CORONA : Confidence in forecasting of natural attenuation as a risk-based groundwater remediation strategy


Prof. Christof Holliger, Dr. Theo Smits, Christiane Devenoges

Funding agency

5th Framework Programm of the EU

Project period

November 2001 – April 2005


Prof. D. Lerner, Univ. of Sheffield, U.K., Prof. P. Grathwohl, Univ. Tübingen, Germany, Prof. P. Bjerg, DTU Kopenhagen, Denmark, Dr. A. Gargini, Univ. Ferrara, Italy, Dr. H. Rijnaarts, TNO Apeldoorn, The Netherlands, Dr. J. Datel, Charles Univ., Prague, Republic Czech, Prof. R.M. Kalin, Univ. Belfast, U.K., Prof. R. Schotting, TU Delft, The Netherlands, Dr. P. Morgan, GeoSyntec, Runcorn, U.K.


The general objective of the CORONA project was the development and optimization of a numerical model allowing the prediction of natural degradation processes occurring in groundwater systems. The project hypothesizes that a common pattern of biodegradation activity can be found in most groundwater pollution plumes. Certain zones – which we will call the CORONA – will have better conditions for biodegradation, and these zones will have more rapid degradation and make a significant contribution to the overall rate of mass loss for the entire plume. We can identify two basic scenarios,

  1. an active, oxidizing fringe to a plume mainly controlled by dispersion, and
  2. an anaerobic core controlled by the interplay between pollutants, environmental conditions and micro-organisms.

The project will collect high resolution data from the coronas at six different field sites. Parallel laboratory studies will include a model plume for analysis of transverse dispersion and detailed study of a corona zone. There will be microbiological and microcosm studies in support. Our part of the project was to elucidate the microbial processes occurring at a site contaminated with chloroethenes and another one contaminated with ammonium.


Possible mechanisms involved at the first site that were taken into consideration were anaerobic oxidation of chloroethenes with iron(III) as electron acceptor and reductive dechlorination. Whereas nothing is known about the bacteria involved in the first process, quite a large amount of information exists about the reductive dechlorination. With microcosm columns operated under different conditions, we obtained indications about the presence of anaerobic chloroethene oxidation in samples from the chloroethene-contaminated site. For reductive dechlorination, we continued to develop a molecular tool enabling the detection of actively dechlorinating bacteria. The ammonium-oxidizing populations present at the ammonium-contaminated site were identified and quantified with Real Time PCR and the use of PCR primers specific for the 16 rRNA genes of the corresponding bacteria.

Using sand samples from borehole MLS3 of the chloroethene-contaminated site, we filled three columns enriched with anaerobic Fe(III)-coated sand with different set-ups. One column was fed with only vinyl chloride (VC), one with VC and acetate, and one with VC and mercury chloride. The columns were operated for 21 months and analyzed periodically. Reductive dechlorination and methanogenesis took place in the column fed with VC and acetate, and disappearance of VC without obvious product formation was observed in the other two columns. Stable isotope analysis showed that in the VC-only and VC + HgCl2 fed columns, the enrichment factor ε for 13C in VC was between -1 and -5, while in the VC and acetate-fed column, the ε-value was between -13 and -18. Literature values for reductive dechlorination of VC range between -20 and -25. The absence of the dechlorination product ethene in the VC-only and VC + HgCl2 fed columns and the clearly different ε-value indicate that the disappearance of VC in these two columns is due to another process, possibly anaerobic microbial VC oxidation. The ε-value in the VC and acetate-fed column was probably the result of two processes, reductive dechlorination and anaerobic VC oxidation. Iron speciation and molecular analyses of certain zones after dismantling the columns showed that in the VC and acetate-fed column, most of the Fe(III) was gone, while in the VC-only column, the concentration of Fe(III) was in the same range as the initial iron concentration. This implied a VC oxidation without concomitant iron reduction, which was confirmed by second-generation cultures started with column material from the VC-only column. Molecular analysis by real-time PCR showed the presence of around 107 copies of 16S rRNA genes per gram column sand in both columns, being slightly higher in the VC and acetate-fed column. In both columns, the absence of Dehalococcoides spp. was shown by real-time PCR, and by the absence of Dehalococcoides-related peaks in T-RFLP analysis. By T-RFLP, the presence of completely different bacterial communities could be shown.

For the development of a method to determine the transcriptional activity of reductive dehalogenating bacteria in soil, pure cultures of Desulfitobacterium hafniense TCE1 were followed over time, and samples for DNA and RNA extraction were taken regularly. Real-time PCR with the DNA and reverse-transcribed RNA were done using Eubacterial primers to obtain the ratio of the number of ribosomes per cell calculated as copy number of rRNA divided by (rRNA gene copy number/2) under the assumption that D. hafniense TCE1 contains two copies of the 16S rRNA gene on its chromosome. This ratio varied between 100 and 12000 for actively growing cells, and decreased to 1-2 after 1 week of stationary phase cultivation. After one month, the rRNA copy number could not be determined anymore. The method is well reproducible, as three DNA extracts from the same culture maximally gave a factor of two difference between the determined copy numbers, whereas three RNA extracts for which three reverse transcriptase steps were done yielded as well only a factor of two difference, both between the individual RNA extracts and the reverse transcriptase steps. The next step will be to test the method on mixed enrichment cultures.

Finally, groundwater samples were obtained from the REXCO site in order to study the presence of aerobic (AOB) and anaerobic (Anammox) ammonium-oxidizing bacteria. High concentrations of ammonium and the presence of nitrate and nitrite at the fringe of the plume indicated that next to aerobic ammonium oxidation, also anaerobic ammonium oxidation could take place. PCR done with primers specific for β-proteobacterial AOB showed the presence of AOBs over the vertical profile of the fringe, while Anammox bacteria were detected in a narrow range below the aerobic zone. Cloning and sequencing of Anammox-specific PCR fragments showed sequence identities of 93-95% to planctomycetes able to catalyze Anammox. Quantification of the populations by real-time PCR showed that AOBs accounted for up to 50% of the total eubacterial community around the aerobic fringe. Below they accounted for approximately 25%, and populations of Anammox bacteria for 10 and 25%. Anammox bacteria have up to now mainly been detected in wastewater treatment plants and in aquatic environment. This is the first report of significant numbers of Anammox bacteria present in an aquifer.