INTERURBAN ACCIDENT RATES BY ROAD TYPE AND
Transcription
INTERURBAN ACCIDENT RATES BY ROAD TYPE AND
INTERURBAN ACCIDENT RATES BY ROAD TYPE AND GEOMETRIC ELEMENTS Dr. D. O’Cinneide J.C. Murphy T. Ryan University College Cork, Ireland 1. INTRODUCTION The identification of accident rates by road type enables a comparison of road types and is important in a time where information on the best allocation of investment funds in roads is vital in order to utilise the resources of the European Union and each country to best advantage. This paper summarises the findings of a four year detailed investigation into road geometric elements and accident rates which was part-funded and supported by the Irish National Roads Authority 1. The resulting national accident rates are the first results of analysis carried out on the Irish National Road Authority (NRA) Needs Study Database. Previously, accident rates in Ireland were mainly based on default UK accident rates. The derived accident rates are compared to published International accident rates. The NRA Database allows research which previously could not have been conducted on the entire network to take place. The purpose of the database is to store detailed information of the network and provide data for the assessment of investment needs and the development of national transport policy. This computerised National Roads Database / Geographical Information System database contains detailed information on each uniform section of the National Road Network such as geometric details, accident data, and traffic volumes. A comparison of the relative safety of the various road types in the interurban (speed limit > 40mph) part of the network was carried out by calculating and comparing accident rates. These rates were expressed in accidents per 100 million vehicle kilometres of travel. Geometric design standards are largely based on logically derived relationships rather than the results of accident studies or database research. Road type is the predominant factor in geometric design choice which affects accident rates. The choice of road type affects not only the speed of vehicles, the number of junctions or interchanges, the type of passing manoeuvre permitted as well as various geometric design features. In order to provide the highest feasible standard of road, the cost of the various road types are balanced against the expected benefits. The accident savings are vital part of such an economic assessment. © Association for European Transport 2004 2. ACCIDENTS – WHAT THEY ARE & HOW THEY ARE RECORDED Standardisation of accident definitions is an essential step in the process towards internationally comparable accident rates. The accident definitions 2 in this study and in the NRA Database analysis NRA (2001) are consistent with Grime (1987) and the Hopes Report (1993). Incongruities arise due to the differing international definitions and reporting procedures. For example, a serious injury accident in Austria is defined as one which results in a person being hospitalised and unable to work for at least 24 days. Yet some countries do not even require hospitalisation for an accident to be described as a “serious injury accident”. (Berns & Bruhning, 1998). In Ireland, accidents are recorded by the police and entered into a database by the National Roads Authority. The management of such an accident database is an arduous task with many incongruities arising. For example; “either a lack of data or imprecision in the grid coordinates” (NRA, 2002). Human error in data recording and transferring is minimised by the use of unique accident ID numbers assigned by the police. The accident analysis carried out in this study was based on accident data from the National Road Needs Study Database. This database only includes injury accidents. Data checks performed include those to ensure no multiple entries occur and to identify accidents occurring at change points of road sections. This paper examines the results of two studies of the NRA Database; the first study concluded in 2002 and examined accident data for the period 1990 to 1998, the second study concluded in 2003 and examined accident data for the period 1996 to 2000. The results produced by these two studies permitted comparisons of the derived accident rates. Table 1 shows the lengths and number of accident by road type for the Interurban 3 sections of the National Network. The variations between the road lengths and the years for which accident data was available in the two studies was as a result of the updating of the National Road Network database, i.e. sections that were eligible for inclusion, or the road type to which sections were assigned, may have been altered. Interurban Length – National Network (km) Road Type Number of Accidents Study 1 Study 2 1990 - 1998 1996 – 2000 3020 5102 3459 1021 869 2719 1621 Improved Wide 2-Lane 226 369 690 809 All 2-Lane Undivided 4189 4258 8511 5889 Motorway 103 114 68 149 Dual Carriageway 150 157 697 592 3-Lane (All) 87 87 244 147 4530 4616 9520 6777 Study 1 Study 2 Unimproved 2-Lane 2943 Improved Standard 2-Lane TOTALS Table 1 - Interurban Lengths and Accidents occurring by Road Type, unadjusted for underreporting. © Association for European Transport 2004 Table 2 shows the percentage of reported accidents by accident type occurring on each of the different road types on the Interurban National Network as identified by Study 2. This table also shows the values contained in the newly published Irish Cost Benefit Analysis Guidelines for COBA 11(NRA 2004), which took into account the findings of this study. % Fatal Road Type Study 2 NRA 2004 % Serious Study 2 NRA 2004 % Minor Study 2 NRA 2004 Unimproved 2-Lane 8.2 24.8 67.0 Improved Standard 2-Lane 12.8 23.6 63.5 Improved Wide 2-Lane 9.5 26.0 64.5 All 2-Lane Undivided 9.7 9.7 24.6 24.6 65.7 65.7 Motorway 6.7 6.0 13.4 13.3 79.9 80.7 Dual Carriageway 6.8 6.8 14.9 14.9 78.4 78.4 3-Lane (2 lane side) 14.8 11.0 27.2 24.2 58.0 64.8 3-Lane (1 lane side) 13.6 11.0 30.3 24.2 56.1 64.8 ALL ROADS 9.4 23.6 66.9 Table 2 - Percentage of Fatal, Serious and Minor Injury Accidents occurring by Road Type, unadjusted for underreporting. From the percentages in Table 2, it is clear that there are few major variations in the composition of the accident totals for each of the three categories of Undivided Two-Lane road. The main variation is the percentage of accidents proving fatal on Improved Standard Two-lane roads. At almost 13%, this appears high in comparison with both the Improved wide and Unimproved Standard categories (almost one third higher). Serious and minor injury accidents exhibit roughly the same distribution pattern, with approximately 25% of accidents involving serious injury, and 65% involving minor injury in the cases of all three road types. It is likely that the actual percentages are affected by underreporting rates. It can also be seen that the results of Study 2 have been taken into consideration in the newly published Irish Cost Benefit Analysis Guidelines. © Association for European Transport 2004 3 INTERURBAN ACCIDENT RATES FOR THE IRISH NATIONAL ROAD NETWORK The carriageway types examined in this paper are Motorway, DualCarriageway, Three-Lane, Undivided Improved Wide Two-Lane 4, Undivided Improved Standard Two-Lane 5 and Undivided Unimproved Two-Lane 6. Interurban road accident rates were calculated by severity, non-injury accidents were not included in the study. The units of the accident rates were accidents/100 million vehicle kilometres travelled. This unit was chosen after the examination of the international statistics and a regression analysis on the most significant variables affecting accidents. The network was divided into homogenous sections on the basis of Carriageway Type, Traffic Volumes, and Speed Limit (Urban /Interurban) for the purpose of analysis. Accident Rate = Total (No. of Accidents)x 108 Total(veh.km) Table 3 and Table 4 show the results of the analysis on accident rates for Study 1 and Study 2 respectively. When viewing these rates, it is important to bear in mind the relative lengths and vehicles km of each road type available for analysis. It is also important to note that accident underreporting has a large effect on the accident totals, and therefore on the accident rates presented here. Road Type Length (km) Vehicle Km 2943.2 Accident Totals Accident Rates* Fatal KSI ALL Fatal KSI ALL 34354352930 465 2069 5102 1.35 6.02 14.85 1020.6 20466979552 334 1198 2719 1.63 5.85 13.29 225.6 5297259391 91 303 690 1.72 5.72 13.03 4189.3 60118591873 890 3570 8511 1.48 5.94 14.16 Motorway 102.7 6774919891 3 28 68 0.10 0.41 1.00 Dual Carriageway 150.2 9259379575 68 235 697 0.73 2.54 7.53 87.3 1937031081 38 110 244 1.96 5.68 12.60 4529.5 78089922420 999 3943 9520 Unimproved 2-Lane Improved Standard 2-Lane Improved Wide 2Lane All 2-Lane Undivided 3-Lane (All) TOTAL Table 3- Study 1 - Interurban Accident Rates by Road Type (*Accidents per 108 Veh. km of travel) unadjusted for underreporting. © Association for European Transport 2004 Accident Totals Fatal KSI All Accident Rates * Fatal KSI All Road Type Length (Km) Vehicle Km Unimproved 2-Lane 3019.9 23025641926 284 1141 3459 1.23 4.96 15.02 869.3 12444924688 208 591 1621 1.67 4.75 13.03 369.2 6270711306 77 287 809 1.23 4.58 12.90 4258.5 41741277920 569 2019 5889 1.36 4.84 14.11 Motorway 113.9 5231463495 10 30 149 0.19 0.57 2.85 Dual Carriageway 156.6 6569210644 40 128 592 0.61 1.95 9.01 49.9 773051607 12 34 81 1.55 4.40 10.48 Improved Standard 2Lane Improved Wide 2Lane All 2-Lane Undivided 3-Lane (2 lane side) 3-Lane (1 lane side) 37.4 600852747 9 29 66 1.50 4.83 10.98 TOTAL 4616.3 54915856414 640 2240 6777 Table 4- Study 2 - Interurban Accident Rates by Road Type (*Accidents per 108 Veh. km of travel) unadjusted for underreporting. If the rates of accident underreporting are assumed to remain constant, regardless of road type and year, then valid comparisons can be made between the relative safeties of the individual road types on the basis of the accident rates. It can be seen from Table 3 and Table 4 that there is an increasing trend in safety as the road type is upgraded. Divided roads are significantly safer than undivided roads. Motorways have by far the lowest accident rates (the short length of Motorways and the small number of Motorway accidents in both studies should be noted), followed by Dual Carriageways, then Three-Lane roads (roads with climbing lanes) and TwoLane roads. Dual Carriageways are the second safest road type, showing the second lowest rates in all three categories of accident. The median barrier of both Dual Carriageways and Motorways may have contributed significantly to their safety when compared with Two-Lane and Three-Lane Undivided roads. The control of access, presence of hard shoulders and consistency of design are also likely to make telling contributions. Although Two-Lane Undivided roads proved the most dangerous in terms of the All-accident rate, Three-Lane roads showed a high Fatal accident rate, both for the one-lane and the two-lane sides. The short length of the ThreeLane road sections used in the calculation of the rates must be noted, as it may have contributed to the high Fatal accident rate when compared to the other road types. This high rate may be due to a higher percentage of head on collisions and right turn accidents on Undivided Three-Lane roads. Perhaps the most surprising feature of the All-accident rates is the seemingly small reduction in the All-accident rate when going from Three-lane roads to Dual Carriageways. This increase in safety is more pronounced when the KSI and Fatal accident rates for both are compared. © Association for European Transport 2004 Study 1 (see Table 3) showed an increase in Fatal accident rates on TwoLane Undivided roads as the road is improved; from Unimproved (the safest) to Improved Standard to Improved Wide (the most dangerous). This differs from the results obtained in Study 2, which showed improved wide Two-Lane roads as the safest Two-Lane road type in terms of Fatal and All accident rates. Both studies showed a decrease in the All accident rate as the quality of Two-Lane Undivided roads is improved. When compared with Unimproved Two-Lane roads, Improved Two-Lane road types appear to have a significantly higher Fatal accident rate, a similar KSI rate and a lower All Accident rate. The higher Fatal rate may result from higher speeds on Improved roads. The principal difficulty when examining accident numbers and rates is the level of underreporting accidents. Reporting rates of accidents increase with increasing accident severity. It is reasonable to assume the need of larger adjustment factor for minor injury accidents than serious injury accidents. Inconsistencies and a lack of clarity in many cases make it difficult to compare the underreporting rate of injury accidents for different countries. It is estimated, on the basis of evidence from a variety of countries and sources, that the rate of road accident reporting in Ireland for serious injury accidents could be anywhere between 50% and 75%. The reporting rate of Minor injury accidents could be as low as 25% and as high as 50%. Therefore there is a large margin for error in applying any corrective factor to minor injury accident totals. Table 5, below, shows the estimates of reporting rates in Ireland as well as the relevant adjustment factor. These underreporting adjustments are then implanted in Table 6 for Study 1 and for Study 2. Accident Severity Reporting Rate Estimate Reporting Rate Adjustment Factor Fatal 100% 1.00 Serious 75% 1.33 Minor 50% 2.00 Table 5 - Summary of Accident reporting estimates. Road Type Unimproved 2Lane Improved Standard 2Lane Improved Wide 2Lane All 2-Lane Undivided Motorway Dual Carriageway 3-Lane Undivided Adjusted Accident Totals (Study 1) Fatal KSI All 465 2598 11273 334 1483 5833 Adjusted Accident Rates *(Study 1) Fatal KSI All 1.35 7.56 32.81 1.63 28.50 208 717 2777 Adjusted Accident Rates *(Study 2) Fatal KSI All 1.23 6.18 26.32 5.77 22.32 91 373 1480 1.72 7.04 27.94 77 356 1400 1.23 5.68 890 4454 18586 1.48 7.41 30.92 569 2498 10238 1.36 5.98 3 31 145 0.04 0.46 2.15 10 37 275 0.19 0.70 68 290 1611 0.73 3.13 17.40 40 157 1085 0.61 2.39 38 134 517 1.96 6.91 26.69 21 77 245 1.53 5.59 TOTAL 999 4909 20859 640 2768 11842 Table 6- Interurban Accident Rates by Road Type – Adjusted for underreporting (see Table 5) (*Accidents per 108 Veh. km of travel) 22.33 24.53 5.25 16.52 17.82 © Association for European Transport 2004 7.25 Adjusted Accident Totals (Study 2) Fatal KSI All 284 1424 6060 1.67 4 HOW DO THEY FIT IN WITH RATES FOUND ELSEWHERE? 4.1 Comparison of study results with the accident rates used in Ireland “Motorways are considered a safe but expensive road option, requiring a high traffic volume to be considered cost effective. Dual Carriageways appear the next safest carriageway type after Motorways. Although Dual Carriageways require higher investment costs than single carriageways, the additional cost can be outweighed by the increased benefits associated with the higher travel speed and better accident record than single carriageway roads” National Road Needs Study, NRA (1998). Accident rates were generated by this study for the four road types in the National Road Network in Ireland. These rates were compared with default accident rates used in the economic assessment of road schemes in Ireland (NRA, 1998) and the UK (DMRB, 2004). The accident rates for each road type are shown in Table 7. The accident rates found in the University College Cork studies were also adjusted for underreporting based on estimates made from a comprehensive review of literature pertaining to the problem of accident underreporting. The recently issued NRA (2004) COBA 11 figures reflect the results found by the University College Cork Studies. Road Type DMRB, 2004 COBA 11 10.7 NRA Technical Report, 1998 7.0 Study 1 Unadjusted Adjusted Accident Accident Rates Rates 1.0 2.1 Study 2 Unadjusted Adjusted Accident Accident rates rates 2.8 5.2 NRA, 2004 COBA 11 Motorway 2 lane 3.7 Dual Carriageway 12.1 12.0 7.5 17.4 9.0 16.5 9.4 2lane Modern 2-Lane 26.0 Roads Modern 2-Lane 17.0 13.3 28.5 13.0 22.3 21.2 14.2 Roads (HS) Older 2-Lane 28.8 Roads 15.0 26.3 20.0 14.9 32.8 Other 2-Lane 44.5 Roads Modern Wide 217.0 18.0 13.0 27.9 12.9 22.3 Lane Roads Three-Lane 12.6 26.7 10.7 17.8 12.5 Undivided 8 Table 7 – Comparison of Irish National Interurban Accident Rates (*Accidents per 10 Veh. km of travel) The relative safety of road types in terms of safety is the same for the University College Cork Studies and the recently issued NRA (2004) COBA 11 figures. It can be seen in Table 7 that the default accident rates are higher than the unadjusted rates found by this study. The NRA (2004) COBA 11 figures are similar to the unadjusted accident rates found by both Study 1 and Study 2. The accident rate for Motorways is more in keeping with the actual situation in Ireland than the UK DMRB (2004) COBA 11 values and the NRA (1998) Technical Report. For almost all road types examined, the default © Association for European Transport 2004 rates presented by the DMRB and in the Technical Report of the Needs Study are lower than the rates obtained in this study adjusted for underreporting. While the results presented here show that the default rates seem to underestimate the rate of accident occurrence on Irish National Roads, it is difficult to be definitive, as so little is known on the extent of the problem of accident underreporting in Ireland, or indeed internationally. The NRA (2004) COBA 11 guidelines produced accident rates which are in keeping with the finding of Study 1 and Study 2. The lower rates for Motorways found in both Study 1 and Study 2 could be attributed to the short length of motorway in the network and the relatively small number of accidents occurring. 4.2 Comparison of Accident Rates in Different Countries A list of “safest” countries cannot easily be identified, due to the various rates which could be used to identify “safeness”. A list drawn up by the number of road deaths per motor vehicle may yield a totally contradictory answer to a list based on the number of road deaths per million vehicle kilometres. Also, if only interurban sections were examined further results may be found. Correction factors can sometimes be applied to international figures in order to make standardised comparisons. It is important to include the lengths, vehicle kilometres and various accident totals in the same table as the calculated accident rates, to keep the rates in context. Comparisons of accident data on the IRTAD website provide an international road traffic and accident database which contrasts rates in various countries. From these values and the detailed notes that accompany them, it can be seen that accident rates from different countries cannot simply be compared without question; variations in definitions and underreporting are important factors when contrasting accident rates. Figure 1 shows the injury and fatality rate for 2002 various countries on all roads expressed per 108 vehicle kilometres of travel. Ireland can be seen to be mid-ranking in terms of fatality rates and fourth in terms of injury accidents. It must be noted however that underreporting of injury accidents would influence these figures greatly. Injury Accidents per 1 billion Veh.Km Fatalities per 1 billion Veh.Km 90 80 70 60 50 40 30 20 10 0 8 Figure 1 – Fatalities and Injury Accidents per 10 Veh. km of travel (IRTAD, 2002) © Association for European Transport 2004 It is very difficult to compare accident rates in different countries by road type because of differences in geometric layout, intersection types, traffic composition, etcetera. The absolute volume of traffic on the different types of road is also relevant. Direct comparisons are most relevant for motorways. Figure 2 shows and fatality rates for Interurban roads and Motorways reported by IRTAD (2002) expressed per 108 vehicle kilometres of travel. Again Ireland can be seen to be mid-ranking in terms of fatality rates. It should also be noted that these results cannot be directly compared with the number of fatal accident occurring. In t e r u r b a n F a t a l i t i e s p e r 1 b i l l i o n v e h . k m Motorw ay Fatalities per 1 billion veh.km 20 18 16 14 12 10 8 6 4 2 0 Figure 2 – Interurban and Motorway Fatalities per 108 Veh. km of travel (IRTAD, 2002) Comparing the results found by the University College Cork Studies to Figure 1 and Figure 2 it can be seen that the Injury accidents are slightly underestimated, whereas the fatality accidents are overestimated. In particular for Motorways, Table 6 showed the fatality rate in the order of 0.19 and Table 7 showed the accident rate at 5.2. The fatality rate for interurban roads is again overestimated in Figure 2 as can be seen from Table 6. Similar results were also found by EuroRAP (2002). “The safest roads everywhere in Europe are undoubtedly the motorways - though again Portugal shows the worst record with 14.1 deaths per billion vehicle-km. Italy is next with 10.2, followed by Austria (8.9), Belgium (7.2), France (5.4) and Finland (5.0). Lower-rate countries are Germany (4.5), Denmark (4.3), Ireland (4.0), Switzerland (3.3), Netherlands (3.3), Sweden (3.2) and Britain (2.0)”. These fatality rates for Ireland are closer to the Motorway all accident rate (Table 6). © Association for European Transport 2004 5 THE RELATIONSHIP BETWEEN GEOMETRIC ELEMENTS AND SAFETY A regression analysis was also carried out into the relationship between geometric elements and safety using the NRA database. This mirrored that carried out in 1975 on a similar database in Ireland. (Hearne, 1976). As with Hearne’s analysis, this work was not intended to result in the creation of a predictive equation, rather to examine the extent to which each of the variables included influence the occurrence of road accidents. The geometric elements considered were hard shoulders, medians, intersections, verges, grade, curves and passing sight distance. Other influencing factors on accident rates such as volume, speed and length were also considered. The results of the stepwise linear regression on Two-Lane Undivided roads showed that the coefficients obtained indicated that vehicles kilometres of travel was the most important of the variables examined in terms of the number of accidents which occur on a given two-lane undivided section of the Irish National Primary or Secondary Network. Higher values of vehicle kilometres of travel on a given road section appear to increase the likelihood of accidents occurring on the section, as would be expected. The variable ‘lane plus shoulder width’ was the second most important variable, indicating that, as the carriageway width increases, the number of accidents tends to decrease. Neither lane width nor hard shoulder width was found to be statistically significant when examined individually in the case of this model. Roadside events (the number of junctions, developments and lay-bys present at the roadside) were shown to be slightly less important than the combined width of the travel lane and hard shoulder. A second regression analysis dealt only with unimproved or standard rural sections of Two-Lane Road. This identified vehicle kilometres as the most important feature followed by the number of roadside events. Lane width was the third most important feature, with the accident number increasing with increasing lane width (to a maximum lane width of 3.75m). Hard shoulder width was the fourth most important of the variables analysed. As the Hard shoulder width is increased, the accident numbers on the section decreased. Further detailed analysis was carried out into the impacts of volume and lane and shoulder width on accidents, resulting in the conclusion that as hard shoulder width is increased, there is a reduction in the rate of accident occurrence. Lane widths of 3.25-3.50m should be avoided on Undivided Roads (Just below minimum lane width recommended by the NRA) 3.503.75m seems the optimal range for lane widths. © Association for European Transport 2004 6 CONCLUSION Any database, such as the NRA Needs Study Database used in this study, which is continuously updated, results in problems in the analysing of the impact of geometrics on accident rates since the details of the accidents are linked to sections with details that do not describe the actual road section on which the accident occurred. Details of the road sections are constantly updated, with Traffic Volume and Carriageway Type details, etc. of upgraded sections removed as the sections are physically altered. Care must be taken to ensure that this risk is minimised and information is checked thoroughly. Further work needs to be carried out to determine accident reporting rates so that more reliable accident statistics can be calculated. Also, the recording of accident details such as the direction of travel and accident type (e.g. opposite direction) would be of considerable use. International comparisons require the standardisation of accident reporting and the formalisation of principal definitions as well as a more detailed breakdown of vehicle kilometres travelled and lengths of sections used. Table 8 show a comparison of road safety when road types are compared with motorways, using the results of Study 2. It can be seen that in general Motorways are 5 times safer than Undivided Two-Lane Roads and 3 times safer than Dual Carriageways and Three-Lane Undivided roads. Looking at Fatal accidents it can be seen that Motorways are 7 times safer than Undivided Two-Lane Roads and 3 times safer than Dual Carriageways but only 8 times safer than Three-Lane Undivided roads. It is clear that road safety is improved as the road is upgraded. All Accidents Motorway 2 Lane 1 1 1 Dual Carriageway 2 Lane 3 3 3 Three-Lane Undivided 8 8 3 All 2-Lane Undivided 7 9 5 Improved Wide 2 Lane 6 8 4 Improved Standard 2 Lane 9 8 4 Unimproved 2 Lane 6 9 5 Table 8 – Comparison of safety to Motorways (Study 2) Carriage Type Fatal KSI A detailed examination into the impacts of lane width and shoulder width on accidents shows that lane widths of 3.25 to 3.50m should be avoided on undivided roads while a lane width from 3.50 to 3.75m is optimal for safety. To minimise accidents on undivided roads, the hard shoulder width should be in the range 2.50 to 3.0m. It was also found that lane width and road type (e.g. motorway) were the most influential factors on accident rates. Hard shoulder presence reduces the rate of All Accidents on Two-lane Undivided roads. The results were similar to those obtained by Rohloff (2000), Hedman (1990), Hearne (1976), and the Transportation Research Board (1978). © Association for European Transport 2004 Bibliography Berns, S and Bruhning, E. (1998) Definitions and Data Availability: Compilation and Evaluation of A-level Roads and Hospitalised Victims in OECD Countries and Accident and Injury Definitions, Operational Committee of IRTAD and Federal Highway Research Institute (BASt) Special Report: June 1998. Design Manual for Roads and Bridges (2004) Volume 13, Economic Assessment of Road Schemes, COBA 11, DMRB, UK. EuroRAP (2002) Key Findings, European Road Assessment Programme. Website: http://www.eurorap.org/inside.htm Grime, G, (1987) Handbook of Road Safety Research, Butterworth & Co. Ltd, U.K. Hearne, R. (1976) Selected Geometric Elements and Accident Densities on the National Network, The Irish National Institute for Physical planning and Construction Research, Ireland. Hedman, K. O., (1990) Road Design and Safety, Proceedings of Strategic Highway Research Program and Traffic Safety on Two Continents. VTI Report 315A. Gothenburg,1989. HOPES (Horizontal Project for the Evaluation of Safety) (1993) Drive II Report, Commission of the European Communities, DRIVE II, 1991-95 IRTAD (2002) Selected Risk Values for the Year 2002, Website: http://www.bast.de/htdocs/fachthemen/irtad//english/we2.html Monash University (2004) Comparison of International Fatality Rates, Australia, Website: http://www.general.monash.edu.au/muarc/fatals/fatals.htm National Roads Authority (2004) Guidelines for Cost Benefit Analysis – COBA 11, National Roads Authority, Ireland. National Roads Authority (2002) High Accident Locations: 1996-2000; The Inter-urban National Route Network RS 461, National Roads Authority, Ireland. National Roads Authority (1998) National Road Needs Study National Roads Authority, Ireland. National Roads Authority (1998) National Road Needs Study- Technical Report, National Roads Authority, Ireland. National Roads Authority (2001) Road Accident Facts Ireland, 2000 Edition, RS 462, National Roads Authority, Ireland. © Association for European Transport 2004 Rohloff, Michael (2000), Re-use of Hard-Shoulders at Federal Motorways, Proceedings 2nd International Symposium on Highway Geometric Design Mainz, Germany. Transportation Research Board (1978). Cost and Safety Effectiveness of Highway Design Elements, National Co-operative Highway Research Programme Report (NCHRPT) 197, UK . © Association for European Transport 2004 Notes 1 Acknowledgement: The University College Cork project was part funded and supported by the National Roads Authority, particular assistance was given by F. Crowley and F. Trace. 2 Accident definitions: Slight/Minor Injury – no death or serious injury occurs, injuries occur such as sprains, bruises etc. Serious Injury - no deaths but serious injury occurs. A serious injury is defined as an injury for which a person is detained as an ‘inpatient’ in hospital or where any of the following occurred; fractures, concussion, internal injuries, crushing, sever cuts or lacerations, severe shock requiring medical treatment. Death/Fatal – involves one or more fatalities as a result of the accident and death occurs within 30 days. Killed or Seriously Injured (KSI) – involves one or more fatalities or one or more serious injuries. 3 Urban sections are defined as sections with a speed limit of 40mph (64kmph) or less, Interurban sections as those where the speed limit exceeds 40mph. 4 Improved Wide Two-Lane roads are defined as having a carriageway width of greater than 4.0metres and a hard shoulder width of greater than 1.0metre. 5 Improved Standard Two-Lane roads are defined as having a carriageway width of less than or equal to 4.0metres and a hard shoulder width of greater than 1.0metre. 6 Unimproved Two-Lane roads are defined as having a hard shoulder width of less than or equal to 1.0metre. © Association for European Transport 2004