wpeD.jpg (1891 bytes)

FRESHWATER

Home Freshwater Resources Conference Programme
 

Protecting Groundwater Resources using

Geographic Information Systems -

Case study: Regional Waste Site Identification, Western Cape

by Julian Conrad, CSIR, South Africa


Abstract

As the pressure on our scarce water and land resources grows, increasing sophistication in water resource protection and management becomes necessary. Although South Africa is not as fortunate as other countries with regard to regional groundwater resources, it does have local resources, that if properly managed on a sustainable basis, could play a major role in meeting the country’s water requirements. Thus, in this arid country, which receives well below the world average rainfall, protection of our groundwater resources is a crucial concern. There are many and varied activities that can cause pollution of groundwater resources. One such activity is the disposal of waste, be it general domestic or hazardous waste.

Due to the issues mentioned above, South Africa’s railway company, Spoornet, approached the CSIR to identify, via a desk top study, broad areas in the Western Cape Province which are potentially suitable for the development of a regional waste disposal site. Spoornet requires this information as part of a pre-feasibility study aimed at assessing possible options and strategies regarding the establishment of a "waste-by-rail" scheme in the Western Cape Province. The data processing was carried out using a Geographical Information System (GIS).

The approach used in this project was firstly to exclude all areas where the development of a waste disposal site would not be permitted. Examples of such exclusionary criteria are close proximity to residential areas, airfields, mountainous areas, nature reserves, indigenous forests, geological faults, the coast, dams or rivers. Once these areas had been identified, the remaining areas were then rated according to the geology, depth to groundwater, soil texture and soil depth. From the combination of these factors, favourable areas were identified. There are a number of factors that could not be included in the study, such as agricultural potential, land use, location of archaeological/historically important sites and areas with mineral rights. In addition, economic, social and political factors were not taken into account in this project.

This pre-feasibility study provides a clear indication of areas potentially suitable for the development of a regional waste disposal site for the Western Cape Province. The result obtained is subject to various assumptions, and area suitability may change when a different set of assumptions are made. Should this become necessary the GIS overlay technique used is flexible enough to allow rapid re-evaluation of area suitability.

1- INTRODUCTION

Waste disposal issues in South Africa.

Waste disposal is an issue that needs to be seriously addressed in this country. There are many sites in SA that are not operated legally and do not meet the minimum requirements as set out by DWA&F. Many sites are nearing the end of their acceptable life span.

The combination of high population growth, urbanisation, the limited number of permitted sites, the pressure on our environment, especially water resources, points to waste disposal in SA being of critical importance. In addition the establishment of new sites, that meet all the minimum requirements is a very lengthy process, and entails great cost, thus has to be based on well chosen sites with a high level of public involvement, and an adoption of a win-win strategy for overcoming the NIMBY attitude. All these factors, point to a crisis situation that urgently needs to be addressed.

Overseas, the concept of moving waste significant distances to an appropriate and accepted regional site is well established. In many cases transportation by rail has advantages over other means of transportation. Transportation distances of waste by rail of 200 to 300 km’s are not uncommon internationally. In one instance waste is transported over a distance of 1300 miles to a regional landfill (Dunckley, 1996).

South Africa has an extensive rail network, with a large percentage of the towns and cities in South Africa having rail facilities. Spoornet is committed to applying and expanding the Waste-by-Rail initiative. One Waste-by-Rail operation is already in place: waste from Cape Town is collected at a transfer station, sorted, compacted into wire-bound bales and transported by rail to a regional landfill 25 kms away. This project became operational in August 1995 and has the capacity to process more than 1000 tons of waste per day.

Waste Disposal in the Western Cape

Some 1 531 000 T/a of waste is generated in the Cape Metropole alone, of which 73% is disposed to 15 operating waste disposal sites. However, 50% of these sites are expected to be closed by the year 2001 and the rate of increase of waste generation is in the order of 10 % p.a. (Parsons, 1997). Recycling and composting play and effective and important role in reducing the volume of waste to be disposed.

Only 26 permitted sites are currently operated in the province. An additional 20 sites are due to be closed and a further 9 sites have been identified as problem sites. Feasibility assessments of 40 sites are being carried out (pers.comm. L.Eichstradt, DWA&F).

GIS approach

The CSIR was appointed by Spoornet in April 1996 to proceed with a "desk top" GIS based study to identify areas potentially suitable for developing a macro waste disposal site for the Western Cape Province.

The choice in the best methodology to use in such a site selection project is a complex matter. There are many references in the literature regarding different approaches for suitability analyses. The methodology used in this project was developed by a group of CSIR specialists in the fields of hydrogeochemistry, environmental, mathematical and spatial modelling. There are basically two components to the methodology used:

  1. all the exclusion criteria for a waste disposal site were considered, and
  2. the remaining non-exclusionary areas were ranked in terms of suitability.

2 - LANDFILL SITING PROCEDURES

Siting a landfill requires a substantial evaluation process in order to identify the most suitable location, that is, a location which meets the requirements of government regulations and minimises economic, environmental, health and social costs. Evaluation processes or methodologies are structured to make the most use of available information and to ensure that the results obtained are reproducible so that outcomes can be validated and defended.

Hopkins (1977) extensively evaluates and compares methods for generating land suitability maps. He defines eight methods; gestalt, ordinal combination, linear combination, non-linear combination, factor combination, cluster analysis, rules of combination, and hierarchical combination. Due to the complementary characteristics of several of the methods, application of more than one method is recommended in carrying out land suitability analysis.

DRASTIC (Noble 1992) and LeGrand methods (Canter et al. 1987) are examples of site evaluation procedures that focus on a single domain. They are used to evaluate groundwater pollution potential from a proposed landfill site. Examples of more general procedures include interaction matrices (Camp Dresser and McKee, Inc. 1984) and the weighted rankings method (Morrison 1974), which are impact assessment techniques used to evaluate the various impacts of proposed landfill sites. These procedures result in an impact rating which is interpreted as the relative suitability of each potential landfill site (Siddiqui et al. 1996).

Several methods of site selection have become available which make use of digitised data. Examples of these methods are the methods of intrinsic suitability used by the Minnesota Pollution Control Authority (Noble 1992) and George Noble’s method (Noble 1992). However, these methods use GIS capabilities only for screening out unsuitable areas. There are factors that still need to be considered that are not exclusion factors. These factors can best be evaluated by means of a ‘rating and weighting’ procedure.

Once potential areas have been identified, they can be evaluated and classified according to the analytic hierarchy process (AHP) or a similar process. Since its introduction in the late 1970s, AHP has been applied in a wide variety of practical settings to model complex decision problems. Its ability to rank and assess decision alternatives quantitatively has led to applications in many diverse areas. AHP is becoming popular in decision making studies where competing objectives are involved. Although the objective of this project was to highlight potentially suitable areas, such a ranking of those areas was not carried out. It may be worthwhile considering the AHP, or a similar process, to rank the available suitable areas, although due to data limitations, the method should only be used for preliminary site ranking. Examples of data pertaining to social attitudes and political issues are often unavailable in a form suitable for GIS.

3 - METHODOLOGY

The optimum methodology for this project was derived by CSIR specialists using information from literature reviewed (Gebhardt and Jankowski 1986, Al-Bakri 1988, Andreottola 1989, Frantzis 1993, Ball 1994, Minor and Jacobs 1994, Siddiqui 1996, plus others), and experience in other site selection projects. The criteria required for the project were determined and available data sets were acquired.

Following on from determining the relevant data sets, it was decided that two components would form the basis of the methodology. The first would be to exclude all the areas that would almost certainly not be acceptable for a waste disposal site and the second would be to rank and weight the non-exclusionary areas. In addition a few of the factors that will be important in the next phase, i.e. the ground-truthing and more detailed investigation phase of waste site location, have been identified.

3.1 - EXCLUSIONARY CRITERIA

These criteria and the relevant exclusion distances were obtained, as far as was possible, from the guidelines of the Department of Water Affairs and Forestry’s Waste Management Series, ‘Minimum requirements for waste disposal by landfill’ (1994). Where explicit information was not available from DWA&F’s document, the information was derived from overseas guidelines, as well as local expert opinion. Some of the data sets used included: geology, geological structure, soil type, topography, rivers, lakes, dams, coastal areas, roads, railways, airfields, power lines, forests and conservation parks.

Relevant exclusionary criteria that were not considered at this stage of the analysis include:

  • the 50 year flood level of rivers
  • agricultural potential and land use
  • location of archaeological/historically important sites
  • areas with mineral rights
  • servitudes
  • atmospheric factors (e.g. areas with very high wind velocities)
  • visibility impacts (e.g. scenic areas that are visible from major roads and railways)

These data could not be included because they are either not available in a usable digital format or not of sufficient accuracy or not relevant for this pre-feasibility study. These criteria will, therefore, have to be considered in a further phase once closer identification of possible sites is reached.

3.2 - NON-EXCLUSIONARY CRITERIA

The second component of the analysis considered the remaining areas that had not been excluded. These areas needed to be rated according to suitability. The four criteria for which data could be obtained and that were seen as important were; geology (lithology), depth to groundwater, soil texture and soil depth. Each of these categories were rated and then a weighted equation developed for the final result. The DRASTIC (1987) approach was used as a guideline for determine rating and weighting values. The final results from the equation were then divided equally into five classes of suitability. The numerical approach taken in the rating was that the lower the score the more suitable the area was for a waste disposal site.

• Geology

The geology category refers to the consolidated or unconsolidated rock which can serve as an aquifer. An aquifer is defined as a subsurface rock unit which will yield sufficient quantities of water for use. The flow system within the aquifer is affected by the aquifer medium. The route and path length which a contaminant must follow are governed by the flow system within the aquifer. The path length is an important control in determining the time available for attenuation processes. The aquifer medium also influences the amount of effective surface area of materials with which the contaminant may come into contact within the aquifer. The route which a contaminant will take can be strongly influenced by fracturing or by an interconnected series of solution openings which may provide pathways for easier flow. In general, the larger the grain size and the more the fractures or openings within the aquifer, the higher the permeability and the lower the attenuation capacity of the aquifer media.

The rating carried out for this project is in order of increasing pollution potential. An example of two rocks types at the extremes of the rating scale would be:

    • massive shale, which only yields small quantities of water, has high potential attenuation capacity, and thus low pollution potential rating.
    • limestone (karstic), typically large, open, interconnected cavities and fractures occur, allowing the rapid movement of groundwater and thus a high pollution potential rating.

The ratings for geology have been carried out using values from DRASTIC (1987) approach.

Depth to groundwater

Depth to groundwater is important, primarily because it determines the depth of unsaturated material through which a contaminant must travel before reaching the aquifer, and it may help to determine the contact time with the surrounding media. The depth to water is also important because it provides the maximum opportunity for oxidation by atmospheric oxygen. In general, there is a greater chance for attenuation to occur as the depth to water increases because deeper water levels imply longer travel times. The presence of low permeability layers which confine aquifers will also limit the travel of contaminants into an aquifer. The depth to groundwater data set was obtained from the National Groundwater Maps. The depth to groundwater was divided into 3 classes and rated, so that the deeper the water level, the more favourable the location.

Groundwater occurs in either unconfined, confined or semi-confined conditions. In an unconfined aquifer, the water table represents the uppermost elevation where the openings in the soil or rock material are filled with water. Where present, unconfined aquifers are the uppermost aquifer near the ground surface, and as a result, these aquifers commonly are susceptible to groundwater pollution. Confined aquifers have more natural protection from contaminants infiltrating from the ground surface and are less vulnerable to pollution. Semi-confined aquifers exhibit characteristics ranging from unconfined to confined aquifers.

For the purposes of this study, a differentiation of the different aquifer types was not made, as it was felt that the depth to water level was sufficiently significant, however in a follow up phase the aquifer type would be important.

• Soil Texture

The soil refers to the uppermost portion of the unsaturated (vadose) zone characterised by significant biological activity. Soil is commonly considered the upper weathered zone of the earth which averages a depth of 2 metres or less from ground surface. Soil has a significant impact on the amount of recharge which can infiltrate into the ground and hence on the ability of a contaminant to move vertically into the vadose zone. The presence of fine-textured materials such as silts and clays can reduce relative soil permeabilities, increases attenuation potential and restricts contaminant migration.

The ratings for soil texture were classified according to a soil textural classification chart. Essentially the higher the clay content within the soil the more favourable the soil, in terms of this project.

• Soil Depth

There are two considerations regarding soil depth or thickness. The first is that the thicker the soil, the greater the potential for attenuation processes such as filtration, biodegradation, sorption and volatilisation. Thus if a soil is thin, it is considered ineffective for contaminant attenuation and the groundwater pollution potential is very high. For this study soils are considered thin or absent if they are less than 350 mm thick. The second consideration is that soil plays an important role in waste site design/management, with the more soil available, the more favourable the site. Fortunately the two components can be rated similarly. However, in terms of this project the soil depth rating was based essentially on the first consideration. The soils have been given a high rating because it is assumed that the soils will be mostly removed prior to laying down the waste.

• Weighting equation

The equation used for obtaining the final scores of non-excluded areas is as follows:

Score = 5 ( G * GW ) + ( ST * SD )

where:

G = geology

GW = depth to groundwater

ST = soil texture

SD = soil depth.

The scores resulting from this weighting and rating were then divided linearly into five categories, with the lowest score representing the most suitable area and the highest score the least suitable.

• Assumptions

The following assumptions were made in determining the above equation:

    • Leachate and gas management will be implemented;
    • Soil possibly removed during construction;
    • The waste site has a 50-100 year life span;
    • The site has an areal extent of approximately 4 km2.

4 RESULTS

The results obtained clearly show the exclusion zones. The 10 km buffer around cities and towns are clearly evident. Where possible the actual shape of the town was used in the buffering. However, this factor needs further work to determine the growth potential for the city/town in the next 50 to 100 years. It may well be more or less than the arbitrary 10 km chosen in this analysis. Landfill odour, dust and other factors will ultimately determine the acceptable distance from any residential area.

Not all relevant data were available for this project. A number of categories were identified, for which the data has never been captured digitally or on maps. Examples of data sets not used in this project are; historic sites, 1 in 50 year flood lines, contributing areas to surface dams, sole source aquifers and agricultural potential.

5 CONCLUSIONS

This pre-feasibility study provides a clear representation of areas potentially suitable for the development of a regional waste disposal site for the Western Cape Province. The result obtained is subject to various assumptions and area suitability may change when a different set of assumptions are made. Should this become necessary, the GIS overlay technique used is flexible enough to allow rapid re-evaluation of area suitability. "What-if" scenarios can quickly be established.

Not all factors impacting on waste disposal site selection could be included in this analysis. Aspects such as the existence of archaeological sites, site accessibility, land potential and cost etc. should be considered in further detailed site selection procedure.

It is critical that use is made of information technologies, such as GIS, for assessing potential environmental impacts. This case studies shows how a computing technology can be used to identify, on a regional scale, potential areas suitable for the disposal of waste. These suitable areas can then be managed on a sustainable basis, and impacts on social and environmental economies minimised. This technology also provides an excellent platform for the distribution of information to the public, during their involvement in the participation and decision making process.

6 ACKNOWLEDGMENTS

The permission of Spoornet to present this paper is gratefully acknowledged.

7 REFERENCES

Al-Bakri D., Shublaq W., Kittaneh W., and Al-Sheikh Z., 1988. Site Selection of a Waste Disposal Facility in Kuwait, Arabian Gulf. Waste Management & Research 6, 363 - 377.

Andreottola G., Cossu R., Serra R., 1989. A Method for the Assessment of Environmnetal Impact of Sanitary Landfill. Assessment of Environmental Impact, 367 - 389.

Ball J., 1994. Practical Issues to Consider Prior to Waste Disposal Site Selection and Development. Municipal Engineer, May 1994 21 - 27.

Camp Dresser and McKee, 1984. Cumberland County Landfill siting report. Camp Dresser & McKee Inc., Edison, N.J.

Canter L., Knox R., and Fairchild D., 1987. Groundwater Quality Protection. Lewis Publishers, Chelsea, Mich.

Department of Water Affairs & Forestry, 1994. Minimum requirements for waste disposal by landfill. Waste Management Series Volume 1. Pretoria 1994.

DRASTIC, 1987. A standardized system for evaluating ground water pollution potential using hydrogeologic settings. United States Environmental Protection Agency.

Dunkley W., 1996. Waste by Rail. Wastecon’96.

Frantzis I, 1993. Methodology for Municipal landfill Sites Selection. Water Management & Research 11, 441 - 451.

Gebhardt K., and Jankowski J., 1987. Preliminary Landfill Siting and Related Analysis Using Simple Modelling Techniques. Engineering Geology, 23, 291 - 306.

Hopkins L. D., 1977. Methods for generating land suitability maps: A comparative evaluation. American Institute of Planners Journal, 386 - 400.

Minor D., and Jacobs T., 1994. Optimal Land Allocation for Solid- and Hazardous- Waste Landfill Siting. Journal of Environmental Engineering, Vol 120, No 5.

Morrison T. H., 1974. Sanitary landfill site selection by the weighted rankings method. MSc thesis, Univ. of Oklahoma, Norman, Okla.

Noble G., 1992. Siting landfills and other LULUs. Technomic Publishing Company, Inc. Lancaster, Pa.

Parsons R., 1997. Waste disposal in and around the Cape Metropole. Cape Metropolitan Council.

Siddiqui M. Z., Everett J. W and Baxter E. V., 1996. Landfill Siting using Geographical Information Systems: A demonstration. Journal of Environmental Engineering, June 1996, Vol 122 No. 6, 515 - 523.