Non-Destructive Cyclic Analysis of Sealing Ability of Well Cement for Seasonal Underground Hydrogen Storage

Abstract

Underground hydrogen storage (UHS) is one potential solution that could provide a steady source of clean energy to the globe. Given their infrastructure, depleted hydrocarbon reservoirs may be a suitable storage option. However, ensuring wellbore integrity is a significant challenge when storing hydrogen in such reservoirs. In this study, 3.81 × 7.62 cm cement samples were cured for 12 and 18 months and were cyclically exposed to hydrogen for three 28-day cycles at 10.34 MPa and 50 °C. The pressure increment was achieved at the rate of 2.06 MPa/hr. The cement’s porosity, permeability, and ultrasonic velocity were tested before and after each cycle. To investigate the changes in the surface structure and elemental composition, scanning electron microscopy (SEM) was conducted. The results illustrate increased porosity and permeability, but the ultrasonic velocity changes were insignificant. The SEM images do not exhibit any change in the microstructure. However, energy dispersive spectroscopy (EDS) mapping exhibited mineral dissolution. This study demonstrates how cyclic exposure to hydrogen will affect the integrity and the sealing ability of aged cement, which will be an essential factor to consider while repurposing existing oil and gas wells to hydrogen injectors or producers for UHS applications in depleted hydrocarbon reservoirs.

1. Introduction

An essential component of the energy transition to a low-carbon, more sustainable future is hydrogen. It is the perfect fuel for a low-carbon energy system because it is a clean energy source [1]. The decarbonization of the energy system can be aided by producing hydrogen using renewable energy sources like solar and wind power. By doing so, it can be ensured that electricity is available when needed and assists in balancing the system [2,3]. These considerations underscore the need for comprehensive research, such as the study presented here, to address the challenges and potential benefits of storing hydrogen in depleted hydrocarbon reservoirs.

Because hydrogen has the potential to lower CO₂ emissions and support a low-carbon energy sector, research on hydrogen as a clean energy source has attracted a lot of attention [4]. Hydrogen is produced via various techniques, including electrolysis and methane fusion, underscoring the significance of large-scale, secure storage. Maximum hydrogen recovery is made possible by naturally occurring sealing qualities in underground storage solutions like salt and rock caverns. However, they have a limited capacity and are only available in certain areas. Although saline aquifers guarantee excellent hydrogen purity and have a huge storage capacity, they also produce a lot of water and have inadequate geological classification. On the other hand, depleted reservoirs have the advantage of already existing infrastructure and well-characterized formations, but they also carry the possibility of leakage, which might be expensive to maintain the infrastructure [5].

Depleted hydrocarbon resources are the most appropriate because hydrogen can be produced and delivered similarly to natural gas, and existing gas infrastructure can be converted to hydrogen use. Hydrogen storage in abandoned oil and gas reservoirs is an affordable solution. However, it is important to note that this approach also has certain disadvantages, including the possibility that residual gas or oil would lower the purity of the hydrogen, and the hydrogen may harm the reservoir’s top seal [6]. The different challenges associated with underground hydrogen storage in depleted hydrocarbon reservoirs are presented in Figure 1. However, difficulties with wellbore integrity could be the leading cause of problems during the injection and withdrawal of hydrogen [7].

The well integrity issues are primarily related to the steel casing and the annular cement. Hydrogen embrittlement (HE) is frequently observed in high-strength steels with tensile strengths exceeding 1000 MPa. It is triggered by the accumulation of hydrogen atoms in trapping sites, especially near grain boundaries. This type of damage can be reversed and usually requires an outside stressor. In contrast, low-strength steels with a tensile strength below 1000 MPa are particularly susceptible to hydrogen-induced cracking (HIC), which is brought on by internal pressure from the production of hydrogen molecules. HIC does not require external stress and causes irreparable damage. Hydrogen blistering (HB) primarily affects low-strength steels or materials with soft zones, also below 1000 MPa, and happens when hydrogen molecules gather in trapping sites near the surface [7].

Previous studies have indicated that the impact of CO₂ on cement seals is relatively minimal, suggesting limited degradation and a minor effect on the integrity of wellbores [8,9]. However, recent investigations have shed light on the role of geochemical processes between cement and CO₂, revealing that these interactions can enhance porosity within the cement structure, consequently leading to degradation [10,11,12]. This degradation occurs as CO₂ dissolves into the surrounding formation brine, resulting in the formation of carbonic acid, which in turn initiates the breakdown of the cement [13,14,15]. Chemical reactions, particularly those associated with carbonation driven by CO₂, can further compromise the strength of the cement and increase its permeability [16,17].

Unlike CO₂, hydrogen’s low density and viscosity may lead to rapid upward flow through cement seals, affecting their integrity and increasing the risk of leakage [18]. Understanding the alterations that cement undergoes during underground hydrogen storage (UHS) remains limited, with only a few studies below addressing this issue. Notably, high-pressure conditions and interactions between hydrogen and rock formations can contribute to the chemical degradation of cement during UHS operations [19,20]. Nevertheless, the potential for hydrogen leakage through various channels during production cycles, particularly in older wells, may gradually lead to cement deterioration, elastomer failure, caprock sealing failure, fluid migration through preexisting leakage routes, casing microbiological corrosion, hydrogen blistering, and hydrogen embrittlement, thereby compromising the overall integrity of the wellbore [21].

Fernandez et al., Hussain et al., and Al-Hadrami et al. studied the effects of H₂ exposure on neat Class H cement, which indicated increased permeability and porosity, and a decrease in the strength of cement [22,23,24]. Experimental investigations by Cracolici et al., Al Yaseri et al., and Vanessa et al. demonstrated no significant alterations in cement properties after exposure to hydrogen under reservoir conditions, suggesting potential suitability for underground hydrogen storage under specific reservoir conditions [25,26,27]. Investigations by Aftab et al. [28] demonstrated that no significant geochemical or structural alterations to cement were revealed due to exposure to hydrogen under simulated reservoir conditions, indicating potential security for underground hydrogen storage under specific conditions. Studies by Hussain et al. [29] and Nasiri et al. [30] investigated the well integrity of H₂ storage using computed tomography and scanning electron microscopy, where no changes in the microstructure were found. However, debonding is a leakage pathway for H₂ migration in depleted hydrocarbon reservoirs. Experimental results by Ugarte et al. [31] suggest increased cement strength and alterations in porosity distribution after exposure to molecular hydrogen under specific conditions, potentially enhancing wellbore integrity in underground hydrogen storage projects.

All the tests performed to assess well integrity for underground hydrogen storage (UHS) have utilized standard cement formulations (1.96 g/cc), typically cured for 3–7 days. However, the depleted hydrocarbon reservoirs designated for hydrogen storage will employ aged cement, which has been in the well for several years. Therefore, this study employed cement samples cured for a prolonged duration of 12 and 18 months to better simulate the conditions of cement encountered in actual UHS scenarios. Studies in the literature have mentioned that the compressive strength of cement increases and then remains stagnant for a curing period of 6 months [32], but after an elongated period, it decreases [33,34]. Another notable distinction is that, unlike Carbon capture and storage (CCS) applications where CO₂ is intended to be stored for extended periods, underground hydrogen storage (UHS) involves a cyclic seasonal process. Therefore, to comprehensively understand the impacts of UHS on well integrity, it is imperative to investigate the effects of cyclic exposure to hydrogen on cement and on its sealing ability. In the long term, interactions between hydrogen, cement, and the surrounding rock formations in the presence of formation fluids could exacerbate cement weakening and compromise containment measures. Thus, a comprehensive understanding of these processes is crucial for ensuring the long-term integrity and effectiveness of UHS operations.

This study aims to provide insights into the physical and chemical characteristics of cement, addressing concerns about its ability to form a reliable barrier for hydrogen containment during subsurface storage. This paper highlights the importance of performing non-destructive testing on the same cement samples for an accurate comparison of experimental results that will shed light on the effects of hydrogen on cement. This will aid in understanding the sealing ability over different aging timeframes and contribute to the development of safe and efficient underground hydrogen storage technologies.

2. Materials and Methods

2.1. Materials

Neat Class H cement, commonly used in oilfield operations until a depth of 8000 ft, where high pressure and temperature conditions prevail, with a water-to-cement ratio of 0.45, was used to make cement slurries 1Y1, 1Y2, 18M1, and 18M2 with a density of 1.85 g/cc. The composition of Class H cement is presented in the table. Class H Portland cement showcases robust hydration characteristics. Upon mixing with water, it reacts to form products like calcium silicate hydrate (C–S–H), which contributes significantly to the strength of the cement matrix. The heat of hydration is also a crucial parameter, as it affects the long-term durability and performance of the cement in adverse conditions.

Table 1. Composition of Class H Cement [35].

Clinker	Percentage
C3S	47.13
C2S	28.27
C4AF	17
C3A	0.65
CaSO4	4.72
Na2O	0.09
MgO	1.09
K2O	0.45
TiO2	0.18
MnO	0.07
ZnO	0.01
SrO	0.07
Free Lime	0.26

shows the clinker percentual composition of Class H cement.

LW1 and LW2 are field blends applied at a total vertical depth (TVD) of 2286 m for the intermediate casing with a density of 1.6 g/cc. LW1 and LW2 contain additives such as silica flour, extenders, fluid loss agents, and defoaming agents. The slurries were blended following the API-RP-10B-2 regulations. The cement was then poured into cylindrical molds of a 3.81 cm diameter and 7.62 cm length and cured for 12 and 18 months in an isothermal water bath at 50 °C (Figure 2a,b). All the cement batches were cured with four samples in a batch. Hydrogen with a purity of 99.99% from Airgas was used.

2.2. Methods

The experimental setup used in this study is shown below (Figure 3). The experimental conditions in this study simulated the reservoir conditions for UHS with a hydrogen storage pressure of 10.34 MPa at a temperature of 50 °C. The pressure and temperature conditions were estimated by referring to studies in hydrogen storage that mention the optimum operating conditions [36,37]. The cement specimens underwent exposure to hydrogen gas within a core holder apparatus utilizing an Ametek Chandler Engineering 2-cylinder Quizix pump (Chandler Engineering, Broken Arrow, OK, USA), with constant pressure maintained via pressure ramp cycling scheduler software. Pressure ramping occurred for 5 h at a rate of 2.06 MPa/hr, followed by a 28-day constant pressure set at 10.34 MPa. Subsequently, pressure reduction was executed over 5 h using the scheduler. The pressure of the system was monitored using a data acquisition system consisting of the Omega Digital Transducer Application software (version 2.3.1.306) and pressure transducers. This pressure cycling protocol was repeated for three cycles spanning 28 days each. Before each exposure cycle, a vacuum pump was employed to remove the air in the core holder. After the exposure cycles, the samples were placed inside air-tight containers to eliminate the possibility of reacting with the moisture in the air.

Computed tomography was employed to identify any pre-existing cracks or deformations within the specimens. The Ceretom^® NL3000 transportable small-bore CT scanner (Neurologica, Danvers, MA, USA), which has a maximum slice thickness resolution of 1.25 mm, is used to acquire the images. Because the CT scanner has eight detector rows, it may acquire eight picture slices simultaneously during a complete rotation. The scanner runs on tube voltages between 100 and 140 kV and currents between 1 and 7 mA. The scanner has a maximum scanning range of 64 cm and a rotation time range of 2 to 6 s. The thickness of each acquired picture slice is 1.25 mm. The image acquisition was performed before and after each pressure cycle, and the images were analyzed using ImageJ software to detect any cracks in the cement. Following each exposure cycle, a comprehensive analysis was conducted, encompassing computed tomography, scanning electron microscopy coupled with energy dispersive spectroscopy (SEM-EDS), evaluation of porosity and permeability, and measurement of ultrasonic velocities in the samples to calculate the dynamic elastic properties.

2.2.1. Porosity and Permeability Measurement

The porosity of the cement samples was determined using the Ultra-porosimeter (200A) (Core Laboratories, Tulsa, OK, USA) and the double-cell method for grain volume based on Boyle’s Law. Since helium has strong diffusion qualities and is an inert gas with less gas adsorption effects, it was used as the gas for the porosity measurement [38].

Using the pulse decay method in Autolab 1500 (New England Research, White River Junction, VT, USA), the permeability of cement samples was assessed in compliance with API-RP-40 requirements [39]. The samples were placed within a core holder as part of the experimental setup. The core holder was connected to two gas reservoirs—upstream and downstream—with pressure transducers installed to measure differential pressure. Both reservoirs were filled with gas until pressure equilibrium was reached, and the samples were initially subjected to a pore pressure of 1.72 MPa (250 psi). After that, the pressure upstream was raised, which allowed gas to pass through the sample and into the reservoir downstream. This study’s confining pressure was 10.34 MPa (1500 psi). The transducers recorded the resulting pressures, and the permeability of the samples was then determined. Nitrogen gas was used in this study to determine the permeability.

2.2.2. Dynamic Elastic Property Measurement

AutoLab 1500 (New England Research, White River Junction, VT, USA), was used in the investigation to determine the P- and S-wave ultrasonic velocities. This was accomplished by inserting cement samples in a high-pressure core holder and connecting them to ultrasonic transducers, which measured the wave velocities as they traveled under various confining pressures. Using a cyclic confining pressure regime, the experiment delivered an initial pressure of 3.7 MPa (500 psi), increased it progressively to 10.34 MPa (1500 psi), and then returned to the initial pressure. Dynamic elastic characteristics like Young’s modulus and Poisson’s ratio were computed using known equations by evaluating the ultrasonic velocity measurements [40].

$$E_d={\displaystyle \frac{\rho V_s^2(3V_{ p}^2-4V_s^2)}{V_p^2-V_s^2}}$$

(1)

$$v_d={\displaystyle \frac{V_{ p}^2-2V_s^2}{2{(V}_p^2-V_s^2)}}$$

(2)

where

$E_d$: Dynamic Young’s modulus

$\rho$: Density of the sample

$V_S$: S-wave propagation velocity

$V_p$: P-wave propagation velocity

$v_d$: Dynamic Poisson’s ratio

2.2.3. SEM-EDS Analyses

Scanning electron microscopy (SEM) images of the cement samples were obtained using a Hitachi S-4300 E/N (Hitachi, Santa Clara, CA, USA), considering their micropore structure. A variable pressure backscattered electron detector operating at a 10 kV accelerating voltage is featured in the Hitachi S/N 4300(Hitachi, Santa Clara, CA, USA). To reduce charging on the sample surface, the variable pressure is adjusted. The microstructural characteristics of the cement samples’ matrix (grain) were described using the SEM images. Measurements of the surface morphology, grain exposure of the mineral, and microfractures between grains were made on designated sections of all the samples before and after they were exposed to hydrogen. Each sample had three designated sections where SEM analyses were performed. SEM magnifications of 40, 100, 300, and 500 were used.

The energy dispersive spectroscopy (EDS) method can yield semi-quantitative data but is primarily utilized for qualitative material examination. An EDS system is typically included with SEM instrumentation to enable the chemical analysis of features seen in the SEM display. A simultaneous SEM and EDS analysis is beneficial when conducting a failure analysis since a spot analysis becomes vital to reaching a reliable result. The signals generated by an SEM/EDS system comprise X-rays utilized for chemical identification and quantification at measurable quantities and secondary and backscattered electrons used in image formation for a morphological analysis. The sample surface conditions affect the EDS detection limit; the smoother the surface, the lower the detection limit. Major and minor elements with concentrations greater than ten weight percent (major) and minor elements with concentrations between 1 and 10 weight percent can be detected using EDS. Since bulk materials have a 0.1 weight percent detection limit, trace elements (concentrations less than 0.01 weight percent) cannot be detected by EDS. For the three designated sections of each cement sample, EDS mapping was used to determine the changes in the elemental composition of the designated sections. Figure 4 represents the flowchart of the methodology used in this study.

3. Results and Discussion

A CT scan test was conducted on the samples to identify any crack(s) before and after each treatment cycle. The CT scan image results before conducting exposure cycles demonstrated that none of the cement samples contained any cracks. The image shows the absence of any cracks/fracture generation on the cement samples after the exposure cycles (Figure A1 of Appendix A). Due to their higher water-to-cement ratio, the LW1 and LW2 samples demonstrate more bubbles than 1Y1, 1Y2, 18M1, and 18M2. Using a micro-CT scan can be helpful for the further investigation and confirmation of this point.

3.1. SEM-EDS Analyses

The SEM images of the cement samples are presented in

Table 2. SEM Analyses of Cement Samples.

Batch	Before Exposure	After 1st Cycle	After 2nd Cycle	After 3rd Cycle
1Y1
1Y2
LW1
LW2
18M1
18M2

. The spots were marked to obtain an SEM image of the exact location and ensure accuracy. The cement samples’ SEM images were captured at 40, 100, 300, and 500 magnifications. The images presented are of 300 magnifications. The analyses of the samples highlighted no significant differences in microstructure before and after the exposure cycles.

Energy dispersive X-ray spectroscopy (EDS) is often used with an SEM to convert characteristic X-rays to electrical voltages to qualitatively and semi-quantitatively describe the distribution. Using SEM–EDS, qualitative and semi-quantitative elemental information can be obtained at an individual point (point analyses) and across an area (elemental maps). An SEM-EDS analysis reports detected elements in weight or atomic percent (at %). The EDS analysis report is presented in Figure 5, Figure 6 and Figure 7 in terms of at % for this study.

The samples 1Y1, 1Y2, 18M1, and 18M2 showed an overall reduction in the at% of Oxygen and Calcium. The overall Oxygen decrease of 3.1%, 2.2%, 2.1%, and 3.4% was observed in the samples 1Y1, 1Y2, 18M1, and 18M2, respectively. The Oxygen content increased for samples LW1 and LW2 by 1.3% and 2%, respectively, and there was no significant change in the Calcium content. The at% of Carbon increased by 3.5–5% in samples 1Y1, 1Y2, 18M1, and 18M2. However, there was an overall reduction in the at% of Carbon for samples LW1 and LW2. This difference in change can be attributed to different additives in the LW1 and LW2 samples. The change in the elemental information indicates hydrogen-induced reactions leading to the dissolution of the compounds in the cement matrix. This equation can give a possible explanation.

C − S − H + 2H⁺ ↔ Ca²⁺ + SiO_xOH_x

(3)

Due to mineral dissolution and solid particles breaking down into ions, the resulting ions are not visible in the mapping because the ions are smaller than microscopic or at deficient concentrations, not producing detectable signals for EDS mapping. Hence, a decrease in Calcium at% is observed in the EDS mapping.

3.2. Porosity

The porosity of the samples before and after exposure cycles is presented in Figure 8. There was an overall increase in the porosity of the samples after the cyclic exposures, with the most notable increment observed in samples after the first cycle for 1Y2, 18M1, and 18M2 samples. An overall increase in the porosity of LW2 is observed, and there is an overall reduction in LW1 porosity values. However, the changes in the LW1 and LW2 samples are not significant. After the first exposure cycle, approximately 29% and 21% change is observed in the porosity values of sample 18M1 and 1Y2, respectively. In contrast, sample 18M2 exhibits a porosity increase of 18% after the first cycle. Through micro-CT imaging, Al Yaseri et al. [26] evaluated the changes in porosity and concluded that there was a minimal decrease in porosity that can be attributed to the precipitation of the calcite and other high-density minerals in the cement. The observations were made through a spot analysis calculating the total porosity of Class G cement, whereas, in this study, effective porosity was measured for Class H cement.

The results confirm the occurrence of geochemical reactions within the cement samples when exposed to hydrogen. The minimal decrease in porosity suggests a near-equilibrium between dissolution and precipitation rates, resulting in insignificant porosity changes. Hydrogen-driven reduction in sulfate and ferric iron can disintegrate cement minerals, namely hematite and ettringite. As a result, sulfurides and ferrous iron are created, eventually precipitating iron oxide and sulfide minerals. However, because the minerals involved make up a minor part of the material, these reactions did not significantly affect the cement’s porosity. Our study’s central hypothesis is that these redox reactions achieved chemical equilibrium. However, we also recognize that their kinetics are constrained and that catalysis—from hydrogenotrophic bacteria, for example—is necessary for them to proceed. Thus, it can be inferred that the geochemical interactions (dissolution) involving hydrogen and cement remain relatively modest after the second and third exposure cycles.

3.3. Permeability

The permeability of the samples before and after treatment is presented in Figure 9. Permeability is a more sensitive indicator of changes in the properties of the sample attributes than porosity. The permeability of the samples increased after the cyclic hydrogen exposure, indicating changes in the pore network. Unlike the porosity results, the permeability of the samples showed a significant increase after each exposure cycle on all the samples. Sample 1Y1 exhibited the most increase in permeability values, while 18M1 demonstrated the least increase in permeability. After three cycles, the permeability of 18M1 increased from 0.14 μD to 1.1 μD. For 18M2, the permeability increased from 0.21 μD to 1.79 μD. Samples 1Y2, LW1, and LW2 have a minimal increase in permeability after the second exposure cycle. The dissolution of pore throats can lead to significant changes in permeability, whereas the corresponding modifications in porosity may be minimal. Nasiri et al. [30] evaluated field samples’ permeability before and after hydrogen injection, where they concluded an increase in permeability of 0.0442 mD when Class G cement was exposed to pure H₂ gas. However, considering this change in permeability in our results, the resulting values are considered appropriate for tight cement being able to make a good seal. Teodoriu et al. [41] mentioned in their study that cement is considered to make a good seal if its permeability is less than 0.01 mD. Further long-term hydrogen exposure of cement samples is required to evaluate the effect of hydrogen in permeability altercations.

3.4. Modulus of Elasticity and Poisson’s Ratio

The calculated dynamic Young’s modulus (E) and Poisson’s ratio (ν) before and after exposure cycles are presented in Figure 10 and Figure 11. The elastic properties of all samples exhibited a decrease following exposure to all the treatment conditions. These results are consistent with the porosity and permeability results; with the increase in porosity and permeability, the Modulus of Elasticity decreases [42]. The overall reduction in Young’s modulus in samples 1Y1, 1Y2, 18M1, and 18M2 after three exposure cycles is about 3%, 5%, 3%, and 4%, respectively. Samples LW1 and LW2 experienced the most significant decrease in their Modulus of Elasticity after hydrogen exposure. This can be because the samples with a higher water-to-cement ratio have less strength, which, in turn, can decrease the Modulus of Elasticity of the samples. A similar study concluded by Al Yaseri et al. [26] reported a slight increase in their samples of 6% in their study, indicating no major impact on wellbore integrity or the cement seal. The decrease in Young’s modulus of the cement indicates that the samples have become less brittle, making them less susceptible to cracking during cyclic loading.

However, the Poisson’s ratio changes observed in the samples were insignificant. The Poisson’s ratio values before and after hydrogen exposure were in the range of 0.2–0.24, which is similar to the values reported in the literature, confirming cement as a low-brittle material [26].

4. Conclusions

The experimental investigation focused on assessing the impact of cyclic hydrogen exposure on the mechanical and microstructural properties of different formulations of cement samples under simulated reservoir conditions. Significant insights were gained through a comprehensive analysis encompassing computed tomography, scanning electron microscopy coupled with energy dispersive spectroscopy, porosity and permeability evaluations, and measurement of ultrasonic velocities.

The results indicate that the cement samples exhibited minimal cracking or fracture generation after exposure to hydrogen gas, as confirmed by CT scan imaging. The SEM-EDS analysis revealed changes in elemental composition, indicating hydrogen-induced reactions leading to the dissolution of cement compounds. Notably, a reduction in at% of Oxygen and Calcium alongside an increase in Carbon content suggests a complex geochemical interaction involving hydrogen. Porosity measurements demonstrated increased sample porosity after cyclic hydrogen exposure, with significant increments observed particularly after the first exposure cycle. This suggests ongoing geochemical reactions within the cement matrix, albeit approaching equilibrium after subsequent cycles. Permeability results showed a notable increase in post-exposure cycles, indicating alterations in the pore network due to dissolution phenomena. Additionally, the dynamic Young’s modulus decreased after exposure cycles, making the cement less brittle and susceptible to cracking under cyclic loading conditions.

It can be concluded that cement with different additives performed differently when exposed to hydrogen. This study can be improved by including kinetics for the dissolution and precipitation of minerals and microbial kinetic equations. Furthermore, using reactive transport simulations, the diffusive movement of hydrogen inside the cement might be considered. Further investigations are warranted to assess the long-term effects of hydrogen exposure on cement properties. Extended exposure durations beyond the scope of this study would provide valuable insights into the sustainability and durability of cement structures under continuous hydrogen exposure. As discussed in [43,44] in the context of CCS, the occurrence of a self-sealing mechanism that can fill fractures and open spaces in the cement through the growth of secondary minerals can be relevant in UHS, and future studies can be conducted to investigate its effects. Advanced techniques such as micro-CT scanning can enhance understanding of microstructural changes induced by hydrogen exposure. This would facilitate a more detailed examination of crack initiation, propagation, and their implications on cement integrity.