SPE Production & Operations
Volume 22, Number 3, August 2007, pp. 368-376

SPE-99520-PA

Gas Storage and Operations in Single-Bedded Salt Caverns: Stability Analyses

View full textPDF ( 2,798 KB )

DOI  More information 10.2118/99520-PA http://dx.doi.org/10.2118/99520-PA

Citation

  • Han, G., Bruno, M., Lao, K., Young, J. and Dorfmann, L. 2007. Gas Storage and Operations in Single-Bedded Salt Caverns: Stability Analyses. SPE Prod & Oper  22 (3): 368-376. SPE-99520-PA.

Discipline Categories

  • 4.3 Underground Gas Storage

Summary

Bedded salt formations are located throughout the U.S., providing valuable storage capacity for natural gas and other hydrocarbons. To increase gas-storage capabilities and provide operators with improved geotechnical design and operating guidelines for these caverns, stability analyses of single-bedded salt caverns have been completed and are described in this paper. This work is a part of integrated efforts initiated and sponsored by the U.S. Department of Energy (DOE), the Gas Technology Institute (GTI), and Pipeline Research Council International, Inc.

Numerical geomechanical models have been developed to investigate single-cavern deformation and bedding-plane slip for a variety of cavern configurations. A viscoplastic salt model has been developed based on an empirical creep law developed for the Waste Isolation Pilot Plant (WIPP) Program and combined with a Drucker-Prager model for damage and failure. The nonsalt materials are described with either a traditional Mohr-Coulomb model, or an elastic model, depending on layer properties. 

A baseline model with specified geometric dimensions is first selected and subjected to 1-year cyclic pressure operations. The amount of damage around the cavern wall and roof is evaluated and used as a comparison in the study. Then, the operations are extended to 15 years to study cavern stability for long-term gas storage and operations. In addition to the baseline model, parametric studies have been performed to investigate cavern damage as a function of salt roof thickness, overburden stiffness, interface properties, and cavern geometries. Each cavern simulation includes 1 year of pressure cycling with a minimum, mean, and maximum cavern pressure of 6.1 MPa (884.5 psi), 8.8 MPa (1,276 psi) and 14.9 MPa (2,160.5 psi), respectively. Different operation conditions (e.g., hydrostatic, cyclic, and direct-pressure drawdown) are compared and evaluated in terms of cavern stability. 

These analyses can be a basis to selecting the best salt cavern candidate for gas storage and operations as well as helping to assess critical cavern design parameters for thin-bedded salt formations.

Introduction

The DOE forecasts global natural gas consumption increasing nearly 70% between 2002 and 2025, with the strongest growth coming from Asia, Eastern Europe, and the former Soviet Union (Donnelly 2005). Gas demand in mature economies, such as North America and Europe, also remains strong. The increase in consumption and strong demand pose significant challenges to gas reservoir development, gas transportation, and storage. Currently, there are three main types of natural underground storage facilities for natural gas: depleted reservoirs, aquifers, and salt caverns. Salt caverns are typically much smaller than depleted reservoirs and aquifers and, therefore, hold much less gas volume. In 2001, the U.S. total gas storage capacity was approximately 2.38´1011 m3 (8.4 Tcf), 82% of which was stored in depleted gas reservoirs, 15 percent in aquifers, and 3% in salt caverns (EIA 2001). Despite the fact that depleted reservoirs are the dominant storage method for natural gas, it is estimated that salt caverns represent 15% of daily deliverability (a measure of the amount of gas that can be withdrawn from a storage facility) in the U.S. (EIA 2001). Salt-cavern storage holds the advantages related to higher deliverability—lower-cushion (or base) gas requirement, less development cost, efficienty to initiate the gas flow and refill.

View full textPDF ( 2,798 KB )

History

  • Original manuscript received: 1 February 2006
  • Meeting paper published: 15 May 2006
  • Revised manuscript received: 3 October 2006
  • Manuscript approved: 6 October 2006
  • Version of record: 20 August 2007