EXPECTED SILICA SCALING FROM REINJECTION WATERS AFTER INSTALATION OF A BINARY CYCLE POWER STATION AT

1
EXPECTED SILICA SCALING FROM REINJECTION WATERS
AFTER INSTALATION OF A BINARY CYCLE POWER STATION
AT BERLIN GEOTHERMAL FIELD, EL SALVADOR, CENTRAL
AMERICAMarlon R. Castro, Dina L. López.
Jaime A. Reyes, Antonio Matus, Francisco
E. Montalvo Carlos E. Guerra.Instituto de
Ingeniería, Universidad Autónoma de Baja
California, Mexicali, B C, MexicoDepartment of
Geological Sciences, Ohio University, Athens,
Ohio. LaGeo, S. A. de C.V, El Salvador, C. A.
2
Background

• One of the two most important geothermal fields
  in El Salvador is Berlin Geothermal Field,
  started operations in 1992. The 100 waste water
  generated during the process of generation of
  electrical energy is reinjected back to the
  reservoir. The supersaturated reinjected waters
  precipitate silica within the reservoir, plugging
  the pores, and decreasing the capacity of the
  reinjection wells with time.
• In April 2004, the Binary Cycle project was
  initiated at the Berlin Geothermal Field. In this
  project, the residual thermal heat contained in
  the steam separated water will be used to
  generate additional electrical energy. The inlet
  stream temperature will be 180 ºC and the outlet
  stream will be at 140 ºC.
• The objective of this paper is to present results
  of our simulations of silica scaling at Berlin
  Geothermal Field under different scenarios of
  mixing between reservoir and the reinjected
  waters at the new temperature conditions, and to
  compare with previous results of mixing of
  reinjection waters at 175 ºC which is the
  reinjection temperature at the present time
  (Castro et al., 2006).

3
Fig. 1. Berlin geothermal field and tectonic
setting of the study area
4
Previous modeling of silica scaling

• Castro et al. (2006) reported the results of the
  simulations indicate that between 0.612 to 0.679
  g of quartz precipitate per kg of reinjected
  water when the reinjection waters mix with
  reservoir waters at 200 ºC and 250 ºC,
  respectively. As the reinjection waters are
  supersaturated, they precipitate the largest mass
  of quartz during the first step of the
  simulation.
• Also calculated the percentage of pores clogged
  per year assuming a reservoir thickness of 200 m,
  a 10 porosity, and three different radius for
  the volume that could be affected by the
  precipitation of quartz. The radius of the
  clogged volume is probably less than 15 m. 2
  decease in porosity per year was obtained for a
  15 m radius. A radius higher than 15 m should
  result in less than 2 per year clogging of the
  pores and should not affect significantly the
  permeability and absorption capacity of the
  wells. This modeling results are coincident with
  field observations of injection decline (Romero,
  2005).

5
MODELING PROCESS

• The computer programs SOLVEQ and CHILLER were
  used in our simulation work.
• The reinjection waters restored to a temperature
  of 140 C and pH 5.75 (H2SO4 was added) for
  well TR-1A were mixed with restored reservoir
  waters (well TR-9) in different proportions
  adding reservoir water to 1 kg of reinjection
  water up to a proportion of only 10 of
  reinjection water.
• The compositions of reservoir waters at two
  different temperatures were used 250 ºC and 200
  ºC. The simulations were done assuming water-rock
  interaction and fractionation of minerals.
• For the simulations, the mineral assemblage
  observed at Berlin Geothermal Field production
  zone and that were included in the
  thermodynamical data base of CHILLER was used.
  This assemblage includes the minerals albite,
  anhydrite, calcite, chlinochlore, galena,
  hematite, magnesite, muscovite, pyrite, and
  quartz.

6
(No Transcript)
7
RESULTS AND DISCUSSION

• The simulations results are presented in several
  graphs showing the evolution of the precipitation
  of quartz as a function of the mixing fraction
  between reinjected and reservoir waters.

8
Fig. 3. Silica results for the mixing of restored
reinjection water at 140 ºC with reservoir water
(TR-9 water) with water-rock interaction assuming
fractionation.
9
Fig. 3. Silica results for the mixing of restored
reinjection water at 140 ºC with reservoir water
(TR-9 water) with water-rock interaction assuming
fractionation.
10
Fig. 4. Quartz precipitation after modeling the
mixing of restored reinjection water of TR-1A
well at 175 ºC (actual condition) and, 140 ºC
with pH 5.75 (future condition) with reservoir
water (TR-9 well) at 250 ºC and 200 ºC with
water-rock interaction and assuming
fractionation. Quartz precipitation, a) mass and
b) volume.
11
Fig. 5. Percent of obstructed porosity after
modeling the mixing of restored reinjection water
of TR-1A well at 175 ºC (actual condition) and,
140 ºC with pH 5.75 (future condition) with
reservoir water (TR-9 well) at 250 ºC and 200 ºC
with water-rock interaction and assuming
fractionation. Reservoir water temperature a) 250
ºC, b) 200 ºC.
12
Effect of quartz precipitation on the porosity of
the reservoir close to the well and comparison
between reinjection water at 140 C and pH
5.75 with 175 C
13
Validation

• We do not have data yet to validate our modeling
  results for the reinjection at 140 C. However,
  we have data for the past behavior for the
  reinjection temperature of 175 C. We can
  consider a two year period for reinjection wells
  TR-8 and TR-14, from 01/01/97 to 02/01/99.
• Considering the initial reinjection flow rates
  during the period of time before mentioned for
  the two wells (53.3 kg/s for TR-8 and 42.2 kg/s
  for TR-14) the decrease in reinjection rate for
  these two wells is 2.32 and 1.3 for TR-8 and
  TR-14, respectively. The first value falls within
  the range of percentage of clogged volume
  estimated for radius between 10 and 15 m in Table
  2 (2.0 to 4.9). The second value (1.3)
  corresponds to a radius slightly larger than 15 m

14
Fig. 6. Behavior of the absorption capacity of
wells TR-8 and TR-14 during period 01/01/97 to
01/02/99.
15
Conclusions

• The results of these simulations show that even
  when the change in temperature of the reinjection
  brine will be around 35 ºC, the amount of quartz
  precipitated during the future reinjection in
  Berlin Geothermal Field will not be too different
  from the mass of quartz that is precipitating
  with the present reinjection temperature of 175
  ºC. A maximum difference of 15 in the percent of
  clogged pores is expected.
• The simulations show that after 10 years a great
  proportion of the well will be clogged in all the
  scenarios modeled and that probably the well will
  not be very useful (43 to 50 decrease in
  porosity). This result can be understood if we
  consider that the mass of silica reinjected per
  unit time will be the same that is reinjected
  nowadays. As the pH will be lowered to 5.75 units
  from 6.5, the precipitation of silica within the
  plant and piping system will be avoided. Then,
  silica will precipitate inside the reservoir
  increasing the clogging of the pores and
  decreasing even further the water absorption
  capacity of the wells, like simulation showed.
• In this modeling work, the kinetics of silica
  precipitation has not been considered because the
  programs used are chemical equilibrium programs.
  However, according to the silica precipitation
  experiments (Molina Padilla et al., 2005), once
  the induction time has finished, silica
  precipitation is fast. Consideration of the
  kinetics of silica precipitation as well as
  modeling of water movement and solute transport
  should be the next step to better understand and
  predict the behavior of silica scaling in Berlin
  Geothermal wells and other geothermal fields of
  the world.

16
Thanks for your attention