Assessment of Aquifer Vulnerability in South Africa
Report No: 1432/1/07
Dec 2007
Executive summary

This report presents the main findings of the WRC-funded four-year project on “Improved methods for vulnerability assessments and protocols for producing vulnerability maps, taking into account information on soils”, abbreviated as AVAP. The project fulfils one of the key recommendations of a strategy on groundwater protection research which was prepared by Sililo et al. (2001). For detailed information, the reader is referred to the enclosed CD which includes all deliverables of the project (Appendixes 1 to 25).

Aquifer vulnerability to contamination comprises two components: unsaturated zone vulnerability and saturated zone vulnerability. For the unsaturated zone, AVAP defined vulnerability as the ease with which groundwater (at the water table) may become contaminated by a contaminant source at the surface or in the unsaturated zone. For the saturated zone, vulnerability is defined as a function of the period of time after contaminating activities have ceased that a given contaminant can be detected in groundwater plus the volume of the aquifer throughout which the contaminant is above a preset concentration. The decision-maker could focus on either one of these vulnerabilities or in combination depending on the type of decision to be made. AVAP developed new approaches for assessing aquifer vulnerability.

Unsaturated zone vulnerability
For the unsaturated zone, the “AQUISOIL” (Aquifer Vulnerability Soil Assessment) approach and a modified DRASTIC approach, called “EUZIT” (Excel-based Unsaturated Index Tool) were developed. Both approaches can be used for assessing vulnerability at the water table.

AQUISOIL focuses on the soil zone as the “first line of defence” but can also be applied to the whole unsaturated zone down to the water table. AQUISOIL comprises three types of vulnerabilities which are rated for each single layer and weighted according to thickness; i.e. chemical, hydraulic and climatic vulnerability. With regards to chemical vulnerability, a new soil classification was developed which relates sorption to selected soil properties (pH, clay and organic carbon content, CBD extractable Fe and Al, and the sum of exchangeable cations for more than 170 soil samples representing major kinds of diagnostic horizons and materials in the South African soil classification), by means of so-called chemical envelope equations. These envelopes predict the upper limit of contaminant sorption to be expected at a specified value of each soil property and are based on the application of quantile regression (also referred to as stochastic frontier analysis) which provides a systematic method of determining ranges of a determinant variable over which sorption can be either potentially maximal or inevitably minimal. Chemical vulnerability classes have been defined accordingly for the retention of the following categories of contaminants: cationic (metals), anionic and non-polar organic. Hydraulic vulnerability is derived from a permeability index which is based on soil texture and the presence or absence of a cemented layer or horizon. The climatic vulnerability uses a leaching or recharge index calculated with one or other of the classical algorithms that take into account factors such as rainfall, surface runoff, profile storage, etc. AQUISOIL has been incorporated in a spreadsheet programme.

EUZIT is a spreadsheet-based modified DRASTIC (Aller et al., 1987) approach. The rating of the unsaturated zone is based on a combination of factors that contribute to the likelihood of contaminants reaching the saturated zone following the path of aquifer recharge. The factors considered are the unsaturated zone thickness, hydraulic properties (vertical hydraulic conductivity), flow mechanism (preferential and matrix), travel time, recharge (based on well established methods), slope, contaminant sorption and decay (degradation). Any of these factors can be disabled depending on site-specific conditions and the weightings for each factor are user defined. EUZIT also allows for a multi-layer unsaturated zone. The rating scores for each factor are combined in the main menu to yield the rating of the impact of vadose zone on aquifer or groundwater vulnerability. The EUZIT programme draws the rating for sorption from the chemical vulnerability part of the AQUISOIL spreadsheet programme. The vulnerability of an aquifer to contamination originating at the soil surface can be assessed using data that are readily available, that can be calculated or that can be estimated using the information provided in the database of EUZIT.

Saturated zone vulnerability
For the saturated zone, two approaches are proposed to determine its vulnerability to contamination: one approach that uses (numerical) geochemical and reactive transport models and a second, generic, approach that generalise contaminant transport by an analytical equation. AVAP concentrated on the first approach and evaluated for two common aquifer types in South Africa: coastal primary and dual porosity aquifers the aquifer vulnerability to inorganic contamination from landfill leachate, industrial effluent and acid mine drainage, assuming identical physical properties and contamination at the water table. PHREEQC-2 (Parkhurst and Appelo, 1999) was used for the calculations. Vulnerability of the saturated zone to contamination is expressed in terms of spatial impact and persistence of the contaminant. Realistic spatial and temporal scales are derived from the linking of PHREEQC-2 with a single or multi-species groundwater flow/transport model such as MT3DMS (Zheng and Wang, 1999). For the modelling of organic contamination, PHT3D (Prommer, 2003), BIOPLUME III and the LNAPL Guide (American Petroleum Institute, 2004) can be used for simulating reactive transport of BTEX (considered representative of petroleum hydrocarbons), and the BIOCHLOR code (Aziz et al., 2000) for chlorinated ethenes.

GIS-based vulnerability assessment
A Geographical Information System (GIS) enables the compilation of vulnerability maps combined with other infrastructural information so that the results can be easily related to by planners and decision makers. Two GIS approaches were developed to determine unsaturated zone vulnerability as part of the AVAP project: EUZIT incorporated in a GIS and the modified UGIf model.

EUZIT was integrated in a GIS to assess the variation in groundwater vulnerability across an area. By building the input datasets into a GIS and converting them to grids the EUZIT algorithms can be applied in GIS using the “map calculator” functions. Within the GIS it is also possible to carry out a sensitivity analysis to optimise the weightings (and ratings) per factor. Depending on the scale of the application and the degree of spatial variability, the possibility does exist, however, that the initial spatial variability is “lost” due to the relatively broad ranges within the rating classes. One could refine the rating classes to enhance spatial variability in vulnerability. Further work is thus recommended on refining the ratings based on the scale of the study and data availability and on scientifically motivating the weightings.

The UGIf model, which estimates recharge fluxes of organic pollutants (BTEX) in an urban environment (Thomas et al., 2001), was modified to include screening level models for vulnerability assessment such as the Attenuation Factor Model (Rao et al., 1985), the Leaching Potential Index Model (Meaks and Dean, 1990), the Ranking Index Model (Britt et al., 1992), and a simple approach to assessing intrinsic vulnerability of conservative contaminants. The three screening level algorithms of UGIf require a combined grid containing attributes of average recharge rate (m/day), soil moisture or volumetric water content, vadose zone depths (m), and the retardation factor values. The UGIf model was made suitable for South African conditions through the incorporation of representative land types. The software packages required to run UGIf model are ArcView GIS 3.x version (ver 3.1 or 3.2 or 3.3) and its extension Spatial Analyst.

Both EUZIT and UGIf deal exclusively with the vulnerability of the unsaturated zone. Both approaches are encouraged, as they present decision-makers with a spatial representation of unsaturated zone vulnerability.

Case studies
Two study sites were selected to illustrate the use of AVAP’s approaches to the assessment of groundwater vulnerability: the Goedehoop irrigation site near Secunda, and the Coastal Park waste disposal site near Cape Town. It was found that the unsaturated zone vulnerabilities are relatively high for both aquifers. From the smaller travel times and higher leaching potential indices as derived from the AQUISOIL and EUZIT spreadsheet tools and the UGIf model it can be concluded that the unsaturated zone vulnerability to dissolved organic contaminants (e.g. BTEX) of the primary Cape Flats aquifer at Coastal Park is higher than the vulnerability of the dual porosity, weathered zone aquifer at the Goedehoop site. This means that contaminants generated at the surface are likely to reach the water table and pollute the aquifer sooner at the Coastal Park waste disposal site. The saturated zone vulnerability in terms of the spatial impact of inorganic and organic contamination is expected to be higher at the Coastal Park waste disposal site whereas the persistence is expected to be higher at the Goedehoop site. More detailed modelling, accounting for site specific physical properties, of the Cape Flats aquifer at the Coastal Park waste disposal site is needed to enable a better comparison of the saturated zone vulnerabilities of both aquifers.

Aquifer vulnerability and decision-making
Aquifer vulnerability assessments form an important input to managing the risk of water resource degradation. A framework to support decision-making was developed to assist groundwater vulnerability assessment practitioners in understanding the role of their assessments in groundwater management and to assist them in the selection of AVAP’s approaches to groundwater vulnerability assessment. The framework highlights the fact that groundwater vulnerability assessments serve as input to contaminant risk assessments, which will in its part, contribute to a cost benefit analysis. It is the outcome of the cost benefit analysis which will ultimately inform decision-making. To guide the groundwater vulnerability assessor, a table is presented which briefly summarises AVAP’s assessment approaches in terms of their limitations, applicability, accuracy, ease of use, and cost.

Copies of this report are available from the Foundation price £25.00 less 20% for FWR members.