Theodore I. Urbancic

Space-time correlations ofb values with stress release

A total of 1503 events for a 2-month period associated with amN 2.6 rockburst is investigated for possible space-time correlations between low magnitude (−1.1 to −0.4)b values and several estimates of stress (static stress drop, apparent stress, and dynamic stress). Spatial variations of decreasingb values were found to be well correlated with increasing stress release estimates for time intervals prior to the rockburst and following the aftershock sequence. The strongest correlation tob value was with the dynamic stress drop, having correlation coefficients of 0.87 and 0.79 for the two intervals, respectively. The rockburst was found to actually occur at the intersection of the spatial coordinates corresponding to the largest gradient inb value. Based on these correlations, we conclude that the low magnitude seismicity is an indicator of the stress state within the rock mass, and can be used to study and forecast stress patterns in the vicinity of an impending major event. Time variations, however, did not show the same clear correlations and these are discussed in terms of departure from steady state conditions. Regardless, our results favour the use ofb values in a spatial, context rather than in a time analysis approach, and we consider thatb values provide valuable information regarding the changing stress conditions within the seismogenic volume.

Recent advances in seismic monitoring technology at Canadian mines

We provide an overview of the current status of seismic monitoring instrumentation employed in Canadian underground mines. Based on several case studies, we outline how passive seismic monitoring techniques are being used to evaluate fractures and stress conditions associated with ore extraction at depth. It is shown that induced microseismicity allows for the remote monitoring of active fractures, delineating modes of failure with advancing excavation fronts, and identifying variations in principal stress orientations during sequential stages of mining. Advances into the characterization of excavation zone of influence through deformation state analysis and the use of seismic hazard analysis to evaluate the potential for ground instability are also discussed.

Mining-induced microseismicity: Monitoring and applications of imaging and source mechanism techniques

The study of microseismicity in mines provides an ideal method for remote volumetric sampling of rock masses. The nature and uniqueness of microseismic monitoring is outlined in the context of acquisition hardware and software requirements. Several topics are used to highlight the potential for novel applications of microseismicity and to outline areas where further study is required. These topics reflect some of the current interest areas in seismology, namelyb values and source parameters, fault-plane solutions, modes of failure and moment tensor inversion, imaging and seismicityvelocity correlations. These studies suggest potential correlations between zones of high seismic velocity, high microseismic activity and maximal stress drops, which can be interpreted spatially to be the locations of highly stressed ground with a potential for rock bursting. Fault-plane solutions are shown to be useful in determining the slip potential of various joint sets in a rock mass. Source parameter studies and moment tensor analysis clearly show the importance of non-shear components of failure, andb values for microseismicity appear to be magnitude-limited and related to spatial rather than temporal variations in effective stress levels.

Non‐similar frequency‐magnitude distribution for M < 1 Seismicity

A series of 1503 induced seismic events ranging in magnitude from −2 to 1 as computed based on a locally calibrated seismic moment scale were located within a well-defined volume at depth. The seismicity analysis was carried out over 2 months and included events recorded before, during, and following the after-shock period of a mN 2.6 rockburst. A non-similar pattern was clearly apparent for both cumulative and non-cumulative frequency-magnitude distributions, showing a relative enhancement of seismicity between magnitudes −0.5 and 0. This supports the existence of non-similar scaling at these magnitudes. The time variation of the b-value, as computed for a limited self-similar magnitude domain (−1.1 to −0.4), did not appear to be related to the rockburst. Instead, it is suggested that the rockburst perturbed the local stress field and consequently the distribution of the very small events. No correlations were found between b and space clustering.

Microseismicity derived fault‐Planes and their relationship to focal mechanism, stress inversion, and geologic data

General stress and faulting trends have been retrieved through the analysis of mining induced microseismic events (M < 0) at two sites, related to an mN 2.6 rockburst, and an excavation at depth. A comparison of results obtained through principal component analysis (PCA) of seismicity, focal mechanism, and stress inversion, with in-situ measurements of stress and structural mapping data, show that: under stable stress conditions, the P, B, T, and stress inversion axes are consistent with in-situ measurements of stresses; stress inversion and PCA fault-planes lie within 10 to 20° of the most significant mapped features at the sites; and the PCA technique provides a robust approach for the determination of fault-planes.

Analysis of mining-induced microseismic events at Strathcona Mine, Subdury, Canada

Rockbursts and mining-induced seismic events have serious socio-economic consequences for the Canadian mining industry, as their mines are extended to greater depths. Automatic multichannel monitoring systems (Electro-Lab MP250s) are routinely, used to detect the arrival times of seismic waves radiated by mining-induced events and sensed on an array of single component transducers installed throughout a mine. These arrival times are then used to locate the events and produce maps of areas of high activity for use in mine planning and design. This approach has limitations in that, it does not allow a detailed analysis of source mechanisms, which could be extracted if whole waveform signals are recorded and analyzed.A major research project, sponsored by the Natural Sciences and Engineering Research Council of Canada (NSERC) with the collaboration of the Canadian mining industry, is aimed at enhancing existing mine seismic monitoring technology in Canada, in order to carry out more advanced processing of data to obtain fundamental scientific information on mining-induced seismic events This paper describes preliminary results from seismic monitoring experiments carried out in a hard rock nickel mine in Sudbury, Canada. Existing seismic monitoring instrumentation was enhanced with a low cost microcomputer-based whole waveform seismic acquisition system. Some of the signals recorded during this experiment indicate anisotropic wave propagation through the mine rock masses, as observed by the splitting of shear waves and the relative arrival of two shear waves polarized in directions which may be related to the structural fabric and/or state of stress in the rock mass. Analysis of compressional wave first motion shows the predominance of shear events, as indicated by focal mechanism studies and is confirmed by spectral analysis of the waveforms. The source parameters were estimated fro typical low magnitude localized microseismic events during the initial monitoring experiments. The seismic moment of these events varied between 106 N.m and 2.108 N.m. with a circular source radius of between 1 m and 2 m with an estimated stress drop of the order of 1 MPa.

Fracture coalescence as a mechanism for earthquakes: observations based on mining induced microseismicity

Abstract Microseismic events, with magnitudes less than zero, recorded within a volume at depth in an underground mine are used to evaluate failure conditions associated with a magnitude 2.9 event (rockburst). The approach incorporated several independent methods of analysis, including the principal component analysis (PCA) of seismicity, focal mechanisms, stress inversion, underground in situ stress measurements and structural mapping, and three-dimensional numerical modelling derived stresses. A good correlation was found between stress inversion and PCA derived microseismic failure planes and structural mapping, and between the in situ principal stress orientations and those obtained through stress inversion and numerical modelling. A few days prior to the large event and during the aftershock sequence, a change in microseismic failure plane orientation to a shallower dip angle than mapped underground was found to coincide with the numerical modelling derived plane of maximum slip potential. This orientation, along with measured fracture characteristics (persistence and spacing) and an increase in event clustering into the plane, suggests that the formation of the large event plane of rupture was likely related to the coalescence of mapped fractures. An observed reduction in stress release accompanying the increased tendency towards planarity prior to the large event indicates that the development of a large rupture surface occurred under an increased effective shear stress build up as part of the earthquake generation process, and, the process of coalescence can be considered as the breakage of barriers (new fracture growth) accompanied by an increase in the effective shear stress on the remaining asperities (pre-existing fractures).

Effects of Rupture Complexity and Stress Regime on Scaling Relations of Induced Microseismic Events

Source parameter scaling relations are examined for microseismic events (−2.4 ≤ M ≤ −0.3) occurring within highly and moderately stressed and fractured rock masses at Strathcona mine, Sudbury, Canada. Insight into scaling is provided by waveform complexities, calculated rupture velocities, and maximum shear stresses based on in situ and numerical modelling data. The importance of normal stress on the failure process is also considered. Our results show that a strong dependence exists between stress release and seismic moment. An observed positive scaling in excess stress release (Δσ/2 − σ a ) is consistent with the concept of overshoot. Rupture velocities ranging from 0.2 to 0.5β and waveform complexities less than 1.5 suggested that overshoot was related to healing behind a slowly advancing rupture front. Scaling in seismic efficiency paralleled that in apparent stress, implying that seismic stress release estimates arc quasi-independent of the maximum shear stress. High levels of normal stress further supported the importance of high resisting stress in the observed overshoot behaviour and its role in the failure process.

Influence of source region properties on scaling relations forM<0 events

Excavation induced seismic events with moment magnitudesM<0 are examined in an attempt to determine the role geology, excavation geometry, and stress have on scaling relations. Correlations are established based on accurate measurements of excavation geometry and methodology, stress regime, rock mass structure, local tectonics, and seismic locations. Scaling relations incorporated seismic moments and source radii obtained by spectral analysis, accounting for source, propagation, and site effects, and using Madariagas dynamic circular fault model. Observations suggest that the interaction of stresses with pre-existing fractures, fracture complexity and depth of events are the main factors influencing source characteristics and scaling behaviour. Self-similar relationships were found for events at similar depths or for weakly structured rock masses with reduced clamping stresses, whereas a non-similar behaviour was found for events with increasing depth or for heavily fractured zones under stress confinement. Additionally, the scaling behaviour for combined data sets tended to mask the non-similar trends. Overall, depth and fracture complexity, initially thought as second order effects, appear to significantly influence source characteristics of seismic events withM<0 and consequently favour a non-similar earthquake generation process.

Source parameters of mining-induced seismic events: An evaluation of homogeneous and inhomogeneous faulting models for assessing damage potential

Source parameter estimates based on the homogeneous and inhomogeneous source models have been examined for an anomalous sequence of seven mine-induced events located between 640 and 825 m depth at Strathcona mine, Ontario, and having magnitudes ranging betweenmN 0.8 and 2.7. The derived Brune static stress drops were found to be similar to those observed for natural earthquakes (∼30 bars), whereas dynamic stress drops were found to range up to 250–300 bars. Source radii derived from Madariagas model better fit documented evidence of underground damage. These values of source radii were similar to those observed for the inhomogeneous model. The displacement at the source, based on the observed attenuation relationship, was about 60 mm for three magnitude 2.7 events. This is in agreement with slip values calculated using peak velocities and assuming the asperity as a Brune source within itself (72 mm). By using Madariagas model for the asperity, the slip was over 3 times larger than observed. Peak velocity and acceleration scaling relations with magnitude were investigated by incorporating available South African data, appropriately reduced to Canadian geophysical conditions. The dynamic stress drop scaled as the square root of the seismic moment, similar to reported results in the literature for crustal earthquakes. This behavior suggests that the size of the asperities responsible for the peak ground motion, with respect to the overall source size, follow distributions that may be similar over a wide range of magnitudes. Measurements of source rupture complexity (ranging from 2 to 4) were found to agree with estimates of overall source to asperity radii, suggesting, together with the observed low rupture velocities (0.3 β to 0.6 β), that the sources were somewhat complex. Validation of source model appropriateness was achieved by direct comparison of the predicted ground motion level to observed underground damage in Creighton mine, located within the same regional stress and geological regime as Strathcona mine. Close to the source (<100 m), corresponding to relatively higher damage levels, a good agreement was found between the predicted peak particle velocities for the inhomogeneous model and velocities derived based on established geomechanical relationships. The similarity between asperity radii and the regions of the highest observed damage provided additional support for the use of the inhomogeneous source model in the assessment of damage potential.