Unravelling radon emission from rock damage mechanisms: new laboratory insights

""Positive radon anomalies are commonly used as a tool to predict dynamic failure in the crust, as stated by the widely-used dilatancy model for earthquake prediction. The model foresees that the formation and propagation of cracks prior to rupture will serve to create new surface area and hence increase radon emanation. However, this simplified picture is not only occasionally contradicted by negative radon anomalies, but sometimes negative anomalies are measured at the seismic source, whilst contemporaneous positive anomalies are recorded at monitoring stations located hundreds of kilometres away. Mysteriously, alternating increases and decreases or no significant variations have also been recorded. Hence, despite early promise, radon emanation does not appear to be a very compelling warning signal and many questions remain unresolved. Although numerous studies have investigated field radon emanation from rocks under natural stress conditions, only few have attempted to experimentally relate rock deformation with radon change. . Here we present, for the first time to our knowledge, an original experimental set-up consisting of a new accumulation chamber made of a drypack material, a polyester-aluminum-polyethylene bag, containing the rock specimen and connected in a closed-loop configuration to the RAD 7. The drypack-chamber is loaded under a uniaxial press, without being damaged during rock deformation. Radon gas concentration emitted from the samples is simultaneously and continuously monitored by the system; therefore, the originality of the experiments consists in a real time monitoring of radon emission changes during the whole process of rock deformation. Additionally, the drypack-chamber is surrounded by a heating belt that allows us to investigate the effect of high temperature conditions (up to 90 °C) on radon emission. It needs to be stressed that the radonometer is alpha-spectrometry based and is able to measure simultaneously 222Rn (radon) and 220Rn (thoron) activity concentrations. The great advantage of determining thoron is the possibility of recording any activity change in the closed system within the length of a single run cycle (30 minutes), due to the short half-life of 220Rn (55.6 seconds) that reaches any new equilibrium conditions in about 5 minutes.. Results demonstrate that radon exhalation from the rock specimen drastically increases with increasing the experimental temperature. This finding allows us to discriminate low variations in radon emissions when rock samples are uniaxially loaded under the press. Uniaxial compressive tests performed on highly porous (47 %) volcanic tuff from Vico Volcanic District (central Italy) shows that the radon emission decreases with increasing load. Such a variation is proportional to a reduction of porosity associated with the compaction of the tuff sample. When the specimen fails, a drastic increase of emission is verified.. Moreover, long-term experiments (in the order of weeks) performed by maintaining a constant load on the tuff specimen show that the deformation increases with time. The radon signal coherently continues to decrease up to the complete closure of pores. At that point, a further increase of load does not induce any additional compaction and any radon emission change, with a constant exhalation up to the failure. This has strong implications in the field of radon anomalies prior to earthquakes, showing that the seismic event can be preceded by no significant radon anomalies, as already documented in literature. . Our experimental investigation sheds light on several apparently contradictory signals recorded by radon monitoring stations near active faults and volcanoes. Results demonstrate that emanation rate is governed by the prevalent deformation mechanism. In terms of seismic and volcanic hazard, the deformation mechanism of rocks under stresses should be carefully considered to properly interpret data from geochemical field monitoring. . . ""