HERBERT HEN KEL NGU -BU LL 439, 2002 - PA GE 45

The Bjorko geothermal energy project

HER BE RT HENKEL

Henkel, H.2002:The Bjorko geothermal energy project. Norg es geologiske undersekelse Bulletin 439, 45-50.

Impact crater formation results in changes in the physical properties of rocksth at can be geop hysically determined and th erefore are of importance for the mapp ing of imp act structuresat depth .Especially electrical prop erti es have a direct bearing on th e remaining porosity found in impact structu res in crystalline target rocks. Several studies made in the Dellen, 5iljan and Lockne craters show typical changes in rock physical prop ert ies. Although a signifi­ cantly increased porosity in the crater basement seems to occur, th e hydraul ic conductivity appears to be rath er small.The total volume of brecciated rocks in impact struct ures may be very large and that volume could , after reac­ tiv ation of th e fractur es,be used asa heat exchange structure for geoth ermal energy retrieval,exploiting th e norm al geothermal gradient at greater depths. The Bjo rko str uctu re is a c. 10 km-diameter impact crater located ju st west of Stockholm in lake Malaren. Studies made so far on islands show the characteristic features of intense brecciation and increased porosity known from other craters.The Bjorko Energy Project is designed to assess th e potential for geoth ermal energy retrieval by map­ ping the struct ure at depth with geop hysical met hods and by drilling. The project is financed by th e Swedish Nation al Energy Admi nistration and runs for 2 years,starting in Octo ber 2000.

Herbert Henkel, Dept. of Land and Water Resources Engineering , Royal Institute of Techn ology, SE-lOO 44 Stockholm, Sweden; Em ail: [email protected].

Introduction 70 " @ impac t c r a t e r Impact craters are caused by bodies in th e solar system pro ba b le . possib le impa ct c r a t e r :45:." la rg e circular s t r u cture whose orbits have been disturb ed, either by the gravity of impa c t bee ccta s ite Jupiter, namely asteroids, or by the gravity of stellar encoun ­ + ve ry s m a ll c ra t e r o 500 ters, namely comets .When their new orbits crossthat of the ----::-=-~- km Earth, they may eventually collide with our planet, resulting in a catastrophic explosion where th e kinetic energy of th e projectile is transferred to the Ea rth asth ermal and mechan­ ical energy.The cratering event transformsterre strial materi­ als into various shock-metamorphic states.Theoretically, it is envisaged that the shock-induced transformation s occur in a hemisph erical region around th e point of impact,with out­ ward decreasing intensity. The shock-induced transforma­ tion s of the target material th us grade from th e formation of vapour th rough melti ng to brecciation. An extensive account of impact cratering has been presented by Melosh (1989). Large impacts result in impa ct craters, which in th e H . H e n ke l & L.J . Pe s o n e n 1992 2 5 " ?on" 55 " active geological system at the Earth 's surface are soon Fig.1 . Impact struct uresin Fennoscandia (from Henkel & Pesonen 1992). eroded or buried beneath sediments.Craters larger th an c. 4 Since 1992, a few more impact craters have been confirmed in Finland. km in diameter evolve as comp lex str uctures with an ele­ vated central rise.The transition to thi s complex fin al crater structure involves massive radial flow of material, which will add to the shock-induced fractu ring of th e target material.In Fracturing related to impact craters in Fennoscandia, some 20 large (Cretaceous and old er) and 4 crystalline environments small (postglacial) impact craters are presently know n. An The concept of fractu ring has not previously attracted even larger number of such structure s is suspected to occur impact researchersas much asthe shock-metamo rphi c tran­ but remain to be confirmed as impact craters. In th e map sitions proper,the crater formation process or the comp osi­ (Fig. 1).from Henkel & Pesonen (1 992).an overview is given tion of imp act-gen erated rock typ es. Thus, th ere are so far of the situation 10 years ago. onlya few stu diesdealing with this subject.The deep drilling This vast fracturing in connection with a large impact at th e Puchezh Katunki structure (Masaitis et al. 1994) is one crater in a crystallin e environment is the research target for example - where impact-induced fracturing is observed to theBjorko energy project. considerable depth along th e enti re c. 5 km depth of th e NG U -BU LL 43 9 , 2002 - PAGE 46 HERBERT HENKEL

drillhole in the central rise. The deep drilling at th e Siljan str uct ure has not explic itly been devoted to mapping t he extent of fractur ing, but shock-induc ed planar deformation features in quartz are reported to occur down to 2.8 km depth (Juhlin 1991). Fracturing in crystalline rocks normally results in increased poro sity, and th e additi onal pore water con tent 20 decreases the elect ric resistivity of th e material. It would t herefore be possib le to map t he extent of impact-induced 10 fracturing w it h geophysical methods th at rely on electr ic conductivity,a concept that has a parallel application in frac­ -;-1 ---,---,....,....,...,.,,.,..-.,....,...,.-:-:-:-m---:-'~~~~,;----!--..!...,-,-...,.,..,.,, o , """I •. '."j ""1 tu re zone mapping.The relationship between fracture zones 10· 10' 10' 10' and changes in physic al propert ies has been explo red in r e si s tivit y, ohm m several con texts, for example in conn ect ion wi t h radwaste storage. It is a well established fact t hat the magnetisation Fig.3 . The resistivity distribution of the central rise of the Bjorko struc ­ (Henkel & Guzrnan 1977) and t he elect ric resistivity both ture (after Henkel 1992). The cumulative frequency is marked cf. The broken line indicates the approxi mate lower resistivity limit of normal show a decrease in fract ure zones (Henkel 1988). crystalline rocks. This knowl edge has been app lied to impact crater st ruc­ tures in four ways:The in-situ measurement of electr ic resis­ tivi ty using VLF-R techn ique; th e study of fracture freq uency tivity attains a minimum over the ring of the crater and a rel­ using airborne VLFdata; a study of t he relationsh ip between ative maximum over the 9 km-diameter region wi th mag­ in-sit u elect ric resisti vity and porosity; and finally a study of netic impa ct melt. The resistivity traverse is shown in Fig. 2 t he relationship bet ween in-sit u resistivity and fracture fre­ (from Henkel 1992). qu ency on a detailed scale. The central rise of the then suspected Bjorko impact In the Del/en impact structure, a traverse wi th in-situ structure was also studied with respect to the distrib ut ion of electric resistivity measurements with the VLF-R technique electri c resistivity. It was found th at the resistivity systemati­ was performed in conn ect ion wi t h regula r mapp ing activi­ cally is below 10' ohm m, wh ich is non -typical for normal ties in the area.The im pact occurred in a relat ively homoge­ low-porosity crystalline rocks such as granite (Fig. 3, from neous gne issic granite (Ljusdalgranite) wit h num erou s out­ Henkel 1992). crops on w hich th e measurement array was locat ed. It was In the cent ral uplift of the Siljan impact structure, th e found that th e electric resistivit y begins to decrease already spacing of a one-directional set of linear airborne VLFanom­ outs ide the present erosion al (topographic) crater edge. alies was studied and compared wi t h the spacing of sim ilar Towards t he crate r interior, outcrops becom e scarcer and exterior anomalies. Such anomal ies have previously been eventually the re is on ly a glacial till cover.The elect ric resis- found to be caused by water-bearing fracture zones. In th e

DELLEN DELLEN TE TE

SILJAN CENTRAL UPLIFT

ohm m ohm m 104t:t:tJ:ttl~tt~~~lltttttii1]~t:fttttjjjjj±ttt±±±:r10 4

DELLEN CENTRAL MELT SHEET

TE I I I TE I SILJAN 20 km o 20 km SILJAN

Fig.2. Resistivity traverses acrossth e Dellen and Siljan structures,for comparison (from HenkeI1 992l.The Dellen structu re isc.20 km in diameter, has a flat floor and contains impactites (tagamite and suevite).The Siljan structure isc.75 km in diameter,and is com plex with a ring-shaped crystalline uplift which is c.20 km in diameter.RB denotes the location of the Palaeozoic ring basin of the Siljan structu re.TEmarks the present topographic (erosional) edge of the str uctures. HERBERT HEN KEL NGU -BU LL 439, 2002 - PAG E 47

Fig. 4. Correlation between surface frac­ tu re frequency and in-situ surface electric 6 10000 --:------.....-., resistivity (from Backstrorn 2001). The -­""

Similarly, impact crater structures result in gravity, mag­ ular orientation of th eir foliati on.An ENE-WSW- striking dyke netic and electromagnetic anomali es, which can be mod­ with magnetic signature can be seen cutting through th e elled once the physical property cont rasts of the rocks have northern part of th e Bjorko stru cture . It is offset in several been constrained by measurements . segments within th e structure, and provi des a useful marker feature for structural modelling. Inspection of thin-sections The Bjorko structure from the cent ral rise region reveal numerous imprints that Bjorko is located in the eastern part of Lake Malaren, and its are typical for impact-ex posed rocks, like sets of planar anomalous character is related to the occurrence of non­ deformation features in quartz, kink banding in biotite and metamorphosed sandstone of supposed Jotnian age feldspars,and a general fragmented appearance on a crystal (Gorbatschev & Kint 1961).In Hoden et al. (1993), the struc­ scale. Sim ilar featur es were found in sediment c1 asts over­ tur e was suggested to be caused by a meteorite impact.The lying the fractured basement. These have th erefore been structure is ca 10 km in diameter, with a centrally located hill interpreted as re-deposited ejecta. c. 2 km in diameter. Ba sed on seismic mapping, the sand­ The entire cent ral part of the stru cture is intensely frac­ stone was found to encircle the central rise in a ca 1700 sec­ tured on the cm scale. This brecciation is also observed at tor to th e southeast of it, with an estima ted thickness vary­ several localitie s near the edge of the structure. In th e cen­ ing bet ween 40 and 280 m based on refraction seismic inter­ tral rise, VLF-R data indicate abundant low resistivity down pretations. Drilling on th e island of Midsommar, however, to a depth of a few hundred metre s.VESmeasuremen ts indi­ revealed a sandstone thickness of over 900 m. The erosion cate similar prope rt ies down to ca 600 m depth, and MT level of the structure is unknown and its age is estimated to measurements indicate decreased electri c resitivity down to about 1.2 Ga based on K-Ar data from a c1 aystone bed a depth of c.7 km. imm ediately above the granitic basement (Floden et al. 1993). The stru cture is also tilted down to th e southe ast. Geothermal energy and the energy Most of it is covered with water except for the cent ral por­ situation in Sweden tion, some islands within th e ring,and parts of the edge.The On a small scale, geothermal energy is frequently used for nature of the target at th e time of impact is unknown ­ heating of single family houses in many places in Sweden. some sandstone may already have existed.To the south , the On a medium scale,only one production facility isoperating. crystalline basement is dominated by gneissic rocks (parag­ This is located in Lund in southern Sweden and is based on neisses and gneissose granites-tonalites) with dist inct E-W th e explo itation of warm water at moderate dept h.The heat strike and steep dip.To th e north,th ese rocksare intr uded by exchange volum e is a sandstone formati on at c. 1 km depth. younger granites (Stockholmgranite) and have a more irreg- The geothe rmal gradient in crystalline rocks is c. 15 K km ',

PO ROSITY ------I N C RE A S IN G PO R 0 STY ------;> ca 0. 5 % , - - ...... / ' I ,/ , I' ,- ..... I l '\ ( ' / - -: --- I ./

FRACTURE ZONE IMPACT BR EC CIATI ON NO RMAL CRYST ALL IN E + ROC KS FRACTURE ZO E

THERMAL GRADIENT ------IN CR EAS ING THERMAL GRADI ENT ------~ ca 15 K km-I ~~ ~~~3~'"3== \ \ ~ \ , ' f ,/ ,'- '/ I ...... '. -:. .... / I ,/ ...... ,> ., ./ - t > \ / / , '/ - -, '\---- I /' I; -/ i ~t;fJ ,,- \. "

SEDIME NTARYCOV ER RADIOACTIVE ROCKS SEDI MENTARY COV ER + RADIOAC TIVE ROC KS

Fig. 5. Cartoon show ing th e imp ortant features representing th e potenti al for geothermal energy in crystalline rocks.The geothe rmal gradient across sedimentary cover rocks depends on their th ermal condu ctivity. The cited norm al geothermal gradient of 15 K krn-l was obtained from th e deep drilli ng at 5iljan (Juhlin 1991l. HERBERT HEN KEL NG U-B U LL 439, 2002 - PAG E 49

know n from th e deep drill ing at Siljan.There are no anom ­ alous thermal st ruct ures known in th e Balt ic Shield, and therefore geothermal energy has for long been disregarded as an important energy resource. The climatic sit uation in Sweden causes a large demand for energy for heating dur­ ing more th an 6 mo nths of th e year.This demand has been covered by elect ric energy, oil and gas combu stion, and to a lesser extent by the burning of waste and biomass. ln many places, a large-scale energy infrastructure wit h district heat­ ing has been establis hed. Presently, the poli tical decisions to reduce and phase out nuclear power and to reduce CO, emissions are not viable as no sufficiently large-scale alterna tives have been develop ed. Geothermal low-temperat ure energy, however, is present everyw here and wo uld be an obvious resource if an efficient heat exchange st ructure coul d be found or created at depth in the crystalline basement.It should also be located close to Sodertalje an exist ing energy infrast ruct ure. The extraction of geothermal energy in ordi nary crys­ Fi9.6. Location of the Bjorko st ructure in relation to the existing energy talline rocks has been tested at Fjall backa (Wallrot h et al. infrastruct ure in th e Stockhol m region. The approximate edge of the Bjorko str uct ure (circle) and its c. 2 km-diameter central uplift are 1999), and included hydraulic fracturing in order to create a marked.The three small rings NW,Sand SW of Stockholm mark th e loca­ heat exchange volume at depth . The exchange efficiency, tions of major district heati ng plant s. The study area is west of however, could not be broug ht to accept able levels and th e Stockholm within the coordi nate frame with markers every 5 km. project was not develop ed fu rth er.It th erefore seems neces­ sary to fi nd already brecciated rock volumes at dept h and to brecciated volume may exceed 250 km'.Assuming a 1% dif­ explore or increase th eir hydraulic conduct ivity. In essence, ference in porosity in a hemisphe rical volume wit h 5 km two such types of structure can be envisaged: Fracture radiu s and a temp eratur e difference of 40° C, th e energy zones, with an essent ially planar extent of fractu red rock, or conten t of th e struct ure is over 4 000 TWh, mo re th an 10 impact craters, with an assumed, essenti ally hemispherical times th e annual energy use in Sweden. The structu re also exte nt ofth e fractured rock.Geoth ermal energy pro specting lies wit hin th e realm s of th e distri ct heating systems of t he should th erefore aim at mapping th e extent of brecciation Sto ckho lm region (Fig. 6) and its energy potent ial could of such st ruct ures and to estimate th eir potent ial as heat cover 70 % of the energy demand for heating on a long­ exchange volum es.In addition to t he fracturing,which even­ te rm basis. tu ally may provid e a sufficient ly efficient heat exchange vol­ In th e fo llowing, brief descript ions are given of th e ongo­ ume, the actual geothermal gradient, the radiogenic heat ing activiti es. prod uction, and t he effect of a shieldin g cover rock Compilation ofexisting data and com plem ent ary mea­ sequence are impor tant factors for th e evaluation of th e surements are performed within a 20 km-diamete r regio n potential for geothermal energy in crystall ine environments around th e Bjorko st ruct ure. These include: airborne geo­ (Fig.s). physics (magnet ic total inte nsity,VLF, and gamm a radiation), data abo ut th e terrain shape (elevation and bathymetric The Bjorko energy project data, to form a 50 grid ), gravity measurements and petro­ With t his background, th e idea was formulated to test th e physical measurement s on rock samples from outcrops and Bjorko st ruct ure for its potenti al as a heat exchange volume drillcores (density, magnetic susceptibility, remanent mag­ for geothermal energy retri eval. An applicati on was for­ netizat ion, elect ric cond uct ivity ). warded to th e Swedish Natio nal Energy Adm inistration by a Measurement s of new data, especially on th e elect rical research team from th e Royal Inst itut e of Techn ology and prop erti es of rocks and rock volumes.These inclu de: VLF-R th e Stockholm University. The project explicit ly aims at measurement s to est imate th e near-surface electric resistiv­ developing too ls fo r energy pro specti ng by int egrated ity variation, MT measuremen ts for mapping of t he electric analysis of relevant surface data and th eir calibration by resistivity at depth, fractu re frequency st udies in type drilling.The most importa nt aspect to start wit h is to find t he regio ns (cent ral part, ring - edge and exterior) to provid e a local relationship between elect rical properti es and porosity basis for th e correlation between elect rical and fractu re do wn to several kilo met res depth. propert ies. The geot hermal energy potential of th e 10 km-diameter Drilling of 2 or 3 drillholes down to markers ob served in Bjorko str ucture is dramatic.The porosity variation as esti ­ MT measurement s and accompanying tests.These includ e: mated from electric resistivity data is 0.5 - 4.5 % and th e chemica l analysis of t he water in drillholes, estimation of th e N GU -BU LL 439, 2 002 - PA GE 50 HERBERT H EN KEL

eroded shield environm ents, with special emphasis on electr ic hydraulic conductivity around drill holes, measurement of resistivity. Tectonophysics 216, 63-89. the local geothermal gradient and the components of th e Henkel, H. & Guzrnan, M. 1977: Magnetic features of fracture zones. stressfield. Geoexpl ora tion 1S,173-181. Henkel, H.& Pesone n, L.J . 1992:Impac t craters and craterform str uct ures Modelfing of the collected data to provide information in Fennoscandia. Tectonophysics 216, 1-19. on the thermal and hydraulic potential of the structure.This Juhlin, C. 1991: Scient ific summary report of the Deep Gas Drilling includes: esti mation of th e radiogenic heat production, Project in the Siljan ring impact structure.Vattenfall Utveckl ings AB, modelling of gravity, magnetic VES, MT and fracture fre­ Repo rt u(G) 1991/14.Alvkorleby, Sweden, 257 pp. Masaitis, V.L., Mashcak, M.5., Naumov, M.V. & Raikhlin, A.1. 1994: Large quency data. astroblems of Russia,VSEGEI, St.Petersburg, 32 pp. The expected results of this energy prospecting proj ect Melosh, H.J. 1989: Imp act Crateri ng. A Geolo gic Process. Oxford are to describe the Bjorko structure in 3-d, estimate its University Press, 245 pp. potenti al for geothermal energy retr ieval, identify possible Wallro th,T., Eliasson,T.& Sundquist,U. 1999:Hot dr y rock research exper­ iments at Fjallbacka, Sweden. Geothermics 28, 617-625. relations hips between surface data and hydraulic properties at depth, and suggest further approaches to verify th e obtained mod elling results.

Conclusions Geothermal energy resources must be a target for renewed research,even in regions with low geoth ermal gradient .This is moti vated by two factors: the need for heating at high lat­ itudes and the local nature of this energy resource (apart from th e political goals to reduce both CO, emissions and th e use of nuclear energy). Meteorite impact structures are suggested to be suitable target s for geothermal energy studi es asthey may prov ide a very large volume of fractu red rock in comp arison to frac­ ture zones. The physical property that is most likely a key factor in mapping the extent of fracturing at depth is the electric resistivity, and therefore efforts are directed towards estab ­ lishing and mode lling th e relation s between th e volume fracture frequency and the electrical properties that are measurable from the surface. The physical cond iti on that is necessary in order to establish a heat exchang e system at depth is the hydraulic condu ctivity that can be created and mainta ined with artifi­ cial methods. How large volumes of fract ured rock react involves issues th at dem and further experiments and research.

Acknowledgements I wo uld like to thank Lars Kirkhusmo and Knut Ellingsen for th eir helpful and con structive reviews of th e manuscrip t.

References Backstrorn. A. 2001: Electr ic resistivi ty, magnetic suscept ibil ity and frac­ ture frequ ency of im pactites and basement rocks of th e Lockne crater. M.5c. Thesis, Department of Geology and Geochemistry, University of Stockh olm , 55 pp. Hode n,T., Sbderberg, P.&Wickman, F-E. 1993: Bjorko - a possible Middle Proterozoic im pact structure west of Stockho lm. Geologiska F6reningen i Stockholm F6rh andlin gar 11S-I , 25-38. Gorbatschev, R.& Kint , O. 1961: The Jotnian Malar sandstone of th e Stockho lm region, Sweden. Bulletin of the Geological Institutions of the University of Uppsala 40, 51-68 . Henkel ,H. 1988:Tecton ic studies in the l.ansjarv region .Swedish Nuclear Fuel and Waste Manageme nt Co, Technical Report 88-07, 66 pp. Henkel, H. 1992: Geophysical aspects of meteorite impact craters in