Before beginning any modeling project, the goals of the project should be clearly defined. Useful evaluations of flow and transport can be undertaken using analytical models, often via a spreadsheet. It is good practice to start with pencil and paper and see how many questions can be answered before building a complicated numerical model. If nothing else, careful calculations provide a baseline to evaluate the basic functionality of future, more complex models.

3.1  Conceptual Models

Most environmental projects for contaminated groundwater sites are required to prepare a conceptual model in order to comply with government regulations. The conceptual model defines basic details of the contaminated aquifer; its geometry, depth, thickness, range of hydraulic conductivities, observed hydraulic heads, and features of the system that influence hydraulic behavior such as surface water bodies and pumping wells. For smaller contaminated sites, it is likely that the details of the conceptual model are adequate to meet regulatory requirements, but do not provide sufficient detail for the modeling to effectively predict the behavior of the system, and thus to design the remediation plan. Conceptual models that describe thick layers with homogeneous properties are an indicator that more thorough review may improve the model.

3.2  Local Geology

Familiarity with local geology is a necessity for any subsurface modeling. Boreholes for water supply wells, geotechnical investigation, and other applications are abundant, often with publicly available logs and published interpretations. It is useful to review reports from other groundwater projects in the area to glean information about subsurface properties and conditions, and to learn from problems encountered by others doing similar work.

3.3  Structural Geology

The data required to identify structural geologic features are essential but not sufficient for many contaminated sites. Consequently, faults may not be identified, and the bedrock surface is often characterized as a monotonic flat surface or typical depth, while in actuality it may have a complex topography. The bedrock surface is extremely important at sites with dense non-aqueous phase liquid (DNAPL) contaminants because the DNAPL sinks to low permeability layers and further migration is controlled by the topography of the surface. Faults can be found virtually everywhere in the world. For example, areas of the Texas Gulf Coast are affected by surface deformation related to creeping faults although that part of North America is tectonically quiet. Often small faults are identified at contaminated sites by noting locations of steeper hydraulic gradients. The Test Site example presented in this book contains a small normal fault.

3.4  Stratigraphy

The sparsity of data at contaminated sites also presents challenges to delineating the stratigraphy. However, more stratigraphic data has been collected in recent years as often stratigraphic complexity is proving to be a key limiting factor to successful remediation. Remediation methods that involve extracting and injecting groundwater can be limited by baffling of flow by low permeability stratigraphic barriers. Thin layers of fine-grained soils rich in clay and organic matter can have a high capacity to trap groundwater contaminants. After the more permeable portions of the aquifer have been remediated, these contaminants can diffuse back into the clean groundwater, causing an unexpected rebound in concentrations. Layers like this can be easily missed using older technologies such as auger and rotary drilling. Electronic borehole logging and high-resolution direct push sensing tools can be used to create logs with the necessary level of detail. However, stratigraphic principles are required to make a meaningful determination of how these details extend into the space between boreholes.  Guidance documents such as Schultz et al. (2017) contain detailed information and workflows on how to make such stratigraphic assessments.

For oil and gas reservoirs, much characterization is done by considering and comparing sedimentary depositional facies (e.g., Shepherd, 2009). For example, the relative homogeneity of aeolian (sand dune) facies leads to better reservoir connectivity than the layered and highly dissected nature of deep‑water channelized turbidites. This topic is worthy of its own textbook and is not discussed here except to say that environmental projects can benefit from the same line of inquiry. Again, contaminated groundwater sites are often of limited extent or the funding to collect the necessary data is lacking, rendering such stratigraphic analysis impossible.