ENGINEERING

Corrosion Inhibition
for Severely Corrosive Gas Wells✫

M.C. Place Jr.*

ABSTRACT occurred about five months after initiation of produc-

tion. As the result of these failures, the technology
The production of the Mississippi deep sour gas presented discussed in this paper was developed.
severe corrosion problems. This paper outlines the tech- This new technology has been very successful.
niques used to solve these problems with a successful From 1974 to 1976, over 5700 joints of tubing have
continuous inhibition system and the application of this been inspected through the use of calipers and sur-
technology to a wide variety of severely corrosive produc- face tubing inspections following well workovers. Only
tions. The keys to a successful inhibition system are
three pits, which exceeded 20 mils in depth, were de-
corrosion inhibition, phase behavior, system solvency, and,
where necessary, desulfurization. tected and none was deeper than 40 mils. Since 1976,
the results have been equally successful.
KEY WORDS: carbon dioxide, inhibition, inhibition system, A successful continuous inhibition system has
phase behavior manifold characteristics that are interrelated:
INTRODUCTION (1) It must have proper phase behavior,
(2) It must be of predictable and uniform quality,
In 1969, Shell discovered the Thomasville Field (3) It must be handleable,
about 15 miles southeast of Jackson, Mississippi.1 (4) It must be able to be dehydrated,
Several fields in this area were developed in the (5) It must have a low-vapor pressure,
Smackover, and the production environments were (6) It must have sufficient sulfur solubility,
very demanding on materials technology (Table 1). (7) It must be nonscaling, and
Each well was equipped with a circulating string (8) It must prevent corrosion.
(tubing) and did not have a packer (Figure 1). The The reason for the initial failures was attributed to
casing was the pressure string. From the start of pro- vaporization of the oil carrier-inhibitor mixture under
duction, oil and inhibitor (an inhibition system) were bottom-hole conditions. This resulted in severe corro-
circulated in each well to prevent sulfur deposition and sion until uphole conditions were reached where
corrosion. condensation of the inhibition system occurred. Above
In spite of continuous inhibition, severe corrosion this point, corrosion was minimal.
caused two tubing strings to part and caused a hole in Thus, with the problem defined, the principal ap-
the third tubing string. The first hole in the tubing proach to a solution was directed toward finding an oil-
inhibitor system that yielded a sufficient liquid phase
✫
Submitted for publication October 1991. under flowing bottom-hole conditions at reasonable
* Shell Offshore Inc., 701 Poydras St., New Orleans, LA 70130. circulation rates (less than 40 barrels per million
0010-9312/92/000077/$3.00/0
CORROSION–Vol. 48, No. 4 © 1992, National Association of Corrosion Engineers 341
ENGINEERING

TABLE 1 environments and CO2 source wells. These topics will

Deep Mississippi Well Conditions be developed by the review of the initial continuous
corrosion inhibition system development, the final cor-
Depth: Greater than or equal to 23,000 ft
rosion inhibition system development, and other
Bottom-hole pressure: Greater than or equal to 23,000 psig applications, respectively.

Bottom-hole temperature: Greater than or equal to 204°C(400°F) INITIAL CONTINUOUS CORROSION

Gas composition:
INHIBITION SYSTEM DEVELOPMENT
Hydrogen sulfide 29.0-45.2%
Carbon dioxide 2.0-9.2% The first system that incorporated the concept of
Methane 54.1-67.3% phase behavior was a corrosion alleviation success
Ethane Trace
Propane plus None but less successful from an operating and economic
Water vapor Saturated perspective. The system was a mixture of a lube oil
base bright stock (a semisolid pure paraffin), which
Water: possessed the required phase behavior (Figure 2) and
Salinity Fresh to greater than 100,000 mg/L
chlorides a slightly more dense mixture of aromatics and lighter
Rate 6-30 barrels per million standard paraffins to reduce the viscosity. The resulting mixture
cubic feet was pumpable, dehydratable, and provided corrosion
protection at reasonable circulation rates for wells with
flowing bottom-hole pressures below 10,000 psig
(Figure 3).
The wells precipitated sulfur (up to 25 lb per stan-
dard million cubic feet). This initial inhibition system
had sufficient sulfur solubility at downhole tempera-
tures and pressures to dissolve the sulfur and to
prevent it from plugging the tubing. However, sulfur
extraction from this inhibition system was never ac-
complished. Sulfur levels were controlled to less than
5 wt% with blowdown. The system was also contami-
nated with significant quantities of tar, which was
believed to be the reaction product between either el-
emental sulfur and/or H2S and inhibitor fragments.
Autoclave tests indicated that the inhibitor
concentration should be greater than the sulfur con-
centration. Although this was tested in the field, it was
never accomplished because the high inhibitor con-
centrations greatly aggravated the tar problem. (The
inhibitor reacted with the sulfur to form tar, and this
was the initial, unintended sulfur removal system.)
In practice, this early system had two fatal flaws:
(1) the wells in which it was used developed severe
annular deposits—up to 3/4-in. thick in three months
FIGURE 1. Schematic of typical well completion. —that required expensive workovers (and the loss of
a well) and (2) the newer fields had initial flowing bot-
standard cubic feet [MMSCF]). This initial system ig- tom-hole pressures approaching 20,000 psig and
nored (7) above—it must be nonscaling. It was this sustained flowing bottom-hole pressures well in ex-
tunnel vision toward phase behavior, with less atten- cess of the 10,000 psig practical limit of the bright
tion to the other required inhibition system stock system.
characteristics, that led to the more intractable annular The annular deposits or scale (Table 2) were a
deposit problem. mixture of inorganic and organic materials, or bricks
This paper will discuss the parameters controlling and mortar as they were called. The inorganics were
phase behavior, the impact of phase behavior on easily identified as small amounts of iron sulfide and
other inhibition system requirements, the solution to predominantly drilling fluid solids. The drilling fluid sol-
the annular deposit problem, the impact of the annular ids were relatively easy to keep out of the system
deposit solution on the total inhibition system, and the through rigorous filtration and revised completion and
application of this technology to other corrosive sour displacement procedures.

342 CORROSION–APRIL 1992

 ENGINEERING

The organic part of the deposits presented a diffi-

cult problem. Essentially, structural analyses were
economically prohibitive. However, the presence of ni-
trogen (inhibitor was the only source of reactive
nitrogen containing compound) and sulfur in a pre-
dominantly carbon-hydrogen compound indicated that
the inhibitor and some form of sulfur may be the keys
to understanding the chemistry of the annular deposit
problem.

FINAL CORROSION INHIBITION SYSTEM

DEVELOPMENT

The problems were attacked on several interre-

lated fronts concurrently:
The corrosion problem was addressed with con-
tinued phase studies to develop a carrier oil that would
not vaporize at 20,000 psig and 204°C (400°F) in the
presence of sour gas.
The annular deposit problem was addressed on
two fronts:

—In the field, a laboratory device (Figure 4) that

simulates well conditions was built, and a series of
statistically designed experiments was run based on
varying levels of temperature, tar content, oil flow
FIGURE 2. Relationship of bright stock and sour gas at 177°C
rate, and suspended solids. (350°F).
—In the research facilities, controlled laboratory ex-
periments to generate tar from various controlled
sources of nitrogen, sulfur, and oils were conducted
to determine the combinations that led to the scale-
forming material.

Corrosion Alleviation
A good corrosion inhibitor is essential. The one
used in this system was an olefin with an imidazoline
ring. The inhibitor was thoroughly tested in rocking au-
toclaves (Figure 5). Table 3 shows some typical data.
Undiluted inhibitors will provide a liquid phase in
relatively small quantities (four to six barrels per stan-
dard million cubic feet of sour gas). However, in the
undiluted form, the inhibitors do not have all the char-
acteristics required for a good inhibitor system;
namely, it must be handleable, and, at all tempera-
tures, it must be dehydratable, and it must have
sufficient sulfur solubility.
It is also 10 times more expensive than carrier
oils; thus, the system inventory is 10 times more ex-
pensive, and this inventory—well annular volume,
pipeline volume, well site storage, and central system
processing—was essentially independent of circula-
tion rate. The total system volume approached 15,000
barrels. At $30 per barrel, system inventory requires
an investment of $450,000; at $400 per barrel (undi-
FIGURE 3. Relationship of bright stock—diluent mixture and
luted inhibitor), the system inventory requires an sour gas at 177°C (350°F).
investment of $6,000,000.

CORROSION–Vol. 48, No. 4 343

ENGINEERING

TABLE 2
Scale Analysis from Well No. 1 (March 1976)
Percents of Total Scale
Organic Residue Inorganic Residue After 800°C
Oils Sulfur After CHCl3 and CS2 Barium
(CHCl2 (CS2 Total Total (100%
Depth Soluble) Soluble) Organic Nitrogen Sulfur Inorganic Calcium(A) Barite) Iron(A) Silicon(A)

2362 39.2 4.0 30.4 0.4 2.1 26.3 4.0 5.3 2.1 2.6
3937 37.1 3.4 48.5 1.9 7.4 11.0 1.1 2.8 2.2 0.8
5827 29.4 2.0 60.6 2.6 8.7 8.0 0.6 2.0 1.6 0.4
7717 12.9 13.5 66.8 3.4 9.8 6.8 0.1 1.4 1.7 0.3
9513 19.3 5.1 69.2 3.9 12.0 6.4 0.4 1.3 1.9 0.4
11,245 24.7 3.4 50.3 2.6 9.0 21.6 1.1 5.4 3.5 1.3
13,104 27.6 4.9 52.1 2.7 9.0 15.4 0.9 3.8 2.9 1.2
14,107 28.1 3.3 53.9 2.8 8.5 14.7 0.9 3.7 2.9 1.8
14,680 36.1 3.3 51.2 2.8 7.7 9.4 0.5 2.4 1.9 0.8
16,096 15.4 7.7 36.9 1.8 7.2 40.0 2.0 9.6 6.0 4.0
17,766 18.9 3.8 28.5 1.3 7.7 48.8 2.0 12.2 4.9 2.9
(A)
No crystalline structure present.

Because inhibitors are effective in low concentra-

tions, they are not used in their neat form (Table 3). As
long as there is a hydrocarbon liquid phase present,
the inhibitors will distribute themselves between the
liquid and the gas phases. However, because of their
relatively high boiling point, they will concentrate in the
liquid phase at 204°C (400°F) and 20,000 psig with an
H2S partial pressure approximating 6000 psig. The in-
hibitors were tested (Figure 5) to simulate downhole
conditions of pressure, temperature, inhibitor concen-
tration, gas composition, and water salinity. The
inhibitor provided excellent resistance to general cor-
rosion, pitting corrosion, and crevice corrosion.
FIGURE 4. Scale test device. Providing the oil carrier was more difficult. A
carrier oil system must have (1) phase behavior prop-
erties that prevent its bottom-hole vaporization, (2) low
enough viscosity to be pumpable, filterable, etc., at
cold surface conditions, (3) solvency to dissolve any
precipitated sulfur, to dissolve the inhibitor, and to pre-
vent annular deposits of sulfur-corrosion inhibitor
reaction products, (4) low enough density to be sepa-
rated from produced water, and (5) low enough vapor
pressure to prevent system loss in the gas-liquid sepa-
ration process.
First, the system must not completely vaporize at
bottom-hole conditions. The initial system used a very
high-viscosity pure paraffin (semisolid lube oil stock)
diluted with a low-viscosity aromatic-paraffin mixture.
The diluent completely vaporized at bottom-hole con-
ditions and was no help in corrosion alleviation. Its
only purpose was to provide for acceptable surface
handling of the heavy lube oil stock.
The lube oil stock gave the required liquid pres-
ence in bottom-hole conditions. Since the diluent
provided no liquid phase, the calculated circulation
FIGURE 5. High-pressure and temperature corrosion test
rates did not consider the diluent. Laboratory and field
installation.
data suggest that continuous inhibition with 200 to

344 CORROSION–APRIL 1992

 ENGINEERING

TABLE 3
Oil Soluble Inhibitor in Sour Brine Exposure at Thomasville Test Site
Crevice Under Scale
Attack, Attack, mpy
Weight Loss Corrosion, General Pitting, mpy (Three (Three
Test % % Temp. Pressure mpy mpy Coupons per Coupons per
No. Inhibitor Demulsifier °C (°F) (psig) (Six Coupons per Test) (Six Coupons per Test) Test) Test)

1. 0 0 204 (400) 18,000 14 14 16 35 39 43 25 50 50 200 150 150 0 0 50 50 150 150

2. 2 0 204 (400) 19,000 12 12 12 13 13 14 0 0 0 0 0 0 0 0 0 50 150 150
3. 2 0.1 204 (400) 19,000 9 10 10 10 10 11 50 50 50 50 100 100 0 0 0 0 50 100

1. a. Test facility, Figure 5.

b. Nominal gas composition 35% H2S, 10% CO2, and 55% CH4.
c. Tests run for seven days in sour brine—5% NaCl + carrier oil and inhibitor.

300 lb of inhibited liquid per standard million cubic feet Aromatics sufficiently reduce viscosities and have
of gas is sufficient to alleviate corrosion. Note that lower vapor pressures than paraffins for comparable
none of these hydrocarbons are pure single-compo- viscosities, but they have higher densities. Because
nent systems. Thus, the dew point is the point where higher dilutions are required to get the needed density
only the largest molecules condense. for dehydration, higher circulation rates are necessary
The lube stock-diluent mixture had sufficient for corrosion control. As will be discussed later, aro-
phase behavior (Figure 3) to be marginally acceptable matics have superior, all-important solvency.
for the lower pressure wells (flowing bottom-hole pres- The final system was a mixture of equal volumes
sure below 10,000 psig), but it would not handle the of special No. 6 oil and a cyclic paraffin. The No. 6 oil
anticipated initial 20,000 psig flowing bottom-hole was selected because it had both acceptable and pre-
pressures of the Southwest Piney Woods Field and dictable phase behavior data (uniform crude oil to
the Harrisville Field. refinery and constant refining specifications). The cy-
A search for a new carrier oil centered on clic paraffin was selected because it had low vapor
deasphaltened oils (DAO), but it became apparent pressure (initial boiling point 301°C [575°F]), lower
that these paraffinic oils would not give the needed density than aromatics (less dilution of the No. 6 oil re-
phase behavior under bottom-hole conditions. quired to achieve a dehydratable fluid), and as a cyclic
The aromatics had superior phase behavior; how- compound, it had higher solvency than normal paraf-
ever, the best material was pitch, the high molecular fins (necessary to prevent asphaltene precipitation
weight residue after crude oil has been distilled and from the No. 6 and to prevent annular deposits).
the still bottoms vacuum flashed. No. 6 oils were also
very good because they are a blend of pitch and a sol- ANNULAR DEPOSITS
vent. These oils were semisolids at room temperature.
A low viscosity (lower viscosity than 20 SAE mo- The combined results of the scale test device and
tor oil), high molecular weight aromatic oil (cat cracker controlled laboratory autoclave experiments revealed
slurry oil) had phase behavior superior to the semi- that the elimination of annular deposits is directly
solid pure paraffins (Figure 6). The problem with these related to elemental sulfur and inhibition system sol-
oils is that they have a specific gravity of approxi- vency.
mately one; thus, they cannot be separated from the
produced water. Consequently, these oils had to be di- Scale Test Device
luted. The annular deposits occurred in a gas-free
Diluent volume is always controlled by gravity and region of the well annulus and often were under very
never by viscosity. Small diluent volumes will reduce little pressure (for flowing bottom-hole pressures be-
viscosities sufficient for pumping, etc., long before low 8800 psig, there was no pressure at the surface).
they reduce the density sufficiently for dehydration. Because the annular deposit mechanism appeared to
Paraffin diluents have the lowest density; thus, lower be pressure independent and nongas-related, an at-
diluent volumes are required. However, paraffins usu- mospheric pressure-liquid only system (no sweet or
ally have higher vapor pressures (more loss in the sour gas) was designed and built to simulate the an-
separation process), but the use of paraffinic oils with nular deposits (Figure 4).
asphaltenic oils (required for phase behavior) often re- The scale test device was a success. In concept,
sults in an unstable solution. DAO are a result of large the device was very simple. It had a feed tank(s) in
dilutions of crude oil with hot, low molecular weight which varying controlled mixtures of contaminated in-
paraffins (propane). The paraffin precipitates the hibition system were placed. A test specimen was
asphaltenes leaving a DAO. made from API tubing with one end closed. For

CORROSION–Vol. 48, No. 4 345

ENGINEERING

known (for example, sulfur was from either hydrogen

sulfide or elemental sulfur). Second, a chromato-
graphic column was developed that would separate
the large molecules so they could be examined more
selectively. The results of this work were revealing.
Hydrogen sulfide would sometimes attach at the
double bonds present in the inhibitor molecule and
perhaps also the carrier oil; in the presence of H2S
only, no sulfur-sulfur bonds with the inhibitor were de-
tected. However, when elemental sulfur was reacted,
sulfur-sulfur bonds were present. In simple terms, the
elemental sulfur vulcanized the inhibitor molecules (or
molecule fragments) very similarly to the way sulfur
vulcanizes natural rubber. With this information, des-
ulfurization of the inhibitor systems became a priority.

FIGURE 6. High-pressure inhibition system. Sulfur Removal

Sulfur removal was accomplished by scrubbing
temperature control, the tubing was internally filled the solution with sodium hydrosulfide similar to a pro-
with a heat medium oil and equipped with a heater cess used by a German company.2 The key to
and temperature controller. The tubing was placed in desulfurization was separation of the basic aqueous
an annulus made from API casing. The casing was solution and the inhibition system. Highly asphaltenic
equipped with an inlet at the bottom and an overflow. oils form emulsions, and caustic solutions enhance
The contaminated inhibition system was delivered this tendency. The result can be a very severe emul-
by an individual metering pump. The actual device sion when the two solutions are mixed to extract the
had 4 tanks, 12 metering pumps, and 12 test dissolved sulfur from the inhibition system. For ex-
specimens. This arrangement allowed several combi- ample, when these solutions were mixed in a
nations of rate, composition, and temperature to be centrifugal pump, it took 25 wt% demulsifier to break
tested concurrently. In the worst scaling systems, it the emulsion. This was not an acceptable operating
generated over 3 lb of scale (1500 g) per square foot procedure.
on an individual test specimen in a one-week test, and An overview of the inhibitor surface treatment
it differentiated between the scaling tendencies of a facilities is shown in Figure 7. The problems of dehy-
wide variety of systems. dration discussed above and desulfurization are
The initial tests were a set of statistically designed interrelated. The key to this system is the ability to
experiments with tar content, drilling mud content, separate the basic solution from the inhibition system.
temperature, and rate as variables. One of the pre- The stronger the basic solution, the more sulfur the
mises of the set of statistically designed tests was that solution can hold as polysulfide and, thus, the better
tar was the major source of the scale; this was par- the desulfurization.
tially accurate. In the worst scaling systems, more tar Conversely, the stronger the basic solution, the
left the test cell than was in the feed even though large more difficult the phase separation is to achieve.
quantities of scale were deposited from the inhibition Through laboratory experiments, a solution of 3.5%±
system; tar was being formed concurrent with scale sodium hydrosulfide, 6.5%± sodium hydroxide, and a
deposition. For this reason, the statistically designed mixing ratio of 90 percent inhibition system and 10
tests did not solve the annular deposit problem and percent aqueous basic solution gave an acceptable
will not be further reviewed. However, information compromise between effective desulfurization, reason-
gained in the statistically designed tests gave consid- able costs (caustic and demulsifier consumption), and
erable insight into the problem, and, as will be phase separation.
discussed later, the test device was essential in con- Two other factors were also important for effective
firming the problem resolution. demulsification. First, mixing between the basic aque-
ous solution and the inhibition system must be
Laboratory Autoclave Experiments thorough but gentle; motionless mixers adequately
Several mixtures of salt water, hydrogen sulfide, handled this task.
carrier oils, elemental sulfur, and inhibitors were Second, the demulsifier must be carefully se-
heated in autoclaves and analyzed by infrared spec- lected. There are three selection criteria: (1) it must
troscopy. The key to unlocking the mortar structure not adversely affect the performance of the inhibitor,
was twofold. First, the source of all components was (2) a single demulsifier must perform in the separation

346 CORROSION–APRIL 1992

 ENGINEERING

of both the produced water-inhibition system and the unwarranted complexity. (Most oilfield chemicals par-
basic aqueous solution-inhibitor solution, and (3) ex- tially degrade at the 177°C [350°F] to 204°C [400°F]
cess demulsifier must not overtreat or stabilize the bottom-hole temperatures in contact with water and
emulsion. sour gas.) Thus, a single demulsifier helped reduce
The first criterion needs little explanation. Inhibi- the system complexity.
tors and demulsifiers are both surface active In addition, the separation of the inhibition system
chemicals, and demulsifiers often reduce the effective- and the produced water was difficult because of its
ness of inhibitors and sometimes essentially neutralize density, viscosity, and chemical properties; the sepa-
the inhibitors. The corrosion problem is severe, and ration of the basic aqueous solution and the inhibitor
the inhibitor’s effectiveness must not be impaired. system was extremely difficult. Whenever this system
The author experienced complete neutralization was upset, it was essential for the operator to know
of an acid inhibitor when a demulsifier was added to that an additional demulsifier would not aggravate the
the hydrochloric acid system to prevent emulsions problem.
when acidized wells were returned to production. The With the surface facilities (Figure 7), few operating
tubing failed in some acidized oil wells because of the problems occurred, and they usually resulted from up-
added demulsifier’s effect on the acid inhibitor. The set conditions such as the following:
autoclave tests confirmed that the demulsifier slightly
reduced the effectiveness of the inhibitor to control pit- —Live steam getting into the sour-water tripper; the
ting (Table 3), but the overall results (corrosion trace boiler chemicals in the steam caused emul-
inhibition and demulsification) were acceptable. There sions.
is no substantial difference between Test 2 and Test 3 —Too high of fluids temperatures in the dehydration
except an increase in pitting when a demulsifier was tanks result in boiling.
added (Table 3). —Too low of fluids temperatures in the dehydration
For the second criteria, the chemistry of the sys- tanks impeded separation (82°C [180°F] was about
tem was already complex with inhibitors for corrosion optimum—as cool as the fluids would separate and
alleviation, oils for phase behavior, and the need for as hot as the tank gaskets could stand).
system solvency. Two separate demulsifiers, an in- Through the use of the laboratory device (Figure
hibitor, and three degradation products (two 4), Table 4 shows the scale developed from undesul-
demulsifiers and one inhibitor) were deemed an furized inhibitor system and the scale developed when

FIGURE 7. Schematic continuous inhibition system.

CORROSION–Vol. 48, No. 4 347

ENGINEERING

TABLE 4
14-Day Scale Tests—Desulfurized Oil
Inhibitor Temp. Desulfurized Oil Untreated Oil
System gal/Day °C (°F) Grams Scale %S Grams Scale %S

A 4 104 (220) — — 1500 1.8

B 4 104 (220) 5.6 0.73 502 1.1
B 7 49 (120) 0.2 0.68 9 1.2
B 7 77 (170) 5.0 0.68 17 1.2
B 7 104 (220) 25.0 0.68 70 1.2
B 7 132 (270) 85.0 0.68 200 1.2

this same system was desulfurized. Note that this de- Reserve solvency is important because the pro-
sulfurization process only removes unreacted duced oil is steam stripped (heated) and dehydrated
dissolved sulfur. (heated) before it is desulfurized (Figure 7). Some of
the sulfur will react with the inhibitor during these hot
Solvency operations. The system must have enough reserve
Solvency is important for eliminating annular de- solvency to keep these reaction products from precipi-
posits. The laboratory test device demonstrated that a tating.
scaling system can be turned into a nonscaling sys-
tem with a solvent that keeps all the inhibitor-sulfur MONITORING
reaction products dissolved (Table 5). The addition of
cat cracker slurry oil and Dutrex† eliminated scale for- Monitoring is multifaceted. First, the inhibition sys-
mation. Dutrex, like cat cracker slurry oil, is also an oil tem must be monitored to confirm that the treated
rich in aromatics. Because there is a dilution effect, inhibition system going to the wells possesses the
the system was also tested with identical amounts of wanted properties. The wells must be monitored to
less-effective diluents to demonstrate the importance confirm that they are not corroding and that they are
of solvency. free of annular deposits.
The key variables to monitor in the treated inhibi-
TABLE 5 tion system are as follows:
Scale Formation Tests • inhibitor concentration,
with Solvent Cut-Back Inhibition System(A) • free sulfur concentration,
Tar, % Scale • total sulfur concentration,
Inhibitor System Feed/Overflow GM/ft2 • density,
A 0.45/1.53 117.3* • reserve solvency, and
A + 30% SB60(1) 0.20/0.50 92.2 • aqueous phase concentration.
A + 30% D. O. #2(1) 0.40/0.54 99.7 The individual tests for each of these variables
A + 30% C.C.S.O(2) 0/0 11.0** are highly system dependent and can be run in a vari-
A + 30% Dutrex-177† Trace/0 11.4**
ety of ways.
* Plugged annulus and stuck tubing in casing. Well monitoring addresses only two issues: corro-
** Tubes appeared to have no scale.
(A)
Test conditions: 104°C (220°F), 4 gal/day, 14-day run time.
sion and annular deposits. Corrosion results were
(1)
Analyses reported above. Subsequent results have been
Component Species SB60 D.O.#2 equally successful. Monitoring the wells for annular
Paraffins 11 41 deposits is more interesting.
Napthenes
A pump test was designed and computerized to
1 ring 23 14 predict the severity of annular deposits. Because oil
2 rings 31 12 viscosity changes exponentially with temperature and
3 rings 18 5 heat-transfer calculations in downhole circulating sys-
4 rings 7 —
5 rings 1 —
tems are not precise, no attempt was made to predict
Aromatic and Compounds pump pressures during circulation tests.
with thiophenes 9 28 The designed system eliminated all variables ex-
100 100 cept annular deposits. From previously pulled wells,
(2)
Cat cracker slurry oil annular deposit profiles were developed, and pump
pressure changes were related to changes in annular
†
Trade name. deposits.

348 CORROSION–APRIL 1992

 ENGINEERING

The procedure is as follows: For a newly worked required a sidetrack and one failure required a redrill.
over well, the inhibition system is pumped down the In addition to the six failures, the redrill experienced a
open annulus at a constant rate with the well flowing failure in three months after initiation of production and
until a stable pump pressure is achieved. When a with batch inhibition. Since continuous inhibition was
stable system has been achieved, the inhibition sys- initiated in these wells in 1976 using the described
tem leaving the bottom of the well annulus equals technology, there have been no failures, and produc-
what is being pumped into the well, and the tempera- tion is continuing without any evidence of further
ture gradients are at steady state. From a known corrosion.
(calculated) flowing bottom-hole pressure and well
configuration, a computer model calculates a pseudo Eustace Field
viscosity. These tests are periodically repeated with This technology has also been used in the
identical circulation rates. Based on the following as- Eustace Field since 1980. In this application, the flow-
sumptions, the only change in the system is caused ing bottom-hole pressures were low enough that a
by annular deposits. very high-solvency inhibition system could be used,
and it was more economical to remove precipitated
—The inhibition system has the same physical sulfur with blowdown than desulfurization. This inhibi-
properties from test to test. tion system had sufficient solvency to keep all
—At a constant well production rate and pump injec- inhibitor-elemental sulfur reaction products dissolved
tion rate, heat transfer in the well bore (between the and, thus, prevent any annular deposits.
tubing and the annular fluid and between the annular The Eustace Field is located in Henderson
fluid and the formation) is constant. Thus, the aver- County in East Texas. The reservoir is a normally
age viscosity is constant. pressured Smackover carbonate at a depth of 12,500
—Adjustments can be made for changes in flowing ft. The bottom-hole temperature is 143°C (290°F). The
bottom-hole pressure. gas composition is 33% H2S and 6% CO2, and the bal-
Figure 8 shows the results of these pump tests ance is hydrocarbons. The average production is 6
before desulfurization. Figure 9 shows the results of million standard cubic feet per day (MMSCFD) with
identical pump tests on the same well following desul- 250 BPD condensate through 2 7/8-in. production tub-
furization of the inhibition system superimposed on ing, which is hung open-ended through the perforated
Figure 8. Note that the cumulative volumes of daily in- interval. Approximately 17 lb of sulfur per MMSCF of
hibition systems used in the well are approximately gas is precipitated.
equal. The slight decline in the pump pressures in Fig- A caliper survey was run after 18 months of pro-
ure 9 reflects a decline in flowing bottom-hole duction in one of the wells. No pits in excess of 20
pressures. percent of the tubing wall thickness were detected.
There is a radioactive sleeve in another well that has
OTHER APPLICATIONS also indicated nil corrosion. X-ray and ultrasonic in-
spections of the wellhead have also failed to reveal
This technology has been applied successfully in any corrosion. Likewise, subsequent calipers have
many corrosive gas fields. confirmed that there is no significant corrosion.

South State Line Field Tuscaloosa-Woodbine

The technology was successfully implemented in Another successful application of this technology
the South State Line Field. Table 6 shows well condi- has been in Tuscaloosa-Woodbine Production in
tions. Because required inhibition system volumes South Louisiana. The wells had initial bottom-hole
were low (three barrels per million standard cubic pressures of 17,000+ psi. Maximum flow rates were
feet), a once-through system was used. The once- 20 million cubic feet per day at a flowing bottom-hole
through system consisted of inhibitor dissolved in 90 pressure of 14,000 psi. The gas contains 5% CO2, 10
percent special No. 6 oil and a 10 percent diluent (to to 20 ppm H2S, and 15+ barrels per million of conden-
enable the mixture to be pumped). The wells pro- sate. Based on a caliper survey, after six months of
duced sufficient condensate to reduce the produced production (20 MMSCF/day), one well did not experi-
hydrocarbon liquids density to a level acceptable for ence any downhole corrosion.
oil-salt water separation. Formation water was highly However, CO2 corrosion is more severe in turbu-
saline. lent areas, and severe damage in the wellhead at a
The South State Line Field was discovered in bend just upstream of the choke was observed. The
1970; two additional development wells were drilled. In gas velocity was 18 fps, CO2 partial pressure was 700
these three wells—inhibited by batch techniques— psi, the wellhead temperature was 115°C (240°F), and
there were six tubing failures by 1975. One failure the well production had decreased to 15 MMSCF/day

CORROSION–Vol. 48, No. 4 349

ENGINEERING

FIGURE 8. Well no. 1 pump tests; pump rate 1.75 bpm ± FIGURE 9. Well no. 1 pump tests; pump rate 1.75 bpm ± .
(undesulfurized).

Jackson Dome CO2 Source Field

This inhibition technology was also tested on the
TABLE 6 Jackson Dome CO2 source wells. The reservoirs are
South State Line Conditions approximately 16,000 ft deep; bottom-hole tempera-
Gas Composition ture is approximately 149°C (300°F), and the wells are
Methane 45.3% slightly geopressured—10,500 psig. The two tested
Hydrogen Sulfide 29.1% reservoirs were the Smackover and the slightly deeper
Carbon Dioxide 11.8%
Norphlet. The gas analyses are shown in Table 7.
Other (N2 and C2+) 13.8%
Three production tests were run on a single well.
Bottom-Hole Temperature 177°C (350°F) The principal reasons for the first two tests were to de-
Bottom-Hole Pressure 9000± psig termine well productivity and to determine the severity
Production Rate 2 - 5 million cubic feet per day
of corrosion problems. The third test was to confirm
that the wells could be produced without plugging
problems.
TABLE 7 Initially, the Norphlet was tested for 30 days and
CO2 Source Wells the Smackover was tested for 21 days. The inhibition
Gas Composition system successfully controlled corrosion. There was a
plugging problem in the Smackover production test.
Norphlet Formation
This problem was successfully resolved, and the solu-
Carbon Dioxide 99%± tion was confirmed in a subsequent 29-day test. The
Nitrogen 0.5%±
plugging problem will be discussed later.
Methane 0.4%±
Hydrogen Sulfide 0-50 PPM As mentioned above, the test well had two zones
that were scheduled for testing. The Norphlet was
Smackover Formation
tested first since it was the deeper zone. Because it
was the first zone to be tested and because it was felt
Carbon Dioxide 98.6%±
to be the more corrosive, it was selected as the zone
Nitrogen 0.2%±
Methane 0.4%± for an extended corrosion control test.
Hydrogen Sulfide 0.7%± The Norphlet was tested for 30 days at a rate of
Carbonyl Sulfide 0.02%± about 20 MM/day. A multirate test was also run with
rates between 10 MMSCF/day and 28 MMSCF/day for
several hours.
gas with 210 barrels of condensate, 700 barrels of Following the test, the tubing was pulled and in-
brine, and 70 barrels of injected inhibition system. The spected. All tubing conformed with mill specifications
general corrosion rate in the failure mode was only for new pipe. Subsequently, six joints at the following
2 mpy. For high brine production wells, the amount of approximate depths were split open for more detailed
inhibition system rate circulated needed to be in- inspection: (1) 4 1/2-in. joint at 180 ft, (2) 4 1/2-in. joint
creased. The increase was done arbitrarily from at 4140 ft, (3) 4 1/2-in. joint at 8300 ft, (4) 4 1/2-in. joint
approximately 5 bbls/MMSCF to 7 to 10 bbls/MMSCF. at 11,400 ft, (5) 2 7/8-in. joint at 14,310 ft, and (6) 2

350 CORROSION–APRIL 1992

 ENGINEERING

FIGURE 10. Well 1-R.

7/8-in. joint at 15,900 ft. No visible corrosion was

present although the 4 1/2-in. pipe had some surface
roughness that may have been minor corrosion. Por-
tions of the above four joints (1), (2), (5), and (6) were
examined more carefully in a laboratory.
Absolutely no corrosion was found in the bottom
portion of the tubing. There was some minor pitting in
the 4 1/2-in. tubing near the surface. General pitting of
2-3 mils appeared to be present and one 5-mil pit was FIGURE 11. Smackover test desulfurized oil.
found. Using the conservative bases that the corrosion
occurred in the well and continued at a constant rate
of 5 mils/month and that the tube is replaced when the
maximum pit reaches half the pipe wall, a 5-year tub-
ing life for the upper part of tubing was projected. With
refinements in the corrosion control systems and ex-
pected retardation of the corrosion after corrosion
products are present, a 5- to 10-year tubing life ap-
pears reasonable for the upper part of the string and a
longer life for the lower portions of the tubing.

FIELD PROBLEMS

Application of this technology resulted in very few

field problems. These problems revolved around in-
hibitor solubility and aqueous phase carry-over into
the inhibitor system.
In order to keep inorganic solids out of the annu-
lus, the inhibition system was carefully filtered before it
was injected into the well annulus. The inhibitor starts
to precipitate at about 49°C (120°F) and will be re-
moved if the system is not above 49°C (120°F) during
filtration.
Trace caustic in the inhibition system caused
problems in two applications.
The first of these was Ridgway 1-R in the South-
west Piney Woods Field. As previously reported,3 this
well had a salt-plugging problem that was alleviated by
FIGURE 12. Smackover test nondesulfurized oil.
periodic water injection.

CORROSION–Vol. 48, No. 4 351

ENGINEERING

The salt-plugging problem was not caused by the inhibition system was used, no productivity loss
inhibition system. The solution to the salt problem in occurred (Figure 12).
conjunction with the inhibition system caused another
scaling or plugging problem. CONCLUSIONS
The Smackover formation brine contains CaCl2
(calcium chloride). The return fluids were contami- ❖ In a wide spectrum of highly corrosive gas wells,
nated with CaCl2. In the presence of CO2 and caustic, corrosion can be economically alleviated with a well
calcium carbonate precipitates as follows: designed and operated continuous inhibition system.
❖ For very corrosive gas wells that do not precipitate
CaCl2 + CO2 + 2NaOH — CaCO3 + 2NaCl + H2O
elemental sulfur, the inhibition system must be de-
The calcium concentration of the produced fluid is signed to incorporate the right combination of
shown in Figure 10. Note that when the well is corrosion inhibition, phase behavior, solvency, den-
inhibited with the desulfurized inhibition system, the sity, and viscosity.
calcium concentration of the produced fluid is low. ❖ For very corrosive gas wells that precipitate el-
When the inhibition system was replaced by a new emental sulfur, a combination of increased solvency
(never used and never desulfurized) inhibition and sulfur removal must be added to the above sys-
system, the calcium concentration increased signifi- tem.
cantly. ❖ Trace aqueous residuals from desulfurization with
Obviously, the change in inhibition system from a highly basic aqueous solutions may result in calcium
desulfurized one with trace caustic to an identical inhi- carbonate precipitation in wells that produce both cal-
bition system without caustic does not affect the cium and a source of carbonate (carbon dioxide or
calcium concentration of the produced brine. The trace bicarbonate). This problem was solved with appropri-
caustic in the desulfurized inhibition system did, how- ate water additions to the inhibition solution.
ever, precipitate CaCO3 in the tubing.
This precipitation of CaCO3 in the tubing caused ACKNOWLEDGMENTS
the calcium concentration in the produced water at the
surface to drop (Figure 10); the tubing restrictions, be- The author thanks Shell Oil Company for granting
cause of CaCO3 precipitation, were confirmed with permission to publish this work.
computer analyses of well flow characteristics. The
problem was solved by adding slightly saline water to REFERENCES
the inhibition system prior to injection.
1. M.C. Place Jr. “Corrosion Control—Deep Sour Gas Production,” H2S
A similar phenomenon was discovered in testing Corrosion in Oil & Gas Production. A compilation of Classic Papers.
a Smackover Formation CO2 source well (Table 7). R.N. Tuttle, R.D. Kane, eds.(Houston, TX: NACE, 1981), p. 829.
2. T.W. Hamby Jr. “Development of High Pressure Sour Gas Technology,’’
When a desulfurized inhibition system from the H2S Corrosion in Oil & Gas Production—A compilation of Classic Pa-
Thomasville system was used, the well flowed for two pers. R.N. Tuttle, R.D. Kane, eds.(Houston, TX: NACE, 1981), p. 682.
3. M.C. Place Jr., J.T. Smith, “An Unusual Case of Salt Plugging in a High-
weeks before severe loss of productivity was experi- Pressure Sour Gas Well,” 59th SPE Annual Technical Conference and
enced (Figure 11); when a nondesulfurized identical Exhibition, paper no. SPE 13246 (Houston, TX: SPE, 1984).

352 CORROSION–APRIL 1992

 • Corrosion Research Calendar •
The NACE Research Committee is responsible for disseminating informa-
tion on upcoming corrosion research meetings, symposia, and confer-
ences. All pertinent information including the time, date, and location should
be sent to W.K. Boyd, 2005 Big Bend Way, Henderson, NV 89014. NACE-
sponsored events are denoted by asterisks.

1992
June 7–11 — Offshore Mechanics October 5–7 — Corrosion Deforma-
and Arctic Engineering Conference— tion Interactions—Fountainebleau,
Calgary, Canada; Contact D.G. France; Contact T. Magnin, Universite
Morrison, Shell Development Co., P.O. de Lille I Lab. de Metallurgie Physique,
Box 481, Houston, TX 77001; 713/663- URA CNRS234, Batiment C6, F-59665
2409; fax 713/663-2233. Villeneuve d'Ascq, France; 33-
20434944; fax 33-20434040.
June 10–12 — Foamed Polymers:
Processing & Production—San Di- October 25–30 — First Pan-Ameri-
ego, CA; Contact Program Division, can Corrosion Congress—Mar de
Technomic Publishing Co. Inc., 851 New Plata, Argentina; Contact Asociacion
Holland Ave., Box 3535, Lancaster, PA Argentina de Corrosion, Comision
17604; 800/233-9936; fax 717/295- Organizador (INGAR), Avda. 3657,
4538. 3000 Santa Fe, Argentina; 54-42-34451;
fax 54-42-50944.
June 10–12 — Biocompatible Sur-
faces: Design Characterization & October 26–29 — The Microstruc-
Applications—Miami, FL; Contact Pro- tures and Mechanical Properties of
gram Division, Technomic Publishing Aging Materials—Detroit, MI; Con-
Co. Inc., 851 New Holland Ave., Box tact TMS, Meetings Department, 420
3535, Lancaster, PA 17604; 800/233- Commonwealth Drive, Warrendale, PA
9936; fax 717/295-4538. 15086; 412/776-9050; fax 412/776-
3770.
June 15–17 — International Sympo-
sium on Automation in Fatigue and November 1–5 — Environmental Ef-
Fracture Testing and Analysis— fects on Intermetallics, Ceramics, and
Paris, France; Contact Dorothy Savini, Composites—Chicago, IL; Contact
Symposium Operations, ASTM, 1916 R.H. Jones, Pacific Northwest Labora-
Race St., Philadelphia, PA 19103-1187; tory, P.O. Box 999, MS P8-15, Richland
215/299-5413. WA 99352; 509/376-4276.
June 28–July 2 — Seventh World November 16–17 — ASTM Sympo-
Conference on Titanium—San Diego, sium on Microbiologically Influenced
CA; Contact the Minerals, Metals, and Corrosion Testing—Miami, FL; Con-
Materials Society (TMS), 420 Common- tact Jeffery Kearns, Allegheny Ludlum
wealth Drive, Warrendale, PA 15086; Corp., Technical Center, Alabama Pa-
412/776-9000; telex 910 380 9397; fax cific Ave., Brackenridge, PA 15014-
412/776-3770. 1597; 412/226-6236.
June 28–July 3 — Advances in Cor- November 16–17 — ASTM Sympo-
rosion and Protection—Manchester, sium on Application of Accelerated
UK; Contact Conference Secretary Corrosion Tests to Service Life Pre-
(ACP), Corrosion and Protection Cen- diction—Miami, FL; Contact G.
tre, UMIST, P.O. Box 88, Manchester Cragnolino, 512/522-5539, or N.
M60 1QD, UK; 44-061-200-4848; fax Sridhar, 512/533-5538; Southwest Re-
44-061-200-4865. search Institute, Center for Nuclear
Waste Regulatory Analyses, 6220
* July 12–17 — Second International Culebra Rd., San Antonio, TX 78228-
Symposium on Electrochemical Im- 0510.
pedance Spectroscopy—Santa Bar-
bara, CA; Contact D.D. Macdonald, 1993
517 Deike Bldg., Penn State University,
University Park, PA 16802; 814/863- February 21–25 — Fifth International
7772; fax 814/863-4718. Symposium on Experimental Meth-
ods for Materials Science Research—
* September 7–11 — Corrosion Asia— Denver CO; Contact W.H. Hofmeister,
Singapore; Contact NACE, P.O. Box Dept of Mat. Sci. & Engg, Box 1593,
218340, Houston, TX 77218-8340; 713/ Station B, Vanderbilt University, Nash-
492-0535; telex 792310 NACEHOU; ville, TN 37235; 615/322-7053, fax 615/
fax 713/492-8254. 343-0466.
Continued on next page