Microbiology of Olkiluoto Groundwater, 2004-2006 (pdf) - Posiva

POSIVA 2008-02Microbiology of Olkiluoto Groundwater2004 – 2006Karsten PedersenFebruary 2008POSIVA OYOlkiluotoFIN-27160 EURAJOKI, FINLANDPhone (02) 8372 31 (nat.), (+358-2-) 8372 31 (int.)Fax (02) 8372 3709 (nat.), (+358-2-) 8372 3709 (int.)

POSIVA 2008-02Microbiology of Olkiluoto Groundwater2004 – 2006Karsten PedersenMicrobial Analytics Sweden ABFebruary 2008Base maps: ©National Land Survey, permission 41/MYY/08POSIVA OYOlkiluotoFI-27160 EURAJOKI, FINLANDPhone (02) 8372 31 (nat.), (+358-2-) 8372 31 (int.)Fax (02) 8372 3709 (nat.), (+358-2-) 8372 3709 (int.)

ISBN 978-951-652-161-2ISSN 1239-3096The conclusions and viewpoints presented in the report arethose of author(s) and do not necessarily coincidewith those of Posiva.

Posiva-raportti – Posiva ReportPosiva OyOlkiluotoFI-27160 EURAJOKI, FINLANDPuh. 02-8372 (31) – Int. Tel. +358 2 8372 (31)Raportin tunnus – Report codePOSIVA 2008-02Julkaisuaika – DateFebruary 2008Tekijä(t) – Author(s)Karsten Pedersen,Microbial Analytics Sweden ABToimeksiantaja(t) – Commissioned byPosiva OyNimeke – TitleMICROBIOLOGY OF OLKILUOTO GROUNDWATER, 2004-2006Tiivistelmä – AbstractThe microbiology of shallow and deep groundwater in Olkiluoto, Finland, was analysed for almost threeyears from 2004 to 2006. The extensive sampling and analysis programme produced a substantial database,including 60 analytical datasets on the microbiology of Olkiluoto groundwater, which is described andinterpreted here. One part of this database comprises 39 complete analytical datasets on microbiology,chemistry, and dissolved gas composition assembled on four sampling campaigns from measurements from16 shallow observation tubes and boreholes ranging in depth from 3.5 to 24.5 m. The second part of thedatabase contains 21 datasets on microbiology and chemistry covering 13 deep boreholes ranging in depthfrom 35 to 450 m. In addition, the database contains 33 completed analyses of gas covering 14 deepboreholes ranging in depth from 40 to 742 m. Most of these analyses were completed before the onset ofONKALO construction, and the remaining samples were collected before ONKALO construction hadextended below a depth of 100 m; therefore, this dataset captures the undisturbed conditions before thebuilding of ONKALO.Shallow groundwater in Olkiluoto contained dissolved oxygen at approximately 10% or less of saturation.The presence of aerobic and anaerobic microorganisms, including methane-oxidizing bacteria, has beendocumented. The data confirm earlier suggested processes of oxygen reduction in the shallow part of thebedrock. These microbial processes reduce intruding oxygen in the shallow groundwater using dissolvedorganic carbon and methane as the main electron donors. Microbiological and geochemical data stronglysuggest that the anaerobic microbial oxidation of methane (ANME) is active at a depth down toapproximately 300 m in Olkiluoto, as has been suggested previously, based on interpretations ofgeochemical data. However, proof of the presence and activity of ANME microorganisms is needed beforethe existence of active ANME processes in Olkiluoto groundwater can be accepted. It appears as thoughANME is limited to the 0–300 m depth interval due to a lack of sulphate at depths below 300 m. Thisimplies that the rate of sulphide production by ANME processes at depths of 300 m and above is limited bythe rate of methane transport from deeper layers.The construction of ONKALO will probably influence the ANME processes. These processes may,therefore, need detailed modelling when there are applicable data regarding how the ANME processesreact to the construction of ONKALO. Future sampling and analysis will reveal whether ONKALOconstruction has influenced biogeochemical conditions in the surrounding groundwater. If such aninfluence is found, it will, hopefully, be possible to model the underlying reasons for this influence and topredict its continuation, based on the obtained data.Avainsanat - KeywordsATP, bacteria, dissolved gas, methanogens, microorganisms, oxygen, shallow groundwater, sulphatereducingbacteriaISBNISBN 978-951-652-161-2Sivumäärä – Number of pages156ISSNKieli – LanguageISSN 1239-3096English

Posiva-raportti – Posiva ReportPosiva OyOlkiluotoFI-27160 EURAJOKI, FINLANDPuh. 02-8372 (31) – Int. Tel. +358 2 8372 (31)Raportin tunnus – Report codePOSIVA 2008-02Julkaisuaika – DateHelmikuu 2008Tekijä(t) – Author(s)Karsten Pedersen,Microbial Analytics Sweden ABToimeksiantaja(t) – Commissioned byPosiva OyNimeke – TitleOLKILUODON POHJAVEDEN MIKROBIOLOGIA, 2004-2006Tiivistelmä – AbstractOlkiluodossa on tutkittu jo vuodesta 2004 alkaen matalien ja syvien pohjavesien mikrobiologiaa. Vuosien2004-2006 laaja näytteenotto- ja analyysiohjelma on tuottanut huomattavan määrän tuloksia, joitakäsitellään ja tulkitaan tässä raportissa. Ensimmäinen osa pitää sisällään 39 perinpohjaista mikrobiologista,kemiallista ja kaasuanalyysia, jotka ovat peräisin neljästä eri matalien kallioreikien japohjavesiputkien näytteenottokampanjasta. Kampanjat toteutettiin yhteensä 16:sta matalasta kallioreiästäsekä pohjavesiputkesta, joiden syvyydet vaihtelivat 3,5 - 24,5 m välillä. Toinen osa pitää sisällään 21mikrobiologista ja kemiallista näytettä, 13:sta eri syvyisestä kairanreiästä, syvyysväliltä 35-450 m. Lisäksikuvaillaan ja käsitellään 33 kaasuanalyysia, jotka on otettu 14:sta kairanreiästä syvyysväliltä 40-742 m.Suurin osa analyyseistä tehtiin ennen kuin ONKALOn rakentaminen alkoi ja loput ennen kuin ONKALOsaavutti 100 m syvyyden, minkä johdosta tulokset edustavat Olkiluodon mikrobiologista luonnontilaa.Matala pohjavesi Olkiluodossa sisältää liuennutta happea noin 10 %:a tai vähemmän. Lisäksi aerobisia jaanaerobisia mikro-organismeja (ml. metaania hapettavat bakteerit) on havaittu. Tämän tutkimuksentulokset tukevat aiempaa käsitystä hapen kulumisesta aivan kallion yläosassa ja maaperässä. Raportissakäsitellyt mikrobiologiset prosessit pelkistävät happea käyttämällä liuennutta orgaanista hiiltä ja metaaniapääasiallisena elektronien luovuttajina. Mikrobiologinen ja hydrogeokemiallinen data viittaa siihen, ettämetaanin anaerobinen mikrobinen hapettuminen (ANME) Olkiluodossa on aktiivista noin 300 msyvyyteen saakka. Tähän on viitattu myös aiemmin perustuen hydrogeokemiallisen datan tulkintaan.Ennen kuin ANME prosessien esiintyminen Olkiluodossa voidaan hyväksyä, tarvitaan todisteita ANMEmikro-organismien olemassaolosta ja aktiivisuudesta. Tämän tutkimuksen perusteella voidaan sanoa, ettäANME prosessit ovat rajoittuneet 0-300 m syvyysvälille, koska sulfaattia ei esiinny tarpeeksi -300 malapuolella. Tämä viittaa siihen, että sulfidin tuottoa rajoittaa syvemmältä kallioperästä kulkeutuvametaanin määrä.ONKALOn rakentaminen tulee todennäköisesti vaikuttamaan ANME prosesseihin. Nämä prosessit voivattarvita yksityiskohtaista mallintamista. Tulevaisuuden näytteenotot ja analyysit selventävät, onkoONKALOn rakentaminen vaikuttanut ympäröivien pohjavesien biogeokemiallisiin olosuhteisiin. Josmuutoksia prosesseissa havaitaan, niiden syyt pyritään mallintamaan ja mahdollinen jatkuvuusennustamaan olemassa olevan datan perusteella.Avainsanat - KeywordsATP, bakteerit, liuenneet kaasut, metanogeenit, mikro-organismit, happi, matala pohjavesi, sulfaattiapelkistävä bakteeriISBNISBN 978-951-652-161-2Sivumäärä – Number of pages156ISSNKieli – LanguageISSN 1239-3096Englanti

1TABLE OF CONTENTSABSTRACTTIIVISTELMÄPREFACE ................................................................................................................ 51 INTRODUCTION .................................................................................................. 71.1 Research, development, and technical design programme: TKS-2003.......... 71.1.1 TKS-2003 Geogases and microbes present at Olkiluoto .......................... 71.1.2 TKS-2006 Hydrogeochemistry ............................................................... 91.2 This work......................................................................................................... 91.3 Microbes – what are they?............................................................................ 101.3.1 Bacteria................................................................................................... 111.3.2 Archaea................................................................................................... 121.3.3 Unicellular fungi....................................................................................... 131.3.4 Unicellular animals .................................................................................. 141.3.5 Unicellular photosynthetic organisms...................................................... 141.3.6 Viruses .................................................................................................... 181.4 Microbial processes ...................................................................................... 201.4.1 Closed systems....................................................................................... 211.4.2 Open systems ......................................................................................... 221.4.3 Microbial oxidation–reduction processes – “behind the scenes”............. 241.5 The microbe’s dilemma – death or survival .................................................. 262 MATERIALS AND METHODS ............................................................................ 272.1 Sampling groundwater from shallow observation tubes and boreholes........ 272.1.1 Sampling point descriptions .................................................................... 272.1.2 Packer control tests with nitrogen ........................................................... 272.1.3 Sterilization of borehole pumps............................................................... 282.1.4 Test for reproducibility of groundwater chemistry and microbiology overtime ......................................................................................................... 292.1.5 Sample collection for microbiological analyses....................................... 292.2 Sampling groundwater from deep boreholes ................................................ 292.2.1 Packer-equipped deep boreholes sampled for microbiology .................. 302.2.2 Sampling, transport, and extraction of deep groundwater samples ........ 302.3 Physical parameters, chemistry, and gas content of the sampledgroundwater .................................................................................................. 322.3.1 Field measurements of physical parameters in shallow boreholes andgroundwater observation tubes............................................................... 322.3.2 Analysis of dissolved oxygen in shallow groundwater using Winklertitration .................................................................................................... 332.3.3 Chemical analyses of shallow and deep groundwater ............................ 332.3.4 Sampling and analysis of dissolved gas ................................................. 342.4 Microbiological analyses ............................................................................... 362.4.1 Determining total number of cells............................................................ 362.4.2 ATP analysis ........................................................................................... 362.4.3 Determining cultivable aerobic bacteria .................................................. 372.4.4 Preparing media for most probable numbers of cultivable anaerobicmicroorganisms....................................................................................... 38

22.4.5 Inoculations and analysis for anaerobic microorganisms........................ 382.4.6 Inoculations and analysis for aerobic methane-oxidizing bacteria .......... 392.4.7 Quality controls for the most probable number analysis ......................... 423 RESULTS ........................................................................................................... 433.1 Analysis of physical and chemical parameters ............................................. 433.1.1 Field measurements of physical parameters .......................................... 433.1.2 Chemical analyses of groundwater ......................................................... 443.2 Sampling, extraction, and analysis of gas..................................................... 493.2.1 Dissolved gas in shallow groundwater – comments on the methods...... 493.2.2 Dissolved gas in deep groundwater – comments on the methods.......... 493.2.3 Distribution of gases in Olkiluoto groundwater........................................ 513.3 Analysis of biological parameters ................................................................. 603.3.1 Sterilization of borehole pumps............................................................... 603.3.2 Comparison of sampling using the SOLINST sampler and using theborehole pump ........................................................................................ 613.3.3 Test for reproducibility of groundwater microbiology over time............... 613.3.4 Tests for reproducibility of the pressure vessel method.......................... 613.3.5 Biomass determinations.......................................................................... 623.3.6 Cultivable heterotrophic aerobic bacteria................................................ 633.3.7 Most probable number of metabolic groups of bacteria .......................... 644 DISCUSSION...................................................................................................... 754.1 Sampling procedures for shallow groundwater............................................. 754.1.1 Selection of sampled shallow groundwater boreholes ............................ 754.1.2 Sampling of shallow groundwater ........................................................... 764.1.3 The oxygen blockage packer test ........................................................... 774.1.4 Sterilization of borehole pumps............................................................... 774.1.5 Comparison of sampling using the SOLINST sampler and using theborehole pump ........................................................................................ 774.2 Sampling procedures for deep groundwater microbiology............................ 784.3 Evaluating the analysis methods .................................................................. 794.3.1 Analysis of physical parameters.............................................................. 804.3.2 Chemical parameters .............................................................................. 804.3.3 Microbiological parameters ..................................................................... 814.4 Geochemical conditions of the investigated aquifers.................................... 864.4.1 Physical parameters................................................................................ 874.4.2 Chemistry dissolved solids................................................................... 884.4.3 Origins and amounts of dissolved gases in Olkiluoto groundwater......... 924.5 Specialists, generalists, opportunists, and antagonists in the world ofmicrobes ....................................................................................................... 954.6 Microbial processes in shallow groundwater ................................................ 964.6.1 Aerobic processes................................................................................... 974.6.2 Anaerobic processes............................................................................... 984.7 Microbial processes in deep groundwater .................................................... 984.7.1 Aerobic processes................................................................................... 984.7.2 Anaerobic processes............................................................................. 1004.8 Relevance of microbiological processes to ONKALO................................. 1044.8.1 Oxygen reduction and maintenance of anoxic and reduced conditions 1044.8.2 Bio-corrosion of construction materials ................................................. 1054.8.3 Bio-mobilization and bio-immobilization of radionuclides, and the effectsof microbial metabolism on radionuclide mobility.................................. 106

3REFERENCES ........................................................................................................... 107A. APPENDIX........................................................................................................ 115

71 INTRODUCTIONThe subsurface biosphere of Earth appears to be far more extensive and metabolicallyand phylogenetically complex than previously thought (Amend and Teske 2005). Adiverse suite of subsurface environments has been reported to support microbialecosystems, extending from a few meters below the surface to thousands of metersunderground (Pedersen 2000a, 2001). The discovery of a deep biosphere (Pedersen1993) will have several important implications for underground repositories for spentradioactive wastes (Pedersen 2002). The main effects of microorganisms in the contextof a KBS-3 type repository (Anonymous 1983) for radioactive waste in the bedrock ofOlkiluoto are:Oxygen reduction and maintenance of anoxic and reduced conditionsBio-corrosion of construction materialsBio-mobilization and bio-immobilization of radionuclides, and the effects ofmicrobial metabolism on radionuclide mobility1.1 Research, development, and technical design programme: TKS-2003Because of the potentially important effects of microorganisms, as listed above,microbiology research initiatives form part of both the Finnish and Swedish radioactivewaste disposal programmes. The first comprehensive Swedish state-of-the-art report onmicrobiology in radioactive waste disposal was published in 1995 (Pedersen andKarlsson 1995). Sweden, unlike Finland, has not yet selected a disposal site, so theSwedish programme has mainly attempted to build our understanding of microbialprocesses in general. Relevant microbiology research in Finland, on the other hand, cannow be more site related, because the disposal site, Olkiluoto, has been selected. TheFinnish programme started extensive microbiological site-related investigations inOlkiluoto in 2004, with the aims set forth in the research (tutkimus), development(kehitys), and technical design (suunnittelu) (TKS) programme (Posiva Oy 2003). Thisprogramme summarized previous work and presented the current (as of 2003) model ofmicrobiology and geogases in Olkiluoto groundwater. The TKS-2003 section of theprogramme, dealing with microbes and geogas, is briefly summarized below; thissection initiated the microbiological investigations that are covered in the present report.1.1.1 TKS-2003 Geogases and microbes present at OlkiluotoTKS-2003 supplied the following background to the microbiological programme inOlkiluoto: Microbes were found in all groundwater studied in the Finnish site selectioninvestigations performed from 1996 to 2000 at depths of between 60 and 900 m(Haveman et al. 1998, 2000). Sulphate-reducing bacteria (SRB) were the most abundantspecies found in the Olkiluoto groundwater (at depths of 200 m and below), and tendedto be associated with groundwater at an intermediate depth range of approximately 250–330 m (Table 1-1). The deeper, saline groundwater (below 400 m) contained very small

8Table 1-1. Total number of cells (TNC) and the most probable number of metabolicgroups of microorganisms in Olkiluoto groundwater sampled from 1996 to 2000. IRB =iron-reducing bacteria, SRB = sulphate-reducing bacteria, AA = autotrophicacetogens, HA = heterotrophic acetogens, AM = autotrophic methanogens, and HM =heterotrophic methanogens.BoreholeDepth(m)cells mL 1TNC IRB SRB AA HA AM HMOL-KR3 243–253 510000 1500 >16000 7.8 330 - a -OL-KR8 302–310 280000 NT b 16000 - - - -OL-KR10 324–332 650000 7 >16000 22 9200 0.45 0.45OL-KR3 438–443 700000 460 420 - 930 - -OL-KR9 470–475 150000 33 92 - 110 - -OL-KR9 563–571 620000 NT 1.7 - - - -OL-KR4 861–866 170000 - - - 4.9 - -a Below detection limit (0.2 cells mL 1 ).b Not tested.amounts of SRB and iron-reducing bacteria (IRB). The populations of SRB and IRBseemed to be high, particularly in the transition zone between sulphate-rich andsulphate-poor groundwater, in which E h conditions changed from sulphidic to methanic.Results of the earlier preliminary investigation phases at Olkiluoto indicated that thesaline groundwater contained massive amounts of dissolved gases (despite the fact thatthe sampling techniques were not very representative). The amounts of dissolved gases,such as methane and hydrogen, were high, especially in the deep saline groundwater,and some of the saline groundwater samples contained methane close to the saturationlimit (Gascoyne 2000). The total content of dissolved gases displayed a fairly coherentincreasing trend with depth, indicating that the current gas sampling system wasrelatively reliable (Pitkänen et al. 2003). Large variations were also observable in singlesamples, for example, in the results for the deep samples from borehole KR4 at a depthof 860 m (900 and 1900 mL L 1 ), reflecting uncertainty in the quantitative results. Themain reason for uncertainty was considered to be the variable amount of waterrecovered during sampling (Gascoyne 2000). Isotopic and chemical data suggested thatbacterial, thermogenic, and abiogenic formation were all potential mechanisms forhydrocarbon (HC) formation (Pitkänen et al. 2003). Microbial analysis by Haveman etal. (1998, 2000) also suggested ongoing methanogenesis occurring below the sulphaterichzone, as indicated by a few low 13 CH4 values. Methane concentrations were severalhundreds of mL L 1 in deep saline groundwater at Olkiluoto. Bacterial methaneformation was evident deep in the bedrock, but insufficient isotopic data on dissolved

9inorganic carbon (DIC) and hydrocarbons impede detailed evaluation of the magnitudeof methanogenesis and its effect on the carbonate system. The calculations suggested alevel of few mL L 1 for bacterial methane production (Pitkänen et al. 2003). Thehydrocarbon data indicated that the principal sources of methane and otherhydrocarbons were thermal processes. However, it was unclear whether thesehydrocarbons were formed by the thermal decomposition of organic matter or byhydrothermal reactions between carbonate or graphite and hydrogen.1.1.2 TKS-2006 HydrogeochemistryThe TKS report, TKS-2006 (Posiva Oy 2006), described the plans for continuedmicrobiology research, and some of the findings of this work are reported here.Microbial processes play important roles in aerobic respiration, methane formation, andsulphate reduction in Olkiluoto groundwater (Andersson et al. 2007b). The results of themicrobiological studies carried out between 2004 and 2006 are presented in the presentreport. Microbiological analysis will continue as part of the sampling campaigns inselected deep boreholes, i.e., as part of the gas investigations. Samples will also betaken from shallow boreholes and groundwater observation tubes every second yearstarting in 2008, to check whether construction has caused any changes in microbeconcentrations. One important aim of the research is to investigate whether constructionat the ONKALO site has influenced the microbiological populations and their activity atdepth. From the point of view of long-term safety, sulphate reduction could harmcopper canisters, and it is particularly important to obtain information on the activity ofsulphate reducers in groundwater close to the disposal depth.The gas data have been further evaluated (Pitkänen and Partamies 2007), and the resultssuggest a need to obtain additional data using improved sampling techniques andanalysis methods, especially from the repository depths and below. In particular,methane formation is an issue that must be evaluated. Gas will continue to be sampledas part of the ongoing monitoring programme and from new boreholes. Special attentionwill be paid to the quality of the gas analyses (e.g., by preventing contamination of thesamples with air) and to the possibility of obtaining additional isotope data (e.g.,regarding helium and hydrogen isotopes) from the gases.1.2 This workAt Olkiluoto, investigations to establish the baselines for subsurface geochemical(Pitkänen et al. 2007) and microbial conditions were performed during the 1996–2000site investigation period (Haveman et al. 1998, 1999, 2000; Haveman and Pedersen2002a). Since then, a new series of deep groundwater samples has been collected from21 sections distributed among 13 deep Olkiluoto boreholes, using the PAVE pressuresampling vessel according to the method of Haveman et al. (1999), and analysed. Thesenew deep groundwater samples were collected over the two years from 10 October 2004to 28 November 2006 from depths of 34 to 900 m. To fill gaps in our knowledge of theshallow groundwater environment, a series of investigations of shallow boreholes inOlkiluoto was performed concurrently with the new Olkiluoto deep groundwater

10investigations. Samples were collected on four different occasions from 16 shallowboreholes ranging in depth from 4 to 24.5 m. The sampling periods were 36 May 2004,1014 October 2005, 2428 April 2006, and 9–13 October 2006. The results of the firsttwo investigations were reported in Posiva working reports (Pedersen 2006, 2007). Allthe results obtained from the deep and shallow groundwater investigations from 2004 to2006 have now been merged and interpreted, and the outcome is reported here.As a guide for readers not so familiar with the science of microbiology, I will firstbriefly introduce the microbial world. The general textbook on microbiology, BrockBiology of Microorganisms (Madigan et al. 2006), is recommended for those who wishto deepen their knowledge of microorganisms.1.3 Microbes – what are they?A microbe is a living entity that contains all functions needed to perform a life cycle,such as feeding, growth, and reproduction, in a single cell. Microbe size variessignificantly, ranging from approximately 0.2 m in diameter in the smallest bacteriumto 1 mm or more in some unicellular animals and plants. The largest known bacterium isthe sulphur-oxidizing microbe Thiomargarita namibiensis, which reaches a maximumdiameter of 0.75 mm (Schulz et al. 1999).The tree of life, based on analysis of the gene 16/18S rRNA, is depicted in Figure 1-1; itdisplays the phylogenetic relationships between the main known and characterizedorganism groups found on Earth. The organisms cluster in three major domains, viz.Bacteria, Archaea, and Eukarya. All organisms in the domains Bacteria and Archaeaare microbes, and most branches of the domain Eukarya are microbial as well. In fact,multicellular organisms are only represented in the three branches comprising animals,plants, and fungi. Microbes can be found virtually everywhere in the tree of life,accounting for most of the diversity of life on our planet. Much microbial diversity isbiochemical, unlike multicellular life in which the diversity is largely morphological.The enormous biochemical diversity among the microbes explains their hugeadaptability to almost any environment on the planet where temperature allows life.Microbes are usually divided into five different groups, based mainly on a mix ofmorphological, biochemical, and molecular criteria. The most important criteria foreach of the five groups, and their relevance to a high-level radioactive waste (HLW)repository, are given below. Viruses constitute a sixth group of microbes that differfrom the other five in their total dependence on a host for reproduction. They cannot befitted into the molecular tree of life shown in Figure 1-1. Viruses display no signs of lifeoutside their host cells.

11Figure 1-1. The phylogenetic relationships between all main organism groups on theplanet can be revealed by comparing their 16S rDNA and 18S rDNA genes, coding forthe ribosomes, which are the protein factories of the cell (Woese et al. 1990). Redrepresents microbes adapted to high temperatures (60–113C), many of which utilizehydrogen as a source of energy. Yellow represents microbes that can live in saturatedsalt solutions (25–30% NaCl). Green represents the proteobacteria, the group thatincludes many microbes found in the Fennoscandian Shield aquifers. Methanogensliving at low or intermediate temperatures (0–60C) appear in light blue; theseconstitute an important group in most underground environments. The bulk of thedomain Bacteria is indicated in blue while the domain Eukarya is indicated in lightbrown.1.3.1 BacteriaA typical bacterium is a very robust organism that generally survives extremely well inthe niche for which it is adapted. It is isolated from its surrounding environment by acell membrane (Figure 1-2) and a cell wall. The sack-like cell membrane containsvarious structures and chemicals that allow the bacterium to function. Key structures arethe nucleotides and the genetic code (DNA), which store information needed for cellfunction, and the cytoplasm, which contains the machinery of cell growth and function.Bacteria are adapted to various conditions and, as a group, the bacteria can handle allpossible combinations of environmental conditions. This is reflected in the speciesdiversity of the domain Bacteria (Figure 1-1), which comprises many millions of

12species, as reflected by environmental ribosomal rDNA sequencing (Pace 1997).Approximately 10,000–15,000 of these microbes have been characterized (Dworkin etal. 2007); the rest remain molecular imprints on the environment of organisms, imprintsthat await exploration and characterization. This vast diversity of unknown speciesrepresents uncertainty with respect to unknown microbial processes that might beimportant for nuclear waste disposal. One obviously undesired species, for example,would be one that would, under repository conditions, produce large quantities ofradionuclide-chelating agents. In contrast, anaerobic methane oxidisers would be verybeneficial, as they would help keep the groundwater redox potential (E h ) at a low andnegative value.There appear to be several overriding characteristics that unify many of the mainbranches of the domain Bacteria (Figure 1-1). The ability to photosynthesize is a typicalcharacteristic of green bacteria (cyanobacteria) and some proteobacteria; because oftheir need for light, these groups are not naturally represented in groundwater. Someother groups are also naturally absent, such as the pathogenic microbes (e.g.,Chlamydia) and all obligate parasitic microbes (mostly among the proteobacteria) thatgenerally require a multicellular host. Representatives of the remaining branches havebeen reported in various underground environments (e.g., Amend and Teske 2005).Fennoscandian Shield rocks are generally cold to moderately warm for the first 2 km ofdepth. The rock temperature at repository depths is some 15–20C, so thermophilic (i.e.,heat-loving) organisms will not be common there before waste disposal. It is uncertainto what extent thermophilic Bacteria and Archaea will invade and/or multiply in arepository area in which the temperature will fall from 80C to 50C over the first 3000years. They certainly can be found active in all naturally occurring high-temperaturegroundwater. The consensus today is that thermophiles will appear in significantnumbers in a warm repository.1.3.2 ArchaeaMicroorganisms in the domain Archaea (Figure 1-1) were regarded as bacteria untilmolecular data revealed that they belong to a domain that differs completely from thoseof all bacteria and all plants, animals, and fungi. A unifying characteristic of organismsin this domain is their ability to adapt to what are called “extreme conditions”. Differentspecies of Archaea are active under different conditions. Some Archaea like very hotconditions (Stetter 1996). For example, the optimum temperature for the growth of thegenus Pyrolobus is 105C and it survives in temperatures of up to 113C. Remarkably,this species “freezes” to death when the temperature goes below below 90C. Manyother genera of Archaea grow best at approximately 100C. The temperature of theHLW repository will consequently not exceed the temperature range within which lifecan exist. Some genera of Archaea are adapted to extreme pH levels, as low as 1 or upto 12, and some may even survive at more extreme pH levels (Pedersen et al. 2004). Agroup that is important for an HLW repository is the methanogens (Figure 1-3, Figure1-4); these produce methane gas from hydrogen and carbon dioxide, or from short-chainorganic carbon compounds, such as formate, methanol, or acetate.

13Figure 1-2. A cross-section of the bacterium Gallionella ferruginea (Hallbeck andPedersen 2005) produced using a transmission electron microscope (TEM). Thismicrobe is very common in groundwater seeps on the walls and floors and in ponds inthe Äspö Hard Rock Laboratory (HRL) tunnel in Sweden (Anderson and Pedersen2003). The cell wall gives the microbe its form and rigidity, while the cell membranecontrols the transport of nutrients into and wastes out of the cell. The nucleic acid DNAconstitutes much of the interior of the cell and carries information necessary for cellfunction and reproduction. This organism is a chemolithotroph that uses ferrous iron(Fe 2+ ) as a source of energy. This energy is used to reduce carbon dioxide to cellcarbon constituents, just as photosynthetic plants do, but using iron energy instead ofsolar energy. The visible structures in this cell do not look very different from those of abacterium that uses organic carbon as a source of energy and for building cellconstituents. The differences are almost completely on the molecular, biochemicalscale, a scale that is not resolved by the TEM. The diameter of this cell is approximately1 m (photograph: Lena Bågenholm and Lotta Hallbeck).1.3.3 Unicellular fungiThe fungi belong to the domain Eukarya (Figure 1-1) and represent great morphologicaland biochemical diversity. There are data in the scientific literature that demonstratefungi to be natural inhabitants of intra-terrestrial environments (Reitner et al. 2005). Theunicellular fungi include yeast, which can commonly ferment many different organiccompounds to form carbon dioxide, organic acids, alcohols, and hydrogen. Some of

14these organic acids, for example, citric acid, are excellent chelating agents and aretherefore undesired in a repository in the case of a canister failure. Mould is anothergroup of fungi regarded as unicellular, despite their ability to form multicellular mycelia(i.e., networks of threads); each cell in a mycelium is capable of a complete life cycleand therefore falls into the microbe category. Some yeasts are capable of performinganaerobic metabolism (i.e., of living without oxygen) and are small, typically no biggerthan a few m or more, which makes them suitable for life in the narrow aquifers ofhard rock. Recent investigations of groundwater from the Äspö Hard Rock Laboratory(HRL) in Sweden (Ekendahl et al. 2003) identify yeast as a natural part of thesubterranean biosphere in Fennoscandian Shield igneous rock aquifers (Figure 1-5).This finding introduces uncertainty regarding repository performance with respect tofungal chelating agents and their influence on radionuclide migration.1.3.4 Unicellular animalsUnicellular animals belong to the domain Eukarya. They are found in all taxonomicbranches except the fungi and plant branches (Figure 1-1). Their natural presence indeep groundwater remains to be established. Some unicellular animals, particularly theflagellates, are so small (a few m) that they are difficult to distinguish from largebacteria and yeasts. Their obvious function in deep groundwater ecosystems would beas grazers of other microbes (Figure 1-6). Many unicellular animals feed on organismsof the domains Bacteria and Archaea.1.3.5 Unicellular photosynthetic organismsUnicellular photosynthetic microbes are found in several branches of the domainBacteria and also in the plant branch of the domain Eukarya (Figure 1-1). The domainArchaea does not contain any known true photosynthetic organisms. The process ofphotosynthesis requires light (Figure 1-6), which is not available underground, except inartificially illuminated vaults and tunnels. Mosses, cyanobacteria, and some otherphotosynthetic organisms have been observed in the Äspö HRL tunnel and willcertainly occur where there is light in a repository during the open phase. Theseorganisms fix carbon dioxide as organic carbon and therefore add some organicsubstances to the repository environment. Their activity in open deposition tunnels,however, is not foreseen to interfere with the long-term performance of the HLWrepository.

15Growth rate (h -1 )0.200.150.100.05AGrowth rate (h -1 )0.250.200.150.10C00.050 10 20 30 40 50 60Incubation temperature (°C)Figure 1-3. Methanobacteriumsubterraneum is a genus of Archaeaisolated from the Äspö HRL aquifersand characterized (Kotelnikova et al.1998). Its temperature (A), pH (B), andsalt (C) requirements include values ofthese parameters typical in therepository. This species is thus likely tobe an important inhabitant of therepository when the temperature isbelow 50C.Growth rate (h -1 )00.500.400.300.200.1000 0.25 0.50 0.75 1.00 1.25 1.5NaCl (M)_B6 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10pHFigure 1-4. Autofluorescent image of Methanobacterium subterraneum. Methanogenscontain a unique molecule, coenzyme F 420 , that takes part in methane formation. Themore active the methanogenesis of the cells, the more F 420 is present. This moleculefluoresces turquoise when irradiated with ultraviolet light. The methanogens on theimage were consequently very active in the pure culture from which this specimen came.

16Figure 1-5. Scanning electron micrographs of yeast strains isolated from Äspö HRL(Ekendahl et al. 2003). Using sterile syringes and needles, groundwater was sampleddirectly from fractures and boreholes and placed in appropriate culturing media.Growth of yeast and fungi occurred frequently. The isolated yeasts depicted wereunique, representing new species having growth demands that correlated with theenvironmental conditions in groundwater at the repository depth of 500 m. The strainsshown are: a) strain J1 (enlargement 8000, bar = 2 m), b) strain J2 (9000, bar = 2m), c) J3 (enlargement 8000, bar = 2 m), d) strain C (9000, bar = 1 m, arrowshows typical bud scar), and e) strain 5e (6750, bar = 2 m; the arrow indicatesexopolymeric material). Images are reproduced from Ekendahl et al. (2003).

17Figure 1-6. Unicellular plants and animals are also microbes. All functions they needin order to live are contained in a single cell. The yellow-brown diatom in the imagemakes dextrose out of light, water, and carbon dioxide. The cell is not entirelywatertight, and some of the sugar leaks out between the shell halves. Bacteria sense thisand gather around the diatom to consume the crumbs from the algal “dining table”.The little round object in the top left corner is a small, unicellular animal with twoflagella with which it swims. It swims fast and hunts bacteria to eat them. The big blobin the bottom left is an amoeba. An amoeba is an unicellular animal with a cellmembrane but no definite shape. It flows over surfaces as would a sack of potatoes; the“potatoes” push in the direction the amoeba wants to move, thereby rolling the wholesack in the desired direction. It typically engulfs bacteria, which it ingests (From:Pedersen 2004).

181.3.6 VirusesA virus is a non-cellular genetic element that uses living cells for its own reproduction.Viruses can be found in one of several different states. Outside cells, a virus is asubmicroscopic particle that contains a nucleic acid surrounded by a shell of proteinscalled a capsid. In this state, the virus is lifeless and does not carry out any biochemicalreactions. The main function of the capsid is to carry the genetic material, the nucleicacids, from one host cell to another. When the genetic material enters a new host cell,viral reproduction occurs. The genetic material of the virus takes over the cellmachinery and produces many new copies of the virus. When a virus enters and infectsa bacterium the result is usually disastrous for the bacterium. When the virus hasmultiplied, what is left of the cell lyses and breaks up and many new viruses are readyto search for a new host to infect and kill. This is called a lytic infection. Sometimes, thegenetic material of the virus can be incorporated into the genetic material of the hostcell. When the host cell multiplies, so does the genetic material of the virus. Thisprocess is called a lysogenic infection.Viruses that infect microorganisms have been found around the world, including insome of the most extreme environments on Earth, such as hot spring water (Rachel et al.2002), Antarctic lakes (Lisle and Priscu 2004), and deep-sea hydrothermal vent systems(Ortmann and Suttle 2005). The presence of prokaryotes in deep intra-terrestrial andsub-seafloor environments to a depth of at least 3.3 km has been established (Amendand Teske 2005; Lin et al. 2006), but viruses have so far not been reported. Previousstudies (Pedersen 2001) of groundwater from deep granitic aquifers revealedmicroorganisms in numbers of 10 4 to 10 6 cells mL –1 , which is at or above the lowerlimit for the replication of prokaryotic viruses (Wiggins and Alexander 1985). Anabundant diversity of viruses has recently been discovered in granitic groundwater fromdepths of 69 to 455 m in the Äspö HRL, Sweden. Fluorescent microscopy counts werein the range of 10 8 to 10 10 virus-like particles L –1 groundwater; these counts generallyexceeded the microbial counts by a factor of 10, which is a ratio typical of ecosystemscontaining active viruses and microorganisms in surface environments (Maranger andBird 1996; Suttle 2005). At concentrations of 10 10 virus-like particles L –1 groundwater,viruses contribute significantly to the pool of colloids. This effect has previously beenoverlooked, because of our ignorance of the presence of viruses in deep groundwater.Using transmission electron microscopy, four distinct main viral morphologies werefound in Äspö HRL groundwater encompassing polyhedral, tailed, filamentous, andpleomorphic forms that could be further divided into 12 distinct morphological subgroups,in accordance with recent assessments of prokaryotic virus diversity(Ackermann 2007). In addition, a tailed virus that infects the indigenous sulphatereducingbacterium Desulfovibrio aespoeensis (Motamedi and Pedersen 1998) wasisolated and subsequently detected in significant numbers in some groundwater samples(Figure 1-7). The presence of active lytic viruses in deep groundwater is a directindicator of virus–microbe, predator–prey interactions in intra-terrestrial ecosystems.The infection of microorganisms by viruses may contribute to the transfer of DNAbetween host cells, implying that viral transduction is important for the diversificationof intra-terrestrial microorganisms.

19Figure 1-7. The morphologies of viruses, isolated from deep groundwater, that werelytic for Desulfovibrio aespoeensis (Motamedi and Pedersen 1998) growing in amedium for sulphate-reducing bacteria, shown in transmission electron micrographs aand b. Images c and d show viruses (indicated by arrows) at the surface of a bacterium.Images were taken using 70,000 magnification in ac and 45,000 magnification in d.Images a and b are of a virus isolate denoted E, c of isolate D, and d of isolate B. Thescale bar represents 100 nm in ac and 500 nm in d. (Photograph: Hallgerd Eydal).The lack of large microbial biomass in intra-terrestrial environments has usually beentaken as evidence that any microorganisms present there are inactive or metabolizingextremely slowly (Kerr 2002). The new results regarding virus presence at the ÄspöHRL offer an alternative explanation of what viruses control active microbialpopulations in deep intra-terrestrial environments. Viruses encounter the cell walls oftheir hosts by chance, and attach to them before infection. The infection rate thusdepends on the numbers of both viruses and available hosts. As the number of hostsdecreases in response to lysis, the number of potential host cells also decreases, as willvirus replication and abundance. If the rate of microorganism growth equals the rate ofinfection and lysis, the overall number of microorganisms will remain within a rangedefined by the infectivity of the viruses.Viruses are completely dependent on active and growing host microorganisms for theirreproduction. The number of viruses has been demonstrated to be significantly relatedto bacterial turnover in samples from deep Mediterranean sediments (Danovaro et al.2002), to bacterial activity in sediments from Nivå Bay in Denmark (Middelboe et al.2003), and to the number of host cells in the Adriatic Sea aquatic system (Corinaldesi etal. 2003). High ratios of viral to bacterial numbers have been observed at the Äspö HRLand are indicative of viruses actively infecting microorganisms that must bemetabolically active. It confirms previously obtained energy source assimilation data(Pedersen and Ekendahl 1992a) and recent ATP analysis data (Eydal and Pedersen

202007), both of which suggested that the investigated microorganisms were in a state ofactive growth. Predator–prey relationships may be present in deep groundwatercontaining active and growing microorganisms, just as they are in many surfaceenvironments. However, as many intra-terrestrial environments are stagnant with low orno advective flow of water, intra-terrestrial microorganisms may be growth limited dueto low access to energy over time. The observed metabolic rates may thus be muchslower than in surface water, but the low numbers could be a result of predation ratherthan of starvation.1.4 Microbial processesMicrobiological decomposition and the production of organic material depend on theenergy sources and electron acceptors present (Madigan and Martinko 2006). Organiccarbon and methane and reduced inorganic molecules, including hydrogen, are possibleenergy sources in the repository environment. During the microbial oxidation of theseenergy sources, microbes preferentially use electron acceptors in a particular order (asdepicted in Figure 1-8): first oxygen, and thereafter nitrate, manganese, iron, sulphate,sulphur, and carbon dioxide are utilized. Simultaneously, fermentative processes supplythe metabolizing microorganisms with, for example, hydrogen and short-chain organicacids. As the solubility of oxygen in water is low, and because oxygen is the preferredelectron acceptor of many bacteria that utilize organic compounds in shallowgroundwater, anaerobic environments and processes usually dominate at depth in thesubterranean environment.The reduction of microbial electron acceptors may significantly alter the chemistry ofgroundwater. Dissolved nitrate is reduced to gaseous nitrogen, solid manganese and ironoxides are reduced to dissolved species, and the sulphur in sulphate is reduced tosulphide (Figure 1-8). In addition, the metabolic processes of some microorganismsproduce organic carbon, such as acetate, from the inorganic gases carbon dioxide andhydrogen, while other microorganisms produce methane from these gases; all theseprocesses generally lower the redox potential, E h . Most of these microbiologicallymediated reactions will not occur in a lifeless groundwater environment lacking thecascade of biochemical reactions going on inside the cell membranes (Figure 1-2) ofmicroorganisms. The mere presence of sulphide in a low-temperature graniticgroundwater provides indisputable evidence of microbiological sulphate reduction.However, concentrations of reduced electron acceptors alone will not reveal when,where, and at what rate the individual microbial processes take place. Hence, robust,sound, and reproducible methods for estimating the rate at which microbial processesoccur have had to be applied. Methods for analysing microbial process rates have beendeveloped and tested under open and closed in situ conditions in the Äspö HRL situated450 m underground. Groundwater that contained microorganisms, coming from afracture adjacent to the laboratory, was circulated under in situ pressure and chemistryvia flow cells that mimicked the conditions of fractured rock. The focus wasdetermining the rate of the reduction of sulphate to sulphide and the rate of theproduction of acetate from hydrogen and carbon dioxide. Changing from an open to aclosed system resulted in significant changes in the biogeochemistry (Hallbeck andPedersen 2008). The conceptual difference between closed and open systems, whichexplains this change, is presented next.

21HydrolysisO 2H 2 OOrganic polymersMonomersAerobic bacteriaCO 2NO 3 N 2HydrolysisOrganic acids,alcoholsOligo- andmonomersFermentativebacteriaAcetateH 2 + CO22DenitrifyingbacteriaManganesereducingbacteriaIron-reducingbacteriaSulphate-reducingbacteriaSulphur-reducingbacteriaMn 4+ Mn 2+Fe 3+ Fe 2+SO24 S 2S 0 S 2CO 2CO 2CO 2CO 2CO 2MethanogensCH 4Syntrophic bacteriaAcetogenic bacteriaH 2 + CO 2AcetateFigure 1-8. Possible pathways for the flow of carbon in the subterranean environment.Organic carbon is respired with oxygen, if present, or else fermentation and anaerobicrespiration occur with an array of different electron acceptors.1.4.1 Closed systemsThe usual way to culture microorganisms in the laboratory is by using batch cultures. Aculture vessel is supplied with all constituents necessary for growth, and is inoculatedwith the microbe of interest. A typical batch growth curve can be registered (Figure1-9). First, there is an adaptation phase during which the cells adjust to the conditions inthe culture vessel. Then the cells start to divide and grow exponentially to high counts,doubling their number over constant time intervals. Finally, growth is arrested whensome limiting component is used up, or when a toxic component is formed andaccumulates to too high a concentration (e.g., alcohol, in fermentation cultures). Figure1-9 indicates that the cells are basically active only during the exponential growthphase. The batch culture represents a closed system with no input or output ofcomponents. It is a superb tool for many research purposes in the laboratory, but it doesnot mimic the life of microbes in natural environments. The environment generallyconsists of a huge number of open systems with continuous input and output of matterbetween them. Models of microbial processes in the repository should therefore bebased on continuous culture conditions, as described below, rather than on batch cultureconditions.

2210 000 000Living cellsper ml1 000 000Stationary phaseDeclination phase100 00010 000Log phaseRelative activityper cell1 00010011001TimeFigure 1-9. A schematic representation of microbial growth in a closed batch culture.The microbes are basically active only during the exponential growth phase, when theydouble in number within specific time periods. The doubling time can be as short as 15min for some easily cultivated microbes or may be many hours for more recalcitrantmicrobes.1.4.2 Open systemsHard rock aquifers can be considered open systems. A particular fracture will containwater of a composition that reflects the origin of the water and the various reactionsbetween the solid and liquid phases occurring along the flow path. A new compositionmay be the result of two fractures meeting and their waters mixing. Though theseprocesses may be slow, there is a continuum of varying geochemical conditions in hardrock aquifers at repository depth, and the repository, with all the alien substances addedby construction, will add variance to these conditions. Microbes are experts at utilizingany energy in the environment that becomes thermodynamically available forbiochemical reactions. A slow but steady flow of organic carbon from the surface or aflow of reduced gases, such as hydrogen and methane from the interior of the planet orhydrogen from iron corrosion processes, will ultimately be the driving forces of theactive life of deep aquifer microbes in and around an HLW repository.

2310 000 000Living cellsper mlEnergy availabilityover time decreases1 000 000Energy availabilityover time increases100 00010 000Relative activityper cell1 00011001001TimeFigure 1-10. The graph is a schematic representation of microbial growth in an open,continuous culture system. The microbes are continuously active at a constant level,except for periods when there is a decrease in energy availability over time. Thedoubling time of the population can be very long and, if growth is counteracted by viralpredation, the numbers observed will remain relatively constant.The continuous growth of microbes can be studied in the laboratory using a chemostat,in which the culture vessel is continuously supplied with energy via a slow inflow ofnutrients. The inflow is balanced by an outflow that removes waste products and somecells. Though the number of microbe cells will therefore remain constant in thechemostat, the microbes that remain will be active (Figure 1-10). Unlike a batch system,a chemostat system is open, as it incorporates both an influx and outflow of matter. Thecontinuous culture conditions of the chemostat are applicable to any hard rock aquiferexperiencing a flux of matter through the continuous mixing of groundwater of varyingcompositions. Though the flows may be very slow in such aquifers, they will besignificant over geological time scales.The open, continuous culture system concept can be used when interpretingmicrobiology data for groundwater, such as the number of cells in a chosen groundwatermeasured at various times. If we apply the batch concept (Figure 1-9), we wouldconclude that the microbes are not growing and are inactive because we do not registerany increase in cell numbers over time. In contrast, with the continuous culture concept(Figure 1-10), it can be predicted that the microbes will be active and growing slowlyunder constant environmental conditions over the time period studied. This predictionrequires the existence of processes that counteract an increase in cell numbers due to

24growth. Viruses that attack and infect microbes (1.3.6) may neutralize cell growth. Theiractivity results in the lysis of infected cells and in the production of new viruses. Thisprocess, which occurs in most surface environments, has recently been found in thedeep aquifers of the Äspö HRL (1.3.6).A special case is the possible occurrence of microbes that grow attached to aquifersurfaces, a phenomenon repeatedly observed in groundwater from deep hard rockaquifers (Ekendahl and Pedersen 1994). Such biofilms will increase their cell numbersuntil they reach steady state, as previously described for the continuous growth ofunattached microbes. A comparison of the hypothetical cell numbers and activities ofattached versus unattached bacteria in a 0.1-mm wide fracture was previously done (seeTable 4.5 in Pedersen 2001). It demonstrated the potential importance of attached versusunattached microorganisms in underground environments. The studied microorganismsattached to artificial surfaces generally exhibited greater activity per cell than did theunattached microorganisms. Taken together with the cell numbers, there were up to fiveorders of magnitude more activity on the surfaces than in the groundwater. It is still anopen question whether attached bacteria are common and active on aquifer rocksurfaces under pristine conditions.1.4.3 Microbial oxidation–reduction processes – “behind the scenes”The biological oxidation–reduction processes presented in Figure 1-8 are commonlyrepresented by general stoichiometric summary reactions. However, the actualbiochemical reaction pathways are always much more complicated, often including acascade of biochemical enzyme-catalysed reactions inside the living cells that arestrictly controlled by the genetic code (i.e., DNA) of the individual cells. In addition,feedback and substrate-level control mechanisms may also be active. It is important tounderstand the biochemistry underlying summary reactions, otherwise the output ofbiological process models may be wrong. The reduction of the sulphur in sulphate tosulphide is used below to demonstrate the difference between a summary reaction andthe full biochemical process.Consider the summary equation for sulphate reduction with lactate as the electrondonor:2CH 3 CHOHCOO + SO 4 2 2CH 3 COO + 2CO 2 + 2H 2 O + S 2 (Eq. 1-1)The reaction would seem to indicate that lactate reacts directly with sulphate resulting inthe formation of acetate, carbon dioxide, and sulphide. This, however, is very far fromwhat actually happens. In fact, lactate and sulphate never make contact, but rather aredealt with by the bacterium in two separate biochemical pathways inside the cell (Figure1-11). Lactate is split into acetate and formate via pyruvate, and the formate is thenoxidized to carbon dioxide. This process involves three enzymes (i.e., lactatedehydrogenase, pyruvate formate lyase, and formate hydrogenolyase) and an oxidizedproton–electron transport molecule denoted nicotinamid adenine nucleotide (NAD + ).The electrons released are used as electron donors in reducing sulphate via a membraneboundrespiration chain. Consequently, the oxidation of the organic carbon is not

25Cell membraneLactateX4H 28H +Sulphate-reducingbacteriumoutside8eH 2-aseDNAinsideLactateLDHPyruvateGenetic controlAcetate + FormateCO 2+ H 2SO 42XH 2 SH +Cyt c 3Hmce FeSproteinATP2e SO 42ATP-sulfurylaseSO23ATP6e SulphitereductaseADPAPSH 2S(excreted)Figure 1-11. Electron transport and energy conservation in sulphate-reducing bacteria.In addition to external hydrogen (H 2 ), H 2 originating from the catabolism of organiccompounds such as lactate and pyruvate can fuel hydrogenase with electrons viaenzymes such as lactate dehydrogenase (LDH). The protons released by thehydrogenase feed the proton gradient across the cell membrane, and this gradientchanges ADP to ATP via ATP-ase. The enzymes hydrogenase (H 2 -ase), cytochrome c 3(cyt c 3 ), and a cytochrome complex (Hmc) are periplasmic proteins located between theouter and inner membranes of the cell. A separate protein functions to shuttle electronsacross the cytoplasmic cell membrane to a cytoplasmic iron-sulphur protein (FeS) thatsupplies the adenosine 5´-phosphosulphate (APS) reductase (forming SO 3 2 ) andsulphite reductase (forming H 2 S). The process of sulphate reduction is controlled by thegenetic code (i.e., DNA) of the bacterium. Environmental conditions are scanned bybacterial sensors that send messages to the DNA, which turns the sulphate reduction onand off under favourable and unfavourable conditions, respectively.directly connected to the reduction of the sulphate. This is an important point valid forall processes depicted in Figure 1-8. Sulphate reduction will occur only when the cellneeds to get rid of electrons that have generated a proton gradient across the cellmembrane. From this, it should be clear that the rate of sulphate reduction cannot bedetermined solely from concentrations of sulphate and lactate. Rather, the needs of thecell are what determines whether sulphate reduction will occur and at what rate. If, forexample, the cell lacks a crucial element needed to synthesize an enzyme involved insulphate reduction, the process will not proceed at all, even if there is plentiful lactateand sulphate in the environment around the cell.

26Turning to other processes, such as iron, manganese, and nitrate reduction, acetateformation, and methanogenesis, will reveal other biochemical pathways, each of whichis unique to the respective process. An endless array of more or less intricately linkedprocesses can be found by anyone looking into a microbiology textbook, such as BrockBiology of Microorganisms (Madigan and Martinko. 2006). A full understanding ofmicrobial oxidation–reduction processes is thus very complex. It becomes necessary todevelop model approximations and simplifications that do not violate the rules ofmicrobial biochemistry, otherwise the model output may be wrong. Later in the presentreport (4.7), the problems of understanding microbial biochemistry in groundwater willbe brought up in relation to the obtained results.1.5 The microbe’s dilemma – death or survivalIn periods of inactivity due to lack of energy and necessary nutrients, or due to otherenvironmental constraints, such as desiccation or slowly decreasing water activity,microbes can do one of two things: die or enter one of many possible dormancy states.Different species have different ways of addressing the problem of unfavourableconditions for active life. The most resistant form of survival is the endospore formedby certain gram-positive and sulphate-reducing bacteria. An endospore displays nomeasurable signs of life yet, after many years of inactivity, it can germinate into anactively growing cell within hours. It resists desiccation, radiation, heat, and aggressivechemicals far better than does the living cell.The endospore is the most resistant survival states of any known life form, but there aremany other survival strategies among the microbes, strategies more or less resistant toenvironmental constraints. Transforming into morphologically specific survival states isan advantage when the environment changes. However, in response to mere nutrient andenergy deficiency, many microbes simply shut down their metabolism to an absoluteminimum level at which they may survive for many years. Most such responses result inthe shrinkage of the cell to a fraction of its volume under optimal growth conditions.What all these survival strategies share is that the cell is active at an absolutely minimallevel – or displays no activity at all. It is thus possible that certain microbes may surviveinitially harsh conditions in a repository, including radiation, desiccation, heat, and highpH, until conditions for growth again become favourable. However, if the conditions areso difficult that all survival forms die off, and if the pore size of the environment doesnot allow for transport of microbes, as in highly compacted bentonite, then it is possiblethat specific environments in the repository may remain free of microbes once theoriginal microbe population has disappeared. It is at present uncertain whether this willindeed be the case.

272 MATERIALS AND METHODS2.1 Sampling groundwater from shallow observation tubes andboreholesSamples were collected on four different occasions from 16 shallow boreholes rangingin depth from 4 to 24.5 m (Table A-1). The sampling periods were 36 May 2004,1014 October 2005, 2428 April 2006, and 9–13 October 2006. Descriptions of thefirst two rounds of sampling and associated investigations have been published asPosiva working reports (Pedersen 2006, 2007). All sampling sites were pumped outusing an immersed borehole pump for at least 1.5 h prior to any field measurements orsample retrieval (Table A-1). The pump and tubing assembly was sterilized forapproximately 2 h in an 11-ppm chlorine dioxide solution (FreeBact 20; XINIX, Märsta,Sweden) in a 100-L plastic barrel. The pumps were soaked in the chlorine dioxidesolution and the solution was also pumped through the tubing.2.1.1 Sampling point descriptionsThe sites sampled at Olkiluoto (PVP1, PVP3A, PVP3B, PVP4A, PVP4B, PVP13,PVP14, PVP20, PR1, PP2, PP3, PP7, PP8, PP9, PP36, and PP39) penetratedgroundwater, which was present in either the overburden (PVP) or water-conductingfractures in the bedrock (PR, PP) (Figure 2-1, Table A-1). Several of the boreholes andtubes selected for the 2004 sampling campaign turned out to be problematic due to badpackers, collapsing rock, similar or little water (PVP3A, PVP3B, PVP4B, PP3, PP7, andPP8). These were abandoned and new sampling points were selected for the threesubsequent field campaigns. Overburden extended down to a depth of approximately 13m and is composed of sand and silt with an organic soil layer approximately 0.8 mthick. Bedrock groundwater samples extended to a depth of 24.5 m. The local bedrockat Olkiluoto is Precambrian, composed of metamorphic rocks (predominantlymigmatitic mica gneisses) and intruded by igneous rocks (granodiorites, coarse-grainedgranites, and granitic pegmatites). Local land use above the aquifers ranges fromundisturbed forest to open areas cleared for repository construction. Further details canbe found in the Posiva 2006 site description report (Andersson et al. 2007a).2.1.2 Packer control tests with nitrogenIn the spring 2004 field week (Pedersen 2006), oxygen was detected in all boreholes. Itwas impossible to determine whether the sampling procedure had introduced thisoxygen, or whether it was actually present in the groundwater that entered theboreholes. Therefore, an inflatable packer that allowed passage of sampling tubes andthe wire for the pump was constructed and tested in the fall 2005 field week. Nitrogenwas flushed, starting before the pumping, through half the number of boreholes thatwere being sampled at a rate of approximately 1 L of nitrogen gas per minute. In thatway, the gas above the water level was replaced with nitrogen while the groundwater

28Figure 2-1. Map showing the sampling points for shallow boreholes. Boreholes markedwith blue squares were only sampled in 2004. See text for details.level was being lowered due by the pumping (Table A-1). This prevented oxygen fromentering the sampled groundwater from the atmosphere in the borehole.2.1.3 Sterilization of borehole pumpsThe procedure for sterilizing the pumps and tubing was tested by pumping sterile waterthrough the equipment after a completed sterilization. The pumps were first soaked in achlorine dioxide solution for 2 h as described above. The barrel was then washed with500 mL of analytical grade water (AGW) (MilliQ-unit in the ONKALO laboratory).The adapter used for microbiology sampling was installed on the orifice of the pumptube. Two L of AGW water was then pumped through the system. Thereafter, 10 L ofAGW water was added to the barrel and pumped (1 L min 1 ) out through the samplingadapter. The adapter was removed after 2 min and the remaining 8 L of water werepumped out (8 L min 1 ). This procedure simulated the pumping out of a borehole beforestarting sampling (See Table A-1). The adapter was flushed with 1 L of sterile AGWwater and was then remounted. Finally, 5 L of sterile AGW water was pumped throughthe tubing, after which a full sampling for microbiology was performed according to theprocedures used in the field.

292.1.4 Test for reproducibility of groundwater chemistry and microbiology overtimePumping out a borehole results transports water from the surrounding aquifers outthrough the borehole. It was deemed important to test for sensitivity in the data obtainedfrom prolonged pumping. The borehole PVP4A had a yield of approximately 4 L min 1 ,and this borehole was selected for a reproducibility test in April 2006. It was sampledtwice within a 6-h interval, which allowed 1440 L of groundwater to be pumped out ofthe borehole between the sampling occasions (Table A-1).2.1.5 Sample collection for microbiological analysesGroundwater was collected in the spring 2004 and fall 2005 field weeks using a SolinstModel 425 Discrete Interval Sampler (Solinst Ltd., Georgetown, ON, Canada)immediately after the pumping period was finished and the pump was hoisted out of theborehole (Figure 2-2). The sampling depth coincided with the depth of the boreholepump (Table A-1). Two different diameter samplers (26 mm and 51 mm) were used,depending on the diameter of the observation tube or borehole used. Prior to sampling,all exterior and interior fittings of the Solinst sampler were sterilized with a 20-ppmchlorine dioxide solution (FreeBact 20, XINIX) and then rinsed with sterile, autoclavedAGW water to prevent microbial contamination of the groundwater. To collect in situgroundwater from the required depth interval, the sampler was kept pressurized to 2bars with N 2 gas until it was at depth; then it was de-gassed (i.e., vented to the surface)allowing the ambient water in, and finally re-pressurized once the sampler was full priorto surface retrieval. Water from the sampler was then dispensed to the variouscontainers for the analyses described below. Pressurizing the sampler to a pressure atleast double that of the highest water pressure experienced by the sampler ensured that itremained closed until reaching the sampling depth.While taking samples from the overburden holes (PVP) in the fall 2005 field campaign,it was noted that the water sampled using the Solinst sampler was in some cases slightlymore turbid than that sampled using the borehole pump. This effect was not observed inthe bedrock holes (PR, PP). It was assumed that the hoisting of the pump and thelowering of the Solinst sampler may have caused some hydrodynamic disturbance thatincreased the concentration of suspended material in the borehole. Microorganismsattach to particles, which could create some uncontrolled variability in the data.Therefore, in fall 2005 a comparison was made in the PVP20 borehole in which waterwas sampled twice for microbiology, first using the with the borehole pump and thenthe Solinst sampler. The assumed effect was confirmed. Therefore, groundwater wastaken directly from the pump to sample tubes and bottles in the 2006 spring and fallsampling campaigns.2.2 Sampling groundwater from deep boreholesDeep groundwater was sampled for the analysis of chemistry, microbiology, and gas, asdescribed below, using the PAVE system.

302.2.1 Packer-equipped deep boreholes sampled for microbiologyA total of 21 samples for microbiological analysis were taken between October 2004and November 2006 (Table 2-1) from 13 boreholes (Figure 2-3). The depth range of theboreholes was from 34.6 m down to 449.6 m.2.2.2 Sampling, transport, and extraction of deep groundwater samplesThe groundwater was sampled using the PAVE system. The procedures formicrobiological analysis using PAVE have been evaluated with the appropriate qualitycontrols (Haveman et al. 1999). Before sampling groundwater from a deep boreholeFigure 2-2. Groundwater from PP36 is transferred from the SOLINST sampler tovarious microbiology analysis tubes by the team from Microbial Analytics Sweden AB.section, the PAVE pressure vessel’s lower compartment was filled with argon ornitrogen and the movable piston was moved to the top of the pressure vessel. The gaspressure was set to approximately 5 bars. The borehole section to be sampled waspacked off with inflatable rubber packers. The PAVE system, consisting of a membranepump and one or several sterile, evacuated, closed pressure vessels connected in aseries, was lowered into the borehole. Groundwater was pumped from the packed-offzone, past the closed pressure vessels, and out of the borehole. Groundwater parameters(i.e., pH, E h , conductivity, O2, temperature, and the drill-water marker uranine) weremonitored on-line until they stabilized. The uranine tracer had to indicate that drillwatercontamination was below 2.5% before sampling could start. At that point,samples for field and laboratory analysis for hydrogeochemical characterization were

31collected (Table A-2) and analysed (Table A-3). After this phase, the pressure valve ofthe PAVE system was opened; groundwater pressure then pushed down the piston in thesampler to fill the sampler with groundwater. The valve was left open for several hoursto allow water to flow through the sampler, after which the pressure vessel was closedagain and raised out of the borehole. The pressure vessels were shipped cold and arrivedat the laboratory in Göteborg the morning after sampling (within 24 h of samplecollection). At the laboratory, the vessel was opened and the groundwater removed.Fifteen numbered, sealable, sterilized anaerobic glass tubes (no. 2048-00150; BellcoGlass, Vineland, NJ, USA), sealed with butyl rubber stoppers (no. 2048-117800) andsealed with aluminium crimp seals (no. 2048-11020, Bellco Glass), were each filledwith 1012 mL of sampled groundwater. Media inoculation started immediately afterremoving the groundwater from the sampler, and work with each sample was completewithin 2–4 h of removal of groundwater from the pressure vessel.Figure 2-3. Map showing the deep boreholes sampled. Red squares indicate thesampled boreholes listed in Table 2-1.

32Table 2-1. Identification information for the deep boreholes sampled for microbiologicalanalyses.Borehole Posiva no. Sampledsection(m)Mid elevation,z(m)SampledateOL-KR-2 KR2-329-1 328.5–330.5 306.2 2004-12-20OL-KR-6 KR6-98-8 98.5–100.5 73.7 2006-10-16OL-KR-6 KR6-125-6 125–130 94.1 2006-06-26OL-KR-6 KR6-135-8 135–137 101.8 2006-08-22OL-KR-6 KR6-422-5 422–425 328.4 2006-05-11OL-KR-7 KR7-275-1 275.5–289.5 249.4 2005-03-01OL-KR-8 KR8-77-1 77.0–84.0 57.3 2005-10-25OL-KR-8 KR8-302-2 302.0–310.0 260.7 2006-06-06OL-KR-10 KR10-326-2 326.0–328.0 316.0 2006-06-19OL-KR-10 KR10-115-1 115.5–118.5 106.0 2005-02-21OL-KR-13 KR13-362-2 362.0–365.0 294.0 2004-10-12OL-KR-13 KR13-362-3 362.0–365.0 294.0 2006-03-14OL-KR-19 KR19-526-1 525.5–539.5 449.6 2004-11-08OL-KR-27 KR27-247-1 247.0–264.0 193.5 2004-11-09OL-KR-27 KR27-503-1 503.0–506.0 391.7 2005-01-17OL-KR-31 KR31-143-1 143.0–146.0 122.4 2006-10-24OL-KR-32 KR32-50-1 50.0–52.0 34.6 2006-01-10OL-KR-33 KR33-95-1 95.0–107.0 70.6 2006-01-24OL-KR-37 KR37-166-1 166–176 111.6 2006-11-28OL-KR-39 KR39-108-1 108.0–110.0 88.2 2006-05-30OL-KR-39 KR39-403-1 403.0–406.0 344.8 2006-04-032.3 Physical parameters, chemistry, and gas content of the sampledgroundwater2.3.1 Field measurements of physical parameters in shallow boreholes andgroundwater observation tubesField measurements were made in a 1-L container at the surface while groundwater wasbeing pumped to the surface. The measurements and sampling for chemistry were doneat the end of the pumping period (Table A-1). The temperature of the groundwater was

33measured using a pIONeer 10 portable pH meter equipped with a pHC5977 cartrodecombined pH electrode (pH range 0–14, ± 0.5 at zero; temperature range –10 to 110°C,± 0.3°C) (Radiometer, Labora, Stockholm, Sweden). Redox was measured using thesame pH meter, but equipped with a MC3187Pt combined platinum electrode with anAg/AgCl reference system, range –2000 to 2000 mV (± 0.01% of reading)(Radiometer). The dissolved oxygen concentration was measured using two differentmeters and electrodes: 1) a pIONeer 20 portable oxygen meter equipped with aDOX20T-T oxygen probe with a concentration range of 1–20 mg/L (0–200% ± 1%)(Radiometer), and 2) an HQ10 Hach Portable LDO Dissolved Oxygen Meter, Cat No.51815-00 (Hach, Stockholm, Sweden). The probes were calibrated in situ per themanufacturer’s instructions. The dissolved oxygen was measured in a series of fivemeasurements made over one year to analyse for seasonal variations; the samplingmonths were October 2005 and April, May, July, and October 2006.2.3.2 Analysis of dissolved oxygen in shallow groundwater using WinklertitrationOxygen was analysed in the laboratory using a modified Winkler method as describedin detail in Carritt and Carpenter (1966). Briefly stated, three approximately 115-mL,glass-stoppered Winkler bottles (Figure 2-4) were flushed with at least three volumes ofgroundwater from the pump to remove all oxygen from atmospheric sources. Thenmanganese ions were precipitated directly in the field in an alkaline medium, formingmanganous hydroxide. This hydroxide was oxidized by present dissolved oxygen in thesample according to:2Mn(OH) 2 + O 2 2MnO(OH) 2 (Eq. 2-1)The manganese hydroxide was dissolved in the laboratory with acid and reduced byiodine ions (Figure 2-4), as follows:MnO(OH) 2 + 4H 3 O + + 3I Mn 2+ + I 3 + 7H 2 O (Eq. 2-2)Finally, the I 3 ions produced were determined by titration, with thiosulphate ions andsoluble starch used as the titration indicator, as follows:2S 2 O 3 2 + I 3 S 4 O 6 2 + 3I (Eq. 2-3)2.3.3 Chemical analyses of shallow and deep groundwaterWater samples were transferred from the investigation site to the Teollisuuden VoimaOy (TVO) laboratory directly after sampling. The chemical analyses were performed byTVO according to their protocols, or were subcontracted to external laboratories.Groundwater samples for laboratory analysis were collected during pumping beforestopping the pump (Table A-1) in a 5-L plastic canister (for testing for Br , Cl , F ,SO 4 2 , S tot , pH, and conductivity), 1-L glass bottles (for testing for alkalinity, acidity,

34Figure 2-4. Winkler bottles with acid-dissolved precipitations. The samples from PP39were free of oxygen. For comparison, oxygen-containing tap water is shown to theright.DIC/DOC), and 1-L nitric acid-washed glass bottles (for testing for metals).Groundwater samples for sulphide analysis were collected in three 100-mL Winklerbottles. All the water chemistry samples were partly filtered with a membrane filter(0.45 µm), bottled, and preserving chemicals were added to part of the samplesaccording to Table A-2. Analysis methods, detection limits, and uncertainties of themeasurements are presented in Table A-3.2.3.4 Sampling and analysis of dissolved gasShallow groundwater was sampled in triplicate in nitrogen-flushed 120-mL glass bottlesequipped with butyl rubber stoppers (no. 2048-117800; Bellco Glass) and sealed withaluminium crimp seals (no. 2048-11020). The vacuum pressure in the bottles was set to10 2 mBar 2–4 h before sampling. Water from the pump was led via poly-ether-etherketon(PEEK) tubing through a syringe into the bottles, which were filled withapproximately 100 mL of groundwater. In the laboratory, the bottles were attached tothe extraction unit (Figure A-1) and the samples were transferred to the extraction unitcylinder. The transfer time was approximately 20–30 min. Thereafter analysis was

35Table 2-2. List of deep packer-equipped boreholes sampled for analysis of dissolvedgas; -N 2 and -Ar appearing after the borehole code indicate the gas used in the pressurecompartment.BoreholeSampledsection(m)Mid elevation,z(m)SamplingdateAnalysisdateOL-KR-2-N 2 596.5–609.5 560 2006-02-28 2006-03-07OL-KR-6-N 2 422–425 328 2005-08-02 2005-08-24OL-KR-6-N 2 135–137 116 2006-08-22 2006-08-28OL-KR-6-N 2 135–137 102 2005-09-27 2005-09-27OL-KR-6-N 2 125–130 94 2006-06-26 2006-07-02OL-KR-6-N 2 120–125 90 2005-11-02 2005-12-12OL-KR-6-N 2 98.5–100.5 74 2006-10-16 2006-10-24OL-KR-6-N 2 98.5–100.5 73 2005-12-27 2006-01-13OL-KR-7-N 2 284–288 257 2006-04-25 2006-05-11OL-KR-7-Ar 220–230 197 2005-04-25 2005-08-22OL-KR-7-N 2 220–230 197 2005-04-25 2005-08-23OL-KR-8-N 2 302–310 261 2006-06-06 2006-06-09OL-KR-8-N 2 77–84 57 2005-10-25 2005-12-12OL-KR-8-N 2 77–84 57 2006-08-15 2006-08-28OL-KR-8-N 2 556.5–561 490 2006-04-27 2006-05-11OL-KR-10-N 2 326–328 316 2006-06-19 2006-06-21OL-KR-10-N 2 259–262 249 2005-04-04 2005-06-26OL-KR-10-Ar 326.5–328.5 316 2005-04-04 2005-08-23OL-KR-10-N 2 326.5–328.5 316 2005-04-04 2005-08-23OL-KR-13-N 2 362–365 294 2006-03-14 2006-03-27OL-KR-19-N 2 110–131 101 2005-09-05 2005-10-05OL-KR-19-N 2 455–468 433 2005-10-31 2005-12-12OL-KR-22-N 2 390–394 320 2006-03-01 2006-03-07OL-KR-22-N 2 147–152 116 2005-12-13 2006-01-13OL-KR-22-N 2 147–152 102 2006-08-17 2006-08-28OL-KR-29-N 2 320–340 293 2005-06-06 2005-08-23OL-KR-29-N 2 800–800 742 2005-04-16 2005-08-23OL-KR-30-N 2 50–54 40 2005-08-04 2005-08-24OL-KR-31-N 2 143–146 122 2006-10-24 2006-10-26OL-KR-33-N 2 95–107 71 2006-01-24 2006-01-26OL-KR-37-N 2 166–176 112 2006-11-28 2006-11-30OL-KR-39-N 2 403–406 345 2006-04-03 2006-04-06OL-KR-39-N 2 108–110 88 2006-05-30 2006-06-09

36performed as described in the Appendix (page 149). Water samples from all boreholesand observation tubes sampled in spring and fall 2006 were analysed for dissolved gas(Table A-1).Deep groundwater (Table 2-2) was sampled using the PAVE sample vessel, which wasattached to the extraction unit in the lab. Groundwater transfer typically took 5 min.Thereafter analysis was performed as described in the Appendix (page 149).2.4 Microbiological analyses2.4.1 Determining total number of cellsThe total number of cells (TNC) was determined using the acridine orange direct count(AODC) method as devised by Hobbie et al. (1977) and modified by Pedersen andEkendahl (1990). All solutions used were filtered through sterilized 32-mm-diameter,0.2-µm-pore-size Minisart CA syringe filters (Sartorius, GTF, Göteborg, Sweden).Stainless steel analytical filter holders, 13 mm in diameter (no. XX3001240; Millipore,Billerica, MA, USA), were rinsed with sterile, filtered, AGW (Millipore Elix 3;Millipore, Solna, Sweden). Samples of 1 mL were suction filtered (20 kPa) onto 0.22-µm-pore-size Sudan black-stained polycarbonate isopore filters, 13 mm in diameter(GTBP011300, Millipore, Solna, Sweden). The filtered cells were stained for 5 minwith 200 µL of an acridine orange (AO) solution (SigmaAldrich, Stockholm, Sweden).The AO solution was prepared by dissolving 10 mg of AO in 1 L of a 6.6 mM sodiumpotassium phosphate buffer, pH 6.7 (Pedersen and Ekendahl 1990). The filters weremounted between microscope slides and cover slips using fluorescence-free immersionoil (Olympus, Göteborg, Sweden). The number of cells was counted under blue light(390–490 nm) and using a band-pass filter for orange light (530 nm), in anepifluorescence microscope (Nikon DIPHOT 300; Tekno-Optik, Göteborg, Sweden).Between 400 and 600 cells, or a minimum of 30 microscopic fields (1 field = 0.01mm 2 ), were counted on each filter.2.4.2 ATP analysisThe ATP Biomass Kit HS for determining total ATP in living cells was used (no. 266-311; BioThema, Handen, Sweden). This analysis kit was developed based on the resultsof Lundin et al. (1986) and Lundin (2000). Sterile and “PCR clean” epTIPS with filters(GTF, Göteborg, Sweden) were used in transferring all solutions and samples to preventATP contamination of pipettes and solutions. Light may cause delayed fluorescence ofmaterials and solutions, so all procedures described below were performed in a darkroom and all plastic material, solutions, and pipettes were stored in the dark. A new 4.0-mL, 12-mm-diameter polypropylene tube (no. 68.752; Sarstedt, Landskrona, Sweden)was filled with 400 µL of the ATP kit reagent HS (BioThema, Handen, Sweden) andinserted into an FB12 tube luminometer (Sirius Berthold, Pforzheim, Germany). Thequick measurement FB12/Sirius software, version 1.4 (Berthold Detection Systems,Pforzheim, Germany), was used to calculate light emission as relative light units per

37second (RLU s 1 ). Light emission was measured for three 5-s intervals with a 5-s delaybefore each interval, and the average of three readings was registered as a singlemeasurement. The background light emission (I bkg ) from the reagent HS and the tubewas monitored and allowed to decrease to a value below 50 RLU s 1 prior to registeringa measurement. ATP was extracted from 100-µL aliquots of sample within 1 h ofcollection, by mixing for 5 s with 100 µL of B/S extractant from the ATP kit in aseparate 4.0-mL polypropylene tube. Immediately after mixing, 100 µL of the obtainedATP extract mixture was added to the reagent HS tube in the FB12 tube luminometer,and the sample light emission (I smp ) was measured. Subsequently, 10 µL of an internalATP standard was added to the reactant tube, and the standard light emission (I std ) wasmeasured. The concentration of the ATP standard was to 10 7 M. Samples with ATPconcentrations close to or higher than that of the ATP standard were diluted with B/Sextractant to a concentration of approximately 1/10 that of the ATP standard. Mixturesof reagent HS and B/S extractant were measured at regular intervals to control forpossible ATP contamination. Values of 1600 ± 500 amol ATP mL 1 (n = 10) wereobtained with clean solutions, while solutions displaying values above 1600 amol ATPmL 1 were disposed of.The ATP concentration of the analysed samples was calculated as follows:amol ATP mL 1 = (I smp I bkg ) / ((I smp + std I bkg ) – (I smp – I bkg)) 10 9 / sample volumewhere I represents the light intensity measured as RLU s 1 , smp represents sample, bkgrepresents the background value of the reagent HS, and std represents the standard(referring to a 10 7 M ATP standard).This ATP biomass method has been evaluated for use with Fennoscandian groundwater,including Olkiluoto groundwater, and the results were recently published (Eydal andPedersen 2007).2.4.3 Determining cultivable aerobic bacteriaPetri dishes containing agar with nutrients were prepared for determining the numbersof cultivable heterotrophic aerobic bacteria (CHAB) in groundwater samples. This agarcontained 0.5 g L 1 of peptone (Merck), 0.5 g L 1 of yeast extract (Merck), 0.25 g L 1 ofsodium acetate (Merck), 0.25 g L 1 of soluble starch (Merck), 0.1 g L 1 of K 2 HPO 4 , 0.2g L 1 of CaCl 2 (Merck), 10 g L 1 of NaCl (Merck), 1 mL L –1 of trace element solution(see Table 2-3 D), and 15 g L 1 of agar (Merck) (Pedersen and Ekendahl 1990). Themedium was sterilized in 1-L batches by autoclaving at 121°C for 20 min, cooled toapproximately 50C in a water bath, and finally distributed in 15-mL portions in 9-cmdiameterplastic Petri dishes (GTF, Göteborg, Sweden). Ten-times dilution series ofculture samples were made in sterile analytical grade water (AGW) with 0.9 g L 1 ofNaCl; 0.1-mL portions of each dilution were spread with a sterile glass rod on the platesin triplicate. The plates were incubated for between 7 and 9 d at 20°C, after which thenumber of colony forming units (CFU) was counted; plates with between 10 and 300colonies were counted.

382.4.4 Preparing media for most probable numbers of cultivable anaerobicmicroorganismsMedia for determining the most probable number of microorganisms (MPN) ingroundwater were formulated based on previously measured chemical data fromOlkiluoto. This allowed the formulation of artificial media that most closely mimickedin situ groundwater chemistry for optimal microbial cultivation (Haveman and Pedersen2002a). Media for the nitrate-reducing bacteria (NRB), iron-reducing bacteria (IRB),manganese-reducing bacteria (MRB), sulphate-reducing bacteria (SRB), autotrophicacetogen (AA), heterotrophic acetogen (HA), autotrophic methanogen (AM), andheterotrophic methanogen (HM) metabolic groups were autoclaved and anaerobicallydispensed, according to the formulations outlined in Table 2-3, into 27-mL, sealableanaerobic glass tubes (no. 2048-00150; Bellco Glass), sealed with butyl rubber stoppers(no. 2048-117800), and sealed with aluminium crimp seals (no. 2048-11020).All culture tubes were flushed with 80/20% N 2 /CO 2 gas and then filled with 9 mL of theappropriate media. For IRB, 1 mL of hydrous ferric oxide (HFO), prepared from FeCl 3 ,was added to each culture tube. The final concentration of the iron solution was 0.44 M.For MRB, 2 mL of 135 mM MnO 2 solution (Lovley and Phillips 1988) was added. TheHM media also contained 20 mL L 1 of 100 g L 1 NaCOO, 3 mL L 1 of 6470 mMtrimethylamine, 4 mL L 1 of methanol, and 20 mL L 1 of a 20 g L 1 solution ofNaCH 3 COO. The HA medium also contained 20 mL L 1 of 100 g/L NaCOO, 3 mL L 1of 6470 mM trimethylamine, and 4 mL L 1 of methanol. The final pH was adjusted tobetween 6.5 and 7.5 with 1 M HCl or 1 M NaOH.2.4.5 Inoculations and analysis for anaerobic microorganismsInoculations for NRB, IRB, MRB, SRB, AA, HA, AM, and HM were performed in thelaboratory less than 2 h after sampling for shallow groundwater and the next morningfor deep groundwater samples. After inoculating, the headspaces of only the AA andAM cultures were filled with H 2 to an overpressure of 2 bars; all MPN tubes wereincubated in the dark at 20°C for 813 weeks. After incubation, the MPN tubes wereanalysed by testing for metabolic products or substrate consumption. Nitrateconsumption was determined using a DR/2500 spectrophotometer (HACH, Loveland,CO, USA) with the chromotropic acid method (HACH method no. 10020) for water andwastewater (0.2–30 mg/L NO 3 -N). The production of ferrous iron by IRB wasdetermined using the 1,10 phenanthroline method (HACH, method no. 8146). HACHmethod no. 8034, based on periodate oxidation, was used in a similar way to determineMn 2+ concentrations in MPN tubes for MRB. SRB were detected by measuring sulphideproduction using the CuSO 4 method according to Widdel and Bak (1992) on a UVvisible spectrophotometer (Genesys10UV, VWR, Stockholm, Sweden). Methanogenswere detected by measuring the production of methane in the culture tube headspace.The methane was analysed using a Star 3400CX gas chromatograph (Varian,Stockholm, Sweden) using a flame ionization detector (FID) at an oven temperature of65°C and a detector temperature of 200°C. The methane gas was separated using aPorapak-Q column (2 m 1/8 inch diameter; Agilent Technologies, Varian, Stockholm,Sweden) and analysed on the FID with nitrogen as the carrier gas (confer Appendix,

39page 149). Acetogens were detected by means of acetate production using an enzymaticUV method (Enzymatic Bioanalysis Kit no. 10 139 084 035; Boehringer Mannheim/R-Biopharm, Food Diagnostics, Göteborg, Sweden) with a UV visible spectrophotometer(per SRB detection). Product formation at a concentration twice or above that of theuninoculated control tubes was taken as positive for all MPN analyses except nitrate, forwhich a 50% reduction in nitrate concentration, compared with that of uninoculatedcontrols, was taken as a positive result.The MPN procedures resulted in protocols with tubes that scored positive or negativefor growth. The results of the analyses were rated positive or negative compared withcontrol levels. Three dilutions with five parallel tubes were used to calculate the MPNof each group, according to the calculations found in Greenberg et al. (1992).2.4.6 Inoculations and analysis for aerobic methane-oxidizing bacteriaSets of MPN tubes were prepared for samples using a nitrate mineral salts (NMS)medium (Whittenbury et al. 1970) prepared as follows: 1.0 g L 1 of KNO 3 , 1 g L 1 ofMgSO 4 7 H 2 O, 0.2 g L 1 of CaCl 2 2 H 2 O, 1 mg L 1 of CuCl 2 2H 2 O, 7 g L 1 ofNaCl, 1 mL L 1 of an iron solution made of 0.5 g of ferric (III) chloride in 1000 mL ofAGW, 1 mL L 1 of a trace element solution according to Table 2-3D and 2 mL L 1 of aphosphate buffer solution made of 3.6 g Na 2 HPO 4 , and 1.4 g NaH 2 PO 4 in 100 mL ofAGW. The pH was adjusted to 6.8–7.0. Cultural conditions were optimized to supportthe growth of both types I and II methane-oxidizing bacteria (MOB) by adding 1 mg L 1of copper chloride dihydrate. This is because the soluble and particulate methanemonooxygenase (s/pMMO) common to all known MOB is controlled by a copperinducibleregulatory pathway.MPN inoculations were completed at the ONKALO laboratory within 2 h of samplecollection for all shallow borehole samples. Five parallel dilution tubes were used foreach dilution. All transfers were performed aseptically using new sterile syringes andneedles. After each transfer, the tubes were vortexed to achieve homogeneity. Controltubes contained nitrate minimal salt medium and 1 mL of filtered groundwater. Afterinoculation, filter-sterilized (using 0.2-µm Millipore filters) methane was injected intothe headspace of each tube to 1 Bar overpressure. The tubes were then incubatedhorizontally in the dark at 20C. Growth of cells was detected after between 2 and 4weeks, as judged by turbidity compared with that of negative controls and theconcomitant production of CO 2 via methane oxidation in turbid tubes. MPNcalculations were made using a combination of positive tubes in a 3-tube dilution series(i.e., 15 tubes) according to Greenberg et al. (1992). The detection limit was

40Table 2-3. A-G. Compositions of anaerobic media used for MPN cultivation of differentmetabolic groups of anaerobic microorganisms. All components were anoxic.A) Ready medium Metabolic group aComponent (mL/L) NRB IRB & MRB SRB AA & HA AM & HMBasal medium (Table B) 925 940 860 860 890Trace elements (Table C) - - 10 10 10Trace elements (Table D) 1.0 1.0 - - -Vitamins (Table E) 1.0 1.0 - - -Vitamins (Table F) - - 10 10 10Thiamine stock (Table G) 1.0 1.0 1.0 1.0 1.0Vitamin B 12 stock (Table G) 1.0 1.0 1.0 1.0 1.0Fe stock (Table G) - - 5.0 5.0 5.0Resazurin (Table G) - - 2.0 2.0 2.0Cysteine hydrochloride (Table G) - - 10 10 10NaHCO 3 (Table G) 30 30 60 60 60Yeast extract (Table G) 1.0 1.0 10 10 10NaCH 3 COO (Table G) 25 25 - - -Lactate (Table G) 5.0 - 5.0 - -KNO 3 (G) 10 - - - -Sodium sulphide (0.2 M) - - 7.5 10 10a NRB = nitrate-reducing bacteria, IRB = iron-reducing bacteria, MRB = manganese-reducing bacteria,AA = autotrophic acetogens, HA = heterotrophic acetogens, AM = autotrophic methanogens, HM =heterotrophic methanogensB) Basal medium Metabolic group aComponent (g) NRB, IRB, & MRB SRB AA & HA AM & HMAGW 1000 1000 1000 1000NaCl 7 7 7 7CaCl 2 *2H 2 O 1.0 1.0 1.0 0.28KCl 0.1 0.67 0.67 0.67NH 4 Cl 1.5 1.0 1.0 1.0KH 2 PO 4 0.2 0.15 0.15 0.15MgCl 2 *6H 2 O 0.1 0.5 0.5 0.5MgSO 4 *7H 2 O 0.1 3.0 - -MnCl 2 *4H 2 O 0.005 - - -Na 2 MoO 4 *2H 2 O 0.001 - - -a NRB = nitrate-reducing bacteria, IRB = iron-reducing bacteria, MRB = manganese-reducing bacteria,AA = autotrophic acetogens, HA = heterotrophic acetogens, AM = autotrophic methanogens, HM =heterotrophic methanogens

41Table 2-3. Continued.C) Trace element solutionComponentAGWNitrilotriacetic acidFe(NH 4 ) 2 (SO 4 ) 2 *6H 2 ONa 2 SeO 3CoCl 2 *6H 2 OMnCl 2 *4H 2 ONa 2 MoO 4 *2H 2 ONa 2 WO 4 *2H 2 OZnSO 4 *7H 2 OAlCl 3NiCl 2 *6H 2 OH 3 BO 3CuCl 2 *2H 2 OAmount1000 mL1500 mg200 mg200 mg100 mg100 mg100 mg100 mg100 mg40 mg25 mg10 mg10 mgF) Vitamin mixture for SRB, AA, HA, AM, andHMComponentAmountSodium phosphate buffer 10 1000 mLmMpH 7.1p-Aminobenzoic acid10 mgNicotinic acid10 mgCalcium D(+) pantothenatePyridoxine dihydrochlorideRiboflavinD(+)-biotinFolic acidDL-6-8-thiotic acid10 mg10 mg10 mg5 mg5 mg5 mgD) Non-chelated trace elementsComponentAmountAGW987 mLHCl (25% = 7.7 M)12.5 mLFeSO 4 *7H 2 O2.1 gH 3 BO 330 mgMnCl 2 *4H 2 O100 mgCoCl 2 *6H 2 O190 mgNiCl 2 *6H 2 O24 mgCuCl 2 *2H 2 O2 mgZnSO 4 *7H 2 O144 mgNa 2 MoO 4 *2H 2 O36 mgE) Vitamin mixture for NRB, IRB, and MRBComponentAmountSodium phosphate buffer 10 mM 100 mLpH 7.14-Aminobenzoic acid4 mgD(+)-biotin1 mgNicotinic acid10 mgPyridoxine dihydrochloride15 mgCalcium D(+) pantothenate5 mgG) Stock solutionsComponentNaHCO 3Thiamine chloridedihydrochloride in a 25 mMsodium phosphate buffer, pH3.4Cyanocobalamin (B 12 )KNO 3NaCH 3 COOYeast extractFe(NH 4 ) 2 (SO 4 ) 2 *6H 2 O,initially dissolved in 0.1 mLof concentrated HClAmount84 g L 1100 mg L 150 mg L 1100 g L 1100g L 150 g L 12 g L 1Resazurin500 mg L 1Cysteine-HCl50 g L 1Sodium lactate solution 50%

422.4.7 Quality controls for the most probable number analysisThe reproducibility of the sampling and analysis procedures was tested using theSwedish PAVE analogue, the PVB pressure vessel. The 353.5–360.0-m section of theForsmark site investigation borehole KFM06A was sampled on 14 March 2005 usingtwo PVB samplers installed at the same time. It was also deemed important to testreproducibility over time in borehole sections that were expected to harbour stable andreproducible populations. This was done in two boreholes at the MICROBE site(Pedersen 2005a) in the Äspö HRL tunnel, denoted KJ0052F01 and KJ0052F03.Groundwater from the borehole sections was sampled using PVB samplers on twooccasions, 26 October 2004 and 9 February 2005. The PVB samplers were attached tothe flows from each borehole section, and groundwater was circulated overnight underin situ pressure, temperature, and chemistry conditions. Early in the morning, thesamplers were closed, detached, and transported to the laboratory in Göteborg; analysisstarted the same afternoon, before 14.00. All parts of this procedure resembled thesampling of sections in the Olkiluoto deep boreholes, except that in this case thesamplers were not operated remotely from the ground surface; instead, personnelstanding next to the PVB samplers in the tunnel manually operated the samplers usingadjustable spanners.

433 RESULTS3.1 Analysis of physical and chemical parametersThe results of all field and laboratory measurements of physical and chemicalparameters are presented in Appendix A, Table A-4. All references to specific physicaland chemical data in this results section refer to this table. For some parameters, such asthe analysis of oxygen and E h using the HACH field instruments, results were onlyobtained for the shallow groundwater samples. The shift from sampling shallowgroundwater using the SOLINST sampler and the pump to sampling deep groundwaterusing the PAVE system partly meant applying different sampling methods. Forexample, shallow groundwater gas was analysed from samples in glass bottles while thedeep groundwater was analysed in samples from the PAVE pressure vessel. However,the differences between sampling procedures do not imply biased data, and shallow anddeep groundwater data are presented and interpreted together. All available data foreach analysed parameter in both shallow and deep groundwater are presented in thefollowing figures.3.1.1 Field measurements of physical parametersThe pH ranged from 4.8 to 8.2. All values below pH 6 were found in groundwater froma depth of less than 10 m (Figure 3-1), except for samples from borehole PP36 (12.1 m),which had a stable pH of 5.8 over all sampling periods. Most of the shallowgroundwater samples had pH values from 6.5 to 7.5, while the deep groundwater hadpH values from 7 to 8.2.The conductivity ranged from 10 to 10000 mS m 1 (Figure 3-2), with the exception ofgroundwater from boreholes PR1 (sampled 2006-10-11) and PP39 (sampled 2006-04-24), which were diluted to below and at the detection limit, respectively. Theconductivity increased exponentially with depth within a range of approximately plus–minus five times the observed average value for each depth.Oxygen was found in several shallow groundwater samples (Figure 3-3) but was absentfrom deep groundwater (see section 3.2). Two different methods were used to analyseoxygen in shallow groundwater, one electrochemical and one wet chemistry method.The HQ10 HACH Portable LDO dissolved oxygen meter was used in the field startingin fall 2005. The data from spring 2004 were obtained using a membrane electrode thatis more difficult to operate than the LDO electrode is. The membrane electrode needsfrequent calibrations that can be difficult to perform in the field, while the LDOelectrode is calibrated once per year and is very stable. On-line electrodes were used toanalyse oxygen in deep groundwater (data not presented). The oxygen values obtainedin shallow groundwater in spring 2004 were generally higher than those obtained in theremaining three field campaigns. Although the membrane electrode was carefullycalibrated, some caution should be used when comparing 2004 oxygen data withoxygen data from 2005 and onwards, as the different types of electrodes used may haveintroduced a bias. Titrating oxygen using the Winkler method was introduced in spring2006. The LDO electrode results correlated well with the Winkler data (Figure 3-4).

44The LDO electrode results were very well correlated with the Winkler results at highoxygen concentrations. The exception was for borehole PP9 (sampled 2006-04-27), buton this occasion there was an unusual gap of approximately 4 h in sampling timebetween the LDO electrode measurement (1 st ) and the Winkler sample (2 nd ), due toproblems with heavy turbidity from a dissolving bentonite packer; this delay may haveintroduced more oxygen. Therefore, the Winkler data (4.24 mg O 2 L 1 ) were muchhigher than the LDO data (2.35 mg O 2 L 1 ) on this occasion. Otherwise, all LDO datawere similar to or somewhat higher than the Winkler data. The LDO electrode is lessprecise at values below 0.5 mg O 2 mL 1 , and those data should be taken asapproximations. The Winkler analysis is reliable over a large range, extending from thedetection limit of 0.05 mg of O 2 mL 1 to oversaturated samples. In conclusion, both theelectrode and the Winkler analyses revealed rapidly decreasing oxygen values withincreasing depth. Small amounts of oxygen remained in the groundwater below 10 m(Figure 3-3), except for the problematic PP9 sample mentioned above.The measurement of E h in shallow groundwater should be regarded as a relativeanalysis and the E h values obtained should not be directly compared with E h valuesobtained in deep groundwater as other electrodes and measurement conditions werevalid there. The E h values over depth in shallow groundwater were very scattered,displaying only a very weak decreasing trend with increasing depth (Figure 3-5).The four field campaigns were performed in April and October. The concentration ofdissolved oxygen was expected to vary seasonally, and when dissolved oxygen wasrepeatedly analysed in summer 2006 this could be confirmed (Figure 3-6). Theconcentration of oxygen decreased in summer and increased in fall and spring.3.1.2 Chemical analyses of groundwaterThe general trend was for dissolved solids to increase with depth (Table A-4), asreflected by the conductivity measurements (Figure 3-2). The concentration of dissolvedorganic carbon (DOC) is of special interest for microbiological interpretations, as DOCcan be expected to relate to microbiology. When analysed, no correlation was foundbetween DOC and depth (Figure 3-7). Instead, the DOC values were scattered frombelow the detection limit of 1.8 mg DOC mL 1 up to 39 mg DOC mL 1 . One sampledisplayed an exceptionally high DOC value of 196 mg DOC mL 1 . This was from theshallow PVP1 observation tube that was completely flooded by snow meltwater untilthe day before sampling (2006-04-27). This was not persistent contamination, as theDOC value was less than a tenth of that six months later (2006-10-12). Theconcentrations of ferrous iron and sulphide displayed inversely related trends, withdecreasing ferrous iron and increasing sulphide values with depth (Figure 3-8). Theferrous iron concentration was up to ten times higher in shallow than in deepgroundwater. The dissolved sulphide concentration was at or below the detection limitdown to a depth of 70 m and peaked at a depth of approximately 300 m.

45110Depth (m)10010004.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5pHFigure 3-1. The pH in groundwater samples from Olkiluoto against depth.110Depth (m)10010000.1 1.0 10.0 100.0 1000.0 10000.0Conductivity (mS m -1 )Figure 3-2. The electrical conductivity in groundwater samples from Olkiluoto overdepth.

4605Depth (m)1015200 1 2 3 4 5 6 7O 2 (mg L 1 )O 2 HACHO 2 WinklerFigure 3-3. The concentration of dissolved oxygen in shallow groundwater analysedusing the HQ10 HACH Portable LDO dissolved oxygen meter and by means ofWinkler titration in the laboratory (the average values of three titrations are shown).10.0O 2 Winkler (mg L 1 )1.00.10.00.0 0.1 1.0 10.0O 2 LDO electrode (mg L 1 )Figure 3-4. The correlation between oxygen as measured using the DOX20T-Tmembrane electrode (2004) or the HQ10 HACH Portable LDO dissolved oxygenmeter and by means of Winkler titration in the laboratory (the average values of threetitrations are shown). The line denotes identical values.

470246Depth (m)81012141618-200 -100 0 100 200 300 400 500E h HACH-electrode (mV)Figure 3-5. The relationship between E h , as analysed using the pIONeer 10 portablepH-E h , meter and depth.65O 2 (mg L 1 )4321010 - 14 thOctober 200524 - 28 thApril 200618 - 23 rdMay 20067 - 12 thJuly 20069 - 13 thOctober 20061 2 3 4 5 6 7 8 9 10 11 12Time (months)PR1PP2PP9PP36PP39PVP1PVP4APVP13PVP14PVP20Figure 3-6. The seasonal variation of dissolved oxygen in shallow Olkiluotogroundwater analysed using the HQ10 HACH Portable LDO dissolved oxygen meter.

48110Depth (m)10010001 10 100DOC (mg L 1 )Figure 3-7. The concentrations of dissolved organic carbon (DOC) in groundwatersamples from Olkiluoto over depth.110Depth (m)10010000.01 0.10 1.00 10.00(mg L 1 )S 2Fe 2Figure 3-8. The concentrations of dissolved ferrous iron and sulphide in groundwatersamples from Olkiluoto over depth.

493.2 Sampling, extraction, and analysis of gasThe total amount of extractable gas per L of Olkiluoto groundwater is shown in Figure3-9. This total gas is equal to the sum of the amounts of several gases analysedseparately. The amount of each analysed gas extracted per L of Olkiluoto groundwateris produced by dividing the total extracted gas per L by the ppm values shown in TableA-5 and Table A-6, resulting in µL of gas per L of groundwater. The ppm values inTable A-5 and Table A-6 thus indicate the proportions of each gas per L of extractedgas, while the scatter plots of gas versus depth (Figure 3-14 to Figure 3-20) indicate theamounts of each gas per L of groundwater. Please note this distinction, as it is veryimportant for understanding how the gas analysis results are reported and interpreted.3.2.1 Dissolved gas in shallow groundwater – comments on the methodsIn the sampling procedure, 120-mL glass bottles were used, typically to collectapproximately 100 mL of groundwater from which gas was extracted (Table A-5). Theamount of extractable gas ranged from 2.3 mL L 1 in borehole PVP4A-1 (2006-04-27)to 7.2 mL L 1 in borehole PP39 (2006-10-11), which is a relatively small amountcompared with what could be extracted from deeper groundwater using the PAVEpressure vessel (Table A-6). The precision of repeated extractions is reflected by thestandard deviations, which were in the 11–87% range. Although these standarddeviations are high, the results must still be regarded as good, given that the gasvolumes extracted were quite small. Using the PAVE vessel for deep gas samples,which contained more water and gas than did the shallow samples, increased theprecision to 10% as judged from the reproducibility tests made at the Äspö HRL (TableA-12).The oxygen present in shallow groundwater samples could have two different origins. Itcould be dissolved oxygen present in the groundwater at the time of sampling, or itcould be oxygen that entered the sample during the extraction procedure. The transfer ofa 100-mL groundwater sample from the sample bottle to the gas extractor (Figure A-1)took 20–30 min, and during this time small amounts of oxygen may have entered thesample. If the oxygen results obtained using the gas extractor are compared with dataobtained using the very reliable Winkler method (Figure 3-10), it is clear that someoxygen did enter the samples during extraction. The oxygen values obtained using gaschromatography, presented in Table A-5, are thus artefacts that in most cases should notbe taken into consideration. Therefore, the values in Table A-4 should be used fordissolved oxygen.3.2.2 Dissolved gas in deep groundwater – comments on the methodsThe analysis of gas sampled using the PAVE pressure vessel and reported here is amethodology under continuous development at Microbial Analytics Sweden AB. Theequipment was technically improved and the sampling, extraction, and analysisprocedures were adjusted in 2005. The precision and reliability of the analyses werethus better in 2006 than in 2005. The analysis procedures must be further developed andsuch development may include constructing a new version of the PAVE pressure vessel.

50However, the data obtained so far are still very valuable if interpretations andconclusions take the following method comments into consideration.The PAVE pressure vessel has a piston that separates the groundwater sample from thelower compartment filled with argon or nitrogen gas. This gas will balance the pressureof the groundwater when sampled, which reduces large shifts in pressure in the samplethat would result in degassing. There have occasionally been some problems withleakage of pressure gas into the sample, which elevates the argon or nitrogenconcentration of the sample. This effect is difficult to track. The best way to judge asample result is to evaluate it in relation to several other results for samples from similardepths. A large discrepancy between a particular sample result and the average result forsamples from the same depth region indicates a sampling artefact. One obvious suchcase is that of the OL-KR22 sample from a depth of 320 m sampled on 2006-03-01(Table A-6). This sample contains much more nitrogen than do all other samples frombelow a depth of 300 m (Figure 3-11), and the amount of extracted gas exceeds that inall other samples from adjacent depths by 3 to 4 times. It is obvious that this sample washeavily contaminated with nitrogen from the lower compartment. As a result, all othergases in this sample were diluted, so the results underestimate the actual values of allother gases by approximately 3 to 4 times. The reproducibility of the PAVE samplingand analysis method was tested twice using two samples from boreholes OL-KR7(2005-04-25) and OL-KR10 (2005-04-04) (Table A-6). The reproducibility was notvery good for unknown reasons. It appears as though the PAVE samplers collectdifferent amounts of gas, possibly due to differences in the volumes of sample and theirpositions in the borehole sampling equipment. This problem is being studied on anongoing basis. As of summer 2007, however, there were still no satisfactoryexplanations of or solutions to this problem. More extensive discussion of the problemswith sampling and analysing gas using the PAVE system, and of the representativity ofthe gas results obtained with it, can be found elsewhere (Gascoyne 2000; Pitkänen andPartamies 2007).The time from sampling to extraction and analysis should preferably be as short aspossible. Technical problems in the laboratory made it impossible to extract samplesfrom April to August 2005. The current standard is to have the sample extracted withina week of the sampling day and this goal was, with some exceptions, achieved in 2006.If samples are kept for long periods in the PAVE system, there is a risk that gasdiffusion processes may change the gas composition; moreover, microbial activityinside the sample may also have a significant effect on results by producing andconsuming hydrogen, methane, and carbon dioxide. To minimize microbial impact onthe samples, they are kept refrigerated until analysis, which reduces the rate ofmicrobial processes. Finally, anaerobic corrosion of the stainless steel in the PAVEcontainer may generate hydrogen, which will of course distort the hydrogen values. Thismay explain the unexpectedly high hydrogen values in some of the samples fromboreholes OL-KR6 and OL-KR8 (Table A-6). Alternately, the anomalously highhydrogen data may be due to incomplete filling of the sample vessels, as suggested byPitkänen and Partamies (2007).During the extraction process, there were problems with air entering the sample, whichwas detected as the presence of oxygen. As deep groundwater generally contain ferrousiron and sometimes sulphide (Table A-4), oxygen should not be present, because these

51two ions are not stable in oxygenated water. Such air leakage was considerable in 2005,and most of the leakage was tracked to the unit used to connect the sampler to the gasextractor. A new type of connector unit was developed in late 2005, and the aircontamination was immediately reduced ten-fold from approximately 10% to 1% (TableA-6). The remaining 1% has been more difficult to handle. In 2007, approximately halfof the samples analysed were free of detectable air, analysed as the presence of oxygen.All deep groundwater samples are back calculated to the gas concentrations the sampleshad before air contamination.The sum of all analysed gases in ppm should theoretically be 1,000,000 ppm,representing 100%. The results shown in Table A-6 indicate that this was achievedmostly within the 2% range. This indicates that the gases selected and analysed forwere actually the dominant gases. If a major gas had not been analysed for, the sum ofall gases would be less than 100%. The sum of the analysed gases was compared withthe amount of extracted gas; these two values were also comparable, as reflected in thetotal percentage of gas.3.2.3 Distribution of gases in Olkiluoto groundwaterThe extracted gas was composed of five major and five minor gases. The distribution ofthe major gases nitrogen, methane, carbon dioxide, and helium is shown in Figure 3-11;argon concentrations were very scattered (Table A-6) and are omitted from thesefigures. There were three distinct gas composition profiles that could be related todifferent depth layers. The shallow gas down to a depth of approximately 20 m wascomposed mainly of nitrogen and a smaller but still significant amount of carbondioxide. Intermediate-depth gas from depths of approximately 20 to 300 m wasdominated by nitrogen. At depths below 300 m, methane concentrations increasedsignificantly, making it the dominant gas in most deep samples. The results for a depthof 320 m in borehole OL-KR-22 were distorted by a leaking PAVE cylinder piston, andthe methane concentration was most likely higher in the groundwater than the resultssuggest. The proportions of the minor gases hydrogen and carbon monoxide were notparticularly correlated with depth; the exception was the minor gas carbon monoxide,the content of which was higher in the shallow and intermediate-depth groundwater thanin the deep groundwater below a depth of 300 m (Figure 3-12). All two-carbonhydrocarbons were absent from the shallow groundwater, except for that from boreholesPVP1 and PVA1 (Table A5). The hydrocarbon gases analysed for appeared in theintermediate-depth groundwater and the proportion of ethane increased significantly inthe deep groundwater (Figure 3-13).The average total amount of dissolved gas increased exponentially with depth (Figure3-9). In the shallow groundwater, volumes of 25–70 mL of gas L 1 groundwater 1 werefound. The amounts then increased up to a maximum of 1380 mL of gas L 1groundwater 1 in the deepest groundwater sample from borehole OL-KR29, at a depthof 742 m.The concentration of dissolved nitrogen per L of groundwater increased with depth, butits concentration range was narrow compared with those of other gases. Nitrogenconcentration increased approximately 20-fold with depth, rising from 15 to 250 mL of

52nitrogen L 1 groundwater 1 (Figure 3-14). In comparison, the noble gas heliumincreased approximately 1000-fold over the depth range analysed, rising from 30 to20000 µL of gas L 1 groundwater 1 . The trend was for average helium concentrations toincrease exponentially over most of the depth range analysed, except in the deepestsample (Figure 3-15). Methane displayed a two-layer profile with values between 1 and1000 µL of gas L 1 groundwater 1 down to a depth of 300 m (Figure 3-16). At thisdepth, there was a distinct 100-fold increase in the methane concentration to 100,000 µLof gas L 1 groundwater 1 ; the methane concentration then increased ten-fold by a depthof 742 m, the depth of the deepest sample analysed. Overall, the concentrations ofmethane were distributed over a ten-million-times range. The concentration of the lastof the major gases analysed, carbon dioxide, decreased approximately ten-fold from theshallow groundwater samples to a depth of approximately 20 m (Figure 3-17).Thereafter, the average concentration decreased slightly, except in the deepest sample,which had a high concentration relative to the other deep (300–560 m) groundwatersamples. The average concentration of dissolved hydrogen per L of groundwaterdisplayed a weak increasing trend with depth, but the data points were very scattered(Figure 3-18). Carbon monoxide concentrations did not change with depth, but were, aswith hydrogen, scattered (Figure 3-19). However, average ethane concentrationsincreased exponentially with depth (Figure 3-20).

530100200Depth (m)30040050060070080010 100 1000 10000Gas (mL L 1 groundwater 1 )Figure 3-9. The total amount of extractable dissolved gas in Olkiluoto groundwater.1210GC O 2 (mL L 1 )864200 1 2 3 4 5Winkler O 2 (mg L 1 Figure 3-10. The relationship between oxygen concentrations in groundwater samplesas analysed using gas chromatography (GC) and using Winkler titration. The dottedbands denote 95% confidence intervals.

54Depth (m)PVP1 - 4PVP1 - 4PVP13 - 6PVP13 - 6PR1 - 6PR 1 - 6PVP14 - 9PVP14 - 9PVP4A - 10PVP4A - 10PVP4A - 10PP36 - 11PP36 - 11PVP20 - 13PP39 - 14PP39 - 14PP9 - 15PP9 - 15PP2 - 14PP2 - 14PVA1 - 20PVA1 - 20KR30 - 40KR8 - 57KR8 - 57KR33 - 71KR6 - 73KR6 - 74KR39 - 88KR6 - 90KR6 - 94KR19 - 101KR6 - 102KR22 - 102KR37 - 112KR22 - 116KR6 - 116KR31 - 122KR7 - 197KR7 - 197KR10 - 249KR7 - 257KR8 - 261KR29 - 293KR13 - 294KR10 - 316KR10 - 316KR10 - 316KR22 - 320KR6 - 328KR39 - 345KR19 - 433KR8 - 490KR2 - 560KR29 - 7420 200 400 600 800 1000Major gas components (mL L 1 gas 1 )MethaneCarbon dioxideHeliumNitrogenFigure 3-11. Stacked values of the major components of the extractable gas fromshallow and deep Olkiluoto groundwater. Blue borehole designations indicate cases inwhich analysis was done twice, once with argon (1 st bar) and once with nitrogen (2 ndbar) in the pressure vessel.

55Depth (m)PVP1 - 4PVP1 - 4PVP13 - 6PVP13 - 6PR1 - 6PR 1 - 6PVP14 - 9PVP14 - 9PVP4A - 10PVP4A - 10PVP4A - 10PP36 - 11PP36 - 11PVP20 - 13PP39 - 14PP39 - 14PP9 - 15PP9 - 15PP2 - 14PP2 - 14PVA1 - 20PVA1 - 20KR30 - 40KR8 - 57KR8 - 57KR33 - 71KR6 - 73KR6 - 74KR39 - 88KR6 - 90KR6 - 94KR19 - 101KR6 - 102KR22 - 102KR37 - 112KR22 - 116KR6 - 116KR31 - 122KR7 - 197KR7 - 197KR10 - 249KR7 - 257KR8 - 261KR29 - 293KR13 - 294KR10 - 316KR10 - 316KR10 - 316KR22 - 320KR6 - 328KR39 - 345KR19 - 433KR8 - 490KR2 - 560KR29 - 7420 50 100 150 650 700Minor gas components (µL L 1 gas 1 )Carbon monoxideHydrogenFigure 3-12. Stacked values of the minor gas components carbon monoxide andhydrogen in the extractable gas from shallow and deep Olkiluoto groundwater. Blueborehole designations indicate cases in which analysis was done twice, once with argon(1 st bar) and once with nitrogen (2 nd bar) in the pressure vessel.

570100200Depth (m)30040050060070080010 100 1000 10000Nitrogen (mL L groundwater )Figure 3-14. The amount of extractable dissolved nitrogen gas in Olkiluotogroundwater.0100200Depth (m)30040050060070080010 100 1000 10000 100000Helium (µL L 1 groundwater 1 )Figure 3-15. The amount of extractable dissolved helium gas in Olkiluoto groundwater.

580100200Depth (m)3004005006007008001 10 100 1000 10000 100000 1000000CH 4 (µL L 1 groundwater 1 )Figure 3-16. The amount of extractable dissolved methane gas in Olkiluotogroundwater.0100200Depth (m)3004005006007008001 10 100 1000 10000 100000 1000000CO 2 (µL L 1 groundwater 1 )Figure 3-17. The amount of extractable dissolved carbon dioxide gas in Olkiluotogroundwater.

590100200Depth (m)3004005006007008000.1 1.0 10.0 100.0Hydrogen (µL L 1 groundwater 1 )Figure 3-18. The amount of extractable dissolved hydrogen gas in Olkiluotogroundwater.0100200Depth (m)3004005006007008000.1 1.0 10.0 100.0CO (µL L 1 groundwater 1 )Figure 3-19. The amount of extractable dissolved carbon monoxide gas in Olkiluotogroundwater.

600100200Depth (m)3004005006007008000.01 0.10 1.00 10.00 100.00 1000.00 10000.00C 2 H 6 (µL L 1 groundwater 1 )Figure 3-20. The amount of extractable dissolved ethane gas in Olkiluoto groundwater.3.3 Analysis of biological parametersConsecutive tests were performed in the 2004, 2005, and 2006 field seasons to developand test the quality of the sampling procedures. The procedures for sterilizing the pumpand samplers were analysed and the data obtained using the SOLINST tube sampler andusing the pump were compared. The influence of pumping time on the results was alsostudied.3.3.1 Sterilization of borehole pumpsTesting the AGW water in the ONKALO laboratory revealed TNC counts that weresignificantly different from zero (Table 3-1). ATP readings confirmed that there wassome biomass in this water-producing unit, which was expected, as such systems are notsterile. However, by using proper cleaning procedures and a UV lamp in the water tank,the bacterial numbers in AGW systems can be kept very low. AGW water sterilized inan autoclave had TNC and CHAB readings that were not significantly different fromzero. An extremely small amount of ATP was detected but, at such a low concentrationlevel, the ATP analysis was very sensitive to even the smallest contamination. Thesterilized pump came directly from the field and had been in use for several years. Evenso, the sterilization testing produced very good results, with values just above zero,significantly lower than had been found in the ONKALO laboratory’s AGW system. Itcan thus be safely concluded that the sterilization procedures worked properly and thatthe sampling pump systems did not cross-contaminate the sampled boreholes orsamples.

61Table 3-1. Results of the sterilization tests of the borehole pump and the analysis of theAGW water.Measurement aAGW waterproduced in theOnkalo laboratorySterile AGW waterfor washing pumpsand samplersSterile AGW waterafter pumping andsamplingTNC (cells mL 1 ) 12000 (2200) b 110 (120) 7000 (3700)ATP (amol mL 1 ) 9725 (3209) 472 (214) 3351 (86)CHAB (cells mL 1 ) - d 3 (6) 43 (15)NRB (cells mL 1 ) - - 0.4 (0.1–1.7) cIRB (cells mL 1 ) - -

62respectively (Greenberg et al. 1992). There was a small bias towards higher numbers inthe PVB sampler denoted 1. The maximum discrepancy between the samples wasobserved for SRB and equalled a factor of two, well within the 95% confidenceintervals of the MPN analysis. The TNC determinations and ATP analysis alsodisplayed good reproducibility; this included the sampling procedure, transportationlogistics, and MPN inoculation, cultivation, and analysis for each physiological groupof microorganisms, i.e., TNC, CHAB, and ATP. The second test explored thereproducibility of two different analytical rounds on groundwater from two differentborehole sections and of repeated sampling over time. This test also included the effectsof different personnel involved and different preparations of chemicals and media. Ingeneral, groundwater from the two borehole sections had very different result profilesthat reproduced well (Table 3-4). The MPNs of MRB, SRB and AM differed mostbetween the sampling times, while the MPNs of AA and HM were highly reproducible.Table 3-2. Comparison of groundwater from the shallow overburden borehole PVP20as sampled using the SOLINST borehole sampler (S) and directly from the pump (P).Measurement a PVP20-S PVP20-P PVP20S /PVP20PDepth (m) 12.80 12.80 -TNC 10 4 (cells L 1 ) 32 (4.3) b 15 (7.6)2.1ATP 10 4 (amol L 1 ) 10.6 (0.77) 7.61 (0.39)CHAB 10 4 (cells L 1 ) 0.22 (0.042) 0.20 (0.023)CHAB of TNC (%) 0.68 1.33 0.5NRB (cells mL 1 ) 130 (50–390) c 2 (0.9–8.6) 65.0IRB (cells mL 1 ) 2.2 (0.9–5.6) 0.4 (0.1–1.7) 5.5MRB (cells mL 1 ) 8.0 (3–25) 2.3 (0.9–8.6) 3.5SRB (cells mL 1 ) 5 (2–15) 3 (1–12) 1.7AA (cells mL 1 ) 1600 (600–5300) 170 (70–480) 9.4HA (cells mL 1 ) 30 (10–130) 30 (10–120) 1.0AM (cells mL 1 )

63450 m in groundwater from borehole OL-KR19 (Table A-10). The overall average TNCover depth in deep groundwater was 5.7 10 4 cells mL 1 , which was almost ten timeslower than the average TNC over depth in shallow groundwater. The average TNC overdepth in deep groundwater did not display any trend with depth.Table 3-3. The total numbers of cells, ATP, and most probable numbers of variousphysiological groups of microorganisms in groundwater sampled using two differentPVB samplers, taken simultaneously on 14 March 2005 at the same location inborehole KFM06A, section 353.5360.0 m.Analysis a andphysiological groupSampleKFM06A:1 KFM06A:2 KFM06A:1/KFM06A:2TNC 10 4 (cells mL 1 ) 7.2 (1.7) b 5.2 (1.7) 1.4ATP 10 4 (amol mL 1 ) 1.51 (0.07) 0,95 (0.05) 1.6IRB (cells mL 1 ) 30 (10–120) c 23 (9–86) 1.3MRB (cells mL 1 ) 13 (5–39) 30 (10–130) 0.44SRB (cells mL 1 ) 0.8 (0.3–2.4) 0.4 (0.1–1.7) 2.0AA (cells mL 1 ) 30 (10–130) 24 (10–94) 1.3HA (cells mL 1 ) 24 (10–94) 24 (10–94) 1.0AM (cells mL 1 )

643.3.7 Most probable number of metabolic groups of bacteriaThe stacked number profiles of MPN values in shallow groundwater (Table A-8 andTable A-9) remained similar from season to season and were borehole specific in thecase of groundwater from several boreholes (Figure 3-25). The spring 2004 values areexcluded from the stacked MPN figures, as most of these values refer to boreholes thatwere not analysed again. Groundwater from boreholes PP2 and PP9 had low MPNvalues in all three seasons while samples from boreholes PR1 and PP39 had the higheststacked MPN values in all shallow groundwater samples. The spring 2006 value forborehole PVP1 groundwater was much higher than in the other two seasons, due to theabove-mentioned flooding event. There was no clear difference in the stacked MPNvalues between overburden (PVP) and shallow rock (PR, PP) boreholes.Table 3-4. The most probable numbers of various physiological groups ofmicroorganisms in two different boreholes sampled on 14 October 2004 and 9February 2005.Physiological group a KJ0052F01:1 KJ0052F01:2 KJ0052F01:1/KJ0052F01:2IRB (cells mL 1 )

65at a depth of approximately 328 m (Figure 3-26). The NRB analysis was introduced intothe sampling programme in 2005, so values are missing from some of the stacked MPNbars for deep groundwater. This may partly explain why the stacked bars for boreholesKR2, KR7, KR10, KR19, and KR27 are shorter than the remaining bars, whichincorporate NRB data.The MPN of NRB over depth displayed a range over four orders of magnitude in theshallow groundwater samples. The highest NRB value was found at a depth of 328 m inborehole OL-KR6 (Table A-11). The MPN of IRB was low in most samples with a fewvalues above 10 cells mL 1 in the shallow groundwater. The deep groundwater samplesdisplayed a peak relative to the other MPN values of IRB, with four IRB valuessignificantly above the detection limit at a depth of approximately 300 m. In the case ofMRB, the situation was similar to that of IRB, but with several more values above 10and 100 cells mL 1 in shallow and intermediate–depth groundwater, respectively. As forIRB, four of the MRB values peaked at a depth of approximately 300 m. The MPN ofSRB followed the trends of IRB and MRB with scattered values in shallow groundwaterup to approximately 1000 cells mL 1 and four values above the detection limit at adepth of approximately 300 m.The MPN results for AA and HA displayed similar patterns. The data were scatteredover a range of four orders of magnitude in the shallow groundwater. At a depth ofapproximately 300 m, there was a peak in the MPN values as was also observed forNRB, IRB, MRB, and SRB. There were some detectable AM and HM in shallowgroundwater and there were very few detectable methanogens at depth. As in all otherMPN analyses, peak values were observed at a depth of approximately 300 m.The MPN analysis of MOB was only performed on shallow groundwater. Thismetabolic group of microorganisms was present in most samples analysed, with a peakobserved in borehole PVB3B water in spring 2004. This borehole was turbid as a resultof dispersed bentonite from the packer of the casing, which may have influenced theresults.

660100Depth (m)2003004005003.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.010 Log(TNC) (cells mL 1 )Figure 3-21. The distribution of total number of cells (TNC) versus depth in Olkiluotogroundwater.0100Depth (m)2003004005003.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.010 Log(ATP) (amol mL 1 )Figure 3-22. The concentrations of ATP distributed over depth in Olkiluotogroundwater.

67024Depth (m)68101214160.0 1.0 2.0 3.0 4.0 5.010 Log(CHAB) (cells mL 1 )Figure 3-23. The distribution of cultivable heterotrophic aerobic cells (CHAB) versusdepth in shallow Olkiluoto groundwater.0100Depth (m)2003004005000.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.010 Log(CHAB) (cells mL 1 )Figure 3-24. The distribution of cultivable heterotrophic aerobic cells (CHAB) versusdepth in Olkiluoto groundwater.

68PVP1, 3.9 m2005-10-112006-04-272006-10-12PVP13, 5.6 m2005-10-122006-04-262006-10-12PVP14, 9.0 m2005-10-132006-04-262006-10-10PVP4A, 10.2 m2005-10-122006-04-272006-04-272006-10-10PVP20, 12.8 m2005-10-132005-10-132006-10-10PR1, 6.0 m2005-10-102006-04-252006-10-11PP36, 12.1 m2005-10-102006-04-252006-10-09PP39, 14.1 m2005-10-112006-04-242006-10-11PP2, 14.7 m2005-10-122006-04-242006-10-11PP9, 14.7 m2005-10-132006-04-272006-10-090 2 4 6 8 10 12 14 16 18 20Stacked 10 log(MPN) (cells mL 1 )MOBHMAMHAAASRBMRBIRBNRBFigure 3-25. Stacked values of most probable numbers of various physiological groupsof microorganisms in shallow Olkiluoto groundwater. NRB = nitrate-reducing bacteria,IRB = iron-reducing bacteria, MRB = manganese-reducing bacteria, SRB = sulphatereducingbacteria, AA = autotrophic acetogens, HA = heterotrophic acetogens, AM =autotrophic methanogens, HM = heterotrophic methanogens, and MOB = methaneoxidizingbacteria.

69Borehole, depth (m)KR32, 34.6KR8, 57.3KR33, 70.6KR6, 73.7KR39, 88.2KR6, 94.1KR6, 101.8KR10, 106.0KR37, 111.6KR31, 122.4KR27, 193.5KR7, 249.4KR8, 260.7KR13, 294.0KR13, 294.0KR2, 306.2KR10, 316.0KR6, 328.4KR39, 344.8KR27, 391.7KR19, 449.60 2 4 6 8 10 12 14 16 18 20Stacked 10 log(MPN) (cells mL 1 )HMAMHAAASRBMRBIRBNRBFigure 3-26. Stacked values of most probable numbers of various physiological groupsof microorganisms in deep Olkiluoto groundwater. NRB = nitrate-reducing bacteria,IRB = iron-reducing bacteria, MRB = manganese-reducing bacteria, SRB = sulphatereducingbacteria, AA = autotrophic acetogens, HA = heterotrophic acetogens, AM =autotrophic methanogens, and HM = heterotrophic methanogens.0100Depth (m)2003004005000 1 2 3 4 510 Log(NRB) (cells mL 1 )Figure 3-27. The distribution of nitrate-reducing bacteria (NRB) versus depth inOlkiluoto groundwater.

700100Depth (m)2003004005000 1 2 3 4 510 Log(IRB) (cells mL 1 )Figure 3-28. The distribution of iron-reducing bacteria (IRB) versus depth in Olkiluotogroundwater.0100Depth (m)2003004005000 1 2 3 4 510 Log(MRB) (cells mL 1 )Figure 3-29. The distribution of manganese-reducing bacteria (MRB) versus depth inOlkiluoto groundwater.

710100Depth (m)2003004005000 1 2 3 4 510 Log(SRB) (cells mL 1 )Figure 3-30. The distribution of sulphate-reducing bacteria (SRB) versus depth inOlkiluoto groundwater.0100Depth (m)2003004005000 1 2 3 4 510 Log(AA) (cells mL 1 )Figure 3-31. The distribution of autotrophic acetogens (AA) versus depth in Olkiluotogroundwater.

720100Depth (m)2003004005000 1 2 3 4 510 Log(HA) (cells mL 1 )Figure 3-32. The distribution of heterotrophic acetogens (HA) versus depth in Olkiluotogroundwater.0100Depth (m)2003004005000 1 2 3 4 510 Log(AM) (cells mL 1 )Figure 3-33. The distribution of autotrophic methanogens (AM) versus depth inOlkiluoto groundwater.

730100Depth (m)2003004005000 1 2 3 4 510 Log(HM) (cells mL 1 )Figure 3-34. The distribution of heterotrophic methanogens (HM) versus depth inOlkiluoto groundwater.024Depth (m)68101214160 1 2 3 4 510 Log(MOB) (cells mL 1 )Figure 3-35. The distribution of methane-oxidizing bacteria (MOB) versus depth inshallow Olkiluoto groundwater.

754 DISCUSSIONThe microbiology of shallow and deep groundwater in Olkiluoto was analysed byMicrobial Analytics Sweden AB and Göteborg University for almost three years from2004 to 2006, and previously by Göteborg University between 1996 and 2000 (Table1-1). The extensive sampling and analysis programme from 2004 to 2006 has produceda very substantial database, including 60 analytical datasets on the microbiology ofOlkiluoto groundwater. This database comprises 39 complete analytical datasetsassembled on four sampling campaigns from measurements from 16 shallowobservation tubes and boreholes ranging in depth from 4 to 24.5 m. The database alsocontains 21 analytical datasets covering 13 deep boreholes ranging in depth from 35 to450 m. In addition, the database contains 33 completed analyses of gas covering 14deep boreholes ranging in depth from 40 to 742 m. Most of these analyses werecompleted before the onset of ONKALO construction, and the remaining samples werecollected before ONKALO construction had extended below a depth of 100 m;therefore, this dataset captures the undisturbed conditions before the building ofONKALO. Future sampling and analysis will reveal whether ONKALO constructionhas influenced biogeochemical conditions in the surrounding groundwater. If such aninfluence is found, it will, hopefully, be possible to model the underlying reasons forthis influence and to predict its continuation.The following discussion first deals with the selection of sampling points, procedures,and methods and the quality-control tests done. Then the research results will beevaluated and interpreted, after which a descriptive model of the microbiologicalprocesses deemed most important will be presented. Finally, the outcome of thereported research will be discussed with reference to ONKALO.4.1 Sampling procedures for shallow groundwaterFour microbiology and gas sampling campaigns have been performed in the shallowgroundwater of Olkiluoto: the first was in May 2004, the second was completed inOctober 2005, and the last two were completed in April and October 2006, respectively.The methods and techniques used generally worked well. Several activities wereconducted specifically to test and possibly improve the methods. The sampling andanalysis procedures were adjusted as deemed necessary to improve the quality andreproducibility of results. The strategies underlying the selection of boreholes andmethods and the outcome of the method tests are discussed next.4.1.1 Selection of sampled shallow groundwater boreholesThe boreholes selected for the 2004 and the 2005–2006 sampling campaigns differ.Four boreholes were used in both the 2004 and 2005–2006 sampling campaigns(PVP1A, PVP4A, PR1, and PP2, but six boreholes were abandoned after 2004 andreplaced with new ones in fall 2005. The selection of boreholes was changed for severalreasons. First, one borehole sampled in 2004 collapsed (PP8) and two becamecontaminated with dispersed bentonite from the packer of the casing (PVP3A andPVP3B). Two were found to have closely related chemistry and microbiology profiles

76(PVP4A and PVP4B), so one of them was abandoned (PVP4B). Finally, two of the2004 boreholes became inaccessible after 2004 (PP3 and PP7), due to the ongoingconstruction of ONKALO and new buildings. The new selection of boreholes made forthe 2005 field campaign was retained for the remaining three sampling campaigns. Fiveoverburden and five shallow rock boreholes were selected, to capture the largestpossible distribution of the content of dissolved solids. These particular boreholes werealso selected to ensure that the ONKALO construction would not interfere with repeatedsampling activities in the future (Figure 2-1). Such repeated field activities should beable to return datasets that represent a wide selection of Olkiluoto shallow groundwaterenvironments over time. It will be possible to continue monitoring these boreholes inthe future and evaluate whether the construction of ONKALO has influenced theshallow groundwater microbiology and gas content.4.1.2 Sampling of shallow groundwaterSampling of shallow boreholes for microbiology differed in many respects from thesampling of deep boreholes. The most obvious difference is that specific fractures in thedeep boreholes were isolated with packers and pumped out for several weeks beforesampling. The shallow boreholes were not supplied with packers and could only bepumped out for a few hours; instead, they were open and in contact with the air. It is ageneral practice to pump out a borehole before sampling. This ensures that standinggroundwater containing dissolved air, precipitation, and dust from the ground surface isremoved before a sample is taken. Therefore, groundwater was sampled after 1.5 h ormore of pumping. The boreholes extended to various depths, usually several metres,beneath the surface of the groundwater table. The groundwater flowing into a specificborehole during pumping may therefore be of several different origins. For example,one component may originate from very shallow groundwater layers, while a secondone may originate from a deeper inflow location. All this is related to the preferentialflow paths of the aquifers in the sampled ground. For consistency, the total depth of ashallow borehole is used here when discussing relationships between measuredparameters and depth.The stability of the chemical conditions during prolonged pumping was tested inborehole PVP4A in spring 2006. Samples were collected for physical, chemical, andmicrobiological analyses at times separated by 6 h of pumping. The volume of thegroundwater pumped out of the borehole between the sampling occasions wasapproximately 1440 L. When compared as ratios between the first and second samplingoccasions, the results indicated very small differences in most of the analysedparameters (indicated in blue in Table A-4). The only parameter that changedsignificantly was the ferrous iron content, which decreased from 3.05 mg L 1 to 1.64 mgL 1 . The groundwater chemistry conditions appeared to be very stable in this borehole.Although this test was only done once for one borehole, it can yet be concluded that thechemical conditions in the shallow boreholes were borehole specific and reproducibleover the seasons. A comparison of data from each borehole over the whole samplingperiod supports this conclusion. For example, comparing the amounts of total dissolvedsolids (TDS), which represents the sum of all dissolved species analysed for separatelyas well, indicates good reproducibility per borehole over time. It can be concluded that

77the applied pumping methodology rendered reproducible results with respect to mostphysical and chemical parameters.The deep boreholes were sampled using the PAVE system, which has up to three closedcontainers that collect pressurized samples. This procedure was initially used for theshallow boreholes, using the SOLINST borehole sampler (using a borehole sampler isoptional, as one can collect samples directly from the pump tubing). However, it had tobe demonstrated that the pumps could indeed be sterilized between pumping out thedifferent boreholes. It was also deemed important to test the difference between pumpedsamples and samples collected using the SOLINST sampler, as discussed below.4.1.3 The oxygen blockage packer testThe shallow boreholes were open and in contact with air. It was therefore deemedpossible that air might have mixed with the sampled water during pumping, as oxygenwas found in most samples in the May 2004 investigation. A packer system was testedin the October 2005 investigation to determine whether the pumping was enough tokeep air from contaminating the samples. The results indicated that oxygen did not mixwith the water, irrespective of whether or not the packer was used (Pedersen 2007). Itwas concluded that a packer is not needed to hinder oxygen in air from mixing with thesampled groundwater.4.1.4 Sterilization of borehole pumpsThe use of chlorine dioxide (FreeBact 20; XINIX) for pump sterilization worked verywell (Table 3-1). The MPN of microorganisms obtained after sterilization was belowthe detection limit of 0.2 cells mL 1 for all analyses except NRB, which produced anumber just above the detection limit. It can thus be concluded that the sampling pumpsystems did not cross-contaminate the sampled boreholes.4.1.5 Comparison of sampling using the SOLINST sampler and using theborehole pumpIn the case of borehole PVP20, the microbiology results obtained from samples madewith the borehole pump were generally equal to or lower than those obtained fromsamples made with the SOLINST sampler, except in the case of MOB (Table 3-2). Thelargest difference was found in the case of NRB and AA. When sampling with theSOLINST sampler, it was observed that the groundwater became slightly more turbidafter retrieving the borehole pump and lowering the SOLINST sampler. The increase inturbidity was only observed in overburden boreholes. This difference most likelyresulted from hydrodynamic disturbance caused by raising and lowering the pump andsampler in the boreholes. Sediment and colloids that became suspended in groundwaterdue to disturbance during sampling would certainly harbour attached microorganisms,which would subsequently increase the biomass estimates in turbid as compared withnon-turbid groundwater. The greater ATP biomass value in borehole PVP20S than inborehole PVP20P may be attributed to the fact that the ATP analysis method extractedATP from both planktonic and biofilm microorganisms on particles in the turbid water.In the case of total counts, the situation was similar, with higher numbers evident in the

78SOLINST sample than in the pump sample. Groundwater from borehole PVP20S wasalso associated with the detection of the greatest number of metabolic groups. Thiswould again be due to higher numbers of microorganisms in the SOLINST samplescaused by the presence of more sediment particles with attached microorganisms.In choosing whether to use the SOLINST or pump method for sampling, it can beargued that the SOLINST method gives higher microorganism numbers related toparticles in the groundwater. These are of course true results, in that these organismswere indeed present and possibly active in the sampled borehole. At the same time, theSOLINST method introduced uncertainty into the results, as it is impossible toreproducibly cause turbidity in the boreholes. For comparative purposes, it was deemedbetter to sample from the borehole pump in the field activities in 2006 and in the future.Otherwise, results from overburden borehole samples may overestimate the planktoniccell numbers compared with results from bedrock borehole samples that were free ofturbidity caused by sampling activities in the borehole.It has been demonstrated that attached microorganisms in deep groundwaterenvironments significantly outnumber planktonic microorganisms (Pedersen andEkendahl 1992a, b; Pedersen 2001). However, it can be assumed that the planktonicnumbers and diversity reflect the numbers and diversity of attached microorganisms.High activity and growth of attached microorganisms will result in an increased numberof microorganisms that slough off due to hydrodynamic forces or that migrate fromgrowing colonies. The investigations of shallow groundwater were intended to build ourunderstanding of seasonal variation and of the future impact of ONKALO ongroundwater biogeochemistry. It was thus deemed more important to use reproduciblemethods than methods that may yield the highest possible numbers as a result ofmanipulating the particle content of the analysed groundwater. The two first fieldactivities reported here, those of May 2004 and October 2005, formed a solid methodand technology basis for the design of future field activities. The only major changemade in the 2006 investigations was that samples were taken directly from the pumpsfor microbiology analysis. The SOLINST sampler was used one more time, to obtainsamples for the analysis of groundwater gas content and composition, but only tocompare the results obtained using SOLINST with those using a glass bottle samplingmethod, as discussed below (4.4.3).4.2 Sampling procedures for deep groundwater microbiologyThe reproducibility test for groundwater samples taken simultaneously in boreholeKFM06A from the 353.5360.0-m depth section in Forsmark using the PVB samplingsystem indicated very good reproducibility (Table 3-3). The difference between the twosamples was generally represented by a difference of only a single tube in the MPNanalyses. The 95% confidence interval for MPN analyses using five parallel tubesequalled approximately 1/3 and 3 times the obtained value (Greenberg et al. 1992). Themaximum difference between the two samples was 1.6, so the two samples were notsignificantly different. In comparison, there were significant differences between thegroundwater samples analysed in the different Äspö HRL boreholes (Table 3-4).Groundwater from borehole KJ0052F01 yielded significantly higher values in allanalyses than did groundwater from borehole KJ0052F03. The results from the Äspö

79HRL had previously indicated the heterogeneity of microbial populations in fracturedrock. Boreholes KJ0052F01 and KJ0052F03 are only separated by a maximum of 50 mof rock (Pedersen 2000b); however, they intersect fractures with differenthydrogeochemistry characteristics, resulting in groundwaters with very differentmicrobiological profiles. In conclusion, the reproducibility test demonstrated theanalytical protocols for microbiological analyses to be reproducible between samples.The reproducibility of the sampling and analysis methods used for groundwater fromborehole sections over a 3.5-month interval was also tested. This test requiredgroundwater samples from borehole sections that did not change their microbiologyprofiles over time. The Äspö HRL MICROBE site was suitable for such a test, becausethere is a record of groundwater analysis results extending back to the May 1999drilling (Pedersen 2000b). For the reproducibility testing, six consecutive samplings formicrobiological analysis were performed over 20 months (Pedersen 2005b), includingthe two results presented here. It was found that the results per borehole section werereproducible over time, as shown in Table 3-4. With a few exceptions, the results didnot differ significantly between the sampling times. The two different groundwatersamples tested were, however, significantly different. Taking all these results intoconsideration, it must be concluded that testing over time indicated extremely goodreproducibility. The MPN procedures and the analyses appeared to be very robust andreproducible both between samples (as shown in Table 3-3) and over sampling timesand between boreholes as shown in Table 3-4.The stability of the microbiological analyses over prolonged pumping was tested inborehole PVP4A in spring 2006, as was done for physical and chemical parameters asdescribed above (section 4.1.2). Samples were collected for microbiology analysis attimes separated by 6 h of pumping. The volume of the groundwater pumped out of theborehole between the sampling occasions was approximately 1440 L. When comparedas ratios between the first and the second sampling occasions, the results indicatedvarying differences in the analysed parameters (indicated in blue in Table A-7). TheCHAB results displayed the largest difference, decreasing 13.5-fold between the firstand second sampling occasion. The MPN of HA also decreased significantly, by afactor of ten. These differences may, however, be due to a change in the microbiologicalcomposition of the groundwater, rather than to any variability in the methods used. Thepumping may have had a filtering effect on the unattached cells, which would result indecreasing values, as was the case for all analysed parameters except AA. The ATPvalue also decreased, which suggests that the volume of active biomass actually wasreduced somewhat. Still, the variability of the PVP4A data appeared reasonablycoherent compared with the variability between boreholes (Figure 3-25).4.3 Evaluating the analysis methodsAn array of analytical methods was used in this research to explore the microbiology ofthe Olkiluoto site groundwater. Correctly interpreting the results of the performedanalyses depends on a basic understanding of the possible limitations of the methodsand of the overlaps and gaps in the dataset. Some methods, like those used to measuretemperature, produce reliable data with clear meaning in most situations, while

80interpreting and understanding the results of other methods, such as MPN analysis,require a thorough knowledge of the underlying rationale of the procedures and of theirlimitations, sensitivity, and precision. In the following sections, the methods used willbe evaluated in the context of the results obtained.4.3.1 Analysis of physical parametersThe temperature (Table A-1), pH (Figure 3-1), and conductivity (Figure 3-2) ofOlkiluoto groundwater have long been analysed, and there should be few problems withthe results of these analyses. In contrast, dissolved oxygen in shallow groundwater hasnot been analysed on an ongoing basis, and the dataset presented here for oxygen is thusimportant for the Olkiluoto site. Three methods were used to measure dissolved oxygen:electrochemical analysis in the field with electrodes (Figure 3-3), chemical analysis inthe laboratory with titration (Figure 3-3), and gas chromatography in the laboratory(Figure 3-10). The results of the electrochemical analysis agreed well with those oftitration (Figure 3-4). The titration method was more reliable at low oxygenconcentrations below approximately 0.5 mg O 2 mL 1 , because electrodes are difficult tocalibrate for concentrations close to zero. The gas chromatography method did not workwell, due to problems with air intrusion in the glass sample vessels during extraction.The electrochemical method has the advantage of being easy to perform by personnel inthe field, while Winkler titration requires a chemist in the laboratory to titrate thesamples. When large datasets are required for routine data collection, as is the case withseasonal variation in many boreholes (Figure 3-6), the electrochemical method is themost cost effective.4.3.2 Chemical parametersChemical analysis is routine work in Olkiluoto and has recently been reviewed(Pitkänen et al. 2007), and so will not be evaluated again here. In general, Pitkänen et al.(2007) concludes that much work and experience have been gained over time regardinghow to obtain high-quality data. This experience was available and used when thehydrochemical data used in the present report were collected. The concentrations ofDOC, ferrous iron, sulphide, and sulphate in groundwater are of special interest formicrobiological investigations, because these chemical species play significant roles inmicrobial processes (Figure 1-8).Several groundwater gases can be produced and consumed by microorganisms.Methanogens produce methane from hydrogen and carbon dioxide and acetogens canproduce acetate from hydrogen and carbon dioxide. Microbial metabolism of organiccarbon generates carbon dioxide and some microorganisms metabolize methane tocarbon dioxide. Consequently, research into microbial processes in groundwaternecessitates the development of analytical methods for detecting gases in environmentaland laboratory samples. Dissolved groundwater gases in deep Olkiluoto groundwaterhave been sampled using the PAVE system and analysed since 1997. Previously, gashad been sampled using glass and aluminium vessels (Gascoyne 2005). Seventy-onedeep groundwater samples taken using PAVE between 1997 to 2005, together withassociated analyses, have been evaluated in detail elsewhere (Pitkänen and Partamies

812007), so that evaluation is not repeated here. Briefly stated, the authors report that theamount of gas is notable at depth and that the major gases are nitrogen and methane.The gas composition closely follows the stratification of redox conditions, a significantshift observable at a depth of approximately 300 m. Furthermore, they conclude that gasformation is of substantial importance for repository safety and that is essential to obtainmore data regarding hydrogen, methane, hydrocarbons, dissolved inorganic carbon,fracture calcites, and microorganisms. Further studies should examine the interfacebetween the methanic and sulphidic systems below and above a depth of 300 m,respectively. The work presented in the present report is a first step in the directionidentified by Pitkänen and Partamies (2007).In the present research, two methods have been used to collect gas in shallow and deepgroundwater for subsequent extraction and analysis, as described in the Appendix (seepage 149). Evacuated glass bottles were used for shallow groundwater analysis and thePAVE system was used for deep groundwater sampling and analysis. As a rule ofthumb, the larger the water sample and the more the gas, the better the precision anddetection obtained. Samples of deep groundwater from Olkiluoto have usuallycontained large volumes of gas, while shallow groundwater has contained less dissolvedgas (Figure 3-9). The glass bottle sampling method still worked well because theuncertainty of using single samples and extractions was compensated for by usingtriplicates of independent samples, extractions, and analyses. In future research, theglass bottle methods used for shallow groundwater can be improved by using largerbottles. An attempt to use the SOLINST sampler for gas did not turn out well, due tocontamination of the sample with the nitrogen used to open and close the sampler(2.1.5). A similar problem was occasionally encountered with the PAVE samplingequipment, where the gas in the pressure compartment, i.e., nitrogen or argon,contaminated the samples (3.2.2). Finally, there was a problem with the aircontamination of samples due to poor (i.e., not vacuum-tight) seals on the originaldevice used to connect the PAVE pressure vessel and the extraction equipment. Thisproblem was later solved by constructing a new device with vacuum-tight seals.4.3.3 Microbiological parametersThe microbial biomass in granitic rock aquifers of the Fennoscandian Shield has beenanalysed in terms of total and viable numbers for almost two decades (Pedersen 2001);total number estimates have ranged from 10 3 to 10 6 cells mL 1 , while viable numberestimates have ranged from 10 0 to 10 5 cells mL 1 . Between 0.00084 and 14.8% of thetotal numbers have been cultivated and detected using most probable number (MPN)methods (Haveman and Pedersen 2002a). Although low viable numbers have beendetected relative to the total numbers observed, in vitro radiographic and radiotracerestimates have suggested that the absolute majority of the total cells observed usingmicroscopy was viable (Pedersen and Ekendahl 1990, 1992a, b). Consequently, therewas a significant gap between estimates of potentially viable total numbers andevidently viable cultivable numbers. Hence, a method for estimating the total amount ofviable biomass in groundwater was sought. A recent investigation found that analysingthe ATP concentration in shallow and deep Fennoscandian groundwater (includingOlkiluoto groundwater) using a commercial assay supplied needed information about

82the metabolic state and biovolume of the bacteria present (Eydal and Pedersen 2007).The assay appeared robust and reliable and had a detection range that took in allsamples analysed. The analysed ATP concentrations were found to correlate both withthe microscopic counts and with the volume and metabolic status of the investigatedpure culture and groundwater cells. The results suggested that bacterial populations indeep groundwater vary significantly in size, and that metabolic activity is a function ofprevailing environmental conditions.ATP was first analysed in Olkiluoto groundwater in fall 2004. When ATP was analysedconcomitantly with TNC, a good correlation was obtained (Figure 4-1). As ATP is anenergy transport compound present in all living cells (confer Figure 1-11), measuring itsconcentration indicates the biovolume and metabolic state of the biomass in any system.A groundwater containing many active cells should thus have a higher ATPconcentration than one containing few such cells. If the cells are large, this will increasethe content of ATP per cell. It has been demonstrated that the ATP/TNC ratio is a goodindicator of the metabolic activity of cells in groundwater (Eydal and Pedersen 2007).The average ATP/TNC ratio for 109 shallow Olkiluoto groundwater determinations was1.02, and for 166 deep Fennoscandian shield groundwater determinations was 0.43. Anyratio higher than these two in shallow or deep groundwater, respectively, thus suggeststhat the microbial population analysed is more active than average, while a lower-thanaverageratio suggests that the population analysed is less active than average.The MPN methods for enumerating microorganisms in deep groundwater was first usedfor analysing methanogens and acetogens in Äspö HRL groundwater (Kotelnikova andPedersen 1998). Later, the methods was further developed, and it has been used toanalyse more types of microorganisms in deep groundwater from Finland (Haveman etal. 1999; Haveman and Pedersen 2002a), including from Olkiluoto (Table 1-1), andfrom the natural nuclear reactors in Bangombé, Gabon, Africa (Haveman and Pedersen2002b). The methods have been modified and changed over time. As the numbers ofsamples and types of organisms analysed have increased, the manual preparation ofsingle tubes, as used for analysing methanogens and acetogens in Äspö HRLgroundwater (Kotelnikova and Pedersen 1998), has had to give way to methods thatcould handle the approximately three thousand MPN tubes (2.4.4) needed during eachof the Olkiluote field investigations of shallow groundwater.The expression “the great plate count anomaly” was coined by Staley and Konopka(1985) to describe the difference in orders of magnitude between the numbers of cellsfrom natural environments that form colonies on agar media (CHAB) and the numberscountable by means of microscopic examination (TNC). In general, only 0.01–0.1% ofbacterial cells sampled from various environmental aquatic systems produce colonieswhen using standard plating techniques so, as expected from the relevant literatureresults, there were no correlations between TNC and CHAB data for Olkiluotogroundwater (Figure 4-2). The anaerobic cultivation methods presented here representthe culmination of almost 10 years of development, testing, and adaptation for deepgroundwater. The success and usefulness of these methods are reflected in themaximum MPN cultivability of 30% of the TNC in the sample from the borehole OL-KR6 422–425 m section and the 0.01–30.25% MPN cultivability range in allgroundwater samples (Table A-10). The use of multiple, liquid anaerobic media (Table2-3) has obviously overcome much of the discrepancy found between TNC and

83cultivations that use agar media only. However, it should be understood that there maystill be microorganisms in the groundwater not cultivable using the applied methods.One example is that of anaerobic methane-oxidizing bacteria (ANME), which as of thetime of writing have escaped successful cultivation by the world microbiologycommunity. ANME have been observed in environmental samples but their successfulcultivation in the laboratory has yet not been described in the literature.The CHAB and MOB were analysed under aerobic conditions, unlike all othercultivation methods, which were performed under anaerobic conditions. Many bacteriaare known to be facultative anaerobes, i.e., they can switch from aerobic respirationusing oxygen to anaerobic respiration using nitrate and often also ferric iron andmanganese(IV) as alternative electron acceptors (Madigan and Martinko 2006).Microorganisms in groundwater must be adapted to anoxic conditions but, if oxygenshould appear, it is advantageous for the microbe to be able to switch to oxygenrespiration. Indigenous groundwater microorganisms should consequently be detectableas both CHAB and NRB, while contaminants from the surface should have a smallertendency to do so. Comparing the CHAB data to the NRB data indicates a reasonablygood correlation (Figure 4-3), suggesting that the microorganisms analysed as CHABwere generally indigenous.Some of the metabolic groups analysed using MPN may overlap in numbers. At theonset of this investigation it was unclear whether AA and HA would differ in numbers.The acetogens are known to be a diverse group of organisms that may switch betweendifferent metabolic states (Drake et al. 2002). Comparing the MPN numbers of AA andHA indicates that they correlated well, although there was a clear tendency for AA tooutnumber HA in several samples (Figure 4-4). Similarly, it is known that one organismcan have the abilities to reduce both iron and manganese (DiChristina and DeLong1993). Comparing the IRB with the MRB numbers indicates that MRB tended tooutnumber IRB in several samples (Figure 4-5). More research will be needed beforewe have a full understanding of the potential differences between AA and HA, andbetween IRB and MRB numbers.

847.06.510 Log(ATP) (amol mL-1 )6.05.55.04.54.0ATP = 0.6309+0.7795*x; 0.95 Conf.Int.3.53.03.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.010 Log(TNC) (cells mL -1 )Figure 4-1. The relationship between the total number of cells (TNC) and theconcentration of ATP in Olkiluoto groundwater. Dashed lines denote the 95%confidence interval.7610 Log(TNC) (cells mL1 )5432100.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.010 Log(CHAB) (cells mL 1 )Figure 4-2. The relationship between the total number of cells (TNC) and the numbersof cultivable aerobic heterotrophic bacteria (CHAB) in Olkiluoto groundwater.

8554NRB = 0.8179*CHAB 0.180310 Log(NRB) (cells mL1 )32100.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.010 Log(CHAB) (cells mL 1 )Figure 4-3. The relationship between the numbers of cultivable aerobic heterotrophicbacteria (CHAB) and the most probable numbers of nitrate-reducing bacteria (NRB) inOlkiluoto groundwater. Dashed lines denote the 95% confidence interval.4.03.510 Log(HA) (cells mL1 )3.02.52.01.51.00.50.00.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.010 Log(AA) (cells mL 1 )Figure 4-4. The relationship between AA and HA in Olkiluoto groundwater samples.The line denotes a one-to-one relationship.

864.03.510 Log(IRB) (cells mL1 )3.02.52.01.51.00.50.00.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.010 Log(MRB) (cells mL 1 )Figure 4-5. The relationship between IRB and MRB in Olkiluoto groundwater samples.The line denotes a one-to-one relationship.4.4 Geochemical conditions of the investigated aquifersEstablishing baseline conditions in Olkiluoto before the start of excavations forONKALO was deemed very important (Andersson et al. 2007b; Pitkänen et al. 2007).These baseline conditions now include the microbiological data from 2004 to 2006 citedin the present report as well as microbiological data from the earlier site investigations(Table 1-1). Seasonal variations may have an effect on microbial processes that will besignificant in shallow groundwater, but will rapidly diminish with increasing depth. Thepossible anomalous mixing of groundwater at a depth of approximately 300 m due tothe long-term pumping of open boreholes has been established (Pitkänen et al. 2007). Itwas speculated that shallow groundwater may flow through open boreholes to thesubhorizontal hydrogeological zone HZ20, which would act as an outflow route forgroundwater in Olkiluoto (Figure 7-7, Andersson et al. 2007b). This type of mixing mayexpose microorganisms to anomalous conditions at a depth of approximately 300 m.What geochemical parameters are then important for microbiology? Temperatureinfluences the reaction rates of most chemical processes. The rate of microbialprocesses thus increases with increasing temperature. The pH is less important formicroorganisms, as they generally have large ranges of several pH units within whichthey can be active. Conductivity reflects the amount of dissolved solids and, as with pH,microorganisms generally have a relatively large range of dissolved solidsconcentrations within which they can be active (see Figure 1-3 for an example of thetemperature, pH, and NaCl ranges for a deep groundwater microorganism). Thepresence and concentration of oxygen has a profound influence on microbial diversity

87and activity. Many deep groundwater microorganisms are killed by oxygen, but thosethat can use oxygen as an electron acceptor in their metabolisms use it more efficientlythan the electron acceptors they utilize under anaerobic conditions (Figure 1-8). Thepresence and concentrations of oxidized and reduced electron acceptors and donors,presented in Figure 1-8, are important to understand. For example, DOC can be used asa source of energy and carbon by heterotrophic microorganisms, while autotrophicmicroorganisms produce DOC, for example, acetate production by AA. Finally, E h is animportant parameter. Microbial processes tend to lower the E h of any system in whichthey are active by consuming oxygen and producing reduced electron acceptors such asferrous iron and sulphide. Next, the conditions in the investigated groundwater, withrespect to the above parameters, are discussed.4.4.1 Physical parametersThe average shallow groundwater temperature was 7.1C in spring and 9.1C in fall(Figure 4-6). The difference in temperature between seasons was most pronounced atdepths of less than 10 m, where the water temperatures in shallow boreholes such asPVP1 and PR1 differed by up to 6C. The pH range was almost three units, 4.7–7.7, inthe shallow groundwater, and stabilized above 7 at depth (Figure 3-1). The effect of thedifferent pH values on microbial processes will be indirect, as pH influences manygeochemical parameters such as mineral dissolution and precipitation, carbon dioxidesolubility, and various solid–aqueous phase equilibria. Microbial processes producecarbon dioxide from the respiration of DOC. Less carbon dioxide will precipitate ascalcite at low pH in dilute shallow groundwater than in deep groundwater where the pHand dissolved solids concentration are higher. This is reflected in the carbon dioxideconcentrations, which were much higher in shallow than in deep groundwater. The inputof biodegradable organic carbon from surface plant and animal ecosystems can beassumed to be higher in shallow than in deep groundwater; thus, the production rate ofcarbon dioxide by microorganisms will be higher in shallow than in deep groundwater.The concentration of oxygen decreased rapidly with depth in the shallow groundwater(Figure 3-3). Many microorganisms prefer oxygen for their metabolisms, so oxygen isthe first electron acceptor to disappear in groundwater. Thereafter, the microorganismsuse other electron acceptors that, when reduced, will force the E h towards more negativevalues. In shallow groundwater, the measured E h was lower at low oxygenconcentrations (Figure 4-7). This correlation supports the general model ofmicroorganisms as important moderators of E h in groundwater, just as they are in manyother systems, for example, in aquatic sediments (Madigan and Martinko 2006).The variation in temperature over the seasons should correlate with the input of organicmaterial to shallow groundwater. In winter and fall, the input will be low as will be thetemperature. This will decrease the rate of microbial processes with a concomitantreduction in the consumption of oxygen. In summer and early fall, the input of organicmaterial will increase as will the temperature. Microbial processes will speed up and theconsumption of oxygen should increase. Such seasonal variation in oxygen was alsonoted over a one-year cycle (Figure 3-6). In summer 2006, only two of a total of 10boreholes had significant concentrations of oxygen. One of these was the shallowborehole PVP1, where the groundwater surface almost coincides with the sampling

88depth in summer, and oxygen easily dissolves. Oxygen has not been routinely measuredin the shallow Olkiluoto groundwater programme lately, so the reproducibility over theyears is impossible to test. It is thus strongly recommended that oxygen be added to theshallow groundwater analysis protocols. The intrusion of oxygen into the ONKALO siteis undesired. If any such intrusion should occur, its extent would probably be seasonrelated. However, microbial processes will continue to reduce oxygen in thegroundwater and form a biological E h front that will possibly fluctuate up and downwith the season. The fluctuation range will probably be relatively narrow, as judgedfrom the results of previous experiments. Research conducted during the construction ofthe Äspö HRL tunnel could not confirm the intrusion of oxygen via a 70-m-long faultthat was intersected by the tunnel (Banwart et al. 1994, 1996). More than 15 years havepassed since that large-scale experiment was finished, and oxygen has still not reachedthe tunnel. The current explanation of this is that continuous microbial processes havereduced oxygen in the intruding groundwater, as explained above.4.4.2 Chemistry dissolved solidsThe concentrations and distribution of the electron acceptors oxygen, nitrate,manganese(IV), ferric iron, and sulphate are important to analyse, as these speciesdetermine what microbial processes are or are not possible with respect to availableelectron acceptors. Oxygen was discussed above, and ferric iron and manganese(IV) aresolids that cannot be analysed in groundwater. When reduced, water, carbon dioxide,nitrogen, ferrous iron, manganese(II), sulphide, methane, and acetate are formedaccording to the reactions in Figure 1-8. The chemical analyses can detect the presenceof DOC, the oxidized electron acceptors nitrate and sulphate, and the reduced electronacceptors sulphide, ferrous iron and manganese(II).The distribution of DOC was scattered over the analysed depth (Figure 3-7), with noclear trends over depth. However, when compared with the concentration of ATP(Figure 4-8) a weak trend was evident, high concentrations of ATP being correlatedwith high concentrations of DOC (r = 0.98, p < 0.05, n = 24). This strongly suggests arelationship between microbial activity and DOC concentration in Olkiluotogroundwater, which is perfectly in line with our understanding of microbial processes.Heterotrophic microorganisms consume DOC and autotrophic ones produce DOC andthey all contain ATP. It is, however, impossible to conclude from concentrations onlywhich of these two processes dominates.The concentration of nitrate was below the detection limit in most samples (Table A-4).This implies either that nitrate was not present in the analysed deep groundwater at all,or that it is consumed as soon as it appears from unknown sources. The general presenceof high numbers of NRB (Figure 3-27) suggests that nitrate is turned over immediatelyif it appears. It is important to realize that knowing the concentrations of theconstituents of a microbial process is not in itself enough to predict the relevance of theprocess. Turnover rates are needed, as exemplified here by nitrate. Sulphate wasscattered over a large concentration range at depths down to 400 m, after which thesulphate concentration approached zero (Figure 4-9). Some of the shallowestgroundwater samples were very dilute and had low sulphate concentrations as well. Thisprofile implies that microbial sulphate reduction is currently possible to a maximumdepth of 400 m at Olkiluoto.

890246Depth (m)810121416182 4 6 8 10 12 14Temperature (C)SpringFallFigure 4-6. Temperature distribution over fall and spring in shallow Olkiluotogroundwater. The dashed lines indicate the average spring (blue) and fall (red)temperatures.500400300E h (mV)2001000-100-2000 1 2 3 4 5 6 7O 2 (mg L 1 )Figure 4-7. The relationship between E h and dissolved oxygen in shallow groundwateranalysed using the pIONeer 10 portable pH meter and the HQ10 HACH PortableLDO dissolved oxygen meter, respectively.

90The concentration of manganese(II) was scattered over depth, displaying somewhathigher values in shallow than in deep groundwater (Table A-4). Ferrous iron andsulphide displayed different profiles with depth (Figure 3-8). Ferrous iron displayed atendency to decrease exponentially with depth, while sulphide was low in all samplesanalysed in this work, except those from a depth of approximately 300 m. If allavailable ferrous iron data are plotted, it becomes clear that the concentration range inshallow groundwater is ten times the range in deep groundwater (Figure 4-10). There isno clear trend in ferrous iron concentration in the deep groundwater. The peak insulphide concentration at approximately 300 m is very obvious in the scatter plot of allOlkiluoto data (Figure 4-11). This peak is almost 100 times the background sulphideconcentrations at all other analysed depths. This profile is indicative of intensivesulphate reduction at a depth of 300 m but, before discussing this indication, data ongases are needed.1000000ATP (amol mL 1 )1000001000010001 10 100DOC (mg C L 1 )Figure 4-8. The relationship between dissolved organic carbon (DOC) and ATP inOlkiluoto groundwater.

910100200300Depth (m)4005006007008009000 100 200 300 400 500 600Sulphate (mg L 1 )Figure 4-9. The distribution of sulphate in Olkiluoto groundwater. All availableOlkiluoto data between 1992 and 2006 have been included in the scatter plot.0100200300Depth (m)4005006007008009000.00 0.01 0.10 1.00 10.00 100.00Fe 2+ (mg L 1 )Figure 4-10. The distribution of ferrous iron in Olkiluoto groundwater. All availableOlkiluoto data between 1992 and 2006 have been included in the scatter plot. The valueof observations of ferrous iron below the detection limit was set to 0.005 in the scatterplot.

920100200300Depth (m)4005006007008009000.00 0.01 0.10 1.00 10.00Sulphide (mg L 1 )Figure 4-11. The distribution of sulphide in Olkiluoto groundwater. All availableOlkiluoto data between 1992 and 2006 have been included in the scatter plot. The valueof observations of sulphide below the detection limit was set to 0.005 in the scatter plot.4.4.3 Origins and amounts of dissolved gases in Olkiluoto groundwaterThe origins of the gases observed in Olkiluoto groundwater depend on variousmechanisms, as discussed elsewhere (Gascoyne 2005; Pitkänen and Partamies 2007).These reports suggest that nitrogen mainly originates from the entrapment ofatmospheric nitrogen during groundwater recharge. However, this process would notexplain the excess of nitrogen at depth. It is much more likely that the nitrogen ingroundwater originates from crustal degassing of the bedrock, as outlined below. Thehighest amount of nitrogen gas found in the present investigation was 247 mL nitrogenL 1 groundwater 1 in borehole OL-KR29 at a depth of 742 m (Figure 3-14), with a totalgas volume of 1380 mL L –1 groundwater 1 (Figure 3-9) at atmospheric pressure. Thiscorresponds to 10.6 mmol L –1 or 0.31 g nitrogen per L of groundwater. The solubility ofnitrogen gas in water at atmospheric pressure at 10C is 0.024 g kg -1 (Aylward andFindley 2002), which corresponds to 0.9 mmol L –1 or 21 mL L –1 . The amount ofnitrogen dissolved in the OL-KR29 groundwater sample was almost 12 times higherthan its solubility would permit at atmospheric pressure. In reality the amount ofnitrogen is even higher, because other shallow groundwater gases, in particular carbondioxide and oxygen, will reduce the solubility of nitrogen during the entrapment ofatmospheric nitrogen. The numbers used here are thus conservative. It is deemed veryunlikely that the observed excess of dissolved nitrogen in the deep Olkiluotogroundwater originates from atmospheric entrapment during groundwater recharge.Most dissolved nitrogen must instead originate from deep crustal sources. The increase

93in nitrogen concentration is exponential over depth (Figure 3-14), which suggests thatdiffusion is the major process transporting nitrogen from deep crustal sources. Actually,only a few of the shallowest groundwater samples have nitrogen concentrations at orbelow the solubility limit (21 mL L 1 ) at atmospheric pressure (Figure 3-14); theabsolute majority of the gas samples had nitrogen concentrations above this limit,supporting the suggested deep crustal origin.There are four helium reservoirs on Earth, namely, the air, crust, and upper and lowermantle (Apps and van de Kamp 1993). Since helium cannot be retained in theatmosphere by gravity, its concentration in air is very low (5.24 ppm by volume). Thehelium in air comes mainly from out-gassing of the continental crust and degassing ofthe mantle. Helium is present as a mixture of two stable isotopes, 3 He and 4 He, inabundances of 1.38 10 –4 % and 99.999862%, respectively. 3 He is mainly of primordialorigin but is also produced by beta decay of 3 H to 3 He, though this reaction is rare. 4 Heis produced by radioactive decay of the uranium- and thorium-series radionuclides.Helium is constantly produced in the crust and mantle by means of these reactions(Marshall and Fairbridge 1999). Consequently, the rate of diffusion of helium to theatmosphere is controlled by its production rate at depth, as inferred by the exponentialincrease with depth of helium in the analysed groundwater (Figure 3-15).The total amount of dissolved gas in Olkiluoto groundwater increased exponentiallywith depth down to the deepest level examined in this work, which was 742 m (Figure3-9). This profile suggests that most of the analysed gases, except carbon dioxide, aremigrating from deep underground towards the surface, as discussed above in the case ofhelium. The three major gases present were carbon dioxide, nitrogen, and methane(Figure 3-11). Carbon dioxide comprised 20–50% of the extracted gas in samples fromshallow groundwater. Thereafter, nitrogen dominated down to a depth of approximately300 m, at which point methane started to account for a significant part of the dissolvedgas, becoming the dominant gas in samples from 320 m and deeper. The concentrationsof dissolved methane and ethane increased markedly by approximately 100 times inanalysed groundwater from depths below 300 m (Figure 3-16, Figure 3-20). Theseobservations are in line with previous results (Pitkänen and Partamies 2007).There are two possible hypotheses explaining of the sharp shift in methaneconcentration at a depth of approximately 300 m. Hypothesis 1 (H1): If there is a flowof groundwater containing low concentrations of methane and ethane inhydrogeological zone HZ20 (Andersson et al. 2007b), it would replace high-methaneconcentrationgroundwater and a rapid drop in the concentrations would result, just as isobserved. Hypothesis 2 (H2): There is a process at a depth of approximately 300 m thatconsumes methane and possibly also ethane. Methane and ethane are accompanied byother gases on its diffusion towards the surface from their respective origins. Helium isan inert gas and thus cannot be consumed or precipitated in any way. A dilution effectof flowing groundwater according to H1 should result in the equal dilution of bothhelium and methane, the ratio of which should not change. If H1 is valid, then themethane/helium ratio should be approximately the same over most of the analysed depthrange. Helium diffuses somewhat faster than methane does, so a slight increase in theratio can be assumed under H1. Inspecting this ratio over depth reveals that the ratiodecreases distinctly by approximately 1000-fold from 300 m up to 200 m (Figure 4-12).

94This effect is less pronounced for ethane/helium ratio. It seems clear that H2 is valid, atleast for methane. However, the methane/helium ratio can also drop if the concentrationof helium should increase for some unexpected reason. Plotting nitrogen, which likehelium is also an inert gas, against helium results in a ratio that increases significantly atdepths of less than 300 m (Figure 4-13). This indicates that helium is decreasing inconcentration relative to nitrogen; it could also indicate that the nitrogen concentrationis increasing, except that such an increase was not observed at depths of less than 300m. If helium actually is decreasing in concentration due to its more rapid diffusion, thenthe methane/helium ratio in Figure 4-12 actually underestimates the methaneconsumption.0100200Depth (m)3004005006007008000.001 0.010 0.100 1.000 10.000 100.000Gas ratioCH 4 / HeC 2 H 6 / HeFigure 4-12. The methane/helium and ethane/helium ratios for Olkiluoto groundwatergas samples.

950100200Depth (m)3004005006007008001 10 100 1000 10000N 2 / HeFigure 4-13. The nitrogen/helium ratios for Olkiluoto groundwater gas samples.Some methane/helium ratios from a depth of approximately 100 m are larger than somefrom depths of 200–300 m. This could be due to anaerobic methane production bymicroorganisms or may simply reflect the very heterogeneous character of the fracturedrock mass sampled, in which several different types of groundwater mix at the samedepth.4.5 Specialists, generalists, opportunists, and antagonists in the worldof microbesThe evolution of the microbial world has been continuous for almost four billion years.Over this time, microbes have evolved and adapted to all the environments on ourplanet where life is possible. To exist, life needs energy, water, and a temperature rangebetween 20C and +113C. The phylogenetic tree depicted in Figure 1-1 reflects theenormous diversity of microbes. Over this evolutionary process, several importantstrategies have been developed. Some microorganisms have become specialized for lifein a very narrow range of conditions. One extreme example is the heat-loving microbesthat live in hot springs at temperatures close to 100C. Many actually “freeze” to deathwhen the temperature drops towards room temperature. Other microorganisms haveevolved to be very general in their required environmental conditions. They survive insoil and water and can tolerate relatively large ranges of pH, salinity, and temperature.The drawback of specialization is that specialists are restricted to a very narrow niche,but the advantage is little competition, as most other microorganisms will die if theyenter the specialist niche. The generalist, on the other hand, will encounter hardcompetition with many other microorganisms, which may require fine tuning of their

96characteristics, such as the ability to grow rapidly when conditions become favourable.If you can grow faster than other microbes, you have an advantage, of course. Manymicroorganisms are thus opportunistic: they wait for favourable conditions when theycan prosper and multiply. While waiting, they may enter various dormancy states, suchas spore formation or starvation, in which they can remain for many years. Finally,some microorganisms have developed ways to compete with other organisms byproducing substances that kill their antagonists. Some bacteria can initiate chemicalwarfare; the actinomycetes are specialists in this, as they produce many differentantibiotics, including streptomycin. Then we have the viruses, which do not strictlyspeaking constitute life, but can be very powerful microbe-killing agents. Takentogether, these different strategies create ecosystems of microbes that are in ecologicalbalance over the long term, although some species may occasionally take over anddominate when given a chance.The implications of the above microbial strategies for Olkiluoto and any otherunderground environment should be obvious. In a completely stagnant groundwatersystem, transport rates are limited to diffusion, which is a very slow process over longdistances (i.e., in the meter range or more) but very quick in the micrometer range.Microbial processes in such systems will be very slow. Opportunistic microbes instagnant systems can wait for many years for conditions to change. If hydrodynamicconditions change to a flow situation in which different groundwaters mix, it is verylikely that opportunists will respond with rapid growth and microbial processes willspeed up significantly for as long as the new flow conditions last. The same thing willhappen during the construction of a deep underground tunnel such as ONKALO, orwhen boreholes are drilled and pumped. Boreholes left without packers, and waterconducting fractures intersected by the tunnel, will “short circuit” various fractures andmobilize substances that microbes need.When opportunists are given good growth conditions, the ATP concentration willincrease together with the TNC and numbers of cultivable microbes. Highconcentrations of organic material should trigger growth of opportunists, as discussedabove. The relationship between ATP and DOC (Figure 4-8) suggests that this isoccurring in Olkiluoto groundwater. By comparing the biomass and DOC data fromdifferent boreholes and depths in Olkiluoto, it is possible to identify “hot spots” wheremicrobial processes are occurring at a rate exceeding the average rates in Olkiluoto.4.6 Microbial processes in shallow groundwaterThe shallow groundwater of Olkiluoto is obviously in close contact with plant andanimal life on the surface. There is an input of rainwater to the ground that willtransport dissolved organic material from degradation processes in the surface soils intoshallow groundwater. Oxygen from the air will dissolve in the recharging rainwater andfollow it into the ground. Life processes in the topsoil and deeper in the overburden willactively degrade particulate and dissolved organic material, which will reduce theoxygen. This is a continuous biological process with a clear seasonal variation: freezingconditions in winter will slow down the processes significantly, though the rechargewill also stop when the ground freezes. In spring, meltwater will intrude and oxygen

97transport into the ground will peak, as was observed in 2006 (Figure 3-6). The shallowgroundwater environment can consequently alternate between aerobic and anaerobicconditions, which most microorganisms are able to handle. Generalist microbes such asNRB and many IRB and MRB can switch from using oxygen as an electron acceptorwhen available, to using nitrate, ferrous iron, or manganese(IV) when needed. Suchorganisms are denoted facultative anaerobes. Other microbes can initiate fermentativeprocesses when oxygen disappears.4.6.1 Aerobic processesThe degradation of organic material with oxygen is a rapid process that is moreenergetically favourable than is degradation with other electron acceptors (cf. Figure1-8). Therefore, oxygen-reducing aerobic processes will dominate as long as oxygen isavailable. This is reflected by the oxygen profiles in shallow Olkiluoto groundwater.Except for a few boreholes, the oxygen concentration was 10% or less of saturation(Figure 3-3), which suggests that aerobic biological processes are consuming oxygen inshallow Olkiluoto groundwater.In addition to the organic material from the surface, methane migrating from biogenicand thermogenic methanogenesis in deeper layers plus methane produced in overburdenlayers such as wetlands can contribute to oxygen reduction by microorganisms. Theshallow groundwater investigations have documented the presence of MOB that oxidizemethane with oxygen (Figure 3-35). Active MOB populations are expected to reducethe oxygen concentration. Inspecting the relationship between the MPN of MOB andthe concentrations of dissolved oxygen in shallow groundwater revealed a clearrelationship (Figure 4-14). There was more MOB in groundwater with lowconcentrations of oxygen than in groundwater with high oxygen concentrations. Sixsamples contained no detectable oxygen and had a range of different MOB numbers.Comparing MOB with the amount of dissolved methane also revealed a relationship inthe case of most samples (Figure 4-15). If the three outliers in parentheses in the figureare excluded, the numbers of MOB would appear to be higher in samples low inmethane and vice versa. The interpretation of these observations is complex, becausethey represent snapshots of ongoing processes. It is clear, however, that low methaneand oxygen concentrations coincided with high numbers of MOB in several samples.This suggests that aerobic methane oxidation is an important process in removingoxygen from intruding oxygenated recharge water. Some water samples were anaerobic,containing various numbers of MOB. It must be noted that MOB are obligate aerobes,unlike ANME consortia, which are strictly anaerobic in their nature (Boetius et al.2000). It can be hypothesized that these groundwaters may have contained oxygenbefore the sampling occasion, and that the oxygen was reduced before sampling. Whatwas observed were remaining MOB populations in various stages of adjustment tooxygen-free conditions. These opportunistic MOB were possibly in various states ofreducing their populations into dormancy until the next time oxygen appeared. Thishypothesis can be tested if time series are performed over seasons, as was done withoxygen in 2006 (Figure 3-6).

984.6.2 Anaerobic processesThe MPN analyses demonstrated the presence of most of the anaerobic microorganismstested for, but the variability in numbers and diversity was large. Samples fromboreholes PP2 and PP9 and two samples from PVP14 contained very low numbers, asdepicted in Figure 3-25. In contrast, samples from borehole PP39 and two samples fromPR1 had among the highest stacked MPN values observed. The reasons for thesedifferences are difficult to define in detail. It is obvious, however, that theenvironmental conditions in different boreholes were reflected by the presence andprobably also the activity of various microorganisms. Borehole PVP1 had a very highstacked value in spring 2006, the reason being a flooding event that lifted the DOCvalue ten times or more (196 mg L 1 ) above the values in most other boreholes analysed(Table A-4). Consequently, an extreme event preceding sampling of this borehole inspring 2006 was clearly reflected in the microbiological analyses. This is good exampleof an opportunistic outburst of microbial activity and multiplication. Borehole PP39groundwater had the second highest concentration of DOC found in the shallowgroundwater, which may explain why it also had among the highest stacked MPNvalues. As discussed before in relation to Figure 4-8, a clear positive correlation existedbetween ATP and DOC concentrations. High concentrations of ATP reflect many andactive microorganisms (Eydal and Pedersen 2007). A high DOC input to shallowgroundwater will rapidly increase the reduction rate of oxygen, and the microbialecosystem will switch to anaerobic processes as soon as oxygen has disappeared.4.7 Microbial processes in deep groundwater4.7.1 Aerobic processesOxygen will not reach very deep in groundwater due to the reducing activity of shallowgroundwater microorganisms discussed above. However, if oxygen should penetratedue to some extreme event, the processes described for shallow groundwater willoperate in deeper groundwater, and the intruding oxygen will soon be reduced to waterby groundwater organisms.

9910,001,00O 2 (mg L 1 )0,100,010,000 1 10 100 1000 10000MOB (cells mL 1 )Figure 4-14. The relationship between the numbers of methane-oxidizing bacteria(MOB) and the concentration of dissolved oxygen analysed with the Winkler titrationmethod.1000CH 4 (µL L 1 groundwater 1 )10010( )( )1( )0.1 1.0 10.0 100.0 1000.0MOB (cells mL 1 )Figure 4-15. The relationship between the numbers of methane-oxidizing bacteria(MOB) and the concentration of dissolved methane.

1004.7.2 Anaerobic processesThe MPN analysis results indicated the presence of NRB, AA, and HA in all samplesanalysed for these metabolic groups (Table A-11). IRB, MRB, and SRB were foundpredominately in samples from the first 100 m and at the 300-m level (Figure 3-28,Figure 3-29, Figure 3-30). Methanogens were found sparsely distributed throughout thedepth range (Figure 3-33, Figure 3-34). The production of sulphide by SRB in Olkiluotogroundwater is important, because sulphide would have the potential to corrode thecopper canisters used to store spent nuclear fuel. The production of acetate by AA isalso important, because acetate can be utilized by SRB, thereby contributing to theamount of produced sulphide. The safety analysis of any future repository requiresdetailed information regarding how much sulphide can be formed under variouscircumstances in the deep aquifers surrounding a repository and in the near field of sucha repository. The microbial and inorganic processes involved in sulphur transformationscan be summarized in the following conceptual model of the coupled reactions that leadto sulphide production.Microbial processesAA: H 2 + CO 2 acetate (Eq. 4-1)IRB: acetate + Fe 3+ Fe 2+ + CO 2 (Eq. 4-2)SRB: acetate + SO 4 2– (+ H 2 ) H 2 S + CO 2 (Eq. 4-3)SRB: DOC + SO 4 2– H 2 S + CO 2 (Eq. 4-4)AM + SRB: CH 4 + SO 4 2– H 2 + CO 2 + SO 4 2– H 2 S + CH 2 O (Eq. 4-5)Inorganic processes (pH > 6.5)H 2 S + 2FeOOH S 0 + 2Fe 2+ + 4OH (Eq. 4-6)H 2 S + Fe 2+ FeS + 2H + (Eq. 4-7)3FeS + 3S 0 Fe 3 S 4 + 2S 0 3FeS 2 (Eq. 4-8)In a hypothetical aquifer in Olkiluoto rock, the model suggests that AA can produceacetate from hydrogen and carbon dioxide at a rate determined by the inflow ofhydrogen (Eq. 4-1). The acetate produced can be utilized by IRB as a source of carbonand energy; as a result, ferrous iron and carbon dioxide are formed from ferric ironminerals and acetate, respectively (Eq. 4-2). Sulphate-reducing bacteria oxidize theacetate produced by AA to carbon dioxide, while sulphate is reduced to sulphide (Eq. 4-3). Several genera of SRB can oxidize acetate, but Desulfovibrio species need hydrogento be able to utilize acetate. If degradable organic carbon (i.e., DOC and TOC) isavailable, SRB will produce sulphide and carbon dioxide from this energy and carbonsource (Eq. 4-4). A special type of sulphate reduction is coupled to anaerobic methaneoxidation (Eq. 4-5). This reaction is common in many marine sedimentary environments(Boetius et al. 2000), but has not yet been demonstrated in deep groundwater. If present,it would have a significant impact on any sulphide production model, because the

101analysed concentration of methane in deep Olkiluoto groundwater is generally muchhigher than the analysed hydrogen concentration. This possibility has been discussed byPitkänen and Partamies (2007). The above microbial reactions result in the productionof sulphide, ferrous iron, acetate, and carbon dioxide. Hydrogen sulphide produced viaequations 4-3 to 4-5 may reduce iron in minerals such as goethite, resulting in theformation of elemental sulphur and ferrous iron (Eq. 4-6). Together with hydrogensulphide, the ferrous iron produced via equations 4-2 and 4-6 can form iron sulphide(Eq. 4-7). This is a solid compound, and the dissolved sulphide that reacts with ferrousiron in equation 4-7 will precipitate from the groundwater. Finally, pyrite may form (Eq.4-8) when oversaturation occurs. Pyrite formation has been found to occur rapidly insurface sediments following seasonal variations in sulphide concentrations (Howarth1979). It was concluded that the rate of sulphate reduction may be grosslyunderestimated if pyrite formation is ignored. Equations 4-1 to 4-8 may explain theobservations reported here. However, the bedrock at Olkiluoto provides a very reducingenvironment, and iron oxyhydroxides have only been observed at very shallow depths,despite the scattered observations of dissolved ferrous iron at depth (Figure 4-10) whichsuggest the ongoing reduction of iron oxyhydroxides. The assumed limited availabilityof iron oxyhydroxides at depth may explain why very high sulphide concentrations areobserved at a depth of 300 m (Andersson et al. 2007b). The rate of sulphate reductionwill then significantly over-ride the rate of ferric iron reduction.It was previously concluded in this report that merely knowing the concentrations ofchemical markers provides insufficient information with which to judge whether or nota microbial process is taking place. As exemplified by Figure 1-11, the rationale behindand regulation of rates of microbial processes are complicated. Above all, if thereactions are occurring at similar rates, the result will be steady-state concentrations ofdissolved ferrous iron and sulphide within a fairly narrow range of values. Inspectingthe concentration profiles of ferrous iron (Figure 4-10) and sulphide (Figure 4-11)reveals a clear peak in the sulphide concentration at approximately 300 m, but it isimpossible to resolve any peak of iron. If the model above is correct, ferrous ironconcentrations should approach zero, according to Eq. 4-7, when sulphideconcentrations increase. If the ferrous iron/sulphide ratio is plotted against sulphide, aclear relationship is evident (Figure 4-16). The figure indicates that the ferrous ironconcentration decreases relative to the increase in sulphide concentration. The ironconcentration decreases almost ten times faster than the sulphide concentrationincreases. This must be due to the effect of equation 4-7 and because the rate ofsulphide production (Eq. 4-3 to 4-5) is faster than that of ferrous iron (Eq. 4-2 and 4-6).Small ferrous iron/sulphide ratios thus indicate samples and sites in Olkiluoto where themicrobial production rate of sulphide is much faster than that of ferrous iron. Ferrousiron can be produced both by microbial processes and via equation 4-6. Plotting theferrous iron/sulphide ratio versus depth enables the identification of sampling pointswith a high sulphide concentration relative to iron. These points are all located at the300-m level and have 100 times or more dissolved sulphide than ferrous iron. Thesepoints in Olkiluoto can be concluded to harbour very active microbial populations. Thenext question then concerns the microbial processes going on at these points.Before answering this question, we must first consider anaerobic methane-oxidizingmicroorganisms (ANME). For a long time, scientists have observed profiles of methane,

102sulphate, sulphide, and carbon dioxide in anaerobic aquatic sediments that stronglysuggested the presence of active ANME (Zehnder and Brock 1980; Thomsen et al.2001). It was not until very recently, however, that the microorganisms behind thisprocess were identified (Boetius et al. 2000). It was demonstrated that two organismsco-operate in oxidizing methane: methanogens first oxidize methane to hydrogen andcarbon dioxide (Eq. 4-9), after which sulphate reducers sweep up the hydrogen andcarbon dioxide and produce hydrogen sulphide (Eq. 4-10). Both types of organisms gainreducing power from the reactions used to synthesize organic molecules, with carbondioxide as the carbon source. To do this, the two organisms must be in very closeproximity; typically, the methane oxidizers are surrounded by sulphate reducers in smallaggregates.Methane oxidizer: CH 4 + 3H 2 O 4H 2 + HCO 3 – + H + (Eq. 4-9)Sulphate reducer: 4H 2 + SO 4 2– +H + HS + 4 H 2 O (Eq. 4-10)Sum reaction: CH 4 + SO 4 2– HS – + HCO 3 – + H 2 O (Eq. 4-11)From Figure 3-16 and Figure 4-9 it is obvious that strong methane and sulphategradients meet in several locations at a depth of 300 m in Olkiluoto. Furthermore, it isobvious from determinations of ATP levels and of the MPNs of various physiologicalgroups of bacteria, that microbial abundance and activity both peak at these samplelocations. Finally, sulphide concentrations are also very high at the same locations. Ofthe sites evaluated and discussed here, KR6-328 m, KR10-316 m, and KR13-294 mhave the greatest potential for pronounced anaerobic methane oxidation; these threelocations have ferrous iron/sulphide ratios of 0.1, 0.01, and 0.05, respectively (TableA-4). They also have high concentrations of ATP and DOC and high MPNs of NRB,SRB, AA, and HA, relative to those of other deep groundwater samples. The last pieceof evidence needed is proof of the presence of ANME in groundwater at these locations.Ongoing investigations are focusing on this task using DNA technology and availablegenetic information (Thomsen et al. 2001).That acetogens may be active is suggested by the presence of hydrogen in groundwatersamples from Olkiuoto (Figure 3-12). Samples from deep layers have highconcentrations of this gas (Andersson et al. 2007b), but most of the samples reported onhere were from depths that were too shallow to harbour these high hydrogenconcentrations (Figure 3-18). In addition to a deep source of hydrogen, it is possible thatthe ANME process may leak hydrogen to acetogenesis, if AA are located close enoughto the ANME aggregates. This possibility is still speculative, and successful isolation ofANME in pure laboratory cultures will be needed in order to conduct detailed studies.The MPNs of sulphate reducers and methanogens in Olkiluoto groundwater weregenerally at or below the detection limits for each type of microorganism, unlike whatwas observed previously (Table 1-1). The protocols used to cultivate these organisms(Table 2-3) have worked very well at other Fennoscandian sites, such as the SwedishForsmark and Laxemar investigation sites and the Äspö HRL (the MPN of SRB hasreached 10,000 cells mL 1 in some groundwater samples from these sites). The sameprotocols should work well for samples from Olkiluoto as well. However, it could bethat Olkiluoto groundwater is dominated by ANME, and that our cultivation protocols

103did not detect them. So far, pure cultures of ANME have not been described in theliterature. ANME may be so strongly linked and interdependent that they cannot becultivated separately in the SRB and AM media used. Alternately, the drilling andpumping out of many new boreholes in Olkiluoto may have disturbed the microbialpopulations and reduced their numbers. Such an effect was observed at the Äspö HRL,in the case of SRB in particular (Pedersen 2005b). That would explain why the numbersof SRB were significantly lower in the 2004–2006 samples than in the 1996–2000samples (Table 1-1). Finally, changes in the PAVE system sampling methodology mayhave introduced this difference. Previously, the pressure vessels were not pumped outafter being opened. Later, in 2004–2006, the microbiological sampling vessels began tobe pumped out for some hours after being opened. However, it is not obvious how thisdifference in sampling methodology could have influenced microbial numbers; testswill be required to explore this point.10000.001000.00100.00Fe 2+ /S 210.001.000.100.010.000.00 0.01 0.10 1.00 10.00Sulphide (mg L 1 )Figure 4-16. The relationship between the ferrous iron/sulphide ratio and sulphide.

1040100200300Depth (m)4005006007008009000.00 0.01 0.10 1.00 10.00 100.00 1000.00 10000.00Fe 2+ /S 2Figure 4-17. The relationship between the ferrous iron/sulphide ratio and depth.4.8 Relevance of microbiological processes to ONKALOThe introduction to this report identified three main effects of microorganisms in thecontext of a KBS-3 type repository for radioactive waste in Olkiluoto bedrock. Theresearch, results, and conclusions presented here constitute an important baseline forunderstanding how microbiological processes may interact with ONKALO and a futureHLW repository. The evaluated dataset from 2004 to 2006 comprised 60 sets ofmicrobiological analyses coupled to analyses of physical and chemical parameters andthe amounts of dissolved gases over the 4–450 m depth range. Continuousmicrobiological research can now focus on processes deemed significant on the basis ofthis report, as outlined next. The relevance of microbiological processes to ONKALOcan be evaluated as follows.4.8.1 Oxygen reduction and maintenance of anoxic and reduced conditionsShallow groundwater in Olkiluoto contained dissolved oxygen at approximately 10% orless of saturation. The presence of aerobic and anaerobic microorganisms, includingmethane-oxidizing bacteria, has been documented. The data suggest that microbialprocesses reduce intruding oxygen in the shallow groundwater, DOC and methane beingthe main electron donors. Biological processes are temperature dependent and seasonalvariation was expected and could be documented. Construction of ONKALO may causethe opening of discrete fractures leading towards the tunnel wall, resulting in increased

105inflow to the tunnel [It is possible that oxygen could reach deeper groundwater, carriedalong by intrusive shallow groundwater that penetrates to greater depths via suchdisturbed fractures.] However, it can be hypothesized that opportunistic microbialprocesses could mitigate this oxygen effect if proper electron donors are available. Thecontinuation of the shallow groundwater research programme is recommended. Newdata can be compared with the data presented here and with data produced by thehydrogeochemical monitoring programme in Olkiluoto. Significant drawdown effects, ifany, caused by ONKALO construction should be detectable with this programme. Inaddition, sampling ONKALO boreholes in the upper part of the tunnel in time serieswill increase our understanding of how microbial processes in shallow fractures react totunnel construction.Fractures opened by the construction of ONKALO will allow for increased water flow.This will allow opportunistic microbes to become activated, resulting in differentmicrobial processes that will influence the geochemistry. After repository closure, whenthe groundwater table has been restored and stabilized, the rates of microbial processeswill again be reduced. The prediction of long-term evolution of hydrogeochemicalconditions in the vicinity of ONKALO and the repository requires data pertaining to thesurrounding groundwater. It is important to understand how these conditions have beeninfluenced by the construction of ONKALO. What new conditions will persist for along time and what conditions will return to their original pre-construction states?Detailed knowledge of microbial processes is needed for such modelling work, becausethese processes influence several important geochemical conditions. A very importantgeochemical parameter strongly influenced by microbial processes is the E h .4.8.2 Bio-corrosion of construction materialsMicrobiological and geochemical data strongly suggest that the anaerobic microbialoxidation of methane (ANME) is active at a depth of approximately 300 m in Olkiluoto.It appears as though ANME is limited to the 0–300 m depth interval due to a lack ofsulphate at depths below 300 m. This implies that the rate of sulphide production in theANME process at a depth of 300 m is limited by the transport rate of methane fromdeeper layers. The construction of ONKALO will probably influence the ANMEprocesses. If groundwater that contains sulphate is drawn down to the bottom of thecompleted ONKALO tunnel, the ANME processes at that level could speed up, in linewith what was discussed for opportunists (4.5). Sulphate seems to be the onlycomponent needed by ANME that is missing at depth. As groundwater drawdowncurrently seems to be boosting the sulphide concentration more than 100-fold at a depthdown to 300 m, it is very important to monitor this process because sulphide cancorrode copper. The fact that the conditions necessary for ANME growth will againbecome limited, extending no deeper than 300 m after tunnel closure and backfilling,will be beneficial for the long-term safety of the repository. This matter may call fordetailed modelling when data are available regarding how the ANME processes react tothe construction of ONKALO.A programme of research into ANME should be initiated. The study of these still poorlyunderstood microorganisms is in the forefront of microbiological research. New tools in

106molecular biology, such as DNA technology, are needed for such research. ANMEsamples can be collected on site in ONKALO, using the PAVE system and monitoringboreholes. The polymerase chain reaction (PCR) method of DNA analysis can be usedto detect DNA sequences specific to ANME microbes, and thus map the distributionand diversity of ANME in Olkiluoto groundwater over the construction period. AnyANME microbes present can be quantified using the real-time polymerase chainreaction (RT-PCR) method. Analysing m-RNA, by applying RT-PCR to copy DNA,can be used to detect possible ongoing ANME activity. In addition, cultivation methodswill be developed and improved. If ANME can be brought into the laboratory, muchnew knowledge of ANME processes can be gained and applied to the evolution ofONKALO groundwater.Marked sulphide production appears to be ongoing at the 300-m level in Olkiluoto.When sulphide comes in contact with air, sulphuric acid may form, which is corrosivefor metals and concrete. The extent and limiting factors of this process in ONKALO arecurrently unknown but, as the consequences include the deterioration of “shotcrete” andthe concrete sealing of fractures, they should be explored.As a safety measure for employees and construction workers, extra caution should betaken if the ONKALO tunnel should pass through sulphide-producing rockenvironments that may have high sulphide concentrations, since hydrogen sulphide gasis very toxic and lethal to humans.4.8.3 Bio-mobilization and bio-immobilization of radionuclides, and the effectsof microbial metabolism on radionuclide mobility.It is well known that microbes can mobilize trace elements (Pedersen 2002). First,unattached microbes, including viruses, may act as large colloids, transportingradionuclides on their surfaces with the groundwater flow (Moll et al. 2004). Second,microbes are known to produce ligands that can mobilize trace elements and that caninhibit trace element sorption to solid phases (Kalinowski et al. 2004, 2006). One groupof microorganisms produces very powerful bioligands, usually denoted pyoverdins,which have a very strong binding affinity for many radionuclides (Johnsson et al. 2006;Essén et al. 2007; Moll et al. 2007). Pyoverdin-producing microbes have been found inshallow Olkiluoto groundwater and in the slime that grows on the tunnel walls ofONKALO. It is important to investigate whether the microbial production of bioligandsin deep groundwater may exceed the safety limit for a repository. Groundwater samplesfrom ONKALO can be analysed for DNA signatures typical of pyoverdin producerssuch as Pseudomonadaceae and Shewanella. The direct interaction betweenradionuclides and any pyoverdins that may be present in deep groundwater should alsobe investigated. Biofilms in aquifers may also influence the retention processes ofradionuclides in groundwater (Anderson et al. 2006).

107REFERENCESAnonymous. 1983, Final storage of spent nuclear fuel, KBS-3 Final Report, Vol. I-IV, I–IV 1–100. Swedish Nuclear Fuel and Waste Management Co., StockholmAckermann, H.W. 2007, 5500 Phages examined in the electron microscope, Archive ofVirology, 152 227–243.Amend, J.P. and Teske, A. 2005, Expanding frontiers in deep subsurface microbiology,Palaeogeography, Palaeoclimatology, Palaeoecology, 219 131–155.Anderson, C., Pedersen, K. and Jakobsson, A.-M. 2006, Autoradiographic comparisonsof radionuclide adsorption between subsurface anaerobic biofilms and granitic hostrocks, Geomicrobiol J, 23 no. 1, 15–29.Anderson, C.R. and Pedersen, K. 2003, In situ growth of Gallionella biofilms andpartitionating of lanthanids and actinides between biological material and ferricoxyhydroxides, Geobiology, 1 no. 2, 169–178.Andersson, J., Ahokas, H., Hudson, J.A., Koskinen, L., Luukkonen, A., Löfman, J.,Keto, V., Pitkänen, P., Mattila, J., Ikonen, A.T.K. and Ylä-Mella, M. 2007a, 3. Surfaceconditions, Olkiluoto site description 2006, 41–51. Report Posiva 2007-3Andersson, J., Ahokas, H., Hudson, J.A., Koskinen, L., Luukkonen, A., Löfman, J.,Keto, V., Pitkänen, P., Mattila, J., Ikonen, A.T.K. and Ylä-Mella, M. 2007b, 7.Hydrogeochemistry, Olkiluoto site description 2006, 225–290.Report Posiva 2007-3,OlkiluotoApps, J.A. and van de Kamp, P.C. 1993, Energy gases of abiogenic origin in the Earth'scrust, in: Howell, G. (ed) The future of Energy gases, U.S. geological SurveyProfessional Papers, 1570 81–132. United States Government Printing Office,WashingtonAylward, G.H. and Findley, T.J.V. 2002, SI Chemical Data 5th edition, John Wiley &Sons Australia, Ltd, Milton, AustraliaBanwart, S., Gustafsson, E., Laaksoharju, M., Nilsson, A.-C., Tullborg, E.-L. andWallin, B. 1994, Large-scale intrusion of shallow water into a granite aquifer, WaterResour Res, 30 1747–1763.Banwart, S., Tullborg, E.-L., Pedersen, K., Gustafsson, E., Laaksoharju, M., Nilsson,A.-C., Wallin, B. and Wikberg, P. 1996, Organic carbon oxidation induced bylargescale shallow water intrusion into a vertical fracture zone at the Äspö Hard RockLaboratory (Sweden), Journal of Contaminant Hydrology, 21 no. 1–4, 115–125.Boetius, A., Ravenschlaug, K., Schubert, C.J., RIckert, D., Widdel, F., Gleseke, A.,Amann, R., Jörgensen, B.B., Witte, U. and Pfannkuche, O. 2000, A marine microbialconsortium apparently mediating anaerobic oxidation of methane, Nature, 407 623–626.

108Carritt, D.E. and Carpenter, J.H. 1966, Comparison and evaluation of currentlyemployed modifications of the Winkler Method for determining dissolved oxygen in seawater; a NASCO Report, Journal of Marine Research, 24 286–318.Corinaldesi, C., Crevatin, E., Del Negro, P., Marini, M., Russo, A., Fonda-Umani, S.and Danovaro, R. 2003, Large-scale spatial distribution of virioplankton in the AdriaticSea: Testing the trophic state control hypothesis, Appl Environ Microbiol, 69 2664–2673.Danovaro, R., Manini, E. and Dell'Anno, A. 2002, Higher abundance of bacteria than ofviruses in deep Mediterranean sediments, Appl Environ Microbiol, 68 1468–1472.DiChristina, T.J. and DeLong, E.F. 1993, Design and application of rRNA-targetedoligonucleotide probes for the dissimilatory iron- and manganese-reducing bacteriumShewanella putrefaciens, Applied and Environmental Microbiology, 59 4152–4160.Drake, H.L., Küsel, K. and Matthies, C. 2002, Ecological consequences of thephylogenetic and phisological diversities of acetogens, Antonie van Leeuwenhoek, 81no. 1203–212.Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H. and Stackebrandt, E. (eds)2007, The Prokaryotes, Volumes 1–7, 3rd edition, 1 SpringerEkendahl, S., O'Neill, A.H., Thomsson, E. and Pedersen, K. 2003, Characterisation ofyeasts isolated from deep igneous rock aquifers of the Fennoscandian Shield, MicrobEcol, 46 416–428.Ekendahl, S. and Pedersen, K. 1994, Carbon transformations by attached bacterialpopulations in granitic ground water from deep crystalline bed-rock of the Striparesearch mine, Microbiology, 140 1565–1573.Essén, S.A., Johnsson, A., Bylund, D., Pedersen, K. and Lundström, U.S. 2007,Siderophore production by Pseudomonas stutzeri under aerobic and anaerobicconditions, Appl Environ Microbiol, 73 no. 18, 5857–5864.Eydal, H.S.C. and Pedersen, P. 2007, Use of an ATP assay to determine viablemicrobial biomass in Fennoscandian Shield groundwater from depths of 3–1000 m, JMicrobiol Meth, 70 363–373.Gascoyne, M. 2000, Dissolved gases in groundwaters at Olkiluoto, Posiva WorkingReport 2000-49.Gascoyne, M. 2005, Dissolved gases in groundwaters at Olkiluoto, Posiva WorkingReport 2005-56.Greenberg, A.E., Clesceri, L.S. and Eaton, A.D. 1992, Estimation of Bacterial Density,Standard methods for the examination of water and wastewater 18th ed, 9–49.American Public Health Association, Washington

109Hallbeck, L.E. and Pedersen, K. 2005, Genus I. Gallionella Ehrenberg 1838 166AL, in:Brenner, D.J., Krieg, N.R. and Staley, J.T. (eds) Bergey's Manual of SystematicBacteriology, Part C: The Proteobacteria; The Alpha-, Beta-, Delta-, andEpsilonproteobacteria, 2 880–886. Springer, United States of AmericaHallbeck, L.E. and Pedersen, K. 2008, Characterization of microbial processes in deepaquifers of the Fennoscandian Shield, Applied Geochemistry (accepted for publication),Haveman, S.A., Nilsson, E.L. and Pedersen, K. 2000, Regional distribution of microbesin groundwater from Hästholmen, Kievetty, Olkiluoto and Romuvaara, Finland,POSIVA report 2000-06.Haveman, S.A. and Pedersen, K. 2002a, Distribution of culturable anaerobicmicroorganisms in Fennoscandian shield groundwater, FEMS Microbiol Ecol, 39 129–137.Haveman, S.A. and Pedersen, K. 2002b, Microbially mediated redox processes innatural analogues for radioactive waste, J Contam Hydrol, 55 161–174.Haveman, S.A., Pedersen, K. and Ruotsalainen, P. 1998, Geomicrobial investigations ofgroundwater from Olkiluouto, Hästholmen, Kivetty and Romuvaara, Finland,Geomicrobial investigations of groundwater from Olkiluouto, Hästholmen, Kivetty andRomuvaara, Finland, POSIVA 98-09 1-40. POSIVA OY, HelsinkiHaveman, S.H., Pedersen, K. and Routsalainen, P. 1999, Distribution and metabolicdiversity of microorganisms in deep igneous rock aquifers of Finland, GeomicrobiologyJournal, 16 277–294.Hobbie, J.E., Daley, R.J. and Jasper, S. 1977, Use of nucleopore filters for countingbacteria by fluorescence microscopy, Appl Environ Microbiol, 33 1225–1228.Howarth, R.W. 1979, Pyrite: Its rapid formation in a salt marsh and its importance inecosystem metabolism, Science, 203 49–51.Johnsson, A., Arlinger, J., Pedersen, K., Ödegaard-Jensen, A. and Albinsson, Y. 2006,Solid-Aqueous Phase Partitioning of Radionuclides by Complexing CompoundsExcreted by Subsurface Bacteria, Geomicrobiol J, 23 621–630.Kalinowski, B.E., Johnsson, A., Arlinger, J., Pedersen, K., Ödegaard-Jensen, A. andEdberg, F. 2006, Microbial mobilization of uranium from shale mine waste,Geomicrobiol J, 23 157–164.Kalinowski, B.E., Oskarsson, A., Albinsson, Y., Arlinger, J., Ödegaard-Jensen, A.,Andlid, T. and Pedersen, K. 2004, Microbial leaching of uranium and other traceelements from shale mine tailings at Ranstad, Geoderma, 122 177–194.Kerr, R.A. 2002, Deep life in the slow, slow lane, Science, 296 1056–1058.Kotelnikova, S., Macario, A.J.L. and Pedersen, K. 1998, Methanobacteriumsubterraneum, a new species pf Archaea isolated from deep groundwater at the Äspö

111Ortmann, A.C. and Suttle, C.A. 2005, High abundances of viruses in a deep-seahydrothermal vent system indicates viral mediated microbial mortality, Deep-SeaResearch, 52 1515–1527.Pace, N.R. 1997, A molecular view of microbial diversity and the biosphere, Science,276 734–740.Pedersen, K. 1993, The deep subterranean biosphere, Earth-Science Reviews, 34 243–260.Pedersen, K. 2000a, Exploration of deep intraterrestrial life – current perspectives,FEMS Microbiology Letters, 185 9–16.Pedersen, K. 2000b, The microbe site. Drilling, instrumentation and characterisation,SKB International Progress Report, IPR-00-36, IPR-00-36 1–25.Pedersen, K. 2001, Diversity and activity of microorganisms in deep igneous rockaquifers of the Fennoscandian Shield, in: Fredrickson, J.K. and Fletcher, M. (eds)Subsurface microbiology and biogeochemistry, 97–139. Wiley-Liss Inc., New YorkPedersen, K. 2002, Microbial processes in the disposal of high level radioactive waste500 m underground in Fennoscandian shield rocks, in: Keith-Roach, M.J. and Livens,F.R. (eds) Interactions of microorganisms with radionuclides, 279–311. Elsevier,AmsterdamPedersen, K. 2004, Strolling through the world of microbes, Microbiome, HindåsPedersen, K. 2005a, The MICROBE framework: Site descriptions, instrumentation, andcharacterization, Äspö Hard Rock Laboratory. International Progress Report IPR-05-05, 1–85. Swedish Nuclear Fuel and Waste Management Co, StockholmPedersen, K. 2005b, MICROBE Analysis of microorganisms and gases in MICROBEgroundwater over time during MINICAN drainage of the MICROBE water conductingzone, Äspö Hard Rock Laboratory. International Progress Report IPR-05-29, 1–26.Swedish Nuclear Fuel and Waste Management Co, StockholmPedersen, K. 2006, Microbiology of transitional groundwater of the porous overburdenand underlying shallow fractured bedrock aquifers in Olkiluoto 2004, Finland, POSIVAWorking Report 2006-09, 1–40.Pedersen, K. 2007, Microbiology of transitional groundwater of the porous overburdenand underlying shallow fractured bedrock aquifers in Olkiluoto 2005, Finland. October2005–January 2006, POSIVA Working Report 2007-20, 1–46.Pedersen, K. and Ekendahl, S. 1990, Distribution and activity of bacteria in deepgranitic groundwaters of southeastern Sweden, Microbial Ecology, 20 37–52.Pedersen, K. and Ekendahl, S. 1992a, Incorporation of CO 2 and introduced organiccompounds by bacterial populations in groundwater from the deep crystalline bedrockof the Stripa mine, Journal of General Microbiology, 138 369–376.

112Pedersen, K. and Ekendahl, S. 1992b, Assimilation of CO 2 and introduced organiccompounds by bacterial communities in ground water from Southeastern Sweden deepcrystalline bedrock, Microbial Ecology, 23 1–14.Pedersen, K. and Karlsson, F. 1995, Investigations of subterranean microorganisms –Their importance for performance assessment of radioactive waste disposal, SKB TR95-10 1–222. Swedish Nuclear Fuel and Waste Management Co., StockholmPedersen, K., Nilsson, E., Arlinger, J., Hallbeck, L. and O'Neill, A. 2004, Distribution,diversity and activity of microorganisms in the hyper-alkaline spring waters of Maqarinin Jordan, Extremophiles, 8 151–164.Pitkänen, P., Ahokas, H., Ylä-Mella, M., Partamies, S., Snellman, M. and Hellä, P.2007, Quality review of hydrochemical baseline data from the Olkiluoto site, POSIVA2007-05, Posiva Oy. 133 p., OlkiluotoPitkänen, P., Luukkonen, A. and Partamies, S. 2003, Hydrochemical interpretation ofbaseline groundwater conditions at the Olkiluoto site, POSIVA 2003-07.Pitkänen, P. and Partamies, S. 2007, Origin and implications of dissolved gases ingroundwater at Olkiluoto, POSIVA 2007-04.Posiva Oy 2003, Nuclear waste management of the Olkiluoto and Loviisa power plants.Programme for research, development and technical design for 2004–2006, TKS-2003.Posiva Oy 2006, Nuclear waste management of the Olkiluoto and Loviisa power plants.Programme for research, development and tehcnical design for 2004–2006, TKS-2006.Rachel, R., Bettstetter, M., Hedlund, B.P., Häring, M., Kessler, A., Stetter, K.O. andPrangishvili, D. 2002, Remarkable morphological diversity of viruses and virus-likeparticles in hot terrestrial environments, Arch Virol, 147 2419–2429.Reitner, J., Schumann, G.A. and Pedersen, K. 2005, Fungi in subterraneanenvironments, in: Gadd, G.J. (ed) Fungi in biogenchemical cycles, 788–1002.Cambridge University Press, CambridgeSchulz, H.N., Brinkhoff, T., Ferdelman, T.G., Hernándes Mariné, M., Teske, A. andJorgensson, B.B. 1999, Dense population of a giant sulfur bacterium in Namibian shelfsediments, Science, 284 493–495.Staley, J.T. and Konopka, A. 1985, Measurements of in situ activities ofnonphotosynthetic microorganisms in aquatic and terrestrial habitats 39:321–346, AnnuRev Microbiol, 39 321–346.Stetter, K.O. 1996, Hyperthermophilic procaryotes, FEMS Microbiol Rev, 18 145–148.Suttle, C.A. 2005, Virus in the sea, Nature, 437 356–361.

113Thomsen, T.R., Finster, K. and Ramsing, N.B. 2001, Biogeochemical and molecularsignatures of anaerobic methane oxidation in a marine sediment, Appl EnvironMicrobiol, 67 no. 4, 1646–1656.Whittenbury, R., Philips, K.C. and Wilkinson, J.F. 1970, Enrichment, culturing,isolation and some properties of methane-utilizing bacteria, J Gen Microbiol, 61 205–218.Widdel, F. and Bak, F. 1992, Gram-negative, mesophilic sulphate-reducing bacteria, in:Balows, A., Truper, H.G., Dworkin, M., Harder, W. and Schleifer, K.-Z. (eds) Theprokaryotes, 4 3352–3378. Springer-Verlag, New YorkWiggins, B.A. and Alexander, M. 1985, Minimum bacterial density for bacteriophagereplication: implications for significance of bacteriophages in natural ecosystems, ApplEnviron Microbiol, 49 19–23.Woese, C., Kandler, O. and Wheelis, M.L. 1990, Towards a natural system oforganisms: proposal for the domains Archaea, Bacteria and Eukarya, Proceedings of theNational Academy of Science, 87 4576–4579.Zehnder, A. and Brock, T. 1980, Anaerobic methane oxidation: Ocurrence and ecology,Appl Environ Microbiol, 39 no. 1, 194–204.

114

115A. APPENDIXTable A-1. Sampling data for four consecutive sampling campaigns in the shallowboreholes in 2004–2006.Borehole Measuredboreholedepth(m)Pump andsamplinglevel(m)Date and time(y-m-d/time)Groundwaterlevel(m)Yield(L/min)NotesPR1 6 5 2004-05-03/14:30 3.50 12.0 start2004-05-03/16:20 4.85 12.02004-05-03/19:00 4.98 12.0 stop2004-05-03/07:50 3.60 12.0 start2004-05-04/08:40 5.07 12.0 stopPR1 6.0 4.0 2005-10-10/10.35 3.47 4.82005-10-10/10.35 3.84 4.52005-10-10/10.35 3.85 4.5 clear water2005-10-10/10.35 3.78 4.5 samplingPR1 14.72 4.0 2006-04-25/06.30 2.84 5.5 pump start2006-04-25/06.35 3.10 5.52006-04-25/08.05 3.26 5.62006-04-25/09.25 3.32 - samplingPR1 14.72 4.0 2006-10-11/06.20 3.08 - pump start2006-10-11/06.35 3.29 3.82006-10-11/08.00 3.40 3.82006-10-11/09.25 3.45 3.8 samplingPP2 24.5 10 2004-05-04/16:00 2.10 6.0 start2004-05-04/17:00 5.67 5.0 stop2004-05-05/07:50 2.09 5.5 start2004-05-05/08:30 5.65 5.5 stopPP2 14.90 5.0 2005-10-12/08.00 2.05 4.8 pump start2005-10-12/08.40 3.93 4.82005-10-12/09.30 - - sampling2006-10-09/12.40 - - samplingPP2 14.80 6.0 2006-04-24/07.50 1.95 - pump start2006-04-24/08.00 3.20 3.62006-04-24/09.40 3.37 3.62006-04-24/10.45 3.40 3.6 samplingPP2 14.72 4.0 2006-10-11/10.15 2.42 - pump start2006-10-11/10.30 3.07 2.12006-10-11/12.00 3.19 2.22006-10-11/13.45 - 2.12006-10-11/14.50 - - sampling

116PP3 14.5 10 2004-05-05/09:50 0.55 7.2 start2004-05-05/11:00 7.0 stopPP7 16.5 10 2004-05-05/12:20 1.14 0.85 start2004-05-05/13:35 7.65 0.802004-05-05/15:35 stopPP8 7.0 6.0 2004-05-05/16:40 3.95 start2004-05-05/17:00 4.10 5.402004-05-05/17:40 4.12 stop2004-05-06/08:05 3.93 5.50 start2004-05-06/08:45 4.13 5.5 stopPP9 14.70 6.0 2005-10-13/12.25 1.31 0.432005-10-13/08.35 3.95 0.44 roily water2005-10-13/12.55 4.15 0.40 samplingPP9 14.70 6.0 2006-04-26/10.15 0.80 - pump start2006-04-26/10.25 1.39 0.612006-04-26/11.50 3.46 0.58 roily water2006-04-26/14.05 3.53 0.57 samplingPP9 14.70 6.0 2006-10-11/07.10 0.78 0.255 pump start2006-10-11/07.20 1.21 0.2552006-10-11/08.05 1.26 0.1802006-10-11/10.30 - 0.1852006-10-11/12.00 - 0.190 roily water2006-10-11/15.15 - - samplingPP36 12.05 6.0 2005-10-10/09.30 4.31 7.2 pump start2005-10-10/12.05 - -2005-10-10/13.05 - - clear water2005-10-10/13.40 - - samplingPP36 - 4.0 2006-04-25/06.45 3.47 5.04 pump start2006-04-25/06.50 3.64 5.042006-04-25/07.55 3.68 5.042006-04-25/12.35 3.70 4.70 samplingPP36 11.00 5.0 2006-10-09/08.40 3.27 - pump start2006-10-09/10.50 2.43 6.22006-10-09/11.55 - 6.02006-10-09/12.05 - 6.2 samplingPP39 14.10 6.0 2005-10-11/08.30 1.32 0.4 pump start2005-10-11/10.00 - - roily water2005-10-11/12.50 - - samplingPP39 14.04 5.0 2006-04-24/08.10 1.08 - pump start2006-04-24/08.20 2.01 0.672006-04-24/09.55 2.08 0.65

1172006-04-24/11.05 2.10 0.642006-04-24/14.10 2.14 - samplingPP39 14.10 6.0 2006-10-11/07.30 1.57 - pump start2006-10-11/07.45 3.20 0.92006-10-11/09.00 3.40 0.92006-10-11/11.45 - 0.92006-10-11/12.45 - - samplingPVP1 4.0 3.0 2004-05-03/16.35 1.05 start2004-05-03/15.45 1.05 2.02004-05-03/17.30 3.05 1.82004-05-03/18.45 2.85 0.25 stop2004-05-04/08.00 1.06 0.25 start2004-05-04/10.00 2.90 0.25 StopPVP1 3.90 2.50 2005-10-11/07.40 0.69 2.82005-10-11/09.50 1.19 2.4 brownwater2005-10-11/12.10 1.01 2.0 samplingPVP1 3.92 2.50 2006-04-27/06.35 0.43 - pump start2006-04-27/06.55 1.25 2.62006-04-27/07.45 1.19 2.6 brownwater2006-04-27/10.20 1.19 2.4 samplingPVP1 3.93 2.50 2006-10-12/06.20 0.70 -2006-10-12/06.50 1.32 3.4 pump start2006-10-12/08.00 1.80 3.02006-10-12/09.00 1.90 3.02006-10-12/09.25 - - samplingPVP3B 4.5 3 2004-05-03/12:20 0.60 4.0 start2004-05-03/13:55 2.60 3.0 run2004-05-03/15.30 2.65 2.5 stopPVP3A 8.5 7.5 2004-05-03/09:50 0.70 start2004-05-03/10:00 2.962004-05-03/10:35 6.90 2.72004-05-03/12:15 7.08 2.52004-05-03/13:30 7.00 StopPVP4A 10 9 2004-05-04/13:15 1.25 start2004-05-04/13:20 6.02004-05-04/14:45 7.30 6.0 stopPVP4A 10.20 5.0 2005-10-12/07.40 1.14 4.8 pump start2005-10-12/09.55 - 4.52005-10-12/12.10 - - sampling

118PVP4A 10.36 6.0 2006-04-27/06.20 0.85 - pump start2006-04-27/06.35 2.85 4.52006-04-27/08.00 3.27 4.42006-04-27/08.35 3.36 3.9 sampling 12006-04-27/14.45 3.16 - sampling 2PVP4A 10.20 5.0 2006-10-10/11.10 1.17 6.0 pump start2006-10-10/11.20 3.04 5.52006-10-10/12.50 - -2006-10-10/13.20 - - samplingPVP4B 9.5 8 2004-05-04/13.45 1.80 start2004-05-04/14:55 6.70 0.702004-05-04/17:10 0.70 stopPVP13 5.60 4.0 2005-10-12/10.40 1.51 1.06 roily water2005-10-12/10.40 1.87 0.80 clear water2005-10-12/10.40 1.99 0.85 samplingPVP13 5.64 3.0 2006-04-26/06.35 0.892006-04-26/06.45 1.95 1.10 pump start2006-04-26/08.00 1.91 1.102006-04-26/09.05 1.93 1.102006-04-26/11.00 1.64 0.76 samplingPVP13 5.60 4.0 2006-10-12/10.40 2.49 0.75 pump start2006-10-12/10.40 - 0.702006-10-12/10.40 2.98 0.702006-10-12/10.40 - - samplingPVP14 9.0 5.0 2005-10-14/07.45 2.29 5.22005-10-14/08.50 - 5.22005-10-14/09.35 - 5.2 samplingPVP14 9.08 6.0 2006-04-26/06.20 1.34 - pump start2006-04-26/06.30 3.18 5.12006-04-26/08.10 3.39 4.92006-04-26/08.45 - - samplingPVP14 9.08 5.5 2006-10-10/06.25 4.6 4.2 pump start2006-10-10/07.30 - 1.82006-10-10/08.30 - 1.82006-10-10/09.30 - 1.82006-10-10/10.15 - - samplingPVP20 12.80 5.0 2005-10-13/07.53 0.71 0.4 pump start2005-10-13/09.20 0.42005-10-13/09.30 samplingPVP20 - - 2006-04-26/00.00 - - frozenPVP20 12.80 5.0 2006-10-10/07.45 1.14 1.05 pump start

1192006-10-10/09.45 - 1.062006-10-10/12.30 - -2006-10-10/14.35 - -2006-10-10/15.20 - - sampling

120Table A-2. Pre-treatment of the groundwater samples.Parameter Container (L)N 2 - shielding/ FilteringPreserving chemicals and details LaboratoryConductivity,density, pH, NH 41 0.5 HDPE -/- - TVOAlkalinity,acidity1 0.5 Duranbottlex/x Sampling with titration sampler. TVOS 2–3 0.1 measuring0.5 mL Zn(Ac)2 and 0.5 mLx/xbottle0.1 M NaOH. Sampling with sampler.TVOCl, Br, SO 4 , S tot 1 0.25 HDPE -/x TVOF 1 0.25 HDPE -/x TVOFe 2+ , Fe totAdding a ferrozine reagent to Fe 2+6 0.05x/x samples in nitrogen atmosphere.measuring bottleSampling with sampler.TVOSodium1 0.05fluorescein measuring bottle-/xTVODIC/DOC1 0.05 brownglass bottle-/x Sampling with sampler. TVONa, Ca, K, Mg, 1 0.25 PE,Fe, Mn, SiO 2 acid washed-/x 2.5 mL conc. HNO 3 / 250 mL TVOPO 4 1 0.25 HDPE -/x 2.5 mL 4 M H 2 SO 4 / 250 mL TVOSr, B tot1 0.1 HDPE,acid washedN tot , NO 2 , NO 3 1 0.25 HDPE -/xH-2, O-18H-3C-13 / C-14Sr-87 / Sr-86Rn-222S-34 (SO 4 ),O-18 (SO 4 )Uranium, U-238Uranium,U-234/U-2381 0.125Nalgene bottle1 0.25 glassbottle1 0.2 brownclass bottle1 0.125Nalgene, acidwashed1 0.01Ultimagoldsolution bottle1 Nalgene1 1 PE,HCl-washed1 1 PE,HCl-washed-/x 1 mL suprapur HNO 3 / 100 mL VTTRaumanymp. lab.-/- Bottle is filled to the brim. GTK-/xTheNetherlandsUppsala-/- GTK-/- Sampling time is written down. STUK-/--/--/x-/xSample volume depends on SO 4 -concentration. Zn(Ac)2 is added tosample according to Posiva watersampling quide.50 mL conc. HCl / 1 L.Filters are sent to HYRL for analyses.50 mL conc. HCl / 1 L.Filters are sent to HYRL for analyses.PE = polyethylene, HDPE = high-density polyethyleneLaboratories:TVOTeollisuuden Voima OyVTTVTT Technical Research Centre of FinlandRauman ymp. lab.Rauman ympäristölaboratorioUppsalaÅngström-laboratory, University of Uppsala, SwedenGTKGeological Survey of FinlandWaterlooEnvironmental Isotope Lab, University of Waterloo, CanadaThe NetherlandsCentre for isotope research, Groningen, The NetherlandsSTUKRadiation and Nuclear Safety Authority, Helsinki, FinlandHYRLDepartment of Radiochemistry, University of Helsinki, FinlandWaterlooHYRLHYRL

121Table A-3. Methods and detection limits for groundwater chemistry.Parameter Apparatus and method Detection limit Uncertainty of themeasurementpHpH meter 0.05ISO-10532ConductivityConductivity analyser5 µS/cm 5%SFS-EN-27888Density Posiva water sampling guide /1 0.001 g/cm 3Sodium fluorescein Fluorometry 1 µg/l 15 µg/l: 0.8%200 µg/l: 1.2%275 µg/l: 0.4%AlkalinityTitration/Posiva water0.05 mmol/L 5%sampling guide /1AcidityTitration/Posiva water0.05 mmol/L 10%sampling guide /1DOC/DIC SFS-EN 1484 0.1 mg/LFe totSpectrophotometry/ Posiva 0.01 mg/L 5%water sampling guide /1Fe 2+Spectrophotometry/ Posiva 0.01 mg/L 5%water sampling guide/1Fe tot ,, MnK, NaSiO 2CaMgICP/OES0.002 mg/L0.5 mg/L0.01 mg/L0.1 mg/L0.02 mg/LSrB totICP-MS 0.5 µg/L2 µg/LClTitration/Posiva water5 mg/L 2.5%sampling guide/1BrIC, conductivity detector. SFS- 0.5 mg/L 4.2%EN ISO 10304-1FISE/ Posiva water sampling 0.1 mg/L 5%guide/1PO 4Spectrophotometer0.012 mg/L ± 24%S 2–SFS-EN 1189SpectrophotometerSFS 30380.01 mg/L 0.07 mg/L: 36%0.17 mg/L: 17%0.53 mg/L: 10%1.25 mg/L ± 3.2%SO 4IC, conductivity detector. SFS-EN ISO 10304-1S tot H 2 O 2 oxidation +IC 0.2 mg/LNH 4Spectrophotometer0.002 mg/L 4%SFS 3032Total nitrogen, N totNitrate, NO 3Nitrite, NO 2HPLCSFS3031HPLCInternal method n:o 10SpectrophotometerSFS3029:19760.20 mg/L3.0 mg/L 3.0–5.0 mg/L: 12%>5.0 mg/L: 7%0.010 mg/L 0.010–0.10 mg/L:10%>0.10 mg/L: 8%18 O MS < 0.1‰18 O (SO 4 ) MS 0.5‰

122Table A-3 (continued). Methods and detection limits for groundwater chemistry.3 H Electrical enrichment + homemade Proportional Gas counter(PGC) detection method0.2 TU 100 ± 2,20 ± 0.5 and1.00 ± 0.10 TU2 H MS 1‰13 C (DIC) MS 0.3 pM 0.05‰14 C (DIC) AMS 0.1 pM86 Sr/ 87 Sr MS 0.003‰34 S (SO 4 ) MS 0.1 mBq/L 0.2‰Rn-222 Liquid scintillation counting / 2 5–10%U(tot) jaAlfaspectrometer0.2 mBq/LU-234/U-238 ASTM D3648-95, 1995References1 Paaso, N. (toim.), Mäntynen, M., Vepsäläinen, A. ja Laakso, T. 2003. Posivan vesinäytteenotonkenttätyöohje, rev.3, Posiva Työraportti 2003-02.2 Salonen L. and Hukkanen H., Advantages of low-background liquid scintillation alphaspectrometryand pulse shape analysis in measuring 222Rn, uranium and 226Ra in groundwatersamples, Journal of Radioanalytical and Nuclear Chemistry, Vol. 226, Nos. 1–2, 1997.

Table A-4. Physical and chemical data for the sampled groundwater.Bore-holeSampling date(Y-M-D)Depth(m)T(oC)pHConductivity(mSm –1 )TDS (mgL –1 )AlkalinityHCO 3(mg L –1 )Acidity(Meq L –1 )DOC(mg CL –1 )DIC(mg CL –1 )O 2HACHelectrode(mg L –1 )O 2Winkler(mg L –1 )E hHACHelectrode(mV)PR1 2004-05-04 6 4.6 5.0 12 78 23.7 2.80 469PR1 2005-10-10 6 11.2 5.2 12 110 30.5 0.94 21.5 0.08 402PR1 2006-04-25 6 3.9 5.3 15 120 33.6 1.12 6.50 16.4 0.42 0.44 134PR1 2006-10-11 6 10.2 5.5 0 98 27.5 0.88 6.10 13.3 0.21 0.23 224PP2 2004-05-05 14.7 6.2 7.2 80 575 271.0 0.45 81.5PP2 2005-10-12 14.7 7.3 7.4 81 650 282.0 0.24 59.9

PVP1 2006-04-27 3.9 3.6 4.8 16 160 76.9 2.24 196.00 3.4 2.11 0.44 126PVP1 2006-10-12 3.9 10.4 5.3 11 93 20.1 0.57 19.20 5.5 5.95 4.31 208PVP3A 2004-05-03 7.8 5.6 6.8 59 413 179.0 1.73 247PVP3B 2004-05-03 3.8 6.6 6.5 59 382 123.0 0.25 315PVP4A 2004-05-04 9.6 6.4 7.0 73 553 286.0 1.04 216PVP4A 2005-10-12 10.2 9.3 7.2 80 750 336.0 0.44 64.4

OL-KR8 2005-10-25 57.3 7.3 65 490 221.0

Table A-4. Continued.Borehole Sampling date Depth(m)E h Pt-probe(mV)SO 42–(mg L –1 )S 2–(mg L –1 )Fe 2+(mg L –1 )Ntot(mg L –1 )NO 2–(mg L –1 )NO 3–(mg L –1 )NH 4(mg L –1 )Cl -(mg L –1 )F -(mg L –1 )Br -(mg L –1 )PR1 2004-05-04 6 33.00 0.54 4 0.10

PVP1 2006-10-12 3.9 23.00

OL-KR8 2006-06-06 260.7 470.00 0.08 0.10 0.21

Table A-4. Continued.Borehole Sampling date Depth (m) SiO 2(mg L –1 )Na(mg L –1 )K(mg L –1 )Ca(mg L –1 )Mg(mg L –1 )Mn(mg L –1 )Sr(mg L –1 )B(mg L –1 )U(µg L –1 )PR1 2004-05-04 6 4 1.1 7 3.7PR1 2005-10-10 6 13.4 6 1.7 8 4.0 0.32 0.025PR1 2006-04-25 6 11.3 7 1.6 10 4.8 0.38 0.032 0.02 0.6PR1 2006-10-11 6 12.3 5 1.7 8 3.8 0.27 0.023 0.03 0.7PP2 2004-05-05 14.7 51 7.5 79 18.3PP2 2005-10-12 14.7 22.4 46 7.8 134 23.2 1.12 0.3PP2 2006-04-24 14.7 21.6 40 7.2 94 19.1 1.13 0.4 0.05

PVP1 2006-10-12 3.9 19.6 7 2.0 5 3.1 0.07 0.02 0.03 2.1PVP3A 2004-05-03 7.8 54 7.3 41 12.0PVP3B 2004-05-03 3.8 61 6.8 32 11.0PVP4A 2004-05-04 9.6 29 5.9 94 15.1PVP4A 2005-10-12 10.2 24.6 57 10.6 160 19.3 1.77 0.2PVP4A:1 2006-04-27–0 h 10.2 21.2 31 6.4 97 15.7 1.36 0.3 0.04 0.1PVP4A:2 2006-04-27–6 h 10.2 20.9 32 6.9 99 16.5 1.39 0.3 0.04

OL-KR8 2006-06-06 260.7 11.0 1800 7.9 810 160.0 0.62 8.4 0.63 0.6OL-KR10 2005-02-21 106 11.0 543 2.5 62 11.0 0.05 0.6 0.64OL-KR10 2006-06-19 316 10.0 1790 7.8 990 58.0 0.55 9.9 1.30

Table A-5. Gas data for the sampled shallow groundwater. Rows in italics with “-SD%” after the borehole code show the standard deviationsin percent of the average. The number of samples (n) was 2 for the April 2006 analyses and 3 for the October 2006 analyses.Borehole PumplevelDepthZ-upSample andextractionVolumewaterExtractedgasExtractedgasOxygen(ppm)Hydrogen(ppm)Helium(ppm)Argon(ppm)(m)(m) date (mL) (mL) (mL L –1 )PR1 4 6 2006-04-25 91.5 6.2 66.5 35450 24.4 0 n.a.PR1-SD% 4 6 2006-04-25 7.0 37.9 31.4 61.6 41.5PR1 4 6 2006-10-11 104.7 6.4 61.0 10650 52.5 0 n.a.PR1-SD% 4 6 2006-10-11 2.8 36.5 37.8 61.0 43.8PP2 6 14.7 2006-04-24 97.5 2.9 28.5 82400 31.7 0 n.a.PP2-SD% 6 14.7 2006-04-24 6.5 86.8 82.6 45.7 82.9PP2 4 14.7 2006-10-11 104.0 3.1 29.4 55695 37.3 0 n.a.PP2-SD% 4 14.7 2006-10-11 3.5 24.5 22.6 31.5 12.6PP9 6 14.7 2006-04-26 93.5 4.0 42.8 259500 13.2 0 n.a.PP9-SD% 6 14.7 2006-04-26 0.8 0.0 0.8 80.4 8.1PP9 6 14.7 2006-10-09 97.3 2.4 24.6 43600 42.7 0 n.a.PP9-SD% 6 14.7 2006-10-09 4.6 22.0 18.2 28.0 29.2PP36 4 11 2006-04-25 102.5 5.4 53.2 30350 10.8 0 n.a.PP36-SD% 4 11 2006-04-25 9.0 35.7 43.9 34.2 19.1PP36 5 11 2006-10-11 95.0 3.9 41.8 69000 40.8 0 n.a.PP36-SD% 5 11 2006-10-11 6.9 53.4 57.0 62.5 73.1PP39 5 14 2006-04-24 82.0 4.0 48.3 22950 15.1 0 n.a.PP39-SD% 5 14 2006-04-24 5.2 9.0 14.1 44.1 22.1PP39 6 14 2006-10-11 97.7 7.2 73.6 10650 13.6 0 n.a.PP39-SD% 6 14 2006-10-11 5.8 28.0 24.8 61.0 22.1PVP1 2.5 3.9 2006-04-27 101.5 5.0 49.3 49600 11.6 0 n.a.132

PVP1-SD% 2.5 3.9 2006-04-27 2.1 0.0 2.1 1.1 6.7PVP1 2.5 3.9 2006-10-12 99.3 5.3 53.8 29333 21.4 793 n.a.PVP1-SD% 2.5 3.9 2006-10-12 1.2 28.6 29.8 14.3 22.7 156PVP4A-1 6 10.2 2006-04-27–0 h 96.0 2.3 23.4 40250 20.1 0 n.a.PVP4A-1-SD% 6 10.2 2006-04-27–0 h 5.9 15.7 9.9 59.6 1.4PVP4A-2 6 10.2 2006-04-27–6 h 95.5 4.1 43.0 17600 20.2 0 n.a.PVP4A-2-SD% 6 10.2 2006-04-27–6 h 0.7 65.5 66.1 53.0 47.4PVP4A 5 10.2 2006-10-11 101.3 3.6 36.0 31900 24.8 0 n.a.PVP4A-SD% 5 10.2 2006-10-11 2.3 17.9 19.7 32.3 12.9PVP13 3 5.6 2006-04-26 101.0 3.6 35.7 37000 17.9 0 n.a.PVP13-SD% 3 5.6 2006-04-26 1.4 15.7 17.1 81.4 11.1PVP13 4 5.6 2006-10-12 99.0 4.5 45.5 8120 20.0 0 n.a.PVP13-SD% 4 5.6 2006-10-12 5.3 29.6 32.4 105.7 49.2PVP14 6 9.1 2006-04-26 83.5 3.2 37.7 30000 18.3 0 n.a.PVP14-SD% 6 9.1 2006-04-26 2.5 11.2 8.7 0.9 4.6PVP14 5.5 9.1 2006-10-10 95.0 4.7 49.7 23400 26.7 0 n.a.PVP14-SD% 5.5 9.1 2006-10-10 1.1 31.8 32.3 21.1 18.7PVP20 12.8 12.8 2006-10-10 100.3 3.8 38.5 27340 30.8 0 n.a.PVP20-SD% 12.8 12.8 2006-10-10 6.8 28.5 33.9 124.5 25.2PVA1 ONKALO 20 2006-04-28 98.0 4.0 40.8 55150 24.8 0 n.a.PVA1-SD% ONKALO 20 2006-04-28 2.9 0.0 2.9 3.2 0.0PVA1 ONKALO 20 2006-10-11 99.3 4.0 39.9 8873 69.3 0 n.a.PVA1-SD% ONKALO 20 2006-10-11 1.5 31.9 31.1 48.9 51.6133

Table A-5. ContinuedBoreholeSample andextractiondateNitrogen(ppm)CO(ppm)CO 2(ppm)CH 4(ppm)C 2 H 6(ppm)C 2 H 2–4(ppm)total gas(%)PR1 2006-04-25 431500 44.1 529500 2905 0 0 99.9PR1-SD% 2006-04-25 39 14.6 27 35 0.1PR1 2006-10-11 461667 29.0 472000 258 0 0 94.5PR1-SD% 2006-10-11 23 22.0 19 17 3.4PP2 2006-04-24 706500 96.2 239000 4650 0 0 103.3PP2-SD% 2006-04-24 36 68.8 86 82 0.6PP2 2006-10-11 706334 41.4 202218 2044 0 0 96.6PP2-SD% 2006-10-11 3 28.6 10 16 1.8PP9 2006-04-26 523500 54.8 249000 327 0 0 103.2PP9-SD% 2006-04-26 32 11.5 4 5 3.1PP9 2006-10-09 501000 125.0 454333 295 0 0 99.9PP9-SD% 2006-10-09 28 16.5 32 18 1.7PP36 2006-04-25 566000 91.5 433000 230 0 0 103.0PP36-SD% 2006-04-25 24 27.0 24 55 2.3PP36 2006-10-11 449667 112.1 453667 262 0 0 97.3PP36-SD% 2006-10-11 65 83.3 57 105 2.1PP39 2006-04-24 542000 35.8 451000 22500 0 0 103.9PP39-SD% 2006-04-24 19 22.1 19 18 0.1PP39 2006-10-11 461667 29.0 472000 258 0 0 94.5PP39-SD% 2006-10-11 23 22.0 19 17 3.4PVP1 2006-04-27 439000 78.6 528500 432 0 0 101.8PVP1-SD% 2006-04-27 0 26.6 2 72 0.9PVP1 2006-10-12 591667 60.6 293333 674 0.1 0 91.6134

PVP1-SD% 2006-10-12 8 69.3 18 37 100 3.2PVP4A-1 2006-04-27–0 h 700000 37.2 306000 4140 0 0 105.0PVP4A-1-SD% 2006-04-27–0 h 10 16.7 13 15 0.8PVP4A-2 2006-04-27–6 h 786000 44.5 246500 2820 0 0 105.3PVP4A-2-SD% 2006-04-27–6 h 23 45.8 66 46 0.5PVP4A 2006-10-11 697667 31.4 265000 2730 0 0 99.7PVP4A-SD% 2006-10-11 6 22.1 12 14 0.2PVP13 2006-04-26 785500 46.5 207500 2370 0 0 103.2PVP13-SD% 2006-04-26 1 40.4 24 17 1.5PVP13 2006-10-12 715667 28.2 228000 1082 0 0 95.3PVP13-SD% 2006-10-12 4 13.8 33 30 10.4PVP14 2006-04-26 795000 41.4 220500 289 0 0 104.6PVP14-SD% 2006-04-26 6 45.0 21 1 0.4PVP14 2006-10-10 800333 12.6 157000 38 0 0 98.1PVP14-SD% 2006-10-10 6 135.4 23 138 0.8PVP20 2006-10-10 769333 28.6 193333 827 0 0 99.1PVP20-SD% 2006-10-10 11 50.0 22 31 0.9PVA1 2006-04-28 716000 32.5 40350 3415 8 0 81.5PVA1-SD% 2006-04-28 0 74.0 13 20 8 0.7PVA1 2006-10-11 926000 43.5 70400 6817 0 0 101.2PVA1-SD% 2006-10-11 4 34.5 16 29 1.7135

Table A-6. Gas data for the sampled deep groundwater.Borhole Upper Lower Depth Sample Extraction Time from Volume Extracted Extracted Analysedlevel level Z-updatedatesampling towater gas gasair(m) (m) (m)analysis (d)(mL) (mL) (mL L –1 ) (%)OL-KR2 596.5 609.5 560 2006-02-28 2006-03-07 7 90 47.0 522.2 0.76OL-KR6 422 425 328 2005-08-02 2005-08-24 22 254 15.5 61.0 11.40OL-KR6 135 137 102 2005-09-27 2005-10-17 20 82 9.5 115.9 2.20OL-KR6 120 125 90 2005-11-02 2005-12-12 40 77 7.5 97.4 4.80OL-KR6 98.5 100.5 73 2005-12-27 2006-01-13 17 201 9.5 47.3 1.10OL-KR6 125 130 94 2006-06-26 2006-07-02 6 76 10.0 131.6 1.89OL-KR6 135 137 116 2006-08-22 2006-08-28 6 65 5.0 76.9 1.13OL-KR6 98.5 100 74 2006-10-16 2006-10-24 8 70 6.2 88.6 1.22OL-KR7 284 288 257 2006-04-25 2006-05-11 16 220 8.6 39.1 5.20OL-KR7-Ar 220 230 197 2005-04-25 2005-08-22 119 30 5.6 186.7 12.10OL-KR7-N2 220 230 197 2005-04-25 2005-08-23 120 74 7.2 97.3 14.40OL-KR8 77 84 57 2005-10-25 2005-12-12 48 183 5.9 32.2 1.85OL-KR8 556.5 561 490 2006-04-27 2006-05-11 14 95 41.1 432.6 0.86OL-KR8 302 310 261 2006-06-06 2006-06-09 3 101 9.2 91.1 0.64OL-KR8 77 84 57 2006-08-15 2006-08-28 13 195 6.8 34.9 1.59OL-KR10 259 262 249 2005-04-04 2005-06-26 83 178 14.0 78.7 12.40OL-KR10 326 328 316 2006-06-19 2006-06-21 2 236 29.5 125.0 2.60OL-KR10-Ar 326.5 328.5 316 2005-04-04 2005-08-23 141 100 9.4 94.0 8.70OL-KR10-N2 326.5 328.5 316 2005-04-04 2005-08-23 141 127 23.6 185.8 0.90OL-KR13 362 365 294 2006-03-14 2006-03-27 13 100 11.4 114.0 0.31OL-KR19 110 131 101 2005-09-05 2005-10-05 30 88.4 7.2 81.4 2.60OL-KR19 455 468 433 2005-10-31 2005-12-12 42 107 25.4 237.4 1.12136

OL-KR22 147 152 116 2005-12-13 2006-01-13 31 231 15.8 68.4 2.94OL-KR22 390 394 320 2006-03-01 2006-03-07 6 244 83.0 340.2 0.00OL-KR22 147 152 102 2006-08-17 2006-08-28 11 95 5.2 54.7 3.38OL-KR29 320 340 293 2005-06-06 2005-08-23 78 65 7.0 107.7 13.30OL-KR29 800 800 742 2005-04-16 2005-08-23 129 163 225.0 1380.4 1.70OL-KR30 50 54 40 2005-08-04 2005-08-24 20 111 4.6 41.4 17.40OL-KR31 143 146 122 2006-10-24 2006-10-26 2 250 8.2 32.8 3.72OL-KR33 95 107 71 2006-01-24 2006-01-26 2 73 8.0 109.6 9.75OL-KR37 166 176 112 2006-11-28 2006-11-30 2 92 4.8 52.2 4.34OL-KR39 403 406 345 2006-04-03 2006-04-06 3 107 15.5 144.9 1.39OL-KR39 108 110 88 2006-05-30 2006-06-09 10 79 4.2 53.2 1.96137

Table A-6. ContinuedBorehole Hydrogen(ppm)Helium(ppm)Argon(ppm)Nitrogen(ppm)CO(ppm)CO 2(ppm)CH 4(ppm)C 2 H 6(ppm)C 2 H 2–4(ppm)total gas(%)OL-KR2 68.6 24400 241 290000 6.30 140 699000 6250.00 0.00 102.0OL-KR6 8.4 54800 3000 671000 8.80 2980 267000 1330.00 0.00 100.0OL-KR6 40.9 1490 35700 954000 26.80 6980 2100 23.80 0.00 100.0OL-KR6 2710.0 1730 4590 905000 72.00 15400 13400 60.10 1.09 94.3OL-KR6 11.6 1130 2440 1010000 16.40 25400 1580 3.97 0.00 104.1OL-KR6 16.0 1030 398 973000 25.80 10900 3090 30.20 0.00 98.8OL-KR6 6.5 1130 4090 973000 9.30 26400 1020 3.90 0.00 100.6OL-KR6 11.2 715 0 981000 15.40 32900 761 4.51 0.00 101.5OL-KR7 5.2 11900 695 953000 6.00 16800 13900 21.20 0.00 99.6OL-KR7-Ar 20.5 30200 51800 904000 20.40 10100 3820 23.80 0.00 100.0OL-KR7-N2 24.7 0 15900 949000 28.40 23500 11000 119.00 0.00 100.0OL-KR8 7540.0 0 1330 936000 27.70 14500 6150 65.90 0.00 96.6OL-KR8 6.5 22500 166 237000 2.30 152 744000 4170.00 0.00 100.8OL-KR8 18.7 8670 893 981000 41.00 8140 2000 24.40 0.00 100.1OL-KR8 27.7 0 4970 953000 15.60 56300 2600 9.08 0.00 101.7OL-KR10 51.9 8030 2500 972000 33.40 5310 12400 30.40 7.31 100.0OL-KR10 21.4 35200 312 440000 14.40 1830 547000 2120.00 0.00 102.6OL-KR10-Ar 5.8 19100 4790 751000 4.30 2040 222000 627.00 0.00 100.0OL-KR10-N2 28.6 18600 10200 654000 4.70 1470 315000 1300.00 0.00 100.1OL-KR13 651.0 30900 1740 769000 5.80 16100 175000 773.00 0.00 99.4OL-KR19 54.0 1630 17100 959000 7.40 21400 841 4.50 0.00 100.0OL-KR19 1320.0 18000 1570 481000 17.40 366 493000 2780.00 0.00 99.8OL-KR22 31.9 898 1810 975000 30.80 33100 14100 57.90 0.00 102.5OL-KR22 33.8 2460 529 986000 6.10 4130 28500 79.80 0.00 102.2138

OL-KR22 10.0 978 3667 942000 8.80 59400 17200 41.80 0.00 102.3OL-KR29 15.6 5330 6430 968000 9.90 2190 16800 1090.00 0.58 100.0OL-KR29 51.8 14900 2660 186000 3.00 1460 779000 15400.00 0.00 100.0OL-KR30 107.0 0 0 976000 32.60 14000 10800 220.00 0.00 100.1OL-KR31 9.7 920 30000 916000 44.30 34200 5320 2.27 1.62 98.6OL-KR33 94.0 0 0 922000 74.80 9940 582 5.68 0.00 93.3OL-KR37 22.3 1400 17100 933000 22.00 55000 2860 5.86 0.00 100.9OL-KR39 18.3 19600 743 408000 12.40 7170 582000 1470.00 0.00 101.9OL-KR39 7.6 0 377 960000 15.10 22600 1080 7.12 0.00 98.4139

Table A-7. Biomass determinations for shallow groundwater in Olkiluoto, sampled over spring and fall seasons. TNC = total number of cells,SD = standard deviation, n = number of observations, CHAB = cultivable heterotrophic aerobic bacteria, MPN = sum of all most probablenumber of cells values (see Table A-8), and n.a. = not analysed for various reasons, for example, inapplicable because the analysis had not yetbeen introduced, sample turbidity, or analytical error.Borehole sampled(Y-M-D)Depth(m)TNC(cells mL 1 )SD n ATP(amol mL 1 )SD n CHAB(cells mL 1 )SD n CHAB/TNC(%)ATP/TNCMPN/TNC(%)PR1 2004-05-04 6.0 57000 57000 6 n.a. 5150 1450 3 9.04 0.29PR1 2005-10-10 6.0 2000000 450000 6 266000 7270 3 11700 117 3 0.59 0.133 1.46PR1 2006-04-25 6.0 820000 84000 3 626000 34700 3 1600 2040 3 0.20 0.763 0.30PR1 2006-10-11 6.0 390000 21000 3 128000 8590 3 827 107 3 0.21 0.328 0.64PP2 2004-05-05 14.7 32000 16000 6 n.a. n.a. - - 0.01PP2 2005-10-12 14.7 110000 53000 6 25100 2440 3 310 399 3 0.28 0.228 0.12PP2 2006-04-24 14.7 55000 1400 3 13300 680 3 553 98 3 1.01 0.242 0.04PP2 2006-10-11 14.7 10000 1100 3 1900 50 3 27 46 3 0.27 0.190 0.10PP3 2004-05-05 14.3 190000 84000 6 n.a. 93 65 3 0.05 0.00PP7 2004-05-05 16.2 31000 23000 6 n.a. 240 150 3 0.77 0.06PP8 2004-05-06 15.2 1500000 160000 5 n.a. 1900 707 2 0.13 0.02PP9 2005-10-13 14.7 200000 46000 6 24500 1770 3 13 6 3 0.01 0.123 0.01PP9 2006-04-27 14.7 n.a. 109000 1400 3 70 46 3PP9 2006-10-09 14.7 220000 5400 3 104000 4610 3 677 508 3 0.31 0.473 0.10PP36 2005-10-10 12.1 110000 45000 6 26900 1080 3 37 6 3 0.03 0.245 0.08PP36 2006-04-25 12.1 370000 29000 3 220600 10600 3 427 60 3 0.12 0.596 0.09PP36 2006-10-09 12.1 400000 15000 3 146000 13600 3 740 42 2 0.19 0.365 0.05PP39 2005-10-11 14.1 580000 560000 6 198000 10800 3 913 93 3 0.16 0.341 0.14PP39 2006-04-24 14.1 540000 140000 3 90400 3110 3 553 64 3 0.10 0.167 0.21PP39 2006-10-11 14.1 410000 42000 3 170000 5220 3 983 145 3 0.24 0.415 1.09140

Table A-7. continued.Borehole sampled(Y-M-D)Depth(m)TNC(cells mL 1 )SD n ATP(amol mL 1 )SD n CHAB(cells mL 1 )SD n CHAB/TNC(%)ATP/TNCMPN/TNC(%)PVP1 2004-05-04 3.9 1500000 400000 6 n.a. 15100 5130 3 1.01 0.00PVP1 2005-10-11 3.9 2500000 670000 6 685000 59900 3 917 218 3 0.04 0.274 0.01PVP1 2006-04-27 3.9 1100000 18000 3 7920000 307000 3 4400 3200 3 0.40 7.200 2.67PVP1 2006-10-12 3.9 200000 16000 3 624000 61800 3 530 125 3 0.27 3.120 0.10PVP3A 2004-05-03 7.8 150000 66000 6 n.a. 1060 482 3 0.71 - 0.00PVP3B 2004-05-03 3.8 41000 53000 6 n.a. 6690 3610 3 16.32 0.17PVP4A 2004-05-04 9.6 96000 55000 6 13200 3 690 539 3 0.72 0.01PVP4A 2005-10-12 10.2 660000 14000 6 30200 4160 3 1240 59 3 0.19 0.046 0.20PVP4A 2006-10-10 10.2 9500 1400 3 1860 170 3 0 0 3 0.00 0.196 0.14PVP4A:1 2006-04-27 10.2 7600 1300 3 11400 2800 3 2330 115 3 30.66 1.500 3.05PVP4A:2 2006-04-27 10.2 7800 1100 3 6320 630 3 173 60 3 2.22 0.810 2.02PVP4A:1/2 2006-04-27 10.2 0.97 1.80 13.5PVP4B 2004-05-04 8.0 51000 19000 6 n.a. 79200 17300 3 155.29 0.01PVP13 2005-10-12 5.6 120000 17000 6 79400 10300 3 1400 106 3 1.17 0.662 0.73PVP13 2006-04-26 5.6 17000 1400 3 12800 560 3 83 40 3 0.49 0.753 2.28PVP13 2006-10-12 5.6 17000 1400 3 12300 790 3 23 12 3 0.14 0.724 0.62PVP14 2005-10-13 9.0 90000 80000 6 3570 480 3 57 30 3 0.06 0.040 0.02PVP14 2006-04-26 9.0 20000 2600 3 3620 670 3 10 7 3 0.05 0.181 0.05PVP14 2006-10-10 9.0 9300 1900 3 4520 280 3 30 26 3 0.32 0.486 3.85PVP20S 2005-10-13 12.8 320000 43000 6 106000 7740 3 2220 415 3 0.69 0.331 0.56PVP20P 2005-10-13 12.8 150000 76000 6 76100 3890 3 1970 225 3 1.31 0.507 0.14PVP20 2006-04-26 12.8 ICE- COVEREDPVP20 2006-10-10 12.8 n.a. - 3 367000 51500 3 783 196 3 - - -141

Table A-8. The most probable numbers of nitrate-, iron-, manganese-, and sulphate-reducing bacteria (NRB, IRB, MRB, and SRB, respectively)in shallow groundwater of Olkiluoto. L and U limits are the 95% confidence values. n.a. = not analysed for various reasons, for example,inapplicable because the analysis had not yet been introduced, sample turbidity, or analytical error.Boreholesampled(Y-M-D)Depth(m)NRB(cellsmL 1 )LlimitUlimitIRB(cellsmL 1 )PR1 2004-05-04 6.0 n.a. 0.4 0.1 1.7 0.0 3.0 1.0 12.0PR1 2005-10-10 6.0 24000.0 10000.0 94000.0 1.3 0.5 3.8 0.8 0.3 2.4 23.0 9.0 86.0PR1 2006-04-25 6.0 1700.0 700.0 4800.0 2.3 0.9 8.6 70.0 30.0 210.0 24.0 10.0 94.0PR1 2006-10-11 6.0 2400.0 1000.0 9400.0 0.2 0.1 1.1 1.3 0.5 3.8 3.0 1.0 12.0PP2 2004-05-05 14.7 n.a.

Table A-8. Continued.Bore-Holesampled(Y-M-D)Depth(m)NRB(cellsmL 1 )LlimitUlimitIRB(cellsmL 1 )PVP1 2004-05-04 3.9 n.a. 0.4 0.1 1.7 9.0 4.0 25.0 0.8 0.3 2.4PVP1 2005-10-11 3.9 160.0

Table A-9. The most probable numbers of autotrophic acetogens (AA) and methanogens (AM), heterotrophic acetogens (HA) and methanogens(HM), and methane-oxidizing bacteria (MOB) in shallow groundwater from Olkiluoto. L and U limits are the 95% confidence values. n.a. = notanalysed for various reasons, for example, inapplicable because the analysis had not yet been introduced, sample turbidity, or analytical error.Boreholesampled(Y-M-D)Depth(m)AAcellsmL 1LlimitUlimitHAcellsmL 1LlimitUlimitPR1 2004-05-04 6.0 160.0

Table A-9. ContinuedBoreholesampled(Y-M-D)Depth(m)AAcellsmL 1LlimitUlimitHAcellsmL 1LlimitUlimitPVP1 2004-05-04 3.9

Table A-10. Biomass determinations for deep groundwater in Olkiluoto. TNC = total number of cells, SD = standard deviation, n = number ofobservations, CHAB = cultivable heterotrophic aerobic bacteria, MPN = sum of all most probable number of cells values (see Table A-11),and n.a. = not analysed for various reasons, for example, inapplicable because the analysis had not yet been introduced, sample turbidity, oranalytical error.Boreholesampled(Y-M-D)Sectionupper–lower(m)Midelevation,z(m)TNC(cellsmL 1 )SD n ATP(amolmL 1 )SD n CHAB(cellsmL 1 )SD n CHAB/TNC(%)ATP/TNCMPN/TNC(%)OL-KR2 2004-12-20 328.5–330.5 306.2 120000 24000 6 206000 40500 3 n.a. 1.717 0.70OL-KR6 2006-05-11 422–425 328.4 100000 7800 2 66800 1520 3 120000 5600 3 120.00 0.668 30.25OL-KR6 2006-06-26 125–130 94.1 2700 1400 3 5720 1390 3 503 187 3 18.63 2.119 20.00OL-KR6 2006-08-22 135–137 101.8 23000 3900 3 15300 410 3 16300 4730 3 70.87 0.665 22.55OL-KR6 2006-10-16 98.5–100.5 73.7 4500 1300 3 3090 170 3 1090 21 3 24.22 0.687 1.08OL-KR7 2005-03-01 275.5–289.5 249.4 74000 15000 6 15800 210 3 n.a. 0.214 0.02OL-KR8 2005-10-25 77.0–84.0 57.3 130000 27000 6 6960 440 3 497 21 3 0.38 0.054 0.16OL-KR8 2006-06-06 302.0–310.0 260.7 11000 2800 3 5730 320 3 820 123 3 7.45 0.521 1.89OL-KR10 2005-02-21 115.5–118.5 106.0 140000 21000 6 20800 10900 3 n.a. 0.149 0.22OL-KR10 2006-06-19 326.0–328.0 316.0 110000 15000 3 25000 1800 3 1390 21 3 1.26 0.227 0.32OL-KR13 2004-10-12 362.0–365.0 294.0 110000 27000 6 82200 9580 3 n.a. 0.00 0.747 0.07OL-KR13 2006-03-14 362.0–365.0 294.0 27000 14000 6 15580 200 3 16300 3400 3 60.37 0.577 3.56OL-KR19 2004-11-08 525.5–539.5 449.6 150000 25000 6 n.a. n.a. 0.17OL-KR27 2004-11-09 247.0–264.0 193.5 29000 14000 6 36400 4590 3 n.a. 1.255 0.05OL-KR27 2005-01-17 503.0–506.0 391.7 21000 4000 6 n.a. n.a. - 0.12OL-KR31 2006-10-24 143.0–146.0 122.4 19000 2600 3 10929 940 3 4230 351 3 22.26 0.575 6.20OL-KR32 2006-01-10 50.0–52.0 34.6 26000 7600 6 23400 1890 3 2430 379 3 9.35 0.900 5.30OL-KR33 2006-01-24 95.0–107.0 70.6 40000 15000 5 6970 490 3 10 10 3 0.03 0.174 0.37OL-KR37 2006-11-28 166–176 111.6 14000 730 3 8020 200 3 3070 473 3 21.93 0.573 21.66OL-KR39 2006-04-03 403.0–406.0 344.8 22000 7600 3 7440 260 3 180 25 3 0.82 0.338 1.60OL-KR39 2006-05-30 108.0–110.0 88.2 21000 960 3 13080 750 3 3150 778 2 15.00 0.623 1.83146

Table A-11. The most probable numbers of nitrate-, iron-, manganese-, and sulphate-reducing bacteria (NRB, IRB, MRB, and SRB,respectively), autotrophic acetogens (AA) and methanogens (AM), heterotrophic acetogens (HA) and methanogens (HM), and methaneoxidizingbacteria (MOB) in deep groundwater from Olkiluoto. L and U limits are the 95% confidence values. n.a. = not analysed for variousreasons, for example, inapplicable because the analysis had not yet been introduced, sample turbidity, or analytical error.Boreholesampled(Y-M-D)Sectionupper–lower(m)Midelevation,z(m)NRB(cellsmL 1 )LlimitUlimitOL-KR2 2004-12-20 328.5–330.5 306.2 n.a. 300 100.0 1300.0 300.0 100.0 1300

Table A-11. Continued.Boreholesampled(Y-M-D)Sectionupper–lower(m)Mid elevation,z(m)AA(cellsmL 1 )LlimitUlimitOL-KR2 2004-12-20 328.5–330.5 306.2 110.0 40.0 300.0 130.0 50.0 390.0 0.4 0.1 1.7 2.3 0.9 8.6OL-KR6 2006-05-11 422–425 328.4 24.0 10.0 94.0 80.0 30.0 250.0

149ANALYSIS OF DISSOLVED GASES IN GROUNDWATERSamples can consist of any combination of gas only, gas and groundwater in separatephases, or groundwater containing dissolved gas, all in closed pressure-safe containers.The sample is transferred to a vacuum container and any gas in the water is boiled offunder vacuum (i.e., water vapour pressure) at room temperature (Figure A-1). After thisextraction, the gas is compressed and transferred to a 10 mL syringe (SGE AnalyticalScience, Victoria, Australia) and the volumes of extracted gas and water are measured.The captured gas is subsequently transferred to a 6.6-mL glass vial stoppered with abutyl rubber stopper sealed with an aluminium crimp seal (Figure A-1). Large gassamples can be transferred to a 27-mL vial instead. The vial is evacuated and flushedtwice with nitrogen, in two cycles, and is left at high vacuum (10 4 Bar). Coppersulphate (dehydrant) is added to adsorb any traces of water remaining in the gas (watercauses troublesome baseline drifts in the gas chromatographs). The vials are storedunder water. Any leakage will result in the blue, dry copper sulphate turning pink.Figure A-1. The 500-mL cylindrical gas extractor with a 6.6-mL sample vial attached,lower right. Gas collection syringes are visible below the metallic SGE 6-port valves.The blue box is the manometer used to measure pressure in the samples. The graycylinder is the cryo-trap for removing moisture from samples.Air contamination during extraction is difficult to avoid. New adaptor equipment hasbeen developed and found to be very efficient. Apparently, any remaining

150contamination can be explained by problems with the PAVE samplers. Currently,a fewa 100 uL air is intruding the extraction procedure, which does not occur when dummiesof pure nitrogen are extracted. There is no oxygen in the analysed deep groundwatersamples obtained using PAVE, so air contamination was subtracted from the resultsbefore recording the data.Figure A-2. The Varian Star 3400CX gas chromatograph is standing closest to themanometers, in the centre of the image. The blue KAPPA V gas chromatograph isvisible behind the Varian.Uncertainties of the used methodsVolumes of 1–1000 µL are injected into the gas chromatograph. The volume used isadjusted according to the sensitivity range of the particular instrument and detector.Several injections are usually needed to determine the proper amount of each gas toinject.The precision of the methods used is the subject of ongoing testing at our laboratory.Recently, we attached three samplers to one groundwater circulation at boreholeKJ0052F01 at the MICROBE laboratory at Äspö (see the SKB International ProgressReport IPR-05-05 for a detailed description of this laboratory; Pedersen 2005a). Thepressure vessel used, the PVB sampler, represents the Swedish analogue of the FinnishPAVE sampler. The results are given in the last section of this method description. Themain conclusions were:

151 The precision of the extractions is currently approximately ± 6% (Table A-12). The uncertainty of the instruments and repeated injections is low, typically 0–4% (Table A-13). The calibration gases used have a maximum accepted mixing uncertainty of ±2%. In total, the analytical uncertainty is currently a maximum of ± 12%.Set-up and calibrationsTwo gas chromatographs are currently in use, as shown in Figure A-2.The chromatographs are calibrated and tested using the four “gas mixtures describedbelow. Multiple points are used for the Varian Star 3400CX gas chromatograph (Varian,Palo Alto, CA, USA), while the KAPPA V gas chromatograph uses single-pointcalibrations. Calibration gases are analysed immediately before analysis of samples, andthe calibration results are used in calculating the concentrations of the gases in thesamples.Special gas 1 (Linde Gas, Pullach, Germany), AGA, certificate no: 28810-3:He 25,700 ppmH 2 964 ppmO 2 10,900 ppmNitrogen962,436 ppmSpecial gas 2 (Linde Gas), AGA, certificate no: 28757-1:Ar 1000 ppmCH 4 2740 ppmCO 2 1040 ppmCO 9.75 ppmNitrogen995,210 ppmSpecial gas 3 (Linde Gas), AGA, certificate no: 28749-1:C 2 H 6 253 ppmC 2 H 4 257 ppmC 2 H 2 248 ppmC 3 H 8 252 ppmC 3 H 6 238 ppmNitrogen998,752 ppmSpecial gas 4 (Linde Gas), AGA, certificate no: 30008-1:H 2 24.6 ppmCO 24.9 ppmNitrogen999,950 ppm

152Analysis of gasLow concentrations of hydrogen (20 ppm) were analysed on a Varian Star 3400CXgas chromatograph using a thermal conductivity detector (TCD) with an oventemperature of 65°C, a detector temperature of 120°C, and a filament temperature of250°C. The hydrogen gas was separated using a Porapak-Q column (2 m 1/8 inchdiameter; Agilent Technologies) followed by a molecular sieve 5A column (6 m 1/8inch) with argon as the carrier gas. Calibration gases 1 and 2 are used.The detection limit of the instrument with a 250-µL injection loop is 5 10 –9 L (20ppm).Carbon monoxide was analysed on a KAPPA-5/E-002 analyser gas chromatographequipped with a 156 1/16-inch stainless steel HayeSep column in line with a 31 1/8-inch stainless steel molecular sieve 5A column, which was subsequently attached to areductive gas detector (RGD). Nitrogen was used as the carrier gas. The sample wasinjected into a 1000-µL injection loop. The sample usually had to be diluted to reach thedetection range of the instrument. This instrument has the most sensitive carbonmonoxide detector on the market. These results were compared with those obtainedusing the Varian Star 3800CX analyser and reported when they agreed.The detection limit of the instrument with a 0.1-mL injection loop is 10 –12 L (1 ppb).Helium was analysed on a Varian Star 3400CX gas chromatograph using a thermalconductivity detector (TCD) with an oven temperature of 65°C, a detector temperatureof 120°C, and a filament temperature of 250°C. The helium gas was separated using aPorapak-Q column (2 m 1/8 inch diameter) followed by a molecular sieve 5A column(6 m 1/8 inch) with argon as the carrier gas.The detection limit of the instrument with a 250-µL injection loop is 5 10 –9 L (20ppm).Nitrogen was analysed on a Varian Star 3400CX gas chromatograph using a thermalconductivity detector (TCD) with an oven temperature of 65°C, a detector temperatureof 120°C, and a filament temperature of 250°C. The nitrogen gas was separated using aPorapak-Q column (2 m 1/8 inch diameter) followed by a molecular Sieve 5A column(6 m 1/8 inch). Argon or helium can be used as the carrier gas. The results obtained

153using argon were compared with those obtained using helium and reported when theyagreed.The detection limit of the instrument with a 250-µL injection loop is 25 10 –9 L (100ppm).Oxygen was analysed on a Varian Star 3400CX gas chromatograph using a thermalconductivity detector (TCD) with an oven temperature of 65°C, a detector temperatureof 120°C, and a filament temperature of 250°C. The oxygen gas was separated using aPorapak-Q column (2 m 1/8 inch diameter) followed by a molecular sieve 5A column(6 m 1/8 inch) with argon as the carrier gas.The detection limit of the instrument with a 250-µL injection loop is 25 10 –9 L (100ppm).Argon was analysed on a Varian Star 3400CX gas chromatograph using a thermalconductivity detector (TCD) with an oven temperature of 65°C, a detector temperatureof 120°C, and a filament temperature of 250°C. The argon gas was separated using aPorapak-Q column (2 m 1/8 inch diameter) followed by a molecular sieve 5A column(6 m 1/8 inch) with helium as the carrier gas. Argon was very difficult to separatefrom oxygen. The strategy used was to analyse the total amount of oxygen and argonwith this configuration; then the result was reduced by the amount of oxygen analysed,using argon as the carrier gas.The detection limit of the instrument with a 250-µL injection loop is 25 10 –9 L (100ppm).Carbon dioxide was analysed on a Varian Star 3400CX gas chromatograph using aflame ionization detector (FID) with an oven temperature of 65°C and a detectortemperature of 200°C. The carbon dioxide gas was separated using a Porapak-Q column(2 m 1/8 inch diameter) and transformed to methane using a 10% Ni 2 NO 3“methanizer” fed with hydrogen gas (9.375 1/8 inch diameter, temperature 370°C).Carbon dioxide was finally analysed as methane on the FID with nitrogen as the carriergas. This configuration used a 156 1/16-inch stainless steel HayeSep and an FIDdetector.The detection limit of the instrument with a 250-µL injection loop is 0.1 10 –9 L (0.4ppm).Methane was analysed on a Varian Star 3400CX gas chromatograph using a flameionization detector (FID) with an oven temperature of 65°C and a detector temperatureof 200°C. The methane gas was separated using a Porapak-Q column (2 m 1/8 inchdiameter) and analysed on the FID with nitrogen as the carrier gas. This configurationused a 156 1/16-inch stainless steel HayeSep and a FID detector.High concentrations of methane, above 1%, require very small injection volumes,with nitrogen as the carrier gas, on the FID. The use of a small injection volume

154increases the uncertainty of the results. Therefore, the sensitivity of the analysis wasreduced as required by analysing methane with helium as the carrier gas and using theTCD. The results obtained using an FID were compared with those obtained using aTCD and reported when they agreed.The detection limit of the instrument with a 250-µL injection loop is 0.1 10 –9 L (0.4ppm).Ethane, ethane + ethylene were analysed on a Varian Star 3400CX gas chromatographusing a flame ionization detector (FID) with an oven temperature of 65°C and a detectortemperature of 200°C. The ethane, ethaneand ethylene, gases were separated using aPorapak-Q column (2 m 1/8 inch diameter) and analysed on the FID with nitrogen asthe carrier gas. This configuration used a 156 1/16-inch stainless steel HayeSep and aFID detector.Ethene and ethylene cannot be separated using the present configuration (i.e., aPorapack-Q column).The detection limit of the instrument with a 250-µL injection loop is 0.1 10 –9 L (0.4ppm).Reproducibility testsThree pressure samplers were attached on 2005-11-24 to one groundwater circulation atborehole KJ0052F01 at the MICROBE laboratory, at a depth of 450 m at the ÄspöHRL. They were left overnight at a flow rate of 30 mL/min and detached in the morningof 2005-11-25. The samplers were transported to the laboratory in Göteborg andextracted on 2005-12-13. The extraction data are shown in Table A-12.The volume of water obtained was a function of the pressure in the gas compartment ofthe pressure vessel. The variability was 2%. The variability of the volume of gasextracted, reduced by the water volume variability, was 6%. This should be thevariability of the extraction, but as the variability of the pressure vessel was unknown,this number simply represents a maximum value. The volumes extracted and analysedvaried by approximately 6% as well. The air contamination was small, less than 0.2 mLper extraction, the lowest amount of contamination being 0.053 mL.

155Table A-12. Measured and calculated variables for three pressure sampler replicatesattached to a groundwater circulation at the MICROBE laboratory.Measured/calculated variable KJ52F01-1 KJ52F01-2 KJ52F01-3 Average (±SD%)Volume of water 176 168 170 171 (± 2%)Volume of extracted gas 10.4 9.0 10.8 10.1 (± 8%)1. Volume of extracted gas /L 61.2 53.6 61.4 58.7 (± 6.2%)2. Volume of analysed gaswith air /L3. Volume of analysed gaswithout air /L60.1 52.4 60.8 57.8 (± 6.6%)59.4 52.1 59.7 57.1 (± 6.2%)4. Air contamination, % 1.13 0.59 1.74 1.15 (± 41%)Volume of air in the extractedgas, L118 53 188 120 (± 46%)The reproducibility of repeated injections from the sample vials is shown in Table A-13.This variability ranges from 0 to 3.8%. If the extraction uncertainty is 6% and theinjection precision is a maximum of 4%, then we have 10% uncertainty in the analysisprocedure.Table A-13. Repeated injections into the gas chromatograph, a and b, for analysis ofcarbon gases.Gas KJ52F01-1 a KJ52F01-1 b Average (± SD%)Carbon dioxide 63.1 64.1 63.6 (± 0.8%)Methane 333 317 325 (± 2.5%)Ethane 0.35 0.37 0.36 (± 1.4%)KJ52F01-2 a KJ52F01-2 b Average (± SD%)Carbon dioxide 31.5 33 32.3 (± 2.3%)Methane 263 284 274 (± 3.8%)Ethane 0.12 0.120 0.12 (± 0%)KJ52F01-3 a KJ52F01-3 b Average (± SD%)Carbon dioxide 57.5 59.3 58.4 (± 1.5%)Methane 379 407 393 (± 3.5%)Ethane 0.18 0.19 0.19 (± 2.6%)The analysis results are shown in Table A-14. In general, the table shows decreasingvariability with increasing amounts of gas analysed. The obtained variability can haveseveral explanations. First, the variability may of course be a result of analytical errors.

156Second, variability in the status of the pressure containers may influence the variabilityof the gas data. Third, it was assumed that three pressure samplers in series wouldcollect identical gas concentrations if those concentrations remained stable over time inthe flowing groundwater. This assumption has not yet been demonstrated. On thecontrary, multiple analyses from the MICROBE site suggest an inherent variability indissolved gas concentrations in the MICROBE groundwater (Pedersen 2005a). It mayactually be that gas concentrations vary from volume to volume of groundwater in anaquifer.Table A-14. Measured gas components for three pressure samplers attached to agroundwater circulation at the MICROBE laboratory. The data refer to the ppm of eachgas in the extracted gas (not in the groundwater).GasKJ52F01-1(ppm)KJ52F01-2(ppm)KJ52F01-3(ppm)Average (±SD%)Hydrogen 87.2 18.2 24.3 43.2 (± 72%)Helium69,000 79,100 95,00081030 (±13.2%)Argon 2170 2150 5190 3170 (± 45%)Nitrogen894,000 885,000 865,000881300 (±1.4%)Carbon monoxide 21.5 15.9 35.3 24.2 (± 33.7%)Carbon dioxide 1030 587 937 851 (± 22.4%)Methane 5450 4910 6170 5510 (± 9.4%)Ethane 5.680 2.260 2.970 3.637 (± 35%)Ethene + Ethylene

1 (1)LIST OF REPORTSPOSIVA-REPORTS 2008POSIVA 2008-01 KBS-3H design Description 2007Jorma Autio, Pekka Anttila, Lennart Börgesson, Torbjörn Sandén,Paul-Erik Rönnqvist, Erik Johansson, Annika Hagros, MagnusEriksson, Bo Halvarsson, Jarno Berghäll, Raimo Kotola, IlpoParkkinenISBN 978-951-652-160-5POSIVA 2008-02 Microbiology of Olkiluoto Groundwater, 2004–2006Karsten Pedersen, Microbial Analytics Sweden ABISBN 978-951-652-161-2: