THE TOPOGRAPHIC APPROACH TO LOCATING HIGH-YIELD WELLS IN CRYSTALLINE ROCKS - DOES IT WORK

Despite convincing evidence that fracture trace mapping can be extremely useful in locating sites for high-yield wells, well drillers persist in using the "lay of the land" to select well sites. This paper examines: (1) the value of topographic variables in predicting the productivity of wells in crystalline rocks; and (2) whether topographic data for well sites can increase the accuracy of well yield predictions based on fracture trace data alone. Topographic and fracture trace data for 29 wells in the Georgia Piedmont and Blue Ridge area were studied. Regression analysis revealed that the LeGrand (1967) point value (TOPO) explained only 0.3% of the variation in the productivity of the 29 wells. More explanation was afforded by the less subjective topographic variables, distance from the well to the nearest valley or draw axis (DISTV), and the difference in elevation between the nearest valley or draw axis and the well site (ELEV). Together these variables still explained only 15.3% of the variation in well productivity. Brook (1988) showed that a single fracture trace variable, distance from the well to the nearest fracture trace intersection (DISTIN), explained 59% of the variation in the productivity of the 29 wells and that the two variables well depth and DISTIN explained 81% of the variation in well productivity. Addition of the topographic variables TOPO, DISTV, and ELEV to the models of well productivity developed by Brook (1988) did not greatly improve the explanation. None of the added variables were significant at the 0.05 confidence level. Results, therefore, indicate that well productivity is only poorly related to topographic characteristics, perhaps explaining why so many of the wells drilled in crystalline rocks using the "lay of the land" have a low yield or are dry regardless of their topographic locations.