Chemical Geology, 71 (1988) 211-222 211

Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands

METHANE-HYDROGEN GAS SEEPS, ZAMBALES OPHIOLITE,

PHILIPPINES:
DEEP OR SHALLOW ORIGIN?

T.A. A B R A J A N O ' , N.C. S T U R C H I O ' , J . K . B O H L K E 1, G.L. L Y O N 2, R.J. P O R E D A 3'* a n d

C.M. S T E V E N S 1

~Argonne National Laboratory, Argonne, IL 60439 (U.S.A.)

2Institute of Nuclear Science, DSIR, Lower Hutt (New Zealand)
3Isotope Laboratory, Scripps Institution of Oceanography, University of California at San Diego,
La Jolla, CA 92093 (U.S.A.)

(Revised and accepted July 7, 1988)

Abstract

Abrajano, T.A., Sturchio, N.C., Bohlke, J.K., Lyon, G.L., Poreda, R.J. and Stevens, C.M., 1988. Methane-hydrogen
gas seeps, Zambales Ophiolite, Philippines: Deep or shallow origin? In: M. Schoell (Guest-Editor), Origins of Meth-
ane in the Earth. Chem. Geol., 71: 211-222.

Isotopically anomalous CH4-rich gas escapes at low flow rate and ambient temperature from seeps in serpentinized
ultramafic rock in the Zambales 0phiolite, Philippines. The major components of the gas are CH4 (55 mole% ) and
H2 (42 mole% ); the C H 4 / C O 2 ratio is > 1800and the CH4/He ratio is 9.2.10 4. The • 13C-valueof the CH4 is - 7.0 + 0.4%o
(PDB), ~ 8%o higher than the highest published values for CH4 in other natural gases and hot springs, but similar to
values commonly attributed to mantle carbon. The 3He/4He ratio is 5.70.10 -6, 4.1 times the atmospheric ratio,
indicative of a substantial mantle He component. The JD-values of CH4 and H2 are - 136 and - 590%o,respectively,
consistent with equilibration temperatures of 110-125°C. Carbon and He isotopic data could be consistent with
derivation of the Zambales gas directly from a reduced mantle. However, phase equilibria and H isotope data indicate
that the gas also could have been produced by reduction of water and carbon during low-temperature serpentinization
of the ophiolite.

1. I n t r o d u c t i o n
T h e i m p o r t a n c e o f deep E a r t h c a r b o n as a
source o f significant h y d r o c a r b o n deposits is a
subject o f m u c h i n t e r e s t a n d debate. A c c o r d i n g
to c o n v e n t i o n a l t h e o r y , h y d r o c a r b o n deposits
(oil a n d gas) are p r o d u c e d b y diagenetic alter-
*Present address: Department of Geological Sciences, Uni-
versity of Rochester, Rochester, NY 14672, U.S.A.
a t i o n o f s e d i m e n t a r y organic m a t t e r as a con-
s e q u e n c e o f b a c t e r i a l activity or deep burial a n d
h e a t i n g in s e d i m e n t a r y b a s i n s ( H u n t , 1979;
Schoell, 1983). In c o n t r a s t , it has b e e n pro-
p o s e d t h a t vast a m o u n t s o f p r i m o r d i a l CH4 a n d
o t h e r h y d r o c a r b o n s are s t o r e d in t h e E a r t h ' s
m a n t l e a n d could be a source o f significant de-
posits in t h e c r u s t ( G o l d a n d Soter, 1980; Giar-
dini et al., 1982). It is also possible t h a t
s u b d u c t e d s e d i m e n t a r y organic c a r b o n is pres-
e n t at m a n t l e d e p t h s a n d m i g r a t e s to shallower

0009-2541/88/$03.50 © 1988 Elsevier Science Publishers B.V.

212
regions in the crust (cf. Javoy et al., 1982; Marty
and Jambon, 1987).
A number of studies related to deep Earth hy-
drocarbons are currently being pursued. For ex-
ample, the Siljan Ring structure (Sweden) has
recently been drilled to test whether accumu-
lations of mantle CH4 have formed underneath
(Fish, 1987; Jeffrey and Kaplan, 1988). Some
studies have indicated that CH4 of apparent
mantle derivation occurs in some volcanic rocks
{e.g., Atlantic "popping" rocks ) and is actively
being emitted from submarine hydrothermal
vents along mid-ocean ridges (H~kinian et al.,
1973; Pineau et al., 1976; Welhan and Craig,
1979, 1982). Compositional variations of mag-
mas erupted in different tectonic settings have
been rationalized in the context of heteroge-
neously distributed reduced fluid species in the
mantle (Woermann and Rosenhauer, 1985;
Green et al., 1987). Intrinsic fo2 measurements
made in recent years indicate that redox con-
ditions closer to iron-wiistite (IW) than
quartz-fayalite-magnetite (QFM) fo2 buffer
curves may prevail in some portions of the
mantle (e.g., Arculus and Delano, 1981 ). Ther-
mochemical calculations indicate that CH4 and
H2 are likely to be the dominant fluid species
under these reducing conditions (e.g., Eggler
and Baker, 1982; Woermann and Rosenhauer,
1985; Green et al., 1987).
It also has been proposed that abiogenic CH4
can form as a product of fluid-rock interaction
processes within the crust (French, 1966; Frost,
1985; Abrajano and Pasteris, 1983; Abrajano,
1984; Holloway, 1984; Apps, 1985). Conditions
under which CH4 can be produced occur where
there are adequate supplies of carbon and a
suitable reductant (such as H2, Fe 2+ or Fe°).
Methane or H2 gas seeps or significant ground-
water concentrations of these reduced gas spe-
cies have been reported in numerous localities
with no clear relation to organic sources (Stoll,
1959; Hahn-Weinheimer and Rost, 1961; Be-
techtin, 1961; Thayer, 1966; Karshavin and
Vendillo, 1970; Petersil'ye et al., 1970; Wood,
1970; Gephart et al., 1979; Petersil'ye and Pri-
pachkin, 1979; Barnes et al., 1981; Abrajano and
Pasteris, 1983; Buntin, 1983; Neal and Stanger,
1983; Abrajano, 1984; Sofer, 1985; Coveney et
al., 1987; Fritz et al., 1987). In many cases, these
reduced gas occurrences are closely associated
with mafic-ultramafic rocks. Methane gas seeps
are not found in all mafic-ultramafic terranes,
however, perhaps because they do not all main-
tain the appropriate conditions for CH4 pro-
duction or storage. The direct derivation of'
reduced gases from the underlying mantle can-
not be precluded in many of these localities be-
cause critical isotopic data are not available.
In this paper, we report chemical and iso-
topic data for a reduced gas occurrence in the
Zambales Ophiolite, Philippines, and present
arguments pertaining to the origin of the gas.
The Zambales Ophiolite consists of fragments
of Eocene oceanic crust that formed in prox-
imity to an island arc. The obducted ophiolite
now occupies a fore-arc setting with respect to
the Manila Trench subduction system (Fig. 1 ).
The sampled gas seeps, known locally as Los
Fuegos Eternos (LFE), issue from fractures in
well-exposed partially serpentinized ultramafic
rock. The nearest outcrop of sedimentary rocks
is ~ 15 km west of LFE. Available geophysical
data indicate that the exposed ultramafic ter-
rane extends to at least several km depth. The
Zambales gas seeps were first described by Stoll
(1959) who noted yellow-orange flames ex-
tending as much as 30 cm high and gas seeping
with slight whispering and rustling sounds.
Thayer (1966) reported the composition of the
Zambales gas as analyzed by the San Jose Oil
Company (see Table I ). When we first visited
Los Fuegos Eternos in 1981, the flames were
only a few centimeters high and were distrib-
uted in two separate areas that are ~ 20 m 2each.
When one of the sites was sampled for this study
in 1986, the flames had been extinguished, pos-
sibly by an unusually heavy rainstorm.
2. R e s u l t s
Three gas samples were collected through
copper tubing into evacuated stainless-steel

I I I I
A-~
| I I I I I
120 ° 122 ° f24 °
,o,
Fig. i.Location m a p of the Zambales Ophiolite. Insert shows
schematic cross-section through major tectonic features re-
lated to the ophiolite terrane.
containers. Samples LFE-2 and LFE-3 were
collected from rock fracture seeps ~ 1 m apart
whereas LFE-I was collected from a bubbling
seep in water-saturated sandy soil covering ser-
pentinized rocks. All three gas samples were
collected in one of the two seep areas noted in
Section 1. Bulk gas compositions were deter-
mined mass spectrometrically. Carbon, H, and
He isotopic analyses followed procedures de-
scribed in detail elsewhere (Lupton and Craig,
1975; Lyon and Cox, 1978; Rust and Stevens,
1980; Hulston et al., 1981; Poreda et al., 1986).
Chemical and isotopic analyses of the gas sam-
ples are listed in Table I. The close similarity
between the chemical compositions reported by
Thayer (1966) and those from the present study
suggests that no significant change in bulk gas
composition has occurred for the past twenty
or more years (Table I). The ethane contents
213
I of these gases are reported here for the first time.
Air contamination accounts for <2% of sam-
18°N ples LFE-2 and LFE-3, whereas LFE-I is dom-
inated by air.
17 °
Determinations of J13C (PDB) of CH4 in the
three samples made by three independent lab-
oratories yield an average value of - 7.0%o with
16 °
a standard deviation of + 0.4%o. This value is
~ 8%o more positive than the highest published
15 °
values for CH4 in natural gas seeps and hot
springs, suggesting an unusual mode of origin
14 ° for the Zambales gas. Indeed, this J'3C-value is
indistinguishable from the commonly accepted
15 o
carbon isotopic composition of the mantle based
on isotopic studies of diamonds and mantle-de-
rived magmas (Deines, 1980; Kyser, 1986). The
3He/4He ratio of the gas is 5.70-10 -8 (4.1 times
the atmospheric ratio, RA), indicating a sub-
stantial mantle He component. The JDcs4 -val-
ues in LFE-2 and LFE-3 are identical within
errors of measurements. The average JDcs4,
- 136%o, is within the range of CH4 produced
thermogenically from sediments and is quite
distinct from JD-values of CH4 produced mi-
crobially either by carbonate reduction or ace-
tate dissimilation (e.g., Schoell, 1983; Whiticar
et al., 1986). The JDH~-Values of - 5 8 1 and
-599%o in these two gas samples are identical
within analytical errors. The hydrogen is ex-
tremely depleted in D, but less so than reported
for other H2 occurrences thought to be related
to serpentinization (Neal and Stanger, 1983;
Coveney et al., 1987). The j13Cco 2 value mea-
sured for LFE-2 ( - 3 2 % 0 ) is too low to be in
isotopic equilibrium with CH4. This C02 com-
ponent may be attributed to shallow oxidation
of CH4 or stripping of soil gas during ascent of
the CH4-H2 gas.
3. D i s c u s s i o n
3.1. Origin of the Zambales gas
Derivation of the Zambales gas entirely from
sedimentary organic-rich sources appears to be
precluded by its relatively low CH4/H2 ratio, its
214

TABLE I

Composition of Zambales gas

Sample 1A 1B LFE-1 LFE-2 LFE-3

Reference [1 ] [1 ] [2] [2] [2]

H2 (vol.%) 41.4 45.6 8.4 42.3 42.6

CH4 52.6 53.8 13.0 55.3 54.8
C2H4 tr. 0.07 0.04
C2H~ 0.04 0.15 0.12
N2 + CO 60.6 1.5 1.8
02 0.5 0.2 16.5 0.09 0.08
Ar 0.73 0.01 0.02
CO2 0.03 < 0.01 0.03

H e ('l) (ppm) 5.1 6.9

Ne !*~ 0.04 0.07

J13CcH, (%0) - 7 . 1 "e _ 7.34 ":~ _ 6.93 ":~

-- 7.50 .4 -- 6.11.4
5Dcm -- 122 .3 -- 118 "3
_ 137 TM _ 134 TM
¢~DH~ - 581"4 _ 599-4
(~13Cco z _32"~

aHe/4He(.l~ 5.72-10 '; 5.69-10 ~

RA 4.07 3.98

References: [1 ] = T h a y e r (1966); [2] = t h i s work.

Laboratory: " S c r i p p s Institution of Oceanography;
"2Argonne National Laboratory; "3Global Geochemistry Corp.;
"4Department of Scientific and Industrial Research, New Zealand.

low concentrations of ethane and CO2, its high
J'3CcH4, and its high 3He/4He ratio. We there-
fore consider two possible abiogenic models for
the origin of the gas, in which organic compo-
nents may be present but are not predominant.
3.1.1. Mantle origin for the Zambales gas. The
observed He isotopic composition of the Zam-
bales gas (4.1RA) indicates that a major portion
of the He is mantle-derived. However, the ob-
served 3He/4He ratio is significantly lower than
that observed in fumaroles from volcanic areas
south of the ophiolite (7.4RA) ( R.J. Poreda and
H. Craig, pers. commun., 1988) and the Zam-
bales gas seeps occur in an area totally devoid
of recent magmatic activity. The presence of a
significant magmatic He component in the
Zambales gas without accompanying magmatic
activity in the gas seep locality is an important
constraint in understanding the origin of the
major components of the gas.
Craig and coworkers suggested the use of
CH4/3He vs. 3He/4He plots in constraining the
origin of CH4 in natural gas fields (e.g., Poreda
et al., 1986; Craig and Horibe, 1988). In Fig. 2,
we compare the CHa and He isotope abun-
dances in the Zambales gas with similar data
compiled from other studies (Wakita and Sano,
1983; Poreda et al., 1986; Craig et al., 1987; Jen-
den et al., 1988; Wakita et al., 1988). It is not-
able that the Zambales gases plot along a two-
component mixing trend defined by gases col-
lected in gas seeps and volcanoes in Japan (e.g.,
Wakita and Sano, 1983; Craig and Horibe, 1988;
Wakita et al., 1988). The end-members in this
mixing trend can be interpreted as: (1) a crus-
tal component that has a relatively high CH4/
3He ratio and a low 3He/aHe ratio; and (2) a
 215

Ca)
- 80

- 60
Q~
T
%
"1- 10
(..) -40

(.9
0 • ZAMRAEES GAS //
I O,O JAPAN GAS
[ ] SACRAMENTO GAS -20
@m L~
• • JAPAN VOLCANO
• • EPR
[ ] REDUCED E P l
8 • ~ ATLANTIS
W MARIANAS • I I I0 I |
2 4 6 8 I 12 14

I l l l I I LO G CHJ3He
5 6 7

LOG 4He/~He
(b)

Fig. 2. The ratios of 4He/3Heand CH4/3Hein the Zambales

gas shown with data from other localities. Closed circles for ~ o
the Japan gas are for samples collectedfrom the Green Tuff
formation. Mixing of gas similar to those from Japan vol- - 60

canoes (V) or EPR ( • ) with crustal gas with high CH4/

3He ratios will produce gas compositions in the field de- I
J 0
0 o/n
/ -OlD
D ~,J
~--~--4t
freed by the Japan gases. The "mantle" end-member could - 40 ~ O/ ~ / I

be represented by Japan Green Tuff Formation gases ( • )

as well (Wakita et al., 1988).
• \D~ /

mantle component that has a high 3He/4He ra-

tio and a relatively low CH4/3He ratio. Accord-
ing to Fig. 2, the Zambales gas could be
interpreted as a mixture of gases from sources
similar to those proposed for the J a p a n gases.
It is noteworthy that the mantle component in
this model is not required to contain a signifi-
cant amount of CH4.
Fig. 3a and b includes analyses for most of
the same samples shown in Fig. 2 b u t projected
in CH4/3He vs. j13C and 3He/4He vs. 613C dia-
grams. In contrast to the inference made from
Fig. 2, the same two end-member mixing model
does not appear to explain adequately the car-
bon isotopic composition of the Zambales gas.
The J13Ccm of the Zambales gas is distinctly
less negative than that predicted by the two end-
member mixing model used for the J a p a n gases.
The Zambales gas data seem to require either
(1) a third component that has a mantle-like
J13C b u t relatively high CH4/3He ratio or (2) a
wider range of CH4/3He ratios for the mantle
lal
LOG 4HePHe
Fig. 3. a. The ratio of CH4/3He plotted against J13Cc m (in
%0) b. The ratio of 4He/3He plotted against JlZCcm (in
%o). Mixing of an EPR-type C H4 ( • ) gas with isotopically
light crustal CH4 will produce gases with •13CcH4-Values
more negative than - 15%o. All symbolsare the same as for
Fig. 2.
end-member than indicated by submarine hy-
drothermal plumes and mid-ocean ridge basalt
( M O R B ) data (cf. M a r t y and Jambon, 1987;
Wakita et al., 1988). The latter suggestion could
be consistent with the primordial CH4 hypoth-
esis advanced by Gold (e.g., Gold and Soter,
1980).
The CH4/3He ratios of mantle fluids are not
well known and could be variable. Variations in
the CH4/3He ratio could be due to (1) varia-
tions in the ratio of total carbon to 3He a n d / o r
(2) temporal or spatial variability in the man-
216
tle redox state. In a fluid phase with constant
'~He and total carbon contents, the CH4/3He ra-
tio increases as the oxygen fugacity decreases.
Quantitative reduction of CO2 to CH4 in a
MORB or E P R (East Pacific Rise) -type fluid
(e.g., from the QFM oxygen buffer to the IW
oxygen buffer) could increase the CH4/3He ra-
tio by as much as three orders of magnitude (re-
duced EPR in Figs. 3 and 4) (cf. Marty and
Jambon, 1987). If mantle conditions were suf-
ficiently reducing for this to occur, the isotopic
composition of the CH4 would equal the bulk
mantle value for carbon ( - 7 to -5%c ). The
higher 513CcH4 and CH4/3He ratio in the Zam-
bales gas, relative to a MORB- or EPR-type gas,
is thus consistent with this reduction trend (Fig.
3a and b), although still higher total carbon
content than present in EPR-type gases may be
required to explain the high CH4/3He ratio in
the Zambales gas. The Zambales gas can there-
fore be derived from reduction of a MORB- or
EPR-type gas, but could also represent a man-
tle fluid with a higher C/3He ratio than these
gases. If derived directly (i.e. chemically un-
modified) from a region in the mantle contain-
ing graphite, the Zambales gas should have had
a higher C H J H 2 ratio than observed (i.e. es-
sentially pure CH4) {e.g., Eggler and Baker,
1982). Furthermore, very reducing conditions
(at or around IW ) in the source region are im-
plied if the Zambales gas was derived directly
from the mantle.
The bulk 5D (CH4 + H2) for the Zambales gas
( - 355%~ ) is distinctly lower than that of typ-
ical mantle 5D-values ( - 8 0 to -50%0 ) (Kyser,
1986), implying that a closed-system mantle
origin for hydrogen is unlikely. The most likely
explanation for the relatively low bulk 5D of the
Zambales gas is the isotopic equilibration of CH4
and H2 with groundwater.
In summary, a direct mantle derivation for
the Zambales gas could be consistent with the
chemical and isotopic data if significant post-
mantle fractionation or isotopic exchange/con-
tamination at low temperatures is also invoked.
The high C H J 3 H e of the Zambales gas corn-
pared to the mantle end-members proposed by
others cannot be a result of addition of sedi-
mentary hydrocarbon to a mantle gas because
this would have had the effect of significantly
lowering the ~13Ccm -value. Reduction of an
EPR-type gas or a mantle gas with a slightly
higher than normal C/3He ratio could produce
a gas with the requisite CH4/3He ratio of the
Zambales gas. If both the C and He were de-
rived from the mantle, then the ratio of total
carbon to 3He in the mantle may be variable. If
the CH4-H2 gas was derived directly from the
mantle, reducing conditions (at or around IW)
would have been required in its mantle source
region.
3.1.2. Shallow crustal origin for the Zambales gas.
The growing number of reports of H2-CH4 gas
seeps in serpentinized mafic-ultramafic rocks
suggests a possible correlation between reduced
gas formation and serpentinization. Thermo-
chemical calculations (Olsen, 1963; Abrajano
and Pasteris, 1983; Abrajano, 1984; Apps, 1985;
Frost, 1985) have shown that very reducing
conditions (fo2 several log units below QFM)
could prevail during serpentinization in a rock-
dominated system (i.e. a system with limited
fluid recharge). The net reaction of interest may
be represented by:
15 olivine (Fogo) + 20.5H2 0
magnetite + 7.5 serpentine + 4.5 brucite + H2
(1)
At equilibrium, this reaction restricts oxygen
fugacities to values close to the iron-magnetite
fo2 buffer and within the stability field of some
native metal alloys (Moody, 1976; Abrajano,
1985; Frost, 1985). Attainment of such reduc-
ing conditions during serpentinization is evi-
denced by the common occurrence of native
metal alloy phases (e.g., Fe-Ni, Fe-Co ) and re-
duced sulfide assemblages in many serpentin-
ized terranes including the Zambales Ophiolite
(Hahn-Weinheimer and Rost, 1961; Kanehira

et al., 1964; Krishnarao, 1964; Chamberlain et
al., 1965; Ramdohr, 1967; Eckstrand, 1972;
Dick, 1974; Deutsch et al., 1977; Grudinin and
Sekerin, 1981; Abrajano and Pasteris, 1983;
Abrajano, 1984; Frost, 1985). Native iron, which
is indicative of even more reducing conditions
than the native metal alloys, has been found in
the Zambales Ophiolite (Abrajano and Pas-
teris, 1983). Serpentinization conditions could
be reducing enough for the saturation and exso-
lution from an aqueous phase of H2 gas with
very low water content (Abrajano, 1984; Frost,
1985). If a source of carbon also is available,
CH4 could exsolve with H2 as well. The ratio of
CH4 to H2 in the liquid and vapor phases would
be governed, in part, by the availability of car-
bon in the system.
The possible serpentinization temperatures
of the Zambales Ophiolite as estimated from a
variety of methods range from 30 ° to 350°C
(Abrajano, 1984; Sturchio et al., 1989). The up-
per temperature stability limit of the assem-
blage serpentine+brucite relative to olivine
(Fo93) is ~355°C at 1-2 kbar (Moody, 1976).
Equilibration temperatures calculated from co-
existing native metal and alloy phases in Zam-
bales range from ~100 ° to ~ 3 5 0 ° C whereas
sulfide assemblages appear to have equilibrated
at the lower end of this range (Abrajano, 1984).
Geothermometric estimates based on Wenner
and Taylor's (1973) serpentine-water oxygen
and hydrogen fractionation factors are consis-
tent with temperatures in the range of 30 ° C to
~ 250°C (Sturchio et al., 1989).
Temperatures calculated from ZIDcH4-H2 in
the Zambales gas using fractionation factors
from Richet et al. (1977) are l l 0 ° C (LFE-2)
and 125°C (LFE-3), consistent with the ser-
pentinization temperatures noted above. Even
if the JDcH4-H2 values used in these estimates
do not represent equilibrium fractionation, the
temperatures noted may still represent mini-
m u m formation temperatures. Equilibration
rates between H20 and H2 are rapid enough that
some isotopic reequilibration of H2 with H20
can be expected above 100°C (Truesdell and
217
Hulston, 1980; Kiyosu, 1983; Lyon and Huls-
ton, 1984). The rate of deuterium-hydrogen
exchange between CH4 and H20 is unknown,
but studies in geothermal systems indicate that
this may be slower than H2-H20 exchange (e.g.,
Truesdell and Hulston, 1980; Kiyosu, 1983;
Lyon and Hulston, 1984). Isotopic equilibra-
tion between H2 and H20 results in a decrease
in JD in H2 and an increase in 5D in H20 with
decreasing temperature. Hence, retrograde re-
equilibration of H2 and H20 could only lead to
a larger apparent ADcH4-R2, and consequently
a lower apparent equilibration temperature.
An important point which must be addressed
in any shallow derivation model for the Zam-
bales gas is the high 3He/4He ratio relative to
crustal He reservoirs. One possibility consis-
tent with a serpentinization origin for the Zam-
bales gas is that primitive He trapped in the
mantle-derived mafic-ultramafic rocks of the
Zambales Ophiolite was released to the fluid
phase during serpentinization. However, no
measurement of 3He/aHe ratio or the amount
of He trapped in the unaltered ultramafic rocks
in the Zambales Ophiolite has yet been made,
so it is necessary to use assumed values to quan-
tify this process. For example, it may be possi-
ble to reconcile the 3He/4He ratio observed in
the Zambales gas (4.1RA) with an assumed ini-
tial mantle-like value (7-8RA) through the lo-
cal accumulation of radiogenic 4He from decay
of U and Th. Assuming a U concentration of 5
ppb (Seitz and Hart, 1973; Tatsumoto, 1978),
a T h / U ratio of 3, and an initial 3He/aHe of
1.1-10 -5, then in 40 Myr. (approximate age of
the ophiolite) the 3He/4He ratio would de-
crease in a closed system by 33% or 83%, re-
spectively, for initial 4He concentrations of 10- v
or 10 -s cm 3 S T P g-1 (Ozima and Podosek,
1983 ). If the above assumptions are applicable,
then all the observed He could have been de-
rived from the ophiolite rocks. An alternative
possibility that the added 4He component could
be atmospheric He that was initially dissolved
in serpentinizing water at the recharge zones is
not consistent with the H e / N e ratio in the gas
218
( 100-130 ); this value is much higher than what
could be expected from the mixing proportions
between atmospheric and mantle He required
by the observed 3He/4He ratio.
The high (~13CcH 4 of the Zambales gas also
can be reconciled with a model of CH4 forma-
tion during serpentinization. Carbon contained
in the ultramafic rocks that underwent serpen-
tinization presumably has the requisite 513C.
Mantle carbon in mafic-ultramafic rocks can
take the form of solid phases (e.g., graphite,
magnesite) or fluid species (e.g., CO2, CO, CH4)
trapped as fluid inclusions. Graphite has been
identified in partly serpentinized Zambales ul-
tramafic rocks but its origin cannot be ascer-
tained. Indeed, some of the observed graphite is
almost certainly of low-temperature origin as it
occupies axial zones in lizardite-chrysotile mesh
structures. Because of the large isotopic frac-
tionation expected between CH4 and graphite
or between CH4 and CO2 or CO at serpentini-
zation temperatures, the conversion of "man-
tle" carbon from the ultramafic rocks (assumed
to have an initial ~flaC of -7%o ) to CH4 in the
gas phase would need to have been virtually
quantitative. Alternatively, mixing with a high-
613C carbon component ( > -7%c~ ) initially
dissolved in the serpentinizing fluid could make
the requirement of quantitative conversion of
intrinsic carbon to CH4 unnecessary. Dissolved
carbon as organic compounds, carbonate or bi-
carbonate from soil is probably too light isotop-
ically to fulfill this role, so that one may have
to further postulate involvement of sedimen-
tary marine carbonate. It is even possible for
dissolved limestone (513C = - 2 to + 2 %o ) to be
the sole source of carbon in the CH4 gas if cal-
cite precipitation from the fluid also occurs prior
to formation of CH4. The relatively high pH
( > 9) that generally prevails in groundwaters
in ultramafic terranes, as well as the increased
temperatures expected at greater depths, could
result in the precipitation of carbonates in de-
scending recharge water. Although the closest
exposures of calcareous sedimentary units are
~ 15 km to the west of the gas seeps and none
are known to be present immediately beneath
them, coupled dehydration-decarbonation re-
actions within the accreted sediments (Fig. 1 )
may be invoked as a possible source of dissolved
carbon.
We noted on p. 217 that the CH4/H2 ratio in
the gas phase produced during serpentinization
will depend on the availability of carbon. How
much carbon is required to produce a gas with
CHa/H2 > 1? Reaction (1) can be modified to
examine the effect of carbon on the CH4/He ra-
tio of the product gas as follows:
15 olivine (Fog0) +20.5H20+xC
magnetite + 7.5 serpentine
+4.5 brucite+xCH4 + (1 - 2x)H2 (2a)
or
15 olivine (Fo9o) + (20.5- 2x ) H2 O + xCO2
magnetite + 7.5 serpentine
+4.5 brucite + xCH4 + ( 1 - 4 x ) H 2 (2b)
For a gas with a C H J H 2 ratio of 1 to form from
reactions (2a) or (2b), the number of moles of
carbon that reacts to form CHt should equal the
number of moles He produced by the accom-
panying serpentinization reaction [x = 0.33 in
(2a), x=0.20 in (2b) ]. If all the carbon that
formed the CH4 component of the Zambales gas
was carried by the serpentinizing water as dis-
solved CO2, then this water would have to have
contained ~ 1 mole% total dissolved COz [re-
action (2b) ]. This amount of dissolved COz was
estimated assuming that the Zambales gas was
formed as a single reaction batch where only
the CO2 dissolved in the amount of water re-
quired by reaction (2b) was converted to CH4.
If dissolved C02 from a larger reservoir of water
than that required by reaction (2b) could be
converted to CH4, a lower concentration of dis-
solved CO2 is necessary. Hence it is conceivable
that the serpentinizing solution could have
supplied all the carbon that formed CH4. If all
CH4 carbon originated as elemental carbon in
the ultramafic rock, ~ 1800 ppm C must have
been contained initially in the rock [reaction

(2a) ]. This estimate requires the assumptions
that the Zambales gas was produced as a single
reaction batch and that only the amount of ol-
ivine specified by reaction (2a) released the
carbon that formed CH4. If carbon from other
sources in the rock (e.g., fluid inclusion, graph-
ite, etc. ) could preferentially contribute to CH4
formation, <1800 p p m of intrinsic carbon
would be required to produce gas with a CH4/
H2 ratio of 1. It is also possible that all CH4 car-
bon in the gas could have originated from the
rock if H2 produced by the serpentinization re-
action can migrate and reduce intrinsic carbon
in a larger volume of ultramafic rock.
In summary, the serpentinization model
stipulates that water of crustal origin reacts with
mafic-ultramafic rocks and the low [o~ result-
ing from the hydration and oxidation reactions
leads to the stabilization of reduced gas species.
Any carbon present during these reactions (i.e.
intrinsic carbon in the rocks or dissolved car-
bon in solution) is thus converted to CH4. The
CH4/H2 ratio in the Zambales gas can be rec-
onciled with derivation of CH4 solely from either
the serpentinizing water or ultramafic rock.
Helium and carbon isotopic signatures of the
Zambales gas are consistent with gas deriva-
tion, at least in part, from trapped mantle com-
ponents in the mafic-ultramafic rocks.
Independent mineralogical and isotopic esti-
mates of serpentinization temperatures are
consistent with the CH4-H2 isotopic equilibra-
tion temperatures of the Zambales gas, further
corroborating a low-temperature origin.
4. Conclusions
We conclude from the foregoing discussion
that a direct mantle origin and a serpentiniza-
tion origin both can be made compatible with
the chemical and isotopic composition of the
Zambales gas. The He isotope data appear to
preclude total derivation of the He in the Zam-
bales gas from the mantle without addition of
radiogenic 4He. The J13C-value for CH4 is un-
usually high, suggesting that CH4 carbon was
219
derived from either a mantle reservoir or a sed-
imentary carbonate reservoir, or some mixture
of the two. Mantle-derived carbon could be
either present-day mantle CH4 emanation or
CH4 produced from mantle carbon initially
trapped in the ultramafic rocks. A sedimentary
marine carbonate source of carbon requires a
precise balance of carbonate dissolution and
precipitation to produce the appropriate value
of • 1 3 C c H 4 . The hydrogen isotope data indicate
a minimum equilibration temperature of 110-
125°C for CH4 and H2, consistent with other
independent estimates of serpentinization
temperatures. The bulk JD-value ( - 355%o ) of
the gas is too low to represent an unmodified
mantle value. If a substantial fraction of the CH4
in the Zambales gas was derived directly from
the mantle, its occurrence in an obduction en-
vironment with no obvious relation to present-
day volcanism is important and should be in-
vestigated further. If CH4 in the gas seep rep-
resents the product of a crustal serpentinization
process, it will be of interest to find out the ex-
tent to which this mode of CH4 production oc-
curs worldwide. Isotopic studies of similar
reduced gas occurrences are highly recom-
mended.
Acknowledgements
Work supported by the U.S. Department of
Energy, Office of Basic Energy Sciences, under
contract W-31-109-Eng-38, and through Na-
tional Science Foundation grant EAR-8313681
(J.D. Pasteris, Washington University). Sup-
port for the He isotope measurements was pro-
vided by a grant from the Mantle Geochemistry
Program, Earth Sciences Division of the Na-
tional Science Foundation to H. Craig and R.
Poreda, Isotope Laboratory, S.I.O. The Phil-
ippine Bureau of Mines and Geosciences pro-
vided field logistical support. T. Engelkemeir is
thanked for performing analyses of bulk gas
composition. P.D. Jenden and H. Wakita pro-
vided preprints of their papers and R. Coveney,
Z. Sofer and M. Schoell reviewed the manu-
220
script. D. Abrajano and C. Wesolowski are
thanked for assistance in drafting and manu-
script preparation.
References
Abrajano, T.A., 1984. Igneous petrology and low tempera-
ture geochemistry of the Acoje massif, Zambales
Ophiolite, Philippines: Mineralogical, isotopic, and
mixed-volatile equilibria. Ph.D. Dissertation, Washing-
ton University, St. Louis, Mo., 475 pp.
Abrajano, T.A. and Pasteris, J.D., 1983. The origin of na-
tive iron and iron alloys in serpentines (Trans. Am.
Geophys. Union), 64:886 (abstract).
Apps, J.A., 1985. Methane formation during hydrolysis of
mafic rocks. Lawrence Berkeley Lab., Berkeley, Calif.,
Univ. of Califi, Annu. Pep., 84: 13-17.
Arculus, R.J. and Delano, J.W., 1981. Intrinsic oxygen fu-
gacity measurements: techniques and results for spinels
from upper mantle peridotites and megacryst assem-
blages. Geochim. Cosmochim. Acta, 45: 899-913.
Barnes, I., Kistler, R.W., Mariner, R.H. and Presser, T.S.,
1981. Geochemical evidence on the nature of the base-
ment rocks of the Sierra Nevada, California. U.S. Geol.
Surv., Water-Supply Pap. No. 2181.
Betechtin, A.G., 1961. Mikroskopische Untersuchungen im
Platinerzen aus dem Ural. Neues Jahrb. Mineral., 97: 1-
34.
Buntin, T.J., 1983. An oxygen, iron, and carbon geochem-
ical study of selected Merensky Reef potholes and their
possible role in Merensky Reef petrogenesis. M.A. The-
sis, Temple University, Philadelphia, Pa., 224 pp.
Chamberlain, J.A., McLeod, C.R., Traill, R.J. and Lach-
ance, G.R., 1965. Native metals in the Muskox intru-
sion. Can. J. Earth Sci., 2: 188-215.
Coveney, R.M., Goebel, E.D., Zeller, E.J., Dreschhoff,
G.A.M. and Angino, E.E., 1987. Serpentinization and
the origin of hydrogen gas in Kansas. Bull. Am. Assoc.
Pet. Geol., 71: 39-48.
Craig, H. and Horibe, Y., 1988. Helium isotopes and meth-
ane in Japanese volcanoes and natural gases. (Unpub-
lished manuscript. )
Craig, H., Horibe, Y., Farley, K.A., Welhan, J.A., Kim, K.R.
and Hey, R.N., 1987. Hydrothermal vents in the Mar-
iana Trough: results of the first ALVIN dives. Eos
{Trans. Am. Geophys. Union), 68:1531 (abstract).
Deines, P., 1980. The carbon isotopic composition of dia-
monds: relationship to diamond shape, color, occur-
rence, and vapor composition. Geochim. Cosmochim.
Acta, 44: 943-961.
Deutsch, E.R., Rao, K.V., Laurent, R. and Sequin, M.K.,
1977. New evidence and possible origin of native iron in
ophiolites of eastern Canada. Nature (London), 269:
684-685.
Dick, H.J.B., 1974. Terrestrial nickel-iron from the Jose-
phine Peridotite, its geologic occurrences, association,
and origin. Earth Planet. Sci. Lett., 24: 291-298.
Eckstrand, O.R., 1972. The Dumont serpentinite: a model
for control of nickeliferous opaque mineral assemblage
by alteration reactions in ultramafic rocks. Econ. Geol.,
70: 183-201.
Eggler, D.H. and Baker, D.R,, 1982. Reduced volatiles in
the system C-O-H: implications to mantle melting, fluid
formation and diamond genesis. In: S. Akimoto and
M.H. Manghnani (Editors), High Pressure Research in
Geophysics. D. Reidel, Boston, Mass., pp. 237-250.
Fish, F., 1987. Siljan drilling experiment: scientific results.
Eos (Trans Am. Geophys. Union ), 68:468 (abstract).
French, B.M., 1966. Some geological implications of equi-
librium between graphite and C - H - O gas phase at high
temperatures and pressures. Rev. Geophys., 4: 223-253.
Fritz, P., Frape, S.K. and Miles, M., 1987. Methane in the
crystalline rocks of the Canadian Shield. In: P. Fritz
and S.K. Frape (Editors), Water and Gases in Crystal-
line Rocks. Geol. Assoc. Can., Spec. Vol., 33: 211-223.
Frost, B.R., 1985. On the stability of sulfides, oxides, and
native metals in serpentinite. J. Petrol., 26: 31-63.
Gephart, R.E., Leonard, L.S., Spane, F.A. and Strait, S.R.,
1979. Pasco Basin hydrology. In: Hydrogeological Stud-
ies Within the Columbia Plateau, Washington: An In-
tegration of Current Knowledge, Ch. 3, Rockwell
Hanford Operations, Richland, Wash., RHO-BWI-ST-
S.
Giardini, A.A., Melton, C.E. and Mitchell, R.S., 1982. The
nature of the upper 400 km of the earth and its potential
as the source for non-biogenic Petroleum. J. Petrol. Geol.,
5: 173-190.
Gold, T. and Soter, S., 1980. The deep-earth gas hypothe-
sis. Sci. Am., 242: 154-170.
Green, D.H., Falloon, T.J. and Taylor, W.R., 1987. Mantle-
derived magmas-roles of variable source peridotite and
variable C - H - O fluid composition. In: B.E. Mysen
(Editor), Magraatic Processes: Physicochemical Prin-
ciples. Geochem. Soc., Washington, D.C., Spec. Publ.
No. 1, pp. 139-154.
Grudinin, M.I. and Sekerin, A.P., 1981. Native iron in ul-
tramafic rocks of the Shamanskiy pluton. Dokl. Akad.
Nauk S.S.S.R., 245: 160-162.
Hahn-Weinheimer, P. and Rost, F., 1961. Akzessorische
Mineralien und Elemente im Serpentinit von Leupolds-
grtin (Mtinchberger Gneismasse). Geochim. Cosmo-
chim. Acta, 21: 165-181.
H~kinian, R., Chaigneau, M. and Cheminde, J.L., 1973.
Popping rocks and lava tubes from the Mid-Atlantic rift
valley at 36°N. Nature (London), 245: 371-373.
Holloway, J.R., 1984. Graphite-CH4-H20-C02 equilibria
at low-grade metamorphic conditions. Geology, 12: 455--
458.
Horibe, Y., Kim, K.R. and Craig, H., 1987. Hydrothermal
methane plumes in the Mariana back-arc spreading
center. Nature (London), 324: 131-133.

Hulston, J.R., Taylor, C.B., Lyon, G.L., Stewart, M.K. and
Cox, M.A., 1981. Environmental isotopes in New Zea-
land hydrology, 2. Standards, measurement techniques,
and reporting of oxygen-18, deuterium and tritium in
water. N.Z.J. Sci., 24: 313-322.
Hunt, J.M., 1979. Petroleum Geochemistry and Geology.
W.H. Freeman, San Francisco, Calif., 617 pp.
Javoy, M., Pineau, F. and All~gre, C.J., 1982. Carbon geo-
dynamic cycle. Nature (London), 300:171-173.
Jeffrey, A.W.A. and Kaplan, I.R., 1988. Hydrocarbons and
inorganic gases in the Gravberg-i well, Siljan Ring,
Sweden. In: M. SchoeU (Guest-Editor), Origins of
Methane in the Earth. Chem. Geol., 71:237-255 (this
special issue).
Jenden, P.D., Kaplan, I.R., Poreda, R.J. and Craig, H., 1988.
Origin of nitrogen-rich natural gases in the Great Val-
ley, California: Evidence from stable isotope ratios of
helium, carbon and nitrogen. Geochim. Cosmochim.
Acta, 52: 851-862.
Kanehira, K., Banno, S. and Hasimoto, M., 1964. Notes on
rock-forming minerals (28). Finding of awaruite (na-
tive nickel iron) from serpentinite near the city of Koti,
Sikoku. J. Geol. Soc. Jpn., 70: 272-277.
Karshavin, V.K. and Vendillo, V.P., 1970. Thermodynamic
equilibrium and conditions for existence of hydrocar-
bon gases in a magmatic process. Geochem. Int.,7: 797-
803.
Kiyosu, Y., 1983. Hydrogen isotopic compositions of hy-
drogen and methane from some volcanic areas in north-
eastern Japan. Earth Planet. Sci. Lett., 62: 41-52.
Krishnarao, J.S,R., 1964. Native nickel-iron alloy, its mode
of occurrence, distribution, and origin. Econ. Geol., 59:
443-448.
Kyser, T.K., 1986. Stable isotope variations in the mantle.
In: J.W. Valley, H.P. Taylor and J.R. O'Neil {Editors),
Stable Isotopes in High Temperature Geological Pro-
cesses. Mineral. Soc. Am., Washington, D.C., pp. 140-
164.
Lupton, J.E. and Craig, H., 1975. Excess 3He in oceanic
basaltsevidence for terrestrialprimordial helium. Earth
Planet. Sci. Lett.,26: 133-139.
Lyon, G.L. and Cox, M.A., 1978. Stable isotope analysis of
hydrogen-methane mixtures. N.Z. Inst.Nucl. Sci.,Rep.
INS-R-254.
Lyon, G.L. and Hulston, J.R., 1984. Carbon and hydrogen
isotope compositions of N e w Zealand geothermal gases.
Geochim. Cosmochim. Acta, 48: 1161-1172.
Marty, B. and Jambon, A., 1987. C/3He in volatilefluxes
from the solid Earth: implications for carbon geodyn-
amics. Earth Planet. Sci. Lett.,83: 16-26.
Moody, J.B., 1976. An experimental study of the serpentin-
ization of iron-bearing olivines.Can. Mineral., 14: 462-
478.
Nagao, K., Takaoka, K. and Matsubayashi, 0., 1981. Rare
gas isotopic compositions in natural gases in Japan.
Earth Planet. Sci. Lett.,53: 175-188.
221
Neal, C. and Stanger, G., 1983. Hydrogen generation from
mantle source rocks in Oman. Earth Planet. Sci. Lett.,
66: 315-320.
Olsen, E., 1963. Equilibrium calculationsin the system Mg,
Fe, Si, O, H and Ni. Am. J. Sci.,261: 943-956.
Ozima, M. and Podosek, F.A., 1983. Noble Gas Geochem-
istry.Cambridge University Press, Cambridge, 367 pp.
Petersil'ye,I.A.and Pripachkin, V.A., 1979. Hydrogen, car-
bon, nitrogen,and He in gases from igneous rocks. Geo-
khimiya, 7: 50-55.
Petersil'ye, I.A., Koslov, Y.K., Belyayev, K.D., Sholokh-
nev, V.V. and Dokuchayeva, V.S., 1970. Nitrogen and
hydrocarbon gases in ultramafic rocks of the Sopcha
stock of the Monchegorsk pluton, Kola Peninsula. Dokl.
Akad. Nauk S.S.S.R., 194: 200-203.
Pineau, F.,Javoy, M. and Bottinga, Y., 1976. I~C/~2C ratios
of rocks and inclusions in popping rocks of the Mid-
Atlantic Ridge. Earth Planet. Sci. Lett.,29: 413-421.
Poreda, R.J.,Jenden, P.D.,Kaplan, I.R.and Craig,H., 1986.
Mantle He in Sacramento basin natural gas well.Geo-
chim. Cosmochim. Acta, 50: 2847-2853.
Ramdohr, P., 1967. A widespread mineral associationcon-
nected with serpentinization. Neues Jahrb., Mineral
Abh., 107: 241-265.
Richet, P., Bottinga, Y. and Javoy, M., 1977. A review of
hydrogen, carbon, nitrogen, oxygen, sulphur and chlo-
rine stable isotope fractionation among gaseous mole-
cules.Annu. Rev. Earth Planet. Sci.,5: 65-110.
Rust, F.E. and Stevens, C.M., 1980. Carbon kinetic iso-
topic effectin the oxidation of methane in hydroxyl. Int.
J. Chem. Kinetics, 12: 371-377.
Schoell, M., 1983. Genetic characterization of natural gases.
Bull. Am. Assoc. Pet. Geol., 67: 2225-2238.
Seitz, M.G. and Hart, S.R., 1973. Uranium and boron dis-
tributions in some oceanic ultramafic rocks. Earth
Planet. Sci. Lett., 21: 97-107.
Sofer, Z., 1985. An unusual occurrence of light hydrocarbon
gases in a well offshore Northeast Palawan Island, the
Philippines. Precambrian Res., 30:179-188.
Stoll, W.C., 1959. Natural fires in South Lawis river valley,
Zambales, P.I. Philipp. Geol., 7: 167-170.
Sturchio, N.C., Abrajano, T.A., Murowchick, J. and Mueh-
lenbachs, K., 1989. Serpentinization of the Acoje mas-
sif, Zambales ophiolite, Philippines: hydrogen and
oxygen isotope geochemistry. Tectonophysics (in press).
Tatsumoto, M., 1978. Isotopic compositions of lead in
oceanic basalt and its implication to mantle evolution.
Earth Planet. Sci. Lett., 38: 63-87.
Thayer, T.P., 1966. Serpentinization considered as a con-
stant volume metasomatic process. Am. Mineral., 51:
685-710.
Truesdell, A.H. and Hulston, J.R., 1980. Isotopic evidence
on environments of geothermal systems. In: P. Fritz and
J.Ch. Fontes {Editors), Handbook of Environmental
Isotope Geochemistry, Vol. 1, The Terrestrial Environ-
ment, A. Elsevier, Amsterdam, pp. 179-226.
222
Wakita, H. and Sano, Y., 1983.3He/4He ratios in methane-
rich gases suggest magraatic origin. Nature (London),
305: 792-794.
Wakita, H., Sano, Y., Urabe, A. and Nakamura, U., 1988.
Origin of methane-rich natural gas in Japan: Formation
of gas field due to large scale submarine volcanisms.
{Unpublished manuscript.)
Welhan, J.A. and Craig, H., 1979. Methane and hydrogen
in East Pacific Rise hydrothermal fluids. Geophys. Res.
Lett., 6: 829-831.
Welhan, J.A. and Craig, H., 1982. Abiogenic methane in
mid-ocean ridge hydrothermal fluids. In: W.J. GwiUiam
(Editor), Deep Source Gas Workshop, Technical Pro-
ceedings. U.S. Dep. of Energy, Morgantown, W.Va., pp.
122-129.
Wenner, D.B. and Taylor, H.P., 1973. Oxygen and hydro-
gen isotope studies of serpentinization of ultramafic
rocks in oceanic environments and continental ophiol-
ite complexes. Am. J. Sci.,273: 207-239.
Whiticar, M.J., Faber, E. and Schoell, M., 1986. Biogenic
methane formation in marine and fresh-water environ-
ments: C02 reduction vs. acetate fermentation. Geo-
chim. Cosmochim. Acta, 50: 693-709.
Woermann, E. and Rosenhauer, M., 1985. Fluid phases and
the redox state of the Earth's mantle: extrapolations
based on experimental, phase-theoretical and petrolog-
icaldata. Fortschr. Mineral., 63: 263-349.
Wood, B.L., 1970. Metamorphosed ultramafites and asso-
ciated formations near Milford Sound, N e w Zealand.
N.Z.J. Geol. Geophys., 15: 88-128.