Conceptual Model Development

4.1 Introduction

4.1.1 General

The development of a conceptual model is one of the most important aspects of a groundwater modelling study, for the conceptual model is the basis of analytical and numerical models that are formulated to replicate field groundwater conditions.

This section describes the nature and formulation of conceptual models. This section provides an overview of the information and processes that need to be considered in constructing a conceptual model for groundwater flow and contaminant transport models, respectively.

This section also discusses potential errors in the development of a conceptual model which could significantly affect the modelling results and should be avoided.

Anderson and Woessner (1992) provide guidance on the development of conceptual models for groundwater flow modelling. Zheng and Bennett (1995) give useful guidance on developing conceptual models for contaminant transport modelling.

4.1.2 Definition

A conceptual model is a simplified representation of the essential features of the physical hydrogeological system, and its hydraulic behavior.

In scientific terms, a conceptual model is a hypothesis which is formulated on the basis of the available data, experience and the professional judgment of the modeller.

4.1.3 Formulation & Use of the Conceptual Model

The conceptual model is based on an initial literature review, data collation and hydrogeological interpretation. The conceptual model should:

• Reflect the modeller's concept of the natural system.
• Represent the crucial factors/processes influencing groundwater flow and contaminant transport.
• Include sufficient detail to address the questions and modelling objectives (refer Section 2)
• Enable critical questions to be answered by the examination of the mathematical model based on the conceptual model.

Many simplifying assumptions need to be made partly because a complete reconstruction of the field system is not feasible, and partly because there are rarely sufficient data to completely describe the system in comprehensive detail.

The conceptual model should be developed using the principle of simplicity (or parsimony), i.e., the model should be as simple as possible, while retaining sufficient complexity to: (i) adequately represent the physical elements of the system; (ii) reproduce the system behavior to be studied; and (iii) facilitate answering the questions related to the modelling objectives (MDBC, 2001).

The modeler should avoid over-simplification, which results in a model that is incapable of simulating observed conditions adequately. At the same time, incorporation of too much complexity may result in a non-transparent and unwieldy model which is not suitable to study a relatively simple problem.

Like all hypotheses, the conceptual model has to be tested. This can be done by converting the conceptual model into a mathematical model and calibrating the model against field observations. If the mathematical model cannot be calibrated satisfactorily this may indicate a flaw in the conceptual model and the conceptual model will have to be revisited and modified (see Box 7 in Figure 2-1).

4.1.4 Limitations of Conceptual Models

A conceptual model rarely explains all field observations and the development of the conceptual model must be an iterative process; it should continually be updated as new data become available, as the understanding of the system is improved, or as questions and modelling objectives evolve. In other words, a conceptual model represents a "work-in-progress" which should be continually challenged by the modeller throughout the modelling project.

A conceptual model may be imperfect and may even be wrong because (NGCLC, 2001):

• The information used to define a problem is incomplete
• Incorrect assumptions are made in developing the conceptual model (e.g. can it be assumed that a sand and gravel deposit identified in three investigation boreholes extends laterally below the whole extent of the site, or have three separate sand and gravel lenses been penetrated?)
• The conceptual model ignores key processes influencing the process to be simulated (e.g. irreversible sorption in solute transport is ignored)
• The physical and/or chemical processes occurring are poorly understood.

4.1.5 Documentation

The conceptual model should be described in a stand-alone Conceptual Model Report or in a specific section of the Groundwater Modelling Report.

The conceptual model should be described in words and supported diagrams, figures, graphs, and tables. The conceptual model is usually presented visually as a series of 2-D cross-sections or as a 3-D block diagram, with supporting text and data tables that describe and quantify the components and features of the model. Examples of visual illustrations supporting conceptual models for different mining projects are shown in Figure 4-1 to 4-4 (see section 4.2).

The data used for development of the conceptual model should be readily available to the reader, either in the Conceptual Model Report , in a preceding section of a more comprehensive modelling report, or in a stand-alone document (data report) accessible to the reviewer.

The preparation of plans, contour maps, cross-sections and block diagrams is essential in the development of a conceptual model as it enables others to gain a rapid understanding of the system. Visualization of important aspects of the conceptual model may also highlight data gaps and inconsistencies and provides a method for checking that any assumptions make sense in the light of existing data (NGCLC, 2001).

Simplifications and assumptions should be documented and supporting information provided to justify that assumptions are reasonable. The conceptual model is a simplification of the complexity of field conditions. Hence, conceptual cross-sections and/or block diagrams typically have a greater degree of abstraction than those prepared for detailed field investigations.

In more complex projects, several illustrations will be required to provide a good illustration of the conceptual model. For example, a combination of plan view and cross-section may be needed to adequately illustrate a complex 3D groundwater flow field (see figures 4-3a/b). In projects with a contaminant transport component (e.g. Rum Jungle mine in figures 4-3a/b) separate illustrations may be required for groundwater flow and contaminant transport.

In describing and documenting the conceptual model the following should be included: (NGCLC, 2001):

• What is known and understood about the site. Supporting data, variability, and calculations should be provided
• What is not known or not understood about the site and whether it is thought to be important.
• What are the uncertainties in the data. This should also include a list of further data requirements ("data gaps") and proposed sampling or investigations to obtain essential information
• What has been assumed about the system. Justifications for decisions and any supporting data or calculations should be provided.
• What has been ignored or simplified in order to answer questions. Justifications for decisions and any supporting data or calculations should be provided.

The following should be documented as part of the documentation supporting the description of the conceptual model:

• Site description and history of resource development (if applicable).
• Hydrology & climate.
• Regional geology, including geological origin and spatial distribution of major bedrock units.
• Site geology, including type and spatial distribution of overburden (sediments), spatial distribution of main lithologies (bedrock); degree and depth of bedrock weathering, description of bedrock quality (geotechnical bedrock logging),
• Structural geology, including presence/alignment of main structures (faults, synclines/anticlines, bedrock contacts etc.) and local and/or regional anisotropy (e.g. from stress analysis, fracture mapping, pumping tests etc.).
• Maps of groundwater elevations and piezometric surfaces and hydrographs of groundwater elevations.

The conceptual model report (or section in a more comprehensive modelling report) should identify critical data gaps and provide recommendations on how to fill those data gaps.

4.2 Examples of Conceptual Models

Figures 4-1 to 4-4 provide examples of visual illustrations supporting different conceptual models of groundwater flow and contaminant transport for mining projects.

Figure 4-1: Conceptual model for groundwater flow at the historic Mt. Morgan mine site (reproduced from Wels et al., 2006).

Figure 4-2: Conceptual model for groundwater flow in mountainous terrain in northern British Columbia.

Figure 4-3a: Conceptual model for Groundwater Flow and Contaminant Transport at the Rum Jungle mine site in plan view (Robertson GeoConsultants Inc., 2012).

Figure 4-3b: Conceptual model for Rum Jungle mine site showing hydrostratigraphic units in cross-section (Robertson GeoConsultants Inc., 2012).

Figure 4-4: Conceptual model illustrating complex relationships between limnological and geochemical controls on pit lake water quality (Reproduced from Bowell et al., 2002).

More details on all four examples are provided in the following sections of conceptual model development.

4.3 Conceptual Models for Groundwater Flow

4.3.1 General

This section summarizes the information and processes that need to be considered for the development of a conceptual groundwater flow model. A comprehensive check list of specific information that should be considered for the conceptual model is provided in Appendix D.

In developing (or reviewing) the conceptual model, the modeller (or reviewer) should ask what evidence there is to support the conceptual model (e.g. head measurements, hydraulic testing), and whether the conceptual model for this project site is plausible compared to experience at nearby sites or sites in similar setting (reference or analogue site).

4.3.2 Hydrogeologic Setting

The extent of data collection and review related to the hydrogeological setting depends on the local site conditions. For example, a proposed groundwater extraction project in a shallow aquifer consisting of fluvio-glacial sediments may require an in-depth review of the geological origin of sediments and seasonal climate conditions while a proposed underground mine in deep bedrock will require a more detailed review of bedrock geology and structural analysis.

At sites with project history (e.g. existing mine sites, brownfield sites and/or groundwater aquifers with past abstractions) a detailed review of past resource development(s) and the resulting response of the aquifer system is a critical component of model conceptualization.

4.3.3 Model Domain

The conceptual model domain must be defined. In most cases, the domain of the conceptual model will be the same as the domain used for numerical models. The conceptual model domain may be larger than the numerical model domain, for example if a project is modeled using several "sub-models" to cover different aspects of a single site (i.e. separate numerical models for an open pit and a TSF).

The size of the model domain depends on the scale of the project (local, intermediate, regional) and the spatial extent of anticipated impacts. Common model domains used for groundwater modelling in large groundwater resource projects include:

• "Aquifer" model, which uses the known (or inferred) spatial extent of the main aquifer of interest to define the model domain
• "Watershed" model, which uses the watershed (or sub-watershed) in which the project is located as a convenient model domain,
• "Local" model, which defines the model domain based on the specific project component(s) to be studied (e.g. an open pit or a tailings impoundment)

The watershed model is the de facto default model for many natural resource projects. However, this model domain may not always be the appropriate model domain. For example, the scale of a watershed (or even sub-watershed) model may be too large to adequately model a local groundwater flow or contaminant transport problem. A watershed model with a size of >100 km² usually does not have the spatial resolution to predict contaminant concentrations in groundwater discharging in specific salmon spawning areas. The conceptual model documentation should justify the scale of the model domain and discuss the potential implications for scale-dependent modelling results (e.g. contaminant transport).

4.3.4 Model Boundaries

The conceptual model documentation should justify the selection of boundaries. The most common boundaries used for groundwater modelling in the larger groundwater resource projects include:

• Watershed divide (representing lines of flow divergence)
• Streams and/or valley centers (representing lines of flow convergence)
• Large water bodies such ocean, lakes, rivers (representing areas with constant and/or known hydraulic head)
• Geological boundaries such as bedrock contact or faults (representing large-scale features of known or assumed hydraulic behavior)
• No-flow conditions perpendicular to streamlines

The conceptual model documentation should identify the monitoring data (or other field evidence) used to select a certain boundary condition. If no data or observations are available, the rationale should be given for the assumed model boundary and the implications of the assumption on modelling results.

The conceptual model documentation should discuss potential changes in the model boundary condition(s) over time due to natural variations (seasonal flow field, climate change) and/or project development. For example, the use of a no-flow boundary to represent a surface watershed divide may be adequate for current conditions but may not be correct for modelling of a large open pit (or underground development) which can result in significant drawdown (cone of depression) reaching hundreds of meters (or even kilometers) beyond watershed divides. In this case the model boundaries would have to be set at sufficient distance such that current and future stresses can be modeled without artificial boundary effects (see case study in Appendix C2 for such an example).

4.3.5 Hydrostratigraphic Units and Hydraulic Properties

The conceptual model formulation should include consideration of (and the report should describe) the presence and spatial distribution of major hydrostratigraphic units and their hydraulic properties. Hydrostratigraphic units are a specific geological material (or a group of materials) that has sufficiently similar hydraulic properties that they can be considered a hydraulic unit for the purpose of a hydrogeological study.

In more complex settings the spatial distribution of hydrostratigraphic units should be visualized using cross-sections and/or 3D block diagrams (e.g. see figures 4-1 and 4-2).

A conceptual model should subdivide the main hydrostratigraphic units at a site into: main aquifer units (i.e. where most groundwater flow occurs); aquitards (with limited groundwater flow); and aquicludes (with insignificant groundwater flow). The selection of hydrostratigraphic units must be justified and documented. If different units are lumped together (e.g. due to lack of data and/or model convenience) provide justification.

The conceptual model should include consideration of (and the report should describe) the hydraulic properties of major hydrostratigraphic units, including, if applicable:

• Hydraulic conductivity (K)
• Storage parameters specific yield (Sy)
• Specific storage (Ss and effective porosity (n_eff ).

The conceptual model formulation should include consideration of (and the report should describe) how properties are determined for each unit. Preferably, hydraulic properties selected in the conceptual model should be based on site-specific hydraulic testing. The conceptual model formulation should include consideration of (and the report should describe) the source and uncertainty of available field data and, to the extent possible, provide summary statistics of the various hydraulic properties.

At complex sites, a visual representation of the available field data used for the conceptual model is recommended (e.g. scatter plot of K versus depth; histogram of K statistics by bedrock lithology).

Natural porous media and fractured rock have significant "heterogeneity" (variability in space) and "anisotropy" (variability with direction) of aquifer hydraulic conductivity and other properties. The conceptual model should include consideration of (and the report should describe) the heterogeneity and anisotropy of the main hydrostratigraphic units of the model which are defined as follows:

• Heterogeneity is the variation of key parameters such as hydraulic conductivity within a hydrostratigraphic unit. A heterogeneous unit has a large range of hydraulic conductivity with no discernible spatial pattern (e.g. a debris flow or moderately fractured bedrock). The degree of heterogeneity is scale-dependent; it has a strong effect on small-scale flow paths and interactions with sources and sinks and boundary conditions. In fractured bedrock, flow occurs mostly in fractures and not in the unfractured matrix, thus the heterogeneity of fractured rock aquifer refers to variation in fracture properties and interconnectivity.
• Anisotropy is the preferred spatial orientation of hydraulic conductivity and resulting preferred direction of flow within a material. Examples of anisotropy include permeable bedding planes within sediments, or preferred orientation of permeable interconnected fractures in rock. The degree of anisotropy must be defined in the conceptual model, and will likely be adjusted in mathematical model calibration.

Strong anisotropy should not be introduced into the model if not clearly supported by geologic data. The conceptual model documentation should discuss evidence of heterogeneity and anisotropy in the field data (or lack thereof) and explain how heterogeneity and/or anisotropy are accounted for in the model. If heterogeneity and/or anisotropy are not accounted for in the conceptual model, this should be justified and potential implications for modelling results should be discussed in the model documentation.

4.3.6 Groundwater Recharge

Recharge is defined as the downward flow of water reaching the water table, adding to groundwater storage (Healy, 2010). Groundwater recharge occurs through diffuse and focused mechanisms (see Figure 4-5). "Diffuse recharge" is distributed over large areas in response to precipitation infiltrating the soil surface and percolating through the unsaturated zone to the water table (Healy, 2010). Diffuse recharge is also referred to as local recharge or direct recharge. "Focused recharge" is the movement from surface water bodies, such as streams, canals, or lakes to an underlying aquifer. Focused recharge is also referred to as indirect recharge or leakage.

Figure 4-5: Vertical cross-section showing infiltration, drainage, aquifer recharge, and inter-aquifer flow (Reproduced from Healy, 2010).

Figure 4-2 illustrates different sources of diffuse recharge that were included in the conceptual model for a mountainous site in northern B.C., including precipitation and snowmelt runoff from the mountain sides and summer snowmelt/ice melt runoff from a glacier.

Natural resource projects typically include "artificial recharge" (also referred to as "anthropogenic recharge") such as seepage from flooded open pits or underground workings, mine waste facilities (e.g. tailings impoundments, mine rock piles, storage dams etc.), recharge from land application areas, and/or recharge from injection wells.

Figure 4-1 shows an example of artificial recharge from a flooded open cut at the Mt Morgan mine site which was partially backfilled with reprocessed tailings. A detailed analysis of water levels and water quality indicated that the contaminated groundwater stored in the backfilled open pit was a major source of seepage to the downgradient aquifer system. This artificial recharge represented a major aspect of the conceptual model for the Mt Morgan mine site.

Recharge is usually expressed as a volumetric flow, in terms of volume per unit time (L³ /T), such as m³ /d, or as a flux, in terms of volume per unit surface area per unit time (L/T), such as mm/yr. These guidelines recommend that diffuse recharge values be described in terms of flux (in mm/yr) to allow a direct comparison to precipitation data. In the context of water balance (see below) recharge should be described in terms of volumetric flow (in m³ /day).

Recharge from precipitation is the primary inflow to groundwater systems in many parts of British Columbia and therefore requires careful consideration in formulating a conceptual model. The conceptual model should account for the seasonal behavior of precipitation (i.e. snowmelt and rainfall) and evapotranspiration and how these seasonal variations influence groundwater recharge. A visual comparison of observed seasonal variations in groundwater levels with daily or monthly precipitation data (and snowmelt data for sites with significant snowpack accumulation) over a period of at least one year is recommended for this analysis.

When formulating a conceptual model, the following factors that may influence recharge from precipitation, should be considered:

• Slope
• Ground conditions
• Vegetation
• The spatial distribution of recharge
• The influence of slope aspect, forest cover, elevation, etc. on snowmelt.

Recharge cannot be measured directly and must be estimated. Healy (2010) provides a comprehensive review of the different methods available for estimating recharge. The most widely used methods for estimating recharge include:

• Water-budget methods
• Water table fluctuation (WTF) method
• Baseflow analysis
• Chloride mass balance (CMB) method.

The method(s) most suitable for estimating recharge for a given project depend on local site conditions and available data. Consult Healy (2010) to determine the most suitable method of estimating recharge for a given project. This book also provides a good discussion of the assumptions and limitations of the various methods.

Recharge from precipitation is often one of the greatest uncertainties in groundwater modelling studies. These guidelines recommend that different techniques be used to estimate groundwater recharge due to precipitation. This will provide a measure of the uncertainty in recharge estimate(s). This uncertainty in recharge will have to be evaluated further during mathematical model calibration and sensitivity analysis (see Section 6).

Formulation of the conceptual model should include consideration of the presence, magnitude, and duration of any "artificial recharge" events observed/anticipated in the past, present and/or in the future life of the project. Recharge estimates should be documented (in the conceptual model report) for major artificial recharge processes known to be active (or anticipated in the future) at the site. The recharge flux in waste rock piles, tailings, or other artificial covered and uncovered materials must be estimated (typically independently of the groundwater model) or measured directly to allow proper simulation of local flow conditions and response to stresses.

Focused recharge from surface waters (e.g. streams, lakes etc.) can also be an important recharge mechanism to groundwater, in particular, in drier parts of British Columbia. Recharge from surface water(s) is also referred to as "leakage" in the context of groundwater modelling and is discussed further in the section on groundwater-surface water interaction (see below).

4.3.7 Groundwater Flow Regime

The conceptual model should represent and the supporting documentation should provide a qualitative and pictorial description of the groundwater flow regime, including:

• The main direction(s) of groundwater flow, including a description of groundwater flow from the main recharge areas to the main discharge areas (including major internal sinks such as pumping wells).
• The location of the groundwater table (depth to water) and flow field (water table map for unconfined aquifers and potentiometric map for confined conditions.
• The horizontal and vertical hydraulic gradients in different parts of the aquifer.
• Estimates of travel times/residence times of groundwater (using Darcy calculations).
• Conceptualization of groundwater flow through the various hydrostratigraphic units (which units carry the majority of flow, which units impede flow etc.)
• Variations in groundwater levels over time (e.g. seasonal or ongoing abstraction).

Figure 4-3a illustrates the groundwater flow regime for the historic Rum Jungle mine site (Robertson GeoConsultants Inc., 2011). At this abandoned mine site, the groundwater flow regime is strongly influenced by seepage from the "Main Heap" which has resulted in local mounding of the groundwater table. Seepage from this and other heaps flows towards the East Finnis River which runs through the site. Note that the groundwater flow regime is also influenced by the flooded open pits which intersect the bedrock aquifer.

The conceptual model documentation should describe observed temporal trends in groundwater levels (preferably by using time trend plots) and discuss the cause(s) for these trends. The conceptual model documentation should discuss whether the groundwater flow field can be considered in a dynamic steady-state or whether transient aspects have to be considered. If a steady-state conceptual model is recommended, the limitations and implications for this simplifying assumption should be discussed (see section 5.2.2 for more detail).

4.3.8 Groundwater Discharge

In formulating the conceptual model, the following groundwater discharges should be considered (and documented):

• Seeps and springs discharging to ground surface
• Groundwater discharge into surface waters (lakes, streams)
• Groundwater abstraction (pumping)
• Seepage interception systems (drains, interception wells)
• Inflow to open pit and/or mine workings
• Evapotranspiration (ET).

Groundwater discharge can be measured either directly (e.g. using flow meters, seepage meters, seep surveys etc.) or indirectly (using observed hydraulic gradients and/or water quality).

Figure 4-1 shows an example of an abandoned mine site where groundwater discharge was an important aspect of the conceptual model. The conceptual model for this site included (i) discharge of highly contaminated seepage from an upgradient mine waste pile into a flooded open pit, (ii) discharge of highly contaminated seepage into several sumps (seepage interception) and (iii) discharge of impacted groundwater to the receiving environment (Dee River).

Groundwater discharge is usually expressed as a volumetric flow, in terms of volume per unit time (L³ /T), such as m³ /d. These guidelines recommend that groundwater discharge be described in terms of volumetric flows (in m³ /d). ET should be expressed in terms of unit flux (in mm/yr), to allow direct comparison with climate data, except in discussions of water balance (see below) where ET should be described in terms of volumetric flow (in m³ /day).

The conceptual model documentation should identify the main groundwater discharge processes in the study area, describe their seasonal behavior and discuss the factors influencing groundwater discharge. To the extent possible, the various groundwater discharge components should be quantified using actual measurements (preferred) or estimates.

Groundwater discharge to surface waters is of special concern in the context of groundwater resource projects (as a potential pathway for contaminants). This groundwater discharge mechanism is described in more detail in Section 4.3.9 on groundwater-surface water interaction (see below).

Evapotranspiration (ET) can be an important groundwater "discharge" mechanism, in particular in warmer and drier climates which exhibit a net negative water balance (i.e. ET >> precipitation). Evapotranspiration is defined as the removal of water by the combined effects of direct evaporation from the ground surface and transpiration by plants from the underlying root zone. ET is only active in the root zone which is usually limited to the upper 1 to 5 meters. In groundwater recharge areas (e.g. high elevation areas, hill sides) the groundwater table is typically below the root zone and ET is commonly accounted for implicitly (i.e. by adjusting recharge). In groundwater discharge areas (e.g. stream valleys, wetlands), the groundwater table is close to surface and the influence of ET on the groundwater budget can be significant and may have to be accounted for explicitly.

4.3.9 Groundwater - Surface Water Interaction

Groundwater-surface water interaction is a critical aspect in the assessment of the environmental impacts of proposed groundwater resource projects. Exchange of groundwater and surface water occurs in most watersheds and is governed by the difference between water-table and surface water elevations (Winter et al., 1998) and geology. If the water table is higher than stream water level, groundwater discharges to the stream and the stream is referred to as a "gaining stream" (Figure 4-6a). If stream water level is higher than the water table, the stream is a "losing stream" and surface water flows into the subsurface (Figure 4-6b). If the water table is below the bottom of the streambed the stream is disconnected (or "perched") (Figure 4-6c).

Figure 4-6: Schematic showing (a) gaining stream, stream stage is below water table; (b) losing stream, stream stage is above water table; and (c) losing stream disconnected from aquifer (Reproduced from Winter et al., 1998).

At a regional scale, losing streams are commonly found in arid or semi-arid climates whereas gaining streams are commonly observed in humid regions. Streams in any climatic setting can have reaches that gain and reaches that lose stream water, depending on surface topography and subsurface geological conditions (see Wei et al, 2010 for an example of the Kettle River at Grand Forks as a stream with reaches that are losing and reaches that are gaining). Streams are dynamic and groundwater-surface water interactions tend to be strongly influenced by seasonal runoff and even individual storm events. Interaction of groundwater with lakes tends to be less dynamic but can also show distinct seasonal patterns.

For those projects where a stream (or lake) has been identified as a VEC, a detailed review of the stream (lake) morphology, hydraulic gradients, stream flow and stream water quality is required to determine the groundwater-surface water interaction.

The conceptual model documentation should discuss the main aspects of groundwater-surface water interaction, with special emphasis on streams and/or lakes which have been identified as high VECs. This discussion should include:

• Description of the nature of the interaction (i.e. losing and/or gaining stream/lake).
• Identification of the main reaches where groundwater discharges to surface water (i.e. gaining reaches).
• Estimation of groundwater recharge (from streams/lakes) and groundwater discharge (to streams/lakes).
• Discussion of transient aspects of groundwater-surface water interaction (seasonal variations, storm events).
• Discussion of quantity/quality of monitoring data.

Remaining uncertainties in groundwater-surface water interaction and their implications for the modelling objectives should be discussed in the conceptual model reports.

4.3.10 Groundwater Budget

A groundwater budget is commonly a fundamental component of the conceptual model of a groundwater system. A groundwater budget provides a quantitative link between the different aspects of the conceptual flow model, i.e. groundwater recharge (inflow), groundwater flow across different hydrostratigraphic units, and groundwater discharge (outflow).

A groundwater budget should be compiled as part of conceptual model development. The groundwater budget should include the specified domain of the conceptual model (see above) and should provide estimates for the following water budget components:

• Groundwater Inflows:
• Groundwater inflow from upgradient boundary ("underflow")
• Recharge from precipitation
• Artificial recharge (irrigation, injection, seepage)
• Groundwater Outflows:
• Groundwater discharge to surface (seeps and springs)
• Groundwater discharge to mine units
• Groundwater abstraction (pumping)
• Evapotranspiration (in groundwater discharge areas)
• Groundwater outflow at downgradient boundary ("underflow").

The groundwater budget is used to identify a plausible range of flows based on available data and professional judgment. The emphasis of the water budget is not on providing a "closed" water balance with a single estimate for each component (which would imply a greater precision than can be achieved at the conceptual level). Instead, a range of flow estimates should be provided in the groundwater budget for each water budget component that reflects the general uncertainty in a given estimate.

Table 4-1 provides an example of a groundwater budget which was developed for the conceptual model of the Rum Jungle mine site (see Figure 4-3a/b). This conceptual groundwater balance provides estimates of low and a high groundwater flows considering the range (and uncertainty) in both recharge (from precipitation and hydraulic conductivity of the main aquifer units.

Table 4-1: Conceptual groundwater budget for Rum Jungle mine site (Robertson GeoConsultants Inc., 2012).

In most cases, a steady-state water budget is adequate for a conceptual model formulation or report documentation. A transient water budget (which accounts for transient changes in inflows/outflows and storage) is only required at the conceptual modelling stage when a transient aspect of groundwater flow is the focus of the modelling study (e.g., how would baseflow in a river change over time due to proposed pumping of the new municipal well?).

For mining projects, a water balance is typically an important component of the mine planning process. The mine water balance is usually prepared by a hydrologist and its focus is often surface water. Nevertheless, the mine water balance may provide valuable information to assist in developing a groundwater budget (e.g. recharge estimates). The groundwater modeller (and reviewer) should ensure that the mine water balance and groundwater budget are compatible (i.e. use the same assumptions for processes included in both).

4.4 Conceptual Model for Contaminant Transport

4.4.1 General

For those natural resource projects which require consideration of contaminant transport a conceptual model that includes a model (representation) of contaminant transport is required. A conceptual contaminant transport model is the basis for making the decision on whether a numerical model is required or whether a simplified assessment of contaminant transport (e.g. particle tracking, analytical model or simple mass balance modelling) is adequate. It follows, that a conceptual model of contaminant transport should be developed even if contaminant transport is ultimately not modeled using a numerical transport model.

A conceptual model of contaminant transport should be consistent with the conceptual model of groundwater flow. For example, seepage from a tailings dam that introduces a contaminant to the aquifer should be included as a source of recharge in the conceptual groundwater flow model. Similarly, any seepage interception (e.g. recovery wells, collection drains) should be included in the conceptual groundwater flow model as groundwater discharge.

Although there are many similarities, it is convenient to distinguish between a conceptual model for groundwater flow and a conceptual model for contaminant transport. As defined here, the conceptual model for contaminant transport focuses on the transport aspects of the problem and complements the conceptual model of groundwater flow.

This section summarizes the information and processes that need to be considered in formulating and developing a conceptual contaminant transport model. A check list of specific information that should be considered for the conceptual model is provided in Appendix D.

In developing (or reviewing) the conceptual model, the modeller (or reviewer) should ask what evidence there is to support the conceptual model features (e.g. geochemical controls), and whether the conceptual model for this project site is plausible compared to experience at nearby sites or sites in similar setting (reference or analogue site).

4.4.2 Vulnerability Assessment

The first step in the development of a conceptual contaminant transport model is to determine whether there is a potential for an impact due to contaminant transport. This process is also referred to as "vulnerability assessment" and involves three steps:

1. Identify VECs which are located downgradient of the proposed project site (e.g. fish-bearing streams/lakes and/or drinking water wells).
2. Determine whether the proposed project has the potential to release contaminants of concern (CoCs) to groundwater which may exceed applicable standards.
3. Determine whether any CoC(s) has the potential to reach (via a groundwater pathway) identified VECs.

The last step is also referred to as "pathway analysis". The pathway analysis requires a review of existing data (or the conceptual model) of the groundwater flow regime to determine whether there is a groundwater pathway from the identified source to the identified VEC. The travel time and/or attenuation mechanisms (e.g. dilution, dispersion, degradation) are not considered at this screening level assessment.

If these conditions are met, a conceptual model for contaminant transport is required. In principal, all potential CoCs should be included in the conceptual contaminant transport model. In many cases, the conceptual model (and subsequent numerical transport model) is first developed for the CoC most likely to exceed an applicable standard or guideline (often a conservative solute such as sulphate). The transport results for this CoC are then used to infer the fate and transport of other CoCs (often making conservative assumptions to simplify the analysis).

The conceptual model documentation should describe the key findings of this vulnerability assessment, including a graphical representation (plan view and cross-section) of the location of the potential source(s) of CoCs and the VECs at risk and the potential groundwater pathways along which the CoCs may reach the VECs.

Figure 4-3a provides a simplified conceptual model of contaminant transport for the abandoned Rum Jungle mine site. This figure illustrates the main sources of contaminants and provides a summary of the key CoCs and their observed concentration ranges.

4.4.3 Source(s) of CoCs

Once potential CoCs and a potential impact to VEC(s) have been identified, the past, current and potential future source term(s) of the CoCs in question need to be defined, including:

• The nature and spatial extent of the source area (e.g. mine rock pile, tailings impoundment, heap leach pile, pit lake, backfilled underground mine, backfilled open pit etc.)
• The time history of CoC release:
• The CoC concentrations as a function of time
• The CoC mass (i.e. concentration x flow) as a function of time.
• Any geochemical controls on the CoC release from the source (e.g. redox conditions, precipitation/dissolution etc.)
• Any hydraulic controls on the CoC release from the source (e.g. covers and/or liners controlling seepage).

At a brownfield site with existing contamination, the definition of a source term is initially based on a review of relevant site history (including historical maps, plans and records) and available site characterization data, in particular plume delineation (see figure 4-3a). At a proposed new project site, the definition of a source term is primarily based on the project description (e.g. proposed future mine plan(s) and predicted source terms (based on hydrological and geochemical testing and modelling).

The determination of appropriate source terms is a critical step in formulating a conceptual transport model. The uncertainty in source term predictions (in particular for complex mining projects) can be significantly greater than uncertainty in groundwater flow and requires extensive geochemical test work. Unless good field data are available (e.g. at brownfield sites with existing contaminant plumes) significant uncertainty about the source terms remains even after extensive testing. Uncertainty should be discussed and quantified in the conceptual model documentation. In many cases, conservative assumptions have to be made about the source terms in order to test the resilience of the receiving environment and to safeguard against remaining uncertainty.

4.4.4 Applicable Transport Processes

The conceptual model documentation should describe the applicable transport processes for the CoCs including:

• Conservative transport processes
• Advection
• Dispersion and
• Dilution.
• Reactive transport processes:
• Retardation (sorption)
• Precipitation/dissolution
• Degradation.

For more details on these transport processes the reader is referred to Section 9.

The conservative transport of a CoC is a function of the groundwater flow field (hydraulic gradients) and the physical transport properties of the aquifer (i.e. effective porosity and dispersivity, see Section 9). At project sites with an existing contaminant plume (e.g. brownfield sites) these transport parameters may be estimated from an analysis of the historic plume migration (often requiring numerical modelling to reconstruct the historic plume). At most proposed natural resource project sites, these transport parameters need to be estimated from the literature and/or similar analogue sites.

The conceptual model documentation should describe the site-specific data available for the transport parameters (effective porosity and dispersivity) and provide estimates of these conservative transport parameters for different hydrostratigraphic units. It is recommended that a range of values be provided for these transport parameters that reflect the uncertainty in these estimates. If no site-specific information is available on transport parameters, an attempt should be made to use field data from other sites with similar geological material. Literature values should only be used as a last resort. In any case, the selection of transport parameters should be clearly documented and justified.

The conceptual model documentation should also discuss the influence of aquifer heterogeneity and anisotropy (in particular in bedrock) on conservative transport, in particular the potential for transport along preferred flow paths such as high-permeability channels in heterogeneous porous media and faults in fractured bedrock (see Section 9 for more details).

The conceptual model documentation should discuss the potential for dilution of a CoC by mixing of impacted groundwater with clean groundwater and/or surface water (prior to reaching an identified VEC). Caution is required when invoking dilution in a groundwater system. Dilution assumes complete mixing along the groundwater flow path which requires significant distances (up to kilometers) and may never occur in heterogeneous systems. For example, recharge from precipitation may provide a significant volume (and potential dilution) to a contaminant plume in an aquifer. This recharge water may preferentially reside near the water table and provide little dilution to the contaminant plume which is migrating at greater depth in the aquifer. The use of dilution (and the implicit assumption of complete mixing) has to be evaluated and justified, if adopted in a conceptual transport model.

Most CoCs do not behave conservatively in the groundwater system. For those "non-conservative" CoCs, the reactive transport processes should be evaluated. The conceptual model should describe the relevant reactive transport processes such as sorption, precipitation/dissolution, and/or degradation. Figure 4-4 provides an example of a project in which geochemical controls represent a key aspect of the conceptual model.

Reactive transport processes are solute-specific and most are site-specific (taking into consideration local aquifer properties such as redox conditions, organic content etc.). Therefore consideration of reactive transport processes requires a data review of solute-specific information and often collection of site-specific data (e.g. site-specific field monitoring of water quality and/or laboratory testing using local soils and groundwater). The conceptual model should describe and evaluate the available evidence to support consideration of reactive transport processes.

In environmental assessments of natural resource projects, reactive transport processes such as sorption and/or precipitation/dissolution are often deliberately ignored because of the uncertainty of the reactive transport parameters (e.g. K_d ) at the field scale. This approach typically produces more conservative results (i.e. more protective of the environment) and is therefore often preferred by the regulators.

However, neglecting reactive transport parameters in the formulation of a conceptual model does not always lead to conservative transport predictions. For example, consider the presence of a residual contaminant plume (say a metal) which should be remediated by pump and treat. If the metal is known to sorb to the aquifer material (K_d >> 0) a significant mass of the metal will be adsorbed to the aquifer material. Assuming conservative transport could therefore result in significant underestimation of the time (and pump volumes) required to clean up the residual plume.

For the above reason, reactive transport processes should always be considered during development of a conceptual model. For each CoC, the known reactive transport processes should be reviewed to determine whether they may apply at the project site and whether they need to be included in the numerical model. If reactive transport processes (for a known reactive CoC) are not included in the numerical model, this decision should be justified.

4.4.5 Sinks of CoCs

The conceptual model should include reasonable representations of the physical processes that may remove the CoC from the groundwater system, including:

• Seeps and springs discharging to ground surface
• Groundwater discharge into surface waters (lakes, streams)
• Groundwater abstraction (pumping)
• Inflow to open pit and/or mine workings
• Evapotranspiration (ET)
• Wetlands.

These physical "sinks" usually represent a subset of the groundwater discharges that may be included in the conceptual flow model. The conceptual model should account for where in the study area impacted groundwater is, or is expected to discharge to surface water, and/or be removed from the groundwater system via pumping (or other passive collection system).

For sites with existing contaminant plumes, available monitoring data should be reviewed and analyzed to determine the CoC concentration in groundwater discharging to surface or collected by wells/drains etc. In addition, the flow rates for these groundwater discharges should be reviewed to determine the contaminant load (equal to the concentration multiplied by the flow) for each CoC removed from the aquifer system.

For proposed project sites, CoC concentrations and contaminant loads discharging to the surface or planned to be collected should be estimated (to the extent practical). The conceptual model documentation should discuss uncertainties in these estimates and any potential changes over time.

4.5 Methods of Conceptual Model Development

The development of a conceptual model requires a review and synthesis of a great deal of information and data, which is often spatially distributed and/or stored in large databases.

The following methods should be considered for model formulation:

• List potential data needs and associated potential data sources.
• Compile a database, preferably one that can be linked and viewed within GIS (e.g. ACCESS).
• Visualize temporal data in time trend plots (e.g. water levels, water quality).
• Visualize spatial data in maps and cross-sections (e.g. topography, geology, boreholes, water table maps, location of source areas, VECs, etc.) using suitable software (GIS, ACAD).
• Use high-resolution air photography or satellite imagery and internet-based mapping services (e.g. Google Maps).
• Supporting analytical calculations (Darcy calculations, travel times, mixing calculations etc.) to provide initial estimates of flow and CoC loads.
• Review of literature (e.g. estimation of flow and transport parameters not available from site-specific field work).

To the extent possible, an integrated software platform should be used for data management (including QA/QC) and spatial visualization (e.g. using GIS) to maximize the use of available data and to minimize errors in data transfer.

Examination of high-resolution air photos (or internet based mapping services where local air photos are not available) during conceptual model development is recommended. These air photos should be geo-referenced to allow at least some level of remote "ground truthing" of spatial data (e.g. locations of critical monitoring wells, zones of groundwater discharge etc.).

4.6 Potential Errors in Conceptual Modelling

The development of a conceptual model is one of the most important steps in groundwater modelling. At this stage of the modelling process, the modeller will have to make decisions on what processes to include (or exclude) and what simplifying assumptions should be made to achieve the modelling objective(s). Those decisions will strongly influence the mathematical model and ultimately the modelling outcome. It follows that errors in the conceptual model can propagate through the remainder of the modelling study, and if not detected early on, can potentially lead to invalid modelling results and conclusions. For this reason, a conceptual model should always be checked carefully for potential errors.

4.6.1 Inadequate Field Data

In some cases the available field data are either missing or inadequate to allow the development of a credible conceptual model. Inadequate field data is a common problem in natural resource projects, in particular for large projects (large foot-print) and/or projects in remote locations where field work is costly, and/or site access is difficult. Another problem in natural resource projects (in particular at the design stage) is that the field data were oftentimes collected for another purpose (say geotechnical study) and may not be adequate for the purpose of groundwater modelling.

If critical data gaps are identified, these gaps need to be filled by additional field work before the conceptual model can be completed and the modelling study can proceed. In some cases such additional field work can be avoided (or at least postponed) by: (i) using two alternative conceptual models that cover the uncertainty due to the data gap; and/or (ii) using conservative assumptions in the conceptual model.

For example, the presence of a fault may have been identified at a project site (from drilling and/or structural analysis) but no hydraulic testing is available to determine whether this fault should be considered a high-K conduit (and a potential preferential flow path for a contaminant) or a barrier to groundwater flow and contaminant transport. This data deficiency should be identified as a critical data gap during conceptual model development.

The preferred course of action is to complete additional drilling and/or hydraulic testing to determine the hydraulic properties of this structure. If that is not feasible, two alternative conceptual models could be developed (one assuming the structure is highly permeable and another assuming it is a low-permeability barrier). Both conceptual models would have to be implemented into a mathematical model and calibrated. If both scenarios are plausible, the more conservative scenario should be assumed for model predictions related to environmental impact assessment.

4.6.2 Conceptual Model is Inconsistent with Existing Data

Conceptual models may be inconsistent with existing data. This error is quite common because the conceptual model is preliminary and qualitative, in particular during the initial modelling stages when the conceptual model has not yet been rigorously tested against field observations. Inconsistencies between the conceptual model and field observations are often recognized during model calibration because incorrect conceptual models can result in significant difficulties in calibrating a model to the field data.

Typical errors in the conceptual model that could result in consistent biases between simulated and observed heads (or concentration) include:

• Incorrect selection (spatial distribution) of hydrostratigraphic units (e.g. presence of a high-permeability channel).
• Incorrect boundary condition (e.g. use of a constant head boundary for a small lake that is not hydraulically connected to the groundwater).
• Omission of a key source or sink of groundwater (e.g. a drain system in a tailings dam not documented).
• Omission of a key transport process (e.g. irreversible sorption of a contaminant).
• Incorrect historic source term (e.g. timing and magnitude of historic seepage from a tailings dam).

The modeller should carefully examine the calibration results for consistent bias in residual errors (in space and time) and/or consistent "outliers" which could be indicative of problems with the conceptual model (see Section 7.3 for more details). The modeller should continually challenge his/her conceptual model and be "open-minded" to attempt changes in aspects of the conceptual model during model calibration that carry significant uncertainty.

4.6.3 Conceptual Model Misses a Key Process

The development of a conceptual model requires the modeller to identify the key processes that need to be included in the numerical model. By virtue of this selection there is a risk that a key process has been overlooked and is not included in the model. Examples of key processes that may not be apparent from the monitoring data and therefore may be overlooked during conceptual model development include:

• Preferential flow and/or transport in heterogeneous aquifers (debris flow sediments; glacio-fluvial sediments, fractured bedrock).
• Preferentially oriented flow and/or transport in anisotropic aquifers (fractured bedrock).
• Complex reactive transport processes (e.g. non-linear sorption, redox sensitive transport).
• Change in reactive transport conditions over time (e.g. change in redox conditions, depletion of sorption sites etc.)

To avoid this problem, the modeller should review possible flow and transport processes that may be active at the project site. For those processes that are uncertain, preliminary sensitivity analyses (using analytical solutions or a simple numerical model) can be carried out to determine whether these processes can be neglected for the purpose of the study or whether they should be included. If the project outcome is sensitive to those assumptions, additional data should be collected and the conceptual model should be refined. If additional data collection is not feasible (or does not resolve the uncertainty) conservative assumptions should be adopted.

4.6.4 Simplifying Assumptions are Non-Conservative

Conceptual models require the use of simplifying assumptions to reduce real-life complexity to a manageable level that can be incorporated into a model. The precautionary principle should be applied when making simplifying assumptions for environmental assessments, i.e. simplifying assumptions should always be conservative in the sense that they result in more, rather than less, predicted impact. A classic example is the assumption of conservative transport for contaminants which are known to be reactive (e.g. metals) which would tend to result in the prediction of greater water quality impacts.

The effect of simplifying assumptions on model predictions is not always straightforward and the modeller (or reviewer) should always confirm that a given simplifying assumption in fact results in more conservative predictions, in particular in the context of environmental impact assessments.

In Section 4.4.3 we described an example where the assumption of conservative transport did not result in a conservative transport predictions (i.e. residual contaminant mass sorbed to aquifer material was ignored). A similar error is sometimes made when conceptualizing the flushing of a contaminant plume from a source term, for example pore water in a tailings impoundment. In this case, the assumption of conservative transport would result in much faster flushing of the contaminant from the tailings pore space than might occur in reality, perhaps due to continued dissolution of a contaminant precipitated out during tailings deposition. If the geochemical conditions in the tailings are not well-understood, a more conservative approach would be to assume a constant concentration in the tailings pore water.

Another example of a simplifying assumption that is commonly misinterpreted is the assumption of a homogeneous aquifer. The assumption of a homogeneous aquifer greatly simplifies the modelling and may be adequate for predicting groundwater flow. However, this assumption may not be conservative when predicting the breakthrough of a contaminant plume. Some may argue that aquifer heterogeneity spreads the contaminant plume and hence dilutes the plume, thus suggesting that the assumption of a homogeneous aquifer is conservative. The opposite is in fact true when predicting contaminant breakthrough at a discrete location. Heterogeneity can result in strongly preferential flow path (in particular in fractured bedrock) with very limited dilution and dispersion over great distances (up to kilometers).

4.6.5 Conceptual Model is Not Updated

A conceptual model is always a "work-in-progress" and may require significant updating during the model study. Tight timelines and/or budget constraints often force the modeller to continue using a model that is based on an incorrect conceptual model (e.g. as evidenced by poor model calibration and/or new field data).

There is also a natural tendency to resist the complete rethinking of a conceptual model and "stick to" an existing model simply because the modeller invested a lot of time and effort into building this model. This tendency does not necessarily imply bad intent on part of the modeller but instead is a natural tendency of people that have invested significant effort. This problem is more commonly observed with very complex models (which typically cost a lot more time and money to construct). This issue is one of the main disadvantages of using very complex models.

One of the best ways to ensure that the modelling process "stays honest" and to ensure that the conceptual model is updated when required, is to provide for model review throughout the modelling process (see Section 2 and 11). Even an informal, internal peer review provides additional dialogue and may uncover problems with the conceptual model that were hitherto not recognized by the modeller.

4.6.6 Conceptual Model Does Not Consider Future Processes

Most model studies are motivated by the need to predict a future behavior that cannot be readily observed today. The lack of knowledge about future processes can sometimes lead to an incomplete, or even missing, conceptual model of those processes. For example, a future change in land use may result in significant changes to local recharge to groundwater and hence the aquifer behavior. Another example that has received much attention recently is the potential for climate change which may affect recharge to groundwater system.

The extent to which future processes should be considered in the conceptual model depends on the project and in particular the time-frame of modelling. Natural resource projects (whether mining or large groundwater extraction projects) tend to have a time frame of decades and associated contaminant transport problems may require models that simulate transport for up to a century or longer.

For those extended time periods, an effort should be made to address potential future changes to the aquifer system (whether induced by the project or otherwise) in the conceptual model, at least qualitatively. In some cases, selected sensitivity runs for potential future "upset" conditions may be required (using an analytical or numerical model) to determine their potential influence on long-term predictions.

4.7 Case Study Example

4.7.1 Case Study 2 - Underground Mine

The conceptual model for the underground mine project includes delineation of hydrostratigraphy, hydrogeologic properties of those strata, relationships between groundwater and surface water features, regional hydrologic (climatic) conditions, the mine development and closure plan, and an understanding of potential receiving environments. Figure 4-7 presents a cross-sectional illustration of the key component of the groundwater system. The reader is referred to Appendix C for additional details regarding this Case Study.

Based on the complexity of the site and level of risk associated with valued ecosystem components, a 3D MODFLOW model was developed. The model was conceptualized using drilling records and geologic data from a number of sources, including:

• Geologic mapping/model,
• Exploration drilling and testing, and
• Residential well logs from MOE's WELLS database

Although continuum models, like MODFLOW, do not address discrete bedrock fractures, the modeller notes that the applicability of an equivalent porous media (EPM) approach to simulations of flow and transport in heterogeneous, fractured bedrock settings has been well proven and documented in other studies elsewhere. However, the justification did not appear to be based on assessments of the conditions of the bedrock units at the site. The assumption of equivalent porous medium became a significant point of uncertainty from the perspective of the government reviewers. Prior field investigations were acknowledged and data was incorporated into the conceptualization and calibration procedures. This data was, however, limited spatially and temporally.

Figure 4-7: Conceptual model used for MODFLOW modelling.

The conceptualization of the hydrostratigraphy and groundwater flow is generally well presented in this Case Study. However, the results are prefaced by the extremely limited dataset on which model conceptualization and calibration were based. Furthermore, major assumptions of equivalent porous medium for fractured bedrock, horizontal isotropy of hydraulic conductivity of the various fractured bedrock units and effective porosities did not appear to be justified by data or assessments from the site. These issues introduced significant uncertainties to assessing how the proposed mine will affect changes in groundwater flow and contaminant travel times, and impacts of these changes to local high VEC water users' streams.

As recommended in these guidelines (see Figure 2-2), consultation with government technical staff during the development of the modelling objectives, the conceptual model, and the data collection requirements could have helped to identify these issues and questions much earlier in the modelling process.

Review Questions

1. Which of the following should be included in the discussion and documentation of a conceptual flow model?
1. The presence and spatial distribution of major hydrostratigraphic units and their hydraulic properties.
2. A description and quantification of recharge and discharge processes, including seasonal variations.
3. The main aspects of groundwater-surface water interaction, with special emphasis on streams and/or lakes which have been identified as high VECs.
4. A groundwater budget.
5. All of the above
2. To test a conceptual model, one should:
1. Translate the conceptual model to an analytical model and use literature values for hydrogeologic parameters.
2. Translate the conceptual model to a mathematical model and calibrate it using site-specific field measurements.
3. Translate the conceptual model to a mathematical model and use hydrogeologic parameter values from a similar site.
4. Accept the conceptual model as correct and perform several sensitivity runs.
5. A and D
3. Why might a conceptual model be wrong?
1. The information used to define a problem is incomplete.
2. Incorrect assumptions are made in developing the conceptual model (e.g. can it be assumed that a sand and gravel deposit identified in three investigation boreholes extends laterally below the whole extent of the site, or have three separate sand and gravel lenses been penetrated?).
3. The available information may be conflicting (e.g. laboratory leaching test results indicate that a given contaminant should be leached from a contaminated soil, whereas no evidence for this contaminant is found from groundwater sampling).
4. The physical and/or chemical processes occurring are poorly understood.
5. All of the above.
4. Acceptable graphical methods for representing a conceptual model for groundwater flow and solute transport include:
1. 2D cross sections, 3D block diagrams, piezometric surface maps, graphs of temporal/spatial trends of water levels and/or water quality.
2. Location map, structural geology map, land usage maps, bore hole fence diagrams.
3. Table of assumed hydrostratigraphic units with literature values presented.
4. A and B.
5. B and C.
5. To determine if a conceptual model is adequate, the reviewer should ensure that:
1. The conceptual model includes all relevant processes and that any omissions or simplifications are justified.
2. The conceptual model is plausible and consistent with literature values for hydrogeologic parameters.
3. The conceptual model is plausible and consistent with site-specific field data.
4. A and B.
5. A and C.

---

Proceed to Section 5: Mathematical Model Selection