Interfacial Evolution of Cement and Steel in CO2 Dissolved Solution Under High Temperature and High Pressure

Chengqiang Ren; Ye Peng; Bing Li; Shuliang Wang; Taihe Shi

doi:10.1515/htmp-2015-0125

Article Open Access

Interfacial Evolution of Cement and Steel in CO₂ Dissolved Solution Under High Temperature and High Pressure

Chengqiang Ren , Ye Peng , Bing Li , Shuliang Wang and Taihe Shi

Published/Copyright: October 24, 2015

Published by

Become an author with De Gruyter Brill

Submit Manuscript Author Information

From the journal High Temperature Materials and Processes Volume 35 Issue 8

Abstract

The experiments were operated for the cylindrical sample (cement/steel) in high temperature and high pressure (HTHP) CO₂ environment to simulate surrounding CO₂ attack in oil and gas well. The interfacial evolutions between well cement and casing steel were measured, including mechanical property, structure alteration, chemical change and electrochemical character. The interfacial behaviors are attributed to the competition of hydration and degradation of Portland cement. The damage at the interface was faster than the cement bulk deterioration by carbonation. Thus, the interface provided a potential flow leakage pathway for the HTHP gas and fluid in the well, so improving interfacial stability between well cement and casing steel is the key issue to long-term zonal isolation.

Keywords: Portland cement; interface; leakage pathway; CO2

Introduction

More and more oil and gas wells containing acid gases, including CO₂ and H₂S, have been widely found [1]. The long-term zonal isolation is a critical issue for the annulus between well bore and formation in the producing well and the abandoned well. If the integrity of this annulus is destroyed in CO₂ or H₂S geological storage environments, the hazardous gas and hydraulic flow can leak from the underground due to the high pressure, which results in well blowout and environmental pollution. Duguid pointed that there should be three flow pathways in this annulus, which includes pores in oil well cement, defects between casing and cement, and pathways between cement and formation [2].

Field investigations have shown serious degradation of cement in the down hole. In recent years, great efforts have been made to explain the degradation mechanism of oil well cement in various environments [3, 4].

However, a reliable barrier to leakage of flow from the subsurface depends not only on the good cement itself but also on the excellent bonding of cement to casing and formation. The integrity of the two interfaces, cement–casing and cement–caprock, should be the most important issue in acid gas-rich environments [5]. Moreover, the interface alteration between cement and caprock was especially concerned recently [6, 7]. It is necessary to reveal interfacial development of the cement–casing by diversity aggressive environments. Therefore, we launched researches on the stability of cement–casing interface. This work focuses on the investigation into the interface by simulating surrounding CO₂ attack in oil and gas well.

Experimental

Materials and environment

Class G Portland cement produced by Jiahua Enterprise Co. Ltd in China was used to prepare specimens, whose mineral compositions contain CaO 64.77%, SiO₂ 22.43%, Al₂O₃ 4.76%, Fe₂O₃ 4.10%, MgO 1.14%, Na₂O(K₂O) 0.24% and SO₃ 1.67% in weight fraction.

N80 steel, with chemical compositions of Mn 1.19%, Cr 0.036%, Ni 0.028%, Mo 0.021%, C 0.24%, Si 0.22, P 0.013%, S 0.004% and balance of Fe, was machined to cylindrical rod with the size of 10 mm diameter and 40 mm height. Screw thread was designed at one end of the steel.

The cement paste with W/C (water/cement) of 0.44 was used throughout the experiments. After strongly stirring, it was filled in the mold to prepare specimens. As schematically shown in Figure 1(a), the specimen of 50 mm diameter and 100 mm height was originally designed by embedding the cylindrical rod N80 steel to simulate the CO2 diffusion from caprock and toward the cement–casing interface through the cement bulk.

Figure 1:

Schematic structure of the specimen. (a) Sample for exposure and (b) sample for property test.

The curing of cement was operated at 80°C and 8 MPa pressed by N₂ gas for 24 h. After demolding, the sample was immersed into oxygen-free formation water with compositions of K⁺ 2.37 mg L, Na⁺ 66.6 mg L, Ca²⁺ 54.0 mg L, Mg²⁺ 4.18 mg L, Cl^– 101 mg L, SO₄^2– 43.7 mg L and HCO₃^– 133 mg L. The exposure experiment was operated at 80°C and 8 MPa CO₂ in an high temperature and high pressure (HTHP) autoclave in the laboratory. The periodic measurements were launched at 0, 10, 20, 30 and 40 days for physical and chemical properties. The cement bulk covering the screw thread of the steel rod was carefully peeled off. The diagram of the tested sample is shown in Figure 1(b). All reported data were calculated from five repeated samples statistically.

Mechanical test

Pullout test was carried out on a WDW-1000 materials testing machine to determine the bond strength of oil well cement to casing steel. A specially designed loading frame was adapted for the test under pullout loading, which ensured delamination at the interface. The maximum tensile force was recorded and converted into the interface shear strength according to eq. (1):

(1)τ=Fmax2πrl

where F_max (kN), r (m) and l (m) are ultimate pullout load, radius and length of steel, respectively. The compressive strength of cement bulk was operated on the WDW-1000 materials testing machine also.

Electrochemical test

Three-electrode system, including working electrode (cement/steel sample), reference electrode (Ag/AgCl in saturated KCl solution) and counter-electrode (ring Pt rounding the sample), was prepared to explore the electrochemical behavior. An Autolab PGSTAT302N electrochemical potentiostatic galvanostat was adapted to measure the electrochemical impendence spectroscopy (EIS) at open circuit potential by using alternating voltage amplitude of 20 mV over the frequency range of 10 mHz to 100 kHz.

Scanning electron microscopy and x-ray diffraction

JSM-6490LV scanning electron microscope was used to observe the microstructure of the fresh surface of cement near the interface, and the element analysis in observed zone was detected by GENESIS 2000 XMS energy-dispersive X-ray spectroscopy equipment. The products in this area were characterized by a DX-2000 Rigakudmax X-ray diffractometer with a copper Kα X-ray source. The scanning range of 2θ started from 15° to 70°.

Results

Mechanical property

The bonding strength of the interface between oil well cement and casing steel during different exposure times is presented in Figure 2. It increased with exposed time during the first 20-day exposure. However, it decreased when the exposure time extended to 40 days. An important result was found that the interface completely unstuck at 40th day due to the low value of 0.20 MPa.

Figure 2:

Interfacial bonding strength at different exposure times.

The compressive strength of the cement bulk was pictured in Figure 3. After 40 days exposure, the compressive strength, 13.5 MPa, was 47% of the peak value obtained at 20th day.

Figure 3:

Compressive strength of cement bulk at different exposure times.

The response of hardened cement to stress depends on the chemical compositions and physical structure, which varies with the complex reactions with environments. Traditionally, compressive strength and interfacial bond strength are widely considered to be the most important properties of cement annulus for zonal isolation under load, such as high gas pressure and thermal stress. The pathways for fluid leakage appear in the event of damage induced by stress. The higher compressive strength and interfacial bond strength are thus expected.

By comparing the changes of bonding strength of the interface and compressive strength of the cement bulk, the mechanical interlocking was destroyed. It should be drawn a conclusion that the degradation rate of the interface is faster than the cement bulk by carbonation.

Composition and structure

The elemental compositions in cement 100 μm thick near the interface were detected by energy-dispersive spectroscopy. The weight percent of elements are listed in Table 1. All products included Ca, Si, Al, C and O, where C and O could be used to distinguish corrosion degree. At 20 days, C was almost undetected, so the aggressive medium has not reached to the interface. After 40 days exposure, corrosion medium played role to the interface, where C and O were 13.13% and 35.84%, respectively.

Table 1:

Elemental composition in cement near the interface.

Sample	Elements (mass %)
	C	O	Al	Si	Ca
20 days	0.22	29.19	1.41	10.03	59.15
40 days	13.13	35.84	1.72	7.8	41.50

To further understand the degradation mechanism, x-ray diffraction (XRD) was applied to investigate the phase transformation. The result is shown in Figure 4. It is well known that the hydration product of cement is mainly composed of calcium hydroxide (CH: Ca(OH)₂), hydrated calcium silicate (CSH: xCaO-ySiO₂-zH₂O) and ettringite (AFt: 3CaO · Al₂O₃ · 3CaSO₄ · 30–32H₂O). At 20th day, both CH and CSH were presented. AFt was too little to be detected in this case. It means the compositions of cement paste are the hydration products. The most notable change was appearance of CaCO₃ accompanied by decrease of CH at 40th day, and the form of CSH was changed because of the change of peak intensity. It indicates corrosion has been occurred at the interface, and preferential carbonation of calcium hydroxide by CO₂ attack is found.

Figure 4:

XRD of cement near the interface. (a) 20th day and (b) 40th day.

Figure 5 shows the microstructure in cement 100 μm thick near the interface after 20 and 40 days exposure. At 20th day, integrated structure was found, whereas perforated feature was appeared at 40th day. The structural integrity was damaged.

Figure 5:

Microstructure of cement near the interface. (a) 20th day and (b) 40th day.

Electrochemistry

The EIS spectra and the corresponding equivalent circuits are shown in Figure 6. Two loops of capacitive impedance were judged from the curve at 40th day. The plot at 20th day yet indicated Warburg’s impedance at intermediate and low frequencies.

Figure 6:

Nyquist plot (a) and equivalent circuit of sample at different exposure times: (b) 20th day and (c) 40th day.

In the intermediate and low frequencies, the response reflects the cement–steel interfacial electrochemistry (a constant phase element Q_in and a charge transfer resistance R_in) [8]. The Warburg resistance (Z_W) superimposes in this region while the passivation occurs on the surface of steel, which should be predicted by the E-pH diagram of Fe–H₂O, because the pH exceeds 12.5 when the hydration deeply processes. Warburg resistance disappears at 40th day due to the neutralization of pore water alkalinity by CO₂.

The parallel circuit (R_hf and Q_hf) in high frequency is elucidated to account for the dielectric properties of cement bulk [8]. The last parameter R_s in the equivalent circuit model is the solution resistance in the pore of cement.

The parameter, R_in, reflects the interfacial stability. By fitting the curves, R_in was calculated as 1.31×10⁵ and 1.82×10² Ω at 20th and 40th days. The degradation of interface was obvious. Also the decreases of R_s and R_hf were observed, which was in favor of improving electrochemical reactivity of the interface by aggressive substances.

Discussion

On the basis of all physical and chemical results mentioned above, the interfacial change can be divided into two stages. One is the first 20 days, and the other is the following 20 days.

The mechanical property is mainly determined by the silicate gel in the cement, since CSH is the main binder phase [9], which has a cross-linking structure and makes up the bulk of the paste matrix. Similar to a composite material, the small molecular compounds play a vital role of filling, for instance, the hydration product of CH and degradation product of CaCO₃, they attribute to merely limited contribution to the mechanical performance but great role of decreasing porosity.

Although the cement slurry solidified after curing for 24 h, the continuous hydration was still in progress, which promotes the mechanical performance until the corrosion effect of the medium becomes dominant, because the carbonation of cement is diffusion-controlled process. The deduction is clearly evidenced by the interfacial bond strength as it increased drastically before the interface touched the aggressive medium in the first 20 days. The longer curing time is used, and the better structure and property are achieved. The same conclusion has also been implied by literatures [10–12]. They considered one of the main factors to the significantly higher permeability was the shorter curing time as reported by Barlet-Gouedard et al. [13]. In the similar experimental conditions, Kutchko et al. allowed the cement to hydrate for 28 days, so that the carbonation was limited after long-term exposure to CO₂-rich brine [10, 11].

Similar to the polymerization reaction, further hydration of CSH occurs as long as free H₂O exists:

(2)∑i=1nCxiSyiHzi+nH2O→CXSYHZ+nCH

As a result, the elevated polymerization degree (n) accelerates the microstructure integrity due to the cross-linking in three-dimensional space. The late product C_XS_YH_Z is therefore more resistant to physical change.

Among many reported reasons [14], chemical adhesion and friction are more close to the interfacial bonding mechanism. The chemical adhesion is due to the role of the polar group in hydration products to Fe. The friction is caused by the long-chain CSH surrounding the steel. The high hydration degree not only increases the length of chain but also results to volume shrinkage, which increases friction force.

The most excellent mechanical properties appeared at 20th day in this case, which lies on the hydration rate and hydration degree of cement in the given condition. Therefore, it may be observed at days, weeks or months in different cases.

The principal hydration products were CSH and CH. CSH represents a complex compound variable over quite a wide range because of nonstoichiometric elements. It is not a well-crystallized material but a cross-linking material. The high concentration of CH enhances the pH to 12.5 or higher. Therefore, the high alkaline environment pushes the steel into passivation region.

After chemical attack, the XRD suggests high alkalinity consumption caused by acid medium, which leads to almost disappearance of CH and formation of CaCO₃:

(3)Ca(OH)2+2H++CO32−→CaCO3+2H2O

When the pH drops below 11, the passive layer is destroyed [15], which corresponds to the disappearance of Warburg resistance. Nevertheless, CSH may be partly decomposed by carbonation:

(4)CXSYHZ+mH2CO3→∑j=1mCxjSyjHzj+mCaCO3+mH2O

The degradation reaction decreases the polymerization degree of silica gel to some degree. In fact, from the results of XRD, it is easy to find that CH is more preferential for carbonation by comparing with CSH.

The reaction followed by the CaCO₃ precipitation is well known as leaching:

(5)CaCO3+2H++CO32−→Ca(HCO3)2

where Ca(HCO₃)₂ is dissoluble. Thus, the porous structure is observed. However, the slow degradation of CSH maintains major cross-linking structure of cement.

On the one hand, corrosion products form a weak layer, which weakens chemical adhesion. On the other hand, the degradation of CSH decreases the chain length, which reduces the surrounding effect to the steel, the mechanical friction, thus, goes down.

Barlet-Gouédard has predicted the cement debonding from casing even before chemical degradation of cement [16], which is manifested by this experimental result.

Though Carey et al. observed FeCO₃ and CaCO₃ filling in groove [17], they cannot contribute to the interfacial bonding, and further they should dissolve in acid solution to provide pathways to flow. The interfacial bond strength is considered to be the most important property of cement annulus for zonal isolation under load, such as high pressure and thermal stress. The pathways for flow leakage appear in the event of damage induced by stress. A 0.1–0.3 cm thick carbonate zone near the cement–casing interface was observed in CO₂-rich reservoir in West Texas [18]. The cement was still integrated, but the cement–casing interface lost stability. It has been calculated that 0.01–0.3 mm gap increased the effective permeability of multiphase flow up to 10⁶ times to the integrated interface [19]. Thus, establishing the standard for the interfacial test and improving stability of the interface are the most important works in the future to ensure the long-time zonal isolation.

Conclusions

The interface between oil well cement and casing steel was researched in HTHP CO₂ environment in order to simulate the CO₂ diffusion from caprock and toward the cement–casing interface through the cement bulk. It was found that the interface lost complete stability before the integrity damage of the cement bulk coated on the casing steel. The corrosion at the interface resulted in porous structure of cement and debonding to the steel. The pathway for flow leakage is provided at the interface. Two aspects, the standard for evaluation of the interfacial integrity and the method for mitigation of interfacial degradation, should be done preferentially during processes and application of oil well cement in HTHP aggressive solution.

Funding statement: Funding: Financial supports by open fund (PLN1306) of State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation (Southwest Petroleum University), National Natural Science Foundation of China (51374180) and Key Lab of Material of Oil and Gas Field (X151515KCL08) are acknowledged.

References

[1] E. Lécolier, A. Rivereau and N. Ferrer, SPE Drill Compl., 25 (2010) 90–95.10.2118/99105-PASearch in Google Scholar

[2] A. Duguid, SPE, 2008, paper no. 119504.Search in Google Scholar

[3] B.G. Kutchko, B.R. Strazisar, D.A. Dzombak, G.V. Lowry and N. Thaulow, Environ. Sci. Technol., 41 (2007) 4787–4792.10.1021/es062828cSearch in Google Scholar

[4] B.G. Kutchko, B.R. Strazisar, S.B. Hawthorne, C.L. Lopano, D.J. Miller, J.A. Hakala and G.D. Guthrie, Int. J. Greenh. Gas Con., 5 (2011) 880–888.10.1016/j.ijggc.2011.02.008Search in Google Scholar

[5] H.B. Jung, D. Jansik and W. Um, Environ. Sci. Technol., 46 (2012) 283–289.Search in Google Scholar

[6] M. Wigand, J.P. Kaszuba, J.W. Carey and W.K. Hollis, Chem. Geol., 265 (2009) 122–133.10.1016/j.chemgeo.2009.04.008Search in Google Scholar

[7] H.E. Mason, W.L. Du Frane, S.D.C. Walsh, Z. Dai, S. Charnvanichborikarn and S.A. Carroll, Environ. Sci. Technol., 47 (2013) 1745–1752.10.1021/es3039906Search in Google Scholar

[8] R. Vedalakshmi, R.R. Devi, B. Emmanuel and N. Palaniswamy, Mater. Struct., 41 (2008) 1315–1326.10.1617/s11527-007-9330-1Search in Google Scholar

[9] F. Zhang, S. Wu, Y. Fang, J. Zhou and Z. Li, J. Wuhan Univ. Technol.-Mater. Sci. Ed., 29 (2014) 507–512.10.1007/s11595-014-0949-9Search in Google Scholar

[10] B.G. Kutchko, B.R. Strazisar, N. Huerta, G.V. Lowry, D.A. Dzombak and N. Thaulow, Environ. Sci. Technol., 43 (2009) 3947–3952.10.1021/es803007eSearch in Google Scholar

[11] B.G. Kutchko, B.R. Strazisar, G.V. Lowry, D. Dzombak and N. Thaulow, Environ. Sci. Technol., 42 (2008) 6237–6242.10.1021/es800049rSearch in Google Scholar

[12] G. Dotelli and C.M. Mari, Mater. Sci. Eng. A, 303 (2001) 54–59.10.1016/S0921-5093(00)01886-4Search in Google Scholar

[13] V. Barlet-Gouédard, G. Rimmelé, B. Goffé and O. Porcherie, SPE, 2006, paper no. 98924.Search in Google Scholar

[14] K. Lundgren, Mag. Concrete Res., 59 (2007) 447–461.10.1680/macr.2007.59.6.447Search in Google Scholar

[15] G. Blanco, A. Bautista and H. Takenouti, Cement Concrete Comp., 28 (2006) 212–219.10.1016/j.cemconcomp.2006.01.012Search in Google Scholar

[16] V. Barlet-Gouédard, G. Rimmelé, O. Porcherie, N. Quise and J. Desroches, Int. J. Greenh. Gas Con., 3 (2009) 206–216.10.1016/j.ijggc.2008.07.005Search in Google Scholar

[17] J.W. Carey, R. Svec, R. Grigg, P.C. Lichtner, J. Zhang and W. Crow, Energ. Proced., 1 (2009) 3609–3615.10.1016/j.egypro.2009.02.156Search in Google Scholar

[18] J.W. Carey, R. Svec, R. Grigg, P.C. Lichtner, J. Zhang and W. Crow, Int. J. Greenh. Gas Con., 13 (2007) 75–85.10.1016/S1750-5836(06)00004-1Search in Google Scholar

[19] S. Bachu and D.B. Bennion, Int. J. Greenh. Gas Con., 3 (2009) 494–501.10.1016/j.ijggc.2008.11.002Search in Google Scholar

Received: 2015-5-21

Accepted: 2015-9-6

Published Online: 2015-10-24

Published in Print: 2016-9-1

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Articles in the same Issue

https://doi.org/10.1515/htmp-2015-0125

Keywords for this article

Portland cement; interface; leakage pathway; CO2

Creative Commons

BY-NC-ND 3.0