Optiwells-2 Synthesis report

Zusammenfassung

Objective of this synthesis report is to summarise the main achievements of the OPTIWELLS-2 project. Based on a preparatory phase OPTIWELLS-1 (2011-2012), the main project phase OPTIWELLS-2 (2012-2015) included the development of two different optimisation modelling methodologies (data-driven, process-driven) for minimising a well field’s specific energy demand whilst satisfying both, water demand and water quality constraints. Chapter 2 gives a short overview on the technical background on pipe hydraulics and the general methodology used within the project. The general workflow of the testing and application for the three case study well fields investigated within OPTIWELLS-2 is summarised in Chapter 3. For the first two case studies (Chapter Fehler! Verweisquelle konnte nicht gefunden werden. and Fehler! Verweisquelle konnte nicht gefunden werden.), a process-driven modelling approach was used, which enabled the assessment of three different management strategies (smart well field management, pump renewal or a combination of both) on the specific energy demand. This approach was more time and data-demanding (Chapter 2.5) compared to the data-driven approach used for the third case study (Chapter Fehler! Verweisquelle konnte nicht gefunden werden.). The cross-case analysis (Chapter 4) showed, that the energetic prediction accuracy of process-driven modelling (Chapter 4.1.3) was improved significantly by using pump characteristics derived from audits instead of relying on manufacturer data, whilst including steady-state well drawdown compared to assuming a static water level in the production well was much less important. This can be explained by the fact, that well drawdown contributed to less than 3% of the required pump head (Chapter 4.1.1), whilst the offset between audit and manufacturer pump characteristics is much more relevant because of pump ageing during long usage periods (up to 40 years). The data-based modelling approach used for Site C has yielded energy consumption forecasts with a similar accuracy, but is more robust as it relies on operational data, thus requiring no calibration.

Publikation
Kompetenzzentrum Wasser Berlin gGmbH
Rustler, M.
Rustler, M.
Data scientist

My research interests include reproducible research, data management and programming (R & Python).

Sonnenberg, H.
Sonnenberg, H.
Researcher