Three-Dimensional Magnetotelluric (MT) Modeling of the Northern Negros Geothermal Project, Central Philippines
The latest magnetotelluric (MT) survey was conducted in the Northern Negros Geothermal Project (NNGP), which is one of the geothermal fields being developed in the Philippines, from December, 2010 to April, 2011. 66 new MT soundings were added to the previous MT dataset. The new stations were located mainly in the southeastern and southern regions to define the extent of drilled high-temperature resource in these areas. Phase tensor analysis show that the MT data in general is only 1-D in the short period range of <1 s and becomes 3-D at longer periods. 1-D, 2-D and 3-D modeling were performed on the MT dataset after stripping it for distortion based on the phase tensor and correcting for static shifts using Transient Electromagnetic (TEM) data. The resistivity structure from all models show three main layers: a >100 ohm-m resistive top layer, a middle <10 ohm-m conductive layer and a >20 ohm-m moderately resistive bottom layer. The highly resistive top layer is associated with the relatively fresh volcanic deposits of the Canlaon Volcanics (CnV). Correlating the 3-D resistivity structure with subsurface data from the drilled wells shows that the thick conductive layer overlaps with the low-temperature alteration minerals such as smectite while the moderately resistive bottom layer coincides with the high-temperature alteration minerals like illite and epidote. These observations are also consistent with the measured well temperatures wherein the elevated temperatures drilled beneath the Pataan sector coincide with the shallow occurence or doming portion of the bottom resistive layer. Tracing the shallow occurrence of the bottom resistive layer revealed a northeast extension to the drilled resource beneath Pataan. The delineated resource area in Pataan is about 3 to 7 km². Other possible high-temperature areas are located within the Upper Hagdan and Hardin Sang Balo sectors. However, resolution of the resistivity structure is not well pronounced in these areas due to limited data coverage.