Subsea Re-Abandonment Strategies for High-Risk Wells to Minimize Leakage Risk in a Depleted Field for CO2 Storage

CO2 storage in a depleted field comes with the risk that is associated with wells integrity which is often defined as the ability to contain fluids with minimum to nil leakage throughout the project lifecycle. The targeted CO2 storage reservoir in offshore Malaysia has existing abandoned exploration/appraisal, and development wells. With a view of developing such CO2 storage sites, it is vital to maintain the integrity of the abandoned wells. High-risk characterized wells need to be analyzed and remedial action plan to be defined by understanding the complexity involved in restoring the integrity. This will safeguard CO2 containment for decades. Abandoned exploration/appraisal wells in the identified field are >40 years old and were not designed to withstand CO2 corrosion environment. Downhole temperature and pressure conditions may have further degraded the wellbore material strength elevating corrosion susceptibility. The reservoir simulation predicts that the CO2 plume will reach to these abandoned wells during the initial phase of total injection period. Single well was selected to assess the loss of containment through the composite structure along the wellbore and to determine the complexity in resorting the well integrity. CO2 leakage rates through all possible pathways were estimated based on numerical models and the well is characterized for its risk. For unacceptable leakage risk, the abandoned well needs to be re-entered to restore the performance of barriers. Minimum plug setting depth (MPSD) and caprock restoration considers original reservoir pressure(3450psia) anticipating the pressure buildup upon CO2 injection and is derived based on fracture gradient and maximum horizontal stress. This paper elaborates unique challenges associated with locating abandoned wells that are submerged below seabed. Top and side re-entry strategies are discussed to overcome challenges. Based on past abandonment scheme, leakage rate modeling calculates estimated leakage rate of ~460SCFD at higher differential pressure of around 3036psia at shallowest barrier and ~15SCFD for differential pressure of 1518psia at deepest barrier. Sensitivity analysis has been carried out for critical barrier parameters (cement permeability, cracks, fractures) to the containment ability and improving understanding of quality of barriers, uncertainties, and complexities for CO2 leakage risk. The paper proposes two(2) minimum plug setting depths (3550ft & 3750ft) derived based on fracture gradient and maximum horizontal stress. Perforate-wash-cement (PWC) and section milling were compared for operational efficiencies to achieve caprock restoration. for MPSD out strategic options to restore well integrity by remediating casing/cement barriers at by performing best fit abandonment technique to contain CO2 in the reservoir. Well integrity risk is assessed for existing plugged and abandoned (P&A) wells in a carbon storage site. Optimized remedial actions are proposed. Quantification of all the uncertainties are resolved that may affect long-term security of CO2 storage site.

Converting Development Wells to Observations Wells to Aid Reservoir and Overburden Monitoring for CO2 Sequestration in a Depleted Carbonate Field

Measurement, Monitoring & Verification (MMV) is crucial to ascertain both containment and conformance in Carbon Capture & Storage (CCS) projects. The magnitude of parameters to be monitored along with the technologies to be adopted could be very cost intensive and impact overall project Net Present value (NPV). To rationalize the associated costs and maximize the value propositions of existing infrastructure, the development wells in depleted field provide the opportunity to reduce the MMV cost by converting them into observation wells. However, the wells are to be analyzed for their strategic location in the reservoir, fit for purpose plug & abandonment plan and the apt technologies that can be implemented for both reservoir & overburden monitoring. Development wells in the identified depleted field are 30-40 years old and were not designed considering high CO2 concentration. In consequence, the possibility of well leakage due to accelerated corrosion channeling, cracks, along the wellbore cannot be ignored and requires careful evaluation. Rigorous process has been adopted in assessing the feasibility for converting existing producers into observation wells. Wells basis of designs disparity between the producer and the required observation well governs the selection for conversion to observation wells or plugging and abandonment. The reservoir simulation and coupled modelling predict that CO2 plume will reach all wells penetrating the storage reservoir during the initial injection phase. Out of 9 available producers, 2 strategically located wells have been evaluated for conversion based on end injection reservoir pressure of ∼3450psi. Quantitative CO2 leakage through the observation wells has been numerically computed based on all possible pathways for risk characterization. The permeable/perforated zones in these two wells are to be isolated along with the cap-rock restoration technique at deepest depth of ∼4000ft TVDSS. This will ensure the wells are safe & accessible for monitoring CO2 plume migration, CO2 leakage and well integrity by analyzing acquired DAS-VSP, DTS, DPS data and well logs. This paper elaborates unique challenges associated with identifying strategic wells for conversion to observation wells. Minimum plug setting depths, ranging from 3720-3880ft TVDSS, for abandonment of 9 development wells are derived based on fracture gradient and maximum horizontal stress. 2 observation wells require deeper plug setting depth to make caprock accessible at ∼4000ft TVDSS to be restored by utilizing either perforate-wash-cement (PWC) or section milling. Based on the subsurface illumination modelling, deployment of fiber-optics sensors in observation wells promises a cost-effective solution for monitoring CO2 plume migration and leakage by acquiring 4D DAS-VSP survey. Conversion of producers to observation wells promises cost effective MMV application for CO2 plume migration and leakage monitoring along with periodic temperature, pressure, and CO2 concentration measurement in overburden.

CO2 Leakage Rate Forecasting Using Optimized Deep Learning

Geologic CO2 Sequestration (GCS) is a promising engineering technology to reduce global greenhouse emissions. Real-time forecasting of CO2 leakage rates is an essential aspect of large-scale GCS deployment. This work introduces a data-driven, physics-featuring surrogate model based on deep-learning technique for CO2 leakage rate forecasting. The workflow for the development of data-driven, physics-featuring surrogate model includes three steps: 1) Datasets Generation: We first identify uncertainty parameters that affect the objective of interests (i.e., CO2 leakage rates). For the identified uncertainty parameters, various realizations are then generated based on Latin Hypercube Sampling (LHS). High-fidelity simulations based on a two-phase black-oil solver within MRST are performed to generate the objective functions. Datasets including inputs (i.e., the uncertainty parameters) and outputs (CO2 leakage rates) are collected. 2) Surrogate Development: In this step, a time-series surrogate model using long short-term memory (LSTM) is constructed to map the nonlinear relationship between these uncertainty parameters as inputs and CO2 leakage rates as outputs. We perform Bayesian optimization to automate the tuning of hyperparameters and network architecture instead of the traditional trial-error tuning process. 3) Uncertainty Analysis: This step aims to perform Monte Carlo (MC) simulations using the successfully trained surrogate model to explore uncertainty propagation. The sampled realizations are collected in the form of distributions from which the probabilistic forecast of percentiles, P10, P50, and P50, are evaluated. We propose a data-driven, physics-featuring surrogate model based on LSTM for CO2 leakage rate forecasting. We demonstrate its performance in terms of accuracy and efficiency by comparing it with ground-truth solutions. The proposed deep-learning workflow shows promising potential and could be readily implemented in commercial-scale GCS for real-time monitoring applications.