial cognition of the roadsides and a vision at a distance of

11 Guidelines for Road Safety Engineering|Part I

the road alignment, especially under poor visibility conditions such as might be encountered in night, rain or fog. Even in central urban areas where a high degree of traffic management is required to cope with high traffic flow levels and at quite complex junctions, there is a conspicuous lack of information and guidance employed on the road surface. Globally the use of lane markings and directions is used to guide traffic into the most appropriate lane when Figure 9 - Road without center lane or edge approaching major junctions. This practice is lines almost non-existent in Pakistan. 1.2.8 Traffic signals Traffic signals, if they do not already suffer from a lack of maintenance, are often inconspicuous incorrectly located, are insufficient for the number of lanes and volumes of traffic, do not appear to have sufficiently illuminated aspects and are often precariously mounted and installed. Particularly on the arterial urban roads, where 3 or more lanes are present, the small size and the location of the traffic lights are inadequate and as a result they are not visible at great distance, especially in heavy Figure 10 - Damaged traffic signals in Karachi traffic, when vehicles ahead may obstruct the view. The result is that drivers may be led to ignore the signs or to brake suddenly. The authorities in Pakistan do not appear to be following any standards or guidelines related to the design or layout of signals with regard to i) location of traffic poles, ii) mounting requirements, iii) size of signal heads, iv) sequence of phasing and v) pavement marking required on the signalized intersections. Moreover, signal timing is also based upon anecdotal experience rather than any specific study or measurement of actual traffic demand. 1.2.9 Temporary traffic management Traffic management for road works is invariably very poor, bordering on the non-existent, in Pakistan. The work site is hardly ever signed or protected (it is often that sites where road construction is taking place to find that traffic is permitted to travel into the construction area), advance warning is usually absent and the transition to the diversion is often very sharp. In general, standard diversion routes are not adopted and traffic is left to find its own way around the construction site.

12 Guidelines for Road I Safety Engineering|Part I

In addition, road works and temporary diversions are not lit and cannot be seen at night-time, signing is not retro- reflective and cannot be discerned in poor light and there is usually confusion with respect to what constitutes the recommended carriageway and what constitutes the work area. In every major project, a Traffic Management Plan is supposed to be developed and submitted to the client by the design consultant. However, during the Figure 11 – Road works poorly signalised implementation stage the approved Traffic Management Plan, if one exists, is invariably ignored. As a result, works sites are very dangerous places for both the traffic and those who have to work within them. 1.2.10 Vulnerable road users Pedestrians have a hard time in Pakistan in both urban and rural settings with very little attention given to the provision of good facilities for walking. It is almost as if it is taken for granted that pedestrians will find a way to their destination and therefore no special facilities need to be provided. Worse still they are expected to find a way across lanes of traffic where, in some places, this can be quite a considerable hazard. Figure 12 - No separation between road vehicles and pedestrians Even in recently built towns, e.g. Islamabad, the pedestrians – and other VRUs – are not properly considered in road planning and design, i.e. pedestrian paths are interrupted, not effective (and therefore not used), not contiuous or even not present. 1.2.11 Traffic calming Traffic calming techniques can be observed rarely in Pakistan. The most common features are speed humps that are often used in a confused manner, i.e. along trunk roads or as isolated measures, without any clear planning or applied strategy. In addition, existing speed humps are often not signalised by markings and/or warning signs, thus being not visible, particularly at night.

13 Guidelines for Road Safety Engineering|Part I

Their profile is usually circular. Flat- topped humps are also frequent, whereas very few sinusoidal profiles have been observed. Again, there is no uniform standard on their application or to cover their design.

Figure 13 - Typical speed hump in Pakistan

14 Guidelines for Road I Safety Engineering|Part I

15 Guidelines for Road Safety Engineering|Part I

2 Review of existing standards and guidelines

2.1 Road design standards currently used In Pakistan, there is no legislation obligating road designers to apply specific standards. Currently the most designers use US standards (AASHTO, ‘A Policy on Geometric Design of Highways and Streets’, 6th Edition, 2011, commonly referred to as the ‘Green Book’). Almost all the roads recently built follows these standards, even at the provincial level. Moreover, it has been observed that also the rehabilitation of old roads, built during the period from 1960 to 1980, is carried out following AASHTO standards. Recently, it has also been observed that the use of Chinese standards is increasing, especially for road projects funded by the Chinese government (e.g. Pakistan Economic Corridor, main arterial highway connecting China border to Gwadar Port). However, these standards are derived mainly from AASHTO standards. A review of ToRs for the design of new roads and rehabilitation of existing road alignments confirms the use of AASHTO standards. It is worth noting, however, that a quick analysis of a sample of documents indicates that some parameters of the AASHTO guidelines are often misinterpreted (or deliberately altered?) with respect to the AASHTO guidance. The actual implementation of these standards is therefore in some cases still far from an acceptable level. The challenge faced by road engineers in implementing AASHTO standards in Pakistan is that they are complex standards and were developed for a completely different road environment. As regards signing and marking, the current standard in Pakistan is the ‘Manual of Signs, Signals and Road Markings’ developed by NTRC in 1989. By many local practitioners, it is expected that this outdated manual will be reviewed and made consistent with current international practices.

2.2 New geometric design standards In order to tackle the critical issues that have risen up in the use of AASHTO standards in Pakistan, at the date of these Guidelines, the NHA is developing an important project for the drafting of new ‘Geometric Design Standards & Parameters for National Highway System of Pakistan’. In order to harmonize the road network in Pakistan for all local and international needs, the new Geometric Design Standards will be country-specific and will address a spectrum of road types, varying from multi-lane motorways to single carriageway roads.

16 Guidelines for Road I Safety Engineering|Part I

A new functional classification of roads is expected and, besides the main Geometric Design Standards, it is expected that the following manuals are also produced: • Manual for road survey • Manual for roadway signage • Manual for pavement marking • Training manuals This document will therefore be the main reference for road designers in Pakistan, who will finally be able to rely on a set of standards tailored to the local situation. In order to provide an overview of the design standards currently in use in some developed countries, a benchmark of the main geometric parameters is provided as an annex to this document.

2.3 Asian Highway design standards In order to address the road safety problem along the Asian Highway Network4 the ESCAP secretariat, conducted a study during 2015-2017 on the development of technical standards on road infrastructure safety facilities. Among others, the objective was to develop detailed design guidelines for the selected road infrastructure safety facilities for the Asian Highway Network. The outcome was a comprehensive document that addresses these facilities from both the road planning and design perspective. The design standard involves both active provision of road infrastructure facilities and avoidance of undesirable practices or design. The design standard, as developed in the abovementioned study, consists of two components5: 1. Design Standards, containing mandatory requirements6; 2. Detail Design Manual, consisting of recommendations pertaining to the design standards.

4 The Asian Highway Network consists of eight core routes that substantially cross more than one sub-region and a number of other routes within sub-regions or ESCAP member countries. The network was formalized through an Intergovernmental Agreement that entered into force in July 2005 (UN ESCAP, 2004). 5 A draft version of both documents is available here: http://www.unescap.org/events/expert-group-meeting-road-infrastructure-safety-facilities- asian-highway 6 In the study it is proposed that these would form Annex IV of the International Agreement on the Asian Highway Network 17 Guidelines for Road Safety Engineering|Part I

The approach followed in drafting the standards is really innovative and addresses road safety from a holistic perspective incorporating modern concepts such as ‘self-explaining roads’7 and ‘forgiving design’8. The purpose of the standards is basically to propose a series of road infrastructure safety facilities that, if implemented, would allow the risk to be reduced and the star rating9 to be increased compared to the baseline scenario10. The guidelines are organised in seven parts as follows: 1. Road infrastructure 2. Intersections 3. Roadside areas 4. Pedestrians, slow vehicles and traffic calming 5. Delineation, pavement markings and lighting 6. Road signage 7. Tunnels The document is therefore an important supplement (and source of valuable information) to the drafting Pakistani geometric design standards.

2.4 CAREC Road Safety Engineering Manuals The Asian Development Bank (ADB) has recently financed a technical assistance for enhancing road safety for CAREC countries11. In the frame of this project three road safety engineering manuals have been produced: 1. Road Safety Audit (March 2018) 2. Safer Road Works (March 2018) 3. Roadside Hazard Management (April 2018)

7 The concept of self-explaining roads encourages road designs that promote road-users to adopt appropriate speeds and behaviour. This subject touches on consistency of alignment design and a well-defined road hierarchy, and should be introduced into the design standard wherever applicable. 8 Forgiving designs aim at giving road-users adequate rooms for errors and limiting the severity of injuries in case of a crash. 9 Star ratings are the indexes used by iRAP to assess the safety of road users. They are based on road inspection data and provide a simple and objective measure of the level of safety that is ‘built-in’ to the road for vehicle occupants, motorcyclists, bicyclists and pedestrians. 10 The ‘baseline’ scenario is the one based on the existing Asian Highway Standards as stipulated in the Annex II to the Intergovernmental Agreement (i.e. Asian Highway classification and design standards). According to the study ‘baseline’ scenarios are in the high risk 1- and 2-star ranges (in a scale 1 to 5). 11 The Central Asia Regional Economic Cooperation (CAREC) Program is a partnership of 11 countries and development partners working together to promote development through cooperation, leading to accelerate economic growth and poverty reduction. Member countries of CAREC are: Afghanistan, Azerbaijan, China, Georgia, Kazakhstan, Kyrgyz Republic, Mongolia, Pakistan, Tajikistan, Turkmenistan and Uzbekistan. 18 Guidelines for Road I Safety Engineering|Part I

These reports cover three key topics in the field of road safety and are therefore practical points of reference for all practitioners in the region. The manuals are mainly addressed to CAREC road projects12, but this does not mean that their use can easily be extended to the entire road network.

Figure 14 – CAREC Road Safety Engineering Manuals

2.5 NHA Road Safety Audit guidelines In August 2018, the NHA issued the first national guidelines for road safety auditing. This is an important step forward in the prevention of road crashes in Pakistan and reflects the NHA’s commitment to contribute to national targets for reducing road deaths and injuries. NHA’s ambitious goal is to make Road Safety Audits (RSAs) a routine activity in the road planning and design process. The document describes policy, procedures and guidelines for planning, undertaking and documenting RSA. It also defines requirements and responsibilities for conducting RSA. It is organized in the following sections: • RSA – An overview • RSA policy • RSA phases • RSA of existing roads – Road safety assessments • Land use development RSA • RSA process

12 CAREC focuses investment and other activities on the development of six competitive transport corridors that link north, south, east, and west through the pivot of Central Asia. The corridors reflect trade flow patterns and will speed the movement of people and goods across the region. Critically, they also connect the mainly landlocked CAREC countries to wider regional and global networks. Pakistan is crossed by Corridors 5a,b and 6c, both of which follow the Peshawar – Islamabad – Lahore - Karachi route. 19 Guidelines for Road Safety Engineering|Part I

• RSA team • Writing RSA report • Responding to RSA report A series of appendices are also provided at the end of the manual to provide prompts for use while auditing, a sample RSA decision tracking form, potential design enhancements to address typical RSA findings on national highways and glossary of terms used in the manual. The RSA manual is primarily developed for RSAs to be conducted on national highways. However, it is suitable for use on the entire Pakistani road network, including networks of Provinces and Territories, by any person with a responsibility for, or an interest in, road safety (e.g. design consultants, police, academics, provincial government officers, students, researchers, etc.). Figure 15 – NHA RSA guidelines

20 Guidelines for Road I Safety Engineering|Part I

21 Guidelines for Road Safety Engineering|Part I

3 Sustainable safety principles for road design

3.1 Safe System principles Safe System principles underpin these Guidelines and the road infrastructure safety management strategy. The identification and removal or treatment of road elements that may contribute to crash occurrence or crash severity is a key component of the Safe System approach to road safety. Adopting a Safe System approach to road safety recognises that humans, as road users are fallible and will continue to make mistakes, and that the community should not penalise people with death or serious injury when they do make mistakes. In a Safe System, therefore, roads (and vehicles) should be designed to reduce the incidence and severity of crashes when they inevitably occur. The Safe System approach requires, in part13: • Designing, constructing and maintaining a road system (roads, vehicles and operating requirements) so that forces on the human body generated in crashes are generally less than those resulting in fatal or debilitating injury. • Improving roads and roadsides to reduce the risk of crashes and minimise harm: measures for higher speed roads including dividing traffic, designing ‘forgiving' roadsides, and providing clear driver guidance. In areas with large numbers of vulnerable road users or substantial collision risk, speed management supplemented by road and roadside treatments is a key strategy for limiting crashes. • Managing speeds, taking into account the risks on different parts of the road system. Road safety engineering is therefore a cornerstone of this strategy. Infrastructure treatments, in fact, can primarily reduce the probability of a crash occurring and secondly to reduce a crash’s severity should it occur. To a lesser extent road safety engineering can even ensure that rescue services can reach a crash site promptly (e.g. providing the motorways with emergency median openings and shoulders).

13 Cf. Australian Transport Council (2006) – National Road Safety Action Plan 2007 and 2008 22 Guidelines for Road I Safety Engineering|Part I

Figure 16 – Safe System approach (source: Safer Roads, Safer Queensland; 2015)

3.2 Sustainable Safety principles

‘In a sustainably safe road traffic system, infrastructure design inherently and drastically reduces crash risk. Should a crash occur, the process that determines crash severity is conditioned in such a way that severe injury is almost excluded.’ From: Naar een duurzaam veilig wegverkeer [Towards sustainably safe road traffic], Koornstra et al., 1992. The concept of Sustainable Safety was launched in the early 1990s in the Netherlands with the ambition stated above. In 2006 this concept was adopted and relaunched by SWOV, a Dutch Institute for Road Safety Research, in order to adapt it, where necessary, to new knowledge and developments (Wegman & Aarts, 2006). The Sustainable Safety vision, which is one of the pillars on which the Safe System approach is built, aims to prevent crashes and, if this is not possible, to reduce crash severity in such a way that (severe) injury risk is almost excluded. These objectives are aimed for by means of a proactive approach informed by prior study of the traffic situations in which serious, injury-producing crashes can occur. The next stage involves two options: either the circumstances are changed in such a way that the crash risk is almost totally removed, or, if this is inevitable, serious crash injury risk is eliminated.

23 Guidelines for Road Safety Engineering|Part I

In the analysis of and approach to preventing crashes or reducing the severity of consequences of dangerous situations, human capacities and limitations are the guiding factors. The central issue is that people, even if they are highly motivated to behave safely while using the road, make errors that may result in crashes. In addition, man is physically vulnerable and this has consequences for injury severity when a crash occurs. Taking into account these human characteristics as the starting point, sustainably safe road traffic can be attained by an integral approach to the components ‘man’, ‘vehicle’ and ‘road’. Focusing on the road infrastructure, this means that the road has to be designed such that it meets human capacities and limitations. Given the fact that people make errors, do not always comply with rules and, moreover, that they are vulnerable, it is essential that ‘gaps’ in the traffic system are prevented in order to avoid a breeding ground for crashes. According to the Sustainable Safety vision, in order to prevent serious unintentional errors, the environment and the task demands that this environment entails have to be adapted to a level that the majority of road users can cope with. This produces, as it were, desirable behaviour almost automatically: the road user knows what to expect (i.e. ‘self-explaining road’), and possible errors can be absorbed by a forgiving environment (i.e. ‘forgiving roads’). This also makes the breeding ground for intentional or unintentional violations less fertile (e.g. speeding would be less likely, as the road environment itself suggests the most appropriate speed). The vulnerable human has to be protected in traffic by the environment by means of structures that absorb the kinetic energy released in a crash. To this end, the mass of vehicles sharing the same space needs to be compatible. If this is not possible, then speeds need to be lowered. This system is embedded in a traffic planning taxonomy of fast traffic flows on the one hand and access to residences on the other. Between these two extremes, traffic has to be guided in good, sustainably safe ways. With this slightly adapted vision on sustainably safe road traffic, SWOV finally arrives at the five central principles: • Functionality • Homogeneity • Predictability • Forgivingness • State awareness A short description of these principles is given in the Table below. The first four principles have strict connection with road infrastructure and road design and are detailed in the following paragraphs.

24 Guidelines for Road I Safety Engineering|Part I

Table 2 – The Sustainable Safety principles

Sustainable Safety principles Description

Functionality of roads Monofunctionality of roads as either through roads, distributor roads, or access roads, in a hierarchically structured road network

Homogeneity of mass and/or speed and Equality in speed, direction, and mass at direction medium and high speeds

Predictability of road course and road user Road environment and road user behaviour by a recognizable road design behaviour that support road user expectations through consistency and continuity in road design

Forgivingness of the environment and of Injury limitation through a forgiving road road users environment and anticipation of road user behaviour

State awareness by the road user Ability to assess one’s task capability to handle the driving task (source: Wegman & Aarts, 2006) 3.2.1 Functionality The first approach to the functional categorization of roads dates back to 1963 when the report Traffic in Towns was published (Buchanan, 1963). This report contained a comprehensive vision for the design of towns and villages in a highly motorized society. A distinction was presented between roads having a traffic flow function (‘distributor designed for movement’), and roads that give access to destinations (‘access roads to serve the buildings’). Elaboration of these ideas resulted in a proposal for a route hierarchy, built up from primary, district and local distributors and access roads to destinations (Figure 17). All roads are grouped into one of these classes, depending on the character of the traffic (i.e. local or long distance) and the degree of land access that they allow. Typically, road users use a combination of arterial, collector, and local roads for their trips. Each type of road has a specific purpose or function: some provide land access to serve each end of the trip; others provide travel mobility at varying levels, which is needed en route (Figure 18). There is a basic relationship between functionally classified highway systems in serving traffic mobility and land access. Arterials provide a high level of mobility and a greater degree of access control, while local facilities provide a high level of access to adjacent properties but a low level of mobility. Collector roads provide a balance between mobility and land access.

25 Guidelines for Road Safety Engineering|Part I

Figure 17 - Functional categorization of Figure 18 - Relationship of functionally classified road roads systems in serving traffic mobility and land access (Buchanan, 1963) (adapted from AASHTO, 2011)

The Sustainable Safety vision builds upon the hierarchy of roads as described above. Based on the functional usage, roads have to be unequivocally distinguishable in the function that they perform (‘monofunctionality’). Motorized traffic should be directed to arterial roads (flow function), causing roads with an access function to be burdened minimally with motorized traffic. Roads with a distribution function (collectors) should direct motorized traffic coming from roads with an access function as quickly as possible to roads with a flow function and vice versa. This principle is meant to prevent unintended use of the infrastructure thus minimizing the number of potential conflicts with severe consequences. On the contrary mixing functions leads to conflicting road design requirements and, hence, to unclear road design for road users, resulting in higher risks. A road network functions properly if function, design and usage (behaviour) are well tuned. There is no reason to discard this principle of sustainably safe road traffic: a functional road network categorization is one where each road or street fulfils only one function – either a flow function, or a distribution function, or an access function (e.g. the ideal arterial road is a motorway, whereas an ideal access road is a 30 km/h street).

26 Guidelines for Road I Safety Engineering|Part I

Figure 19 – Chaotic and unsafe mix of function along N5 highway

Figure 20 – The flow function of the M2 motorway is clear and makes traffic safer

This framework is, generally, accepted in most of the countries and forms part of road design handbooks and categorization plans. In Pakistan, too, this is a well-known concept, but unfortunately it is not always put into practice. However, examples of good practice are also present here. At urban level, the road network of Islamabad is a typical example of hierarchic network, where the three functions are pretty well respected (see Figure 21). The Capital's master plan is in fact a forward-looking and innovative urban experiment developed at the beginning of the 1960s, where the road network also benefits from the regular and well-defined structure of the city. By contrast, in many areas of the country, these principles are not observed thus resulting in dangerous function mix. Typical examples are housing schemes that are increasingly developing along the main arterial roads. These areas, which can accommodate up to thousands of people, often have direct access to the main road, thus mixing different functions and creating dangerous points of conflict (see Figure 22).

27 Guidelines for Road Safety Engineering|Part I

Figure 21 – Aerial view of the regular Islamabad road network

Figure 22 - Access to a big housing scheme along N5 highway 3.2.2 Homogeneity The homogeneous usage of the road aims to avoid large differences in speeds, directions and masses at moderate and high speeds, thus reducing crash severity when crashes cannot be prevented. The corresponding idea is that it is beneficial for road safety when there is little variation in the speeds of close-moving vehicles travelling in the same direction. Worldwide the safest roads are the Figure 23 – Animal drawn vehicle in a trunk road in Pakistan: differences in speed and mass pose serious motorways, based on the number road safety problems of casualties per kilometre driven as the safety indicator. Although driving speeds are the highest, they are

28 Guidelines for Road I Safety Engineering|Part I

relatively uniform and is little variation in direction (e.g. no crossing traffic) and vehicle mass (no pedestrians, cyclists, mopeds or slow-moving vehicles). The 30 km/h zones and residential areas are also relatively safe despite considerable variation in the direction and mass of traffic participants. In these cases, the increased safety is attributable to low driving speeds and small speed variations between different road users. The principle of homogeneous use leads, for example, to operational requirements for directional separation on arterial and collector roads. For intersections, operational requirements are derived from the starting principle to eliminate collisions with high speed and mass differences. Pedestrians, cycles and mopeds should not be present at the points of access of arterial roads. Speed differences should be reduced to acceptable levels at collector roads where mass differences are allowed functionally. In this frame SWOV suggests a system of ‘safe speeds’ (see Table 3) taking into account that speed limits and travel speeds should not be higher than safe crash speeds and that is useful to distinguish between urban and rural areas (although the difference is not always clear for road users).

Table 3 – Safe-speed system

Location Safe travel speed (km/h) Rural road sections Arterial road 120 Collector road with physical separation of driving directions 80 without physical separation of driving directions 70 Local road 40/60 Rural intersections Collector road and local road without vulnerable road users 50 with vulnerable road users 30 Urban road sections Arterial road 50/70 Collector road 40/50 Local road 30 Urban intersections Collector road 50 Local road 30 (source: adapted from Wegman & Aarts, 2006)

29 Guidelines for Road Safety Engineering|Part I

Speed is therefore a very important factor to be taken into account. At lower speeds a driver will have greater opportunity to react and avoid a crash. Speed also affects the severity of crashes. Higher speed crashes involve more kinetic energy (kinetic energy is proportional to the speed squared) and the more energy that is dispersed in a crash, the more severe it tends to be. The Figure 24 shows the expected consequences for three of the main crash types at different speed. As speed increases, the fatality risk increases very sharply for each of the crash types. This leads to several guiding principles for survivability: • Where conflicts between pedestrians and cars are possible, the speed at which most will survive is 30 km/h (red line) • Where side impacts are possible at intersections (e.g. cross roads and T- intersections), the speed at which most will survive is 50 km/h (green line) • Where head-on crashes are possible (e.g. where there is no median separation), the speed at which most will survive is 70 km/h (blue line) The previous ones are therefore the already mentioned safe crash speeds. On this basis potential frontal impacts with crash speeds exceeding 70 km/h have to be excluded. This means that the direction of travel on roads with speeds of 80 km/h or higher will need to be separated in such a way that vehicles cannot hit each other head on.

Figure 24 - Crash types and indicative fatality risk at speeds (source: Wramborg, as cited by AfDB, 2014)

30 Guidelines for Road I Safety Engineering|Part I

3.2.3 Predictability The predictable usage is aimed at preventing human error by offering a road environment to the road user that is recognizable and predictable (i.e. ‘self- explaining’). This indicates permissible road user behaviour and makes the behaviour of other road users more predictable. Accordingly, within a given road category, the road and traffic characteristics have to be as uniform as possible and designed homogeneously because, from a road user perspective, a considerable amount of uniformity is desirable. This principle aims in practice to ensure that the road user can recognize the road type by its road characteristics (recognisability), which makes the road course and the behaviour of other road users more predictable (predictability). Unexpected traffic situations in fact simply cost more time for road users to detect, to perceive, to interpret, to assess, and to elicit the correct behaviour or response. This also means that transitions from one road category to another require the necessary precision and time from road users to adapt their behaviour. In short, we can state that for the recognisability of roads it is important that they: • are distinguishable, and • evoke and support correct expectations. According to some research (Van Schagen et al., 1999), only a limited number of characteristics can be used for distinguishing road categories. These characteristics must be: i. continuously perceivable ii. practical iii. not disadvantageous for road safety In practice, to facilitate recognisability, the number of road classes should be restricted and their design and layout as uniform as possible within each category. Road users will then have a better idea of what sort of driving behaviour is expected of them and be better able to anticipate the driving behaviour of other road users. With ‘self-explaining’ roads, road users will know at which speed to drive, whether to expect traffic from side roads, and whether vulnerable users are likely to be on the road. In practice the aim of ‘self- explaining’ roads is to lower the workload (or mental load) of drivers. This will have a positive influence on the performance of the driving task. The requirements for recognition and predictability are: • avoid unpredictable behaviour by clear designing, marking and signing • make road categories recognizable • limit the number of design elements each category and make them uniform For Sustainable Safety, the limitation of the number of road categories produces the largest contribution to the recognition. This assumes that the

31 Guidelines for Road Safety Engineering|Part I

differences between the categories are large, and within each category are small. A small set of the operational requirements should ensure the predictability of the traffic situations: • speed limits • longitudinal lane/direction road markings • separation of directions • pavement, irregularity of the surface • presence of hard shoulders and obstacle-free zones (emergency lane on motorways) • types of intersections/accesses allowed • expected road user types The above set of items must be clear for each road category.

Figure 25 – Private accesses along an arterial road in Pakistan: manoeuvres associated with their presence are not ‘predictable’ for users

Figure 26 – How to make a road ‘self-explaining’ using marking and signing (source: IRF)

32 Guidelines for Road I Safety Engineering|Part I

3.2.4 Forgivingness The starting principle is that road users make errors and that the environment should be sufficiently forgiving for road users to avoid the severe consequences of these errors. The first step towards making the road user environment forgiving is to make road shoulders sustainably safe. This activity mainly takes place on rural roads (arterial and collector) where the speed is supposed to be higher. A forgiving road has a cross section that is sufficiently wide, has sufficient bearing capacity and obstacle-free shoulders, and is adapted to acceptable risks14 to third parties or risks to car occupants. If this is not feasible and if the danger zone cannot be removed in another way, it is recommended the use of a protective feature (i.e. vehicle restraint systems). It is important to underline that the use of restraint systems is just the last solution to protect road users from roadside hazards. Designers should therefore preliminarily check alternative solutions, namely: • remove the hazards; • make the hazard safe (e.g. by changing the design of the median and verges of the road); • replace the hazards with a passive safe structure (e.g. posts and columns). Vehicle restraint systems represent in fact a hazard in themselves and should therefore only be installed if it is more dangerous to drive off the road than to drive into the vehicle restraint systems. Restraint systems shall be therefore installed where there are one or more hazards within the ‘safety zone’ (or ‘clear zone’). A ‘safety zone’ is an obstacle-free area with flat and gently graded ground, thus providing road users with sufficient space and the right conditions to regain control over their vehicles in case of a run-off15.

14 As regards what constitutes acceptable risk, generally it is translated for shoulders on the basis that if a vehicle leaves the road it should not hit any obstacles causing severe injury. 15 Further details on the ‘clear zone’ concept and, more in general, on the roadside hazard management, are available in the CAREC Road Safety Engineering Manual 3 (CAREC, 2018c) 33 Guidelines for Road Safety Engineering|Part I

Figure 27 – Safety (clear) zone definition

Figure 27 – Example of a ‘safety zone’

According to these principles fixed roadside objects should be designed such that crashes at high speeds cannot result in severe injuries. Here, international criteria for vehicle restraint systems (‘performance classes’) have been established (e.g. European Standards EN 1317-1 to 7, US MASH, etc.). The fact that there are still many road crash victims following impact with protective devices, raises questions as to whether the Figure 28 – Crash against a poorly installed barrier currently used criteria require (N5 highway, Taxila, 15/11/2017 - 2 fatalities) revision, or in turn, the decision to implement a protective device in certain circumstances.

34 Guidelines for Road I Safety Engineering|Part I

Safe shoulders along collector roads are however a difficult subject. Often, the free space is not sufficiently wide, nor has it sufficient bearing capacity, nor is it obstacle-free for protective devices to work in a safe way. In addition, it is not yet general practice in Pakistan - like in many other countries - to protect roadside obstacles on rural collector roads.

3.3 Categorization of roads and network design Categorizing roads is a core activity for sustainably safe infrastructure. The initial three categories (see section 3.2.1) are generally detailed in more classes in order to take into account local circumstances (e.g. distinction can be made between inside and outside urban areas). As stated in the section 0, the influence of the design and the environment on driver anticipation is crucial. For this reason, according to the function, to each category of road must be associated, not only a typical cross-section and a design speed, but also a mode of operation, defining the ‘rules’ to use the specific infrastructure. The layout of a road should therefore be appropriate to its function and ‘automatically’ enforce the desired speed. In Pakistan this is presently explicitly the case for motorways, but not for all other road categories. The Tables below summarizes the essential characteristics of the main road categories, both in rural and urban areas.

35 Guidelines for Road Safety Engineering|Part I

Table 4 – Characteristics of main road categories in rural environment

Road Characteristics Example category Arterial road - speed limit 120 km/h Motorway - grade-separated interchanges - U-turns not allowed - physical separation - at least 2x2 lanes - emergency lane - lighting only at interchanges and other peculiar points

Arterial road - speed limit 90/100 km/h Trunk road - grade-separated interchanges - U-turns not allowed - physical separation - 2x2 lanes - emergency bays and/or semi hard shoulder - lighting only at interchanges and other peculiar points

Collector - speed limit 80 km/h road - physical carriageway Dual separation carriageway - priority road, 2x2 lanes - closed to (light-) mopeds and bicycles - junctions designed as roundabouts or priority crossroad with traffic lights - limited number of U-turns - limited number of connections to access roads - emergency bays or semi- surfaced shoulder Collector - speed limit 70 km/h road - non-physical driving direction Single separation carriageway - priority road, 1x2 lanes - closed to (light-) mopeds and bicycles - junctions equipped with speed reducing provisions or designed as roundabout - limited number of connections to local roads - emergency bays and semi- surfaced shoulder

36 Guidelines for Road I Safety Engineering|Part I

Road Characteristics Example category Local road - speed limit 40/60 km/h - non-physical driving direction separation - 1x2 lanes - at grade intersections in the form of a roundabout or a three or four-arm crossroads - presence of private accesses

Table 5 – Characteristics of main road categories in urban areas

Road Characteristics Example category Arterial road - speed limit 90/100 km/h Urban - grade-separated Motorway interchanges - U-turns not allowed - physical separation - at least 2x2 lanes - emergency lane - service roads (optional) - lighting

Arterial road - speed limit 50/70 km/h Transit - junctions designed as corridor roundabouts or priority crossroad with traffic lights - U-turns not allowed - physical separation - at least 2x2 lanes - service roads (optional) - lighting

Collector - speed limit 50 km/h street - physical carriageway Dual separation carriageway - priority road, 2x2 lanes - junctions designed as roundabouts or priority crossroad with traffic lights - limited number of U-turns - limited number of connections to access streets - presence of parking areas, preferably outside the carriageway - presence of footpaths 37 Guidelines for Road Safety Engineering|Part I

Road Characteristics Example category - lighting Collector - speed limit 40/50 km/h street - non-physical driving direction Single separation carriageway - 1x2 lanes - junctions designed as priority crossroad with or without traffic lights - presence of parking areas, preferably outside the carriageway - presence of footpaths - lighting Local street - speed limit 30/40 km/h - non-physical driving direction separation - 1x2 lanes - at grade intersections in the form of a three or four-arm crossroads - presence of private accesses - presence of parking slots - presence of traffic calming measures - presence of footpaths (but in case of 30 km/h areas, the space can be shared by pedestrians and motorists) - lighting

Once the characteristics of each road category have been defined, it is particularly important to define the principles for designing the network and, in particular, the interconnection nodes. The matrix below distinguishes between homogeneous nodes connecting roads of the same type and non-homogeneous nodes connecting roads of different types. While in the first (‘homogeneous’ nodes) connections are always allowed that transfer flows from one road to another, in the second (‘non-homogeneous’), for safety and functional reasons, the realization of traffic flow connection may not always be allowed. Therefore, some nodes, where there is a large difference between the hierarchical levels of the confluent roads, should be considered ineligible.

38 Guidelines for Road I Safety Engineering|Part I

Table 6 – Matrix of possible intersection nodes

Arterial Collector Local Homogeneous nodes

1 2 - Permitted nodes

Arterial

Not permitted nodes

2 3 3

Collector Type 1: System interchange (e.g. cloverleaf)

Type 2: Service interchange (e.g. half - 3 3 cloverleaf) Local Type 3: At-grade intersection

Where connection is allowed, it is possible to distinguish different types of node according to whether or not intersection conflict points may occur at the node. In the case of a junction where the roads are all with separate carriageways, intersection conflict points are not allowed and the connection will be solved with an interchange (type 1 junction). Where one of the roads converging on the node has a single carriageway, at-grade manoeuvres may be permitted on that road, while the crossing of the main currents must be solved by separating the levels (type 2 node). Where the two roads considered to belong to types for which the intended cross-section is a single carriageway, the intersection may be solved at-grade (type 3 node). The concepts illustrated are generally present in the most widely used road design standards. In Table 7, as an example, is summarized the mode of operation of the highways according to the Asian Highway Design Standards. As regards Pakistan, the classification currently in use divides roads into very specific categories, even though the modes of operation are not well defined and, above all, not always easily distinguishable on the roads in operation16.

16 NHA classifies the roads in four main categories: 1. Motorways and Expressways are four or six lane divided highways. The access of the Motorways is fully controlled, while the access of Expressways is partially controlled. 2. Primary Roads are basically the National Highways and Provincial Roads on the primary routes. They are further split into three categories: • Primary I (P-I), • Primary II (P-II), and • Primary III (P-III), depending on the number of lanes and pavement of shoulders. 3. Secondary Roads (S-I) are Provincial Roads that serve as feeder roads for the primary routes. 4. Tertiary Roads (T-I) are basically the collector roads for the secondary network. 39 Guidelines for Road Safety Engineering|Part I

Table 7 - Mode of operation for each Asian Highway class17

Class Primary Class I Class II Class III Description Access- 4 or more controlled 2 lanes 2 lanes lanes highways Mode of operation Controlled-access Full No18 No18 No At-grade intersections Not Yes18 Yes18 Yes permitted Overtaking on opposing Not Not Yes Yes lane permitted permitted Pedestrians Not Yes18 Yes18 Yes permitted Slow vehicles19 Not Yes18 Yes18 Yes permitted (source: adapted from UN ESCAP, 2004)

3.4 Speed management 3.4.1 Overview The issue of speed is central to road safety, and its management requires work across several road safety sectors. Effective speed management comprises a series of interventions that can have great benefits for road safety. According to the World Health Organisation (WHO, 2008), addressing speed management policies and programs plays a critical role in improving a country’s road safety record. As seen in section 3.1, the role of speed has been recognized as one of the most important elements of a ‘safe system’ and its relevance to each of the pillars of the United Nations Decade of Action for Road Safety has been noted. According to Organisation for Economic Co-operation and Development (OECD, 2006), tension exists between the perception that speed is a sign of improved efficiency and a tool for progress versus the negative consequences of speeding. Indeed, it is noted that individuals and the community as a whole perceive the benefits and dis-benefits of lower speeds differently. So, for communities, the overall cost of crashes is high, but for the individual the risk may still be quite low. Arguably the environmental benefits (e.g. air pollution and noise) of reducing speeds for individuals (perhaps with the exception of a reduced fuel bill) are less apparent than for society at the aggregate level.

17 Asian Highway Design Standards do not apply to built-up areas 18 Unless specially designed for controlled-access operation 19 Bicycles, low power motorcycles, agricultural vehicles, mopeds, animal-drawn carts, animal herds 40 Guidelines for Road I Safety Engineering|Part I

Speeding has been widely recognized to fall into one of two categories: • Excessive speed: drivers exceeding the posted speed limit • Inappropriate speed: drivers choosing a speed that is not safe for the given conditions Both types of speeding can potentially raise the likelihood of a crash occurring through increasing the stopping distance by increases to: • The distance the vehicle travels between the initial perception of an event requiring the vehicle to stop and the actual motor action taken to stop or slow the vehicle. • The stopping distance of the vehicle at a given speed. Figure 28 illustrates the stopping distance required at various speeds.

Figure 28 - Braking distances (source: Australian Transport Safety Bureau, as cited by WHO, 2008)

Speed also affects the severity of crashes. Higher speed crashes involve more kinetic energy (kinetic energy is proportional to the speed squared) and the more energy that is dispersed in a crash, the more severe it tends to be. It is the scale of this energy exchange that determines the severity of injury. The likelihood of being involved in a serious or fatal crash increases significantly with even small increases in vehicle speed as shown in Figure 29. Reading across the x-axis (bottom) of this graph, it can be seen that an increase in mean speeds of 5% leads to an increase in injury crashes of 10% (black dotted line), and a 20% increase in fatal crashes (black continuous line). Similarly, if mean speeds were to decrease by 5%, then a reduction in all injury crashes of 10% and a reduction in fatal crashes of 20% should be expected.

41 Guidelines for Road Safety Engineering|Part I

Figure 29 – Change in number of injured on change in mean speed (source: WHO, 2008)

Speed differential (differences in speeds travelled by different vehicles) is also crucial factor influencing crash rates and outcomes. This is particularly the case in urban areas; however, it is also an issue on high-speed roads. Speed differential is strongly linked to fatality rates on rural roads and urban arterials (OECD, 2006). Research on urban roads indicates that crash rates increase as the proportion of drivers who exceed the speed limit increases. Research in Australia (Kloeden et al., as cited in OECD, 2006) shows that faster drivers have a higher crash risk as shown in Figure 30 (as average speed driven by a driver increases along the x-axis, relative crash risk increases particularly strongly for urban roads). The same Figure indicates that slower drivers do not have a higher crash risk.

Figure 30 - Relative injury accident rate on urban roads and rural roads for vehicles going faster and slower than average speed (source: Kloden et al., as cited in OECD, 2006)

42 Guidelines for Road I Safety Engineering|Part I

Increasing the speed at which activities can be undertaken is generally taken as a benefit to society. In transportation, speed is seen as a key element as it reduces journey time and, as a consequence, it can reduce business costs and allow people to enjoy activities before or after the journey for longer. Smooth and quick journeys are often seen as an indication of an advanced and efficient transportation system. However, there are also significant dis-benefits associated with speed. The severity and frequency of crashes are closely linked to increased speed (see section 3.2.2). This is particularly the case in urban areas. There are a number of factors that drivers will take into account when choosing the speed to travel, as indicted in Figure 31, and the posted speed limit is only one of them. In accordance with the multi-faceted nature of the problem, effective treatment of speed also needs to be multi-faceted in order to adequately address the complex reasons behind speeding.

Driver factors

Education Vehicles promotion factors

Driver Speed Road factors zone/limit speed choice

Enforcement Traffic sanctions conditions

Crash and injury risk

Figure 31 – Factors affecting speed choice (source: Oxley and Corben, as cited in WHO, 2008)

Even though some research has shown that speed limits have a positive influence on actual speeds, however, it should be stressed that other research has indicated that changing the speed limit alone has little effect. In places where speed limits have been changed and no other action taken, the change in average speed is only about one quarter of the change of the speed limit (DETR, 2000). Any changes in speed limits should ideally be accompanied by appropriate enforcement, engineering and educational measures. International best practice suggests that the best results are likely to

43 Guidelines for Road Safety Engineering|Part I

be achieved when engineering, educational and enforcement interventions are implemented to compliment and reinforce speed limit adherence. In particular engineering refers to the design of the road and any physical interventions which have a direct impact on driver speed choice. Interventions can range from the manipulation of the geometry of the road (curve radius, incline and sight distance), to fixed interventions such as road humps. Engineering measures can also be used to prevent crashes or reduce the severity of crashes that do occur. In the previous sections has already been described as a road can be self-explaining (i.e. engineering measures can enhance the degree to which speed limits are credible and understood by drivers and riders; see section 0) or forgiving (i.e. roads can be designed such that when a crash happens, crash forces are effectively dissipated between the vehicle and the road; see section 3.2.4). Roads can be also self-enforcing, i.e. at particular locations physical measures (e.g. speed humps or chicanes) can ensure that speeds are low (i.e. ‘traffic calming’ measures). These locations include areas where there is a high concentration of vulnerable road users, for example in residential areas and outside schools or busy shopping areas. Speed management is a complex issue and, as seen in Figure 31, there is no single solution to the problem of excessive and inappropriate speeds and a package of countermeasures is necessary. The countermeasures should complement one another and increase the overall effectiveness of, and compliance with, the speed limit. An effective speed management policy targets both inappropriate and excess speed using engineering, enforcement and educational measures to reduce speeds that will ultimately result in fewer speed-related crashes and reduced severity of crashes that do occur. Speed management requires therefore a systematic approach incorporating all of the following elements: • Setting and signing speed limits: Speed limits need to be appropriate for the road to which they apply and should reflect the road function, traffic composition, frontage development and road design characteristics. The driver should always know what the speed limit is. The conventional way to achieve this is to use traffic signs and road markings. • Road engineering measures: The road infrastructure can be designed such that roads are self-enforcing (at certain locations road designs such as speed humps may be used to make it difficult for motorists to travel at higher speeds than desired) and self-explaining (road designs are intuitive and clear such that road users understand the speed limit and the speed limit is credible). In addition, roads should be also forgiving (the road and vehicle in combination protect the road user from serious or fatal injury at legally permitted speeds), • Education: The provision of information and education for drivers is also a very important activity. If drivers understand the importance of speed limits, it is more likely that they will comply with them. • Police enforcement: Police enforcement is necessary to deter intentional speed violations. 44 Guidelines for Road I Safety Engineering|Part I

This document focuses on speed limits and engineering measures, a brief overview of which is presented in the next paragraphs. 3.4.2 Setting speed limits Speed limit setting has traditionally reflected attempts to achieve a balance between safety and mobility. However, countries that recognize their poor safety record and are committed to reducing road deaths and injury are shifting this balance in favour of safety. According to the Safe System approach principles (see section 3.1), some countries are now setting speed limits with reference to the limits of human injury tolerance, that is, to a level that will not usually result in death or serious injury to road users when crashes occur. Guidelines for setting limits according to Safe System principles have been already presented in section 3.2.2. It is recommended to consider the ‘safe speeds’ reported in Table 3 as ‘default’ speed limits or, better, as ‘national speed limit regime’. Nevertheless, they should be adapted to local situation according to factors described in the Box 1.

Box 1. Factors to consider when setting speed limits (source: WHO, 2008)

• Traffic mix and the different types of vulnerable road user. • Crash history, severity (injury) and crash rate (per vehicle kilometre of travel) where possible. Road alignment (both vertically and horizontally). Crash prone stretches of road should have lower limits. • Road shoulder width and pavement quality – narrow shoulder widths (especially those with poor pavement quality) can run an increased risk of ‘loss of control’ crashes. Therefore, speed limits should be lower for these conditions. • Road delineation – edge and centre-line marking, reflectors and guideposts on the edge of shoulders and advisory speed limits. Where roads have poor visual definition, the speed limits should be lower to enable time for driver judgements. • Road and lane widths should be adequate (i.e. at least two lanes with a minimum lane width of 3.4 metres). Narrow lane widths offer little margin of error and therefore speed limits must not exceed that required by drivers to keep consistently within a lane. • The intensity of land development abutting a carriageway – in built-up areas, there is a dual risk of poor visibility and more varied activity of people and vehicles entering the road environment, and therefore speed limits should be lower. • The type of intersections and the nature of traffic control measures at intersections. While all types of intersection present increased risk to road users – and roads other than motorways should have lower limits – poorly marked intersections require even lower speeds leading up to them than other, more clearly marked intersections or roundabouts. • Traffic volume and traffic flow – lower speed limits in areas of high traffic volume can be used to smooth traffic flows, making for better network efficiency and environmental benefits, as well as improved safety. • Types and standards of vehicles allowed to access – roadways that vulnerable road users such as cyclists are allowed to use should have lower limits than those that only allow four- wheeled (or above) motor vehicles. • The free travel speed of the road. • The ability to overtake safely (within sight distance) at the posted speed.

45 Guidelines for Road Safety Engineering|Part I

Once the appropriate speed limit has been determined for a road or a section of road, steps must be taken to ensure drivers adopt the appropriate speed. The ‘default’ speed limits, in many countries, are not signposted. Nevertheless, in countries like Pakistan that need to tackle the problem with determination, it is advisable that they would be clear to existing and new drivers (including visitors) entering the road network or a specific road (e.g. entering a motorway). In specific cases alternative (to default) speed limits can be used. These limits may include: • linear speed limits (including transition/buffer speed limits) i.e. along lengths of roads and streets • area-wide residential or commercial speed limits, with signs at entry point to the designated area • time based speed zones: o school speed zone, usually twice daily time-based lower limits for an hour or so at school starting and finishing times o seasonal speed zone, for example at holiday resorts in busier summer months when vehicular and pedestrian traffic is greater • variable speed limits (limits that change under certain conditions or times of day, e.g. in wet conditions) • heavy vehicle speed limits It is recommended, however, not to continuously change the limit as this can lead to unexpected and inconsistent behaviour. If on a road section it is necessary to reduce the speed in many locations, it is advisable to adopt a lower limit than the default one on the entire stretch. As regards the lower limits for heavy vehicles, it is worth to remind, that the use of different speed limits for different vehicle categories on a section of road, could create the opportunity for substantial turbulence within traffic and may increase the frequency of overtaking manoeuvres, which can in themselves lead to increased crash risk. If there is to be a lower limit, it is suggested that this is a consistent amount below general limits, whether default or signed, on all rural roads. Speed differential is a major cause of crash risk on higher speed roads (WHO, 2008; see also 3.2.2). Signs are the primary way of communicating the speed limit of a road to drivers. Careful consideration needs to be given to the use and placement of signs to ensure that all drivers on a given section of road are given consistent information. The recommended intervals are as follows: • urban areas: 400 m • rural areas: o motorways and other arterial roads: 4/5 km o other roads: 2/3 km In addition to post-mounted speed limit signs next to the road, markings on the road can also be used to show the current speed limit.

46 Guidelines for Road I Safety Engineering|Part I

Figure 32 – Speed limits marked on the road pavement in Spain 3.4.3 Road engineering measures to reduce speed There is a large range of engineering treatments that have been shown to be of varying use in speed management. These measures are described in significant detail in various manuals; however, a broad overview of available treatments is given below. Practical examples will be also duly described in Part II of these Guidelines. Treatments include engineering or re-engineering the road to encourage lower speeds, or make the road and its environment ‘self-enforcing’ or ‘self- explaining’. They generally relate to so-called ‘traffic calming’ measures, which include a range of physical features that have been developed by road safety and traffic management engineers to encourage, or force, drivers to drive more slowly. Many of these treatments have the effect of making it feel uncomfortable to drive in excess of the legal or recommended speed. Measures can be grouped in three main methods: a) Narrowing b) Horizontal deflection c) Vertical deflection Of course, they can be combined to address complex situations where the individual measure would be ineffective. A typical example is the treatment at the entrance of settled areas where a ‘gateway’20 is generally accompanied by other measures aimed to alert the drivers and force them to adapt the speed.

20 Gateways are devices used to mark a threshold – usually to a village or higher risk location on the road – where lower speeds are required from drivers. Gateways rely on highly visible vertical treatments to capture driver/rider attention and usually include signs, pavement markings, speed limits, but also architectural and rural treatments such as picket fencing or gates, earth mounds and rock walls. 47 Guidelines for Road Safety Engineering|Part I

3.4.3.1 Narrowing Wider roads invite drivers to select higher travel speeds. This may be because the perceived margin for error is greater. So, narrower road widths tend to slow traffic speeds. Actually, road narrowing cannot be considered as a speed- reducing device in itself, but it can act as a reminder or encouragement to drive slowly or calmly. It uses a psycho-perceptive sense of enclosure to discourage speeding. Figure 33 – Road narrowing in an urban street in the USA Where traffic capacity is not a problem, the cross section of the road should change where the built-up area begins and the width of lanes can be reduced. In many cases, it is possible to change a four-lane road into a two- lane road through the built-up area. The new layout can be obtained adding a raised island located along the centreline of a road that narrows the travel lanes. Fitted with a gap to allow pedestrians to walk through at a crosswalk, they are also very useful as ‘pedestrian refuges’. If it is not advisable to reduce the capacity, a simple ‘optical’ narrowing can be effective too. In this case the cross-section remains unchanged, whereas the shoulder is changed (e.g. building a sidewalk with kerbstone, planting trees or bushes, etc.). 3.4.3.2 Horizontal deflection Horizontal deflection measures use forces of lateral acceleration to discourage speeding. Generally, all horizontal shifts may be classified as chicanes, more or less pronounced. In case of presence of a central island, they mainly differ because of its width. A chicane consists of a displacement of the road axis with a significant deflection of the trajectory. Figure 34 – Chicane with a central island in France This is a traffic calming measure used to emphasize the transition between the rural environment and the built-

48 Guidelines for Road I Safety Engineering|Part I

up area. It is generally located after the sign of beginning of settlement and strengthens the image of the entrance into the settlement. Depending on the site configuration, two types of chicanes are possible: • Chicane with a central island: • Chicane without a central island (simple or double): suitable in case the speed is already substantially lowered before the chicane The extreme case is represented by a roundabout that, of course, accomplishes also other functions. Roundabouts are effective in reducing the severity of crashes at an intersection because they require traffic to deviate from a straight path and therefore slow down to undertake the manoeuvre. The reduced speeds of travel through an intersection that a roundabout can achieve, together with the non-right-angle nature of side-impact crashes because of the geometry of the roundabout, result in reduced crash severity. 3.4.3.3 Vertical deflection Much can be achieved by road design and the simple physical countermeasures described above. However, some drivers will still drive very fast despite the signs. The hazards these drivers create can only be tackled by strong physical measures such as: • speed humps • speed tables • rumble strips Figure 35 – Speed hump in France These vertical measures use forces of vertical acceleration to discourage speeding. In general, these measures, except for rumble strips, aim to reduce the speed to 30 km/h. Therefore, for safety reasons, CERTU (CERTU, 1994) recommends implementing these measures only if the approaching speed of the 85th percentile of users is less than 60 km/h. If this value is exceeded, other preliminary measures (i.e. narrowing, horizontal deflection) must be implemented in the previous road section. These measures should not be stand-alone, but they should be implemented in series. 3.4.4 Separation of vulnerable road users Speed should be limited to ensure that vulnerable road users are not exposed to risk of serious injury. If this is not possible, separating the vulnerable road users from motorized traffic is an alternative.

49 Guidelines for Road Safety Engineering|Part I

Pedestrian fencing is useful for improving the safety of pedestrians by directing larger flows of pedestrians away from random crossing locations (particularly in busy pedestrian crossing locations) to safer crossing points, which may be equipped with treatments such as speed humps or raised platforms in the roadway, or a set of traffic signals. Refuge islands and medians can assist pedestrians in crossing the road by allowing a staged crossing and simplifying decision-making. Kerb extensions can also improve pedestrian safety by reducing the crossing distance, and the area and time in which the pedestrian is at risk. This is particularly helpful for older or disabled pedestrians who may have difficulty choosing a safe gap in traffic at a conventional crossing point. In many situations in rural (and urban) areas there will not be any footpath provision for the large numbers of pedestrians walking from point to point. They will often be forced to walk on the carriageway. Provision of a walking path is a highly effective means of removing the pedestrians from a medium to high- speed carriageway. Bicycles should be separated from motorized traffic as well, if at all possible.

50 Guidelines for Road I Safety Engineering|Part I

51 Guidelines for Road Safety Engineering|Part I

4 Road infrastructure safety management

Safer roads are the Pillar 2 of the UN Decade of Action for Road Safety 2011- 2020 (WHO, 2010) and are one of the aspects of the Safe System approach as discussed in section 3.1. As seen at length in Chapter 3, safe roads are roads that are self-explaining and forgiving of mistakes to reduce the risk of a crash occurring and to protect road users from fatal or serious injury should a crash occur. This requires roads and roadsides to be designed, built, and maintained to reduce the risk and severity of crashes. To achieve this objective, a comprehensive strategy for road infrastructure safety management needs to be put in place. Without it, in fact, there is a real risk of implementing disconnected interventions in a framework of high inefficiency and consequent possible waste of resources. The strategy, which is based on the Safe System principles, is outlined in the Figure 36 and is described in detail in the following paragraphs. A very good benchmark in this field is the European Directive 2008/96/EC of the European Parliament and of the Council of 19 November 2008 on road infrastructure safety management. This Directive mandates all EU member States to establish and implement a well-defined set of procedures relating to abovementioned strategy. This Directive will be referred to several times in the following paragraphs.

Construction • RSIA at feasibility • Post-opening RSA stage • RSI (proactive) • Pre-opening RSA • RSA at preliminary • Data analysis and and detailed design treatement of crash stages locations (reactive) Design Operation

Figure 36 - Road safety management approaches throughout the project life-cycle (adapted from AfDB, 2014)

52 Guidelines for Road I Safety Engineering|Part I

4.1 Road safety strategies The treatment of crash locations and correcting safety problems across the road network which may result in road trauma by undertaking Road Safety Audits (RSA) and Inspections (RSI) form an integral part of a Safe System. These methods belong to two main road safety strategies: • Proactive approach, assessing a design or concept before it is built (and thus before crashes happen), or the safety of an existing road once built based on safety features present to identify any treatment to reduce the likelihood or severity of a crash. • Reactive approach, responding to an existing crash problem. Effective road safety management programs should exercise an optimal balance between reactive and proactive strategies. 4.1.1 Proactive approach A proactive approach focuses on the evolving ‘science of safety’, that is, what is known about the evolving specific safety implications of road design and operations decisions. The proactive approach applies this knowledge to the roadway design process or to the implementation of improvement plans on existing roads to diminish the potential of crashes occurring prior to the road being built or reconstructed. Conducting RSAs is an example of a proactive road safety strategy, but also RSIs can be conducted following this approach. The advantages of a proactive approach include: • Crash prevention: it is not necessary for crashes to occur before crash prevention measures are taken (‘preventing is better than curing’). • Lower costs: changing plans is easier and less costly than to implement an improvement plan on a road open to the public. 4.1.2 Reactive approach A reactive approach to road safety is based on the analysis of existing crash data. Road safety improvements proposed are considered in reaction to identified safety problems brought to light by crashes that have occurred after the road has been designed, built, and opened to the traveling public. Traditional reactive road safety engineering processes include such activities as information collection and management (crash information systems), identification of problem locations on the road network, analysis, development and implementation of countermeasures. A ‘blackspot programme’ is an example of reactive approach to crash frequency and/or severity reduction. Limitations of the reactive approach are as follows: • It requires the identification of high crash locations before improvement plans may be developed and implemented.

53 Guidelines for Road Safety Engineering|Part I

• The supporting crash data is often dated, incomplete and/or insufficient to support accurate diagnosis and intervention. • It may also be costlier, since improvement plans are necessarily implemented on a road already built and open to public. Despite these limitations, no road safety management system can be considered complete without a reactive component as it is a powerful tool for addressing existing safety problems.

4.2 Road safety impact assessment According to the European Directive 2008/96/EC ‘road safety impact assessment’ (RSIA) means a ‘strategic comparative analysis of the impact of a new road or a substantial modification to the existing network on the safety performance of the road network’. The RSIA should be carried out at the initial planning stage before the infrastructure project is approved. It should indicate the road safety considerations which contribute to the choice of the proposed solution. It should further provide all relevant information necessary for a cost-benefit analysis of the different options assessed. Any new infrastructure project realignment or change to existing infrastructure that substantially affects the performance of the national road network should be assessed. RSIA is required only where the anticipated effect on the main road network is substantial. Small projects generally do not require assessment. RSIA should be carried out at the initial planning stages of a project and should be used as one of the tools for project selection. This assessment should consider the safety implications of the different alternatives as well as the option to not proceed with the project. As the project design progresses the RSIA should be regularly reviewed to ensure that the road safety implications of all design revisions are considered. RSIA is an integral part of the design process and can be carried out within the design team. The assessment team, generally comprising at least two individuals, should include at least one experienced road design engineer and at least one experienced road safety engineer/auditor. In the absence of competence in RSIA within the design team, an assessment team should be sourced from elsewhere and should join the design team for this specific task. It is important to note that a RSIA is not a separate audit of the project carried out by an independent team. It is an on-going task within the design process and generally carried out within the design team. If an external assessment team is brought in to provide road safety expertise, then that team shall be viewed as temporarily part of the design team. In this view RSIA does not replace or preclude RSA, which is carried out by a team independent of the design process. RSA is described in the following paragraph.

54 Guidelines for Road I Safety Engineering|Part I

The RSIA should be carried out while the project is still at concept stage. At this stage the assessment explores the road safety implications of each option being considered, including the Do-Nothing and Do-Minimum options. The assessment should provide all relevant information necessary for comparison of the options and selection of the solution, including a comparative analysis of the road safety implications of each alternative considered and an evaluation of the road safety benefits and dis-benefits arising from each alternative. The objective of RSIA is to consider the proposed project from a road safety point of view, to compare the impact on road safety of each proposed option and to determine which would give the best road safety outcome. With every project there is the possibility that the existing situation would be preferable to any of the options considered, and so it is essential that this alternative is also considered in the assessment. Road safety impact is only one of the aspects considered by a design team when selecting the preferred option. It is important that the reasoning behind the conclusions of the impact assessment is made clear, so that it is given due weight in the selection process. This should minimise the risk of collisions occurring in the future either as a result of planning decisions or as a result of unintended effects of the design of road schemes. A methodology for RSIA is set out in the Irish guidelines (NRA, 2016) and is reported in Box 2.

Box 2. RSIA methodology (adapted from NRA, 2016)

i. Define the project and its objectives: • Clarify the objectives of the project (e.g. to increase capacity, to remove traffic from a village, to eliminate poor alignment, to provide an amenity, etc.). • Clarify whether the major objective of the scheme is to address road safety issues. ii. Define the study area and the area of influence of the project: • Clarify the extents of the surrounding road network where any of the proposed options would affect the operation of the network. Check the likely changes to drivers’ route choice and choice of travel mode or time, and thus the likely effects on traffic patterns. • The entire study area shall be examined when assessing each proposed option, so that like can be compared with like. iii. Establish the existing road safety problems: • Examine existing collision statistics and carry out an analysis. • Establish any patterns in the collisions and any high collision locations, either stretches of road or single sites at junctions or other conflict points. • Establish any patterns over time of day or year, or any patterns involving road user type. Examine any road safety reviews that may have been carried out previously on all or part of the area. • Collision statistics for national roads are available from NH&MP. For provincial projects, it is recommended to contact the local authorities (e.g. Police, Communication & Works Departments, etc.) who might have information on

55 Guidelines for Road Safety Engineering|Part I

crashes that they have attended. iv. Road safety objectives: • Define the road safety objectives of the scheme. • This will usually include addressing the existing road safety problems, but there may be further objectives, such as improving pedestrian access to an amenity or improving public transport access. v. The options, including Do-Nothing and Do-Minimum: • Examine the drawings of each proposed option for the project. • Include the existing Do-Nothing situation which would prevail if no works at all were to be implemented. • Include the Do-Minimum situation, where the very minimum possible is to be implemented, such as provision of signs, surface overlay and any committed schemes. • Visit the site to visually establish the alignment of each proposed option and the surrounding topography. A site visit is important as it may identify existing arrangements or patterns of use that may not be evident in the drawings and other information examined. • Examine both existing and proposed traffic flows. It may be necessary to establish peak times of use for certain parts of the network, such as access to schools or sports grounds or weekly markets, so that the appropriate flows can be examined. • Patterns of use of all road users must be considered. In general pedestrians and other vulnerable road users are affected more acutely than other road traffic by both changes in road alignment and changes to available routes of travel. vi. Analysis of impacts on road safety of the proposed alternatives: • The main element of the assessment is the comparison of the road safety effects of each alternative proposal. This must include Do-Nothing and Do-Minimum options. • The effects on the entire study area must be examined for each proposed option. Where proposed alternatives differ in scale and cover differing lengths or areas of the existing network, the remainder of the road network outside the proposed works must be included in the analysis. The assessment area must be the same for all options being compared. • An assessment of the effects of each alternative must be carried out in terms of predicted collisions. Quantitative indicators can be used such as collision rates and collisions per junction type. • To assess the likely collision occurrence in the proposed options, it is recommended to use established local collision rates in the surrounding area for equivalent road types. If these are unavailable, then the collision rates for road types at national level should be used. • To establish the economic collision cost of each option, in absence of a Pakistani assessment, it is suggested to refer to average values based on local GDP21. • All effects on traffic flow and traffic patterns must be considered. Any projected change in modal split as a consequence of the proposals is important as this may not only affect the mix of vehicle category within the traffic flow, but may also impact on patterns of pedestrian and cycle travel and locations where conflicts

21 According to iRAP the value of a fatality is about 70*GDP/Capita, whereas the value of a serious injury is 17*GDP/Capita (McMahon & Dahdah, 2008) 56 Guidelines for Road I Safety Engineering|Part I

with other vehicles occur. • Seasonal and climatic conditions such as the likelihood of flooding and foggy conditions should be considered, as this might differ between options. • The possibility of seismic activity should also be considered. vii. Comparison and ranking: • Comparison of the alternatives should not only give a qualitative list of benefits and dis-benefits, but should also include a cost benefit analysis of the road safety aspects. • The options, including the Do-Nothing and Do-Minimum option, should be ranked in terms of road safety considerations, giving an order of preference and an indication of the magnitude of difference between options. If one option, or a group of options, shows considerably more or less benefit than the others then this should be highlighted. Conversely, if there is little difference in road safety terms between two or more of the proposals then these should be given the same ranking.

4.3 Road safety audits The RSA process was initiated when road safety engineers realised that they were carrying out collision remedial schemes on relatively new roads. Adopting the principle of ‘preventing is better than curing’, they decided to use some of the safety experience that they had gained from the remedial work to design safety into new road schemes. Since then the concept grew over the years from an informal check of new schemes to the current system of RSA as an essential integral part of design and construction procedures. According to the European Directive 2008/96/EC ‘road safety audit’ means an ‘independent detailed systematic and technical safety check relating to the design characteristics of a road infrastructure project and covering all stages from planning to early operation’. Accordingly, RSA should be carried out for all road infrastructure projects and should form an integral part of their design process at least at following stages: • Preliminary design • Detailed design • Pre-opening • Early operation Practical instructions on how to carry out and manage an RSA can be found on the CAREC RSA manual (ADB, 2018a). It is here important to highlight that an RSA should be carried out by an auditor appointed in accordance with the following provisions: • She/he should have undergone an initial training resulting in the award of a certificate of competence • She/he should have relevant experience or training in road design, road safety engineering and crash analysis • She/he should not at the time of the audit be involved in the conception or operation of the relevant infrastructure project

57 Guidelines for Road Safety Engineering|Part I

It is therefore quite obvious that auditor approval is a crucial task for any road authority. As stated in the CAREC manual, it is suggested that every country establishes its own national register of auditors. In Pakistan, besides a national register, which ideally should be administered by NHA, different registers can be established in each province or territory. In any case it is of paramount importance to establish clear requirements to approve applicants. Firstly, it is strongly recommended that application would be limited only to civil engineers (or equivalent) that have at least experience in road design and/or traffic planning. In addition, as recommended by the CAREC RSA manual, it is recommended that an applicant, in order to be registered as an RSA team leader (i.e. senior auditor), should: • have completed an approved road safety audit training course, • have a minimum of 3 years’ practical experience in a road or road safety-related field, and • have completed at least five road safety audits under the guidance of a senior road safety auditor, of which at least three of the five audits must be at a design stage. Finally, to be registered as a team member, an applicant should: • have completed an approved road safety audit training course, and • have a minimum of 2 years’ practical experience in a road- or road safety-related field According to this scheme is very important to promote the involvement of local engineers in auditing. In the first period (2/3 years) the audits will be necessarily leaded by international auditors, but only following this approach the local practitioners can gain sufficient experience to lead an audit in the future. The RSA training courses should be officially recognised by national authorities (e.g. NHA, MoC, etc.), should be at least of 5 days’ duration, presented by an experienced road safety auditor, and should include a final exam. Only applicants passing successfully the exam should be allowed to be registered. The course should contain a program of presentations that should be not limited to the audit process but should describe in detail the whole road safety engineering process. Technical topics may include safety in geometric design, vulnerable road users, signs, delineation, safety at road works, and roadside hazard management.

58 Guidelines for Road I Safety Engineering|Part I

Box 3. Case study: RSA&I training course at NUST University

A 7-day Road Safety Auditing & Inspection training course was held in the NUST University (Islamabad) in February 2018. The course was approved and funded by the ADB in order to support the Government of Pakistan to: • build capacity to identify, treat and eliminate high crash and hazardous locations as identified in the National Highway Authority Road Safety Action Plan 2018-2020, and • establish a team of informed and skilled focal persons to coordinate dissemination and piloting road safety engineering best practices. The course was designed to provide road engineers with a theoretical and practical knowledge of main road safety engineering tools and specifically to build the knowledge and skills required to conduct road safety audits and inspections. It referenced the strategic context for safe roads within the draft National Road Safety Strategy 2018-2030, with particular reference to Pillar 2 – Safe Roads. Specific objectives were as follows: • To train and qualify a first kernel of road safety auditors • To train road authorities officials on road safety engineering practices to identify and treat high crash cluster locations • To train road authorities officials on preventive road safety inspections • To share international best practices in safe road design • To discuss main issues concerning road design standards and their current implementation in Pakistan The course included theory sessions, best practices and case studies taking into account the local practice, small group sessions, participant workshops, site visits, submission of a Road Safety Inspection Report and a final examination. The training team – contracted by the Danish firm NTU - included both international and national experts, covering various aspects of road safety (engineering, road crash investigation, crash data analysis, institutional aspects). The course was articulated over: • 9 days (including a weekend) • 10 modules (both theoretical and practical) • workshops • homework • final examination The Training Course program was as follows: • Day 1: o Module 1: Road safety national and international framework o Module 2: Crash investigation • Day 2: o Module 3: Road safety engineering process • Day 3: o Module 4: Conflict studies o Module 5: The management system of crash data • Day 4: o Module 6: Road safety audits and inspections

59 Guidelines for Road Safety Engineering|Part I

• Day 5: o Module 7: Site visit • Days 6/7: o Self-directed preparation of the Road Safety Inspection Report for examination assessment and preparation for the course examination • Day 8: o Module 9: The safe road design • Day 9: o Module 10: Principles of traffic engineering o Final examination A total of 26 participants completed the course, including representatives from National Highway Authority (NHA), Capital Development Authority (CDA), Pakhtunkhwa Highways Authority (PKHA), Communication and Works Departments of Governments of KP and Balochistan, Engineering Consultancy Services Punjab (ECSP), NUST, University of Karachi. Only 4 participants did not pass the final exam.

Group photo

Workshops on site (left) and in the classroom (right)

4.4 Road safety inspections RSI should be undertaken in respect of the roads in operation in order to identify the road safety related features and prevent crashes. According to the European Directive 2008/96/EC ‘road safety inspection’ means an ‘ordinary periodical verification of the characteristics and defects that require maintenance work for reasons of safety’.

60 Guidelines for Road I Safety Engineering|Part I

The terminology of this definition gives an indication of the scope of the RSI as follows: • The term ordinary indicates that an in-depth, forensic investigation is not expected. • The measures to be carried out in response to the inspection are described as maintenance work; this suggests that major changes to the layout of the road, entailing high cost, are not envisaged as counter- measures; however, engineering works are often required to remediate the issues. • The term periodical indicates the need for inspections to be repeated at intervals, rather than being a once-off event. Their frequency should be sufficiently frequent to safeguard adequate safety levels for the road infrastructure in question22. RSI is an on-going process, with roads being subject to re-inspection at regular intervals. The iRAP programme can be considered as an evolution of the RSI as described here. This methodology turns the road features detected during the inspection in quantitative attributes that then, using a quite complex algorithm, provide a synthetic judgment about road safety. In this case, the inspection is therefore not limited to the identification of road safety issues, but it is aimed to identify all road characteristics that, to a different extent, influence the likelihood of a crash and its severity. More details about iRAP and the techniques developed under this programme are reported in the Box 4. RSI responds to the safety implications of changing conditions on the road network. The road environment is dynamic; it is not fixed over its design life. Roadside features are added or removed, materials forming the road deteriorate and are replaced, new developments are built on the road frontage altering access conditions and changing traffic flows. Changes also occur to our understanding of road safety and road design standards; certain engineering designs that would have been considered safe in the past are today considered unacceptable. In respect of the timing, the inspection should be carried out: • By day, in both directions. An inspection by night is advisable if collision records show an unexpected share of crashes during the night. • At times of normal operation of the road. Unless otherwise required, avoid times when the road environment conditions are abnormal, such as when special events are occurring. However, if events are frequent (occurring at least weekly), and if the conditions during those events are considered to affect road safety, then the route should also be driven

22 Irish guidelines (NRA, 2014) suggest the following maximum periods between inspections: • 5 years for motorways and other dual-carriageway arterial roads • 3 years for other roads 61 Guidelines for Road Safety Engineering|Part I

under those conditions. School-times and commuter congestion are examples of factors which may need consideration if they apply to the route and are significant in safety terms. Off-peak conditions should however always be considered. The RSI should be conducted by a team of at least two experienced practitioners. The benefits of a team approach are that reports are likely more balanced, and the likelihood to miss some issues are is lower. An additional consideration is that it is often not practical to carry out an RSI alone. It is recommended that at least one team member (to be appointed as Team Leader) would be an approved road safety auditor according to the scheme described in the previous paragraph. In addition, it is strongly recommended that all team members would be independent of the maintenance and operation of the road. The inspection is intended to be a fresh, independent look at the road and therefore it is not recommended that the inspector have had a role in the design or maintenance of the route, within the three-year period prior to the inspection. It does not mean that the inspector cannot be an employee of the road authority or road operator. The inspection is carried out with a vehicle and, given its ‘ordinary’ nature, it is not needed to stop for in-depth investigations. It is strongly recommended to equip the vehicle with a high-resolution video- camera provided with a GPS device. It allows in fact an easy positioning of the identified issues on a map (see Figure 37). The camera must be tight installed on the windscreen through a suction cup or similar (the camera cannot be handled).

Figure 37 – Screenshot of a software to manage videos from a camera provided with GPS (on the top right it is visible the tracking of the travelled route)

62 Guidelines for Road I Safety Engineering|Part I

Box 4. iRAP methodology

About iRAP iRAP was formed in England in 2006 and was granted charity status in 2011. It is an umbrella organization for road assessment programmes worldwide (e.g. EuroRAP in Europe, AusRAP in Australia, usRAP in the US, ChinaRAP in China, PakRAP – soon – in Pakistan) and facilitates the development of road assessment work in low- and middle-income countries. iRAP assumed that reliable crash data would not be available in these countries and developed new ‘proactive’ techniques to overcome these shortcomings. Today iRAP works in partnership with government and non-government organisation to (i) inspect high-risk roads and develop targeted road safety plans, (ii) build local capability providing training, technology and support, and (iii) track road safety performance so that funding agencies can assess the benefits of their investments. Road inspection Using vehicles equipped with one or more cameras, inspections focus on more than 50 different attributes that are known to influence the likelihood of a crash and its severity. These attributes include intersection design, the number of lanes and markings, roadside hazards, footpaths and bicycle facilities.

Example of road attributes

Star Ratings Star Ratings are based on road inspection data and provide a simple and objective measure of the level of safety which is ‘built-in’ to the road for vehicle occupants, motorcyclists, bicyclists and pedestrians. Five-star roads are the safest while one-star roads are the least safe. According to iRAP, every extra star rating results in a halving of crash cost on terms of the number of people

63 Guidelines for Road Safety Engineering|Part I

who are killed and seriously injured. Star Ratings can be completed without reference to detailed crash data (proactive approach). The process offers an opportunity for road owners to set a performance-based star rating target for all road users23. PakistanRAP To help reduce the risk of deaths and serious injuries in Pakistan, the ADB engaged iRAP to assist the NHA to develop the Pakistan Road Assessment Program (PakistanRAP). The initiative includes developing Star Rating and Safer Roads Investment Plans for some 14,000 km of national highways and building local capacity to perform assessments and make use of the data produced to guide policy and investment. To accelerate knowledge exchange and capacity building, the ChinaRAP team, which has extensive experience in implementing the Highway Safety to Cherish Life project in China, is playing a key role in the initiative. A pilot phase of the initiative involved assessment of National Highway 5 (N-5) which is approximately 3,500 carriageway kilometres in length and introductory training for NHA staff. The following phases involve assessment of a further some 10,000 carriageway kilometres of national highways and more in-depth training for local engineers.

Star rating results of N-5 highway

23 On 21 November 2017 WHO Member States agreed 12 road safety performance targets. Among them, Target 3 states: ‘By 2030, all new roads achieve technical standards for all road users that take into account road safety, or meet a three star rating or better’ 64 Guidelines for Road I Safety Engineering|Part I

4.5 Treatment of crash locations The treatment of crash locations involves a step-by-step process starting with the identification and ranking of sections of the network in operation which have a high collision concentration. This analysis is therefore based on crash data (i.e. reactive approach). In other words, it means to analyse and rank sections of the road network which have been in operation for more than three years24 - and upon which a large number of crashes have occurred - in order to recognise the causes, select possible countermeasures and identify those having a potential for improvement. The treatment of crash locations is therefore a multi-stage process that follows the steps listed below: a) Initial desktop study b) Detailed desktop study c) Site visit d) Diagnosing the crash problem e) Treating the crash problems f) Prioritising the schemes on the network g) Monitoring and evaluation 4.5.1 Initial desktop study The initial desktop study is based on the crash analysis. The availability of comprehensive and accurate data about crashes is therefore crucial, because the whole process of investigating, analysing and effectively treating crash locations relies on them. Data about road and traffic characteristics of the crash locations are also important. Good data permits, amongst other things (Austroads, 2009): • crash locations to be accurately pinpointed • the sequence of events in a crash to be appreciated • the contributing factors in a crash or a group of crashes to be identified, so that treatment can be directed at those factors • common factors across a number of crashes to be identified • the cost consequences of a single crash, all crashes at one location or several crashes with common factors to be identified • several crash sites to be ranked, so that treatment can be applied to the worthiest sites first Unfortunately, the data set currently available in Pakistan does not allow most of the above analysis, but it is sufficient to implement the basics of this strategy.

24 Three years is the minimum period to assure statistical reliability 65 Guidelines for Road Safety Engineering|Part I

The desktop analysis should be carried out taking into account only collisions resulting in personal injury because the reporting level of material damage collisions is variable and the information recorded – if any – is minimal. It is recommended that the number of collisions be taken into account as a method, which means that the number of fatalities and injuries should be ignored25. This approach is based on the following assumptions: • The use of the number of casualties as a criterion may distort the network ranking, potentially making critical road sections with few crashes but many fatalities and/or injuries26. • The target of this methodology is the infrastructure. The road does not have any influence in certain variables determining the number of casualties (i.e. number of vehicle occupants) or the severity of the crash (i.e. seat belt/helmet wearing, age of occupants, presence of air bags and/or other vehicle safety devices, etc.). In order to classify the road network, the following indexes can be used: • Collision rate (per volume of traffic)27 • Frequency of collisions (or collisions per km of road) The difference between the two above indexes is that the second takes no account of exposure (i.e. traffic). Of course, the accuracy of the collision rate is dependent on the accuracy of traffic volume information. In a first phase, where the overall objective is to reduce crash numbers, the frequency is a valid index and its use is recommended in case the traffic data are not reliable and/or are not available for the whole network to be analysed. The above indexes must be referred to homogeneous road sections, that is: • Layout and traffic flow must not change significantly in any single route (e.g. sections can contain either motorway links or non-motorway links, but not both). • The whole section must fall along the same numbered road to ensure that crash data can be retrieved using only road number and chainage. • The design and operation must be as uniform as possible, so that the conditions along the route are similar and the crash index reflects a reasonable average over all parts of the section. • Sections should start and end at relevant junctions. Besides the above criteria, it is essential to assess lengths that minimise the impact of year-on-year variability in crash numbers and present a stable

25 The exception to this ‘rule’ would be a site with a much higher than expected proportion of fatal and serious crashes. For example, on a high-speed road with one fatal and three serious injury crashes but no minor injury crashes. 26 As an example, on 11th November 2014, along the Therry bypass (Khairpur, Sindh), a passenger bus collided a truck. In just one crash 59 casualties occurred. Can a single crash define a site as hazardous? 67.8799:;:7<;/>?@A × CDE 27 !"##$%$"& ()*+ ,+( 10/0+ℎ. 34 = FGH.IH J@>; × KKLM × A7NO? 9?

longer-term estimate of crash risk. Research conducted by the Transport Research Laboratory (TRL) show that the number of crashes on busy networks over a three-year data period for sections of at least 5 km provide a reasonably robust estimate of risk. Sections less than 5 km tend to show greater year-on- year variability in crash numbers in addition to being more likely to change risk rating from one period to another and were therefore less reliable when compared over time. For motorways and dual and single carriageways these differences were significant up to section lengths of 10 km (EuroRAP, 2018). Minimum thresholds of 10 km for motorways and dual carriageways and 5 km for single carriageways are therefore recommended. Following this method, this initial study results in a risk mapping that identifies high collision sections which are then subject to a detailed desktop study.

Figure 38 – Risk map of motorway and national road network of France 4.5.2 Detailed desktop study Per each high collision section, a detailed review of the collision data should be undertaken to identify: • Clusters of crashes (i.e. blackspots) • Factors which can explain how the various road users failed to cope immediately prior to the collisions Unfortunately, the second analysis requires information that in Pakistan is not yet fully available. The desktop analysis must be therefore necessarily

67 Guidelines for Road Safety Engineering|Part I

supplemented by a site inspection carried out following a proactive approach (see following paragraph). If data allow this kind of analysis, they should furtherly clustered by common crash-type or factors like common weather or daylight, common speed range, etc.28 This analysis should aim to highlight factors common to a number of the collisions at the site in the view that the prevalence of crashes at only some locations, and the clustering of crash-types at a single location usually indicates that there are common causes for the crashes. The aim of this analysis is therefore to identify sites which have ‘treatable’ collision problems amenable to road engineering treatments. Figure 39 shows a site with a dominant right turn accident problem, along with other more individual accident types.

Figure 39 – Site with a predominant right-turn collision problem (source: DoE, 1996)

If there is not a dominant crash-type, development of a remedial treatment can be very difficult. Indeed, it may be that no engineering measure is applicable to the problems at that site. Alternatively, there may be two or more major crash-types, with two or more different remedial engineering treatments called for. 4.5.3 Site visit Once the collision data and other relevant data for a particular location have been studied, it is necessary to carry out a site visit. The site visit should only take place after the initial collision study has been completed. This should avoid the pre-judgement of collision problems that can happen if the site is visited prior to the collision data been examined.

28 For this kind of analysis, it would be very useful to examine the individual police crash report forms. Unfortunately - in Pakistan like in many other countries - these records are not made easily available to practitioners. 68 Guidelines for Road I Safety Engineering|Part I

The primary purpose of the site visit is to identify any environment and traffic deficiencies which may have contributed to the recorded crash history. To ensure that road deficiencies are identified it is essential that site inspections are carried out in an extremely systematic and purposeful manner. If the crash data show a cluster of a particular crash-type, this means several people are misreading the situation as they approach, drive through, turn at, walk across, or otherwise negotiate the location. The questions that always should be done are therefore: • What is causing them to do this? • What is misleading or difficult to deal with? These issues cannot be identified only from crash data, crash report forms or photographs. It is essential to carry out a site visit always taking into account the point of view of different road users (‘role play’). Basic principles of the site visit are as follows: • Look at the site at in relation of the results of data analysis • Carry out the visit in conditions similar of those when crashes occur (e.g. same time, peak/off-peak, light/dark, wet/dry, etc.) • View the location as the road users in a crash may have seen it • Drive through the location repeating the manoeuvres featured in the crash data • Observe ‘dangerous movements’ • Take notes, photos and videos 4.5.4 Diagnosing the crash problems The objective of this activity is - basing on crash analysis and site visit - to identify these common causes and to counter them by applying appropriate remedial engineering treatments. It is not possible to define specific rules to identify the causes. One type of crash may have different causes at different locations. A list of possible contributing factors for different types of crashes is provided in Table 8. Table 8 - Some possible contributing factors for different type of crashes

Crash type Possible contributing factors Right angle - Restricted sight distance. collisions - High approach speeds. - ‘See through’ effect on a minor approach. - Obscured control sign, control lines or signal lanterns. - The presence of the intersection is not otherwise evident (at time of day). - Traffic volumes too high for Give Way or Stop controls (inadequate gaps).

69 Guidelines for Road Safety Engineering|Part I

Crash type Possible contributing factors Right turn - Restricted visibility. collisions with - Queued oncoming right turners block visibility. oncoming - Insufficient number of gaps in oncoming traffic. traffic - Too many lanes of oncoming traffic to filter across. - Complex intersection layout. Straight ahead - Queued right turn vehicles further ahead. rear end - Traffic signals around curve or over crest. collisions - Other unexpected cause of delay further ahead. - Inadequate skid resistance or pavement drainage. - Wrong offset timing of linked signals. - ‘See through’ effect of consecutive traffic signals. - Inadequate inter-green phase on signals. - Presence of parked cars. - Unstable flow on high speed road. Right or left turn - Turning vehicles where they are not expected (e.g. just before or rear end just after signals). collisions - A left turn slip lane permitting high speed turns. Hit fixed object - Islands not visible. crashes - Complex layout. - Reasons as for Run-off-road crashes Crashes - Unexpected parked vehicle in traffic lane. involving a - Edge line not visible. parked vehicle - Lanes too narrow. Side-swipe - Lanes too narrow (for traffic composition, speed, curvature of collisions road, angle of lanes). - Lane lines, edge lines not visible. - Presence of parked cars or other obstruction. - Unexpected lane drop or merge area. - Inadequate direction information. Head-on - Lanes too narrow (for traffic composition, speed or curvature of collisions road). - Centreline not visible. - Severity of curve cannot be judged. - A hidden dip or crest. - Insufficient overtaking opportunities. - Road surface deficiencies. Run-off-road - Narrow lanes or narrow seal. type crashes - Severity of curve cannot be judged. - Edge of the road is not evident. - Gravel shoulders do not allow recovery of control. - Alignment of road is deceptive. - Inadequate skid resistance or pavement drainage.

70 Guidelines for Road I Safety Engineering|Part I

Crash type Possible contributing factors Pedestrian - Too much traffic for adequate gaps. crashes - High speed, multi-lane and two-way traffic. - Complex or unexpected traffic movements. - Traffic hidden by parked cars, other objects or excessive landscaping. - A marked crossing which is not evident to drivers. - Long signal cycles which encourage pedestrians to disobey signals. - Inappropriate device or lack of devices for mix of pedestrians (e.g. disabled). - Inadequate lighting. Railway level - Location of crossing is not evident. crossing - Impending presence of train is not evident. crashes - Form of control is not accurately identified (or is inconsistent). - Driver’s attention distracted by intersection or other feature. - Obscured control devices. (source: Austroads, 2009) 4.5.5 Treating the crash problems 4.5.5.1 Approaches to crash reduction Once the crash patterns and causes are identified, the issues can be approached using different strategies. In general, the following four approaches can be referred to: i. Single site: Crashes occurring at an individual site are examined in detail, common crash types are identified and measures introduced to treat the problems identified. An example could be the introduction of a chevron sign on a bend with a loss of control accident problem. ii. Route action: Sections of road are identified and treated together. An example of this could be the introduction of edge line markings along a route with a record of vehicles leaving the road, or the introduction of traffic calming and gateways through a village on a national route. iii. Mass action: Groups of sites are identified with common crash causes and a single measure introduced at all the sites. An example of this approach could be to identify a series of sites with wet skid accidents and apply anti-skid surfacing to those sites. The mass action approach is particularly useful in rural areas on local roads with small crash numbers. iv. Area-wide action: Parts of an urban area are identified where crashes of a particular type can be identified, but are not confined to single sites or routes. An example of this could be the introduction of traffic calming to reduce pedestrian accidents in an urban area. These approaches illustrate the various ways in which an authority can tackle its road safety problems. Each approach has a different emphasis and is likely to produce a variation in the type of crash likely to be treated.

71 Guidelines for Road Safety Engineering|Part I

Figure 40 – Possible approaches to crash reduction (source: DoE, 1996)

4.5.5.2 Selecting the countermeasures Having identified the elements of the road and traffic environment which contributed to the crashes, the next step involves consideration of countermeasures. For a solution to be effective, it must be applied to a particular problem which it is known to affect. It must be an effective countermeasure. The aim of countermeasure development is to: • select countermeasures which, on the basis of professional judgement and experience, can be expected to reduce the number or severity of crashes of the type(s) dominant at the location • check that adopted countermeasures do not have undesirable consequences, either in safety terms (e.g. lead to an increase in the number or severity of another crash-type) or in traffic efficiency or environmental terms • be cost-effective, i.e. maximise the benefits from the whole program of expenditure over a number of sites • be efficient, i.e. produce benefits which outweigh the costs. There are several criteria for countermeasure selection, including (Ogded, as cited in Austroads, 2009): • technical feasibility: can the countermeasure provide an answer to the safety problems which have been diagnosed and does it have a technical basis for success? • economic efficiency: is the countermeasure likely to be cost-effective and will it produce benefits to exceed its costs? • affordability: can it be accommodated within the program budget; if not, should it be deferred, or should a cheaper, perhaps interim solution be adopted? • acceptability: does the countermeasure clearly target the identified problem and will it be readily understandable by the community?

72 Guidelines for Road I Safety Engineering|Part I

• practicability: is there likely to be a problem of non-compliance, or can the measure work without unreasonable enforcement effort? • political and institutional acceptability: is the countermeasure likely to attract political support and will it be supported by the organisation responsible for its installation and ongoing management? • legal conformity: is the countermeasure a legal device, or will users be breaking any law by using it in the way intended? • compatibility: is the countermeasure compatible and consistent with other strategies, either in the same locality or which have been applied in similar situations elsewhere? The key to the selection of countermeasures is to concentrate on the particular crash-types which have been identified in the diagnosis phase and which are amenable to treatment with road or traffic engineering measures. If this relationship between detected issues and countermeasures fails, the investment is destined to be not cost-effective. Often there will be a number of alternative remedial treatments which could be applied, either individually or in combination. The final choice about which countermeasure(s) to select requires road safety engineering experience and judgement about the factors which have led to the crashes. A catalogue of possible countermeasures to be implemented in Pakistan is reported in the Part II of these Guidelines. 4.5.5.3 Assessing the crash reduction Once the countermeasure is selected its effect in crash reduction must be determined as well as its implementation cost (even if approximated). Several studies have been performed to estimate the safety impact of various types of road infrastructure improvements. Many existing Crash Modification Factors (CMF)29 are derived from these evaluation studies, like before-and-after analysis of actual countermeasures implementation (see, for instance, Elvik et al., 2009)30. An alternative method, more pragmatic, guesstimates collisions savings comparing the current crash patterns with a situation where the measure is already implemented. In practice, it is assumed that unless something is done at the site the existing pattern of crashes will be repeated over time. The collision record is therefore re-examined, and each crash assessed to

29 A CMF is a multiplicative factor used to compute the expected number of crashes after modifying the road characteristics at a specific site (e.g. by implementing a given countermeasure). A CMF higher than 1 is assumed to increase the likelihood or severity of a target crash type, while if lower than 1 decreases crash likelihood/severity. 30 A useful reference is SafetyCube, a research project funded by the European Commission under the programme Horizon 2020 (www.safetycube-project.eu). Under this project an innovative road safety Decision Support System (DSS) providing detailed interactive information on a large list of road crash risk factors and related road safety countermeasures has been developed (www.roadsafety-dss.eu). 73 Guidelines for Road Safety Engineering|Part I

determine whether that particular crash would have happened if the proposed measure had been implemented. An application is presented in Box 5.

Box 5. Assessment of collision savings (example)

There is an urban priority junction with a history of 10 injury collisions in 3 years, distributed as follows: - 4 overshoots from side road - 3 right turns from main road - 3 pedestrians crossing main road It is proposed to install traffic signals.

How can the reduction of crashes be estimated? The following table can be used as a reference guide.

Type of measure Collision reduction Measure totally eliminating the risk factor - 75/100% (e.g. segregation of different traffic streams) Measure partially eliminating the risk - 35/75% factor (e.g. new traffic signal) Measure not eliminating the risk factor, - 25/35% but making users aware of its presence (e.g. new markings)

With reference to the above ranges, the following reductions can be assumed:

Manoeuvre Potential % collision Potential collision reduction saving Cross over 75% x 4 3.0 Right turn 50% x 3 1.5 Pedestrian 33% x 3 1.0 Total 5.5

It may be needed to add collisions generated by the scheme (e.g. shunts at traffic signals).

74 Guidelines for Road I Safety Engineering|Part I

4.5.6 Prioritising the schemes on the network31 Having decided upon treatments for high collision sites, the road authority should prioritise all potential schemes within their network32. This should be done using an economic appraisal that essentially concerns with the economic efficiency of alternative proposals. With this appraisal economic costs of a proposal are compared with the economic benefits. The outcome is not only an assessment of the quality of an individual project (i.e. there is a net economic benefit, such that the community is economically better off by implementing the project than by not doing so), but also - and above all - an indication of which project (or set of projects) is the best. While the assessment of costs is rather simple and does not require any additional information, the assessment of benefits is more complex. Benefits are assessed by assigning a monetary value to the crashes saved. There is in fact a cost incurred to the community when a road crash takes place. Collision costs can be classified into: • lost output due to death or injury • human costs of pain, grief and suffering • resource cost in terms of: o hospital o emergency service o damage to property o insurance costs At this stage, no studies are available to estimate this social cost for Pakistan. For the time being, therefore, we can only rely on proxy analyses. iRAP provides a statistical evaluation of life that can be applied worldwide, along with typical fatal to serious injury ratios (McMahon & Dahdah, 2008). The recommended values for middle and high-income countries are as follows: • Value of fatality = 70 x GDP per capita • Value of serious injury = 25% of the Value of fatality It should be borne in mind that these figures should in any case be related to the ‘average crash’ in Pakistan. In practice, the total cost of the crashes has to be calculated using the above method and then divided by the number of injury crashes. Once costs and benefits are estimated, it is possible to compare them and decide the best schemes to be implemented. The most efficient means of prioritising is to use the First Year Rate of Return (FYRR) that is a simple way of calculating whether a scheme can be justified in

31 This methodology mainly refers to a ‘single site’ approach 32 The described methodology can be applied also to a single site in case it is necessary to choose between alternative options 75 Guidelines for Road Safety Engineering|Part I

economic terms33. The method can be used to rank proposals at different sites, but also to rank different options at a site. The FYRR is calculated using the formula: Y&&Z)# ["##$%$"& %)0$&\% VWXX = ][ℎ+4+ ["%* The options with the highest FYRR are then chosen for implementation. However, there should be a note of caution here in that high FYRR values are generally achieved with schemes costing little money but saving fewer crashes (the temptation to solve every problem by putting up a sign should be avoided!). The number of crashes saved within each option should therefore be taken into account. It is suggested that a minimum predicted crash reduction of 25-30% should be achievable for each site. 4.5.7 Monitoring and evaluation Monitoring is the systematic collection of data about the performance of road safety treatments after their implementation. Evaluation is the statistical analysis of that data to assess the extent to which the treatment (or a wider treatment program) has met crash reduction objectives. Post-implementation monitoring is essential to ascertain the positive and negative effects of a treatment and thus improve the accuracy and confidence of predictions of that treatment’s effectiveness in subsequent applications. There is a duty to ensure that the public does not experience additional hazards as a result of treatments and this duty carries with it an implied need to monitor what happens when a scheme is introduced. The purposes of monitoring a treatment are to (IHT, as cited in Austroads, 2009): • assess what changes have occurred in crash occurrence and whether safety objectives have been met • assess the treatment’s impact on the distribution of traffic and the speeds of motor vehicles • call attention to any unintended effects on traffic movements or crash occurrence • assess the effects of the treatment on the local environment • learn of the public’s response to the treatment: its acceptability in general and any concerns about safety in particular.

33 It should be noted that there is a more rigorous method of economic assessment of proposals known as the Net Present Value (NPV). The NPV method takes a longer-term view than FYRR and calculates benefits over a 10 or 15-year scheme life, using discount factors to assess costs and benefits at year 1 prices. It may be more appropriate to use the NPV method for higher cost schemes. 76 Guidelines for Road I Safety Engineering|Part I

There are three elements to monitoring and evaluation (County Surveyors’ Society, as cited in Austroads, 2009): • Pay careful attention to a site immediately after treatment in case things go badly wrong. • Assess the effects over a longer time period, say three years, to attempt to determine the influence of the treatment on crashes or other performance measures. This requires careful statistical analysis, correcting for external factors and bearing in mind that crash frequencies may be so low that any observed changes in crashes may not be statistically significant. • Focus, over this longer time period, upon the crash-types which the treatment was intended to correct and assess whether these have declined. The first parameter to be monitored is, of course, the frequency of crashes. The main problem is that to have reliable before-after comparisons a three-year period of analysis is needed. An early indication of safety benefits can be obtained from other observations like: • traffic and speed surveys • conflict studies34

34 A ‘conflict study’ is a technique that involves observation of the conflicts, or ‘near crashes’, experienced by road users. Conflicts can be defined as ‘situations involving one or more road users who approach each other such that there is a risk of collision if their movements remain unchanged’. In a conflict study the numbers of conflicts are recorded and graded according to the scale of severity ranging from controlled evasive manoeuvres to extreme emergency action. 77

Guidelines for Road I Safety Engineering|Part I

References

AASHTO (2011) – A Policy on Geometric Design of Highways and Streets – Washington, USA ADB (2018a) – CAREC Road Safety Engineering Manual 1 / Road Safety Audit ADB (2018b) – CAREC Road Safety Engineering Manual 2 / Safer Road Works ADB (2018c) – CAREC Road Safety Engineering Manual 3 / Roadside Hazard Management AfDB (2014a) – Road Safety Manuals for Africa – Existing Roads: Proactive Approaches AfDB (2014b) – Road Safety Manuals for Africa – Existing Roads: Reactive Approaches AfDB (2014c) – Road Safety Manuals for Africa – New Roads and Schemes: Road Safety Audit Ahmed, A. et al. (2018) – Road Safety Audit Policy, Procedures & Guidelines – National Highway Authority, Islamabad, Pakistan Austroads (2009) – Guide to Road Safety Part 8: Treatment of Crash Locations – Sydney, Australia Buchanan, C. (1963) - Traffic in towns; A study of the long term problems of traffic in urban areas - London, UK CERTU (1994) - Les ralentisseurs de type dos d’âne et trapézoïdal [Speed humps and speed tables] (Lyon, France) Department of Environment (1996) - A Guide to Road Safety Engineering in Ireland – Dublin, Ireland DETR (2000) – New directions in speed management: a review of policy – London, UK DHV (2005) – Sustainable safe road design – A practical manual – The World Bank, Washington, USA Elvik, R., Hoye, A., Vaa, T., Sorensen, M. (2009) - The Handbook of Road Safety Measures. Second Edition - Emerald Group Publishing Limited, Bingley, UK EuroRAP (2018) – RAP Road Risk Mapping Manual: Technical Specifications – Basingstoke, UK McMahon, K. & Dahdah, S. (2008) – The true cost of road crashes. Valuing life and cost of a serious injury – iRAP, Basingstoke, UK NRA (2014) – Road Safety Inspection Guidelines - Volume 5 Section 2, Part 2, NRA HA 17/14 - Dublin, Ireland NRA (2016) – Road Safety Impact Assessment Guidelines – Volume 5 Section 2, Part 2, NRA HA 18/16 - Dublin, Ireland

79 Guidelines for Road Safety Engineering|Part I

NTRC (1989) – Manual of Signs, Signals and Road Markings – Islamabad, Pakistan OECD (2006) – Speed management – Paris, France OECD (2008) – Towards Zero: Ambitious Road Safety Targets and the Safe System Approach – Paris, France PIARC (2015) – Road Safety Manual – A manual for practitioners and decisions makers on implementing Safe System infrastructure! – Paris, France UN ESCAP (2004) – Intergovernmental agreement on the Asian Highway Network – Shanghai, China UN ESCAP (2017a) - Development of road infrastructure safety facility standards for the Asian Highway Network (draft version) – Bangkok, Thailand UN ESCAP (2017b) - Recommended Detail Design Guidelines on Road Infrastructure Safety Facilities for the Asian Highway Network (draft version) – Bangkok, Thailand Van Schagen, I. N. L. G., Dijkstra, A., Claessens, F. M. M. & Janssen, W. H. (1999) - Recognition of sustainably safe road types - SWOV Institute for Road Safety Research, Leidschendam, The Netherlands Wegman, F. & Aarts, L. (2006) - Advancing Sustainable Safety: National Road Safety Outlook for 2005-2020 - SWOV Institute for Road Safety Research, Leidschendam, The Netherlands WHO (2008) - Speed management: a road safety manual for decision-makers and practitioners – Geneva, Switzerland WHO (2010) – Global Plan for the Decade of Action for Road Safety 2011 – 2020 - Geneva, Switzerland WHO (2018) – Global status report on road safety 2018 – Geneva, Switzerland

80 Guidelines for Road I Safety Engineering|Part I

Annex – Road design standards’ benchmark

Road design standards analysed:

Standard Reference Region/ Country Asian UN ESCAP - Recommended Detail Design Guidelines on Asia Highway Road Infrastructure Safety Facilities for the Asian Highway Standards Network – April 2017 TEM UNECE – TEM standards and recommended practice – Europe February 2002 AASHTO AASHTO - A Policy on Geometric Design of Highways and USA Streets – 2011, 6th Edition Austroads Austroads – Guide to Road Design Part 3 – Geometric Australi Design – September 2016 a DMRB Design Manual for Roads and Bridges: Volume 6 – Road UK Geometry ICTAAL Ministry of Public Works - Instruction sur les Conditions France Techniques d’Aménagement des Autoroutes de Liaison [Instruction on technical design requirements for rural motorways] – Circular of 12th December 2000 ARP Ministry of Public Works – Aménagement des routes France principales [Construction of main roads] – August 1994 Instrucción de Ministry of Development - Instrucción de Carreteras – Spain Carreteras Norma 3.1-IC – Trazado [Road instruction – Standard 3.1- IC – Alignment] - Order FOM/273/2016 of 16th February 2016 Italian Ministry of Infrastructures and Transports – Norme Italy Standards funzionali e geometriche per costruzione delle strade [Functional and geometric standards for road construction] – Ministerial Decree 5th November 2001

The comparative analysis refers to two typical cases: • dual carriageway motorway • single carriageway rural road

81 Guidelines for Road Safety Engineering|Part I

Motorways (design speed: 110/120 km/h)

Standard Region/ Category Minimum radii (m) Max. Cross-section Country grade Horizontal1 Vertical Lane Shoulder width (m) Median (%) width (m) width (m) Crest Sag Outer Inner Asian Highway Asia Primary N/A N/A N/A N/A 3.50-3.75 >1.00 N/A >1.00 Standards road TEM Europe Motorway 650 12,000 4,444 6 3.50-3.75 3.00 - >4.00 AASHTO USA Freeway 667 N/A N/A 3-4 3.60 3.00 1.20-2.40 >3.00 Austroads Australia Motorway 667 N/A N/A 3-6 3.50 2.50-3.002 1.00 N/A DMRB UK Motorway 510 N/A N/A 3 3.65 3.30 0.70 >3.10 (D2M)

ICTAAL France L1 600 12,500 4,200 5 3.50 2.50-3.00 1.00 N/A Instrucción de Spain Motorway 700 N/A N/A 4 3.50 2.50 1.00-1.50 >2.00 Carreteras (A-120) Italian Standards Italy Motorway 667 8,000 4,800 5 3.75 3.00 0.70 >2.60 (A)

1 With superelevation (max. 8%) 2 Depending on traffic volume

82 Guidelines for Road I Safety Engineering|Part I

Rural roads (design speed: 80 km/h)

Standard Region/ Category Minimum radii (m) Maximum Cross-section Country grade Horizontal3 Vertical Lane width Shoulder (%) (m) width (m) Crest Sag Asian Highway Standards Asia Class II road 200-400 N/A N/A N/A 3.50 >1.00 AASHTO USA Collector road 229 N/A N/A 6-9 3.00-3.604 0.60-2.405 Austroads Australia Intermediate 153 N/A N/A 4-9 3.10-3.505 1.50-2.505 speed rural road DMRB UK Rural all- 255 N/A N/A 6 3.65 1.00 purpose road (S2) ARP France All-purpose 240 3,000 2,200 5 3.50 2.00 road (R80) Instrucción de Carreteras Spain Conventional 265 N/A N/A 5 3.50 1.50 road (C-80) Italian Standards Italy Secondary rural 252 3,000 3,000 7 3.75 1.50 road (C1)

3 With superelevation (max. 8%) 4 Depending on design volume and speed 5 Depending on design volume

83