Prediction of reinjection effects in the Cerro Prieto geothermal system

0375-6505/84 $3.00 + 0.00 Pergamon Press Ltd. © 1984 CNR.

Geothermics, Vol. 13, No. 1/2, pp. 141 - 162, 1984. Printed in Great Britain.

PREDICTION OF REINJECTION EFFECTS IN THE CERRO PRIETO GEOTHERMAL SYSTEM C. F. T S A N G , D. C. M A N G O L D , C. D O U G H T Y and M. J. L I P P M A N N Earth Sciences Division, Lawrence Berkeley Laboratory, University of California, Berkeley, CA 94720, U.S.A. (Received 17 November 1982)

Abstract--The response of the Cerro Prieto geothermal field to different reinjection schemes is predicted using a two-dimensional vertical reservoir model with single- or two-phase flow. The advance of cold fronts and pressure changes in the system associated with the injection operations are computed, taking into consideration the geologic characteristics of the field. The effects of well location, depth and rates of injection are analysed. Results indicate that significant pressure maintenance effects may be realized in a carefully designed reinjection operation.

NOMENCLATURE Q, Q~ Q20 S ": 0

= = = = = =

Injection well flow rate (kg/s). Production area flow rate (kg/s). Two-dimensional equivalent injection well flow rate (kg/s). Distance from the injection well to the boundary of the production area (m). Angle of the whole production area seen by the injection well (rad). Angle of the modeled section of the production area seen by the injection well (rad).

INTRODUCTION Reinjection o f separated geothermal brines and condensate was initially considered as a possible means o f disposing o f large quantities o f these waste fluids. It was soon realized that the m o r e i m p o r t a n t advantages o f reinjection are its potential capability o f maintaining reservoir pressures and enhancing the extraction o f heat f r o m the reservoir rocks, thus prolonging the commercial life o f geothermal fields. H o w e v e r , one factor o f great concern that has prevented large-scale reinjection at m a n y sites is the fear o f premature arrival at the p r o d u c t i o n wells o f the cold temperature front associated with the injected water. A properly designed reinjection operation is required to avoid this. In particular, well locations, depths and rates o f injection must be planned with specific consideration o f the characteristics o f the geological f o r m a t i o n s in the field. This paper presents calculations predicting cold temperature front m o v e m e n t s and pressure changes in the C e r r o Prieto reservoir system for a n u m b e r o f different injection cases. PREVIOUS STUDIES Interest in reinjection at the C e r r o Prieto geothermal system was reported in the First Cerro P r i e t o S y m p o s i u m . T w o papers presented at that meeting addressed the problems o f reinjection, the first on chemical studies o f reinjection effects (Rivera et al., 1980) and the second a hypothetical study o f the influence on the producing field o f cold temperature fronts resulting f r o m injection operations (Tsang et al., 1980). This latter study assumed the reservoir to be one-layered, with injection carried out in different areas o f the field. Based o n the average reservoir parameters k n o w n at that time, it was shown that the cold temperature f r o n t would not reach the nearest p r o d u c t i o n well for a considerable time ( > 60 years). 141

142

C . F . Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann

It was soon realized that the Cerro Prieto reservoir was far from being a one-layer reservoir system. At the Second Cerro Prieto Symposium a generic study of two-layered reservoirs was presented (Tsang et al., 1981). Effects of injection in the upper reservoir on the lower one and vice versa were calculated. The influence of an opening in the shale layer separating the two reservoirs was also calculated. All these calculations still assumed a very simplified model of the Cerro Prieto reservoir. Since August 1979, the Comisi6n Federal de Electricidad has been reinjecting 165°C untreated brines into well M-9. The maximum injection rate was reported to have been approximately 80 t/h, or 20 kg/s, and the depth of injection was in an interval between 721 and 864 m. Neighbouring production wells, such as M-29, opened at about 1100 m depth, were monitored in order to detect changes in temperature, pressure and enthalpy of the produced fluids. There has been no report that injection has caused any changes in the characteristics of these wells. This injection test was discussed in a numerical modeling paper presented at the Third Cerro Prieto Symposium (Tsang et al., 1981). A rather realistic geological model of the area near M-9, based on the stratigraphic analysis of Lyons and van de Kamp (1980), was used in the 1981 study. This model showed an upper aquifer of about 400 m thickness at the injection level of M-9 and a lower aquifer of 180 m average thickness, representing what is commonly known as the A or c~ reservoir, at the production level of M-29. The two aquifers were assumed separated by a 20 m thick less permeable layer. Based on this geological model, detailed numerical calculations indicate that over the injection-test period (about 1.5 years by 1981), no significant effects due to injection should be expected at the production wells. This conclusion is consistent with the field experience. P R E S E N T STUDY AND A P P R O A C H Based on experiences gained from previous studies, the present paper attempts to predict long-term reinjection effects at Cerro Prieto using a recently developed geologic model of the field. Such predictions, with proper short-term validations, will give an estimate of both the beneficial and adverse effects of long-term reinjection at this site and will also help in designing reinjection strategies, including well location, depth and flow rate. In our calculations, the stratigraphy of Cerro Prieto developed by Halfman et al. (1983) is used. Due to the lack of full three-dimensional geological information, we will model only a vertical cross-section of the system. Figure 1 presents a two-dimensional multilayered model that fits Halfman's stratigraphy of the western part of the field along a line through wells M-9, M-29 and M-10. This layered model, on which our reservoir calculations are based, includes several major features o f the geology of the area, such as the variations in thickness and depth of the various layers. The model has a closed boundary 1225 m southwest of well M-9, which is SW

M-9

M-IO

M-29

NE

600 E x

-"

6oo 1000 1200 1400 1600 1225

m

.

.

.

1500

.

220

m

1750

m

m

Fig. 1. Vertical two-dimensional model used for the calculations. The shaded areas represent less permeable layers. The diagonally hatched area is the production zone.

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

143

assumed to be associated with the strike-slip Cerro Prieto fault. The intent of our study is to calculate the pressure and temperature distribution in the cross-section being modeled when reinjection is carried out at different locations and depths. Figure 2 shows a discretized version of part of our two-dimensional vertical section. This grid will be used for mass and heat flow calculations based on an Integrated Finite Difference Method discussed below under Methodology. Figure 1 shows a multilayered reservoir model, which consists of an upper aquifer, the ~ reservoir, and the 13 reservoir, separated by less permeable layers. The production region at Cerro Prieto corresponds approximately to a 1.5 × 1.5 km area. In the vertical two-dimensional section, the zone being produced is represented by the 1.5 km long diagonally hatched area in the a reservoir. 600 800

111111111111 IlUlllIIIII

E v 1000

~ 12001400 1600

niinnlWnhi i IlIIIIIIlllI l IIIIIIIINtLL i

400

800

1200

1600

2000

DISTANCE (m) Fig. 2. The central portion of the vertical two-dimensional model (given in Fig. 1) showing the spatial discretization used to calculate the heat and mass flows. The four alternate injection regions are shown cross-hatched. The diagonally hatched area is the production zone.

We will calculate the temperature and pressure changes in the production region resulting from the injection of cooler water (165°C) into well M-9 (220 m southwest of well M-29) or into 300 m wide hypothetical reinjection regions centered 595 m southwest of well M-29. Two different depths of reinjection will be considered: one in the upper aquifer and the other in the ct reservoir. The four reinjection regions are indicated as cross-hatched zones in Fig. 2. It is apparent from Fig. 2 that the mesh is finer in the region around well M-9 and coarser elsewhere. This will tend to introduce some numerical dispersion which will artificially spread the thermal front as the injected water moves from well M-9 into the coarser parts of the mesh. However, considering the general nature of this study, such a dispersive effect is not expected to alter our overall conclusions. E Q U I V A L E N T I N J E C T I O N RATE IN A VERTICAL SECTION A major problem in studying a three-dimensional system using a vertical two-dimensional model is that o f representing the equivalent injection rate. This is still an open problem and requires further study. For our present paper, the following approach is proposed. Figure 3 shows schematically an areal view of the production field represented by a 1.5 x 1.5 km area. The vertical two-dimensional section which we are to study is represented by the zone between the two broken lines, chosen arbitrarily to be 150 m wide, with a fluid extraction rate in the production region of Qp(150/1500) = Qp/lO, if we neglect edge effects. The two-dimensional flow rate for an injection well having a flow rate Qi, located at distance, S, southwest of the production area, is estimated as follows. First, assume that Qi/2 of the injected flow rate goes towards the production area in response to the lower pressure there, i.e. half the injected fluid

144

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann

/A j/'

Production Area

/"

_v_ I n j e c t i o n Well

modeU.ed section of field W

Fig. 3. A schematic areal view of the Cerro Prieto field and the vertical section modelled. Used to determine the twodimensional injection flow rate. flows towards and half flows away from the production zone. Thus Q,/2 is contained in the angle, y, between lines stretching from the injection well to points A and B in Fig. 3. Then, the injected fluid entering the vetical section of interest will be proportional to the angle, 0, between lines stretching f r o m the injection well to points V and W. This flow rate is (Q,/2) (0/y). Therefore, for the entire model which extends on both sides of the injection well, the flow rate to be used should be Q,O/3'. This expression has the proper limits for an injection well either very close or very far f r o m the production area. Table 1 shows the weighing factor 0/3' for different distances between production and injection zones. This technique is used in our calculations and will be further investigated in a future study to determine its validity as well as its limitations. Table 1. Two-dimensional flow rate determination Q2d = O/'y(Q.), Q. = 200 kg/s = 30% Qp S(m)

O/y

Q2d(kg/s)

1 25 50 100 220* 595* 2000

0.99 0.81 0.65 0.45 0.26 0.14 0.10

198 -162 130 90 52 28 20

*Cases simulated in this paper. METHODOLOGY Two computer codes, developed at Lawrence Berkeley Laboratory, were employed to predict the effects o f reinjection at C e r r o Prieto. P r o g r a m P T (for p r e s s u r e - t e m p e r a t u r e ) (Bodvarsson, 1982) is an expanded and revised version of code CCC used in earlier Cerro Prieto reinjection studies. It models single-phase (liquid) heat and mass transport in permeable media, employing the Integrated Finite Difference Method (IFDM) which permits the analysis of threedimensional systems with complex geometry. The code has been validated against analytic and semi-analytic solutions and has also been carefully verified against a series of field experiments. It has been applied extensively to many thermohydrological problems. The other code, SHAFT79 (Pruess and Schroeder, 1980), also developed at Lawrence Berkeley Laboratory, is a two-phase ( l i q u i d - v a p o u r ) I F D M code that models heat and steam - water flow in three-dimensional porous media. Recent developments enable it to model

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

145

fractured porous media as well. It has been validated against a number of analytic results and experimental data, and has been applied to the study of several geothermal development problems. These programs have been applied to calculate several hypothetical cases of long-term reinjection at Cerro Prieto. These are summarized in Table 2. All cases are calculated for Qj = 0.3 Qp. For the single-phase (liquid) calculations, we shall assume that the principle of superposition holds and the injection effects are calculated over an injection period of 30 years. Any temperatures and pressures obtained will be predicted changes due to long-term injection. On the other hand, for two-phase (steam-water) calculations, we cannot assume that the principle of superposition holds. Thus, both a 9 year production period and a subsequent 5 year injection period with ongoing production are simulated. Table 2. Cases calculated Injection distance from edge of production zone Single-phase calculations Case 1 220 m (into M-9) Case 2 220 m (into M-9) Case 3 595 m (300 m injection zone) Case 4 595 m (300 m injection zone) Case 5 220 m (into M-9)

Injected layer

Upper aquifer ct Reservoir Upper aquifer ct Reservoir Upper aquifer

Case 6

220 m (into M-9)

ct Reservoir

Case 7

220 m (into M-9)

ct Reservoir

Other conditions

D

D

m

A break is assumed in the intervening layer separating upper aquifer and a reservoir A break is assumed in the intervening layer separating ct and 13 reservoirs A break is assumed in the intervening layer separating upper aquifer and a reservoir

Two-phase calculations First, production, is simulated for 9 years, then injection into M-9 as production continues Case 1 220 m (into M-9) Upper aquifer -Case 2 220 m (into M-9) tt Reservoir Case 3 220 m (into M-9) Upper aquifer A break is assumed in the intervening layer separating upper aquifer and ¢t reservoir

INITIAL CONDITIONS AND PARAMETERS USED The material parameters of the different layers used in the calculations are shown in Table 3. These are reasonable values for Cerro Prieto based on information obtained to date. The initial temperature and pressure conditions over the vertical multilayered system are obtained by equilibrating a vertical column in the mesh shown in Fig. 2, assuming constant temperature closed-flow boundaries on top and at the bottom. By assuming the upper boundary to be at Table 3. Material properties and total production/injection rates

Upper aquifer Intervening layer a Reservoir Intervening layer 13 Reservoir

Permeability (md)

Compressibility ( x 10-'°) (Pa")

Porosity

50 0.5 50 5 50

2 5 2 5 2

0.18 0.40 0.22 0.40 0.22

Production rate Qp = 670 kg/s. Injection rate Q, = 0.30 Qp ~ 200 kg/s.

146

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann

225°C and the lower boundary at 325°C, pressure and temperature profiles are obtained as shown in Fig. 4. These match reasonably well with field measurements from the production region. These column-equilibrated pressure and temperature values were assigned to the entire mesh, and equilibration was carried out for up to 60 years. Changes after 30 years were minimal. Thus, the temperature and pressure distributions after 30 years of equilibration were used as the initial conditions for all our calculations. The error introduced, due to a further 30 year equilibration period, is estimated to be about 1 psi and 0.5°C.

600~ 8 o o I\

......

600 ~-,;-

PRESSURE ( P S I ) 1000 1400 r---~ .....

~\

'\\ \

1800 2200 ,. . . . . . . . . .

\ [- . . . . . . . . . . . . . . . . . . . .

E v 1 004 "it-~. 120~ I.i C3 1401

160b 225

250 275 300 TEMPERATURE (DEG C)

325

Fig. 4. Pressureandtemperatureprofilesusedasinitialconditions ~rthecalculations.

RESULTS

--

SINGLE-PHASE CALCULATIONS

The calculated temperature and pressure changes for each single-phase reinjection case listed in Table 2 are presented as contour plots in Figs 5 - 12. The pressure increases in response to reinjection are quickly established and then change little with time, so only one pressure distribution (after l0 years o f injection) is shown for each case. On the other hand, the temperature varies with time, so the calculated temperature changes after 10, 20 and 30 years of injection are shown. In the plots the less permeable layers between the aquifers are shaded, and the Mo9 injection interval and the location of M-29 (the production well closest to the injection location) are indicated by vertical bars. The 300 m wide injection region is indicated by a rectangle.

600 80( v T I--

0.

1001 120q 140t

160~ 400

800

1200

1600

DISTANCE ( m )

Fig. 5 (A).

2000

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

,,....

E

v

1

"r I--

o.

1

w r,,,

1

400

800

1200

1600

2000

DISTANCE ( m ) ~t3t3

E "r O. hJ

400

800

1200

1600

2000

DISTANCE (m)

v -,t-

1

o..

1 1 1

400

800

1200 DISTANCE

1600

2000

(m )

600 800 E

1000

Z

1200 w

1400

1600

400

800

1200 DISTANCE

Figs. 5(B-E).

1600 (m)

2000

147

148

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann 600r F

800 ~ E v

5~o

I 0 0 0 ~ = : ........ ~

.

~

/'

~ /'

~

.

720~ / / ~

~ ', ',

tm 1400630 Ds 1600

. 400

.

.

.

.

800

1 200

DISTANCE

600,

?"-" lpn

1600

2000

(m

ia 600

800!

ps

\

i

I

i

100( 120(

a 140( 6 0 0 IL 400

'~ 800

51q 1200 DISTANCE

1600

I 2000

(m)

Fig. 5. Pressure changes after 10 years of injection in single-phase calculations. ( A - G ) Cases 1 - 7 , respectively. The vertical lines (horizontal rectangle) indicate the location of the injected interval (zone) and the open interval in well M-29. Contour interval: 30 psi.

Pressure changes (after 10 years of injection) Case 1. The pressure increase due to injection into the upper aquifer through well M-9 (Fig. 5A) is not confined to the upper aquifer, but penetrates through the less permeable layers into the a and 13 reservoirs. The less permeable layers retard the pressure response somewhat so, at a given lateral distance from M-9, the pressure change decreases as one goes from the upper (injected) aquifer to the lower 13 reservoir. The effect o f the closed southwest boundary of the field (Fig. 1) is shown by the shape of the contour lines to the left of M-9. The asymmetry o f the pressure-change contours with respect to the injection well (M-9) is due to the reflection of the pressure pulse o f f that closed boundary. Case 2. The pressure changes resulting from injection into the ct reservoir through M-9 (Fig. 5B) show the same general characteristics as those o f Case 1. The less permeable layers retard the pressure response but do not completely confine it to the layer into which injection is carried out. The pressure reflection o f f the closed boundary is also evident. These features are c o m m o n to all the single-phase calculations we have done. Cases 3 and 4. Figure 5 (C, D) shows the pressure changes due to injection into the upper aquifer and the a reservoir, respectively, through a 300 m wide injection zone whose center is 595 m southwest o f well M-29. Note, from Table 2, that the injection rate into this zone is smaller than it was for Cases 1 and 2 (injection into M-9). This reflects the fact that, when the injection region is farther away from the production zone, less of the injected fluid flows into the two-dimensional section o f the production zone considered by our model. However, even in these cases a significant pressure increase is seen in the production zone.

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

149

Cases 5 - 7. The effect of a break in either intervening layer is shown in Fig. 5 ( E - G). Different pressure changes are observed in each case, but in all of them the pressure is readily transmitted through the breaks. A comparison of Cases 1 - 7 shows that, in all cases, reinjection causes a pressure increase throughout the multilayered reservoir systems considered in our model. The closed boundary to the southwest further enhances these pressure increases. It is to be noted that these calculations are based on liquid-phase systems. In the case o f s t e a m - w a t e r systems, the high compressibility of the two-phase fluid will result in much lower values for the calculated pressure increases (see next section). However, the qualitative conclusions above still hold. Temperature changes Case 1. The thermal response to reinjection into the upper aquifer through well M-9 (Fig. 6 A - C ) is the formation of a cool region that steadily grows with time and sinks due to the 600 .A

~

800

1_.ooYEARS

X\

JJlJ /

1000

illlllli llliliilkLllIlllllh,li E'! iv :,,,i,llLiiiiHVl lldllhHUILi,,,iH]Itttltllltiillt llHI,,il [,' ' ll 'l,lilIE,liI,,,tJ' "' '%HIHiil 1400 I '"'

1600/'

i

i

i

i

J

A

I

I

4O0

800

1200

1600

2000

400

800

1200

1600

2000

400

800

1200

1600

2000

I

600 800 A

E

1000

"1" I--

1200

laJ t',,

1400 1600

1 1 1 1

DISTANCE (m) Fig. 6. Temperaturech~gesduetoi~ectionintotheupper ~ui~rthrough well M-9insingl~phasecalculations (Casel).(A)Aflerl0years;(B)a~er20years;(C)a~er30years. Contourintervahl0°C.

150

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann

higher density of the cooler injected water. The less permeable layers slow the downward movement of the cool water but do not stop it entirely. After 10 years of reinjection the cool water has just reached the top of the c~ reservoir; after 30 years it has spread through it and just penetrated the top of the 13 reservoir. Case 2. The temperature response to injection into the ~ reservoir through M-9 (Fig. 7A - C) shows that after only 10 years of reinjection, temperature changes have reached the upper aquifer and the 13 reservoir, and have extended into the production zone of the ct reservoir. After 30 years, much larger temperature changes have reached the ct and 13 reservoirs than in Case 1 (with injection into the upper aquifer). 6 0 0 1 ~

I

10 YEARS

800

1000 1200 1400 160OIL 400

L

~

,

,

,

~

,

,

800

1200

1600

2000

800

1200

1600

2000

,

600 80C

Tv -l-

in i.i r~

100C 120C 140C 160G

4OO

c

1

30 YEAR:

800" 1

0

0

0

~

1

2

0

0

~

1400 i 1600 J ~ 400

t

800

1200

1600

i

t

i

2000

DISTANCE (m) Fig. 7. Temperature changes due to injection into the ct reservoir through M-9 in single-phase calculations (Case 2). (A) After 10 years; (B) after 20 years; (C) after 30 years. Contour interval: 10°C.

Case 3. After 10 years of injection into a 300 m wide injection zone in the upper aquifer centered 595 m from well M-29, the temperature changes (Fig. 8A) are confined to the upper aquifer. After 20 years (Fig. 8B) a small temperature change has reached the ct reservoir, but it

B00~60

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System ~

0

*

C [

' 400

' 800

400

800

1200

1600

2000

400

aoo

1200

1600

2000

151

10 YEARS

1 000 1200 1400 1600''

'

' 1200

J

' 1600

'

' 2000

600 800

E

v -r I-O. bJ

1000 1200 1400 1600

600 800 1000 1200 1400 1600

DISTANCE ( m )

Fig. 8. Temperature changes due to injection into the upper aquifer through a 300 m-wide injection zone centered 595 m southwest of well M-29 in single-phase calculations (Case 3). (A) After l0 years; (B) after 20 years; (C) after 30 years. Contour interval: 10°C.

is far from the production zone. After 30 years (Fig. 80), the cool water still has not reached the a reservoir production zone or the [~ reservoir. Case 4. Even after 30 years of injection into the ct reservoir through the 300 m wide injection zone, the temperature changes have just barely extended into the production zone of the a reservoir (see Fig. 9A - C). There is a temperature decrease in the 13reservoir after 30 years, but it is largely limited to the region under the injection zone. Case 5. This case considers the effect of a gap in the intervening layer separating the upper aquifer and the a reservoir as injection is carried out into the upper aquifer through well M-9 (Fig. 10A - C). After only 10 years of injection the discontinuity in the lower permeability layer has a strong effect (cf. Figs 6A and 10A). With a continuous intervening layer (Case 1) the temperature change has barely penetrated the top of the ct reservoir; with a break in the layer the cool water extends well into the ct reservoir. After 30 years, the largest temperature decrease

152

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann

6°°FA

..........

~ _ _

lo YeaRs

800 I 1 0 0 0 1 ......

-

-

1

~

...............I~UJi.................................................................................................

,, ool\\

;o, o

1,oolNllilJINIIJJIFH]lilNIJiliPJI 1600 f , 400

,

,

,

J

,

,

,

~

800

1200

1600

2000

400

800

1200

1600

2000

400

800

1200

1600

2000

R f't t~

"" :I: I-o.

1

1

W 1

600 800 1000 1200 1400 1600

DISTANCE

(m)

Fig. 9. Temperature changes due to injection into the ct reservoir through a 300 m-wide injection zone centered 595 m southwest o f well M-29 in single-phase calculations (Case 4). (A) After l0 years; (B) after 20 years; (C) after 30 years. C o n t o u r interval: 10°C.

is found in the ct reservoir, rather than in the upper aquifer and, in general, the cool region has moved farther down towards the ~ reservoir. Case 6. In this case, the effect of a break in the intervening layer between the ct and 13 reservoirs is studied, as colder water is injected into the ct reservoir through well M-9 (Fig. I I A - C ) . After 10 years of injection, the gap in the layer has very little influence on the temperature changes (cf. Figs 7A and 11A). After 20 and 30 years, more cool water has flowed into the l~ reservoir than in Case 2 with its continous intervening layer. However, even after 30 years, the break has only a minor effect on the overall shape and extent of the cooler region. Case 7. The gap between the upper aquifer and the et reservoir considered in Case 5 is again assumed, with injection into the ¢t reservoir through M-9 (Fig. 1 2 A - C ) . Although the temperature changes do propagate into the upper aquifer through the break in the layer, the overall effect of the gap on the shape of the cool region is much less dramatic when injection is

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

400

800

1200

1600

2000

400

800

1200

1600

2000

400

800

1200

1600

2000

153

E "1" I-

0.

1

1,¢

1

lv~v

6O0 80O 1000 1200 1400 1600 DISTANCE

(m)

Fig. 10. Temperature changes due to i~ection into the upper aqui~r through well M-9 when there is a break in the

intervening layer between the upper aqui~r and the a reservoir, in single-phase calculations (Case 5). (A) A~er l0 years; (B) after 20 years; (C) a~er 30 years. Contour interval: 10°C.

carried out below the gap (this case) than when injection is done above it (Case 5). This is because the cooler injected water, denser than the native hot water, tends to sink due to gravity. The extent of the cold temperature front after 30 years of reinjection varies from case to case. Moving the reinjection zone farther away from the production zone both laterally (Cases 3 and 4) and vertically (Cases 1 and 3) results in smaller temperature changes in the production zone of the ct reservoir and in the 13 reservoir. The break in the intevening layer between the upper aquifer and the et reservoir (Cases 5 and 7) strongly affects the downward propagation of the cold temperature front from the upper aquifer into the a and 13 reservoirs, but has less influence when fluid is injected into the reservoir, below the break. The gap in the intervening layer between the et and 13 reservoirs (Case 6) located farther away from the reinjection area has only a small influence on the overall advance of the cold temperature front.

154

C. F. Tsang, D. C. ( Mangold, C. Doughty and M. J. Lippmann 6O0

A

[

1o YEARS

8OO

100C

J

1200 ~'

1400

--10 ° C

1 6 0 0 I, 400 600

I I':i'hll "i

800

1200

•B

1600

2000

I

2G YEARS

800 j=

1000 1200 1400 1600

t

400

600

I

t

800

i

I

1

1200

IC

~

1600

[

~

I

I

2000

30YEARS

800 1000 1200 1400 16001 400

800

1200 DISTANCE

1600

2000

(m)

Fig. l I. Temperature changes due to injection into the ct reservoir through M-9 when there is a gap in the intervening layer between the ct and I~ reservoirs, in single-phase calculations (Case 6). (A) After l0 years; (B) after 20 years; (C) after 30 years. Contour interval: 10°C.

RESULTS - - T W O - P H A S E C A L C U L A T I O N S In the two-phase calculations, we cannot assume the principle of superposition because of the nonlinear characteristics of the phenomena involved. Thus, instead of calculating pressure and temperature changes as in the single-phase cases described above, we have to calculate actual temperature, pressure and steam saturation values. Production is first simulated for 9 years, then the three cases of reinjection listed in Table 2 are performed as production continues. The results are presented as contour plots of pressure and temperature changes from the initial conditions shown in Fig. 4. Vapour saturation distribution in the system is also plotted; initially, the vapour saturation was assumed to be zero throughout the model. Below, the results of the initial 9 years of production and the subsequent 5 years of reinjection are presented for each of the two-phase cases studied.

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System 600[ A

L

155

10 YEARS

/

8ooi1000 120C 1400 1600

c

i]lll]lllll[[li[lilllllillliJIHJlllllllilllIII llLihLlltlitl!liiLIIIIlllLlllllllli i

400

600

i

800

i

i

1200

B

i

i

1600

2000 20 YEARS

~

800 E v

1000 1200 -

i~ ~ " ~ = ~

_~'~'~'~"~"~'~"'~j__100 , C

IililHI!![/',E

w

1400

llllllil]lltlltLiiftiit a

1600-L--.-J--~

40O

800600t C

~

1200

800 ~

L

1600

2000

L

----30 YEARS

I o o o[ ........................

1

i

14001 1600 ~' ~ 400

illlllllllll ~

-

1

o

L L , 800 1200 D ] STANCE

oc J

L

1600

~

L

2000

(m)

Fig. 12. Temperature changes due to injection into the e reservoir through M-9 when there is a gap in the intervening layer between the upper aquifer and the ~ reservoir, in single-phase calculations (Case 7). (A) After 10 years; (B) after 20 years; (C) after 30 years. Contour interval: 10°C.

Production simulation The calculated temperature, pressure and saturation changes are given in Figs 13 and 14 after 3 and 9 years of production, respectively. The pressure change after 3 years (Fig. 13A) is concentrated in the production region as expected, but also penetrates to the other layers. The effect of the southwestern boundary of the field is also reflected by the shape of the contours left of well M-29. The initial pressure and the temperature distribution with depth (Fig. 4) represents water near its saturation point in the ~ reservoir; the drop in pressure from production causes flashing to occur there as the pressure falls below the steam saturation pressure (Fig. 13C). This leads to widespread boiling and, eventually, to the lowering of the temperature seen in the lower ~ reservoir (Fig. 13B) as pressure continues to drop. The higher temperature regions directly below the production zone are due to the upward flow of hotter fluids toward the production zone. In particular, the steam from the two-phase

156

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann 600

A

1

PRESSURE CHANGE

800 1000 1200 1400 1600 400

800

1200

1600

600

2000

\ TEMPERATURE

CHANGE ]

800

|

E

v

1000

"r F0,. laJ r',,

1200 1400

0 e

-5 ° C 1600 400

600

800

1200

C

1600

[

2000

SA,TURATION

800 1000 1200

..................................................................................................

Ilil!lHiEii Jllf rIlIlifJjJtfi li JIIJdl Ji!ltliPi!Jtirl!JJ i l

I JJJ *

'

1400 . ~ . ~ - - - ~ 16001

::lul ................................................................................................

L

400

,

'

I

It~

,

rll

~

~

8oo

J

-L

12oo

~

,

160o

,

,

,

2000

Dl STANCE (m) Fig. 13. Pressure, temperature and steam saturation responses after 3 years of production from tile ct reservoir in twophase calculations. (A) Pressure changes (in psi); (B) temperature changes (°C); (C) steam saturation.

reservoir is migrating upward and condensing in the cooler liquid of the lower region of the ct reservoir. This condensation releases latent heat raising the temperature of this region. Temperature increases as a consequence of production, caused by upward migration and condensation of steam, have been observed in several geothermal fields as well as in numerical studies of field behaviour (Bodvarsson et al., 1982). The results after 9 years of production show a further development of these phenomena. The pressure changes (Fig. 14A) have extended farther into the 13 reservoir and show a more pronounced effect of the closed boundary to the west. To the northeast, a constant pressure boundary (Fig. 1) allows fluid and heat recharge, but to the southwest the closed boundary does not allow it, thus enhancing the growth of the two-phase zone there (Fig. 14C). The region of increased temperature above the 13 reservoir (Fig. 14B) moves to the west along with the twophase zone, because of steam condensation effects.

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

157

600 800 1000 1200 1400 1600

400

800

1200

1600

2000

400

800

1200

1600

2000

600 800 v

1000

"r In t=J El

1200 1400 1600

l

6°°fC

SATURATION

800I 1000 i 12OO 1400 1600

400

800

1200 DISTANCE

1600

2000

(m)

Fig. 14. Pressure, temperature and steam saturation responses after 9 years of production from the ~ reservoir in twophase calculations. (A) Pressure changes (in psi); (B) temperature changes (°C); (C) steam saturation.

Reinjection simulation Case 1. The results of 5 years of injecting 30°70 of the mass produced into the upper reservoir through well M-9 are shown in Fig. 15. The pressure contours (Fig. 15A) show significant effects in both the upper aquifer and the a reservoir, despite the low permeability layer separating them. For example, the pressure drop at well M-29 has decreased approximately 50 psi since injection began. However, in the 13 reservoir, more than 500 m below the injection zone, the pressure decline due to production continues. The high compressibility of the twophase zone makes the pressure increases due to injection very slow. The temperature contours (Fig. 15B) show the region of injected cooler water very clearly. It is apparent that the denser colder water is drawn to the producing zone located in the a reservoir. The vapour saturation (Fig. 15C) decreases in the production zone, but is relatively unchanged in the layers below it.

158

C. F. Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann 60c' 80q

I00~ 120~

1401 160b

400

800

1200

1600

2000

400

800

1200

t600

2000

600 800 E

1000

-r I--

o.. 1200 ,,-t

1400 1600

E0 0EC

l

8o, I

SATUR ATJON

I

r

120C

1400 1600

~ 400

800

~ 1200

1600

2000

D] STANCE (m i Fig. 15. Pressure, temperature and steam saturation responses after 5 years of injection into the upper aquifer through M-9, with continuing production from the ct reservoir in two-phase calculations (Case l). (A) Pressure changes (in psi); (B) temperature changes (°C); (C) steam saturation.

Case 2. The pressure response to injection in the ct reservoir through well M-9 (Fig. 16A) shows the strong influence o f the injected water in the liquid regions of the tx reservoir and of the upper aquifer, but the high compressibility o f the two-phase zone around the production region and in the I~ reservoir tends to diminish the pressure increase there. Significant temperature changes (Fig. 16B) have occurred in the production zone. The vapour saturation (Fig. 16C) decreases in the ¢t reservoir and in the intervening layer just below M-9 and M-29. The lower part of the 13 reservoir returns to a one-phase liquid condition. Case 3. The effect of a gap in the intervening layer separating the upper aquifer and the ct reservoir (after the production period was simulated assuming a continuous layer) is shown for injection into the upper aquifer through well M-9 (Fig. 1 7 A - C ) . The pressure effects are clearly seen to be greater in the ct reservoir than for the case of a continuous intervening layer (cf. Figs 15A and 17A). For example, the pressure at M-29 has dropped approximately 20 psi

Prediction of Reinjection Effects in the Cerro Prieto Geothermal System

400

800

1200

1600

2000

400

800

1200

1600

2000

159

600 800 E

1000

-1I.Q. 1 2 0 0 I,IJ

1400 1600

600

C

1 SATURATION

800 1000 1200 1400 1600

400

800

1200 DISTANCE

1600

2000

(m)

Fig. 16. Pressure, temperature and steam saturation responses after 5 years of injection into the a reservoir through M-9, with continuing production from the a reservoir in two-phase calculations (Case 2). (A) Pressure changes (in psi); (B) temperature changes (°C); (C) steam saturation. less than it did in Case 1. In the upper reservoir, however, the pressure decline is greater than it was in Case 1. The temperature contours show cooler waters entering the production zone through the gap in the invervening layer (Fig. 17B). This development occurs earlier here than for the case of a continuous intervening layer (cf. Fig. 15B), due to the higher permeability channel now available. There is also a slightly greater contraction of the 0.1 saturation curve in the production region. The influence of the gap on the saturation in the intervening layer between the ct and ~ reservoirs and in the [3 reservoir just begins to be apparent after 5 years (Fig. 17C). Thus, even a relatively small break in the intervening layer can have a measurable effect on the pressure and temperature distributions after only 5 years of injection. A comparison of these three two-phase cases shows that reinjection causes pressure increases in the production region when fluids are injected either in the upper aquifer or in the tz

160

C. [ \ Tsang, D. C. Mangold, C. Doughty and M. J. Lippmann 600 A ,oo

~

[ PRESSURE CHANGE

~

....

,ooo

12oo 1 6 0 0 _u 400

lllllllt ,

, 800

~

, 1200

+

i 1600

,

, 2000

,

600 800 E

1000 1200

W

1400

1600

400

800

1200

1600

l

60°fC

2000

SATURA+,ON

800~ 1000~,,,~

lau

ittt j n +

1200~ 1400~

1 6 0 0 ~, 40O

800

1200

1600

2000

DISTANCE (m) Fig. 17. Pressure, temperature and steam saturation responses after 5 years of injection into the upper aquifer through M-9 with continuing production when the intervening layer between the upper aquifer and the t~ reservoir is discontinuous, in two-phase calculations (Case 3). (A) Pressure changes (in psi); (B) temperature changes (°C); (C) steam saturation.

reservoir. However, these increases are much smaller than in the single-phase liquid cases discussed in earlier sections. This is due to the high compressibility of the two-phase zones, which diminishes the pressure changes. Nevertheless, a definite pressure-increase effect is seen even in these cases. There are also declines in the saturation levels in the production region near the injection intervals. Temperature reductions in the producing (¢t) reservoir are important only if reinjection is carried out in the same reservoir or when significant breaks exist in the lower permeability layers between the produced and injected reservoirs. The results discussed in this section depend strongly on initial reservoir conditions which are only very roughly known. Thus, further calculations much beyond the 5 year injection into the two-phase reservoir may not be so meaningful. Our proposal is to check calculated results against actual field reinjection data for a period of 1 - 5 years in order to validate the numerical

Prediction o f Reinjection E f f e c t s in the Cerro Prieto G e o t h e r m a l S y s t e m

161

m o d e l s a n d initial c o n d i t i o n s e m p l o y e d . T h e v a l i d a t e d m o d e l a n d c o n d i t i o n s m a y t h e n be used to m a k e l o n g e r - t e r m p r e d i c t i o n s . CONCLUSIONS T h e m o s t recent geologic m o d e l o f the C e r r o P r i e t o g e o t h e r m a l system, d e v e l o p e d f r o m welllog analysis, was used to calculate the expected effects o f r e i n j e c t i o n at d i f f e r e n t l o c a t i o n s relative to the f i e l d ' s p r o d u c t i o n zone. This is p a r t o f a series o f r e i n j e c t i o n studies m a d e on this g e o t h e r m a l field. T h e C e r r o P r i e t o reservoir system is c o n s i d e r e d to be m u l t i l a y e r e d , with an u p p e r c o l d e r aquifer, an i n t e r m e d i a t e (u) g e o t h e r m a l reservoir a n d a lower (~) reservoir. R e i n j e c t i o n into the u p p e r a q u i f e r a n d ct reservoir are the two alternatives s t u d i e d in this p a p e r . I n j e c t i o n l o c a t i o n s are a s s u m e d to be 220 m o r 595 m s o u t h w e s t o f the edge o f the p r o d u c t i o n area, the f o r m e r c o r r e s p o n d i n g to the p o s i t i o n o f well M-9. B o t h single-phase (liquid) a n d t w o - p h a s e ( s t e a m - w a t e r ) c a l c u l a t i o n s are c a r r i e d o u t using n u m e r i c a l m o d e l s P T a n d S H A F T 7 9 , d e v e l o p e d at the L a w r e n c e Berkeley L a b o r a t o r y . T h e results o f o u r s t u d y show t h a t significant p r e s s u r e - s u s t a i n i n g effects can be o b t a i n e d in the p r o d u c t i o n zone by reinjecting 30% o f the fluid m a s s e x t r a c t e d . T h e b r e a k t h r o u g h o f the injected c o l d w a t e r into the p r o d u c t i o n zone is s t r o n g l y d e p e n d e n t on the l o c a t i o n o f the i n j e c t i o n wells. B r e a k t h r o u g h is very significant if large-scale i n j e c t i o n is c a r r i e d o u t near well M-9. O n the o t h e r h a n d , it seems to be safe to inject 30% o f the fluid p r o d u c e d if r e i n j e c t i o n is c a r r i e d o u t 595 m (or farther) f r o m the p r o d u c t i o n area. G a p s in the l o w - p e r m e a b l i t y layers b e t w e e n the injected a n d p r o d u c e d reservoirs have a significant effect on the a d v a n c e o f t h e r m a l fronts into the e x p l o i t e d zones. In c o n c l u s i o n , we w o u l d r e c o m m e n d t h a t , b e c a u s e o f the significant benefit in pressure m a i n t e n a n c e in the reservoir, r e i n j e c t i o n be c a r r i e d o u t in a carefully p l a n n e d a n d c a r e f u l l y m o n i t o r e d f a s h i o n . E a r l y results over 1 - 5 years m a y .be used to v a l i d a t e o u r a s s u m p t i o n s a n d models. O n c e v a l i d a t e d , the m e t h o d can be used to p r e d i c t reservoir b e h a v i o u r with considerably more confidence. Acknowledgements--We would like to acknowledge discussion with, and assistance from, S. Halfman and K. Pruess of Lawrence Berkeley Laboratory. We also thank our colleagues of the Comisi6n Federal de Electricidad for providing some of the data used in this study. This work was supported by the Assistant Secretary of Conservation and Renewable Energy, Office of Renewable Technology, Division of Geothermal and Hydropower Technologies of the U.S. Department of Energy under Contract No. DE-AC03-76SF00098.

REFERENCES Bodvarsson, G. S. (1982) Mathematical modeling of the behavior of geothermal systems under exploitation. Ph.D. thesis, Univ. Calif. Lawrence Berkeley Lab. Rep. LBL-13937. Bodvarsson, G. S., Pruess, K., Lippmann, M. J. and Bjornsson, S. (1982) Improved energy recovery from geothermal resources. J. Petrol. TechnoL 34, 1920- 1928. Halfman, S. E., Lippmann, M. J., Zeiwer, R. and Howard, J. H. (1983) A geological interpretation of the geothermal fluid movement in the Cerro Prieto field, Baja California, Mexico. Am. Ass. Petrol Geol. Bull. Lyons, D. J. and van de Kamp, P. C. (1980) Subsurface geological and geophysical study of the Cerro Prieto geothermal field, Baja California, Mexico. Lawrence Berkeley Lab. Rep. LBL-I0540. Pruess, K. and Schroeder, R. C. (1980) SHAFT79, User's Manual. Univ. Calif. Lawrence Berkeley Lab. Rep. LBL-I0861. Rivera R., J., Mercado G., S. and Tsang, C. F. (1980) Preliminary studies on brine reinjection at the Cerro Prieto field. Presented at the First Syrup. Cerro Prieto Geothermal Field, San Diego, California, September 1978. Geothermics 9, 209-212. Tsang, C. F., Bodvarsson, G. S., Lippmann, M. J. and Rivera R., J. (1980) Some aspects of the response of geothermal reservoirs to brine injection with application to the Cerro Prieto field. Presented at the First Symp. Cerro Prieto Geothermal Field, San Diego, California, September 1978. Geothermics 9, 213- 220.

162

C. F. Tsang, D. C. Mangold, C. D o u g h t y a n d M. J. L i p p m a n n

Tsang, C. F., Mangold, D. C. and Lippmann, M. J. (1981) Simulation of reinjection at Cerro Prieto using an idealized two-reservoir model. Presented at the Second Symp. Cerro Prieto Geothemal Field, Mexicali, October 1979. Geothermics 10, 195 - 200. Tsang, C. F., Mangold, D. C., Doughty, C. and Lippmann, M. J. (1981) The Cerro Prieto reinjection tests: Studies of a multilayer system. Proc. Third Syrup. Cerro Prieto Geothermal Field, San Francisco, California, March 1981. Univ. Calif. Lawrencea Berkeley Lab. Rep. LBL-II967, 5 3 7 - 547.