Rational Design Approach

An alternative method of calculating the condition and residual life of composite pavements has been developed and validated in Portsmouth – advancing techniques of pavement management in the process

Around 70% of principal roads in Portsmouth consist of concrete base structures beneath asphalt overlays. For normal maintenance purposes the structural condition of these highways could be estimated to an adequate degree of accuracy using conventional methods of assessment. But in Portsmouth the contractual environment of a highway maintenance PFI project has been introduced.

Responsibility for substantial short-term investment and long-term maintenance has been transferred to Portsmouth City Council’s (PCC) PFI concessionaire Ensign Highways. Both PCC and Ensign have had good reason to look much more closely at developing accurate measurement and analysis of pavement condition.

Around five years ago, as the Portsmouth PFI was getting under way, the ‘local’ team of PCC, Ensign and the latter’s delivery partner, Colas, began investigations in partnership with the Colas Group’s research and development department at the company’s Campus for Science & Technology, in France. The results are a new analytical Rational Design Approach (RDA) for calculating the residual life of all types of pavements in the UK – and greater knowledge and understanding of pavement management techniques.

‘RDA builds an accurate and realistic picture of the condition of individual layers and pavement structures overall, and allows data to be used in a more intelligent and useful way, in many different applications,’ said Mott MacDonald technical director and pavement management team leader Iain MacGregor.

Mott MacDonald were instrumental in setting up the first highway maintenance PFI in Portsmouth, particularly the contract’s system of service and repayment mechanisms. These were set up based on the number of heavy vehicles using Portsmouth’s roads, plus lane availability and a system of performance measures including a Network Condition Index (NCI) and Pavement Condition Index (PCI).

The PCI was key due to the weighting of the repayment mechanisms, which gives greatest importance to the condition of Portsmouth’s principal road network. The contract was set up with PCI linked to calculated residual life of pavements, measured using deflectograph surveys. Results would be interpreted with standard Pandef analysis software.

‘Use of deflectograph analysed with Pandef is an adequate way of determining the residual life (RL) of asphalt pavements. It also gives a rough estimation of the RL of roads consisting of bituminous material over concrete base structures,’ explained Colas principal engineer Bashir Abdulrahim. ‘However, we needed an alternative method for the Portsmouth PFI contract. Accurate measurements were needed for ensuring correct levels of structural repair were carried out and for long-term maintenance planning.’

Colas’s efforts were not aimed at reducing the quantity of work to be carried out during the project’s five-year, £60 million ‘core investment period’ (from 2004–2009). The Portsmouth PFI contract allows for changes in pavement survey and analysis, providing a ‘parallel running period and review’ is carried out. The contract conditions state that the minimum amount of work to be carried out is set at the level determined by the standard method of calculation. This quantity of activity cannot be reduced. If the new technique shows less work is needed, the NCI is lifted instead.

‘It was in all parties’ interests to have good confidence in the pavement analysis system and use of the latest innovations,’ said Mr Abdulrahim. ‘It was important for PCC to know that its roads were being improved correctly. For Colas, as the long-term steward responsible for maintaining Portsmouth’s roads at the condition set by the NCI for the following 25 years, accurate analysis is vital for future resurfacing programmes. Best value, minimal disruption within Portsmouth and long-term success can only be achieved with well-informed knowledge of the residual life of the pavements concerned.’

As PCC’s professional advisors, Mott MacDonald were given the job of validating the RDA method, which Colas developed from a French approach to pavement analysis.

Iain MacGregor explained: ‘Pavement design in the UK is generally based on TRL work of the 1970s and 80s, based on observations of pavement behaviour under various loads which produce stress and strain criteria to judge pavements against. Analytical methods, such as RDA, are usually proven by comparison with empirical methods, and while the French method is different in detail from UK approaches, it essentially uses the same principles.

‘We reviewed RDA, to see if the technique could apply to the UK, by doing research into how RDA compares with other methods in Europe; then we ran analyses of a number of case studies using the different software to investigate whether identical pavement structures could be modelled using each method, and we found they could be. Results were comparable.’

Colas’ RDA system is based on an internationally recognized French method of pavement design produced by France’s LCPC (Laboratoire Central des Points et Chaussees) and SETRA (Service dEtudes Techniques des Routes et Autoroutes). Both the LCPC/SETRA technique and its UK counterpart make use of the same Burmister model – a linear multi-layered elastic model for calculating stresses and strains
from given loads – verified from observed empirical values of pavement deflection or deformation.

There are subtle differences between the two methods, but the important task for Mott MacDonald came down to validation of RDA by comparing results obtained from French and UK software based on the same Burmister model of analysis. For this, Mott MacDonald used Shell BISAR – the most commonly used software for analytical pavement design in the UK. ‘The conclusion of the comparison was that in all cases it was possible to replicate the examples given by Colas in both sets of software and produce the same raw output of stresses and strains, also that the operation of both types of software is intuitive and broadly similar,’ states Mott MacDonald’s evaluation report.

‘Our broad conclusion was that RDA could be used to measure and analyse pavement condition in Portsmouth,’ said Mr MacGregor. ‘However, further work was then needed from Colas to show the magnitude of difference between RL results from RDA and Pandef analysis of deflectograph measurements made in Portsmouth, and to prove the difference was repeatable.’

This work has now been done, and according to Colas’s Bashir Abdulrahim, the overall results have consistently shown Pandef to give overly conservative estimations of the residual life of pavements in Portsmouth. In effect, Colas have proven less structural road repair work was needed to meet the required NCI, which has been raised as a result, with positive implications for levels of maintenance needed in future. Furthermore, pavements are now being managed with considerably higher levels of intelligence for the long term in Portsmouth.

‘Whereas Pandef only gave estimations of RL, the RDA method allows use of outputs for the design of structural inlays and a whole range of different applications,’ said Mr MacGregor. ‘It has certainly altered our thinking on pavement management in highway PFI contracts.’