Real Time Control of a Well Field with a Groundwater Model

The risk that drinking water wells close to industrial areas and rivers pump contaminated groundwater can be minimised by developing an optimal pumping strategy in space and time.

Groundwater models are used to model the groundwater flow in the area around the drinking water wells. An improved groundwater model can be obtained by inverse modelling, which means that the model is consistent with measured hydraulic conductivities, time series of hydraulic head data and other experimental information like e.g. tracer tests. Such an improved model is a basis for predicting groundwater flow, contaminant mass transport and heat transport in the area of the drinking water well in the (nearby) future, probably for a changed pumping regime. As a consequence, it also serves as a basis for a decision on a pumping strategy in the (nearby) future. One of the objectives of such a strategy may be that not too warm water is pumped and that the water has a low risk of being contaminated. At the same time, new experimental information (e.g. from devices that measure groundwater temperature, electrical conductivity and pressure in real time) can be included and be used to improve the models by filter methods (Kalman filtering or variational methods).

We develop a 3D groundwater flow model of the Limmat Valley around the Hardhof area, where an important amount of groundwater is pumped for human consumption. A groundwater flow model is built that is consistent with measurements, including historical information of groundwater levels. Also a 3D solute transport model and a 3D heat transport model of the area are developed. In addition, an algorithm is developed to automatically update and improve the groundwater flow model with new measurement data. Finally, the prediction of groundwater flow, solute transport and heat transport on the basis of the improved models is used to optimise a pumping strategy, according to the criteria of the water works. Basically, it means reducing the risk of pumping contaminated water.

Researchers involved at the ETH:
Dr. Harrie-Jan Hendricks Franssen, IHW, ETH Zürich
Prof. Dr. Fritz Stauffer, IHW, ETH Zürich
Prof Dr. W. Kinzelbach, IHW, ETH Zürich

Other project partners:
Wasserversorgung Zürich
TK Consult AG, Zürich