temperature and pressure in the svartsengi ... - Orkustofnun

TEMPERATURE ANDPRESSURE INTHE SVARTSENGI GEOTHERMAL RESERVOIRGabriel Gebreigzlabhier.*UNU Geothermal Training ProgrammeNational Energy AuthorityGrensasvegur 9, 108 ReykjavikICELAND·Permanent address:Geothermal Energy Exploration ProjectGeological Survey of EthiopiaMinistry of Mines and EnergyP.O.Box 486, Addis AbabaETHIOPIA

3ABSTRACTThiS report ls a study of the distribution of an d changesin temperature and pressure in the high temperaturegeothermal field of Svartsengi. Use ls made of t h e variousmeasurements of the 12 individual wells at differentperiods and the cross-sections made from thes e measurementa. Consideration is given to the effect of drilling,redrl111ng and reinjection on the t emperature profiles ofthe wells and the cross - sect 10ns and appropriate selectionsare made of reliable and repr esentative profiles ofthe many measurements. Bol11ng conditions of th e reservoirand evolution with time is studied. Some possibilitiesthat can cause the reservoir to be coo11ng down arepresented and discussed. They are boiling effect s, lateralcold inflow, episoidal cold inflow and decr e ased lateralinflow of hot fl u ids.This study shows that there is a sUbstantial cooling andpressure drawdown due to exploitation deep in thereservoir. It is found that temperature of the fluid isfairly stable during exploitation. However, in the timebetween 1982 and 1983 significant t emperature changes areobserved In the reservoir. Th e reason for the cooli ng ofthe reservoir is found to be most li kely a recharge of coldinflow.

5LIST OF CONTENTSPag eABSTRACT .... . ...• . .......•....•...•..•..• .. . .. . .. ..... .. .. 3INTRODU CTION1.1 Scope of the report1.2 Aims and objectives1 11 11 . 3 Geographical setting of Svartsengl . ..•...... ........ 121.4 Data used in the report 131.5 Limits to the study................................ 132 GENERAL DESCRIPTION OF THE FIELD2.1 Geological setting 1 42.2 Chemical studies .. •.. •••••••••••..• .. ••.•..•..•. .•. • 142.3 Resistivity studies of the field.................... 172.4 Tectonic activity of the Reykjanes peninsula........ 172.5 Hydrological characteristics of the geothermal system 172.6 Proposed reservoir model by Kjaran et al .••....... ,. 192.7 General temperature conditions .. ......•....•........ 222.8 The power plant..................................... 222.9 History of development •.•••.••.. ..•... .....•.....••. 243 INSTRUMENTATION3 . 13.23.3Pressure elementTemperature element .•..••.•..•..•...................Limits of accuracy .....•............................2629294 INDIVIDUAL WELLS ...........•........•..•..••.•...•..•... 305 SPATIAL DISTRIBUTION OF TEMPERATURE AND PRESSURE5 • 1 Vertical distribution· . . . . . · . .. . ... 485 . 1 . 1 Cross-section A· · · .. ·485 • 1 . 2 Cross- section B ·. .. . . . · . . ·. 535 • 1 .3 Cross - section C· . · .. . 535 • 1 .4 Cross-section D . . . . ·. . . . · 575 . 1 .5 Cross- section E· . . · . ·. . . 57

65.25.35.1.6 Cross - sectIon F ... . ........................... 63Horizontal distribution5 • 2. 1 DistributIon at a5.2.2 Distribution at a5.2.3 Distribution at aCooling between 1982 and........... ..................depth of 700 m ...... ........depth of 1000 m .............depth of 1 1 00 m .............1983 In the reservoir ......63666666726 EVOLUTION WITH TIME............... .. . .. ................. 7~7 CAUSES 0, TEMPERATURE CHANGES7.1 Boiling 777.2 Lateral cold inflow.......... . ...................... 787.3 Episodic cold inflow................................ 787.4 Decreased lateral inflow of hot fluid ............ , .. 797.5 Calibration of instruments ........ ............... ... 797.6 Discussion ..... ... ... ..... ........................ .. 808 CONCLUSIONS............................................. 81ACKNOWLEDGEMENTS .•...•..•••.•••••••••••••••••••••••..•.••. 82RHERENCES •.••.••••••..• . ....•••.•...••..•.•••..•..•.....• 83LIST OF '1GURESFig. 1 Western Reykjanes peninsula and location of Svartsengihigh temperature field ...... ... ....... .. ...... 12Fig. 2 Location of wells. faults and dykes in the Svarts -engi production field.............................. 15Fig. 3 A simplified geological cross - section of the produc ­tion field in Svartsengi. Line of cross - sectionshown in Fig. 2 •...•..•..•..•.•••.••••••••••••••••• 16Fig . 4 Resistivity at 600 m depth below sea level at thewestern part of the Reykjanes peninsula ......... ... 18

7Fig. 5 Conceptual flow model from the Reykjanes peninsula.. 20Fig. 6 History of exploitation of the Svartsengi reservoirfrom 1977 to 1982 • .. .•.••..••.•.....•..•.•••.•••... 21Fig. 7 Simplified temperature profile of the Svartsengigeothermal field.......... .. ....................... 23Fig. 8 Amerada recording gauges for pressure andtemperature........................... .. ........... 27Fig. 9 Cross-sections of Amerada pressure and temperatureprobes 28Fig.lO Svartsengi well SG - 1. Temperature profile in 1976.. 31Fig.lt Svartsengl well SG - l, Pressure profile in 1976 . .... 31Flg.12 Svartsengi well SG - 2. Temperature in 1976 and 1981with corresponding saturation curves 32Fig.13 Sva r tsengi well SG - 2. Pressure profiles in 1976and 1981 ......•..•............•.................. . . 32Fig.14 Svartsengi well SG - 3. Temperature profiles in 1976.1981. 1983 and corresponding saturation temperatureprofiles............................ . .. . ........... 34Fig.15 Svartsengi well SG - 3. Pressure profiles in 1977.1981 and 1983 ....... . ............. . ............ . .. . 3~Fig.16 Svartsengi well SG - 4. Temperature profiles in 1976and 1982 ..••.....•..••..•........•.••.••.•••.•..•.. 35Fig.17 Svartsengl well SG - 4. Pressure profiles in 1982 .... 35Fig.18 Svartsengi well SG - 5 . Temperatu r e profile in 1976and correspond i ng saturation profile .•.... • .. ... ... 36

8Flg.19 Svartsengi well SG - 5. Temperature profile 1n 1982and corresponding saturation profile............... 36Fig.20 Svartsengi well Sa-5. Temperature profile in 1983and corresponding saturation profile ............ . . . 37Flg.21 Svartsengl well SG - 5. Pressure profiles 1n 1976,1982 and 1983 . • ...... . .....•..•.................... 37Fig.22 Svartsengi well sa - 6. Temperature profiles 1n 1981.1982, 1983 39Fig.23 Svartsengi well SG-6. Pressure profiles 1n 1980,1981, 1982 and 1983 .. • .•.•......... . • .. .... ........ 39Fig.24 Svartsengi well 5G-7. Temperature profiles in 1980,1982 and 1983 ......... . . ......... . .............. . . . 40Fig.25 Svartsengi well SG - 7. Temperature profiles 1n 1982,1983 and corresponding saturation profiles 40Fi g.26 Svartsengi well 5G - 7. Pressure profiles in 1980,1982 and 1983 ...................................... 41Flg.27 Svartsengi well SG - B. Temperature profiles in 1982and 1983 with corresponding saturation profiles .... 41Fig.28 Svartsengi well SG-8 . Pressure profiles in 1982and 1983 ...•......•....•............•..•......•.•.. 42Fig.29 Svartsengi well SG-9 . Temperature profiles in 1982,and 1983 with corresponding saturation curves 42Fig.30 Svartsengi well SG - 9. Pr essure profiles in 1980,1982 and 1983 .. .. . .. .. .. . .... .. .. . . . ... . .. . .. . ..... 43Fig.31 Svartsengi well $G - 10. Temperature profile in 1982and corresponding saturation profile............... 43

9Fig.32 Svartsengi well SG-l0. Pressure profiles in 1980and 1983 ... .. ... . ........... ... . ................... 44Fig.33 Svartsengi well SO - 11. Temperature profiles in 1982and 1983 with corresponding saturation profiles .... 46Fig.34 Svartsengi well SG-1'. Pressure profiles in 1982 and1983 •••••••.•...••••••••••••...••....••.••••••••••• 46Flg .35 Svartsengi well SG-12. Temperature profiles In 1982and 1983 with corresponding saturation profiles •... 47Fig.36 Svartsengi well 50 - 12. Pressure profiles in 1982and 1983 ........................................... 47Flg.37 Location of cross - sections A, B, C, D, E, and F •.•• 1j9Flg.38 Cross - section A. Initial temperature distribution.. 50Flg.39 Cross-section A. Temperature distribution in 1982.. 51Fig.40 Cross - section A. Temperature distribution In 1983.. 52Fig.41 Cross - section B. Temperature distribution in 1982.. 54Fig.42 Cross-section B. Temperature distribution in 1983.. 55Fig.43 Cross - section C. Initial temperature distribution.. 56Fig.44 Cross-section D. Temperature distribution in 1982.. 58Fig.45 Cross-section D. Temperature distribution in 1983.. 59Fig.46 Cross - section E. Te.mperature distribution in 1982 .. 60Fig.47 Cross-section E. Temperature distribution in 1983.. 61Fig.48 Cross-section E. Cooling between 1982 and 1983 ..•.• 62

10Fig.~9 Cross - sectIon F. Temperature distribution in 1982.. 64Fig.50 Cross-section F. Temperature distribution in 1983 65Fig.51 Temperature distribution at 700 m depth in 1978 .... 67Fig.52 Temperature distribut i on at 700 m depth in 1982 ..•• 67Fig . 53 Temperature distribution at 700 m depth in 1983 68Fig.54 Temperature distribution at 1000 m depth in 1978 , . , 68Fig.55 Temperature distribution at 1000 m depth in 1982 69Fig.56 Temperature distribution at 1000 rn depth in 1983 ... 69Fig . 57 Temperature distribution at "00 m depth In 1978 10Fig.58 Temperature distribution at 1100 m depth in 1982 ... 70Fig.59 Temperature distribution at 1100 m depth In 1983 71Fig.60 Cooling between 1982 and 1983 at 100 m depth 71Fig.61 Cooling between 1982 and 1983 at 800 m depth .. . .. . . 12Fig.62 Cooling between 1982 and 1983 at 1000 m depth ..•.. . 13Fig.63 Cooling between 1982 and 1983 at 1100 m depth •..... 13Fig.64 Relationship between pressure drawdown and totalmass withdrawn from the Svartsengi reservoir .•..•.. 15

111 INTRODUCTION1.1 Scope of the reportThe author fulfIlled one of the requirements of the UNUGeothermal Training Programme Fellows by producing thiswork during the last ten weeks of the six month trainingduring the summer of 19B3 at the National Energy Authorityof Iceland, which was financed both by the United NationsUniversity and the Government of Iceland.The Training Programme commenced with a six weekintroductory lecture course on the relevant aspects ofgeothermics delivered by specialists from the N.E.A. andthe University of Iceland and complemented by appropriatefield excursions. The author further had specializedlectures, seminars, tutorials, field practices andsupervised readings on borehole geophysics and reservoirengineering for six weeks, and additional lectures andseminars during the two week field excursions to geothermalfields of Iceland.The author found it viable to take on this project as apractical exercise because of the availability of thetools involved and in view of the immediate applicabilityof the methods in the geothermal fields of his homecountry.1.2 Aims and objectivesThe Svartsengi geotherma l reservoir has been observed to becooling down. There are various possibilities for causingthis. Identifying the real ones and locating the optimumzone for feasible exploitation is important to ensure thatthe power plant will keep on producing for a long time.This problem can be tackled by studying the temperatureand pressure in one, two, or in three dimensions. Theindividual well measurements and compilations of thesemeasurements into cross-sections (usually vertical and

12••• ••••o:/.4"m :() Known oreos of klw tniltlwity...-.. ~ ~.)'lIjonn Pnlsuto~ ""INC zone• HiQh t."'*"'ture enc.,- Main tectonic iaJ"- Eruption '''tuNS•Fig. 1 Western Reykjanes peninsula and l ocation of Svarts ­eng! high temperature field (from Franzson. 1983).horizontal planes) are used 1n the study of the behaviourof the reservoir as a function of time. Boiling conditionsas a function of time are also included .1.3 Geographical setting of SvartsenglThe Mid-Atlantic Ridge surfaces on to Iceland at thesouthwestern end of the Reykjanes peninsula where itcrosses the country in a northeasterly direction. The riftprocess on thE: Reykjd lles peninsu l a Is characterized by anENE-WSW trendlng seismic zone below 2-3 km depth, but onthe surface by NE - SW fissure swarms arranged en echeleon

1 3to the former (Georgsson. 1981; Franzsson, 1983).intersection of the two types is the Svartsengitemperature field located (Fig. 1).Atanhigh-1.4 Data used in the reportThe temperature and pressure measurements used In thisstudy were all taken by the Geothermal Division of the NEAin the period between '16.07.12 and '83.06.16. Thedifferences In the well conditions at the time of measure ­ment range from the wells having been closed for as longas one year I up to measurements made right after production.1.5 Limits to the studyThe work Is restricted to the time and spatial distributionof temperature and pressure at depths of 500 m and belowin the Svartsengl reservoir.

1 42 GENERAL DESCRIPTION OF THE FIELDLocation of wells, faults and dikes in the Svartsengl fieldIs shown in Fig. 2 . The Svartsengi high-temperature area 1nIceland is a part of the Reykjanes geothermal system. Thegeothermal wells at Svartsengi yield 235°C water in 1983which Is 5°C less than the 240°C before production began.The salinity Is roughly two - thirds that of sea water(Thorhallsson, 1979) .2.1 Geological settingThe geological succession of the field consists of lavasuccess ions wit h intervening hyaloclastlte formations(Franzsson, 1983) . Some 20 - 40 % of the succession below800 ID depth consists of intrusions . The reservoir acquifersare predominantly controlled by intrusives and sometectonic fractures/faults. The alteration pattern can besubdivided into distinct episodes where low temperatureconditions are seen to have prevailed prior to hightemperatureconditions. Two distinct episodes of h1ghtemperatureactivity are observed, where the latter onerises separately to higher elevation. The dominant alterationpattern conforms reasonably well to the presentthermal regime in the field (Franzsson, 1983 ) .2.2 Chemical studiesChemical analysis of the geothermal water indicates about67 % sea water, but deuterium measurements give only about57 % sea water (Arnason, 1976). The difference isexplained by flashing of the water in the conceptual modelof Kj aran et 01.(1979). It 1s assumed that the rechargearea for the geothermal water is located in the vicinityof lake Kleifarvatn.

u JHD - BJ~2300 HF,83.05.0702 AAI>-Y""'~ Line of section11fIrS-12 S-IO 1/It11S-2 11/ Concealed fault",/ 40 fhraw in meters1 J •• ~S-3/" Concealed dyke, J/1• S -9 aorehole-1 •, 11110 100 200 300m7, , 11J/I 1I!11I111111tNIII \t I"/" S-5 --::;I>- S-7 \\, ~f:. -- :=t"'-=---- ! 1di/f/40;8011,\ "'-- _-;:"""111IJ11""" - :::-- I'IS-8S- I /I/It /- f -J/40f / rI101I1 /"S-IIrI-; 11J//25)11'11111/1111Fig. 2 Location of wells. faults a nd dykes in the Svartsengiproduction field (from Franzson, 1983).

1 6~JHD-8 J - Z300 Hr.L'J:J 8 ~~0703 AAADepth[m)5009 11 7 8 4 5 6 12 :3 2100'-~----++---+4-~-+--~~~BI u"'"u'">-" t.J>,,'"!.J>....'" ::>"....z1000t 11 H1500\-rM ,'"~" .... -,,~;~"" ..... ,>,,'" ~:,'I;.... ''''-' /i:, ,I'-.!.--I!, ," :, " \' ,, -, -, '" ,~,

1 72.3 Resistivity studies of the fieldSurface hydrothermal manifestations are meager within theSvartsengi field and only scattered over about 4 squarekl1ometers. The resistivity surveys (Fig. 4) show thegeothermal area to be elongated NE - SW 1n the northern partbut ESE-WSW in the southern part (Georgsson, 1981), Acontinuous E- W striking low resistivity zone whichcoincides fairly closely wi th the Reykjanes seismic zonetha t has been interpreted as a plate boundary. Study ofcorrelation between resistivity salinity and temperatureseem to confirm that the low resistivity can only beexplained by anomalously high heat flow . This indicatesthat explol table geothermal energy may be found 1n theuppermost 1 - 2 km along the plate boundaries on theReykjanes Peninsula even far away from surface manifestations.2.~ TectoniC activity of the Reykjanes peninsulaOne can see the main zone of earthquakes and location ofquake epicenters on the area from Fig. 1. The tectonicactivity increases the permeability and thus creates somesort of a drainage system along the zone of earthquakes.Some geothermal areas like Svartsengi are situated wheresurface fissure swarms intersect the main zone of earthquakes.2.5 Hydrological characteristics of the geotherma l systemThe water percolates down from the infiltration area andflows along the permeable earthquake zone. On its way itwarms up and mixes with intrusive seawater. Where thesurface fissure swarms intersect the earthquake zone. thepermeability becomes great enough to allow free convection,thus forming a geothermal area.

c:J~5nmc::J5-anm5- Lines, of equcl..... resishviliso..... ........... , ....... .~::::::3 ~ a.12 mHiQh tempercture fieldoooo, ,le or"~REYKJANES PENINSULARES I STIVITY AT 600 M.A.S.LFig. 4 Resistivity at 600 m depth below sea level at thewestern part of the Reykjanes peninsula(from Georgsson et al., 1983) .

1 9Fresh water is pumped from shallow wells at a distance of 4to 5 km from the plant. There Is a fresh water lens ofonly 45 m thickness in porous surface lavas. The lens IsfloatIng on top of the seawater below. The pumping is donewith great care in order to avoid pumping "salt ll waterwhich could happen if the drawdown of the fresh watersurface in the well Is too much .2.6 Proposed reservoir model by Kjaran et al. (1979)This conceptual model for the regIonal geothermal system isa hybrid convection model for the geotherma! area itselflinking together reservoir performance and well performanceto give an estimate of the reservoir capacity.Free convection exists to some extent in most geothermalreservoirs (Ellasson, 1973). increasing considerably thenormal heat flow to the surface. The heat flux has topenetrate the caprock and it can be shown that it is almostimpossible that heat transport be only by conduction.There has to be some heat trans port by discharge of hotwater and steam to keep the convection going. If there issuch an outflow from the reservoir it must be expected tobe balanced by an equal amount of inflow . This inflow isassumed to be the infiltration at lake Kleifarvatn,flowing along the main earthquake zone, as shown in Fig. 5.The recharge and discharge flow and the convection are thepremise for the hybrid convection model shown in Fig. 5.From this figure one can wr i tethree continuity equations at eachequations are: continuity of mass,down six e quat ions:point E and D. Theenthalpy and chlorideconcentration. There are six equations and twelve unknowns.So six unknowns have to be assumed. At Svartsengi theseparameters were either measured or estimated thus enablingthe authors to solve for the remaining unknowns. Thenatural heat loss can then be calculated as the differenceof the upward and downward heat flow. The calculated valuewas about 300 MW, which was three times the necessary power

;-n va o-vv ~ 900 . S.P".KL.l:=J 82 .010037 Sy .J'" oA 1140 m osl1 SvortsengiPA~IOOO kg/m'uK A = 0 .2.10 - 5 m/sec- - -COi~-sea waTer - - - - -( d22?222?lP~'" "Ill"~ qt??cc?> \.\KB -2000 2,5U"0.3'10-'m/ secPB~ 190kp/cm 2 P B ~835 kg/m'BJ. J. X A --7 :z A' A X -Xr--CoTemp.'c500 601000 1201500 1802000 2402500 300Pressurekp/cm250.610014619022882.0 1. 003 7.Layer 3k-oFig. 5 Conceptual flow model from the Reykjanes peninsula(from Kjaran et al., 1980).

... T [IIKFII€9ISlOFoVI VII2.O

22consumption of the geothermal heat ing plant. Fig . 6 showsthe discharge of wells and drawdown in the Svartsengireservoir from 1976 to 1982.2.7 General temperature conditionsThe general temperature picture of the field is describedby Fig. 7. There is a 300 m layer of warm water systemhaving a temperature of 1I0-60 D C which can be seen frommeasurements of wells which water level Is 15-20 m belowground level. The geotherrnal system between the depths of600 m and 2000 m had a temperature of at least 240°C whenthe undisturbed water level was between 64 and 73 m before1976. In other words the warm water system overlies thegeothermal system at all locations except in the westernpart of the field (Wells SG-2, 3 and 10) where the latteris vented to the atmosphere through fumaroles. There thesubsurface temperature follows the boiling point curve downto ~OO-500 m where a temperature of about 2~OoC is reached.2.8 The power plantThe power plant at Svartsengi is the first geothermal plantof its kind in the world (T horhallsson, 1979). The thingthat makes it unique is that a high temperature brine isused as a heat source for space heating and also for thegeneration of electricity. called co-generation. thusmaking better use of the geothermal energy than isordinary for high-temperature geothermal installations.This plant was constructed in several stages. The geothermalfluid is exploited by using heat exchangers forthe space heating of all the communities in the westernReykjanes peninsula and the international airport atKeflavik (Thorhallsson. 1979). It was first built togenerate 50 MWt (megawatt thermal) and 2 MWe (megawattelectric) for a district heating system. A 75 MWt a nd 6 MWeaddition was later completed. The heat is extracted byflashing the brine in two stages to60 0 c and using the

2 3~ JHO-HSP - 2300 G9~ 83.091236 AADepth[m]250SIMPLIFIED TEMPERATURE PROFILE OFTHE SVARTSENGI GEOTHERMAL FIELDo 100 200 300 [OCJo+--.--~----~----~~~----------~Warm water systemTemperature 40-60°CWater level in wells15 -20 m below ground leve I5007501000-1250-The geothermal systembelow 600 m de plhTemperature co. 240°CUndisturbed water level01 64-73 m deplhbefore 1976150017502000Fig. 7 Simplified temperature profile of the Svartsengigeothermal field.

flash steam for heat and electricity generation. The pilotplant studies were the basis for the design of the firststages in constructing the plant Nevertheless a lot ofdetails. materials and components had to be tested beforethe main plant was designed and constructed. All the samethe power plant has been able to supply thermal energy tothe different communities according to the requirement atany time since the commencement of the programme.2.9 History of developmentThe highlights of the historical development of t he projectare as follows (Bjornsson. 1983):1971 - 1972 Wells SG-2 and SG - 3 were drilled to depths of 240and 400 m respectively. The results turned out to be veryencouraging proving there was a high temperature fieldwith a brine temperature of 235°C.1974 The Orkustofnun (NEA) built aSvartsengi. Two wells weredrilled. SG-4pilot plant at(1700 m) and SO-5(1525 m).1974-1975 Tests were carried out by the Orkustofnun (NEA)in the pilot plant which later became the basis for thedesign of power plant 1.1976 Preliminary power plant of 3 MWt was commissioned onNovember 6.1977 First unit of power plant 1 of 8 MWt was commissionedin November.1978 First 1 MWe turbogenerator was commissioned in April.Second unit of power plant 1 of 12.5 MWt was c ommissionedin November. Well SG-6 was drilled to a depth of 1734 m.1979 Third and fourth units of power plant of 2 X 12.5MWt were commissioned. The second 1 MWe turbogenerator was

25commissioned. Well SG - 7 was drilled to a depth of 1438 m.1980 Third 6 MWe t urbogenerator was commissioned. WellsSG - 8, SG-9, SG-l0, and SC - 1t were drilled to depths of1603, 994, 425 and 1'41 m respectively.1981 First unit of power plant 2 of 25 MWt was commisionedin May. Second and third units of power plant 2 of 2 X 25MWt were commissioned 1n September.1982 The tota l installed capacity in power plants 1 , 2 and3 became about 46 MWt • 2 MWe, 75 MWt and 6 MWerespectively. The last well to date, namely SG-12 wasdrilled to a depth of 1488 m. Relnjection was done for 25days in the months of September and October.

263 INSTRUMENTATIONThe Amerada mechani caI gauges were used for both tempera ­ture and pressure measurements used in this researchproj ect.There are two main types of mechanical gauges, namely theAmerada temperature gauge, which has a bourdon tube wherethe boiling pressure of a special fluid Is recorded, andthe Kuster gauge which has a bimetal sensor where thetemperature torsion of the bimetal indicates the tempera -ture. When using the mechanical gauges data Is nottransmitted to the surface, but is recorded inside theprobe on a clock driven recorder. As many as 20 - 30measuringwell.pOints can be recorded during one run 1n theThe Amerada gauges consist of three basic parts, therecording section, a clock and either a pressure sensor ora temperature sensor (Figures 8 and 9).3.1 Pressure elementThe active element in the pressure element i s a helicalbourdon tube, fixed at one end and free to rotate at theo t her. The interior of the tube is subjected or exposed tothe pressure in the well. The resulting rotation of thefree end of the bourdon tube is transmitted directly to arecording stylus without the use of gears or levers. Thestylus records on a metal chart made of thin metal coatedon one side with a special paint. The paint rendersfriction extremely low and the scribed lines are so easilyvisible that the chart scanner user can measure chartdeflections to obtain the accuracy the gauge is capable ofproducing. The chart is carried on a removable cylindricalchart holder, the position of which is controlled by aclock.

27~ J H D-HS~ - 9 0 00 Gl:iL.J::..J 83 .09. 12 42Clock I, . ,WIRELINESO CKETCOMPLETERECORDINGPRESSUREReeord1naSecti onti ,Ireadscrew:'Chart:Ho lderf,StylusAssemblyOUTERHOUS I NGGAUGEPr" ssureElementTEMPERATUREELEMENTMax.ReadingTherm.011TrapThe rm.WellBOTTOM HOLE RECORDINGPRESSURE OR TEMPERATURE GAUGERPG-3 1 1,;" Dia . RPG-4 1" Dia.~lg.8 Amerada recording gauges for pressure andtempera t ure.

;T=1 JHD·HS~·9000 GGL.:J::J 83 .09.1241OutlrHoulin, ----0Support,..St'l'IUI ___ _BULBSund.rdFnl R.spond inlTEMPERATUREELEMENTClrryin,M" dmumR.,illlrln,ThlrmomltlrRECOROING SECTION(GAUGE)PRESSUREELEMENTFig . 9 Cross-sections of Amerada pressure and t emperatureprobes.

293.2 Temperature elementThe bourdon tube in the temperature element is sealed.Variations 1n temperature create different boilingpressure inside the bourdon tube. The vapour pressure ofthe enclosed liquid is directly related to its temperature,making the rotated position of the free end of the bourdontube an usable measure of the temperature of the element.This rotation is recorded on the gauge chart as describedabove.3.3 Limits of accuracyThe repeatabl11ty of a properly maintained gauge is betterthan -0.1 % of full range of the pressure element in use,while the absolute accuracy Is 0 . 2% . Temperatures above70 0 C affect the strength of most Bourdon tubes socalibrations are necessary to maintain the accuracy of theinstrument. The sensitivity of the gauge 1s 0.2% of fullscale deflection.The absolute accuracy of the temperature gauge is usuallyassumed to be 2°C but this is related to the calibrationand operation of the instruments.

304 INDIVIDUAL WELLSThere have been drilled 12 wells In the high temperaturefield of Svartsengi (Fig. 2). The temperature changes withrespect to time and the boiling conditions are discussedIn this chapter.Well SG -l Is a shallow well with relatively low tempera ­ture. It was drilled for the purpose of using its water Indrilling the other wells. The depth of the well Is 250 mand the temperature is less than 50°C. A peak in tempera -ture Is measured at the ground water level of 25 m depth(Fig. 10). This indicates a considerable horizontal flowIn this shallow acquirer, most probably a run off from themain geothermal field (Stefansson and Stelngrimsson.1981). The maximum temperature measured In this well is47°C at the bottom at 250 m depth. The temperature at thisdepth in the nearest well (SG-5) is about 200°C, showingthe abrupt demarcation of the geothermal system at thislocation and depth. The cold ground water reaches down tolittle more than 200 m depth, but beneath that the temperatureincreases more steeply, which might be a sign that ahotter fluid can be found at greater depth at thislocation.The pressure gradient in the well is hydrostatic (~ig. 11)corresponding to a mean density of 1022 kg/~. The densityof seawater at 35°C is 1021 kg/m 3 .Well SG-2 is a shallow well (about 250 m deep) located inthe area where the geothermal system is vented to theatmosphere. It was initially boiling up to at least 115 m.Now It Is boiling from top to bottom. Unlike many otherwells ofincreasedresult ofthe Svartsengi geothermal field, this well hasin temperature with time (~ig. 12). This is theincreased boiling In the top of the reservoirdue to the pressure drawdown In the reservoir. The boilinglevel Is continuously migrating to larger depths, whichresults In higher pressure and temperature of the steamcap in the top of the system. Between 1976 and 1981 the

SVARTSENG I ,WEL L SG-1SVAR TSENG IWEL L 8G- 1TEMPER ATURE Le]31 ~ ~ 3~ ~ ~ D ~ ~ ~ 41 .2 43 4 4 ~ ~ Uo I I , , , , , , , , , , , ! ! t I•, •PRESSURE (kg/cnl'J8 10 12 14 1(1 18 20 22 24""~"""."'"n,urW0'",~'"".'"n,urrw0'"'"'"'"'",~'"_I'EASURED ' 76.07 . 17 8.S.'00 "J__ i--L-" __ i--L-" __ i--L-" __ i--L-" __ i--L-" __ L'"'"-+-MEASURED ' 111.01.17 S.S .,"J-

32~~• ..~, .J ..3 • w_ o"- • •Cl)~ 0", • , ·00 • >0• •0 0 c.;;- 0~ ~~•N, ,. , .... -,--.--......- , I ...... -.......•0 00 .•~ ~ , ,• •0("'J Hlrl30< < "•• •I I • c0 0~ ,N 0• • • • •~• •~~ ---'.• c• •~N--- -_..;;~1; ~~c] •• 00 •, •, •~ • •~~e. • c- N • "(!) I 0•• 0Z(!) w.0WCI)Cl) •• " 28 -- •~~ • ,f-..J 00N- O00~..:--c00"" •0 0 < ~ ~ • • • • •0

33temperature at the bottom of the well has increased by 15°C(Fig. 12), and the corresponding pressure increase is 1.3bar (Fig. 13).Well SG-3 Is located close to well SG-2. It is ~02 m deepand has a bottom temperature of about 225°C as measured In177.06.14. At that time, a two-phase zone reached up toapproximately a depth of 250 m (see Fig. 15). In 1981 theboiling level has migrated down to 375 m (FIg. 15) and thetemperature has decreased by about 15°C. Further drawdownIn the reservoir has resulted in a pure steam cap down tothe bottom of the well. Both temperature and pressure haveincreased in the period from 1981 to 1983 as can be seen inFig. 14 and 15.Well SG - 4 was initially drilled to 1713 m depth, and for along time it was a production well with 1680 m cleardepth. Casing damage occured in the well in 1980 and theattempt of repairing it by redrilling was unsuccessful. Thedrill bit went out of the casing. Presently the well isopen to 540 m depth and cannot be utilized for powerproduction. The top of the geothermal system is fairlywell defined by the temperature profiles in the well(Fig. 16). The steep temperature increase between 400 mand 600 m reflects the impermeable layer at the top of thereservoir at this depth. The initial temperature in thiswell was close to 240°C and the temperature was ratheruniform at least down to 1700 m depth. The pressuregradient is hydrostatic and boiling conditions have notbeen observed in the well during static conditions(Fig.17).Well SG-5 is 1519 m deep. There has been a temperaturechange of 3°C deep in the well. The pressure drawdownbetween the years 1976 and 1983 has been about 13 kg/cm3(Fig. 21). Boiling levels were at 300 and at 400 rn in theyears 1982 and 1983 respectively (Fig. 19 and 20). whereasno boiling is recorded in 1976 (Fig . 18).

~. - .34•~~, ~~g0•~/N, •INI ,• I j 0N• • N,&;.• -c') ON /b ,to I , . ,ON• Zto~~ I, W(I) (I) •W->- ..J ~N/ C>: ..J ,. ,• •~~N , ~ i ,«W ~.>3 w_ , • • •(I)~","-/ • • ~,- "~8 8N ~~: • >~• '"~• -" , ~~0 0•0w w w• ~~ ~ ·0•, r • • ~ ~ •> ~~I I I" ,'Trnot..- ,......,-.- · ~T""' - i, ..-rp~~"• •0~ ~ 8 • • g • ~ g0~ g0 0N N• ,N N N 0- - - - , • • •~N N" • " • • •~CW) Hld3C1;r..~,~~L.....u......l. .........L...o......L..o ~!" "/,//0., "; • •~0• , ,•0i1, - ----0,.-" ,• " ;: · • • •J·-.- •• ,"'• ,~ • ·,-C) [Lj ,:i .; j, N0 to I L -0 N / ...- • ~" ,Zto w , , ;;j•~,W(I) 0~~ •w .,(I)~ ~ •,•>-..J : ..J ~2 • • • • J~ ~ ~~!! w «W"• • 3 w. • , (I) , ~ - • , ~ ~ ~ • •0• " 00, !" ! 0 >••• ~ ~ ;; ,§..; ..; 1 • >• 0u u 1 •0 0~ u w • • ~J, ~ > >~ ~ , •0 ~~ • • • ui •~ • ~ • ·I• I • • I I ," •>• -~~ 0~ : I 1 ...... ' , ,,, , " •0 0 0• ~ • • ~ ~ • ~ • g 0 00 ,0 N •~,N N N• • • • •N, , "[W] Hld30• •~•~

-".(!)'W (!)rJ)rJ)f­a...J WrJ) 3-".(!) ,Z(!)WrJ)rJ)f-...Ja...J3rJ)G~u"~~ ,

SVARTSENGIWELL SG-5TEMPERATURE CC)85 IOI!§ '26 He 1&6 186 20& 225 24& 28 5 2815 306 3215o j , , , , , , , , ! 1 , l'"."'"SVART8 ENGIWELL 8G-5TEMPERATURE ["C)1110 170 liD 190 200 2 10 220 230 2

SVARTSENGIWELL SG-5SVARTSENGIWELL SG-5n,uTEMPER" TURE le]190 200 210 220 230 2~O 260 2eO 270 280 290 300 310o I ! I 1 , I , , , , , ! I'"'"'"'"i=.ooo•~,~.."'""''''_I1EASUREO '83. 06.05 es_ S.9.T. ACC. lO tEAS. PRESSUA[ '83 .01l .. 0S liS",,-L~-L~-L"-"-"-"-"-"-"-L-~~~-"-"-"-"~-"~-LFlg. 20 Svart~engl well se-s . Temperature prorlle In 1983and corresponding saturation prorlle.,'"'"'"."'"'"'"."n,u .";: t 000• w01100,~"",...'""'""1100,~,""''''''PRESSURE (kg/c..n12 ~ ~ ~ ~4 ~ 80 9 n ~ ~ 100 I~ \10j , , , , , , , , , , , , l_ MU_EO ' 83 .011.05 BS_MEASUAEC '12.03 . 1' 119/ 81>9' P_t3. 1 bQ ~_IIEA!UAEO '7t1.07. 17 8 .S, • II.!.. 14. 4 .JL-"-,-"-,-"-",-__ ,-"-"-__ -,-"-,-"-"-""-,-,-,-"--, .. Iflg.21 SVllrtsengl IIell SG-5 . Pre.ssur e p r o rlles In 1976 ,1982 and 1983 . ...•w...,

38Wel l SG - 6 was 1731i ID deep before it was redrilled in 1982to a depth of 19511 rn. Apparent changes in temperatureprofiles in 1982 are due to the cooling effect of thisredrilling. The cooling effects are therefore badlydocumented in this well. Deep In the well it has cooleddown by !j°C between the measurements of '81.12.09 and'83.06.15 (Fig. 22). In the same period there has been apressure drawdown of about 4 k g/cm2 (Fig. 23). In 1983 theboiling level has gone deeper than 500 m.Well SG - 7 is 1438 ID deep. There has occured a cooling ofabout SOC (Fig . 25) and a pressure drawdown of 3 kg/cm2between the '82.03.15 and '83.05.04 measurements(Fig. 26). But between the ' 80.10 . 08 and '83.05 . 04 datedmeasurements there has been a cooling of 4 to 7°C (Fig. 24)and a pressure drawdown of 7 kg/cm2 (Fig . 26). In 1980,1982 and 1983 the boiling levels were at 400, 450 and 500m respectively (Fig. 25) .Well SG-8 has a depth of 1603 m and has undergone a changeof 7°C cooling (Fig. 27) and a pressure drawdown of2 kg/cm2 between the measurements dated '82.03.16 and'83.03.16 (Fig. 28). According to the measured pressure onthese dates the boiling level was 425 and 450 mrespectively (Fig. 27).Well SG - 9, which is 994 m deep , experienced a cooling of9°C at depth (Fig. 29) with a pressure drawdown of2 kg/cm2 between '82 . 03.17 and '83.05.04 (Fig. 30). At thelatter date it was boi l ing at a level of about !l50 mwhereas there is no indication of a ny boiling atthe formerdate.Well SG - l0 is 425 m deep and close to wells SG - 2 and SG - 3.At this location, the reservpoir is boiling down to thebottom of the well. The steam cap is down to !lOO m. Thetemperature has changed moderately in the period 1980- 1982(Fig. 31), The form of the temperature profiles indicate,however, that the boiling level Is migrating downwards atthis location . This is further strengthened by the

SVARTSENGIWELL SG-6SVARTSENGIWELL SG - 6,=1'''~

40•~~• s" • •"• ~ ,• 0 0~ •;~••, , • •• ,-"0nu. 0t!) I L~8 " "Zt!) W " · "• •wO) • ,~~ 0 0) ~. " •0>-..J ~• ~ •-.~ :g " ""'..J w. 0 ~ «W • ~ ~. • >3 w. • ~· • 00)·~. · , , • • 00~ "• • " •" ,• •0~1~ 8 0• , • • ~ ~ • • ~ •~ "2 • • ~t t • ! >j~~••~, ....,-..-, • I I I~0 g0~ • !ig . ~ 8 8 • 8 ~ !i • 8•~!i i 8• • " - , 2 8C·]• •Hld30 ~0--"08 •~•••~• •& ~0• • Si! • •-" ,t!) I La · ·"zt!) w-• , wO) i .~~ • 0), ~->-..J

SVARTSENGI\JELL SG-7SVARTSENGIwELL SG-8,"o I "" "PRESSURE [kg/c.IJ46M~'51105 95 1011 1115 125TEMPERA TURf ['C)170 . '0 .eo 200 2 10 220 230 2 40 250 2aO 270 280 290 300 310 320 3300 " .' ,I,I,I,I,I,L.l.I,I,I,I .I.L~~,~.00~""i= 1000~Wa,~," 001800l'100-1.-I£ASURED '83.06.0" liS lPo~ 1 0.1-11.5 rJ...... I£ASUR£O '82 03 t 6 IS/SlOG_IIEIISUREO '10 1008 IS/OJ(I__tIUSUR[ O '80 03 0 .. aS/o'JoI£AIlIRED 'eo 01 28 8H2000 ... _L-----..1.-...... ..l.-.. -t- .. ...L.....- ~ _L . _LI,~~.00 -'00-' 00-~ 1000-"Wa1200-'''00-II!IOO -'1800 J~,\\I \!I1 c...... _ ........... ' 113 . 03.18 balM" Po-e. l 0b00- r__ II(ASUREO '82.03.1091/91>0 l P .. -,II 8-18 I).... 8.1I.T. IICC. TO HEAl. PRES5UIIE ' 8303.111..... 5.11 T . ACC. TO I'IEA5. PRES8URE ' 1112 . 03.1112QOO 1_. L .. I '. L. _L

.. ,~~" "O..J..--- I "j,. ooj1,oo~j. 00.. '"•SV ART SENGI\.JELL 8G-8PRESSURE [ko/c rrl)110 60 70 80 9(JI _• • l _ ...... 1 _ •.. 1 . • L100 110 120l . 1 _ . _ I'''0[tn,~,".."~'"' 0000'SVARTS ENG I\.JELL SG-9TEMPERllrURE Le)." '"':K!1l 230 2 40 260'" '" '" '"'t'"~ 1000~WaIlQOI~~wa~,'"I~OO'"~.~ ,"" _I'£"'_EO ' 83 03 . 18 !lS/GO' Po· G. 12 b ........ 11E ... _EO '8203. 1889/81>1>' 1>0-14.8- 18.0 b .....000.. "..... _'"-".01 '83.05 . 0

430NNN~~•-••;;• •~0l•0 ~0•, 0-0 EN

8VART8ENGIWELL 8G-10n,~,•wnPRESSURE [kg/c.t.1, , e a 10 12 14 18 11 2'0 22 24 2'8 211 30"""'"12I J"' ~"'1"'1225~,250 "1275 -I3OO ~3251'" ~'"~ 14211 1I45°1 _ MEASURED '1I3.03 . 02BS/SI>G'"... I1! ASlflED '15Q. 1 1.27 aS/GG'"1\\I~Flg . 32 Svartsengl well 50 - 10. Pressure profile.!! In 1980and 1983.

45pressure profiles in Fig. 32. In 1980 a pure steam capreaches down to approximately 300 m. but In 1983 the lowerlevel of the steam cap Is at least at 350 m depth(Fig. 32).Well SG-l 1 is 1141 m deep. It has cooled down by about 13°Cat depth between the dates '82.03.18 and '83.05.05(Fig. 33). The boiling level in 19 82 right after completicnof the well was at a depth of 400 m, but due to thecoolIng of th e reservoir fl uid the boiling leve l !n 1983 Isalmost at the same level as In 1982 Inspite of a drawdownIn the reservoir during that time.Well SG-12 was initially drilled as an relnjectlon well toa depth of 11167 m. Its flow has been tested and anexperiment on injection has been carried out with thiswell. Between 1982 and 1983 ( Fig 35) a cooling seems to betaking place in the well. but the experiment on injectionmight have disturbed these measurements. The drawdown inthe reservoir at this place seems to be of the order of 7bars between 1982 and 1983 (Fig. 36). No boiling effectshave been detected in this well.

SVARTSENGIWELL SG-IITEMPERA lURE rc).85 11105 2O!i 2115 22.5 2315 2~15 21515 2815 2715 2l1li 2915O! r ~ 1 1 1 1 • 1 1 • 1 1,"" "M ..SVARTSENGIWELL SG-II"PRESSURE (kg / col]~ ..,. ..~'"'".00'"'"'00n,00''00n,' 00'"'00~~ 1000~Wo'''',~~r~~W0'00"' ~,;"'1j1000'000..... 11E",l.REO '83.06.06 8S 15 P ......... ~.aoo-i ... ' .•. T . ...cc. TO IE"S\.REO PREUI.A£ ' 83.06.06_..... MD.8l.REO ' 112 .03." 18111>9 15 P_''''-I7 .' Ocr... '.I.T. "CC. TO IE"S . PII[tltRE '12.03.18~_~_~~...L" 00"00"00..... 11E"SUllEO '83.06.015.015 8S P ..- 13. 15-'15 Ocr_1E~SUIIEO '12.03.18 IS/SPG 15 1' .._11.15_17.1 b..,.,.001I-"~L-__"-~~~-L~-"-"-"C-~L-~"-~-LFig.)) Svart.5engl wIl l 50-11 . Temperlture profiles In 1982and 1983 with corresponding saturst lon profiles.Flg.3Q Svsrtsengl well SO-11. Press ure profiles In 1982 IInd 1983.

47N,~• ~, •8 •"•a'i:i•L-N • •(!)- , I " Z I •• • ~,W(!) u , , LU)(f) • •>- ~g • iO::.:..:J ~•N" ·"LJ ~g , • ~ ,>W w ii! 0en3 «N~g• " ",8 , • ; "•g0L~• ,•~ 0 0 0 •~• • • • NU ~ ~" " " •~~ ~ ~L, •+ ! >+ "" •• •~• ~ ~ 8 § 8 ~ 8 8•~C·] Hld;ao•" • • •• N• • •"8• ••0L0• •0•• L 0~n~ • • -N 8 , L L0 ! • (!)- Lg • • " · ·, Z • I • • • N L~. •W(!)•~ "• ~ • ~ 8 8 " •8 !! 8 • ~• , • • ~~CW] H1d30•N~~

485 SPATIAL DISTRIBUTION OF TEMPERATURE AND PRESSURE INTHE RESERVOIRLocation of wells and cross-sections Is shown In relationwith the locatIon of the power plant In Svartsengi (Fig.37). In order to show both the vertical and the horizontaldistribution of the temperature. several croas-sectionshave been selected together with planar mapsdepths.at various5.1 Vertical distributionThe wells I n Svartsengi are more or less alligned in theSW - NE direction. Cross- section A will therefore be theprinCipal one (Fig. 37). In addition five other crosssectionshave been selected In order to see better thethree dimensionalthermal system. Weindividually.variationwill nowof temperature In the geodiscusseach cross- section5.1.1 Cross-section AThe location is shown in Fig.distribution is shown for 1976initial state of the reservoir),37. and the temperature(which is close to the1982 and 1983. BeforeexplOitation, temperatures in excess of 2~0°C were found inwells SO - ~ and SO-5 (Fig. 38), but in 1982 such tempera ­tures were only observed in the SW part of the field inwells SO-ll, 9, 8 and 7 (Fig. 39). In 1983 temperaturesof 240°C cannot be found in the reservoir (Fig. 40). Thehottest part of the reservoir at this time is 230 - 235°C.There seems to be a temperature inversion in the SW partof the field. and the hottest part atlarge depth seems tobe around wells SO-5 and SO-6. The disappearance of the2~OoCfluid can be associated with the utilization of thefield. Production started from the northern part of thefield, and up to 1981,the production was only from wellsSG-3, 4, 5 and 6. This means that the mass taken out of the

;-r-::I J HO - HS" - 2 300 Gq\//Ll:.JS3.0a . 1017 AASVARTSENGI H,T AREALocat ion of drillholes_ /\ I "~ ~ 2 :: LlN£-CJ \/ /5G3\~, • }\_-_---- ~. __ f/J..,..,.1:',,', ," ·1":,, ,~- ,;;-- . ' "" ,::.',' '>, ... ,\ I , -:-:' ~ SI.> -5 .-:'' t;:"'""'". I.~ " '~ ::>~NI/I~0",5G;'~.? ?Dv ~

50~JHO-HSP-2300GGL.:.t.J83.08. 1019 U.SVARTSENGICross section- AInitial stot eSG-# SG-5 SG-4oDepth,9> 192 ,0500 22~ r-..... Z'=750ZZ6"0I1I1240V~6o-

EI]JHO-HSI> . 2300 GGI 83.08.1020 A"ASVARTSENGI 1982Cross section- ASG-# SG - II SG-9 SG-8 SG-7 SG - 5 SG - 4 SG-6 SG-12 SG-3 SG-2 SG-K)Dep th[m)0."In.. , , ,~,1~2 . 7I 18 :1.3~,202..8 2~ . 0208.8~,204.7 209.1210:4 211 .0113.8 211 .2,~,1~1.0211 8~201. 8 0215.1,~,219.6 222.6 217. 2 1770 20 5.0 1:10.0,,; • 0~\B21"~2a.' 194.'211 8500231. 1 H.750- •,,;'"' • •UO.5- 220 g51.1 2 59.5 U3. 4 0311.3 239.7 ,~ 234.2 226 . 5 " 0241.0'"m. 221410001250241.4 2)8.' U7.0 226 .5238,2Z..o.1 242.0 237. 8~ . 232 .2z" .~Z ~ L1240 2385 2,..8 232 . 4,,; ,,; 24L1 2~UU8.S238 .1•~ 242.1 2(0,,.." "•• "2 42-1 232.4 259.01500 242.1 239.2,,;241 .3 • ,,; 2l!l.ZI

GJ;]JHO-HSI>-2300 GG83 0 81021 AJt.SVARTSENGI 1983Cross section- A'-"'"SG-# SG-Ii0Depth[m]ZOI.5SG-9-_. ,,,,,2'2.0500 - V 2211.'1/ fnt.: Z32.8IU7.3 232 .8750 - \2~\1000 -232.'2213 233.0221.:5u n"U7.lm .;"• •001250 - • •15001750 -2000"• •........SG-S SG-7 SG-5 SG-4 SG-6 SG- 12 SG - 3 SG-2' 71.' ZOZ .7 ZIO.'z,z.o204.1 2131ZI3.1/ m..21 4.3211.2 2202 Z224 159. 8•226.4 Z2l 3 213.3 il;2n. ' 2)50 228 6 237.9 Z2l.l 220234.7 2)4.' 230 . 23S .1 226. 22 34.' 234.; 231 8 2)7.8 2194234.9 2 3 4.1 2 34.1 237.9 221 .'234 , 2n.8 235' n ll.4 2U.l234.9 233.8 23,.:5 2311 .4 U:5.]'- 234 , 233.0 U,., 2'U 222 .''- 23 :5.0,;; 2!2 .4 235.5 tU .323:5.0r-Z26.1 23l. 3 228.422'.8........ 230235.0 m 235.4- 0••• UlO•234 . 7•2311. 12)'.0 ~ .- 230 •2 35.4 00 •0"m~•-----.;",eO. 3'975~,SG- IOFig.~O Cross-section A. Temperature distribution in 1983 .

53reservoir during 1976- 1981 Is from wells which are to thenortheast of the power plant. On the other hand theproduction from wells to the southwest started in 1981and has been of the order of 300 kg/s sl~ce then (Fig. 6).In 1982 the 240 0 c hot fluid Is not found In wells SG-lI andSC - 5 (Fig. 38), but is measured In wells in the SW part ofthe field (Fig. 39). This part of the reservoir had atthat time been I n production for about a year. We aretherefore tempted to assume that temperatures up to 240°Cwere found In this part of the reservoir prior to production.In 1983. however, the whole reservoir has a tempera ­ture lower than 240°C (Fig. 40). The temperature distributionin 1983 (Fig. 110) suggests the upflow zone in thesystem to be near to wells SG-5 and SG-6 and furthersuggests there is a lateral flow at 600-800 m depth fromwells SG - 5 and SG - 6 towards wells SG - 9 and SG - 11.5.1.2 Cross-section 8It is almost in the N- S direction (Fig. 37). It shows onlythree wells. namely SG-5. SG - 6 and SG-12. The temperaturedistribution for 1982 is shown in Fig. 41 and for 1983 inFig. 42. These figures are in agreement with thehypothesis that the upflow zone can be near wells SG - 5 andSG - 6. The cooling observed in well SG - 12 can be due to thereinjection experiment performed in September and October1982, or alternatively increased cold recharge from thenorth.5.1.3 Cross-section CIt is almost in E-W direction in the northern part of theproduction field ( Fig. 37). Only the initial temperaturedistribution is shown in Figure 43 . This figure reflectsthe fact that the natural discharge to the surface Is theboiling area around wells SG-2, SG - 3 and SG-lO. At shallowdepth the temperature is higher there than elsewhere Inthe production field.

54rn.JHO.HS~.2300 GG83.08. 1013 Ail.SG - #0SG-12SVARTSENGI 1982 -Cross section - 8,SG-6SG-5Depth[m]14H190 4 17 .. 197.0203·°,09.6"0.0 203.0 219.2«,..500 199.4- 226/_ '1. '1.0229:~220.'233.4 227.:5'1.30234.2 230.0 I-750221.4 2. 3:5.:5 232.3-,,=>'>226. :5 '(, 237.0 236.9V1000, 232 .2 238.:5 237.81250232 .4 238.8 238 .:5228 .5 238 .8238.72. 29.6 238.8 NN233.4 239.0/ ~ '"234.3 N.,239,21500 /N0239. 2.- 0lli175020000 '"N.,Flg.~l Cross-sectlon B. Tem perature distribution in 1982.

55~ JHD-HSP-2300 GC!L...C.J 83.08.1014 AASG-#Depth[m]500SG-12SVARTSENGI 1983Cross section - BSG-6SG-5Or-----+----------------r------------+TemperatureDistribution(OC)213 .3 2302. 25 .3220---~~ __ __222.4352 35.42: 7.9750226.2219.4238 .1237.8221.9237.91000225.3236. 4235.5225,3236.4235..51250225.5222.3236 .1235.51500'":g'"1750 -"'228 .42. 2 8 .8/235.3235.4235.423!1 .0234.711'" C>'" -"'2000C> '"_Ill236 .1Flg.42 Cross- section B. Temperature distributi on in 1983·

56~ JHO·HSI:I-2300 GGL:..t.J 83 .08. 1011 A 'ASVARTSENGICross section - CInitial stateSG-# SG-4oDepth92 .0152.0192 .0500 210.0750"'-236.0240 ,0~ I.O242\1000 242.01250242 .0 N...242 .0242 .0242 .01500 241 .0SG-6SG-3,SG-2,SG-10200.0 I • •203 .0IU203.0210.0)~~IS . O 197o 216.0.., -'2. 225.0.-. '/.0_22 •. 0'1.30,..,0" NN0 0,.., N ·-237.3 ... ...237.9237.6237.3237.3237 .5237. 4236 .8234.6.-0

575.1.~ Cross-section DIt is an E-W section in the southern part of the productionfield (Fig. 37). The temperature distribution in 1982 Isshown In Fig. 44 and the distribution In 1983 Is shown inFig. 45. Quite large temperature changes are taking placein this part of the reservoir during one year. In 1982 thetemperatures are fairly uniformly distributed In this E- Wcross - section. A possIble temperature inversion Is notedIn wells SG-1 and SG - TT . Th e temperature distribution of1983 suggests that substantial cooling has taken place.particularly In well SG - TT. and this raises the Questionwhether this cooling is through well SG-lT. As seen InFig. 33 the possibility of down flow in 1983 In well SG - Ttcannot be ruled out from the present available data. On theother hand. the temperature profile from 1982 does notindicate a downward flow at that time. Temperatureinversions seem to be present in wells SG - 1 and SG - 9 in1983 and the temperatures in these wells are in general1-10 o C colder than in 1982. The cooling in well SG - 11 is13°C between 1982 and 1983.The lower end of the cemented casing in SG-l1 is at 582 mdepth. and a large604 m depth. Itacquifer was noticed during drilling atis possible that a downflow from theacquirer at 604 m caused the cooling of the reservoir atlarger depth. This situation might be at hand in some otherwells in this part of the field. The temperature profilein well SG - 8 from '83.03.16 (Fig. 21) is In an agreementwith an internal down flow in that well.5.1.5 Cross-section EIts location is shown in Fig. 31. and the temperaturedistribution in 1982 and 1983 is shown in Fig. 46 and 41respectively. The temperature variation within thereservoir in this SW - NE section is rather smooth at each

58~ JHO-HSD - 2300 GGL.:...t.JS3 .08.I009 A'ASVARTSENGI 1982Cross section - DSG - # SG-9 SG-II SG-70112.1 194.0SG-8185.3Depth[m]180.9 198.9 209.120~ . O 215,5 226.6---239.35002 371220219 .2 '2.'3 0---232.2 40.0750'2. 40 238.0 240 .91000...~0'"1250 -'"240.9 2:39.4 241.124 1.71500 '"1750~0N'"232 .4~

59rnJHO-HSI:>-Z300 GG.83 .08 .1010 AACROSS SECTION-DTEMPERATURE DISTRIBUTION IN 1983 IN ·CSG- # SG-9 SG-II SG-7 SG-80Depth[m]212 .0 Z01.:5 202,7 171,~204 .1;;220 .7 209.9 h,o., 217.2220500 228.8 30226.4.........230232.8 227.0 23:5.0\2JS "-.., 233.92328 221.3 234.5 234.7750 "-232 .8 227.3 234.5 234.9233.0 227.3 234.1 234.91000 - 229.5 227.3 233.8 234.9g'" 0227.3'" 233.0-'" ~1250 0,.;1500~017502000--'"233.B 234.9234 .9232.4 23:5.0"-1'-'28.1 235.0235.0'" 02,35 .0'"

;r=:I JHO-HSP- 2:w

I"jT'=I~ HD-H S P - 2 500 GC;L..t.J1 5.01 . 1OO1 AASVARTSENGI1983Cross section - ESG - #SG-IISG-7SG - 120Depth[m 1500.20t.~209 .9202 .7220.21 ~9 e225 2 13. 3227023!>.Ot- 22~750221.32 27. :5234 50'0234~226 2:2 19. 4227.)2 3 4. 122 1.91000227.3ZHe22!> :5125015001750" 0" 0•227.30 •" 0•-"2:338233-0232.4228 I08•225 :5222 .522.2.3Z 2 8 .4228. 85B2000'"Fi g. 4 7 Cross - section E . Tem peratur e distrib u t i o n i n 1983.

;T'=l JHD-HSp-230Q GGLJ:::J 83.09. 1006 AaoSG-IISVARTSENGICross section - ECooling from 1982 to 1983SG-7SG-12'" NDepth[m]500750- 2 .65.64,44 .9m .7.'6.42 . 44 .35.5r 6 .4 -5-9.8-13.9- 4 .8r 0 .12 .01000125012 . 112 .812.67. 08 .07. 78 .38. 64 .66:9_7. 16.07 .6150010f- 4 .35 .0••17502000Fig.~8 Cross-section E. Cooling be tween 1982 and 1983.

time , but the change in temperature from 1982 to 1983 isQuite noticeable (Fig. 48). The cooling in all three wellsis in excess of 5°C and at the bottom of well SG - 11 thecooling is 13°C. Fig. 48 indicates that the cooling of theSvartsengi reservoir is Quite general, though the largestcooling is observed in well SG - 11. Furthermore, thelargest cooling is observed at depth in the reservoir.5.1.6 Cross-section FIt is almost parallel to cross - section E, but locatedfarther east in the production field (FIg. 37). Thetemperature distributions for 1982 and 1983 are shown inFig. 49 and 50 respectIvely. Temperature variations withinthe reservoir are Quite moderate at each time, but acooling of approximately 5°C is observed both in thesouthern and in the northern end of this cross- section,i.e. at wells SG-8 and SG-12. Temperatures in wells SG - 5and SG - 6 have not changed significantl y in the same time.Fig . 49 and 50 might indicate that a cold recharge isentering the geothermal reservoir both from the south andfrom the north . The effect of reinjection into SG - 12 in1982 might of course mask this picture. Significantcooling at large depth has not been confirmed in well SG-5and SG - 6. This can indicate that the upflow zone in theSvartsengi reservoir is located near these wells.5.2 Horizontal distributionThe horizontal distribution of temperature has been studiedat three levels, namely at 700, 1000 and 1100 m depth.This was done for 1978, 1982 and 1983. Data available from1978 or before can be regarded as close to the initialstate of the reservoir, but the situations in 1982 and 1983reflect the effect of full production in the field.

:r:I JHO-H$P - 2300 GG~ 83.08. 1023 Ab.SVARTSENGI 1983Cross section - F'" ~SG- #art"SG-8D ept h 204.1[m]500~ 219_6-+--237.1."'3 0 __SG-5 SG -6197. 0207.62172:223.8224.1227.5203.02296SG - 12150.0 .199.4220.57 50 -1239_724'7:0

f"i'T"':lJHO~HSP-2]OO GGL.:..L...J 83.08.1024 U.SVARTSENGI 1983Cross section- F5G-#Depth[m]5005G-8 SG-5 SG-6 SG-120~1--+---------~--------+---------1-----------!71.~204.'lIT.'1 203.0220.226 .4 221.3 291 ,,230I- 15. 8213 • UI.' U3.4 "'", ~ll .. 32262750234. ' 231.8 U55 219 4234.7 no.'""234. ' 234. ' Bro uu1000 234.' 2l5.5 2385 U5.312502H.9 235.5 238 8 U5. 3,234. '2l5~U81 U2 5235.0 UB8 Ut l235.0 U'O 22841500 235.0 1592.;235.0- -.; 2)92 00 •- •L0 •• 0 •2000 ..-" •1750.. •"U ••V> '"Fig.50 Cross-section F. Temperature distribution in 1983.

665.2. 1 Distribution at 700 mThe temperature dIstribution in 1978,1982 and 1983 areshown In Figures 51. 52 and 53. In the beginning ofproduction In 1978 for well SO-4 and In 1982 for wells8G - 7. 80 - 8 and SO - g, temperatures close to 240 0 C areobserved at this depth . In 1983 the highest temperaturesat this level are approximately 235°C. The l evel at 700 mdepth is selected here to represent the top of thegeothermal r eservoir in Svartsengi.5.2.2 Distribution at 1000 mThis level is considered to represent the mean level In theSvartsengi reservoir . Temperature distributions In 1978,1982 and 1983 are shown In Fig. 54, 55 and 56 . In theinitial state (Fig. 54) t e mperatures of 240 0 C can beexpected in wells 8G-4 and SC - 5, and most probabl y also tothe SW of these wells. In 1982 the hottest part of thereservoir is in the south where 240°C is still recorded inwells SO-7, SO-8 and SO-11 (Fig. 55). One year later thehottest place in the reservoir is in the SE and thetemperature there is approximately 235°C.5.2.3 Distribution at 1100 IDThis level was chosen in order to represent the deepestcommon level in the production interval of the reservoir.Due to the different depths of the production wells. itwas necessary to select a level where all wells can berepresented. The temperature distributions at this levelfor the years 1978, 1982 and 1983 are shown in Fig. 57, 58and 59. The picture obtained is almost identical to thechanges observed at 1000 ID depth. The hottest part of thereservoir in 1983 is the eastern part including wells SO-8,SO-5 and SO-6. From these considerations and others

67r;r.:I~HO ·Hs~-noo GGL.:..J::J ljJ .OII .09!i7 AASVARTSENGI H.T AREALocation of drillholesSG-12•.SG- IQ ;..700m1978so·,•NISG-S•230..00 '00fig . 51 Telllperature distribution at 100 III depth in 1978.r;r::IoH)- tI S ~ - 2 30 0 G4L.:..J::J e3.08.loOO ,0.SVARTSENGI H.T. AREALocotion of drllthol"SG-12•n.room1982so·,•'"'NI-111111, T ..... '......

68r,n JHD-H$P-2300 GGL:.l:..J 83.D8.I003 All.SVARTSENGI H.T AREALocation of drillholesSG-125(;-10• • '"700m1983SG-4•• w·, 11'"N, II1 j I Tempera ture ~ 235°CFIB·53 Te~p.ratur. distribution at 700 m depth In 19 8~ .r;r> ',','0',"0',,",300 G!,L.:...tJ . . AASVARTSENGI H.T. AREALocation of drillholes~NIw-,• •• SG~Iw-,•00'"' "" .SG-12•.SG-IO :.,"".,".' ., "••I: "SG-2 ::•• I'SG':, ,II' ::'.,. ;:'".,! .'I, "" I';, ,:I' h" ":::;" , .1000m1978rll.5~ Temperature distribution at 1000 m depth In 1978.

69["j'T'=l JHO·HS ~- 2.100 GGL.:..t.J I 3 . 0 8 . 1001 AASVARTSENGI H.T AREALocation of drillholesSG-12•SG -IO•2.32 235 SG-2SOl••1000 In1982SO·,·Z38. ..SC:5.H'.~NI0 « '00 ~.Flg . 55 r'lIIperatlJre dl~tr .lbut.1on at 1000 III depth In 1982.rj1"':I JHO-HS~ -2.300 GGL..t:..I 83.08 .1004 U.SVARTSENGI H.T. AREALocation of drillholes1000 m1983SG·,SG-3••~NISG"• n..SG .... I'"00 '00 =-~ ~ ~ = !,.

70r,r:lJHO·H'" - Z 300 O(;GL.:..t:.J 1I3 .08. 0'99 All.SVARTSENGI H.T. AREALocation of drillholesSG-12•.5(;-10 ~. "56-2 ,•• I'5G·31I 0 0m1978. so ..'"~NI50·.SO· 7• •.SG~ I, 00=-0.56- 8I: " :;, " "'.. ",I I' ";'.',IIIII Temperature ~240°CHg. 51 Ternperature distribution at 1100 m depth In 1978 .~ JHD - H81> - 2300 GOL.:..J:.J U . OII , I002 4:lSVARTSENGI H.T. AREALocation of dnllholesSG-12•,". SG-6"..,.;-----!~"••5G'3235.",~' ~,5G - IO•50·'.'•1100 m1982N,,1 1111' Temperature ~ 240°CFlg.58 Temperature distribution at 1100 m depth In 1962.

7 1r,r::l JHO-HS P- 2 300 GG~ a3 . 0e . I003 A'ASVARTSENGI HT AREALocation of drillholes~NISG-'0.SG~I'", ., '00"""SG-12o"SG-IO ;o "5(;·2 ::•• I'S G

72already mentioned it is found natural to assume that theupflow zone In Svartsengi Is closer to these wells thanother wells In that field.5.3 Cooling between 1982 and 1983 in the reservoirThe cooling of the field Is studied by considering thehorizontal coolIng distribution at four different levels,namely at 700,800,1000 and 1100 m depths. This one yearcooling is shown In Figures 60, 61, 62 and 63. A commonfeature of this representation Is that the maximum coolIngeffects are observed in the SW part of the field with themaximum coolIng at well 5G-11. At depth there is a smallor no cooling registered In wells 50-5 and 50- 6. The threewells in the NW corner of the production field areunfortunatelyrepresentation.too shallow to contribute to thisIf the cooling of the Svartsengl reservoir is due to coldrecharge to the geothermal system, it seems that this coldrecharge can come from all direct Ions except from theeast. The most prominent cooling seems to come from theSW.r,r:I~ HD ·HS ~· UOO OGt..:l:..J 83. 08 .099-' A'ASVARTSENGI H.T AREALocation of drillholes•50·'••5G':55(; -10 ~800m1982 - '83tNI050· '• • . 1IO~ - 10.700"" -SG- B·6.150.'' ",fig.61 Cool1ng be tween 1982 and 1983 at 800 m depth.

73~ JHD·HSJ> · 2300 GOt..:J:.J 83 .0'.0"' A).NISVARTSENGI H.T. AREALocation of drillholeso .50· ,50"• • 8~~;~8 \50- '••~97.5SG -125o.,SG-IQ •,;$C-2 ;: "•• ,ISG

7~6 EVOLUTION WITH TIMEThe production from the Svartsengi reservoir started inOctober 1976. The history of the productIon now coversabout 2600 days. Fig. 6 shows the first 2000 days of thisproductIon. During the first two years the mass flow ratefrom the reservoir Is 50 kg/s, but In 1979 the flow rateincreases to approximately 100 kg/so The next increase inmass flow rate Is in 1981, when the winter flow Isincreased to about 300 kg/so The productIon rate has beenat this level during the last three years. In the summertime the mass withdrawn from the reservoir Is about halfof the winter flow rate. The total mass taken out of thereservoir In Svartsengi Is now about ~xl010 kg.The pressure drawdown In the reservoir has beensubstantial. Fig. 6 shows that in 1982 the water level inthe reservoir is 90 m lower than at the start of produc ­tion. There is a simple relation between the pressuredrawdown in the reservoir and the total mass wi thdrawnfrom the reservoir as shown in Fig. 6~.The Svartsengi geothermal reservoir is a typical liquiddominated reservoir where the ther mal fluid in theformation is in liquid phase. The temperature of the fluidin the undisturbed state was close to 2~OoC. However,during the natural surface discharge in the area wherewells SG - 2, SG - 3 and SG-l0 are located, the system wasboiling and two - phase conditions are still present atshallow depths. During the exploitation of the field. wherethe pressure in the system has decreased with time, theboiling at shallow level around wells SG - 2, SG - 3 and SC - l0has increased wi th time. These cond i t ions were descr i bedin Chapter 4 (see Fig. 12 , 13, 14 , 15, 31 and 32) . Thelower level of the boiling zone is now at approximately500 - 550 ID depth.The top of the main reservoir in Svartsengi is, however, atapproximately 600 m depth. The caprock in the geothermalsystem is most likely the hyaloclastite formation as shown

75rn JHD·HS~-2300GG83.09.1274 A"ASVARTSENGITOTAL MASS WITHDRAWL ASFUNCTION OF DRAW DOWN,lO"10 101110'/V-/il10·~14'"-• Measure ment- Best line through pointsI 10 100 draw down [m]Fig.54 Re lationship betwee n pressure drawdown and totalmass withdrawn from the Svart sengi reservoir.

76In Fig. 3. Significant temperature changes were notrecognized until 1983 In the main liquid part of thereservoir I.e. beneath 600 m depth. However as hasbeen shown In the previous chapters. there were signs ofsmall changes prior to 1983. These changes In temperaturewere not signIficantly larger than the absolute accuracyof the instruments used for temperature measurements.Therefore, significant temperature changes 1n thereservoir could not be detected until 1983.At present. the response of the Svartsengi reservoir toutIlizatIon has been both a decrease In pressure andtemperature. The pressure decline Is considered to bepredictable Crom natural hydrological considerations. Thetemerature decline, however. is still a matter ofdiscussion. as the mechanism of cooling has not beendefined. The purpose of the present study is to put someconstraints on the cooling process presently observed Inthe reservoir.

777 CAUSES OF TEMPERATURE CHANGESExploitation of the Svartsengl reservoir has caused thewithdrawal of about 40 million tonnes of geothermal fluid(Thorhallsson, personal communication, 1983), This hascaused about 13 bar pressure drawdown In the reservoir . Adecrease In pressure causes increased boIling in the top ofthe reservoir. The recharge of cold water can increase aspressure decreases In the reservoir. Episodic cold inflowfrom the overlying ground water can take place In associationwith earthquakes. The decrease In the temperaturecould also be caused by a decrease In the lateral flow ofhot fluid. This chapter will deal with some of thepossible cooling processes in the Svartsengl reservoir.7.1 BoilingThe salini ty of the geothermal fluid of the Svartsengireservoir is roughly 2/3 that of sea water. The saturationtemperature of this salt solution is higher than thesaturation temperature of pure water by a maximum of 2°C ata depth of about 2000 m (Michaelides, 1981; Grant, 1982)Little or no correction is thus needed when using thesteam tables.Exploitation has caused a pressure drawdown which makes itpossible for the reservoir fluid t o boil at greater depththan it could initially. In other words. the drop inpressure caused by the exploitation has increased theboiling In the upper part of the reservoir. During boiling.thermal energy is used in changing the state of the fluidfrom liquid to vapour and this causes cooling in thereservoir. However, as the level of boiling is migratingdownwards, hotter and hotter water comes into boilingconditions. This boiling sustalns the steam cap locatedaround wells SG-2. SG-3 and SG-l O. The effect of theboiling Is therefore to increase the temperature of thesteam cap. The temperature of the steam z one was about

78210 - 215°C in 1983. but was approximately 190-20QoC priorto exploItation of the reservoir. As the largest coolIngIn the Svartsengi reservoir 15 observed at a great depth,where the pressure in the reservoir is much greater thanthe saturation pressure for 240 o c, boiling can be excludedas being the cause of the cooling observed In Svartsengi .7 .2 Lateral cold inflowThe fall in pressure has allowed additional recharge fluid,more cold than hot. to flow into the reservoir. The factthat cooling Is observed In many wells In Svartsenglfavours the possibIlity of lateral cold inflow into thereservoir. However, the hydrological model used so far tosimulate the pressure drawdown In the Svartsengl reservoirassumes the recharge to be only from one direction, andthe other three directions are impermeable boundaries tothe system. Furthermore, up to 1982 there could not bedetected any recharge to the reservoir (Vatnaskil, 1982).If the recharge of cold water had started in 1982, thereneed not be any discrepancy between temperature measurementsand reservoir calcuculations. At present thepossibility of cold inflow is considered to be one of themost likely explanations of the cooling mechanism of thereservoir.7.3 Episodic cold inflowThe reservoir lies along the main zone of earthquakeepicenters on the Reykjanes peninsula, as mentionedearlier. It is probable that fluid recharg e occurs fromepisodic cold inflow of surface groundwaters. Episodic coldinflow into the reservoir in connection with seismicactivity is considered a possible explanation for thecooling .

797.4 Decreased lateral inflow of hot fluidAs described previously the temperature distribution in theSvartsengi reservoir indicates a lateral flow in thereservoir at 700-1000 m depth. The direction of flowshould be to the SW. If this natural flow of hot fluid isdecreasing due to pressure decline in the production field.this could show up as coolIng In wells downstream in thisflow. The available data do not contradict this descrip -tian of the cooling, but the result is quite similar tothe recharge of cold water.7.5 Calibration of instrumentsAnother possible explanation of the coolIng in theSvartsengi reservoir could be that the instruments usedfor temperature measurements were not properly calIbratedand that the cooling is simply a measurement error.possibility has been looked into, and it is foundprior to 1982, the calibration of the measurementThisthattoolswas not satisfactory and at this time the accuracy was notbetter than ...... 5°C. However, as many different instrumentshave been used, the confidence in temperature deter minationis higher, and an error of the order of 3 - 5°C should havebeen detected . After 1982, the ac c uracy of the temperatureinstruments used in Svartsengi should be better than...... 2°C . The cooling in well SG-11 is therefore far frombeing within the reasonable accuracy of the measuringtools and is therefore considered significant. The coolingin wells SG - 7, SG - B and SG-9 is close to the accuracy ofmeasurements as it was before 1982. This cooling cantherefore be considered to be Significant . In o t her wellsit is not possible to state that temperature changes havetaken place .

80In the treatment of thetemperaturedataIn this reporttemperature changes ofless thanS'Chave not beenregarded as sIgnificant.7.6 DiscussionIn the previous paragraphs five possible cooling processeshave beendiscussed.It was found that boiling cannot be responsible for thecoolIng observed 1n Svartsengi. All the other processesmentioned can be the cause of the coolIng, and it ispossible that a combination of all the above mentionedpossibilitiescoolIng.1s the real explanation of the observedFrom the data presented in this report, the possibility ofa cold recharge at about 1000 m depth In the neighbourhoodof well $G-11 seems to be the most plausible explanation.

818 CONCLUSIONS1. The Svartsengl reservoir is a liquid dominated reservoirwith a temperature of 235-240 oc.2. The main reservoir Is below 600 m depth, and ahyaloc!astlte formation acts as the confIned cap rock tothe reservoir. Natural discharge to the surface causesboIling above 500 m depth. and a local steam zone atshallow level is mIgrating downward as pressure decreasesin the reservoir due to exploitation.3. Prior to production, temperatures In excess of 240°Cwere recorded at several places In the reservoir. In 1983the highest temperatures found In the reservoir arebetween 235 and 23 7 °C.4. Between 1982 and 1983 significant cooling Is recorded inwell SC-TT and almost significant coolIng in wells SG - 1.SG-8 and SG-9. The cold water inflow seems to be at around1000 m depth near to well SG - l 1.5. The pressure draw down in the Svartsengi reservoir in1983 is about 9 bar and the total mass withdrawal from thereservoir from 1916 to 1983 is 4.1010 kg. A simplerelation has been established between the total masswithdrawalreservoir.and pressure decline In the Svartsengi

82ACKNOWLEDGEMENTSThe author would like to take th i s opportunity inexpressIng his gratItude to the United Nations UniversItyfor awarding the fellowship and the National EnergyAuthority of Iceland for creatIng the learn1ng situat i on.The active help and constant encouragement of the ProjectCoordinator of the UNU. Dr. Ingvar Sirs!r Frldlelfsson.and the continuous supervision and instruction of theDeputy Director of the Geothermal Division of the NEAtDr. Valgardur Stefansson. without whom this report wouldnot have been a possibility . Is highly apprecIated.The rich experience imparted by the teaching staff of theUNUis acknowledged.Special thanks are given to the geophysical logging teamfor giving the author practical instruction on the variouslogging applications used in Iceland.The author is also grateful to Benedikt Steingrimsson forcarefully reading the manuscript and giving valueablesuggestions, to Steinar Thor Gudlaugsson and Hel gaTulinius for introducing him to the computer softwarefacilities, to Gudjon Gudmundsson for collecting some ofthe measurements . to Audur Agustsdottir for drafting manyof the figure s and to all the Icelanders who made his workand life in Iceland more meaningful.Last but not least the author would like to thank SigurjonAsbjornssonwork.for creating a conducive atmosphere for the

83REFERENCESArnason, B .• 1976: Groundwater systems In Iceland tracedby deuterium, Soc. Se!. Islandlca. v. 42, 236 pp.Bjornsson, G •• 1983: The power plant at Svartsengi,Technical features and operating experiencePresented at The World Bank In Washington DC,June 28,1983, 21 pp.Ellasson, J., 1973: Convective ground water flow. Instituteof Hydrodynam i cs and Hydraulic Engineering. TechnicalUniversity of Denmark. Series Paper 3.El ias s on, J. t S. St . Arnalds and S.P. Kjaran, 1977:Svartsengi. Straumfraedileg rannsokn a jardhitasvaedi.Orkustofnun, report OS ROD 7718, OS SFS 7702.Franzsson. H •• 1983: The Svartsengl high-temperatur e field,Iceland. Subsurface geology and a l teration: GeothermalResources Council Transactions. 7. 1~1 - 1~5Franz s son. H .. 1983: Svartsengi. hola SG - 12. Borun.jardlog. ummyndun og vatnsaedar. OS - 83003/JHD- 02. 55 pp.Georgsson. L .• 1981: A resistivity su r vey in the plateboundar1es 1n the western Reykjanes peninsula.Iceland. Geothermal Resources Cou nc11 Transact10ns,5, 75 -78.Geo r gsson. L. and Tul1n1us H .• 1983: V1dnamasmael1ngar autanverdum Reykjanesskaga 1981 og 1982. Or kustofnunreport OS-83049/JHD- 09, 70 pp.Grant. M .• Donaldson. loG. and B1xles P.F .• 1982: GeothermalReservo1r Engineering. Academ1c Press, New York.369 pp.

84Kjaran, S.P., Halldorsson G. K. , Thorhallsson, S . andEllasson J . , 1979: Reservoir engIneerIng aspects ofSvartsengl geothermal area. Geothermal ResourcesTransactions, 3, 337-339 .Kj a ran, S.P., Ellasson, J . and Halldorsson , G. K • • 1980:Svartsengl. Athugun a vinnslu jardhita . Orkustofnunreport OS800021/1980 , ROD10 - JH D17 .Mlchaelldes , E. E . • 1981: ThermodynamIc properties of geothermalf l uIds . Geot hermal Resources Counc i lTransactions , S . 361-364 .Stefansson . V. and Steingrimss on, B., 1981 : Geotherma lLogging, An introduction to techniques and i n t erpretation. Orkustofnun report OS800 17/JHD09,Reykjavik , 117 p .Thorhallsson, S.,19 79 : Combined generation of heat andelectricity from a geothermal brine a t Svartseng iin SW-Iceland . Geothermal Resources CouncilTransactions , 3, 733-736 .Vatnaskil, 1982 : Svartsengi . Vatnsbordslaekkun me d vinnslu .Report to Hltaveita Sudurnesja by Vatnaskil Ltd . •July 1982 .