Suffolk County Council (SCC) is responsible for some 6,500 km of county road. There are also approximately 400 carriageway kilometres of trunk road in the county from which SCC was the maintaining agent until 31 March 1997. The 400 km of trunk road and 1,430 km of county road comprise the Priority 1 salting network. This was treated by 48 routes covering an average of 38 salted kilometres per route in the two hour maximum treatment time. Eleven routes were in the agency areas of Ipswich, Lowestoft, Bury St. Edmunds and Haverhill. The remaining 37 routes operated from nine county depots, three of which have salt barns ranging in capacity from 3,500 - 4,500 tonnes.

Prior to the change in trunk road agency, winter maintenance on both trunk roads and county roads was planned and managed as an integrated operation. Although dual carriageway trunk roads were largely covered by dedicated routes, overall efficiency was boosted by integrating some slip roads and most single carriageway trunk roads with county roads to produce combined routes. The loss of the trunk road agency meant that there was an inescapable requirement to develop a new set of salting routes. In addition, savings which had previously been achieved through sharing resources, facilities and other fixed costs were lost. The county road service therefore faced an increase in cost at a time when the real value of the winter maintenance budget had declined in several successive years. This has been compounded by the depletion of a £750,000 reserve fund to virtually nil over two winters which were not substantially worse than average. Around the time that the need for a major review became apparent, SCC became aware of the GeoRoute software for routeing in winter maintenance vehicles and other municipal operations.

SCC had already considered (and rejected) a proposal from a winter maintenance consultant to undertake a one-off optimisation. Given a background of change, there was no certainty that any new scenario would remain applicable for long enough to break even. The prospect of a set of tools to develop routes, amend them as needed and explore different scenarios and service levels was a much more attractive proposition.

Consequently, PSR Group was commissioned to undertake an initial trial involving two routes in the agency area of Haverhill. This was completed in February 1996. The results clearly demonstrated the capabilities of the software, albeit in a small urban area. The next stage was to test the software with larger volumes of data and the very different network topography of rural Suffolk.

A six month trial began in September 1996. For this, the software was installed at SCC’s offices and SCC staff undertook the optimisation work.

Attention was first given to the county town of Ipswich. In Ipswich, there are some 167 km of Priority 1 salting network served by five routes. The routes had been developed manually some years before. A major depot relocation meant that new routeing was inevitable.

In Ipswich (and indeed for all other optimisations) the first step was to define the network to be salted and replicate the existing routes. This enabled parameters such as vehicle speeds to be adjusted to reproduce actual route timings. At this stage, GeoRoute was able to improve efficiency by re-sequencing the existing routes (i.e. the same roads covered by each route but in a different order).

The next step was to move the depot location and create entirely new routes. This can involve different options for treatment times, and disposition and capacity of vehicles. Routeing can also give priority for certain roads to be treated in a specific time window (e.g. the first hour of a two hour treatment time). In Ipswich, it was not economic to increase the size of spreaders from the present 4 cubic metre units. Therefore, the new routeing still comprised five routes but these were better balanced in terms of treatment time and salt usage and allowed key roads to be treated in the first hour.

Attention then turned to a larger trial area. This comprised the "East Area" of the county, bordering the North Sea and stretching from Lowestoft in the north to Felixstowe in the south. It forms one of three discrete areas for delivery of highway management services in Suffolk.

It became apparent that this area was on the limit of the capacity of GeoRoute. This was not due to the extent of the geographic area as such, but to the representation of the road network and the intermediate points which give it shape. Further developments of GeoRoute have subsequently allowed much of the minor road network to be filtered out which allows larger areas, and hence cross boundary operation, to be considered.

The Priority 1 salting network in East Area originally comprised 309 km of county roads and 123 km of trunk roads, treated by 12 routes in a two hour treatment time. This gave an average of 36 salted kilometres per vehicle with an average efficiency of 58.5%. The county roads therefore comprised 8.6 equivalent routes on a pro rata basis.

A base review of the Priority 1 network and small changes in cross boundary salting left 346 km to be treated. Initial work showed that this could be achieved by eight routes in a two hour maximum treatment time. Allowing up to two and half hours to completion of salting allowed the number of routes to be reduced to seven with an average of 49 salted kilometres per route.

Although the residual county road network was less conducive to efficient routeing, an average efficiency of 57.3% was achieved. This included the effects of closing a small satellite depot which, if applied to the existing routes, would have reduced the efficiency from 58.5% to 56.3%.

Details of the proposed winter maintenance service were reported to SCC’s Transport Committee on 11 September 1997 and were approved by Members. As well as increasing the maximum allowed treatment time, the package of measures included the closure of two minor depots. One of these was an agency depot and involved the transfer of salting operations from the agent to the county council. Closing such minor depots can have an adverse effect on routeing efficiency but this is outweighed by other savings. The greatest savings come in quantum steps by reducing the number of gritters. Once a practical minimum number of gritters has been reached, other factors can be brought into play to see whether the service can be optimised within this figure and within the other overriding parameters such as treatment time. Options include concentrating operations at depots with salt barns to maximise the use of barn stored salt, exploring depot closure and seeking to ensure that the most important or difficult sections of road are treated early within the overall treatment time. These benefits normally come at the expense of routeing efficiency and the real benefit of using route optimisation software is to enable an optimum balance of all these issues to be achieved, rather than pursuing a holy grail of optimising absolute routeing efficiency in isolation.

A further issue touched upon in the Transport Committee report was the probable introduction of bridge weight restrictions as a result of capacity assessments of weak bridges across the county. These are likely to have an impact within the next two years on the routes which have been carefully crafted this year. While such restrictions will undoubtedly increase the cost of the service and reduce the potential efficiency, the work done to date on route optimisation will enable new routes to be developed quickly and different options to be considered.

The cost of the optimisation exercise, including staff time and the purchase of the software, is estimated to be of the order of £70,000. Some of this cost would have been incurred anyway if a manual re-routeing had been undertaken. Even without taking this into account, a pay-back is expected within two years. The cost is little more than that of employing a specialist consultant for a one-off optimisation. Since most of the cost is in purchasing the software and setting up the initial data, we are now well equipped to deal with whatever changes might be sprung on us.

Andrew joined Suffolk County Council in 1990 as Group Manage (Maintenance) and has been responsible for planning all aspects of highway maintenance since that time.

Prior to this he spent 11 years with Hertfordshire County Council engaged on a range of activities but with an emphasis on highways maintenance planning and implementation.