Economic, Environmental and Social Impact of Changes in Maintenance Spend on Roads in Scotland Summary Report
Appendix B Key aspects of methodology and assumptions for the quantified analyses
Various assumptions have been made in order to complete the quantitative analysis. This Appendix describes those assumptions considered to have most significance in terms of the results.
B.2 Assumptions on scaling
For the local road network, assumptions were made in order to scale the results to the full economic impact of the Scenarios. A sample of Local Authorities (8) was analysed in detail. These were selected to be representative of the 32 Local Authorities in terms of the differences across the network (ranging from rural to city networks). In order to scale the results from 8 to 32 Local Authorities, a methodology was developed to account for the different authority and road types, the existing road conditions and the different amounts of travel across the network.
Full details of the sampling approach and the scaling methodology are provided in the detailed report (Transport Scotland, 2012b)(Transport Scotland, 2012b).
For Scenario 1 (retaining 2010/11 funding levels), the trunk road analysis was based on the budgets given for 2010/11 in September 2010. Similarly, the local road analysis was based on budgets reported for 2009/10 by Audit Scotland (Audit Scotland, 2011) and increased for 2010/11 using a 2.5% increase. In practice, budgets have since been further reduced for both networks compared with those used in the Scenario. The implication is that there is already a long term disbenefit to society if current investment levels continue.
For both networks, the reduction in budget for each Scenario has been allocated differently across different activities (see Section 3.1 and Appendix A). In each case this results in the budget reduction for carriageway maintenance being significantly greater than the overall budget reduction. The effect of this change has been assumed to be linear with the 20% and 40% funding reductions.
The significance of the above effect is shown particularly in the trunk road results. The NPV for Scenario 3 with a 40% funding reduction is predicted to be marginally less negative than for Scenario 2 with a 20% funding reduction because:
- The overall budget reduction generates a benefit (expenditure reduction) of over twice as much for Scenario 3 compared with Scenario 2[9]
- The increase in vehicle operating costs, the component of the impacts which represent the most significant reduction in benefit, is based on the effect of carriageway maintenance reductions, which increase from 44% (20% funding reduction Scenario) to 76% (40% funding reduction Scenario) (see Appendix A). The carriageway maintenance reductions with an overall 40% reduction are 73% more than for the Scenario with a 20% overall reduction in funding (i.e. significantly less than twice as much difference).
The effect is exacerbated further by the methodology to predict performance of the network described in Appendix B (Section B.5.1).
The effect is not seen in the local road analysis because the reduction in carriageway maintenance budgets from the 20% funding reduction Scenario to the 40% funding reduction Scenario (from 35% to 69%) is nearly twice as much, so the impact on vehicle operating costs is more significant.
B.4 Travel data
Travel data was taken from published figures for Scotland (Scottish Government, 2011). The total travel on the networks, for each road type, was distributed equally across the parts of the network in different condition (i.e. no account was taken of condition or location of a link within the network).
B.5 Vehicle operating costs
The predictions of conditions for each network and each budget Scenario have been derived using existing condition projection models. WDM Ltd provided predictions of the future condition of the trunk road network to Transport Scotland, and similar information for the local road network to SCOTS. The models for the networks are conceptually similar but differ in the technical implementation.
Standard UK models take no account of changes in road condition on vehicle operating costs, assuming that conditions across the road network in the future are similar to today. To assess the impact of road conditions on vehicle operating costs, the World Bank HDM-4 models have been used with parameters set (where applicable) in accordance with vehicle types used in the UK and with standard requirements for appraisals in Scotland (e.g. predicted traffic growth, fuel price increases and vehicle efficiency gains).
A specific anomaly regarding future network conditions was identified by the trunk road analysis. Both reduced expenditure Scenarios include a return to current expenditure levels in year 15 (i.e. 2025) and then a 2.5% per annum increase until year 20 (i.e. 2030). This resulted in the predicted condition of the trunk road motorway network becoming slightly better for the 40% funding reduction Scenario between years 15 and 20 than for the Scenario with a 20% funding reduction, which in turn is predicted to be slightly better than the Scenario with a constant (2010/11) budget. The effect is small and, accounting for the level of confidence in the analyses, considered insignificant but further contributes to the effect of the calculated NPV for the 40% funding reduction Scenario being slightly less negative than that for the 20% funding reduction Scenario, which has been discussed in Section B.3.
A number of detailed assumptions were made for modelling the effect of road maintenance funding on future network condition. Some of the more significant assumptions were:
- Standard and consistent unit rates were assumed for works across both networks
- For trunk roads there was a consistent proportional allocation of the works budgets to different work types (i.e. if x% of the budget is currently allocated to road resurfacing as a proportion of the overall spend on carriageway maintenance, then this is continued through the analysis period)
- For local roads there was a consistent proportional allocation of the maintenance works budgets to different road hierarchies (i.e. A, B, C and U class roads)
- There was no effect on condition from the impact of unplanned maintenance
- Future budgetary regimes and condition drivers were assumed to be similar to past experience. In particular, the Scenario with a 40% reduction in overall maintenance funding considered cuts to carriageway budgets of 56% for trunk roads and 69% for local roads. These are significantly beyond the range of previous budget variations. Further, the models were developed primarily as a tool to predict performance over the next 10 years, and differences between predicted and actual performance will be increased for longer term projections.
B.6 Travel time costs
The effect of reduced maintenance funding on travel time costs was assessed using the outputs from the condition projection models and standard costs of travel time as required by standard requirements for appraisals in Scotland. The evidence on which the effect was based was past research in the UK and is considered primarily applicable for higher speed roads (Cooper, Jordan, & Young, 1980). The analysis has therefore only been applied to trunk roads and A class local roads.
B.7 Skid related accident costs
Skid resistance data from the last 10 years was reviewed and suggests that there have been improvements to skid resistance over time which have levelled out in recent years. A simple link with historic road resurfacing budgets was also established.
Recent accident figures were assessed and the potential increase in accidents was estimated based on an assumption that skid resistance might deteriorate for the 40% funding reduction Scenario, given the evidence of historic budgets and skid resistance above. The most recent risk models for accidents on roads with poor skid resistance were adopted (Coyle & Viner, 2009).
B.8 Delay costs at roadworks
The effect on the total delays to traffic at roadworks sites was quantified using standard UK delay cost models in accordance with requirements for transport appraisals in Scotland.
B.9 Emissions costs and air quality
B.9.1 Methodology
The calculation of Carbon Costs considered carbon dioxide emissions from 3 sources:
- CO2 emissions from vehicles travelling over the network under normal running conditions;
- CO2 emissions from vehicles as they travel through roadwork sites (associated with traffic delays); and.
- The CO2 emissions from road maintenance activities (embodied Carbon).
The methodology for calculating the CO2 emissions from vehicles under normal running conditions used the predicted condition of the network for each budget Scenario. The methodology built on that described in Section B.5.1. Using the fuel consumption outputs and CO2 output data from HDM-4, the total emissions were calculated based on the number and types of vehicles travelling over parts of the network with different levels of roughness.
The CO2 emissions from vehicles delayed through roadwork sites were quantified using UK traffic delay cost models in accordance with standard requirements for appraisals in Scotland. The QUADRO (Queues and Delays at Roadworks) model provided a CO2 cost output based on user defined traffic management arrangements.
A number of representative (notional) maintenance schemes were modelled using QUADRO and the resulting CO2 costs scaled by the number of notional schemes that would be required to treat the length of network predicted to be in need of maintenance, based on the predicted network conditions.
The CO2 emissions from maintenance works were based on default values used in the asPECT model used in the LifeCycle Carbon Assessment of asphalt materials in compliance with PAS2050 (British Standards Institution, 2008)(British Standards Institution, 2008).
From the predicted mass of CO2, masses of other pollutants were derived based on typical combustion equations and fuel type (e.g. diesel and petrol).
B.9.2 Key assumptions
All key assumptions in Section B.5.2 apply to the emissions analysis.
B.10 Asset valuation
B.10.1 Methodology
Valuation of the road network was based on the replacement cost of the network calculated using standard maintenance cost unit rates for different types of road (supplied to this study).
Depreciated road asset value was assessed using the predictions of network condition over time. For roads in an as new condition, no depreciation was assumed. For roads showing some signs of deterioration, a reduction in service life was derived which enabled an assessment of the depreciation to be made. The assessment was made for each length of the network so that a total value of depreciation could be assessed.
Valuation figures were supplied by WDM Ltd in different forms for local and trunk roads. For local roads, the predicted accumulated depreciation was provided. For trunk roads, the predicted depreciated replacement cost was provided.
B.10.2 Key assumptions
Assumptions given in Section B.1 are also applicable for the valuation methodology. In addition, it should be noted that there are numerous other assets on the network besides carriageways (e.g. bridges, streetlights and other road furniture). No assessment is made of the depreciation of these assets.