Geothermics 39 (2010) 382–390

Contents lists available at ScienceDirect

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

Sustainability of the Hatchobaru geothermal field, Japan

Tetsuya Yahara, Hiroyuki Tokita ∗
Geothermal Department, West Japan Engineering Consultants, Inc., Fukuoka 810-0004, Japan

a r t i c l e i n f o a b s t r a c t

Article history: The Hatchobaru power plant Unit No. 1 (55 MW) has been operating since 1977 and Unit No. 2 (55 MW)
Received 29 May 2009 since 1990. The mean capacity factor of the power plant has reached about 90%. Considering that the
Accepted 1 September 2010 long-term operation of the plant, over 30 years for Unit No. 1 and nearly 20 years for Unit No. 2, has been
Available online 27 October 2010
maintained with such a high capacity factor, sustainable development in terms of economic production
has been achieved. To maintain a stable operation, systematic reservoir monitoring and reservoir simu-
Keywords:
lation studies have been conducted. The monitoring of changes in reservoir pressure, temperature and
Geothermal
gravity indicates that the reservoir is currently approaching a stable state. Results of a simulation study
Hatchobaru
Reservoir monitoring
suggest that the sustainable power output of the Hatchobaru reservoir is approximately 120 MW, and
Reservoir simulation each productive fault has the capacity to produce enough steam to generate from 11 to 55 MW. Therefore,
Reservoir management it would be possible to maintain the rated power output of 110 MW by optimizing well alignments so
Sustainability that the mass production can be kept within the sustainable productivity of each fault, and the injected
Japan water does not cool the production zones.
© 2010 Published by Elsevier Ltd.

1. Introduction two-phase pipelines from wellheads to independent separators

located in the plant area. The two pipeline networks are connected
The Hatchobaru geothermal field is located in central Kyushu, near the power plant to allow part of the fluids from the wells sup-
Japan. The power plant started commercial operation in 1977 (Unit plying Unit No. 2 to flow into the pipeline network for Unit No.
No. 1, 55 MW), and in 1990, its total capacity was expanded by 1. Unit No. 1 has been in operation for 32 years and Unit No. 2
adding a second unit of 55 MW. The total rated power output of for 19 years. The mean capacity factor of the power plant reached
110 MW makes Hatchobaru the largest geothermal plant in Japan; about 88% in fiscal year 2007, indicating stable plant operation. Con-
its double-flash system was the first one in the world. The primary sidering the development area of the Hatchobaru field, the power
turbine inlet pressures are 0.6 MPa and 0.7 MPa for Units 1 and 2, density is estimated to be around 37 MW/km2 , which is much
respectively. Secondary pressure at the flasher is about 0.1 MPa. higher than many other geothermal fields in the world (Tokita et
Wellhead pressure of production wells averages 0.8 MPa. Separated al., 2006b).
water and excess condensate are injected in to the reservoir at
about 90 ◦ C and at atmospheric pressure. 2. Reservoir management
Fig. 1 depicts the Hatchobaru geothermal field. The reservoir is
water-dominated with temperatures in the range of 240–300 ◦ C From 1990 to 1994 i.e., within 4 years of the commissioning
and the underground fluid movement is strongly controlled by of Unit No. 2 a significant power decline from 110 MW to 80 MW
faults. The fluids supplied to the power facilities are tapped from was experienced (see Fig. 2). After the commencement of Unit
depths of between 1000 m and 2300 m. These fluids reside in five No. 2 operation, the trend of power output can be classified into
main productive faults: Komatsuike fault, Komatsuike sub-fault, three periods: (A) a decline caused by the decrease of mass flow
NE3 fault, NE4 fault and Hatchobaru fault. The total mass extrac- from production wells and lower capacity of the injection wells,
tion rate from the reservoir (20 production wells) is currently about (B) a recovery caused by the drilling of additional production and
700 kg/s and the rate of total mass injection (15 injection wells) is injection wells, and (C) a stable period (Saito et al., 2006). In the
about 500 kg/s. Currently, the northern sector of the field is used decline period, integrated reservoir management consisting of sys-
for injection and the southern sector for production. The fluids tematic reservoir monitoring and a reservoir simulation study was
necessary to drive each of the units are transported through two undertaken in order to define the best approach for recovering and
maintaining the power output.
Fig. 3 shows the history of integrating reservoir simulation tech-
∗ Corresponding author. Tel.: +81 92 781 2835; fax: +81 92 751 5044. nologies and systematic monitoring for reservoir management that
E-mail address: [email protected] (H. Tokita). has been adopted for more than 20 years in the Hatchobaru field to

0375-6505/$ – see front matter © 2010 Published by Elsevier Ltd.

doi:10.1016/j.geothermics.2010.09.001
 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390 383

Fig. 1. The Hatchobaru geothermal field; the filled square in the inset shows the location of the geothermal field. The contours in the background denote the local ground
surface elevation in meters above sea level.
Figure modified from Yahara et al. (2006).

To predict the future power output of the plant, a numeri-

cal simulation technology has been developed in which reservoir,
wellbore, and transportation pipeline models are coupled. Fluid
transport and changes in the characteristics of the reservoir are
calculated using the reservoir simulator TOUGH2 (Pruess, 1991;
Pruess et al., 1999). The reservoir simulator is coupled to a multifeed
wellbore simulator (MULFEWS) that calculates fluid flow within the
wellbore and reservoir assuming radial flow from a homogeneous
porous reservoir around each of the feed zones (Tokita and Itoi,
2004; Tokita et al., 2005b). These two simulators are then coupled
to a two-phase pipeline flow simulator (TPGS) that calculates pres-
sure drops within the pipeline network from wellheads to main
Fig. 2. Power output at the Hatchobaru power plant. separators (Lima et al., 2004; Tokita et al., 2005a, 2006b). Mass
Figure modified from Thermal and Nuclear Power Engineering Society (2008). and energy transport can therefore be calculated from the reser-
voirs to the turbines to predict the future performance. In addition,
maintain the rated power output. Reservoir management is based this can improve the accuracy of forecasts compared to models
on reservoir monitoring and computer simulations that are con- that couple only the reservoir and wellbore simulators, because
ducted periodically to recalibrate the reservoir model and to ensure the effect of changes in wellhead pressures of production wells on
an accurate forecast of future performance of the reservoir and its pressure distribution in pipelines can be calculated. The coupled
production characteristics. approach to reservoir simulation has contributed to a stable power
plant operation and to a sustainable reservoir management.

3. Reservoir monitoring

To aid in reservoir management, monitoring of changes in reser-

voir pressure, temperature, gravity, chemical characteristics of
discharged fluids and tracer behavior has been periodically con-
ducted. Monitoring of gravity changes has been carried out since
1990 covering most of the Hatchobaru exploitation area. Gravity
measurements are carried out using a Sintrex CG-3 gravimeter
(Tagomori et al., 1996). Fig. 4 shows downhole pressure measured
at observation well 2H-11, and gravity measured at the gravity
monitoring point GM36. The well 2H-11 is located in the south-
ern part of the production zone, and the gravity monitoring point
GM36 in the southeastern part of the production zone. It can be
Fig. 3. Reservoir management for the Hatchobaru field (Tokita et al., 2006b). observed that both the pressure and gravity gradually decreased
384 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390

Fig. 5. Changes over time in silica temperature and chloride concentration in fluids
produced from well H-13.
Figure modified from Asai (2002).

center of the production zone for Unit No. 1. After the commence-
ment of Unit No. 2 operation, the silica temperature decreased by
40 ◦ C within 5 years, and the chloride concentration increased by
500 mg/l due to the increasing migration of injected water into
the well. Since then, the silica temperature has approached a sta-
ble value, and the chloride concentration has gradually decreased.
The decrease of the chloride concentration suggests that shallower
Fig. 4. Comparison between pressure changes in 2H-11 observation well and gravity water of low chloride concentration around the well migrated into
changes at GM36 (Saito et al., 2006). the well; the migration was accelerated by the pressure drawdown
that occurred after the commencement of Unit No. 2. Relocation
of injection wells and drilling of make-up production wells in the
after the commencement of the Unit No. 2 operation, with the southeastern area where pressure drawdown and injected water
changes becoming more moderate around 2000. Good correlations returns occurred helped in recovering the power output of the
are observed between the changes in pressure and gravity. It is plant. The power output since 1997 has been above 90 MW.
considered that the gravity is affected by changes in mass, which Design and development of a proper injection system that min-
occur due to production and injection operations (Saito et al., 2006). imizes the temperature decline in production wells and secures
Reservoir pressure is also related to mass inflows into and outflows the injection capacity is of great importance for sustainable plant
from the geothermal reservoir. Downhole monitoring of changes in operation. Investigation and monitoring have been conducted at
the reservoir pressure and temperature has been conducted contin- both newly drilled and operating production and injection wells.
uously since 1988 using capillary tubing systems that are installed For instance, analysis of cuttings, temperature logs, injection tests,
in several observation wells. Results of pressure monitoring indi- production logs, and tracer tests have been conducted for newly
cate that a complicated change in pressure usually occurs due to drilled wells. In addition, the monitoring of wellhead pressures and
fluid production and injection during the plant operation (Tsuru two-phase flow rates of production wells, flow rates of injection
and Tokita, 1994). A pressure drawdown of more than 0.5 MPa in wells, downhole pressures and temperatures in observation wells,
the production zones for Unit No. 2 was detected soon after this steam flow rates, water flow rates and specific enthalpies of dis-
unit commenced commercial operation. The reservoir pressure also charged fluids and chemical analyses of gas, steam and water have
declined by 0.4 MPa in the production zones for Unit No. 1, and been conducted for the operating wells (Yahara et al., 2006). These
propagated into the corresponding injection zones. data provide the basic information for diagnosing the causes for
Although the reservoir temperature had already decreased as a declines in productivity or injectivity of wells.
consequence of the injection since Unit No. 1 commenced opera- Productivity declines of wells were analyzed to understand
tion, the temperature decrease accelerated after the commissioning the causal mechanisms. For example, the pressure interference
of Unit No. 2 due to the increased return of injected water to the pro- between production wells tapping the reservoir along the Hatcho-
duction zones of both units. Results of tracer tests indicate that the baru fault was induced by overproduction (Yahara et al., 2008).
injected water partially returns to the production zones through Another example is that periodic fluctuations in mass productivity
faults, and that the rapid return from some injection wells is the that occurred in the production well H-30 located near the injec-
cause of the steep temperature decline. tion zones were caused by an increase in injection flow rate; this
The temporal and spatial changes in reservoir temperature indicated the need for a reduction in the injection flow rate (Yahara
are estimated by applying the silica, Na–K and Na–K–Ca geother- et al., 2005).
mometers and the results of chemical analyses of discharged fluids After a decline in power output started in 1990, injection wells
sampled from production wells every 6 months (Asai, 2002). The were drilled in the reservoir along the Komatsuike fault away
increase in chloride (Cl) concentration in the discharged fluid from the production area to minimize the influence of injected
also suggested that the decline in temperature (and hence power water upon the production wells. In addition, side-track drilling
output) was due to a large inflow of injected water of high Cl con- of injection wells that lost injectivity due to silica deposition in
centration into the production zone. Fig. 5 shows changes over time the formation was carried out for recovery of injection capacity.
in silica temperature and Cl concentration of the fluids discharged The injection capacity, however, did not increase because of pres-
from well H-13 that is located along the Komatsuike-sub fault in the sure interference between injection wells. The change of water
 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390 385

southwestern margin of the Kuju-Beppu graben, which is a low-

gravity anomaly zone. However, the graben is characterized by
positive gravity anomalies in some areas, reflecting the existence of
horst structures within it (Tamanyu, 1993). The Hatchobaru area is
located in a positive gravity anomaly zone, which is a horst adjacent
to the Shishimuta depression zone. The geologic column is divided
into four units, Kuju volcanic rocks (Middle-Upper Pleistocene),
Hohi volcanic rocks (Lower Pleistocene), Usa group (Miocene) and
basement rocks (pre-Tertiary), in descending order (Fujino and
Yamasaki, 1985).
The interpretation of fluid isotope data indicates that the
geothermal fluids are derived almost entirely from meteoric water.
The recharge zone is considered to be located around the south part
of Mt. Kuroiwa. Here, the meteoric water flows downward through
Fig. 6. Change of water level in the injection area (Yahara et al., 2006).
the faults and through the related high-angle fracture networks
associated with the boundary rim of the Shishimuta depression
level in the injection wells after drilling or workover is shown in zone. The meteoric water permeating into the deeper zone changes
Fig. 6. its chemical composition through water/rock interactions and is
The water level in the well tapped in the reservoir along the heated up to about 300 ◦ C by a heat source related to the Quaternary
Komatsuike fault began rising in 1992 and reached its highest point volcanism. The heat source of the area is considered to correspond
in 2002. This rise of water level implies that the injection flow rate to to a residual magma chamber beneath Mt. Kuroiwa, where the heat
this reservoir exceeds its injection capacity. Therefore, it was con- source produces enough conductive heat to increase the tempera-
cluded that to secure injection capacity, other injection zones that ture at the top of the granitic basement to 300–350 ◦ C (Momita et
did not affect the existing injection zones needed to be developed. al., 2000). The heated fluids then migrate upward through the per-
Based on this idea, side-track drilling of injection wells that targeted meable passages associated with the Komatsuike fault (Shimada
deeper zones along the Komatsuike fault began in 2003. The use et al., 1994; Shimada and Fujino, 1995). Several isotope indices
of these side-tracked boreholes has led to a water level decline of such as B/Cl, As/Cl and 87 Sr/86 Sr suggest that the deep geother-
about 100 m in the original reservoir, confirming the effectiveness mal fluids ascend through the Tertiary Hohi volcanic rocks and the
of the strategy for securing injection capacity. Usa group which overlie the granitic basement. Other geothermal
reservoirs within the granitic basement have been confirmed by
4. Reservoir model the wells drilled in the southwestern part of the Hatchobaru area.
The B/Cl and 87 Sr/86 Sr ratios are quite distinct from those of the
Fig. 7 shows a model of the geothermal system of the Hatcho- above-mentioned fluids. There is sufficient evidence to suggest that
baru geothermal field. The Hatchobaru field is situated in the the parental fluid of the present production zones in the Hatcho-

Fig. 7. Model of the Hatchobaru Geothermal Field (Momita et al., 2000).

386 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390

Fig. 8. Plan view of the grid used to model the Hatchobaru Geothermal Field (Tokita et al., 2006a).

baru field is stored in deep zones beneath Mt. Kuroiwa. However, as it migrates toward production zones at deep levels. Compar-
we believe that this deep reservoir, even lying within the granitic ing the simulated temperature distributions in 1998 and 2005, it
basement rocks, supplies the hydrothermal system with hot flu- appears that the degree of cooling in the production zone has been
ids through migration along the Hohi volcanic rocks and Usa group mitigated.
which are overlying the granitic basement rocks (Momita et al.,
2000). 5. Reservoir sustainability
The numerical model of the reservoir covers 16.5 km2
(3.70 km × 4.45 km) as shown in Fig. 8. In the vertical direction, When the power output declined significantly after commis-
the top and bottom of the numerical model correspond to eleva- sioning Unit No. 2, several numerical case studies were conducted
tions of about 1100 m above sea level and 1400 m below sea level, using the reservoir simulator to understand the reason for this rapid
respectively. The 2500 m of total model thickness is divided into power decline, and to evaluate the sustainable power output. The
nine layers with layer thicknesses ranging from 100 to 400 m. The results suggest that the average pressure drawdown in the produc-
top two layers are designed to represent the topography. The total tion zones for Unit No. 2 and Unit No. 1 is 0.9 MPa and 0.5–0.7 MPa,
number of grid blocks is 7425. The boundary conditions and dis- respectively. These pressure changes occurred within 3 years of
tribution of rock properties such as density, porosity, permeability, Unit No. 2 commissioning, because of an increase of the mass
etc. were varied in a trial-and-error calibration process to repro- extraction from 420 kg/s to 560 kg/s (Kawazoe and Tokita, 1993).
duce the temporal pressure and temperature changes during plant In addition, the model showed that to mitigate cooling effects, the
operation. In addition, temporal changes in the fluid density and the most appropriate location for injection would be a distance of more
fluid behavior in the model were constrained by the corresponding than 500 m from the production zones (Tokita et al., 1995) There-
changes in gravity over time and tracer behavior. The fracture mod- fore, injection was relocated to the northwest to maintain sufficient
eling technique, called the Multiple Interacting Continua (MINC) distance from production zones. The power output was predicted
method (Pruess, 1991), was applied for modeling faults represent- to decline to around 80 MW within 8 years if make-up production
ing the dominant fluid flow pattern. wells were not drilled. The mass flow rate tapped from the NE3
Fig. 9 shows simulated temperature distributions and the zone fault was predicted to decline and the corresponding power output
containing injected water for years 1984, 1991, 1998 and 2005. to fall from 43 MW at that time to around 24 MW before attaining
This figure indicates that the injected water zone expanded dra- stabilization, while that for other four productive faults showed
matically after the commissioning of Unit No. 2 in 1990, and the only a slight decline. Thus, the sustainable power potential of the
size of the zone increased further from 1991 to 1998. Although the NE3 fault is inferred to be around 24 MW (Tokita and Haruguchi,
propagation of the injected water front will continue in the future 1998).
at deep levels, the cooling effects are believed to be smaller than Because the Hatchobaru reservoir is considered to mainly con-
when injection was conducted at shallower levels. This is because, sist of five productive faults (Fig. 1), the sustainable power potential
which is actually corroborated by the monitored silica tempera- of the reservoirs can be approximately evaluated by integrating that
ture (Tokita et al., 2000), temperature of injected water increases of each productive fault. The Komatsuike-sub fault and Hatchobaru
 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390 387

Fig. 9. Simulated temperature distribution and the zone containing injected water (Tokita et al., 2000).

fault currently have more or less stable to slight overproduction assumed in this case study are required to maintain nearly 120 MW
conditions. However, the sustainable power output of the Komat- from beginning of the calculation, it will be easy to sustain around
suike and NE4 faults is unknown. Accordingly, a numerical case 120 MW over 9 years. Therefore, it is suggested that the sustain-
study was conducted assuming additional production wells in all able power potential of the Hatchobaru reservoirs is approximately
productive faults except the NE3 fault. Three scenarios were con- 120 MW. It will also be possible to maintain the rated power out-
sidered in which a total of 17, 26 and 31 make-up production put of 110 MW if the targets for production make-up wells were
wells were assumed to produce from these faults. Both the number selected in such way as not to overproduce the faults beyond their
of wells and the spacing between them (100–200 m) were speci- sustainable power potential (Tokita et al., 2000, 2006b).
fied regardless of the actual number of wells that could be drilled Fig. 11 shows the sustainable power potential of each of the five
from the existing drilling pads. These scenarios showed that the productive faults, which ranges from 11 to 55 MW. It was also found
simulated total power output temporally reached 190–265 MW, that concentrating replacement wells into a particular fault should
depending on the number of assumed make-up wells; however, in be avoided in order to maintain the balance of mass production. The
all three scenarios a power decline was experienced until it sta- simulated results showed that the rated power output of 110 MW
bilized at around 120 MW, as shown in Fig. 10. The production could be sustained with make-up wells drilled in the productive
period is 9 years, which may not be sufficient to determine the sus- faults and operated at mass extraction rates that do not exceed
tainable potential. Since a smaller number of make-up wells than

Fig. 10. Prediction of sustainable power output of the Hatchobaru Geothermal Field Fig. 11. The sustainable power potential of each of the productive faults.
(Tokita et al., 2000, 2006b). Figure modified from Tokita and Haruguchi (1998).
388 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390

Fig. 12. Changes in the reservoir deduced from the results of gravity monitoring: (a) before commencement of Unit No. 2 operation, (b) immediately after commencement
of the Unit No. 2 operation, (c) initial decline period, (d) recovery period (e) second decline period, and (f) relatively stable period.
Modified from Saito et al. (2006).

their sustainable production potentials. The simulated results show range of −20 to 50 ␮gal. The gravity change in the injection area was
that geothermal fluids are continuously supplied by injection and small compared to that in the production area, where the gravity
influx from lateral boundaries. The computed mass inflows from widely reduced by around 10 ␮gal. However, this gravity change in
these sources are 46% and 48% of the produced mass, respectively. the production area had already slowed, reflecting the stabilizing
Thus, the mass deficit due to production is estimated as only 6% reservoir pressure in this period. According to the results of a quasi-
of the produced mass, which indicates that the mass balance is three-dimensional inversion of the gravity changes from 1994 to
approximately maintained (Tokita et al., 2006b). 2000, the underground mass in the production area is likely to have
The gravity changes also indicate that underground mass in the stabilized in recent years (Saito et al., 2006). According to Ehara and
production area is likely to have stabilized in recent years. Measure- Nishijima (2004), the repeat gravity measurements can be used to
ments of gravity and pressure indicate a close correlation between estimate the mass recharge in the reservoirs during exploitation
variations of these two parameters. Therefore, the gravity changes of the Hatchobaru geothermal field. The mass balance estimate
are considered to reflect the changes in reservoir mass behavior. based on change in gravity suggests that fluid recharge together
Fig. 12 summarizes changes in the reservoir deduced from the with injection is approximately equal to the produced mass.
results of gravity monitoring (Saito et al., 2006). In this figure, pres- On average, the specific enthalpy in the reservoirs is predicted to
sures changes in production and injection zones are represented as decrease in the next 5 years by 3.3 kJ/kg/y, equivalent to a temper-
variations in the water level for the periods (a) before commence- ature drop of about 0.8 ◦ C per year. This rate of temperature drop
ment of Unit No. 2 operation, (b) immediately after commencement is in agreement with the recorded changes in silica temperatures
of the Unit No. 2 operation, (c) initial decline period, (d) recovery of existing wells (e.g., well H-13, as shown in Fig. 5). Accordingly, it
period, (e) second decline period and (f) relatively stable period. is predicted that the power output will slightly decrease with time
The water level rose due to increased injection mass flow rate in due to cooling of the production zones. Considering constraints for
the injection zone and began to fall due to increased production expansion of the development area (it is only a few km2 in size,
mass flow rate in the production zone in period (b). The water and because of the national park that surrounds the Hatchobaru
level fell over a wide area in period (c) because of the effects of field further expansion is not allowed), it is very difficult to com-
pressure drawdown that resulted from the increase of production pletely avoid migration of injected water into the production zones
mass flow rate. The pressure drawdown in the reservoirs caused in the current development area. What should be done, therefore, to
the power output to decline to around 80–90 MW. The water level minimize the cooling in the production zones, is to optimize the dis-
began to recover during period (d) and the power output recovered tribution of injected water in the reservoirs. Based on appropriate
to around 100 MW by the addition of make-up wells and relocation reservoir management, it will be possible to keep the present power
of injection wells to mitigate cooling in the production zone. There- output stable by adding a make-up well every 1 or 2 years, which is
after, the water level fell again in the production zone in period (e). economically acceptable (Tokita et al., 2006b). According to Rybach
Mitigation of cooling and pressure drawdown, and the approximate and Mongillo (2006), in relation to geothermal resources, especially
mass balance in the reservoirs was attained in period (f). The gravity to their exploitation, sustainability means the ability to sustain the
changes from early 2000 to late 2001 in period (f) showed a small production level over long times. Therefore, sustainable produc-
 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390 389

Fig. 13. Optimum alignment of make-up production and injection wells (Tokita et al., 2006b). Alignment is not evident, zones are. Detected concentration ratio of fluorescein
and 1,5-naphthalens disulufonate (−).

tion of geothermal energy secures the longevity of the resources at above sea level, simulations indicate that the reservoir has a sus-
an appropriate production level. We think that an appropriate pro- tainable capacity of about 120 MW. The rated power output of
duction level should be economically acceptable. This means that 110 MW could, therefore, be sustained by appropriate location and
make-up wells are allowed to be added to maintain the appropriate operation of wells.
production level as long as the costs for drilling them are accept-
able in terms of economical exploitation, even if mass and energy Acknowledgements
balances in the reservoirs have not strictly been attained. From this
view point, the exploitation conditions of the Hatchobaru reser- We highly appreciate and express our special thanks to the staffs
voirs are sustainable under the present mass production rate and of the geothermal group and the Hatchobaru power plant of Kyushu
well locations. Electric Power Company Inc. for their cooperation in the data pro-
Fig. 13 shows the basic design for the optimized alignment of vision and technical advice and for their permission to publish this
make-up production and injection wells. Simulation results indi- paper.
cate that the entire reservoir can be classified into three regions
in terms of production decline: a region where pressure draw-
References
down has been caused by production; a region where the reservoirs
have been cooled due to excessive invasion of injected water; Asai, A., 2002. Utilization of geochemistry to reservoir management at the Otake-
and a region suffering from both pressure drawdown and cool- Hatchobaru geothermal field (geochemistry for geothermal development).
ing. Although these impacts to the reservoirs still remain, reservoir Chinetsu 39 (1), 50–62.
Ehara, S., Nishijima, J., 2004. Study on sustainable development of geothermal energy
management based on coupled simulations successfully makes from geophysical monitoring of geothermal reservoir – a case study of repeat
their impact small. Basic strategy for the sustainable development gravity measurement. Journal of the Geothermal Research Society of Japan 26
is to avoid and to mitigate overproduction and cooling in produc- (2), 181–193.
Fujino, T., Yamasaki, T., 1985. Geologic and geothermal structure of the Hatchobaru
tion zones. It is, therefore, important to optimize the alignment field, Central Kyushu, Japan. Geothermal Resource Council Bulletin 14, 11–15.
of production and injection wells. Make-up production and injec- Kawazoe, S., Tokita, H., 1993. Reservoir management of the Hatchobaru Geothermal
tion wells should be placed so as to keep the total mass production Field. In: Proceedings of the 15th New Zealand Geothermal Workshop, Auckland,
New Zealand, pp. 291–296.
within the sustainable productivity of each fault, and to ensure that Lima, E., Tokita, H., Tsukamoto, S., 2004. Operation of the Hatchobaru geothermal
the injected water does not migrate into production zones (Tokita field based upon a coupled reservoir-well and piping network numerical sim-
et al., 2006b). ulator. In: Proceedings of the KenGen Geothermal Conference, Nairobi, Kenya,
CD-ROM, pp. 58–64.
Momita, M., Tokita, H., Matsuda, K., Takagi, H., Soeda, Y., Tosha, T., Koide, K.,
6. Conclusions 2000. Deep geothermal structure and the hydrothermal system in the Otake
Hatchobaru geothermal field, Japan. In: Proceedings of the 22nd New Zealand
The Hatchobaru reservoirs including production faults have Geothermal Workshop, Auckland, New Zealand, pp. 257–262.
Pruess, K., 1991. TOUGH2—A general-purpose numerical simulator for multiphase
been stably producing water and steam equivalent to 90 MW or
fluid and heat flow. Report LBL-29400. Lawrence Berkeley National Laboratory,
more. Although the reservoirs have been affected by production Berkeley, California, 102 pp.
and injection, the reservoir pressure, temperature, and gravity Pruess, K., Oldenburg, C., Moridis, G., 1999. TOUGH2 User’s Guide, Version 2.0. Report
monitoring data indicate small declines in these parameters sug- LBNL-43134. Lawrence Berkeley National Laboratory, Berkeley, California.
Rybach, L., Mongillo, M., 2006. Geothermal sustainability – a review with iden-
gesting that the current exploitation conditions are sustainable. tified research needs. Transactions of the Geothermal Resources Council 30,
Assuming that the exploitation depth is limited to around −1400 m 1083–1090.
390 T. Yahara, H. Tokita / Geothermics 39 (2010) 382–390

Saito, H., Tagomori, K., Honda, M., Kasagi, J., Fujise, Y., Ehara, S., 2006. Reservoir Tokita, H., Lima, E., Hashimoto, K., 2005a. A Middle-term power output prediction
behavior inferred from quasi-3D analysis of gravity changes in the Hatchobaru at the Hatchobaru field by coupling multifeed wellbore simulator and fluid-
geothermal field. Journal of the Geothermal Research Society of Japan 28 (2), gathering pipeline simulator to reservoir simulator. In: Proceedings of World
181–197. Geothermal Congress 2005, Antalya, Turkey, CD-ROM, 8 pp.
Shimada, K., Fujino, T., 1995. Deep-seated geothermal resources in Kyushu Island, Tokita, H., Itoi, R., Hamada, Y., Senjyu, T., 2005b. Development of a multi-feed well-
Japan. Transactions of the Geothermal Resources Council 19, 301–319. bore simulator MULFEWS and its application to field data analysis at Hatchobaru.
Shimada, K., Tagomori, K., Fujino, T., Kitakoga, I., 1994. Deep geothermal structure Journal of the Geothermal Research Society of Japan 27 (1), 11–26.
and targets of future development in the Hatchobaru geothermal field. Chinetsu Tokita, H., Lima, E., Itoi, R., Motoyama, T., 2006a. Development of technique
31 (4), 301–319. for the coupled simulator of reservoir, wellbore, and field-gathering pipeline
Tagomori, K., Ehara, S., Nagano, H., Oishi, K., 1996. Study on reservoir behavior networks to provide optimum operation and management of geothermal
based on gravity changes in the Hatchobaru geothermal field. Journal of the power plant. Journal of the Geothermal Research Society of Japan 28 (1), 77–
Geothermal Research Society of Japan 18 (2), 91–105. 93.
Tamanyu, S., 1993. Geotectonic developing history in the Beppu-Kuju graben, Cen- Tokita, H., Lima, E., Itoi, R., Akiyoshi, M., Senjyu, T., 2006b. Application of cou-
tral Kyushu, Japan. Memoirs Geological Society of Japan 41, 93–106. pled numerical reservoir simulation to design a sustainable exploitation of the
Thermal and Nuclear Power Engineering Society, 2008. The present conditions and Hatchobaru geothermal field. In: Proceedings of Renewable Energy 2006, CD-
a trend of geothermal power generation in Japan, 2008. Report of Thermal and ROM, 4 pp.
Nuclear Power Engineering Society, 99 pp. Tsuru, Y., Tokita, H., 1994. Monitoring of the reservoir pressure and temper-
Tokita, H., Haruguchi, K., 1998. Strategy of reservoir management based on numer- ature in the Hatchobaru geothermal field, Japan. In: Proceedings of 15th
ical simulation in the Hatchobaru geothermal field. In: Proceedings of Asian PNOC-EDC Geothermal Conference 1994, Manila, The Philippines, pp. 95–
Geothermal Energy ‘98 on Current and Future Geothermal Energy Development 102.
in Asia, Subic, The Philippines, pp. 89–96. Yahara, T., Saito, H., Itoi, R., Motoyama, T., 2005. Mechanism of mass productivity
Tokita, H., Itoi, R., 2004. Development of MULFEWS a multi-feed wellbore simulator. fluctuations in Hatchobaru production Well H-30. Journal of the Geothermal
In: Proceedings of the 29th Workshop on Geothermal Reservoir Engineering Research Society of Japan 27 (4), 321–336.
Stanford University, California, CD-ROM, 6 pp. Yahara, T., Saito, H., Itoi, R., Takeshima, N., 2006. Reservoir management for sustain-
Tokita, H., Yahara, T., Kitakoga, I., 1995. Cooling effect and fluid behavior due to ing injectivity and productivity of wells in the Hatchobaru geothermal field. In:
reinjected hot water in the Hatchobaru geothermal field, Japan. In: Proceedings Proceedings of the 4th International Workshop on Earth Science and Technology,
of the World Geothermal Congress 1995, Florence, Italy, pp. 1869–1874. Fukuoka, Japan, pp. 375–382.
Tokita, H., Haruguchi, K., Kamenosono, H., 2000. Maintaining the rated power output Yahara, T., Kiyota, Y., Itoi, R., Takeshima, N., 2008. Application of lumped parameter
of the Hatchobaru geothermal field through an integrated reservoir manage- model for evaluating pressure interference influence in geothermal reservoir
ment. In: Proceedings of World Geothermal Congress 2000, Kyushu-Tohoku, management. Journal of the Geothermal Research Society of Japan 30 (2),
Japan, pp. 2263–2268. 131–144.