Lower Bound HPMVs Analysis of Pavement Impacts

Transcription

1 Lower Bound HPMVs Analysis of Pavement Impacts

2 Lower Bound HPMVs Analysis of Pavement Impacts Prepared By Adele Jones Asset Manager - Infrastructure Opus International Consultants Ltd Napier Office Opus House, 6 Ossian Street Private Bag 6019, Hawkes Bay Mail Centre, Napier 4142 New Zealand Reviewed By Telephone: William Gray Facsimile: Service Excellence Leader - Central Region Date: 29 April 2013 Reference: Status: FINAL (Version 5) Approved for Release By Trent Downing Work Group Manager - Information Management Opus International Consultants Ltd 2012

3 Contents 1 Executive Summary Introduction Methodology Inputs Required from Other Work Streams Methodology Loading Impact Assessment TERNZ LB HPMV Pro-forma Design Inputs WiM Data Inputs ESA Calculation Spreadsheet Assumptions Loading Impact Summary Review of NZ Pavement Strengths Measure of Pavement Strength CAPTIF Research of ESA and Pavement Strength Relationship State Highways Pavement Strength Local Authority (LA) Roads Pavement Strength Pavement Strength Summary Pavement Effects Loading Impact on Pavements Loading and Pavement Assumptions Original VDM Methodology Literature Review of Pavement Effects Pavement Effects Summary Conclusions and Recommendations Loading Impact Pavement Strength Analysis Pavement Effects General Recommendations References Acknowledgements Appendix A ESA Calculation Spreadsheets for WiM sites (n=4) Appendix B CAPTIF Research of Equivalent Standard Axles and Pavement Strength Relationship December 2012 i

4 Appendix C Loading Effects on Pavement Design December 2012 ii

5 1 Executive Summary The NZ Transport Agency (NZTA) has introduced the concept of a Lower Bound High Productivity Motor Vehicles (LB HPMV), which will result in increased freight productivity while having minor or no impact on roading infrastructure in terms of load limits on structures and impact on pavements. The base assumption is that because the LB HPMVs will be carting the same overall freight task, the overall number of trips will reduce, potentially resulting in less heavy vehicles on the road. The purpose of this report is to review the new LB HPMV proforma vehicles and assess whether the loading impact on the pavement is neutral when compared with the existing heavy vehicle traffic fleet. It also provides an assessment of the pavement effects based on the loading impact outcomes. The impact of the addition of LB HPMVs has been assessed using the Equivalent Standard Axle (ESA) 4 th power law, which associates pavement wear with distress caused by vertical loads. The latest Weigh in Motion (WiM) data from five sites on state highways around New Zealand was used as the base traffic fleet mix and compared with a fleet mix including LB HPMVs. Using this approach, findings show that there is a slight reduction in overall ESA loading for the 50 tonne LB HPMV, based on assessed industry Base Case take-up. This confirms that the addition of LB HPMVs to the existing fleet mix produces a neutral impact in terms of pavement loading, using this approach. Overloading above 50 tonnes (up to 53 tonnes) was also reviewed using the same method and findings show there is a small loading increase, based on assessed industry Base Case take-up. However, due to the strict penalties imposed on HPMV permit holders there is unlikely to be any significant overloading by LB HPMV operators. It is important to note that the use of a blanket percentage change in ESA loading based on WiM site traffic data is not necessarily the best way to represent the loading impact across all roads. It is unlikely that all roads will get the same change in loading. The take-up forecast shows that most of the take-up will be on urban and line haul routes (75% take-up), with only approximately 20% take-up likely on rural local roads. Therefore, the loading impact assessment included in this report could be considered the upper bound of impact for many local authority roads. An assessment of pavement strengths (SNP) across New Zealand showed that pavement strengths for roads with higher traffic volumes (Average Daily Traffic > 4,000 vehicles per day) are generally higher, indicating that these roads are less likely to be impacted by changes in loading than lower trafficked roads. The local soils and geology also affect the ability of pavements to carry traffic loading. Based on SNP, approximately 20% of state highway pavements and 30% of LA road pavements are characterised as weaker (SNP < 2.4) and more vulnerable to any increase in pavement loading. The overall risk of increased pavement deterioration as a result of LB HPMVs is assessed to be low. As the impact of the LB HPMVs was confirmed to be neutral using the 4 th power law approach and assessed Base Case take-up, theoretically there will be no resulting pavement impact in terms of rutting in the subgrade. Dynamic loading impacts resulting in April

6 shear failure and pavement surface damage have not been quantified but are unlikely to be significant. Indications are that the areas where take-up of LB HPMVs is most likely are urban and line haul routes. These generally encompass the more highly trafficked stronger pavements (i.e. state highways), which are less susceptible to changes in loading. However, both state highway and LA road impacts will be more dependent on localised conditions. There are parts of all networks that are vulnerable to the any loading change due to soft subgrades, poor quality pavement materials and road alignment. If the take-up significantly increases from that assessed, it is possible that weaker pavements (SNP < 2.4) may be more susceptible to the LB HPMV loading. The risk of the take-up being higher than assessed is low. The impacts of any change in take-up would need to be assessed against the productivity gains. It is recommended that a further review be completed on the outcomes of a number of applicable NZTA and Austroads research projects that are currently being completed, to determine any applicable outcome in terms of LB HPMVs impact on pavements and surfacings. April

7 2 Introduction The Land Transport Rule: Vehicle Dimensions and Mass Amendment 2010 (VDM Rule Amendment), allows for High Productivity Motor Vehicles (HPMVs) to operate under permit at weights and lengths greater than previously allowed, on approved roads within New Zealand. The purpose of this change was to improve freight efficiency across the country by achieving fewer trips to move the existing freight task, potentially resulting in less heavy vehicles on the road. Although a significant number of vehicles are now operating under such permits, only limited routes have been opened up for HPMV use due to capacity issues with weak structures and pavements. Therefore, the NZ Transport Agency (NZTA) has introduced the concept of a Lower Bound HPMV (LB HPMV), which will result in increased freight productivity while having minor or no impact on roading infrastructure in terms of load limits on structures and impact on pavements. The NZTA has proposed that LB HPMVs can be achieved with modifications to the existing fleet and the introduction of a new proforma design for vehicle mass and length. It is also proposed that these vehicles will be allowed general rather than restricted access across the network. However, initially these vehicles would be Permitted and a review of any impacts completed at a later time (maybe up to five years) prior to any change in regulations. The new LB HPMV proforma vehicle designs must comply with a revised bridge formula which is an extrapolation of the existing Class 1 bridge formula for weights above 44 tonnes and they must not generate any more pavement wear than the existing standard vehicles that they will replace, for the same freight task (i.e. individual HPMVs may have higher impact per vehicle but fewer trips will be needed to carry the same freight). The purpose of this report is to review the new LB HPMV proforma vehicles and assess whether the loading impact on the pavement is neutral when compared with the existing heavy vehicle traffic fleet, carrying the same total freight task. It also provides an assessment of the pavement effects based on the loading impact outcomes. 3 Methodology This methodology covers the requirements of Work Stream 2 Analysis of Pavements included in the scoping document NZTA s Preparation for the Introduction of Lower Bound HPMV. 3.1 Inputs Required from Other Work Streams Transport Engineering Research New Zealand Limited (TERNZ) was commissioned by the NZTA to complete Work Stream 3 to develop a new proforma design for a LB HPMV that will conform to the requirements set out the VDM Rule Amendment. The outcomes from TERNZ s review provide a significant input into this work stream. The objective for the new LB HPMV configurations was to produce an Equivalent Standard Axle (ESA) per tonne of April

8 payload the same or less than the current vehicle fleet at their maximum loading allowance when using the 4 th power rule, thus producing a neutral impact on pavements. Stimpson & Co have been commissioned by the NZTA to complete Work Stream 4 to complete an economic analysis including determining the level of take-up by the road transport industry. The outcomes from this assessment provide an input into the loading impact assessment completed as part of Work Stream Methodology The original Work Stream 2 methodology submitted to and approved by the NZTA, was based around the methodology for assessing additional pavement costs (VDM Methodology) 1 resulting from Opus International Consultants 2010 report VDM Rule Amendment Impact on State Highway Pavements. This was to provide consistency in reviewing the pavement impacts of LB HPMV against previous analysis completed on full HPMV pavement impacts. However, during the completion of this project, this methodology has been modified in agreement with NZTA and the final methodology used is outlined below. Confirming loading impact Using the latest Weigh in Motion (WiM) data from the five WiM sites in New Zealand and the ESA calculator spreadsheet, confirm that the loading impact is neutral in the 4 th power law case. This has been considered for both a nominal 50 tonne LB HPMV and an overloaded (up to 6% above nominal 50 tonne) LB HPMV scenario. This has been completed for three industry take-up scenario outcomes from Work Stream 4. Data Requirements: NZTA to supply national WiM data Output: ESA calculator spreadsheets for each of the five WIM sites. Review of NZ pavements strengths A review of New Zealand pavement strengths is to be completed in order to make an assessment of the weaker pavements that may be more susceptible to any increase in loading. For the purpose of assessing the strengths of pavements across New Zealand, this report uses the Adjusted Structural Number (SNP). Information on NZ State Highways is held within NZTA s State Highway RAMM database. From this information a review of the pavement strength characteristics (based on SNP) of the state highway pavements is completed. From this we can assess the length of highway which may be impacted by loading changes due to weak pavements. For many local authority roads there is no pavement strength data held in RAMM therefore, the pavement strength for LA roads has been reviewed using two methods: 1 Hunter, E & Patrick, J (May 2010). Vehicle Dimension and Mass Amendment 2012 Methodology for Assessing Additional Pavement Costs from HPMV Loading on an Approved Route. Opus International Consultants Ltd, Napier. April

9 Using data collected at the LA Long Term Pavement Performance (LTPP) sites monitored by NZTA. There are 84 sites across 21 LAs. This provides a sample of the full local road network across the country, for which data is consistently collected and recorded. Pavement strength data from RAMM has been obtained from a number of LAs for which Opus has access to RAMM databases. The LAs included in this review provide a reasonably representative cross section of LAs throughout New Zealand, with a variety of different traffic volume and geological characteristics. These methods have been used to determine the pavement strength characteristics across the LA road network and to assess the percentage of LA pavements which may be impacted by loading changes due to weak pavements. Data Requirements: NZTA to supply access to NZTA State Highway RAMM database and local authority LTPP site data for pavement strength Output: An assessment of the pavement strengths of state highway and local authority roads. What are the pavement effects? From the VDM methodology, there are a number of pavement and surfacing factors which may be impacted by increased HPMV loadings. These include: Planned maintenance Reactive maintenance Pavement and surfacing design changes Vulnerable areas high risk curves and intersections All of the above factors are mostly dependent on the axle loadings. If we can show that LB HPMVs have a neutral loading impact for the pavement, then for the 4 th power law it is likely that these vehicles will only impact the pavement and surfacing in vulnerable areas. The impact on vulnerable areas will be dependent on the location of the additional axle on each Lower Bound HPMV. Feedback from the consultants completing Work Stream 3 will be required to indicate the best configurations for minimising the impact on vulnerable areas. It should be noted that the vulnerable areas impact assessment from the original methodology is based on practitioners knowledge and is very dependent on individual vulnerable area site conditions. Therefore, a literature review of dynamic loading effects of changed loading and configurations that contribute to shear failure and pavement surface damage has also been carried out. In particular, this looks at the impact of changing a tandem axle to a tridem axle in the LB HPMV proforma designs. Output: Confirmation that axle loading impacts are neutral or otherwise for Lower Bound HPMV. April

10 4 Loading Impact Assessment 4.1 TERNZ LB HPMV Pro-forma Design Inputs Transport Engineering Research New Zealand Limited (TERNZ) has been commissioned by the NZTA to complete Work stream 3 to develop a new proforma design for a LB HPMV that will conform to the requirements set out the VDM Rule Amendment. The TERNZ draft report 2 concludes that there are only two vehicle configurations that have the axle group weight capacity to allow additional gross weight. These are the truck and trailer and the B-train. The LB HPMV pro-forma designs developed in TERNZ s report are based on the existing pro-forma HPMV designs. For both the truck-trailer and the B-train it is possible to increase the Gross Combination Weight (GCW) to 50 tonne using a longer vehicle (approx. 22.3m), without increasing pavement wear using the R22T23 and B1233 combinations. It was assumed that the pattern of overloading for the LB HPMVs will be similar to that for existing vehicles. Thus the R22T23 vehicle is assumed to have a GCW of tonnes and the B1233 is assumed to have a GCW of tonne. The current pro-forma design for the truck and trailer (R22T22) is shown in Figure 1. To make it a valid LB HPMV requires the following in addition: The rear axle group on the trailer must be a tridem group (making it an R22T23). Other axle groups may be tridem group but this is not a requirement. For 50t GCW, the distance from the first-to-last axle must be a minimum of 20m, for 49t it must be a minimum of m, for 48t a minimum of 18.75m, for 47t a minimum of m and for 46t a minimum of 17.5m. All other axle combinations must be checked for compliance with the bridge formulae and axle group weight limits must be specified such that it is not possible to exceed the bridge formula while complying with the axle group limits. Figure 1 Current 22.3m pro-forma truck and trailer (R22T22) 2 de Pont, J. (June 2012). Lower Bound HPMVs Vehicle Configurations (draft report). TERNZ Ltd. April

11 There are current three pro-forma B-train (B1233 and B1232) designs as shown in Figure 2 - Figure 4. The LB HPMV pro-forma could use the dimensional envelopes of any of these three designs with the following additional conditions: The trailer axle groups must be tridems For 50t GCW, the distance from the first-to-last axle must be a minimum of 20m, for 49t it must be a minimum of m, for 48t a minimum of 18.75m, for 47t a minimum of m and for 46t a minimum of 17.5m. All other axle combinations must be checked for compliance with the bridge and axle group weight limits must be specified such that it is not possible to exceed the bridge formula while complying with the axle group limits. Figure 2 22m pro-forma B-train (B1233) Figure m pro-forma B-train with 5.68m tractor (B1232) April

12 Figure m pro-forma B-train with 5.70m tractor (B1232) The outcomes from the TERNZ draft report that have been used in this review of loading impact are summarised in Table 1. Table 1 Profroma LB HPMV ESAs Vehicle Configuration Load State Average Weight (tonnes) Axle Gp Axle Gp Axle Gp Axle Gp GCW R22T23 Laden R22T23 Tare B1233 Laden B1233 Tare If we limit the laden weight to a maximum of 50 tonnes, the laden ESA for the R22T23 becomes 3.22 and the laden ESA for the B1233 becomes WiM Data Inputs There are six WiM sites in New Zealand collecting axle loading data for use nationally in traffic monitoring. These are all located on State Highways as follows: State Highway 1 at Drury near Auckland State Highway 2 at Te Puke in the Bay of Plenty State Highway 1 at Tokoroa in South Waikato State Highway 35 near Gisborne State Highway 5 at Eskdale in the Hawke s Bay State Highway 1 at Waipara in Canterbury The Hamamanaua WiM site on SH35 in the Gisborne region is the latest WiM site to be introduced, data collection started in November The collected data was not included in the most recent WiM report published in April The WiM data used for this loading ESA April

13 impact review is the annual data provided for the 2011 year; therefore it excludes the SH35 WiM site in Gisborne. All sites are continuously collecting individual vehicle records, and statistics normally downloaded weekly. The loading impact review was completed using the full fleet mix included in the WiM data and has been completed as a separate review for each of the five WiM sites included. This permits the impact of any varying loading effects across the country to be assessed. It should be noted that there are a number of limitations in using the WiM data. These are as follows: The data is from sites which are all on generally higher trafficked rural State Highways and may not necessarily be representative of the traffic mix across all New Zealand roads. The data provided has an accuracy tolerance of ±10% for gross loads and ±15% for axle group loads. The data does not separately identify permitted overweight vehicles. This means that information on existing HPMVs will be contained within the WiM data. It should also be noted that the 2011 WiM data contained significant portions of the heavy vehicle fleet that were overweight (i.e. total gross weight greater than 44 tonne). The data cannot distinguish between single and dual tyres. It is assumed that steer axles are single tyred and all others are dual tyred. Therefore, any subsequent calculation of ESAs will be based on assumed axle group types. The classification used by NZTA in their 2011 summary report and the count data for each of the five sites included in this review is summarised in Table 2. Table 2 Summary of 2011 WiM Data Annual Traffic Counts Type Pat Class Vehicle Configurations Veh Class Drury Tokoroa Te Puke Waipara Eskdale R11 20 o-o (wb m, gw >= 2.5t) MCV R11 21 o--o (wb >3.2m, gw >= 2.5t) MCV R11T1 30 o-o--o HCV R12 31 o--oo HCV R21 34 oo--o HCV A o-o--oo HCV R12 T1 42 o-oo--o HCV R21 T1 44 oo-o--o HCV R22 45 oo--oo HCV April

14 Type Pat Class Vehicle Configurations Veh Class Drury Annual Traffic Counts Tokoroa Te Puke Waipara Eskdale R13 47 o--ooo HCV o-o-o-o-o HCV R12 T11 52 o--oo-o--o HCV A o-oo--oo HCV o--o-----ooo HCV A111 T12 61 o-o--o-o--oo HCV o--oo--o-o-o HCV R12 T12 63 o--oo-o--oo HCV R21 T12 65 oo--o-o--oo HCV R22 T11 66 oo--oo-o--o HCV R22 T2 68 oo--oo--oo HCV A o-oo--ooo HCV A122 T11 74 o-oo--oo-o--o HCV R22 T12 77 oo--oo-o--oo HCV o--o--o MCV o--oo HCV o--o--oo MCV o--oo---o HCV o--oo--oo HCV oo--ooo HCV o--o--oo--o-o HCV A oo-oo--ooo HCV A o--ooo---ooo HCV R12 T22 or B o-oo--oo--oo B-train or T&T HCV oo--o--oo--oo HCV A o-oo-oooo HCV o--oo--oo--ooo HCV A oo-oo--oooo HCV A o--ooo---oooo HCV B o-oo--ooo--oo HCV R22 T oo--oo-oo--oo HCV B oo-oo--ooo-oo HCV R22 T oo-oo--oo-ooo HCV B o-oo-ooo-ooo HCV B oo-oo-ooo-ooo HCV B o-oo-ooo-oooo HCV Total April

15 4.3 ESA Calculation Spreadsheet Assumptions An ESA calculation spreadsheet from the original VDM Methodology has been produced for each of the five WiM sites using the traffic counts included in Table 2. The average ESA for each vehicle type is aggregated up to a total ESA for the existing fleet. This loading is then compared with a revised vehicle fleet mix which includes the new LB HPMV Proforma vehicles. The same ESA calculation process is completed and the two ESA loading impact outcomes are compared to confirm whether the revised traffic fleet mix has any increase in overall loading impact. The spreadsheet allows calculation of this impact for the 4 th power law case. This spreadsheet incorporates a number of assumptions as outlined below. Efficiency gain Each of the LB HPMV vehicles can carry more freight due to increased weight limits and thus to transport the same amount of freight, the overall number of trips will reduce, potentially resulting in less heavy vehicles on the road. This assumption has been incorporated into the spreadsheet. Traffic mix The main state highways carry a full range of commodities most of which will not change as a result of the new LB HPMV loads. Therefore, NZTA s WIM data summarised in Table 2 was used to determine the existing traffic mix, including vehicle types and their weights. Existing Fleet ESA/Vehicle The average ESA per vehicle configuration has been based on those included in the original ESA calculation spreadsheet (calculated from previous WiM data) and provided in the TERNZ report. There are a number of new vehicle configurations included in the 2011 WiM data which were not included in the original ESA calculation spreadsheet. The average ESA/vehicle for these configurations has been estimated based on other similar configurations/vehicle classes. Changes to these estimated values had minimal impact on the overall change in loading calculated in the spreadsheet, as they stay the same for both the existing and new fleet mixes. The spreadsheet used to calculate the increase in ESA is based on WIM data where the existing traffic loading in ESA takes into account unloaded, partially loaded, and fully loaded truck travel to determine an existing average ESA per heavy vehicle. Calculating the new total ESA for the road network incorporating the new LB HMPV vehicles is detailed in the formula below: New Total ESA = (current average ESA per vehicle)*(number of vehicles that have not changed to the new LB HMPV plus the number of unloaded trips of the LB HMPV) + (new fully loaded ESA per LB HMPV)*(number of fully loaded LB HMPV vehicles) The current average ESA per vehicle in the equation above has been reduced in value to take account of the reduction in number of fully loaded vehicles in the existing fleet that have not changed to LB HPMV. Percentage take-up The percentage take-up to the new LB HPMV loading has a direct impact on the increase in pavement loading. The overall percentage take-up for all LB HPMVs (R22T23 and B1233) has been based on the outcomes of the Stimpson Business April

16 Case, November This presents three loading take-up scenarios: Base Case takeup is 52%, Pessimistic Case is 17% and Optimistic Case is 66%. The total take-up is estimated to be over five years, however for the purposes of this loading impact review the total take-up for each case has been used. The take-up forecasts show that most of the take-up will be by non-rural and line haul vehicles (75% take-up for the Base Case ), with only limited take-up likely on rural local roads (20% take-up for the Base Case ). Including all three take-up scenarios allows for review of the sensitivity of loading change based on a change in take-up. New maximum allowable weights The ESA calculation spreadsheet assumes that those existing vehicles that are near their maximum weight will choose to adopt the new HPMV limits. The first loading scenario used a maximum allowable weight for LB HPMV s of 50 tonnes. A second scenario was calculated where the LB HPMVs were assumed to be approximately 6% heavier than the new mass limits. 4.4 Loading Impact Summary Table 3 summarises the loading impact outputs from the ESA calculation spreadsheets for each of the five WiM sites reviewed for the three industry take-up scenarios ( Pessimistic Case 17%, Base Case 52% and Optimistic Case 66%). It also shows the impact of the LB HPMV nominal gross weight of 50 tonne as well as the overloaded scenario increased by 6% as discussed above. The spreadsheets for each WiM site (50t, 52% Base Case take-up scenario) are included in Appendix A. Table 3 Summary of ESA Calculation Spreadsheets for WiM Sites WiM Site 17% takeup Nominal 50t LB HPMV 52% takeup 66% takeup 6% Overloaded LB HPMV 17% takeup 52% takeup 66% takeup Drury -5.2% -1.0% 0.7% -4.1% 2.6% 5.3% Tokoroa -7.9% -1.6% 0.9% -6.1% 3.8% 7.8% Te Puke -6.4% -1.4% 0.7% -5.0% 3.0% 6.2% Eskdale -7.7% -1.6% 0.8% -6.0% 3.6% 7.5% Waipara -7.7% -1.5% 1.1% -5.9% 4.2% 8.2% Table 3 shows that for the 52% Base Case take-up there is actually a slight reduction in loading across all WiM sites for the 50t LB HPMV. There is a very minor increase in loading for the 66% Optimistic Case. This confirms that the addition of LB HPMVs to the existing fleet mix produces a neutral impact in terms of pavement loading, based on the 4 th power law approach. For the overloaded case, there is a small loading increase for 52% take-up, and a further increase for the 66% take-up scenario, which demonstrates a potential impact if operators do not conform to the new proforma LB HPMV weight limits under higher take-up scenarios. 3 Appendix Two - Stimpson, D. (27 November 2012). Business Case for Lower Bound High Productivity Motor Vehicles. Stimpson & Co, Wellington. April

17 It is worth noting at this point that more rigorous overloading implications exist for HPMVs than for regular Class 1 vehicles operating at 44 tonnes or less. In terms of the Land Transport (Offences and Penalties) Regulations 1999, there is a tolerance of up to 1.5 tonnes for any weight recorded or calculated where the legal maximum weight exceeds 33 tonnes but does not exceed 60 tonnes. Overloading above this tolerance results in an infringement fine (a maximum of $10,000 for up to 13,000kg exceedence). A higher mass HPMV will have the administrative concessionary enforcement tolerance applied, which is 300kg on a front axle, 500kg on any other axle, axle group or gross. If any of these concessionary tolerances are exceeded, the permit is voided and standard vehicle enforcement practices will apply, including infringement fees. This supports better control of overloading for the LB HPMV case, and therefore it is probable that there will be minimal impacts from overloading. April

18 5 Review of NZ Pavement Strengths 5.1 Measure of Pavement Strength With reference to Cenek et al (2011), the pavement is a semi-infinite continuum comprising layers of materials with often greatly differing properties and behaviour under load. Also, light loads have a shallower influence than heavy loads. Therefore, a uniform basis is required for representing pavement strength. For the purpose of assessing the strengths of pavements across New Zealand, this report uses the Adjusted Structural Number (SNP). The SNP of a section of pavement is a single parameter used to provide a representation of the load-bearing ability of that pavement. The bigger the SNP number the greater the load bearing capacity of the pavement. SNP can be used as an approximate indicator for the capacity or structural life of pavements, provided that: (i) rutting is the governing distress mechanism; (ii) the majority of the rutting occurs in the subgrade rather than the overlying layers; (iii) the treatment length 4 is correctly defined and relates to a uniform sub-section; and, (iv) the appropriate percentile (rather than average) SNP is determined that corresponds to the percentage of road in a terminal condition which would trigger rehabilitation. In reality pavements are subjected to many other distress modes and therefore there are limitations with this method of assessing pavement structural capacity. However, SNP has been used for this review as data is relatively available and it provides a simple method of analysis that can be widely applied. It is also the currently adopted parameter for deterioration modelling in New Zealand and was used as part of the original methodology for assessing the impact of HPMVs on pavements. Other limitations in the approach of using SNP including: SNP had its origin in the AASHO Road Test in the late 1950 s before the advent of analytical methods. Research reported by Stevens et al (2009), showed that the number of ESA to a terminal rutting condition using the Austroads subgrade strain criterion apparently ranges over two or three orders of magnitude for a given SNP value. There are several methods for evaluating structural strength, including SNP and two methods for modified structural number (SNC). Cenek et al (2011) have shown that the variation in structural number with displacement is very similar for all three methods, although SNP appears to give lower values on weak (low structural number) pavements. In New Zealand SNP is typically derived from falling weight deflectometer (FWD) surveys. There is variability and possibly a lack of consistency in terms of timing of 4 A treatment length is a discreet length of pavement with the same condition and age characteristics April

19 FWD testing from year to year. This may result in seasonal variations of pavement strength data. The data also provides a snapshot in time of the pavement in one particular condition (i.e. wet, dry etc). 5.2 CAPTIF Research of ESA and Pavement Strength Relationship An accelerated loading test was undertaken in 2002 at the Canterbury Accelerated pavement Testing Indoor Facility (CAPTIF) to compare the wear generated by different levels of loading (Arnold et al 2005). The pavement consisted of five different segments that were subjected to 1,000,000 load cycles in two parallel wheel paths. The axle load on one wheel path was 8.2 tonnes while the load on the other was 12 tonnes. Key findings from this study can be summarised as follows: The relationships between SNP and pavement life are best when using the lower 10 th percentile value of SNP for the road section of interest. There is a relationship between SNP and the damage law exponent, n, the lower the value of SNP, the higher the damage exponent as shown in Appendix B. This relationship indicates that pavements with an SNP less than 2.4 will have a damage exponent higher than n=4 and where axles are overloaded this will have greater impact. It is understood that new research 5 is currently being completed to further review this relationship for axle loads less than the standard axle. This should be reviewed at some stage in future for applicability to the LB HPMV case. However, for the purposes of this report it is assumed that an SNP less than 2.4 indicates that there will be more impact of increased loading on these pavements. It is also worth noting that based on measured SNP values research by Cenek et al (2011) indicates a lower limit 10 th percentile SNP value with a limiting SNP of 1.8. In terms of FWD derived SNP, an SNP of less than 2 is associated with central deflections greater than 3mm. For NZ pavements with non-volcanic subgrades, deflections greater than 3mm would be uncommon and thus a lower limit SNP value of 1.8 appears reasonable. 5.3 State Highways Pavement Strength There are 10,894km of state highways across New Zealand, which make up approximately 12% of New Zealand s roads but account for around half of the 36 billion kilometres travelled each year. In order to review the strength of these state highway pavements, we have reviewed the SNP data from the State Highway RAMM database. This SNP data is generally back calculated from Falling Weight Deflectometer (FWD) data collected during on-site testing. The data analysed includes the latest data for all treatment lengths across the state highway network, where this is available. Approximately 144,000 data results were used in this review. 5 The relationship between vehicle axle loadings and pavement wear on local roads NZTA Current Research Project RRT6. April

20 Figure 5 shows the location of SNP data across the state highway network, based on the year the data was collected and recorded in RAMM. This shows that some of the data included in this analysis dates back to 1998, but the majority of data is from the last 10 years. It provides a snapshot of the pavement condition at the time of testing, however the pavement may have since deteriorated or been rehabilitated. Because of this, there may be some discrepancy between the results of this analysis and the actual strength of existing state highway pavements at the present time. Figure 5 State Highway SNP data by year recorded A comparison of the SNP data across a variety of different traffic volume scenarios has been achieved by reviewing the data against the old National State Highway Strategy (NSHS) Hierarchy categories which are currently available in RAMM. Although there is now a new classification system for State Highways, the old hierarchy allows us to group the strength of pavements by different traffic volume groups around the country. Table 4 shows the hierarchy as defined in the State Highway National Pavement Condition Report The R2, R3 and R4 roads make up a total of 85% of the network length. April

21 Table 4 NSHS Hierarchy Classification Figures 6 and 7 illustrate the breakdown of these results by NSHS hierarchy. Figure 6 indicates that R2, R3 and R4 roads have the highest frequency of SNP data. For all hierarchies the results are generally well distributed, although the data for motorways is skewed towards the higher end of the SNP scale. This indicates good pavement strength, which is to be expected. Figure 6 Frequency of SNP by NSHS Hierarchy Figure 7 SNP Distribution by NSHS Hierarchy April

22 As an indication of likely weaker pavements, approximately 20% of all state highway SNP results have a value of less than 2.4 so will potentially be more susceptible to loading impact. Figures 8 and 9 show the location of SNP results that are less than or equal to 2.4 compared with SNP results of 2.4 or greater. There is a reasonable spread of results across the state highway network for both cases, although the south island and northland appear to have generally higher SNPs. Figure 8 SNP results 2.4 Figure 9 SNP results > 2.4 As an indication of the lower limit of pavement strength for all state highways, the 10 th percentile SNP value is Individual road hierarchies show that R3 and R4 highways have the weakest pavements with 10 th percentile SNPs of 1.77 and 1.74 respectively, which is to be expected. This compares well with the lower limit SNP values determined in research by Cenek et al (2011). 5.4 Local Authority (LA) Roads Pavement Strength There are 83,200km of LA roads across New Zealand, which make up approximately 78% of New Zealand s roads and are administered by 78 councils. There has been significantly more difficulty in obtaining pavement strength data for LA pavements for a number of reasons including: April

23 Limited access to LA RAMM databases due to sensitivities around obtaining this without having to seek direct approval from LAs SNP data is not collected and/or recorded by all LAs There is a large component of the LA network that is unsealed (38% of the 83,200km nationally) which is generally not tested for pavement strength and is more likely to have pavement strength variation over time Long Term Pavement Performance (LTPP) data There are 84 Long Term Pavement Performance (LTPP) sites across 21 LAs, which are monitored by NZTA. These LTPP sites were chosen to ensure they provided good representation across a range of environments, traffic classes, pavement types/strengths, pavement age/condition, urban/rural and maintenance regime (with or without maintenance). Although the LTPP data provides a limited sample in relation to all LA roads, it has the advantage of providing data which is consistently collected and recorded. An assessment of pavement strengths for the LTPP sites has been completed based on the NSHS road hierarchies (i.e. using the same traffic volume bands) to enable ready comparison with state highway results. Using this information and extrapolating it across the national length of LA roads can give an indication of the pavement strengths across all LA roads. The SNP provided in the LTPP dataset is a representative SNP value for each site and has been derived from back analysis of FWD testing completed on all sites in Individual SNP data readings were not provided in NZTA s LTPP database. The data included in this analysis provides a snapshot of the pavement condition at the time of testing, however the pavement may have since deteriorated or been rehabilitated. Because of this, there may be some discrepancy between the results of this analysis and the actual strength of existing state highway pavements at the present time. The SNP data from the 84 LA LTPP sites is shown in Figure 10. SNP values are lowest for those roads with traffic volumes of less than 1,000 vehicles per day. 30% of these sites have pavement strengths below the indicative SNP of 2.4 which equates to n=4 damage exponent. SNP values for roads with higher traffic volumes are all substantially higher and improve with increasing traffic volume, indicating that roads with higher traffic volumes have more robust design and construction requirements. As an indication of the lower limit of pavement strength for roads with an AADT less than 1,000 vehicles a day, the 10 th percentile SNP value is April

24 Figure 10 SNP Distribution by Traffic Volume for LTPP Sites A straight extrapolation of this data across the national local road length indicates that lower trafficked local roads are most at risk of increased loading impact Individual LA RAMM Data A subset of New Zealand LAs for which we could access RAMM has been reviewed for SNP data. Table 5 shows the LAs included in the review. A number of these did not have any SNP data recorded in their current RAMM database. Limitations of this individual LA data review include: The SNP results have not all been calculated using the same methodology. Some of the methods for calculating SNP in RAMM can give poor approximations. The best method available is the RAMM FWD Pavement Strength method, which is based on Tonkin & Taylor s methodology. This methodology recommends adjusting SNP results based on geology. The SNP data has not been adjusted for subgrade variation factors. Some of the SNP data is based on FWD test points and is relatively detailed, while other data is provided in the format of a representative SNP value per treatment length. Testing has been completed at different times and using different test suppliers. Some LAs complete FWD testing on sites prior to rehabilitation, which may skew results to a lower limit as these are generally weaker pavements so will likely have lower SNP values. It is uncertain whether this is the case for the LAs represented in this review. Most LAs only have SNP results for a limited sample of their network. This is partially because many LAs have a significant portion of unsealed roads, which are generally not tested. April

25 Table 5 Local Authorities RAMM Data Local Authority SNP Data in RAMM Total Network length (km) Percentage of Network Sealed Network Length Represented by SNP data Typical AADT North Island Auckland Area* Yes % ,000+ Basis of SNP Data Representative value for each treatment length Central Hawke s Bay DC Yes % 56% 100-2,000 Individual FWD test points South Waikato DC Yes % 20% Individual FWD test points Wairoa DC Yes % 18% 100-1,000 Individual FWD test points Western Bay of Plenty DC Yes % 100-4,000 Individual FWD test points South Island Ashburton DC No % 0% 100-4,000 N/A Gore DC No % 0% 100-2,000 N/A Mackenzie DC No % 0% 100-1,000 N/A Marlborough DC Yes % 11% 100-5,000 Individual FWD test points Southland DC Yes % 11% 100-2,000 Representative value for each treatment length Timaru DC No % 0% 100-8,000 N/A Waitaki DC No % 0% 100-1,000 N/A *Auckland Area includes Auckland CC, Franklin DC, Manukau CC, Papakura CC, and Waitakere DC. These councils are now amalgamated. SNP data is either based on a representative value for each treatment length or based on individual FWD test points (i.e. a number of values per treatment length) Figure 11 shows the LA roading network areas included in Table 5, where SNP data has been recorded in RAMM. The map identifies the SNP data points, showing geographically the extent of SNP data. As also indicated in Table 5, some networks have limited coverage of SNP data across their network. Note that the Southland network GIS mapping was not available for inclusion at the time of release of this report, therefore the area (Northern Southland) of SNP data included has been indicated in red hatching against the full Southland district in black hatching. April

26 Figure 11 Location of LA SNP data Results from the analysis of these LAs are shown in Figures 12 and 13. These figures show substantially more variability in SNP results across the LAs reviewed than within the LTPP data, as would be expected. This is both in terms of the amount of data obtained and the overall strength of pavements. April

27 Figure 12 Frequency of SNP data for Individual LAs Figure 13 SNP Distribution for Individual LAs Western Bay of Plenty (WBOP), South Waikato and Wairoa have 87%, 75% and 59% of SNP results less than 2.4 respectively and lower limit 10 th percentile values of 1.19, 1.50 and 0.43 respectively. In particular, the distributions for WBOP and South Waikato are heavily weighted to lower SNP values due to the soil types in the region (volcanic ash subgrades giving higher FWD deflections), and the raw SNP data is normally adjusted (increased) for predictive modelling and other analysis purposes. For example, in WBOP for predictive modelling the SNP values typically increase by due to the soil type. Figures 14 and 15 provide a review of the data from all councils grouped together into traffic volume bands. Because of the impact of the volcanic soils in WBOP and South Waikato, this same analysis has been completed with the data from these regions excluded. This is included in Figures 16 and 17. April

28 Figure 14 Frequency of SNP data for all LAs by Traffic Volume Figure 15 SNP Distribution for all LAs by Traffic Volume Figure 16 Frequency of SNP data for all LAs by Traffic Volume (excluding volcanic Subgrades) Figure 17 SNP Distribution for all LAs by Traffic Volume (excluding volcanic Subgrades) Again the SNP values for roads with higher traffic volumes are substantially higher and improve with increasing traffic volume, indicating that roads with higher traffic volumes have more robust design and construction requirements than higher trafficked roads have. What is interesting is that the lowest trafficked roads (ADT < 100 vehicles a day) have better strength than the mid-traffic volume bands (ADT 100-4,000 vehicles per day). This is perhaps because many of these lower trafficked roads are in fact urban streets that have been also been well constructed with longer pavement design lives. With the SNP values from regions with volcanic ash subgrade (Western Bay of Plenty and South Waikato) excluded, the distributions for roads with ADT 100 to 4,000 show a higher pavement strength and a similar overall distribution to other traffic volumes. This comparison emphasises the need to ensure that the SNP data being reviewed is a good overall representation of the pavement strength. In taking the RAMM data at face value, the results are skewed towards lower strength pavements. April

29 For all LA approximately 45% of all SNP results are less than 2.4 and the lower limit 10 th percentile is For the LAs excluding those with volcanic ash subgrade 28% of all SNP values are less than 2.4 and the lower limit 10 th percentile value is This correlates with the LTPP data review which shows that 30% of pavements have SNP values less than 2.4. Overall, this review of a selection of LAs reflects the fact that local soils and geology can play a significant part in the ability of pavements to carry loading. It also confirms the variability of results and the limited ability to create a one size fits all solution for increased pavement loading, based on such a limited dataset. However, it does indicate that higher trafficked local roads are stronger than lower trafficked roads, and the roads most at risk of loading impact are those with an ADT 100 to 4,000 vehicles a day. This does not directly correlate with the outcomes of the LA LTPP data review. Both analyses show that higher volume roads are unlikely to be affected by increased loading, but there is some disparity in the conclusions drawn over the rest of the traffic volume spectrum. 5.5 Pavement Strength Summary The State Highway RAMM database included SNP data across the SH network. Although some of the data dated back to 1998 and may not wholly reflect the current pavement strength, it provides a reasonable indication of the strength of existing SH pavements. The data analysed by road hierarchies shows generally well distributed results and good pavement strength. However, results indicate that up to 20% of state highway pavements may be more susceptible to any increased loading, based on approximately 21% of all state highway SNP results having a value of less than 2.4. The weakest pavements are on lower trafficked R3 and R4 roads. Pavement strength data for Local Authority (LA) roads was more difficult to obtain and results were variable. However, indications are that higher trafficked roads are generally significantly stronger than the lower trafficked roads. Results indicate that up to 30% of LA pavements may be more susceptible to any increased loading, based on approximately 28-30% of all LA SNP results (excluding volcanic subgrades) having a value of less than 2.4. April

30 6 Pavement Effects 6.1 Loading Impact on Pavements The loading impact assessment completed in Section 4 has used the traditional approach to pavement design in New Zealand, which is based on the number of Equivalent Standard Axles (ESA) using the 4 th power law. The key assumption is that any axle group that causes the same maximum surface deflection as a Standard Axle causes the same damage as the Standard Axle. Therefore, this loading impact only associates pavement wear with distress caused by vertical loads. This method of assessing pavement impact does not take account the following: Dynamic loading effects (i.e. surface deflection measurements are static). The viscoelastic nature of some pavement materials, as well as the development of pore pressures within granular and natural soil layers would be expected to be affected by the loading and unloading speed. The changing performance of the material layers in the pavement (i.e. it treats the pavement as a single entity) under loading. Therefore, the impact of the LB HPMVs on pavements should be assessed based on a number of pavement and surfacing deterioration factors as outlined below: (i) Rutting of the Subgrade resulting from increased vertical loading, assessed based on ESA loading. (ii) Shear Failure occurring in the near surface layers of the pavement, which is impacted by dynamic loading. (iii) Pavement Surface Damage mainly caused by tyre scuffing forces. This may contribute to shear failure where water proofing of the pavement is reduced by scuffing of the surfacing. 6.2 Loading and Pavement Assumptions Based on the analysis completed in Sections 4 and 5, the following assumptions are used in this pavement impact review: The significant change for existing vehicle configurations to become proforma LB HPMV configurations is the rear tandem axle set changing to a tridem axle set (i.e. B1232 becomes B1233 and R22T22 becomes and R22T23). Based on assessed Base Case take-up and the assumption of no change to the total freight task, there is no increase in ESA loading on pavements for the new vehicle fleet including 50t LB HPMVs. Because the loading impact based on the 4 th power law is neutral based on assessed Base Case take-up, this indicates that rutting of the subgrade will not be impacted. April

31 Therefore, the assessment of pavement impacts below is more focussed on shear failure and pavement surface damage. The majority of pavements in New Zealand are granular pavements with thin bituminous surfacing, with the exception of highly trafficked pavements (e.g. motorways) which tend to be structural asphaltic concrete. Based on findings from Section 5 (snapshot of SNPs around NZ), approximately 20% of state highway pavements and 30% of LA road pavements are characterised as weaker and would be more vulnerable to any potential increase in pavement loading. 6.3 Original VDM Methodology The original VDM Methodology includes a number of pavement and surfacing factors which may be impacted by increased HPMV loadings as follows: Planned maintenance Reactive maintenance Pavement and surfacing design changes Vulnerable areas high risk curves and intersections For maintenance activities associated with rutting, there is no expected pavement effect as a result of LB HPMVs. However, for some factors such as reactive maintenance and vulnerable areas there may also be an impact in terms of shear failure and pavement surface damage. A summary of the assessed pavement impact of LB HPMVs, based on this methodology, is included in Table 6. Table 6 Summary of Pavement Impacts using Original VDM Methodology Pavement Impact Factor Planned maintenance Reactive maintenance Pavement and surfacing design changes Governing Loading Impact Rutting in subgrade (ESA increase = 0) Shear failure Surface damage Rutting in subgrade (ESA increase = 0) Shear failure Surface damage Rutting in subgrade (ESA increase = 0) Areas of Impact Chip seal surfacings may have shortened lives in areas of higher stress (see vulnerable areas below). Weaker pavements (SNP < 2.4) may be more susceptible to shear failure. Based on findings from Section 5 (snapshot of SNPs around NZ) 20% of SHs and 30% of LA roads may have increased reactive maintenance. Increase in edge break resulting from longer vehicles cutting corners on curvilinear alignments. Even in an increased loading situation pavement design changes can be minimal. Austroads 2004 Figure 8.4 Assessed Pavement Impact Nil to small - some resurfacing may need to be advanced. Possible minimal increase Nil to small - some resurfacing may need to be April

32 Pavement Impact Factor Governing Loading Impact Shear failure Surface damage Areas of Impact design chart, shows a 10% increase in ESA loading results in a maximum of 7mm increase of in design pavement depth for granular pavements with thin bituminous surfacings as detailed further in Appendix C. This is negligible when compared to existing design and construction tolerances. Resurfacing options for new pavements may need to be reviewed in areas of higher stress (e.g. intersections). Assessed Pavement Impact changed to more robust solutions. Vulnerable Areas Shear failure Surface damage High stress & high speed curves may have increased planned and reactive maintenance approx 12% of SH network (4434 high risk curves in SH database), no information obtained for LA roads. Intersections with sharp/low speed turning movements may have increased scuffing damage, and increased risk of pavement shear. Possible minimal increase 6.4 Literature Review of Pavement Effects A number of research projects have been completed in recent years, focusing on the impact of changed loading in terms of traditional vertical loading, dynamic loading and scuffing force effects on pavements. The outcomes of some of the more significant research reports applicable to this review of the possible impact of LB HPMVs are discussed below Influence of Multiple Axle Loads on Pavement Performance Austroads Publication No. AP T184/11 outlines interim findings of research that examines the effects of axle group type on pavement performance, focussing on Australasian flexible pavement types. One of the reasons for this research is to assist industry (vehicle designers and operators) in the development of more efficient heavy vehicles which will maximise payload without increasing the wear to established road infrastructure. In general terms, the objective of this research study was to investigate improved methods for assessing the pavement damage caused by different multiple axle group loads, and to develop a framework that can be used to quantify this pavement damage for use in the Austroads flexible pavement design processes. Of particular applicability to the LB HPMV case, was the testing completed on unbound pavements, which sought to assess the deformation performance of a typical unbound April

33 granular pavement and subgrade structure under full-scale accelerated loading. Single (40kN), tandem (60KN and 80kN) and tridem axles (90kN) were run over the pavement. Although the overall deformation was slightly less for the tandem axle (80kN) than tridem axle, no definitive conclusions can be drawn to relate this to the LB HPMV case. Insufficient performance data was collected during the research project to allow the development of new design methods. Austroads have established an additional research project, TT1614 Pavement wear effects of heavy vehicle axle groups, to expand the data collected, and to undertake the analysis required to develop the design framework. This is yet to be reported on and should be reviewed for applicability in New Zealand once published. One question that this research does not appear to address yet is whether the dynamic effects of the LB HPMV caused by an increasing number of tyres per vehicle could generate enhanced pore pressure effects in the near surface basecourse materials. Such effects could shorten the life of pavements, particularly where lower quality basecourse materials are present in the pavement Pavement Surface Damage Caused by Tyre Scuffing Forces Land Transport NZ research report 347 (completed by TERNZ) investigated pavement surface damage resulting from tyre scuffing in locations with tight alignments which require heavy vehicles to complete low-speed turns. This study showed that the amount of scuffing force depends on the axle weight, axle group spread, road curvature (increasing turn radius), the tyre configuration, inflation pressure, the use of self-steering axles, and on the type of vehicle. The pertinent conclusions from this report that particularly impact on LB HPMV outcomes are: Scuffing forces increase with increasing axle weight Scuffing forces increase with increasing axle group spread. When laden to the maximum legal weight limits, tridem axle groups produce higher scuffing forces than tandem axle groups even though the tridem axle groups have less weight per axle. In terms of the LB HPMV configuration changes, the changes from current configurations are shown in Tables 7 and 8. These show that the new LB HPMV configurations (R22T23 and B1233) have heavier axle weights on nearly all axles. Also the rear axle is changed from a tandem axle to a tridem axle. April

34 Table 7 Truck and Trailer Existing and LB HPMV Configurations Vehicle Load Average Weight (tonnes) ESA Configuration State Axle Gp 1 Axle Gp 2 Truck Axle Gp 3 Axle Gp 4 Trailer GCW R22T22 Laden R12T22 Laden R22T23 (LB) Laden Table 8 B-Train Existing and LB HPMV Configurations Vehicle Load Average Weight (tonnes) ESA Configuration State Axle Gp 1 Axle Gp 2 Axle Gp 3 Axle Gp 4 GCW B1222 Laden B1232 Laden B1233 Laden B1233 (LB) Laden Research Report 347 concludes that for the same vehicle configuration scuffing forces are proportional to load. So for the truck-trailers, the trucks are the same configuration but with higher weight so the scuffing forces on the drive axles will increase. For the rear trailers (truck trailers and B-trains) there are two competing effects. The tridem axle group generate more scuffing than the tandem but the axle loads are less which produces less pavement wear. The R22T23 combination was not investigated because none existed at that stage, however a comparison of the B1233 and the B1232 was made. For smaller turn angles, the B1233 actually generated lower peak scuffing forces than the B1232. The crossover point was at about 90 degrees. The indications from this research are that the changes made to LB HPMVs are likely to be relatively neutral in terms of scuffing compared with the existing HCV configurations they will replace. However, depending on how the configurations are actually loaded there may be a small impact on scuffing Relationship Between Vehicle Axle Loadings and Pavement Wear on Local Roads NZTA research is currently being completed to provide reliable evidence on the wear characteristics of New Zealand local road pavements from accelerated pavement loading studies at CAPTIF and validated with field data from the nationwide LTTP sites. The objective of this research is to provide a comprehensive picture of the load wear relationships for New Zealand roads. Previous research here and in Australia has only considered loads above the current legal limit on State Highways. However, that research has indicated that there may be a significantly different relationship on local roads and below the legal limit. This research will fill in the gaps below the legal limit and on local roads. The research outputs will be a Power Law model that has been tested across the full April

35 range of New Zealand roads and a full range of loads used. The results will be published as NZTA research report. At this stage there have been no published findings from this research, however it will have direct applicability to the LB HPMV case and should be reviewed in this context when published Modelling Extreme Traffic Loading Effects NZTA research (Cenek et al, 2011) on modelling extreme traffic loading effects is currently being finalised. The draft report presents findings of a study aimed at establishing whether the pavement deterioration and pavement distress models for roughness, rutting and cracking progression incorporated into pavement management systems, such as NZdTIMS, could be modified so they reliability predict the condition of a pavement after it has been exposed to sudden extreme traffic loading. Although there is some relevance of this research to the review of impacts from LB HPMVs, it is important to note that the impact for the LB HPMV loading scenario is apparently neutral, based on the 4 th power law, so modelling outcomes should be unchanged. However, it may be worth reviewing the outcomes of the final NZTA research report for applicability to the LB HPMV loading scenario. 6.5 Pavement Effects Summary The loading impact assessment using the 4 th power law has provided the basis for the pavement impact in terms of rutting in the subgrade. As the impact of the LB HPMV vehicles was confirmed to be neutral using this approach for the assessed Base Case take-up, theoretically there will be no resulting pavement damage. If the take-up significantly increases, it is possible that weaker pavements (SNP < 2.4) may be more susceptible to the LB HPMV loading. The risk of the take-up being higher than assessed is considered to be low. The impacts of any change in take-up would need to be assessed against the productivity gains. In terms of the impacts of dynamic loading resulting in shear failure and pavement surface damage, there is no conclusive pavement impact. There are parts of all networks that are likely to be more vulnerable to the change due to soft subgrades, poor quality pavement materials and road alignment. High stress curves and intersections may also be impacted due to the change in axle configuration from tandem to tridem resulting in a change in scuffing forces. However, research indicates that the changes made to LB HPMVs are likely to be relatively neutral in terms of scuffing compared with the existing HCV configurations they will replace. This is dependent on the way axles are loaded and the turning angles. April

36 7 Conclusions and Recommendations 7.1 Loading Impact ESA Calculation Spreadsheet Outcomes The ESA calculation spreadsheet used in this analysis provides a simplified approach to determining the loading impact of the addition of LB HPMVs to the existing traffic fleet. There are a number of assumptions built into the spreadsheet which could be further reviewed to provide a more accurate calculation of the loading impact. It is recommended that if the ESA calculation spreadsheet is going to be used for further HPMV loading impact assessments in future, it should be reviewed and updated Industry Take-up The overall assessed industry Base Case take-up (based on Stimpson Business Case, November 2012) has been incorporated into the loading impact calculations. This take-up forecast shows that most of the take-up will be on urban and line haul routes (75% take-up), with only approximately 20% take-up likely on rural local roads. Because of this, it is possible that the loading impact on most local authority roads will be less than assessed based on WiM site traffic data. Therefore, the loading impact assessment included in this report could be considered the upper bound of impact for many local authority roads Percentage Change in Loading As indicated in above, the use of a blanket percentage change in ESA loading based on WiM site traffic data is not necessarily the best way to represent the loading impact across all roads. It is unlikely that all roads will get the same change in loading. Some roads may not have any vehicles change to LB HPMV (e.g. low volume LA roads), while other routes (most likely urban and line haul) may have substantial take-up. A further review at a more detailed regional or network level would provide a more accurate reflection of the actual loading impact on road pavements LB HPMV versus Demand Based Loading Increases It is worth pointing out that the LB HPMV will result in a net change in trafficking based on a change to vehicle configurations. It is assumed that LB HPMVs will be carting the same overall freight task, and therefore overall trips will reduce. They will be replacing existing heavy vehicles and travelling on the same routes. Therefore, the effect of LB HPMVs cannot be compared to the effects of demand based loading changes, which potentially present a far greater loading increase and impact on the pavement. Weaker pavements are very susceptible to increased loading resulting from a change in traffic use characteristics, such as resulting from land use changes (e.g. dairy conversions, forestry harvesting). It is understood that this is a significant issue for many local authorities who are currently experiencing pavement deterioration due to demand based changes. April

37 7.2 Pavement Strength Analysis The pavement strength analysis completed in this report was based on a desk top study of existing data in RAMM databases maintained by NZTA and various LAs as well as LTPP site data. As detailed in Section 5, there were a number of limitations with this approach and the outcomes of this analysis provide an indication only of New Zealand pavement strengths. In particular, the review of LA roads has covered only a small portion of the LA roads across the country and further review and testing of pavement strength at a network level would be required to confirm pavement strengths. Further, as SNP represents vertical loading outcomes (i.e. SNP represents rutting in the subgrade), this review does not necessarily take into account other pavement strength parameters such as individual pavement layer strength. 7.3 Pavement Effects The overall risk of increased pavement deterioration as a result of LB HPMVs is assessed to be low. As the impact of the LB HPMVs was confirmed to be neutral using the 4 th power law approach and assessed Base Case take-up, theoretically there will be no resulting pavement impact in terms of rutting in the subgrade. Dynamic loading impacts resulting in shear failure and pavement surface damage have not been quantified but are unlikely to be significant. Indications are that the areas where take-up of LB HPMVs is most likely are urban and line haul routes. These generally encompass the more highly trafficked stronger pavements (i.e. state highways), which are less susceptible to changes in loading. However, both state highway and LA road impacts will be more dependent on localised conditions. There are parts of all networks that are vulnerable to the any loading change due to soft subgrades, poor quality pavement materials and road alignment. 7.4 General Recommendations Although indications are that LB HPMVs will have a neutral impact on pavements at assessed Base Case take-up, it would be prudent to complete monitoring of reactive maintenance post LB HPMV take-up. NZTA already has a monitoring regime in place since the introduction of HPMVs, which reviews impacts on the LTPP sites across the country, on both State Highway and LA roads. This monitoring would be appropriate to assess the impacts of the LB HPMVs also, including any change in shallow shear pavement repair quantities, increased edge break repairs on curves and scuffing of surfacing in low speed turning environments. This will be particularly important in areas where pavement strength is lower as these areas are most likely to be impacted. It is also recommended that a further review be completed on the outcomes of a number of applicable NZTA and Austroads research projects that are currently being completed as discussed in this report, to determine any applicable outcome in terms of LB HPMVs impact on pavements and surfacings. April

38 8 References Arnold,G., Steven,B., Alabaster,D., Fussell,A. (2005): Effect on Pavement Wear of an Increase in Mass Limits for Heavy Vehicles Stage 3, Land Transport New Zealand Research Report 279, Wellington, New Zealand. Arnold, G.,Steven, B., Alabaster, D., Fussell, A. (2005). Effect on pavement wear of increased mass limits for heavy vehicles concluding report. Land Transport New Zealand Research Report 281. Austroads Ltd. (2011).The Influence of Multiple Axle Loads on Pavement Performance: Interim Findings. Austroads Publication No. AP T184/11. Austroads (2004). Pavement Design - A Guide to the Structural Design of Road Pavements. Austroads, Sydney. Cenek, PD, Henderson, R., McIver, I., Patrick, J. (2011) Modelling of Extreme Traffic Loading Effects. DRAFT NZ Transport Agency research report. de Pont, J. (June 2012). Lower Bound HPMVs Vehicle Configurations (draft report). TERNZ Ltd. Hunter, E & Patrick, J (April 2010). VDM Rule Amendment Impact on State Highway Pavements and Addendum 2 VDM Rule Amendment Impact on State Highway Pavements. Opus International Consultants Ltd, Napier. Hunter, E & Patrick, J (May 2010). Vehicle Dimension and Mass Amendment 2012 Methodology for Assessing Additional Pavement Costs from HPMV Loading on an Approved Route. Opus International Consultants Ltd, Napier. NZ Government. (March 2012). Land Transport (Offences and Penalties) Regulations NZ Transport Agency (April 2012). Annual Weight-in-Motion (WiM) Report Salt G.; Henning T.F.P.; Stevens D.; and Roux D.C. (2010) Rationalisation of the structural capacity definition and quantification of roads based on falling weight deflectometer tests. NZ Transport Agency Research Report no.401. Stevens D.; Salt G.; Henning T.F.P. & Roux D.C. (November 2009). Pavement Performance Prediction: A Comprehensive New Approach to Defining Structural Capacity (SNP). Paper for TRANSIT NZIHT 10th ANNUAL CONFERENCE, Rotorua. Stimpson, D. (27 November 2012). Business Case for Lower Bound High Productivity Motor Vehicles. Stimpson & Co, Wellington. Taramoeroa, N., de Pont, J. (2008). Characterising pavement surface damage caused by tyre scuffing forces. Land Transport New Zealand Research Report 374. April

39 9 Acknowledgements We wish to acknowledge the contribution of Dr Greg Arnold (Technical Manager Pavements, Road Science), in reviewing loading impact analysis, including the ESA calculation spreadsheet outputs, and providing confirmation that new research is currently being completed to review the relationship between vehicle axle loadings (including axle loads less than the standard axle) and pavement wear on local roads. April

40 Appendix A ESA Calculation Spreadsheets for WiM sites (n=4) LB HPMVs 50tonne, 52% Base Case take-up scenario DRURY WiM SITE Type R11 R11 Vehicle WIM Configurations o-o (wb m, 20 gw >= 2.5t) o--o (wb >3.2m, gw 21 >= 2.5t) Vehicles converted to LB HPMV Numb. Passes - for 1 week or more Sum ESAs Average ESA per Vehicle* No. Vehicles Changed to LB HPMV New traffic Fleet (including 50t LB HPMVs) % of the heaviest vehicles changed (55% loaded) % Weight Increase applied New ESA for the heavy vehicles changed % Uptake to HMPV Number of vehicles needed to cart same load MCV % 0% Sum ESAs (all vehicles includes those not changed) MCV % 0% R11T1 30 o-o--o HCV % 0% R12 31 o--oo HCV % 0% R21 34 oo--o HCV % 0% R11 T11 40 o--o-o--o HCV % 0% A o-o--oo HCV % 0% R12 T1 42 o-oo--o HCV % 0% R21 T1 44 oo-o--o HCV % 0% R22 45 oo--oo HCV % 0% R13 47 o--ooo HCV % 0% o-o-o-o-o HCV % 0% R12 T11 52 o--oo-o--o HCV % 0% A o-oo--oo HCV % 0% o--o-----ooo HCV % 0% A111 T12 61 o-o--o-o--oo HCV % 0% o--oo--o-o-o HCV % 0% R12 T12 63 o--oo-o--oo HCV % 0% R21 T12 65 oo--o-o--oo HCV % 0% R22 T11 66 oo--oo-o--o HCV % 0% R22 T2 68 oo--oo--oo HCV % 0% A o-oo--ooo HCV % 0% A122 T11 74 o-oo--oo-o--o HCV % 0% R22 T12 77 oo--oo-o--oo HCV % 0% o--ooo-o--oo HCV % 0% o-oo--oo-o--oo HCV % 0% A123 T11 89 o-oo--ooo-o--o HCV % 0% o--o--o MCV % 0% o--oo HCV % 0% o--o--oo MCV % 0% o--oo---o HCV % 0% o--oo--oo HCV % 0% oo--ooo HCV % 0% A121 T o-oo--o-o--o HCV % 0% o--o--oo--o-o HCV % 0% A oo-oo--ooo HCV % 0% A121 T o-oo--o-o--oo HCV % 0% A o--ooo---ooo HCV % 0% R12 T22 or B1222 o-oo--oo--oo B-train 751 HCV2 or T&T % 0% A122 T2 752 o--oo-oo--oo HCV % 0% oo--o--oo--oo HCV % 0% A o-oo-oooo HCV % 0% o--oo--oo--ooo HCV % 0% A oo-oo--oooo HCV % 0% A o--ooo---oooo HCV % 0% B o-oo--ooo--oo HCV2 LB HMPV % 29% R22 T oo--oo-oo--oo HCV2 LB HMPV % 29% B oo-oo--ooo-oo HCV % 0% R22 T oo-oo--oo-ooo HCV % 0% B o-oo-ooo-ooo HCV % 0% B oo-oo-ooo-ooo HCV % 0% B o-oo-ooo-oooo HCV % 0% B oo-oo-ooo-oooo HCV % 0% Total % ESA Loading increase: *Values in bold itallics are assumed based on similar weight class/configuration vehicles. Changes to these ESA/veh have limited impact on the overall %ESA increase due to low vehicle numbers in these classes Existing Traffic Fleet April

41 Tokoroa WiM SITE Type R11 R11 Vehicle WIM Configurations o-o (wb m, 20 gw >= 2.5t) o--o (wb >3.2m, gw 21 >= 2.5t) Veh converted to LB HPMV Number of Passes - for 1 week or more Sum ESAs Average ESA per Veh* No. Vehicles Changed New traffic Fleet (including 50t LB HPMVs) % of the heaviest vehicles changed (55% loaded) % Weight Increase applied New ESA for the heavy vehicles changed % Uptake to HMPV Number of vehicles needed to cart same load MCV % 0% Sum ESAs (all vehicles includes those not changed) MCV % 0% R11T1 30 o-o--o HCV % 0% R12 31 o--oo HCV % 0% R21 34 oo--o HCV % 0% R11 T11 40 o--o-o--o HCV % 0% A o-o--oo HCV % 0% R12 T1 42 o-oo--o HCV % 0% R21 T1 44 oo-o--o HCV % 0% R22 45 oo--oo HCV % 0% R13 47 o--ooo HCV % 0% o-o-o-o-o HCV % 0% R12 T11 52 o--oo-o--o HCV % 0% A o-oo--oo HCV % 0% o--o-----ooo HCV % 0% A111 T12 61 o-o--o-o--oo HCV % 0% o--oo--o-o-o HCV % 0% R12 T12 63 o--oo-o--oo HCV % 0% R21 T12 65 oo--o-o--oo HCV % 0% R22 T11 66 oo--oo-o--o HCV % 0% R22 T2 68 oo--oo--oo HCV % 0% A o-oo--ooo HCV % 0% A122 T11 74 o-oo--oo-o--o HCV % 0% R22 T12 77 oo--oo-o--oo HCV % 0% o--ooo-o--oo HCV % 0% o-oo--oo-o--oo HCV % 0% A123 T11 89 o-oo--ooo-o--o HCV % 0% o--o--o MCV % 0% o--oo HCV % 0% o--o--oo MCV % 0% o--oo---o HCV % 0% o--oo--oo HCV % 0% oo--ooo HCV % 0% A121 T o-oo--o-o--o HCV % 0% o--o--oo--o-o HCV % 0% A oo-oo--ooo HCV % 0% A121 T o-oo--o-o--oo HCV % 0% A o--ooo---ooo HCV % 0% R12 T22 or B1222 o-oo--oo--oo B-train 751 HCV2 or T&T % 0% A122 T2 752 o--oo-oo--oo HCV % 0% oo--o--oo--oo HCV % 0% A o-oo-oooo HCV % 0% o--oo--oo--ooo HCV % 0% A oo-oo--oooo HCV % 0% A o--ooo---oooo HCV % 0% B o-oo--ooo--oo HCV2 LB HMPV % 29% R22 T oo--oo-oo--oo HCV2 LB HMPV % 29% B oo-oo--ooo-oo HCV % 0% R22 T oo-oo--oo-ooo HCV % 0% B o-oo-ooo-ooo HCV % 0% B oo-oo-ooo-ooo HCV % 0% B o-oo-ooo-oooo HCV % 0% B oo-oo-ooo-oooo HCV % 0% Total SUM % ESA Loading increase: *Values in bold itallics are assumed based on similar weight class/configuration vehicles. Changes to these ESA/veh have limited impact on the overall %ESA increase due to low vehicle numbers in these classes Existing Traffic Fleet April

42 TE PUKE WiM SITE Type R11 R11 Vehicle WIM Configurations o-o (wb m, 20 gw >= 2.5t) o--o (wb >3.2m, gw 21 >= 2.5t) Veh converted to LB HPMV Number of Passes - for 1 week or more Sum ESAs Average ESA per Veh* No. Vehicles Changed % Uptake to HMPV % of the heaviest vehicles changed (55% loaded) % Weight Increase applied New ESA for the heavy vehicles changed Number of vehicles needed to cart same load MCV % 0% Sum ESAs (all vehicles includes those not changed) MCV % 0% R11T1 30 o-o--o HCV % 0% R12 31 o--oo HCV % 0% R21 34 oo--o HCV % 0% R11 T11 40 o--o-o--o HCV % 0% A o-o--oo HCV % 0% R12 T1 42 o-oo--o HCV % 0% R21 T1 44 oo-o--o HCV % 0% R22 45 oo--oo HCV % 0% R13 47 o--ooo HCV % 0% o-o-o-o-o HCV % 0% R12 T11 52 o--oo-o--o HCV % 0% A o-oo--oo HCV % 0% o--o-----ooo HCV % 0% A111 T12 61 o-o--o-o--oo HCV % 0% o--oo--o-o-o HCV % 0% R12 T12 63 o--oo-o--oo HCV % 0% R21 T12 65 oo--o-o--oo HCV % 0% R22 T11 66 oo--oo-o--o HCV % 0% R22 T2 68 oo--oo--oo HCV % 0% A o-oo--ooo HCV % 0% A122 T11 74 o-oo--oo-o--o HCV % 0% R22 T12 77 oo--oo-o--oo HCV % 0% o--ooo-o--oo HCV % 0% o-oo--oo-o--oo HCV % 0% A123 T11 89 o-oo--ooo-o--o HCV % 0% o--o--o MCV % 0% o--oo HCV % 0% o--o--oo MCV % 0% o--oo---o HCV % 0% o--oo--oo HCV % 0% oo--ooo HCV % 0% A121 T o-oo--o-o--o HCV % 0% o--o--oo--o-o HCV % 0% A oo-oo--ooo HCV % 0% A121 T o-oo--o-o--oo HCV % 0% A o--ooo---ooo HCV % 0% R12 T22 or B1222 o-oo--oo--oo B-train 751 HCV2 or T&T % 0% A122 T2 752 o--oo-oo--oo HCV % 0% oo--o--oo--oo HCV % 0% A o-oo-oooo HCV % 0% o--oo--oo--ooo HCV % 0% A oo-oo--oooo HCV % 0% A o--ooo---oooo HCV % 0% B o-oo--ooo--oo HCV2 LB HMPV % 29% R22 T oo--oo-oo--oo HCV2 LB HMPV % 29% B oo-oo--ooo-oo HCV % 0% R22 T oo-oo--oo-ooo HCV % 0% B o-oo-ooo-ooo HCV % 0% B oo-oo-ooo-ooo HCV % 0% B o-oo-ooo-oooo HCV % 0% B oo-oo-ooo-oooo HCV % 0% Total SUM % ESA Loading increase: *Values in bold itallics are assumed based on similar weight class/configuration vehicles. Changes to these ESA/veh have limited impact on the overall %ESA increase due to low vehicle numbers in these classes Existing Traffic Fleet New traffic Fleet (including 50t LB HPMVs) April

43 ESKDALE WiM SITE Type R11 R11 Vehicle WIM Configurations o-o (wb m, 20 gw >= 2.5t) o--o (wb >3.2m, gw 21 >= 2.5t) Veh converted to LB HPMV Number of Passes - for 1 week or more Sum ESAs Average ESA per Vehicle* No. Vehicles Changed New traffic Fleet (including 50t LB HPMVs) % of the heaviest vehicles changed (55% loaded) % Weight Increase applied New ESA for the heavy vehicles changed % Uptake to HMPV Number of vehicles needed to cart same load MCV % 0% Sum ESAs (all vehicles includes those not changed) MCV % 0% R11T1 30 o-o--o HCV % 0% R12 31 o--oo HCV % 0% R21 34 oo--o HCV % 0% R11 T11 40 o--o-o--o HCV % 0% A o-o--oo HCV % 0% R12 T1 42 o-oo--o HCV % 0% R21 T1 44 oo-o--o HCV % 0% R22 45 oo--oo HCV % 0% R13 47 o--ooo HCV % 0% o-o-o-o-o HCV % 0% R12 T11 52 o--oo-o--o HCV % 0% A o-oo--oo HCV % 0% o--o-----ooo HCV % 0% A111 T12 61 o-o--o-o--oo HCV % 0% o--oo--o-o-o HCV % 0% R12 T12 63 o--oo-o--oo HCV % 0% R21 T12 65 oo--o-o--oo HCV % 0% R22 T11 66 oo--oo-o--o HCV % 0% R22 T2 68 oo--oo--oo HCV % 0% A o-oo--ooo HCV % 0% A122 T11 74 o-oo--oo-o--o HCV % 0% R22 T12 77 oo--oo-o--oo HCV % 0% o--ooo-o--oo HCV % 0% o-oo--oo-o--oo HCV % 0% A123 T11 89 o-oo--ooo-o--o HCV % 0% o--o--o MCV % 0% o--oo HCV % 0% o--o--oo MCV % 0% o--oo---o HCV % 0% o--oo--oo HCV % 0% oo--ooo HCV % 0% A121 T o-oo--o-o--o HCV % 0% o--o--oo--o-o HCV % 0% A oo-oo--ooo HCV % 0% A121 T o-oo--o-o--oo HCV % 0% A o--ooo---ooo HCV % 0% R12 T22 or B1222 o-oo--oo--oo B-train 751 HCV2 or T&T % 0% A122 T2 752 o--oo-oo--oo HCV % 0% oo--o--oo--oo HCV % 0% A o-oo-oooo HCV % 0% o--oo--oo--ooo HCV % 0% A oo-oo--oooo HCV % 0% A o--ooo---oooo HCV % 0% B o-oo--ooo--oo HCV2 LB HMPV % 29% R22 T oo--oo-oo--oo HCV2 LB HMPV % 29% B oo-oo--ooo-oo HCV % 0% R22 T oo-oo--oo-ooo HCV % 0% B o-oo-ooo-ooo HCV % 0% B oo-oo-ooo-ooo HCV % 0% B o-oo-ooo-oooo HCV % 0% B oo-oo-ooo-oooo HCV % 0% Total SUM % ESA Loading increase: *Values in bold itallics are assumed based on similar weight class/configuration vehicles. Changes to these ESA/veh have limited impact on the overall %ESA increase due to low vehicle numbers in these classes Existing Traffic Fleet April

44 Waipara WiM SITE Type R11 R11 Vehicle WIM Configurations o-o (wb m, 20 gw >= 2.5t) o--o (wb >3.2m, gw 21 >= 2.5t) Veh converted to LB HPMV Number of Passes - for 1 week or more Sum ESAs Average ESA per Veh* No. Vehicles Changed New traffic Fleet (including 50t LB HPMVs) % of the heaviest vehicles changed (55% loaded) % Weight Increase applied New ESA for the heavy vehicles changed % Uptake to HMPV Number of vehicles needed to cart same load MCV % 0% Sum ESAs (all vehicles includes those not changed) MCV % 0% R11T1 30 o-o--o HCV % 0% R12 31 o--oo HCV % 0% R21 34 oo--o HCV % 0% R11 T11 40 o--o-o--o HCV % 0% A o-o--oo HCV % 0% R12 T1 42 o-oo--o HCV % 0% R21 T1 44 oo-o--o HCV % 0% R22 45 oo--oo HCV % 0% R13 47 o--ooo HCV % 0% o-o-o-o-o HCV % 0% R12 T11 52 o--oo-o--o HCV % 0% A o-oo--oo HCV % 0% o--o-----ooo HCV % 0% A111 T12 61 o-o--o-o--oo HCV % 0% o--oo--o-o-o HCV % 0% R12 T12 63 o--oo-o--oo HCV % 0% R21 T12 65 oo--o-o--oo HCV % 0% R22 T11 66 oo--oo-o--o HCV % 0% R22 T2 68 oo--oo--oo HCV % 0% A o-oo--ooo HCV % 0% A122 T11 74 o-oo--oo-o--o HCV % 0% R22 T12 77 oo--oo-o--oo HCV % 0% o--ooo-o--oo HCV % 0% o-oo--oo-o--oo HCV % 0% A123 T11 89 o-oo--ooo-o--o HCV % 0% o--o--o MCV % 0% o--oo HCV % 0% o--o--oo MCV % 0% o--oo---o HCV % 0% o--oo--oo HCV % 0% oo--ooo HCV % 0% A121 T o-oo--o-o--o HCV % 0% o--o--oo--o-o HCV % 0% A oo-oo--ooo HCV % 0% A121 T o-oo--o-o--oo HCV % 0% A o--ooo---ooo HCV % 0% R12 T22 or B1222 o-oo--oo--oo B-train 751 HCV2 or T&T % 0% A122 T2 752 o--oo-oo--oo HCV % 0% oo--o--oo--oo HCV % 0% A o-oo-oooo HCV % 0% o--oo--oo--ooo HCV % 0% A oo-oo--oooo HCV % 0% A o--ooo---oooo HCV % 0% B o-oo--ooo--oo HCV2 LB HMPV % 29% R22 T oo--oo-oo--oo HCV2 LB HMPV % 29% B oo-oo--ooo-oo HCV % 0% R22 T oo-oo--oo-ooo HCV % 0% B o-oo-ooo-ooo HCV % 0% B oo-oo-ooo-ooo HCV % 0% B o-oo-ooo-oooo HCV % 0% B oo-oo-ooo-oooo HCV % 0% Total SUM % ESA Loading increase: *Values in bold itallics are assumed based on similar weight class/configuration vehicles. Changes to these ESA/veh have limited impact on the overall %ESA increase due to low vehicle numbers in these classes Existing Traffic Fleet April

45 Appendix B CAPTIF Research of Equivalent Standard Axles and Pavement Strength Relationship Road traffic consists of a range of vehicle types, wheels and loads. A method intrinsic in pavement design and deterioration modelling is to combine all traffic into one type. This one type of traffic is commonly referred to as an Equivalent Standard Axle (ESA). The standard axle is defined as a single axle with dual wheels that carries a load of 8.2 tonnes (40kN half dual tyred axle as tested at NZTAs Pavement Test Facility CAPTIF). To calculate the number of ESA for any given traffic distribution, equation (1) is used as given in the Austroads Pavement Design Guide (Austroads, 2004): where, ESA = Axle load Axle load reference ESAs = number of standard axles needed to cause the same damage as one pass of the actual axle load (Axle load, Equation 1). Axle load = actual axle load in kn (or total axle group weight). Axle load reference = reference load depending on the axle load group below. n (1) Axle Group Type Load (kn) Load (kg) Single axle with single tyres (SAST) Single axle with dual tyres (SADT) Tandem axle with dual tyres (TADT) Triaxial with dual tyres (TRDT) Quad-axle with dual tyres (QADT) n = damage law exponent (commonly = 4, although different exponents are used depending on pavement strength, SNP). Research at CAPTIF on the effects of Mass Limits found that the fourth power relationship is not valid for all pavement types. A range of relationships were determined between SNP (determined by two different FWD methods) and the damage law exponent, n. Results showed that the exponent is nearly 1.0 for strong well built pavements (high SNP) and as high as 8.0 for weak pavements (low SNP) indicating their brittle nature (i.e. weak pavements fail quickly if their shear strength is exceeded). Practitioners would agree with this trend as local roads that are weak would fail quickly with the introduction of new vehicles with increased mass limits. Further, the expected trend is that the strong pavements (high SNP) on the busiest state highways would feel little structural impact with the introduction of increased mass limits. April

46 Two methods were used at CAPTIF to calculate the SNP. One method involved a regression equation using FWD deflection data. The regression equation has since changed and therefore the CAPTIF data was revisited to determine SNP using the new equation (2): SNP = 112(D 0 ) (D 0 -D 900 ) (D 0 -D 1500 ) (2) Where deflections are in microns, after standardising to 40 kn plate load and subscripts are offsets in mm from the plate centre. The end of life for the CAPTIF pavements was a rut depth/vertical surface deformation of 15 mm. Based on this criteria and the new equation for calculating SNP a relationship between damage exponent, n and SNP was determined as detailed in Figure 5 6. Although the CAPTIF results match expectations they are from a very limited number of pavement types and should be used cautiously. Damage Exponent, n Old Eqn from CAPTIF y = x R 2 = New Eqn n = 51.62SNP R 2 = SNP Figure 5 SNP versus damage exponent (n) For the ESA damage exponent of 4 the correlating SNP is 2.4, based on the new equation from CAPTIF. This indicates that those pavements with a SNP lower than 2.4, may potentially be more susceptible to loading impact than those with an SNP of 2.4 or more. 6 Opus International Consultants Ltd. VDM Rule Amendment Impact on State Highway Pavements. April 2010, pp 21-22, Figure 4.2.1a. April

47 Appendix C Loading Effects on Pavement Design Granular Overlay Depth Change The common method of determining the granular overlay depth is to assume a design of a new pavement where the: Granular Overlay Depth = Depth of New Pavement minus the Depth of the Old/Existing Pavement. Table D-1 below shows a worst case scenario for the increase in thickness due to a 10% increase in ESA based on the design chart from Austroads as shown above. As the percentage increase in ESAs is the same only the subgrade CBR affects the increase in depth required. Subgrade CBR Existing ESA New LB HPMV ESA Depth Required (Existing) (mm) Depth Required (LB HPMV) (mm) Increase in Depth (mm) E E E E E E E E E E E E E E April