Geology Reports⌕ Search

Geology topics

R.O. Fournier

Publications and source records attributed to R.O. Fournier.

At least 19 recordsLinked to original sources
USGS2012

The transition from hydrostatic to greater than hydrostatic fluid pressure in presently active continental hydrothermal systems in crystalline rock

Fluid flow at hydrostatic pressure (P h ) is relatively common through fractures in silicic and in mafic crystalline rocks where temperatures are less than about 350–370°C. In contrast, pore-fluid pressure (P f ) > P h has been encountered at the bottom of 3 geothermal exploration wells that attained temperatures >370°C (at Larderello, Italy, at Nesjavellir, Iceland, and at The Geysers, California). Chemical sealing by deposition of minerals in veins appears to have allowed the development of the high P f encountered in the above wells. The upper limit for the magnitude of P f that can be attained is controlled by either the onset of shear fracturing (where differential stress is relatively high) that reopens clogged veins, or the hydraulic opening of new or old fractures (at relatively low values of differential stress). The brittle-plastic transition for silicic rocks can occur at temperatures as high as 370–400°C in tectonically active regions. In regions where high-temperature geothermal systems develop and persist, it appears that either strain rates commonly are in the range 10 −12 to 10 −13 , or that silicic rocks in the shallow crust generally behave rheologically more like wet quartz diorite than wet Westerly granite.

Geophysical Research Letters↗
USGS2012

Chemistry and geothermometry of brine produced from the Salton Sea Scientific drill hole, Imperial Valley, California

The December 29–30, 1985, flow test of the State 2–14 well, also known as the Salton Sea Scientific drill hole, produced fluid from a depth of 1865–1877 m at a reservoir temperature of 305° ± 5°C. Another flow test at a depth of 3170 m produced brine contaminated by drilling fluid and diesel oil. Therefore we focus on the first flow test. Samples were collected at five different flashing pressures. The brines are Na-Ca-K-Cl-type waters with very high metal and low SO 4 and HCO 3 contents. Compositions of the flashed brines were normalized relative to the 25°C densities of the solutions, and an ionic charge balance was achieved by adjusting the Na concentration. The composition of the preflashed reservoir fluid was calculated using enthalpy-chloride relations applied to the normalized and charge-balanced brines. The calculated total dissolved solids in the preflashed reservoir fluid ranges from about 24.8 wt %, assuming insignificant thermal losses from the erupting fluid before sampling, to 26.0 wt %, assuming a 10% enthalpy loss by conduction of thermal energy through casing and surface piping. The preferred total dissolved solids of the reservoir fluid is 25.05 wt %. The calculated specific density of the preflashed reservoir fluid at 305°C and 1870 m depth ranges from 0.9980 (no thermal loss prior to sampling) to 1.0107 ± 0.0023 g cm −3 (10% thermal loss). Of the various cation geothermometers that are now in common use, the Na-K-Ca method gives a temperature (310°C) closest to the measured temperature (305°C) in the production horizon. Calculated Na/K geothermometer temperatures, using equations suggested by different investigators, range from 326° to 364°C. The Mg/K 2 method gives a temperature of about 350°C, Mg/Li 2 about 282°, and Na/Li 395°–418°C.

Journal of Geophysical Research Solid Earth↗
USGS2007

Hydrothermal systems and volcano geochemistry

The upward intrusion of magma from deeper to shallower levels beneath volcanoes obviously plays an important role in their surface deformation. This chapter will examine less obvious roles that hydrothermal processes might play in volcanic deformation. Emphasis will be placed on the effect that the transition from brittle to plastic behavior of rocks is likely to have on magma degassing and hydrothermal processes, and on the likely chemical variations in brine and gas compositions that occur as a result of movement of aqueous-rich fluids from plastic into brittle rock at different depths. To a great extent, the model of hydrothermal processes in sub-volcanic systems that is presented here is inferential, based in part on information obtained from deep drilling for geothermal resources, and in part on the study of ore deposits that are thought to have formed in volcanic and shallow plutonic environments.

Book chapter↗
USGS2002

Results of weekly chemical and isotopic monitoring of selected springs in Norris Geyser Basin, Yellowstone National Park during June-September, 1995

Each year at Norris Geyser Basin, generally in August or September, a widespread hydrothermal 'disturbance' occurs that is characterized by simultaneous changes in the discharge characteristics of many springs, particularly in the Back Basin. During the summer season of 1995, water samples from eight widely distributed hot springs and geysers at Norris were collected each week and analyzed to determine whether chemical and isotopic changes also occurred in the thermal waters at the time of the disturbance. In addition, Beryl Spring in Gibbon Canyon, 5.8 km southwest of Norris Geyser Basin, was included in the monitoring program. Waters discharged by four of the monitored hot springs and geysers appear to issue from relatively deep reservoirs where temperatures are at least 270 C and possibly higher than 300 C. At the time of, and for several days after, the onset of the 1995 disturbance, the normally neutral-chloride waters discharged by these four features all picked up an acid-sulfate component and became isotopically heavier. The acid-sulfate component appears to be similar in composition to some waters discharged in 100 Spring Plain that issue from subsurface regions where temperatures are in the range 170-210 C. However, the two monitored springs that discharge acid-chloride-sulfate waters in the 100 Spring Plain region did not show any significant chemical or isotopic response to the annual disturbance. Beryl Spring, and two neutral-chloride hot springs at Norris that appear to draw their water from reservoirs where temperatures are 250 C or less, also did not show any significant chemical or isotopic response to the annual disturbance. After the start of the annual disturbance, chloride concentrations in water sampled from Double Bulger Geyser in the Back Basin increased from about 800 ppm to about 1500 ppm, nearly twice as high as any previously reported chloride concentration in a thermal water at Yellowstone. The isotopic composition of that water precludes an origin of the high chloride by evaporation at atmospheric pressure. One way to account for the unique chemical and isotopic composition of this highly concentrated wateris by recirculation of water that had gone through one cycle of adiabatic cooling during upflow (decompressional boiling) back down into the hydrothermal system, where it is reheated to greater than 220 C. This previously boiled water then undergoes additional cycles of decompressional boiling during subsequent upflow. Another way the unique chemical and isotopic composition of Double Bulger water might evolve is by excess boiling in the formation that results from a decrease in fluid pressure within the channels of upflow. The annual disturbance at Norris Geyser Basin generally appears to be triggered by a cyclic up and down movement of the boilingpoint curve within the hydrothermal system in response to changes in the potentiometric surface of the cold water that is adjacent to, and interconnected with, that hydrothermal system. Annual disturbance phenomena that are easily recognized at Norris Geyser Basin may not be easily recognized elsewhere in Yellowstone National Park because (1) the neutral-chloride waters at Norris ascend directly from higher-temperature and higherpressure reservoirs (270 to >300 C at Norris compared to 180-215C at Upper and Lower Geyser Basins) that are capable of producing massive amounts of high-pressure steam, and (2) the clay that makes hot spring and geyser waters become turbid at Norris, heralding the start of the disturbance, comes from acid altered rocks that are widely distributed at intermediate depths at Norris, and that are rare in other geyser basins.

Wyoming↗
USGS1999

A geochemical reconnaissance of the Alid volcanic center and geothermal system, Danakil depression, Eritrea

Geological and geochemical studies indicate that a high-temperature geothermal system underlies the Alid volcanic center in the northern Danakil depression of Eritrea. Alid is a very late-Pleistocene structural dome formed by shallow intrusion of rhyolitic magma, some of which vented as lavas and pyroclastic flows. Fumaroles and boiling pools distributed widely over an area of ~10 km2 on the northern half of Alid suggest that an active hydrothermal system underlies much of that part of the mountain. Geothermometers indicate that the fumarolic gases are derived from a geothermal system with temperatures >225??C. The isotopic composition of condensed fumarolic steam is consistent with these temperatures and implies that the source water is derived primarily from either lowland meteoric waters or fossil Red Sea water, or both. Some gases vented from the system (CO2, H2S and He) are largely magmatic in origin. Permeability beneath the volcanic center may be high, given the amount of intrusion-related deformation and the active normal faulting within the Danakil depression.Geological and geochemical studies indicate that a high-temperature geothermal system underlies the Alid volcanic center in the northern Danakil depression of Eritrea. Alid is a very late-Pleistocene structural dome formed by shallow intrusion of rhyolitic magma, some of which vented as lavas and pyroclastic flows. Fumaroles and boiling pools distributed widely over an area of approx. 10 km2 on the northern half of Alid suggest that an active hydrothermal system underlies much of that part of the mountain. Geothermometers indicate that the fumarolic gases are derived from a geothermal system with temperatures >225??C. The isotopic composition of condensed fumarolic steam is consistent with these temperatures and implies that the source water is derived primarily from either lowland meteoric waters or fossil Red Sea water, or both. Some gases vented from the system (CO2, H2S and He) are largely magmatic in origin. Permeability beneath the volcanic center may be high, given the amount of intrusion-related deformation and the active normal faulting within the Danakil depression.

Geothermics↗
USGS1996

Compressive and tensile failure at high fluid pressure where preexisting fractures have cohesive strength, with application to the San Andreas fault

In thrusting and strike-slip situations, when the maximum principal horizontal stress S 1 acts nearly normal to a fault (a misoriented fault, such as the San Andreas), pore-fluid pressure > the lithostatic load, P f > S v , is required to reactivate movement on that fault. P f > S v may be achieved without causing hydraulic tensile fracturing if (1) previously existing cracks have regained cohesive strength by chemical processes, (2) subcritical crack growth has been blunted, and (3) the least principal horizontal stress S 3 nearly equals S v . Where P f > S v has been attained within a misaligned fault, increasing the stress difference ( S 1 – S 3 ) at constant P f > S v will not lead to shear failure, while a decrease in ( S 1 – S 3 ) can lead to shear failure of that fault. However, where the cohesive strength of material in a broad misaligned fault zone is less than that of the surrounding intact rock, increasing ( S 1 – S 3 ) while P f > S v can result in shear failure of fractures at near optimum angles to S 1 , but confined within this weak fault zone. If this faulting results in the local short-lived attainment of P f > S v (cataclastic deformation and frictional heating overcoming dilation) and a simultaneous decrease in ( S 1 – S 3 ), this combination of effects can trigger movement along the main trace of the misaligned fault. When increasing P f results in hydraulic failure, anisotropy in tensile strength or fracture toughness resulting from foliation within faults allows fractures to propagate along the planes of weakness rather than across the foliation perpendicular to S 3 .

Journal of Geophysical Research B: Solid Earth↗
USGS1994

Correlation of gold in siliceous sinters with 3He 4He in hot spring waters of Yellowstone National Park

Opaline sinter samples collected at Yellowstone National Park (YNP) were analyzed for gold by neutron activation and for other trace elements by the inductively coupled plasma optical emission spectroscopy (ICP-OES) method. No correlation was found between Au and As, Sb, or total Fe in the sinters, although the sample containing the highest Au also contains the highest Sb. There also was no correlation of Au in the sinter with the H 2 S concentration in the discharged hot spring water or with the estimated temperature of last equilibration of the water with the surrounding rock. The Au in rhyolitic tuffs and lavas at YNP found within the Yellowstone caldera show the same range in Au as do those outside the caldera, while thermal waters from within this caldera all have been found to contain relatively low dissolved Au and to deposit sinters that contain relatively little Au. Therefore, it is not likely that variations in Au concentrations among these sinters simply reflect differences in leachable Au in the rocks through which the hydrothermal fluids have passed. Rather, variations in [H 2 S], the concentration of total dissolved sulfide, that result from different physical and chemical processes that occur in different parts of the hydrothermal system appear to exert the main control on the abundance of Au in these sinters. Hydrothermal fluids at YNP convect upward through a series of successively shallower and cooler reservoirs where water-rock chemical and isotopic reactions occur in response to changing temperature and pressure. In some parts of the system the fluids undergo decompressional boiling, and in other parts they cool conductively without boiling. Mixing of ascending water from deep in the system with shallow groundwaters is common. All three processes generally result in a decrease in [H 2 S] and destabilize dissolved gold bisulfide complexes in reservoir waters in the YNP system. Thus, different reservoirs in rocks of similar composition and at similar temperatures may contain waters with different [H 2 S] and [Au]. The [H 2 S] in a subsurface reservoir water is difficult to assess on the basis of analyses of hot spring waters because of uncertainties about steam loss during fluid ascent. However, the same processes that result in low [H 2 S] in reservoir waters also tend to result in decreases in the ratio of 3 He 4 He(R) "> 3He4He(R) dissolved in that water. Values of R relative to this ratio in air ( Ra ) attain values > 15 in YNP thermal waters. To date, all of the thermal waters at YNP that have R Ra "> RRa values <9 have been found to deposit sinters with relatively low gold concentrations. These include all of the thermal waters that discharge from 180–215°C reservoirs at Upper, Midway, and Lower Geyser Basins within the western part of the Yellowstone caldera, and thermal waters at Norris Geyser Basin, outside the Yellowstone caldera, where some of the waters flow directly to the surface from a reservoir where the temperature is about 300°C. A high 3 He 4 He "> 3He4He ratio in thermal water discharged at the surface does not guarantee high gold concentrations in the sinter deposited by this water. Boiling with loss of steam (the gas phase takes a separate route to the surface) during rapid upflow from the shallowest reservoir to the surface decreases the [H 2 S] and total He dissolved in the residual liquid without appreciably changing the 3 He 4 He "> 3He4He ratio. This is because the isotopic composition of the He of the initial bulk fluid is unchanged and there is too little time for much radiogenic 4 He to build back into the liquid during this rapid ascent from the near-surface reservoir. However, if boiling with phase separation and loss of steam occurs deep in the system, the 3 He 4 He "> 3He4He ratio in the residual liquid, now depleted in H 2 S and total He, will be susceptible to dilution with radiogenic 4 He that is acquired during the longer residence time underground. Some or all of the Au that comes out of solution when an initial gold bisulfide complex breaks down as a result of loss of H 2 S may be swept up to the surface as solid (probably colloidal) particles, depending on the rate of flow of the mixture of water and steam, and the geometry of the channel. Where colloidal silica also forms as a result of this boiling, free Au apparently becomes attached to the colloidal silica and deposits where the silica deposits.

Wyoming↗
USGS1991

Conditions leading to a recent small hydrothermal explosion at Yellowstone National Park

Porkchop Geyser, in Yellowstone National Park, was the site of a small hydrothermal explosion on September 5, 1989. In the early 1960s, this was a quiescent spring with an intermittent seeping discharge. Infrequent geyser eruptions 3-5 m high started in 1971, and in 1985 the geyser began erupting as a perpetual spouter 6-9 m high. Perpetual spouting at the latter height continued until just before the catastrophic hydrothermal event when the geyser column suddenly rose to a height of 20-30 m, followed immediately by the explosive ejection of sinter blocks up to 1.88 m in maximum dimension and formation of an irregular crater 13.9 m long and 11.7 m wide. The ejected blocks show a variety of siliceous deposits indicative of changing environments of deposition with time, and possibly of prior hydrothermal explosive activity at this site. Water samples from Porkchop were collected and analyzed once in the 1920s, again in 1951, ten times between 1960 and mid-1989, and once in January 1990 after the explosion. Chemical geothermometry shows an increase in the temperature of last water-rock equilibration of about 60 to 70 °C from 1962 through 1989. This may have been the result of drawing water progressively from different and hotter regions of a single reservoir or mixing waters from two different reservoirs with a progressively larger proportion of water being drawn from the hotter reservoir from 1962 through 1989. Boiling of ascending fluids coming from hotter subsurface regions resulted in an increase in the proportion of steam to water discharged by the geyser. A constriction at the exit of the geyser tube throttled the flow of water and steam and allowed water in shallow cavities adjacent to the geyser tube to become heated to the boiling point at pressures greater than normal hydrostatic. We hypothesize that a sudden breaking loose of this constriction, likely triggered by a seasonal increase in subsurface boiling throughout Norris Basin, allowed water and steam to be discharged from Porkchop much more rapidly than previously. This resulted in a drop in pressure within the geyser tube, causing water in adjacent connected chambers to become superheated. An ensuing rapid flashing of superheated water to steam within relatively confined spaces resulted in the hydrothermal explosion.

Wyoming↗
USGS1990

Recent crustal subsidence at Yellowstone Caldera, Wyoming

Following a period of net uplift at an average rate of 15??1 mm/year from 1923 to 1984, the east-central floor of Yellowstone Caldera stopped rising during 1984-1985 and then subsided 25??7 mm during 1985-1986 and an additional 35??7 mm during 1986-1987. The average horizontal strain rates in the northeast part of the caldera for the period from 1984 to 1987 were: {Mathematical expression}1 = 0.10 ?? 0.09 ??strain/year oriented N33?? E??9?? and {Mathematical expression}2 = 0.20 ?? 0.09 ??strain/year oriented N57?? W??9?? (extension reckoned positive). A best-fit elastic model of the 1985-1987 vertical and horizontal displacements in the eastern part of the caldera suggests deflation of a horizontal tabular body located 10??5 km beneath Le Hardys Rapids, i.e., within a deep hydrothermal system or within an underlying body of partly molten rhyolite. Two end-member models each explain most aspects of historical unrest at Yellowstone, including the recent reversal from uplift to subsidence. Both involve crystallization of an amount of rhyolitic magma that is compatible with the thermal energy requirements of Yellowstone's vigorous hydrothermal system. In the first model, injection of basalt near the base of the rhyolitic system is the primary cause of uplift. Higher in the magmatic system, rhyolite crystallizes and releases all of its magmatic volatiles into the shallow hydrothermal system. Uplift stops and subsidence starts whenever the supply rate of basalt is less than the subsidence rate produced by crystallization of rhyolite and associated fluid loss. In the second model, uplift is caused primarily by pressurization of the deep hydrothermal system by magmatic gas and brine that are released during crystallization of rhyolite and them trapped at lithostatic pressure beneath an impermeable self-sealed zone. Subsidence occurs during episodic hydrofracturing and injection of pore fluid from the deep lithostatic-pressure zone into a shallow hydrostatic-pressure zone. Heat input from basaltic intrusions is required to maintain Yellowstone's silicic magmatic system and shallow hydrothermal system over time scales longer than about 105 years, but for the historical time period crystallization of rhyolite can account for most aspects of unrest at Yellowstone, including seismicity, uplift, subsidence, and hydrothermal activity. ?? 1990 Springer-Verlag.

Bulletin of Volcanology↗
USGS1990

Double-diffusive convection in geothermal systems: the Salton Sea, California, geothermal system as a likely candidate

Much has been published about double-diffusive convection as a mechanism for explaining variations in composition and temperature within all-liquid natural systems. However, relatively little is known about the applicability of this phenomenon within the heterogeneous rocks of currently active geothermal systems where primary porosity may control fluid flow in some places and fractures may control it in others. The main appeal of double-diffusive convection within hydrothermal systems is-that it is a mechanism that may allow efficient transfer of heat mainly by convection, while at the same time maintaining vertical and lateral salinity gradients. The Salton Sea geothermal system exhibits the following reservoir characteristics: (1) decreasing salinity and temperature from bottom to top and center toward the sides, (2) a very high heat flow from the top of the system that seems to require a major component of convective transfer of heat within the chemically stratified main reservoir, and (3) a relatively uniform density of the reservoir fluid throughout the system at all combinations of subsurface temperature, pressure, and salinity. Double-diffusive convection can account for these characteristics very nicely whereas other previously suggested models appear to account either for the thermal structure or for the salinity variations, but not both. Hydrologists, reservoir engineers, and particularly geochemists should consider the possibility and consequences of double-diffusive convection when formulating models of hydrothermal processes, and of the response of reservoirs to testing and production.

Geothermics↗
USGS1990

Interpretation of Na-K-Mg relations in geothermal waters

When using a Na-K-???Mg triangular diagram as an aid in the interpretation of a geothermal water, the estimated temperature of last water-rock equilibration may change by as much as 50??C, depending on which of the many Na/K geothermometers one assumes is correct. A particular geothermometer may work well in one place and not in another because of differences in the mineralogy of the phases that are in contact with the reservoir fluid. The position of the full equilibrium line that is used for geothermometry and for assessing degrees of departure from equilibrium also changes as the assumed K/???Mg geothermometer equation changes. The degree of ambiguity can be evaluated by utilizing the results of all the recently published Na/K geothermometers on a single Na-K-???Mg triangular plot.

Conference Paper↗
USGS1988

Fluid-inclusion evidence for previous higher temperatures in the Miravalles geothermal field, Costa Rica

Heating and freezing data were obtained for liquid-rich secondary fluid inclusions in magmatic quartz, hydrothermal calcite and hydrothermal quartz crystals from 19 sampled depths in eight production drill holes (PGM-1, 2, 3, 5, 10, 11, 12 and 15) of the Miravalles geothermal field in northwestern Costa Rica. Homogenization temperatures for 386 fluid inclusions range from near the present measured temperatures to as much as 70°C higher than the maximum measured well temperature of about 240°C. Melting-point temperature measurements for 76 fluid inclusions suggest a calculated salinity range of about 0.2–1.9 wt% NaCl equivalent. Calculated salinities as high as 3.1–4.0 wt% NaCl equivalent for 20 fluid inclusions from the lower part of drill hole PGM-15 (the deepest drill hole) indicate that higher salinity water probably was present in the deeper part of the Miravalles geothermal field at the time these fluid inclusions were formed.

Geothermics↗
USGS1986

Geochemical evaluation of the geothermal resources in the San Marcos region, Guatemala

The chemical and isotopic compositions of hot springs in the San Marcos region of Guatemala are internally consistent with a hydrologic model in which a deep 240°C reservoir and one or more shallow 195–200°C reservoirs are present. Variations in hot-spring water compositions results from a combination of boiling, mixing with cold, dilute water, and chemical re-equilibration with decreasing temperature. The recharge water for the deep 240°C reservoir is isotopically heavier than the local meteoric water and probably comes from many kilometers to the west or southwest. The water in the shallow reservoir is a mixture of the 240°C water with about 20 ± 5% of cold, locally derived meteoric water. After mixing, the water in the shallow reservoir re-equilibrates with reservoir rock at 195–200°C. In some places additional mixing with cold water occurs after water leaves the shallow reservoir.

San Marcos region↗
USGS1985

YELLOWSTONE MAGMATIC-HYDROTHERMAL SYSTEM, U. S. A.

At Yellowstone National Park, the deep permeability and fluid circulation are probably controlled and maintained by repeated brittle fracture of rocks in response to local and regional stress. Focal depths of earthquakes beneath the Yellowstone caldera suggest that the transition from brittle fracture to quasi-plastic flow takes place at about 3 to 4 km. The maximum temperature likely to be attained by the hydrothermal system is 350 to 450 degree C, the convective thermal output is about 5. 5 multiplied by 10**9 watts, and the minimum average thermal flux is about 1800 mW/m**2 throughout 2,500 km**2. The average thermal gradient between the heat source and the convecting hydrothermal system must be at least 700 to 1000 degree C/km. Crystallization and partial cooling of about 0. 082 km**3 of basalt or 0. 10 km**3 of rhyolite annually could furnish the heat discharged in the hot-spring system. The Yellowstone magmatic-hydrothermal system as a whole appears to be cooling down, in spite of a relatively large rate of inflation of the Yellowstone caldera.

Conference Paper↗
USGS1983

A method of calculating quartz solubilities in aqueous sodium chloride solutions

The aqueous silica species that form when quartz dissolves in water or saline solutions are hydrated. Therefore, the amount of quartz that will dissolve at a given temperature is influenced by the prevailing activity of water. Using a standard state in which there are 1,000 g of water (55.51 moles) per 1,000 cm 3 of solution allows activity of water in a NaCl solution at high temperature to be closely approximated by the effective density of water, p e , in that solution, i.e. the product of the density of the NaCl solution times the weight fraction of water in the solution, corrected for the amount of water strongly bound to aqueous silica and Na + as water of hydration. Generally, the hydration of water correction is negligible. The solubility of quartz in pure water is well known over a large temperature-pressure range. An empirical formula expresses that solubility in terms of temperature and density of water and thus takes care of activity coefficient and pressure-effect terms. Solubilities of quartz in NaCl solutions can be calculated by using that equation and substituting p e , for the density of pure water. Calculated and experimentally determined quartz solubilities in NaCl solutions show excellent agreement when the experiments were carried out in non-reactive platinum, gold, or gold plus titanium containers. Reactive metal containers generally yield dissolved silica concentrations higher than calculated, probably because of the formation of metal chlorides plus NaOH and H 2 . In the absence of NaOH there appears to be no detectable silica complexing in NaCl solutions, and the variation in quartz solubility with NaCl concentration at constant temperature can be accounted for entirely by variations in the activity of water. The average hydration number per molecule of dissolved SiO 2 in liquid water and NaCl solutions decreases from about 2.4 at 200°C to about 2.1 at 350°C. This suggests that H 4 SiO 4 may be the dominant aqueous silica species at 350°C, but other polymeric forms become important at lower temperatures.

Geochimica et Cosmochimica Acta↗