Fracture Detection in Geothermal Wells Drilled in Volcanic Rocks

The Phlegrean Fields, close to Naples, are the site of important geothermal activity. The formations are volcanic and mostly tuffites. They are originally very tight but the geothermal alteration locally produces fractures with large increase in permeability. The lack of geological markers makes well-to-well correlation quite difficult. Thus the local detection of fractured zones in each well is very important for the evaluation of its potential. The Mofete 8 D well is a typical example. A rather complete logging program was run for fracture detection. Standard methods turned out to be disappointing. However several non-standard detectors were found to be very consistent and, later on, in excellent agreement with the analysis of cuttings. They are derived from the Dual Laterolog, the SP, the Temperature log and, most particularly, the Acoustic Waveforms from the Long Spacing Sonic. The Dual Laterolog and the Temperature Log indicate invasion by fresh and cold mud filtrate; the SP behaves as in a typical Sand-Shale sequence. Sonic Waveforms were first analyzed by a purely empirical method derived from consistent log patterns. A practical algorithm compares the total energy measured in each of the two fixed time windows located the one before, the other after the fluid arrivals. The altered zones (i.e. fractured and permeable) are clearly shown by a complete reversal of the relative energy of these two windows. A more scientific method was then applied to the Waveforms; it is based on both logging experiments and physical considerations. The energy carried by the tube wave is separated by a frequency discrimination: it correlates very well with formation alteration, thus also with the other indicators including the empirical Waveform method. It should have two advantages: – It should permit at least a semi quantitative permeability evaluation – It seems to be promising in other formations: non-volcanic geothermal wells and even hydrocarbon-bearing rocks. 10 refs., 6 figs