Water samples were taken at the KAUST RO plant, the WDRC pilot plant and three other full-scale desalination installations in Saudi Arabia. The water was characterized using selected microbiological parameters, being conventional (heterotopic place count (HPC), total coliforms, Escherichia coli) and more novel and sensitive methods (adenosine tri-phosphate (ATP, a measure for bacterial activity), as well as total and intact bacterial cell concentrations (TDC using flow cytometry) and supporting parameters (pH, conductivity, residual chlorine and temperature). Selective samples were used to quantify the bacterial growth potential (“food for the bacteria”), applying a flow cytometer based easily Assimilable Organic Carbon (AOC) assay. Hypothesized was that no or very low bacterial numbers would occur after RO filtration in the plants due to the high rejection properties of the RO membranes and the produced water exceptionally low mineral and nutrient content. Key findings are that the (i) RO permeate contains bacterial cell concentrations exceeding 1.0 × 103 cells/mL. The highest percentage of cells are intact and active, based on the ATP and total cell counts (ii) advanced microbial parameters ATP and TDC enabled to detect and quantify bacteria numbers and activity while the less sensitive conventional plate counts based techniques did not, (iii) flow cytometer-based growth potential measurements indicate the presence of 8 µg AOC/L in the RO permeate. A typical last step in drinking water production is chlorination, effectively inactivating all the bacterial cells. The origin of the bacterial cells and the biodegradable nutrients enabling the bacterial growth in the RO permeate is not clear. There is a clear need to assess the origin of the nutrients and bacteria found in the RO produced water. It is not expected to be passing the RO membrane.

The article focuses on the non-agent construction of the ASKRIPTION of aussehen (er sieht mitgenommen aus, see basically Lasch 2016a) in the center of the considerations. On the basis of a corpuslinguistic evaluation of the KERN corpus at the DWDS it is shown that and why this construction belongs to the repertoire of linguistic patterns, which are especially established for descriptions in narrative texts.