Carbon Dioxide Sequestering Using
FRASER GOFF * and K.S. LACKNER ‡
EES-1, MS D462; Los Alamos National Laboratory; Los Alamos, NM 87545
T-3, MS B216; Los Alamos National Laboratory; Los Alamos, NM 87545 (Environmental Geosciences, Volume 5, Number 3, 1998 89-101)
ABSTRACT magnitude greater than the annual production of CO2 by vol-
Fossil fuels continue to provide major sources of energy canoes and metamorphic processes and an order of magni-
to the modern world even though global emissions of CO2 tude greater than the consumption rate of CO2 by natural
are presently at levels of ≥19 gigatons/year. Future antipol- geologic processes (Kerrick et al., 1995). During the last two
lution measures may include sequestering of waste CO2 as centuries, CO2 levels in the atmosphere have increased ~30%
magnesite (MgCO3) by processing ultramafic rocks. Com- (Ramanathan, 1988) raising legitimate concerns about global
mon ultramafic rocks react easily with HCl to form MgCl2 warming and the terrestrial carbon cycle (Sabine et al., 1997;
which is hydrolyzed to form Mg(OH)2. CO2 would be trans- Weart, 1997).
ported by pipeline from a fossil fuel power plant to a seques- Various schemes have been devised to reduce CO2 emis-
tering site and then reacted with Mg(OH)2 to produce ther- sions while allowing continued consumption of fossil fuels
modynamically stable magnesite. Huge ultramafic deposits (Blok et al., 1992). In one of these schemes, Lackner et al.
consisting of relatively pure Mg-rich silicates exist (1995) described two chemical processes to sequester CO2 b y
throughout much of the world in ophiolites and, to a lesser formation of carbonate minerals. Sequestering CO2 involves
extent, in layered intrusions. Peridotites and associated ser- reaction with divalent cations (principally Mg and/or Ca)
pentinite are found in discontinuous ophiolite belts along derived from natural mineral deposits by either direct car-
both continental margins of North America. Serpentinites bonation at high temperature or by reaction in aqueous solu-
and dunites comprise the best ores because they contain the tion. Both sequestering processes are thermodynamically
most Mg by weight (35 to 49 wt-% MgO) and are relatively favorable, but details of actual implementation require much
reactive to hot acids such as HCl. Small ultramafic bodies further investigation (Lackner et al., 1995; 1997; Butt et al.,
(~1 km3) can potentially sequester ~1 gigaton of CO2 or 1996).
~20% of annual USA emissions. A single large deposit of Mg and Ca comprise ~2.0 and 2.1 mol-% of the earth’s
dunite (~30 km3) could dispose of nearly 20 years of current crust, respectively, primarily bound in silicate minerals
USA CO2 emissions. The sequestering process could provide (Brownlow, 1979). Although molar abundances are similar,
Mg, Si, Fe, Cr, Ni, and Mn as byproducts for other industrial Mg silicates contain more reactive material per ton of rock
and strategic uses. Because "white" asbestos (chyrsotile) is a due to the lower molecular weight of Mg. Two types of Mg
serpentine mineral, CO2 sequestering could dispose of some silicate minerals occur in relatively pure deposits and have
waste asbestos. The cost and environmental impact of ex- thermodynamic and chemical properties desirable for CO2
ploiting ultramafic deposits must be weighed against the waste processing. These minerals, forsterite and serpentine,
increased costs of energy and benefits to the atmosphere and occur as peridotites and serpentinites in ultramafic rocks.
climate. The object of this paper is to briefly describe CO2 sequester-
ing by magnesite formation, discuss the distribution and
Key Words: carbon dioxide, environmental geology, geo- geochemistry of ultramafic deposits, estimate the CO2 se-
chemistry, global warming, mining, ultramafic rocks, waste questering potential of some Unites States resources, and
disposal. consider some industrial and environmental impacts that
would result from large-scale ultramafic mining.
Numerous resource evaluations show that worldwide re- CO2 SEQUESTERING IN SOLID FORM
serves of fossil fuels can provide mankind’s energy needs for In the presence of CO2, serpentine and forsterite are both
many centuries (e.g., United Nations, 1995). The major ob- thermodynamically unstable. Given time, they are trans-
stacle to consuming these resources is growing emissions of formed into magnesite plus additional materials that contain
CO2 to the atmosphere. The United States produces ~5 giga- the silica and water that made up the stoichiometric balance
tons of CO2 annually, which is ~25% of the present global of the original minerals. These carbonation reactions are
output of ≥19 gigatons. These quantities are an order of exothermic and lower the free energy of the system. The en-
ergy release is 64 and 95 kJ per mol of CO2 for serpentine and disassociates into Mg(OH) 2 and MgCl2 in aqueous solution.
forsterite, respectively (Robie et al., 1979). This should be In this cycle, it is possible to recover virtually all of the HCl.
compared to the heat of combustion of 394 kJ that is released The most serious loss mechanism is the formation of alkali
in the formation of one mol of CO2. Because they are ther- chlorides. However, the alkali content of peridotite and ser-
modynamically favored, magnesite and silica are common i n pentinite rocks is very small, usually ≤0.5 wt-% (see below).
serpentinized ultramafic rocks (Barnes et al., 1973; O’Hanley, A major design concern is to accomplish these processing
1996). The formation of these silica-carbonate rocks is pro- steps by using only energy that is available from heat
moted by natural CO2-rich fluids permeating the mineral sources within the processing scheme.
deposits. The resulting magnesite is stable and is not likely Fortunately, the carbonation reaction is quite exothermic
to release the bound CO2 again. and, in principle, can provide all the energy necessary t o
One approach to CO2 disposal is to accelerate this natural recover the HCl. To harness this energy, the carbonation i s
process and form magnesite on a rapid scale. The resulting performed in a gas-solid reaction between Mg(OH) 2 and CO2.
mineral is environmentally safe and provides a permanent In the aqueous alternative, the reaction rates may be faster,
storage for the large volumes of CO2 resulting from power but because of the high degree of dilution, the heat of reac-
generation. The cost of such a process can be held low be- tion would be lost for practical purposes.
cause the reaction is exothermic and, if properly designed, The CO2 would be pipelined to the disposal site from a
does not require the additional input of energy. power plant. Separating and shipping of CO2 has similarities
The carbonation of serpentinite is broken up into several with all other CO2 disposal methods. It is not further dis-
steps (Figure 1). First, the mineral ore is mined and ground cussed here because its implementation will differ greatly
to a powder. To improve the carbonation reaction kinetics, between various plant designs. The emphasis of this present
the magnesium is extracted from the ore and put in the form work has been on the downstream disposal process. In par-
of magnesium hydroxide. The extraction is accomplished ticular, we have demonstrated experimentally that the gas-
with HCl, which dissolves the mineral forming MgCl2, and solid carbonation reaction goes virtually to completion i n
silica, which is readily precipitated along with iron oxide <30 min at pressures of 50 bar and temperatures of ~500 t o
that was present in the original minerals. This process was 600°C. Such a process would already be economically feasi-
described in the 1940s and 1950s when Mg shortages were ble if one takes advantage of the fact that the pipelined CO2
driving research into alternative extraction technologies is already pressurized. However, based on experience in the
(Houston, 1945; Barnes et al., 1950). The resulting MgCl2, analogous sulfation chemistry of calcium oxide, it is ex-
which is always hydrated, can be hydrolyzed at 200 to 250°C pected that a more careful investigation of the reaction
to form Mg(OH)Cl and HCl (Kelley, 1945). The Mg(OH)Cl mechanisms will lead to a significantly improved implemen-
Coal Mine Power Plant CO2
earth moving Coal 33% CO2 collecting Pipeline
190 kt/day 9 kt/day conversion 24 kt/day compressing 24 kt/day
efficiency FIGURE 1: Processing stream for CO2
<1000 km disposal (Lackner et al., 1997); the upper
branch shows the carbon flow from en-
MgCO3 46 kt/day Carbonation ergy source (coal mine) through the
power plant and the sequestration unit.
Solid-Gas The large amount of earth moved reflects
497 MW 513 MW Reactor the overburden of a typical surface mine.
thermal thermal 550° C It is assumed that the CO2 gas is delivered
to the disposal site through a pipeline. The
bottom half of the diagram represents the
Serpentine Rock MgCl2 HCl recovery Mg(OH)2 disposal process (see text). Kt. Kilotons.
Mine 48-62 kt/day 52 kt/day 100-260° C 32 kt/day
35-45% Silica HCl endothermic H2O
MgO 21-35 kt/day 40 kt/day 608 MW 3 kt/day
Fe2O3 MgCl2 7 kt/day
5 kt/day 0.5 kt/day
tation of the reaction (Lackner et al., 1997).
The cost of mining and grinding the ore can be roughly es- A
timated from the analogous processing steps in copper min-
ing. It would amount to ~$8/ton of CO2. Very simple esti- EUR AFR
mates suggest that an additional $8/ton of CO2 would be a
reasonable goal to aim for in the cost of downstream process-
ing. To achieve such favorable costs would require that the
overall process does not demand significant amounts of ex-
ternal energy either in the form of heat or recompression of
gases. The $8/ton of CO2 would pay for a $300 million plant,
plus operating and personnel costs, plus makeup of lost HCl.
To set the scale, for a power plant with 44% conversion effi-
ciency, $16/ton of CO2 amounts to $0.012 per kW hr (Lack- SA
ner et al., 1995).
RESOURCES Ophiolite Belts
The magnesium-rich, ultramafic rocks (primarily peri-
dotites and serpentinites) that are candidate ores in the se-
questering process are distributed throughout the world. B Skaergaard
There are at least nine types of ultramafic rock associations,
but they occur in magmatic-tectonic settings too varied t o
document here (Coleman, 1977). The most voluminous and
widespread ultramafic rocks are the alpine (“metamorphic”)
peridotites that form the basal sequence of ophiolites, slabs
of oceanic crust uplifted and eroded along subduction zones
and plate boundaries. The basal peridotites represent de-
tached slices of the Earth’s upper mantle exposed by these
tectonic processes (Dickinson et al., 1996). Because they
occur mostly along the upper plate of present and past sub- Chuacus
duction zones, ophiolites are found as belts throughout most
of the world, having discontinuously exposed dimensions of
as much as 1000 × 100 km. Within North America, ophiolite FIGURE 2: (A) Polar projection of the world showing generalized
belts are found along the Appalachian mountain chain locations of ophiolite belts (from Coleman, 1977). (B) Locations of
stretching from the southeast USA into Quebec and New- ophiolite belts and major layered intrusions of North America. Example
ophiolite bodies mentioned in text are B, Baltimore Complex; BM,
foundland and along the Cordilleran mountain chain stretch- Belvidere Mountain; CM, Canyon Mountain; G, Green Mountain; TS,
ing from Alaska through British Columbia to California Twin Sisters (from Coleman, 1977 and references in text).
(Figure 2). Smaller belts are found in Guatemala and in the
When examined more closely, the basal ultramafic rocks i n the first formed minerals of crystallization (primarily Mg-
ophiolite belts are found to be elongate ribbons and frag- rich silicates) settled by gravity toward the bottom of the
ments that parallel regional geologic structures (Figure 3). intrusions. As a result, layers of peridotite as thick as a few
The tectonic processes that create ophiolites and expose hundred meters can be found in exposures as long as several
elongate fragments of the upper mantle are complex and usu- tens of kilometers. The largest such body is the famous
ally take several million years to complete (Coleman, 1977; Bushveld Complex in South Africa but other well-known
Harper, 1984). Individual exposures of ultramafic rock may bodies occur at the Stillwater, Sudbury, and Skaergaard loca-
occupy hundreds of square kilometers or may be as small as tions in North America (Figure 2). Of these three, the late
hand samples incorporated into fault zones. Archean (2.7 Ga (billion years ago)) Stillwater Complex i n
The second most voluminous class of ultramafic rocks oc- Montana has the largest exposures of gravity-settled peri-
curs in large, layered intrusions at many localities worldwide dotite in North America (~48 km long; Czamanske and Zien-
(Hess, 1960; Cawthorn, 1997). These magma bodies gener- tek, 1985).
ally had initial compositions similar to mantle basalt and
were intruded into shallow levels of the Earth’s crust (usu- CALIFORNIA STUDY REGION
ally ≤8 km). The larger bodies tend to be Precambrian age This present investigations focuses on California ultrama-
(≥650 Ma (million years ago)). Because of their great initial fic bodies (Figure 3), because their basic geology is well-
volumes, these magmas cooled slowly within the crust and known, their distribution and volume is significant, their
Klammoth- BM Burro Mountain
J Region DP Del Puerto
T J Josephine
KR Kings River
NI New Idria
PS Point Sal
Co S SM San Mateo
ast Lake FIGURE 3: Map of California (modi-
Tahoe T Trinity fied from Jennings, 1977) showing loca-
V Vulcan Peak tions of ultramafic provinces and some
W Wilbur Springs example sites described in text. The
Vulcan Peak peridotite in extreme
San southern Oregon is actually part of the
Francisco greater Josephine ophiolite, the largest
DA in North America. The ophiolites have
SM DP been deformed and exposed during
prolonged subduction of the Pacific
KR Plate (west) beneath the North Ameri-
Monterey can Plate (east) during late Paleozoic to
B early Tertiary time. Since ~30 Ma, the
rocks have experienced considerable
right-lateral shear along the present San
r Andreas transform zone.
0 50 100 150 ARIZON
proximity to population and power manufacturing centers i s 1964). The body is part of an elongate slab of ophiolite
favorable, and their previous exploitation is well-established whose ultramafic part (up to 300 m thick and ~40 km2) i s
(Coleman, 1996). variably serpentinized (Evarts and Schiffman, 1982).
California ultramafic bodies occur in four locales: the The contact between ultramafic rock and underlying ma-
Coast Ranges, Big Sur, Sierra Nevada foothills belts, and the rine deposits of the Franciscan Complex (Jurassic to Eocene)
Klammoth-Trinity region (Saleeby, 1982; Harper, 1984; is relatively flat and sharp (Figure 4A). The extreme base of
Dickinson et al., 1996). It is well beyond the scope of this the ultramafic body consists of strongly foliated antigorite
report to review the age and tectonic history of each province schist grading upward into sheared, serpentinized harzbur-
in detail, but each formed during subduction-related events gite and local zones of dunite. The western crest of the de-
that occurred from ~300 to 50 Ma (Paleozoic to Early Terti- posit consists of material that is only 5 to 40% serpentinized
ary). Two ultramafic bodies in the Coast Ranges belt were and contains ≥45 mol-% MgO (Table 1). The east portion of
chosen for preliminary study: the Del Puerto body because i t the deposit is more highly serpentinized (Maddock, 1964;
contains a large mass of relatively unserpentinized ultrama- Himmelberg and Coleman, 1968).
fic rock, and the Wilbur Springs body because it is mostly Within the deposit, cross-cutting faults are pervasive. A
serpentinized. set of high-angle, northwest-trending faults and fractures
host magnesite veins, pods, and eroded spring deposits
Del Puerto Ultramafic Body (Bodenlos, 1950). Considerable geochemical research has
The Del Puerto ultramafic body lies 60 km due east of San shown that the magnesite forms from groundwater alteration
Jose in the California Coast Ranges (Himmelberg and Cole- of the ultramafic rock and transport of Mg to favorable sites
man, 1968; Evarts, 1977). The ultramafic and surrounding for precipitation (Barnes et al., 1967; 1973). Obviously, the
rocks are well explored as they have been mined for Mg host rocks are completely compatible with the magnesite
(magnesite), Mn (pyrolusite, MnO2), cinnabar (HgS), and waste that would be generated by CO2 sequestering.
chromite (Hawkes et al., 1942; Bodenlos, 1950; Maddock,
FIGURE 4: (A) Photograph looking north of the
west side of the Del Puerto ultramafic body (UM)
overlying marine rocks of the Franciscan complex
(F). Note sharp vegetation contrast between units.
This boundary is a thrust fault developed during
late Cretaceous-early Teriary subduction. Ul-
tramafic rocks change from highly foliated
antigorite schist at the base to relatively massive
harzburgite and dunite near the top. Fracturing is
pervasive and serpentinization varies from 5 to 60
% within this zone. The ultramafic mass thickens
to the east (right). (B) Photograph looking north
of the Wilbur Springs serpentinite along Kilpepper
Creek, 1 km west of Complexion Spring. The
serpentinite, which is over 100 m thick at this
location, is sheared and fractured into blocky
rubble. Only ~5 % of the original peridotite min-
erals remain unaltered.
Although the magnesite has been largely mined out, the and “ranchettes” of single-family homes have been built i n
haul roads, shafts, pits, and dumps are still visible. Since San Antonio Valley on the west margin of the ophiolite. The
World War II, the region has been mostly used for cattle impact of renewed mining for carbonate waste disposal
ranching and hunting clubs. The ultramafic rocks are covered would be carefully examined by the local public. Historic
primarily with brush of manzanita and live oak with scattered (and dilapidated) mining infrastructure is visible all over the
pines. The area may be reached by paved roads from the west, Del Puerto body.
east, and north. A small county park is located on the north-
eastern margin of the area. More recently, small subdivisions
TABLE 1. Chemical compositions (major elements in wt-%) of ultramafic rocks from the Del Puerto and Wilbur Springs ultramafic
Del Puerto Ultramafic Body
DPS-AVEd UM96-26a Wilbur Springs Serpentinite
a b c
Sample No. UM96-19 UMDP-AVE 66R22 c
66R20 Serpentinized Serpentinized UM96-3a UM96-13a UMWS-AVEe
Type Dunite Peridotite Dunite Harzburgite Peridotite Lherzolite Serpentinite Serpentinite Serpentinite
SiO2 38.22 40.50 ± 1.89 39.0 44.9 38.8 40.8 39.75 41.73 40.6 ± 1.2
TiO2 0.002 0.000± 0.001 0.02 0.02 0.013 0.011 0.080 0.065 0.040± 0.02
Al2O3 0.71 0.40 ± 0.33 0.04 0.91 0.47 0.032 2.28 2.59 1.99 ± 0.40
Fe2O3 4.54 4.10 ± 1.32 2.8 0.80 3.23 5.49 6.80 4.58 5.06 ± 1.3
FeO 4.68 4.25 ± 1.15 5.0 7.0 4.74 1.88 1.38 3.53 2.94 ± 1.0
MnO 0.159 0.134± 0.11 0.11 0.12 0.13 0.14 0.147 0.128 0.138± 0.15
MgO 44.24 42.92 ± 1.18 46.1 43.0 42.5 34.7 34.61 36.85 36.2 ± 1.1
CaO 0.00 0.00 0.00 1.50 0.53 5.76 0.17 1.87 0.42 ± 0.1
Na2O 0.00 0.00 0.00 0.02 0.013 0.00 0.00 0.00 0.00
K2O 0.00 0.00 0.23 0.08 0.15 0.00 0.00 0.00 0.00
P2O5 <0.005 <0.005 0.03 0.03 0.051 <0.005 <0.005 <0.005 <0.005
NiO 0.396 0.34 ± 0.03 0.35 0.32 0.27 0.116 0.34 0.30 0.32
Cr2O3 1.18 0.53 ± .25 0.44 0.47 0.60 0.413 0.45 0.36 0.40
CO2 ––– f ––– 0.21 <0.05 0.18 ––– ––– ––– –––
H2O(+) 5.02 6.23 ± 2.42 5.6 1.0 8.43 11.3 13.76 8.50 12.6 ± 1.1
H2O(-) ––– ––– 0.50 0.09 0.45 ––– ––– ––– –––
TOTAL 99.15 99.40 100.4 100.3 100.6 100.9 99.77 100.5 100.7
MgO/SiO2 1.16 1.06 1.18 0.96 1.09 0.850 0.871 0.883 0.892
Mol-% MgO 52.5 48.2 51.1 52.7 45.6 36.7 35.9 37.8 37.6
ρ(g/cm3) 3.25 2.95 2.83 3.22 2.85 2.69 2.55 2.85 2.65
% Serpentinite 10.0 4 5 ± 30 58.5 10.0 56.1 90 100 60 95 ±5
Analysis from Goff et al. (1997); LOI (loss on ignition) listed as H2O(+).
Average of nine dunite and harzburgite samples with std. dev. (1s) from Goff et al. (1997); LOI listed as H2O(+).
Analysis from Himmelberg and Coleman (1968).
Average of eight peridotite samples from Himmelberg and Coleman (1968).
Average of 15 samples with std. dev. (1s) from Goff et al. (1997); LOI listed as H2O(+).
(–––). not analyzed.
Wilbur Springs Serpentinite tinite body and is the surface expression of a small geo-
The Wilbur Springs serpentinite mass is located ≥200 thermal reservoir (Goff and Janik, 1993). Drilling for
km NNE of San Francisco on the eastern side of the geothermal resources in the 1960’s failed to find suffi-
Coast Ranges (McLaughlin et al., 1989). The serpen- ciently high temperatures for electrical generation (reser-
tinite is the preserved base of an extensive, north-trending voir temperature is ≤140°C at >2000 m). The presence of
sheet of ophiolite that is ~50 km long and averages 2 to 6 low-temperature mineral springs, which occur sporadically
km wide. The thickness of the deposit varies from a few throughout most of the serpentinite, indicates that modern
tens of meters on the west to several hundred meters on day serpentinization is happening by reactions with
the east (Figure 4B). groundwaters. Complexion Spring (≤20°C), near the
In contrast to the Del Puerto deposit, our examinations heart of our study area, precipitates brucite (Mg(OH)2) and
show that most outcrops have very little preserved peri- has a pH ≤12 (Barnes et al, 1972).
dotite minerals (≤5% overall), although original textures The southern margin of the serpentinite also hosts sev-
are occasionally well-preserved. The rocks are serpenti- eral small cinnabar mines, last worked in the early
nized harzburgites that are pervasively faulted and sheared 1950’s, and one small gold mining district that has not
throughout much of the deposit. Typical samples from been worked since World War I (Peters, 1991; Goff and
15 widespread locations in the mass contain ~36 mol-% Janik, 1993). Homestake Mining Company re-explored
MgO (Table 1). The serpentinites overlie deformed Fran- the deposit in the 1980’s but decided to drop their lease
ciscan Complex rocks on the west and south and are over- due to low tonnage of gold-bearing rock. Wilbur Springs
lain by Great Valley sequence marine sediments (Jurassic- proper is a small but thriving hot spring resort first de-
Cretaceous) on the east. veloped before the turn of the century (Goff and Janik,
A WNW-trending group of hot springs in an 8-km-long 1993). The resort now caters to people who seek quiet,
zone occurs at the extreme southern edge of the serpen- natural surroundings.
Most of the land occupied by the serpentinite belongs wt-% and ≤5 wt-%, respectively. Total Ti, Na, K, and P con-
to the U.S. Bureau of Land Management or to a few cattle tents are usually ≤0.3 wt-%. The Fe is relatively reduced (Fe3+
ranches. Indian Valley on the west side of the mass con- /Fe2+ ≈ 0.3). Peridotites in layered intrusions generally con-
tains a reservoir that is used for recreational purposes, tain less magnesia and more of the other oxides due to higher
when there is water. The serpentinite hosts scrubby vege- proportions of pyroxenes and minor presence of plagioclase.
tation consisting of manzanita, buckthorn, live oak, scat- Contents of Cr, Ni, and Mn are roughly equivalent to alpine
tered pines, and rare cypress trees which can be nearly peridotites.
impenetrable to humans on foot. The area may be reached Most peridotites are partly to completely reconstituted
by dirt road from the south, east, and northwest. into hydrated Mg-rich silicates (serpentine and related min-
erals). The three serpentine minerals (lizardite, chrysotile,
GEOCHEMISTRY OF antigorite) are isochemical with very similar sheet-like struc-
ULTRAMAFIC ROCKS tures. The resulting serpentinites may contain some relict
Mineral contents and chemical analyses of “fresh” peri- olivine and pyroxene, but more often they contain only ser-
dotite bodies (≥50 modal-% unaltered rock) are listed in nu- pentine minerals (Mg3(Si2O5) (OH)4), magnetite (Fe-rich
merous reports (e.g., Goff et al., 1997). Typical peridotites spinel), and residual chromite, plus brucite (Mg(OH)2), car-
are harzburgites containing ≤90 modal-% of forsteritic oli- bonates (usually magnesite, MgCO3), and free silica (SiO2).
vine and ≥10 modal-% of orthopyroxene with accessory Serpentinites may contain as much as 14 wt-% water. Tex-
chrome spinel ± chrome diopside. Most peridotites contain tural evidence in rocks shows that the hydration of forsterite
39 to 44 wt-% SiO2, 42 to 50 wt-% MgO, 7 to 9 wt-% FeO (as to form serpentine (and brucite) is accompanied by a volume
total Fe), ≤4000 ppm Cr, ≤3000 ppm Ni, and ≤1200 ppm Mn increase of as much as 53%. Thus, the serpentinites are low
(Tables 1 and 2). CaO and Al2O3 contents are generally ≤1.5 density rocks (~2.5 g/cm3) relative to the original peridotites
TABLE 2. Chemical compositions (major elements in wt-%) of ultramafic rocks from selected locations in the United States.
Cyn Mtn Balt. Cplx San Mateo
OR MD CA
a b c d e f g
Sample No. STD UNIM VP-AVE 19VP68 1VP68 BM-AV BMS-AVE CM-AVE BC-AVE h
Type Dunite Dunite Peridotite Dunite Harzburgite Dunite Serpentinite Harzburgite Serpentinite Serpentinite Peridotite
SiO2 40.41 42.52 41.2 39.6 43.4 39.7 33.0 42.1 41.2 41.08 47.7
TiO2 0.005 –––j 0.035 0.03 0.02 0.0 0.0 0.04 0.05 0.054 0.12
Al2O3 0.19 0.19 0.38 0.07 0.25 0.4 0.6 1.70 1.33 1.76 4.82
Fe2O3 1.03 7.68 1.52 1.1 0.52 ––– ––– 2.49 6.17 8.92 2.94
FeO 6.97 ––– 6.93 9.8 7.8 8.9 14.1 4.79 2.43 ––– 6.54
MnO 0.12 ––– 0.12 0.14 0.14 0.23 0.2 0.13 0.12 0.133 0.17
MgO 49.59 48.01 45.5 47.4 45.3 48.3 38.0 35.6 35.09 33.06 29.0
CaO 0.17 0.02 0.70 0.30 0.91 ––– ––– 5.58 1.67 1.76 2.44
Na2O 0.015 0.02 0.006 0.02 0.00 ––– ––– 0.55 0.01 0.00 0.19
K2O 0.001 0.01 0.068 0.06 0.07 ––– ––– 0.03 0.06 0.00 0.02
P2O5 0.002 ––– 0.032 0.04 0.04 ––– ––– 0.02 0.008 0.011 0.01
NiO 0.30 0.37 0.26 0.27 0.25 ––– ––– ––– 0.18 0.33 –––
Cr2O3 0.58 0.13? 0.25 0.30 0.18 1.6 0.7 ––– 0.16 0.37 0.48
CO2 0.08 ––– <0.06 0.05 <0.05 ––– ––– 0.12 0.03 ––– 0.11
H2O(+) 0.44 1.05 2.97 1.1 1.3 1.23 12.5 6.94 11.06 12.46 4.91
H2O(-) 0.06 ––– 0.18 0.10 0.05 ––– ––– ––– 0.31 ––– 0.49
TOTAL 99.96 100.0 100.2 100.4 100.2 100.4 99.1 100.1 99.88 99.99 99.95
MgO/SiO2 1.23 1.13 1.10 1.20 1.04 1.22 1.15 0.845 0.852 0.804 0.608
Mol-% MgO 59.9 57.1 53.2 57.0 54.6 57.9 39.3 40.4 37.7 35.5 35.5
ρ(g/cm3) 3.32 3.3 3.18 3.28 3.27 3.27 2.63 2.95 2.71 2.63 3.07
% Serpentinite4 4 21 9 10 10 100 50 80 100 35
Rock standard STD (Govindaraju, 1994).
Analysis provided by Unimin Corp. (UNIM); Fe as Fe2O3 and LOI (loss on ignition) listed as H2O(+).
Average of 13; all Vulcan Peak analyses from Himmelberg and Looney (1973).
Average of three (Labotka and Albee, 1979); Fe as FeO and LOI listed as H2O(+).
Average of three (Labotka and Albee, 1979); Fe as FeO and LOI listed as H2O(+).
Average of 11 (Thayer, 1977); LOI listed as H2O(+).
Average of four (Morgan, 1977).
Average of eight homogenized pieces (total weight = 2 kg) from single outcrop (Goff et al., 1997); Fe as Fe2O3 and LOI listed as H2O(+).
Average of 26 (Hess, 1960); no analyses for Ni reported.
(–––). not analyzed.
(~3.3 g/cm3). tion causes the average oxidation state of iron to rise (Fe3+
Chemical analyses of serpentinites are widely available i n /Fe2+ ≈ 2). Contents of other transition metals are slightly
the literature (Tables 1 and 2). Because of their lower densi- less than the original peridotites.
ties and high water contents, serpentinites contain substan-
tially less magnesia than peridotites, usually between 32 and Acid Dissolution Experiments
36 wt-% MgO. Formation of magnetite during serpentiniza- Acid dissolution experiments were conducted on various
TABLE 3. Results of dissolution and analysis using 1 g of ultramafic rock sample mixed in hot 1:1 HCl or in
HCl-HNO3-HF. The results are compared with X-ray fluorescence (XRF) analyses. The residues are mixtures of
silica gel, spinels, and pyroxenes. No single reagent can effectively dissolve all components from ultramafic
rocks, but HCl is more effective at dissolving Mg than is the mixed reagent and works better on serpentinites
STD dunite (std value, Table 1) –––a 49.59 929 2360 3990
STD (powder, 3 acids) ––– 8.04 916 2454 157
STD (powder, hot HCl) 59.6 53.05? 951 2530 20
Three Sisters dunite (Unimin, ≤4% serp) ––– 48.01 ––– 2907 886
Three Sisters dunite (XRF) ––– 47.93 930 2830 3820
Three Sisters dunite (ore, hot HCl) 56.5 42.47 747 2524 21
Three Sisters dunite (crushed ore, hot HCl) 63.1 40.38 761 2430 27
Three Sisters dunite (powder, hot HCl) 53.0 46.89 830 2728 127
Green Mtn peridotite (Unimin, 10% serp) ––– 47.65 ––– 3500 1710?
Green Mtn peridotite (XRF) ––– 46.68 1005 140 3832
Green Mtn peridotite (powder, hot HCl) 48.1 41.28 738 2479 657
PCC-1 peridotite (std value) ––– 43.43 930 2380 2730
PCC-1 (powder, 3 acids) ––– 8.04 899 2484 804
JP-1 peridotite (std value) ––– 44.72 930 2460 2970
JP-1 (powder, 3 acids) ––– 7.73 914 2445 709
Del Puerto dunite (XRF, 40% serp) ––– 44.31 951 2754 3160
Del Puerto dunite (powder, hot HCl) 43.4 41.82 828 2398 25
Del Puerto dunite (powder, 3 acids) ––– 11.46 822 2653 33
Del Puerto hzbgite (XRF, 50% serp) ––– 41.63 1050 2560 2794
Del Puerto hzbgite (powder, hot HCl) 46.7 34.64 784 2141 185
Del Puerto hzbgite (powder, 3 acids) ––– 12.79 908 2285 435
UB-N serpentine (std value) ––– 35.21 929 2000 2300
UB-N (powder, 3 acids, 5/96) ––– ––– 981 2064 2145
UB-N (powder, 3 acids, 2/97) ––– 13.71 985 2102 2200
San Mateo serpentinite (XRF, Table 1) ––– 33.06 1045 2514 2688
San Mateo serpentinite (powder, 3 acids) ––– ––– 1014 2312 1998
San Mateo serpentinite (powder, 60°C HCl) 48.8 30.92 765 2385 1440
San Mateo serpentinite (powder, ≤128°C HCl) 46.6 32.75 865 2120 1595
San Mateo serpentinite (powder, ≤208°C HCl) 47.5 32.66 840 2270 2135
San Mateo serpentinite (60°C residue, 3 acids) ––– ––– 103 57 1481
San Mateo serpentinite (60°C residue + leachate) ––– ––– 868 2442 2921
Wilbur Spgs serp (XRF, 85% serp) ––– 35.41 1155 2253 3054
Wilbur Spgs serp (powder, hot HCl) 42.2 34.04 985 1960 1992
Wilbur Spgs serp (powder, 3 acids) ––– 15.19 1039 2137 2849
(–––). not analyzed.
peridotite and serpentinite samples to determine the relative est such body in the USA is the Twin Sisters Dunite (Ragan,
merits of HCl as opposed to a more complex rock reagent 1963), which occurs in the Cascade Range of northwestern
such as HCl-HNO3-HF (Table 3). The results of these experi- Washington (Table 2). This dunite body covers ~90 km2 and
ments show that: is presently mined by open-pit methods for refractory (foun-
• Hot HCl is better at dissolving Mg from ultramafic rocks dry) sand. An examination of a 20-kg sample provided b y
than the three-reagent mixture (≥35 wt-% versus ≤15 wt-% the operator shows that it contains ≤4% serpentine and other
Mg). Residual products (~45 to 60 wt-%) from HCl dissolu- secondary reaction products. The relatively large, unserpen-
tion include silica gel, spinels, and pyroxenes and additional tinized Green Mountain Peridotite (≥5 km2) occurs in the
silicates such as talc, amphiboles, chlorite and sericite. The Appalachians of North Carolina and is also mined primarily
three-acid mixture apparently precipitates MgF compounds for foundry sand. An examination of this material shows that
while dissolving the rock. it contains ≥5% orthopyroxene and ~10% of high-grade
• Hot HCl is slightly better at dissolving Mg from serpentinite metamorphic products including talc and Mg-rich amphibole
than from peridotite, including dunite, because serpentinites (minerals not soluble in HCl). This material is less desirable
contain less non-reactive silicates such as pyroxene. Much of than dunite as a CO2 sequestering ore even though it con-
the Fe in serpentinites occurs as microcrystalline magnetite tains nearly comparable amounts of Mg (Goff et al., 1997).
that is relatively easy to dissolve in HCl. Partially serpentinized peridotite and dunite in large
• Hot HCl is less effective at dissolving many trace metals masses (20 to 80% serpentine) are more common. Examples
from ultramafic rocks than the three-acid mixture. This i s include the Belvidere Mountain prospect in Vermont
especially true for high field-strength elements like Cr and (Labotka and Albee, 1979), the Canyon Mountain and Vulcan
slightly true for Mn. On the other hand Co (not listed i n Peak deposits in Oregon (Himmelberg and Loney, 1973;
Table 2) and Ni mainly reside in olivine; thus, HCl dissolu- Thayer, 1977), the Del Puerto and Burro Mountain bodies of
tion works well for these elements. California (Page, 1967; Himmelberg and Coleman, 1968;
• HCl dissolution at 60°C and atmospheric pressure is nearly Goff et al., 1997), and most peridotite in the Stillwater Com-
as effective as HCl at 200°C and 15 bars for dissolving Mg plex (Tables 1 and 2). The Belvidere, Vulcan Peak, Del
from serpentinite (see results for San Mateo serpentinite, Puerto, and Stillwater bodies contain small zones (≤4 km2) of
Table 2). However, the gain in trace metal dissolution is too relatively unaltered dunite.
small to make this procedure worthwhile. Completely serpentinized peridotite is exceptionally
common in certain areas of eastern and western North Amer-
Best Ultramafic Ores for Carbonate Disposal ica. Ultramafic rocks in the Baltimore Complex contain
Because olivine and serpentine are the most reactive Mg- ≥80% serpentinite and nearly all California deposits consist
rich minerals in the HCl dissolution process, rocks com- of ≥95% sepentinite (Rice, 1957; Morgan, 1977; Goff et al.,
prised solely of these minerals would make the best ores. In 1997). Perhaps the largest body of continuous serpentine
contrast, spinels and pyroxenes are not as reactive; thus, outcrop in the USA occurs at the Josephine Ophiolite of
these residual minerals (and silica) must be separated from northwest California which extends over ≥800 km2 (Harper,
primary reactants during processing. Although Cr-spinel 1984). These deposits, although some are huge, are mostly
(and other potentially valuable metals) may contribute serpentinized harzburgite and are not as desirable as the un-
greatly to the economics of the process, the pyroxenes con- altered dunites.
tribute virtually nothing of value. Thus, peridotites and
serpentinites with little pyroxene comprise the most desir- SEQUESTERING POTENTIAL OF
able ores (Figure 5). TYPICAL ULTRAMAFIC BODIES
Fresh dunite (or any unserpentinized peridotite for that The CO 2-sequestering potential of some USA ultramafic
matter) is relatively uncommon in large quantities. The larg- bodies is compared in Table 4. Volume, bulk density, and
30 30 FIGURE 5: Diagram showing lower right
"Alpine" corner of the Opx-Cpx-Ol ternary and
Peridotite Serpentinized "Alpine" lower half of the Opx-Ol-Serp ternary
Dunite Peridotite and Dunite
and Dunite (modal-%). Because pyroxenes (and ac-
cessory chromite) are not soluble in HCl,
5 5 dunite and thoroughly serpentinized dunite
Wehrlite Best Ore
or harzburgite comprise the best ores for
Cpx Ol Serp CO2 sequestering. Numbers around the
perimeter of diagram are percent Opx.
TABLE 4. Physical-chemical and CO2-sequestering properties of example peridotite/serpentinite bodies (data from Czamanske and Zientek, 1985, Goff
et al., 1997, Hess, 1960, Himmelberg and Coleman, 1968, Himmelberg and Loney, 1973, Lobatka and Albee, 1979, Morgan, 1977, and Ragan, 1963).
WA OR CA VT CA MD CA MT
Area (km2) 91 16 40 2.3 200 100 4? 26
Depth (km) 0.6 0.5 0.3 ≤1 ≥0.2 0.3 0.25 0.5
Est. Vol. (km3) 54 8 12 2 40 30 1 13
Density (g/cm3) 3.3 3.2 2.8 2.9 2.65 2.7 2.6 3.1
Peridotite 29.9 27.4 27.2 29.1 ––– ––– ––– 17.5
Serpentinite –––a ––– 20.9 23.1 21.8 21.2 19.9 –––
Combined 29 27 23 26 21 21 20 17
R(CO2)b 1.91 2.05 2.40 2.13 2.63 2.61 2.76 3.25
Mg (109 tons) 52.0 6.91 7.73 1.5 22.3 17.0 0.5 6.85
CO2 (109 tons) 94.1 12.5 14.0 2.7 40.3 30.8 0.9 12.4
USA (yr)c 18.8 2.50 2.80 0.55 8.05 6.15 0.18 2.48
World (yr)c 4.95 0.66 0.74 0.14 2.12 1.62 0.048 0.65
(–––). not significant.
R(CO2) is the calculated mass ratio of rock processed to CO2 disposed.
Assumes annual USA and World CO2 emission rates of ~5 × 109 and ~19 × 109 tons/yr. respectively.
Mg content for each body were estimated using published would generate ~0.15 tons dissolved Fe per ton of fixed CO2
geologic maps, reports, and chemical analyses. As men- or a total of 2 gigatons Fe for the entire deposit which would
tioned above, fresh peridotites, especially dunites, contain be precipitated as the oxide for feedstock in other industries.
the most Mg per unit mass of rock. Stillwater peridotite con- If completely mined, total Mg in the Del Puerto deposit
tains relatively more pyroxene plus some plagioclase but could fix 14 gigatons of CO2.
less olivine and is the least attractive ultramafic body evalu- Compared with a typical coal mine, the Del Puerto deposit
ated in terms of Mg content. All economic trade-offs for min- is quite large. By itself, the Del Puerto deposit could dispose
ing and processing various ultramafic rocks have not been of all the CO2 emissions associated with a population of 1 0
evaluated but some are discussed below. million for ~70 years. Thus, it would suffice for the local
The sequestering potential of small ultramafic bodies i s region. However, to absorb the CO2 output of the US many
considerable. For example the Belvidere Mountain ultrama- more such mines would be needed. Del Puerto could only
fic body, having an estimated volume of ~2 km3, is capable handle 2.8 years of total US output.
of handling the equivalent of 0.5 year of present total U.S.
CO2 emissions (~5 gigatons/yr; this value includes auto ADDITIONAL CONSIDERATIONS
emissions). The large dunite at Twin Sisters, Washington Strategic Minerals
could dispose of nearly 19 years of U.S. CO2 emissions and Ultramafic rocks contain many mineral resources.
~5 years of global emissions. From another perspective, the Chrome, platinum group metals, nickel, cobalt, and dia-
abundance of Mg in the Earth’s crust (2.0 mol-%) is nearly 6 0 monds come from various ultramafic rocks and their eroded
times greater than the abundance of C (0.035 mol-%; Brown- products, whereas manganese, copper, mercury, and other
low, 1979); thus, it is not surprising that there should be metals are sometimes obtained from within the bodies or
more than enough Mg in ultramafic deposits to sequester from enclosing rocks (Maddock, 1964). Metals such as these
global CO2 emissions. are of strategic importance during desperate political periods
At a deposit such as the Del Puerto ultramafic body, Cali- but may be of environmental concern in large-scale mining
fornia (Evarts and Schiffman, 1982), a ton of sequestered CO2 operations. The metal-bearing residues could be back-filled
would require on average mining, crushing, and dissolving into excavations or could be stockpiled for future use. Some
2.4 tons of ultramafic rock (Table 4, where R = mass ratio of metals such as Cr are bound primarily in chromite and minor
rock processed to CO2 sequestered). Every ton of CO2 would clinopyroxene which are relatively benign in the sequester-
precipitate 1.8 tons of magnesite that would be backfilled ing process. Most of the Ni and Mn is hosted in the olivine
into the existing mine. An additional 1.2 tons of silica and or products formed during serpentinization and are released
residual minerals such as pyroxenes and chromite would be by acid dissolution.
backfilled with the magnesite, although some could be Following the example used above, complete utilization of
stockpiled and sold for other industrial uses. The process the Del Puerto ultramafic body would yield at least 80 mega-
tons of Cr, another 80 megatons of Ni, and perhaps 30 mega- Because ultramafic rocks are exceptionally low in K and P,
tons of Mn. These quantities could dramatically impact the and rich in Mg and Fe, they sustain a unique flora and fauna
economics of existing metals industries but would be a bene- that are noticeable to even the most casual of observers
ficial by-product of CO2 sequestering. (O’Hanley, 1996). Most grasses grow with difficulty on ser-
pentinites; instead one observes a restricted assemblage of
Chrysotile Asbestos thorny brush, pines and cypress trees (Figure 4A). Ultramafic
Serpentinites host commercial deposits of chrysotile or soils tend to be brick red to dark orange from oxidation of
"white" asbestos (Coleman, 1996; O’Hanley, 1996). Because Fe. Near-surface outcrops combine these bright soil colors
most serpentinites contain appreciable chrysotile (usually i n with the greens, grays, and blacks of the host rock. Sharp
noncommercial form), special environmental precautions vegetation and soil contrasts occur along contacts between
may be required during mining. Chrysotile accounts for ultramafic rocks and other formations. Because serpentinite
~90% of the asbestos that has been used historically in the is the "state rock" of California, special care is generally
United States (Ross, 1981), and it is the primary type of as- taken to restore or reseed areas where serpentinites are dis-
bestos used in insulation and many other construction mate- turbed in this state (i.e., Dellinger, 1997).
rials. Chrysotile forms the bulk of asbestos-contaminated
waste resulting from removal activities. CONCLUSIONS
The health risks associated with chrysotile have been the Steady increases in global emissions will require new
focus of extensive scientific and public debate (Alleman and technologies to capture and immobilize waste CO2. Conver-
Mossman, 1997). Many of the arguments deliberate whether sion of CO2 into thermodynamically stable magnesite is one
chrysotile can cause mesothelioma (a rare type of cancer) i n of many technologies under current examination. Abundant
humans. Although this issue is largely unresolved, the risk resources of Mg-rich peridotite and serpentinite exist within
appears to be much less than that posed by amphibole asbes- the United States and many other countries. Peridotite and
tos (Mossman et al., 1990; Hume and Rimstidt, 1992). serpentinite are relatively soluble in HCl; thus, huge quanti-
Chrysotile asbestos continues to be removed from buildings, ties of Mg can be easily dissolved for further chemical uses.
albeit at a lower level than in years past. This asbestos- Engineering and technology advances could lead to con-
containing material must be disposed of in landfills with struction of coal- or gas-fired power plants in which waste
special precautions prescribed by the U.S. Environmental CO2 is fed to a sequestering plant adjacent to an open-pit
Protection Agency (e.g., the use of a >15-cm cover of asbes- ultramafic mine. Peridotite, serpentinite, and waste asbestos
tos-free material). Waste chrysotile asbestos could be used would be consumed whereas magnesite and silica residues
as feedstock in the sequestering of CO2 if it was relatively would be backfilled into the pit. Byproducts would include
uncontaminated by other materials. Fe, Cr, Ni, Mn, and possibly other metals.
Retrofitting all existing fossil fuel plants for benign CO2
Large-scale Open Pit Mining disposal is surely an impractical task, but the global com-
Open pit mining on the scale envisioned here would have munity must eventually deal with the CO2 dilemma. CO2
profound economic and environmental impact. Projected sequestering in magnesite is one of many technologies that
costs would probably be similar to open-pit copper mining, may eventually reduce or stabilize emissions. The environ-
as mentioned above. Precedents for open-pit mining of ser- mental impact of large-scale ultramafic mining and CO2 se-
pentinite presently exist at New Idria and at other mines i n questering, with associated increases in energy costs, would
the California Coast Ranges (Coleman, 1996). have to be weighed against the counterimpact of continued
The McLaughlin gold mine (Homestake Mining Company) CO 2 emissions to the atmosphere and the risk of accelerated
was constructed in a similar geologic and physiographic climate change.
environment to the Wilbur Springs serpentinite in the late
1970’s. The McLaughlin mine is located ~30 to 40 km south ACKNOWLEDGMENTS
of the Wilbur Springs region and was over 100 m deep and We thank the following people: G. Guthrie (Los Alamos
1000 m long during maximum exploitation (Sherlock et al., National Laboratory [LANL]) for input on asbestos; D.
1995). The gold has now been mined out and the site is be- Counce, E. Kluk, and M. Snow (LANL) for various analyses;
ing reclaimed. Processing of stockpiled ore will yield 19.2 D. Bergfeld (LANL) for graphics; A. Adams (LANL) for thin
tons of gold before Homestake abandons the site in the year sections; P. Canelli (UNIMIN Corp.) for peridotite samples
2003 (Field, 1996). A large open-pit mine designed for car- from two active mines; California Division of Mines and
bonate waste disposal would probably have considerable Geology for maps and information; and R. G. Coleman (Stan-
community support in this economically depressed area. ford University) for good advice. Initial reviews were pro-
However, the environmental impact would have to be fully vided by M. J. Aldrich and B. Carey (LANL). Final reviews
evaluated before any development proceeded. were obtained from R. G. Coleman and two anonymous indi-
viduals. This research was funded by the Program Directorate
Ultramafic Rocks and Ecology Energy Technologies PDET (E. Joyce; LANL).
Goff, F. and Janik, C. J. (1993). Gas geochemistry and guide
for geothermal features in the Clear Lake region, Califor-
REFERENCES nia. In J. J. Rytuba (Ed.), Active geothermal systems and
Alleman, J. E., and Mossman, B. T., 1997, Asbestos revisted. gold-mercury deposits in the Sonoma-Clear Lake volcanic
Sci Am, 277, 70-75. fields, California (pp. 207-261). Littleton, CO: Society of
Barnes, I., LaMarche, V. C., and Himmelberg, G.R., (1967). Economic Geology Guidebook Series, vol. 16.
Geochemical evidence of present-day serpentinization. Goff, F., Guthrie, G., Counce, D., Kluk, E., Bergfeld, D., and
Science, 56, 830-832. Snow, M. (1997). Preliminary investigations on the CO2
Barnes, I., Rapp, J. B., O’Neil, J. R., Sheppard, R. A., and Gude, sequestering potential of ultramafic rocks. Los Alamos,
A. J.,III, (1972). Metamorphic assemblages and the direc- NM: Los Alamos National Laboratry Rep. LA-13328-MS.
tion of flow of metamorphic fluids in four instances of Govindaraju, K. (1994). Geostandards Newsletter (special
serpentinization. Contrib Mineral. Petrol, 35, 263-276. issue), 18, 15.
Barnes, I., O’Neil, J. R., Rapp, J. B., and White, D. E. (1973). Harper, G. D. (1984). The Josephine Ophiolite, northwestern
Silica-carbonate alteration of serpentine: Wall rock altera- California. Geol Soc Am Bull, 95, p. 1009-1026.
tion in mercury deposits of the California Coast Ranges. Hawkes, H. E., Wells, F. G., and Wheeler, D. P. (1942). Chro-
Econ Geol, 68, 388-398. mite and quicksilver deposits of the Del Puerto area,
Barnes, V. E., Shock, D. A., and Cunningham, W. A. (1950). In Stanislaus County, California. U.S. Geol Surv Bull, 936-D,
Utilization of Texas serpentine (pp. 5-52). Austin TX: 79-110.
University of Texas, Publication No. 5020, Hess, H. H. (1960). Stillwater igneous complex, Montana: A
Blok, K., Turkenburg, W. C., Hendriks, C. A., and Steinberg, quantitative mineralogical study. Boulder CO: Geological
M. (Eds.) (1992). Proceedings of the first international Society of America Memoir 80.
conference on carbon dioxide removal. Oxford: Pergamon. Himmelberg, G. R., and Coleman, R. G. (1968). Chemistry of
Bodenlos, A. J. (1950). Geology of the Red Mountain magne- Primary Minerals and Rocks from the Red Mountain-Del
site district, Santa Clara and Stanislaus Counties, Califor- Puerto Ultramafic Mass, California. U.S. Geol Surv Profes-
nia. Calif J Mines Geol., 46, 223-278. sional Paper, 600-C, C18-C26.
Brownlow, A. H., (1979). Geochemistry. Englewood Cliffs, Himmelberg, G. R., and Loney, R. A. (1973). Petrology of the
NJ: Prentice-Hall. Vulcan Peak alpine-type peridotite, southwestern Oregon.
Butt, D. P., Lackner, K. S., Wendt, C. H., Conone, S. D., Kung, Geol Soc Am Bull, 84, 1585-1600.
H., Lu, Y.-C., Bremser, J. (1996). Kinetics of thermal dehy- Houston, E. C., (1945). Magnesium from olivine. New York:
droxylation and carbonation of magnesium hydroxide. J American Institute of Mines and Materials Engineering,
Am Ceramics Soc, 79, 1982-1998. Tech. Pub. 1828, Class D, No. 85.
Cawthorn, R. G. (Ed.) (1997). Layered Intrusions. Amster- Hume, L. A., and Rimstidt, J. D. (1992). The biodurability of
dam: Elsevier. chrysotile asbestos. Am Mineral, 77, 1125-1128.
Coleman, R. G. (1977). Ophiolites. Berlin: Springer-Verlag. Jennings, C. W. (1977). Geologic map of California. Sacra-
Coleman, R. G. (1996). New Idria Serpentinite: A land man- mento: California Division of Mines and Geology, Geol.
agement dilemma. Environ Eng Geosci, 2, 9-22. Data Map 2, 1:750,000 scale (color).
Czamanske, G. K., and Zientek, M. L. (1985). The Stillwater Kelly, K. K. (1945). Energy requirements and equilibria i n
Complex, Montana: Geology and guide. Butte, MT: Mon- the dehydration, hydrolysis, and decomposition of mag-
tana Bureau of Mines and Geology, Spec. Publ. 92. nesium chloride. Washington, DC: U.S. Department of Inte-
Dickinson, W. R., Hopson, C. A., and Saleeby, J. B. (1996). rior, Bureau of Mines Tech. Paper 676.
Alternate origins of the Coast Range Ophiolite (Califor- Kerrick, D. M., McKibben, M. A., Seward, T. M., and Caldiera,
nia): Introduction and implications. GSA Today, 6, 1-10. K. (1995). Convective hydrothermal CO2 emission from
Dellinger, M. (1997), The Lake County-Geysers effluent pipe- high heat flow regions. Chem Geol, 121, 285-293.
line and injection project. Geotherm Resour Counc Bull, Labotka, T. C., and Albee, A. L. (1979). Serpentinization of
26, 218-223. the Belvidere Mountain ultramafic body, Vermont: Mass
Evarts, R. C. (1977). The geology and petrology of the Del balance and reaction at the metasomatic front. Can Min-
Puerto ophiolite, Diablo Range, central California Coast eral, 17, 831-845.
Ranges. In R. G. Coleman, and W. P. Irwin, (Eds.), North Lackner, K. S., Wendt, C. H., Butt, D. P., Joyce, E. L., Jr., and
American Ophiolites (pp. 121-139). Portland: Oregon De- Sharp, D. H. (1995). Carbon dioxide disposal in carbonate
partment of Geology and Mineral. Industries Bull. 95. minerals. Energy, 20, 1153-1170.
Evarts, R. C., and Schiffman, P. (1982). Submarine hydro- Lackner, K. S., Butt, D. P., and Wendt, C. H. (1997). Magnesite
thermal metamorphism of the Del Puerto ophiolite, Cali- disposal of carbon dioxide. Proceedings 22nd interna-
fornia. Am J Sci 283, 289-342. tional conference on coal utilization and fuel systems,
Field, D. (1996). Homestake Charts its Course at Lower (pp. 419-430). Washington DC: Coal Slurry Technology
Lake’s McLaughlin Mine. Lake County Record-Bee, p. 2. Association.
Maddock, M. E. (1964). Geologic map and sections of the Thayer, T. P. (1977) The Canyon Mountain Complex, Oregon,
Mount Boardman Quadrangle, Santa Clara and Stanis- and some problems of ophiolites. In R. G. Coleman and W.
laus counties, California. Sacramento: California Divi- P. Irwin (Eds.), North American ophiolites (pp. 93-105).
sion of Mines and Geology, Map Sheet 3, 1:62,500 (color). Salem: Oregon Department of Geology and Mineralogical
McLaughlin, R. J., Ohlin, H. N., Thormahlen, D. J., Jones, D. L., Industries Bull. 95.
Miller, J. W., and Blome, C. D. (1989). Geologic map and United Nations. (1995). 1993 energy statistics yearbook.
structure sections of the Little Indian Valley-Wilbur New York: United Nations.
Springs geothermal area, northern Coast Ranges, Cali- Weart, S. R., (1977) The discovery of the risk of global warm-
fornia. Washington DC: U.S. Geological Survey Misc. In- ing. Phys Today, Jan., 34-40.
vest. Map I-1706, 1:24,000 scale, 2 sheets (color).
Morgan, B. A. (1977). The Baltimore Complex, Maryland,
Pennsylvania, and Virginia. In R. G.Coleman and W.P.
Irwin (Eds.), North American ophiolites (pp. 41-49). Port-
land: Oregon Department of Geology and Mineralogical
Industries Bull. 95.
Mossman, B. T., Bignon, J., Corn, M., Seaton, A., and Gee, J. B.
L. (1990). Asbestos: Scientific developments and implica-
tions for public policy. Science, 247, 294-301.
O’Hanley, D. S., 1996, Serpentinites: Records of tectonic and
petrological history. Oxf Monogr Geol and Geophys, 34, 1-
Page, N. J. (1967). Serpentinization at Burro Mountain, Cali-
fornia. Contrib Mineral Petrol, 14, 321-342.
Peters, E. K. (1991). Gold-bearing Hot Spring Systems of the
Northern Coast Ranges, California. Econ Geol, 86, 1519-
Ragan, D. M. (1963). Emplacement of the Twin Sisters Dunite,
Washington. Am J Sci, 261, 549-565.
Ramanathan, V. (1988). The greenhouse theory of climate
change: A test by an inadvertent global experiment. Sci-
ence, 240, 293-295.
Rice, S. J. (1957). Asbestos. In L. A. Wright, (Ed.), Mineral
Commodities of California (pp. 49-58). Sacramento: Cali-
fornia Division of Mines Bull. 176.
Robie, R. A., Hemingway, B. S., and Fisher, J. R. (1979).
Thermodynamic properties of minerals and related sub-
stances at 298.15°K and 1 bar (105 pascals) pressure and at
higher temperatures. U.S. Geol Surv Bull., 1452.
Ross, M. (1981). The geologic occurrences and health haz-
ards of amphibole and serpentine asbestos. In D. R. Veblen
(Ed.), Amphiboles and Other Hydrous Pyri-
boles—Mineralogy, (pp. 279-323). Washington, DC: Min-
eralogical Society of America.
Sabine, C. L., Wallace, D. W. R., and Millero, F. J. (1997). Sur-
vey of CO2 in the oceans reveals clues about global carbon
cycle. EOS, 78, 51, 54-55.
Saleeby, J. B. (1982). Polygenetic Ophiolite Belt of the Cali-
fornia Sierra Nevada: Geochronological and Tectonostrati-
graphic Development. J Geophys Res, 87, 1803-1824.
Sherlock, R. L., Tosdal, R. M., Lehrman, N. J., Graney, J. R.,
Losh, S., Jowett, E. C., and Kesler, S. E. (1995). Origin of the
McLaughlin Mine Sheeted Vein Complex: Metal Zoning,
Fluid Inclusion, and Isotopic Evidence. Econ Geol, 90,