Predicting the performance state of a pavement network is often high on the priority list for any local council or road agency when it comes to making sure the pavements are performing to satisfactory conditions.
One of the ways that New Zealand's local authorities have been ensuring this is through the use of the dTIMS forecasting system. dTIMS has been used by many of the local councils and road controlling authorities in some of the largest networks in New Zealand to plan for long term pavement performance and maintenance needs. While this system has been highly applicable to councils that manage large road networks, there has always been a caution used by those whose sealed road networks are relatively small. An example of this is the Central Otago district council which used dTIMS for maintenance and renewal planning of their road network, and many promising results have been achieved.
The Central Otago District Council manages what you would call a 'small' road network. With only 400 kilometres of sealed road length, this is sizably smaller compared to some of the other regions in New Zealand. Typically, one of the main things that had deterred the use of dTIMS is the confusion around its flexibility and applicability for smaller networks with low traffic volumes. Initially, for the Central Otago network, this was the case. “We didn't use dTIMS not just because of the smaller size of the network, but because at the time we didn't think dTIMS was applicable to low volume networks,” says Julie Muir of Central Otago District Council. Much of the time the reason for not using dTIMS has been because the true value that could be gained from using such a tool is not clear, as Muir adds: “We haven't had a rehabilitation programme for 15 to 20 years, and so we didn't know we would get the value that we did from dTIMS.”
So what are some of the reasons for making use of such a system as dTIMS?
dTIMS can aid with making predictions of the condition of a road network over long time periods, where alternative methods may only be able to predict for shorter periods of time. This has been the main trigger for using dTIMS for the Central Otago road network. “It was possible to make observations of the reseal needs over a 3-year period, and maybe push it to 10-years, but it is not possible to predict what would happen to the network over a 30-year period,” says Muir. It is this superior predictive capability of dTIMS that makes it an excellent tool for developing long-term infrastructure strategies.
As with any predictive tool, the quality of the outputs that dTIMS can deliver depends on the quality of its inputs, and the process of data tidying is critical in ensuring the quality of the performance forecasting within the dTIMS modelling framework. This is something the Central Otago District Council has spent a lot of time ensuring. “
We got pavement life data, and went through old contracts to get reseal dates. We made sure that we put the effort into getting the quality of the data right.
We did FWD tests as part of the recommendations,” says Muir. The dTIMS system was used to forecast the maintenance needs and the amount of spending that was required for the sealed network in Central Otago, and the outcomes of modelling task have been tremendously useful in many aspects. The predictions showed a much reduced resurfacing programme, as much as a 28% reduction in resurfacing.
When these predicted outputs were compared to the actual on-site conditions, it was found that the predictions from dTIMS which identified the sites that needed resurfacing were highly accurate to on-site conditions. The only issue that was seen to come up was around timing of the resurfacings, and the length of treatment that was required, where too long of a treatment length may have some differences in the predicted length that needed to be resurfaced. However, overall, the prediction capabilities are spot on. Refer to the figure that illustrates the condition outcome on the CODC network following different resurfacing strategies. For example for an over-all resurfacing strategy of 20km/yr, one can expect an over-all surface condition of CI= 2, which is still considered to be good.
The result of using dTIMS in terms of the overall investment profile has given the Central Otago District Council much to be happy about, where a 17% reduction in maintenance and renewals needs have been predicted. The council has been all too aware of the possibility of this reduced expenditure exposing the network to danger, and the predicted outputs from dTIMS had been confirmed using site visits. A point of focus for this forecasting exercise was the roughness values on the Central Otago network, and from the predictions it was found that the change in pavement roughness was minimal, where even for the worst case scenario, the predicted roughness values were still within the maintenance guidelines.
When the performance of Central Otago district council's network was compared to New Zealand's overall pavement network performance, it was found that the performance of the Central Otago network was well above the national average, something which indicates that the maintenance works on this network are sufficient and that the network managers are proactively doing what is needed for the network. This achievement was possible only with dTIMS, and Muir says “We have 96% satisfaction rating for our network, which is extremely high.” The excellent performance of the network has brought about an interesting dilemma as it challenged the typical decision making framework for network managers, and Muir adds: “We have never thought should we do less work on sealed roads, it has always been 'should we do more?'” For Central Otago district council this was uncharted territory, and presented an excellent monitoring aspect for the future.
The financial savings the Central Otago district council has managed to achieve on their annual forecast as a result of dTIMS equates to approximately $300,000, which is substantial considering the size of their network. Even after accounting for the costs of licenses and modelling, this saving is significant. Adding to that the fact that there is no need to conduct the dTIMS modelling annually, the final savings that can be made over a monitoring period are considerable. Critical to confirming the outputs from dTIMS is the FWD testing, and while this may be viewed as an unnecessary item, the cost of FWD testing can far outweigh the value that can be gained from the dTIMS modelling. “The FWD test was such good value for money because you can confirm the result of dTIMS,” says Muir.