Seismic Risk Management Process

The seismic analysis step is where the seismic data is used to assess the seismic hazard and to understand the sources of seismicity and response to mining. The assessment of seismic hazard is generally divided into analysis techniques for short term seismic responses (short term seismic hazard) and medium- to long-term seismic hazard assessment techniques. Seismic analysis is also used to investigate seismic sources and mechanisms to increase the understanding of the rock mass response to mining.

Short-term seismic hazard focuses on two components; namely:

• quantifying elelevated seismicity and its decay after blasts and large events, and, analysis of seismic responses often refered to as “Omori analysis”; and
• the detection of change in seismicity resulting in short term elevated seismic hazard, i.e time series analysis.

6.1 Short term seismic hazard assessment

6.1.1 Short term seismic hazard assesment – basic

For the analysis of seismic responses after blasts and large events, the basic approaches include the qualitative assessment of a time series chart (e.g. magnitude-time, daily histogram, cumulative event count) and spatial plot of seismicity (e.g. plan view, 3D view). This style of analysis is done on the vast majority of sites irrespective whether more advanced approaches are also considered. Figure 33 provides an example of a seismic response to a large event. Spatial plot is shown left and an annotated time series is shown right (Vallejos and McKinnon 2010).

Figure 33 A spatial plot a) and cumulative event count time seires b) for a seismic response to a large seismic event (Vallejos and McKinnon 2010)

Additionally, the time of day of events is commonly used to evaluate to what degree seismicity is related to blasting; specifically, to estimate the rate of seismicity that occurs independent of blasting (background rate). Figure 34 provides an example of a diurnal chart (Vallejos and McKinnon 2010).

Figure 34 Diurnal chart of seismic and blast activity (Vallejos and McKinnon 2010)

6.1.2 Short term seismic hazard assesment – advanced

Advanced analysis of seismic responses incorporates seismic source parameters into the assessment and/or aims to quantify the space-time characteristics of a response. Temporal assessment of seismic responses can incorporate the location of events by defining an analysis volume. These are arbitrary volumes often defined by coordinates (Figure 35) or by only considering events within a certain distance from a blast.

Figure 35 Spatial volumes (green frames/polygons) are defined with respect to mine surveys a) or elipsoids around blast locations b) to intoduce a spatial component to temporal analysis

The most common method of incorporating seismic source parameters is to consider a rate or accumulation of a parameter value over time.

Advanced analysis also aims to quantify seismic responses by fitting a temporal model to event occurrence. This model is most commonly the Modified Omori Law (MOL) which was derived from the study of earthquake aftershocks (Omori 1894); for example, Figure 36 shows the MOL which models five responses to blasting (Plenkers et al. 2010). Temporal modelling approaches incorporate the location of events by either considering arbitrary areas of a mine, volumes around blasts, clusters determined visually, or clustered determined by an algorithm.

Figure 36 Application of the MOL to seismic responses following blasting. Seismic event rate is plotted with respect to the fitted model and background rate of seismicity

Time series analysis aims at detecting anomalous seismic behaviour as a precursor to elevated seismic hazard. These types of assessments are often used in conjunction with each other. One of the most common approaches is to examine energy-index and cumulative apparent volume, Schmidt Number and extraction induced seismic strain, over time (Figure 37 and Figure 38). Other approaches include cumulative energy, apparent stress time history, ratio of s-wave energy to p-wave energy, static stress drop, and dynamic stress drop.

Figure 37 Time history of CAV, EI and Scs for about one month before a m 1.2 at a deep gold mine in South Africa (Rebuli and van Aswegen 2013)

Figure 38 Time history of CAV, cumulative production and EISSr for about one month before a m 1.9. EISSr shows a steady drop a few days before the m 1.9 event (Rebuli and van Aswegen 2013)

6.2 Re-entry analysis

6.2.1 Re-entry analysis – basic

The topic of re-entry is covered in two sections. This section focuses on the analysis of data following blasts and large events and the derivation of rules. The application of these rules and real-time monitoring of seismicity and re-entry decisions is discussed under the Control section (10.1.2). For many sites, re-entry protocols are not based on back analysis of seismicity, but rather on choosing conservative re-entry periods. Some sites with enough flexibility in the mining schedule allow long periods before re-entry by sheduling work to be performed in areas further away from recent blasts.

6.2.2 Re-entry analysis – advanced

Re-entry analysis uses short-term assessment techniques to characterise current and/or historical seismic responses. Blanket exclusion rules are developed from historical responses and consider practical operational constraints. These rules are typically conservative by design and use the assessment of a current seismic response to refined exclusions. Cumulative event count: count of the events that occur within a fixed analysis period (e.g. six hours after) and distance (e.g. 50 m) from blast. Re-entry protocol length is based on an arbitrary event percent for historical responses; for example, protocols are lifted when 90% of events occurred for historical responses.

Seismic event rate – Return to background: re-entry protocols are lifted when the rate of seismicity returns to a previously determined background rate. This is evaluated in several ways ranging from qualitative assessment (e.g. when the slope of cumulative events looks similar to that prior to the blast) to quantitative assessment. IMS STAT is an example where such analysis is performed in real time and translated to a traffic light system for real-time assessment for re-entry times.

6.3.1 Medium/Long term hazard assessment – basic

Seismic hazard over the longer term is generally reported using simple statistics. Weekly and monthly reports usually include the total number of events recorded, as well as the number of large/significant events. Sometimes the event count over the previous reporting periods is included for reference.

The cumulative number of events is usually used to assess changes in seismic activity rate over time; higher gradients generally representing periods of elevated hazard. The areas within the mine with elevated hazard are often evaluated with event density plots. Event density isosurfaces (Figure 39) are plotted based on an event count within a certain radius. Sometimes a threshold event magnitude is applied to the event count to reduce the effect of varying seismic system sensitivity across the mine.