Reservoir Modeling of Deep Geothermal Systems - Geotermia

Reservoir Modeling of Deep Geothermal Systems:
the Examples of Guardia Lombardi
Giordano Montegrossi
Consiglio Nazionale delle Ricerche, Istituto di Geoscienze e Georisorse
Via G. La Pira 4
50121 Firenze, Italy
[email protected]
ENI GEOLAB Bolgiano
3 wells investigated (Monteforcuso 1, Monteforcuso 2
and Bonito 1 Dir)
Total core drilled samples: 12 (16.4 m total recovery) :
300
200
150
P Hydrostatic
P Hydraulic Test
100
50
0
500
1000
1500
2000
2500
3000
True Vertical Depth (m)
140
120
100
80
T °C
Pressure (bar)
250
T °C Teoric non-stabilized
T °C meassured non-stabilized
60
40
20
0
500
1000
1500
2000
True Vertical Depth (m)
2500
3000
TOUGH – family codes owe its birth
To Karsten Pruess, just retired !
John Backus
Born:
December 3, 1924
Philadelphia, Pennsylvania
Died:
March 17, 2007
(aged 82)
Ashland, Oregon
Well enthalpy with time
$
$
$
$
$
:
:
:
$
$
cd /home/grb/hengill/graphics
outfile=/home/grb/hengill/2005-model/may02a/forward_ouput_file
grep “OUPUT DATA AFTER” $outfile| cut –c40-50 > j1
grep “WELL10” $outfile| awk ‘{if (NF==12) print $5}’ > j2
pastej1 j2 | awk ‘{print 1975+$1/3.14E+7,$2/1000}’ | xy_plot
debug commands until right
history –5 > draw.well.10
rm j1 j2
# this script plots temperature history of well 5 in layer a
#
outfile=/home/grb/hengill/2005-model/may02a/inverse_tekplot_file
grep “w5ta” $outfile | awk ‘{print $1,$2}’ | sec2date –d 1975 > j1
grep “w5ta” $outfile | awk ‘{print $1,$3}’ | sec2date –d 1975 >> j1
grep “w5ta” $outfile | awk ‘{print $1,$4}’ | sec2date –d 1975 >> j1
'Steady State' or Natural state computation:
Given the Boundary condition from:
Geological investigation,
Geochemical investigation,
Geophysical investigation,
Etc.
We could use this condition to define the property of the system,
simply using the provided geometry and boundary conditions to
populate the computational grid and let the system stabilize.
Then we have to compare the grid results with our point values, and check.....
Output Data Visualization:
Tecplot has a data type ‘cell centered data’ perfectly suited to TOUGH software
packages output
Describe element geometry and add a header to your file like:
VARIABLES= "X" "Y" "Z" "Temperature" "Saturation" "Ph"
ZONE F=FEBLOCK ET=BRICK N= 5824 E= 4590
VARLOCATION = ([4-0049]=CELLCENTERED) T=" 36443. Days"
Top boundary
Condition: heat flow
Bottom boundary
Condition: heat flow
All cell-centered data could be imported in PETREL as point data with attribute !
From the 'Steady State', or Natural State computed, we could try to foresee
what could happen with a producing well.
We have to point out some notes:
Since Pressure gradient is Hydrostatic, the well will not significantly produce without
a pumping system
For economical reason, water flux should be at least 10 - 30 l/min, better in the magnitude of
Hundreds of l/min
We had to use a recharge, that is unknown, and we put as recharge boundary condition
an inlet with the same characteristic of the reservoir from the south direction (along Y axes)
Pressure vs time
Hot water production: nearly 808 Kg/min
280
270
260
250
240
Pressure (bar)
230
220
210
200
190
180
170
160
150
0
0
0,01
0,1
1
Time (days)
10
100
1000
10000
It seems some 'grass', but
correspond to the CO2(aq)
exsolution level
Conclusion:
The production of up to 800 l/min is sustainable, given an efficient recharge
The impact on the reservoir is to lower the pressure from 271 bar to nearly
177 bar
No significative scaling of the well is expected, but geochemical analisys
should be revised
To foresee the effect of the exploitation on reservoir temperature, the
recharge path should be better constrained!