AQUIFERS. Report to the Water Research Commission. J.A. Pretorius, B.H. Usher and R.A. Gebrekristos

Transcription

1 GROUNDWATER MONITORING GUIDELINES FOR DNAPLS IN SOUTH AFRICAN AQUIFERS Report to the Water Research Commission by J.A. Pretorius, B.H. Usher and R.A. Gebrekristos Institute for Groundwater Studies University of the Free State Bloemfontein 9300 Report to the Water Research Commission Project Leader: B.H. Usher WRC 1501/3/08 ISBN Set January 2008

2 The publication of this report emanates from a project entitled: Field investigations to study the fate and transport of dense non-aqueous phase liquids (DNAPLs) in groundwater (WRC Project No K5/1501) DISCLAIMER This report has been reviewed by the Water Research Commission (WRC) and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the WRC, nor does mention of trade names or commercial products constitute endorsement or recommendation for use

3 Groundwater monitoring guidelines for DNAPLs This document provides suggested monitoring guidelines for DNAPL contaminated sites in South Africa. Weaver et al., 2007* (WRC Document TT 303/07) have recently published updated guidelines for sampling in this country, and the concepts contained in that document should also be considered. However, due to the fact that DNAPLs are a class of contaminants with distinctly different properties, several considerations for monitoring borehole placement, construction and sampling are provided in this document. This should be of use to site investigators, geohydrologists, site-owners and regulators (DWAF and local authorities). This document forms part of a series of documents, produced by Water Research Commission project K5/1501 Field investigations to study the fate and transport of dense non-aqueous phase liquids (DNAPLs) in groundwater. The documents in this series include: Executive Summary of the Project Manual for Site Assessment at DNAPL Contaminated Sites in South Africa Groundwater monitoring guidelines for DNAPLs Guidelines for the acceptance of Monitored Natural Attenuation processes in South Africa Handbook for DNAPL Contaminated Sites in South Africa An Introduction to DNAPLs in South Africa: A Citizen Guide Field and laboratory investigations to study the fate and transport of DNAPLs in groundwater All these documents are contained on the CD included with this report. *Weaver, JMC, Cavé, L and Talma, AS (2007) Groundwater Sampling (Second Edition) WRC Report TT303/07. ISBN , Water Research Commission, Pretoria. page i

4 page ii

5 Contents 1. BACKGROUND AND INTRODUCTION Introduction Types and Occurrence of DNAPL contaminants Physical Properties related to DNAPL Migration Transport and Fate of DNAPLs in the Subsurface DNAPL Site Characterisation 6 2. SOUTH AFRICAN AQUIFER TYPES Desk Studies Field Investigations Data Analysis System Design Installation MONITORING BOREHOLES Introduction Drilling Drilling techniques Air Percussion drilling Auger Drilling Core drilling Compatibility of construction materials Borehole construction Monitoring Locations SAMPLING PROCEDURES NAPL samples Water samples Decontamination Sample preservation Sample analysis Sampling procedures First-run sampling procedure Routine Sampling Procedure REFERENCES 36 page iii

6 Tables Table 1: Factors facilitating or inhibiting DNAPL penetration...6 Table 2: Unconsolidated material characteristics to determine in site investigations...8 Table 3: Bedrock properties to determine during site investigations (From Kueper et al., 2003)...8 Table 4: Flow mechanisms of main aquifer systems in South Africa Table 5: Properties of selected South African aquifers and their implications for contaminant transport...10 Table 6: Borehole construction design and the effect it has on measured DNAPL thickness in borehole...16 Table 7: Analytical target groups for organic water analysis Table 8: Inorganic parameters to analyses for during site assessment of a DNAPL contaminated site Figures Figure 1: Residual DNAPL in the unsaturated or vadose zone...3 Figure 2: Summary of groundwater monitoring design steps (Adapted from Sara, 2003) Figure 3: Multiple piezometer installation...21 Figure 4: Depth specific boreholes in multiple...22 Figure 5: Construction of Borehole A or B/C...22 Figure 6: Plan view of schematic location of monitoring boreholes around a plume...24 Figure 7: Cross-section of schematic location of monitoring boreholes around a plume...25 page iv

7 1. BACKGROUND AND INTRODUCTION 1.1 Introduction The potential for serious long-term contamination of groundwater by Dense Non-Aqueous Phase Liquid (DNAPL) chemicals at many sites is largely due to their toxicity, limited solubility (but much higher than drinking water limits) and significant migration potential in soil gas, groundwater, and/or as a separate phase. DNAPL chemicals, especially chlorinated solvents, are among the most prevalent groundwater contaminants identified in groundwater supplies and at waste disposal sites (Cohen and Mercer, 1993). These contaminants have also been identified as priority contaminants in South African aquifers (Usher et al, 2003). The subsurface movement of DNAPL is significantly controlled by the nature of the release, the DNAPL density, interfacial tension and viscosity, porous media capillary properties and, usually to a lesser extent, hydraulic forces. Below the water table, non-wetting DNAPL migrates preferentially through permeable pathways such as soil and rock fractures, root holes and sand layers that provide relatively little capillary resistance to flow. Visual detection of DNAPL in soil and groundwater samples may be difficult where the DNAPL is transparent, present in low saturation or distributed heterogeneously. These factors impede characterisation of the movement and distribution of DNAPL even at sites with relatively homogeneous soil and a known, uniform DNAPL source. The difficulty of site characterisation and therefore monitoring, management and remediation of DNAPL contaminated sites is further compounded by fractured bedrock heterogeneous strata, multiple DNAPL mixtures and releases, etc. South African groundwater systems differ in many ways from those overseas. Monitoring methodologies and requirements that have been developed overseas do not necessarily apply to the South African situation. This could lead to incorrect sampling and interpretation and result in unnecessary expenditure. Many of the requirements as specified in this document deviate from those stipulated in classic text-books or the South African Minimum Requirements documents. These deviations do not contradict well-established methodologies, but modifications have been introduced in accordance with the batneec principle for South African conditions. This should not deter those who want to follow procedures that are more stringent from doing so. Apart from listing monitoring requirements, the intention is to explain: Groundwater and DNAPL behaviour Reasoning behind monitoring Installation of a monitoring system Principles of water sampling Appropriate laboratory analysis 1.2 Types and Occurrence of DNAPL contaminants In general, a DNAPL is defined as a heavier-than-water organic liquid that is only slightly soluble in water. The potential for widespread contamination of groundwater by DNAPLs is substantial because of the extensive production, transport, utilisation and disposal of large volumes of DNAPL chemicals. There are thousands of potential sites in South Africa, as elsewhere in the world, where DNAPLs may have been released to the page 1

8 subsurface in varying quantities. Often DNAPL chemicals are detected at contaminated sites even where DNAPL presence has not been determined. All DNAPLs can be characterised by their physical properties such as density, viscosity and interfacial tension with water, component composition, solubility in water, vapour pressure and wettability. The most prevalent DNAPL types are outlined in Appendix A, with summary information on DNAPL density and viscosity, appearance and usage. The first step in the assessment of a potential DNAPL site is the consideration of the potential chemicals that might be present at the site. A wide variety of chemical products and wastes may comprise a DNAPL. For a chemical (or chemical mixture) to be considered as a DNAPL, it must have a fluid density greater than 1.01g/cm 3, a solubility in water of less than 2% (or mgll) and a vapour pressure of less than 300 torr (Pankow and Cherry, 1996). Thus, the amount of chemical released has to be in excess of the total amount that could be dissolved or be sorbed on the soil materials in the source zone for the NAPL phase to be present. The major DNAPL types include: halogenated hydrocarbons, especially solvents, coal tar and creosote, polychlorinated biphenyls (PCBs), some pesticides and miscellaneous or mixed DNAPLs. Of these types, the most extensive subsurface contamination is associated with halogenated (primarily chlorinated) solvents, either alone or within mixed DNAPL sites, due to their widespread use and properties (high density, low viscosity, significant solubility and high toxicity). 1.3 Physical Properties related to DNAPL Migration The physical properties of the organic compounds affect their behaviour in the subsurface (Fetter, 1999). Some of these physical properties are, for example, interfacial tension and wettability, density, viscosity, solubility, Henry s law constants, capillary pressure and relative permeability. Table 1 summarises some of the important properties of DNAPL compounds and how these affect their behaviour as groundwater contaminants. 1.4 Transport and Fate of DNAPLs in the Subsurface When spilled, DNAPLs will move downward through the unsaturated zone, trapping some (residual) DNAPL in the pore space. If a large volume of DNAPL is spilled and permeability exists in the subsurface, the DNAPL will penetrate past the water table and continue moving downward through the saturated zone due to gravity. The subsequent subsurface migration of DNAPLs is not a function of conventional groundwater transport mechanisms (i.e. advection, dispersion and diffusion), but rather a function of geological structures (i.e. fissures, bedding planes, etc.) and gravity (including the bedrock topography). However, soluble constituents of DNAPLs will dissolve into groundwater and their distribution is a function of the hydraulic gradient, resulting in a groundwater plume down gradient from the contaminant source, as with other dissolved contaminants. The potential depth of DNAPL penetration through the vadose zone and into the groundwater will depend on the properties of the DNAPL, the nature of the DNAPL release and the properties and geological structure within the vadose and groundwater zones. page 2

9 DNAPLs tend to partition among all phases in the subsurface and the four partitioning processes which play a role in the fate and transport of DNAPLs are: Dissolution into water phase Volatisation of dissolved chemicals from water phase to air phase Vaporisation of DNAPL into air phase, and Sorption of dissolved chemicals from water to solids. Aquifer/soil grain Water DNAPL Air Figure 1: Residual DNAPL in the unsaturated or vadose zone (From Kueper et al., 2003). Both residual DNAPL and pools will dissolve into groundwater flowing through the DNAPL source zone, giving rise to aqueous phase plumes. Given the tortuous and sporadic nature of DNAPL occurrence within the source zone, it follows that the associated aqueous phase plumes will exhibit significant spatial variability in terms of concentration. Various factors influence the magnitude of contaminant concentrations obtained from monitoring borehole samples relative to the actual concentrations in the aquifer. The net effect of hydrodynamic dispersion, inborehole dilution, monitoring borehole placement and potential degradation processes is that contaminant concentrations in a sample obtained from a monitoring borehole downstream of a DNAPL source zone may be significantly less than the aqueous solubility of the DNAPL of interest (Kueper et al., 2003). Experience has shown that a DNAPL source may be present upstream of a monitoring well if sample concentrations exceed 1% of the effective solubility of the component of interest (US EPA, 1992). The 1% rule of thumb has been criticised because it does not provide guidance on how far upstream the DNAPL source zone is located. It is clear that a variety of site-specific factors influence the magnitude of sampled contaminant concentrations and that some of these factors cannot be determined (for example, the distance a monitoring borehole is offset from plume centre line and the amount of in-borehole dilution occurring during purging). page 3

10 The 1 percent rule of thumb should not be used in isolation to establish DNAPL presence at a site, but instead, should be used with other converging lines of evidence to establish both the DNAPL presence and to delineate the spatial extent of the source zone (Kueper et al., 2003). DNAPLs are rarely found at contaminated sites as a single component contaminant, but rather as a mixture of contaminants. Such a mixture of contaminants will not dissolve into groundwater at their single component, textbook solubility values. Rather, the various components may compete for the dissolution process. The dissolution of a multi-component NAPL may be described using Raoult s law. Raoult s law states that the effective solubility of a NAPL component in (ground)water is equal to the product of the mole fraction in the NAPL and the single component aqueous solubility of that compound: Ci = misi Where: Ci is the effective solubility of component i; mi is the mole fraction of component i in the NAPL; Si is the single component solubility of component i. In practical terms, the effective solubility of a component is the maximum concentration that could possibly be observed in groundwater. Sample concentrations obtained from monitoring boreholes in the plume will be less than the effective solubility due to in-borehole dilution, hydrodynamic dispersion and other effects. The relative concentration of individual components, however, is dictated by Raoult s law. In terms of the 1% rule of thumb, it is 1% of the effective solubility that is taken as an indicator of possible upstream DNAPL presence, not 1% of the single component solubility. Once DNAPL is present in fractured bedrock, it will slowly dissolve into groundwater flowing through open fractures, giving rise to aqueous phase plumes. The plumes will generally migrate in the hydraulically downgradient direction subject to advection, dispersion, sorption to fracture walls, possible biodegradation and matrix diffusion. As with plume migration in unconsolidated deposits, the chemical composition of the plume will be a function of the chemical composition of the DNAPL. Therefore, the plume would expect to be dominated by higher effective solubility compounds at an early time, gradually shifting later towards higher concentrations of the lower solubility compounds. Like plumes in unconsolidated deposits, plumes in fractured bedrock may eventually reach a steady-state configuration where the leading and side edges of the plume (as defined by a specific concentration level) are no longer expanding (Kueper et al., 2003). Thus, cognisance should be taken of the factors which will influence contaminant concentrations measured in monitoring boreholes, when delineating the dissolved plume. With respect to physical processes influencing plume behaviour, there is one fundamental difference between porous and fractured aquifers. Plumes in fractured formations are subject to a process known as matrix diffusion. Matrix diffusion refers to the process in which solutes dissolved in groundwater diffuse into and out of the rock matrix. If concentrations are higher in the open fracture, the diffusion process will result in dissolved contaminants moving into the rock matrix (forward diffusion). If concentrations are higher in the rock matrix, dissolved contaminants will move out of the rock matrix and into water in the open fractures (back diffusion). Matrix diffusion will occur in all rock types exhibiting a finite matrix porosity. page 4

11 The matrix diffusion process may cause that the dissolved plumes in fractured material migrate slower than the rate of groundwater flow. The rate of plume may be attenuated with rates as high as 100 or more, relative to the rate of groundwater migration (Kueper et al., 2003). The time-scale of remediation in fractured rocks is also often controlled by the back diffusion process and not by the presence of DNAPL in fractures. Even for relatively short initial exposure times, the back diffusion process may continue for many decades. page 5

12 Table 1: Factors facilitating or inhibiting DNAPL penetration (Adapted from Pankow and Cherry, 1996). Factors facilitating DNAPL penetration High DNAPL density Low interfacial tension Low viscosity Large DNAPL volume release Long duration DNAPL release High permeability Vertical and sub-vertical geological structures Factors inhibiting DNAPL penetration Low DNAPL density High interfacial tension High viscosity Small DNAPL volume release Short duration DNAPL release Low permeability Horizontal bedding in sandy aquifers, silt and clay aquitards Horizontal bedding plane partings in sedimentary rocks Typical circumstances Chlorinated solvents, PCB aroclors Surfactants or miscible co-solvents e.g. methanol, ketones, acetone in DNAPL Surfactants or co-solvents in aqueous wastes or groundwater Complex mixtures Chlorinated solvents Disposal in landfill or dams Catastrophic and/or ongoing spills Disposal in landfill or dams Catastrophic and/or ongoing spills Sand, gravel and fractured rock Angled bedding planes Fractures and fissures in aquitards Fractured rocks Typical circumstances Coal tar/creosote Chlorinated solvents at low concentrations in petroleum hydrocarbons Relative pure chemical products Coal tar/creosote PCB aroclors Mixtures with high concentrations high molecular hydrocarbons Small spills and leaks Small once off spills and leaks Unfractured silt and clay aquitards Unfractured rock 1.5 DNAPL Site Characterisation Many site characterisation techniques have been developed during the past few decades. However, it is not possible to apply all of the methods at the same time because of practical reasons. Consideration should be given to the following aspects: DNAPL contaminant properties South African aquifer characteristics Available technology in South Africa Innovative and cost-effective assessment methodologies. DNAPL contaminated sites tend to be more complex than light non-aqueous phase liquid (LNAPL) or aqueous contaminated sites because the physical and chemical characteristics of dense, sparingly soluble contaminants page 6

13 add additional complexity to heterogeneous geology and hydrogeology at most sites. LNAPL contamination is usually constrained to the top of the water table and above and aqueous phase contamination follows the hydrology of the site. The subsurface movement of DNAPL is controlled substantially by the nature of the release, the DNAPL density, interfacial tension and viscosity, porous media capillary properties, gravity and, usually to a lesser extent, hydraulic forces. Obtaining a detailed delineation of subsurface DNAPL, therefore, can be very costly and may be impractical using conventional site investigation techniques. Furthermore, the risk of causing DNAPL migration by drilling or other actions may be substantial and should be considered prior to commencing fieldwork. Although DNAPL contamination may greatly complicate site characterisation, failure to adequately define its presence, fate and transport may result in misguided investigation and remedial efforts. Large savings and environmental benefits may be realised by conducting studies and implementing remedies in a cost-effective manner. Cost-effective DNAPL site management requires an understanding of DNAPL properties and migration processes and of the methods available to investigate and interpret the transport and fate of DNAPL in the subsurface. Several techniques may be applied for the characterisation of a site contaminated with dense nonaqueous phase liquids (DNAPLs). These include: Evaluation of historical records in the vicinity Geophysical methods Vapour analysis Invasive methods such as drilling or excavations Visual and olfactory examination during drilling Aided visual inspections using dyes and/or UV light Aquifer and tracer tests Down-hole geophysics, geochemical and flow logging Soil, water and NAPL sampling and analysis A general assessment of an industrial or waste disposal site can be made to determine the potential presence of DNAPL in the groundwater zone. This general assessment is usually made early in a site investigation programme using existing information (e.g. chemicals used or stored and the activities at the site). Measurement of the DNAPL properties will require the recovery of a sample of DNAPL from the subsurface. If this is not possible but the composition of the DNAPL is known, the density and viscosity may be estimated from the literature sources. DNAPL-water interfacial tension should not be estimated from handbooks; however, as this is a site-specific parameter which is influenced strongly by even small amounts of impurities. The organic carbon partition coefficient is typically obtained from the literature sources, along with the DNAPL vapour pressure. The contaminant half-life depends on site-specific geochemical conditions and therefore, should not generally be taken from handbooks or the literature; this parameter is typically determined through model calibration. page 7

14 Table 2 and Table 3 list the properties of the aquifer media, which will be relevant to establish a conceptual site model and to use the information during the site characterisation. Many of these parameters may be estimated from the literature or ideally measured during field investigations. Table 2: Unconsolidated material characteristics to determine in site investigations (From Kueper et al., 2003) Parameter Porosity Dry bulk density Fraction organic carbon Hydraulic conductivity Displacement pressure Bulk retention capacity Contact angle Hydraulic head distribution Bedding structures Spatial extent of DNAPL source zone Spatial extent of plume Example use of information Plume velocity calculation; Diffusion calculations DNAPL threshold concentration calculation Plume velocity calculation; DNAPL threshold calculation Plume velocity calculation; Design of extraction wells Pool height calculations DNAPL mass estimate Refinement of conceptual model on DNAPL mobility Directions and velocity of groundwater flow Directions of DNAPL migration Guides remedy selection and design Guides remedy selection; risk analysis Table 3: Bedrock properties to determine during site investigations (From Kueper et al., 2003) Parameter Matrix porosity Matrix dry bulk density Matrix fraction organic carbon Orientation of major fracture sets Fracture spacing Fracture porosity Bulk rock hydraulic conductivity Hydraulic head distribution Bulk retention capacity Contact angle Spatial extent of DNAPL source zone Spatial extent of plume Example use of information Diffusion calculations Estimate of remediation timeframe Estimate of (retarded) plume velocity Determine direction of plume migration Directions of DNAPL migration Diffusion calculations Plume velocity calculation Plume velocity calculation Design of extraction wells Directions and velocity of groundwater flow DNAPL mass estimate DNAPL rock-water wetting relationship Guides remedy selection Guides remedy selection; risk analysis 2. SOUTH AFRICAN AQUIFER TYPES A basic understanding of the nature and occurrence of groundwater in South Africa aquifers is a prerequisite for the design of monitoring systems at DNAPL contaminated sites. This section provides a general introduction to the topic and the implication for DNAPL contaminated sites. An aquifer means a geological formation which has structures or textures that hold water or permit appreciable water movement through them. (National Water Act (Act No. 36 of 1998)). page 8

15 The physical properties of an aquifer that have the greatest impact on contamination transport (dissolved phase) are the flow rate and flow mechanism present, and the hydraulic conductivity. The hydraulic conductivity of an aquifer depends upon a number of physical factors including porosity and particle size distribution. In isotropic, intergranular flow systems aquifer characterisation, can predict groundwater flow patterns and attenuation processes with the greatest confidence. By contrast, characterisation of the hydraulic regime in fractured aquifers is complicated by the highly heterogeneous nature of the system. Long-term predictions of plume development are, therefore, likely to be inherently uncertain. For the purposes of this document, the main aquifer systems in South Africa are classified according to the dominant flow mechanisms, flow characteristics and the implications for contaminant transport (Table 5). Four types of flow mechanisms are identified, intergranular (porous) flow, intergranular and fracture flow, fracture flow and karst (dolomitic) flow. Preferential groundwater pathways, including fractures, joints and solution channels result in higher contaminant velocities, which in turn, may lead to rapidly expanding contaminant plumes. Groundwater flow under these conditions is highly unpredictable, making plume characterisation difficult. Intergranular flow is more predictable and the travel rates are lower, providing greater time for degradation to occur and longer exposure of contaminants to active biodegradative/mineral sites. Only about 10% of the South African aquifers are of the type where intergranular (porous) flow is the dominant flow mechanism (DWAF, 1998). In these instances, flow is around grains of sand and clay, which make up the aquifer. Examples of such aquifers are: coastal sands, gravels and other unconsolidated material along the South African coast, and sands and gravels along stream beds. The term, fracture flow, is the description of groundwater movement through a variety of secondary structures in rock. Geologically, these structures may be defined as joints, cracks, fractures and faults. The degree of fracturing of rocks in South Africa is a function of the tectonic history of the rocks, as well as the rock composition. Competent rocks, such as dolerite and quartzite, for instance, fracture more readily than incompetent or ductile rocks, such as dolomite and shale. The degree of fracturing in an aquifer is not necessarily a measure of the degree to which the aquifer can transmit water. Many of the fractures are tight, because of compression forces acting within the earth s crust. Usually, at depths more than 60 metres below the surface, fewer than 1% of the fractures transmit significant amounts of water. Exceptions occur within quartzitic rocks, where significant yields are possible at greater depths (DWAF, 1998). A combination of fracture flow and intergranular (porous) flow mechanisms often exists in a single aquifer. Two examples of aquifers of this type are sandstone and weathered granite. Sandstone has an inherent permeability of its own and water can, to a lesser or greater extent, flow around the grains within the sandstone, depending on their degree of cementation. The fractures are usually the dominant flow mechanism within the sandstone. Granite is an igneous rock and, like all other igneous rocks, impermeable to groundwater flow in its unfractured and fresh state. However, granite weathers comparatively easily. Weathering usually starts along fractures in the granite, eventually affecting large areas within the granitic mass. When weathered, crystals within the rock disengage and water can flow around individual crystals within the granite. Depending on the degree of weathering, the groundwater flow mechanism within granite mass may therefore be dominantly fracture flow or porous flow (DWAF, 1998). Dolomite is a crystalline rock, unstable in acid environments. Dissolution channels that develop along fractures within dolomite may extend to the surface and give rise to a typical karst topography, which has a very significant influence upon recharge characteristics of the aquifer. Dolomitic aquifers that are not protected by overlying geological formations are particularly vulnerable to pollution because of their thin soil cover and high transmissive characteristics (DWAF, 1998). page 9

16 Table 4: Flow mechanisms of main aquifer systems in South Africa. Dominant flow Hydrogeological Porosity type mechanism domains Intergranular Primary Quaternary and Tertiary deposits Intergranular and fractured Dual Sedimentary rock and composite rock regions Fracture flow Fracture Crystalline metamorphic and igneous regions, Intrusive and extrusive rock regions Karst Karstic Sedimentary rock regions and composite rock regions Examples Alluvium, Cape Flats, Kalahari Sands Table Mountain Group, Karoo Group West Rand Group, Basement Karst Belt, Ghaap plateau Rock type Unconsolidated sands Sandstones, shales, arenites Granites Dolomite Table 5: Properties of selected South African aquifers and their implications for contaminant transport Aquifer Issues Implications Sandstones, shales, arenites Intergranular flow High fracture porosity, low matrix porosity Fracture/pore water interaction Multi-layered aquifer due to presence of hydraulic variance in horizons Horizontal and vertical flow important Generally deep aquifer Dolomite Granite Unconsolidated Sand/ gravel Fracture and Karst flow Preferential pathways Rapid flow Low matrix porosity Fracture/pore water interaction Significant variation in vertical and horizontal permeability Large seasonal water table variation Regional variations in behaviour of the aquifer (e.g. solution features) Fracture flow Preferential pathways Rapid flow Low matrix porosity Significant variation in vertical and horizontal permeability Thin aquifer Shallow water table High porosity High organic and clay content Variable lithology and permeability Construction of boreholes due to multilayered system and potential for diving plumes Slow groundwater velocities Vertical hydraulic gradients High cost of investigation in thick aquifers/ deep unsaturated zones Layering may limit vertical dispersion Borehole location in relation to preferential pathways Variable and rapid travel times need to be taken into account in assessing viability of attenuation and frequency of monitoring Short-circuiting of normal flow regime at times of high water table Thick aquifers and deep unsaturated zones may result in high cost of investigation Borehole location in relation to preferential pathways Difficult to be confident about representative nature of monitoring results Range of investigation techniques Thin unsaturated zone Easiest translation of US and other international experience to SA hydrogeology page 10

17 Figure 2: Summary of groundwater monitoring design steps (Adapted from Sara, 2003). The key to successful monitoring is the linking of point information into larger systems, referred to as monitoring networks. Monitoring networks operate on local, regional and national scales. A local monitoring network is intended for the single facility/site, whereas regional monitoring relates to a combination of facilities, such as those usually present at mines, power stations, other large industries and large municipalities. Monitoring on a national scale, could, for instance, be meaningful in terms of salt loads within catchments (DWAF, 1998). Monitoring on all these levels is necessary. However, for the purposes of this document, the emphasis is only on the local (or site specific) monitoring network at a DNAPL contaminated site. The design of monitoring systems for any DNAPL contaminated site should follow a certain sequence of events. The flow chart above is a summary of the steps required when planning a monitoring system. page 11

18 2.1 Desk Studies The first step in planning a monitoring system will comprise mainly of a desk study for the purpose of gathering background information on the site and defining the objectives of the monitoring system. The objective of the monitoring system will determine the level of investigation and further field studies which will be required to implement the chosen monitoring system. The objectives of monitoring of groundwater contamination are to: provide reliable data on the quality and chemical composition of the groundwater, detect and quantify the presence and seriousness of any (DNAPL) contaminants in the groundwater at the very earliest stage possible, detect the possible release or impending release of contaminants from specific source areas, provide a rational comparison between the predicted and actual flow and solute transport rates, and provide an ongoing and reliable performance record for the design and control system(s) for effectively controlling pollution. The collection of background information is made to determine the potential presence of DNAPL in the groundwater zone. This will be used to construct the initial conceptual model of the aquifer system. Typical background information will include: Once this background information is collected and processed, an initial conceptual model can be constructed for the site. The level of field investigation that will be required for the implementation of a monitoring system may now be considered. It is important to understand the: (1) type of chemicals used at a site, but also how these chemicals have been used or are currently being used. Thus with an appreciation of the varied uses of DNAPL chemicals, a site assessor can identify the (2) site activities that should be associated with DNAPL use and disposal. Possible information sources include: interviews with present or former employees; records of chemical purchases, off-site waste disposal or waste received; historical site engineering drawings; and archival aerial photographs. An estimation of the (3) potential depth of DNAPL penetration/migration through the vadose zone and into the groundwater zone is an important component of a site assessment (Pankow and Cherry, 1996). This estimation is often attempted early on in a site investigation and based on general information about site operations and subsurface conditions. The depth of the penetration/migration will depend on: Properties of the DNAPL; Nature of the release; Properties and geological structure within the vadose and groundwater zone. Although it is not possible to predict precisely the extent or rate of DNAPL migration, it is important to recognise, in general terms the factors that will influence the depth of DNAPL penetration in the subsurface. This information is used by the site investigator to select appropriate depths to which the monitoring boreholes must be drilled. page 12

19 2.2 Field Investigations Following development of the initial site conceptual model based on available information collected during the general site assessment, a combination of non-invasive and invasive field methods/techniques will generally be required to advance site characterisation, enable the investigator to refine the site specific geology/geohydrological conceptual model and to define the aquifer systems present at the site. This phase will often include geophysical surveys, soil-gas surveys and drilling and the installation of observation boreholes/ piezometers and aquifer testing. This phase will include the sampling of soil and groundwater to determine the contaminants present and the spatial extent of the contamination in the aquifer system (DNAPL and dissolved phases). The objective of the monitoring/management plan and the complexity of the site will determine the extent of the site characterisation. 2.3 Data Analysis The next step entails data processing and analysis of information gathered during the desk study and the field investigations. Typically, this will include drawing water level elevation maps, geological crosscuts, the determination of aquifer parameters and the refinement of the geological and geohydrological conceptual model. Depending on the objectives and the detail of the study, predictive flow and transport modelling (of the dissolved phases) can be done. The results from the numerical modelling will help in the design of the monitoring network and placement of monitoring boreholes. 2.4 System Design The monitoring system design process is where the target monitoring areas are selected, monitoring borehole positions are proposed and borehole designs are put forward. In terms of regulatory requirements, it is important to monitor the groundwater quality both hydraulically upstream and downstream from a selected site or facility. In DWAF s minimum requirements (1998) for monitoring at waste sites, the following requirement is set for local monitoring networks: Local monitoring networks should extend beyond pollution plumes to allow for the delineation of plumes and investigations into the pollution migration rate. The site or facility design must be taken into account when selecting appropriate borehole positions. The location of specific site features may limit the positioning of monitoring boreholes. The locations of source areas and receptor areas such as ponds, streams or production boreholes also need to be considered. The conceptual model of the site is used to determine the position and construction of the boreholes. The conceptual model should be three-dimensional. The conceptual model should include information on the potential DNAPL source zones (location and extent) and the factors which will influence the free phase penetration/migration and the transport of the dissolved phase. From the geology/geohydrology conceptual models, the extent of the most likely aquifer to be monitored may be delineated. page 13

20 The groundwater flow directions (three-dimensional flownets) and flow rates should be calculated for the aquifer(s). Thus, all reasonable flow paths must be identified. In the case of a fractured aquifer, fracture characterisation will be required to determine the preferred pathways through the system. From all of this, the (3D) monitoring target zone may be delineated. Where more than one aquifer system is present, it is important to ensure that the monitoring boreholes should be representative of all the aquifers. The boreholes/piezometers should be constructed so that vertical and horizontal hydraulic gradients may be established for each of the aquifers. Possible interconnection between aquifers must be investigated. Where multiple aquifer systems are present, depth specific boreholes (well nests or multiple boreholes) are recommended, rather than multiple piezometers in one borehole. At a DNAPL contaminated site the risk of cross contamination between the piezometers is greater than in boreholes constructed, so that only one aquifer system is intersected. The objective of the monitoring network will determine the extent to which multiple aquifer systems will be monitored. For example, if the objective is only to monitor the immediate detection of a contaminant release, only the uppermost unconfined aquifer will be monitored. If the objective is to assess the extent of the contamination from a facility or site, all aquifer systems (localised and regional) should be monitored, including any possible interconnection with surface water bodies. 2.5 Installation Installation of initial monitoring boreholes has to be done once the monitoring system s design is completed and appropriate locations and construction of boreholes has been selected. Boreholes should be installed and constructed according to the design plan (see also following section). Where existing monitoring boreholes are used, it important to verify that these boreholes have been drilled to the appropriate depth, and are constructed so that the target aquifer system is measured. The installation of a monitoring system must be an iterative process, where improvements to borehole design, drilling depths, materials used and locations is done based on the findings of early drilling. It is also recommended that the design and construction of newly installed monitoring boreholes should be verified during the initial phase of monitoring. During this initial phase of monitoring, the site conceptual model can be finalised. Where the site owners are responsible for the long-term monitoring of the site, it is important that the on-site personnel be adequately trained in the use of the database, the sampling equipment and in the interpretation of the data. Depending on the regulatory and/or in-house requirements, information dissemination should take place on a regular and structured basis. This will ensure that timely action can be taken if required and that the monitoring network can be optimised based on data collected. page 14

21 3. MONITORING BOREHOLES 3.1 Introduction Monitoring boreholes are installed to characterise flow directions and rates, groundwater quality, NAPL fluid distributions and media hydraulic properties. Pertinent data may be acquired by conducting fluid thickness and elevation surveys, fluid sampling surveys, hydraulic tests and down-hole surveys. The locations and design of the boreholes are selected based on consideration of the site conceptual model and specific data collection objectives (Cohen and Mercer, 1993). The design and construction of boreholes at DNAPL sites require special consideration of: The effect of borehole design and location on NAPL movement and distribution within the borehole and the surrounding environment; The compatibility of construction materials with NAPLs and dissolved chemicals; and Borehole development options. Inappropriate design may increase the potential for causing vertical DNAPL migration and misinterpretation of fluid elevation and thickness measurements. Table 6 lists the relationship between borehole construction with the mobile DNAPL phase and capillary barriers. The construction of the borehole may influence direct measurements of DNAPL in boreholes and cause potential vertical DNAPL migration in the borehole environment. A borehole that is completed to the top of a capillary barrier and screened from the capillary barrier surface to above the DNAPL-water interface, is most likely to provide DNAPL thickness and elevation data that are representative of formation conditions. It is therefore very important to record all observations during drilling in order to select the correct construction method for the borehole. Furthermore, it is important to note that most of these observations noted above and in Table 6 are relevant to unconsolidated media. Pankow and Cherry (1996) observe that the only conclusion that can be drawn from the observation of DNAPL in a borehole in fractured media, is the presence of DNAPL at the site. The height of DNAPL measured in the borehole may be misleading. The same height of DNAPL in the borehole may be measured for two different scenarios: either the presence of DNAPL corresponds with the general same elevation of DNAPL in the formation, or the height of DNAPL measured in the borehole is from accumulation of DNAPL from a higher elevation in the formation. The borehole may also act as a conduit and DNAPL may short-circuit to the formation below the borehole if the capillary pressure exceeds the entry pressure for a fracture below the borehole. It is therefore very important to ensure that boreholes in fractured media are adequately sealed (with e.g. bentonite) at the bottom and the screen intervals selected according to the geological log, with the purpose of preventing downward mobilisation of DNAPL. page 15

22 Table 6: Borehole construction design and the effect it has on measured DNAPL thickness in borehole. (Adapted from Cohen and Mercer, 1993.) Borehole construction If the borehole screen or casing extends below the top of a DNAPL barrier layer If the borehole bottom is set above the top of a DNAPL barrier layer If the borehole connects a DNAPL pool above a barrier layer to a deeper permeable formation and the borehole causes DNAPL to short-circuit the barrier layer and contaminate the lower permeable formation NAPL which enters a coarse sandpack may sink to the bottom of the sandpack rather than flow through the borehole screen The bottom of the borehole screen is set above the bottom of the sand pack and there is no casing beneath the screen Sandpacks generally should be coarser than the surrounding media If the borehole screen is located entirely within a DNAPL pool and water is pumped from the borehole If the top of the DNAPL pool is under drainage conditions Measurement/observation of DNAPL in borehole DNAPL thickness > pool thickness Exceed by the length of the borehole below the barrier layer surface Measured DNAPL thickness < pool thickness By the distance separating the borehole bottom from the capillary barrier layer upon which DNAPL pools DNAPL elevation and thickness in the borehole are likely to be erroneous The height of the DNAPL column at the borehole bottom will tend to equal or less than the critical DNAPL height required to overcome the capillary resistance offered by the sandpack and/or of the less permeable formation Small quantities of DNAPL may elude detection by sinking down the sandpack and accumulating below the base of the borehole screen Small quantities of DNAPL may elude detection by sinking out of the base of the screen and into the underlying sandpack Screen or sandpack openings that are too small may act as a capillary barrier to DNAPL flow; therefore, DNAPL will not be able to enter the borehole DNAPL will upcone in the borehole to maintain hydrostatic equilibrium causing the DNAPL thickness in the borehole to exceed that in the formation The elevation of DNAPL in a borehole may exceed that in the adjacent formation by a length equivalent to the DNAPL-water capillary fringe height 3.2 Drilling Boreholes are installed to evaluate subsurface stratigraphic, hydrogeologic and contaminant conditions. The selection of drilling locations, depths and methods must be based on available information regarding site conditions and the initial site conceptual model. The potential for causing DNAPL migration by drilling through a barrier layer should be considered before and during drilling and hence minimised. Conventional drilling methods have a high potential for promoting downward DNAPL migration. For example, DNAPL trapped in structural or stratigraphic lows may be mobilised by site characterisation activities; for example, drilling through a DNAPL pool (Cohen and Mercer, 1993). The type of drilling used must cause as little possible disturbance to the aquifer, producing minimal modification to the qualitative characteristics of the subsurface water. page 16

23 Several guidelines for drilling at DNAPL sites are available (e.g. Milan (2000); Pankow and Cherry (1996); Aller et al. (1989); Millison et al. (1989); and Cohen and Mercer (1993). The following is a summary from these references: 1. Avoid unnecessary drilling within the DNAPL zone with an outside-in approach being advisable. The first boreholes are installed on the perimeter of the site (in less contaminated areas or uncontaminated areas). The initial boreholes may be drilled up gradient of the DNAPL source (and/or dissolved-phase plume) and through any possible confining layer to characterise the geology of the site. The appropriateness of this approach must be evaluated on a site-specific basis. 2. Utilise knowledge of site stratigraphy and chemical distribution and carefully examine subsurface materials brought to the surface as drilling proceeds, to avoid drilling through a confining layer beneath DNAPL. In some situations, it may be necessary to drill through actual or possible confining layers at a site. Special precautions should be taken when investigators believe they may encounter a confining layer during drilling. Moreover, if field personnel suspect they may have encountered a possible confining layer while drilling a borehole and an approved plan for drilling through confining layers does not exist, drilling should be stopped immediately and the borehole should be plugged with a bentonite and cement seal above the suspected confining layer and the borehole decommissioned. 3. Minimise the time and the length of the hole, which is open at any time. Any borehole that is not completed as a monitoring borehole should be properly decommissioned. 4. A telescoped casing drilling technique is advised to isolate shallow contaminated zones from deeper zones. Such techniques typically involve drilling an initial borehole partially into the possible confining layer, installing (grouting in) an exterior casing, emplacing grout in the cased portion of the borehole and after flushing the interior of the casing, drilling a smaller diameter hole through the cased off/grouted portion of the borehole (i.e. telescoping casing) through the confining layer. The appropriateness and actual design of telescoping borings and casings should be determined on a sitespecific basis. Telescoping boreholes may be completed as boreholes or piezometers. 5. Select optimum construction materials and grouting methods based on consideration of site-specific chemical compatibility. 6. The drilling method should allow for the collection of representative samples of rock, unconsolidated materials and soil. 7. The drilling and sampling method should allow the geologist to determine where an appropriate location for the screened interval exists. 8. If drilling in a suspected DNAPL pool, consider the use of dense drilling mud to prevent DNAPL from sinking down the borehole during drilling. This should be done at the final stages of the monitoring programme. 9. Drilling fluids used for the installation of one casing should generally not be re-used for the installation of another casing, particularly where contamination is suspected or present. However, drilling fluids (including air), should be used only when minimal impact to the surrounding formation and groundwater can be ensured. 10. The drilling techniques using the auger bit cause a lower risk of cross contamination between the various levels of the aquifer and do not require the introduction of drilling fluids. However, it is not able to drill very durable rock formations or loose ones with large fragments and could influence the local permeability, thus significantly reducing the efficacy of the operation. These situations will often arise in South African aquifers. 11. To prevent surface contamination, sheeting can be placed on the surface to collect any contaminated sludge at the top of the borehole. After drilling has been completed, the sheeting together with the contaminated sludge should be properly disposed of. It is important to have a qualified geohydrologist on site to record all relevant information during the drilling process. Information that needs to be recorded includes the following: page 17

24 Geological logs (Even subtle changes in the lithologies need to be recorded. Permeability changes can greatly affect the migration/penetration of DNAPLs) Soil and rock borings/chips need to be inspected for visible NAPL presence (enhanced techniques such as Sudan IV dye or UV fluorescence can be applied) Type and volumes of drill fluids used Water intersections (depth and quantity) Construction information (depth of hole and casing, borehole diameter, method drilled, date drilled) Use of water, if not solely for monitoring; frequency of abstraction; abstraction rate and whether other water sources are readily available Water quality 3.3 Drilling techniques From the available drilling techniques, the following were selected as suitable techniques for installation of boreholes on DNAPL contaminated sites in South Africa Air Percussion drilling Percussion drilling methods operate by pulverising material at the bottom of a hole by dropping or pounding a bit. This type of drilling involves crushing by impact of the teeth of the drill bit. Most percussion drills are actuated by compressed air (pneumatic percussion). Percussion drills are best suited for drilling brittle, moderately soft to hard rock. In hard rock with percussion drills, you can drill faster and more economically than with other drilling methods. The air blows the cuttings from under the bit and up the annulus to the surface. The air's ability to carry the rock chips depends primarily on high air velocity. If drill penetration rates are high, you may need a higher up-hole velocity to clean the hole. Advantages of air percussion drilling include the speed of drilling through most semi- to well-consolidated formations and no water is required. The use of drilling fluids can be minimised when required. The technique is relatively inexpensive but the equipment is expensive and has high maintenance costs Auger Drilling Hollow-stem augers are widely used in drilling and sampling at contaminated sites. The auger stem is hollow and can be used to sample the formation, for geophysical logging and for borehole completion procedures while the formation is stabilised by the outside of the auger flights. Drilling with hollow-stem augers is similar to solid-flight augers. The augers are rotated into the ground using a rotary drive head on the drill rig. Sampling may be achieved at any depth if a pilot bit is used inside the hollow stem. The pilot bit and rods are withdrawn and a sampling device inserted. This may be either a split-spoon sampler driven into the formation beneath the bottom of the augers to obtain a disturbed core or a thin wall tube pushed beyond the base of the augers to collect an undisturbed sample. When drilling into an aquifer that is under even low to moderate confining pressure or drilling far beneath the water table, the sand and gravel will frequently heave upward into the hollow stem. If a centre plug is used during drilling, then the flow of the aquifer material frequently occurs as the rods are pulled back. When this problem occurs, the bottom of the augers must be cleaned out before further sampling or progress can occur. Unfortunately, the removal of material from the base will normally induce further material to flow up page 18

25 into the augers, thus compounding the problem. Sampling of running sands (or heaving sands) is very complicated. The advantages of auger drilling include: minimal damage to aquifer, no drilling fluids are required; auger flight acts as temporary casing, stabilising hole for borehole construction; a good technique for unconsolidated deposits; collection of relative undisturbed in situ soil samples at greater depth than is possible with test pits; and a continuous core can be collected with the hollow-stem method. The disadvantages of auger drilling include: sand and clay may form a slurry which can cake the outside of the formation, (this can be difficult to remove later and may prevent contamination from being detected in lower hydraulic conductivity layers); contaminants may be smeared down the formation, which leads to bias and inaccuracy in later sampling; cannot be used in consolidated deposits; sampling of running sands (or heaving sands) is very complicated; limited to boreholes less than approximately 50 m in depth; the heat generated by the drilling process may liberate volatiles and if boulders are encountered Core drilling Diamond core drilling is the most versatile of all the methods and it is designed specifically for the resource exploration industry. In core drilling the sample is cut by a diamond armoured drill bit, stored in the inner barrel of a drill pipe, after which the pipe is brought to the surface and the core removed. In wire line diamond drilling, the most widely used and time effective method, the inner tube is hoisted to the surface through the drill rods without the need to remove them from the hole. Advantages of this technique include: its versatility allows it to be done at most surface and underground locations; it is the only method available that provides a core, a complete record of geological structure and rock texture; and it is the only commonly used method that provides reliable samples for accurate geochemical testing. The disadvantages are however, that the technique is the most expensive of all the drilling techniques and the drill rigs are often not available for contamination investigations. Some types of broken and abrasive rock are nearly impossible to core at a reasonable cost (these lithologies literally fall apart during transport up the drill string), but special methods of recovery for these soft lithologies have been designed using tubes and protective sheaths, but the recovery is still generally poor. 3.4 Compatibility of construction materials The compatibility of borehole construction materials with NAPLs and highly contaminated groundwater should also be evaluated during borehole design. Borehole construction materials (e.g. screens, casings, sealants) are subject to degradation or corrosion in the natural environment. Materials exposed to NAPLs may also be degraded or corroded, which may lead to structural failure. This vulnerability applies to materials exposed to these chemicals in both the subsurface and above ground. The EPA (1995), recommends that at sites where the presence of NAPLs is suspected, a materials compatibility review should be conducted. Since the time requirements for either monitoring (or subsurface remediation systems) at contaminated sites are usually longterm, it is economically and technically important that these systems be constructed of materials with known chemical resistance qualities to provide reliable service over many years. There are two types of effects that NAPLs have on materials used in borehole construction (sampling and remediation): 1. The structural integrity of a material may be compromised by corrosion or solvation. page 19

26 2. Dissolved groundwater contaminants from NAPLs can sorb to or leach from monitoring materials, which affect groundwater quality measurements. Another way of viewing these two effects is from a concentration perspective. Sorption to monitoring surfaces may have the greatest effect on water quality measurements when contaminants are present at low dissolved concentrations. Conversely, sorption of contaminants present as NAPLs or in high dissolved concentrations may have a minimal effect on water quality measurements, while the effects on the structural integrity of the materials may be at a maximum. McCaulou et al. (1995) published a Chemical Compatibility Table. The compatibility in this paper is defined as a material s ability to withstand corrosion or degradation under specific experimental conditions. This refers to the effects that NAPLs and high concentrations of dissolved organic compounds have on the structural integrity of materials. During this project, borehole construction materials commonly used (in South Africa) were tested with various DNAPL contaminants and concentrations of the contaminants to saturation. Bentonite was tested as a sealant of a borehole bottom. It was found that under the correct application (the bentonite must be submersed in water); the integrity of the bentonite is not affected by a NAPL such as TCE. However, if the bentonite is not correctly applied, cracks may develop which cause preferential pathways for the NAPL to migrate downwards. The casing material tested, which includes, steel, PVC and HDPE, was inspected visually (and under UV light) and evaluated after exposure to the various NAPL contaminants of various concentrations, after 18 months. The free phase NAPLs sorbed strongly to the PVC material. Some sorbtion was visible of the free phase NAPL to the steel casing material but could easily be removed by shaking the containers of the experiments. Corrosion of the steel casing was hastened by higher concentrations and the free phase NAPL. The PVC material exposed to the free phase NAPL showed softening and swelling where the NAPL sorbed to the material. The HDPE was the only material on which no visible sorbtion and/or corrosion was found. From these results, it is evident that care should be taken to select the correct construction material to ensure long-term stability and to prevent water quality errors during sampling. It is recommended that selection of construction materials be done on a site-specific basis. Once the contaminants of concern are known, the Chemical Compatibility Table of McCaulou et al. (1995) can be used to select suitable materials. A copy of this table is in Appendix B. 3.5 Borehole construction Monitoring boreholes must be of a diameter that will allow easy access to the aquifer, for the purpose of water sampling and for lowering other test instruments. Ideally, diameters between mm are recommended. If casing, screens and/or filters are installed, the diameter of these must allow easy access for monitoring purposes and should in no way block the flow of water through the borehole. The top of a casing in a monitoring borehole should rise beat least cm above the general ground surface to ensure that surface run-off does not flow into the borehole during flood conditions. A concrete block around the top of the casing, to protect the casing and borehole, as well as to prevent surface pollution from flowing down the side of the casing, is essential. Required minimum dimensions for the concrete block are 750 mm x 750 mm x 150 mm. A lockable cap is also recommended to prevent tampering or vandalism to the borehole. page 20

27 In the installation of the borehole, measures to limit short-circuiting of contamination from forming different layers must be implemented. Use of grouts such as bentonite in backfilled or gravel-packed portions of the annulus is recommended and care must be taken when casings are installed that vertical flow alongside the casing is prevented. Piezometer installations are not recommended at DNAPL contaminated sites, due to the low concentration levels measured and the risks of contaminating deeper lying units. Where this is the method of choice, appropriate precautions to prevent cross-aquifer interference must be taken (Figure 3). Where more than one aquifer system is present (Figure 4), it is recommended that boreholes be constructed in a such a way that only one aquifer system is measured at a time, e.g. a shallow borehole sealed at the confining layer and a deeper borehole which should be cased off from the top with solid casing (Figure 5). It is important that specific attention be given to the precise depths of screening or slotted casing to obtain the most representative indication of the water quality and that a sump be installed in boreholes where DNAPLs are suspected. The end of each borehole should be plugged with cement and bentonite plugs to prevent downward migration of any DNAPL. Where gravel/sand packs are used, these should be coarser than the surrounding aquifer media and specific care must be taken that these are sealed off at interfaces to prevent cross-unit contamination. Lockable Casing cap Concrete SOIL Concrete Solid casing SANDSTONE Fracture Zone Gravel pack Slotted casing or screen Piezometer SHALE Solid casing Piezometer SANDSTONE Cement plug Bentonite Cement plug Gravel pack Slotted casing or screen Cement plug Bentonite plug Cement plug Multiple piezometers Figure 3: Multiple piezometer installation page 21

28 Soil DNAPL A B C Mudstone Shale Fracture Zone Fracture Zone Sandstone Solid Casing Plug Figure 4: Depth specific boreholes in multiple Figure 5: Construction of Borehole A or B/C page 22

29 3.6 Monitoring Locations Generally, each distinct zone of contaminant migration and geochemical regime within the contaminated area should be monitored. If the geochemical properties are important; for instance if considering MNA, then if part of a plume of tetrachloroethene is anaerobic with high levels of electron donors available and another part of the plume is aerobic with few electron donors available, degradation of the tetrachloroethene may be very active in the anaerobic zone but nonexistent in the aerobic zone (Pope et al., 2004). For each zone with distinctly different conditions or controls on contaminant migration and fate, the following locations would be monitored: areas hydraulically up-gradient and adjacent to the plume, source area, main body of the plume and distal portions and boundaries of the plume. Up-gradient and adjacent boreholes are used to determine changes in background water quality. These boreholes would also provide confirmation of the plume geometry and the absence of seasonal changes in groundwater or plume flow direction. Boreholes in the contaminant source area monitor changes in source strength over time. Boreholes located down-gradient of the source area, but within the contaminant plume, monitor plume behaviour and changing concentrations over time. These will normally be located along the centre line of the plume. Boreholes located immediately down-gradient from the contaminant plume should provide early warning of any migration of the contaminant plume. Ideally these boreholes will also provide supporting evidence for NA; for example, evidence of an absence of contaminants but the depletion of electron acceptors, (for instance, decrease in nitrate and dissolved oxygen concentrations as advanced evidence of hydrocarbon pollution). So-called sentinel boreholes located between the plume and any identified receptors should be monitored for compliance to water quality guidelines and/or licence conditions and also to assess if remediation targets are met. Figure 6 and Figure 7 are schematic outlays of a typical monitoring network for the purpose of monitoring NA at a contaminated site. The precise location of boreholes must be determined by the likely pathway for dissolved or NAPL contaminants. An iterative approach to the construction of feasible conceptual models of each specific site is therefore of greatest importance to the success of the monitoring programme. The location of sentinel boreholes down-gradient of the source/contaminant plume will need to be determined based on the rate of groundwater/contaminant movement and the distance to potential receptors. The boreholes should be located at sufficient distance up-gradient of the receptor to provide adequate warning/protection. The distance to down-gradient boreholes will also influence the duration of monitoring. The monitoring strategy will, therefore, need to take account of/demonstrate that the borehole locations and the period of monitoring, are appropriate to the expected behaviour of the plume. The use of fate and transport models is therefore recommended for determining borehole locations. page 23

30 Figure 6: Plan view of schematic location of monitoring boreholes around a plume page 24 (From Carey et al., 2000)

31 Figure 7: Cross-section of schematic location of monitoring boreholes around a plume page 25 (From Carey et al., 2000)