SPE 134473

CO2-Induced Changes in Oilwell Cements Under Downhole Conditions: First Experimental Results André GARNIER1; Jean-Benoît LAUDET1, Nadine NEUVILLE2, Yvi LE GUEN2, Dominique FOURMAINTRAUX3, Noureddine RAFAI4, Nicolas BURLION5, Jian-Fu SHAO5

1: TOTAL - Centre Scientifique et Technique Jean Feger, avenue Larribau, 64000 Pau (France) 2: OXAND SA - 49 avenue Franklin Roosevelt, 77210 Avon (France) 3: DFi Ingénierie - 11 bis rue du Glazen Kervarlay, 56680 Plouhinec (France) 4: LERM - 23 rue de la Madeleine, BP 60 136 - 13631 Arles Cedex (France) 5: LML - UMR CNRS 8107, Polytech’Lille, Université Lille1, Cité Scientifique, 59650 Villeneuve d’Ascq, France (France) Copyright 2010, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Florence, Italy, 19–22 September 2010. 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 One of the major technological issues for CO2 injection (for EOR, CCS, etc.) is the long-term behavior of cement-based materials used to ensure the overall sealing performance of the storage wells. When water is present, the CO2 after injection can react chemically with the cement (i.e. carbonation). How do the CO2-enriched formation fluids changes the cement’s chemistry and properties? Could the sealing efficiency of the wells be affected by these changes? The objectives of our experimental program are to assess the kinetics and phenomenology of the changes that occur in different class-G Portland cements exposed to CO2-enriched aqueous fluids at 8 MPa and two different temperatures. The experimental program presented in this paper consists of: a first carbonation test (Test # 1) using neat G cement, at a temperature of 90°C (194° F) and a pressure (supercritical CO2 above water) of 8 MPa (1160 psi); a second carbonation test (Test # 2) using G cement with silica flour (to prevent strength retrogression), at a temperature of 140°C (284°F) and at the same CO2 pressure of 8Mpa. Finally, coupled chemo-mechanical tests (dynamic tests) are underway on similar class-G cement and similar CO2-rich water. All the samples were prepared according to ISO/API specifications. The experimental set-up simulates downhole “static” conditions: the samples were immersed in water in a cell thermally regulated and pressurized by CO2. Cement samples were exposed to CO2-saturated water for various lengths of time (from one week to 3 months) and were characterized using advanced methods for chemical and mineralogical analysis (X-ray tomography, SEM, XRD, TGA-TDA…) and mechanical testing. The main preliminary results show a reactive front (characterized by carbonation) progressing from the fluid-sample interface towards the sample centre. The carbonation front moves faster during Test 2 (at higher temperature) than during Test 1 (at lower temperature). SEM images of Test 2 also show a thin layer of dissolved carbonate at the sample’s surface. The carbonated cement areas exhibit increased density and greater compressive strength. The results of coupled chemo-mechanical tests with injection of CO2-enriched water in samples under deviatoric stress show that the CO2 flow rate in the cement rapidly decreases, finally resulting in carbonation clogging of the cement sample. These results seem consistent with reported field observations.

2 SPE 134473

Introduction

When CO2 is injected into a well, for various purposes (EOR: Enhanced Oil Recovery, CCS: Carbon Capture & Sequestration, etc.) [Noik et al., 1998] the well’s integrity becomes a major concern due to the acidic environment CO2 creates when in contact with water. Well cementing is devoted to providing long-term zonal isolation and offering protection to the casing supporting the wellbore. Long-term durability of the cement-based materials is therefore of paramount importance in this operation to guarantee the tightness of these wells. The cement sheath has to ensure containment of the injected CO2 in a specific geological layer (e.g. depleted reservoir) without any leakage of CO2 outside that selected layer. Cement can be damaged by aggressive environments: (i) failure under mechanical stresses that can induce cracks (ii) chemical degradation in acidic environments that can alter cement properties. The available papers and discussions in conferences show that many experimental set-ups have been developed to increase understanding of the behavior of oilwell cements under severe conditions, with respect to chemical attack by CO2. The main results return contradictory feedback about the CO2 reaction with cement [Mulders, 2007]. So Total decided to launch a research program directed at characterizing the long-term behavior of Portland cement in a CO2 environment. Portland cement (class G or H) was selected as it has been the cement most commonly used by the Oil & Gas industry in well construction for decades. The ongoing research program is focused on two objectives: (i) to assess the risk of injecting CO2 in fields connected to old wells cemented with Portland cement, and (ii) to evaluate the relevance for the industry today of using Portland cement as the base component of most composite cements when drilling new wells in fields where CO2 is already present.

Strategy The experimental program is devoted to oilwell cement behavior at high temperature and pressure conditions with the aim of improving knowledge concerning its long-term behavior in the presence of acid gases, and in terms of phenomenology, reaction kinetics and evolution of the cement’s properties. The strategy implemented includes different sets of experimental tests (Table 1) and numerical modeling. The experimental tests and lab characterization methods have to be suitable to qualify the long-term zonal isolation of cement sheaths in downhole conditions in the presence of CO2, including time effects (long-term and accelerated tests). Tests with samples cured for a short period (≈ 1 week) refer to wells drilled in a formation already containing CO2, other tests with samples cured for several months represent cement sheaths exposed to CO2 after a long period of time. This paper focuses on static and dynamic tests, which are the most advanced ones.

What kind of test? Why? Which characterization? - Neat class-G cement curing (> 1 month) - Class G + silica flour: short- (≈ 1 week) and long-time curing (> 1 month)

Ensure reproducibility, quality control (homogeneity) Assess the impact of the curing time on initial cement properties

• Slurry: rheology, stability • Hardened material: X-Ray tomography

Static carbonation tests Assess phenomenology and the kinetics of damaging processes

• Damaging front progression; X-ray tomography

• Chemical evolution (sample and leachants) • Mechanical properties (micro-hardness,

triaxial test, UCS) • Transport properties (Porosity, Permeability)

Dynamic tests (coupled chemo-mechanical test, CO2-rich aqueous fluids flux in cement samples under triaxial stresses)

Assess mechanical and transport properties

• Mechanical properties: evolution of deformation with time (creep test: time-dependent behavior )

• Transport properties : flowrate, permeability) Characterization tests on completely carbonated cement samples [Ollivier, 1997] [Fernández Bertos et al. 2004], and completely lixiviated samples

Assess the mechanical and transport properties of completely carbonated cement samples and completely lixiviated samples

• Chemistry: chemical analyses, mineralogy, SEM

• Mechanics: triaxial tests, time dependent – creep – tests

Table 1: Experimental strategy of the research program – Ongoing lab experiments

Material and Method used for Static Tests Material: sample preparation and control

Portland cement was prepared with the usual additives (we used Schlumberger: D153 “antisettling”, D604AM “dispersant”, D47 “defoamer”) according to ISO/API specification and hardened in cylindrical molds (diameter = 20 mm, height = 240 mm) in

SPE 134473 3

curing chambers using lime-saturated water. The first material used is neat class-G cement; the second is class-G cement with addition of silica flour (35% bwoc). Adding silica flour produces secondary hydrates by pozzolanic reaction with the lime resulting from primary hydration (SiO2+Ca(OH)2→C–S–H), thereby avoiding strength retrogression during the hardening period due to α‐C2SH formation at 140°C [Hewlett, 1991] [Taylor, 1997] [Meller et al., 2005]. A carefully followed QA/QC program requires that, before being molded, samples should display neither free water nor sedimentation nor visible heterogeneity by X-ray tomography (e.g. air bubbles) so that all samples are homogeneous and similar. The cement slurries were cured for more than 28 days under different conditions: the neat class-G cement at 90°C and at atmospheric pressure (for carbonation test at 90°C); the class G + silica flour cement at 140°C and 20.7 MPa (3000 psi) for the first 10 days, then at 90°C and at atmospheric pressure (for the carbonation test at 140°C). The curing conditions used for the class-G + silica flour cement represent a cement sheath rapidly exposed to CO2 after well drilling. Other tests are under way with cement cured for several months to represent a cement sheath exposed to CO2 after a long period of time. After the curing periods, the cement samples were controlled by X-ray tomography to avoid any visible heterogeneities within the bulk of the cement (as shown in Figure 1 for 2 samples of class G + silica flour cement). X-ray tomography images are used as reference to assess the impact of test conditions during the tests. Neat class-G cement samples have a porosity of about 30% and contain mainly amorphous phases (associated with calcium silicate hydrates, or C-S-H, ≈ 67% wt.), portlandite (≈ 17% wt.) and katoite (≈ 8% wt.). Samples of class-G cement with silica flour have a porosity of 28% and contain mainly C-S-H (≈ 67% wt.), quartz (≈ 10% wt.) and katoite (≈ 7% wt.). In order to obtain a 1-D progression of the reaction front during the carbonation tests, the cylindrical surface of the samples was protected by PEEK (polyetheretherketone). Method: Experimental Setup Batch experiments were carried out in steel reactors under static conditions. Two sets of experiments (test#1, test#2) were performed with supercritical CO2 pressurized at 8 MPa (1160 psi). One type of cement was used for each temperature:

• Test #1: neat class-G cement at 90°C (194° F); • Test #2: class-G cement with 35% bwoc silica flour at 140°C (284°F).

The experimental setup is the same for static tests #1 and #2. Cement samples were cut after the curing period (diameter = 20 mm, height = 40mm) and positioned 4 by 4 on a 3-tier sample-holder (12 samples in each cell, Figure 2). The sample-holder is placed in a glass beaker filled with distilled water at 90°C (about 350 ml) and the glass beaker is placed in the cell preheated at 90°C. Throughout the entire experiment, the samples are fully immersed in and covered by the water (Figure 2) which is in contact with the CO2 at 8 MPa. They are exposed for only a few seconds to air when removed from the cells and placed in storage flask for characterization. The cell is closed and CO2 is injected into it, regulated at the selected temperatures: 90°C or 140°C. In order to avoid thermal/mechanical stress within the samples while pressure and temperature increase, CO2 pressure is increased gradually and regularly from 0.1 MPa to 8 MPa at a rate of 0.2 MPa / minute (about 15 psi/min). Temperature is raised from 90°C to 140°C for test #2 at a rate of 2°C / minute (no temperature increase for test #1). The injected CO2 dissolves in the water until equilibrium is reached. The CO2 concentration in aqueous solution can be estimated from literature: CO2 solubility in pure water is about 0.7 mol.kg-1 H2O at 90°C (Test #1) and about 0.6 mol.kg-1 H2O at 140°C (Test #2) (Figure 3, Span et al. 1997). The CO2-saturated aqueous solution thus formed reacts with the cement samples. Due to the dissolution of CO2 in the water and the chemical reactions with the samples, the CO2 pressure has to be regularly adjusted during both tests to maintain it at 8 MPa. Throughout the tests, CO2 pressure (see Figure 4) and temperature are recorded. No variation in temperature was observed during exposure of the samples to CO2. At different points in time during the tests, two cement samples are removed from the cell: for test #1 (neat G cement at 90°C) after 7, 36, 65 and 90 days and for test #2 (G cement with 35% silica flour at 140°C) after 4, 12, 21, 31, 55 and 88 days. To prevent any mechanical damage to the samples due to depressurization, pressure and temperature are slowly decreased before opening the

4 SPE 134473

cells. Then, the two samples are weighed (for the mass measurements, samples are swabbed beforehand in a paper towel in order to remove the excess water), and characterized by X-Ray tomography and complete chemical analysis. At the end of each carbonation period in both tests, the fluid in the glass beaker is collected for chemical analyses, and replaced by fresh distilled water at 90°C to ensure that the aggressiveness of the solutions is kept high throughout the duration of the tests.

Results Test #1 – CO2 test at 90°C and 8 MPa with neat class G cement This test is still running. Only the mass evolution and X-Ray tomography results are available for the period up to 65 days of testing. The chemical and mineralogical analyses are ongoing.

X-ray tomography and mass variation increase

Figure 5 shows X-ray density in planar sections crossing the centre of each sample at different times. After 7, 36 and 65 days, only two zones with different X-ray density are visible in the samples: the initial (“un-reacted”) cement and the modified (“reacted”) cement. The modified material has a greater X-ray density than the initial cement. The increase in sample mass with time is consistent with the increase in X-ray density, and maximum mass increase is 6% after 65 days of exposure. The lateral protection (PEEK) is partly effective at 90°C, even though zones of reacted cement are visible at the side of the samples.

Kinetics of the reaction front’s progress

Figure 6 shows the position of the carbonation front in the samples at different times during the test (measured from X-ray tomography images, on the basis of visual contrasts in gray levels). The reaction front progresses linearly with the square root of time for the first 65 days of test #1. The average thickness of the reacted zone is about 4 mm after 65 days of testing. This corresponds to an average kinetics of the reaction front of about 9.3 mm.yr-1/2.

Chemical analyses of the leachant Calcium and silicon released in the CO2-rich solution by each cement sample are respectively 28 and 29 mg after 65 days (Figure 7). Calcium is usually the most soluble ion during cement degradation processes; considering an initial CaO content in the material of about 50%, the cement samples lose about 0.5% of calcium after 65 days of testing, which is low compared to other degradation processes in natural fluids [Neuville, 2008], or pure water ([Kamali et al. (2008) showed that the leached calcium after the same experimentation time is by two orders of magnitude higher in the presence of pure water). This result indicates that the main cement degradation process involved in this test is not calcium leaching, but internal calcium carbonate precipitation, namely carbonation. Test #2 – CO2 test at 140°C and 8 MPa with class-G cement + 35% bwoc. of silica flour

X-ray tomography and SEM observations

X-ray tomography images (Figure 5) show progression of a carbonation front from the outer surfaces to the centre of the samples over the course of the test. The outer ring (i.e. the reacted zone) is characterized by an increase in x-ray density. Only one front is observed, separating two zones of cement with different X-ray densities. At the end of the experiment, after 88 days, the entire sample has reacted. No fracture is visible in the samples during the test.

Alteration is observed on the samples cut from the lateral surfaces, whereas unidirectional attack was expected; this was due to the PEEK ribbon’s ineffective protection. Additional analyses of the X-ray tomography data highlight the initial observations. Figure 8 shows cumulated frequencies of X-ray densities respectively in the initial and reacted cement samples after different exposure times. The proportion of higher X-ray density values (from 1650 to 1850) increases with time of exposure, reflecting the densification of the samples. Mass measurements on carbonated samples are consistent with these results: the maximum mass increase is about 14% after 88 days of exposure.

SPE 134473 5

SEM observations by BSE (Back Scattered Electrons) images (Figure 9) show the presence of three different zones in the samples: an inner zone of initial cement, then a carbonated zone, and finally a thin zone characterized by higher-porosity material close to the sample surface. These layers are too thin to be seen by X-ray tomography.

Kinetics of the carbonation front’s progress

In the same way as for Test #1, X-ray density images were used to locate the position of the carbonation front in the samples at different points in time during the test (Figure 10). The thickness of the reacted cement increases linearly with time: the kinetics of the carbonation process is about 0.2 mm.d-1 (i.e. 73 mm.year-1) for the 88 days of test #2.

Chemical analyses of the leachant and mineralogical analyses of the carbonated cement samples Mineralogical analyses of the carbonated cement samples

In order to assess the evolution of carbonate content, thermogravimetric analyses (TGA) were performed on an initial sample (sample before test #2) and on the carbonated samples removed from the cell after different lengths of time. Each analysis was carried out on powders (about 100 mg) sampled every 2 mm from the surface to the core of each sample. Analysis conditions were as follows: temperature increase from 20 to 1000°C in increments of 10°C/min. Nitrogen was flowed into the chamber (flow rate =60 ml/ min) to avoid additional cement carbonation during the analyses. The results indicate that the initial sample contains no calcium carbonate (CaCO3), while the most carbonated samples (i.e. obtained at the end of the test after 88 days) contain from 30 wt% (in the centre of the samples) to 55 wt% (in the outer part of the samples) of calcium carbonate (Figure 11), mainly aragonite and calcite. The presence of neo-formed calcium carbonate results from the reaction between CO2 and cement hydrates. This explains the increased mass of the samples and the greater X-ray density values observed in their outer part.

Chemical analyses of the leachant

The cumulated masses of calcium and silicon released into the solution by each cement sample after 88 days of testing (Figure 12) are respectively 460 and 167 mg. Considering that their initial CaO content is around 39%, the cement samples lose about 8.8% of calcium after 88 days of testing. This result shows that the process of cement carbonation is brought about mainly by internal carbonate precipitation as observed during test #1. The more porous outer layer, visible only in the SEM images of Figure 9, may explain the greater calcium quantities in the leachant compared to the test at 90°C.

Chemo-Mechanical Tests (dynamic test) Cement that is set in a well can be exposed to a number of mechanical and chemical stresses. To reproduce these conditions, specific dynamic tests were performed to simultaneously apply chemical and mechanical stresses on a sample. Their objective is to assess the impact of these stresses on the transport and mechanical properties of class-G Portland cement at selected downhole conditions and in the presence of CO2. Some tests are still running. The dynamic tests were performed on similar class-G cement samples at 90°C. For all the experiments, neutral water (lime-saturated water) was injected initially into the samples, followed by different fluids: CO2-saturated water or supercritical CO2. The dynamic test (Figure 13) with CO2-saturated water was conducted for 12 days with a hydrostatic stress phase followed by a deviatoric stress phase. Initial conditions were: injection pressure pi = 2.5 MPa; confining stress σ3= 3 MPa; axial stress σ1= 3.5 MPa during hydrostatic phase and 6 MPa during deviatoric phase (deviatoric stress σ1 - σ3 = 3 MPa). Figure 14 shows the evolution of the parameters during the dynamic test, the sample’s axial deformation (ε1) and the total volume of CO2-rich water injected. It appears from the main results that adding CO2 to the water injected throughout the mechanically stressed cement has no significant impact on deformation of the cement (no change in the strain rate is observed). Simply, compaction was observed, as expected due to the increase in deviatoric stress. No change in the cement’s mechanical properties was measured during CO2-rich water injection. The permeability (K) measured during the hydrostatic phase (lasting 4 days) is 1.9x10-3 mDarcy, and decreases during the deviatoric phase to 1.6 x10-3 mDarcy (Figure 14). After 10 days of the experiment, at the beginning of CO2-rich water

6 SPE 134473

injection, the flow rate decreases and permeability can no longer be measured. The decrease in permeability, and ultimate clogging of the sample, is explained by the reaction of a thin layer of cement where the CO2-rich water is injected, whereas most of the sample is not impacted by the injected fluid. Mineralogical and microstructural analyses show that, for the sample area directly exposed to injection of the aggressive fluid, the thickness of the carbonated material is very low: 300μm, measured with a phenolphthalein test (Figure 15) (during the static test at 90°C, the carbonated cement thickness is about 1.2 mm after 12 days, Figure 6). Discussion on main results For both static tests, CO2(g) dissolves in water according to: CO2(g)+2OH-(aq) → CO3

2-(aq)+H2O. Contact of CO2-rich water with cement modifies the equilibrium between cement hydrates and the pore solution [Adenot, 1992]. This induces dissolution of the hydrate phases followed by internal precipitation of calcium carbonates. This process is known as carbonation. For the cement used in test #1 (Portland cement), the carbonation mainly involves reactions with portlandite, followed by the precipitation of calcium carbonate according to the reaction 1 below:

Ca(OH)2(s)+ CO32-(aq) → CaCO3(s)+H2O(aq). Equation 1

During test #1 the kinetics of carbonation was observed to be proportional to the square root of time, which implies that cement carbonation in this case is governed by diffusion processes [Adenot, 1992] [Duguid, 2006]. For the Portland cement with the addition of 35% bwoc silica flour (test #2), the reacted zone resulted from the reactions between CO2(aq) and Calcium Silicate Hydrates, also named C-S-H (portlandite (Ca(OH)2) is absent in this material), according to reaction 2 [Thiery, 2005]:

Cx-Sy-Hz + xH2CO3 → xCaCO3 + ySiO2.tH2O + (x-t+z)H2O Equation 2

This is consistent with the TGA which showed the presence of calcium carbonate (CaCO3) in the outer part of the cement samples. X-Ray tomography results showed that the kinetics of carbonation is proportional with time throughout the duration of the experiment; it can therefore be limited by chemical reaction kinetics. The carbonated front moves faster (and deeper) at 140°C than at 90°C. As the connected porosity exhibits a similar value for both materials (about 30%), and the CO2 solubility is similar in pure water at 90°C and 140°C (Figure 3), the difference may stem from the mineralogical nature of the hydrates that react with the dissolved CO2, their proportion and microstructural assemblage [Rafai, 1992] (e.g. no portlandite in the cement used for test #2 with silica flour and 140°C). Shen et al., 1989 showed that for neat class-G cement, Ca(OH)2 is carbonated into CaCO3, leading to an increase of the solid matrix volume (because the molar volume of Ca(OH)2 is lower than CaCO3). Consequently, the CaCO3 formed from Ca(OH)2 can appreciably reduce the cement’s porosity and permeability by volume expansion (and ultimately limits the carbonation process). Conversely, the reaction of CxSyHz (or C-S-H) with CO2 results in a highly polymerized silica gel associated with a reduction of solid matrix volume despite precipitation of calcium carbonate (that has a lower molar volume than C-S-H). The conclusions of Shen et al., 1989 may explain the differences observed between the two carbonation tests presented in this paper: precipitation of calcium carbonate in the neat class-G Portland cement (test #1) induces a drastic decrease in the transport processes of the chemical species, while with class-G Portland cement with silica flour (test #2), the precipitation of calcium carbonate does not affect the transport processes, allowing continuous carbonation of the samples. The results of the coupled chemo-mechanical (“dynamic”) test are consistent with observations made by Carey et al., 2007 who showed that in situ cement sheath samples were not altered by CO2 attack after 30 years of CO2 injection. The steep decrease of cement permeability still provides protection against significant movement of CO2 through the cement matrix.

Conclusions and Prospects

The static carbonation tests on neat class-G cement at 90°C and class G with silica flour at 140°C resulted in partial or total carbonation of the samples (for the experimental conditions described in this paper). Class-G oilwell cement is thus reactive when exposed to carbonic acid under downhole conditions. However, the behavior is different for each material: the carbonation front progression is controlled by diffusion for neat class-G cement (the front evolves linearly with the square root of time), while it is limited by reaction kinetics for class G + silica flour cement (the carbonation front evolves linearly with time).

SPE 134473 7

In addition to the static tests assessing the phenomenology and the kinetics of carbonation, dynamic tests were performed to assess the evolution of the mechanical and transport properties of oilwell cements in selected downhole conditions in the presence of CO2. The results of the first tests (still running) show that the cement sheath is not mechanically damaged and is still able to fulfill its main function of ensuring the confinement of CO2 in the target reservoir (by improving its transport properties). From the preliminary results, the cement matrix reaction in CO2 environments does not appear to be the driving phenomenon for loss of well integrity; mechanically induced damage to the cement sheaths (hydro frac, micro-annulus, etc.) may give greater cause for concern. References

Adenot, F. Concrete durability: characterization and modelling of the chemical and physical processes of cement degradation. Thèse de Doctorat de l‘Université d’Orléans, 1992 (in French). Benge, G. Improving wellbore seal integrity in CO2 wells. SPE/IADC 119267, 2009. Carey, J.W., Wigand, M., Chipera, S.J., Wolde Gabriel, G., Pawar, R., Lichtner, P.C., Wehner, S.C., Raines, M.A., Guthrie Jr., G.D. Analysis and performance of oil well cement with 30 years of CO2 exposure from the SACROC Unit, West Texas, USA. international journal of greenhouse gas control, 2007.

Duguid, A., The effect of carbonic acid on well cements, PhD Thesis of the faculty of Princeton University 2006. Fernández Bertos, M., Simons, S.J.R., Hills, C.D., Carey. P.J. A review of accelerated carbonation technology in the treatment of cement-based materials and sequestration of CO2, Journal of Hazardous Materials, 2004. Hewlett, P. Lea’s chemistry of cement and concrete. Fourth Edition, edited by Peter C. Hewlett, ARNOLD publishers, 1991. Kamali, S., Moranville, M., Leclercq, S., Material and environmental parameter effects on the leaching of cement pastes: Experiments and modeling, Cement and Concrete Research, 38, 2008 Meller, N., Hall, Ch., Phipps, J.S. A new phase diagram for the CaO–Al2O3–SiO2–H2O hydroceramic system at 200°C. Materials Research Bulletin, 40, 2005.

Mulders Frans. Analysis of abandoned well integrity at a potential CO2 storage site. 3rd Well Bore Integrity Network Meeting, Santa Fe, USA, 2007.

Neuville, N. Study and modeling of physico-chemical degradation of oil wells cementing materials, PhD thesis, Ecole des Mines de Paris France, 2008 (in French).

Noïk, Ch., Rivereau, A., Vernet, Ch., Novel cements materials for high-pressure/high-temperature wells, SPE 50589 – SPE European Petroleum Conference held in The Hague, The Netherlands, 1998.

Ollivier, J.P. Minutes of technical days – AFPC-AFREM durabilité des bétons – Required methods for the measurements of parameters associated to durability, Toulouse, France, 1997 (in french).

Rafai, N., Letolle, R., Blanc, P., Gegout, P., Revertegat, E. Carbonation-decarbonation of concretes studies by the way of carbon and oxygen stable isotopes. Cement and Concrete Reseach, vol.22, 1992.

Shen. J.C., Pye, D.S. Effects of CO2 Attack on Cement in High-Temperature Applications. SPE/IADC 18618, 1989.

Span, R., and Wagner, W.. A new equation of state for carbon dioxide covering the fluid region from the triple-point temperature to 1100 K at pressures up to 800 MPa. J. Phys. Chem. Ref. Data., 25, 1996.

Taylor H.F.W., Cement Chemistry, 2nd edition, edited by Thomas Telford, 1997.

8 SPE 134473

FIGURES:

Figure 1: Examples of RX density images for 2 cement samples used for the test #2 (class G + silica flour), after curing period

Figure 2: Schematic of a cell used for experiments

Glass beaker Tube

Samples

Distilled water

Heating resistance regulation at 90°C or 140°C Supercritical CO2 at 90°C or 140°C and 80 bar

Gas injection Gas exit

3-tier sample-holder

Hounsfield scale: 1100 2600 20mm*40mm

SPE 134473 9

CO2solubility(m

ol.kg‐1H2O)

Temperature (°C)

Test 2Test 1

Figure 3: CO2 solubility in pure water under various temperature and pressure conditions (Span et al. 1997)

Figure 4 : Pressure monitoring for the test #1 (black line) test #2 (grey line)

10 SPE 134473

Figure 5 : Sagittal sections obtained by RX tomography on samples taken at different times in the test #1 (top) and test #2 (bottom). The gray scale (Hounsfield scale) represents RX density scale

0

1

2

3

4

5

0 2 4 6 8 10

Degrade

d thickness (mm)

Time of experiment (√day)

Figure 6: Thickness of carbonated cement as a function of square root of time (test #1: 90°C)

Before test

7 days

36 days 65 days

Hounsfield scale: 1100 2600

Test #1 (90°C)

Test #2 (140°C)

Before test

4 days 21 days 31 days

88 days

SPE 134473 11

0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Cumulated

 qua

ntity

 of ion

s in solution

 (mg for 

each

 sample)

Time of experiment (day)

Calcium Silicon

Figure 7: Total quantity of calcium and silicon in solution in the leachant (test #1: 90°C)

0102030405060708090100

1600 1700 1800 1900 2000 2100 2200

Cumulated

 freq

uency (%

)

RX density

4 days 12 days 88 days Initial material

Figure 8: Cumulated frequencies of RX density in the initial and the carbonated cement samples (test #2: 140°C)

12 SPE 134473

4 days

12 days

31 days

Carbonated frontHigher porouszone

Agressive

 fluid

Figure 9: SEM image (BSE) of carbonated cement samples after 4, 12 and 31 days of test #2 (140°C)

0

5

10

15

20

25

0 20 40 60 80 100

Degrade

d thickness (mm)

Time of experiment (day)

Figure 10: Thickness of carbonated cement as a function of time (test #2: 140°C)

1 mm

SPE 134473 13

0

10

20

30

40

50

60

0 1 2 3 4 5 6 7 8 9 10

CaCO

3 conten

t (% wt.)

Material depth (mm)

Reference sample 4 days12 days 21 days55 days 88 days

Figure 11: CaCO3 content (% wt.) in the initial (“reference”) sample and in the carbonated cement after different lengths of time (test #2: 140°C)

0

50

100

150

200

250

300

350

400

450

0 20 40 60 80 100

Cumulated

 quantity of ions in solution (m

g for each sample)

Time of experiment(day)

Calcium Silicon

Figure 12: Quantity of calcium and silicon in solution in the leachant (test #2: 140°C)

14 SPE 134473

Confining stress

Fluids : unreactive water, CO2

rich fluid

Axial stress

Piston

Cement sample

T = 90°C

S3

S1

Figure 13: Schematic of the dynamic chemo-mechanical test

Lime‐saturated water CO2‐richwater Figure 14: Evolution of parameters and sample response through to clogging: σ1 and σ3 stresses, injection pressure (pi), axial deformation (ε1) and injected volume

SPE 134473 15

Figure 15: Photograph of cement sample after the chemo-mechanical test