Site engineers are often keen to develop site-specific charts for everything from underground stope design to open pit blast design. But, how much data is required to ensure that the new site-specific chart is as good, or better than a generic design chart? In this stope design chart example, a boot-strapping statistical method was developed to evaluate the effect of data base size upon the accuracy of the site-specific chart.
Guidelines were developed to determine the number and type of case studies required to develop a site-specific stability chart for stope design. By analysing trends in logistic regression parameter variance, it was possible to determine the type and number of case studies to develop a site-specific stope stability chart. The analysis indicated that a reliable stable-failure boundary requires at least 150 case histories, of which a minimum of 10 percent should be unstable stope surfaces. Marginal site-specific effects were observed for the operating conditions captured within the database. It has been concluded that the apparent site-specific effects contained in previous literature were attributable to operating conditions inadequately represented in the database. Such operating conditions could induce erroneous stability predictions at any site, and therefore, are not truly site-specific.