Geothermics 38 (2009) 370378

Contents lists available at ScienceDirect

Geothermics
journal homepage: www.elsevier.com/locate/geothermics

Hydrology, hydrochemistry and geothermal potential of El Chichn

volcano-hydrothermal system, Mexico
Yuri A. Taran , Loc Peiffer
Instituto de Geofsica, Universidad Nacional Autnoma de Mxico, 3000, Av. Universidad, Coyoacan, 04510 Mxico, D.F., Mexico

a r t i c l e

i n f o

Article history:
Received 17 April 2009
Accepted 8 September 2009
Available online 1 October 2009
Keywords:
Geothermal
Chloride inventory
Flow measurements
Hot springs
Heat and mass output
Hydrothermal systems
El Chichn
Mexico

a b s t r a c t
Fluid and heat discharge rates of thermal springs of El Chichn volcano were measured using the chloride
inventory method. Four of the ve known groups of hot springs discharge near-neutral NaCaClSO4
waters with a similar composition (Cl 15002000 mg kg1 and Cl/SO4 3) and temperatures in the
5074 C range. The other group discharges acidic (pH 2.22.7) NaCl water of high salinity (>15 g/L). All
ve groups are located on the volcano slopes, 23 km in a straight line from the bottom of the volcano
crater. They are in the upper parts of canyons where thermal waters mix with surface meteoric waters
and form thermal streams. All these streams ow into the Ro Magdalena, which is the only drainage
of all thermal waters coming from the volcano. The total Cl and SO4 discharges measured in the Ro
Magdalena downstream from its junction with all the thermal streams are very close to the sum of the
transported Cl and SO4 by each of these streams, indicating that the inltration through the river bed
is low. The net discharge rate of hydrothermal Cl measured for all thermal springs is about 468 g s1 ,
which corresponds to 234 kg s1 of hot water with Cl = 2000 mg kg1 . Together with earlier calculations
of the hydrothermal steam output from the volcano crater, the total natural heat output from El Chichn
is estimated to be about 160 MWt. Such a high and concentrated discharge of thermal waters from a
hydrothermal system is not common and may indicate the high geothermal potential of the system. For
the deep water temperatures in the 200250 C range (based on geothermometry), and a mass ow rate
of 234 kg s1 , the total heat being discharged by the upowing hot waters may be 175210 MWt.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
Fluid discharge rates of major thermal areas, worldwide, range
from a few liters to several cubic meters of hot water per second. For
example, the total thermal waters discharge from the Yellowstone
Caldera, USA, was estimated from chloride ux measurements
to be 3200 L s1 (Fournier, 1989). At the Mutnovsky geothermal
eld in Kamchatka, Russia, where a 60 MWe power plant is now
operational, the total hot water discharge rate before drilling was
estimated to be about 80 L s1 (Vakin and Pilipenko, 1986). Preexploitation discharge at Wairakei, New Zealand (190 MWe power
plant in operation), was about 100 L s1 (Ellis and Wilson, 1955).
About 340 L s1 was measured by Mariner et al. (1990) for the thermal water discharge along about 1000 km of the Cascade Range, in
NW USA. A hot spring area associated with the Lassen Peak, USA,
hydrothermal system is characterized by about 20 L s1 of measured thermal water ow (Sorey, 1986). Heat output associated
with the hot water discharge (i.e. the advective heat ow) may be
used as a preliminary indicator of the thermal power of a given

Corresponding author. Tel.: +52 55 5622 4145; fax: +52 55 5550 2486.
E-mail address: taran@geosica.unam.mx (Y.A. Taran).
0375-6505/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geothermics.2009.09.002

hydrothermal system (Ellis and Mahon, 1977; Hedenquist et al.,

1988; Pilipenko, 1989; Hochstein and Browne, 2000).
Prior to this study the amount of thermal water discharging
from the volcano-hydrothermal system of El Chichn volcano in the
state of Chiapas, Mexico, was poorly known. Taran et al. (1998) and
Rouwet et al. (2004) reported a very preliminary ow rate of about
100 L s1 for the Agua Caliente hot springs at El Chichn. Taran and
Rouwet (2008) and Mazot and Taran (2009) estimated the heat output from the volcano crater, where the main convective heat ow
is provided by the steam vents, to be about 120 30 MWt.
The purpose of the work reported herein was to determine the
uid and heat output rates from hot springs on the slopes of El
Chichn. The mass and heat output associated with the discharge
of steam from the crater and thermal water from the slopes of
volcano and its distribution over the area may be further used
to quantitatively model and characterize the El Chichn volcanohydrothermal system.
2. El Chichn volcano and its hydrothermal manifestations
El Chichn is a small, complex volcanic edice composed of
domes and a pyroclastic cover of trachyandesitic composition. Its
maximum elevation is about 1100 m above sea level (m asl). Max-

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

371

Fig. 1. Simplied geologic map of El Chichn area based on the Digital Elevation Model showing the main hydrographic network (modied after Layer et al., 2009). Red dots:
hot springs (see Fig. 2). Also shown is the location of the NWSE cross-section given in Fig. 9. (For interpretation of the references to color in this gure legend, the reader is
referred to the web version of the article.)

imum relief on the southwestern side of the volcano is 800 m,

whereas to the ENE it is only 400500 m. This seems to indicate that
the pre-volcanic basement beneath El Chichn is generally dipping
to the SSW.
The basement is represented by NeogenePaleogene sandstones
and siltstones, and Cretaceous to Jurassic limestones and evaporites (Macas et al., 2003; Garca-Palomo et al., 2004). The oldest
exposed volcanic structure is the southern extrusive dome with
a 40 Ar/39 Ar age of about 200 Ka (Layer et al., 2009). The age of a
younger NW dome structure was estimated to be around 90 Ka.
A schematic geologic map of the El Chichn area is given in
Fig. 1.
Geothermal surface manifestations existed at the same general locations as today before the 1982 catastrophic eruption of
El Chichn (Molina-Berbeyer, 1974; Templos et al., 1981). Thermal
springs on the southern slopes of the volcano and steam vents at
the base of the former central lava dome provided evidence for
a boiling hydrothermal system beneath the volcanic edice. The
Comisin Federal de Electricidad of Mexico (CFE) completed an initial geological and geothermal investigation of the area and issued
several internal reports with a limited set of very preliminary data
on the chemical composition of the summit fumaroles and thermal
waters of the Agua Caliente hot springs (Molina-Berbeyer, 1974;
Templos et al., 1981).
After the 1982 eruption a one km wide, 160 m deep crater was
formed that exposed the El Chichn hydrothermal system. The elevation of the crater oor is 870 m asl. The total thermal output
of the crater fumaroles, steam-heated pools, boiling springs and a
warm, large and shallow crater lake with bubbling gas was estimated to be about 120 30 MWt (about 45 10 kg s1 of steam;
Taran and Rouwet, 2008; Mazot and Taran, 2009). A chain of hot
springs off the crater extends along the southern slope from SE to
NW at about 600 m asl, 23 km from the crater oor (Fig. 1). All
springs are located at the upper parts of canyons whose mouths
end at the Ro Magdalena, the only drainage of all thermal waters
discharging from the volcano slopes. The geochemistry of the spring
waters, including trace elements, water isotopes and geothermometry, was reported by Taran et al. (2008).

The study area is in a wet tropical zone with an average yearly

rainfall of about 4000 mm (Taran and Rouwet, 2008). The volcanic
edice is covered by thick vegetation that grew after the 1982 eruption.
3. Methodology
3.1. Flow rate measurements
All thermal springs on El Chichn slopes discharge hot water
from multiple orices or as seepages from walls of altered rocks
or thermal swamps; some vents are located in inaccessible parts
of deep canyons covered by thick vegetation. All these springs
drain into cold water streams whose ow rates vary during the
wet and dry seasons. However, the hot spring discharge rates vary
insignicantly, as indicated by a near constant Cl content of the
hottest vents, independent of the season (Tassi et al., 2003; Rouwet,
2006; Taran et al., 2008). All cold (i.e. non-thermal) streams at El
Chichn are relatively high in Cl, (79 mg L1 ) and high in SO4 (up
to 300 mg L1 ) due to leaching of the anhydrite-rich pyroclastics
deposited by the 1982 eruption (Luhr et al., 1984).
The discharge rate of each group of hot water springs was determined based on measurements made where streams merge into
the Ro Magdalena, the only drainage for all El Chichn thermal
springs. We utilized the chloride inventory method of Ellis and
Wilson (1955), that has been used subsequently by several authors
(e.g. Fournier et al., 1976; Fournier, 1989; Pilipenko, 1989; Mariner
et al., 1990; Ingebritsen et al., 2001; Taran, 2009).
The basis of the Cl-inventory method is the simultaneous measurement of ow rate and chemical composition of streams and
rivers at specic sampling locations. We used a standard FP101
ow probe (Global Water) for determining ow rates in the Ro
Magdalena and all streams entering it from the southern slopes of
the volcano. We measured and calculated ow rates following the
methodology of Rantz et al. (1982). This was facilitated because
near El Chichn the Ro Magdalena can be waded across in many
places. The ve-point method was used with measurements at 0,
0.2, 0.5, 0.7, 1.0 of the total water depth, with 68 vertical proles

1
1
1
1
1
93.2762
93.272
93.2717
93.216
93.211
17.3605
17.3525
17.3417
17.334
17.334
0.3
4.2
0.1
0.1
1.8
129
135
103
101
82
145
390
388
637
725
446
440
258
563
278
2.1
5.8
4.5
5.8
4.0
11
27
16
38
24
113
177
202
306
302
26
33
13
31
26
References: 1: This work; 2: Taran et al. (1998); 3: Taran et al. (2008).
a
Thermal streams draining the corresponding groups of hot springs (see Fig. 2).

238
206
122
264
140
40.6
89.9
63.1
70.6
55.6
8.30
8.25
8.22
8.18
7.82
Nov-08
Nov-08
Nov-08
Nov-08
Nov-08
Thermal streamsa
Ro AS
Ro AT1
Ro AT2
Ro ACn
Ro AC

24.3
26.6
32.0
27.4
27.5

93.2212
93.252
93.2167
93.225
93.2363
93.255
93.255
17.3463
17.3608
17.3528
17.3602
17.3522
17.3763
17.3763
2.4
1.1
0.9
1.2
3.5
6.1
1.6
196
103
228
229
185
0
0
456
525
451
781
900
546
590
13
1504
16
1827
31
2230
15.2
1762
16
1554
0.5
213
38
10,003
48
49
70
77
61
14
40
368
514
623
517
666
95
1181
60
90
109
120
55
33
488
656
704
816
828
702
198
4830
159
191
172
152
107
157
225
7.45
7.74
6.42
5.56
5.98
2.76
2.18

SO4
Cl
B
Mg
Ca
K
Na
SiO2
pH
T ( C)

Jun-08
Jun-08
Jan-97
Mar-05
Apr-98
Jun-04
Jun-04

There are ve groups of hot springs on the volcano slopes (Fig. 2).
The AT1 and AT2 groups are actually made up of two subgroups,
each one with its vent locations along tributary streams valleys
separated by a drainage divide. The second (western) subgroup of
the AT1 springs and a new group that has been called ACn (Agua
Caliente new) were discovered in 2008. Table 1 shows all chemical
data (major species) for the hot springs (samples collected directly
from vents) and for the thermal streams draining the corresponding
hot springs, where these streams enter the Ro Magdalena.
Mixing relationships for hot springs and thermal streams are
given in Fig. 3(a), and for the Ro Magdalena in Fig. 3(b). Both plots
demonstrate that there is a good correlation between Na and Cl for

Hot springs
ACn
AT1n
AC
AT1
AT2
AS2
AS3

4.1. Chemistry of thermal springs and streams

57
50
74
68
49
75
58

Longitude ( W)
Charge balance (%)

Latitude ( N)

4. Results and discussion

Sampling dates

All water samples were collected in plastic bottles after ltration

through a 0.45 m MILLIPORE lter. Samples for cation analysis
were acidied using ultra-pure HNO3 . Major dissolved species (Na,
K, Ca, Mg, Cl, SO4 ) were analyzed by a METRHOM-761 ion chromatograph. Alkalinity (expressed as HCO3 concentrations) was
measured by titration with 0.1 N HCl. The total (dissolved and colloidal) SiO2 was determined by the colorimetric technique using
ammonium molybdate (Giggenbach and Goguel, 1989). We consider the analytical errors to be usually less than 5%.

Hot springs and thermal streams

3.2. Water sampling and analysis

Table 1
Chemical composition (mg/kg) and location of hot springs and sampling sites for thermal streams on the El Chichn slopes.

across the ow, so that at each measurement station we took a total

of not less than 30 current velocities in a given cross-section. The
total relative error of the measured ow rates probably does not
exceed 5%.
On the other hand, at some places the chemistry of the Ro
Magdalena water varies signicantly because of incomplete mixing
with waters from thermal streams (entering exclusively from the
northern and north-eastern side of the river). Averaging the total
Cl and SO4 contents in a river cross-section may lead to errors up to
20%. In contrast, the errors in the chemical uxes for the individual
streams should be signicantly lower (510%) because of a more
complete mixing.

HCO3

Fig. 2. Hydrographic network on the slopes of El Chichn volcano; the Ro Magdalena and Ro Susnubac are the main drainage paths for all streams on the southern
and western slopes (see text and Table 2 for more details). Note that all spring are
located close to the 600 m asl contour (gray line). SP: Soap Pool, a boiling, geyser-like
spring located within the crater.

1
1
2
3
3
3
3

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

Ref.

372

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

373

Fig. 4. Correlation between Ca and SO4 for all waters collected at El Chichn. Streams
and shallow groundwaters waters show a good correlation with an almost equimolar
Ca/SO4 ratio due to leaching of anhydrite (CaSO4 ) from pyroclastic deposits (see text
for details).

Fig. 3. Mixing plots: (a) for the El Chichn hot springs and thermal streams; (b) for
the cold streams and the Ro Magdalena.

all waters, but a very poor correlation exists between Cl and Ca. This
indicates that the source of Na and Cl is the same for all samples.
However, the absence of such correlation for Ca suggests that there
is an additional source of Ca. Taran et al. (1998) reported a wide
occurrence of the so-called Red Waters discharging at the base
of the volcanic edice. The composition of these shallow ground
waters of meteoric origin is formed by the leaching of fresh (1982
eruption) pyroclastics rich in anhydrite (more than 2 wt%; Luhr et
al., 1984); the most concentrated Red Waters contain about 2 g/kg
of (Ca2+ + SO4 2 ).
We conclude that all surface waters in the area are more or less
contaminated by Ca2+ and SO4 2 from the Red Waters, in close to
equimolar proportions. This can be seen in Fig. 4, where the correlation of Ca2+ and SO4 2 is shown for all waters from the hot springs,
thermal streams and the Ro Magdalena. Points for the streams and
the river lie close to the CaSO4 equimolar line, whereas hot springs
also show a strong CaSO4 correlation, but with a slope about two
times steeper than that for the streams. The line Soap Pool (SP) is
related to the boiling, geyser-like springs in the El Chichn crater
which discharges a neutral NaCaCl water with a lower sulfate
concentration (Taran et al., 1998, 2008). Thus, the concentrations
of Ca and SO4 in hot springs could be, at least partially, formed
by the mixing of deep, relatively SO4 -poor, thermal waters with
shallow groundwaters that have leached anhydrite from the 1982
pyroclastic deposits.
4.2. Variation in the chemistry and ow rate of the Ro Magdalena
Table 2 presents changes in the chemistry, ow rates and uxes
of Cl and SO4 of the Ro Magdalena after accepting waters from
streams draining into it (Fig. 2), as well as the compositions, ow

rates and uxes for all tributaries (thermal and cold). The last two
columns show measured and calculated uxes of Cl and SO4 in the
river from the uppermost site RM8 to the lowermost site RM1 (RM
stands for Ro Magdalena); the information is also shown in Fig. 5.
The calculated values are a simple sum of the uxes of the consecutive tributaries before the current measurement site in the Ro
Magdalena and Ro Susnubac. In general, a slight decit of Cl and
SO4 (calculated ux > measured ux) is observed. It could be the
result of errors (ow rate measurements, source heterogeneity and
analysis) and due to inltration (water losses) through the river bed.
In any case, the mass balance is within about 20% of the estimated
error range (see Section 3.1).
4.3. Estimation of the thermal output from hot springs on the El
Chichn slopes
The heat output (convective discharged heat) from a group of
hot springs can be calculated as follows:
W = Q (Hs Ha )

(1)

where Q is the effective mass ow rate (see below) of a group of

hot springs (in L s1 ); Hs is the enthalpy of the hot spring water (in
kJ/kg); and Ha is the reference enthalpy (background) corresponding to that of water at ambient temperature.
Each group of hot springs has a vent with a maximum temperature that always corresponds to a maximum Cl content. The
maximum temperature of the Agua Caliente springs was 74 C
as measured in 2004. This vent discharged water with about
2200 mg/kg of chloride. Ro AC (a stream draining AC group, Site
9, Fig. 2 and Table 2) transports 56 g s1 of Cl. This corresponds
to an effective ow rate of 25 kg s1 of hot (74 C) water with
2200 mg/kg of Cl. Taking into account that the annual ambient temperature around El Chichn is 22 C (Taran and Rouwet, 2008), we
have a heat discharge value of 5.4 MWt for the AC group.
Results for each group of springs are shown in Table 3. The
Ro Magdalena at the lowermost site (RM1) transports 529 g s1
of Cl with a ow rate of 11,020 L s1 . A signicant amount of this
chloride is contributed by shallow groundwater which on average
has about 8 mg/kg of Cl. Accounting for this the total discharge
of the hydrothermal chloride is reduced by about 88 g s1 (i.e.
11,020 L s1 8 mg/kg), and becomes 440 88 g s1 with an error
of 20% in determining the chloride ux in the Ro Magdalena (Section 3.1)

374

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

Table 2
Water chemistry, ow rates and Cl and SO4 uxes (FCl and FSO4 ) for the Ro Magdalena and Ro Susnubac (stations RM1 to RM8) and for all streams between the uppermost
station RM8 and the lowermost station RM1.Cl and SO4 uxes for the Ro Magdalena are shown as measured (M) and calculated (C). The latter values are obtained by summing
the contributions from each stream or group of streams (see text for details).
Site

Number shown in Fig. 2

Na in mg/kg

Ca in mg/kg

Cl in mg/kg

SO4 in mg/kg

RM8a
RM7
RM6
RM5
RM4
RM3
RM2
RM1
Ro AC
Ro ACn
Ro 7
Ro 6
Ro 5
Ro 12
RM10b
Ro AT2
Ro 4
Ro 3
Ro AT1
Ro AS
Ro 1

8
7
6
5
4
3
2
1
9
8
7
6
10
12
13
5
4
11
3
2
1

5
14
33
35
8.2
8.5
15
32
140
264
22
5.8
7
7
8.5
122
8.8
25
206
238
61

107
111
149
143
72
80
79
92
302
306
173
110
169
23
80
202
329
127
177
113
117

9
30
61
56
24
21
52
48
278
563
31
7
8
6
7.0
201
12
9
440
446
62

235
288
305
316
124
135
140
135
646
570
361
209
249
42
54
464
829
871
390
145
255

Flow rate in L/s

2520
2710
2940
3020
8020
9670
10,510
11,020
200
210
80
90
10
3
5010
280
10
10
390
110
490

FCl (g/s) M/C

FSO4 (g/s) M/C

23/23
81/79
179/197
169/200
192/235
369/463
547/556
529/586
56
118
2.5
0.9
0.1
0.0
35
56
0.12
0.1
172
49
30

592/592
780/734
896/864
954/893
994/1183
1305/1444
1471/1460
1488/1585
129
120
29
20
3
0.12
270
129
8
9
152
16
125

RM5 to RM8 are stations on the Ro Susnubac.

Station RM10 is on the Ro Magdalena upstream from its junction with Ro Susnubac (Fig. 2). Ro AC, Ro ACn, Ro AT1, Ro AT2 and Ro AS are streams draining the
corresponding groups of hot springs (see Fig. 2 and Table 3).
b

From Table 3 the sum of the measured chloride uxes from

all thermal groups is 468 47 g s1 with an error of 10% for the
thermal stream Cl ux (Section 3.1). The agreement between the
measured integrated Ro Magdalena chloride ux and the sum of
the measured thermal stream uxes is good. The difference could
be the result of both errors of multiple origins, and the loss of
water through the riverbed. The convective heat output from hot
springs of the El Chichn volcano-hydrothermal system discharging at the slope of the volcanic edice is thus estimated to be about
45 5 MWt (Table 3). This value is about two times lower than was
estimated by Taran et al. (2008).
It should be noted that the anomalous high Cl content in Ro 1
(62 mg L1 ) and Ro 7 (31 mg L1 ) (Table 2) may indicate the existence of two more undiscovered groups of hot springs in the upper
parts of the corresponding canyons. Assuming about 2000 mg L1
of Cl in the waters from these undiscovered groups, their discharge
rate should be about 15 L s1 for Ro 1 (Ro Cambac) springs and
1 L s1 for Ro 7.
4.4. Summary of the solute and gas geothermometry
Taran et al. (1998), Rouwet (2006) and Taran et al. (2008) using
different geothermometers and thermochemical calculations have
shown that temperatures in the El Chichn geothermal system at
depth are in 200260 C range. The application of ionic geothermometers to El Chichn thermal waters is somewhat controversial

and not obvious because of their high Mg and Ca contents. All water
analyses including those given by Taran et al. (1998) and Taran et al.
(2008) are plotted on the ternary NaKMg diagram (Giggenbach,
1988) together with the composition of the crater lake after dissolving 100 g, 10 g, and 1 g of average El Chichn erupted rock in
1 kg of water (Fig. 6); the assumed rock compositions was taken
from Andrews et al. (2008). All data points correspond to immature waters, and show two distinct trends. One for the thermal
springs extends between the 230 C and the 250 C NaK isotherms,
and another for the surface Red Waters and the acidic lake water
sampled in 1983 (Casadevall et al., 1984) lies on the El Chichn
erupted rock trend. We believe that the trend along the 250 C NaK
isotherm corresponds to both re-equilibration of the deep NaCl
hot uid to lower temperatures (dissolution of Mg-rich clay minerals) and mixing of deep uid with surface waters enriched in Ca,
Mg and SO4 . It is accepted that Na and K are more conservative,
whereas Mg and Ca are rapidly affected by lowering of temperature
and degassing (Fournier, 1981; Giggenbach, 1988).
Taran et al. (2008), using the SOLVEQ computer code (Reed,
1982), have shown that saturation indices of all but Mg-silicates
for waters of neutral thermal springs ideally converge to a single equilibrium temperature of about 180 C. For the SP boiling
hot springs located in the crater (Fig. 2) we estimated an equilibrium temperature of 220 C. For the acid AS springs a very good
convergence of saturation indices to a lower equilibrium temperature of 160 C was obtained. Therefore, we consider that 200 C is

Table 3
Estimated thermal water and heat outow rates for hot spring groups on the El Chichn slopes computed based on thermal stream ow rates and Cl in the spring waters.
Q: water ow rates; FCl : chloride ux; H: heat output. Ro Cambac stands for a hot spring or group of springs in the Ro Cambac canyon (Fig. 2) whose exact location is not
known at present.
Group of hot springs

Max. spring temp. ( C)

Cl max (mg/kg)

Agua Caliente (AC)

Agua Caliente New (ACn)
Agua Tibia 1 (AT1)
Agua Tibia 2 (AT2)
Agua Salada (AS)
Ro Cambac

74
57
68
51
75
50?

2200
1830
1760
1550
2000
2000?

Total

Q stream (L/s)
200
210
390
280
110
490
1680

Cl stream (mg/kg)

FCl (g/s)

278
563
440
201
446
62

54
117
169
54
48
26

Q spring (L/s)
25.3
63.8
99.1
34.9
24.1
13.2?

468

260.4?

H spring (MWt)
5.4
9.3
19.1
4.2
5.3
2?
45.3?

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

375

a reasonable estimate of reservoir temperature for the El Chichn

volcano-hydrothermal system.
Perhaps the closest analogue to the El Chichn volcanohydrothermal system is Mt. Pinatubo (Philippines). Before the
catastrophic 1991 eruption, three deep exploration wells were
drilled close to the volcano summit. Ruaya et al. (1992), Deln et al.
(1996) and Stimac et al. (2004) discussed the chemistry of hot acidic
brines from Pinatubo wells and springs applying, among other geochemical tools, the Giggenbach (1988) ternary plot. These brines
were essentially immature, but their compositions plotted along
the NaK isotherms corresponded to temperatures measured in
wells (i.e. 250300 C). These authors also showed that the crater
lake of Mt. Pinatubo in 2002 had a signicant input from the underlying hydrothermal system like during the 19952008 evolution of
the El Chichn crater lake (Taran et al., 1998; Rouwet et al., 2004;
Taran and Rouwet, 2008).
Taran et al. (1998) and Rouwet (2006), based on gas geothermometry (i.e. equilibrium in the CO2 H2 COH2 system) applied
to the steam vents in the volcano crater, suggested 250300 C temperatures and a two-phase uid in the equilibrium zone. Moreover,
the high H2 content of crater fumaroles, with a H2 /H2 O ratio typical for vapor-dominated hydrothermal systems (Giggenbach, 1987;
Taran et al., 1998; Tassi et al., 2003), also indicate the presence of
a steam-dominated or a vapor-enriched two-phase zone beneath
the crater oor. The only known hot springs outside the crater
showing (very weak) gas bubbling are those of Agua Caliente. The
absence of bubbling gas in the neutral hot springs suggests almost
complete degassing of the deep uid through the thermal manifestations (steam vents, steam-heated pools and boiling springs) in the
volcano crater and that the top of the thermal aquifer is centered
within the crater.
5. Geothermal potential and conceptual model of El
Chichn volcano-hydrothermal system
The advective heat transport of the hot-spring systems can be
expressed as (Ingebritsen et al., 2001):
A=
Fig. 5. Chloride and sulfate uxes calculated from the Ro Magdalena current ow
measurements and Cl content (black squares) and as a sum of contributions (bars)
from all tributaries located upstream of the current ow stations (see text for
details).

Fig. 6. NaKMg ternary diagram (Giggenbach, 1988) for El Chichn spring waters.
The rock trend is shown as composition of waters after dissolving 100 g, 10 g, and
1 g of the El Chichn volcanic rock. See text for details.

FCl c(Tg Ta )
Cls

(2)

where FCl is the Cl ux, c the heat capacity of the uid, Tg the uid
temperature at depth (as determined by chemical geothermometry), Cls is the maximum chloride concentration in the thermal
water, and Ta is the ambient temperature.
The measured Cl ux for all thermal springs at El Chichn is
468 g s1 , which corresponds to a ow rate of 234 L s1 of water
with 2000 mg L1 of Cl. Considering a minimum initial water temperature of 200 C, the estimated amount of heat advected from
the deep heat source calculated using Eq. (2) is 175 MWt. This value
should be considered as a maximum since it is based on the assumption that the deep thermal water has about 2000 mg/kg of chloride,
close to the maximum measured Cl in the neutral springs. Only
about a quarter of the deep advected heat (i.e. 45 5 MWt) is transported to the surface by the hot spring waters; the rest is lost by
conduction and other processes as the water moves toward the
discharging vents.
The estimated advected thermal output of 175 MWt would correspond to an average heat ux of about 1750 mWt m2 over an
area of 100 km2 around the volcano crater. This value is close to
the estimated average advective ux for the 2500 km2 Yellowstone
Caldera (i.e. about 2000 mWt m2 ; Fournier, 1989).
The surface manifestations of the El Chichn geothermal system are restricted to a relatively small area and the amount of
the discharged hot water is signicantly higher than the annual
recharge by rainfall inltrating through the bottom of the crater.
The area bounded by the crater rim is about 800,000 m2 . Consider-

376

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

Fig. 7. Mixing plots Na, Ca vs. Cl for all neutral El Chichn springs including the
Soap Pool boiling springs (SP) in the volcano crater. The increment Ca shows the
contribution of Ca from cold shallow groundwaters saturated or close to saturation
with respect to anhydrite (see text for details).

ing an average annual rainfall of 4 m the total volume of rain falling

into the crater is about 3.2 106 m3 per year, which corresponds to
about 100 kg s1 or 2.3 times smaller than the estimated thermal
water discharge rate. This indicates that the seepage through the
crater oor, even when evaporation is neglected, cannot supply the
amount of water being discharged by the El Chichn hot springs and
that the main source of these thermal waters is located at depth.
In Fig. 7 Na and Ca concentrations in the neutral waters of the hot
springs on the volcano slopes and of the SP boiling springs within
the crater are plotted vs. chloride. Since there is a very good correlation between Na and Cl over the entire concentration range, this
mixing plot conrms that all neutral thermal waters at El Chichn
have the same source of Na and Cl.
A good correlation between Cl and Ca is also observed for the
SP boiling springs. As noted in Section 4.1, spring waters have an
additional source of Ca (and SO4 ) in cold, shallow groundwaters
that are saturated or are close to saturation with respect to anhydrite (shown in Fig. 7 as Ca). Fig. 7 also suggests that waters
from El Chichn thermal springs, despite their relatively high chloride contents (about 2000 ppm), may be signicantly diluted, and
that the parent thermal uid may have high salinity as observed
in the Mt. Pinatubo hydrothermal system (Deln et al., 1996;
Stimac et al., 2004). An indirect evidence for the high-salinity parent thermal water of the El Chichn deep aquifer may be the
chloridetemperature plot shown in Fig. 8. The data points on this
graph correspond to all thermal vents sampled between 1995 and
2008 (except the Agua Salada springs) and the chemical compositions given in Taran et al. (1998), Tassi et al. (2003), Rouwet (2006)
and Taran et al. (2008).
Fig. 8 shows that there is a good correlation between Cl and temperature that may be the result of shallow mixing between about
100 C boiling water rising to the surface and loosing steam quasiadiabatically, and shallow groundwaters. In that case, the chloride
concentration in this boiling end-member can be estimated to be
about 3000 mg/kg and the heat ow from the thermal aquifer will
be 3000/2000 = 1.5 times lower, i.e. about 120 MWt.
The linear correlation between Cl and temperature (T) with the
best-t regression line (Fig. 8):
Cl = 36T 575
C,

(3)
mg L1 ),

(T in
Cl in
provides another independent way to
estimate the natural thermal output associated with the ow of
thermal waters. Combining Eqs. (2) and (3) and using Cls = 2 g L1 ,

Fig. 8. Correlation between chloride content and discharge temperature (T in C)

for all neutral thermal springs on the slopes of El Chichn.

FCl = 1 g s1 (Tg is replaced by T calculated from Eq. (3)), the

chlorideheat equivalent for the El Chichn hydrothermal system would be about 0.103 MWt corresponding to a chloride ux of
1 g s1 . Therefore, the measured 468 g s1 of chloride discharging
from all El Chichn hot springs (Table 3) corresponds to a natural
heat ow of about 48 MWt, close to the calculated thermal power
obtained using the enthalpies of the spring waters (about 45 MWt;
see Section 4.3).
In the absence of exploration drilling and hence lack of subsurface geological and hydrological data (as was the case for Mount
Pinatubo; Deln et al., 1996) and of details on local tectonics,
it is difcult to develop a conceptual model of the El Chichn
volcano-hydrothermal system. The tectonic and structural analysis of the El Chichn area by Garca-Palomo et al. (2004) did not
reveal any system of faults or fracture zones associated with the
hot springs and did not explain the unusually restricted location
of the springs on the southern slopes of the volcano at approximately 600 m asl (Fig. 2). On the other hand, the high rate of hot
spring water discharge may indicate the presence of a large, fractured, high-permeability geothermal reservoir at depth. Table 4
compares selected hydrothermal systems of the world in terms
of their mean chloride uxes, Cl concentrations in the hot spring
waters, estimated deep temperatures, measured thermal output of
hot springs, and calculated advective heat ows. The El Chichn
volcano-hydrothermal system is characterized by extremely high
chloride ux considering the relatively small area over which thermal manifestations are distributed (10 km2 including the volcano
crater; see Figs. 1 and 2).
The cross-section in Fig. 9 shows a very preliminary model for
the El Chichn volcano-hydrothermal system having the following
main features:
1. The heat source is located at an unknown depth and may
be represented by the hot plumbing system of the volcano, which includes its main magmatic conduit, the roots
of the volcanic domes and (possible) intrusive bodies. Before
the 1982 eruption, the El Chichn trachyandesitic magma
was thought to have been near water saturated at a temperature of 800850 C, a pressure of about 2000 bars (Luhr,
1990), and a corresponding depth of 67 km. According to seismic data (Jimnez et al., 1999) a magma chamber, if present
beneath the volcanic edice, would be at more than 8 km
depth. Further geophysical exploration is needed to resolve this
issue.

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

377

Table 4
Chloride uxes and mean advective heat ow through selected hydrothermal systems.
System

Cl ux (g/s)

Cl concentration in
hot spring waters
(mg/L)

Source temperature
(Tg ) based on
geothermometry ( C)

Natural heat owa

(MWt)

Advective heat owb

(MWt)

References

Yellowstone USA
Cascades, USA 4149 N
El Chichn, Mxico
Wairakei, New Zealand 1951
Rotorua, New Zealand
Lassen Peak, USA
Mutnovsky, Russia

12701737
721
468
300
52
42
13

400450
Variable
2000
1610
700
2400
220

340350
Variable
200
260
320
240
270

16002100
82
45
78
31
7.4
34

53006100

175
188
99
18
92

1, 2
3
4
5
6
2, 7
8

References: 1: Fournier (1989); 2: Ingebritsen et al. (2001); 3: Mariner et al. (1990); 4: This work; 5: Glover and Mroczek (2009); 6: Hedenquist et al. (1988); 7: Sorey (1986);
8: Vakin and Pilipenko (1986).
a
Hot springs.
b
See text.

2. The source of hot spring water is a geothermal, water-dominated

aquifer beneath the volcanic edice. There is a high probability
that the parent uid is a brine, maybe acidic, similar to the one
at Mt. Pinatubo. This brine mixes at depth with inltrated meteoric waters. The mixed hydrothermal uid, with temperatures
in the 200250 C range, rises to the surface through a fractured
zone associated with the magmatic conduit of the 1982 eruption
forming a two-phase (steam-water) mixture. The separation of
the main water ow from the upwelling steam-water mixture
occurs above the boiling level, inside volcanic edice.
3. The ow of neutral thermal waters except the SP boiling springs
within the crater is most likely sub-horizontal, along the contact
between the low-permeability basement that gently dips to the
SSW, and the overlying more permeable volcanics. This assumption is based on the observation that all, even the hottest, springs
discharge degassed water. The location of the thermal springs, at
similar elevations and distances from the crater (Fig. 2), suggests
radial ow along that contact.
4. The Agua Salada springs discharge parent brine that has mixed
with shallow groundwater and has been heated by the ascending
steam. One of the probable scenarios is a shallow acidication of
the brine by steam-heated waters due to H2 S oxidation. The brine
rises through a separate system of fractures, probably associated
with old caldera-forming faults or with the magmatic conduit of
the NW dome (Fig. 9). Based on such a model the bubbling gas or
steam vents should exist within this group of springs at higher
levels of the canyon.
5. The adiabatic boiling of water with an initial temperature of
200 C (brinemeteoric water mixture) at depth, and 100 C at

the surface should produce about 15 wt% of steam. Taking into

account the total observed hot water discharge of 234 kg s1 ,
the corresponding discharge of hydrothermal steam through the
crater oor should be about 41 kg s1 . Mazot and Taran (2009)
estimated the total CO2 ux from the volcano crater to be about
370 t d1 , or 4.3 kg s1 based on the measured CO2 ux from the
crater lake. Taking into account that the CO2 concentration in
El Chichn crater fumaroles is on average 10 wt% (Taran et al.,
1998; Tassi et al., 2003; Capaccioni et al., 2004), the associated
steam ux should be close to 45 kg s1 . The two independent
estimates are surprisingly close.
6. The large natural convective heat ow from the El Chichn
volcano-hydrothermal system of about 160 MWt (hot spring
discharge: 45 MWt + steam output from the crater: about
120 30 MWt) is an indication of its high potential, though the
geothermal system might not be a good prospect for commercial
development. Two main issues should be taken into account, the
high probability of nding acidic saline uids at depth and the
active volcanism of the area.
Note added to the proof
After the manuscript was accepted D. Rouwet and F. Sandoval
discovered bubbling steam-heated pools and weak fumaroles at the
upper part of the AS canyon.
Acknowledgments
This study was supported by the CONACyT (Mexico) grant #
81571. We thank George Chelnokov for the eld assistance and
Ruben Bernard for his help with the water analyses. Comments and
suggestions by Jim Stimac and two anonymous reviewers helped
to improve the manuscript. Careful and constructive suggestions
by the editors M. Lippmann and S. Garg are deeply appreciated.
References

Fig. 9. Hypothetical SENW section across the El Chichn volcano.

Andrews, B.J., Gardner, J.E., Housh, T.B., 2008. Repeated recharge, assimilation and
hybridization in magmas erupted from El Chichon as recorded by plagioclase
and amphibole phenocrysts. J. Volcanol. Geotherm. Res. 175, 415426.
Capaccioni, B., Taran, Y.A., Tassi, F., Vaselli, O., Mangani, C., Macias, J.L., 2004. Source
conditions and degradation processes of light hydrocarbons in volcanic gases:
an example from El Chichon volcano (Chiapas State, Mexico). Chem. Geol. 206,
8196.
Casadevall, T.J., De la Cruz Reyna, S., Rose, W.I., Bagley, S., Finnegan, D.L., Zoller, W.H.,
1984. Crater lake and post-eruption hydrothermal activity, El Chichn volcano,
Mexico. J. Volcanol. Geotherm. Res. 23, 169191.
Deln Jr., F.G., Villarosa, H.G., Layugan, D.B., Clemente, V., Candelaria, M.R., Ruaya, J.R.,
1996. Geothermal exploration of the pre-1991 Mount Pinatubo hydrothermal
system. In: Newhall, C.G., Punongbayan, R.S. (Eds.), Fire and MudEruptions and
Lahars of Mount Pinatubo. Philippine Institute of Volcanology and Seismology
and the University of Washington Press, Philippines, pp. 197212.
Ellis, A.J., Wilson, S.H., 1955. The heat from Wairakei-Taupo thermal region calculated from the chloride output. N. Z. J. Sci. Technol. Sect. B 36, 622631.

378

Y.A. Taran, L. Peiffer / Geothermics 38 (2009) 370378

Ellis, A.J., Mahon, W.A.J., 1977. Chemistry and Geothermal Systems. Acad. Press, New
York, 392 pp.
Fournier, R.O., 1981. Application of water chemistry to geothermal exploration and
reservoir engineering. In: Rybach, L., Mufer, L.J.P. (Eds.), Geothermal Systems,
Principles and Case Histories. John Wiley & Sons, pp. 109143.
Fournier, R.O., 1989. Geochemistry and dynamics of the Yellowstone National Park
hydrothermal system. Ann. Rev. Earth Planet. Sci. 17, 1353.
Fournier, R.O., White, D.E., Truesdell, A.H., 1976. Convective heat ow in Yellowstone
National Park. In: Proceedings of the 2nd UN Symposium on the Development
and Use of Geothermal Resources, vol. 1. U.S. Government Printing Ofce, Washington, DC, USA, pp. 731739.
Garca-Palomo, A., Macas, J.L., Espndola, J.M., 2004. Strike-slip faults and K-alkaline
volcanism at El Chichn volcano, southeastern Mexico. J. Volcanol. Geotherm.
Res. 136, 247268.
Giggenbach, W.F., 1987. Redox processes governing the chemistry of fumarolic gas
discharge from White Island, New Zealand. Appl. Geochem. 2, 143161.
Giggenbach, W.F., 1988. Geothermal solute equilibria. Derivation of NaKMgCa
geoindicators. Geochim. Cosmochim. Acta 52, 27492765.
Giggenbach, W.F., Goguel, R.L., 1989. Collection and analysis of geothermal and volcanic water and gas discharges. New Zealand DSIR Chem. Division Report 2407,
Christchurch, New Zealand, 88 pp.
Glover, R.B., Mroczek, E.K., 2009. Chemical changes in natural features and well
discharges in response to production at Wairakei, New Zealand. Geothermics
38, 117133.
Hedenquist, J.W., Mroczek, E.K., Giggenbach, W.F., 1988. Geochemistry of the
Rotokawa geothermal system: summary of data, interpretation and appraisal
for energy development. Chemistry Division DSIR, Technical note 88/6,
Christchurch, New Zealand, 63 pp.
Hochstein, M.P., Browne, P.R.L., 2000. Surface manifestations of geothermal systems
with volcanic heat sources. In: Encyclopedia of Volcanoes. Academic Press, New
York, NY, USA, pp. 835855.
Ingebritsen, S.E., Galloway, D.L., Colvard, E.M., Sorey, M.L., Mariner, R.H., 2001. Time
variation of hydrothermal discharge at selected sites in the western United
States: implications for monitoring. J. Volcanol. Geotherm. Res. 111, 123.
Jimnez, Z., Espndola, V.H., Espndola, J.M., 1999. Evolution of the seismic activity
from the 1982 eruption of El Chichn volcano, Chiapas, Mexico. Bull. Volcanol.
61, 411422.
Layer, P.W., Garca-Palomo, A., Jones, D., Macas, J.L., Arce, J.L., Mora, J.C., 2009.
El Chichn Volcanic Complex, Chiapas, Mexico: stages of evolution based
on eld mapping and 40 Ar/39 Ar geochronology. Geofsica Internacional 48,
3354.
Luhr, J.F., 1990. Experimental phase relations of water- and sulfur-saturated arc
magmas and the 1982 eruption of El Chichon volcano. J. Petrol. 31, 10711114.
Luhr, J.F., Carmichael, I.S.E., Varekamp, J.C., 1984. The 1982 eruptions of El Chichn
volcano, Chiapas, Mexico: mineralogy and petrology of the anhydrite-bearing
pumices. J. Volcanol. Geotherm. Res. 23, 69108.
Macas, J.L., Arce, J.L., Mora, J.C., Espndola, J.M., Saucedo, R., Manetti, P., 2003. A
550-year-old Plinian eruption at El Chichn Volcano, Chiapas, Mexico: explosive
volcanism linked to reheating of the magma reservoir. J. Geophys. Res. 108 (B12),
25692581, doi:10.1029/2003JB002551.
Mariner, R.H., Presser, T.S., Evans, W.C., Pringle, M.K.W., 1990. Discharge rates of
uid and heat by thermal springs of the Cascade Range, Washington, Oregon,
and Northern California. J. Geophys. Res. 95 (B12), 19,51719,531.

Mazot, A., Taran, Y.A., 2009. CO2 ux from the crater lake of El Chichn volcano
(Mxico). Geofsica Internacional 48, 7383.
Molina-Berbeyer, R., 1974. Informe preliminary geoquimico de los uidos geotermicos del volcan del Chichonal, Chiapas. Comisin Federal de Electricidad, Morelia,
Mich., Mexico, unpublished report 27-74, 24 pp.
Pilipenko, G.F., 1989. Thermal springs of the Karymskii volcanic center in Kamchatka.
Volcanol Seismol. 6, 8596.
Rantz, S.E., et al., 1982. Measurement and computation of streamow. Volume 1,
measurement of stage and discharge, U.S. Geol. Survey Water Supply Paper 2175,
284 pp.
Reed, M.H., 1982. Calculation of multicomponent chemical equilibria and reaction
processes in systems involving minerals, gases and an aqueous phase. Geochim.
Cosmochim. Acta 46, 513528.
Rouwet, D., 2006. Estudio geoqumico comparativo de los sistemas hidrotermales
de los volcanes activos en Chiapas: El Chichn y Tacan. Ph.D. Thesis. IGF-UNAM,
Mexico City, Mexico, 218 pp.
Rouwet, D., Taran, Y.A., Varley, N.R., 2004. Dynamics and mass balance of El Chichn
crater lake, Mexico. Geos. Int. 43, 427434.
Ruaya, J.R., Ramos, M.N., Gonantini, R., 1992. Assessment of magmatic components of the uids at Pinatubo volcano-hydrothermal system, Philippines,
from chemical and isotopic data. Geol. Survey of Japan Report 279, pp.
141151.
Sorey, M.L., 1986. Hot spring monitoring at Lassen Volcanic National Park, California
19831985. In: Proceedings of the 11th Workshop on Geothermal Reservoir
Engineering. Stanford University, Stanford, CA, USA, pp. 141149.
Stimac, J.A., Goff, F., Counce, D., Laroque, A.C.L., Hilton, D.R., Morgenstern, U., 2004.
The crater lake and hydrothermal system of Mount Pinatubo Philippines: evolution in the decade after the eruption. Bull. Volcanol. 66, 149167.
Taran, Y.A., 2009. Geochemistry of volcanic and hydrothermal uids and volatile
budget of Kamchatka-Kuril subduction zone. Geochim. Cosmochim. Acta 74,
10671094.
Taran, Y.A., Fischer, T.P., Pokrovsky, B.G., Sano, Y., Armienta, M.A., Macias, J.L., 1998.
Geochemistry of the volcano-hydrothermal system of El Chichn Volcano, Chiapas, Mexico. Bull. Volcanol. 59, 436449.
Taran, Y.A., Rouwet, D., 2008. Estimated thermal inow to El Chichn crater lake
using the energy-budget, chemical and isotope balance approaches. J. Volcanol.
Geotherm. Res. 175, 472481.
Taran, Y., Rouwet, D., Inguaggiato, S., Aiuppa, A., 2008. Major and trace element geochemistry of neutral and acidic thermal springs at El Chichn volcano, Mexico.
Implications for monitoring of the volcanic activity. J. Volcanol. Geotherm. Res.
178, 224236.
Tassi, F., Vaselli, O., Capaccioni, B., Macas, J.L., Nencetti, A., Montegrossi, G.,
Magro, G., 2003. Chemical composition of fumarolic gases and spring discharges from El Chichn volcano, Mexico: causes and implications of the
changes detected over the period 19982000. J. Volcanol. Geotherm. Res. 123,
105121.
Templos, L.A., Mungua Bracamontes, F., Barrera, V.M., 1981. Observaciones geoqumicas en la zona geotrmica del Chichonal, Chiapas, Mexico. Comisin
Federal de Electricidad, Morelia, Mich., Mexico, unpublished report 33-81, 32
pp.
Vakin, E.A., Pilipenko, G.F., 1986. The Mutnovsky geothermal area in Kamchatka. In:
Exploration and Exploitation of Geothermal Resources in Volcanic Areas. Nauka,
Moscow, Russia, 3645 pp (in Russian).