In this survey we identified some measuring metrics to study the average noise power variations in typical outdoor power substations. Power substations generally have metallic structures and despite the insulation considerations have high electric fields. The physical size of a substation does not allow a complete controlled experiment. We arranged a setup plan to study the noise floor variation in a few substation yards in residential, industrial/isolated (sparsely populated) subdivisions, collected the empirical data sets and compared the achieved results with the known noise constituents that were cited in the literature. A two-weeks measuring window was chosen to control for any possible factors that might confound results. The analysis suggests that the noise floor variation (and hence the link quality) has an underlying dominant semi-deterministic time-dependent constituent in addition to the classical random distribution. Although it is of no surprise that the semi-deterministic component is associated with the location of the substation yard (e.g. residential or industrial), its dynamic range is significant. The methodology, which is adopted in this study, has applications in the analysis of Fixed Wireless Access (FWA), Local Multipoint Distribution Service (LMDS).