CHLORIDES IN OIL-FIELD WATERS 687

Chlorides in Oil-Field Waters

BY C. W. WASHBURNE, WASHINGTON, D, C.

(New York Meeting, February, 1914)

THE waters of many oil fields have been regarded as burieq sea water
which has been retained in the sediments since the time of theirdeposi-
tion. The preservation of connate water through geological time has
seemed improbable to only a few geologists, but there is room for doubt
that buried sea water could remain in the strata during their periods of
deformation and during the many subsequent epochs of the circulation
of meteoric ground-water. Some have even suggested that the calcium
chloride water in the deep mines of the Lake Superior region is ancient
sea water buried in the Algonkian lavas, but it is hard to understand
how such water could be derived from the highly sodic and sulphatic
water of the sea. It would seem more plausible to connect the calcium
chloride with the highly calcic magmas of the greenstones in which it is
found.
Chemical analyses of waters associated,with oil differ widely from
the composition of sea water, requiring extensive alteration of the latter,
if the former is truly its derivative. The first notable difference is the
general absence of sulphate~ from oil-field waters, but this has be.en
explained satisfactorily through reduction by hydrocarbons and organic
matter. Oil reduces sulphates with the production of hydrogen sulphide
(or sulphur), water, and carbonates (or carbon dioxide).
The second striking difference is in the high ratio of chlorine to sodium.
An examination of the analyses of chloride waters from oilfields shows
that they contain a large relative excess of chlorine over that in sea.water. -
Moreover, there is difficulty in finding chemical reactions that areprobable
in nature, by which sea water could be converted into anything like the
chloride waters of oil fields. In the latter a large part of the chlorine
occurs as calcium and magnesium chloride. This condition might
be produced by the removal of sodium carbonate, which, however,
cannot be precipitated in the presence of much calcium or magnesium.
 688 CHLORIDES INOIL-FIEI,D WATERS

Composition of Chloride Wat.ers from Oil Fields

8e.a Water Pennsylvania Kansas.

A B C D E

C1. ............... . 55.29a 62.31a 161.80b 49,285c 30,066. Dc

Br ............... . 0.19· 0.53 0.70 79 ........ -, ...
l. ................ . 0.01 ......... . 8.4 ............
80 4 •••••••••••••.• 7.69 0.03 0.05 40 ............
C0 3 • • • • • • • • • • • • • • • 0.21 0.27 0.00 a ............
B 40 7 •••••••••••••• ......... . 150 .. .........
I} 4,216
;.

Na.: .............. . 30.59 18.35 64.55 } 2,791.

K ................ . 1.11 1.55 5.16
Li. ............... . 0.04
NH 4 • • • • • • • • • • • • • •
Ca ............... . 1.20
0.23
13.86
:::::::::
25.19
:1:::::::::::
1,425
:[::::::::::::
2,792
Ba ......................... . trace
Sr ......................... .. 3.55
Mg............... 3.72 2.53 2.48 2,844 6,310
AhO, .............. )......... . 0.02. 0.00
Fe 20a ............ ··1· ........ · 48
Fe ......................... . .. .. <>: iB"} 56 ;}
SiOz........................ . 0.00 11 _ _ 43_1_,~_85_
100.00 100.00 263.64 II 56,721 1 43,517
Ratio ~~.......... 1.8 3.5 2.5 .' 18.2,± I 7. 5 :I:

a. Percentage of total salts.

b. Grams per kilogram of water.
c. Parts per million (?) .
A. Dittmar's summary of the principal salts in sea water, from .Clarke, Data of
Geochemistry, Bulletin No. 491, U.S. Geologicai Sy,Tvey, p.1l3, 1911. Placed here for
comparison.
B ..Brine from well 2,667 ft. deep at Conneautsville, Pa. Salinity about nine times
that of sea water. Analysis by Robinson and Mabery, quoted by Clarke, loco cit.,
p.174.
C. Brine from depth of 6,300 ft. in well near Imperi\11, Allegheny county, Pa.
Analysis by George Steiger, Journal of the Washington Academy of Sciences, vol. iii,
- p. 423, i913. Well record described by White, Bulletin of the Geological Society of
America, vol. xxiv, No.2, pp. 275-282, June, 1913. The salinity is eight times that
of sea water.
D. lIudson well, Fredonia, Wilson county, Kan. Depth 400 ft. Analysis by
Bailey and Davies, quoted in Water Sy,pply PaperNo. 273, U. S Geological Sy,rvey,
p. 199 (1911). .
E. Flowing salt well in Pennsylvanian strata at Lawrence, Kan., depth 1,400 ft.
Analysis by Bartow and Thompson, Kansas GeologicalSurvey, vol. vii, p. 151 (1902).
 CHLORIDES IN OIL-F1ELD WATERS 689
The folloWing American analyses were gathered by Hofer, l the ··first
four (A-D) being from oil wells near Tarentum, Pa. (analyses by
Charles Lenny), and the last two (E-F) from oil and gas wells in Ohio
(analyses"by W. J. Root). The high percentage of calcium and mag-
nesium chlorides is noteworthy.

A B' C D E F

NsCl..... . . 3,2000 5.0203 4.7152 32.978 \ 71.645 30.128

NaI .............................. -:- ....... '" . " . . . . 0.013 0.024
KCl........ ........... .......... .......... 0.041 .... , ............... .
NH 4Cl. . . ... .......... .......... .......... 0.007 ......... '. . ........ .
CaC1 2 .. .. .. • 0.5587 1. 6462· 0.9657 8.578 18.665 48.969
CaI 2 • • • • • • • • • • • • • • • . • • • • • • • • • • • • • • : . . . . . . . . 0.080 ................... .
CaCO....... .......... ....... .... .......... 2.764 ................... .
MgBr2 ................ \................... , O~ 1154 0.271 0.592

:::~: ~~~:J:~~: ~~~: :

MgCl....... 0.2933 0.3193 0.3132 2.166
MgCO a....... , ...... :. ; ..... :, .......... ,. 1.377.
BaCl.. . . . . . . 0.0380 0.0642 0.0759 0.006
BaCO' ................................. '.. : ·0.004 .................. ..
SrCO, .................................... \ 0.059 I........ : ......... ..
A1.Oo, .... ·} ........... ..... ....... .......... .......... 0.096 0.054
F e. ,..... .
F.eCO, ..' .. ·.. 1 . . . . . . . . . . . . . . . . : . . . . . . . . . . . . . . 0.038 ................ ~ ...
S10,........ .......... .......... .......... 0.102 0.024 0.032
CO,. . . . . . .. .......... . ......... [. . . . . . . . . . 1.953 . . . . . . . . .. . ........ .
CO ............... ~.......... .......... 0.008 .................. ..

Two working hypotheses may be entertained t6account for the great

excess of chlorine over sodium in the water of oil fields. By the first
hypothesis, we may explain the disappearance of sodium by the pre-
cipitation of sodium chloride from ancient sea water after its burial in
the sediments, assuming that the Paleozoic ocean contained less sodium
than the modern seas. . It is improbable that the average water originally
buried in the sediments of oil fields could have been a concentrated
brine from which much sodium chloride had been previously precipitated.
But the dryness of deep sands in oil fields, and the doWnward increase
in salinity, suggest that beneath the zone of ground water there is a
deeper dry zone from which water has been extracted, possibly threugh
the drying influence of the ascending rock gases, which consist mainly of
nitrogen, carbon dioxide, and methane. This process would precipitate
sodium chloride on the surfaces of deep rock pores, leaving a residual
water relatively rich in the soluble magnesium and calcium chlorides,

1 Das Erdol, voL H,P!' I, p. 29 (1909). Gives 25 analyses.

VOL. XLVIII.-44
690 CHLORIDES IN OIL-FIELD W A'I'ERS

which would be driven 'slowly upward by the rising gas. As the water
passed through b~dies of shale the adsorption of sodium chloride on clay
particles, mentioned by .Johnston, 2 might further increase the relative
amount of magnesium and calcium chloride, but'since potassium salts
are adsorbed on clay more than sodium salts 3 the relatively high ratio
of potassium to sodium in oil-field waters indicates that adsorption has
played a minor role in their alteration.
The second and stronger working hypothesis is that the excess
chlorine has been added directly as magnesium and calcium chlorides
in solutions rising from below. These substances are characteristically
high in oil-field waters. Furthermore, it appears that the waters of
the Ohio oil fields are converting limestone into dolomite, indicating
thai the water which accompanied the oil in its invasion of the Trenton
limestone was rich in magnesium, as it still remains. Bownocker states
that the percentage of magnesium in the limestone increases as one ap-
proaches an oil pool, reaching a maximum in the pool itself. Some have
suggested that this is merely a question of the relative porosity of dolomite
and limestone, but the factor of porosity, although very important,
does not explain the whole question. It fails especially to correlate
the phenomena with the characteristically high percentage of magnesium
in the waters of oil fields, and with the occurrences of secondary '-dolomite
in the coastal fields of Louisiana, Texas, and Mexico. The lenses of
dolQmite or of dolomitic limestone which cap many of the salt cores of
the Gulf coast are clearly secondary deposits which grew within the
strata. They were precipitated from the same solutions that· deposited
the sodium chloride, or at least from solutions which followed the channel
made by the plug or core of rock salt. The localized occurrence of
dolomite, which is so intimately associated with oil near the coast of
the -Gulf of Mexico, and in Ohio, Indiana, and Canada, is therefore re-
garded as the direct or indirect product of solutions rich in magnesium
chloride. The bitter brine found in the latter fields is thought to have
produced the local dolomitization of the Trenton limestone.
The sourc~ of the magnesium and calcium chloride as well as of
much sodium chloride may be found theoretically in the emanations
from deep-seated basic rocks rich in magnesium, such as the olivine-
basalt which is so intimately tied with the occurrence of oil in Mexico,
and the peridotite in the dikes of western Kentucky, or the serpentine
at~Byracuse, N. Y., which are the only igneous rocks known near the
northern oil region. Moreover,. it is rock of this general character
which contains diamonds in South Africa, Arkansas, and other places.
2 Discussion of paper by J. A. L. Clerc and J. F. Brezeale, Journal of the Wash-
ington Academy of Sciences, voL iii, p. 198 (1913). ,
$Clarke: Data of Geochemistry, Bulletin No. 491, U. S. Geological 'S'urvey, p. 200
(1911).
 CHLORIDES IN OIL-FIELD WATERS 691

Stelznet, Bonney and others have ,proved that the diamonds of the
Kimberley district crystallized as an integral part of the magma and
that the carbon is of deep~seated origin. In Cuba there is a remarkable
series of occurrences of light oil in serpentine. In similar rock Diller
has found metallic iron which probably contains. carbides. Magriesium-
rich magmas therefore are the most probable source of oil (if it be
volcanic) and of the accompanying chloride solutions.
The acceptance of the idea that modified volcanic emanations have
entered the strata of oil fields simplifies many problems besides that of
the relative excess of chlorine. For instance, it furnishes an adequate
cause for the localization of the great salt-cores of the Texas, Louisiana,
and Chiapas. Harris 4 finds that the salt domes of Texas and Louisiana
. have a linear arrangement suggestive of long intersecting fissures.
He finds that the amount of uplift of the strata is entirely inadequate
to a.ccount for the amount of space occupied by the salt plugs, some of
which have been penetrated by the drill nearly 3,000 ft., without reaching
bottom. He concludes, therefore, that the sart cores are not laccolithic
or plug-like intrusions into the sediments, squeezed up from a great
hypothetical salt bed in some lower formation, but rather that they
have grown by crystallization at the places where they now occur and
have not undergone much deformation. In other words, they are great
cylindrical concretions of salt 1,000 ft. or more across and over 3,000
ft. high. He believes that the salt and the associated hydrocarbons
were gathered by meteoric waters which percolated through the ~edi­
mentary strata and rose along the intersections of fissures, where the
salt was precipitated because of the decrease in temperature and pres-
sure. The decrease in pressure would cause but a negligible precipita-
tion, practically nothing. Temperature is somewhat more effective,.
but Lindgren 5 says: "As the solubility of salt increases ori.Iy slightly
with increase of temperature (35.69 per cent. at 10° C.; 39.12 per cent.
at 100°; 44.90 per cent. at 180°), only the increment could have been
precipitated as the temperature of the ascending current was lowered,
and hence the quaritity of primary salt required by this hypothesis is
incredibly large." Let us assume that.· the solutions cooled as fast as
the normal underground heat gradient, or about 21°C.· in ascending
2,000 ft. This would precipitate about 2 per cent. of the total salt in a

4 Bulletin No. 429, U. S. Geological Survey (1910). The alignment of the salt domes
was first observed by Cl!-ptl!-il1. A. F. Lucas: .Rock-Salt in Louisiana, Trans., xxix, 463,
l"ig. 1 (1899). Captain LUQas was also the first to observe the geological relations of
the salt cores with reference to the occurrence of oil, which he proved by drilling the
famous "Lucas gusher" at"Spindletop. CI. Robert T. Hill: Journal of the, Franklin
Institute, vol. cliv, pp. 143, 225, 263 (Aug.-Oct., 1902). A. F. Lucas: Science, vol.
xxxv, pp. 96h4 (June 21,1912). Economic Geology, vol. vi, No.4, p. 380 (June, 1911).
6 Mineral Deposits, p. 288(1913).
692 CHLORIDES INOIL-FIEJ.,DW ATER:S

solution saturated at JOO°. In other words, the salt cores would rep-
resent only about 2 per cent. of the total amount 'Of sodium chloride
which had risen in the fissures, the rest having been carried beyond the
top of the cores and lost: .The salt cores in this country and Mexico
number several hundred and their total volume is many cubic miles ..
The theory of Harris,· if not modified, requires a supply of roughly 50
ti.mes as many cubic· miles of salt.
Yet the" concretionary" theory of Harris has much in its favor,
since it meets the mechanical requirements of the problem better than
any other. The chemical difficulty can be met if the salt domes have
been the loci of the escape of solutions carrying a common ion, either
of sodium or of chlorine. Analyses of some volcanic waters show an
abundance of sodium chloride, and others of sodium sulphate or car-
bonate, but these do not meet the requirements of the present problem.
From the character of the water and from the presence of secondary
lenses of dolomite one may infer that the precipitation was produced
by the intermingling of concentrated salt solutions with brines rich in
magnesium and calcium chloride. The former were probably derived
mainly from the sedimentary strata as suggested by Harris, but the
latter probably rose from underlying plugs of olivine basalt which failed
to reach the surface. .
Plugs and dikes of this type of rock are common in the coastal plain
of Mexico, where they have about the same relation to the occurrence
of oil as do the salt cores in Texas to the north or in Chiapas to the
south of them. The chlorine probably escaped from the magma as
free chlorine, as hydrochloric acid, and as chlorides of ammonium,
calcium, magnesium, and sodium. If the igneous plugs are not too
deeply buried beneath the plains it is probable that large quantities
of hydrochloric acid passed up through the overlying fractures made
by the intrusion of the plug. This is the most effective precipitant of
common salt, and in the case of the fields about the Gulf of Mexico,
the evidence of volcanic action is so strong as to suggest that hydro-
chloric acid may have played an important role in the precipitation of
the salt above the voI.canic plugs. Also magnesium and calcium chlo-
rides would be quite effeQtive to precipitate rock salt, and if the ascending
igneous emanations passed through many strata before reaching the
horizon of the salt cores it is certain that the hydrochloric acid would be
all converted into chlorides, largely into the chlorides of calcium and
magnesium. For our present argument it is immaterial whether the
chlorine ions were attached mainly to hydrogen,to calcium, or to mag-
nesium, Acco!ding to Nernst's law, any of these probable. chlorides
would precipitate sodium chloride from a concentrated solution. In
this way maSses of salt derived from the sediments would be formed
 CHLORIDES IN OIL-FIELD WATERS 693

where the sedimentary solutions mixed with the rising solutions of

volcanic chlorides, i.e., directly above thedeeply buried plugs.
Thus the volcanic hypothesis offers a simple explanat~on of the salt
and dolomite problem of oil fields, and the relative excess of chlorine
in the associated water may be regarded as a possible indication of the
entrance or volcanic emanations, which in most cases have been greatly
modified by their long ascent through the rocks .
. ,,
DISCUSSION

A. F. LUCAS, Washington, D. C.-I have long held the opinion

that igneous rock may underlie the salt domes of Louisiana, Texas,
and elsewhere, resembling laccoliths, batholiths, or sills, and an un-
successful attempt to prove this was made at my instance by the drill-
ing of the Knapp well at Belle Isle, La., which passed through over
3,000 ft. of rock salt (the first time to my knowledge that a salt dome
has been pierced), then entered into sedimentary, striking first an
oil sand giving about one barrel of red paraffine oil, then limestone,
calcite, or dolomite (not positively ascertained),. until a total depth
was reached of about 3,300 ft., when the well was lost by the collapse of
the casing.
Atth~ extreme bottom the drill was working on very hard formation,
which it could not penetrate, although repeated attempts were made,
nor was it possible to secure samples of this rock, which, from' its ex-
treme hardness, I surmised to be possibly some igneous rock, or perhaps
iron pyrite.
EUGENE COSTE,Toronto, Canada.-I agree entirely with Mr. Wash-
burne's view that the strong salinity of these waters in the oil fields is due
to solfataric volcanic emanations; the composition of these waters, not
. only the large excess of chlorine in them but the excess of magnesium and
calcium, differentiates them entirely from sea waters and they can only
be compared with magmatic waters. In the Texas and Louisiana oil
fields we certainly have the· clearest examples of solfataric action which
we have in the oil fields of theU nited States-Imean in the salt domes of
Louisiana and Texas, where secondary products, not only of salt in large
bodies, but enormous bodies of dolomite, gypsum, and sulphur, have been
deposited in vertical chimneys by solfataric hot waters accompanied with
hydrocarbons and sulphuretted hydrogen; the salts being impregnated
with natural gas, sulphuretted hydrogen,and oil, and the salt waters and
the oils being quite hot yet in some cases.
A.C." LANE, Tufts College, Mass.-Mr. Washburne's paper brings
forward some interesting points; and.his main thesis, that these are de-
rived from within (juvenile); is one well worthy of careful consideration.
Does he not; however, assume that the early sea water had a character
 694 CHLORIDES IN OIL-FIELD WAT$RS

too much like that at present? Judgingfrom the average water now re-
ceived by the ocean, either according to Murray of Clarke, there must
have been a gradual change in its composition, as Hunt, GoesSman and I
. have pointed out, not only in increasing salinity but in the relative pro-
portions of the different salts.
I t would be interesting for him to extend. his collections of analyses to
, a wider range of geological horizons. Those cited are Paleozoic. In a
paper on tho chemical evolution of the ocean, yet unpublished, I have
shown reason to believe that while underground circulation tends to run
up the ratio 'of N a : Cl, the character of truly stagnant water varies with
its geological age.
The trouble with following his suggestions and connecting the calcium
chloride waters of the Lake Superior mines with the highly calcic magmas
of the greenstones associated is that the calcium chloride waters occur
just the same in the felsitic conglomerates iqterbedded and also in far-
removed sandstones, e.g., near Whitefish lake and Ontonagon, in the
Michigan iron mines, and in the Storm King granite. If we are to look
to an earth sweat for these universal waters, then I think it would be
better to assume the chloride merely to be slowly diffused from the in-
terior, picking up its bases en route. But a serious difficulty there is that
we have, for instance, under the oil-field waters referred to by Washburne,
the St. Peter sandstone, with a water which is much fresher, apparently
cutting off any prospect of a juvenile source, which otherwise I should be
delighted to accept.
The objection to Washburne's theory is the fact, if it be a fact, that
different geological horizons have different chemical characters, and that
if not too much disturbed by circulation they can be recognized thereby.