Challenging current practices for management of pollution in separate stormwater discharges

Urbanization, climate changes and the wish for healthier natural eco-systems have created a worldwide push for better stormwater management practices. For separate sewer systems, stormwater may adversely affect receiving waters as the runoff from built-up urban surfaces disrupts the natural flow regime and acts as a vector for diffuse urban pollutants.

Stormwater Control Measures (SCMs) are applied to mitigate such effects by alleviating peakflows and providing treatment before discharge. Planning efforts for better stormwater management are influenced by legislative and regulatory frameworks, which encourage holistic management. However, in the EU, diffuse urban pollutants seem to be overlooked in practice.

Also, stormwater quantity and quality management efforts are encouraged under separate regulatory frameworks, which mean that coordination is challenged. When planning for better stormwater management, an additional challenge is how to address the variation in pollutant concentration levels. As it is impossible to treat all stormwater runoff, flush-based management strategies can be applied to intercept pollutant flushes.

The concept of first flush, where the initial part of the runoff holds the majority of the transported pollutants, is a long-term standing assumption used when designing urban drainage systems. However, the relevance of the first flush phenomenon in separate sewer systems is under discussion, and new approaches may be needed for working with pollutant flushes.

Furthermore, variations in pollutant levels are generally not considered when assessing if SCMs are suitable for ensuring that runoff discharges do not compromise Water Quality Standards (WQS) in receiving waters. This PhD thesis investigated the current practice of pollution management for separate system discharges of urban stormwater: (1) by analyzing the legislative and regulatory framework for its ability to limit pollutants in these discharges and for promoting coordination of stormwater quality and quantity management; (2) by scrutinizing the assumption of first flush found in the literature and exploring alternative analysis approaches to work with pollutant concentration variations; and (3) by investigating impacts of pollutant variations for assessments of SCM performance.

To analyse the regulatory framework for managing stormwater pollutants, reviews were carried out using Denmark as a case study on (i) national implementation of relevant EU directives and (ii) the practical application of stormwater regulation as implemented in discharge permits. It was found that a legislative gap for how to condition stormwater discharge permits means that regulation relies heavily on requirements based on Best Available Technology (BAT).

Also, missing standards and procedures give very little room for innovative SCMs to be applied. Assessments of effects from discharges on chemical status in the recipients are not carried out, and in general, there is a much larger focus on regulating for hydro-morphological impacts than biological or physicochemical impacts.

Furthermore, it was found that separate stormwater discharges are not included in holistic plans for river basin management and that opportunities for coordinating initiatives from the Water Framework Directive and the Flood Directive are limited. To scrutinize the flush-based management strategies, a review was carried out on flush phenomena in the scientific literature as well as on the application of first flush based strategies in SCM design guides.

It was found that the multitude of different definitions of flush complicate comparisons of studies and generalisations of findings. While impacting factors of flush could be determined in simple systems, such as a single rooftop, SCMs are often being installed right before recipient discharges downstream of complex systems.

Here at catchment outlets, assessments of flush signals showed high variability, indicating that general first flush assumptions cannot be made. These were, however, found in the guidelines applied for SCM design in several countries. Furthermore, a new method for analysing flush variations was explored based on Functional Data Analysis (FDA), which successfully performed automated flush signal classification.

To investigate the impacts of pollutant level variations on SCM performance assessments, a simple conceptual model was set up for four SCM archetypes. The model was run with time series of runoff flow as well as time series of concentrations for three different indicator pollutants (Cu, BaP and Diuron) that were created with different resolutions.

It was found that using coarse input resolution to the model, and thus not considering the dynamic nature of pollutant fluxes, can affect the assessment of compliance with WQS as well as influence the choice of SCM. Applying high input resolution is especially critical for pollutants that are present in concentrations close to limit values.