-
Geothermics 34 (2005) 99118
Interpretation of a well interference test at theChingshui
geothermal field, Taiwan
Kai C. Fan, M.C. Tom Kuo, Kan F. Liang,C. Shu Lee, Shin C.
Chiang
Department of Mineral and Petroleum Engineering, National Cheng
Kung University, Tainan, TaiwanReceived 29 August 2003; accepted 5
November 2004
Available online 22 January 2005
Abstract
Production in the liquid-dominated Chingshui geothermal field is
largely from a fractured zonein the Jentse Member of the Miocene
Lushan Formation. The geological data strongly indicate
apossibility of linear-flow geometry on a field-wide scale. This
was confirmed by re-analyzing theresults of a multiple-well
interference test performed in 1979. Radial and linear-flow models
wereused in this process. An evaluation of computed reservoir
transmissivities and well capacities indicatedthat a linear model
fitted the interference test data significantly better than a
radial model. The linear-flow model that was developed for the
Chingshui reservoir was also instrumental in obtaining animproved
estimation of the geothermal fluid reserves (i.e., fluid-in-place).
2004 CNR. Published by Elsevier Ltd. All rights reserved.
Keywords: Geothermal reservoir; Well tests; Chingshui;
Taiwan
1. Introduction
Taiwan is located at the western rim of the Circum-Pacific
margin, one of the majorgeothermal and volcanic belts in the world.
The Taiwanese island lies on a convergent andcompression boundary
between the Philippine Sea and Eurasian Plates. The collision
ofthese two tectonic plates results in frequent earthquakes and
explains the presence of nu-
Corresponding author. Tel.: +886 6 2757575; fax: +886 6
2747378.E-mail address: [email protected] (M.C. Tom
Kuo).
0375-6505/$30.00 2004 CNR. Published by Elsevier Ltd. All rights
reserved.doi:10.1016/j.geothermics.2004.11.003
-
100 K.C. Fan et al. / Geothermics 34 (2005) 99118
Nomenclature
a formation diffusivity (m2/s)A area (m2)b width of fractured
reservoir (m)ct compressibility of fluid (Pa1)FIP fluid-in-place
(m3)h formation thickness (m)k permeability (m2)kh
permeabilitythickness product (m3)P pressure (Pa)Pi initial
pressure (Pa)P pressure change (Pa)PDl dimensionless pressure in
linear-flow modelq well volumetric flow-rate at reservoir
conditions (m3/s)qm well mass flow-rate (kg/s)Q well capacity
(kg/s)t time (s)tDb dimensionless time in linear-flow modelx
distance between the production and observation well (m)xD
dimensionless distance between the production and observation
well
Greek letters viscosity (Pa s) specific volume at reservoir
conditions (m3/kg) porosityh porositythickness product (m)
Conversion factors1 bar 105 Pa1 h 3600 s1 darcy (0.9869 1012
m2)
merous volcanoes and geothermal areas. Close to a hundred
hot-spring locations have beenidentified in Taiwan and have been
classified as volcanic or non-volcanic hot springs. Thenon-volcanic
hot springs are found in both the sedimentary province and the
metamorphicterrains of the island (Fig. 1). Table 1 summarizes the
characteristics of these two typesof hot springs in Taiwan, such as
reservoir temperature, predominant lithology, type ofpermeability,
and fluid chemistry.
The Chingshui geothermal field is located in the northeast
sector of Taiwan, in themetamorphic terrains (Fig. 1). Geothermal
exploration in this area began in 1973 (Leeet al., 1980), and
consisted of geological, geochemical and geophysical
investigations
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 101
Fig. 1. Thermal springs in Taiwan (Chen, 1985).
Table 1Characteristics of thermal springs in Taiwan
Hot-spring type Volcanic area Sedimentary province, metamorphic
terrain
Reservoir temperature 200300 C 100200 CReservoir permeability
Fractures in sandstones
and andesitesFractures in sandstones and metamorphic rocks
Major ions SO42; Cl Na+; HCO3pH 25 89
-
102 K.C. Fan et al. / Geothermics 34 (2005) 99118
(e.g., Su, 1978; Tseng, 1978; Cherng, 1979; Hsiao and Chiang,
1979; Lee et al., 1980),as well as drilling of a number of gradient
(less than 500 m deep) and deeper exploratoryand production wells
(Cherng, 1979).
Pressure buildup and interference tests were conducted for an
initial assessment of thefield in 1979 (Chang and Ramey, 1979;
Chiang et al., 1979). According to Chang and Ramey,two short,
preliminary interference tests were carried out to determine
whether detectablepressure responses were observed. A third, 11-day
long interference test was completed inNovember 1979. Its main
objective was to determine the transmissivity and coefficient
ofstorage of the reservoir, parameters that are needed to estimate
reservoir deliverability andfluid reserves.
Based on the results of these exploration and reservoir
evaluation studies, a 1.5 MWpower plant was installed at Chingshui
in October 1977 (Lee et al., 1981). The plant wasreplaced by a
larger unit (3 MW), which came on line in July 1981. During its
first yearof operation, the average power output was only 1.18 MW.
That average dropped morethan 50%, to 0.52 MW, during the third
year because of a sharp decline in the productivityof the wells. In
1993, the plant ceased operations when the average power output
wasonly 0.18 MW. Fig. 2 shows the field deliverability of
geothermal fluids and the poweroutput at Chingshui from 1981 to
1993. There was no re-injection of spent geothermalfluids.
The maximum measured reservoir temperature at Chingshui was
about 225 C.Table 2 depicts the chemistry of the geothermal fluids
produced at a flowing well-head pressure of 3.92 bars (4 kg/cm2);
their pH ranged between 8.5 and 8.8. Non-condensable gases in the
produced fluid, primarily CO2, amounted to over 10% byvolume. Scale
deposits of CaCO3, NaHCO3, and SiO2 were identified during
wellworkovers. Mineral scaling was one predominant reason for the
decline of well producti-vities.
The purpose of the study presented here is the re-interpretation
of the 1979 interferencetest data in order to calculate the
reservoir fluid-in-place, as part of a project to evaluatethe
feasibility of resuming commercial exploitation of the Chingshui
geothermal system.A conceptual linear-flow model based on
geological data was considered in the test
datare-interpretation.
Table 2Chemistry of Chingshui geothermal fluids
Water phase Concentrations (ppm)pH K+ Na+ Ca2+ Mg2+ HCO3 Co32 Cl
SO42 SiO28.8 36 1149 1.0 0.8 2768 186.0 16.0 23.6 3708.5 36 1095
0.6 0.2 2619 92.5 18.3 32.0 342
Steam phase Composition of non-condensable gases
(vol.%)Steam:non-condensable gases (vol.%) CO2 H2S
Residue89.70:10.30 97.89 0.46 1.65
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 103
Fig. 2. Production history of the Chingshui geothermal
field.
2. Geology
The Chingshui geothermal field is an area of hot springs along
the Chingshui River,approximately 13 km southwest of Lotung (Fig.
3). Dark-gray and black slates predom-inate. They are from the
Miocene Lushan Formation, which can be divided litholo-gically into
the Jentse, Chingshuihu, and Kulu members. In general, the Jentse
Mem-ber is composed mainly of metasandstones intercalated by
slates, while the underlyingChingshuihu and Kulu members consist
mostly of slates (Tseng, 1978; Chiang et al.,1979).
The Chingshui geothermal area is located on a monocline
structure, which is cut internallyby numerous thrust faults that
are lightly curved, and essentially trend NESW, parallel tothe
bedding; the most important ones are the Tashi, Hsiaonanao and
Hanhsi faults, shown inFig. 4 (Su, 1978; Hsiao and Chiang, 1979).
In the field itself, along the Chingshui River, isthe normal, NS
striking Chingshuihsi fault. Active tectonic movements most likely
created
-
104 K.C. Fan et al. / Geothermics 34 (2005) 99118
Fig. 3. Location map of the Chingshui geothermal area.
the numerous faults and well-developed fractures around the
Chingshui geothermal area.The best developed fractures in the
slates occur near the most convex part of the Hsiaonanaofault,
along the Chingshui River.
There is clear evidence that the geothermal reservoir is
fracture-dominated. As a result ofthe poor porosity and
permeability of the slates, faults, joints, and other extensive
fracturesprovide the conduits for the geothermal fluid flow. The
predominant joints, which are alignedalmost perpendicular to the
strike of the strata, are densely developed within the sandy
JentseMember. Fig. 5 shows the rose diagram for 67 joints measured
at an outcrop of the JentseMember located near the Chingshui
geothermal field (Tseng, 1978). The most prominent set
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 105
Fig. 4. Geological map of the Chingshui geothermal area showing
the Chingshuihu, Jentse, and Kulu membersof the Miocene Lushan
Formation. Triangles and rectangles indicate the up-dipped sides of
the reverse faults andthe direction of dip of the normal fault,
respectively.
of joints strikes northwest and dips between 65 and 80 to the
southwest. A less conspicuousset strikes northeast and dips steeply
northwest.
The trend of the Chingshui River runs almost parallel to that of
the joints. Its bed has cutthrough the slates, which present
well-developed fractures. There are numerous hot springsand
fumaroles along the river within the geothermal field. It is
reasonable to infer that theriver bed is the area where the major
open fractures reach the surface.
Subsurface data indicate that geothermal production at Chingshui
comes largely from afracture zone within the steeply dipping Jentse
Member (Hsiao and Chiang, 1979). Structuralanalyses show that this
member presents predominant, well-developed, steeply dippingjoints
that strike between N 25W and N 40W. Outcrops near the area of the
thermalmanifestations also reveal that the faults run parallel for
almost 100150 m, striking betweenN 30W and N 35W (Tseng, 1978).
-
106 K.C. Fan et al. / Geothermics 34 (2005) 99118
Fig. 5. Rose diagram for 67 joints in the Chingshui geothermal
area (Tseng, 1978).
3. Well drilling, completion and development
The drilling fluid used at Chingshui was bentonite slurry
treated with chrome lignosul-fonates. In order to prevent wellbore
cave-ins, the mud was always maintained at a specificgravity of
1.10 to 1.25 and at a Marsh funnel viscosity of 40 s. Heavy
circulation lossesoccurred when the drill bit penetrated major
fracture zones. During completion, fresh waterwas injected into the
well to wash out the drilling mud and remaining drill cuttings.
The casing program for the production wells was: 50.8 cm (20
in.) conductor, 34 cm(13 38 in.) surface casing, 24.4 cm (9 58 in.)
production casing, and 17.8 cm (7 in.) or 11.4 cm(4 12 in.) slotted
liner. The slotted liners hung between 490 and 1048 m depth,
depending uponthe depth of the high-temperature production zone;
the length of the liners varied between950 and 2160 m. All casings,
except the liners, were cemented. Table 3 summarizes the
com-pletion and capacity data for all the Chingshui production
wells (i.e., 4T, 5T, 9T, 12T, 13T,14T, and 16T). Well-production
capacities were measured using the James method (1966).
4. Interference test
Identification of the prevalent reservoir fluid flow model is a
pre-requisite for reachinga correct evaluation of well test data,
and thus of fluid-in-place, and for developing drillingand field
management programs.
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 107
Table 3Well capacity and completion data for wells in Chingshui
geothermal reservoir
Well Elevation(m asl)
Well capacity Well completion
Steam rate(l03 kg/h)
Hot water rate(l03 kg/h)
Total flow-rate(l03 kg/h)
Total depth(TD) (m)
Depth ofliner shoe(m)
Depth oflinerhanger (m)
Temperatureat TD (C)
4T 257.95 28.6 98.1 126.7 1505 1503 498 2015T 269.54 6.0 34.0
40.0 2005 1998 493 2209T 260.67 18.7 55.3 74.0 2079 2074 490 20512T
260.67 6.9 40.0 46.9 2003 1998 1048 22313T 269.54 10.4 60.1 70.6
2020 2015 505 21914T 281.50 22.0 66.0 88.0 2003 1995 947 21516T
272.58 30.3 85.9 116.2 3000 2990 830 225
In this paper we describe a new interpretation of the data of
the November 1979 wellinterference test. In the original
interpretation by Chang and Ramey (1979) the authorsassume a radial
flow model. We will show that the use of a linear-flow model is
moreappropriate for the Chingshui reservoir.
As mentioned earlier, the geothermal reservoir presents
well-developed joints and faults.All the wells drilled into the
Jentse Member, the producing formation, have rather high-angle
inclinations of up to 35, and all are deviated almost parallel to
the joints observedat the surface (Hsiao and Chiang, 1979). The
deviations in some wells changed abruptly,turning sharply back in a
direction almost parallel to the joints, with loss of mud
circulationwhenever a highly productive fractured zone was
encountered.
On the basis of the trend of the fractures in the production
zone, a linear-flow model shouldmore accurately represent the
geothermal reservoir and the flow regime between wells thana radial
flow model, which assumes either primary porosity or a random
distribution andorientation of joints and fractures.
During the 1979 test, well 16T was put into production, and
pressure responses wereobserved in wells 4T, 5T, 9T, 12T, 13T, and
14T (Chang and Ramey, 1979). Fig. 6 showsboth surface and
bottom-hole locations of these wells. Since the drill bit in all
the wellsdrifted in the direction of the geologic structures, the
distance between the bottom-holelocations had to be estimated in
order to interpret the interference test data, the distancesbetween
wells corresponding to the distance between pairs of feed
zones.
The data relative to the 11-day interference test are presented
in Table 4. Hot waterproduction rate, measured in a weir, ranged
from 80,000 to 83,500 kg/h during thetest. The total fluid (water +
steam) production rate was calculated from the hot waterproduction
rate using energy-balance considerations for flashing water. For
the 11-dayinterference test, the wellhead pressure, water
production rate and total fluid produc-tion rate of flowing well
16T stabilized at 3.59 bars, 80,000 kg/h, and 105,000
kg/h,respectively. Wellhead pressures were monitored at all the
observation wells, but datafrom wells 5T and 13T appear to be
unreliable because of equipment malfunction. Thetheory and
interpretation of the interference test results will be presented
in Sections5 and 6.
-
108K.C.F
anetal./G
eothermics34(2005)99118
Table 4Interference test in Chingshui geothermal field (Chang
and Ramey, 1979)Time (h) Observation wells Flowing well
4T 9T 12T 14T 16 T
WHPa (bar) Pb (bar) WHP (bar) P (bar) WHP (bar) P (bar) WHP
(bar) P (bar) WHP (bar) Hot waterrate(103 kg/h)
Total fluid rate(103 kg/h)
0.0 11.86 0.00 9.51 0.00 12.89 0.00 9.17 0.00 17.79 0.0 0.018.5
11.79 0.07 9.45 0.07 12.76 0.14 9.17 0.00 4.76 24.0 30.842.5 11.58
0.28 9.31 0.21 11.17 0.34 8.96 0.21 4.00 83.5 108.766.5 11.45 0.41
9.17 0.34 12.55 0.34 8.62 0.55 3.86 83.1 108.490.5 11.45 0.41 8.96
0.55 12.41 0.48 8.62 0.55 3.86 83.1 108.4
114.5 11.38 0.48 8.96 0.55 12.34 0.55 8.48 0.69 3.86 82.0
107.0138.5 11.31 0.55 8.96 0.55 12.27 0.62 8.34 0.83 3.86 82.4
107.5162.5 11.31 0.55 8.89 0.62 12.20 0.69 8.27 0.90 3.72 82.4
107.8186.5 11.24 0.62 8.83 0.69 12.13 0.76 8.20 0.97 3.72 81.0
106.0210.5 11.17 0.69 8.76 0.76 12.07 0.83 8.20 1.03 3.65 80.0
104.8234.5 11.17 0.69 8.76 0.76 12.07 0.83 8.07 1.10 3.59 80.0
105.0258.5 11.10 0.76 8.69 0.83 12.07 0.83 7.93 1.24 3.59 80.0
105.0
a WHP: wellhead pressure.b P: pressure change.
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 109
Fig. 6. Well locations and inferred temperature distribution (C)
at 1500 m depth in the Chingshui geothermalarea (from Chang and
Ramey, 1979).
5. Theory
There are a number of books and reports discussing the theory
and practice of ana-lyzing data from tests performed in different
types of wells (groundwater, oil and gas,geothermal). The test data
can be analysed by means of curve fitting techniques or
-
110 K.C. Fan et al. / Geothermics 34 (2005) 99118
computer-aided approaches (e.g., McLaughlin et al., 1995; Horne,
1995; OSullivan et al.,2005).
5.1. Radial ow model
The most commonly used analytical solutions for interpreting an
interference test isthe Theis solution (Theis, 1935) and the line
source solution (van Everdingen and Hurst,1949), for use in
groundwater and petroleum engineering, respectively. The line
sourcesolution corresponds to an infinite-acting, isotropic
reservoir, and assumes a constantproduction/injection rate in which
only one (liquid) phase is involved. This is themodel used by Chang
and Ramey (1979) in their analysis of the 1979 interference
testdata.
5.2. Linear-ow model
We have developed a conceptual linear-flow model of the
Chingshui geothermal reser-voir, based on the geological data of
the area (see previous sections). Fig. 7 is a sketch ofthe
linear-flow model in which the geothermal reservoir is represented
by a parallelepiped.Fluid flow is parallel to the main strike of
the joints and the lateral boundaries of the prism.The
cross-section of the parallelepiped is assumed to be a rectangle
with a height h anda width b. The production well is represented by
a planar source. The diffusivity equa-
Fig. 7. Sketch of a linear-flow model for a fractured
reservoir.
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 111
tion governing fluid flow in an infinite linear reservoir for
the constant-rate case is givenby:
P
t= a
2P
x2(1)
subject to the following initial and boundary conditions:t = 0,
P = Pi for x 0
and
t > 0,P
x= q
2bhkfor x = 0 and lim
xP(x, t) = Pi for x
where a= k/ct, the formation diffusivity.Miller (1960)
investigated unsteady influx of water in linear reservoirs for the
constant-
rate case and an infinite-acting reservoir using the equations
given above. Miller adaptedthe Carslaw-Jaeger (1959; p. 75)
solution of heat conduction to pressure drawdown in
linearreservoirs as follows:
p(x, t) = pi q2khb
[2at
exp
( x
2
4at
) x erfc
(x
2at
)](2)
Millers solution [Eq. (2)], is similar to that developed by
Jenkins and Prentice (1982) toanalyze aquifer tests in fractured
rocks assuming linear-flow conditions. The Miller solutionwas
previously applied to a steam reservoir for interference analysis
(Ehlig-Economideset al., 1980). To apply Millers solution to a
fractured hot water reservoir using SI units, thefollowing
dimensionless variables are defined:
PDl = 2khPq
(3)
xD = xb
(4)
tDb = khthctb2
(5)
where k is permeability (m2); h is formation thickness (m); P is
pressure change (Pa); q isvolumetric well flow-rate at reservoir
conditions (m3/s); is viscosity (Pa s); x is distancebetween the
production and observation wells (m); b is width of the fractured
reservoir (m); is porosity; ct is compressibility of fluid (Pa1);
and t is time (s). The well volumetricflow-rate at reservoir
conditions can be calculated by multiplying the well mass
flow-ratewith the specific volume of geothermal fluid at reservoir
conditions, or, q= qm where qmis well mass flow-rate (kg/s); and is
specific volume of geothermal fluid at reservoirconditions
(m3/kg).
-
112 K.C. Fan et al. / Geothermics 34 (2005) 99118
Table 5Type-curve matching using van Everdingen and Hurst (1949)
solutionaRadial flow model Observation wells
4T 9T 12T 14T
Match pointPD (P= 0.6895 bars) 0.95 0.77 0.78 0.43tD/r
2D (t= 100 h) 1.5 0.95 1.2 0.75
Distance (m) 175 300 90 330kh (darcy m) 9.24 7.49 7.59 4.18h (m)
425 185 1650 108
a qm = 105,000 kg/h, = 1.188 103 m3/kg, = 0.12 103 Pa s, ct =
1.45 104 bar1.
Eq. (2) can then be written in terms of the dimensionless
variables defined above:
PDlxD
= 2
tDb
x2Dexp
( x
2D
4tDb
) erfc
(xD
2tDb
)(6)
Eq. (6) can be used to calculate a loglog type-curve, PDl/xD
versus tDb/x2D, for linear-flow.If practical units (i.e., darcy,
bar, and hour) are used instead of SI units, Eqs. (3) and (5)
become:
PDl = 0.0007106khPq
(3a)
tDb = 0.0003553khthctb2
(5a)
The CarslawJaegerMiller solution assumes that one half of the
produced fluid comesfrom each side of the production plane/well;
for this reason, Fig. 7 shows only one side fora production well
along the direction of fluid flow and the main strike of the
joints.
6. Interpretation of interference test data
Fig. 8 is a match of the well 4T pressure versus time data
against the vanEverdingenHurst radial flow solution. Fig. 9 shows
similar matches for wells 9T, 12Tand 14T. Table 5 summarizes the
curve matching results for all the well pairs assuming theradial
flow model.
Fig. 10 is a match of the well 4T pressure versus time data
against the Miller linear-flowsolution. Fig. 11 shows the match for
wells 9T, 12T and 14T. The linear-flow type-curvematching results
for all pairs are given in Table 6.
As can be seen from Tables 5 and 6, porositythickness products
obtained from the radialflow model varied over a wide range, from
108 to 1650 m, while the range obtained fromthe linear-flow model
was significantly smaller (i.e., 279956 m). To apply the
linear-flowmodel, the width of the Chingshui geothermal reservoir
was estimated to be around 300 m,
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 113
Fig. 8. Well 4T. Type-curve match using the radial flow
model.
Table 6Type curve matching using Miller (1960)
solutionaLinear-flow model Observation wells
4T 9T 12T 14T
Match pointPDl/xD (P= 0.6895 bars) 4.8 1.6 2.9 1.4tDb/x2D (t=
100 h) 13 3 7.5 3.4
Distance (m) 175 300 90 330kh (darcy m) 85.5 48.9 26.6 47.1h (m)
468 396 956 279
a qm = 105,000 kg/h, = 1.188 103 m3/kg, = 0.12 103 Pa s, ct =
1.45 104 bar1, b= 300 m.
based on the bottom-hole locations of the production zones of
the seven wells (4T, 5T, 9T,12T, 13T, 14T, and 16T) shown in Fig.
6.
7. Discussions
The main objective of the study of the Chingshui field was to
identify a model thatmost appropriately describes the geology and
flow behavior of the fractured geothermal
-
114 K.C. Fan et al. / Geothermics 34 (2005) 99118
Fig. 9. Wells 9T, 12T, and 14T. Type-curve match using the
radial flow model.
reservoir. A realistic model is an important tool in our
analysis of interference test data andin estimates of the
geothermal fluid-in-place.
The capacity of a geothermal well depends on a number of
factors, such as permeability-thickness product, well skin, and
well completion. According to Darcys law, undersimilar conditions
of well completion and skin, well capacity is proportional to
thepermeabilitythickness product.
In our interpretation of the interference test data, we applied
both the radial and the linear-flow model, and compared their
results. The permeabilitythickness products estimatedwith these two
models were compared and correlated to well capacities in order to
selectthe appropriate model. As shown in Table 7, the kh products
estimated using the linear-flow
Table 7Comparison of well capacity with permeabilitythickness
product
Well number Well productivity totalfluid rate, Q (10 3kg/h)
Permeabilitythickness product, kh (darcy m)Linear-flow model
Radial flow model
4T 126.7 85.5 9.249T 74 48.9 7.4912T 46.9 26.6 7.5914T 88 47.1
4.18
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 115
Fig. 10. Well 4T. Type-curve match using the linear-flow
model.
model appeared to correlate with well capacities, whereas this
was not the case for the radialmodel.
Figs. 12 and 13 illustrate a least-square fit of well capacities
versus permeabilitythickness products for the radial and
linear-flow models, respectively. The small valueof the sample
correlation squared regression coefficient (i.e., R2 =0.1362)
correspond-ing to the radial model (Fig. 12) indicates that the two
parameters are not well correlated.On the other hand, the same
coefficient is quite high (i.e., R2 = 0.9146) for the
linear-flowmodel (Fig. 13). In this case, the regressed equation
would be quite useful in predictingwell capacity from the
permeabilitythickness product, i.e.,
Q = 1574kh (7)where Q is well capacity (kg/h) and kh is the
permeabilitythickness product (darcy m).
Because the thickness of Chingshui geothermal reservoir is not
known, we were unableto separate the average permeabilitythickness
product of 52 darcy m (or the averageporositythickness product of
525 m) into effective permeability (or porosity) and net thick-ness
with accuracy. However, a range of thickness may be considered to
estimate effectivepermeability and porosity values. For example, if
the net reservoir thickness is assumed to be2500 m, then the
effective permeability and porosity would be approximately 21
millidarcysand 0.20, respectively, for this liquid-dominated
system. A porosity of this order of mag-
-
116 K.C. Fan et al. / Geothermics 34 (2005) 99118
Fig. 11. Wells 9T, 12T, and 14T. Type-curve match using the
linear-flow model.
Fig. 12. Regression lines between well capacities and
permeabilitythickness products obtained using the radialflow
model.
-
K.C. Fan et al. / Geothermics 34 (2005) 99118 117
Fig. 13. Regression lines between well capacities and
permeabilitythickness products obtained using the linear-flow
model.
nitude appears high, so the indication is that the net reservoir
thickness may exceed2500 m.
The main importance of the porositythickness product obtained
from interference test-ing is that it can provide an estimate of
the fluid-in-place. Based on the isotherms map(Fig. 6), the area of
the Chingshui geothermal reservoir is estimated to be around 2
km2.The fluid-in-place can be calculated by a volumetric method
using the following equation
FIP = hA (8)where FIP is fluid-in-place (m3); h is
porositythickness product (m); and A is area (m2).Therefore, the
fluid-in-place for Chingshui geothermal reservoir is slightly
higher than109 m3 (i.e., 525 m 2 km2).
8. Conclusions
The application of a linear-flow model in the interpretation of
data from an interfer-ence test carried out in the Chingshui
geothermal field was discussed. The analysis showedthat the data
can be matched to type curves for either the radial or the
linear-flow model.However, when the values of permeabilitythickness
products are correlated with the wellproductivities, the
correlation is poor for the radial flow model and very good for
thelinear model. As initially indicated by geologic and well
drilling information, the in-terference data analysis confirms that
the flow regime in the reservoir is predominantlylinear.
The amount of reservoir fluid-in-place is significant (more than
109 m3); however, un-less scaling is controlled, the long-term
electricity generating capacity of the Chingshui
-
118 K.C. Fan et al. / Geothermics 34 (2005) 99118
geothermal system may remain at the 12 MW level. Scaling control
and re-injection ofspent geothermal liquids would help increase the
productive lifetime of the geothermal field.
Acknowledgments
This research was funded by the Energy Commission, Ministry of
Economic Affairsand National Science Council of Taiwan
(NSC-93-2623-7-006-006-ET). The authors aregrateful for the
valuable comments and suggestions of the reviewers, especially Dr.
SabodhGarg and Dr. Marcelo Lippmann, who helped to improve the
paper.
References
Carslaw, H.S., Jaeger, J.C., 1959. Conduction of Heat in Solids,
2nd ed. Oxford, University Press.Chang, C.R.Y., Ramey, H.J., 1979.
Well interference test in the Chingshui geothermal field. Paper
presented at
Fifth Geothermal Reservoir Engineering Workshop, Stanford
University, Stanford, CA, pp. 6476.Chen, C.H., 1985. Chemical
characteristics of thermal waters in the central range of Taiwan.
R.O.C. Chem. Geol.
49, 303317.Cherng, F.P., 1979. Geochemistry of the geothermal
fields in the slate terrane. Geotherm. Resour. Counc. Trans.
3, 107111.Chiang, S.C., Lin, J.J., Chang, C.R.Y., Wu, T.M.,
1979.A preliminary study of the Chingshui geothermal area, Ilan
Taiwan. Paper presented at Fifth Geothermal Reservoir
Engineering Workshop, Stanford University, Stanford,CA, pp.
249254.
Ehlig-Economides, C., Economides, M.J., Miller, F.G., 1980.
Interference between wells in a fractured formation.Geotherm.
Resour. Counc. Trans. 4, 321324.
Horne, R.N., 1995. Modern Well Test Analysis, 2nd ed. Petroway
Inc, Palo Alto, CA.Hsiao, P.T., Chiang, S.C., 1979. Geology and
geothermal system of the Chingshui-Tuchang geothermal area,
Ilan,
Taiwan. Petrol. Geol. Taiwan 16, 205213.James, R., 1966.
Measurement of steam-water mixtures at the speed of sound to the
atmosphere. N. Z. Eng. 20
(1), 437441.Jenkins, D.N., Prentice, J.K., 1982. Theory for
aquifer test analysis in fractured rocks under linear
(nonradial)
flow conditions. Ground Water 20 (1), 1221.Lee, C.R., Lee, C.F.,
Cheng, W.T., 1980. Application of roving bipole-dipole mapping
method to the Chingshui
geothermal area, Taiwan. Geotherm. Resour. Counc Trans. 4,
7376.Lee, C.S., Chang, P.T., Hsu, J.B., 1981. A model approach for
the geothermal fields in slate terrain of Taiwan.
Geotherm. Resour. Counc. Trans. 5, 181184.McLaughlin, K.L.,
Barker, T.G., Owusu, L.A., Garg, S.K., 1995. DIAGNS: An interactive
workstation-based
system for well test data diagnostics and inversion. Proceedings
World Geothermal Congress, Florence, Italy,vol. 4, pp.
29412944.
Miller, F.G., 1960. Theory of unsteady-state influx of water in
linear reservoirs. Report. Stanford University,Stanford, CA, p.
45.
OSullivan, M.O., Croucher, A.E., Anderson, E.B., Kikuchi, T.,
Nakagome, O., 2005. An Automated Well TestAnalysis System (AWTAS).
Geothermics 34, 325.
Su, F.C., 1978. Resistivity survey in the Chingshui prospect,
I-lan, Taiwan. Petrol. Geol. Taiwan 15, 255264.Theis, C.V., 1935.
The relation between the lowering of the piezometric surface and
the rate and duration of
discharge of a well using ground-water storage. Trans. Am.
Geophys. Union 16, 519524.Tseng, C.S., 1978. Geology and geothermal
occurrence of the Chingshui and Tuchang districts, Ilan. Petrol.
Geol.
Taiwan 15, 1123.van Everdingen, A.F., Hurst, W., 1949. The
application of the Laplace transformation to flow problems in
reservoirs.
Trans. AIME 186, 305324.
Interpretation of a well interference test at the Chingshui
geothermal field, TaiwanIntroductionGeologyWell drilling,
completion and developmentInterference testTheoryRadial flow
modelLinear-flow model
Interpretation of interference test
dataDiscussionsConclusionsAcknowledgmentsReferences