Professional Documents
Culture Documents
1. INTRODUCTION
An ability to predict both the quantity of fluid that can be
produced and its thermodynamic state (pressure,
temperature, enthalpy, gas content, salinity, etc.) is essential
for estimating the total usable energy of a geothermal
resource. Numerical reservoir simulators can be utilized to
calculate the thermodynamic state of the fluid at the
underground feed-zone(s) at which the fluid enters the
wellbore. The computation of the well-head fluid properties
from a given underground state (or vice-versa) requires the
use of a wellbore simulator.
Garg, et al.
In two-phase water/steam flow, pressure and temperature
are not independent of each other. For any reasonable value
of effective thermal conductivity , the downhole flowing
enthalpy may be adjusted to yield the appropriate pressure,
and hence temperature, distribution in the wellbore, and
flowing wellhead enthalpy. Matching the pressure/
temperature distribution in the wellbore and flowing
wellhead enthalpy does not constrain the heat loss and
downhole enthalpy. Since the flowing downhole enthalpy is
not a measured quantity, it is not possible to determine a
unique value for heat loss in the presence of two-phase
flow. It is, therefore, appropriate to eliminate effective
thermal conductivity as a free parameter, and use a constant
value for (= 4 W/m-C).
Roughness factor,
Angle with
Vertical
(Degrees)
0.000
11.759
= 0.012 kg/kg
= 0.0011 kg/kg
= 0.09 for depths < 350 m
= 0.09 + 0.0062 (depth350)
for 350 m < depth < 400 m
= 0.40 for depths > 400 m
= 0.00 mm for all depths
100
200
= 1102 kJ/kg
Given the fluid state at the bottom of the cased interval and
the mass discharge rate, a wellbore simulator such as
WELBOR can be used to compute the fluid state along the
wellbore and at the wellhead. The principal parameters that
may be varied to match the measured conditions along the
wellbore (pressure and temperature) and at the wellhead
(pressure, temperature, steam and liquid flow rates, liquid
salinity, gas content of steam) are (1) flowing enthalpy,
salinity and gas content at the bottom of the cased interval,
(2) holdup parameter, , and (3) interior roughness factor, .
Measured
Depth
(meters)
277.4
901.6
Temperature
(Degrees Celsius)
27
68
212
227
300
400
500
600
700
Internal
Diameter
(mm)
384
315
800
900
10
12
14
16
18
20
22
24
Pressure, bars
Garg, et al.
single Caithness well) were obtained while these wells were
discharging in a more or less stable manner. Available data
for these wells include the NCG and salt content of the
discharge stream. By way of contrast, data for the Japanese
wells were obtained during short term discharge tests. No
NCG and salinity measurements are available for the
production fluid from these short term tests. Also, discharge
rate and wellhead enthalpy measurements from the short
term tests may not be very accurate; these errors are likely
to be most problematical for wells with very low discharge
rates. Although flow data for the Japanese wells in the
dataset are liable to be less accurate than for the Unocal
wells, it was decided against discarding these data in that
the Japanese well test data are typical of relatively high
quality measurements taken during the exploration phase.
102
10
FROTH FLOW
Nl
REGION I
ITI
ON
TR
AN
S
1
BUBBLE FLOW
PLUG FLOW
101
101
SLUG FLOW
HEADING
10
Qs
[1 Qs ]K + Qs [1 l (1 K ) / g ]
(2)
Sg g
m = Sl l + S g g
(3)
(4)
Q f Qs
(5a)
Sg K 1
(5b)
103
AS g g vg
MIST FLOW
102
Ng
Qf =
REGION III
REGION II
(1)
Rn =
dwM
Am
Fr =
M2
d w A2 g m2
We =
Sl vl
( Sl vl + S g vg )
d w2 g g Fr
(6a-d)
Garg, et al.
Here d w is the inside well diameter, is the liquid surface
tension, and m is the mixture viscosity.
m = Sl l + S g g
1.00
(7)
0.95
K 0.90
0.85
0.80
2.0
3.5
4.0
(8)
(9)
The computed results for wellhead pressure for all the high
mass velocity (mass velocity = total discharge rate / pipe
cross-sectional area > 687 kg/s-m2. The low mass velocity
profiles are considered separately in Section 4.) profiles are
compared with data (i.e. wellhead pressures computed by
matching downhole pressure profiles in Section 2) in Table
1. The K(Z) fit yields satisfactory agreement between data
and computed wellhead pressures (Table 1); the root-meansquare error is only 0.67 bars.
S .V . = ( Ki +1 Ki )2
3.0
2.5
(10)
i =1
Garg, et al.
Discharge Profile A-4
0
Pressure, bars
Saturation pressure
Computed (best match)
Computed (holdup correlation)
100
P-comp
9.51
9.22
10.75
15.15
11.23
9.66
14.17
10.59
11.79
12.12
11.79
10.86
10.89
9.59
10.26
9.77
10.26
7.68
Choke @ 3.7m*
Choke @ 88.9m
7.11
Choke @7.6m**
6.61
8.15
12.12
10.10
14.56
13.59
12.43
**
***
Well
A-1
A-2
A-4
A-6
A-7
A-8
A-9
A-10
A-11
A-12
A-13
A-14
A-16
A-19
A-20
A-21
B-5
C-6
KE1-4a
KE1-4b
KE1-17a
KE1-19Sa
KE1-19Sb
KE1-22b
GH-11a
GH-11b
GH-20a
GH-20b
GH-20c
200
P***
0.57
0.44
0.22
0.45
1.62
0.09
0.39
0.80
0.63
0.02
0.39
0.12
0.53
0.02
0.14
0.75
0.71
1.87
NA
NA
0.47
NA
0.24
0.78
0.11
0.41
0.42
0.57
0.73
300
400
500
600
700
800
900
8
10
12
14
16
18
20
22
24
Pressure, bars
1.11
The computed results for wellhead pressure for all the low
mass velocity (mass velocity = total discharge rate / pipe
cross-sectional area < 650 kg/s-m2.) profiles are compared
with data (i.e. wellhead pressures computed by matching
downhole pressure profiles in Section 2) in Table 2. The
K(Z) fit shown in Figure 5 yields satisfactory agreement
between data and computed wellhead pressures (Table 2);
the root-mean-square error is 0.73 bars.
Garg, et al.
Discharge Profile A-4
A-18
C-1
C-2
C-3
C-4
C-5
B-3
B-4
B-13
KE1-9
KE1-11
KE1-17b
KE1-22a
S-2
CS-1
P-data
8.19
10.35
11.58
11.74
12.43
13.88
16.27
18.70
11.76
8.59
18.66
7.11
7.82
1.15
3.97
P-comp
8.99
10.35
11.69
10.92
12.15
14.64
16.70
17.96
10.57
8.74
17.29
7.38
8.22
2.50
4.06
Total (15 profiles)
100
200
P**
Vertical Depth, meters
Well
0.80
0.00
0.11
0.82
0.28
0.76
0.43
0.74
1.19
0.15
1.37
0.27
0.40
1.35
0.09
200
250
300
350
400
450
Spinner, rev/s
0.04
(11)
6. CONCLUDING REMARKS
The principal goal of the present work is to use high quality
data from flowing geothermal wells to devise new liquid
holdup correlations for geothermal applications. To make
the problem tractable, it was decided to at first develop a
holdup correlation for only the cased section of geothermal
wells. The holdup correlation presented in the preceding
sections displays considerable promise for simulating twophase flow in the cased section of geothermal wells. Future
plans call for the formulation of a holdup correlation for the
open hole/slotted liner section of geothermal wells.
(l Sl + g S g )
600
900
150
(l Sl vl + g S g vg )
500
800
P = P-data P-comp
vm =
400
700
**
f = m(vm vc ) + c
300
(12)
ACKNOWLEDGEMENTS
This manuscript has been authored by a contractor of the
U.S. government under DOE Contract DE-AC0799ID13727. Accordingly, the U.S. Government retains a
nonexclusive, royalty-free license to publish or reproduce
the published form of this contribution, or allow others to
do so, for U.S. Government purposes. We also wish to
express our thanks to Unocal Corporation and Caithness
Operating Company LLC for making their proprietary well
data available for this study.
REFERENCES
Ansari, A. M., Sylvester, N. D., Sarica, C., Shoham, O., and
Brill, J. P.: A Comprehensive Mechanistic Model for
Upward Two-Phase Flow in Wellbores, SPE
Production & Facilities, 143165, May (1994).
The smoothed spinner data for well A-4 are compared with
the computed spinner response in Figure 6. The spinner
data for well A-4 (Figure 6) exhibit an anomalous response
Garg, et al.
Bankoff, S. G.: A variable density single-fluid model for
two-phase flow with particular reference to steamwater flow, Journal of Heat Transfer, ASME
Transactions, Series C, Vol. 82, 265272, 1960.