News

Abstract

This paper presents results of design and evaluations of various Al-rich cements under supercritical hydrothermal and CO₂ field conditions. Results of short-to-long term exposure tests up to 30 days under hydrothermal and up to 9 months under supercritical CO₂ environments are reported including mechanical properties, morphological, and phase changes for calcium-aluminate cement-based blends. Performance of aluminum-rich cement formulations is compared to that of Ordinary Portland Cement (OPC) with silica blend. The results confirm that properties of the blends with lower calcium content persist under supercritical CO₂, while the OPC blend undergoes severe degradation with the loss of mechanical properties. For alkali activated and chemical blends partial carbonation occurs through calcium carbonation. Electrochemical corrosion tests showed that Al-rich cement formulations can provide Ni-alloy corrosion protection under supercritical conditions. All tested calcium-aluminate-cement based blends outperformed OPC/silica under the CO₂-rich well conditions.

Abstract

Cement Integrity Assessment Using a Hydration-Coupled Thermo-Mechanical Model

Portland cement is commonly used in wells to provide zonal isolation in the annulus. A damaged cement sheath can expose the casing to corrosive fluids and open a leakage pathway to shallow freshwater aquifers and atmosphere. The leakage can manifest itself as sustained casing pressure (SCP) or lead to gas accumulation in shallower formations. The impact of pressure and temperature variation on cement stress has been widely studied in the literature. However, the hydration reactions of cement are not usually included in the mechanical models. This leads to incorrect assumptions about the initial state of stress in cement immediately after curing.

In this work, we have developed a 3D well integrity model that incorporates the cement hydration process. The model is verified using laboratory experiments on cement stress evolution. The model calculates the water consumption during the hydration reactions to predict the pore pressure change in cement. The evolution of cement's mechanical properties with the hydration degree is captured using a homogenization model. A case study is designed to represent a typical low-enthalpy geothermal well in the Netherlands, using well designs and inputs from publicly available data. The cement stresses are tracked over the life of the well, to understand the magnitude of the stress cycles and to assess the potential long-term damage to the cement sheath.

The results show that the pore pressure drop due to cement hydration causes an increase in shear stress in the cement sheath. The pore pressure drop during hydration can debond the cement from the formation. The level of destressing in cement is a function of cement properties, formation stiffness, and the depth of the top of cement. When placed against softer formations, the stress drop in cement is more muted leading to a better seal. During the temperature cycles, the shear stress in cement changes in a cyclical manner. Depending on the magnitude of the stress cycles, damage can be accumulated in the cement sheath. The stress evolution in cement can also vary depending on the presence of external water (formation permeability).

The modelling technique presented in this work provides a robust methodology to estimate the magnitude of cyclical stresses in the cement sheath. This is a critical input to design cement recipes that can withstand load cycles throughout the lifetime of the well. The results of this work indicate the need to assess the integrity of cement at various depths and against various formations. It may not be possible to guarantee the seal efficiency against all formations, however risk analysis can be conducted using the presented model to assess the seal integrity of critical locations in the well profile.

Abstract

Cements for Use in Supercritical Geothermal Wells at 400oC

Efficiency and economics of geothermal energy production could be significantly improved if high temperature (HT) wells could be reliably exploited. Ten times higher energy output could be achieved with supercritical geofluid compared to the conventional geothermal wells. This paper presents the design and characterization of geothermal cement formulations for supercritical wells. Performance of calcium-silicate-, calcium-aluminate-, and aluminum- based cements after exposure to super critical water (scH₂O, 400^oC, 25.5 MPa) for up to 30 days is presented. Mechanical properties, including compressive strength, Young’s modulus, compressive toughness are reported along with the water-fillable porosity. Microstructural development and phase compositions and transitions are linked to the mechanical and physical properties of different formulations. Common OPC/silica flour-based cement experienced slow increase in porosity and deterioration of mechanical properties during the prolonged exposure to scH₂O. Several alternative formulations demonstrated persistent mechanical properties with stable phase compositions or phase transitions that did not have a negative effect on their performance.

Abstract

Results of High-Temperature Cement Blends Exposure in Newberry Well, Oregon

This paper discusses the results of cement exposure tests in a deep (~3 km) high-temperature (HT) Newberry well in Oregon. To conduct the tests cement-exposure tools were fabricated, and various HT cement formulations were exposed to the well conditions (300^oC) for a period of 3 months. The paper presents changes in mechanical and physical properties of the exposed samples along with their microstructural alterations and phase transitions. Results of the cement formulations performance are directly correlated with their compositions and composition alterations. The tested cement formulations included reference OPC/silica HT blends, cements with secondary cementitious materials, calcium-aluminate-based cements (CAC), and calcium-free cement. Based on XRD analyses, SEM imaging coupled with EDX elemental composition determination, and thermal gravimetric tests it was shown that an Ordinary Portland Cement (OPC) formulations undergo severe carbonation in the 3 months of exposure with calcium carbonate being the only crystalline product that forms in the cement. Although mechanical properties of the HT OPC-silica blend persisted due to the cement matrix densification with calcium carbonate, its long-term performance is questionable with the likely further formation of calcium bicarbonate and its eventual dissolution, and erosion of remaining amorphous silica after longer exposure times. Partial carbonation of calcium-phosphate based cement and other CAC-based blends was also observed. Calcium free, aluminum-based cement did not experience any carbonation. Possible solutions for deep geothermal well cementing are discussed.

Abstract

In the transition from fossil fuels to sustainable, renewable energy sources, development of geothermal energies sources provides an important contribution to ensure the green shift. Per today, most geothermal wells operate at a temperature range of 150-300 ⁰C, but in order to increase power production, a transition to more supercritical conditions (~500 ⁰C) is needed. It is generally accepted that Ordinary Portland Cement (OPC)is not suitable for supercritical wells.

In this paper we report mechanical properties and pressure cycling-test (function tests) results of two alternative Calcium-Aluminate-Cement (CAC)-based systems at 300oC and 400oC hydrothermal conditions.

Down-scaled simplified functional tests to mimic well-start up have been performed to evaluate cement sheath integrity under cycling pressure and constant temperature . The setup consists of a downscaled section of a wellbore with casing-cement-rock, and with the use of a pressure shaft inside the casing pressure cycling experiments are performed to investigate the cement sheath integrity and degradation under the increasing pressure loadings. Lower strength and Young’s modulus CAC/silica flour blend failure pressure was higher than that for CAC/silica/phosphate blend. Both CAC formulations outperformed OPC/Silica HT blend.”

Abstract

Geothermal heat is expected to become an important source of sustainable energy for the Netherlands and geothermal wells are expected to last for decades and provide an acceptable return on investment. The well cement in geothermal wells experiences a unique stress condition due to the injection/production of cold/hot water in a doublet. The impact of these cyclical loads on cement is not known. Critical state constitutive models such as the Modified cam-clay (MCC) are the most suitable to describe cement’s mechanical behavior. In this work, we present a new experimental protocol to measure the MCC model parameters for well cement and to quantify the plastic damage accumulation under realistic cyclical conditions relevant to the Dutch geothermal wells. The protocol was tested on class G cement.

The experiments were carried out in a triaxial apparatus and consisted of (1) unconfined compressive strength (UCS) tests; (2) hydrostatic compression tests to find the initial size of the yield surface (p₀) and the slope of the swelling and compression lines (κ and λ); (3) triaxial tests on both the dry and wet side of the yield surface; and finally (4) cyclic loading tests under conditions relevant to a mid-enthalpy doublet. The effect of cyclic loading on cement integrity is quantified by measuring the inelastic deformation and elastic moduli after each cycle followed by a triaxial test.

The UCS tests on class G cement show peak stresses of 34-41 MPa, and critical state stresses (CSS) of approximately 10 MPa. The triaxial tests performed on the dry and wet sides display a strain softening and hardening behavior, respectively, as predicted by the MCC model. CSS values from all the tests align well along the newly defined critical state line (CSL) for class-G cement. The CSL has a slope M = 1.5 and an intercept (p) of -3.7 MPa, which is close to the anticipated tensile strength of cement. Hydrostatic tests give p₀ of 23-29 MPa and κ and λ of 0.0046 and 0.02.  The cyclic test simulating the producer well shows negligible plastic deformation after 10 cycles. Conversely, the deep injector well shows cement damage accumulation.

Presentation Abstract

Modelling stress evolution in well cement during hydration

The integrity of cement/casing bond in a plug or an annulus must be maintained to ensure zonal isolation in wells. The initial stress state of cement after curing is a critical parameter that sets the safe operating window for the subsequent pressure and temperature changes in a well. Recent experimental studies indicate that cement plugs experience a loss in pore pressure and radial stress during hydration. The drop in radial stress could result in cement debonding from the casing.

In this work, we have developed a methodology to model the stress evolution in cement plugs during hydration. The model begins with the slurry state of cement and calculates the water consumption and void creation over time as the hydration reactions progress. The void volume change due to chemical shrinkage is imported into a coupled mechanical model that calculates the pore pressure drop and the resulting change in radial stress. The results of the proposed modelling methodology are verified using lab experiments from the literature.

The results provide new insights in understanding cement behavior under lab and field conditions. Under most scenarios, the cement pore pressure drops to the water saturation pressure which leads to evaporation of the remaining pore water. This pore pressure drop controls the radial stress change, according to the theory of poro-elasticity. For a plug set under an initial pressure of 40 MPa, the radial stress drops to 20 MPa after 40 hours of curing. For an initial pressure of 20 MPa, radial stress decreases to 8 MPa after 25 hours. This stress drop can cause the cement to debond from the casing, if the fluid pressure below the plug exceeds the final radial stress. The bulk shrinkage of cement can also be estimated using the model. Bulk shrinkage is equal to chemical shrinkage until a solid skeleton is formed in cement. Afterwards, bulk shrinkage is driven by the pore pressure change and the physical boundary conditions.

The proposed methodology in this work demonstrates that the pore pressure and stress drop in cement can be effectively modelled. This allows us to determine the initial stress state of cement after curing to better understand the integrity of the annular cement and cement plugs, under field conditions. Furthermore, understanding the mechanisms responsible for the cement stress drop helps in designing better cement recipes that can maintain higher initial tress levels.

Abstract

Assessing the Impact of Temperature Cycles on well Cement in Low Enthalpy Geothermal Wells

Geothermal energy is an important candidate to aid the transition to a low-carbon economy. Lowenthalpy geothermal wells are particularly important to jurisdictions with a low geothermal gradient, such as the Netherlands. In order to be economically viable, low-enthalpy wells should be relatively cheap to drill and last for several decades, longer than typical oil and gas wells. The required longevity for these wells presents new challenges in terms of well integrity, particularly related to cement and casing damage.

Geothermal wells experience temperature cycles as they are suspended for various operational reasons. This leads to a change in the casing/cement temperature along the well, for hundreds or thousands of cycles during its lifetime. The temperature cycles lead to changes in cement stress, particularly due to the differential thermal expansion of the casing/cement/formation system. In order to ensure a long operational life, the cement formulation should be designed to be able to withstand the cyclical stresses induced by the temperature cycles. A damaged cement sheath can expose the casing to corrosive fluids and open a leakage pathway to shallow freshwater aquifers.

In this work, we have developed a 3D well integrity model that incorporates the cement hydration process. The model is previously verified using laboratory experiments. A case study is designed to represent a typical lowenthalpy geothermal well in the Netherlands, using well designs and inputs from publicly available data. The cement stresses are tracked over the life of the well, to understand the magnitude of the stress cycles and assess the potential long-term damage to the cement sheath.

The results show that the pore pressure drop due to cement hydration causes an increase in shear stress in the cement sheath. The pore pressure drop during hydration can debond the cement from the formation, barring the use of expansive agents in the cement recipe. During the temperature cycles, the tangential and axial stresses in cement change more than the radial stress. Axial stress in the cement sheath is particularly sensitive to temperature changes due to the axial movement of the casing. The shear stress in the cement sheath cycles between 16 and 18 MPa (2,300 and 2,600 psi) as the temperature fluctuates between 80 and 30 °C (176 and 86 °F), at a depth of 2000 m for the particular conditions in the present case study.

The modelling technique presented in this work provides a robust methodology to estimate the magnitude of cyclical stresses in the cement sheath of low-enthalpy geothermal wells. This is a critical input to design cement recipes that can withstand thousands of load cycles throughout the lifetime of a geothermal well.

Geothermal Rising Conference

Abstract

Modelling stress evolution in well cement during hydration

The integrity of cement/casing bond in a plug or an annulus must be maintained to ensure zonal isolation in wells. The initial stress state of cement after curing is a critical parameter that sets the safe operating window for the subsequent pressure and temperature changes in a well. Recent experimental studies indicate that cement plugs experience a loss in pore pressure and radial stress during hydration. The drop in radial stress could result in cement debonding from the casing.

In this work, we have developed a methodology to model the stress evolution in cement plugs during hydration. The model begins with the slurry state of cement and calculates the water consumption and void creation over time as the hydration reactions progress. The void volume change due to chemical shrinkage is imported into a coupled mechanical model that calculates the pore pressure drop and the resulting change in radial stress. The results of the proposed modelling methodology are verified using lab experiments from the literature.

The results provide new insights in understanding cement behavior under lab and field conditions. Under most scenarios, the cement pore pressure drops to the water saturation pressure which leads to evaporation of the remaining pore water. This pore pressure drop controls the radial stress change, according to the theory of poro-elasticity. For a plug set under an initial pressure of 40 MPa, the radial stress drops to 20 MPa after 40 hours of curing. For an initial pressure of 20 MPa, radial stress decreases to 8 MPa after 25 hours. This stress drop can cause the cement to debond from the casing, if the fluid pressure below the plug exceeds the final radial stress. The bulk shrinkage of cement can also be estimated using the model. Bulk shrinkage is equal to chemical shrinkage until a solid skeleton is formed in cement. Afterwards, bulk shrinkage is driven by the pore pressure change and the physical boundary conditions.

The proposed methodology in this work demonstrates that the pore pressure and stress drop in cement can be effectively modelled. This allows us to determine the initial stress state of cement after curing to better understand the integrity of the annular cement and cement plugs, under field conditions. Furthermore, understanding the mechanisms responsible for the cement stress drop helps in designing better cement recipes that can maintain higher initial tress levels.

US Rock Mechanics/Geomechanics Symposium

Paper

Cement Formulations for Super-Critical Geothermal Wells

This work is a part of a joint international effort to develop “sustainable geothermal well cement for challenging thermo-mechanical conditions” (Test-Cem) under the umbrella of Cofund GEOTHERMICA. The paper describes an exploratory study of cementitious composites for applications in geothermal wells under super-critical conditions. Preliminary screening of a wide range of blends was done on formulations mixed at room temperature, cured under hydrothermal conditions at 85oC (overnight) followed by overnight hydrothermal curing at 300oC and the final curing under the super-critical conditions at 400oC and pressure of 25.5 MPa (3,700 psi). The tested formulations included high-temperature Portland cement modified with silica (CaO-SiO₂ system), P₂O₅-Na₂O-CaO-Al₂O₃-SiO2 – based systems with varied Al/Si ratios, and P₂O₅-Na₂O-CaO-Al₂O₃-SiO₂-Fe₂O₃ phosphate systems, Na₂O-CaO-Al₂O₃-SiO₂-MgO-Fe₂O₃ and Na₂O-CaO-Al₂O₃-SiO₂-MgO Mg-containing systems, and Na₂O-Al₂O₃-SiO₂ cement system.

Mechanical properties of the blends and crystalline phase transitions were determined after the initial curing and exposure to the supercritical conditions to understand effect of phase transitions upon materials performance. Formulations with short-term compressive strength development of no less than 6.9 MPa (1,000 psi) and stable phase compositions were identified for further testing to fully evaluate their true potential.

Cement Formulations for Super-Critical Geothermal Wells