NRA Pavement Cost Model manual

Published Project Report NRA Pavement Cost Model manual T Buckland

Transport Research Laboratory Creating the future of transport PUBLISHED PROJECT REPORT NRA Pavement Cost Model Manual T Buckland Prepared for: National Roads Authority, Ireland Project Ref: 11110073 Quality approved: T Buckland (Project Manager) V Ramdas (Technical Referee) Transport Research Laboratory 2012

Disclaimer This report has been produced by the Transport Research Laboratory under a contract with National Roads Authority, Ireland. Any views expressed in this report are not necessarily those of National Roads Authority, Ireland. The information contained herein is the property of TRL Limited and does not necessarily reflect the views or policies of the customer for whom this report was prepared. Whilst every effort has been made to ensure that the matter presented in this report is relevant, accurate and up-to-date, TRL Limited cannot accept any liability for any error or omission, or reliance on part or all of the content in another context. When purchased in hard copy, this publication is printed on paper that is FSC (Forest Stewardship Council) and TCF (Totally Chlorine Free) registered. Contents amendment record This report has been amended and issued as follows: Version Date Description Editor Technical Referee 0.1 15/10/11 First Draft TB VR 1.0 19/04/12 Final version (incorporating client comments) TB VR i

Table of contents Abstract 1 1 Introduction 2 1.1 Development and licensing 2 1.2 Installation 2 2 Model and data operation 3 3 Modelling processes 4 4 Data setup 5 4.1 Condition parameter selection 5 4.2 Network selection 5 4.3 Network class 5 4.4 Carriageway Type 5 4.5 Treatment selection 6 4.6 Missing Data 6 5 Data ageing 7 5.1 Longitudinal Profile 7 5.2 Rut depth 7 5.3 Sideway-Force Coefficient 8 5.4 Texture 8 6 Homogenisation 9 7 Treatment identification 10 7.1 Model treatments 11 7.2 Treatment options 11 7.2.1 Do Something option 11 7.2.2 Do Minimum option 12 7.3 Intervention levels 12 8 Scheme creation 13 9 Costing 14 9.1 Works costs 14 9.2 User delay costs 14 9.2.1 Closure Type 14 9.2.2 Working pattern 14 9.3 Residual value 15 9.4 Discount rate 15 9.5 Total scheme costs 15 ii

9.6 Economic indicator and savings 15 10 Scheme prioritisation and selection 17 10.1 Prioritisation by budget 17 10.2 Prioritisation by condition 18 10.3 Treatment effects 19 10.4 Treatment resets 19 11 Committed works 21 12 Acknowledgements 22 13 References 22 Appendix A 23 iii

Abstract Road authorities require pavement information and decision support tools to assist them in the management of their asset, particularly in determining the budgets and maintenance that are needed on their network. This report covers the modelling methodology and processes that have been developed and implemented in a pavement cost model for the National Roads Authority, Ireland. The model has been developed so that the reference data and imported network data link into a workspace, which is then used as the calculation space during an analysis. A user can choose to create a range of different analyses in order to provide answers and associated sensitivity analysis to issues such as how much budget is needed to maintain the network in a steady state. The modelling process is detailed within this report, along with the rules and algorithms used. This includes the steps of ageing data and identifying treatments through to selecting and costing maintenance schemes in the chosen network. 1

1 Introduction This document is the manual for the NRA Pavement Cost Model (hereinafter referred to as the model) developed by the Transport Research Laboratory (TRL) for National Roads Authority (NRA), Ireland. This manual covers the modelling methodology and processes, database formats and rules and algorithms used by the model. A separate User Guide (Buckland, 2011b) has been produced which is designed to guide the user through the operation of the model, setting up analyses and viewing the outputs. 1.1 Development and licensing The software has been developed using Microsoft Visual Basic 6 and is provided to NRA as a stand-alone executable that can be installed on local machines as required. The software must not be redistributed (outside of NRA) or modified in any way without permission of TRL. The software connects to a Microsoft Access database and therefore Microsoft Access (and a license) is required to run the software. 1.2 Installation The NRA Pavement Cost Model comes with an automated installation package, which installs the appropriate files and folders on the destination computer. To run the installation, double-click on setup.exe. This will begin the installation and extract the necessary files. During the installation it may provide a prompt that: A file being copied is older than the file currently on your system. The installation will ask if you want to keep the current file and it is recommended that you click Yes. The destination file is in-use. Please ensure all other applications are closed. Click Ignore. It will then prompt that if you ignore a copy error, the file will not be copied. Do you want to ignore the error? Click Yes. Once the installation has completed there will be a new Program item under the Start menu for the model, entitled NRA Cost Model. 2

2 Model and data operation The model has been designed so that the reference data and imported NRA data link into a workspace. The workspace is the database that the model connects to through the model and is used to house all the tables created during run time. When the model is connected to the workspace, the user can create any number of analyses within that workspace. These analyses can be setup to use different configurations in order to test a range of scenarios. For each successful analysis completed in the model a separate output database can be created to store the results of the analysis. As long as the output databases are not overwritten and their location remains the same as when created, they can be used within the model to view the analysis results at any point in the future. The interaction of the model with the various databases can be seen in Figure 1. For further information on the how the model operates and for instructions on configuring an analysis and completing a run there is a separate User Guide (Buckland, 2011b). Reference Data Contains default lookup values for items such as deterioration rates, carriageway types and unit rates. Imported NRA Data Contains actual NRA network data for items such as route characteristics, condition parameters and traffic data. Workspace Contains the linked reference and imported data, and is where all the run time tables are created and analysed. Modelling Analyses are setup and run. Output Data Contains the output data and results from the completed analysis. Figure 1: Model interaction with input and output data 3

3 Modelling processes A summary of the modelling process implemented in the model is shown in Figure 2. Data Setup Data Ageing Data Homogenisation Treatment Identification Scheme Creation Evaluate Scheme Options Scheme Costing Loop through each year being analysed Scheme Prioritisation Scheme Selection Year End Calculations Figure 2: Modelling Processes When the user chooses to run an analysis the process described in Figure 2 is undertaken. This allows the model to operate in a consistent manner for each analysis, for any number of years. The principal output from this modelling process is a road maintenance programme that best meets the particular constraints for the analysis using whole life costing. 4

4 Data setup In setting up an analysis the user is configuring a set of variables that are used in two ways: 1. In advance of the run to determine the input data to use; 2. During the run to determine the calculations and options that are carried out. An exhaustive list of all the descriptions for the analysis configuration and setup options can be found in the model User Guide (Buckland, 2011b). 4.1 Condition parameter selection An analysis must be configured to use at least one condition parameter from those available: Longitudinal Profile, a measure of the unevenness of the road surface; Rut depth, a measurement of the deformation of the upper pavement layers; SFC, the Sideway-Force Coefficient, a measure of skid resistance of the pavement surface; Texture depth, a measure of the texture of the road surface that can be related to the skid resistance. 4.2 Network selection A network can be chosen for analysis from the available data, which can be the entire network or a selected subset of the network. In order to choose a specific subset of the network there is a hierarchy of selection options from which data can be selected. In order of hierarchy, the selection options are: County; Carriageway Type; Road; Route ID. 4.3 Network class Some of the rules in the model (e.g. deterioration rates) require the model to know which parts of the network are primary and which parts are secondary. This is determined from a lookup table in the reference database (tbl_lookup_network_class). If any definitions change in the future then this table can be updated to reflect the change. At the time of development the N1 to N50 were classed as primary routes and the N51 to N87 as secondary routes. 4.4 Carriageway Type Some of the rules in the model (e.g. deterioration rates) require the model to know which parts of the network are of which carriageway type (Single/Dual/Motorway). This is determined from a lookup table in the reference database (tbl_lookup_cway_type) and the default data can be seen in Table 5. 5

4.5 Treatment selection An analysis must be configured to use at least one of the available treatments from those available in the configuration. The model has been designed so that the user can add a treatment to the list of those currently available. A detailed description of the process for adding a new treatment is found in the user guide. 4.6 Missing Data If the user has opted to configure the analysis using only actual data then only records that have a surveyed measurement for each selected condition parameter selected will be kept. All other records will be dropped from the analysis. Otherwise, if the user has opted to fill missing data with defaults then where any of the selected condition parameters has a missing value for any record it will be filled with a default value. The default values are held in the reference database and have been chosen so that any default value itself should not trigger a treatment intervention (if recommended deterioration rates are used). The default values recommended are: Longitudinal Profile: 0.1 IRI (International Roughness Index) Rut depth: -10 mm SFC: 99 CSC (Characteristic SCRIM Coefficient) Texture: 5 mm Once a treatment occurs on a record that has been populated with a default condition value, then (if applicable) that value will be reset according to the reset rule for the treatment. 6

5 Data ageing Condition projection rules are utilised within the model in two main areas: 1. Existing condition data for the selected network is projected to the start year of analysis. 2. All condition parameters used within the model are aged up to the year of their treatment (if any). Following a treatment the appropriate condition parameters are reset and then the data is aged again until the next treatment. This continues until the end of the analysis period. The same condition rules are used in the two different parts of the analysis. For all condition parameters the deterioration rates can be specified separately for both for primary and secondary network. For each network the rates can also be specified by the three carriageway types of single, dual and motorway. 5.1 Longitudinal Profile The longitudinal profile data can be deteriorated in two ways, one of which must be chosen by the user during the configuration of the run: By a user specified amount per year; By a user specified percentage increase per year. 5.2 Rut depth The rutting data can be deteriorated in three ways, one of which must be chosen by the user during the configuration of the run: By a user specified amount per year; By a user specified percentage increase per year; By using a road type based relationship. The road type based relationships are of the form: where a, b and c are co-efficients for the rutting prediction and i is the year to predict. The default values for the carriageways are given in Table 1. Table 1: Default road type relationship coefficents Road Type a b c Single Carriageway 0.1114 1.175 0.0015 Dual Carriageway 0.2235 1.175 0.0030 Motorway 0.2500 1.220 0.0030 7

5.3 Sideway-Force Coefficient The SFC data can be deteriorated in two ways, one of which must be chosen by the user during the configuration of the run: By a user specified amount per year; By a user specified percentage increase per year. 5.4 Texture The texture data can be deteriorated in two ways, one of which must be chosen by the user during the configuration of the run: By a user specified amount per year; By a user specified percentage increase per year. 8

6 Homogenisation The user can choose to homogenise data that is statistically similar through the use of an algorithm within the model. Such pavement lengths are grouped together and assigned the same condition value. Each condition indicator is evaluated for homogeneity independently from the rest. The user can setup the homogenisation options at the time of configuring the run. This will also include whether data homogenisation is permitted across changes in the construction or not. The homogenisation algorithm operates as follows: All individual road lengths are initially considered as potentially homogeneous lengths. If the user has chosen to homogenise across construction breaks then condition data for the entire length of each road will be the starting point for the algorithm. If the user has selected to not allow homogenisation across changes in construction then condition data for each road will be segmented into groups which have the same construction; they will then be the starting points for the algorithm. The average condition value d dj N j N = = j= 1 is calculated in each length, where d j is the j th condition measure in the length and N is the number of condition measures in the length. j= N 1 2 Calculate the statistic R = + 1 d j) 2 j= 1 j=n 2 (d j (dj d). If N <= 60 then the value of R is compared to tabulated values (see Table 5). If it satisfies certain conditions then the length is accepted as homogeneous and the value d is assigned as the pavement condition of the entire length. If N > 60, then calculate U = N 2 1 N 2 length is accepted as a homogeneous as above. j=1 (1-R). If -1.282 < U < 1.282 then the If the length is not accepted as homogeneous the algorithm enters into an iterative process of splitting the subsection into two smaller parts and testing each for homogeneity. The location i of N readings - at which a subsection is split is determined by maximising the statistic G N = i )} = j i 2 = { lj(dj d (l j is the i(n i) j 1 length of the reading). If the length of the subsection is less than twice the homogeneity limit then the mean condition value is assigned to the subsection. The iterative process continues with each new subsection being analysed independently and potentially split many more times until either all subsections are homogenised or all fall below the minimum allowed length (which is userdefined). 9

7 Treatment identification For each year of the analysis the model calculates when the next treatment would be needed for each condition defect in the run by considering the current condition value, the threshold level and the method of deterioration. Any lengths that require treatment in the current analysis year for one or more defects are selected for consideration as schemes. There are two different analysis periods used within the model which relate to the rules for treatment identification: Programme period: this is the number of years for which a maintenance programme will be determined (shown by the loop in Figure 2); Treatment evaluation period: this is the length of the period over which costs are considered for any scheme in the programme period. If the number of years in the programme period is greater than one, then each year in the programme period acts as a base year in turn, and all maintenance identified in each base year is evaluated for a number of years beyond the base year, which is the treatment evaluation period (as shown in Figure 3). First Programme Period Year Treatment Identification Scheme Creation Loop through each year in programme period Evaluate future costs within treatment evaluation period Loop through each year in treatment evaluation period Analyse schemes for selection in current year Figure 3: Treatment years assessment loop 10

7.1 Model treatments The model has been developed with four default treatments: Overlay; Inlay; Surface dressing; Patching. In addition, the user can enter any other treatments of their choice to be considered alongside all existing treatments. A treatment is identified once a condition parameter passes the respective intervention threshold for that defect. The treatment that is triggered depends on the condition parameters and its value; this also determines whether it is a Do Minimum or not. It is possible for more than one treatment to be required and this can result in one treatment taking precedence over the others. If multiple treatments are flagged as needed, the treatment with the greatest hierarchy is undertaken; the hierarchy of the treatments can be managed from the configuration form. The four default treatments in the list above are ordered in descending order of the hierarchy they were given at the time of development. If a length of road is in need of a treatment that is not permitted on that particular carriageway, then the next available treatment with a greater hierarchy will be chosen. 7.2 Treatment options Two treatment options will be analysed for each scheme. They are: Do Something; Do Minimum. For each year of the programme period in turn a Do Something option and a Do Minimum option are analysed for each pavement scheme. When those options are analysed and costed in subsequent years (equal to the treatment evaluation period) only Do Something treatments are analysed. 7.2.1 Do Something option Treatments required in year 1 (the first year when the scheme was identified) are determined by comparing the condition value to the appropriate intervention level for that measurement. Within a scheme, some lengths may not have exceeded thresholds but may be close to the threshold (i.e. they will exceed threshold with a user defined look ahead period). Lengths that will exceed the given thresholds within a user set look ahead period (i.e. future number of years) will also be considered for maintenance. The whole life costs of the treatment options being considered are calculated by generating a maintenance profile over the treatment evaluation period based on intervention thresholds and the minimum intervention period. 11

7.2.2 Do Minimum option The Do Minimum option consists of the minimum treatment required to keep the condition within a scheme functional and safe for at least one year. For example, if any lengths are in extremely poor condition in year 1 (i.e. below the minimum standards to maintain safety), those lengths will be treated to maintain safety standards. Treatments in the following year will be determined in the same way as treatments for the Do Something option. 7.3 Intervention levels For all condition parameters, intervention levels can be defined by the user when setting up the run, and treatments will be considered when a parameter value exceeds one of the intervention levels. For each parameter, separate levels can be specified for investigation (do something) levels and safety (do minimum) levels. Recommended default values for the intervention levels are found in a lookup table in the reference database. 12

8 Scheme creation Lengths of road that have exceeded the minimum intervention period following the last treatment are considered for maintenance. Schemes will be created by joining together individual lengths of road based upon the following methodology: All lengths that need treatment will have been identified previously (see section 7); Lengths of treatments needing maintenance that span either different counties or carriageway types will be split accordingly; Committed works are included alongside any treatments identified by the model. If committed works overlap with a model scheme and the user has chosen not to allow the limits of the committed works to be extended then the model identified length will be dropped. If the user has allowed the committed works to be extended then its limits will be extended to encompass any overlapping model schemes; These initial treatment lengths form the initial schemes; Any schemes that end within a user defined length of another scheme are joined; If any scheme is less than the minimum scheme length then it is dropped for consideration in the current year; If any scheme is greater than the maximum scheme length then it is split into the required number of schemes to comply with the limits. 13

9 Costing The area of pavement requiring treatment is given by: Area requiring treatment = (Length requiring treatment) x (Carriageway width) All costs are calculated using the base year costs; all in-year costs during the analysis are the same (obtained from the reference database). 9.1 Works costs Works costs and durations use the units cost per metre squared and are dependent on the treatment and carriageway type. The unit rates and durations for treatments can be updated through the configuration screen. The total works cost will be calculated by multiplying the work costs per unit area by the area requiring treatment. The duration of a treatment will be calculated (in hours) by multiplying the duration per unit area by the area requiring treatment. Recommended default values for the costs and durations (at the time of development) can be seen in Table 7. 9.2 User delay costs User delay costs are used in the model to provide a representation of delay that is experienced by road users during each maintenance intervention. Lookup values for the user delay costs are held in the reference database and use the unit cost per vehicle per duration hour and are dependent on the carriageway type, closure type, traffic level and working pattern. 9.2.1 Closure Type The options are working pattern are shown in Table 2. The user can choose a default closure type for each treatment in the configuration screen. Table 2: Closure types Closure Type Contraflow Lane Closure Shuttle Working 9.2.2 Working pattern The options are working pattern are shown in Table 3. The user can choose a working pattern for low, medium and high traffic levels for each carriageway type in the configuration screen. 14

Table 3: Closure working patterns Working Pattern Night Off-peak All day (24hr) 9.3 Residual value The residual value of a scheme is the remaining value at the end of the analysis of any interventions carried out in the treatment evaluation period. In order to allow comparison between different treatment profiles calculated within the treatment evaluation period the remaining value in the pavement following the last intervention is subtracted from the total cost. This allows an equal comparison between a scheme that leaves the length in poor condition and a scheme that intervenes just before the end of the analysis but leaves the length in good condition. There are three options that the user can choose to calculate the residual value in the pavement: 1. Linear method: This method assumes that the cost of the last treatment intervention declines linearly until the next treatment is needed; 2. Double-declining method: This method is based on declining at a rate double the linear rate (NB: this method may not fully depreciate the value of the pavement at the time of the next treatment); 3. Minimum of linear and double-declining method: This method takes the minimum of the linear and double-declining method each year. 9.4 Discount rate Discounting is the method that allows the maintenance interventions in different years to be compared equally, by discounting their costs back to a base year. All future costs will be discounted using the standard formula for discounting: Discounted Cost = where n is the number of years elapsed since the base year of the analysis. discount rate is entered by the user when setting up the run parameters. The 9.5 Total scheme costs The total scheme costs are made up of: Total scheme cost = Works costs + User delay costs Residual value 9.6 Economic indicator and savings All Do Something schemes have a savings and economic indicator calculated which are used in the prioritisation process. They are calculated using the following methods: 15

Savings: The different between the discounted total costs minus the residual value of the Do Something and Do Minimum options; Economic Indicator: Savings / (Difference between the discounted year 1 costs of the Do Something and Do Minimum options). 16

10 Scheme prioritisation and selection The maintenance schemes are prioritised in order to select appropriate schemes against any constraints for a run. There are two methods of prioritisation in the model: 1. By budget; 2. By condition. Regardless of the prioritisation method chosen, the model first determines if any of the schemes on opposite lanes on single carriageway roads overlap. The schemes are treated separately for costing purposes, savings and economic indicator calculations. However, if any of the schemes overlap with another and the overlap for at least one of the schemes is greater than a user specified value then if one of the schemes get selected, the other overlapping schemes will also get selected if any constraints allow it (i.e. if there is enough budget). 10.1 Prioritisation by budget The following method is used if prioritising by budget. For each year in the programme period: Determine the budget available for the current year; If there are any committed works that have been setup to be carried out regardless of any budget constraint they are selected first, and their total in-year cost is calculated; For any remaining schemes, consider the Do Minimum options next; o o If the user has opted to carry out the Do Minimum options regardless of budget they are selected, and their total cost is calculated and added to the previous committed works total to calculate the current in-year spend; If the user did not opt to carry out all Do Minimum options regardless of budget, order the Do Minimum options by their year 1 works cost and cycle through the options, selecting each one in turn until the budget is exceeded; If there is any remaining budget, order the remaining Do Something options by their economic indicator and then year 1 works costs; o o If the user chose to try and spend as much of the available budget as possible (i.e. they might have committed contracts with contractors), select the Do Something option only if the difference between the Do Something and Do Minimum year 1 cost does not exceed the budget. The associated Do Minimum option will then be deselected; If the user does not necessarily want to spend the available budget, select the Do Something option only if the difference between the Do Something and Do Minimum year 1 cost does not exceed the budget and the Do Something option represent an economic saving over the treatment evaluation period. The associated Do Minimum option will then be deselected; 17

o If there are any associated overlapping schemes, consider those Do Something options next, before looping through the remaining Do Something options until the budget is exceeded. 10.2 Prioritisation by condition During the run configuration, the user enters a target for the percentage of the network they will allow to be in poor condition for one or more of the defects being used in the analysis. The target can either be specified for the whole network or separate targets can be specified for each carriageway type. The user enters a target year in which they wish the target to be met. The following method is used if prioritising by condition. Calculate the current poor condition of the network for each defect, ordered by the importance of the defects from the treatment triggers and their treatment hierarchy; o o If the target year is less than the current year then calculate the poor condition for each defect for the current year, so that it can be maintained against the target condition (i.e. if the target year has been exceeded then the model tries to maintain the network in a steady state beyond the target year at a condition equal to the target condition). The improvement needed in the year is the difference between the current poor condition and the target poor condition; If the target year is greater than or equal to the current year then calculate the poor condition for each defect that would exist in the target year if no treatments were performed. In order to meet the target in the target year it is assumed there will be equal improvements in all preceding years, therefore the improvement needed in each year up to the target year is the different between the current poor condition and the calculated poor condition in the target year, divided equally between the number of years from the current year to the target year; If there are any committed works that have been setup to be carried out regardless of any constraints they are selected first, and their total in-year condition improvement calculated; If the user has opted to carry out the Do Minimum options regardless of any constraints, they are selected, and their total in-year condition improvement calculated and added to the previous committed works condition improvement to calculate the current in-year improvement; If there is any remaining improvement required then all other schemes are considered in the order: o o o Do Minimum committed works schemes are looped through and selected until the required condition improvement has been met; Do Minimum non-committed works schemes are looped through and selected until the required condition improvement has been met; All remaining Do Something non-committed works; 18

If the Do Minimum option of the current Do Something scheme was not selected then select the current Do Something if a condition improvement is required; If the Do Minimum option of the current Do Something scheme was selected, only select the current Do Something if: The Do Minimum had a lower improvement for the defect and a condition improvement is still required; or The Do Something improvement is equal to the Do Minimum improvement and the Do Something provides an economic saving; If any of the Do Something options are selected then the Do Minimum option for the same scheme is deselected and the previous improvement is removed from the cumulative improvement total; o If any of these scheme options are selected then the scheme is marked as being considered so that it is not considered again for an improvement of any lower hierarchy defects; If there is any remaining improvement required after all the schemes have been considered then patching short lengths is considered; o o o Patching can be considered for any lengths of the network that were not in schemes already considered (i.e. schemes not considered in the above and lengths not included in any schemes); For any lengths available for patching, cycle through them in order of worst condition first and select the lengths until the required improvement is met; All lengths that are patched are costed based on the area patched and the unit rate of patching; 10.3 Treatment effects The effect of performing a treatment on a scheme is that no further treatment will be considered on that length of road until a number of years equal to the minimum intervention period (except for patching, where the minimum intervention period does not apply). 10.4 Treatment resets Following a treatment being selected one or more condition parameters may be reset. If a treatment is selected then the defects that triggered that treatment will be reset, along with any other defect triggers for treatments of a lower hierarchy. The defect triggers can be amended in the configuration screen. Separate reset values are specified for patching to reflect the lower standard of maintenance being applied and the fact that a future intervention is likely to be needed sooner than if a full treatment was applied. 19

The treatment reset values are held in the reference database (tbl_lookup_defaults) and shown in Table 4. Treatment values must be non-zero in order for all the deterioration relationships to function in the model. Table 4: Default treatment reset values Defect Standard Reset Patching Reset Longitudinal Profile 0.8 3 Rut 0.5 2.5 SCRIM 60 60 Texture 2 2 20

11 Committed works The user has the option of entering committed works during the analysis period. Committed works are entered if a scheme is already committed on a particular length of road. The schemes can be set to occur in a given year within the programme period. The User Guide (Buckland, 2011b) provides details on entering committed works when setting up an analysis. The following points apply to committed works: The user can specify whether to allow the length of the committed scheme to be extended if the modelled has identified any treatment that overlaps; For each committed scheme, the user can specify whether it should be carried out regardless of any budget or condition constraints. If so then the committed works scheme is selected in advance of any prioritisation. 21

12 Acknowledgements The work described in this report was carried out in the Asset Management group of the Transport Research Laboratory. The author is grateful to V Ramdas who carried out the technical review and auditing of this report. 13 References Buckland, T (2011a). NRA Pavement Cost Model: Data and Development Issues. TRL, PPR597. Buckland, T (2011b). NRA Pavement Cost Model: User Guide. TRL, CPR1253. 22

Appendix A Table 5: Default carriageway descriptions and type Model ID Description Carriageway 0 2 Lane Road Single 1 Motorway Mway 2 Dual Carriageway Dual 3 3 Lane Road/2 Lane Side Single 4 3 Lane Road/1 Lane Side Single 5 One Way Forward Single 6 One Way Reverse Single 7 6 Lane Road Dual 8 TPO Single 9 3 Lane Motorway Mway 10 3 Lane Dual Dual 11 Reduced Single Single 12 Wide Single Single Table 6: Homogeneity test values N R value lower bound R value upper bound 0 0 0 1 0 0 2 0 0 3 0 0 4 0.390225 1.609762 5 0.4102 1.5898 23

6 0.4450833 1.554917 7 0.4679571 1.532057 8 0.491225 1.508762 9 0.5121778 1.487822 10 0.531135 1.46889 11 0.5482727 1.451727 12 0.5637959 1.436188 13 0.5778923 1.422092 14 0.5908036 1.4092 15 0.6026533 1.39734 16 0.6135938 1.386422 17 0.6236706 1.376329 18 0.6330139 1.366989 19 0.6417 1.358289 20 0.6498 1.350188 21 0.6573809 1.347381 22 0.6645068 1.335505 23 0.6712391 1.3288 24 0.6775896 1.322452 25 0.683568 1.316448 26 0.6892307 1.310769 27 0.6945852 1.305393 28 0.6996857 1.300291 29 0.704538 1.295434 30 0.7091467 1.290838 31 0.7135161 1.286468 32 0.7176984 1.282286 33 0.721697 1.278303 34 0.7255632 1.274479 24

35 0.7292514 1.270774 36 0.7328125 1.267195 37 0.7362973 1.263746 38 0.7396592 1.260386 39 0.7429 1.257118 40 0.74607 1.253947 41 0.7491219 1.250878 42 0.7520571 1.247913 43 0.7549255 1.245056 44 0.7576796 1.242309 45 0.76032 1.239676 46 0.7628478 1.237158 47 0.7652638 1.234757 48 0.7675688 1.232428 49 0.7697632 1.23022 50 0.771848 1.228136 51 0.7738725 1.226128 52 0.7758865 1.224147 53 0.7778415 1.222196 54 0.779787 1.220276 55 0.7816746 1.218387 56 0.7835045 1.216531 57 0.7853264 1.214709 58 0.7871405 1.212921 59 0.7889475 1.211119 60 0.7906983 1.209303 25

Table 7: Recommended default unit rates and durations Treatment Carriageway Rate ( /m²) Duration (hrs/m²) Surface Dressing Mway 7 0.0027 Dual 7 0.0027 Single 7 0.0027 Inlay Mway 10 0.0057 Dual 10 0.0057 Single 10 0.0057 Overlay Mway 20 0.0411 Dual 20 0.0411 Single 20 0.0411 Patching Mway 15 0.0057 Dual 15 0.0057 Single 15 0.0057 26

NRA Pavement Cost Model manual Road authorities require pavement information and decision support tools to assist them in the management of their asset, particularly in determining the budgets and maintenance that are needed on their network. This report covers the modelling methodology and processes that have been developed and implemented in a pavement cost model for the National Roads Authority, Ireland. The model has been developed so that the reference data and imported network data link into a workspace, which is then used as the calculation space during an analysis. A user can choose to create a range of different analyses in order to provide answers and associated sensitivity analysis to issues such as how much budget is needed to maintain the network in a steady state. The modelling process is detailed within this report, along with the rules and algorithms used. This includes the steps of ageing data and identifying treatments through to selecting and costing maintenance schemes in the chosen network. Other titles from this subject area TRL674 Durability of thin surfacing systems. Part 4: Final report after nine years monitoring. J C Nicholls, I Carswell, C Thomas and B Sexton. 2010 PPR497 GripTester trial October 2009 including SCRIM comparison. A Dunford. 2010 PPR468 PPR458 PPR457 Enhanced levels of reclaimed asphalt in surfacing materials: a case study evaluating carbon dioxide emissions. M Wayman and I Carswell. 2010 Review of UKPMS core functionality the minimum functionality all PMS should embody in the UK. B V Cleave, R A Cartwright, K A Gallagher and T Rasalingam. 2010 SCANNER accredited surveys on local roads in England accreditation, QA and audit testing annual report 2008 09. P Werro, I Robinson, E Benbow and A Wright. 2010 PPR437 Highways Agency 2009 National Falling Weight Deflectometer Correlation trials. S Brittain. 2010 PPR393 Measuring skid resistance without contact 2008 2009 progress report. A Dunford. 2009 Price code: 2X ISSN 0968-4093 TRL Crowthorne House, Nine Mile Ride Wokingham, Berkshire RG40 3GA United Kingdom T: +44 (0) 1344 773131 F: +44 (0) 1344 770356 E: enquiries@trl.co.uk W: www.trl.co.uk Published by IHS Willoughby Road, Bracknell Berkshire RG12 8FB United Kingdom T: +44 (0) 1344 328038 F: +44 (0) 1344 328005 E: trl@ihs.com W: http://emeastore.ihs.com