|
Discussion |
1 Bradfield Hall, Cornell University, Ithaca, New York 14853, USA
2 Water UK, 1 Queen Anne’s Gate, London, SW1H 9BT, UK
3 Department of Civil & Environmental Engineering, Imperial College, London, SW7 2BU, UKt
P. Baveye writes: In the title and abstract of their recent article, Tompkins et al. (2001) advocate engineered, in-situ bioremediation as a viable and practicable option for nitrate removal in Chalk aquifers. Surpringly, this enthusiastic endorsement appears to be severely undermined by the otherwise remarkably thorough and interesting multiscale analysis presented in the body of the article. Indeed, as a conclusion of this analysis, Tompkins et al. (2001, p. 121) point out that the ‘biochemical and hydrological flow processes controlling denitrification activity in fissured aquifers are complex and not yet sufficiently understood at this stage for the in-situ technology to be optimized and verified technically or economically’. They recommend that a systematic study be carried out, using a combined numerical and experimental approach, to develop a fundamental understanding of the mechanisms and factors controlling denitrification in Chalk.
Given the fact that the authors appear to be simultaneously on both side of the fence, it must be unclear to practitioners who would happen to read Tompkins et al.’s (2001) article whether they should feel encouraged to implement in-situ bioremediation strategies in Chalk aquifers, or whether for the time being they should turn to other options, until further experimental results and reliable models become available.
I would like to suggest that if the analysis of Tompkins et al. (2001) is taken one step further, a somewhat clearer picture emerges. The step in question consists of better taking into account the influence of microbial activity on the hydrodynamic characteristics of Chalk aquifers. Tompkins et al. (2001) focussed on what they refer to as ‘biofouling’. They have correctly recognized (cf. Tompkins et al. 2001, fig. 7) that the development of biofilms on the walls of primary and secondary fissures can reduce the groundwater flow rate in individual fissures by restricting their aperture, and can also decrease the overall connectivity of the aquifers. These processes can obviously have serious consequences for the practice of bioremediation and require optimization, e.g., by adjusting the rate of injection of electron donor in the aquifer.
However, the influence of micro-organisms is most probably not limited to biofouling of primary or secondary fissures. Micro-organisms, directly via their cell biomass or indirectly via the release of metabolic by-products, are likely to also block the pores of the Chalk matrix, at least in the regions of the matrix that are adjacent to fissures. One possible mechanism of this ‘bioclogging’ is the formation of gas bubbles, which in other contexts have been shown to be extremely effective at blocking pores (e.g. Baveye et al. 1998; Beckwith & Baird 2001). The experiments carried out by Soares et al. (1989) provide compelling evidence that, in particular, dinitrogen gas bubbles, released by denitrifiers, can reduce significantly the hydraulic conductivity of porous media. In principle, these bubbles should also reduce drastically the interstitial space through which chemical compounds (such as nitrate ions) can diffuse.
In this context, it is more than plausible to consider a hypothetical scenario in which peripheral regions of matrix blocks in Chalk aquifers become severely clogged as a result of denitrifying activity, stimulated by injection of an electron donor. In this scenario, the nitrate ions inside the matrix become virtually sequestered, unlike those in the fissures, which are both readily available to the biomass and easily flushed by the flowing groundwater. After a certain time, even though appreciable amounts of nitrate are still present in the aquifer (in the matrix), the nitrate concentration in the fissure water drops to very low levels, below regulatory standards. At that point, further injection of the electron donor in the aquifer becomes pointless. After injection is stopped, the biomass begins to decay, the dinitrogen gas progressively gets solubilized in the groundwater, the trapped bubbles evanesce, and diffusion of nitrate ions between the matrix and the fissures is resumed. This sequence of processes causes the nitrate concentration in monitoring wells to ‘rebound’ (e.g. Baveye & Bladon 1999) to levels commensurate with those before the start of the bioremediation operation. Such a rebound is clearly demonstrated in the field experiments described by Chevron et al. (1998). After 150 days of pulsed ethanol injection in a nitrate-contaminated chalk aquifer in France, these authors interrupted injection entirely for 100 days and observed that, as a result, groundwater nitrate concentrations at monitoring wells returned to their initial, pre-bioremediation level. The exact cause of this rebound is unknown, even though the authors explicitly suggested gas production as a possible mechanism of bioclogging at the investigated field site.
A consequence of these observations is that bioremediation is likely to be an effective solution to nitrate contamination in Chalk aquifers, but that a number of long-term cycles of electron-donor injection may be necessary to completely alleviate the nitrate problem in any given situation. For the timing and fine-tuning of these injection cycles, measurements of nitrate or nitrite concentrations in observation wells would be inadequate. Instead, one should proceed to routine monitoring of dissolved dinitrogen gas or, preferably (Blicher-Mathiesen et al. 1998), of other, more reliable evidence of denitrifying activity (e.g., enzyme concentration). This monitoring may have to last significantly longer than what is commonly envisaged by the private sector and, consequently, it may make in-situ bioremediation far less appealing economically than it is currently perceived to be. Nevertheless, only case-by-case comparisons with alternative treatments (like pump-and-treat) can determine when and where in-situ bioremediation will be a ‘viable’ technology.
J. A. Tompkins & S. R. Smith reply: We welcome the critique of our paper by Philippe Baveye. Earlier reviews considering in-situ denitrification as a potential treatment option for high nitrate groundwater abstracted from Chalk aquifers were sceptical about the likely success of this approach to nitrate removal (e.g. Hiscock et al. 1991). However, following a detailed programme of research using laboratory scale microcosms, and a modelling approach to predict water flow, nitrate transport and biochemical attenuation in porous aquifer media (Smith et al. 1999), we considered it timely to revisit the literature on denitrification in fractured aquifer rock types. Our original research, in collaboration with WRc in the UK, demonstrated that in-situ bioremediation in Triassic sandstone was a technically feasible, economically viable and environmentally sustainable method for nitrate removal (Smith et al. 1999). The principal conclusion reached by Tompkins et al. (2001) was that a managed denitrification approach was also possible in Chalk groundwaters and we agree with Baveye that bioremediation is likely to be an effective solution to nitrate contamination in Chalk aquifers.
Having accepted the principle of in-situ bioremediation as a feasible option, it is necessary to consider its application on a site-by-site basis because of the unique combination and interaction of hydrogeological and biochemical factors mediating attenuation that control nitrate removal under site specific conditions. In our earlier work with Triassic sandstone, we performed a series of controlled experiments to parameterize a biochemical model of denitrification activity (Smith et al. 1999). Prior to this, there were virtually no data available that were suitable for the purpose of parameterizing biochemical models of denitrification activity in the subsurface environment (Kinzelbach et al. 1990). This was incorporated into a transport model that could operate at the aquifer scale so that the injection of electron donor could be optimized in a managed bioremediation system. In Tompkins et al. (2001), we advocate the same systematic and fundamental experimental and modelling approach to support the designing of site specific management systems for in-situ bioremediation by denitrification in Chalk aquifers. We believe that it is possible to develop a detailed mechanistic model of the interaction between the aquifer hydrology and the biochemical attenuation processes to optimize and configure system design and operation. A paper addressing these specific areas is currently in preparation.
Baveye raises the issue of bioclogging of the matrix adjacent to fissures by gas bubbles, and the associated reduction in hydraulic conductivity, as a process with potentially important implications for in-situ treatment of nitrate by enhanced biological denitrification. Indeed, we have also observed gas production at normal, ambient atmospheric pressures in experimental work on denitrification in subsurface rock (Cartmell et al. 1999). This raises two important points. Firstly, we have calculated that the high pressures, typical of saturated and confined Chalk aquifers used for public water abstraction in the UK, would probably maintain the gases in the aqueous phase. Indeed Blicher-Mathiesen et al. (1998) observed N2 in solution at much lower pressures. This would alleviate bioclogging due to gas production in the majority of the aquifer matrix, although there is a potential issue of gas release at the point of abstraction. The second point concerns the rebound effect associated with nitrate movement from the aquifer matrix. We recognize that nitrate may leak from the matrix into the denitrified fissure water as it transfers across a chemical gradient from zones of high concentration in pore water to low concentration water in the fissures, where biological denitrification takes place. However, we are not suggesting the treatment of whole aquifer systems, as this would be an impractical proposition at this stage. The extent of diffuse nitrate contamination of groundwater recharge, principally from intensively managed agriculture, will require an ongoing commitment to nitrate removal by the Water Industry to comply with current drinking water standards, despite the implementation of management practices by farmers to mitigate leaching losses from agricultural soil. Therefore, we consider the most likely practical approach to in-situ bioremediation will be to continuously treat high nitrate water in the zone surrounding a public water supply borehole as it is drawn towards the point of abstraction. Under these circumstances, leakage of nitrate from the matrix is unlikely to be a major constraint to the treatment process and can be managed by adjusting the stoichiometric ratio of electron donor supply to compensate for the matrix yield. We have argued that partial bioclogging of fissures could be beneficial in enhancing the dispersion of the injected carbon source in Chalk. Restricting the movement of nitrate from the matrix in the enhanced denitrification zone could also have a potential benefit by simplifying the control strategy, which could focus principally on treating the nitrate present in the incoming fissure water. On balance, biofouling of the fissure network is probably a more serious potential concern than clogging of the pores in the aquifer matrix, although we have found no experimental evidence for this when the carbon substrate is in limiting supply. We agree, therefore, that management regimes require optimization depending on individual circumstances by adjusting the rate of injection of electron donor in the aquifer and that this may include pulse-dosing to minimize the excessive accumulation of biofilm in the fissures.
Baveyes suggestion regarding the use of gas monitoring in the feed back control mechanism for substrate pulsing is interesting and helpful. However, controlling the balance between the rate of injection of electron donor with the rate of water abstraction, the nitrate content of the incoming water from the surrounding aquifer and the target nitrate content of the treated water requires understanding of a series of complex and dynamic hydrological and biological interrelations. This can be successfully achieved through a quantitative modelling approach and would form the basis of the management control system for the treatment process.
We believe that in-situ bioremediation is potentially a viable alternative technology to conventional methods of nitrate removal from Chalk groundwaters for potable water supply. Encouragingly, enhanced in-situ biodenitrification is becoming more widely accepted as one of the most promising technologies to recently emerge for treating nitrate contaminated groundwater (ITRC 2000). More work needs to be done, however, to develop the basic tools for predicting treatment performance, system configuration and control in Chalk and other fractured rock aquifers, but we consider this goal is achievable within a relatively short timescale, following the programme of experimental and modelling work we have proposed (Tompkins et al. 2001).
 |
References |
|---|
 TOP References |
|---|
Baveye, P. & Bladon, R. 1999. Bioavailability of organic xenobiotics in the environment: A critical perspective. In: Baveye, P, Block, J.-C, Goncharuk, & V. (eds) . Bioavailability of organic xenobiotics in the environment. NATO ASI Series. . Kluwer Academic Publishers, Dordrecht, The Netherlands227-248.
Baveye, P., Vandevivere, P, Hoyle, B. L, Deleo, P. C. & Sanchez de Lozada, D. 1998. Environmental impact and mechanisms of the biological clogging of saturated soils and aquifer materials. Critical Reviews in Environmental Science and Technology 28, 123-191.[CrossRef][ISI][GeoRef]
Beckwith, C. W. & Baird, A. J. 2001. Effect of biogenic gas bubbles on water flow through poorly decomposed blanket peat. Water Resources Research 37, 551-558.[GeoRef]
Blicher-Mathiesen, G., McCarty, G. W. & Nielsen, L. P. 1998. Denitrification and degassing in groundwater estimated from dissolved dinitrogen and argon. Journal of Hydrology 208, 16-24.[GeoRef]
Cartmell, E., Clark, L, Oakes, D, Smith, S. R. & Tompkins, J. 1999. Feasibility Study of Bioremediation of Nitrate in Aquifer Systems. Final Report. WRc Report No. , CO 4683/1, . WRc Medmenham, Marlow, UK.
Chevron, F., Lecomte, P, Darmendrail, D. & Charbonnier, P. 1998. Réhabilitation de la qualité physico-chimique d’un aquifère contaminé par des nitrates d’origine industrielle – un exemple en région Nord-Pas-de-Calais. L’Eau, L’Industrie, Les Nuisances 208, 31-35.
Hiscock, K. M., Lloyd, J. W. & Lerner, D. N. 1991. Review of natural and artificial denitrification of groundwater. Water Research 25, 1099-1111.[CrossRef][ISI]
ITRC (Interstate Technology and Regulatory Cooperation Work Group), 2000. Emerging Technologies for Enhanced In Situ Biodenitrification (EISBD) of Nitrate-Contaminated Ground Water. ITRC/EISBD-1. Also published on the ITRC Web site: http://www.itrcweb.org.
Kinzelbach, W. K. H., Dillon, P. J. & Jenson, K. H. 1990. State of the art of existing numerical groundwater quality models of the saturated zone and experience with their application in agricultural problems. In: DeCoursey, D.D. (ed) . Proceedings of the International Symposium on Water Quality Modelling of Agricultural Non-Point Sources, Part I. Report No. , ARS-81, . Agric. Res. Serv, US Dept of Agric, Beltsville307-325.
Smith, S. R., Cartmell, E. & Tompkins, J. A. 1999. Nitrate removal from groundwater by in situ bioremediation: Constraints and opportunities. . In: CIWEM National Conference on Making Better Use of Water Resources. The Chartered Institution of Water and Environmental Management, London, 105–117.
Soares, M. I. M., Belkin, S. & Abeliovich, A. 1989. Clogging of microbial denitrification sand columns: Gas bubbles or biomass accumulation? . Zeitschrift für Wasser- und Abwasser-Forschung 22, 20-24.
Tompkins, J. A., Smith, S. R, Cartmell, E. & Wheater, H. S. 2001. In-situ bioremediation is a viable option for denitrification of Chalk groundwaters. Quarterly Journal of Engineering Geology and Hydrogeology 34, 111-125.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
