Professional Documents
Culture Documents
CONTAINING·BRINES
' '
By
GUOHAI LI·U
Bachel~r of Science
East China Institute of Chemical
Technology
Shanghai, China
1982
Master of Science
Central Coal Mining Research Institute
·Beijing, China
. ' 1984
Thesis Approved:
ii
ACKNOWLEDGMENTS
be.
iii
TABLE O~F CONTENTS
Chapter Page
Physical Model • . . . . . . . . . • • 30
Model Formulations . . . . • . • 35
The Interface of the-Gas and Liquid . . 35
Mass Transfer Through Turbulent Film
Layer . . ·. . . . . . . . . . . . . . 42
Mass Tr~nsfer and Ion Migration
Through Diffusion Layer . • . . . , . 51
Mass Transfer Through the Corrosion
Product Layer . . . . , . . . . . . . 53
Kinetics of Surface Reactions . . • . . 53
iv
Chapter Page
Case IV • . . . . . 99
Case V. I I I I I I I10 2
Conclusions . . . . . . . . • . . . . . • • . 112
Recom~endations . . . . . . . . • . 114
Modeling Localized Corrosion . . . . 114
Modeling of the Gas Wells with
Condensate. . .....•. .115
Prediction of the Effect of the
Inhibitor . . , . . • . . . . . . • • • 115
Experimental Research on the
Corrosion Rate Under Slug Flow. . • . 116
v
LIST OF TABLES
Table Page
vi
LIST OF FIGURES
Figure Page
vii
Figure Page
viii
NOMENCLATURE
c~thodic ~afel,constant
activity of i component
F Faraday's constant
ix
fugacity coefficients of gas phase
g gravitational acceleration
H Henry's constap:t
h step size
I ionic strength
Im mixing length
ks wall .roughness
X
mi molality of i component
core radium
Sc Schmidt number
t temperature in degree C
T absolute temperature K
Ui mobility of i species
xi
y wall coordinate
z axial coordinate
xii
PL liquid density
Pg gas density
T shear stress
~ electric potential
xiii
CHAPTER I
INTRODUCTION
1
2
rate.
mathematical treatments.
corrosion.
CHAPTER II
LITERATURE SURVEY
given.
5
6
fittings.
Burke and Hausler 1985, Crolet 1983, Crolet and Bonis 1984,
flow.
corrosive conditions.
better than did C02 partial pressure. This index has been
"
log ic - - AspH + B (2 •1 )
with
As = (2.2)
RD
H2C03 + e ----> H + HC03- (2. 3)
( 2 •4 )
2H ---> H2 ( 2. 5 )
(2 •7 )
( 2 •8 )
12
be returned to later.
H+ + e = Had (2.13)
H+ + Had + e = H2 (2.14)
14
discharge, They also showed that H2S did not change the
reaction.
2.3-2.5).
HzS and COz in solutions does not change the Tafel slopes
and C02 shift the anodic and cathodic potentials toward the
discussed.
diffusion barriers.
consists of FeS2, FeS and Fe9S8, all of. which may be formed
species as follows:
(2.22)
later.
is that the flow and mass transfer rates are di~ferent for
since too many parameters are unknown for the FeC03 film to
understood.
satisfactorily.
mass transfer process for the time and flow rate range
those mentioned.
corrosion rate.
model.
charge transfer.
MODEL DESCRIPTION
Physical. Model
Figure 1.
of the tubing and the liquid moves along the tube wall as a
stronger than that the film can absorb, some liquid in the
30
31
tube wall
I
I ' . I
I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I gas core I
I I
I I
I I
roughness.
since it has been found that the wall shear stress is much
Gas Core ·
Diffusion Layer
discussed later.
product layer. The ferrous ~ons from the wall will diffuse
Model .Formulations
( 3 •2 )
(3.4')
36
(3 .5)
HS- = s= + H+ ( 3. 6)
(3•7)
H2 0 = H+ + OH- ( 3. 8)
(3.9)
[H_S- ] [H+]
Kl. H 2 s = (3.10)
[Hz S]
[ s= 1 [ H+ 1
K2 • H 2 s = ( 3.11)
[HS-]
37
[ HC03 - ] [ H+ ]
Kl,C02 = (3.12)
[ H2 C03 ]
[ C03 = ] [ H+ ]
K2 , co 2 = ( 3.13)
[ HC03- ] /
[H+ ] [OH- l
Kw = (3.14)
[H2 0]
~ z~m~ =0 ( 3.15)
i
variables: H2S1., H2C03, HS-, HCOJ-, s=, C03=, H20, H+, OH-,
Sulfur balance
Carbon balance
Hydrogen balance
are expressed as
TABLE I
Elec- Range of
trolyte A1 Az A3 A4 Validity °C
TABLE II
Elec- Range of
trolyte B1 B3 B4 Validity °C
expression
al. (1982).
TABLE III
I = 0. 5 Smi Zi 2 ( 3. 23)
is involved.
expression
-2748.51/T (3.26)
42
60 bar.
Film Layer
(1990) have made a study of mass transfer from the wall for
dp 1 d
--- + - - (r 1:) =0 ( 3. 27)
dz r· dr
at the wall
r =R T = l:w (3.28)
we have
T
- _- "t·1
= 1 - ( 1 ) 'I ( 3. 29)
Tv Tv
where
y
'I = a (3.30)
44
as
ctu
T = ( llL + llt ) ( 3. 31)
dy
T 1 vt dU+
= -a+ - 1 + ( 3. 32)
Til VL dfl
where,
a u*
a+ = ( 3. 33)
VL
U+ = U/u* (3.34)
u* =( T
w
/p )1/2
·L ( 3. 35)
( 3. 36)
dfl
I + = ( 3. 37)
II
or
amount y', and the reference plane moves with the velocity
(3.40)
46
in this model.
{3.42)
(3.43)
with
1
Uav = JR 2nrUdr
n(R2 -ri 2 ) ri
47
. r: (1 - q ) a dq (3.44)
2
f =
[r: ( 1 - q) u' dq r (3.45)
KL c5 f 1/ 2
Sh = = Ref(-- ) Sc K+ (3.47)
Dl. 2
where,
KL
K+ = u*
(3.48)
48
(3.49)
K+ = 1/Ja+
a1+ 1/ Sc +· A ( y+ ) 3
a+ dy+
::::1/ J
0
1 I Sc + A ( y+ ) 3
[ ( ASc ) - 1/ 3 + (a+ ) ]2 ] - 1
+ ln (3.50)
( ASc ) - 2 I 3 - ( ASc )- 1 I 3 a+ + (a+ ) 2
49
A = 4.5x10-4 is taken.
the following.
(3.51)
where,
8+
ft i = o.079Rec- 1 ' 4 ( 1 + 360 (3.52)
D+
wg + WLE
Pc = Pg (3.53)
wg
4 (Wg + WLE)
uc = (3.54)
1t Pc ( D-28 ) 2
50
B 6. 59Fo
= (3.55)
D (1 + 1400F0 )1/2
where,
·and
ks + = .7 0 D+ N1 o . 4 3 3 Rec - 1 • 3 s (3.58)
where,
(3.59)
through the momentum balance for the liquid film with the
following expr~ssion
also put this layer into the turbul~nt mass transfer layer.
divides the liquid film layer int~ two layers as was done
d 2 Ci d d4>
D~
dy2
+ Zi u~ F--('Ci
dy dy
=0 (3.61)
where,
Di
u~ = R T
(3.62)
dCi d4>
+ Zi Ui FCi =0 ( 3. 63)
dy dy
53
de~ d~ Ii.
D~ + z~ u~ Fe~--.- = (3.64)
dy dy n F
ei = e~ s (3.65)
At y = lh
ei = ei 1 (3.66)
Product Layer
same modifications.
Anode
Cathode
2HzC03 + 2e (3.68)
(3.70)
where,
(3.71)
al. (1961) reported that ia~ has values ranging from 1.1 x
which was used in this model. For the anodic reaction, the
et al. 1985) of
RT
Ee,M = -0.44 + ---- ln[Fe++] (3.72)
2F
reactions:
with
Ks p, Fe s = .[Fe++ ] 1 s= ] (3.76)
expressed as:
( 3. 77)
where,
(3.78)
(3.79)
ia = ic 1 + ic 2 (3.80)
additional requ~rement
59
60
are set close to 1.0 except for s= and cos=, for which a
converging.
Turbule~t Layer
with the boundary conditions ( 3.• 60) and ,( 3. 61) are first
1979) is used.
( 4. 3)
Di C1. ' + Zi Ui F ( C1. <P • ' + Ci • <P' ) = Z1. U1. FCi • <P • ' (4•6)
C1.(j+l) - Ci{j-1)
CJ. ' = (4 . 9 )
2h
<P(j+1) - <P(j-1)
<P, = (4.11)
2h
(4.12)
-<P(j-1)Ziul.FCi" + <P(j+l)ZiUJ.FCJ."
zero.
given in,Figure 3.
as discussed above.
65
START
INPUT
SET DZ
Z =Z + DZ
YES NO
NO
CALC. ELECTROLYTE EQUIL. AT
GAS AND LIQ. INTERFACE
YES
CORROSION RATE CALC.
AND OUTPUT
Partial Pressure
were taken fro• single phase liquid flow while the model is
66
2000~--------------------------------------------
o~~~~~~~~~~~~~~~~~~~~~~~~
0 100 200 300 400 500
Partial Pressure of C0 2 (psia)
f..__,
~
Q) 2000
-+-'
0
0:::
c:
0 _, _..-'
1500 ,
..,,.. .... /
/ \ J""'
Q) /
0 ~/
I.... ,...,. ., /
,......
;""
1....
~ / /
0
u ... , /
/ /
O'l
00
69
mass transfer effect has been found when the liquid flow
rate is above 3.3 ft/s. In fact, all later calculations of
70
(1984), who has pointed out that the DeWaard and Milliams
50
0
/
/
.....Q)0
0::: 30
c
0
"iii
0
L..
L..
0 20
u /
I
/
I
I
10
- - This Model Prediction with U.. = 0.29 ft/s
__ This ,Model Prediction with U.. = 2.9 ft;'s
(l]]]J Bartonicek's
Correlation Corrected to the System pH
o~~~~~~~~~~~~~~~~~~~~~~~~~
0 5 10 15 20 25 30 35
Partial Pressure of HtS (psia)
60
11>
.......
0
0:: 40
c
0
(/)
0
L..
L..
0
u
20
_ _ This 114odel Prediction with u.= 2.9 ft/s.
[[[[[] Expl. Data from Greco ond Wright (1 ~62).
~ExpL Data from Rhodes and Clark {1936)
o~_,,_.,.,.,.,TTTTTTTTTTTT,.OTr<rorrrrrrrr~~~~
0 100 200 300 400 500
Partial Pressure of C0 2 (psia)
this model, although under low liquid flow rate this model
Wright (1962) and Rhodes and Clark (~936), one find~ that
125 psi and room temperature the corrosion rate in the tap
returned to later.
12
10
c
-~ 6
0
"-
"-
0
u
4
0
0 10 20 30 40 50 60 70
Partial Pressure of H2S (psia)
addressed.
300
Q)
.......
0
0::
c 200
.2
C/l
0
,_
L..
0
(..)
100
_ _ This t.Aodel Prediction with UG ... 3 m/s
IIIDJExpl. Data from Enksrud and Sontvedt (1984)
With Peo2 = 1.56 bars.
o~~~~~~~~~~~~~~~~~~~~~~~~~
0.00 0.40 0.80 1.20 1.60
Liquid Velocity (m/s)
13
cu
......
0
0::: 11
c
.Q
(/')
0
I-
I-
810
8~~~~nn~~~~nn,.~rr~~~~~~~~~~
0.00 0.10 0.20 0.30 0.40 0.50
Liquid Velocity (m/s)
00
0
81
potential.
engineer.
Case I
service the pipe in the well has many places that have
case are given in Table IV, V and VI. The model results
82
TABLE IV
Methane 90.94
Ethane 4.37
Propane 1.14
!-butane 0.27
N-butane 0.23
!-pentane 0.13
N-pentane 0.08
Hexanes 0-.11
Nitrogen 0.25
Totals 100.00
83
TABLE V
C.onstituents PPM
Sodium 6490
Calcium 298
Magnesium 38
Barium 4
Iron 36
Chloride 10100
Sulfate 111
Bicarbonat·e 879
Total Dissolved
Solids 17~5,6
84
TABLE VI
WELL CONDITIONS FOR CASE I
Depth ft 9700
f 800
~
..........
<ll
......
0
~ 600
c
0
(/)
0
.._
I...
u0 400
200
O~~TT~~~~TT~~~~TT~rnnn~TTrrrnnn~;nM
0 2000 4000 6000 8000 10000
Depth (ft)
Figure 11. The Corrosion Rate Profile Along Depth for Case I
00
01
86
rate and not very high gas prod~ction rate. The brine
Case II
are listed in Table VII; VIII and IX. The model results
water and gas production rates compared with Case I but the
TABLE VII
Methane 90.10
Ethane. 6.00
Propane 1. 68
!-butane 0.45
N-butane 0.34
!-pentane 0.20
N-pentane 0.12
Hexanes 0.18
Nitrogen 0.22
Totals 100.00
89
TABLE VIII
Constituents PPM
Sodium 127
Calcium 21
Magnesium 0
Barium 3
Iron 0
Chloride 195
Sulfate. 0
Bicarbonate 60
Total Dissolved
Solids 406
90
TABLE IX
WELL CONDITIONS FOR CASE II
Depth ft 9620.
.........
it
~300
Q)
......
0
0:::
c::
0
·~ 200
........
0
0
100
O,,,~~~TTTTTTTTTTrrrrrrrrnrrnrnrn~~,,,_,
0 2000 4000 6000 8000 10000
Depth (ft)
Figure 12. The Corrosion Rote Profile Along Depth for Case II
92
Case III
bound all the field data because the observed field data
uncorroded parts.
Case I. From this two Figures, one can see that the model
two cases, one would notice that these two wells produce
TABLE X
GAS COMPOSITION OF THE WELL
FOR CASE III
Methane 91.60
Ethane 4.39
Propane 1.18
!-butane 0.33
N-butane 0.25
!-pentane 0.14
N-pentane 0.09
Hexanes 0.13
Nitrogen 0.30
Carbon Dioxide 1. 26
Totals 100.00
94
TABLE 4I
WATER ANALYSIS OF GAS WELL
FOR CASE III
Constituents PPM
Sodium 6280
Calcium 454
Magnesium 50
Barium 2
Iron 0
Chloride 10300
Sulfate 196
Bicarbonate 313
Total Dissolved
Solids 17595
95
TABLE XII
Depth ft 9450
(])
.......
0
0::: 400
c::
0
(/)
0
I.-
I.-
0
()
200
O~~~~~~hn~~~TT~~~~~~~~TT~~~rn
0 2000 4000 6000 8000 10000
Depth (ft)
Figure 13. The Corrosion Rate Profile Along Depth for Case Ill
25~------------------------------------------,
.,........ 0
>-
Q..
615 0 0
.....0
Q.)
a:: DO
c
0
'2.._.._ 10 0 0 0 0 0
0
0 o ________ _
DO
------------ ----
5 mo o 0 0 0 0 0 0
0 0
o~~rrnn~TTrnn<TTTT>nnoTTTTrnnoTTTOMn"~rrrl
0 2000 4000 6000 8000 10000
Depth (ft)
Figure 14. The Corrosion Rate Profile Along Depth for Case Ill
with protective film
60~------------------------------------------~
DO
s:'40
Q..
0
~
...........
0
....,
G)
0
a::: 30
c 0
0
'Ci)
0
._
~
u0 20
0
0 0
---- -a---------0-- ---------- ----
10 o
0
0 0
Q;,.,.,~~~TTTTTTTTrrrrrrrrrrrrrnrnrnnono.,,..,,;
0 2000 4000 6000 8000 10000
Depth (ft)
Figure 15. The Corrosion Rate Profile along Depth for Case I
with Protective Film
c.o
00
99
Case I with Case II and concluded 'that f,or the two cases
and III, the ~odel predicts that corrosion will start right
above. For Case III, the field data indicates that the
no simple explanation.
Case IV
TABLE XIII
Methane 90.44
Ethane 5.07
Propane 1. 36
!-butane 0.32
N-butane 0.27
!-pentane 0.15
N-pentane 0.09
Hexanes· 0.14
Nitrogen 0.12
Carbon Dioxide 1. 72
Totals 100.00
101
TABLE XIV
Depth ft 9540
ones.
Case V
are given in Table XV, XVI and XVII. The gas composition
for this well is not reported. _The only known data is that
there is 0.8% of C02 in. the gas analysis. The data listed
0
0 500 1000 1500 2000 2500 3000 3500
Depth (ft)
Figure 16. The Corrosion Rate Profile along Depth for Case IV
with Protective Film
......
0
w
104
TABLE XV
Methane 91.84
Ethane 4.59
Propane 1. 36
!-butane 0.32
N-butane 0.27
!-pentane 0.15
N-pentane 0.09
Hexanes 0.14
Nitrogen 0.12
Totals 100.00
105
TABLE XVI
Constituents PPM
Sodium 45121
Calcium 8880
Magnesium 864
Barium 17
Iron 31.5
Chloride 87500
Sulfate 248
Bicarbonate 205
Total Dissolved
Solids 143834
106
TABLE XVII
WELL CONDITIONS FOR CASE V
Depth ft 13000
predictions).
I
I
I
- - Model Prediction I
4000 - - De Woord's Model Prediction I
I
I
r-.. I
>-
0... /
~3000 /
Q)
/
...... /
0
cr /
/
c /
0
.2.... 2000 /
/
.... /
0 /
u /
/
/
./
/
1000 /
/
04Tno~~~~TnTn'"'"'"~~nononono~~~TnTnonM
0 2000 4000 6000 8000 10000 12000 14000
Depth (ft)
Figure 17. The Corrosion Rate Profile Along Depth for Case V
Without Protective Film
1-'
0
co
109
results.
-
Figure 18 shows the predicted corrosion rate with a
16, one can figure out that the corrosion rate for this
corrosive than those .in Cases I and IV, which had a real
cu
.......
0
0::: 20
c
0
f/)
0
L.
L.
0
()
19
18
174TMTMT~~~~~~~~~~~~~~~TM~~~~~
0 2000 4000 6000 8000 10000 12000 14000
Depth (ft)
Figure 18. The Corrosion Rate Profile Along Depth for Case V
with Protective Film
111
the model predicts the corrosion rate more than 12 MPY, the
Conclusions
well.
112
113
predicted from this model are much nearer the field data,
approach.
114
Recommendations
next period, when the liquid slug flows past this section
of the pipe the liquid with many small bubbles will flow
squeeze some small gas bubbles towards the wall, where the
115
with Condensate
of the Inhibitor
inhibitor.
enormous importance.
REFERENCES
117
118
GUOHAI LIU
Doctor of Philosophy
Biographical: