Executive Summary

Conventionally, sub/non-critical sewers have not been maintained proactively, however with the added emphasis on this statutory duty, water companies now find it economically viable to maintain certain sub-critical pipes proactively. Using often limited historical event data as the only available driver, the next step is for the water companies to determine which of these assets will benefit from proactive maintenance.

In Scotland, where this project is based, the water industry has been historically underfunded in comparison with other parts of the UK and many catchments do not yet have sufficient event data to allow hotspots to be analytically defined based on failure statistics. Additionally, failure patterns may be restricted to the catchment on which the study was undertaken due to the individual characteristics of the network and its interaction with the topography and population of the catchment and therefore the transfer of patterns from one catchment to another is difficult. Furthermore, within Scotland, the water industry is faced with the added complication of small lateral sewers, running from individual property boundaries to the main sewers, which are the responsibility of Scottish Water. In England and Wales, most of this sewer type, is considered private and is therefore the responsibility of the property owner. This adds significantly to the length of network for which Scottish Water is responsible. Furthermore, it is this section of network which is known to produce most serviceability problems.

This project has the aim of augmenting sewerage asset management tools developed at Heriot-Watt University by attempting to link asset age to failure likelihood. However, the history of sewerage infrastructure development and maintenance in the UK means that asset age is often unknown. Based on this consideration, the philosophy which underlies the methodology adopted in this project was that the age of the development on the surface can be used to infer the age of the sub-surface sewerage assets.

Specifically, based on a series of recognised historical maps of Edinburgh, a GIS representation of its surface was constructed where 40258156m2 had a tag which indicated the point in time at which it was originally developed. These tags were then used to infer the age of the sub-surface assets. This then made it possible to analyse the linkages between asset age and failure likelihood.

This analysis found that there was a strong correlation between the inferred asset age and failure likelihood. This was found to be true when failures were considered in terms of “fails per unit of surface area” and “fails per length of pipe”.

This result is important as it will offer asset managers significant assistance in planning sewerage asset maintenance – particularly when combined with the failure likelihood model developed at Heriot Watt University which uses data such as conduit size, population density and hydraulic characteristics.

Despite the significance of the result, the analysis undertaken to date does not indicate why older assets should be more likely to fail or which are most likely to fail. Based on this, a key conclusion of the reported work is that further focussed activity is required in this field of applied research.