Long-term Sealing of CAES Caverns: Stress-Seepage Coupling

Introduction

Compressed air energy storage (CAES) systems rely on underground caverns, such as salt caverns or lined rock caverns, to store high-pressure air. The long-term sealing performance of these storage cavities is critical to maintaining operational efficiency and safety over a 20-year design life. Leakage not only reduces energy recovery but also poses economic and environmental risks. This case study examines a typical CAES project in a salt dome, focusing on the methodologies used to evaluate and ensure long-term gas tightness.

The cavern under analysis is an artificial salt cavern with a volume of 300,000 m³, operating at pressures between 50 and 80 bar. The rock salt formation exhibits viscoplastic behavior, while cyclic pressure loading induces stress redistribution and potential damage. To address these challenges, a coupled stress-seepage analysis was performed to predict the evolution of permeability and leakage paths.

Stress-Seepage Coupling Analysis

A finite element model integrating mechanical deformation and fluid flow was developed to simulate the coupled response of the cavern and surrounding rock mass. The stress-dependent permeability of salt is modeled using empirical relationships calibrated from core laboratory tests. The analysis considers creep deformation, cyclic pressure loading, and the resultant changes in porosity and microcrack density. Results indicate that near-field permeability increases by up to two orders of magnitude near the cavern wall due to dilatancy, but this zone is limited to the first 2-3 m of rock.

Long-term simulations over 20 years show that creep tends to close microcracks after initial spalling, reducing permeability in the far field. However, a residual damaged zone persists within 1 m of the wall. The coupling analysis identified that the maximum leakage rate through the intact rock mass remains below 0.1% of stored air per day, meeting the project's target. A key finding is that the stress path and cyclic loading magnitude significantly influence damage accumulation, highlighting the need for accurate in-situ stress measurements.

Lining Material Selection

To mitigate leakage through the damaged zone, an engineered lining system is applied. For this salt cavern, a two-layer liner was selected: an inner steel liner (10 mm thick) for primary gas containment, and an outer concrete plug with a polymer membrane to seal against the rock. The steel liner is designed to withstand fatigue from daily pressure cycles, with a yield strength of 450 MPa. The concrete plug (C40 grade) provides structural support and distributes stress, while the polymer membrane (ethylene propylene diene monomer, EPDM) offers flexibility and chemical resistance to brine.

Material qualification involved accelerated aging tests at 60°C and 100 bar for 1,000 hours, simulating 20 years of operation. No significant degradation was observed in the EPDM membrane, and the steel liner showed acceptable fatigue life. A key consideration was the interface bond between the concrete and rock salt, addressed by roughening the cavern wall and using a high-alkalinity grout to prevent dissolution. This lining system was estimated to reduce leakage by a factor of 10 compared to an unlined cavern.

Leakage Monitoring Techniques

Continuous leakage monitoring is essential to verify the sealing performance over the 20-year life. The project implemented a multi-tier monitoring system: (1) Mass balance method using flow meters at the compressor and turbine to track air inventory; (2) Pressure decay test during weekly shut-in periods, corrected for thermal effects; (3) Distributed fiber optic sensors (DFOS) embedded in the concrete plug to detect strain and temperature anomalies caused by gas leaks; and (4) Chemical tracers (perfluorocarbons) injected into the cavern, with downstream sampling in surrounding wells to detect even minor crossflow.

Initial commissioning data showed a leakage rate of 0.03% per day, well below the design limit of 0.1%. Over two years of operation, the monitored leakage remained stable, with slight decreases attributed to creep-induced sealing. The DFOS system successfully identified a minor thermal anomaly near a construction joint, which was remediated by injecting a sealant. This case demonstrates that combining multiple monitoring methods provides redundancy and early warning, ensuring that any deterioration can be addressed before compromising the 20-year seal.

Case Study: 20-Year Performance Projection

Using the calibrated coupled model, a forward simulation predicted the sealing evolution for two operating scenarios: base case (daily 12-hour cycles) and worst case (extended shutdown with pressure drop). For the base case, the leakage rate stays below 0.05% per day for 20 years, as the liner degrades only gradually due to fatigue. The worst case (6-month shutdown) caused a temporary increase in leakage to 0.12% during re-pressurization, but after several cycles it re-stabilized. The analysis recommended installing additional fiber-optic cables near the liner interface to monitor early crack initiation.

Real-world validation came from a similar CAES plant in operation for 15 years, where periodic pressure tests confirmed the sealing integrity. The plant's operator reported that the leakage rate has remained under 0.02% per day, with no major liner failure. This aligns with the model predictions, proving that careful design of stress-seepage coupling, material selection, and monitoring can achieve the required 20-year air tightness for underground CAES storage.

Conclusion

The integrated approach combining stress-seepage coupling analysis, robust lining materials, and advanced monitoring techniques enables reliable long-term sealing of CAES caverns. The case study demonstrates that for salt caverns, a steel liner with concrete and polymer membrane offers cost-effective gas containment, while distributed fiber optics and tracer methods provide real-time leakage detection. With proper design and maintenance, the 20-year airtightness requirement is achievable, supporting the role of CAES as a viable long-duration energy storage solution.

Future work should focus on developing self-healing liners and automated monitoring systems to further reduce leakage risk. This methodology can also be adapted to lined rock caverns in other geological formations, expanding the applicability of compressed air energy storage.