sign in sign up
<<
Home , Anhydrite, Salton Trough

PROCEEDINGS, Thirteenth Workshop on Geothermal Reservoir Engineering Stanford Universily, Stanford, , January 19-21, 1988 SGP-TR-113

Evolution Of The Thermal Cap In Two Wells From The Geothermal System, California

Joseph N. Moore and Michael C. Adams

University of Utah Research Institute Salt Lake City, Utah

ABSTRACT terized the upper parts of the geothermal field. Furthermore, because there is The Salton Sea geothermal system is evidence that conditions within the geother- overlain by a thermal cap of low permeability mal system have changed with time (Skinner rocks that restricts the upward movement of et a1 ., 1967; Huang, 1977; and the high-temperature reservoir . McKi bben, 1987), the present borehole Petrographic and fluid inclusion data from temperatures and fluids may not accurately two wells show that the thermal cap in the reflect the conditions associated with the southern part of the field consists of an alteration of the shallow sediments. upper layer of lacustrine and Alternatively, this data can be obtained deposits with low initial permeabilities and from fluid inclusions contained within the a lower layer of deltaic . The authigenic . In this paper we sandstones were incorporated into the thermal describe the results of fluid inclusion and cap as downward percol ating fluids deposited petrographic studies of two boreholes anhydrite and in the pore space of drilled in the southern part of the field the rocks, reducing their permeabilities. (Fig. 1). The samples were provided by During development of the thermal cap, base- Unocal . metal , potassium feldspar and quartz veins were deposited by brines from higher temperature portions of the system. LITHOLOGIC AND MINERALOGIC RELATIONSHIPS INTRODUCTION The rocks within the Salton Sea The Salton Sea geothermal system is geothermal system can be divided into an located near the center of the sediment upper sequence of lacustrine claystone and filled Salton Trough of evaporite deposits and a lower sequence of and . assemblages and textures deltaic sandstones, siltstones and shales formed in response to geothermal activity can (Randall, 1974). Figure 2 details the be categorized as diagenetic or metamorphic 1 ithologies and mineral assemblages found in (McKibben and Elders, 1985). Diagenetic the upper parts of the wells studied by us. processes occur at temperatures of less than These lithologic columns are based on an about 25OoC and have resulted in the recrys- examination of dri11 chip samples col 1ected tallization of the sheet silicates, and the at 6 or 9 m (20 or 30 foot) intervals. deposition of anhydrite, carbonates, and The primary and secondary minerals we sulfides in the pore space of the sediments. observed (Fig. 2) are similar to those found The reduction in the permeabilities of the in other parts of the field (e. g. Muffler shallow sediments caused by the deposition of and White, 1969; McDowell and Elders, 1979, these pore filling minerals has led to the 1980, 1983). With increasing depth, the development of a thick thermal cap, or zone secondary assembl ages are characterized by: of conductive heating, over the reservoir 1) anhydrite, calcite, interlayered il- (Younker et a1 ., 1982). Metamorphic proces- lite/smectite, and interlayered chlorite/- ses occurring in the reservojr., where smectite (interlayered illite/smectite zone: temperatures range from 250° to 365OC, have 0-222 m, Well A; 0-241 m, Well B); 2) led to the development of hornfelsic textures chlorite, illite, interlayered chlorite- and mineral assemblages typical of the / smect ite , cal cite, an hydri te , potass ium greenschist facies (Muffler and White, 1969). feldspar, quartz, sphene, and base-metal Fluids encountered at the depths where these sulfides (chlorite-calcite zone: 222-616 m, metamorphic processes are occurring have Well A; 241-789 m, Well B); and 3) epidote, salinities of up to 25 weight percent (Hel- calcite, chlorite, quartz, potassium feldsp- geson, 1968). ar, albitic plagioclase, anhydrite, illite, In contrast to the deep portions of the and sphene (chlorite-epidote-calcite zone: Salton Sea geothermal system, little data 616-1646 m, we1 1 A; 789-1292 m, We1 1 B). has been presented on the temperatures and Traces of and hematite occur through- salinites of the brines that have charac- out these zones.

-107- FLUID INCLUSION DATA described by Somner et a1 . (1985). Methane, ethane, propane and carbon dioxide were Fluid inclusions in anhydrite, sphaler- detected in primary inclusions in sphalerite ite and vein quartz were studied. In the and quartz. Inclusions in anhydrite con- lacustrine deposits, anhydrite occurs as tained mainly carbon dioxide and methane. rosettes in the claystones and as aggregates Anhydrite suitable for fluid inclusion associated with calcite in the evaporite measurements typically occurs as beds. In the deltaic deposits, anhydrite and fragments up to 2 mn across. Typically calcite the sandstones above 387 m in several generations of secondary or pseudo- well A and 351 m in well B. At greater secondary inclusions are present. Primary depths, anhydrite-rich zones are restricted inclusions containing a small (<1micron), to a few narrow intervals. unidentified birefringent daughter mineral Sphalerite occurs in significant were observed in samples from 168-177 m in amounts (up to several percent) as a cement well B. These inclusions form three dimen- in the sandstones between 341 and 387 m in sional arrays that define growth zones in well A and from 277 to 360 m and 424 to 433 m the cleavage fragments. in well B. In places, the sphalerite The homogenization temperatures of contains inclusions of anhydrite, suggesting fluid inclusions in anhydrite are plotted that it has replaced preexisting anhydrite against depth in Figure 3A. The data show cement. Rocks between depths of 300 and 600 that irrespective of the origin of the m also commonly contain secondary potassium inclusions in these samples, homogenization feldspar after detrital plagioclase. The temperatures vary little within each depth significance of these mineral assemblages is interval, and that the two wells record discussed bel ow. simi 1ar thermal profiles. No pressure Quartz veins containing minor barite, correction has been appl ied to these data chlorite, hematite, and traces of , because of the shallow depths. sphalerite, calcite and anhydrite occur Ice melting temperatures of fluid between depths of 305 and 378 m in well B inclusions in anhydrite (Fig. 38) range from (Fig. 2). Barite, in particular, is an -21.4O to -4.7OC. Liquid was observed in uncommon mineral in the Salton Sea geothermal some of the larger inclusions at tempera- system. In these veins, it is found as tures near -5OOC (first melting), indicating embayed s surrounded by quartz. that the inclusion fluids are enriched in The petrographic relationships suggest CaCl . For reference, melting point that the deposition of anhydrite began prior deprgssions ranging from 4.7O to 20.8OC to both sphalerite cementation and quartz correspond to salinities of 7.4 to 23.2 veining. Thus, fluid inclusions in anhydrite equi Val ent weight percent NaCl (Potter et can potentially provide information on the a1 ., 1978). temperatures and compositions of the brines Fluid inclusions in sphalerite suitable since the early evolution of the thermal for freezing and heating measurements were system. As discussed below, inclusions in found in a few from two depth sphalerite and vein quartz provide informa- intervals in well B. These inclusions are tion associated with the episodic upwelling primary, forming small isolated groups or of brines from deeper parts of the system. three-dimensional arrays that para1 le1 color Most of the fluid inclusion data from the banding. The inclusions yielded homogeniza- Salton Sea geothermal system is of this tion temperatures ranging from 179O to 223OC latter type (Huang, 1977; Freckman, 1978; (Fig. 3A), and ice melting temperatures Andes and McKibben, 1987; Roedder and Howard, (Fig. 38) ranging from -12.0 to -lO.O°C 1987). (16.0 to 14.0 equivalent weight percent All the fluid inclusions examined in NaCl ) . this study were liquid-rich and contained a Data were obtained on primary (refer to small vapor bubble that occupied about 20 Figs. 3A and B) and secondary inclusions of percent of the inclusion volume at room vein quartz. The primary inclusions occur temperature. Most had maximum dimensions in in three dimensional arrays that define the range of 2 to 10 microns. No evidence of growth zones. Homogenization temperatures a separate gas phase was observed in these of these inclusions range from 194O to 242OC inclusions and, with the exception of (Fig. 3A). Because of their small size (1 anhydrite from 168-177 m in'well B, none of to 3 microns) only a few freezing point the inclusions found contained any solid depressions were measured. The ice melting phases. In order to assure the validity of temperatures ranged from -13.9O to -9.2OC our data, fluid inclusions in anhydrite from (17.8 to 13.1 equivalent weight percent three depth intervals were systematically NaCl ) . overheated to determine their stretching Secondary incl usions have homogenization characteristics. These experiments demon- temperatures ranging from 189O to 21OoC and strated that stretching caused by 1 aboratory ice melting temperatures that vary from heating is not likely to have affected the -16.5O to -0.7OC (20.0 to 1.2 equivalent results reported in this study. weight percent NaC1). The high-sal inity In addition, the gas contents of in- inclusions have freezing point depressions clusions in ten samples were analyzed by similar to the brines trapped in anhydrite at mass spectrometry using the techniques these depths and thus may represent locally

-108- derived fluids. The low-sal inity fluids may deltaic rocks must have resulted from the represent condensate mixed with a small influx of brines derived from higher- amount of the higher salinity brines. temperature portions of the geothermal system. These fluids would also have been a DISCUSSION likely source of zinc, lead, and copper. As discussed above, our observations The data described above suggest that suggest that the sphalerite cement found in the evolution of the rocks in the shallow the deltaic sandstones has replaced pre- portions of the of the two wells studied existing anhydrite cement. Reduction of the involved four distinct processes. These sulfate to could have occurred by include: 1) the formation of secondary several different mechanisms. However, mineral s through low- to moderate-temperature reduction by pre-existing sulfides or isochemical reactions; 2) the progressive anerobic bacteria is not likely in this downward cementation of the sandstones by case. Alternatively, the sulfate could have anhydrite and calcite; 3) the influx of been reduced by hydrocarbons carried in metal-bearing brines; and 4) the development solution by the upwelling brines according of permeability. The initial to the reaction (Anderson, 1983): heating of the sediments is reflected in the formation of anhydrite after and the anhydrite + CH4 + Zn+2 recrystallization of the sheet silicates as discussed by Muffler and White (1969). In = sphalerite + Ca+2 + COP + 2 H20 contrast, the anhydrite + calcite cement in the deltaic sandstones could not have formed We infer from the widespread occurrence from the fluids or detrital components of sphalerite cement that matrix per- originally present in these rocks. However, meabilities were relatively high during its the fluids in the overlying lacustrine rocks deposition. In contrast, the formation of would have been enriched in both sulfate and veins implies lower matrix permeabilities bicarbonate. Because calcite and anhydrite and fracture-dominated fluid flow (Elders, display retrograde solubilities, heating of 1979). The occurrance of barite in these downward moving sulfate and carbonate ions veins suggests that some mixing of the pore would have resulted in the depositon of these and vein fluids occurred. As noted by minerals in the sandstones imnediately Barnes (1983) in his discussion of underlying the lacustrine deposits. With Mississippi Valley lead-zinc deposits, the time, continued diffusion of sulfate and car- deposition of barite would most likely occur bonate would have led to a progressive where -bearing solutions encountered a thickening of the anhydrite + calcite sulfate-rich groundwater or where oxidation cemented sandstones. of sulfide to sulfate in solution can occur. The fluid inclusions in anhydrite record As discussed above, sulphate-bearing pore temperature profiles characterized by high fluids must have been present in the gradients through the upper 323 m in well A sandstones in the upper part of the deltaic and 341 m in well B. Using the average section and could have mixed with the rising homogenization temperature of 247OC for vein fluids. Later, these fluids would have samples from a depth of 332-341 m in well B, been purged from the veins as upward flow and a mean annual air temperature of 22.5OC continued, causing dissolution of early (Imperial Irrigation District, 1982), the barite and deposition of sulfides, hematite calculated average gradient to this depth is and additional quartz from further cooling. 0.66OC/m. Younker et a1 . (1982) have shown that such high gradients are indicative of heat transfer by conduction through rocks of SUMMARY AND CONCLUSIONS low permeability. This calculated gradient and the mineralogic re1ationshi ps suggest Fluid inclusion data and petrographic that the thermal cap extends to the base of studies have been used to reconstruct the the anhydrite + calcite cemented sandstones. evolution of the rocks in the shallow In contrast, homogenization temperatures of portions of two wells drilled in the inclusions in well A between 460 and 622 m southern part of the Salton Sea geothermal are nearly constant with depth, implying that system. The initial heating of these rocks heat is transferred through these rocks by resulted in the formation of anhydrite after convection (Younker et a1 ., 1982). gypsum and the formation of secondary The secondary potassium feldspar and phyllosilicat'es. Anhydrite and calcite were sphalerite cement found in the lower part of also deposited directly from the pore fluids the thermal cap could not have been formed as a cement in the underlying deltaic by the fluids that deposited the anhydrite sandstones. Cementation of the sandstones in the sandstones. The replacement of progressed downward, reducing their per- plagioclase by potassium feldspar involved meabilities and leading to a thickening of the addition of K and removal of Na from the the thermal cap with time. Fluid inclusion rocks. Because the K/Na ratio of geothermal homogenization temperatures in anhydrite waters increases with increasing temperature increase systematically with depth through (Fournier and Truesdell, 1973), the forma- the thermal cap. Variations in the salinit- tion of potassium feldspar in the shallow ies of these fluid inclusions demonstrate

-109- that some mixing of downward-percolating Haas, J. L., Jr. (1971) The effect of brines and reservoir fluid took place. salinity on the maximum thermal gradient of However, the velocities of the descending a hydrothermal system at hydrostatic pres- waters were apparantly not high enough to sure. Econ. Geol. 66, 940 - 946. disturb the thermal gradient. The influx of brines from higher Huang, C. I. (1977) Fluid inclusion study of temperature portions of the geothermal system well cuttings from Magmamax No. 2 drillhole, occurred episodically during development of Sal ton Sea geothermal area , Cal ifornia the thermal cap. Movement of these brines (abstract). Econ. Geol’. 72, 730. through the thermal cap resulted in the deposition of potassium feldspar and sphaler- Helgeson, H. C. (1968) Geologic and ther- ite cement and in the formation of quartz modynamic characteristics of the Salton Sea veins. geothermal system. Am. J. Sci. 266, 3, 129- 166. ACKNOWLEDGEMENTS Imperial Irrigation District, Imperial Funding for this project was provided by Valley Annual Weather Sumnary (1982) Monthly the Department of Energy under Contract high, low, and mean temperatures and Number DE-AC03-84SF12196. We would 1 ike rainfall, 1914-1982. Comnunity and Special to thank R. Dondanville and G. Suemnicht of Services, Imperial Irrigation District, El Unocal for providing the samples. M. Somner Centro, CA. 19 pp. performed the gas analyses. McDowell , S. D. , and Elders, W. A. (1979) Geothermal metamorphism of in the REFERENCES Salton Sea geothermal system. In: Geology and Geothermics of the Salton Trough sa edited Anderson, G. M. (1983) Some geochemical by Elders, W. A.), Campus Museum Contribu- aspects of sulfide precipitation in carbonate tion, University of California, Riverside, rocks. In: International Conference on CA. 5, pp. 70 - 76. Mississippi Val ley Type Lead-Zinc Deposits. Proc. Vol . (Edited by Kisvarsanyi , G., McDowell, S. D., and Elders, W. A. (1980) Grant, S. K., Pratt, W. P. and Koenig, J. Authigenic layer silicate minerals in W.), pp. 61 -76. University of Missouri- borehole Elmore 1, Salton Sea geothermal Rolla, Rolla, MO. field, California, U.S.A. Cont. Mineral. Pet. 74, 293 - 310. Andes, J. P., Jr., and McKibben, M. A. (1987) Thermal and chemical history of mineral ized McDowell, S. D., and Elders, W. A. (1983) fractures in cores from the Salton Sea Allogenic layer silicate minerals in scientific drilling project (abstract). Eos borehole Elmore No. 1, Salton Sea geother- Trans. AGU. 68, 439. mal systems. Am. Mineral. 68, 1146 - 1159.

Barnes, H. L. (1983) Ore-depositing reactions McKibben, M. A,, and Elders, W. A. (1985) in Mississippi Valley-type deposits. In: Fe-Zn-Cu-Pb mineral ization in the Salton Sea International Conference on Mississippi geothermal system, , Val ley Type Lead-Zinc Deposits. Proc. Vol . California. Econ. Geol. 80, 539 - 559. (Edited by Kisvarsanyi, G., Grant, S. K., Pratt, W. P. and Koenig, J. W.), pp. 77 - 85. Muffler, L. J. P., and White, D. E. (1969) University of Missouri-Rolla, Rolla, MO. Active metamorphism of upper Cenozoic sedi- ments in the Salton sea geothermal field and Elders, W. A. (1979) The geological back- the Salton Trough, southeastern California. ground of the geothermal fields of the Salton Geol SOC. Am. Bull. 80, 157 - 182. Trough. In: Geology and Geothermics of the . Salton Trough (Edited by Elders, W. A,), Campus Museum Contribution, University of Potter, R. W., 11, Clynne, M. A., and Brown, California, Riverside, Ca. 5, pp. 1 - 19. D. L. (1978) Freezing point depression of aqueous chloride solutions. Econ. Fournier, R. O., and Truesdell A. H. (1973) Geol . 73, 284 - 285. An empirical Na-K-Ca geothermometer for natural waters. Geochim. Cosmochim. Acta. Randall , W. (1974) An analysis of the 37, 1255 - 1275. subsurface structure and stratigraphy of the Sal ton Sea geothermal anomaly, Imperial Freckman, J. T. (1978) Fluid inclusion and Val ley, California. Ph.D. thesis Universitv oxygen isotope geothermometry of rock of California, Riverside, CA. 92 pp. samples from Sinclair No. 4 and Elmore No. 1 borehol es , Sal ton Sea geothermal field , Skinner, B. J., White, D. E., Rose, H. J., Imperial Valley, California, U.S.A. M.S. and Mays, R. E. (1967) Sulfides associated Thesis, University of California, Riverside, with the Salton Sea geothermal . Econ. CA. 66 pp. Geol . 62, 316 - 330.

-1 10- Somner, M. A., 11, Yanover, R. N., Bourcier, W. L., and Gibson, E. K. (1985) Determination of H 0 and COP concentrations in fluid Well A incl6sions in minerals using laser decrepita- 50 tion and capacitance manometer analysis. Anal. Chem. 57, 449 - 453. Younker, L. W., Kasameyer, P. W., and Tewhey, -lM-- J. D. (1982) Geological, geophysical, and thermal characteristics of the Salton Sea geothermal field, California. J. of Volcan. 1-50 and Geothermal Res. 12, 221 - 258.

200 203

250 - 250 - h +f fs i! i! 300 - 303 - f fQ Q D D 350 - 350

403 403

450 450

500 -- 500

550 550

KGRA b; .// ////////// 6M 012 SdO" 1 I Sea kilometers 7M 7M EXPLANATION Lithology Key Shale S~llslone Fig. 1. a 750 ~e~":,~e0 No Sample Well locations in the Salton Sea 0 Sandstone geothermal field. Wells: CS = Mineral Key 803 California State; E = Elmore; MM = Mineral Present I I Not Preienl Magrnamax; S = Sinclair; SP = Sportsman; W = Woolsey. A and B I I Not Analyred are the locations of the wells 850 described in this st-udy. Rhyolite domes: MI = Mullet Island; OB = Fig. 2. Obsidian Butte; RH = Rock Hill; RI = Rock Island. The hachured line Lithologies and mineral shows the boundary of the original distributions in the interlayered Salton Sea Known Geothermal Resource Area (KGRA). Modified illite/smectite and chlorite - from McDowell and Elders (1980). calcite zones of the two wells studied. I/S and C/S refer to interlayered il lite/smectite and chlorite/smectite respectively.

-1 11- -21 -20 -19 -18 -17 .I6 .IS -14 -13 -12 -11 -10 .9 -8 -7 -6 .S -. - M-0 zm- -100

.B 8.. 250 - - L(0

-A A A .. Y - . -... . 300- - 300

h 350- ut_+l -350

vE v f 4m- z -4w f, D8 a, D 410- - - 4SO

500 - - 5co

550 - - 550

6m- -Mo , , , , , , , (b) -21 .20 -19 .I8 -17 .I6 .IS .I4 -13 -12 .I1 .IO .9 4 -7 -6 .5

Fig. 3A.

Homogenization temperature vs. depth of fluid inclusions in anhydrite, sphalerite, and quartz. Symbols: anhydrite = open squares, Well A; filled squares, Well B; + = quartz, Well B; X = sphalerite, Well B. The baselines of the histograms are located at the lowermost limit of the sample interval. Tick marks labeled A and B indicate the depth to the base of the lacustrine deposits in the two wells. The boiling point curve for a 15 weight percent NaCl solution (Haas, 1971) is also shown for reference. B. Freezing point depression vs. depth for fluid inclusions in anhydrite. Refer to Fig. 3A for an explanation of the symbols.

-112-