During periods of heavy rainfall storm sewage volumes can exceed the capacity of combined sewer systems and overflow to surface water bodies. Combined sewer overflows (CSO) cause significant impacts on the water quality and their identification is crucial to plan CSO control programs or to fulfil legal requirements. This paper proposes and demonstrates six different methods to identify the occurrence of CSO based on information on the sewer system alone (methods 1 and 2), in combination with rain data (methods 3 and 4) or in combination with water quality data of the receiving surface water (methods 5 and 6). The methods provide different information on CSO, from occurrence to pollution load and impacts in receiving surface water. The methods introduced have all been applied to the Berlin urban water system. Based on these experiences they are compared considering the effort needed for their application, the required data and the obtained output. It is concluded that certainty of CSO identification can be increased by combining some of the presented methods.

In the city of Berlin regular combined sewer overflows (CSO) lead to acute stress of aquatic organisms in the receiving River Spree and its side channels. Of most concern are oxygen depressions, following the inflow of degradable organic matter via ~180 CSO outlets, along a river stretch of 16 km. For the assessment of the severity of these oxygen depressions, an existing impact-based approach suggested by Lammersen (1997) was combined with information on the local fish fauna. Application of this locally adapted assessment method to seven years of oxygen measurements at a CSO hotspot in the river yielded an annual average of 14 periods with suboptimal conditions for which adverse effects on the fish fauna are expected and 20 periods with critical conditions for which acute fish kills are possible. Further investigation on rain and sewer management data proved that such critical conditions only occurred as a direct result of CSO events, whereas suboptimal conditions are also possible at dry weather and may last up to 32 days (Riechel et al. 2010).

Combined sewer overflows can lead to acute, critical conditions for aquatic organisms in receiving surface waters (Borchardt et al. 2007; FWR 1998; Harremoes et al. 1996; Krejci et al. 2004; Lammersen 1997). Based on the river type of the River Spree, CSO impacts of possible concern were identified to be high ammonia (NH3) and low dissolved oxygen concentrations (DO) (Senatsverwaltung für Stadtentwicklung 2001; Leszinski et al. 2007). For DO, existing continuous measurements from the River Spree from 2000 to 2007 were assessed in detail in the KWB report by Riechel (2009). However, Riechel (2009) neglected NH3 toxicity assessment, since no continuous NH3 measurements were available. The present report aims at filling this gap by estimating the potential for toxic NH3 concentrations in the River Spree with recent data. Based on stormwater impact guidelines for ammonia, critical total ammonium concentrations ([NH4,tot] = [NH4+] + [NH3]) were calculated and compared to continuous NH4,tot measurements in the Berlin River Spree. NH4,tot was measured i) at a heavily CSO impacted river stretch (year 2011) and ii) at a monitoring station several kilometres downstream of the combined sewer area (years 2010 and 2011). The analysis led to the following results: (i) Two years of continuous NH4,tot measurements showed clear increases in NH4,tot due to CSO but no occurrence of critical toxicity levels for cyprinid fish, according to Lammersen (1997) (ii) Maximal observed concentration of ~1.3 mg-N-NH4,tot l-1 was ~5 times smaller than the lowest existing threshold, which would need to be exceeded for 24 h to be considered as critical. The observed maximal concentration peak had a duration of only 3 h. The threshold, corresponding to the 3 h-duration would be even ~8 times higher than the observed ~1.3 mg-N-NH4,tot l-1. (iii) Ammonia toxicity would only be possible if maximal NH4,tot occurred during highest sensitivity of the river due to very high pH > 9. However, it was observed that pH drops significantly during CSO impacts due to low pH in rain water, which makes pH > 9 during CSO very unlikely. Given the results, the risk for ammonia toxicity due to CSO is judged as very low, particularly in comparison with regular problematic DO conditions after CSO events in summer.

During its passage through the City of Berlin (Germany), the quality of the River Spree is strongly influenced by combined sewer overflows (CSO), which lead to critical oxygen concentrations (DO) every year and to occasional larger fish kills. A continuous integrated monitoring concept, using state-of-the-art online sensors, was planned and started in spring 2010. It combines (i) continuous measurements of the quality and flow rates of CSO at one main CSO outlet downstream of the overflow structure and (ii) continuous measurements of water quality parameters at five sites within the urban stretch of the receiving River Spree. The first monitoring results show that continuous water quality measurements in CSO outlets and at downstream river stations are possible at high accuracy, even for comparably complex parameters such as chemical oygen demand (COD). Analysis of measured data confirms the significance of CSO discharges on receiving waters and underlines the value of continuous measurements in describing the local dynamics of the CSO and their impacts on water bodies.