Database development and 3D modeling of textural variations in heterogeneous, unconsolidated aquifer media: Application to the Milan plain

doi:10.1016/j.cageo.2007.09.006

Computers & Geosciences

Volume 35, Issue 1, January 2009, Pages 134-145

https://doi.org/10.1016/j.cageo.2007.09.006 Get rights and content

Abstract

The textural and hydrogeologic properties of loose deposits are closely connected to the sedimentary processes occurring in fluvial and glacio-fluvial systems. The proposed procedure combines geologic knowledge, coded well data logs stored in a hydrogeologic georeferenced database, geographic information system (GIS) and 3D calculation software to reconstruct the detailed distribution of the subsoil's hydrogeologic parameters. The calculations may apply to any subsoil part, bounded by one or more surfaces. This methodological approach may be applied to diverse investigation depths, vertical intervals detailed to varying extents, on a local or regional scale. Employing great quantities of well data logs, recorded in different ways and distributed across large areas, has enabled significant spatial sedimentologic reconstruction. Two groups of parameters were considered: textural (percentages of gravel, sand and clay) and hydrogeologic (hydraulic conductivity and porosity). In terms of sedimentology, the method reconstructs spatial heterogeneity both along cross-sections and for an entire user-defined volume to analyze the sediments’ energy functions and their distribution; deposition rates are proportional to process intensity and distribution with time. In terms of hydrogeology, detailed 3D hydrogeologic structural data provide detailed input for hydrogeologic models in groundwater resource management and planning, such as for transport models.

Introduction

Geologic processes and parameters are often expressed in a qualitative form, and converting them into a quantitative form may be difficult. The textural and hydrogeologic properties of loose deposits are closely connected to sedimentary processes occurring in alluvial and fluvial environments. A quantitative estimate becomes necessary in cases such as using hydrogeologic modeling to manage groundwater resources. Groundwater flow and travel time are dependent on stratigraphic architecture governed by competing processes that control the spatial and temporal distribution of accommodation and sediment supply (Edington and Poeter, 2006). Accommodation is the amount of space in which sediment may accumulate. The groundwater flow and transport models do not necessarily require a sedimentary basin model to be reconstructed, as they consider the deposits in hydrogeologic terms and identify them as a function of their defining parameters. However, as aquifers are heterogeneous in longitudinal and vertical directions, the consequences of spatial variability in aquifer properties are known to be a major limitation in using the models (Gòmez-Hernandez, 2003; Peck et al., 1988). Basin simulation includes spatial and temporal distribution of petrophysical parameters (Bitzer, 1999). It should also be kept in mind that transport behaviors, such as average fluxes and heads and overall plume movements, are well represented in upscaled models, and extreme flow and transport behaviors, such as early and late arrival times, local heads and velocities, are more strongly influenced by local discrete heterogeneities and are not well represented in upscaled systems (Scheibe and Yabusaki, 1998).

Furthermore, hydrogeologic flow models require parameters to be distributed in both space and depth. Generally, hydrogeologic investigations use geologic or seismic cross sections, field data and stratigraphic well data interpretations to determine parameters. Parameterization of aquifer materials is often achieved by assigning mean parameter values to hydrogeologic formations or to hydrostratigraphic units in the aquifers. These values are obtained by means such as pumping tests and field observations. This study's intent is to propose an improved definition of the spatial distribution of hydraulic conductivity and porosity based on differing percentages of heterogeneous materials, such as gravel, sand and clay, in fluvial and glacio-fluvial deposits. This objective can be achieved by quantitative three-dimensional (3D) processing of information in well logs, stored and codified in a hydrogeologic database. The well log data are often used to reconstruct the hydrogeologic sections, but the information is rarely translated into quantitative evaluations. This often has to do with its poor reliability and lack of homogeneity. This study shows that if they are analyzed in large quantities and over extensive areas, they will be considered a useful tool for investigation and quantitative parameterization.

Section snippets

Procedure

The proposed procedure evaluates the well logs, suggests a continuous and systematic method for storing the data, proposes a coding system for the stratigraphic descriptions and calculates the hydraulic conductivity distribution from the well logs as a function of 3D numerical modeling. The procedure links a computer-based hydrogeologic database to a 3D model through several steps.

The procedure is divided into three phases. The first phase is essential for making quantitative evaluations of the

Study area: province of Milan

The area under study is located in the Po Plain (Fig. 1) and is extremely dense in its urban, industrial and agricultural development. The province of Milan is about 1989 km², and 71% of its land is used for agriculture; 189 municipalities are located within the province's borders. A total of 7583 wells were identified in the province area, 3472 of which have a stratigraphic log. The area has both a natural hydrographic network and an artificial one. The natural network consists of the Adda and

3D textural modeling

3D modeling calculations pertain to two aspects: hydrostratigraphic unit reconstructions and groundwater flow models.

Groundwater flow model

The study's third phase connects the 3D hydrogeologic model to the groundwater flow model. This phase pertains to the unconfined portion of the subsurface. The hydrogeologic model is the primary element for setting up a groundwater flow model that respects the existing hydrogeologic complexity. In areas with a highly complicated depositional structure like the Po River plain, neither homogeneous and isotropic deposits nor a sharp division between an aquifer and an underlying one can be defined

Conclusions

This study originated from an attempt to reduce limitations due to a lack of knowledge about both the heterogeneous characteristics of the aquifers and the attenuating mechanisms of organic compounds in the water table that often lead to insufficient modeling applications. Experience has confirmed that the most significant deficiency, particularly in the case of pollutant transport in the water table, is associated with the level of detail with which the aquifer is represented.

The proposed

References (32)

K. Bitzer
Two-dimensional simulation of clastic and carbonate sedimentation, consolidation, subsidence, fluid flow, heat flow and solute transport during the formation of sedimentary basins
Computers & Geosciences
(1999)
J.L. Mallet
Discrete smooth interpolation in geometric modeling
Computer Aided Design
(1992)
T. Scheibe et al.
Scaling of flow and transport behavior in heterogeneous groundwater systems
Advances in Water Resources
(1998)
L. Aller et al.
DRASTIC: a Standardized System for Evaluating Ground Water Pollution Potential using Hydrogeologic Settings. Prepared for US EPA (Environmental Protection Agency) Office of Research and Development, Ada, OK
(1987)
M.P. Anderson et al.
Applied Groundwater Modeling
(1992)
Avanzini, M., Beretta, G.P., Francani, V., Nespoli, A., 1995. Indagine preliminare sull’uso sostenibile delle falde...
P.F. Barnaba
Considerazioni sui lineamenti idrogeologici del sud-est milanese: risorse idriche e falda superficiale (Considerations on the hydrogeological features of the south-eastern milan area: water resources and unconfined aquifer)
Acque Sotterranee
(1998)
G.P. Beretta et al.
Primo bilancio idrogeologico della Pianura Milanese (First hydrogeologic mass balance of the Milan Plain)
Acque Sotterranee
(1985)
A. Bondesan et al.
Regional infiltrability map of the Po and Veneto-Friuli Plain. Memorie Società
Geologica Italiana
(1994)
T. Bonomi et al.
Analysis of factors contributing to groundwater contamination risk in highly developed areas. A case study in the province of Milan
Terra Nova
(1994)

T. Bonomi et al.

Un database per pozzi: Tangram (A database for wells: Tangram)

Quaderni Geologia Applicata

(1995)

T. Bonomi et al.

3D subsoil parameterisation in a plain region of northern Italy

Memorie Società Geologica Italiana

(2002)

A. Cavallin et al.

Coefficenti di trasmissivita’, immagazzinamento e permeabilita’ dell’acquifero milanese con prove in situ (Trasmissivity, storage and hydraulic conductivity of the Milan aquifer using field tests)

Bollettino Associazione Mineraria Subalpina

(1980)

A. Cavallin et al.

Studio idrogeologico della pianura compresa fra Adda e Ticino (Hydrogeological study of the plain included between Adda and Ticino)

Rivista Costruzioni

(1983)

D. Edington et al.

Stratigraphic control of flow and transport characteristics

Ground Water

(2006)

C.W. Fetter

Applied Hydrogeology

(1994)

Cited by (43)

Insights into the hydrogeological framework of the NW Himalayan Karewas (India)
2021, Environmental Challenges
Citation Excerpt :
Geological and hydrogeological models form the backbone of numerical groundwater flow and transport models as demonstrated in many studies (Nury et al., 2009; Chesnaux et al., 2011; Alam and Umar, 2013; Qadir et al., 2016; Sahoo and Jha, 2017; etc). For example, Bonomi (2009) demonstrated the reconstruction of hydrostratigraphic units for the Milan plain in Italy and used the same for groundwater flow modelling of the region. The study shows the usefulness of hydrogeological framework models in assigning quantitative hydraulic parameters to the existent subsurface formations.
Karewas which represent the Quarternary sediments deposited in fluvio-lacustrine/fluvio-glacial environment and peripherally surround the valley of the Jhelum river basin in India are highly devoid of surface water resources. The present study is an investigation of the subsurface regime of watersheds of Jhelum basin, encompassing alluvial and Karewa (fluvio-lacustrine) sediments using various software-based techniques with an aim to understand its groundwater scenario. Geological and hydrogeological framework models were developed to infer vertical and horizontal extent of aquifers. The models predict multiple fine sand aquifers in the alluvial sediments, 0.3 to 48 m thick, encountered at 3 to 10 m below the ground level (m bgl); major potential aquifer of 10 to 50 m thickness in Upper Karewa sediments at a depth of 25 to 35 m bgl and about 20 m thick aquifer in Lower Karewa sediments. Water table location obtained by geospatial (stochastic and deterministic) interpolation of static water levels reveals shallow water table at 0.3 to 10 m bgl within almost the entire alluvium region, deepening to 30 m bgl within the Karewa formations. The confining heads vary from a few meters below the ground surface to a few meters above it within alluvium, and much higher within the Karewas attributable to the location of recharge zones within the tall mountains bordering the study area. The study provides a framework for numerical groundwater modelling of the region. The methodology outlined in this study can be applied to diverse geological environments across local or regional scales to understand the hydrogeological conditions therein.
A regional-scale conceptual and numerical groundwater flow model in fluvio-glacial sediments for the Milan Metropolitan area (Northern Italy)
2020, Journal of Hydrology: Regional Studies
Citation Excerpt :
For the Milan metropolitan area, several studies have been proposed, which combine geological and geophysical data for the characterisation of the textural variations of the fluvio-glacial deposits different at scales. However, the rigorous recognition of specific sedimentological and stratigraphic sequences, the definition of hydrofacies groups, and the validation by means of available datasets, add meaning and robustness to this work, highlight the difference with previous works in this area (Giudici et al., 2000; Bonomi, 2009; Giudici, 2010; Mele et al., 2012, 2013; Comunian et al., 2016). The Milan metropolitan area (Fig. 2a) is a portion (3135 km2) of the Po Plain (Northern Italy) bounded by the Po River to the south, the Adda and the Ticino rivers to the east and west, respectively, and by the Prealps foothills to the north.
The Milan metropolitan area lies on one of the most important aquifer in Italy, heavily exploited for public and industrial water supply. The area, covering 3135 km² in the Po Plain (Northern Italy) with a continental climate, is bounded by the Po, the Adda and the Ticino rivers and by the prealpine foothills. Regional hydrology is characterised by a network of natural and man-made elements, and lowland springs. The sedimentary sequence, from bottom to top, is formed by meandering river plain deposits, the distal fringe of the glacial outwash plains and proximal braid-plain deposits.
This study proposes a general approach for aquifer geometry reconstruction and hydrodynamic parametrization of hydrofacies in fluvio-glacial deposits, and their implementation into a 3D regional groundwater flow model. This approach is based on sedimentologically-defined lithofacies/hydrofacies and their correlation in space to obtain nearly homogeneous subunits starting from available data (i.e. 8628 borehole logs, grain size distributions, well tests) and sedimentological knowledge.
The calibrated 3D FEM groundwater model allows quantifying the main components of the hydrogeological budget at the regional scale, and the fluxes among the different hydro-stratigraphic units. A sensitivity analysis of groundwater levels to the main recharge components suggests importance of anthropogenic disturbances with respect to natural recharge, and that land-use change may impact water resources more than climate change.
Groundwater diffuse pollution in functional urban areas: The need to define anthropogenic diffuse pollution background levels
2019, Science of the Total Environment
Groundwater status in highly urbanized areas is particularly affected by anthropogenic influence due to diffuse pollution deriving from many sources. This makes very often challenging to determine whether the observed groundwater conditions are the result of localized pollutant sources (PS-Point Sources). In the EU legislative framework it is accepted that, when Natural Background Levels (NBLs) of undesirable elements are proven to be higher than specific Groundwater Quality Standards (GQSs), NBLs should be assumed as GQSs. No procedure is instead defined when anthropogenic Diffuse Pollution Background levels (DBPLs) are higher than GQSs and make unfeasible any remediation strategy. Among the many contaminants affecting groundwater, the chlorinated solvents, tetrachloroethylene (PCE), trichloroethylene (TCE) and trichloromethane (TCM) among the organics and hexavalent chromium, among the inorganics, having been widely used in several industries all over Europe, are very often the most prevalent contaminants in soil and groundwater. Aim of this paper is to discuss a multivariate statistical approach to address the issue of identification of anthropogenic Diffuse Pollution Background Levels. With such aim, an area of about 1600 km², including the Functional Urban Area of Milan, was considered and 10 independent geochemical datasets, provided by local and regional agencies, and covering the period 2003–2014 were merged into a single database after homogenization and multiple quality checks. A total of 618,258 chemical analyses from 3477 sampling wells were considered, being all samples collected and analyzed through internally consistent protocols. The analysis enabled to identify five main clusters, having specific hydrogeological characteristics, different temporal profiles and pollutant background concentration levels, which were also found to respond differently to meteo-climatic changes. This study offers a robust knowledge basis for drafting a diffuse pollution management plan of the area.
Pollutant sources in an arsenic-affected multilayer aquifer in the Po Plain of Italy: Implications for drinking-water supply
2017, Science of the Total Environment
Citation Excerpt :
The thickness of the silt/clay units increases from north (the Alpine foothills) to south (the Po River), reflecting the waning transport energy of glacial and post-glacial rivers (Ori, 1993). This geological setting is particularly evident in Lombardy Region (Fig. 1), where a monolithic aquifer of gravel and sand in the northern, higher, part of the plain passes southwards into a multilayer system around the Po River in the lower plain (Bonomi, 2009; Perego et al., 2014). In the multilayer system of lower plain, the deeper aquifers have a sluggish circulation and so longer residence times for groundwater under natural conditions of flow (Martinelli et al., 2014).
In aquifers 160 to 260 m deep that used for public water-supply in an area ~ 150 km² around the town of Cremona, in the Po Plain of Northern Italy, concentrations of arsenic (As) are increasing with time in some wells. The increase is due to drawdown of As-polluted groundwater (As ≤ 144 μg/L) from overlying aquifers at depths 65 to 150 m deep in response to large-scale abstraction for public supply. The increase in As threatens drinking-water quality locally, and by inference does so across the entire Po Plain, where natural As-pollution of groundwater (As > 10 μg/L) is a basin-wide problem.
Using new and legacy data for Cl/Br, δ¹⁸O/δ²H and other hydrochemical parameters with groundwater from 32 wells, 9 surface waters, a sewage outfall and rainwater, we show that the deep aquifer (160–260 m below ground level), which is tapped widely for public water-supply, is partly recharged by seepage from overlying aquifers (65–150 m below ground level).
Groundwater quality in deep aquifers appears free of anthropogenic influences and typically < 10 μg/L of As. In contrast, shallow groundwater and surface water in some, not all, areas are affected by anthropogenic contamination and natural As-pollution (As > 10 μg/L). Outfalls from sewage-treatment plants and black water from septic tanks firstly affect surface waters, which then locally infiltrate shallow aquifers under high channel-stages. Wastewater permeating shallow aquifers carries with it NO₃ and SO₄ which suppress reduction of iron oxyhydroxides in the aquifer sediments and so suppress the natural release of As to groundwater.
Arsenic release and attenuation in a multilayer aquifer in the Po Plain (northern Italy): Reactive transport modeling
2015, Applied Geochemistry
Groundwater As concentrations >WHO limit (10 μg/L) are frequently found in the Po Plain (N. Italy). Although several hypotheses on As mobilization exist (i.e., reductive dissolution driven by peat degradation), the mechanisms of As release and subsequent attenuation acting in the multilayer aquifer in the Po Plain were poorly understood.
The present work aims at implementing a reactive transport modeling of the aquifer system in Cremona, affected by As <183 μg/L, in order to quantify and test the feasibility of As release by the reductive dissolution of Fe-oxides driven by the degradation of peat contained in leaky aquitards and As attenuation downstream by the co-precipitation in iron sulfides.
The model, based on a partial equilibrium approach, revealed that the observed As, Fe and Mn chemistry could be mostly explained by the simultaneous equilibrium between Fe-oxide and sulfate reduction and FeS precipitation and by the equilibrium of rhodochrosite precipitation/dissolution. Model results, together with litholog analysis, supported the assumption of peat as the likely source of organic matter driving As release. The model fitted to measured data showed that the peak in the organic carbon degradation rate at 20–40 m below surface (average of 0.67 mM/y), corresponding to the shallow peaty aquitard and the upper portion of the underlying semiconfined aquifer, is associated with the peak of net release of As (average of 0.32 μM/y) that is followed just downstream by a net precipitation in iron sulfides at 40–60 m below surface (average of 0.30 μM/y). These results support the assumptions of peaty aquifers as drivers of As release and iron sulfides as As traps. The model also outlined the following aspects that could have a broad applicability in other alluvial As affected aquifers worldwide: (a) shallow peaty aquitards may have a greater role in driving the As release since they likely have young and more reactive organic matter; (b) the occurrence of Fe-oxide reduction and FeS precipitation, that represent the As source and sink, together with sulfate reduction occurring simultaneously close to equilibrium may restrict the As mobility limiting the extent of contamination just downstream the source of organic matter that drives its release.
Integrating deterministic lithostratigraphic models in stochastic realizations of subsurface heterogeneity. Impact on predictions of lithology, hydraulic heads and groundwater fluxes
2015, Journal of Hydrology
Citation Excerpt :
These have the advantage of being consistent with known geological relationships (stratigraphic, chronological, lithological, etc.) and/or with established conceptualizations of the geological system of interest (e.g., a certain depositional system). Because of the development of computer based 3-D geological modelling tools (Mallet, 2002; Turner, 2006; Turner and Gable, 2007; Kessler et al., 2009; Zanchi et al., 2009; Royse, 2010), deterministic models have become increasingly popular not only in the oil and mining industry (e.g., Xu and Dowd, 2003; Perrin et al., 2005), but also to support water resources assessment and management (Artimo et al., 2003; Ross et al., 2005; Campbell et al., 2010; Nury et al., 2010; Gill et al., 2011; Raiber et al., 2012; Carreño et al., 2014), and numerical simulation of hydrogeological processes (D’Agnese et al., 1999; Robins et al., 2005; Wycisk et al., 2009; Bonomi, 2009; Blessent et al., 2009; Giambastiani et al., 2012; Turner et al., 2014). In flow and solute transport modelling, in particular, deterministic 3-D geological models are considered valuable in the preparatory stage of the numerical modelling implementation to establish a sound geological framework for the development of the conceptual model (Robins et al., 2005; Bredehoft, 2005).
Realistic representations of geological complexity are important to address several engineering and environmental challenges. The spatial distribution of properties controlling physical and geochemical processes can be effectively described by the geological structure of the subsurface. In this work, we present an approach to account for geological structure in geostatistical simulations of categorical variables. The approach is based on the extraction of information from a deterministic conceptualization of the subsurface, which is then used in the geostatistical analysis for the development of models of spatial correlation and as soft conditioning data. The approach was tested to simulate the distribution of four lithofacies in highly heterolithic Quaternary deposits. A transition probability-based stochastic model was implemented using hard borehole data and soft data extracted from a 3-D deterministic lithostratigraphic model. Simulated lithofacies distributions were also used as input in a flow model for numerical simulation of hydraulic head and groundwater flux. The outputs from these models were compared to corresponding values from models based exclusively on borehole data. Results show that soft lithostratigraphic information increases the accuracy and reduces the uncertainty of these predictions. The representation of the geological structure also allows a more precise definition of the spatial distribution of prediction uncertainty, here quantified with a metric based on Shannon information entropy. Correlations between prediction uncertainties for lithofacies, hydraulic heads and groundwater fluxes were also investigated. The results from this analysis provide useful insights about the incorporation of soft geological data into stochastic realizations of subsurface heterogeneity, and emphasize the critical importance of this type of information for reducing the uncertainty of simulations considering flux-dependent processes.

View all citing articles on Scopus

^☆: This article presents a case study, in the province of Milan, where well data logs are processed to build a 3D reconstruction of aquifer characteristics and gain detailed hydrogeologic structural data as flux and transport model input.

View full text

Database development and 3D modeling of textural variations in heterogeneous, unconsolidated aquifer media: Application to the Milan plain☆

Abstract

Introduction

Section snippets

Procedure

Study area: province of Milan

3D textural modeling

Groundwater flow model

Conclusions

Computers & Geosciences

Computer Aided Design

Advances in Water Resources

DRASTIC: a Standardized System for Evaluating Ground Water Pollution Potential using Hydrogeologic Settings. Prepared for US EPA (Environmental Protection Agency) Office of Research and Development, Ada, OK

Applied Groundwater Modeling

Considerazioni sui lineamenti idrogeologici del sud-est milanese: risorse idriche e falda superficiale (Considerations on the hydrogeological features of the south-eastern milan area: water resources and unconfined aquifer)

Acque Sotterranee

Primo bilancio idrogeologico della Pianura Milanese (First hydrogeologic mass balance of the Milan Plain)