Locating Groundwater at Ta-Kang-Shan and Ping-Ting Sites by Geoelectric Methods

A great population growth and rapid industrial development i n Taiwan have placed ever increasing demands on the water supply. In fact, since the yield of kno,vn aquifers can no longer meet the current needs, new reservoirs must be found for emergency use, especially i n those areas with a short of water . Traditionally, the selection of well sites in Taiw·an was based on lin­ eaments shown i n aerial photographs and on the topographic conditions of a site. As much as this was successful i n the past, the number of potential aquifers today continues to decrease rapidly. Conventional groundwater ex­ ploration methods do not give satisfactory results; and locating an aquifer by geoelectric represents a significant challenge. Ho\\''ever, the combined geo­ electric method and conventional methods provides a better approach for site evaluation, especially in those areas of water shortage. In this paper, it will be sh0\\70 the way by \\,hich geoelectric methods are applied to locate aquifers in poor water storage areas and highlight the propriety of such methods used during this survey. In so doing, useful infor­ mation as to the aquifer characterization will be acquired. Two selected sites in this study have different geologic settings; one in a volcanic area, the other in a hilly area. All the geoelectric features obtained i n these areas not only delineate the water bearing zones clearly but also indicate the depth extent accurately.


INTRODUCTION
Although geoelectric techniques are popular and successful in locating aquifers around the world (Kelly, 1977;Mazac et al., 1990;Yang, 1992), the chances of their success depend on the features of the aquifer. Various factors affecting the recognition of aquifers with 1 Institute of Applied Geology, National Central University, Chungli, Taiwan, R.O.C. TAO, Vol. 7, No.l, A1arch 1996 geoelectric methods include the size of the reservoir, its physical parameters and the contrast in the conductivity between the aquifer and the bedrock. This research deals with the detection of aquifers located in a volcanic and a hilly area. The inherent difficulties in site evaluation in those water shortage areas are overcome by combined use of different geoelectric techniques, namely, the direct current (DC) resistivity method and/or the terrain conductivity method (TCM). A good geophysical interpretation is achieved prior to the drilling of the wells. The results provide us with a profound under standing of the structure of the aquifer zones.

CASE STUDIES
. DC resistivity surveys were carried out at the two selected sites of Ta-Kang-Shan, and Ping-Ting in Taiwan. In order to get a better interpretaton of the field data, TCM measurements were also made during DC soundings at Ta-Kang-Shan for comparsion. The field work and final results at each site are described in the following.

Ta-Kang-Shan Site
Location and geologic setting: Located in the east of Ta -Kang-Shan in southern Taiwan, the Niu-Chou-Pu Chi River to the west and Road 10 to the north ( Figure 1) serve as boundaries. The survey area is about 250,000 square meters. With the whole area covered by the alluvium of Niu-Chou-Pu Chi with a grey mudstone of poor pemeability, intercalated with yello\\l brown siltstones. Although the topography of the study area is mostly of low relief of an a\1erage altitude of 43 m, the difference in the altitude between the ground surface and the valley of Niu-Chou-Pu Chi ranges from I 5 to 20 m. According to past records, the study area belongs to a water shortage area. The purpose of field and geophysical surveys is to map the aquifers for future drilling .   Geoelectric surveys: TCM measurements -were made using a Geonics EM-34 at 30 locations ( Figure 1 ). This loop-loop EM system was operated at intercoil seperations of 10, 20 and 30 m for both vertica] and horizontal parallel dipo1e geometries. The maximum depth of penetration for each intercoil separation is shown in Table 1 The DC resistivity Schlumberger vertical electric. sounding (YES) with a maximum half-spacing of 200 m was carried out at 16 positions ( Figure 1 ). An OYO ES-G2 was used in this sur\1ey. In order to get better interpretation of the lithology of the surface from DC data, the resistivity measurements were made at known outcrops, The final resistivity spectrum is shown in Table 2.

195-196
Qualitati\1e data interpretation: Appa1�ent resistivity data for both the TCM and the DC were depicted and contoured as pseudoresistivity maps. Those points used to describe the apparent resistivity values were plotted at each station at a pseudodepth equal to array spacing (the loop distance in the TCM surveys or the half electrode spacing (AB/2) . in the DC surveys). The apparent resistivity maps for both the TCM data ( Figure 2) and the DC data ( Figure 3) indicate spatial variation (both horizontal and vertical directions) of the overburden resistivities. An increasing coil spacing in the TCM data and/or larger value of AB/2 in the · DC data will reflect deep resistivity features. These figures show that : ( 1) The apparent resistivity of deeper layers have less lateral variation at each depth range.
However, the values corresponding to each depth tend to decrease with an increase in depth (see Figures   ..; .. . ,  Ta-Kang-Shan Chieh-Hou Ya. ng & Ll1n-Tao Tong 7 (2) Figure 3(c) shows no significant variation in apparent resistivity. The average of the apparent resistivity values in this map was 10 ohm-m. The low uniformly distributed resistivity values at this depth imply' that the deep laye.rs are composed of unifo1m mudstone. Quantiti\1e data interpre. tation: The one-dimensional i11version scheme de. veloped pre vious by Zodhy (1989) and Jupp and Vozoff ( 1975) is used in this study to convert the sounding data to geoelectric structure. The geoelectric la)'ers finally can be grouped into 5 layers, namely A, B, C, D and E respectivity. Layer A has a resistivity ranging from 9 to 25 ohm-m which corresponds to the brown muddy top soil. Layer B has a resistivity varying fr om 7 to 14 ohm-rn \\1hich is believed to be mud interlayered b)' tir1e sand layers. I_Jayer C is the highest resistivity zone ( 17 to 31 ohm-m) which corresponds to a mud )ayer intercalated fine sand and pebble, or alternated of sandy mudstone and sandstone. Layer D corresponds to the mudstone with intercalating sandy mudstone that has a resistiv·ity ranging from 8 to 12 ohm-m. The last E layer is composed of the lo\vest resisti\1ity' (2. 5 to 7.1 oh1n-m) bedrock consisting ot' homogeneous mudstone. Based on this resistivity c.lassificatior1, ]ayers A and B can be interpreted as being the alluvial soil of the river with their cotnbined thickness abot1t 2 to 5 m. Local groundwater may be found in this laye1·. The unconformity bet\\1een the la)'ers B and C inay coincide with the erosion surface of an anicent river.

Ping-Ting District
The Ping-Ting District, site located in the Hsiao-Ping-Ting area, Tanshui, in northern Taiwan, covers an area of 33.8 hectares. Because the Nan Kuo corporation was planning to develop a new community in this area, a groundwater survey was carried out to find new aquifers to meet the increasing water demands in the future. The exploration included two steps: a DC survey to dete. rmine the locations for test \\'ells and well loggings and pumping tests at each test well.
Geologic setting: Located in an andesitic area, the site has a maximum variation of altitude of about 125 m. The whole area is covered by light yellow to dark grey \\leathered volcanic detritus with outcrops of andesite either on the site or in the surroundings. In the past, no groundwater was found in 27 boreholes with depths ranging from I 0 to 35 m. The lithofacies from the logs show that the overburden consists of weathered red-brown muddy fine sand with detritus, hard red-brown andesite and grey tuft· .
DC sounding: 15 vertical electric soundings (VES) with a Schlumberger array were carried out in the. site. (Figure 5). The maximt1m half spread spacing \\/as 400 m. Sounding locations numbered 8 to 12 were sequentially distant from the site and became closer to the known spring or wells. The purpose of selecting those locations was to compare the DC results obtained from the site with the. kno\\1n groundwater sources. The OYO ES-G2 was used to collect the sounding data.
Data interpretation: All the VES curves obtained from the sounding locations were K type e. xcept those-at location number 8. Those curve. s are interpreted to be three layers.
sistivity ( 1000 to 1900 ohm-m) corresponds to fresh andesities with a thickne. ss varies from 140 to 181 m. The last layer (C layer) has a resistivity \'aring from 100 to 260 ohm-m at a depth from 170 to 190 m, and it is composed of loose volcanic breccia. This layer may represent the main body· of the aquit' er. Figure 6 shows the interpreted section along VES locations 4, 3, 15 and 2. Comparing the geoelectric sections and the well int· or1nation shown in Figure 7(a) with Figure 7 (b) and also the geologic features of the springs located outside the site indicate that the whole area has the same aquifer. The springs probably· represent the. \\restern extension of the aquifer. From this point of view: the estimated depth of the aquifer on the site is about 160 to 180 m, leading to the conclusion that the proper drilling locations are close to the sounding position number 2.
We. II logging: Two wells named NK-1 and NK-2 were drilled ( Figure 5) to a depth of 250 m and 165 m, respectively. The lithological sequenc.es and logs (Figures 7 (a) and 7 (b)) do match the geoelectric model obtained from the DC data. An obvious drop in temperature is shown 1n the temperture logs at depth of 92 m for NK-1 well and at 1.62 m for the NK-2 well. The presence of groundwater flow Jo\vers the temperture g1·adient at tl1ese depths. The temperture gradients of both wells show the trend of increasing with depth. Caliper 1neasurements at the NK-1 well indicate that the first 31 m has no obvious variation due to casing. An abrupt change in the logging curve from the depth of 30 to 50 m represents a potential well collapse and may indicate that highly fractured layers are present at this depth range. From the depth of· 50 to 146.8 1n . local v<:1riations sho\;yn i11 the curves a1�e p1�esence of massive andesites. No obvious variation in the caliper cttrve was obser\1ed at a depth greater than 146.8 m meaning that the cementation o· t· the layers in this range is quite dit· t· erent from the overburden.
The gamma logs show that there is a distinct boundary in each well, namely, at 169 in for the NK-1 well and at I 05.3 m for the NK-· 2 well. The intensity of the gamma ray does increase \Vith the depth in both \\,ells.