Objectives:

W-SAHaRA focuses on developing stochastic computational methods and experimental procedures for prediction of flow, well catchment extent and contaminant time-of-residence in heterogeneous aquifers under data scarcity and uncertainty, in the presence of a variety of measured data, together with the quantification of the concept of risk associated with such predictions. The problem is attacked from various perspectives: (a) theoretical/conceptual; (b) laboratory and field scale; (c) deterministic and stochastic model development of synthetic and real-world cases; (d) increase of social acceptability of stochastic concepts in groundwater resources management. The developed concepts and models and experimental procedures serve as innovative tools for groundwater monitoring programs.

Scientific achievements:

W-SAHaRA produces enhanced basic understanding of issues related to reliable wellhead protection and risk assessment in randomly heterogeneous aquifers and substantial innovation in the methods and techniques for the development of combined theoretical and experimental tools to monitor and estimate groundwater flow and contaminant travel time in well fields. Scientific achievements of the second year are consistent with the general objectives of W-SAHaRA and are described in the following, according to the methodology adopted.

ü      Nonlocal moment equations: W-SAHaRA provides algorithms and analytical expressions for computing leading statistical moments of hydraulic head, fluxes, residence time and trajectories of contaminants under steady state flow to wells that pump water from a bounded, randomly heterogeneous aquifer. Analytical expressions are offered for two- and three-dimensional convergent (in the mean) flow scenarios of practical interest, when the natural logarithm Y of aquifer transmissivity (or hydraulic conductivity in three-dimensions) is modelled as a statistically homogeneous random field with a Gaussian spatial correlation function. Numerical solutions are offered for general flow conditions, taking full advantage of available information. This is achieved by developing procedures to condition predictors and the associated prediction errors upon hydraulic conductivity and/or hydraulic heads measurements and information rendering the spatial distribution of materials within an aquifer (such as data from geophysical exploration). Combined use of such information allows reducing uncertainty associated to predictions and the influence of each type of data is analyzed. The methodology also provides useful information about the appropriate choice of a sampling strategy within a heterogeneous aquifer. Our two- and three-dimensional solutions are supported by Monte Carlo simulations, which demonstrate (a) their applicability to strongly heterogeneous aquifers, characterized by large values of log transmissivity (or conductivity) variance and (b) their robustness to provide conceptual and theoretical insight into the nature of the solution.

ü      Monte Carlo approach: W-SAHaRA has investigated the capabilities of numerical Monte Carlo techniques to efficiently incorporate different types of data in modeling well protection zones and catchments under uncertain conditions. The influence of different geologic and hydrogeologic settings and processes on capture zones, examples of which are groundwater recharge and complex confining layers, have been analyzed. This will enable one to identify the key hydrogeologic parameters that influence capture zones and aquifer monitoring strategies. Key parameters concern the impact of recharge and boundary conditions on the capture zone. It was found that spatially variable recharge in most cases is not expected to considerably contribute to uncertainty in the well capture zone. Nevertheless, uncertainty in the mean areal recharge affects the capture zone significantly. Temporally variable recharge is important in case the catchment has groundwater residence times that are smaller than the time scale on which the recharge fluctuates. From the other studied key parameters, only uncertainty with respect to prescribed head values on boundaries, had an important impact on the well capture zone estimation.

ü      Laboratory and field experiments: For the experimental site chosen by W-SAHaRA in the Neckar Valley, close to Tuebingen (Germany) further hydrogeological characterization based on water level measurements, borehole core sample analyses, permeameter measurements, borehole flowmeter logs, small scale pumping tests etc. was obtained. A principal deterministic flow and transport model, needed for design and later evaluation of the field tracer experiments, has been set up. To reproduce the flow situation of a well field the tracer tests are run under convergent flow forced gradient conditions. Because different flow directions have to be considered, different tracers are used and an innovative multilevel-multitracer approach has been developed. To obtain a high temporal and spatial resolution, on-line measurement equipment is used (multilevel fiber optic fluorimeters for fluorescent tracers). Locations for additional groundwater wells for pumping, tracer injection and tracer breakthrough measurement were defined, as well as pumping rates, tracer masses and tracer test durations. Using the existing and the new drilled wells, a systematic procedure for additional aquifer parameter characterization has been carried out (water level measurements, flowmeter logs, pumping tests). The tracer test equipment was installed at the test site and a small scale field tracer test was performed. W-SAHaRA also performed a series of tracer tests within the laboratory scale model MARCEAUS (Institut de Mécanique des Fluides of the Université Louis Pasteur in Strasbourg – France). MARCEAUS is a three-dimensional box, which is filled with cubic blocks of sand and gravel to mimic the heterogeneity of an aquifer at the laboratory scale. The model has been equipped with pumping wells, and proper measuring instrumentation (to measure groundwater pressure heads and solute concentration).

ü      Determinstic model of real-world field: W-SAHaRA has identified a test site for deterministic modeling of wellhead protection zones in the area of Bologna (Italy), where a major extraction well field is located and exploited by one of the partners. Geological and hydrogeological reconstruction of the aquifer have been undertaken and its architecture has been identified. The probabilistic reconstruction of facies distribution within the first aquitard of chosen aquifer has been performed and a methodology has been refined to predict well catchments including these data within the models developed for predictors of catchments and contaminants trajectories. To model the groundwater flow in the Bologna aquifer the finite element code FEFLOW has been used. Data assimilation has been completed: data comprise general and thematic maps, distribution of extraction activities within the aquifer, location of monitoring wells, meteoric precipitation, hydrographic network, hydraulic conductivity. Information has been validated, and standardized in electronic archives. Implementation of the deterministic model is at an advanced stage.

ü      Public awareness and data dissemination: W-SAHaRA has launched an internet platform which is going to host the general ideas and main results of the project. W-SAHaRA has produced a critical review of stochastic methodologies for assessing uncertainty in wellhead protection, which has been posted on the Consortium web platform. In addition, a preliminary guidance document describing the implementation of the results of the W_SAHaRA work has been produced. This, along with a concise project brochure (currently in draft form) will provide end users with information on how to practically apply the outcomes from the project. Data obtained from the laboratory experiments are going to be published on the web site and will be freely accessible to users. W-SAHaRA has organized two special sessions at two key international conferences. Results have also been disseminated via publications in technical journals and articles in newspapers. Dissemination of results has also commenced at international conferences, on international journals and with graduation thesis and interdisciplinary Master thesis. Partners involved in the public administration are working in close contact to give large circulation to the concepts and results.

Socio-economic relevance and policy implications:

Wellhead protection zones serve as an effective protection of groundwater against pollution by regulating human activities in areas around drinking wells, therefore having large social and economic impact. Moreover, developing reliable strategies and methodologies for monitoring the groundwater resource is proving to be a major concern in all Europe. The socio-economical relevance of the project is evident, since a very high number of well protection zones have to be delineated in Europe in the next future. The end-users are not only engineers, hydrogeologists and other scientists, dealing with groundwater protection zones, but also officers from agencies, involved in regulations and the assessment of related projects. It is believed that the improved theoretical and experimental techniques and data developed by W-SAHaRA may be a valuable source of tools and information for the scientific community and private and public agencies.
 

Conclusions:

W-SAHaRA provides methodological advances including (1) the development of methods for effectively incorporating various type of (limited) available information in probabilistic analysis of wellhead protection and reduction of predictions uncertainty; (2) the development of simulation tools for well protection in a stochastic framework; (3) the development of innovative monitoring and investigation techniques at the laboratory and field scale, to quantify the effect of heterogeneity at various scales on predictions of groundwater flow and contaminant fluxes. W-SAHaRA is also working in the direction of accelerating social acceptance of stochastic concepts in groundwater management practice.
 

Keywords: Wellhead protection, well catchment, capture zones, random media, moment equations, Monte Carlo, laboratory experiments, field experiments, groundwater monitoring, deterministic models, groundwater, solute transport, conditioning, inverse models, heterogeneity.