In Berlin, drinking water is exclusively abstracted from groundwater. For this reason, well monitoring is important to ensure a good water quality and a high well capacity. The project WellMa, which stands for well management, aims at the optimization of the operation and maintenance of drinking water abstraction wells. During the first work package of the preparatory project phase WellMa1, available data of 615 wells in Berlin were evaluated. Since ochre formations appear most frequently, the context between iron-related clogging and hydrochemistry, well construction and operation was examined. Based on the statistical analysis, TV camera inspections were identified to be the most reliable marker to indicate the iron-related clogging status of a well. The results were classified into four stages and set into relation to relevant chemical, operational and constructional parameters to determine trends and correlations. Outliers lie below or above the detected trends, as given by the generated boxplots. Seven parameters have to be taken into account: iron, manganese and nitrate-nitrogen concentrations, top of the first filter, distance to the next surface water, mean discharge rate and monthly operation hours. Wells, whose outlier parameter based on one of the following criteria, were excluded from further investigations: (i) clogging state 1, which should be lower according to the outlier parameter, at a high remaining specific capacity, (ii) nitrate-nitrogen values that should lead to higher clogging degrees lying slightly out of the range given by the boxplot, (iii) classification of the clogging degree shortly after construction, regeneration or H2O2 treatment and (iv) missing information about clogging state in the BWB well files and missing TV inspection videos to verify deviating values. The analysis demonstrates a complex interaction between the single parameters. High nutrient inflow because of river bank filtration, leakages of the casing, high discharge rates or unconfined conditions superimpose the distance to the surface of a deep screened well. Since the encrustations are mediated by bacteria, the composition and permeability of the biofilm varies depending on the engaged microorganism. Furthermore, the thickness and dimension are influenced by the available time for bacterial growth, the inflow velocity, number of switchings or position of the water levels.
Determination of the clogging potential of drinking water wells in Berlin by an outlier analysis of well management data