Loading...
HomeMy WebLinkAboutAkutan Geothermal Final AK-3 Well Test Report 2017AKUTAN UNALASKA ISLAND UMNAK ISLAND v Neil McMahon Alaska Energy Authority Grant No. 7040050 Mike Weathers U.S. Department of Energy Grant No. DE-EE0000329.005 AKUTAN I t? UNIMAK ISLAND r-. December 4, 2017 Dear Sirs, The City of Akutan (COA) has received the final AK-3 Well Test Report (copy attached), prepared by our Consultant Geothermal Resource Group (GRG). Based on and analysis of the data derived from the well test, it has been determined that: "the AK-3 well as completed does not have an adequate flow rate to confirm a commercial resource at this well site. The permeability of the well is too low for a long-term test." Therefore, it is the conclusion ofthe COA Project Team that the AK-3 well does not meet the objectives set forth under AEA Grant No. 7040050, Amendment 5; "Identification of a resource sufficient to support 2 MW minimum energy requirements set forth for generation of electrical power". Subsequently, and consistent with the terms set forth in AEA Grant No. 7040050; Amendment 5, Task 8, sub para. b and c; the COA would like to proceed with termination of the grant and concurrent release of $269,500 in withheld project funds. Sincerely, Herman J. Tuna Scanlan D.P.A. Project Principal Investigator Cc Mayor Joe Bereskin Robert E. Kirkman — COA Project Manager Attachment/ Akutan Geothermal Project; Well AK-3 Flow Test Report; August 2017 Geothermal Resource Group, Inc. ��Geothermal Resource Group 77530 Enfield Lane, Building E TAPPING THE EARTHS ENERGY Palm Desert, CA 92211 Phone: 760-341-0186 Akutan Geothermal Project Well AK-3 Flow Test Report, August 2017 Initial flow of steam and brine from AK-3, 28 August 2017 17 November 2017 Page 1 of 30 Introduction Well AK-3 was completed on 8 September 2016 to 1955' with a perforated 3.5" liner from 755' set on bottom (Figure 1). It is the third exploratory geothermal well drilled in Hot Springs Bay Valley as part of an Alaska Energy Authority (AEA) grant funded geothermal exploration program managed by the City of Akutan. The AK-3 drilling and testing program was made possible through additional grant funding allocated by the US Department of Energy (DOE). AK-3 was flow tested in August 2017, with the intent of assessing the power production potential of the outflow resource in Hot Springs Bay Valley, located approximately 3.5 miles from the City of Akutan, with access overland from Akutan Harbor by way of a topographic saddle (Figure 2). During the drilling and testing of this third well, as it has been for most of the activities related to the Akutan Geothermal Project, access to the study area was by helicopter. The object of this production test was to measure the flowing mass rate, pressure, temperature and enthalpy of the well, conduct essential downhole surveys of flowing reservoir temperature and pressure, and collect and analyze geochemical samples representative of the reservoir fluid composition. For the testing program, a test unit was designed and built to be transportable under the 1300 lb. lift capacity constraint of the helicopter that was secured for completion of the project. The test unit consisted of a 4" flow line, James tube drop out spool of two sizes 1.5" and 3", two vertical separator (silencer) tanks and a V-notch weir box for measuring fluid flow. The system was designed in conformance to a typical geothermal well test used to measure the enthalpy and volume of the produced fluid. The unit was sized according to the anticipated fluid volume given the resource temperature and wellbore size. Geothermal Resource Group provided the test equipment, planning and supervision of the well test activities. The City of Akutan provided essential project support including test set up, logistical support, housing, and transportation on the island. Page 2 of 30 (� AK-3 v As-" KW sdkam W perm by OM 30-SWWabw 2016 - Net TO SaM 10-vs, hole to 47 '~-8-5/8- co dudor @ 49 7-VY We to M, —2 5-11T,15.5 ppf K55, VFJ @ 225 —5 �155 `SoP e kyw —750- ( I PQ (4.847 hole to OW, m ~-4-112r, 15.5 ppf, K55, VFJ casing to 80p' -1, -1,5 HO (3.8-)hole to 1,OW, -W-OD, Mark and perfMW 8 ppf sa st I RW iI d i Ii d' -2, Figure l: AK-3 wellbom wnfigumtion. Page 3 of 30 22 A i 05 t 2 3 K:bmeters N e Wei 1 Hot Spnngs Bay Sp"m9a. v"tlll¢_ BWCfYl11 tumarobs NW stoke <y".V" `• p.,ns 9 NSA _ NE stoke S d Y HS B - E SAY" HSC of 0"W �j�CM ._/Nruun HS O Ansylp env Tneent Saloons KS '�, prgef3 Wes ` .... r AK sne flat ��.\. Nbp TG.a 'ht pnl - Fumerde e: e �pne.e Figure 2: Akutan Geothermal project location shown, with wells drilled to date. Modified from Stelling, 2015. Operations The equipment and materials needed for the AK-3 test were shipped by way of Coastal Transportation vessels that left from Seattle on 4 or 11 August. The bulk of the equipment arrived at the Akutan dock on 20 August. The Maritime helicopter pilot looked over the items and determined that it would be prudent to have a specialized pilot for movement and set up of the larger items including the separator tanks, which were manufactured in two pieces for Page 4 of 30 transportation. Maritime mobilized the new pilot, and he arrived on 23 August, along with City advising consultant and logistics manager Robert Kirkman. An operations tent was set up near the new well, and all of the equipment was successfully brought to the site slung under the helicopter. Test Supervisors, Leland Davis and Mary Mann arrived midday on 25 August, were briefed on helicopter safety and began work at the site. The main elements of the test unit were assembled on 25-27 August. A tag run with the well downhole logging unit was made on 27 August, to a depth of 1952'. Shut-in wellhead pressure was 20 psi. On 28 August, a static temperature survey was run to total depth, and then final test preparations were made, a safety briefing was conducted, and the well was opened to flow at 15:45. It took about 5 minutes for the fluid to reach the surface. After a few minutes of flowing, the control valve was shut to repair leaks, then opened back up at about 16:39. The Mayor Joe Bereskin came out to the site by helicopter about 20 minutes into the flowing of the well. The flowline temperature decreased to boiling temperature, indicating that the volume of liquid was flashing to steam before reaching the surface, and there was not enough fluid to fill the weir box. The well was shut in at 17:33, and the data logger was left to monitor well head pressure and temperature overnight. On 29 August, upon return to the well site, the wellhead pressure was 60 psi and line temperature 300°F. A second static pressure and temperature survey was run to TD, and final preparations were made to open the well to flow, including putting the PT tool in the lubricator on the well in preparation for running a flowing survey, and preparing the fluid sampling equipment. The well was opened at 12:40 after a safety meeting. A brine sample and a steam sample were taken through the sampling separator and then one steam sample was taken directly from a flow line pom for ease of sampling. Most of the brine was already flashed to steam when it reached the surface. A flowing pressure and temperature survey started at 13:07, with moderate flow of steam to the separator tanks and the pressure fell gradually until there was no pressure on the flowline. The survey was taken to bottom, back up to the casing bottom at 800' and back down to 1550', where the first static survey indicated may be the top of the reservoir. The flow line was closed at 15:30, with the tool in the hole to measure the build-up pressure and temperature in the well. The tool was pulled out of the hole at 17:30, with a wellhead pressure of 40 psi. The surface pressure and temperature were logged overnight. The test supervisors returned on the afternoon of 30 August, to retrieve the fluid samples, the data logger and to pack the well test equipment for shipping. City personnel completed the disassembly and retrieval of the test equipment to Akutan harbor or to the City on 31 August. The test equipment vessels are stored at Akutan Harbor, with some of the elements preserved in the Akutan warehouse at the City dock. Page 5 of 30 400 Flow Test Results Set-up The flow test equipment consisted of a 4" flow line from the wellhead, line pressure and temperature instrument ports, a 4" flow line valve, sampling ports and a Weber separator, adjacent pressure and temperature instruments, 1.5" James tube and lip pressure instruments, expansion to 12" split line to two silencer tanks, outlet of silencers manifold to V-notch weir box, spilling into a ditch to the test pit dug to accommodate the produced fluids (Figure 3 and 4). Figure 3: Test set-up looking east-southeast. Page 6 of 30 Figure 4: Sump built to hold testing fluids, looking northeast. Pressure Temperature Survey Results The initial pressure -temperature surveys run in AK-3 directly after drilling were used to predict the final heat up temperature for the AK-3 well (Figure 5). These predicted temperatures were exceeded in the static pressure -temperature survey run on 28 August for the upper part of the hole, that was subjected to more cooling during drilling, through the shape of the curve with depth is the same. The static water level was 20' below ground level. The highest temperature in AK-3 was 358°F at 405'. Below that was a roll over in temperature to a low of 303°F at 1600 ft. The temperature increased again to 311*17 by 1950'. Page 7 of 30 Ten gwainim (f) 0 25 s0 75 100 125 150 175 200 225 250 275 300 325 350 375 300._'� 1 700 ow 600 • 9W ! t 1000 ti — - Li 1100 1 1200 1300 — — 1400 -1 ...•••• 2016Temp1- 1500 ........ _ _ 2016 Temp 2 _... _.. 2016Temp 3 . ..... ........__.... 60 10 ............._ ... t ... Sunc 2017 Temp ! 17W Pmst1 -----Press2 lawPreswre 3 19W __ Sulk 2017 Pre6sure t 2000 0 100 200 300 400 500 600 700 800 900 Pressure (psi) Figure 5: Static temperature and pressure surveys from 1248 hours after completion of drilling activities in 2016 and the tirst static survey prior to well testing on 28August 2017. After completion of the pre -flowing static survey, AK-3 was opened to flow for one hour. The flow test did not continue because the flow to surface was too weak to register pressure on the 1.5" James tube lip pressure gauge or fill the weir box, so the test was halted. A second static pressure - temperature survey was run the next day, after about 17 hours shut in, and it shows significant Page 8 of 30 differences from the static survey run before it was flowed (Figure 6). The water level had not yet recovered, and the steam gas pressure of about 70 psi is holding the water level at about 960'. The bottom part of the well, below about 1400' was not affected by the flowing of the well, and the temperature remained constant, indicating that there is no permeability in this part of the well. The well was opened to flow once again, with similar results observed at the surface as were experienced on 28 August. A flowing temperature and pressure survey was begun within minutes of opening the well (Figure 7). This survey shows that the fluid in the well bore begins to turn to steam at about 1260', indicated by the fall in temperature, and unstable pressure measurements. Geothermal fluids will lose pressure as they come up the wellbore and 0 to 100% of the liquid in the stream can flash to steam, causing the fluid to sometimes pass through several flow regimes with different pressure loss and flow properties. If liquid is present in the flow, the velocity must be sufficient to lift the water or liquid fallback will kill the well flow. This is exactly what is seen in AK-3 as the steam rises from 1260', there is not enough velocity from the flow to lift the liquid left in the stream, the density increases, and the liquid fallback kills the flow. The well was logged to bottom, then up to the casing shoe at 800' and then down to 1550' where an isothermal temperature gradient seemed to indicate reservoir depth in the first static logs. During the second two passes through the well, the flashing depth moved down the hole, indicating that there is very little fluid recharging into the well as it flows. After the tool reached the hang depth, the flow line was closed. The well at the tool hanging depth of 1550', recovered 66 psi in two hours (Figure 8) and the temperature remained constant. Page 9 of 30 Tamp M 0 s0 100 130 200 250 300 330 400 0 200 400 bbo B00 loon 1200 1400 1600 Static i Tatperenve -Static 2 Figure 6: Static pressure and temperature surveys, number 1 was done before the well was flowed on 28 August, and numberwas run before the well was opened to flow again on 29 August. Page 10 of 30 Tamp (F) 200 210 220 230 240 230 260 270 230 290 300 310 320 330 0 200 400 600 goo tl 1000 1200 1400 1600 —plowft ��M�pfw�oe Temperadae ]800 2000 0 30 100 130 200 230 300 330 400 pressure (psi) Figure 7: Flowing temperature and pressure log for AK-3. Page 11 of 30 Temp (F) 200 210 220 230 240 250 260 270 280 290 300 310 320 330 0 200 400 600 80D .d= S000 1200 1400 1600 1800 2000 0 SO 100 ISO 20D 250 300 350 400 Pressure (psi) 1 — � I ` i � 1 I ♦ Fbwirg Temperature Down 1 — — Fowing Temperature UPI ...... Flowing Temperature Down 2 \ Flowing Pressure Down 1 ._..,..__ __...._...__... -- Flowing Pressure Up .. I — — —Flowing Pressure Down 2 Figure 8: AK-1 flowing logs 29 August Page 12 of 30 330 310 290 270 250 230 210 190 170 150 15M:36 15:36:00 1510:24 16--04:48 16.19.12 16:33:36 16:49:00 17:02:24 1736:48 17:31:12 37.45:36 Time ve ......... . ..... -- --------- - - - Figure 9: Pressure and temperature build up after flowing, hanging at 1550'. Page 13 of 30 Data Collection and Fluid Sampling Surface data collection Pressure and temperature data were logged at the surface at the flowline close to the wellhead. There was instrumentation to collect the James Tube lip pressure, but this measurement is not reliable because of the low flow rate conditions (<1 psi). Prior to flowing, the wellhead pressure was 20 psi. When flowing, the pressure rose to about 60 psi and 303°F, then fell within a matter of minutes to almost 0 psi and 212 F. When shut in, the surface pressure rises to about 60 psi, before starting to fall back to the expected 20 psi static wellhead pressure (Figure 10). 70 60 so n 40 w 30 a 20 10 -- Well 'ng 0 12:00:00 13.12:00 14:24:00 15:36:00 16:487.00 1&00:00 19:12:00 20:24:00 21:36:00 Time Fieurc 10: Line pressure data collected at the surface during the flow and recovery of AK-3 on 29 August. AK-3 Sampling Well AK-3 was targeted to explore for a deeper, hotter reservoir below the known permeable zone in AK-2, encountered at approximately—500-800 ft bgs. The well was cased to 800 ft bgs and completed with slotted liner to 1955 ft depth. No major permeability was encountered during drilling below the cemented casing, though some relatively small mud loss zones were indicated. The PT tests from 28 and 29 August 2017 indicate the temperature below the casing (800 ft below ground surface (bgs)) is >_240°F. The goal for the AK-3 sampling was to collect a two-phase sample, with separated brine and steam from the two-phase flow line, according to industry standards (ASTM, E 1675-95, 1995). 105 Page 14 of 30 Preparations were made to collect the first set of geochemical samples from AK-3 after flowing the well for several hours on 28 August. Planned sampling included geothermal brine, steam condensate, noncondensible gas (NCG), and helium isotopes for laboratory analyses, as well as field measurements of electrical conductivity, pH, and the ratio of NCG to steam in the production flow. Shortly after opening the well, the wellhead pressure, and line pressure and temperature began to quickly decline indicating that the well was not sustaining consistent flow. Before an initial geochemical sample could be collected, flow from the well principally ceased, with only a small amount of steam continuing to release from the well. Additionally, after an initial surge, brine did not consistently flow to the weir box, with the fluid level never reaching higher than —2-3 inches below the weir notch, meaning only flashed steam was discharging from the wellbore. A second attempt at geochemical sampling was made on 29 August. Samples were collected immediately after opening the well to guarantee some fluid collection. While this technique does not follow ASTM Standards for stable sampling of reservoir fluids, the well status dictated that the sample could not be collected under ideal conditions. Samples were primarily collected from a sampling separator located along the flow line, downstream of the throttle valve. The sampling separator was connected with a 3-port configuration —top, side, and bottom of the flow line (Figure 11). The sample port lines were flushed prior to connecting the sampling separator. The separator was operated by matching separator pressure (SP) with line pressure (LP) (measured directly upstream of the sample ports). A brine sample set was collected with only the bottom brine -side sample port open. By the time brine samples were collected, LP/SP had declined to nearly atmospheric pressure (--2.0 psig). Fluids were collected and filtered and preserved with nitric acid for major cation analyses, including additional dilution (4:1) for silica (SiO2). Additional sample was collected and left unpreserved for analyses of major anions and stable isotopes (oxygen-18 and deuterium). The brine was measured in the field for pH (8.02), electrical conductivity (-59.3 mV) and sampling temperature (17.00062.6017). Two samples were collected for NCG laboratory analysis. Samples were collected in evacuated glass Giggenbach gas flasks. One sample was collected from the sampling separator and the other directly from the top of the flow line. The first sample (AK-3a) was collected with only the steam - side top port open to the separator. Since flow from the well was so low, there was no risk of brine carry-over on the flow line, therefore the second sample was collected from a sample port located on the top of the flow line just downstream from the wellhead and upstream from the throttle valve. The samples were collected from a condenser coil connected to the steam sampling port on the separator and flow line. Given the low sampling pressures, the flow line sample was likely more representative of the gas and steam concentrations passing through the flow line, therefore, this Page 15 of 30 sample was analyzed by the lab, with the first sample held as a backup. After the results were received and analyzed the lab was given the authorization to dispose of the second sample. Figure 1 l: Leland Davis preparing for gas sample collection with Webre separator attached with 3-port configuration on flow line. In addition to the NCG analyses, a separate sample was collected for helium isotope analysis. This sample was collected using a straight copper tube. After allowing the sample to pass through the tube for —5 minutes, each end was sealed by crimping shut with a special clamp, downstream end first, then upstream to avoid any atmospheric contamination. Due to high cost of laboratory analysis, this sample was initially held during lab analyses for total NCG, in order to first confirm the presence of geothermal gas. After review of NCG lab results, it appears no geothermal gases were present during sampling and therefore it was decided that additional analysis for He isotopes would likely not provide additional value and therefore the sample was discarded. q� Page 16 of 30 NoName Spring A brine sample was collected from the "No Name Spring", which is located --1/4 mile up the valley from AK-3, and has not been previously sampled or analyzed, only located in the field. Discharge temperature was measured at —195OF from multiple seeps over several meters along the mountain stream bank, a tributary to the main valley creek. A basic estimation was made of spring discharge totaling —25-50 gpm at the surface. Samples were collected in clean poly bottles and carried back to camp where some of the sample was filtered and preserved for major cation analyses, while the rest was left unfiltered for major anion and isotope analyses (Table 1). Table 1. Geochemical sam bottle size le record FiIUUnFitl ana nerd measuremcnu Analysissi LP LT SSP FR Cond. T H sample Name Date Time Sa ler fluid rN For 11F OF si mV H OF AK-3A 29-A -17 13:15 L.DaHs Brine 250 F SiO2 2 215 2 - -59.3 8.02 62.6 AK-3A 29-A -17 13:15 T-D-a s Brine 500 F Cations P1. 2 215 2 - -59.3 8.02 62.6 AK-3A 29-A -17 13:15 C--DD s Brine 1000 OF Anions 2 215 2 - -59.3 8.02 82.8 AK-3A 29-A -17 13:15CD—,, Brne 000 OF 678D 2 215 2 - -59.3 8.02 62.6 AK-3A 29- 17 13:25 L.DaNs Gas Ogg OF NCG 1.5 215 1.5 - - - AK-3A 29-A -17 13:30 L.DaHs Gas Co OF Hefum Isota 1.5 215 1.5 - -AK-3B 29-A -17 13:36 L.DaNs Gas a OF NCG . 1 212 - - Laboratory Analyses Noncondensible Gas Noncondensible gas (NCG) sampling was conducted at or very near the collapse of active now from AK-3. As such, the samples were not collected under ideal conditions. Initially a sample was collected from the 2-phase sampling separator at nearly atmospheric conditions with no measurable brine discharge from the well. A second sample, which was subsequently collected directly from a sample port on the top of the flow line immediately downstream of the wellhead was the sample that was eventually analyzed by the laboratory. The first sample was temporarily held and ultimately disposed of after laboratory results of the second sample confirmed that no significantly measurable geothermal gases were discharging from the well at the time of sampling. Noncondensible gas samples were sent to Thermochem Laboratories in Santa Rosa, CA for Total NCG analysis. Analytes included: • Argon (Ar), oxygen (02), nitrogen (N2), methane (CH4), carbon dioxide (CO2), ammonia (NH3), hydrogen sulfide (H2S), and Total NCG (g/s ratio) The gas laboratory analysis for AK-3a is presented in Table 2. The sample is primarily air (97.31/o), due to sampling occuring after flow collapsed. Even if the sample were to be corrected for air based on oxygen, high nitrogen (N2) (30.3%) indicates the sample contains even more air contamination than reported. Page 17 of 30 W Reported Gas Data Well/Site CO Z HzS NH3 At NZ CH, Hz Dry Gas % by Volume AK-3A 51.4 1 <252.0 13.0 1 0.1 1 30.3 3.1 2.2 Brine AK-3 brine samples were sent to Western Environmental Testing Laboratory (WETLab) in Reno, NV for major anion and cation analyses (Table 2). Major Anions unfiltered/un reserved : Major Cations (filtered/preserved): Stable Isotopes unfiltered/un reserved : pH • Arsenic • Lithium • Oxygen-18 • TDS • Barium • Magnesium • Deuterium • Specific Conductivity • Boron • Manganese • Alkalinity • Calcium • Potassium (Carb/Bicarb/Hydroxide) • Chromium • Sodium • Chloride • Iron • Strontium • Fluoride • Lead • Zinc • Sulfate The laboratory analyses for AK-3a and No Name Spring are presented in Table 3. Although the reservoir temperature of AK-3 is not known, it is clear that the fluid boiled before discharge. Therefore, an attempt was made to correct for steam loss based on the minimum source temperature of the fluid. Assuming the fluid came from below the casing shoe at 800', as would be expected, the minimum temperature is approximately 240aF (116'Q (indicated by the flowing PT log in Figure 8). The result of the steam loss correction for AK-3 samples from 240OF (116'Q to surface conditions 212a17 (100aC) is presented in Table 4. Table 4. Laboratory analysis for the liquid somples AK-3o and No Name Spring collected 28 August 2017.502forAK-30 is corrected for dilution. LneOlYe{LO, A ported •rl MUweWs" rOd Ib [ 6 U l,� !I M • SN)i 11[O, fae O O h MIN CQ ti1 ^�'• 10t-31e SODA 818 558 330 33 01 0.3e m.5 0.74 0.03 37 U It 33 F 03 OA3 OS tx.0 011 N•OYiM 9L3 7A0 UW 740 71 310 230 leA 3 3.5 33 33D L6 31 1300 QO 0.74 1 LI i c1.0 1 037 Page 18 of 30 Table S. Fluid data correcting for steam loss for AK-3a assuming a reservoir temperature of 2407.. Brine Data Corrected for Steam Loss Sampling Temp. lab pH TDS No K G Lf Mg Sr I As B SIO, NCO, SO. I tl I F I rb I NH4 Mn Well eraturc oc m9ft AK-3A 100 1 1128 1 541 1 301 1 21.4 1 66.0 1 0.37 1 0.02 0.7 1 0.03 1 35.9 1 13.2 17.5 22.3 1 62.1 1 0.32 1 0.02 1 0.50 1 0.12 The water from AK-3 does not appear to be representative of geothermal fluid. The low chloride (Cl) and Si02 content of the sample from AK-3, as well as available isotope analysis indicates the fluid is not in equilibrium with rocks at elevated reservoir temperatures. At 240°F (minimum subsurface source temperature), Si02 concentrations are estimated at 66 mg/kg in a liquid in equilibrium with quartz (all other forms of silica would produce higher concentrations); measured Si02 was 14 mg/kg. Correcting this measured value for steam loss, the silica in AK-3a liquid would be 13.2 mg/kg—much lower than the estimated equilibrium concentration and any other thermal waters in the area (Table 5 and Table ). Furthermore, the results indicate low Cl, and previous work indicates that the thermal waters in the area are primarily sodium (Na)-CI brines (Stelling et al., 2015; USGS, 2012). Meteoric water is relatively low in magnesium (Mg) and bicarbonate (HCO3) in AK-3a, suggesting the sample is not meteoric water. Comparison of 2017 samples and previous fluid chemistry The AK-3a and No Name Spring samples are compared to publicly available fluid geochemical analyses for regional springs (HS A3, HS BI, HS C4, HS D2, HS E, Fum Spring), well discharge samples (TG-2, TG-4), fumarole steam condensate samples (AGP-Fum), and meteoric water samples (MW 8, MW 9, MW 10, MW 16) to better characterize the geochemistry of the 2017 samples. The results of the laboratory analysis, and fluid chemistry data for the other regional waters obtained from published data (Stelling et al., 2015 and USGS, 2012), are presented in Table 5. Some of lab results from the previous samples need to be used with caution. For example, the samples TG-2- 254m and TG-4 - 500m, were possibly mixed with drilling fluids, as these samples were air lifted just after drilling. Depending on the source of mixtures, though the absolute concentrations are unreliable, the relative ratios of chemical constituents can be used to determine reservoir conditions (Stelling et al., 2015). Fum-Sp sample appears to be dominated by steam condensate, and is not indicative of thermal water. HS-E is dominated by sea water, and also not representative of a regional geothermal system. by Page 19 of 30 Table 5. Laboratory analysis of AK-3 (sample AK-3a) and No Name Spring. SM for AK3A is corrected for dilution. The publicly available data for regional waters HS A3, HS Bl, HS C4, HS D2, HS E, Fum Sp, TG-2 and TG-4 included from Stelling et al. (2015); the cold water samples AKU12-15, AKU12-16, AKU12-18, AKU12-19, AKUl2-20, AKU12.21 included from .SfT GSScientitic Investigations Keport tuts-523t 26U).-,_..-.... - aeooded ado4 oar4wea/Sft IUMV„ TDs U. rt4 a ca U BSl Sr Ae B Sah - SO, a F_ re M14 a1, Mn .c ma/ke AK-3A 300.0 &28 S56 330 22 6B 036 <OS 0.74 0A3 37 14 10 23 64 03 DAl 0.5 <3A 0.12 NO Name SP;k 92.2 7.40 2400 740 73 110 _ Z.30 - _ 19A - 1 15 31 120 86 31 .-120D «0 0.74 it <IA M37 .--. NS A3.._.. - _..7...._ 840 7.00 - -_ 323 26 12 1.30 0.9 Il _ LIS 172 43 420 1.1 S A3 _ 84.0 328 26 _ 12 1.20 1.0 II ]35 41 410 0.9 S a_1 67.4 6.40 172 16 15 0.61 15 5.9 ]03 116 72 330 0.6 i C4 73A 650 207 16 1e 0.61 1.6 7 133 116 43 2B0 IA _ S DZ _ 59.8 6A0 128 9 11 034 12.0 3A 91 129 26 340 0.9 HSE 67.0 7.30 1660 N 130 3.30 310A 4S 121 161 49S 3M0 OS,msp- __- 2, 178m 1820 - 6.72LZ - 2, 151m [AKU12-15- 1710 7.95 676 16 292 103 03 27 26 70 IM 4, Seem 1610 6.11 420 215 Z751 0.7S 17 6 37 22 32 5745 0.7 O S.3 7.49 42 3 0 5 <OA03 0.9 7 16 6 4 OD OA4 105 6.66 172 i3 3 12 am 2.0 21 35 7 58 DA 0357U12-10 2U12.16 59 7.02 47 4 0 6 m.001 1A a 26 5 4 OA 0.03 0.02 U33-19 14 4 1 1 <OA01 05 0 1 7 0.00.03 0.03 U12-26 5.7 6.71 53 9 1 10 2.1 35 3 13 0.0 0.33 O.Od 6 12 0.1 U12-21 The cations from AK-3a are relatively similar to the other thermal water samples; however, the absolute anion concentrations in AK-3a are much lower producing a high negative charge imbalance (Table 6). The laboratory repeated both the anion and cation analysis for the AK-3a sample and was not able to resolve the missing anions. The charge balance for No Name Spring appears reasonable. Since the brine sample from AK-3a and the water sample from the spring were both collected and analyzed using the same methods, this does not appear to be a sampling or analytical issue. Table 6. Cation and anion balance for 3a and No Name S rin based on data collected durin 28 Au ust 2017 sampling. Sample I Sum of Cations Sum of Anions ICharge Balance 4-3A 1 17.521 2.601 0.74 lNoNamespringl 41.481 35.911 0.07 Most NaCl geothermal brines have a nearly 1:1 mole ratio (or weight ratio correcting for the different molecular weight) of Na:CI. A comparison of Na-CI ratio of the Akutan thermal waters further indicates AK-3a as is an outlier. The Cl concentration in AK-3a is low, and it does not plot near the standard weight ratio line for NaCl brines indicating possible sample contamination. As expected, TG-4, one of the TG-2 samples and HS-E are also outliers, possibly due to sample Page 20 of 30 contamination from drilling fluids and sea water, respectively. However, the other samples all lie close the standard weight ratio line Figure 12 for NaCl brines. i z000 • AK-3A i ■ No Name low Spring ■ HS A3 16W ■ HS 61 ■ HS C4 1400 ■ HS D2 ■ HSE 1200 ■ Fum Sp 000 • TG-2 • TG4 POD % Meteoric Water GOD -----WetgM Patio 400 200 i 0 0 1000 2000 G OOfg) 4000 S000 6000 10, 0 Q i i r i i� r i i i f r i t , r Figure 12. Cl-Na weight ratio comparison plot. AK-3a (low Cl) TG-4 and one of TG-2 (drilling mud contaminated) and HS E (sea water contaminated) are potential outliers. The dashed line represents the weight equivalent of a 1:1 mole ratio. Fluid Chemistry The fluid chemistry is further characterized by plotting results on anion and cation trilinear diagrams. Trilinear diagrams in Figure 13 and 14 allow for comparative ratios between anions or cations, which help to indicate dominant constituents. The regional fluids displayed a range in water type. The fluids from No Name Spring, and samples from TG-2 and TG-4 are neutral to near -neutral Na-Ca-CI type, but AK-3a is less concentrated in Cl than some of the other fluids. The cation trilinear diagram indicates relative cation concentrations for AK-3a are consistent with the other waters. No Name Spring has notably distinct cation concentrations, especially calcium (Ca) than the other hot springs (dominated by potassium (K)) but is more similar to TG-2. Page 21 of 30 The anion trilinear diagram indicates relative concentrations of anions in AK-3a are unique; the absolute anion concentrations are significantly lower (Figure 14). The Fumarole Spring (Fum Spring) is acidic Ca-SO4 type fluid, and appears to be a steam -heated water. The anion concentrations of No Name Spring are also more similar to TG-2 than to the other hot springs. ♦ AK-3A sgR(Ca) a No Name Spring o HS Spring ---- 90 aFum Spring oTG-2 ...._.._jt.,._... 80 �. O oTG-4 ��--* 70 rMeteoric Water .......... `—........ 80 �j 50 K1100 Na1100 10 20 30 40 50 60 70 80 90 Figure 13. Sodium (Na)-Calcium (Ca)- Potassium (K) cation trilinear diagram comparing relative sample ratios. Water samples contain a range of Na-Ca-K ratios, either having high mtios of Ca or Na. Page 22 of 30 0 CI a AFC-3A 10 20 30 40 50 80 70 80 90 o Name Spring S Spring um Spring G-2 G-4 leteonc Water HCO3 Figure 14. Bicarbonate (HCO3)-Chloride (Cl)- Sulfate (SO4) anion trilinear diagram comparing relative sample ratios. The water samples display a variety in dominant anion. Isotope Analysis The isotopes are plotted in comparison to the Global Meteoric Water Line (MWL), and the Aleutian -Alaska Peninsula MWL, which both approximately have the same linear trend, to assess fluid relationships (Figure 15). The AGP-Fum and HS Spring have similar 5 Oxygen-18/8 Deuterium, suggesting limited water -rock interaction. AK-3a and the Fum Spring show elevated S Oxygen-I8 and S Deuterium than typical meteoric waters. Elevated S Oxygen-18 and S Deuterium in AK-3a compared to the meteoric waters further indicates AK-3a is not a meteoric water. AK-3a and Fum Spring could be original meteoric waters which boiled and lost steam at approximately 350-400eF (180-2001C); the samples reflect the water that remained behind (Fum Spring boiled more than AK-3a). Page 23 of 30 -- -- Akutan-Stable Isotopes o .20 r' Aso Qec _ k as _30 ....Steam 220"C Lao[ �pyo�o�. -40 ^ taoc � a -so � ioo c i' X ! -so -16.00 -14.00 -12A0 -10A0 i.00 4.00 .4.00 -2.00 0A0 2A0 b D�� ♦ AK3A • AGP-Fum © HSSpririg ■ Fum Sprig # Metedic Water GlobalMWL — Aleutian-AlasWn Peninsula MWL Figure 15. Available stable isotopes for AK-3a, AGP-Fum, HS Springs, Fum Spring, and regional meteoric water samples. Higher 80-18 than meteoric water concentration is indicative of water rock interaction at higher temperatures Mixing Plots Mixing plots were created to evaluate mixing trends between the fluids in the Akutan region. These mixing plots compare conservative constituents of thermal and non -thermal waters and measured temperatures. Assuming that the meteoric water is dilute and cold, a linear relationship of warm to hot springs and well waters indicates that the relationship between the well waters and surface waters is simple mixing of cold water and thermal waters. The mixing plot comparing SiO2 and measured temperature indicates some sort of mixing trend between the thermal and non -thermal waters (Figure 20). The linear relationships of Cl and Na in mixing plots between meteoric water, hot springs and TG-2 (Figure 16 and Figure 17) indicate that the hot springs in Hot Springs Bay Valley are primarily the result of mixing of shallow thermal water in TG-2 and meteoric water. The No Name Spring is an outlier of the general mixing trend between the other hot springs and meteoric water. No Name Spring has the Na and Cl concentrations of the deep water, but appears to have been conductively cooled on the way to the surface as indicated by the lower silica concentrations. In addition, the cation chemistry of No Name Spring, although immature, suggests elevated temperatures (Na/K geothermometer reservoir estimates between approximately 375-420°F (-190-215°C). Page 24 of 30 N AK-3a is not representative of regional geothermal waters, as indicated by the reported chemistry and stable isotope analysis previously discussed and therefore not included in mixing trend plots. Figure 16. Mixing plot comparing CI versus Si02. HS-E, Fum Sp, TG-2 and TG-4, which do not appear to be representative of the same mixing trend, are not included. The chemistry of No Name Spring is consistent with a thermal out -flow at or above the temperatures observed in TG-2. Assuming the volcano is the source of higher temperature fluids, the chemistry of this spring suggests that it is discharging from a source that was hotter, maybe greater than 400°F (>200°C), and conductively cooled. This conductive cooling could have occurred during lateral outflow. Page 25 of 30 EHSA3 700 W- EKS51 MHSC4 GOD ■HS D2 •TG-2 Soo xmaeodcwarer 1400 3300 20D IOD L4 0 On 200 40LO 60.0 Rao 100.0 I"D 140A 16DD to" 200D Temperature (*C) Figure 17. Mixing is shown between all samples wide from No Name Spring. No Name Spring has similar concentrations of Na as TG-2 (178m), indicating that both are composed of the same geothermal fluid, but the fluid at No Name Spring has undergone conductive cooling. HS E was omitted as it had mixed with seawater, resulting in elevated Na concentrations TG-2 (254m) TG4 chemistry is also likely not repetitive, and therefore not included. Page 26 of 30 q913:0s� P Figure 18. Mixing is shown between all samples aside from No Name Spring. No Name Spring has similar concentrations of C I as TG-2 (178m), indicating that both are composed of the same geothermal fluid, but the fluid at No Name Spring has undergone conductive cooling. HS E was omitted as it had mixed with seawater, resulting in elevated Cl concentrations. TG-2 (254m) TO-4 chemistry is also likely not repetitive, and therefore not included. Geothermometers Based on the data reported in Stelling et al. (2015) the reservoir temperature was estimated to be 450°C (230°C) based on the Na-K-Mg geothermometer applied to the most representative sample from TG-2. Although there appears to be a mixing line between TG-2 and the hot springs (Figure 19) which includes the No Name spring, other assessments indicate that the fluid from this spring is not on a simple mixing line. Na/K geothermometers for TG-2, the sample possibly closest to a geothermal source water, provide a reservoir temperature estimate of approximately 375-450T (190-230°C). The Na/K geothermometer applied to the No Name spring cation chemistry is almost identical at 361-444T (183-229°C). Because the Na/K ratio is not affected by mixing, it appears the thermal source for the No Name spring discharge fluids could be similar to the thermal source of TG-2. b� Page 27 of 30 Na 10K o TG-2 ■No Name Spring OHS Spring MFum Spring %Meteoric Water sgrgmorl000 Figure 19. Na-K-Mg trilinear diagram used to calculate Na-K-Mg geothermometer. No Name Spring confirms the potential mixing line along the 220-2300C Na-K-Mg geothermometer line reported in Stelling et al. (2015). As indicators of subsurface temperatures, SiO2-based geothermometers are reduced by mixing, especially when mixed with cold meteoric water in hot and warm spring samples. A plot of SiO2 versus measured temperature for available hot spring and well data was used to develop a trendline beyond the available data to predict the concentration of SiO2 in equilibrium with quartz in the reservoir, and provide an approximate estimate of reservoir temperature. The extrapolated trendline indicates silica is approximately 343mg/kg in the reservoir, which suggests a reservoir temperature of about 430eF (-220°C) by applying the quartz conductive cooling geothermometer to the hot end -member concentration (Figure 20). The quartz geothermometer applied to the silica concentration TG-2 sample from 584 ft (178 m) which is most directly representative of subsurface conditions (measured temperature 360°F (182°C); Stelling et al., 2015) indicated a temperature of 356°F (—I80°C). Both the cation geothermometers and silica mixing diagram suggest that the temperatures of the thermal source of TG-2 and No Name Spring could be above 392°F (200°C) possibly as high as 446°F (230°C). If the No Name spring is diluted to achieve the current discharge chemistry, as Page 28 of 30 ,05 suggested by the cation trilinear diagram (Figure 13), it could be that it was originally a more concentrated brine than TG-2. --------- --- -- -- - - --...._..- 450 ■tb Name SWtaa ■HSA3 400 ■M591 ■ HS C4 ■HSD2 ■HSE 300 --.___ — ■F=50 •TG-2 25D tl Meteoric Water —__— I •' W-OAM4 •' ISO -------- ----- ------------- Cie ■ ■ .■' eta Sa w 0 00 Sao IUOA 15010 200A 250.0 Temperature rC) Figure 20. The dissolved silica concentration of the various hot springs and cold meteoric water is plotted, yielding a mixing line of geothermal and non -geothermal fluids, shown as the blue line. Extrapolation of this line can be used to determine the temperature of geothermal fluid before mining. The silica saturation curve is plotted in black, and intercepts the mixing line at a silica concentration of 343mg/kg. This silica concentration corresponds with a reservoir temperature of 428OF (220 °C) using the Quartz conductive cooling geothemometer model (Fournier 1982). Geochemistry Summary The lack of evidence for equilibrium with rock at elevated temperatures, negative charge imbalance, dissimilarity with other thermal waters in the area suggest the water sample from AK-3a is not representative of a subsurface thermal aquifer and therefore evaluation of this water does not provide evidence of an underlying geothermal reservoir. Low Mg and HCO3, as well as higher S Oxygen-18 and S Deuterium than typical meteoric waters, indicate AK-3a is also not a meteoric water. No Name Spring has chemistry of a geothermal fluid similar to TG-2, but may have been a more concentrated NaCl brine. Na/K, Na-K-Mg diagram and the silica mixing diagram all suggest that the source fluids of the Akutan thermal waters could be above 392°F (200 °C). No Name spring Page 29 of 30 appears to be conductively cooled. Mixing plots indicate the other hot springs (besides HS-E) are cooled through mixing with meteoric waters. The mixing for the other hot springs could have occurred in the immediate area of the springs. No Name Spring does not follow the same mixing trend as the other hot springs, suggesting cooling by conduction could occur as a result of lateral outflow from the geothermal source which could be greater than 392°C (>200 °C). Results and Recommendations The AK-3 well as completed does not have an adequate flow rate to confirm a commercial resource at this well site. The permeability of the well is too low for a long-term test. Temperatures of the well are adequate for commercial electricity generation, but subsequent drilling or perforation of the shallow zone will be needed confirm the adequacy of the resource. This shallow resource seems to correlate with the same zone that was seen in the nearby Hot Springs Well TG-2 which may help constrain the size of the shallow resource. Aside from electricity generation, the shallow resource near the hot springs may have other uses, such as for heating, greenhouses or other industrial applications, a feasibility study would have to be completed to pursue any option. The new chemistry data from the well was inconclusive, but the sample from the NoName Spring, indicates that resource temperatures could be over 392°C (>200 °C). Given the amount of study that this resource has received since the beginning of the program in 2009, GRG recommends, completing an update of the Conceptual Model of the hydrothermal system that was developed in 2010 (with updates following drilling and subsequent exploration data in 2012). GRG recommends that the latest update focus on the Hot Springs Bay Valley resource area. The conceptual model update report should include an estimated and statistically modified megawatt potential of the shallow resource at HSBV, and a reassessment of the potential for a deeper resource being present, from a relatively accessible surface location near the valley floor. Page 30 of 30 105