Carbon Sequestration Potential of the Coast Range Ophiolite in California

A number of methods of drawing down atmospheric carbon dioxide are being investigated as a result of concerns over the impacts of global warming. Mineral carbonation is a proven manner of sequestering carbon dioxide and functions by binding carbon dioxide with iron-, calcium-, and magnesium-containing minerals—including olivine, pyroxene, and serpentine— to form carbonate minerals. The Coast Range Ophiolite (CRO) in California is a 700 km long body of ultramafic rocks containing vast amounts of olivine and serpentine. By assessing the mineralogical and geochemical characteristics of the minerals in these rocks, I assess the potential quantity of CO2 that could be sequestered in these rocks. Olivine is the most ideal mineral reactant for carbon sequestration, based on abundance and reactivity. My data show that the first 45 m of surface rocks of the CRO contain very little olivine, but petrography shows olivine at least 110 m below the surface, and olivine has been previously identified in various sections of the CRO. Using energy-dispersive x-ray spectroscopy, I showed that the amount of Mg in olivine in the rocks is 24% and the amount of total Mg + Fe + Ca is 38%. Based on these figures and volume, density, and the percentage of rock reacted with CO2, I calculate that the CRO can sequester ~ 6.9x1013 kg CO2 for Mg-only sequestration and ~9.1x1013 kg CO2 for Mg, Fe, and Ca sequestration, which correspond to 14% and 39% of the amount of atmospheric CO2 that would return us to 350 ppm CO2, Thus the CRO, while not capable of eliminating enough CO2 to halt global warming by itself, still can sequester a considerable amount and should be considered a viable option for mineral sequestration.


Introduction
Warm Earth surface temperatures rely on a functioning planetary greenhouse, and CO 2 is a powerful gas component of this greenhouse. Industrial fossil fuel burning has increased atmosphere CO 2 dramatically in recent times: atmospheric scientists measured atmospheric carbon dioxide levels since 1958 and found that atmospheric CO 2 concentrations have increased since then reached a record high 396.78 ppm in May of 2012, as of the latest Mauna Loa Observatory readings (NOAA-ESRL, 2012). With global temperatures having already risen 1.4°F since the start of the 20 th century, and with temperatures likely rise at least another 2 °F-and possibly more than 11 °F-over the next 100 years, the level of CO 2 in the atmosphere is a serious concern (NAS, 2008).
With global population and industrialization increasing, CO 2 levels are unlikely to drop in the near future. The Intergovernmental Panel on Climate Change (IPCC) has modeled possible future atmospheric CO 2 concentrations based on scenarios where certain variables such as population, renewable energy sources as a percentage of total energy, affluence, energy demands, and economic output are altered. Emissions could reach as high as 80 Gt C/yr by 2058 if no action is taken (Mathez, 2009). Both CO 2 levels and global temperatures could reach levels high enough to cause negative impacts on people, agriculture, and wildlife across the world (Fig. 3) (Smith et al., 2009;Thornton, 2012;Doney et al., 2012). The El Niño Southern Oscillation affects ecosystems, agriculture, freshwater supplies, hurricanes and other severe weather events worldwide and is a temperature-sensitive phenomenon, and while it is unclear as to the effects increased temperature will have (Collins et al., 2010), increasing the risk of permanently changing the way those natural systems function could be perilous.

Carbon Sequestration Fundamentals
Carbon sequestration occurs naturally and is responsible for moderating carbon dioxide levels in the atmosphere over geologic time. An enormous reserve of carbon is stored in sediment and rocks in the Earth: Ruddiman (2001) estimates 66 million Gt of carbon are stored in sediment and rocks in the Earth. Pre-industrial levels of atmospheric carbon are estimated to be about 600 Gt. Plainly, if the Earth did not have a way to fix carbon dioxide, atmospheric levels would be orders of magnitude higher (Ruddiman, 2001).
Natural sequestration of carbon begins with carbon dioxide in the atmosphere dissolving in the ocean, and is effected through chemical precipitation of carbonate minerals (often biologically mediated) and subsequent compaction into limestone, and/or deposition of mineralized carbon in organic-rich sedimentary shales, the largest reservoirs of carbon on Earth. In natural waters, the speciation of CO 2 is pH-dependent. At low pH values (<6.4), dissolved CO 2 or carbonic acid, H 2 CO 3 , is the dominant species (pK H = 1.5 at 25°C). At intermediate, near-neutral pH values, bicarbonate, HCO 3 -, dominates (pK 1 = 6.3 at 25°C). The carbonate ion, CO 3 2-, dominates at elevated pH (pK 2 = 10.3 at 25°C) (Fetter Jr., 1980). The carbonate system modulates global biogeochemical cycles in the ocean and in reactive rock environments, providing essential dissolved inorganic carbon to the biosphere.
Carbon is also naturally stored in plant remains that have been buried and preserved over time to become coal, oil, and natural gas. While the quantity of carbon stored this way is volumetrically subordinate to the carbon stored in limestone, it represents the portion of carbon that is currently being re-released into the atmosphere.
Carbonation of mafic and ultramafic rocks on the Earth's surface occurs naturally as well, locking carbon in geological formations over deep time. Natural carbonation of basalt consumes ~1.8 x 10 8 tons of CO 2 per year globally, and is quite a rapid process (Matter and Kelemen, 2009). In Oman, ~10 3 tons of CO 2 km 3 /yr are estimated to be consumed by peridotite carbonation occurring in naturally exposed mantle units of ophiolites (Matter and Kelemen, 2009). Given increased (and increasing) CO 2 generation from anthropogenic sources, the rates of natural sequestration reactions are not sufficient to stabilize Earth's rising temperature.
Human-facilitated carbon sequestration methods include (a) subterranean burial of biomass or biochar (i.e., charcoal created by pyrolysis of biomass), (b) ocean storage of biomass, where biomass is transported down rivers and buried where the river meets the sea (Strand and Bedford, 2009), and (c) injecting aqueous CO 2 below the Earth's surface to be locked in place by carbonation reactions. Iron fertilization of nutrient-limited patches of the global ocean surface has been considered as a means of stimulating primary production (thus drawing down atmospheric CO 2 ), but the practice could be ecologically disruptive, and models have consistently shown that at the limit, the approach could not substantially change the trajectory of global warming (Strong et al., 2009). Mineral sequestration has both the largest potential for storage and the lowest potential for CO 2 leakage among sequestration methods (Herzog, 2009), and is the focus of this study.
The largest ongoing mineral sequestration operation is the Carb Fix project in Iceland (Gislason et al., 2010). The project captures CO 2 emitted from the nearby Hellisheidi geothermal power plant and combines it with water to create a carbonrich fluid. The fluid is then injected into boreholes anywhere from 200 m-2000 m deep, where it reacts with calcium-rich basaltic rocks to form calcium carbonate (CaCO 3 ) (Gislason et al., 2010). Iceland is located on a divergent plate boundary and is composed almost entirely of basalt. As a mafic rock with abundant divalent cations, basalt presents an excellent opportunity for in situ capture and storage.
Assuming 10% porosity of the rocks, of which only 10% will eventually be filled with calcite, 0.01 km 3 volume is available for calcite precipitation. This volume can accommodate 12 million tons (1.10 x 10 10 kg or 10.9 Mt) of CO 2 . At the present CO 2 emission rate of magmatic CO 2 from the geothermal power plant at Hellisheidi Iceland (60,000 tons/year), it would take about 200 years to fill this available pore space (Gislason et al., 2010).
However, there are currently no mineral carbon sequestration facilities in the United States, yet ample opportunity exists. For example, the Central Atlantic Magmatic Province (CAMP) -located off the eastern U.S. coast-is partially suboceanic continental flood basalt stretching from Newfoundland in Canada down to Florida (Goldberg et al., 2009). With such an expanse of flood basalt existing below the ocean, the CAMP presents an opportunity for mineral sequestration using CO 2 dissolved in seawater.
A serpentinite-based mineral sequestration project is being planned in New South Wales, Australia by the GreenMag Group and the University of Newcastle.
Because of the lack of suitable geology for deep injection of fluids, the group plans to construct a mineral sequestration plant that takes advantage of the abundance of serpentinite rocks and soils in the area. Plans are underway for its construction, and with a fledgling carbon tax law, there is economic incentive. The successful operation of this plant would provide empirical proof that serpentinite-hosted carbon sequestration is a viable method of carbon capture (GreenMag Group media release, 2010).
In the United States, numerous teams are engaged in research evaluating mineral sequestration in serpentine-and olivine-rich matrices. The Albany Research Center in Albany, OR has conducted research on serpentine and olivine-based mineral sequestration and proposed that moderately high temperature (185 °C for olivine) carbonate mineralization reactions spurred in olivine-bearing rock can convert between 49% and 91% of gas phase or aqueous CO 2 to carbonate mineral phases .
On the west coast of the United States, continental ultramafic serpentinites and some olivine-rich rocks stretch from Oregon to central California, with enormous potential for carbon storage. These rocks are in close proximity to cities with high populations--San Francisco, Oakland, Sacramento, Napa, Fresno, and San Jose. These cities all demand resources such as electricity, fuel, and cement, all of which are high-volume CO 2 emitters. Additionally, oceanic crust off the Atlantic and Pacific coasts is broadly suitable for seabed carbon sequestration projects, as described theoretically in Kelemen and Matter (2008) and Goldberg et al. (2009). Xu et al. (2004) state that aquifers containing 'basic' silicate minerals with a high proportion of Mg and Ca, such as olivine, serpentine, pyroxenes and plagioclase, have the greatest potential to fix CO 2 as carbonate minerals because they have a high molar proportion of divalent cations and they react rapidly to form carbonate minerals.
In the following work, I synthesize new data for serpentinite rock cores with regional data pertinent to ultramafic rock characteristics, to quantify the total carbonate sequestration potential of the Coast Range Ophiolite in northern California. Cored rocks comprise two, 31 m and 46 m-deep serpentinite sections, and data include lithostratigraphic columns descriptions, petrography of thin sections, x-ray diffraction results, and bulk and spot solid geochemistry data obtained through x-ray fluorescence (XRF) and scanning electron microscopy coupled to energy dispersive x-ray spectroscopy (EDS). These results will anchor quantitative modeling of the carbon storage possible in these rocks. Recent progress in ultramafic rock carbonation reactions provides context for this study, enabling assessment of Coast Range Ophiolite serpentinites as geological repositories of carbon; Coast Range Ophiolite ultramafic rocks may well represent a significant carbon sequestration option for northern California and serve as a model for other locations.

Carbon Sequestration Chemistry
In the presence of CO 2 -rich fluids, with certain temperature and pressure conditions, the following minerals react predictably to from new carbonate minerals (O'Connor, 2005 These are simplified reactions. In reality, the process from ore mineral to carbonate mineral contains multiple steps. The primary step is the dissolution of gaseous carbon dioxide into the aqueous phase via CO 2 (g) + H 2 O (l) → H 2 CO 3 (l) → HCO 3 -(aq) + H + (aq) The second step is the dissolution of the primary cation reactants from the mineral lattices into solution. The following reaction depicts the dissolution of serpentine from Krevor and Lackner (2011) Mg 2 SiO 4 (s) + 4H + → 2Mg 2+ + H 4 SiO 4 (aq) With aqueous bicarbonate and magnesium ions, the final step of the carbonation reaction is ready to proceed.
Mg 2+ + HCO 3 -→ MgCO 3(s) + H + The magnesium and bicarbonate ions react to form magnesite and a proton. The reaction in Eq. 9 will obviously not proceed unless protons are consumed. Proton consumption conveniently takes place during magnesium dissolution in ultrabasic rocks, as seen in Eqs. 7and 8. Thus we have a multi-step process for the carbonation reaction, with a natural proclivity-in ultramafic rocks-to proceed to the right. This is another reason ultramafic rocks are idea candidates for carbon sequestration.

Geologic Setting
The Coast Range Ophiolite (CRO) -the oceanic basement beneath the Jurassic and Cretaceous sedimentary rocks of the Great Valley sequence-is an assemblage of ultramafic, mafic, and minor felsic igneous rocks of Jurassic age scattered in a 700 km long segment (Fig. 4) in Northern California Bailey, 1971). An ophiolite is a section of oceanic lithosphere that has been uplifted and exposed above sea level and often emplaced onto continental crustal rocks.
Geochemical evidence indicates that the CRO formed in a supra-subduction zone. Shervais et al. (2004) propose that the Coast Range Ophiolite formed in multiple stages, beginning with mantle melting at a nascent subduction zone. Further melting induced increased fluid flux from the subducting slab, creating an assemblage of wehrlite and feldspathic peridotite born from refractory melt. As time progressed, the subduction zone matured and stabilized, underlaying the refractory rocks with calc-alkaline and tholeiitic rocks. Finally, the incomplete ophiolite sequence was underthrust by sheet basalts from the subduction of a nearby spreading center. As the spreading center subducted, the leading slab detached and sank into the mantle.
When the tailing slab of the spreading center entered the subduction zone, sediments were scraped off, forming an accretionary wedge (the Franciscan Complex, in this case). The newly formed accretionary wedge provided enough buoyant force to uplift and expose the overlying ophiolite sequence (Shervais et al., 2004). Presently, both obduction and accretionary uplift are plausible sources of CRO emplacement.
Ophiolites expose rocks that are stable kilometers below the ocean floor to shallow crustal, continental environments. When basalts, pyroxenites, and peridotites are exposed to water at elevated temperatures and pressures, either through hydrothermal fracturing, convergent margin volatile cycling, or postemplacement weathering, the pyroxene and olivine in those rocks may be transformed into serpentine minerals. As this is the scenario for the creation of the Coast Range Ophiolite, outcrops of the CRO comprise: mantle peridotite, gabbro, pyroxenite, basalt, and serpentinite.

X-Ray Powder Diffraction
A Terra Field Portable XRD/XRF Instrument (University of Rhode Island) was used to determine the mineralogy of the core samples. The Terra is designed for powder diffraction, requiring that rocks be pulverized before analysis. At least 15 mg of a sample is crushed and sieved to a grain size of 150µm or less, then transferred to the sampling cell. The Terra uses a cobalt anode with 30 kV x-ray tube voltage and 10W x-ray tube power, with the x-ray energy range varying from 3-25 keV.
Laboratory default parameters for the instrument are set for 250 exposures, producing both x-ray diffraction (XRD) patterns for mineral characterization and xray fluorescence (XRF) spectra for summary geochemical data.
The diffraction patterns are analyzed using XPowder ( http://xpowder.com/) software, which compares the diffraction pattern against a database of known mineral patterns and attempts to match based on peak location d-spacings. The user can decide to keep the match or search for specific minerals to attempt to make a match themselves.
Core samples from the field site were generally clay, cobbles, or veins within the cobbles. Isolated samples of each (clay, cobbles, and veins) were run to ensure accuracy of identification of each sample and then a larger sample of cobbles entrained in clay is run to estimate each constituent. At least one sample from each section of core (a core section is approximately 1 m in length) was sampled, and in the case of a noticeable change in mineralogy, multiple samples from a single core may have been obtained. The corresponding depth and depth interval of each section was noted and a core log was compiled.

Thin Section Petrography
Thin sections were made of a number of cobbles and solid intervals within the cores. Thin sections were cut at the Spectrum Petrographics facility (a fee-forservice lab, petrography.com) and analyzed using an Olympus BH-2petrographic microscope, utilizing both plane-and cross-polarized light. Fields of interest identified via petrography were further studied with a scanning electron microscope.

Energy-Dispersive X-ray Spectroscopy
Thin sections were further analyzed on the JEOL 5900 SEM at the Environmental Scanning Electron Microscope facility in [University of Rhode Island]. Geochemical mapping of points (or tagging) and lines across thin section surfaces were performed on a portion of a sample to determine elemental abundances within individual crystals or within the groundmass of the slide.

X-Ray Fluorescence
A Thermo Scientific Niton XL3 Series XRF Analyzer was used to determine elemental abundance within each sample. One sample from each core interval was taken. The elements used for analysis are as follows: Zr, Sr, Rb, Pb, Zn, Ni, Co, Ca, Fe, Mn, Cr, V, Ti, K. Accurate estimation of the volume of ultramafic rocks in the Coast Range Ophiolite (CRO) is not simple since the CRO is present in many disparate sections, but this is a necessary step in calculating total carbon sequestration potential. Wakabayashi and Dilek (2000) quantify the length of the of the CRO as 900 km, with most outcrops not exceeding 10 km in length, and with an average depth of 5 km. If the CRO were to occupy all of this land, we could constrain a maximum volume as a simple l x w x d calculation, to set a maximum volume of 4.5 x 10 4 km 3 . If we compare the idealized maximum area of 9000 km 2 to a map of the Coast Range ( Fig.   4.), we can see that ultramafic rocks certainly do not span its entire length and width-thus 9000 km 2 is an overestimation, but provides a working estimate for a maximum possible area.

Quantifying Carbon Sequestration Potential
An ArcGIS-facilitated area estimation of CRO serpentinite better quantifies the serpentinite extent. Through ArcGIS, it is straightforward to sum the areas of each individual rock type on the USGS geologic map of California (Fig. 7). The total area of serpentinites determined in this way is 3865 km 2 . Given this area and an average depth of 5 km, the approximate volume of the CRO is 19,000 km 3 ; this volume is used below as the basis for the carbonation calculation.

Uncertainty
The calculations made in this paper were based either upon data obtained from previous works or from data obtained from work done specifically for this thesis. There is variable uncertainty with each component used to calculate total sequestration potential, including rock volume, rock density, olivine:serpentine percentage, and magnesium percentage of olivine.
The volume of the CRO contains uncertainty more in the depth component of calculating volume than area. The area component was obtained using reliable GIS data from USGS, while depth was obtained from Wakabayashi and Dilek, (2000).
The value of 5 km depth from Wakabayashi and Dilek (2000) was the only source that provided a depth for the Coast Range Ophiolite. As such, it is subject to question, however, it falls within acceptable values for ophiolite depths and will be used despite the inherent uncertainty. Thus, values for depth were chosen using 5 km as a median for the calculations in tables 22 and 23.

Results and Discussion
Carbon sequestration rate and total potential are dependent on a number of factors: most fundamentally, available carbon dioxide and free reactive cations, (e.g., Mg 2+ , Fe 2+ and Ca 2+ ) are necessary. CO 2 can be readily dissolved in a manufactured fluid, naturally occurring groundwater or seawater. Sufficient cations result from alteration of ultramafic rocks. Sequestration also responds to pressure-temperature conditions, mineralogy, rock chemistry, porosity/permeability, and solution chemistry. Each of these items impacts net carbon sequestration and affects calculations of the total mass of carbon that can be fixed through the idealized reaction of Coast Range Ophiolite ultramafic rocks.
It is important to note that the sequestration outlined in this paper is not that of drawing CO 2 out of the atmosphere to reduce atmospheric CO 2 levels, but rather delineating the CRO as a repository for sequestering future carbon emissions.
Drawing CO 2 out of the atmosphere in an attempt to reduce atmospheric levels of would be unfeasible, as the amount of time it would take to accomplish the level of sequestration desirable would be overwhelmed by the amount of CO 2 generated by human input during that time. Thus, "sequestration" in this sense could more accurately be worded as "storage," however , the term "sequestration" has historically been used to convey what "storage" entails.

Temperature-Pressure Constraints on Carbonation Reactions
Like most processes in nature, carbonation of olivine and serpentine are both greatly aided by the presence of heat. Heating the rocks is a necessary step, as the carbonation reaction is slow at lower temperatures. Natural carbonation occurs wherever there are mafic and ultramafic rocks exposed to either atmospheric or dissolved CO 2 . Kelemen and Matter (2008) showed that 4x10 7 kg yr -1 of CO 2 are sequestered in this manner every year at the Semail Ophiolite. While that is an impressive figure, that rate needs to increase by orders of magnitude to use mineral sequestration as a primary carbon dioxide mitigation process. O'Connor et al.
(2005) showed that olivine reacts most ideally at 185 °C and 150 atm P CO2 , whereas serpentine requires a costly 600 °C preheating procedure that costs too much money and energy to be considered useful. The rate of CO 2 sequestration can be elevated to as high as an estimated 2x10 9 tons of CO 2 /km 3 per year (Kelemen and Matter, 2008).
Thus, olivine rocks should be preheated with hot fluid before they are reacted with any CO 2 . Fortuitously, the carbonation reaction is exothermic, so the reaction of creating carbonate minerals (Eq. 9) can generate enough heat to sufficiently warm the surrounding rocks and aid in boosting the reaction kinetics to keep carbonation reactions continuous and rapid and to prevent the constant need to heat and pump non-CO 2 fluid down into the borehole. With the aid of heat pretreatment and increasing the partial pressure of CO 2 in solution, rates can be increased anywhere from 10 6 to 10 9 times the natural rate (Kelemen and Matter, 2008).

Serpentinite and Peridotite Mineralogy
Minerals in mafic and ultramafic rocks (i.e., rocks rich in magnesium and iron compared to silica) provide cations necessary for mineralization of carbonate. The most common mafic minerals include olivine, pyroxene, serpentine, and amphibole.
under CO 2 -rich conditions, such as might be expected in submarine serpentinization, or a shallow crustal continental setting with active meteoric water inputs.
Olivine (Mg 2 SiO 4 ) is the predominant mineral in peridotite, which comprises the bulk of the Earth's upper mantle. Olivine is one of the two minerals-the other being pyroxene-that commonly transforms into serpentine minerals via exposure to water, heat, and pressure.
Albite (NaAlSi 3 O 8 ) is a common igneous and low grade metamorphic mineral. It is the most sodic form of plagioclase. The albite in the Coast Range Ophiolite region originates from oceanic crust portion of the ophiolite, which is composed mostly of basalt . Albite is also a common mineral found in greenstones, basalts that have undergone low-grade greenschist facies metamorphism at temperatures between 300 and 450 °C and pressures of 1 to 4 kbar (Deer et al., 1992). Many of the rocks in the CRO have been subject to hydrothermal alteration under greenschist facies pressure/temperature conditions . In the landscape, more weather-resistant greenstones protrude from the surrounding serpentine. The basalt, greenstone, and serpentinized peridotite occur together in a regional-scale mélange , and also intermix sporadically in the cores examined here.
Clay and clay minerals: The clay mineralogy of weathered serpentinites is complex.
The heterogeneity of serpentinite textures-mesh, hourglass, blade (Wicks and Whittaker 1977)-and the occurrence of additional rock-forming minerals, like chlorite, add to the complexity of the system (Caillaud et al., 2006). It is not uncommon to see saponite, nontronite, montmorillonite, or vermiculite. The XPowder XRD database does not have a nontronite or a saponite reference file, however XRD data show multiple peaks consistent with the presence of montmorillonite and/or vermiculite. In the CRO, smectites are a result of serpentine weathering and are represented in all tables in this paper as "clay". Vermiculite indicates weathering of chlorite, which occurs in each core also. Often both clay minerals are mixed together, and their diffraction peaks overlap, so it is difficult to parse which mineral is present.
Tremolite (Ca 2 Mg 5 Si 8 O 22 (OH) 2 ) is an amphibole mineral that occurs in low-grade metamorphic ultramafic rocks and is a common accessory mineral for serpentinites (Deer et al., 1992). It is also a regular component of greenstones and as such only occurs in the core samples when other greenstone minerals are present. Tremolite alters to talc, chlorite, and carbonate minerals, depending on geochemical context (Deer et al., 1992).
Calcite (CaCO 3 ) is a product of the calcium-rich and carbon-poor groundwater coming into contact with carbon dioxide either from the atmosphere or from meteoric water. Veins of calcite are strewn throughout the core samples, a result of hydrothermal precipitation of waters enriched in calcite via meteoric influx.
Lizardite (Mg,Fe) 3 Si 2 O 5 (OH) 4 is a polymorph of serpentine and is produced through alteration of olivine and pyroxene. It is the lower temperature and pressure phase of serpentine polymorphs (Moody, 1976). It is impossible to discern antigorite from lizardite in a hand sample, so x-ray diffraction must be utilized. Even with XRD, it is not entirely clear which polymorph is present. XRD most strongly supports the presence of lizardite.

Core Descriptions of Serpentinites from CROMO1 and CROMO2
One or more samples were taken from each core section. Interval sizes are irregular as a result of incomplete core collection, different sized core inserts during drilling, and occasions when auger drilling with no sample retrieval were necessary.

Summary Stratigraphy: CROMO1
0-1.98 m: The first 1.98 m of the Core Shed Well is a soil layer near the surface and records intense biological, physical, and chemical weathering. The bulk composition of this soil is magnetite-rich clay with albite and quartz, most likely from the weathering of nearby greenstones (Fig. 8). Albite is one of the primary minerals composing both basalts and greenstones. When basalt is metamorphosed to greenstones, some minerals get chloritized. Further erosion of chlorite can produce vermiculite (Caillaud et al., 2006), which is suggested by XRD data but would require further testing to confirm.
1.98-2.44 m CROMO1 1,5 represents the first strong presence of lizardite ( Fig. 9) and additionally the end of any evidence of bioturbation (e.g., the first 20 cm contain roots, however no signs of flora or fauna exist beyond the first 20 cm). Core samples have a fine-grained clayey matrix with angular clasts. XRD determined the clay is 12.65 -27.43 m Greenstone minerals tremolite, chlorite, and albite are absent for the next 15 meters. There is a predominance of lizardite and magnetite ( Fig. 12) with some clay (Fig. 13). Lizardite is not in solid cobbles, but rather large, intact, amalgamated pieces that appear to be massed together from tectonic activity; quite possibly the cores sample a shear zone and this is representative fault gouge material. Samples from this site are very friable, with clay intruding where some of the lizardite is schistose with stacked and separated phyllosilicate layers.
27.43 -31.09 m Greenstone and serpentine mix as large cobbles and intact core segments. Hand samples are denser than pure serpentine and lack the greasy luster and platy qualities of serpentine hand samples. Petrography reveals abundance of twinned plagioclase crystals (Fig. 14), which is supported by XRD (Fig. 15). Since entire cobbles from this horizon show concurrent greenstone and serpentinite, it can be attributed not to the mixing of horizons from tectonism, but rather from partial serpentinization during the metamorphosis of basalt to greenstone.
Additionally, there was a vein of calcite in this section at 27.5 m depth, which was isolated and analyzed (Fig. 16). Recall that calcite formation is a result of the natural carbonation process that ultramafic rocks undergo.

Summary Stratigraphy: CROMO2
0.00-1.37 m The topsoil layer is composed predominantly of lizardite and magnetite present in friable pieces situated in a yellow-brown lizardite clay (Fig. 17). Plant roots are present throughout, though less so than at CROMO1. CROMO2 area is considerably drier, so the top soil is not as moist and is more oxidized than CROMO1 (Fig. 18). 5.03 -22.86 m This interval contains an extensive segment of pure serpentinite minerals, with a predominance of lizardite and magnetite and little to no clay minerals or greenstone minerals. There is an exception occurring at 6.10m depth, where there is a small pocket of greenstone clasts (Fig. 19) fragmented along with lizardite clay (Fig. 20). Aside from this interval, the specimens appear identical in hand sample and contain angular pebble-sized lizardite clasts in light grey clay ( Prior to the 1980s, when the Homestake Mining Corporation surveyed the land that is now the McLaughlin Natural Reserve, a series of strategically placed holes were drilled to focus gold-mining activities in nearby silica-rich hydrothermal deposits. These exploration-related cores remain intact, and one deep serpentinite core was sampled for thin section review. The drill site for this core was destroyed during gold mining activities, but was located <5 km to the SE of the CROMO drilling location; this site will for simplicity be called Homestake Mine Pit Borehole (HMPB), and the depth of the sample is about 110 m below the surface.

Deep Core Specimen-Homestake Mine Pit Borehole
Homestake Mine Pit Borehole (HMPB), ~110 m depth: This core sample is very hard and shows no signs of friability or schistosity. The sample resembles hard serpentine in hand sample, and while subtly different in texture, does not appear different from shallower cored sections until examination with petrographic microscope. Fig. 30 shows that strongly birefringent olivine is present (note blues, pinks, purples, yellows, and oranges) and there is also a large pyroxene crystal in the NE corner of the slide with streaks of serpentine running through it. Both the olivine and the pyroxene have been serpentinized to some degree [serpentine minerals and magnetite are scattered throughout the thin section field of view], however much less so than in shallower rocks.

Integrated View of CRO Subsurface Rocks, McLaughlin Locality
Overall, the shallowest 45 m of CRO rock at McLaughlin show nearly complete serpentinization. In addition to serpentinite rocks, there are greenschist facies rocks distributed throughout the cores. Often, greenschist facies rocks (here taken to be metamorphosed mafic igneous rocks containing chlorite, albite and tremolite among other phases) occur in a sample of core alongside serpentinite minerals, with the close juxtaposition of phases a result of the history of repeated convergent margin tectonism in northern California. Peridotite with relict olivine grains does not occur in the shallowest rocks, but does indeed present at 110 m depth (Fig. 30). Between 45 m and 110 m, the mineralogy may shift gradually from more to less serpentinized, or there may be a contact zone of some kind (not detected in cored samples) that causes a sharp transition; only further coring or comprehensive access to archived cores with Homestake will determine the nature of the transition. Presently access to Homestake cores is limited due to a building structure safety issue where they are being housed.
Notably absent from the cores were some of the more common accessory minerals in ultramafic rocks. Talc (Mg 3 Si 4 O 10 (OH) 2 ) is often found associated with serpentinization, however, XRD failed to show any significant proof of its presence, perhaps due to a dearth of excess silica during serpentinization or the absence of CO 2 during any metamorphic changes in the serpentine. Also absent from the XRD results is brucite (Mg(OH) 2 ), which forms during the serpentinization of dunites (Deer et al., 1965). It is possible that brucite was once present, but has since been weathered away or dissolved by either meteoric water or groundwater, or is present in such small amounts that it is undetectable by XRD.

Observations of ultramafic rock cores sampled at the McLaughlin Natural
Reserve indicate that the peridotite has been heavily serpentinized in general, but that relict olivine and pyroxene grains were present and visible under the petrographic microscope.

Geochemistry
The mineralogy of a system controls whether it is an effective sequestration vehicle, but geochemical data determine to what degree the host rock will function effectively in sequestration projects. Ultramafic rocks are defined by their mineral content. The adjective ultramafic implies the abundant presence of magnesium and iron (recall that magnesium-ferric contracts to mafic) and those elements, along with calcium, serve as the primary cations in sequestration reactions.
While I will concentrate my efforts on olivine only, efforts continue to make carbonation of serpentine a more feasible sequestration option. Park et al. (2003), for example, increased the rate of dissolution of magnesium from serpentine by adding a mixture of 1 vol% orthophosphoric acid, 0.9 wt% of oxalic acid and 0.1 wt% EDTA (ethylenediaminetetraacetic acid); they greatly enhanced the Mg leaching process of ground serpentine while preventing the precipitation of Fe(III) on the surface of the mineral particles (Park et al., 2003). When this acidic solvent was used for the aqueous mineral carbonation, the overall process was limited by the rates of dissolution of CO 2 and dissociation of carbonic acid, rather than the dissolution rate of the serpentine (Park et al., 2003). However, the temperature at which these processes occurred is ~500°C and requires too much energy input to make sequestration a viable option.
In fact, serpentinite-based carbon sequestration has a useful advantage over peridotite-based models in one sense: the carbonation of serpentinite results in considerably less volume change than carbonation of peridotite. Because carbonation of serpentinite results in a smaller volume difference, the chance of clogging pore spaces is reduced, allowing CO 2 -rich fluids to percolate over longer time scales in the subsurface during in situ carbonation (Beinlich et al., 2012). Free flow of pore fluids allows the reaction to proceed continuously, though the slow rate of carbonation of serpentinite (Gerdemann et al., 2007) may not result in the generation of sufficient heat, through exothermic reaction to bring the surrounding rocks up to temperature.
The focus here will be on the divalent cation elements in olivine that occur in abundance in ultramafic rocks, i.e., calcium, iron, and magnesium; these elements are particularly good at binding with carbonate anions to form carbonate minerals (O'Connor et al., 2005).
Since the CROMO surface rocks have considerably lower olivine: serpentine ratio and also contain abundant greenstone and clay minerals, they are not ideal for in situ carbon sequestration. However, they still have potential as a source for ex situ sequestration, whereby the rocks are mined, ground up, heated, and exposed to CO 2 -rich fluids and allowed to react. This process is much more controversial, given dollar costs, carbon footprint, and environmental impacts, and is discussed in more detail in section 4.9.1.

Point Geochemistry via SEM-EDS, to Determine Cation Abundances
SEM-EDS was applied to areas of interest on selected thin sections, for which separate mineral phases were analyzed for specified elements. Geochemical data for the ultramafic mineral phases relevant to carbon sequestration are presented below, as average wt% of each element from multiple samples analyzed.
CROMO2-134, 40 m depth: The serpentine phases from this section average 22.0 wt% Mg and 11.3 wt% Fe and less than 1 wt% Ca (Tables 3 and 5). The serpentine from this interval is in the form of a massive matrix-like lizardite ( Fig. 31 and 32) and chrysotile, which grows laterally in vein-like structures (Fig. 31). There is also abundant magnetite scattered throughout the thin section, either as a massivetextured rim-like growth or as octahedral individual crystals (Fig. 31). The magnetite is enriched in iron, with weight percentage averaging 70.4% (Table 4).  (Table 6). Aluminum in these rocks was extremely low; 0.06 wt% Al was the maximum value obtained. The serpentine phases in this section were predominantly of a massive texture with no evidence any vein-like growths. There were two scant magnetite crystals as well (Fig. 33), but they were not analyzed for elemental abundance.
Homestake Mine Pit Borehole (HMPB) at 110 m depth: The thin section from HMPB shows a different mineralogy. As in the petrography image, there are far more pyroxene than olivine crystals. This thin section provided geochemical data for deeper CRO olivine and pyroxene phases, yielding valuable insight into the sequestration capability of the CRO rocks.
Six fields of view were examined via SEM-EDS. The first region of interest (ROI) contained a sample of pyroxene amid a groundmass of serpentine (Fig. 34).
The serpentine was of massive texture and was likely lizardite. The geochemical make-up of the serpentine phase was lower in magnesium (18.4%) and higher in calcium (1.33%) and iron (12.4%) than any of the previous serpentine samples (Table 7).
Pyroxene is visible in two different ROI (Figs. 34 and 35). The pyroxene in  Table 2). The pyroxene crystal analyzed here had higher calcium (22.3%) than any of the serpentine or olivine minerals analyzed. The crystal from the ROI in Fig. 34 also showed minor amounts of iron (3.9%) and magnesium (7.4%) for a total divalent cation percentage of 33.6%. The crystal in  (Table 12) of magnesium (24.5%) than any of the other mineral phases present in any of the ROI. The olivine from the ROI in Fig. 37 contains less Mg (Table 13) than the portion of olivine sampled from Fig. 36, and has relatively the same concentration of Mg as the serpentine phases sampled from all depths.
The other ROI of the slide from Fig. 37 and the other prevalent mineral phase in the HMPB slide is a spinel group mineral enriched in iron, chromium, magnesium, and aluminum (Tables 14, 15, and 18). The spinel began as chromite (FeCr 2 O 4 ) and due to ion replacement, a ferritchromite rind forms that is completely devoid of aluminum (Ulmer, 1974). As mineral assemblages alter, iron may also be concentrated in magnetite (Fe 3 O 4 ), which tends to form rims around the ferritchromite rinds (Ulmer, 1974), which can be seen in the CRO rocks (Figs. 38 and 39). As a result, the spinel takes on the characteristics of a zoned crystal, with chromium and aluminum concentrated in the center and, iron and chromium on the inner edge, and iron with little to no aluminum or chromium on the outer edge (González-Mancera et al., 2007).
Iron and magnesium are both mobilized in the course of serpentinization; the inventories of these elements are lower in the serpentine phase products than they are in the initial olivine phase. During serpentinization and carbonation, minerals with higher Mg# (i.e., nMg/(nMg+nFe), where n is number of moles)) than parent olivine and pyroxene are formed (Evans 2008)   Pyroxene contained minor amounts iron and aluminum; however this cation replacement alters the percentages of magnesium and calcium.

Bulk Solid Geochemical Data from X-Ray Fluorescence
Additional geochemical data was obtained for samples from CROMO1 and CROMO2. The XRF unit operated under atmospheric pressure rather than under vacuum, thus concentration data for elements lighter than sulfur including magnesium and aluminum data are absent.
Given the relatively shallow total depth of coring (to only 30 m and 45 m below surface), no strong pattern emerged from calcium or iron data. One core sample per section was analyzed, thus these data should be taken as spot samples of the entire cored volume. The highest concentration analytes were calcium and iron (Tables 3  and 4). Nickel, manganese, chromium, and titanium were also abundant in the cores, with sections of the core often containing 1,000 ppm or more (see tables 20 and 21). A strong potassium signature in the geochemical data matches up with the XRD evidence of greenstone minerals, mirroring K-inventories in altered clay phases. Potassium and sodium often substitute for one another in feldspar minerals, so tracking potassium in the cores could help indicate where greenstone intercalations are likely to be found in a dominantly serpentinitic set of cores.

Porosity and Permeability Constraints
Maintaining travel paths for the CO 2 -rich fluids requires both creation and Carbonation efficiency can be hindered by the formation of minerals on the surface of olivine grains. Magnetite and/or silica-rich saponite and talc form around the rims of olivine grains, preventing the dissolution of olivine and/or pyroxene.
Eliminating the "protective" barrier from the olivine grains is critical to allowing the dissolution reaction (Eq. 14) to progress.
Curtailing unwanted mineralization on olivine grains is critical. A study by Andreani et al. (2008) showed that silica-rich layers form on olivine grains at the greatest frequency during diffusion-dominated transport of the flowing solution.
Diffusion-dominated transport implies that transport is controlled by a concentration gradient and not by advective forces and, as such, is much slower.
The reduced speed provides sufficient time for the silica-rich layers to form,

Solution Chemistry
The final component of the carbonation process is the composition of formation fluids. Supercritical CO 2 would be the most reactive available solution, but obtaining it would be difficult and maintaining a temperature and pressure regime capable of supporting a supercritical state may not be economically feasible.
Compressed liquid CO 2 is more feasible, but still requires processing to obtain.
Dissolving carbon dioxide into solution is the next best option. O'Connor et al., (2005) conducted a set of experiments designed to determine the optimal combination and concentration of additives to enhance the rate of carbonation.

pH
pH plays a conflicting role in carrier solution chemistry (O'Connor et al., 2005). An acidic solution would better serve a solution targeting olivine dissolution.
As can be seen from Eq. (14), a supply of H + ions is needed for the reaction to proceed. The reaction for carbonation of Mg 2+ ions requires a basic solution, however. As in Eq. (15): For this reaction to proceed to the right, H + ions must be consumed, which in a basic solution is accomplished by free OHions that will react as follows: O' Connor et al. (2005) showed that the addition of sodium bicarbonate (NaHCO 3 ) to the solution reduced the required residence times to achieve a reaction of approximately 80% from 24 hours to 6 hours, while holding all other carbonation constants stable. It was proposed that the addition of NaHCO 3 causes the Mg in olivine to interact directly with bicarbonate ions (Eq. 17).
OH -+ CO 2 (aq) → HCO 3 - Additionally, O'Connor et al. (2005) found that the addition of NaCl also improved the reaction rate, possibly due to Na + ions modifying the surface charge of silicate particles, which expedite ion exchange at the solid liquid interface. Clions may form complexes with magnesium, further increasing solubility of magnesium in solution (O'Connor et al., 2005); interestingly, regional groundwaters very near the McLaughlin locality exhibit some brine characteristics derived from interaction with local Great Valley sedimentary sequences and could serve as a readily available source of these ions (Goff et al., 2001). In general, the maximum concentrations of NaHCO 3 and NaCl found to be soluble in water at ambient conditions are 0.64 M NaHCO 3 and 1M NaCl. Overall, there is now sufficient experimental evidence for enhanced carbonation to serve pilot studies in the field in the CRO.

Calculation of CO 2 Sequestration Potential in the CRO
Using the volume of reactable rock, the mineralogy of the rock, and the chemical compositions of the minerals which compose the rock, the approximate total carbonation yield is calculated here. If the calculation is based solely on the magnesium content of olivine (Goff et al., 2000), a conservative, minimum estimate is reached for the possible amount of CO 2 that can be sequestered. Examples of the calculations involved are provided below (from Kelemen and Matter, 2008).
The sequestration potential of these rocks is increased by including calcium, magnesium, and iron in olivine in the reaction with CO 2 . By repeating the same computation for iron and calcium and summing the three figures, it is possible to obtain a total CO 2 sequestration amount, binding all three cations.
The difference between using only magnesium and using all three cations is 23,600 Mt of CO 2 ; this is a large mass of carbon, but within the same order of magnitude. In this particular example, considering iron and calcium results in a 34% increase in the amount of CO 2 sequestered.
Certain variables that factor into sequestration potential impact the estimated potential sequestration quantities to greater and lesser degrees. Figures   41, 42, and 43 display sequestration potential, with variables such as formation depth, formation density, and olivine percentage in the formation being manipulated. Considering a wide range of variables creates a range of possible CO 2 drawdown figures. These figures constrain a most conservative estimate and also allow for a best-possible scenario. Figure 41 shows sequestered CO 2 mass against density of the formation.
Density of the formation affects sequestration potential in that a higher density calculates to a greater mass of reactable rock, which in turn allows for a greater mass of carbonate to form from the rock. Density is primarily dependent upon the mineral composition of the rock. If peridotite is around 30% olivine and 70% serpentine as is the case in the Semail Ophiolite (Kelemen and Matter 2008), and is the ratio I have chosen for the CRO then the density would equate to ~ 2800 kg/m 3 . Figure 42 shows the olivine fraction of the peridotite against sequestered CO 2 mass. Since olivine is the most reactive of the ultramafic minerals (aside from rare wollastonite), a higher percentage of olivine in the rock will increase the sequestration potential of the rock. A study by Choi et al. (2008) showed a wide range of serpentinization of Coast Range peridotite. Some sites were as low as 18% serpentinized and some were as high as 95% serpentinized. An average of all the values acquired from this study reveals a mean serpentinization of 61.5%. The Semail Ophiolite consists of ~30% olivine peridotite (Kelemen and Matter, 2008), so to again be conservative with estimates, I chose to round the 61.5% obtained from Choi et al. (2008) to 70%, based on review of the new data in this study. I have chosen values as low as 20% and as high as 40% to determine a maximum and minimum percentage of olivine in the CRO rocks.
By considering possible depths to which the Coast Range Ophiolite ultramafic unit extends, I have created a range of volumes for the CRO. A deeper formation simply means more there is more volume and thus more mass of rock to react.
Depth is difficult to determine in the formation, as complex tectonism has confounded the determination of depth for each individual portion of the CRO. Wakabayashi and Dilek (2000) provide an average depth of around 5000 m, considered here as a generous estimate. 5500 m is taken as a maximum depth and 3500 m as a minimum depth, so as to err on the conservative side and to recognize the practical difficulty of drilling 5000 m or more.
After all of the possible scenarios were evaluated, the Mg-only case (Table   22) and the Mg+Fe+Ca case (Table 23) are presented. Furthermore, the maximum possible value and the minimum possible value, based on the parameters set forth in this paper, are hereby established-providing extremely conservative and extremely generous estimates for CRO sequestration potential (Table 24). Note that the upper limit of sequestration potential is only 3.6 times greater than the lower limit, showing that even given variability in host rock and cation reactivity, the sequestration potential is in fact fairly well constrained.
The averages obtained are close in value to the calculations shown in (Eqs. 19 and 20). For Mg-only sequestration, a total of 6.8 x 10 13 kg CO 2 (6.8 x 10 4 Mt, 7.5 x 10 10 tons) is able to be sequestered in the CRO and for Mg, Ca, and Fe, the total amount of CO 2 that can be sequestered is 9.1 x 10 13 kg (9.1 x 10 4 Mt, 1.00 x 10 11 tons). There are currently ~2.9 x 10 15 kg CO 2 in the atmosphere; if we were to draw down all 9.1 x 10 13 kg of CO 2 , we could eliminate 3.1% of the carbon dioxide in the atmosphere. If only Mg reacted, then 2.3 % of the carbon dioxide could be drawn down.

Conclusion
The Coast Range Ophiolite, situated along a 700 km stretch in California, provides an opportunity to study and work with highly reactive rocks that are in disequilibrium with surface conditions. I proposed to evaluate the harnessing of this disequilibrium to react the ultramafic rocks with carbon dioxide, in an effort to sequester atmospheric CO 2 , as has been proposed as a large scale means of geological sequestration of carbon recently (Kelemen and Matter, 2008;Goldberg et al., 2009;Gislason et al., 2010). Geochemical and mineralogical data enabled calculation of the potential of the CRO rocks for carbon sequestration capability.
XRD, SEM-EDS, and petrography determined that surface rocks were rather thoroughly serpentinized, with no visible trace of relict olivine in any of the surface rocks sampled. One deeper serpentinite sample from the vicinity had prevalent olivine and pyroxene. Through the compilation of SEM-EDS mineral geochemistry data, with relevant geochemical data from sources from previous research, the total quantity of CO 2 that may be sequestered in CRO ultramafic rocks was estimated as ranging from 4.60x10 13 to 1.24x10 14 kg CO 2 . For comparison, Kelemen and Matter (2008) estimated the larger Semail Opholite's potential total drawdown as 6.9x10 14 kg CO 2 .
In all, the quantity of carbon dioxide that can be sequestered in the Coast Range Ophiolite alone is significant. Whether the capacity of the CRO is based on magnesium only (2.3% of atmospheric CO 2 ) or magnesium, iron, and calcium (3.1% of atmospheric CO 2 ), the capability of the CRO to serve as a mitigating agent should not be ignored. If we are to return to 350 ppm (as recommended by the IPCC) from present (as of 1/3/2013) 392 ppm, we would need 11% of the CO 2 currently in the atmosphere to be consumed. This is not to say we are capable of pulling that quantity of CO 2 out of the atmosphere, but rather we are able to sequester a quantity that represents a percentage of current atmospheric CO 2 . Industry will continue to produce carbon dioxide, and the best way to currently control emitted CO 2 is to capture and store it, preferably in close propinquity to where it is produced.
Since there are ~2.9x10 15 kg CO 2 in the atmosphere, 3.2x10 14 kg of that quantity need to be sequestered. The CRO represents 14% (Mg-only) -39% (Mg, Fe, Ca) of that total, which is insufficient by itself, but represents a significant value for a moderate sized ophiolite.
While serpentinites and peridotites remain compelling as candidate rock hosts for the geologic sequestration of carbon, research regarding their utility in more than experimental scales is still inchoate. Any field-scale test of carbon sequestration potential in the CRO must consider also environmental impact assessments (the protection of rare serpentine endemic eco-systems is of paramount importance), dangers of carbon leakage from an incompletely sealed subsurface reservoir, and net carbon emissions of the entire process, among other concerns. Given the considerable areas of variably altered peridotite in continental and submarine settings, serpentinite-based carbon sequestration would be an enticing mitigation strategy for anthropogenically enhanced atmospheric CO 2 levels, should the reaction conditions and longevity be sufficiently controllable. Fig. 1. Graph shows CO 2 levels in parts per million in the atmosphere since 1958. Sawtooth pattern is due to seasonal changes as shown in inset plot at lower right (NOAA 2010).      Å shows first slight presence of lizardite. High intensity peak at 3.3419 Å represents quartz, likely a product from erosion of greenstones. Evidence of greenstone exists with albite peak at 3.1965 Å.

Appendix A -Figures and Tables
High quartz:albite ratio shows more stable quartz has survived while albite has been broken down over time. The second highest peak is a clay mineral(s) on the left, also indicative of erosion.            Å and 4.6 Å refer to lizardite, but the horizon is completely dominated by the presence of the clay, which is the tallest peak on the far left.                Heavier elements show up as lighter colored on the image. The fractured crystal with slightly higher relief than the groundmass is a spinel, while the groundmass itself is serpentine and the lightly colored veined mineral is magnetite.           Table 9 Serpentine phase mineral weight percentages from HMPB slide (Fig. 35)

Table 19
Combined averages for the elemental weight percentages of each of the major mineral phases in the core samples from CROMO1, CROMO2, and HMPB.   Table 22 List of values of sequestered CO 2 masses for Mg as the only reactive cation, ordered by sequestration potential.

Appendix C -Ex Situ vs. In Situ Carbon Sequestration Options
In situ and ex situ sequestration are both viable options for the Coast Range Ophiolite. There are certainly benefits and disadvantages to both, meriting exploration.
Ex situ sequestration is analogous to strip mining. During ex situ processing, rock must first be extracted from the underlying rock before it can be processed.
The extraction itself is both a labor-and energy-intensive process, requiring removal of overlying soil and plant life. The amazing biodiversity of serpentine endemic plant and soil communities would make it difficult to justify this approach, particularly in conserved lands of any sort. Given the dearth of serpentine outcrops and soils in specific climates, removal of the endemic flora would prove highly controversial in addition to being environmentally deleterious.
A benefit of ex situ sequestration is that when an entire block of rock is extracted, it can be ground up to a diameter of a specific grain size, thereby increasing the surface area for chemical reactions-but this approach is not without costs. Decreasing the grain size of the rock requires copious amounts of grinding; the grinding process demands energy, which results in CO 2 generation, thus negating the entire motivation behind sequestration.
In the case of the Coast Range Ophiolite, the rocks of interest are at least 100 m below the surface. In related units elsewhere in the region, the olivinerich peridotites may be even deeper beneath the surface. Removing shallow layers of olivine-poor serpentinite would require even more energy and destruction of landscape. Deeply buried olivine-rich lithologies, as at the McLaughlin locality of the CRO, cause the cost of ex situ sequestration to escalate. Quarrying rocks may be more viable in cases of shallowly buried or exposed blocks of peridotites in deserts or other remote locations.
O' Connor et al. (2005) estimated to cost of sequestration for pure olivine at $69/ton CO 2 sequestered. Serpentine carbonation costs were reduced to roughly $78 per ton of CO 2 sequestered, including a necessary heat-treatment step prior to reacting the serpentine with CO 2 . While thermal activation makes serpentine an effective medium of sequestration, the process results in more CO 2 generated than is sequestered by the process, making it impractical (O'Connor, 2005). The report does not mention if the cost and CO 2 generation figures include the process of mining the rocks. Mining is energy-intensive and surely would subtract from the benefits of the process.
Both in situ and ex situ sequestration generate carbonate minerals as a result of the chemical reaction, but ex situ processes allow access to the carbonate. Because in situ sequestration happens so far beneath the surface, access to the minerals is extremely limited. Industrial uses of magnesite, calcite and siderite include uses in antacids, construction materials, pigments, and cement-making.
In situ sequestration is seen as less environmentally deleterious, as it does not require the extraction of rock from the subsurface. Instead, holes of varying depth are drilled, and carbon-rich fluid is pumped deep into the subsurface where it is allowed to react freely with rocks. There are still untallied costs and energy inputs associated with in situ carbonation, yet there are simply fewer steps involved, thus likely less cost in dollars. Part of those costs comes from the numerous holes that need to be drilled. Over a lengthy expanse of ultramafic rocks, thousands of holes would need to be drilled. Cost is also incurred through the necessary heating of rock prior to reaction. This is accomplished by pumping preheated water into a borehole. Costs can be considerably reduced if the exothermic reaction (Eq. 9) of carbonation preheats the surrounding rock and allows continuous heating throughout the process, eliminating the need for additional inputs of thermal energy.