Using a Bayesian belief network to diagnose significant adverse effect of the EU Water Framework Directive on hydropower production in Norway

Abstract

We evaluate the multi-criteria methodology employed by Norwegian authorities to screen and prioritise a large number of hydropower licences due for revision of their environmental and production conditions. We structured the national licence screening methodology using the multi-criteria decision analysis theory implemented in Bayesian Network (BBN) software. We show how BBNs can be used to diagnose the overall importance of different criteria implicit in the licence screening project's methodology. The diagnostic analysis finds that hydropower loss criteria were considerably more import than environmental criteria in explaining which licences were given lower licence revision priority. We also use the BBN to assess the trade-offs between environmental impacts and hydropower. Using the licence ranking by Norwegian authoities, we provide an interpretation of the ‘significant adverse effect’ for Norwegian hydropower of achieving Water Framework Directive. The national-level estimates of significant adverse effect are likely to be high, given the limited number of mitigation measures that were assessed in the national screening study. The paper discusses the importance of further regional and local assessment of all available measures to achieve good ecological potential in Norwegian water courses.