Copyright
©2008 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole must be in writing to NACE
International, Copyright Division, 1440 South creek Drive, Houston, Texas 777084. The material presented and the views expressed in this paper are
solely those of the author(s) and are not necessarily endorsed by the Association. Printed in the U.S.A.
CO2 CORROSION INHIBITOR PERFORMANCE IN THE PRESENCE OF
SOLIDS: TEST METHOD DEVELOPMENT
Anette Pedersen,1 Katerina Bilkova,2 Egil Gulbrandsen3, Jon Kvarekvål
Institute for Energy Technology (IFE)
P.O. Box 40
NO-2027, Kjeller, Norway
ABSTRACT
The background and development of three test methods for CO2 corrosion inhibitors
in presence of solids are described: 1) Inhibitor performance testing in the presence of
suspended solids; 2) Inhibitor performance testing on steel covered with sand deposits; and
3) Assessment of sand aggregate formation with oil in the presence of corrosion inhibitors.
Key words: Corrosion, Inhibition, Carbon dioxide, Carbon steel, solid particles
INTRODUCTION
Carbon steel used in combination with corrosion inhibitors is an economically favorable
alternative for multiphase pipelines compared to the use of corrosion resistant materials.
CO2 corrosion inhibitor formulations contain surface-active compounds that form a
protective layer at the pipe wall. The surfactants also adsorb to other surfaces and
interfaces in the produced fluids, like solid particles and emulsion droplets, etc.1-9
The solid particles may comprise sand and clay from the reservoir, and corrosion products
and mineral scale. Fine solid particles, such as kaolinite clay, can consume a significant
amount of corrosion inhibitor by adsorption.2 This consumption may cause the inhibitor
concentration to drop below the minimum effective concentration, and may lead to
corrosion failure if the depletion of the inhibitor is not properly accounted for.3-4 Solids can
affect detrimentally not only inhibitor performance, but many other aspects of the
petroleum production as well. Sand particles are known to cause erosion corrosion. They
can also deposit in some part of the pipelines or in separators, and cause severe attacks
under the deposit.3 Corrosion inhibitor performance testing in presence of solids has
received little attention so far.
1 Present address: Det Norske Veritas, NO-1322, Høvik, Norway.
2 Present address: Intecha, 170 00 Praha, Czech Republic.
3 Presenting author, e-mail: egil.gulbrandsen@ife.no
1
Paper No.
08632
The present paper describes parts of the results obtained in a recent joint industry project
(see Acknowledgement section), where the objective was to develop test methods and test
protocols for laboratory testing of CO2 corrosion inhibitor performance. The paper focuses
on the background and development of three methods for corrosion inhibitor testing in the
presence of solids.
The first method is used to assess inhibitor performance in presence of suspended solids.
The second method is used to test inhibitor performance under sand deposits. The third
method is used to assess aggregation of sand and oil caused by presence of surface-active
corrosion inhibitors.
The intention of the work was not to compare inhibitors. The inhibitors were used as
example products in the development of the test method and identification of critical
parameters for the testing. The presented test methods may also contribute to the
development of improved inhibitor products.
TESTING OF INHIBITOR PERFORMANCE WITH SUSPENDED SOLIDS
This method was developed for screening tests of inhibitors and determination of the
effective inhibitor concentration in the presence of suspended solids. The testing method is
intended for field cases when solids are continuously produced in amounts that may reduce
inhibitor performance. Kaolinite clay was used as a model solid with well-defined surface
area.
Clay is added prior to inhibitor addition in the test method. This resembles the case where
the solids are already present at the point of inhibitor injection. The effect of oil was also
studied.
Experimental Procedure
All the tests were carried out in glass cells with a volume of 3 liters. Electrodes for
electrochemical corrosion monitoring and pH measurements, temperature probes and gas
inlets/outlets were inserted through stainless steel lids. The glass cell with accessories is
schematically represented in Figure 1. A heating plate was used to control the temperature
of the test solution, and the solution was gently stirred with a magnetic stirrer.
The test specimens were machined from X65 carbon steel. The element analysis and
microstructure of the steel are given in Table 1. Cylinder specimens with surface area of
3.14 cm2 were ground to 1000 mesh with wetted SiC paper prior to use. The specimens
were wetted with isopropanol, cleaned with technical acetone in ultrasonic bath, and rinsed
with ethanol. Then the specimens were blow dried prior to mounting on the specimen
holders.
The test solutions were prepared from technical grade NaCl and distilled water. The pH
was adjusted by addition of AR grade NaHCO3. The test solutions were continuously
purged with CO2 grade 4.0. This gas contains less than 10 ppm of O2, corresponding to less
than 0.5 ppb dissolved in the solution in equilibrium with the gas at
1 bar and 25 ºC.
2
The kaolinite clay was a reference clay, denoted KGa-2, from The Clay Mineral Society,
University of Missouri, USA.9 The clay properties are given in Table 2. The surface area
constituted by the clay is expressed in terms of surface area per unit volume of test brine
(m2/L). The standard kaolinite clay (KGa-2) has a specific surface area of 23.5 m2/g. A
clay concentration of 43 ppm therefore corresponds to an area of 1 m2/L. A scanning
electron image of the clay platelets, which were 0.1 µm–0.5 µm thick and 0.5-2 µm long,
is shown in Figure 2.
The inhibitors used for the testing are listed in Table 3. The inhibitor concentrations are
reported as ppm (i.e. mg/L) of blended product based on total liquid volume (water + oil).
The properties of the oils used in the tests are listed in Table 4; Oil 1 is a black oil, Oil 2 is
a condensate, while Oil 3 is a refined, low-aromatic oil product. The oils were tested with
Inhibitor M. Table 5 gives the partitioning of the inhibitor between the phases.
A potentiostat with a multiplexer was used for electrochemical measurements. The
corrosion rate was monitored by the polarization resistance method (LPR) throughout the
test. LPR measurements were performed in three-electrode configuration, i.e. working
electrode - carbon steel specimen, reference electrode - Ag/AgCl, auxiliary electrode - Ti
ring. The potential ramp for the LPR measurement was –5 mV to +5 mV vs. Ecor, with scan
rate 0.1 mV/s. The corrosion currents was calculated as B/Rp . A constant value of B =
20 mV was used, based on polarization curves and mass loss determinations, as described
in Refs. 13-14, and Appendix A. The corrosion rates are reported as average penetration
rates (1 A/m2 corresponds to 1.16 mm/y). The polarization resistance was compensated for
IR drop, determined by means of impedance spectroscopy (EIS). Potentiodynamic
polarization curves were measured at the end of the test. The cathodic polarization curve
was run from 0 to –250 mV vs. Ecor, and the anodic one from 0 to +150 mV vs. Ecor, both
at scan rate 0.1 mV/s. All the specimens were inspected for localized attack by an optical
microscope after the test.
The tests were carried out at 60 °C, 0.8 bar CO2 and pH 4.5. Most of the tests were
performed in 10 %(w/w) NaCl brine. The solution in the glass cell was deoxygenated for
2-3 hours before the immersion of test specimens. Three specimens were introduced to
each cell; two specimen for electrochemical measurements and one specimen for weigh
loss measurements. The steel specimens were precorroded for 24 hours prior to addition of
clay and inhibitor. Kaolinite clay was added to the cell 10 min before the inhibitor
addition.
In the tests performed with 10 % oil, the oil was de-aerated with 1 bar CO2 for minimum 4
hours and heated to test temperature before it was transferred to the test cell. 300 ml oil
was transferred to the 3-litre test cell after start of precorrosion, and minimum
2 hours before the clay and inhibitor was introduced. A peristaltic pump was used to
transfer the oil. The inhibitor was added to the cell from a 10 % stock solution, diluted with
either oil or water.
3
Results and discussion
Inhibitor testing without oil. Examples of the experimental data from one test without
kaolinite clay and one test with 2.5m2/L surface area of clay (108 ppm) are given in Figure
3. The baseline corrosion rate was approximately 3 mm/y. Kaolinite clay and 30 ppm of
Inhibitor A was added after 24 hours of precorrosion. The inhibitor concentration was
increased to 50 ppm towards the end of the test. The addition of the inhibitor resulted in a
rapid decrease of the corrosion rate. The corrosion rate in the test with no clay continued to
decrease until it stabilized at about 0.06 mm/y. The corrosion rate with clay stabilized at a
higher corrosion rate (0.15 mm/y). This shows that the clay reduced the inhibitor
performance.
Figure 4 summarizes the steady corrosion rates at the end of the tests, and residual
concentrations of the active inhibitor compounds vs. the surface area of clay. The
concentrations of the active compound given as the corresponding concentration of the
formulated product. The residual analysis was carried out by the inhibitor suppliers. The
corrosion rates are the steady corrosion rates for 30 ppm dose rate of the respective
inhibitor.
The data for Inhibitor A illustrates a case when the inhibition effect was remarkably
reduced in the presence of clay (Fig. 4a). The inhibited corrosion rate in absence of clay
was 0.1 mm/y. The residual inhibitor concentration dropped to 11 ppm with 2.5 m2/L clay
surface area. Nevertheless the inhibited corrosion rate increased only slightly to 0.2 mm/y.
With 5 m2/L in the solution, the residual inhibitor concentration decreased to 5 ppm. This
resulted in an increase of the corrosion rate to 0.4 mm/y. Only 2 ppm of Inhibitor A was
left in the solution in the test with 7.5 m2/L clay surface area (323 ppm clay). This
depletion was associated with a dramatic rise of the corrosion rate to 2 mm/y, which was
nearly the same as the uninhibited corrosion baseline. Similar results were obtained with
Inhibitor B (not shown).
Figure 4b shows the results obtained with Inhibitor C. Even if the corrosion rate increased
slightly with increasing clay concentration, the corrosion rate remained below 0.1 mm/y in
presence of even in presence of 7.5 m2/L clay surface area. The residual inhibitor
concentration for this inhibitor was determined both for filtered and unfiltered solution
sample. Despite some discrepancy between the results obtained with two analytical
protocols, both of the data sets indicate that the inhibitor concentration was reduced to less
that 20 ppm for 5 m2/L of clay surface area. This shows that inhibitor performance in
presence of suspended solids depends strongly on formulation.
The results of the tests with increasing dose of Inhibitor B are given in Figure 5. Two tests,
one without clay and one with 5 m2/L clay surface area, are shown in the figure. The
corrosion rate decreased when 10 ppm of Inhibitor B was added for the test without clay.
The corrosion rate continued to decrease with increasing amount of inhibitor until it
stabilized when 80 ppm inhibitor was added. The inhibitor performance was negligible for
10 ppm of the inhibitor with 5 m2/L of clay particle surface. With higher inhibitor
concentrations the corrosion rate gradually decreased. Eventually at 150 ppm, the
corrosion rate dropped to nearly the same value as for the test without clay. This indicates
that it is possible to saturate the clay surface with inhibitor by adding sufficient amount of
it, and in that way obtain adequate inhibitor performance.
4
Adsorption studies of surfactants on clay5 also showed that there was a limit concentration
above which the adsorbed amount of the inhibitor did not increase, thus all the clay surface
was saturated with the inhibitor. It was reported that zeta potential of kaolinite clay
increased strongly with addition of CO2 corrosion inhibitor, which contained cationic
surfactant. Such behavior is consistent with adsorption of cations.2,4 Presence of other
cations like Ca2+ can further reduce performance of cationic inhibitors.2 On the other hand
anionic inhibitors showed low affinity to clay.2 The problem of parasitic consumption of
inhibitor by clay and other fine particles in the produced fluids is a problem of determining
and achieving the optimal dose rate. However, the chemistry of the inhibitor plays a key
role in the problem. It appears possible to select an inhibitor product with acceptable
performance in the presence of solids.
Inhibitor testing with oil present. A summary of the inhibitor test results with oil is given in
Figure 6. All the experiments were carried out with 30 ppm of Inhibitor M. The corrosion
rate generally increased with increasing clay concentration. In presence of clay, the
corrosion rate for the tests with Oil 2 was lower than in the tests with the other oils, or
those without oil. On the other hand, the corrosion rate with the other oils was higher than
the corrosion rate without oil, for most the tests with clay present. The corrosion rate for
Oil 1, and Oil 3 with 5 m2/L clay and more was nearly the same as the uninhibited
baseline. These results indicate that oil can include components that may enhance or
reduce the inhibitor performance, both in absence and presence of clay. The observed
effects may rather be related to residual inhibitor concentration, than a direct interaction of
oil and clay on the steel surface.
The specimen for one test with 10 % of Oil 3 got oil wetted, and the corrosion rate dropped
to extremely low values. The oil-wetting phenomenon is not well understood yet; i.e. it is
not possible to predict when oil wetting occurs.
Conclusions
• The inhibitor performance of some inhibitors was reduced in presence of clay
due to adsorption. The test protocol allows distinguishing among different
inhibitor product in terms of their sensitivity to clay.
• The sensitivity of the inhibitor performance to clay depends on the inhibitor
chemistry. It appears feasible to formulate inhibitors so that the performance is
little affected the present amounts of clay.
• For the inhibitors that were affected by the presence of clay, increased inhibitor
dosage would efficiently saturate the clay surface and adequate inhibitor
performance could be reached.
• Laboratory studies to select inhibitors should include experiments in the
presence of representative solids at the expected concentration.
TESTS OF INHIBITOR PERFORMANCE UNDER SAND DEPOSITS
This test method was developed to evaluate CO2 corrosion performance under sand
deposits. The objective was to assess the risk of galvanic corrosion, formed due to sand
5
deposition on part of the steel surface, and to evaluate the ability of inhibitors to mitigate
this type of corrosion.
The test method is intended for application with continuous sand production or with high
probability of sand production. Sand deposition prior to inhibitor addition resembles the
case of sand deposition under conditions of insufficient inhibition. Inhibitor addition prior
to sand deposition represents the case of sand deposition under conditions of adequate
inhibition (at least at bare surfaces). The effect of oil has not been included in the test
method yet.
Experimental Procedure
The chemicals and specimen material were the same as described in the previous part of
the paper, see Tables 1, 3 and 4. The tests were performed in 10 %(w/w) NaCl aqueous
solution. Some tests were carried out at 60 ºC, which correspond to CO2 partial pressure of
0.8 bar. Other tests were performed at 90 ºC with CO2 partial pressure 0.5 bar.
The sand used in these tests was analytical grade, acid-washed and calcinated silica sand.
The grain size was specified as 200-400 mesh (30-80 micron). However, in-house SEM
examination showed particle diameters of 200–600 µm (Figure 7). Estimating by use of a
sphere model, the geometric area was less than 0.01 m2/g. Surface area measurements by
the BET technique (nitrogen adsorption) indicated ca. 0.1 m2/g, which indicated substantial
porosity and roughness of the sand grains. This is to some extent supported by the SEM
image. The sand layer thickness was about 5 mm in all tests. The total amount of sand in
the cell was approximately 3 g. The sand thus had a total surface area of about 0.3 m2.
The test set-up is shown in Figure 8. A specimen assembly consisting of three specimens
used in most of the tests; one specimen not covered by sand (abbreviated NS - no sand),
and two specimens fully covered by sand (abbreviated as FS1 and FS2 - fully sand
covered). Specimens FS1 and NS were galvanically coupled, while the FS2 specimen was
not galvanically coupled to other specimens. The area of the NS specimen was ca. 4 cm2,
while the areas of each of the FS specimens were 1 cm2. The specimens were molded into
epoxy, and pretreated as described in the previous part of this paper. The sand was
deposited with negligible air ingress by means of a specially designed glass tube device.
The sand was contained between two moveable pistons. In the tests with inhibitor added
before the sand, the sand was saturated with inhibitor by flushing with portions of the
inhibitor containing test solution.
The FS1 and NS specimens were galvanically coupled by means of the ZRA (Zero
Resistance Ammeter) facility of the potentiostat. The galvanic current was logged at
regular intervals. The corrosion current was measured with the polarization resistance
technique in the galvanically coupled mode, with individual ZRA current measurement on
each specimen of the couple. The potential was scanned from -5 mV to +5 mV vs. Eoc of
the couple, at a scan rate of 0.1 mV/s. The anodic dissolution current calculation is
described in Appendix A.
The tests were performed in two different ways. The sand was deposited 1-3 days prior to
inhibitor addition in the test resembling the case of sand deposition under conditions of
insufficient inhibition. In the test resembling the case of sand deposition under conditions
6
of adequate inhibition, the inhibitor was introduced 1-2 days before sand was deposited on
the steel surface. The sand surface was then saturated with inhibitor by exposure to
portions of the inhibited brine before deposition on the specimen surface.
Results
Sand deposited before inhibitor addition. The results for a test where sand was added
before Inhibitor C are shown in Figure 9. Only NS and FS1 were exposed in this initial
test. The baseline corrosion rate without sand was about 2 mm/y. Sand was added 6 hours
after specimen immersion. After one day exposure 30 ppm inhibitor C was added. The
corrosion rate of the NS specimen (not covered by sand) decreased rapidly due to the
inhibition effect of Inhibitor C. The steady corrosion rate for the NS specimen was
0.03 mm/y. The corrosion rate of the FS specimen decreased gradually to about 0.5 mm/y.
Increasing the inhibitor concentration to 50 ppm did not have significant effect on the
corrosion rates.
Sand deposition c