Transient radon release from low- and high- porosity rocks under upper crustal environment

Abstract:

Natural gaseous emissions near large crustal faults and their potential relations with deformation and earthquake have always aroused significant interest. Radon-222, a radioactive gas (half-life of 3.8 days) naturally produced in the Earth’s crust, is used as a tracer of fluid migration in fault zones and is sometimes considered as a potential earthquake precursor. However, this matter is still highly debated. In this study, dedicated rock deformation experiments are conducted on various low-porosity granites and high-porosity volcanic tuffs under controlled upper crustal conditions. The sample is placed in the tri-axial cell under given confining and pore pressures. In practice, pore pressure, controlled by a flux of inert argon, expels the poral gas of the sample, whose radon concentration is continuously measured. First, when pore pressure is applied, radon is expelled from the connected pores and cracks of the sample, leading to a transient peak in radon concentration. This case may correspond to an initially cemented fault conduit which becomes suddenly unsealed. Second, radon concentration increases with confining pressure, but starts to decrease above a given value of confining pressure. This is related to the closing of pores and cracks that may be considered when a fault conduit becomes progressively sealed. Third, when axial stress increases, hence approaching the macroscopic rupture of the sample, initially isolated micro-cracks interconnect. In some cases, this may lead to transient peaks of radon concentration. Finally, after the macroscopic rupture, a significant transient radon peak is systematically detected. The time between the rupture and the transient radon signal mainly depends on the permeability of the sample at the rupture point. This study demonstrates that advective conditions are definitely required to efficiently observe potential transient radon signals, and that, even so, precursory radon signals cannot be systematically observed. Radon appears also as a tracer of what is effectively changing in a rock sample during deformation and in the general context of dynamic permeability changes.