Click on the image above for a larger picture.
A Holistic Approach to Water Management of The Cradle for Humankind World Heritage Site.
2005
J. Groenewald(1), M. Adlem(2), M. Buchanan(3) and G. Eloff(4)
(1) Council for Geoscience, Private bag X112, Pretoria South Africa.
(2) Council for Geoscience, Private bag X112, Pretoria South Africa.
(3) P.O.Box 8947, Cinda Park, Boksburg.
(4) Gauteng Department of Agriculture, Conservation & Environment, PO Box 8769, Johannesburg.
Abstract
South Africa is host to a large succession of sedimentary carbonate rocks within the geological column. They range from Swazian (>3090Ma) to quaternary in age and are variably composed from pure calcium carbonate (limestone) to 46 percent magnesium carbonate (dolomite). The carbonate rocks of interest to this project is the impure limestone or dolomite rocks of the Malmani Subgroup (Fig. 1) which forms part of the Transvaal Sequence of sedimentary and volcanic rocks that was deposited in a generally shallow marine environment. This sequence was only moderately affected by tectonism. The Malmani Subgroup is divided into five formations on the basis of its general lithology, however the two formations in which karstification took place most frequently is the Oaktree and Monte Christo formations which forms the 900 m base of the Malmani Subgroup. These formations are playing host in recent times to the proclaimed World Heritage Site (WHS) for the Cradle of Humankind (COH), where the most productive and important palaeoanthropological foundings were made over the last sixty years.
Mining and industrial activities have caused pollution in the surface and subsurface hydraulic regimes over the last thirty years and has raised alarm to the authorities managing the World Heritage Site.
A project initiated by the Gauteng Department of Agriculture, Conservation and Environment (GDACE) is currently investigating the quality of the water interacting with these very important cave systems. Although this has only recently been finalized, work has already started to establish a baseline database on current water quality by the Council for Geoscience. This is only part of a program that will deal with managing the water for future sustainability not only by consumers and tourists in the area but also for the fauna and flora and the valuable cave systems. The following is merely a presentation of findings to date and proposed methodology for a holistic approach.
Introduction
At the request of the local environmental authority for the COH, Gauteng Department of Agriculture, Conservation and Environment (GDACE), a project was launched to compile a water management strategy and program to ensure the sustainability of the WHS. The main objective of this strategy will be to monitor, manage and remediate any damaging influence from cultural activities.
The Cradle for Humankind is an approximate 51 000 Ha (figure 2) area lying to the North West of Mogale City (Krugersdorp), spanning across provincial borders of Gauteng in the south and North West province to the north. The protection of this area is obvious for several reasons including most importantly, amongst others, conservation and tourism. Due to closing of mines to the south and obvious job losses occurring this area could yield several job opportunities in future balancing the equation. The mere abundance of pristine fauna and flora along with billion year old geological formations and fossils, most certainly warrants protection of this area for our future generations.
The land use in and adjacent to the area includes conservation, tourism, agriculture, small industries and mining. The most significant polluters from these are the mines and agriculture with the smaller industries playing a lesser role. Typically pollutants from these activity includes high salt loads (SO4-, Na+, Cl-, PO4-), heavy metals, and possibly some organic contaminants (DNAPL and LNAPL).
These contaminants are significant in some way or the other mostly in a destructive capacity and thus have to be quantified and qualified to minimize any of its associated processes and effects. This can only be achieved as part of a holistic approach. Water monitoring on surface is usually easy as the flow paths are known, groundwater on the other hand is much more difficult as the flow paths and its dynamics can not be seen, conduit establishment age does not mimic the terrestrial flow paths, they can however be measured and conceptualized. It would therefore be vital to firstly understand the hydrologic and geohydrological regimes of the area as it interacts with the Lithology. Secondly to establish a base line data set, thirdly interpreting and understanding (conceptualizing), fourthly establishing a monitoring network and finally to manage and remediate the contaminated system.
History of Sterkfontein Caves
Sterkfontein is one of the world's most productive and important palaeoanthropological sites. It is the place where the very fist adult ape-man was found by Dr Robert Broom in 1936. This ancient cave system has over the years revealed a sequence of sedimentary infill deposits with fossils dating from about 3.5 to 1.5 million years ago, a period of time which spans the early development of the family of man - the hominids. In addition to almost 500 skull, jaw, teeth and skeletal fossils of these early hominids, there are many thousands of other animal fossils, over 300 fragments of fossils wood, and over 9,000 stone tools which include some of the earliest manifestations of human culture on earth. Some of the youngest deposits in the cave also contain fossils and tools from the period just prior to the emergence of modern humans, the period ca. 100,000 to 250,000 years ago.
The environment inside caves is fragile, and the locations of many caves are kept secret in order to protect them. The younger fossil-bearing sediment at Sterkfontein, about 1.5 to 2 million years old, contain stone tools that are the earliest cultural remains yet found in southern Africa. These are attributed to a more advanced species of hominid, such as Homo habilis ('handy man'). A cranium of H. habilis was excavated from the upper levels at Sterkfontein in 1976. Many scientists consider H. habilis to be the ancestor of H. erectus, who was, in turn, the ancestor of H. sapiens, the single species to which all of humankind belongs.H. habilis appears to have co-existed with a second australopithcine species, a more heavily-built, specialised individual called A. Robustus, which was first described from Kromdraai, a few kilometres north of Sterkfontein, by Broom in 1938. A. robustus and early members of the genus Homo have also been found at Swartkrans and Gladysvale. The extraordinary richness of the fossil sites around the Sterkfontein area led to concentrated efforts resulting in the area being declared a Cultural World Heritage Site.
Possible causes of pollution in the area and most important chemical processes.
Sources of contamination
Sources of contamination
Storm water from mines, industrial and domestic areas
Slime dams and dump leachate
Domestic waste leachate
Treated sewerage waste released into the streams
Local septic tanks and informal pit latrines
Mine water AMD
Agricultural run-off and leachate
Site specific waste dumps (metal in caves etc.)
Leaking fuel storage tanks.
Most important chemistry processes
Dolomite dissolution (after Hodgson et al., 2001)
Dolomite dissociation is given as:
CaMg(CO3)2 0 Ca2+ + Mg2+ + 2CO32-
The solubility product for this reaction in pure water is 1.995 x 10. This is very small compared to the reaction in de-ionised water, where the solubility is 14 mg/L, which would increase with the hydration of CO2 . This id due to the reaction:
CaCO3.MgCO3 + 2HCO3 0 Ca2+ + Mg2+ + 4HCO3-
This indicates the importance of CO2 in natural karst development. Other acids also results in lowering of pH which results in charged metal ions in solution, for example the reaction of dolomite with sulphuric acid:
2H2SO4 + CaCO3.MgCO3 + 2HCO3 0 Ca2+ + Mg2+ + 2H2O + 2CO2 + 2SO4
Solubility also increases with ionic strength, i.e. as the salinity of the water increases, so the solubility of dolomite will increase. This is due to solution non-ideality. Thus in natural developing karst the solubility will be mainly controlled by chemical factors and physical factors such as precipitation and physiography.
The expected stoichiometric composition of CaCO3 to MgCO3 and Si4+ of the dolomites should be 53:43:2 percent (Booysen, 1981).Due to the chert banding in the Monte Christo formation there will be co-precipitation of Silica and also some iron, aluminium and manganese oxides due to the presence of siderite and rodochrosite.
Acid Mine Drainage (after Scott, 1995).
This phenomenon has already been known since 1903 during pumping from the gold mines. The Witwatersrand Supergroup sediments contain varying proportions of sulphide minerals, the predominant sulphide being pyrite (FeS2) and can constitute up to 3 % by weight of the rock. Other traces of suphide includes; pyrrhotite, arsenopyrite, chalcopyrite, galena, cobaltite and gersdorffite. Atmospheric oxidation initiates the reaction, the sulphidic component (S2- ) in pyrite is oxidized to sulphate (SO42- ), whereby acidity (H+) is generated and ferrous iron (Fe2+) ions are released, the reaction:
FeS2(s) + 7/2O2 + H2O → Fe2+(aq) + 2SO42-(aq) + 2H+Acidity (aq)
Once the reaction is initiated ferrous iron is oxidized to ferric iron (Fe3+) the reaction is speeded up by microbial catalysis, the reaction:
Fe2+(aq) + 1/4O2 + H+ Bacteria Fe3+(aq) + 1/2H2 O
The ferric iron is then hydrolysed by water to form the insoluble precipitate, ferrihydrate (Fe(OH)3) plus more acidity, the reaction:
Fe3+(aq) + 3H2 O Fe(OH)3(s) + 3H+
The oxidation of pyrite is the most acidic of all weathering reactions since four moles of acidity is created from one mole pyrite. In addition to reacting with oxygen, pyrite may also be oxidized by dissolved ferric iron to produce additional ferrous iron and acidity, the reaction:
FeS2(s) + 14Fe3+ + 8H2O → 15Fe2+(aq) + 2SO42-(aq) + 16H+Acidity (aq)
Thus once the acid generating reactions have started, oxygen is only necessary for the microbial oxidation, pyrite will continue to be oxidized even in the absence of oxygen by ferric iron. Therefore sealing or flooding of mines will not stop oxidation.
The lowered pH due to above reactions also has the ability to leach other sulphide minerals in the ore like Ni, Pb, Cu, As and leachable oxides which are uranium bearing. These will normally co-precipitate with the iron oxides.
Geology (after Martini et al., 2003)
Stratigraphy
The geological sequence in the Cradle area consists of sediments that have been deposited onto basement granites and greenstones. The Witwatersrand sediments containing gold bearing conglomerate layers were deposited on this basement rock. They were in turn unconformably overlain by the Ventersdorp Supergroup andesitic lavas, which in turn were unconformably overlain by the Transvaal Supergroup dolomites. They are finaly covered conformably by Pretoria Series shales and quartzites.
The Cradle Caves are developed in the Malmani Subgroup of late Archaean age (2,5-2,6 billion years), which have been deposited on the intracratonic Transvaal Basin.
The lithology consists essentially of shallow marine stromatolitic dolostone with a variable amount of chert. The dolomite mineral is typically rich in Fe and Mn (up to 3% combined). Its thickness reaches 1450m in the Sterkfontein area . Based on the abundance of chert, the subgroup has been subdivided into 5 formations.
The Oaktree Formation (180m thick) represents the basal unit, characterized by its very chert poor nature. The overlying unit, the Monte Christo Formation (700m thick) is rich in chert. It has thin but spectacular oolitic beds at its base. The strata dip about 30° to the NW.
Tectonic Activities
Tectonic activity resulted in open folds and block faults in the Witwatersrand sediments, which caused some of the sediments to outcrop to the south of the Cradle area and is where most of the mining activity took place. Folding and faulting is featured throughout the Transvaal dolomites as well and three events can be highlighted: (i) Folding and faulting due to the origin of the Bushveld complex +/- 2000 ma, (ii) folding and faulting during Pilanesburg intrusion +/- 1400-1300 ma and (iii) later movement associated with uplifting.
Dykes and sills from north striking to east-west striking of Pilanesburg and Karoo age are numerous in the area. A large variety of fractures are visible in the dolomite of which the strike and dip varies throughout the area.
Methodology
The following sampling procedure was followed according to Department of Water Affairs and Forestry (DWAF) criteria and standards. Measuring of infield physical parameters (EC, DO, Redox, pH, TDS, Sal and Temp.)
Measure infield Total Alkalinity (mg/l CaCO3) and Sulfates. Sampling for Major ions and Trace elements as follows: Major Ions: 2 x 250 ml plastic bottles, rinsed three times with sample water before filling (physical parameters are measured using 60 ml from one bottle and alkalinity and sulphates with 10 ml respectively). Both bottles are immediately cooled in a cooler bag to +/- 4 degrees and stored in a fridge thereafter until lodged at the Lab (usually within a day).
Trace Elements: 1 x (pre-rinsed with DI water) glass bottle is filled with 100 ml of water via a syringe filter -0.45μm (conducted even under the most strenuous conditions) and 5 ml HNO3 to prevent secondary reactions from developing. Gloves are worn during sampling and equipment is thoroughly de contaminated with cleaning solution and DI water before continuing.
The samples are then analysed by the CGS’s Laboratory where Ion chromatography (IC) is employed to determine seven different anionic species, mass spectrometer (MS) and inductively coupled plasma spectrometer (ICP) is utilized for the trace elements. The major ions are measured in ppm while the minor or trace elements are measured in ppb.
Results
To date 61 samples have been acquired throughout the southern portion of the Cradle area (figure 2). These were all analysed and checked for reliability with a partly developed excel spreadsheet based on work done by Hounslow (1995). Using the charge balance calculation for reliability, it was found that 83 % of the samples were within + or – 10 %. The other 17 % is due to very aggressive mine waters with low pH and possible organics in significant quantities; they also typically have high salt values and no bicarbonates.
All these samples are one off samples and seasonal variations most certainly can cause significant variations. However certain interpretations can already be made.
Three different water bodies were sampled, surface water, cave water bodies and boreholes, although the caves and boreholes are mainly the same water in the dolomite the latter just not allowing direct access. The springs sampled are mainly the end of the line, residence time wise for the groundwater in the dolomite but are essentially groundwater. A basic relationship between surface and groundwater were attempted with no implication as to dynamic processes. This is due to surface water being subject to rapid mixing and gaseous exchange, which is not necessarily the case with groundwater.
Also as a note, no datum levels has been measured up to date and will follow in future, it would seem that the cave water bodies are some times at different levels due to sealing by clay or silty layers making them separate entities (pools) in what seem to be very low energy flow systems. All samples were plotted onto tri-linear diagrams (Piper) and Durov, Expanded Durov, Schoeller and Stiff diagrams to interpret the data and infer relationships.
Surface and mine waters
The four direct mine related samples (Harm-01, Harm bypass-01, T-spruit and Sunken stream), taken from the mine decant point and the immediate downstream points is typical Witwatersrand mine type water (Scott, 1995 and Usher, 2001) as it plots in field IV on the Expanded Durov diagram (Figure 4). The reef that was mined previously is below the dolomite and the water therefore has to move through the dolomite zone before decanting on the surface and therefore dissolution of the dolomite occurs and therefore the high calcium and magnesium (figure 5-Piper) associated with the Acid Mine Drainage (AMD) or sulphide oxidation taking place in the old worked out shafts. The spruit sample plots in field VI indicating water from treated effluent (municipal waste disposal site) and is directly downstream from the sewerage works, typical sulphates along with sodium and chloride values. No mixing with mine water is evident here. The surface samples progressing downstream plot in field V and VI which would indicate mixing from high sulphate mine waters and sodium and chloride treated effluent water. These values seems to gradually dilute as mixed with fresher dolomitic waters along the flow path until it plot in field II of the Expanded Durov Diagram which indicates typical dolomitic water. The samples that plot in field I is probably freshly recharged water. The Stiff diagrams (figure 6), although masked by the high concentration of sulphate scale wise, indicates the same patterns as described above with the Piper and Expanded Durov diagrams and specific signatures are visible for these different type waters.
All indications are therefore that the mine decant water mixes with the downstream surface water although other factors such as effluent water as well as agricultural pollution also contributes to the surface regime. The Schoeller diagram shows the same pattern although not included here.
Way Forward
Phase 1-A desk study compiling all current information including, Climate data, precipitation from relevant stations, stream flow, land use, GIS maps (topography, dtm, geology, catchments etc.), reports from previous authors, in digital format. Basically all relevant information on Geohydrology, hydrology and chemistry of the cradle. Phase 2-All structures and stress directions should be mapped from the airborne geophysical data. DWAF will assist in mapping of local geological structures. Phase 3-Gravity surveys over specific caves to map the direction of passages to link the dynamics between caves. This should be done for the bigger systems to begin with and possible ad hoc work later on for drilling. Completing the baseline data for the project area. This data will then be used for compiling flow nets (piezometric heads) for the modeling and quality maps to assist with the establishment of a monitoring network. A complete borehole census (including springs and caves) of the cradle area to the north. This will include a first phase identification of all boreholes including data on general access, co-operation with land owners, gps – coordinates, casing height, equipment fitted, is sampling possible, water level, owner details etc. Phase 4-During this phase a survey team will measure the accurate positions of the surveyed boreholes captured in phase one with regards to its x, y and z position in Cartesian and geographical coordinates.
Phase 5-Sampling of all available boreholes for base line geo-hydrochemistry data. Phase 6-Natural isotope samples and carbon 14 dating for some points in network, to date or trace possible flow paths. Interpretation of all relevant data acquired to date to construct conceptual model of the cradle.
Phase 7-Possible drilling of extra boreholes to calculate aquifer parameters for input into numerical model. Pump testing of boreholes to obtain hydraulic aquifer parameters. Phase 8-Double ring infiltrometer tests to establish seepage rate of surface water bodies into the subsurface to calculate recharge. Phase 9-Recharge calculations from all acquired data, essential for numerical model. Compilation and construction of numerical model and calibration using FEFLOW and PMWIN software. A possibility (not necessarily required) will be to look at tracer tests to establish certain important flow paths. Phase 10-From the above database and results of the numerical model a network of sampling points covering a grid of the area as well as all relevant “hot spots” will be designed from newly drilled boreholes and existing ones.
Phase 11 -Frequent sampling according to above criteria for at least two years. Compiling a GIS map containing all liable polluters to the Cradle area. Following legal frame work to hold certain liable parties accountable.
Conclusions
The baseline chemistry data acquired to date is a very helpful tool to establish trends for future reference for pollution and state of the environment. Although it is not a full complete picture of the Cradle it does however yield a starting point from where future work can advance and concentrate. More seasonal and temporal data is definitely needed to complete the picture. With this data it is clear that three types of pollution is entering the system, those from the mining activities, Agriculture and small industries and waste disposal sites. Building a suite of data as part of a Geographical information system seems to be the only way to manage a system of this size. Monitoring of the system with properly planned positions from the current data will allow to pinpoint any future changes. Conceptualizing this system is going to be of utmost importance before final interpretations can be made.
References and Bibliography
Bredenkamp, D.B. 1986. Groundwater supply potential of dolomite compartments west of Krugersdorp , DWAF. Technical Report No. GH3440.
Drever, J.E. 1982. The geochemistry of natural waters. Prentice Hall, nglewood Cliffs. ISBN 0-13-272790-0.
Hodgson, F.D.I., Usher, B.H., Scott, R., Zeelie, S., Cruywagen, L-M. and De Necker, E. 2001. Prediction Techniques and Preventative Measures Relating to the Post Operational Impact of Underground Mines on the Quality and Quantity of Groundwater Resources.
WRC Report No. 699/1/01. ISBN 1-86845-701-X
Hounslow, A. W. 1995. Water quality data: Analysis and Interpretation. ISBN 0-87371-676-0.
J.E.J.Martini(1), P.E.Wipplinger(2), H.F.G.Moen(2) and A.Keyser(3). 2004. Contribution to the speleology of Sterkfontein Cave,Gauteng Province, South Africa.
Ligthelm, R. 1993. Die Hidrogeologie van Karstgebiede met Spesiale verwysing na Grondwater Besoedeling Suid van Pretoria. MSc. University of Pretoria.
Nazari, M.M., Burston, M.W., Bishop, P.K. and Lerner, D.N. 1993. Urban Ground-water Pollution: Case Study from Coventry, United Kingdom. Ground Water, Vol. 31, No. 3. pp 417-424.
Scott, R. 1996. Geohydrochemistry and Pollution. 3rd Edition, Institute for groundwater studies.
