• Energy Research

AbstractThousands of abandoned wellbores may lie within the aerial extent of a CO2 storage operation. These wellbores represent a potential leakage pathway and a leaky wellbore needs to be re-completed or otherwise repaired to restore seal integrity and ensure containment of the stored CO2. Due to the high cost of recompleting a well, a sufficient economic incentive exists if a viable seal repair technology is available. In this paper, we examine the use of epoxy nanocomposites as potential seal repair materials that have excellent bond characteristics with both steel and cement when cured in the subsurface environment. Test results show Novolac epoxy nanocomposites incorporating nanosilica, nanoclay or nanoalumina to have acceptable flowability that enable injection in wellbore cracks and significantly higher bond strength compared with standard microfine cement.

AbstractThousands of abandoned wellbores may lie within the aerial extent of a CO2 storage operation. These wellbores represent a potential leakage pathway and a leaky wellbore needs to be re-completed or otherwise repaired to restore seal integrity and ensure containment of the stored CO2. Due to the high cost of recompleting a well, a sufficient economic incentive exists if a viable seal repair technology is available. In this paper, we examine the use of epoxy nanocomposites as potential seal repair materials that have excellent bond characteristics with both steel and cement when cured in the subsurface environment. Test results show Novolac epoxy nanocomposites incorporating nanosilica, nanoclay or nanoalumina to have acceptable flowability that enable injection in wellbore cracks and significantly higher bond strength compared with standard microfine cement.

AbstractStoring carbon dioxide in depleted petroleum reservoirs is a viable strategy for carbon mitigation, but ensuring that the sequestered CO2 remains in the formation is vital to the success of such projects. There is great concern for the development of leakage pathways through annuli between the well cement and the formation or the casing. Predicting the behavior of such potential leakage pathways is critical. Numerical simulations conducted using a reactive transport module match well with experimental studies, but also show the necessity of quantifying the transport and mechanical properties of the leached solid cementitious solids–predominantly silica gel–produced by carbonic acid corrosion of well cement.Bench-top experiments have been performed with the following goals in mind: (1) to investigate the parameter space of relevant corrosion boundary conditions, e.g. pH, CO2 concentration, and calcium ion concentration, (2) to produce samples that can be used to quantify the transport and mechanical properties of acid corroded Class H well cement, and (3) to validate and improve the accuracy of numerical simulations of the reaction of well cement with carbonic acid.Class H cement samples were uniaxially corroded via exposure to a brine of constant composition. Constant composition is ensured by constant renewal of the brine at a rate larger than cement reaction rate. H+, Ca2+ and CO2 total aqueous concentration in the NaCl brine are controlled independently by adding known amounts of NaCl, HCl, CaCl2 and NaHCO3 and by controlling CO2 partial pressure. Microscopic (30X) time-lapse videos were taken of each sample so that corrosion front movements could be accurately measured. These experiments have yielded corrosion front measurements that clearly show that corrosion front advancement is diffusion controlled (i.e., linear as a function of the square root of time). The uniaxial corrosion of these samples has not only allowed for detailed measurements of the corrosion front, but also affords the opportunity to measure the mechanical properties of the corroded samples as a function of depth. The one-dimensional corrosion also allows for measuring the diffusion coefficient of the outer layer of silica gel by low field Nuclear Magnetic Resonance (NMR).Measuring the kinetics under various boundary conditions has validated the modeling results reported by Huet et al.. The measurements of mechanical and transport properties can now be used to improve the predictive power of these simulations by providing much needed information on the exterior layer of corroded Class H well cement. Additionally, these experiments offer experimental validation that the corrosion kinetics are enhanced by the presence of CO2 and open the door to better understanding of the mechanism of, and boundary conditions that might lead to, “pore-plugging” by the corrosion products, which in turn leads to a drastic retardation of the corrosion reaction.

AbstractStoring carbon dioxide in depleted petroleum reservoirs is a viable strategy for carbon mitigation, but ensuring that the sequestered CO2 remains in the formation is vital to the success of such projects. There is great concern for the development of leakage pathways through annuli between the well cement and the formation or the casing. Predicting the behavior of such potential leakage pathways is critical. Numerical simulations conducted using a reactive transport module match well with experimental studies, but also show the necessity of quantifying the transport and mechanical properties of the leached solid cementitious solids–predominantly silica gel–produced by carbonic acid corrosion of well cement.Bench-top experiments have been performed with the following goals in mind: (1) to investigate the parameter space of relevant corrosion boundary conditions, e.g. pH, CO2 concentration, and calcium ion concentration, (2) to produce samples that can be used to quantify the transport and mechanical properties of acid corroded Class H well cement, and (3) to validate and improve the accuracy of numerical simulations of the reaction of well cement with carbonic acid.Class H cement samples were uniaxially corroded via exposure to a brine of constant composition. Constant composition is ensured by constant renewal of the brine at a rate larger than cement reaction rate. H+, Ca2+ and CO2 total aqueous concentration in the NaCl brine are controlled independently by adding known amounts of NaCl, HCl, CaCl2 and NaHCO3 and by controlling CO2 partial pressure. Microscopic (30X) time-lapse videos were taken of each sample so that corrosion front movements could be accurately measured. These experiments have yielded corrosion front measurements that clearly show that corrosion front advancement is diffusion controlled (i.e., linear as a function of the square root of time). The uniaxial corrosion of these samples has not only allowed for detailed measurements of the corrosion front, but also affords the opportunity to measure the mechanical properties of the corroded samples as a function of depth. The one-dimensional corrosion also allows for measuring the diffusion coefficient of the outer layer of silica gel by low field Nuclear Magnetic Resonance (NMR).Measuring the kinetics under various boundary conditions has validated the modeling results reported by Huet et al.. The measurements of mechanical and transport properties can now be used to improve the predictive power of these simulations by providing much needed information on the exterior layer of corroded Class H well cement. Additionally, these experiments offer experimental validation that the corrosion kinetics are enhanced by the presence of CO2 and open the door to better understanding of the mechanism of, and boundary conditions that might lead to, “pore-plugging” by the corrosion products, which in turn leads to a drastic retardation of the corrosion reaction.

Abstract Seal integrity of functional oil wells and abandoned wellbores used for CO 2 subsequent storage has become of significant interest with the oil and gas leaks worldwide. This is attributed to the fact that wellbores intersecting geographical formations contain potential leakage pathways. One of the critical leakage pathways is the cement-shale interface. In this paper, we examine the efficiency of a new polymer nanocomposite repair material that can be injected for sealing micro annulus in wellbores. The bond strength and microstructure of the interface of Type G oil well cement (reference), microfine cement, Novolac epoxy incorporating Neat, 0.25%, 0.5%, and 1.0% Aluminum Nanoparticles (ANPs) with shale is investigated. Interfacial bond strength testing shows that injected microfine cement repair has considerably low bond strength, while ANPs-epoxy nanocomposites have a bond strength that is an order of magnitude higher than cement. Microscopic investigations of the interface show that micro annulus interfacial cracks with widths up to 40 μm were observed at the cement-shale interface while these cracks were absent at the cement-epoxy-shale interface. Fourier Transform Infrared and Dynamic mechanical analysis measurements showed that ANPs improve interfacial bond by limiting epoxy crosslinking, and therefore allowing epoxy to form robust bonds with cement and shale.

Abstract Seal integrity of functional oil wells and abandoned wellbores used for CO 2 subsequent storage has become of significant interest with the oil and gas leaks worldwide. This is attributed to the fact that wellbores intersecting geographical formations contain potential leakage pathways. One of the critical leakage pathways is the cement-shale interface. In this paper, we examine the efficiency of a new polymer nanocomposite repair material that can be injected for sealing micro annulus in wellbores. The bond strength and microstructure of the interface of Type G oil well cement (reference), microfine cement, Novolac epoxy incorporating Neat, 0.25%, 0.5%, and 1.0% Aluminum Nanoparticles (ANPs) with shale is investigated. Interfacial bond strength testing shows that injected microfine cement repair has considerably low bond strength, while ANPs-epoxy nanocomposites have a bond strength that is an order of magnitude higher than cement. Microscopic investigations of the interface show that micro annulus interfacial cracks with widths up to 40 μm were observed at the cement-shale interface while these cracks were absent at the cement-epoxy-shale interface. Fourier Transform Infrared and Dynamic mechanical analysis measurements showed that ANPs improve interfacial bond by limiting epoxy crosslinking, and therefore allowing epoxy to form robust bonds with cement and shale.

Abstract The rate of acid corrosion of Class H Portland cement was measured using time-lapse video over a range of temperature ( T = 30–80 °C) and pH (0–3.7), both for hydrochloric acid and carbonic acid. The process was found to be diffusion-controlled, and the dependence of the slope, S , of corrosion depth vs square root of time was obtained as a function of T and pH. The slope decreases by about a factor of 3, if the leachate is allowed to accumulate on the corroded surface; however, a flow rate as low as 4 cm/h is sufficient to flush the surface and establish an equilibrium rate. With or without accumulation of leachate, the cumulative cation mass loss is proportional to the depth of leaching, indicating that the phases extracted are very similar in both cases. The corrosion rate between 30 and 80 °C is characterized by an activation energy of 39.6 kJ/mol and a power-law dependence on acid concentration, S ∼ [H + ] 0.35 . The resulting equation describes the present results, and agrees within a factor of 2 with rates reported in the literature by other workers. On the basis of this equation, estimates are provided of the leakage rates from a reservoir that is sealed with a sound plug of cement (where escape of the acid would take millions of years) or where an annular gap extends through the caprock. In the latter case, corrosion in small annuli (≤10 μm diameter) is predicted to be concentrated near the bottom (i.e., near the reservoir-caprock boundary), so that penetration of acid into the overlying formation will take centuries.

Abstract The rate of acid corrosion of Class H Portland cement was measured using time-lapse video over a range of temperature ( T = 30–80 °C) and pH (0–3.7), both for hydrochloric acid and carbonic acid. The process was found to be diffusion-controlled, and the dependence of the slope, S , of corrosion depth vs square root of time was obtained as a function of T and pH. The slope decreases by about a factor of 3, if the leachate is allowed to accumulate on the corroded surface; however, a flow rate as low as 4 cm/h is sufficient to flush the surface and establish an equilibrium rate. With or without accumulation of leachate, the cumulative cation mass loss is proportional to the depth of leaching, indicating that the phases extracted are very similar in both cases. The corrosion rate between 30 and 80 °C is characterized by an activation energy of 39.6 kJ/mol and a power-law dependence on acid concentration, S ∼ [H + ] 0.35 . The resulting equation describes the present results, and agrees within a factor of 2 with rates reported in the literature by other workers. On the basis of this equation, estimates are provided of the leakage rates from a reservoir that is sealed with a sound plug of cement (where escape of the acid would take millions of years) or where an annular gap extends through the caprock. In the latter case, corrosion in small annuli (≤10 μm diameter) is predicted to be concentrated near the bottom (i.e., near the reservoir-caprock boundary), so that penetration of acid into the overlying formation will take centuries.