Michael R. Shultz, Colin Plank
8 min
Effective groundwater management, particularly for remediation, requires a precise understanding of the spatial distribution of silt and clay-rich bodies that function as aquitards. These features compartmentalize aquifers, control flow vectors, and can store contaminants via matrix diffusion. Despite this, many environmental investigations still rely on simplified, 'layer-cake' models that treat the subsurface as a homogeneous, isotropic medium. This paper argues that the distribution of these fine-grained elements is not random but follows predictable patterns determined by their depositional environments.
The authors advocate for the use of 'facies models'—predictive frameworks derived from sedimentology—to better interpret stratigraphic data. By analyzing the depositional environment (such as braided fluvial, meandering fluvial, or alluvial fan systems), practitioners can more reliably interpret logs from wells and cone penetrometer testing. For example, understanding how a meandering river evolves allows a geologist to distinguish between localized 'clay plugs' (abandoned oxbow lakes) and laterally extensive floodplain deposits. Failing to apply these models often leads to incorrectly mapping continuous aquitards where they do not exist, or missing them entirely, which ultimately compromises the design of remediation and cleanup strategies.
A critical, yet frequently overlooked, component of these systems is the paleosol—an ancient soil horizon that formed during periods of non-deposition. Paleosols are often interpreted simply as general soil in boring logs, but they can act as significant, widespread aquitards. The authors provide practical diagnostic criteria, such as soil color, pedogenic structure, and the presence of root casts, to help field teams identify these units. Recognizing these features early in a project lifecycle is essential for building a robust Conceptual Site Model (CSM).
The paper concludes that transitioning away from traditional, simplified models is a necessity for modern environmental protection. The authors highlight the need for a synergy between hydrogeologists, who interpret subsurface reality, and modelers, who need to incorporate this geologic complexity into numerical simulations. By focusing on high-resolution data collection and applying geological principles, practitioners can significantly improve the success and cost-efficiency of groundwater remediation projects.
A sophisticated understanding of the three-dimensional distribution of silt- and clay-rich bodies of strata (elements) in aquifers is critical given that they not only have the potential to act as aquitards or semi-confining units and vertically partition groundwater flow into separate aquifer zones, but also provide lateral barriers to groundwater flow, impacting contaminant distribution and groundwater flow dynamics. Additionally, when in prolonged contact with dense nonaqueous phase liquid (DNAPL) or contaminated groundwater, fine-grained elements may become storage zones for contaminant mass via matrix diffusion and thus serve as long-term secondary sources of contamination to groundwater that can confound remediation strategies and render remedy performance projections unreliable. The stratigraphic architecture of aquifer systems, including fine-grained facies architecture, is complex but is not random and can be effectively predicted through application of facies models. This paper reviews depositional models ("facies models") for common depositional environments with a focus on alluvial end-members of braided fluvial, meandering fluvial, and alluvial fan facies models. We examine the facies models from the perspective of aquitards and present case studies to provide an overview of the expected aquitard dimensions and characteristics. The critical yet underappreciated role of the paleosol as a potential aquitard is also examined, and basic criteria for differentiating ancient floodplain clay units with high lateral continuity from other laterally discontinuous clay units are provided.
Alex: And direction matters. A horizontal connection might be appropriate along the ancient shoreline, but not across it. The depositional model tells you which geometry is plausible before you draw the connection.
Sam: How does that logic translate to rivers, where many contaminated groundwater sites sit?
Alex: The review focuses on three end-member environments: braided rivers, meandering rivers, and alluvial fans. Braided rivers have branching channels separated by sediment bars. Their deposits commonly contain abundant sand and gravel, with relatively few lateral barriers.
Sam: Does that make braided deposits a setting where clay can mostly be ignored?
Alex: No. Fine sediment can accumulate outside the active channels on floodplains. If buried and preserved, those deposits can form extensive barriers. Smaller clay deposits also occur within abandoned channels and on bars, but their geometry differs.
Sam: And a meandering river leaves another arrangement because its channel moves sideways.
Alex: As a bend migrates, sand accumulates on its inner side. Eventually, a bend can become cut off, leaving an oxbow lake. That abandoned channel fills with silt and clay, producing what the authors call a clay plug.
Sam: That sounds less like a blanket over an aquifer and more like a wall interrupting sideways flow.
Alex: It can be. The review describes neighboring oil fields in Mississippi separated by an interpreted clay plug. Different fluid contacts and pressure data support hydraulic isolation. It's evidence that abandoned-channel deposits can create substantial lateral barriers.
Sam: Useful evidence for the mechanism, but not proof that every buried clay plug seals a groundwater site.
Alex: The review doesn't establish that. Its point is that origin helps predict geometry and possible function. Floodplain clay may restrict vertical flow over a broad area; a channel plug may redirect flow laterally within a more limited area.
Sam: Where does that distinction actually overturn a groundwater interpretation?
Alex: At a northeastern United States site, a continuous barrier called Aquitard A had shaped monitoring and remediation for decades. The review compares interpretations of three cone penetrometer logs, collected by pushing an instrument into the ground. The older interpretation connected fine-grained intervals into an uninterrupted barrier.
Sam: What made the alternative more than a different drawing through the same observations?
Alex: The central log had a much weaker pore-pressure response than the neighboring clay intervals. Associated boring descriptions supported thin silt beds rather than a substantial clay body. Gravel at the sand's base and an upward-fining profile also supported an erosive channel fill.
Sam: So the alternative was a sand-filled channel cutting through the supposed barrier, supported by several geological clues.
Alex: Yes. That would connect groundwater zones previously treated as separated. It's a concrete warning: a long-established conceptual model can remain wrong if the correlations aren't checked against how sediments form.
Sam: You mentioned ancient soils as overlooked barriers. Why would soil survive inside an aquifer?
Alex: A soil can develop when sediment deposition pauses or slows, then become buried when deposition resumes. That preserved ancient soil is a paleosol. It needn't mark a major climate event; ordinary channel relocation can create the pause.
Sam: What would make someone logging a core recognize one, instead of simply writing down “clay”?
Alex: The review points to varied colors, soil horizons, root traces, and nodules. Such features can help distinguish floodplain deposits from laminated clay filling an abandoned channel. They're clues to origin, not a standalone hydraulic test.
Sam: How thin can one of these soil horizons be and still matter?
Alex: At a desert southwestern United States site, interpreted paleosols less than fifty centimeters thick separated individual alluvial-fan deposits. Wells screened within different fan units showed distinct contaminant concentrations. The authors interpret this as evidence that even thin horizons can compartmentalize groundwater.
Sam: Did that interpretation change anything practical, or just improve the geological description?
Alex: The review contrasts it with an earlier interpretation using kriging, a statistical method for estimating conditions between observations. That representation missed the long, thin, sloping clay connections. An ethanol-injection treatment designed using it reached uncontaminated groundwater and generated a byproduct plume requiring further remediation.
Sam: That's a serious example, but how far should we generalize from these cases?
Alex: My reading is that they're persuasive demonstrations of failure modes, not a measured success rate for the approach. This is a review, with no newly generated or analyzed dataset. It doesn't quantify how often facies-based interpretation improves cleanup outcomes.
Sam: And the templates themselves could become another rigid model if people apply them without checking.
Alex: The authors explicitly warn that real environments form a continuum. A single site can also contain deposits from different environments through time. They advocate high-resolution observations early in investigations, including detailed cores and well logs, to constrain the interpretation.
Sam: Who should read the full document, and where should they start?
Alex: Researchers investigating contaminated aquifers, remediation designers, and groundwater modelers. Start with the case study on the supposed continuous aquitard, then the alluvial-fan case comparing geological and statistical interpretations. Read “Practical Field Application: Clay Encountered when Drilling” for clues that can guide decisions beyond the well itself.
Sam: And what should everyone else carry away?
Alex: Finding clay tells you what's in the borehole. Understanding how it formed helps tell you what it separates.
Sam: Thanks for listening.