1

Introduction

Fractured rock contamination remains one of the greatest challenges to groundwater protection and cleanup (e.g., Steimle, 2002). Efforts to characterize relatively homogenous, unconsolidated soils (e.g., homogenous sand aquifers) and to remediate contaminated groundwater within them have become increasingly routine in recent decades. However, characterization of fractured rock and remediation of contaminated sites have not advanced to the same level. Because of the discrete nature of transport pathways in fractured rock, approaches applied commonly in soils can be ineffective in fractured rock. Contaminants can be transported great distances and at relatively high velocities along discrete channels, and therefore it becomes necessary to characterize both the rock matrix and properties of fractures that control or affect possible contaminant transport or remediation.

Rocks fracture when tensile, compressive, and shear stress conditions exceed their mechanical strengths. Fractures then can be modified as a result of tectonic and thermal stress and strain, as well as physical, chemical, and biological processes such as erosion, dissolution or precipitation of minerals, and degradation from roots. Fractures can be of varying length, aperture, and spacing. Water flow through fractures is commonly distributed unevenly, so the hydraulic significance of individual fractures within a fracture set can vary greatly. Often relatively few fractures present are hydraulically significant (discussion of this topic is found in Chapter 3, Box 3.1). Fractures generally occur in patterns dictated by geology, geomechanics, and geochemistry. In-depth understanding of the various coupled processes that control fracture formation and contaminant transport will enable better characterization, modeling, monitoring, and remediation of the fractured subsurface. This is the focus of this report. This chapter provides an introduction to the committee’s task and a brief qualitative discussion of the technical issues for readers not as familiar with the topic. More technical descriptions are provided in later chapters.

THE CONTEXT

Society engages in many activities that cause contaminants to enter into the subsurface. Contaminants may be released into the subsurface at ground level or through surface waters. Others migrate from infrastructure installed or wastes buried in the shallow or deep subsurface. Contamination of groundwater can occur in all types of rock and with many types of contaminants from many sources. Italicized text in this chapter presents one hypothetical scenario of fractured rock contamination.

Fractured rock systems include intact blocks of rock surrounded by fractures and perhaps other features such as dikes, brecciated layers, or features associated with karst (e.g., Hoek and Bray, 1981). The rock blocks in the system may be crystalline with nominal porosity (i.e., many igneous rocks), or granular with varying amounts of cementation and porosity (i.e., sedimentary rocks). Insufficient characterization of the fractured rock environment, and of fate and transport processes within it, can lead to engineering strategies that fail to meet expectations. Contaminants, in many cases, diffuse quickly from fractures into the surrounding low permeability rock matrix. Diffusion back out of the matrix, however, may occur over decades or centuries, and the matrix effectively becomes a long-term contaminant reservoir. Remediated water in fractures can be then recontaminated as contaminants diffuse back into the fractures. Adequate characterization of the fractured rock environment goes beyond that for the homogenous porous environment and includes accounting for spatial heterogeneities and processes such as dissolution/precipitation, reduction-oxidation reactions, and biodegradation within the different geometries of hydraulic importance. Monitoring and characterization information is required to identify fracture and matrix porosities, and contaminant fate and transport in both fractures and rock matrix.

Uncertainties in characterization data and interpretation can result in inadequate site conceptual models and numerical models of fluid flow and transport. They can impact the effectiveness of monitoring programs and hamper lifecycle planning, mitigation, and remediation activities. These translate into less optimal management of contaminated sites or of sites where contamination is possible given the presence of certain types of infrastructure.

THE COMMITTEE’S TASK

Nearly a decade has passed since advances in the state of the art and state of the practice in fractured rock characterization and remediation have been examined comprehensively with respect to changes in regulatory regimes for implementing national policies. This report presents the findings of such an examination by a study committee of the National Academies of Sciences, Engineering, and Medicine (the National Academies). The study was funded by the National Aeronautics and Space Administration, the U.S. Nuclear Regulatory Commission (USNRC), the U.S. Department of Energy (DOE). The funding agencies have some level of regulatory or management responsibility for sites at which contamination in fractured bedrock is possible or present at depths of hundreds of feet, and extending several thousands of feet laterally. USNRC and DOE also have a role in identifying new sites or types of facilities considered for the long-term disposal of high-level radioactive waste. Some proposals involve disposal of wastes at great depths (i.e., up to 5 kilometers). This report is intended to inform the regulatory regimes or policy related to different types of infrastructure and their impacts on fractured rock systems and considers the long-term legacy of future waste repositories.

The National Academies convened a committee to undertake this study. The Statement of Task provided to that committee is provided in Box 1.1. Based on discussions with the study sponsors, the committee focused its attention on naturally fractured rock (although there are unconsolidated media in which fractures dominate flow regimes). Given the breadth of the committee’s task, the committee made the strategic decision to focus on issues in the area beneath the vadose zone and up to approximately 5 kilometers in depth. Geotechnical, geologic, and hydrologic issues relevant throughout the lifecycle of engineered facilities are addressed. Research directions are suggested that could improve the state of the art, and applications are suggested that could enhance practice, including where better scientific understanding could inform regulations, policies, and implementation guidelines related to infrastructure development and operations. The scope of this study does not include techniques used to intentionally fracture low-permeability rocks for the purpose of extracting hydrocarbons (hydraulic fracturing); those tend to be purpose-specific. Neither does the report include discussion of the infrastructure from which contaminants originate, but rather focuses on their potential impacts to the subsurface geologic and hydrologic environments during the infrastructure lifecycles.

The committee includes researchers and practitioners with expertise in areas such as geohydrologic site characterization, hydrogeology, site-scale geotechnical and hydrologic modeling, contaminant fate and transport modeling, geotechnical and geohydrologic monitoring, environmental engineering, remediation practices, and risk assessment (see Appendix A for the committee member biographies). Committee members relied on their own expertise, the expertise of speakers and guests invited to their meetings, and information gathered during a workshop they organized. Appendix B provides the agendas for the open sessions of the committee meetings and the public workshop.

PREVIOUS NATIONAL ACADEMIES STUDIES

Multiple National Academies studies address issues related to fracture flow, contaminant transport, and subsurface remediation. The report Rock Fractures and Fluid Flow: Contemporary Understanding and Applications reviewed methods and strategies to characterize fracture flow in use from the mid-1970s to the early 1990s (NRC, 1996). A 2001 study on conceptual models of flow transport in the vadose zone built on the 1996 report and addressed development and confirmation of conceptual models that focus on how fractures affect recharge to local groundwater systems through all types of bedrock (NRC, 2001). A third report (2004) reviewed technologies for source remediation in various formations including fractured rock and proposed protocols to aid decision making for remediation strategies. Other NRC reports have stressed the need to establish guidelines to increase long-term direct monitoring of waste containment systems (2007), and a recent report explores alternatives to current groundwater remediation practices at complex sites (2013). Several other NRC reports address subsurface remediation from multiple perspectives (e.g., NRC, 1994, 1999, 2000, 2003), including from that of the effectiveness of remediation approaches. The number of reports reflects the challenges associated with subsurface remediation, aggravated in fractured rock, and indicates that technical gaps in remediation practice remain in

spite of many recent advances. Regulatory guidelines and policies are in need of review, and this report is intended, in part, to describe the states of the art and practice to inform those reviews.

THE HETEROGENEITY OF FRACTURED ROCK

The heterogeneity of fractured rock can require more detailed characterization and analyses than what is required for less variable geologic materials. Rock fracture patterns form in response to stress fields in the Earth’s crust (natural or anthropogenic) and rock type (e.g., sedimentary or crystalline) combined with various geomechanical, geochemical, and geobiological processes (discussed in Chapters 2 and 3). The causative factors of those processes vary over space and time, so heterogeneities and anisotropy in rock and fracture properties will also occur at various scales—from micrometer (grain size) to kilometer (regional) scales. Factors that cause fracturing also influence fracture geometries and their variations—characteristics such as fracture length, aperture, orientation, spacing, extent, intensity—as well as the geometric properties of the rock matrix blocks. Fractures can occur in multiple orientations at a given location as a result of evolving stress fields in the Earth’s crust, as illustrated in Figures 1.1 and 1.2. Fracture geometries can be planar where they form preferentially along or quasi-orthogonal to bedding, as in the relatively undeformed sedimentary rocks pictured in Figure 1.3, but they may be non-planar, as might occur in the folded sedimentary rocks pictured in Figure 1.4.

Fracture geometry is a controlling factor of flow: fractures create discrete pathways for both groundwater flow and contaminant transport. Fractures can be highly conductive, but they can also be flow barriers, for example, where faults contain clay gouge (e.g., Cain et al., 1996; Crawford et al., 2008). Exploration boreholes drilled at a site may not capture features that control contaminant fate and transport, as is discussed in more detail in Chapter 5.

CONTAMINANTS IN FRACTURED ROCK

A variety of contaminants can be detrimental to groundwater quality and the environment. Given their different characteristics, distinct strategies are needed to prevent, mitigate, monitor, and remediate contamination of fractured rock. Indeed, proper understanding of specific contaminants and their interaction with fractured rock is critical to reduce the uncertainties associated with their management. Specific details regarding individual contaminants are beyond the scope of this report, but below are brief discussions of common water soluble and non-aqueous phase liquids. Vapor phase contaminants are also found in fractured rock settings, but are more common in the vadose zone.

Water Soluble Contaminants

Contaminant solubility in groundwater can range from fully soluble to less than parts-per-million levels of solubility. Fully miscible contaminants (e.g., nitrates, acetone, and methanol) readily dissolve in water and can travel with the groundwater great distances from their sources. Dissolved contaminants may react with surrounding geologic materials, precipitate from solution, or exchange electrons or ions with the rock surface and change oxidation states. These reactive dissolved contaminants migrate more slowly than the transporting water and are said to be “retarded” or “attenuated.” Some attenuated contaminants—particularly organic chemicals such as benzene—tend to be sorbed and held by organic matter found in fractures or by the rock matrix (Zytner, 1994). Many weakly soluble organic contaminants (e.g., benzene) can be biodegraded under the right circumstances by naturally occurring microorganisms in fractured rock environment (e.g., Johnston et al., 1994).

Some dissolved contaminants are radioactive and inherently unstable. Like other soluble contaminants, their rate of movement is a function of their chemical properties. Some radioactive dissolved contaminants such as tritium have a short enough half-life that over a period of a few decades, radioactive decay leads to significant reductions in concentration. Others, such as components of spent nuclear fuels, have half-lives spanning many thousands of years.

Non-Aqueous Phase Liquids

Liquids that do not dissolve in water in appreciable amounts are referred to as non-aqueous phase liquids (NAPLs). NAPLs such as dry-cleaning liquids, petroleum products, and cleaning solvents contaminate sites throughout the world. Many NAPLs have components that dissolve in water (e.g., benzene and toluene from gasoline at low concentrations). NAPLs migrate through fractured rock in liquid phases separate from water and flow at different rates (and potentially directions) than water, depending on their viscosities, relative densities, and surface tension with water.

Light non-aqueous phase liquids (LNAPLs; e.g., gasoline) are less dense than water and accumulate primarily above the water table, but can be transported horizontally and vertically in fractures beneath the water table (Adamski et al., 2005). LNAPLs can spread or be trapped beneath and above the water table with changing groundwater levels. This can cause contamination of remediated groundwater when it comes in contact with mobile or residual LNAPL (Newell, 1995). Dense non-aqueous phase liquids (DNAPLs; e.g., trichloroethene) are denser than water and sink below the water table to locations that can be difficult to identify. Most chlorinated solvents are DNAPLs. These include effective but sometimes toxic degreasing agents used widely for industrial cleaning, dry cleaning of clothes, and other applications. DNAPL groundwater contamination can be most difficult to identify and manage and can often go undetected at a site given the variables that affect its transport and storage (Huling and Weaver, 1991). Because DNAPL is denser than the groundwater, it sinks until it encounters a barrier. Figure 1.5 demonstrates the potential distribution of DNAPLs in a karst setting.

FATE AND TRANSPORT

Depending on local geologic, geomechanical, and hydrogeologic conditions, individual fractures or fracture networks can become high-velocity discrete pathways for groundwater and contaminants. Advective travel times in these pathways may be significantly shorter than flow through lower permeability rock matrix materials (Fossum and Horne, 1982). Different contaminant concentrations, velocities, and geochemical and biologic conditions may be found within fractures and matrix.

Advection, addressed throughout this report, is the transport of dissolved chemicals, particles, or dissolved or volatilized gas by a carrier fluid (i.e., groundwater) as it flows through the subsurface. The transport rate and flow direction of substances is not necessarily the same as the velocity and direction of the carrier fluid. As a contaminant from a source (i.e., a storage container) is carried downstream (a plume), the contaminant spreads laterally and vertically (hydrodynamic dispersion) as a result of rock fracture and matrix geometries and local variations in flow velocity. The measure of the spread of contaminant as it travels through the groundwater is the hydrodynamic dispersion. Other factors that affect the behavior of contaminants include fracture network geometry, variations in fracture aperture and roughness, and the local hydrodynamics where fractures intersect (Neuman, 1990). This is discussed more fully in Chapter 3.

Diffusive transport occurs in response to random molecular motion. At the macroscale, diffusive transport reflects concentration gradients and the contaminant moves from zones of higher to lower concentration. Chemical constituents of fluids in fractures initially diffuse quickly into fluids in the rock matrix surrounding fractures because concentration gradients are high. They diffuse more slowly back into the fractures from the matrix because the gradients are much smaller and are present in two directions (away from and toward the fracture). This slow release of chemicals from the rock matrix may occur over long periods (decades to centuries for some substances).

The fate and transport of chemicals in fractured rock are also affected (sometimes profoundly) by chemical and biological processes. Chemical alterations can include precipitation of dissolved constituents from solution, dissolution of solid constituents into groundwater, sorption of chemicals onto the surfaces of rock and into the rock matrix, changes in oxidation state that affect mobility or toxicity, volatilization from liquid to gaseous state, dissolution from gaseous state into liquid state, and chemical transformations into different compounds. These chemical changes, in turn, can be affected by changes in dissolved oxygen, pH, temperature, and other parameters. Biological processes also affect these chemical processes in fractured rock sites as microorganisms, and their byproducts can act as catalysts for chemical reactions. This can result in contaminant transformations, precipitation/dissolution reactions, gas bubble nucleations, and changes in pH and redox conditions. Microorganisms can also form flocs (aggregates of particles) and biofilms that modify fluid flow and may result in clogging. Biological processes are especially critical in terms of biotransformation of organic and inorganic contaminants in the subsurface. In fractured rock formations, biological processes may be significant, depending on the specific site, at both shallow depths and at depths of many kilometers.

KNOWLEDGE FROM RELATED AREAS

This report focuses on the understanding of fractured rock in the context of understanding the behavior of contaminants and contaminant remediation. However, there has been significant research on fluid movement in rock for applications in the energy sector, including that related to oil and gas exploration and production, geothermal energy production, and carbon capture and storage (CCS) (Zimmermann et al., 2011; Rutqvist, 2012; Kim and Moridis, 2015). These energy-related activities occur at depths from 1 to 4 kilometers, so the knowledge gained from research in those areas informs the main topics of this report. Whereas this report does not include lengthy discussion of research and development in the energy sector, relevant references from that sector are provided throughout this report.

REPORT ORGANIZATION

This report examines recent progress and addresses issues raised since the publication of the comprehensive 1996 NRC report Rock Fractures and Fluid Flow: Contemporary Understanding and Applications. Progress has been made characterizing, modeling, monitoring, and remediating the fractured rock environment, but significant challenges remain. These are due, in part, because only a small number of fractures often carry most of the flow in a rock mass. Without understanding the geologic context of a site, it can be difficult to identify and locate those hydrologically important fractures and understand how contaminants may be transported and move between rock fractures and matrix. Modeling the interactions between concurrent physical, chemical, thermal, and biological processes that determine contaminant migration is inherently complex.

Subsequent chapters of this report examine progress in these and other areas related to the Statement of Task:

• Chapter 2 provides background on the characteristics of fractured rock important to contaminant migration.
• Chapter 3 considers fundamental phenomena underlying coupled processes that occur in fractured rock and control the fate and transport of contaminants within fractures, and between the fractures and rock matrix.
• Chapter 4 explores the importance of appropriate hydrostructural conceptual models and the representation of physical processes in contaminant migration models in fractured rock systems, including inherent limitations. The goal of a modeling effort is to examine how, where, and how fast contaminants might move in the subsurface; examine the impact of potential efforts to remediate contaminated sites; and inform where and which characterization and remediation efforts may be most efficient and cost-effective.
• Chapter 5 reviews methods for acquiring information about the subsurface in ways that support conceptual and quantitative modeling of the hydrostructural system and contaminant transport and storage and that determine the effectiveness of engineered systems.
• Chapter 6 explores options available for site remediation and management and provides a set of important considerations for remediation of fractured rock.
• Chapter 7 considers decision-making processes for characterizing, monitoring, and remediating fractured rock sites.

Findings and conclusions are found throughout this report; however, Chapter 8 synthesizes the committee’s overarching findings and recommendations. These stress the importance of interdisciplinary approaches to understanding fractured rock and the interactions and processes within it that control contaminant transport and fate; the importance of developing appropriate hydrostructural conceptual models and applying numerical models that accommodate them; the need for developing appropriate goals for fractured rock site remediation; and the importance of adaptive and observational approaches. Recommendations in this report are written to improve science, engineering, and research. An integrated systems approach for engineering fractured rock sites is emphasized, so recommendations are presented in a broader systems context.