EVIDENCE OF RESERVOIR INDUCED MACROSEISMICITY
Abstract
The published evidence of reservoir-induced macroseismicity is critically reviewed. The evidence is partitioned into three types: (1) evidence of a postimpoundment increase in seismicity; (2) correlation evidence that is composed of a plot of a reservoir variable and a seismicity variable and, in most cases, an auxiliary variable, time; (3) evidence based on the slope of the magnitude-frequency relationship (b value evidence). The basic source for each of these types of evidence is a suitably edited catalog of earthquake occurrences. Recommendations are given regarding catalog preparation and editing. The types of evidence are examined to recognize the effects of common methods of data acquisition and treatment. Postimpoundment change in seismicity can be assessed using certain recommended statistical tests. Recommendations are made regarding construction and interpretation of correlation evidence. The interpretation of correlation evidence is subjective because the assumptions required to construct and interpret the evidence cannot be verified. Evidence based on the b value is shown to be useless as a discriminant to identify induced seismicity from normal seismicity. Eight cases of reservoir induced macroseismicity are reviewed. These eight cases are: (1) Hoover/Lake Mead, (2) Kariba, (3) Kremasta, (4) Koyna, (5) Kurobe, (6) Manic 3, (7) Hsinfengkiang, and (8) Nurek. Three cases (Hoover, Manic 3, Nurek) appear to be well-documented cases of induced seismicity. Two cases (Kurobe, Kremasta) are located in seismically active areas and the seismicity attributed to the reservoir is consistent with the natural local seismicity. These two cases do not appear to be cases of induced seismicity. The three remaining cases (Hsinfengkiang, Kariba, Koyna) present insufficient information to support a conclusion.
Degree
Ph.D.
Subject Area
Civil engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.