The feasibility of IN SITU groundwater remediation as robust low-cost water treatment option.
Feb 2004

Report No 1325/1/04


Groundwater pollution is a worldwide phenomenon and the wide ranging consequences are gradually being realized. Research into methods of groundwater cleanup or remediation has increased and various techniques have been developed and applied. "Ex situ" pump-and-treat systems have been used widely, particularly in the United States of America. However, the success of such approaches has been questioned, considering the high costs involved. As an alternative, in situ treatment techniques, that remove contaminants while the groundwater resides in the aquifer, are being developed.


This report presents information gained from literature and contacts with researchers abroad on the application of in situ treatment techniques. The present investigation represents an initial phase of the original proposal and the following objectives formed the core of the research:

Later phases of the research would entail the design, construction, operation and evaluation of a prototype in situ groundwater denitrification system, which will be to the benefit of the community and the derivation of design criteria for such systems. This, however, does not form part of the present project.

The present investigation focuses to a large extent on the removal of nitrate and, therefore, more detailed information is recorded on denitrification than on other treatment processes.

Nitrate removal

Most methods of nitrate removal that have been applied for in situ groundwater treatment are based on chemical and/or biological denitrification. The methods apply redox reactions, often with biological catalysis, to reduce nitrate to nitrogen gas. The appeal of using denitrification reactions for in situ application lies mainly in the fact that the main products of the reactions are gaseous and do not accumulate as hazardous by-products in the subsurface. Some of the techniques also do not require highly sophisticated technology.

A series of factors influence the denitrification reactions and the most important ones for biological denitrification are listed below:

The efficiency of denitrification may not be of too great concern for in situ treatment systems, as long as the product water complies with drinking water specifications. Porosity and permeability in the aquifer are additional system related factors that affect the efficiency of denitrification.

Ion exchange, reverse osmosis and electro-dialysis have also been used in nitrate treatment plants, but have not been developed for in situ application. These methods remove anionic nitrate without the need for redox conversions, but they result in concentrated water treatment wastes that need to be disposed of.

Review of in situ groundwater treatment

Although in situ groundwater treatment is still being researched and developed, it has shown potential for the removal of a large range of inorganic contaminants, including arsenic (As), cadmium (Cd), chromium (Cr), copper (Cu), mercury (Hg), iron (Fe), manganese (Mn), molybdenum (Mo), nickel (Ni), lead (Pb), selenium (Se), technetium (Te), uranium (U), vanadium (V), nitrate (NO3-), phosphate (PO4-), and sulphate (SO42-). Organic compounds can also be removed by in situ treatment methods.

Treatment methods

The most important in situ groundwater treatment methods for nitrate and many other contaminants can be divided into the following three main groups:

Permeable Reactive Barrier (PRB) techniques (also called "passive treatment walls") involve the physical placement of a barrier, consisting of reactive material, into a trench excavated in the aquifer. In other cases, a chemical reagent (e.g. a reducing agent) is injected into the aquifer to create the reactive barrier. This technique is called In Situ Redox Manipulation (ISRM). The barrier created with ISRM can be placed at much greater depths in the aquifer than with a trench-filled PRB. For these methods to work efficiently, especially for remediation operations, a primary aquifer is needed with well-defined, impermeable lower boundary.

PRB techniques, specifically zero-valent iron walls, are widely used for aquifer cleanup operations in preference to pump-and-treat methods. A wide range of contaminants, including arsenic, cadmium, chromium, copper, mercury, iron, manganese, molybdenum, nickel, lead, selenium, technetium, uranium, vanadium, nitrate, phosphate, and sulphate, can be removed in this way. Field studies of in situ treatment using PRBs have been completed for most of these contaminants.

With regard to biological methods, this study concentrated on denitrification and did not investigate the wide range of organic compounds that may also be treated by this method. In Situ Biological Denitrification (ISBD) methods generally require the introduction of a carbon source into the aquifer that serves as a substrate for the bacteria. This may be in the form of methanol, ethanol, glucose or even sawdust or wood chips injected via one or more boreholes. The Nitredox method is a special case of ISBD, involving the introduction of a carbon source and later oxidation of iron and manganese by-products.

Biological methods are widely used for removal of nitrate and degradable organic compounds. This treatment method is a viable option when the rate of contaminant biodegradation is faster than the rate of contaminant migration.

Electrochemical methods are more complex than PRB or biological methods, but are considered here because they are reputed to be applicable to fractured aquifer environments. Various electrochemical systems, applying enhancements such as electrokinetics and electro osmosis, have been described for groundwater treatment. These systems generally use an electrical current, applied via two in situ electrodes to control the movement and redox chemistry of ions and water in the subsurface. Electrochemical methods can be used as an enhancement of other PRB systems, e.g. by combining electrodes with iron walls to remediate nitrate contaminated groundwater and soils abiotically.

Electrokinetic methods appear to have been developed with a focus on remediating spills or leaks of organic chemical products. Much of the emphasis is on mobilising contaminated pore water and there is an expectation that the abstracted water would require additional ex situ treatment. It was stated in the literature that nitrate removal could be improved by coupling the electrokinetic method with an iron wall. However, no information was provided comparing the efficiency of electrokinetics on its own, or that of a coupled system, with non-electrical methods.

Advantages of in situ treatment

In situ treatment of contaminated groundwater uses the aquifer as a subsurface "treatment plant" to improve the quality of groundwater supplies. This has several advantages over conventional ex situ treatment technologies (pump-and-treat systems), including:

Side effects of in situ treatment

Some in situ treatment methods for removing groundwater contaminants may cause undesirable side effects, of which the most common problem is clogging in the subsurface. In the case of removal of organic compounds, the products of the treatment reaction may cause clogging of the aquifer due to biofilm build-up. If metals are removed by precipitation, the solid precipitates in the aquifer matrix may reduce the permeability of the aquifer over the long term. The specific hydrogeological conditions and the contaminant load will determine the extent of these phenomena. Side effects need to be managed to maintain the efficiency of the scheme and increase the treatment lifetime.

Depending on the environmental conditions, denitrification reactions may not always run to completion, i.e. nitrate nitrogen is not always fully converted to nitrogen gas. This can lead to the accumulation of other undesirable nitrogen species, such as nitrite or ammonium, in the groundwater. Nitrogen immobilisation may also retard the appearance of nitrate, but will not remove the nitrogen source from the aquifer and nitrate may re-develop at a later stage. Where potentially toxic reagents (e.g. dithionite, methanol) are introduced to the subsurface and not fully recovered or consumed by the reactions, groundwater quality may be adversely affected.

Possible side effects of in situ nitrate removal methods include clogging or loss of effective permeability in the aquifer as a result of factors such as:

Design criteria

For the design of any treatment system, both the hydrodynamics of the flow system and the source of the pollution need to be exceptionally well characterised to optimise the system.

The nature of the pollution source will affect the choice of system design. Where a single point source of pollution is affecting a water supply borehole, a simple porous barrier, single ISRM or ISBD injection borehole can be constructed along the flow path between the source and production borehole. If non-point sources are responsible for pollution, production boreholes must be protected by a surrounding gallery of treatment systems e.g. the daisy configuration for ISBD, the Nitredox system, multi-electrode arrangement or a circular barrier wall.

Trench and fill barrier methods are only suitable for shallow flow systems where the barrier can be constructed down to the impermeable bedrock. In deeper flow systems, treatment barriers can be created by borehole injection. In situ treatment methods are more likely to be successful in primary aquifers, where hydrodynamics are more easily understood and greater control over the treatment zone can be exercised than in fractured flow environments.

For nitrate removal systems, management of nitrogen inputs to the subsurface is still required, as treatment of the nitrate in the aquifer does not remove the contamination source. Although the treatment systems generally have a long lifetime, it is still important that the source of contamination be eliminated wherever possible.

Pilot and full-scale in situ denitrification systems worldwide

Most of the field scale in situ groundwater treatment systems use the PRB type technique. In Canada and New Zealand, sawdust and woodchips are used as carbon substrate in PRB systems. Biological denitrification plants in the USA, France and elsewhere, generally use ethanol or methanol as carbon substrate. In one pilot plant in Israel, sucrose was used for this purpose.

One full-scale denitrification plant has been in use for more than a decade. This is the one at Bisamberg, which supplies denitrified water to Vienna, Austria. The plant uses the Nitredox method. Iron and manganese that are mobilized from the aquifer during the denitrification process are re-precipitated in the oxidation step. The capacity of the plant is 60 L/s.

For the purpose of comparison, brief information on certain ex situ treatment systems is also provided in the report.

Potential for in situ treatment applications in South Africa

The literature review shows that in situ groundwater treatment has significant potential for application in South Africa. Internationally, researchers are enthusiastic about the potential of in situ treatment for addressing various problems regarding chemical constituents found in pollution situations but also those in natural groundwater environments. In situ treatment systems for iron and manganese removal have been used successfully for many decades while in situ nitrate removal has been in place for approximately two decades.

The countrywide distribution of nitrate levels in groundwater was compared with the various hydrogeological terrains of South Africa and the towns or rural communities where groundwater constitutes the sole water supply source. This assessment has indicated that there is some agreement between high nitrate concentrations, sole source areas and hydrogeology. High nitrate concentrations are fairly well correlated with aquifers composed of unconsolidated deposits, weathered basalts, and dolomites over an area extending from the Northern Cape Province all the way to northern Mpumalanga, including parts of North West and the Limpopo Province. Areas that are sole source areas with high groundwater nitrate concentrations have priority over other areas for remediation. These include towns such as Marydale, Leliefontein, Reivilo, Rietfontein in the Northern Cape, and others, largely located in the rural parts of Northwest and Limpopo Provinces.

Iron and manganese, occurring naturally in groundwater, may cause significant clogging problems in boreholes when redox conditions change and iron bacteria start multiplying. In primary aquifers, in situ treatment by oxygenation may provide a viable solution, e.g. by using the Vyredox process. This can be applied at Atlantis, where even persistent low levels of iron cause significant borehole clogging problems.

Regulatory requirements

Regulatory requirements for in situ treatment methods are not specifically covered in South African law. However, any in situ groundwater treatment system will be subject to the stipulations contained in the National Water Act (No 36 of 1998) and the National Environmental Management Act (No 107 of 1998). As a minimum, an impact assessment study will have to be carried out for any proposed full-scale in situ treatment scheme. The potential impact on the environment, socio-economic conditions and cultural heritage of all activities that require permission by law must be considered, investigated and assessed before implementation. Compliance with pollution prevention regulations require authorisation from DWAF before any reagents/substrates would be allowed to be injected into an aquifer as well as for the management of any waste products produced if any. However, it is believed that with a thorough approach, the legislative requirements can be fulfilled.

In this regard it may be of great value to learn from the approach in the USA and extracts of regulatory issues that are being addressed in that country are included in the report.

Cost comparison for different treatment methods

For the purpose of assessing economic feasibility, a desk study was conducted for denitrification of the groundwater at the rural town of Marydale. Three different in situ methods and one ex situ were considered. The capital investment for the ex situ plant was between three and seven times higher than for the in situ systems. The largest difference, however, was in the operating costs. Whereas the in situ methods had virtually no operating costs, the ex situ costs each year were nearly as high as the initial capital costs. The permeable reactive barrier system should have the lowest capital costs.

Even if these estimates have certain incorrect assumptions, it is evident from the calculations that it is impossible to have an ex situ treatment system that will be more economical than an in situ system such as the PRB. This also explains why in situ methods have already gained wide acceptance abroad.


The analysis of existing methods and systems worldwide, together with the preliminary cost analysis, shows that in many instances in situ groundwater treatment methods will provide a viable, cost-effective alternative to ex situ water treatment. In view of the success abroad, it is strongly recommended that such systems be tested both at field scale and full scale in South Africa. Particularly those systems that require a low capital investment, low running costs and limited know-how should be tested and installed without delay. For each potential site a feasibility study should be undertaken and cost estimates calculated. Three to five sites should be developed as demonstration units, for technology transfer and obtaining local experience with this technique.

It is recommended that all four methods, i.e. permeable reactive barriers (PRB), in situ redox manipulation (ISRM), in situ biological denitrification (ISBD) and electrokinetics are tested at suitable locations. All methods are suitable for groundwater denitrification, but also for the removal of other contaminants such as heavy metals or organic compounds. All four methods would be viable in the town water supply context but only some of them, e.g. PRBs, would be compatible with the rural setting. The availability of infrastructure and technical know-how would make the Nitredox and Vyredox methods viable for nitrate, iron and manganese removal in town water supply applications.

Local testing of PRBs for the remediation of contaminants such as heavy metals (e.g. chromium) or organic compounds is strongly recommended. In the USA and other countries many PRB remediation systems are successfully in operation. The cost advantages involved make these methods attractive for remediation of spills and other pollution problems and it is recommended that the methods be applied locally to develop the local expertise for wider application. In situ remediation could, for example, be tested at Springbok, where chromium was allegedly spilled into an alluvial aquifer. Whereas PRBs are applicable to shallow alluvial and other porous aquifers, ISRM techniques are recommended for testing on deeper porous aquifers. More details on each are provided below.

Recommended denitrification systems

Permeable reactive barriers (or "treatment walls"): PRBs can be constructed from cheap, readily available materials and would be relatively simple for communities to install with limited training. They provide long-term treatment without maintenance and no power source is required. By-products of the sugar industry could be investigated as a more readily degradable carbon source, for example in rural KwaZulu-Natal, in addition to the more conventional wood chip/sawdust barriers.

In situ redox manipulation: ISRM should be tested for deeper primary aquifers in South Africa. The dithionite chemical reagent should be readily available, since it is used in the pulp and paper industry. Laboratory testing would be recommended for the first phase followed by field-testing for those instances with promising test results. This method can be tested in the Kalahari in areas where the aquifers are not too deep. It could also work in basalt aquifers with sufficient porosity.

In situ biological denitrification: Internationally ISBD is probably the most widely used in situ treatment method for the removal of nitrate from groundwater. The configuration of the injection and abstraction boreholes is flexible and can be adapted to suit the specific treatment problem. Various conventional and organic substrates and cheaper locally available options could be tested for potential application in rural treatment systems. The injection of microbes would not be recommended at this stage.

Electrokinetics: Various electrochemical techniques might be useful as an enhancement for the first two methods above, as the electrochemical techniques do not appear to be completely efficient for nitrate removal on their own. More information would need to be collected on the electrode composition and installation, applied voltages, etc. and extensive testing conducted before electrokinetics could be applied in the field. The electrokinetic method is, however, the only one found that has been claimed to be suitable for fractured rock environments, which would make it worth testing for South Africa. This is the main reason why this method is recommended for testing despite the higher levels of technical skill and know-how that is required.

Nitredox: The Nitredox denitrification method is more complex and possibly too expensive for rural water supply application. However, based on its long successful record at Bisamberg, Austria, it is recommended that this technique be tested in a primary aquifer where the large number of injection and observation points can easily be created by the installation of well points. It is, however, essential to investigate whether the licensing costs in terms of registered patents may prohibit its application. Nitredox denitrification may be considered for town water supply application following a thorough evaluation of the costs involved.

Recommendations for iron and manganese removal systems

Vyredox: Iron and manganese removal is imperative in many high volume abstraction situations to avoid or minimise borehole clogging effects. At towns such as Atlantis, the beneficial effects of iron and manganese removal will offset the costs of implementing in situ treatment. It is recommended that a system similar to the Vyredox system be tested on a South African primary aquifer.

Recommendations for contaminant remediation systems

PRB: These can be constructed from cheap, readily available materials such as foundry wastes or furnace slag. They would be relatively easy to install at mining and industrial sites and would provide a low-cost, long-term treatment option.

ISRM: The ISRM technique is practically as effective and maintenance free as the PRB method but as ISRM can be installed at greater depth it can be applied to a larger range of primary aquifers.