+1(978)310-4246 credencewriters@gmail.com
  

Description

Briefly summarize underlying:

1) Causes of downhole corrosion/scaling experience in a high-pressure, high-temperature gas condensate producer.

2) Remediation steps taken

3) Lessons learned

SPE-169618-MS
Jawad I. Al-Tammar, Saudi Aramco; Michel Bonis, Total; Ho Jin Choi, Former Saudi Aramco Metallurgist,
currently with Samsung C&T Corporation; Yousef Al-Salim, Saudi Aramco
Copyright 2014, Society of Petroleum Engineers
This paper was prepared for presentation at the SPE International Oilfield Corrosion Conference and Exhibition held in Aberdeen, Scotland, UK, 12–13 May 2014.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
Abstract
This paper shares Saudi Aramco’s operational experience and challenges with downhole sweet and sour
corrosion and scaling tendencies in the carbon steel (CS) completions. It also reports the company’s
approach to deal with these issues. Corrosion in wells drilled in sour gas bearing reservoirs commonly
occurs in conjunction with iron sulfide scale deposits in the lowermost section of the production tubing
and liner. Historically, the progression of sour corrosion has been relatively slow, so CS completions are
cost-effective as the frequency of required workovers is low. Nevertheless, although scale deposits are not
confirmed to curtail well productivity they reduce wellbore accessibility, which sometimes necessitates
their removal by rig-less mechanical operation. Scale removal from old completions resulted in failure of
the tubular integrity. In most sour gas wells corrosion starts posing scaling and obstruction problems
inside the tubing far before any leaking risk. It is worth noting that this experience is consistent with field
cases experienced by Total or other operators, as recently reported in the literature. The company
experience with sweet gas producers have shown corrosion tendency to occur at shallow depths with no
observed scale deposits. Corrosion was in the form of localized spaced scratches in the joints and metal
loss in the pin- and box-ends. For these wells, the average life cycle of tubulars exceeds 10 years. The
results from this field study might benefit the world gas producers of Saudi Aramco’s intensive and
diversified experience and of the completion practices that the company has adapted to develop the high
temperature gas condensate fields. It is of particular interest to understand why the corrosion observed on
these sweet wells is significantly lower than predicted by most carbon dioxide (CO2) corrosion prediction
models.
Introduction
Saudi Aramco started the deep gas condensate development program in 1984 by producing a prolific sour
carbonate reservoir (Khuff) that underlay the well-known massive Ghawar oil field. Since that time, the
company gas system was expanded by installing gas processing units to the existing associated gas
processing facilities and by building new nonassociated gas facilities. The produced gas is used locally for
power generation and as a feed stock for the growing local chemical industry. Though reservoir conditions
are not in the range of what is currently named as high-pressure/high temperature (HP/HT) reservoirs
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Saudi Aramco Downhole Corrosion/Scaling Operational Experience and
Challenges in HP/HT Gas Condensate Producers
2
SPE-169618-MS
Downhole Tubulars Material Selection
From the start of the deep sour gas condensate development project, the company elected to complete
wells with the T-95 low alloy CS (0.5% to 1.0% Cr, 0.2% Ni and 0.2% Mo). The T-95 CS completion
was chosen as the company’s cost-effective option for the large scale onshore gas development where the
cost of the alternative choice Alloy 28 (27% Cr, 31% Ni, 3.5% Mo and 1% Cu) was about tenfold. About
95% of the gas wells have been completed with CS tubulars, and the remaining 5% with S13Cr (13% Cr,
4% Ni1% to 2% Mo) for some of the sweet producers. The corrosion resistant alloys (CRA), S13Cr, was
considered in 2001 for the new sweet completions, but shortly was discontinued due to high capital cost,
especially for large scale developments.
The completion strategy implemented by Saudi Aramco has proven sound and effective throughout the
years. The company is still using CS as the cost-effective completion option for large-scale sweet and sour
gas developments. For offshore operations, where reliability and durability play an ultimate role, the
company installed CRAs (Alloy 28 for Khuff reservoir with H2S⫽2% and CO2⫽6%). Likewise, CRAs
are being used by other operators in offshore fields. The author’s understanding is that not much is shared
in the literature of the industry’s CRA completion experience. On the other hand, abrasion resistant
nonmetallic coatings, such as fluoropolymer are also being investigated by the company for trial test.
Corrosion Forming Mechanism
There have been a large amount of studies and published literature on corrosion forming mechanisms in
sweet and sour conditions, the sweet part without H2S has been the dominant area of investigation by far.
Though all detailed mechanisms are not yet fully agreed upon, in particular with H2S, a substantial amount
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
(10,000 psi/300 °F), most conditions are in the high
temperature range and initial pressures also are
quite high, asreservoir temperatures and pressures
were 270 °F to 320 °F and 4,000 psi to 8,500 psi,
respectively. Acid gas contents of the Khuff gas
streams were in the range of 0 – 6% hydrogen sulfide (H2S) and 0.5– 4% carbon dioxide (CO2),
while the sweet sandstone development have zero
H2S and 2– 4% CO2. The produced streams compose of 70 – 80% methane and 3–14% nitrogen.
Produced high API gravity hydrocarbon condensate rates ranges between 20 –300 bbls/MMscfd.
Produced water is a mixture of condensed water
vapor and some interstitial water, which is dependent on flow stripping velocities. The TDS of the
produced formation water reaches up to 300,000
mg/l. The average water production rate is less than
Figure 1—Typical gas well completion sketch.
2 bbls per MMscfd. Wells were completed away
from gas-water contacts to minimize water production. Wells were completed with low alloy API
5CT T-95 carbon steel (CS). Tubing sizes are 7”, 5½” and 4½”, in accordance with well productivity. A
typical well completion is shown in Fig. 1.
The goal of this paper is to first give an overview on corrosion and scaling issues experienced on wells
producing from these sweet and sour reservoirs for more than 20 years. The experience gained by other
operators is also used inasmuch as it consolidates our findings and contributes to our understanding of
what has been observed on these wells.
SPE-169618-MS
3
Depending on the water composition, H2S and CO2 contents and total pressures, the resulting pH
typically ranges from 3 to 6.5. Under these acidic reducing conditions, the primary corrosion product is
Fe⫹⫹, while bicarbonate anions (and also bisulfide with H2S) are also present in the water as corrosion
byproducts. The point is that iron carbonate is a quite insoluble corrosion product and that any of the
various possible iron sulfides is even less soluble, by a factor 1,000 to 100,000.
Under the so-called sweet conditions, i.e., with no or very minor H2S content, an iron carbonate layer
may or may not form on the steel surface as a thin layer depending on the pH of the water and the amount
of CO2 and bicarbonates. The resulting corrosion morphology may then range from a severe uniform one
to a severe localized one or even to a low residual corrosion, depending on the water composition and on
the behavior of this iron carbonate layer. On the other hand, a thick layer is rarely observed over a
corroding surface because of the intermediate solubility of iron carbonate.
On the other hand, when significant amounts of H2S are present, an iron sulfide layer is very quickly
and easily formed over any corroding area because of the very low solubility of any of all iron sulfides
(mackinawite, greigite, pyrrotite, pyrite, etc.). In general, a combination of several of these iron sulfides
is observed on real corrosion products. The point is that a great variety of morphologies of such sulfides
may be observed from very thin to thick ones protruding above a still corroding surface. In other cases,
the iron sulfide layers formed show a poor adherence to the surface from where they originate so they are
transported in suspension and may accumulate at further locations. We will see hereafter that, in Saudi
Aramco gas wells, the thick protruding corrosion layers are those posing the most problems.
As far as the development of the so-called “H2S ⫹ CO2 corrosion” is concerned, depending on the
environmental conditions the iron sulfides formed may lead to a very low residual corrosion below a
protecting layer, to a moderate corrosion below a semi-protective layer or to dramatic localized corrosion
where the iron sulfide layer is assumed acting as a galvanic enhancer. The exact conditions leading to
these various corrosion developments are still not well-known or controversial. In particular, whether
chlorides or salts are promoting or not, the development of a localized corrosion is still a matter of
discussion1. On the other hand, it is generally acknowledged that trace levels of oxygen entering in a sour
environment, or a natural deposition of sulfur from the gas, may lead to very severe localized
corrosion damages. It is also generally recognized that, on average, among a large number of field
experiences, the presence of H2S in addition to CO2 leads to less corrosive environments than when
only CO2 is present1.
It is finally noted that H2S has another well-known and extremely severe damaging effect: the sulfide
stress cracking. This failure mode is, however, not discussed in this paper. It is well acknowledged that
carbon and low-alloy steels are adequate for use in oil field applications as long as they meet ISO 15156
(heat treatment, hardness, nickel contents, cold working, etc.) standards.
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
of intermediate phenomena are well described and acknowledged. Our objective in this paper is not to
discuss these mechanisms in any detail. We only intend to introduce a few basic considerations of interest
and to discuss the experience reported hereafter.
It is agreed that the primary cause of corrosion in gas producing wells is the presence of CO2, H2S, and
water in the flow stream. Ideally, if these gases could be excluded and water maintained at a neutral pH
or higher, the presence of water would cause very few corrosion problems. Similarly both CO2 and H2S
are noncorrosive in the absence of moisture. This being said, the question may come about the critical
levels of water or CO2 and H2S above which corrosion starts becoming an issue. This will certainly be
a matter of thought in some cases discussed hereafter. It is also well acknowledged that CO2 and H2S are
acting as weak acids when combined with water:
4
SPE-169618-MS
Table 1—Tubing life of sour gas producers completed with CS tubulars
NACE MR-0175 – ISO 15156 has classified sweet and sour corrosion conditions based on a partial
pressure of H2S greater than 0.05 psi. This classification has indeed been defined for sulfide stress
cracking resistance in various oil field applications. A limit of 0.05 psi does not necessarily apply to the
transition from sweet corrosion to H2S ⫹ CO2 corrosion: Saudi Aramco has employed a classification for
gas production where sweet or sour streams are defined based on the following criteria:
●
●
●
Gas streams with an H2S/CO2 ratio of â±– 0.05 are classified as sour.
Gas streams with H2S ⱖ100 ppm and an H2S/CO2 ratio of ⱕ 0.05 are classified as mildly sour.
Gas streams with H2S level of ⱕ 100 ppm (~0.05 psi H2S) are classified as sweet.
Sour Gas Corrosion Experience
The initial gas development program was started in 1984 from two sour fields SG and UT with small
number of high gas rate wells. The program was expanded massively in 2000 and onwards to include more
sour and sweet fields. The majority of gas wells have less than 13 years of production life. Downhole
corrosion inhibition and monitoring programs were considered at the early life of the field development
but soon were discontinued for reasons that will be explained in this paper. Well workovers, surveillance,
intervention and producing field data gave insight of the downhole corrosion and scaling tendencies.
Table 1 presents a review of the tubing life cycle of the initial sour gas producers. The completions
survived for up to 20 years before a tubing failure happened, as a result of internal corrosion that would
require a workover rig to replace the production tubing. In some cases the workover operations were
lengthy due to badly deteriorated tubing, which required extended fishing operations.
Corrosion Inhibition
Field wide downhole continuous corrosion inhibition was not part of the well completion design
philosophy of the deep gas fields from the start. A large-scale downhole corrosion inhibition program was
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Gas Stream Classification
SPE-169618-MS
5
considered for implementation at the beginning of the gas development program, but the idea was
abandoned following disappointing results from a testing pilot. Two different downhole inhibition
methods were tested: batch treatments and nitrogen misted squeezes. Ten monthly batch treatments of a
commercial inhibitor mixed with diesel were implemented in gas producer SG-11 and two in SG-14. The
treatment’s effectiveness in SG-11 was evaluated by running a base and a time-lapse caliper log eight
months apart. Although no corrosion was recorded in the well at the end of the trial period, the results from
the pilot test were inconclusive. Two nitrogen misted squeeze treatments were pumped in high gas
producer SG-4, but the treatments were suspended due to high cost and suspected formation damage.
Therefore, the pilot study was concluded and the downhole corrosion inhibition program was abandoned.
Corrosion Monitoring
Early in the sour gas development program the company considered implementing a downhole corrosion
monitoring program using a multifinger (feeler) caliper tool. Such a tool, when run across the completion
from top to bottom, measures the tubing’s inside diameter. The corrosion or scale build up are calculated
after each subsequent run of the tool during the producing life of the well. The caliper program was tried
on five wells but did not last long due to constant tool sticking and malfunctioning during logging
operation, which resulted in erroneous data recording. Loose pieces of scale from the tubing wall
sometimes become lodged between the logging tool feelers, which affected the tool readings. Consequently, the corrosion monitoring program on the sour gas producers was discontinued permanently.
Field Observations
Throughout the producing life of the gas fields, the company constantly strived to collect and maintain any
corrosion and scale related field observations obtained from well surveillance and/or workover operations.
When conducting a surveillance work, the wellbore is drifted sometimes with multiple sizes of gauge rings
conveyed through a slick line unit to ensure wellbore path clearance. Reduction of tubing inside size and
or obstructions will be investigated for possible onset of scale formation. Pieces of dislodged scale are
sometimes recovered during well intervening operations together with the test tools. In some instances,
scale pieces were observed to be produced and accumulated from some wells over time in the well surface
piping upstream of the flow control valve. Figure 2 show photos taken for scale pieces accumulated
upstream of the wellhead choke.
During the producing life of the field, wells will undergo workover operations for several reasons that
may include but are not limited to the following: failed corroded tubing, stuck fish, leaking packer, reentry
to sidetrack and drill into other parts of producing zone, collapsed external casing, failed gas-left
mandrels, etc. Whatever the reason of such workovers, they allowed a close look into the corrosion/scaling
condition of the pulled tubing. Figure 3 shows photos taken during field inspection operations conducted
on removed tubing from sour gas producers. Internal corrosion in the sour Khuff gas producers was
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Figure 2—Scale accumulation in surface piping upstream of choke.
6
SPE-169618-MS
observed to coexist with scale, which is mostly corrosion byproducts that was formed on the spot or
transported and deposited during production2,3.
Field observations showed that downhole corrosion/scaling activities in the sour gas producers
dominate in the lower part of the production string. Corrosion has been found to occur underneath the
scale deposits. Scale initiated as scattered crystalized lumps are initiated from corrosion pits and with time
grow up and join and become larger. Apparently after some time they may cover most of the pipe surface
area. The thickness varied from thin to thick, flat or irregular surface, contingent on the H2S content of
the flow. Generally speaking, scale formation followed a layering pattern, possibly due to depositing at
various times and conditions. The layer nearest to the pipe wall showed mainly iron sulfides. The layer
close to the producing fluids consisted mainly of iron sulfide and other scale types, Table 2. Impurities
of conventional scales and others were also found. The geochemical data in hand shows that most of the
wells produce relatively low to nil quantities of formation water at the wellhead as apparent from the
chloride contents of the produced water. Only a few of these wells produced formation water and they
were produced for short periods. We have therefore no field data that supports the increased corrosion or
scaling tendencies with the increase of formation water production. We also have no field data to
document whether more conventional mineral scales (CaCO3, CaSO4 and BaSO4) as well as FeCO3
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Figure 3—Scale and corrosion deposits in tubing recovered during workovers.
SPE-169618-MS
7
Table 2—Scale composition
Scale Formation Mechanism
Laboratory analyses of many collected scale samples have been carried out; results are shown in Table 2.
The scale is a mixture of iron sulfides (predominantly pyrohotite and troilite), iron carbonate and small
quantities of chlorides. Iron oxides are also found but it is quite possible that these are formed from other
corrosion products, which may have oxidized after exposure to the atmosphere (we have, however, no
strict evidence to fully support this possibility). The source of iron in scale deposits has been the subject
of considerable internal debate. Internal studies and samples recovered from retrieved tubing have shown
that one of the iron sources comes from the corrosion process that occurs in the tubing as corrosion is
observed just below the removed scale. Other sources have not been conclusively identified. Steps to
prevent introducing iron sources into the flow stream, such as ensuring that no iron cross-linked
compounds are used in fracturing fluids, and mixing iron scavengers and corrosion inhibitors with the
treatment fluid mixes, have been consistently implemented throughout the gas development program to
prevent any other sources of precipitation. Testing showed that the source of the iron does not appear to
be from the produced formation water or the reservoir rock as iron sulfides are too insoluble to assure any
significant dissolved iron in the formation water.
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
siderite, may become more dominant deposits in the wellbore tubular with the increase of formation water
production. Some engineers state that scale formation is taking place in the reservoir, however, this
postulate is not very well supported and the experienced well gas rate declines could be attributed to other
reasons, such as reservoir natural depletion, water or condensate loading, etc.
It is also worth mentioning that the synthetic review of field experience on sour oil and gas fields
published in 20064 and 20091 also reports similar findings that corrosion on gas wells is almost uniquely
observed at the bottom of wells producing low to nil amounts of formation water. The fact that this
corrosion is associated with protruding corrosion layers is also widely observed, as these layers generally
include other mineral components, such as iron carbonate, calcium carbonate and others.
8
SPE-169618-MS
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
All of this strongly suggests that it is a corrosion product that comes from tubing, though a minor
contribution from various fluids lost in the formation during drilling, completion and stimulation
operations cannot be strictly excluded. The highly corrosive environment for the tubular at such high
temperatures with acidic gases (H2S, CO2), water, high salinity/chlorides, and probably oxygen that might
be introduced during stimulations, results in the formation of iron compounds that comprise about 90%
of the scales. The presence of calcite and dolomite is also worth noting: calcium and magnesium are
definitely components from formation water. As a consequence, even if the water produced at the
wellhead is essentially condensed water, i.e., with a low salt content, the water at the bottom of the well
is certainly more saline.
As far as the origin of scales is concerned, the final authors’ opinion is that they are mostly if not
uniquely corrosion layers, as other sources of iron can hardly be found. More precisely, these corrosion
scales are formed at the exact location where corrosion occurs, as observed by visual inspection. The
growth of these layers is therefore directly related to the progressing corrosion underneath. It is indeed not
really a surprise that such an insoluble product as iron sulfide easily precipitates with H2S contents in the
range of 1% to 4%.
The practical consequence of this first consideration is that these corrosion layers are not formally
protective, since corrosion continues, although at a quite moderate rate (typically ⬍ 1 mm/y). On the other
hand, these layers assure a bit of protectiveness since it is reported in this paper that their mechanical
removal tends decreasing the remaining service life, i.e., increasing the corrosion rate. Another argument
is that the average corrosion rates of 0.5 to 1 mm/y corresponding to service lives of 10 to 20 years are
far below the corrosiveness that the water in place would have towards a bare steel surface if no corrosion
layer was formed (at least 10 mm/y as calculated from current corrosion models or measured from short
duration corrosion measurements by Linear Polarization Resistance, before scale deposition).
Another key finding of importance in such scales is the presence of calcite and aragonite, as well as
some chlorides: this means that, at this bottom and hot location of gas wells where fresh condensed water
has almost not started to form, the water in place is saline reservoir water. It is strongly believed that this
presence of reservoir water, which has already been reported by other authors, is decisive to the
occurrence of corrosion and consequently to the scale deposition.
The next question is why some reservoir water may be present at this location in a sufficient amount
to induce corrosion, when almost no water and no salts are produced at the wellhead (~2 bbl/MMscf at
wellhead as mentioned earlier, i.e., 10 to 100 times less at the bottom of the well, before water
condensation takes place, when the salinity at the wellhead is 10 to 100 times lower than that of the
formation water). The opinion of the authors is that this is related to low enough gas velocities at the
bottom of the wells (because of the high-pressure present), which allows some slug flow and an
accumulation at the bottom of the tubing of the trace amount of formation water extracted from the
reservoir. As a consequence, reservoir water may remain accumulated at the bottom of the wells despite
very low water production rates as long as the gas velocity is too low to lift it (typically below 2 m/s to
5 m/s depending on the tubing diameter and inclination). This tentative explanation has already been
expressed1 and looks consistent with previous findings from other field cases that the corrosion at the
bottom of the well tends to be reduced or disappears when the gas flow velocity increases due to pressure
decline, while it extends when the rate of production of reservoir water increases. The situation is of
course different when the wells start producing high amounts of formation water until the wellhead. Under
such conditions the corrosion tends to extend over a longer length of the tubing. A slug flow regime is
also no longer needed to assure a massive contact between the formation water and the steel surface.
In summary, it is believed, jointly from this experience and from many other consistent findings
already published4,1, that the scaling at the bottom of the wells are a corrosion layer formed in situ because
of the accumulation of reservoir water under slug flow conditions due to low gas velocities, despite a very
low amount of water being produced. At the opposite, the lack of corrosion almost systematically
SPE-169618-MS
9
Tubing Scale Removal Challenges
Although the presence of scale deposits in gas producers has rarely caused reduction of well productivity,
it has curtailed full wellbore accessibility in many wells, thereby preventing surveillance work for
reservoir management purposes. Historically, in very few cases the presence of scale resulted in well
surveillance tools becoming stuck, wellhead valves becoming plugged and rate being dropped.
Early on in the development, lab tests showed that iron sulfide scale crystallites are bounded together
with high molecular weight hydrocarbons (Bitumen, asphaltene and varnish). The hydrocarbons coat the
scale, thereby making it impossible to dissolve the iron sulfide scale crystals with acid. In 1990, a
particular descaling compound was pumped in two wells to remove the hydrocarbon binder, allowing the
iron sulfide scale crystallites to break apart and be flowed out of the well. This compound was pumped
in the tubing and allowed to soak for 12 hrs. The chemical treatment was cleaned out and followed with
acid treatment, which successfully dissolved the exposed scale crystallites that could not be removed by
the acid or descaling compound alone; however, removing the iron sulfide scale layer exposed the bare
pipe, which was severely corroded and resulted in losing its integrity. Both wells were shut in and worked
over to replace completion. The scale layer appears to offer some degree of corrosion protection and is
strong enough to retain pipe integrity. Following the two treatments, scale removal operations were put
on band until 2006.
An average of six scale removal jobs were performed a year, mostly to allow for setting a whipstock
on the bottom to allow for sidetracking off the main wellbore and drill multilaterals to improve well
productivity. Two methodologies were used to remove scale deposits to gain wellbore accessibility, which
are chemical and mechanical. In most of the chemically treated wells the pumped chemicals were able to
dissolve the solid scale deposits, but the mineral compounds remained in suspension in the wellbore fluids
and re-precipitated at favorable pressure and temperature conditions. In several cases, the re-precipitated
scale traveled down the wellbore into the formation and created formation damage, which lowered well
productivity. This led to abandoning the chemical removal method and pursuing a mechanical removal
option. Mechanical removal is the current method and is performed using different types of tools that are
run on coiled tubing. The main drawback of this methodology is that it is time-consuming and expensive,
because it requires extensive utilization of coiled tubing.
Generally speaking, the scale removal operations resulted in the exposure of the CS surface that was
severely corroded underneath the scale in some areas. About 60% of the jobs resulted in tubing failure that
required rig workover operation to replace the corrosion penetrated tubing6,7,8.
Corrosion/Scale Mitigation Study
Saudi Aramco has recently initiated a research study in collaboration with University scale consultants
from the U.K. to understand the scaling issues and provide problem analysis and recommendations for the
path forward in controlling the scale formation in the sour gas wells. The study identified that preliminary
scale prediction calculations suggest that the iron sulfide deposition can be mitigated or significantly
reduced by effective corrosion management. The study proposed potential lab and modeling work to
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
observed on the upper sections of sour gas wells is assumed to be related to higher gas velocities, inducing
an annular flow regime, i.e., only with a thin layer of flowing liquid on the surface. The flow velocity and
flow regime are therefore the most decisive corrosion factors as long as this mechanism is agreed. The fact
that the water present on the upper section is a diluted one is also of importance.
It is also believed, or speculated, that the protruding and nonprotective characteristic of the iron sulfide
scale formed at the bottom of sour wells is related to the saturation in calcium carbonate of the water
present at that location5. On the other hand, a protective layer is formed on the upper sections when the
water is diluted by the condensing part, i.e., no longer saturated in CaCO3; however, the authors would
like to mention that this hypothesis is not yet supported by any decisive demonstration.
10
SPE-169618-MS
Sweet Gas Corrosion Experience
In 2000, the company embarked on a huge gas expansion development to meet local gas demand. A
significant portion of the new gas fields produced sweet gas that contains nil H2S and 1% to ⫺3% CO2.
For sweet gas production, initially, the company had considered a general industry rule of thumb of
produced hydrocarbon to water ratio of 4:1 that will provide a coating protective layer on the CS tubular,
which will mitigate CO2 corrosion. Knowing that the company gas fields were expected to produce at
ratios of 20/1 up to 100/1, CO2 corrosion would not be a serious concern.
Simulation runs using industry recognized simulation models predicted high corrosion rates in the
range of 80 mpy to 400 mpy for the low and high condensate wells (the acceptable corrosion rate is 15
mpy). Accordingly, the 4-to-1 condensate ratio was not included as adequate protection criteria for the CS
tubular. As a result, the need to study the effects of CO2 corrosion in more detail was identified. Shortly,
a team was formed to verify the aggressiveness of CO2 corrosion for the future sweet gas producing fields,
and to identify the necessary action to be taken to prevent or mitigate sweet gas corrosion. The team
consisted of representatives from various disciplines with the mission of conducting laboratory testing,
simulation modeling and literature review of the impact of CO2 corrosion on gas wells tubular.
In-house laboratory testing results showed that produced condensate did not provide natural protection
on the studied sweet gas systems. Increasing chloride concentration levels in brine resulted only in slightly
higher pitting tendency, and most importantly, CS suffered high corrosion under the simulated sweet gas
downhole conditions.
The team joined an out-of-company industry research, conducted by a Norwegian company called
Institute for Energy Technology (IFE), where 14 corrosion computer simulation models, submitted by
various operators, were benchmarked against actual field conditions. In addition to the benchmarking
results, IFE was provided two scenario models: one with condensed water and another with formation
produced water, using the results from the simulator runs. Eight of the corrosion prediction models were
used. The models were run at four depths in the well. The study concluded that both scenarios would
require either CS with continuous downhole inhibition or use of CRAs such as S13Cr.
An extensive literature search was conducted of numerous technical resources on sweet gas corrosion,
also some team members met with corrosion experts from major operating companies. The consensus was
to conduct an economics feasibility study for two options, either complete wells with CRAs or complete
with CS and continuous downhole corrosion inhibition.
The team’s investigative work supported the simulation results that show CS tubular to fail in 9 –18
months. Based on this premise, and since continuous downhole inhibition was considered to add
complexity to the well integrity and operation, the team recommended using S13Cr tubing for wells with
H2S content of less than 100 ppm and high nickel alloys for wells with H2S â±–100 ppm, but where the
H2S/CO2 ratio is ⱕ 0.5. The first purchase of S13Cr tubular was made based on these guidelines; however,
due to the significant variation in tubular costs, the company issued a directive to defer all additional alloy
tubular purchases and to run CS in all wells. Also, a corrosion monitoring pilot program was initiated for
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
design inhibition treatments that will inhibit iron sulfide from being deposited in the wellbore, but it
anticipates its deposition to take place uphole in the surface facilities. The use of CRAs as tubing materials
might also be a solution; however, none of these two solutions have yet been considered cost-effective for
the large number of wells in Saudi Aramco’s onshore fields. In addition to resolving the iron sulfide
scaling issue, it will still leave a calcium carbonate problem on wells producing some formation water that
will also require management. This clearly makes us wonder of the practicality and cost effectiveness of
controlling scaling in such a complex environment. A solution for such a complex case will be a true
breakthrough in technology, as long as it does not impact too dramatically the operation of a large number
of dispersed wells.
SPE-169618-MS
11
the sweet wells that are completed with CS to
ensure a timely response to any surprises and to
adjust the completion strategy as needed.
The corrosion monitoring program was implemented on all sweet and mildly sour gas producers
using a 40-finger caliper tool, Figure 4, to measure
the change of the tubing inner diameter with time
on any of the 40 fingers due to corrosion during
production. A base log is run prior to putting the
well on production and subsequent time-lapse logs
are run throughout the producing life of the well.
Initially, the time-lapse runs were on a three month
basis, but later the logging frequency was adjusted
in accordance to previous log results.
Inhibitor batch treatments were reinitiated in
sweet gas producers when the caliper logging program was started. At the time, a batch treatment
Figure 4 —Caliper tool.
was pumped in an attempt to coat the tubulars
immediately after a caliper log was run, but the
treatments were also discontinued after a brief period because their effectiveness was suspect when
factors, such as high wellbore temperature (disintegrate the corrosion chemical film) and high gas rate
production (strip out the film), were taken into account.
Field Observations
To-date, the logging campaign has shown that corrosion rates in the gas producers are much lower than
what was initially predicted. The 9 –18 months warning critical stage has already passed by many folds
with no single catastrophic corrosion related completion failure. The production mode was continuous
throughout the elapsed years.
In late 2008 and in the first quarter of 2009, four wells were worked over based on caliper logging data,
which showed a measured localized weight loss of 32% or greater of the tubing wall thickness (the 32%
weight loss was the company set workover criteria). The removed tubing was visually inspected where
several selected joints across the tubing string were picked up and sent to the lab for close inspection. The
pipe cuts were slid in halves and examined by metallurgists. Table 3 presents a summary of the inspection
results from these wells. The table presents the maximum measured localized corrosion pits of the selected
joints. The results, which confirm log measured data, were less than 32% except for one well that had a
localized 56% weight loss in an out of specification short joint. The pup-joint which was upstream of an
R-nipple suffered the greatest metal loss (200-mils of metal loss) because the pup joint had a low
chromium content (0.2%Cr), while API 5CT T-95 carbon steel tubing contained as much as 1%Cr10.
Figure 5 presents a selection of photos taken for retrieved tubing during the highlighted corrosion
related workovers. Two major observations that are worth highlighting: CO2 corrosion took place on
straight aligned track sites and was severe on the pin- and box-ends. These corrosion grooves
developed during the production life of the well as a result of sweet flow influenced the corrosion
effect that took place at diameter changes present at pin- and box-ends and on the caliper feelers
tracks. Apparently, the FeCO3 iron carbonate corrosion film called siderite is weak enough to be
stripped off by the caliper feelers while logging. At pin- and box-ends, it may be assumed that
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Corrosion Monitoring
12
SPE-169618-MS
Table 3—Corrosion wall loss measurements for the removed tubing
corrosion takes place at locations of water accumulation. The exact reasons why no protective layer
remains at this location shall be discussed later9,10.
Figure 6 depicts a typical processed caliper log data at different tubing depths. The caliper log clearly
identifies the extent and severity of corrosion across the tubing circumference by rerecording data points
at every 0.1” of the tubing depth.
Field data showed that sweet corrosion predominantly occurred in the upper part of the tubing string.
Our present analysis of these findings is that:
1. The water produced has the ability to corrode the tubing, since some corrosion is definitely
observed. This corrosive behavior towards bare steel is probably not so far from what was
predicted by models mentioned prevoiusly (5 to 10 mm/y), if we assume a complete water wetting
and no protective layer on the surface. The occurrence of this corrosion at some locations also
indicates that the low water cut of these wells (1% to 5%) does not fully prevent water wetting
at some locations.
2. Subsequently, corrosion does not occur on most surfaces (except the specific locations indicated
earlier, and at much lower corrosion rates than the worst water corrosiveness): this definitely
means that a good protectiveness is gained over these surfaces, i.e., that a protecting carbonate
layer is formed. The author’s opinion is that this is related to the very limited amount of water
present and to an annular flow regime inducing only a very thin water layer on the surface. Such
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Figure 5—Sections of retrieved tubings from sweet gas producers.
13
SPE-169618-MS
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
Figure 6 —Caliper log data from selected depths of a gas well.
14
SPE-169618-MS
In conclusion, the amount of condensed water present on the upper part of the tubing and the flow
regime are the two main factors influencing the corrosion inside these wells. We therefore believe,
consistently with all the facts in hand, that corrosion did remain low because of the low amount of
condensing water from bottom to top, as a result of the particular pressure and temperature profiles of
these wells. In addition, an annular flow regime has caused the water layer to remain very thin and very
easy to become saturated in iron carbonate. It is also possible, though not documented in these fields, that
this condensed water does not contain too much organic acids. We cannot, however, comment on this
possibility as we do not have enough data at this point.
Finally, the main contribution of the flow toward the few corroding areas is certainly to bring the water
to these locations rather than being involved in any mechanical effect toward the corrosion layer, e.g., via
a shearing effect.
Conclusions
1. Downhole corrosion in our sour gas wells occurs in the lower part of the production tubing and
is associated with the growth of a thick corrosion scale. This finding is consistent with many other
cases summarized in the recent literature9, though these cases are from fields with significantly
different H2S and CO2 contents, temperatures and pressures. This corrosion is predicted resulting
from an accumulation of formation water in this part of the tubing when the gas velocity is too
low to lift this water. The flow regime is therefore considered as an important corrosion factor in
sour gas wells.
2. Scale deposits are predominantly corrosion byproducts, where90% of the scales are corrosion
byproducts, iron sulfides (FeSn⫹1), 10% mix of iron carbonates (FeCO3) and iron oxides (Fe3O4)
— possibly formed after exposure to air.
3. Field data shows that scale deposits tend to be thicker on high H2S wells. Scale deposits vary from
patches to continuous layers; apparently this started as localized patches and then grew bigger into
continuous sheets. These scales pose more operational problems due to restrictions of the tubing
diameter than integrity issues due to tubing perforation: The experience proves that such
perforation requires 15 to 20 years on these sour gas wells.
4. About 60% of the scale removal operations resulted in either accelerating failure of the completion or immediate loss of the well integrity and required a workover rig to restore the well. Scale
removal operations were cumbersome and costly. The mechanical scale removal method is the
company prevailing technique.
5. Field data showed that sweet corrosion predominantly occurred in the upper part of the tubing
string, as opposed to H2S ⫹ CO2 corrosion, which is mostly found on the lower part. Pin- and
box-ends suffered significant metal loss due to localized turbulence, which supplies more
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
thin layer of condensed water may become very quickly saturated in FeCO3, even with a moderate
average corrosion rate: The lower the water flow rate, the easier the protectiveness.
3. The specificity of pin-box connection is that water may accumulate in a higher quantity, and is
certainly subjected to turbulent flow. Because of the higher volume/surface ratio, the FeCO3
saturation is more difficult to reach, therefore a lower or even nil protectiveness if the amount of
water is high enough.
4. The particularity of caliper tracks is that they temporarily remove the protective corrosion layer,
allowing a more severe corrosion to initiate. Whether corrosion progresses or stops on these
caliper tracks depends on the winner of the battle between reestablishing a satisfactory protectiveness or initiating enough corrosion to stabilize an anodic/cathodic coupling and/or to maintain
a higher volume of water in the corroded area, which therefire is more difficult to get re-saturated
and well protected.
SPE-169618-MS
6.
8.
9.
10.
corrosive water at these locations than along the tubing length and thereby destabilizes the
protective iron carbonate film.
Well intervention tracks (caliper feelers tracks) did show to initiate sweet corrosion on the sweet
gas wells reviewed in this experience. On the other hand, caliper measurements proved to
accurately monitor metal loss corrosion on un-scaled tubing.
Sweet corrosion did not promote heavy scale deposits because of the higher solubility of iron
carbonate when compared to iron sulfide. CS completions survived for more than 10 years.
The corrosion rate observed on these sweet wells has been much lower than predicted by most
present corrosion models. This difference is certainly due to the combination of a low amount of
condensing water along the production tubing and an annular flow regime, which prevents this
water from significantly accumulating along the tubing length.
Saudi Aramco considers CS completions as the cost-effective option for onshore large-scale
sweet and sour gas developments but CRAs are for offshore developments.
Nonmetallic coatings, such as fluoropolymer, can be useful to minimize sour corrosion scale
buildups, which frequently hamper well interventions, if proved successful in the upcoming
planned field trials. Whether these coatings do or do not provide a long-term resistance and
preserve their adhesion on the steel surface, they are among the key challenges of this solution.
Acknowledgements
The authors wish to thank the management of Saudi Aramco for permission to publish this paper and the
management of Total for the long-term collaborative work with Saudi Aramco on this corrosion topic.
References
1 Bonis, M. “Weight Loss Corrosion with H2S: From Facts to Leading Parameters and Mechanisms,” CORROSION/2009, Paper No. 09564, (NACE International, Houston, Tezas), 2009.
2 Kasnick, M.A. and Engen, R.J. “Iron Sulfide Scaling and Associated Corrosion in Saudi Arabian
Khuff Gas Wells,” paper SPE 17933, Bahrain, March 1989.
3 Choi, H.J., “Field Corrosion Damage Assessment of Carbon Steel Production Tubing at Khuff
Gas Wells,” CORROSION/2006, Paper No. 06603, San Diego, California, March 2006.
4 Bonis, M., Girgis, M., Goerz, K. et al “Weight Loss Corrosion with H2S: Using past operations
for designing future facilities,” Corrosion/2006, Paper No. 06122, (NACE International, Houston, Texas), 2006.
5 Crolet, J.L. and Bonis, M. “Algorithms of the Protectiveness of Corrosion Layers . . .” CORROSION/2010, Paper No. 10365, (NACE International, Houston, Texas), 2010.
6 Mirza, M.S. and Prasad, V. “Scale Removal in Khuff Gas Wells,” paper SPE 53345, Bahrain,
February 1999.
7 Nasr-El-Din, H.A. and Al-Humaidan, A.Y. “Iron Sulfide Scale: Formation, Removal and
Prevention,” paper SPE 68315; Aberdeen, January 2001.
8 Leal, J., Nunez, W., Al-Ismail, S., et al “Novel Mechanical Scale Cleanout Approach to Remove
Iron Sulphide Scale from Tubular,” paper SPE 121404, June 2009.
9 Choi, H.J. and Al-Ajwad, H.A. “Flow Dependency of Sour Corrosion of Carbon Steel Production
Tubing in Khuff Gas Wells,” Paper No. CORROSION/2008 in New Orleans, Louisiana, March
2008.
10 Choi, H.J. and Al-Tammar, J. “Field Corrosion of API 5CT T-95 Carbon Steel Tubing in Saudi
Arabia Gas Wells,” 13th Middle East Corrosion Conference, Paper No.10016, Bahrain, February
2010.
Downloaded from http://onepetro.org/SPEOFCS/proceedings-pdf/14OFCS/All-14OFCS/SPE-169618-MS/1544003/spe-169618-ms.pdf by King Fahd University of Petroleum & Minerals user on 24 March 2021
7.
15

Purchase answer to see full
attachment

  
error: Content is protected !!