Predicting the risk to permit compliance of new sewage treatment works

Abstract

The privatised water industry in England and Wales has the requirement to construct a substantial number of new assets to meet new national and European legislation. In order to do so economically, the risk of performance of new assets has to be matched to the cost of construction. To undertake such risk management, a risk technique has been developed that allows the performance against permit compliance of new processes to be established. The technique is based on statistical analysis of data effectively creating a pure black box model to determine performance. The technique has been further developed to take into account the dependency between influent quality and process performance. The dependency factor selected is the Spearman's Rank Coefficient which is a convenient way to define complex relationships. It can be used to define many dependent variables referring to the same independent variable and as such removes the need for complex mechanistic equations. This technique has been calibrated using two years, existing data and then validated for the next 6 months data of a wastewater treatment works in the North West of England. The results of the validation illustrate the ability of the technique to be used as a predictive tool. The output from the model allows the direct comparison of different process options and permits executive management to make informed decisions about investment against risk to permit compliance.