To replace part of the cushion gas in an underground gas storage by inert gas is of significant economic interest. Injection of inert gas before natural gas was performed as many as ten years ago on the aquifer storage of SAINT-CLAIR-SUR-EPTE (FRANCE). Meanwhile several projects involving the substitution of cushion gas projects involving the substitution of cushion gas already in place have been ... [Show full abstract] studied (FRANCE, USA).
Nevertheless, operators still ask important questions concerning the long-term behavior of a storage containing inert gas as a part of its cushion. These questions most often focus on the effect mixing has on the withdrawn-gas quality.
Therefore, theoretical considerations, supported by field experiments on "mutli-gas" storages operated by GdF, are examined.
In addition, a real-case study is shown, primarily because a 20 % substitution of cushion primarily because a 20 % substitution of cushion is about to begin in the SAINT-ILLIERS (FRANCE) storage. Towards this end, a two-phase, three-dimensional model was used taking into account all physical phenomena and including a precise description of the heterogeneities.
Results in the real-case are given for a long-term study extending from 1965 to 2015, and include the following :
- Color-filled contour and grids plots illustrating inert gas concentrations at different values of time.
- Future concentration histories of the operating zone and of the observation wells.
One significant result shows that the gas quality during production periods is predicted as always compatible with the BTU requirements of the storage operation : the maximum inert gas concentration withdrawn is approximately one percent at most.
The general behavior trends of the SAINT-ILLIERS storage can then be extended to other reservoirs containing inert gas as a part of their cushion. A sophisticated study is required, but the change from base gas to inert gas is manageable and produces a reliable gas quality, even for long-term operations.
A comparison between the cost of natural gas and that of inert gas clearly shows the interest in most cases of replacing part of the cushion gas by inert gas in a depleted field or an aquifer storage reservoir.
This is the motivation for studies of such operations; the first one began in the SAINT-ILLIERS (FRANCE) storage reservoir in June 1989.
The main potential drawback of the process concerns inert gas/natural gas mixing, which may influence withdrawn gas quality. Operators may be faced with questions about the long term behavior of the storage reservoir : will mixing increase over time or not ?
This paper answers that kind of question in the particular case of SAINT-ILLIERS, although many of the results can also be applied to the general case.