Serpentinites offer a highly reactive feedstock for carbonation reactions

Transcription

1 Serpentinite Carbonation for CO 2 Sequestration Ian M. Power 1, Siobhan A. Wilson 2, and Gregory M. Dipple /13/ $2.50 DOI: /gselements Serpentinites offer a highly reactive feedstock for carbonation reactions and the capacity to sequester carbon dioxide (CO 2 ) on a global scale. CO 2 can be sequestered in mined serpentinite using high-temperature carbonation reactors, by carbonating alkaline mine wastes, or by subsurface reaction through CO 2 injection into serpentinite-hosted aquifers and serpentinized peridotites. Natural analogues to serpentinite carbonation, such as exhumed hydrothermal systems, alkaline travertines, and hydromagnesite magnesite playas, provide insights into geochemical controls on carbonation rates that can guide industrial CO 2 sequestration. The upscaling of existing technologies that accelerate serpentinite carbonation may prove sufficient for offsetting local industrial emissions, but global-scale implementation will require considerable incentives and further research and development. KEYWORDS: carbon sequestration, climate change, mineral carbonation, geoengineering, serpentinite INTRODUCTION Carbon sequestration research and technology is motivated by concerns that increasing atmospheric carbon dioxide (CO 2 ) concentrations will drive long-term change to Earth s climate (IPCC 2005). The concentration of atmospheric CO 2 is expected to continue to rise due to the availability of inexpensive fossil fuels and increasing energy demands driven by population growth. Mineral carbonation is an approach for sequestering CO 2 that utilizes the reaction of alkaline earth silicate and hydroxide minerals with CO 2 to form carbonate minerals. It has been an active area of greenhouse gas (GHG) mitigation research since the mid-1990s. This research seeks to capitalize on the large global reserves and wide distribution of ultramafic rocks, such as serpentinite, and the long-term stability of carbonate minerals (Lackner et al. 1995). Global estimates of accessible serpentinized peridotite reserves are in the hundreds of thousands of gigatonnes (Lackner 2002); carbonation of this resource could mineralize the CO 2 produced from the combustion of all known coal reserves (10,000 Gt C; Lackner et al. 1995). GHG mitigation strategies that employ serpentinite include (1) ex situ mineral carbonation reactions in high-temperature industrial reactors and low-temperature 1 Mineral Deposit Research Unit Department of Earth, Ocean and Atmospheric Sciences The University of British Columbia Vancouver, British Columbia V6T 1Z4, Canada of corresponding author: ipower@eos.ubc.ca 2 School of Geosciences, Monash University Clayton, VIC 3800, Australia ELEMENTS, VOL. 9, PP APRIL carbonation of alkaline industrial waste, such as serpentinite mine tailings, and (2) in situ mineral carbonation processes initiated by the injection of CO 2 into serpentinite-hosted aquifers or the injection of CO 2 and seawater into peridotite and serpentinite at higher temperatures (>150 ºC) (FIG. 1). The challenge is to develop carbonation technologies that operate at a scale and rate that are relevant to anthropogenic GHG emissions locally and globally. Serpentinite carbonation occurs naturally under geochemical conditions that are similar to those currently being investigated for the development of industrial and geoengineering strategies for CO 2 sequestration. Carbonate minerals are produced during serpentinization (Evans et al this issue) and during hydrothermal alteration and weathering of serpentinite. This propensity to form carbonate minerals in ultramafic terranes reflects the thermodynamic stability of these minerals in the presence of CO 2. In natural systems, the extent of carbonation may be limited by the availability of CO 2 or by sluggish reaction kinetics (e.g. Zhang et al. 2000; Power et al. 2009; Beinlich et al. 2012). Natural analogues provide valuable insight into the mechanisms of carbonation and the geochemical controls on the scale and rate of carbonation reactions. Thus, understanding the processes and limitations of mineral carbonation in natural systems is a useful guide for the development of serpentinite carbonation for CO 2 sequestration. CARBONATION REACTIONS In nature, carbonate minerals can form during serpentinization or during hydrothermal carbonation and weathering of serpentinites (FIG. 2A). Industrial mineralcarbonation processes are commonly represented by the reaction of olivine or serpentine with CO 2 to form magnesite + quartz ± H 2 O. Similar reaction pathways are preserved in serpentinized peridotites and in carbonate-altered serpentinites. Serpentinization reactions may proceed by the hydration of olivine to produce serpentine + brucite or the hydration carbonation of olivine to produce serpentine + magnesite (FIG. 2A). Subsequent carbonate alteration of serpentinite can produce talc + magnesite and magnesite + quartz (listwanite) assemblages (Hansen et al. 2005) (FIG. 2A). These reactions are relevant to carbon

2 commonly controlled, at least in part, by microbial activity. Microorganisms are ubiquitous in the environment and have a variety of metabolisms that may enhance dissolution of the mineral constituents of serpentinite. For instance, biological effects (e.g. microbial acid production) enhance weathering on Earth by at least 100 to possibly >1000 times (Schwartzman and Volk 1989). Microorganisms may also contribute to carbonate precipitation while playing an important role in the global carbon budget (Ferris et al. 1994). Microbial processes may alter water chemistry by affecting factors such as ph and the concentrations of cations and dissolved inorganic carbon, which can induce carbonate precipitation. Such processes include photosynthesis, ammonification, denitrification, sulfate reduction, and manganese and iron oxide reduction (Ferris et al. 1994; Dupraz et al. 2009). Some of these processes are now known to play an active role in the formation of landscape features, such as hydromagnesite magnesite playas, that result from serpentinite weathering. Mg2Si2O Reactants to products (A) Pertinent reaction pathways at high and low temperatures that lead to carbonation of peridotite and serpentinite. FIGURE 2 E LEMENTS MgCO3 Mineral carbonation: 2.0 magnesite magnesite + quartz MgCO3 + SiO2 CaMgSi2O6 MgCO33 3H22O hydromagnesite + amor. silica Mg5(CO3)4(OH)2 4H2O + SiO2 diopside Mg(OH)2 nesquehonite g & rin tic e o i ab ath e W Mg2SiO4 brucite lansfordite talc + magnesite Mg3Si4O10(OH)2 + MgCO3 ic ot bi forsterite MgCO3 5H2O CARBONATION Carbonate alteration 3.0 Mg33Si22O55(OH)44 artinite serpentine + magnesite MgCO3 Mg3Si2O5(OH)4 + M 4.0 serpentine magnesite + CaCO3 enstatite serpentine + brucite Mg3Si2O5(OH)4 + Mg(OH)2 Mg2CO3(OH)2 3H2O olivine Mg2SiO4 B HYDRATION Tonnes reactant mineral/tonne CO2 Se rpe nti ni z ati on A The mineralogy of both feedstock and carbonate precipitate affects the footprint of an industrial carbon-sequestration operation. The mass ratio of CO2 sequestered to reactant mineral depends on the amount of Mg present in the dypingite + CaCO3 Serpentinite carbonation also occurs as a consequence of weathering at the lower temperatures that dominate at Earth s surface (FIG. 2A). Globally, continental weathering sequesters an estimated 300 million tonnes of CO2 each year (Seifritz 1990). While serpentinite represents only a small proportion of crustal rocks, it contributes a disproportionately high share of weathering products because of its reactivity. The weathering of ultramafic rocks is hydromagnesite + CaCO3 sequestration technologies that use high-temperature carbonation reactors and the geothermal carbonation of serpentinite (FIG. 1). hydromagnesite Conceptual diagram depicting geoengineered systems designed for serpentinite carbonation. Graphical elements are courtesy of the Integration and Application Network, University of Maryland Center for Environmental Science ( umces.edu/symbols). FIGURE 1 Mg5(CO3)4(OH)2 4H2O Cumulates Magnesite precipitation, although thermodynamically stable, is inhibited in near-surface environments by the strong hydration shells surrounding aqueous Mg2+ ions. Consequently, hydrated magnesium carbonate minerals, such as hydromagnesite, dypingite and nesquehonite (FIG. 2), form instead (e.g. Königsberger et al. 1999). The variability in the hydration states (i.e. the H 2O content) of Mg carbonate minerals reflects the efficiency with which natural geochemical systems are able to break the hydration shells surrounding Mg2+ ions. Natural systems that produce hydrated Mg carbonate minerals may therefore provide valuable insight into reaction pathways that could be used to improve the rate and efficiency of mineral carbonation. Hydrated Mg carbonate minerals form most commonly in hydromagnesite magnesite playas and alkaline spring systems, which represent natural analogues to low-temperature mineral carbonation in alkaline wastes, such as serpentinite mine tailings (FIG. 1). artinite + CaCO3 nesquehonite + CaCO3 Gabbros Serpentinized harzburgite + dunite Mélange dypingite Geothermal carbonation of serpentinized peridotite Sheared serpentinite Mg5(CO3)4(OH)2 5H2O Electrical generation and CO2 capture CO2 injection into serpentinite-hosted aquifer lansfordite + CaCO3 High-T carbonation reactors Carbonation of serpentinite tailings 2.0 Tonnes product mineral/tonne CO2 (B) Efficiency of carbonating the major mineral components of serpentinite and peridotite to form Mg carbonate minerals (see text) 116 A PR IL 2013

3 starting mineral. The mass ratio of CO 2 sequestered to product mineral depends on (1) the ratio of CO 2 to MgO in the mineral precipitate and (2) the amount of H 2 O in the carbonate-mineral precipitate (FIG. 2B). For instance, less reactant mineral is required to sequester 1 tonne of CO 2 in minerals with a 1:1 CO 2 :MgO ratio (e.g. magnesite, lansfordite, nesquehonite) than in minerals with lower CO 2 :MgO ratios (e.g. hydromagnesite, dypingite, artinite) (FIG. 2B). Importantly, the mass ratio also depends strongly upon the hydration state of the carbonate mineral. The hydration state of the carbonate mineral impacts the water budget for mineral carbonation as well as the stability of CO 2 fixation, with less hydrated minerals in general being more stable. Thus, reactant mineral feedstocks and carbonation products need to be carefully selected and controlled in order to manage the stability of carbonate minerals, water resources, and the cost of transportation and processing associated with serpentinite carbonation. The rate of CO 2 fixation is related to dissolution kinetics, which favor the reaction of brucite or olivine over serpentine. However, many factors contribute to dissolution rates, including the extent of silica polymerization in alkaline earth silicates, metal cation size, and the availability of reactive surface area (Brantley 2003). The inherently high reactive surface areas of serpentine-group minerals, often enhanced by comminution during mineral processing, contribute to a relatively high dissolution rate for serpentine in industrial wastes. Carbonation rates are dictated by the rates of several processes, including the initial dissolution of silicate and hydroxide minerals, the passivation of mineral surfaces by secondary minerals, the availability of CO 2 in solution, and the dehydration of Mg 2+ ions. NATURAL ANALOGUES Diverse reaction pathways and limits on reaction rates are evident in natural examples of serpentinite carbonation, the study of which provides a framework for developing geochemical and biogeochemical strategies for enhancing carbonation rates in geoengineered systems (FIG. 3). Talc Magnesite and Listwanite Alteration Talc magnesite and magnesite quartz (listwanite) rocks, essentially fossil analogues of mineral-carbonation systems, are common in ultramafic terranes on Earth. Their abundance suggests that conditions favorable for mineral carbonation exist at shallow levels of the crust. While modest carbonation may accompany serpentinization, extensive carbonate alteration typically represents a separate hydrothermal event (Naldrett 1966), producing talc magnesite and magnesite quartz assemblages (FIG. 3A). Carbonate alteration occurs at temperatures ( C) and pressures (equivalent to 3 5 km depth) that may be within reach of deep, industrial CO 2 -injection systems, which currently extend to 2.5 km. In ultramafic terranes, the carbonation reaction is driven by CO 2 infiltration (Naldrett 1966), and the reaction sequences correlate with changes in the nature of permeability (Hansen et al. 2005). Phase-equilibrium calculations indicate that reaction sequences preserved in carbonate-altered serpentinite can be attributed to changes in the CO 2 content of the fluid phase, which could be controlled in a CO 2 injection scenario. Talc magnesite alteration of serpentinite is nearly isovolumetric, because the volume gain of carbonation is largely balanced by the volume loss associated with dehydration (Naldrett 1966). Thus porosity and permeability may be maintained during reaction. In partially serpentinized rocks, talc magnesite alteration of olivine produces rock volume increases that can be preserved as extension veins (Hansen et al. 2005). Sharp reaction fronts in exhumed systems indicate that reaction rate and progress are governed by the rate of CO 2 supply to the reaction site (Beinlich et al. 2012). Listwanite formation is generally associated with a large increase in rock volume and is typically associated with extension vein systems attributed to reaction-induced cracking. This self-cracking mechanism may be desirable for injection of CO 2 into serpentinites, as it would perpetuate mineral carbonation reactions by generating permeability and reactive surfaces. Listwanite Springs & travertines A B C Hydromagnesitemagnesite playas Processes: Limitations: Geoengineered systems: High-T reactions Self-cracking Reactive surface area CO 2 supply High-T & geothermal carbonation Subsurface dissolution Surface precipitation Carbonate precipitation Water and CO 2 supply CO 2 injection into serpentinite-hosted aquifer Weathering (abiotic & biotic) Acid generation (silicate dissolution) Alkaline ph (carbonate precipitation) Magnesite precipitation CO 2 supply Carbonation of serpentinite mine tailings FIGURE 3 Natural analogues for serpentinite carbonation. (A) Listwanite formed under high-temperature and high-pressure conditions showing a self-cracking mechanism (British Columbia, Canada). (B) Alkaline spring in travertine in Oman. (C) Hydromagnesite magnesite playas formed from weathering of serpentinite in British Columbia, Canada. ELEMENTS 117 APRIL 2013

4 Springs and Travertines Springs and travertines are associated with serpentinites and variably hydrated peridotites. Near-surface carbonation of serpentinites and less hydrated peridotites is known to occur in low-temperature spring systems in Italy, California, and Oman (O Neil and Barnes 1971; Kelemen and Matter 2008). Natural carbonation of these Mg-rich rocks produces alkaline springs that form large travertine deposits (FIG. 3B), carbonate fracture and vein fillings, and carbonate chimneys at submarine hydrothermal vents. Weathering of ultramafic rocks produces two types of waters: (1) shallow, circumneutral Mg HCO 3 groundwater, which develops from the interaction of meteoric water with serpentinite and can produce hydromagnesite magnesite playas; and (2) deep, high-ph ( 11) Ca OH groundwater, which develops from Mg HCO 3 water when isolated from the atmosphere during carbonate alteration of serpentinite at depth (Barnes and O Neil 1969). Although both Mg 2+ and Ca 2+ are released from primary forsterite and pyroxenes in peridotite, Mg 2+ is preferentially incorporated into reaction products such as serpentine, brucite, and magnesite, thereby concentrating Ca 2+ in solution. The resultant Ca OH groundwaters are extremely depleted in aqueous CO 2. Thus, upon discharging at the Earth s surface as springs, they react rapidly with the atmosphere by absorbing CO 2 and precipitating Ca carbonate minerals. As an example, in Oman ~10 4 tonnes CO 2 per year are consumed by peridotite carbonation (Kelemen and Matter 2008). The long-term precipitation of calcite from emergent Ca OH spring water can produce extensive carbonate travertine deposits and highlights the capacity for alkaline solutions to absorb CO 2 directly from the atmosphere. The rate of carbon fixation is likely limited by the flow rate of the springs. Spring and travertine systems associated with serpentinites and variably hydrated peridotites represent natural analogues for carbonation in low-temperature, serpentinite-hosted aquifers (FIG. 1). Hydromagnesite Magnesite Playas At near-surface conditions, serpentinite carbonation can proceed in two spatially separated stages: (1) weathering of serpentinite that produces Mg HCO 3 groundwater and (2) Mg carbonate precipitation that typically occurs in closed basins where groundwater discharges. This process is exemplified by the formation of hydromagnesite magnesite playas (FIG. 3C) (Power et al. 2009). Mg carbonate precipitation is mediated by microorganisms and abiotic kinetic processes that include CO 2 degassing and evaporation of groundwaters discharged at the surface. For instance, cyanobacteria are able to induce the precipitation of dypingite through the alkalinization of their microenvironment in wetlands (Power et al. 2009). The playa deposits produced by these wetlands are dominated by hydrated Mg carbonate minerals such as hydromagnesite, which is a common dehydration product of dypingite. In this near-surface environment, serpentinite carbonation is limited by the slow kinetics of silicate-mineral dissolution and carbonate precipitation. In terms of stability, the transformation of hydromagnesite to magnesite, the most stable of the Mg carbonate minerals, occurs either by dissolution precipitation or by dehydration of hydromagnesite. It is estimated that this process requires on the order of tens to hundreds of years (Zhang et al. 2000). Early Holocene glaciolacustrine sediments underlie hydromagnesite magnesite playas in northern British Columbia, which suggests that these natural deposits are stable on the timescale of millennia and that artificial carbon sinks designed to mimic these Mg carbonate deposits should be stable for thousands of years, a period long enough to produce a greenhouse gas benefit. Hydromagnesite magnesite playas highlight the important role that biogeochemical reactions might play in carbon sequestration in alkaline wastes such as serpentinite mine tailings (Power et al. 2009). EX SITU INDUSTRIAL CARBONATION High-Temperature Carbonation Reactors Carbonation in high-temperature chemical reactors (FIG. 1) has received by far the most research activity (Huijgen and Comans 2005; Zevenhoven et al. 2011) and also bears the least resemblance to natural systems. High-temperature mineral carbonation targets the rapid conversion of silicate minerals, including serpentine and olivine, to magnesite and quartz at rates sufficient to neutralize the CO 2 emissions of a coal-burning electricity-generation facility (~ tonnes CO 2 per hour). A mineral-carbonation plant would operate using mineral feedstock delivered from a mine to a chemical reactor that is supplied with high-purity CO 2 captured from power plant emissions. The accommodation of power plant CO 2 fluxes requires rapid reactions and short residence times for mineral feedstock (i.e. on the timescale of hours). Most high-temperature carbonation strategies employ either gas mineral or gas water mineral reactions. Both dry and wet reaction systems have merit. Direct gas reaction offers simple process design and greater potential for recovering heat but suffers from slow reaction rates. The addition of water greatly enhances reaction rates but requires significant energy inputs and the conscription of water resources. Direct aqueous carbonation involves (1) dissolution of CO 2 gas into an aqueous solution, (2) dissolution of a Mg silicate mineral, and (3) precipitation of a carbonate mineral. The release of Mg from silicate minerals is likely a rate-limiting step in mineral carbonation. Serpentine carbonation is slow relative to olivine carbonation but can be enhanced (Balucan and Dlugogorski 2013). However, serpentinite continues to be a focus for mineral carbonation because of the accessibility of large deposits to serve as feedstock. Thermal pretreatment to dehydrate serpentine enhances reactivity substantially. Serpentine dissolution rates are also accelerated with chemical additives, including salts, acids, and complexing agents (Carey et al. 2004). In some hightemperature systems, serpentine carbonation is limited by the rate of CO 2 dissolution into the aqueous solution, as it is during the weathering of mine wastes and in other natural low-temperature environments. Other controls on rate are mineral-surface passivation and kinetic limitations to carbonate precipitation, which may also have relevance to accelerating low-temperature carbonation processes. Hybrid wet dry reaction systems include indirect carbonation, which converts serpentine to magnesium hydroxide (brucite) in an aqueous reactor and then dry-carbonates brucite to magnesite. Dry-carbonation of brucite occurs rapidly, yet is limited by the rate of production of brucite from serpentine or olivine. Similarly, pulverized brucite, such as that found in mine wastes, carbonates rapidly at low temperatures when exposed to fluids with elevated concentrations of dissolved CO 2 (Harrison et al. 2013). Mineral carbonation using high-temperature reactors is technically and energetically feasible; however, the cost of existing technology (~$100 per tonne CO 2 ) exceeds current carbon prices (currently about $5/tonne through the European Union Emissions Trading System). The expense reflects the costs of mining, energy requirements, chemical inputs, and pretreatments to enhance mineral reactivity. These penalties could be reduced if the exothermic nature of mineral carbonation reactions were harnessed. Alternatively, ELEMENTS 118 APRIL 2013

5 carbonation reactors could possibly be used effectively in industries that produce alkaline wastes, which would eliminate the cost of material extraction. Carbonation of Serpentinite Mine Tailings Industrial wastes such as serpentinite mine tailings are rich in the alkaline earth metals necessary for the production of stable carbonate minerals (Renforth et al. 2011; Bobicki et al. 2012). Consequently, these materials are of considerable interest as feedstocks for ex situ mineral carbonation. Serpentinite mine wastes are typically characterized by large reactive surface areas and, until recently, have been stockpiled in significant quantities without consideration for future use (Wilson et al. 2009; Bobicki et al. 2012). The annual production of ultramafic tailings from a large serpentinite-hosted ore deposit can be greater than 10 Mt, and correspondingly large amounts of mine waste accumulate over the life of a mine. For instance, the serpentinite-hosted chrysotile mines of Québec, Canada, produced approximately 2 Gt of chrysotile- and brucite-rich mine wastes, which could be conscripted for CO2 sequestration (Larachi et al. 2010). Hydrated Mg carbonate minerals have long been recognized as low-temperature weathering products of serpentine minerals and brucite (e.g. O Neil and Barnes 1971). These Mg carbonate minerals commonly form in two distinct environments within mine tailings storage facilities (FIG. 4A): (1) in surficial environments as efflorescent crusts and hardpans and (2) at depth within tailings as a cement that forms between mineral grains (Wilson et al. 2009). The surface expression of mineral carbonation is likely related to uptake of atmospheric CO2 into near-surface tailings waters coupled with evapoconcentration of these Mg-rich brines. Significant abundances of hydrated Mg carbonate minerals may develop at depth as carbonated surfaces become buried or as a consequence of ongoing reaction after burial (e.g. Pronost et al. 2012). The scale and rate of passive carbonation of serpentinerich mine tailings can be significant: on the order of 0.1 to 10 kg of CO2 per tonne of tailings per year, which can represent a substantial, but incomplete, offsetting of a mine s greenhouse gas emissions (Harrison et al. 2013). A major limitation on passive carbonation in mine tailings is the sluggish rate of CO2 uptake into tailings waters. Harrison et al. (2013) have demonstrated that a dramatic acceleration over passive rates of carbonation may be achieved by delivering CO2 -rich gas to tailings. Thus, storage facilities for serpentinite mine tailings could be redesigned to take advantage of this by injecting CO2 from onsite energy generation into the tailings (FIG. 4B). A tenfold increase in passive carbonation rates would offset total greenhouse gas emissions for a mine. A hundredfold acceleration would sequester carbon at a rate commensurate with the largest carbon capture and storage demonstration projects currently operating (~1 million tonnes of CO2 per year). CO2 injection into mine tailings could offer valuable insights by acting as a testing facility for injection into geological formations, such as serpentinite-hosted aquifers. Enhanced carbonation of mine tailings could also be achieved using microbial processes that facilitate serpentinite dissolution and Mg carbonate precipitation by creating optimal conditions for growth of selected microbes. The layering of waste sulfides and elemental sulfur colonized by Acidithiobacillus spp. onto serpentinite tailings would increase the rates of silicate mineral dissolution by at least one order of magnitude (Power et al. 2010). Cyanobacteria are able to induce carbonate precipitation, typically by the alkalinization of their microenvironment (Ferris et al. 1994; Power et al. 2011). The utilization of photoautotrophic microbes, such as algae and cyanobacteria, for mineral carbonation is particularly attractive because they use sunlight as an energy source and CO2 as a carbon source. Mine sites hosting serpentinite tailings could be geoengineered to take advantage of bioleaching to produce Mg-rich waters that may be directed into specially designed carbonate precipitation ponds (Power et al. 2011) (FIG. 4C). Photoautotrophs would not only produce carbonate minerals in these ponds; their biomass might also be harvested for biofuel or valuable by-products. Bioleaching of serpentinite tailings and subsequent microbial carbonate precipitation have a geologically stable analogue in natural hydromagnesite magnesite playas and offer a low-energy, low-temperature strategy for sequestering atmospheric CO2. Providing industries that produce alkaline wastes (e.g. serpentinite mine tailings) with incentives to manage their carbon footprint will drive innovation and technology development, creating new approaches beyond those described here. IN SITU INDUSTRIAL CARBONATION Serpentinite-Hosted Aquifers Mineral carbonation occurs naturally in the subsurface as a result of fluid rock interactions within serpentinite, which occur during serpentinization and carbonate alteration (e.g. listwanite). In situ carbonation aims to promote these Serpentinite tailings A A) Passive carbon sequestration at mine sites Tailings deposition Water reclamation pond Water infiltration Carbonate precipitation Impermeable barrier B B) CO2 injection into serpentinite tailings Evapoconcentration and gas exchange CO2 injection C C) Microbially accelerated carbonation Bioleaching Acid-generating substance Carbonate precipitation pond Microbial mats A schematic of (A) passive mineral carbonation in serpentinite mine tailings, (B) an abiotic geoengineering strategy employing CO2 injection, FIGURE 4 E LEMENTS and (C) a geoengineered tailings facility using bioleaching and microbial carbonate precipitation. 119 A PR IL 2013

6 reactions by injecting CO 2 into porous, subsurface geological formations, such as serpentinite-hosted aquifers (FIG. 1). The rate of CO 2 sequestration in this geological environment is controlled primarily by reactive surface area, temperature, ph, and the partial pressure of CO 2. Geothermal heating accelerates serpentine dissolution and may allow for the precipitation of magnesite. Cipolli et al. (2004) modelled CO 2 sequestration in a serpentinite aquifer at 60 ºC and 250 bars pco 2 ; they estimated a CO 2 sequestration rate of 33 g of CO 2 per kilogram of H 2 O per year. The advantage of in situ carbonation is that there is no need to mine, grind, and activate serpentinite, nor transport reactants or products, because CO 2 is injected on site. The major obstacles for implementation are the generation of reactive surface area, loss of porosity, and passivation of reactive surfaces. By analogy with the volume changes recorded in exhumed carbonated serpentinite, porosity losses will be minimized if injection is carried out in fully serpentinized rocks and uses the CO 2 content of injected fluid to promote formation of talc + magnesite rather than magnesite + quartz assemblages (Hansen et al. 2005). Alternatively, serpentine dissolution and magnesite precipitation may occur in separate zones, with hydraulic fracturing limiting reactive surface passivation (Boschi et al. 2009). Technologies that have been employed for CO 2 injection into deep saline reservoirs (e.g. Sleipner, Norway) could potentially be adapted for serpentinite-hosted aquifers. One possible means to further accelerate carbonation rates is to promote the growth of biomineralizing biofilms in the subsurface. The microorganisms in these biofilms could enhance carbonate precipitation and solubility trapping, while providing a physical barrier to minimize CO 2 leakage (Mitchell et al. 2010). Geothermal Mineral Carbonation Kelemen and Matter (2008) envision more vigorous in situ mineral carbonation processes in which the heat production and solid volume gain of carbonation reactions are exploited to accelerate uptake of CO 2 in peridotite and serpentinite (FIG. 1). In one method, a rock volume at depth is hydraulically fractured and heated to optimal reaction temperature. Injection of CO 2 -rich fluid drives carbonation at a rate of up to 1 tonne of CO 2 per m 3 per year. The fluid injection rate employed in this scenario must be controlled to balance exothermic heat generation. This is required in order to maintain the rock mass near optimal-reactionrate conditions (185 ºC, pco 2 > 70 bar, ph ~8 for olivine; Kelemen et al. 2011). Permeability and reactive mineral surfaces would be maintained through reaction-induced cracking, induced by volume increases commensurate with carbonation reactions, such as those observed in listwanite. In this scenario, reaction rates are likely to depend on the rate at which reactive mineral surfaces may be generated as well as the relative abundances of less reactive serpentine and more reactive olivine and pyroxenes in the peridotite. An alternative method relies on the thermal convection of seawater to deliver CO 2 to a high-temperature reaction site below the sea floor. In this second scenario, thermal convection would be promoted through a circulation system consisting of injection and production drill holes, connected at depth by a zone of hydraulically fractured peridotite. Carbonation reactions at depth would thus consume most of the CO 2 that is naturally dissolved in seawater and vent CO 2 -depleted seawater. Maximum carbonation rates of up to 1000 tonnes per year per well are based on optimistic predictions for fluid velocities in drill holes. Although hydraulic fracturing and CO 2 -injection technologies exist, significant new technology development may be required to implement these approaches at the temperatures and depths that have been proposed. SCALE AND CHALLENGES Although serpentinite deposits are sufficiently abundant to offset anthropogenic CO 2 emissions (29 billion tonnes CO 2 per year; Oelkers and Cole 2008), industrial serpentinite carbonation at a commensurate rate is unlikely in the foreseeable future unless considerable incentives are put in place. Kelemen et al. (2011) estimate that in situ geothermal carbonation at a rate of a billion tonnes CO 2 per year would require a million drill holes, which is approximately the total number of producing oil and gas wells in the United States. Proportional deployment of ex situ carbonation would require new mining activities at a scale comparable to the total existing global mining operations (fossil fuel, minerals and metals, aggregate). Alternatively, ex situ carbonation could be deployed alongside appropriate existing industrial processes to offset the greenhouse gas emissions of individual industrial operations (10 5 to 10 6 tonnes CO 2 per year) in the near term. Demonstration projects for serpentinite carbonation at this scale are likely feasible with existing technologies and could be used as a springboard for further research and development. ACKNOWLEDGMENTS We acknowledge funding by the Carbon Management Canada National Centre of Excellence and the Natural Sciences and Engineering Research Council of Canada, and we thank Anna Harrison, Peter Kelemen, and Gordon Southam for their insights. The comments and suggestions by the editors and three anonymous reviewers were helpful in improving this article. This is publication 310 of the Mineral Deposit Research Unit. REFERENCES Balucan RD, Dlugogorski BZ (2013) Thermal activation of antigorite for mineralization of CO 2. Environmental Science & Technology 47: Barnes I, O Neil JR (1969) The relationship between fluids in some fresh Alpinetype ultramafics and possible modern serpentinization, western United States. Geological Society of America Bulletin 80: Beinlich A, Plümper O, Hövelmann J, Austrheim H, Jamtveit B (2012) Massive serpentinite carbonation at Linnajavri, N Norway. Terra Nova 24: Bobicki ER, Liu Q, Xu Z, Zeng H (2012) Carbon capture and storage using alkaline industrial wastes. Progress in Energy and Combustion Science 38: Boschi C, Dini A, Dallai L, Ruggieri G, Gianelli G (2009) Enhanced CO 2 -mineral sequestration by cyclic hydraulic fracturing and Si-rich fluid infiltration into serpentinites at Malentrata (Tuscany, Italy). Chemical Geology 265: Brantley SL (2003) Reaction kinetics of primary rock-forming minerals under ambient conditions. Treatise on Geochemistry 5: Carey JW, Ziock H, Guthrie GD (2004) Reactivity of serpentine in CO 2 -bearing solutions: Application to CO 2 sequestration. 227 th American Chemical Society Meeting, Abstract U1095 Cipolli F, Gambardella B, Marini L, Ottonello G, Zuccolini MV (2004) Geochemistry of high-ph waters from serpentinites of the Gruppo di Voltri (Genova, Italy) and reaction path modeling of CO 2 sequestration in serpentinite aquifers. Applied Geochemistry 19: Dupraz C, Reid RP, Braissant O, Decho AW, Norman RS, Visscher PT (2009) Processes of carbonate precipitation in modern microbial mats. Earth-Science Reviews 96: Evans BW, Hattori K, Baronnet A (2013) Serpentinite: What, why, where? Elements 9: Ferris FG, Wiese RG, Fyfe WS (1994) Precipitation of carbonate minerals by microorganisms: Implications for ELEMENTS 120 APRIL 2013

7 silicate weathering and the global carbon dioxide budget. Geomicrobiology Journal 12: 1-13 Hansen LD, Dipple GM, Gordon TM, Kellett DA (2005) Carbonated serpentinite (listwanite) at Atlin, British Columbia: a geological analogue to carbon dioxide sequestration. Canadian Mineralogist 43: Harrison AL, Power IM, Dipple GM (2013) Accelerated carbonation of brucite in mine tailings for carbon sequestration. Environmental Science & Technology 47: Huijgen WJJ, Comans RNJ (2005) Carbon Dioxide Sequestration by Mineral Carbonation: Literature Review Update Energy Research Centre of the Netherlands, Report ECN-C , 37 pp IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage. Metz B, Davidson O, de Coninck HC, Loos M, Meyer LA (eds) Cambridge University Press, Cambridge, 422 pp Kelemen PB, Matter J (2008) In situ carbonation of peridotite for CO 2 storage. Proceedings of the National Academy of Sciences 105: Kelemen PB, Matter J, Streit EE, Rudge JF, Curry WB, Blusztajn J (2011) Rates and mechanisms of mineral carbonation in peridotite: Natural processes and recipes for enhanced, in situ CO 2 capture and storage. Annual Review of Earth and Planetary Sciences 39: Königsberger E, Königsberger L-C, Gamsjäger H (1999) Low-temperature thermodynamic model for the system Na 2 CO 3 MgCO 3 CaCO 3 H 2 O. Geochimica et Cosmochimica Acta 63: Lackner KS (2002) Carbonate chemistry for sequestering fossil carbon. Annual Review of Energy and the Environment 27: Lackner KS, Wendt CH, Butt DP, Joyce EL Jr, Sharp DH (1995) Carbon dioxide disposal in carbonate minerals. Energy 20: Larachi F, Daldoul I, Beaudoin G (2010) Fixation of CO 2 by chrysotile in low-pressure dry and moist carbonation: Ex-situ and in-situ characterizations. Geochimica et Cosmochimica Acta 74: Mitchell AC, Dideriksen K, Spangler LH, Cunningham AB, Gerlach R (2010) Microbially enhanced carbon capture and storage by mineral-trapping and solubility-trapping. Environmental Science & Technology 44: Naldrett AJ (1966) Talc-carbonate alteration of some serpentinized ultramafic rocks south of Timmins, Ontario. Journal of Petrology 7: Oelkers EH, Cole DR (2008) Carbon dioxide sequestration: A solution to a global problem. Elements 4: O Neil JR, Barnes I (1971) C 13 and O 18 compositions in some fresh-water carbonates associated with ultramafic rocks and serpentinites: western United States. Geochimica et Cosmochimica Acta 35: Power IM, Wilson SA, Thom JM, Dipple GM, Gabites JE, Southam G (2009) The hydromagnesite playas of Atlin, British Columbia, Canada: A biogeochemical model for CO 2 sequestration. Chemical Geology 260: Power IM, Dipple GM, Southam G (2010) Bioleaching of ultramafic tailings by Acidithiobacillus spp. for CO 2 sequestration. Environmental Science & Technology 44: Power IM, Wilson SA, Small DP, Dipple GM, Wan WK, Southam G (2011) Microbially mediated mineral carbonation: Roles of phototrophy and heterotrophy. Environmental Science & Technology 45: Pronost J and 10 coauthors (2012) CO 2 -depleted warm air venting from chrysotile milling waste (Thetford Mines, Canada): Evidence for in-situ carbon capture from the atmosphere. Geology 40: Renforth P, Washbourne C-L, Taylder J, Manning DAC (2011) Silicate production and availability for mineral carbonation. Environmental Science & Technology 45: Schwartzman DW, Volk T (1989) Biotic enhancement of weathering and the habitability of Earth. Nature 340: Seifritz W (1990) CO 2 disposal by means of silicates. Nature 345: 486 Wilson SA, Dipple GM, Power IM, Thom JM, Anderson RG, Raudsepp M, Gabites JE, Southam G (2009) Carbon dioxide fixation within mine wastes of ultramafic-hosted ore deposits: Examples from the Clinton Creek and Cassiar chrysotile deposits, Canada. Economic Geology 104: Zevenhoven R, Fagerlund J, Songok JK (2011) CO 2 mineral sequestration: developments toward large-scale application. Greenhouse Gases: Science and Technology 1: Zhang P, Anderson HLJ, Kelly JW, Krumhansl JL, Papenguth HW (2000) Kinetics and Mechanisms of Formation of Magnesite from Hydromagnesite in Brine. Sandia National Laboratories, Albuquerque, Technical Report SAN J, 26 pp The Economic Geology Anniversary Collection. Available on DVD. NEW RELEASE! This long-awaited DVD brings together the contents of the Economic Geology 50th, 75th, and 100th Anniversary Volumes in a searchable format. The was followed by the two successor volumes in 1980 and All are still in demand and many of Order yours today! segweb.org/store Economic Geology Anniversary Collection Member: $98 Non-member: $140 ELEMENTS 121 APRIL 2013

8