Abstract
In different subsurface energy technologies, traffic light systems (TLSs) have been implemented for limiting the strength of induced seismicity. Despite their widespread application, fundamental assumptions regarding the controllability of induced seismicity were usually not reviewed. This is the focus of the current article, in which we discuss limitations of a TLS in the context of seismicity induced by fluid injection and gas production.
Most existing TLSs are based on a critical earthquake magnitude or vibration level that should be prevented to occur. Operational measures are defined to be taken after an induced earthquake exceeds predefined threshold values. This concept rests on the tacit assumptions that induced earthquakes of a critical strength announce themselves by precursory events of smaller strength and that future earthquakes of a critical strength can be prevented by modifying or stopping subsurface operations. We investigate to what extent these assumptions can be justified by studying observation data from a dozen fluid‐injection operations in geothermal reservoirs as well as from gas production in 26 gas fields in The Netherlands.
In our case studies, whereas fluid injection–induced seismicity generally starts at a low‐magnitude level and exhibits a gradual temporal increase of the maximum earthquake magnitude with the duration of the injection, the largest magnitude event frequently occurs postinjection. The temporal evolution of the seismicity induced by gas production in The Netherlands is less systematic. In some gas fields, seismicity started at a comparatively large‐magnitude level (ML ≥ 2.7) without detectable precursors. A correlation between seismic activity and the gas production rate is only observed in the largest gas field.
Our findings indicate that the precision to what an earthquake of a given strength can be prevented by a TLS has more limitations than typically assumed.
Most existing TLSs are based on a critical earthquake magnitude or vibration level that should be prevented to occur. Operational measures are defined to be taken after an induced earthquake exceeds predefined threshold values. This concept rests on the tacit assumptions that induced earthquakes of a critical strength announce themselves by precursory events of smaller strength and that future earthquakes of a critical strength can be prevented by modifying or stopping subsurface operations. We investigate to what extent these assumptions can be justified by studying observation data from a dozen fluid‐injection operations in geothermal reservoirs as well as from gas production in 26 gas fields in The Netherlands.
In our case studies, whereas fluid injection–induced seismicity generally starts at a low‐magnitude level and exhibits a gradual temporal increase of the maximum earthquake magnitude with the duration of the injection, the largest magnitude event frequently occurs postinjection. The temporal evolution of the seismicity induced by gas production in The Netherlands is less systematic. In some gas fields, seismicity started at a comparatively large‐magnitude level (ML ≥ 2.7) without detectable precursors. A correlation between seismic activity and the gas production rate is only observed in the largest gas field.
Our findings indicate that the precision to what an earthquake of a given strength can be prevented by a TLS has more limitations than typically assumed.
| Original language | English |
|---|---|
| Pages (from-to) | 1145-1154 |
| Number of pages | 10 |
| Journal | Seismological Research Letters |
| Volume | 90 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 2019 |
| Externally published | Yes |