Electrical Resistivity Tomography, VES and Magnetic Surveys for Dam Site Characterization, 2 Wukro, Northern Ethiopia

Accepted Manuscript Electrical Resistivity Tomography, VES and Magnetic Surveys for Dam Site Characterization, Wukro, Northern Ethiopia Tigsitu Haile, Solomun Atsbaha PII: DOI: Reference: S1464-343X(14)00077-6 http://dx.doi.org/10.1016/j.jafrearsci.2014.03.023 AES 2007 To appear in: African Earth Sciences Received Date: Revised Date: Accepted Date: 27 November 2013 24 March 2014 25 March 2014 Please cite this article as: Haile, T., Atsbaha, S., Electrical Resistivity Tomography, VES and Magnetic Surveys for Dam Site Characterization, Wukro, Northern Ethiopia, African Earth Sciences (2014), doi: http://dx.doi.org/ 10.1016/j.jafrearsci.2014.03.023 This is a PDF file of an unedited manuscript that has been accepted for publication As a service to our customers we are providing this early version of the manuscript The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain Electrical Resistivity Tomography, VES and Magnetic Surveys for Dam Site Characterization, Wukro, Northern Ethiopia a Tigsitu Haile * and Solomun Atsbahab School of Earth Sciences, College of Natural Science, Addis Ababa University, P O Box 1176, Adds Ababa, Ethiopia; bSamara University, P O Box 132, Samara, Afar, Ethiopia *corresponding author: *htigistu@yahoo.com, 00251-111-239462 a Abstract Geophysical surveys involving the techniques of electrical resistivity imaging, electrical sounding and 10 magnetics were employed to characterize the ground conditions at a proposed dam site at Hizaeti-Afras, 11 Wukro, North Ethiopia The techniques were utilized to map the depth to the competent formations, their 12 relative suitability for foundation work and the presence and extent of weak zones in the subsurface The 13 work has mapped the different lithologic units of the subsurface and determined the depth to the basement 14 rocks in the area Through correlation of the inverse model resistivity sections of the imaging surveys, the 15 geoelectric section of the sounding survey and the magnetic profile plots with available borehole lithologic 16 logs, it is shown that the results very well supplement the geotechnical point data in addition to providing a 17 wider coverage in mapping areas of weak ground that could otherwise be missed with widely spaced 18 borehole information The combined results of the survey show the proposed dam axis to be unsuitable 19 The power of the electrical resistivity imaging technique and its potential to map the shallow subsurface 20 with adequate resolution are illustrated The result is a strong suggestion that geophysical techniques can be 21 used to assist and extrapolate borehole geotechnical data especially when large area is to be used for 22 development of large infrastructure 23 24 Key words: Dam site selection, Earth fill dam, electrical resistivity imaging, magnetic anomaly, Ethiopia 25 26 27 28 29 30 31 Introduction 32 To increase agricultural productivity, make the agricultural economy more sustainable and improve the 33 livelihood of the population, the Tigray Regional State in northern Ethiopia had a targeted drive to 34 construct a number of small scale/micro dams To this end, more than 50 micro dams have been 35 constructed in the last 20 years, out of which only few dams are functional at the time being About 70% of 36 these have water harvest failure problem Out of these failed dams 60 to 65% are related to geological and 37 geo-technical reasons as indicated by the preliminary assessment of the Tigray Water Resource, Mines and 38 Energy Bureau (Dam Safety Committee, 2010) Although the magnitudes of the problems differ from one 39 project to another, the major technical problems observed in many of the micro dams include: excessive 40 seepage, insufficient inflow, early sedimentation and structural and dam stability problems related to 41 geology and geo-technical characteristics of the dam site At early stage of construction the micro dams in 42 the region were expected to irrigate about 2745ha However due to failure of dams only 1000ha is under 43 irrigation (Abay, 2010) Specifically, what prompted this study is the proposal to construct a new dam at 44 the location of an earlier failed dam at Hizaeti Afras, near Wukro town, in Tigray, North Ethiopia (Tigrai 45 Water Resources Mines and Energy Bureau, 2011) 46 It is an established fact that the successful design and construction of dams require a comprehensive site 47 characterization and a detailed design of each feature The design and construction of embankment dams is 48 complex because of the nature of the varying foundation conditions and range of properties of the materials 49 available for the use in the embankment Specifically, the major risk in the construction of any dam 50 foundation is the uncertainty involved in predicting ground conditions and behaviors Of course, the 51 accuracy of these predictions will improve with increasing effort devoted to the subsurface investigation 52 (Fang and John, 2006; U.S Army Corps of Engineers, 2004) 53 Furthermore, the stability of a structure depends upon the stability of the supporting earth materials The 54 important factors that require consideration are the location and depth of foundation In deciding the 55 location and depth, one has to consider the existence of active faults, failure surfaces, underground cavities, 56 lateral transition of geological formations which are some of the most common subjects of concern for 57 construction (Al-Fares, 2011; Mc Dowell et al 2002) Therefore, the first step should be to conduct 58 detailed geological and subsurface explorations, which characterize the foundation, abutments and potential 59 burrow areas 60 To assist in the mitigation of such dam failure problems and verify the viability of combining geophysical 61 data with geotechnical point data, a geophysical study has been conducted on a proposed dam site at 62 Hizaeti Afras near Wukro town, Tigray Regional State, North Ethiopia The aim is to characterize the 63 ground conditions and examine the suitability of the site for the proposed dam The data from geophysics 64 are also intended to supplement the geotechnical point data and give a better picture of the ground 65 conditions over the site This research work is specially directed at examining the reasons for the failure of 66 an earlier dam at the site and the viability of the proposal to reconstruct the dam on the same site 67 68 Material and Methods 69 2.1 The survey area and the proposed dam characteristics 70 71 The study area, Hizaeti Afras, is located near Wukro town, Tigray Regional State, North Ethiopia (Fig 1) 72 The dam to be constructed is an Earth fill (zoned) dam type with crest length of 564m, dam width of 5.0m 73 at the crest and maximum height of 12m with embankment slopes of 2:1 on downstream and 2.5:1 on 74 upstream sides The size of the watershed area is 43.84km2 of which 19.42km2 is cultivable lands, 75 14.39km2 grazing land and 10.03km2 is miscellaneous; and the total stream length up to the dam axis is 76 16.117km The Hizaeti-Afras catchment is a sub-catchment of the Gereb-Giba watershed area, which 77 drains to Tekeze River 78 79 [Figure 1] 80 81 The hydrologic analysis of 18 years of rainfall data from the nearby Wukro station shows the mean annual 82 rainfall to be 565mm and the calculated Peak Discharge from the catchment to be 142.52m3/s The storage 83 capacity of the reservoir is 1.052Mm3 including a dead storage volume of 0.274Mm3 The project is 84 planned to irrigate 42 hectare of land under dry season irrigation and give supplementary irrigation for 85 more than 85ha 86 Figure shows the general view of the dam site with the reservoir area in the foreground The elevated 87 ground (marked portion) on the farther side is the face of a former dam which has failed on the western side 88 of the picture (north is towards the observer) 89 2.2 Methodology and instrumentation 90 Geophysical surveys have been applied to civil engineering investigations since the late 1920s, where 91 specifically seismic and resistivity surveys were used for dam site studies Nowadays, geophysical 92 techniques are routinely used as part of geological investigations to provide information on site parameters 93 In dam site investigations in particular, geophysical methods are extensively used both on dam construction 94 projects and in the assessment of the condition of existing dam structures These methods help in 95 identifying local areas of concern which have no surface expression Moreover, the methods help to 96 delineate boundaries between residual soils, and weathered and fresh rock It is also possible to locate 97 anomalous foundation features like dykes, cavities, fault zones and buried river channels with these 98 techniques (Anderson et al 2008; Chambers et al 2006; Fell et al 2005; Reynolds, 1997) 99 100 [Figure 2] 101 102 The current study was carried out with the general objective of verifying the suitability of the subsurface 103 geological formations, ground conditions and structures for foundation of the proposed dam based on 104 results from geophysical investigations at the dam site Specifically, the objectives of the work are to map 105 the subsurface geology and geological structures at the dam site, map the subsurface resistivity 106 stratification and to establish the geoelectric section of the dam site, determine the depth to the sound 107 geological formation/bed rock and to correlate, wherever possible, the geotechnical results with findings 108 from geophysical survey and come up with a detailed description of the subsurface that could be used as 109 inputs to the dam design 110 To this end, the geophysical methods employed on the site included Electrical Resistivity Tomography 111 (ERT), Vertical Electrical Sounding (VES), and Magnetic surveys The electrical surveys have been 112 extensively used for geotechnical investigation in determining the subsurface stratification, contact between 113 rock units, locate fault zones, zones of deep weathering and cavities (Loke, 1999) Recent advances in 114 instrumentation, data acquisition and interpretation have resulted in the development of the electrical 115 imaging technique which has made it possible to systematically acquire large volume data in a short time 116 whose interpretation yields a much refined and better picture of the ground The imaging technique is 117 especially suitable in areas of complex geology where there are significant variations in resistivity both 118 vertically and laterally Owing to its high resolution, the technique as such is suitable for geotechnical 119 problems like dam foundation studies It can also be used in the exploration of alluvial deposits where 120 permeable gravel and sand beds can be distinguished from low permeability clays or rock This capability 121 has been applied for foundation materials at dam site where significant alluvial deposits occur (McCann et 122 al., 1987) The other method employed in the survey, the magnetic method, has the application in 123 engineering studies to locate boundaries between different lithologic units and geological structures that 124 display magnetic contrasts such as faults or dykes (e.g Hansen et al., 2005) 125 126 The instrument employed for the ERT and VES surveys is the IRIS SYSCAL Pro R1 Plus Switch-72, IP 127 and resistivity unit working in both “multielectrode” and “rho” modes for the electrical resistivity 128 imagining and sounding surveys respectively A Scintrex made Integrated Geophysical System (IGS-2) 129 proton precession magnetometer working in “total field” mode has been used for the magnetic surveys A 130 GPS has been employed for position location as well as timing the magnetic surveys 131 132 Geology of the study area 133 3.1 Regional Geology 134 The history of sedimentary basin in Tigray began in either the Ordovician or Carboniferous and probably 135 ended in the lower Cretaceous before the eruption of the trap volcanic series (Beyth, 1971) During the 136 Paleozoic, the channel-like basin was flooded followed by the deposition of Enticho sandstone and Edaga 137 Arbi glacial on the basement rock Transgression during the entire Triassic and most of Jurassic, sediments 138 mostly fluvial in origin, the so called Adigrat sandstone was accumulated on the craton (Bosellini et al., 139 1997) Further, transgression of the sea deposited the Antalo limestone The regression of the sea led to the 140 deposition of the Agula Shale A regional NW-SE uplift in the early Cretaceous is responsible for the 141 forced withdrawal of the sea from the entire East Africa This last regression of the sea deposited the Amba 142 Aradom formation, which is composed of siltstone, sandstone and conglomerate Boselleni et al (1997) 143 suggested that the deposition of Amba Aradom formation is post tectonic in nature The stratigraphy of the 144 rock units constituting the region from oldest to youngest is Enticho sandstone and Edaga Arbi glacial, 145 Adigrat sandstone, Antalo limestone, Agula shale, Amba Aradom formation and the Mekelle dolerite 146 147 3.2 Local Geology 148 The local geology of the study area and its surrounding is dominated by quaternary alluvial deposits, Agula 149 shale and limestone units (Levitte, 1970; Fig 3) The alluvial deposits are mainly composed of clay, silt, 150 coarser sediments like sand gravel and boulders The flat plain part of the dam reservoir that extends from 151 the river centre towards the rim of the reservoir is dominated by river deposit material including clay, silt, 152 sand and gravel This layer has variable thickness that varies from 2m to 5m when seen in the river cut 153 154 [Figure 3] 155 156 The Agula shale is an intercalation of shale, marl and limestone This unit is found dominantly at the 157 downstream and northeast side of the study area It is variegated with an alternating grey, brown and 158 pinkish colour In most places the shale unit has friable nature and is lightly fractured The limestone and 159 marl are moderately to highly fractured and jointed The light yellow coloured limestone unit is 160 horizontally bedded and moderately to highly fractured This unit is exposed mainly on the western bank of 161 the dam axis and the reservoir 162 Data acquisition and processing 163 4.1 Data Acquisition and data volume 164 The 2D Electrical Imaging data were acquired along four lines- three of which run almost parallel to the 165 proposed dam axis and one perpendicular to it The three parallel profiles, all oriented east-west, are about 166 120 m apart One of the survey traverses (Profile 1) is purposely designed to pass over the axis of the newly 167 proposed dam which coincides with the axis of the failed dam Interelectrode spacing of 5m was used 168 which, with the 72 electrode IRIS R1 Plus Switch setup, gives a spread length of 355 m for the main 169 sequence and 90 m for each roll along sequence (IRIS Instruments, 2006) A sequence uploaded to the 170 instruments with 648 data points /quadrupoles/ for the main and 297 data points for each roll along 171 sequence was used for data acquisition Accordingly, through a combination of a main sequence and a 172 number of roll along sequences imaging survey Profile has a length of 540 m while the other three have 173 profile lengths of 450 m each With these arrangements, quite large number of data points is acquired with 174 this method substantiating the high data density-high resolution of the imaging technique 175 176 Three VES with interspacing of 165 m were conducted using the Schlumberger expanding spread on the 177 profile that runs right on the dam axis (Profile 1) A maximum half current electrode spacing (AB/2) of 178 220m, believed to be of adequate investigation depth for dam site studies, was employed The magnetic 179 data were collected over the four imaging lines (at a station spacing of 20 m) and additionally over random 180 points well distributed to cover the area on both the upstream and down stream side of the proposed dam 181 axis (at nominal station spacing of 30 m) A total of 264 data points have been collected The desired 182 resolution for shallow depth geotechnical applications of the magnetic method is achieved by analyzing the 183 short wavelengths of the signal Merely scaling a deep investigation tool with respect to wavelength, to 184 adapt the method to near-surface investigations is, however, insufficient It is also important to increase the 185 spatial sampling density and thus reduce aliasing (Hansen et al., 2005; Pellerin, 2002) An increased 186 number of data points /high data volume/ have thus been acquired (Fig 4) The key aspects to such densely 187 sampled data include the obvious factors such as enhanced resolution of the subsurface in addition to the 188 ability to identify noise and multi-dimensional effects, and reduced spatially aliasing of the data, all of 189 which are important for inversion schemes Figure gives the distribution of the survey traverses and the 190 data points of all the methods employed 191 192 [Figure 4] 193 194 4.2 Data Processing, Results and Discussion 195 2D Electrical Resistivity, Electrical sounding and magnetic data processing 196 Data downloaded from the IRIS unit with the ProsysII software are initially filtered automatically and noisy 197 data removed (IRIS instruments, 2006) Manual examination of the data also makes it possible to remove 198 unwanted data sets Filtered main and roll along sequence data for each profile are then joined to form one 199 series whereby the final processing of the data is achieved using the RES2DINV inversion software to 200 produce an inverse mode resistivity section for each profile (Griffiths and Baker, 1993; Loke, 1997, 1999) 201 202 In this, a least square inversion process- an automatic iteration process that fits the modeled data with the 203 calculated one- proceeds to analyze the data (Loke and Barker, 1995) A root mean square error (RMS) of 204 5-8.8% achieved at the end of the iteration process was taken to be acceptable The final output of 205 RES2DINV software, the inverted 2D model resistivity sections, are used for interpretation purpose and to 206 correlate the results of the inverse model sections with results of other methods 207 The vertical electrical sounding data are processed using classical resistivity interpretation software where 208 the layer parameters are developed from which a geo-electric section that represents the geoelectrical 209 stratification of the area under study is constructed For the magnetic data, diurnal correction and removal 210 of the IGRF determined field value (Kearey et al., 2002) are used to determine the total magnetic field 211 anomaly and to produce anomaly map, models and profile plots The average IGRF value determined for 212 the nearby magnetic base station and subtracted to obtain the anomalies is 37231 nT 213 214 Interpretation of results and discussion 215 Interpretation is made based on integration of results from the 2D electrical imaging, vertical electrical 216 sounding, magnetic, and borehole log data Well lithologic log from an existing borehole, dug at the 217 boundary of the study area, has been used for lithological correlation with the vertical resistivity 218 stratification and assist in the interpretation The total depth of the borehole is 78.5m while the depth of 219 investigation of the 2D electrical imaging is 65m; the correlation between the two is therefore made only 220 for the first 65m depth Furthermore, a correlation between the inverse model resistivity section and the 221 magnetic anomaly profile plots is also done 222 223 Further analysis of the electrical and magnetic data has enabled the interpretation of apparent resistivity 224 sliced-stacked depth map, pseudodepth section map, geo-electric section, magnetic anomaly map, 225 analytical signal map, tilt derivative map, Euler depth map and 2D magnetic model which are developed 226 using different interpretation software 227 228 5.1 Electrical Imaging Profiles- Profiles-1, -2 and -3 229 Profile is aligned in a near E-W directon and runs over the proposed dam axis On this profile, electrical 230 imaging data of 1242 data points were acquired on a traverse of about 540m length The inverse model 231 resistivity section shown in Figure 5(a) was obtained with RMS error of 5.1% after just five iterations 232 showing a good data quality The high resolution 2D inverse resistivity model section has been correlated 233 with the lithologic log data of the nearby borehole and the magnetic profile plot for the same survey 234 traverse 235 236 Profile two is surveyed along similar E-W direction and lies within the resorvior area, which is on the up 237 stream side of the dam This profile covers a distance of 450m which is surveyed by one main and one 238 rollalong sequences Accourdingly, data was collected from 945 data points in which, after filtering the 239 noisy data, 900 data points were used for the inversion purpose The resultig inverse model resistivity 240 section is given in Figure 5(b) along with the magnetic profile plot for the same line Profile three was 241 condacted on the downstream side of the proposed dam axis along similar orientation with the above two 242 profiles and the traverse covers a distance of 450m Out of the data collected from 945 data points only 923 243 data points were used for the inversion purpose, again showing good data quality The least square 244 inversion of this data resulted in the inversion model section (Figure 5c) with RMS error of 3.9% 245 246 In all three plots (Figs a, b and c) the different segments of the inversion model section are clustered into 247 three prominent layers: a top thin layer of intermediate resistivity representing the dry top soil of aggregate 248 composition, a second layer of relatively lower resistivity as a response of the moist/saturated weathered 249 limestone layer, and a third layer of high resistivity response representing the competent black limestone 250 These subsurface stratifications match well with the nearby borehole log data and the correlations are 251 shown for the corresponding depths 252 253 [Figure 5] 254 255 The amplitudes of magnetic anomalies along the three parallel profiles drawn on the top part of the 256 corresponding 2D inverse section show considerable variation in their peaks (Fig 5) It is known that a 257 change in value/slope of the plot is an indication of difference in lithology, a contact between different rock 258 units or a variation in depth of the anomalous/causative body while the slope of the plot at these locations is 259 an indication of the contrasts in magnetic response Accordingly, there are a number of contact zones 260 interpreted from the sections The lower peaks depicted in the profile plots are interpretd to indicate the 261 presence of weak zones, areas of weathered, fractured and reworked material with low magnetic response 262 What is interesting in these plots is the certainity with which the high resolution imaging sections and the 263 magnetic anomaly plots are in agreement in depicting these subsurface discontinuties/weak zones that are 264 important from the point of view of the objectives of the survey 265 266 5.2 Electrical imaging cross profile: Profile 267 Imaging Profile is oriented in the N-S direction and is designed to be perpendicular to the three near 268 parallel survey traverses with the objective of examining the presence of structures/weak zones that run 269 parallel to the dam axis and to visualize the lithological stratification in the particular direction The inverse 270 model resistivity section for this line (Fig 6) is distinct from the survey lines that run parallel to the dam 271 axis An extensive low resistivity horizon characterizes of this section The magnetic anomaly profile plot 272 is marked by distictly lower peaks in two areas possibly indicating the presence of zones subsurface 273 weakness extending to larger depths at these locations These structures, being parallel to the dam axis, if 274 not damaging to the life and well being of the dam may contribute to the seepage and loss of water from the 275 dam reservoir 276 277 5.3 Correlation of the 2D inversion model sections 278 It is informative to combine the three parallel imaging profiles as shown in Figure for a better 279 understanding and visualization of the subsurface condition in the study area A physical barrier has 280 necessitated the beginning of survey Profile 1, the traverse line that runs along the intended dam axis, 281 starting 90m ahead of the other two profiles which were conducted on the upstream and downstream side of 282 the dam axis However, all the three profiles have their beginning on the western side of the study area in 283 about the same region, and hence have been arranged accordingly 284 285 As shown in Figure 7, the first layer with high apparent resistivity value is relatively thick in the eastern 286 part of the study area in almost all the profiles with a decreasing coverage from upstream to downstream 287 side, and the thickness of the second layer is significant in the western part as compared to the east on all 288 the profiles However, the thickness of this layer also considerably decreases on the downstream side as 289 compared to the upstream side of the dam axis The third layer has undulating nature in all profiles and it is 536 537 538 Caption to Figures 539 540 Figure Location map of the study area 541 Figure General view of the Hizaeti Afras dam site with the earlier now broken Earth dam on 542 the western side of the area (north towards the observer) 543 Figure Geological map of the study area and its surroundings 544 Figure Location map of the electrical survey traverses and the data points (solid dots) for the 545 546 547 magnetic surveys Figures a, b, and c Interpretation of electrical imaging 2D inverse resistivity sections and magnetic profile plots along Profiles -1, -2 and -3 respectively 548 549 550 Figure Interpretation of geophysical data of Profile 4; a) magnetic anomaly curve and b) 2D inversion model section 551 Figure Correlation of the 2D inverse model resistivity sections 552 Figure Apparent resistivity sliced–stacked depth map of the survey area The orientations of the 553 554 555 profiles from which the data have been extracted are shown on the top slice Figure Geoelectric section along the dam axis (Profile-1) Hizaeti Afras Dam, Wukro, North Ethiopia 556 Figure 10 2D magnetic modeling along Profile-1 557 Figure 11 Magnetic anomaly (a) and analytical signal (b) map and of the surveyed area 558 Figure 12 Magnetic (a) Tilt angle Derivative and (b) Euler Depth Solution map of the study area 559 560 561 562 18 533298 Sudan 543298 553298 Kilte Awulaelo Woreda South Sudan Somalia Kenya Legend Wukro Town Road River Study area Kilte Awulaelo Woreda Grid UTM: zone 37N Projection: Transverse Mercator Datum: adindan Figure 563298 Kilometers 573298 1493422 1503422 1513422 1523422 1533422 1543422 N 572 573 E W Reservoir Area 574 575 576 Figure 577 20 578 579 580 581 582 21 583 584 585 586 587 22 588 589 590 591 23 592 593 594 595 596 597 24 598 599 600 601 602 25 603 604 605 606 607 608 Figure 609 26 610 611 612 613 614 27 615 616 617 618 619 620 28 621 622 623 624 625 626 627 29 Figure 12 628 629 Abstract 630 Geophysical surveys involving the techniques of electrical resistivity imaging, electrical sounding and 631 magnetics were employed to characterize the ground conditions at a proposed dam site at Hizaeti-Afras, 632 Wukro, North Ethiopia The techniques were utilized to map the depth to the competent formations, their 633 relative suitability for foundation work and the presence and extent of weak zones in the subsurface The 634 work has mapped the different lithologic units of the subsurface and determined the depth to the basement 635 rocks in the area Through correlation of the inverse model resistivity sections of the imaging surveys, the 636 geoelectric section of the sounding survey and the magnetic profile plots with available borehole lithologic 637 logs, it is shown that the results very well supplement the geotechnical point data in addition to providing a 638 wider coverage in mapping areas of weak ground that could otherwise be missed with widely spaced 639 borehole information The combined results of the survey show the proposed dam axis to be unsuitable 640 The power of the electrical resistivity imaging technique and its potential to map the shallow subsurface 641 with adequate resolution are illustrated The result is a strong suggestion that geophysical techniques can be 642 used to assist and extrapolate borehole geotechnical data especially when large area is to be used for 643 development of large infrastructure 644 645 646 30 647 648 649 650 651 652 653 Highlights • • • • • High resolution 2D electrical imaging sections and their suitability for geotechnical studies Mapping structures with correlation of 2D resistivity, geoelectric and modeled magnetic sections Advantages of using a combination of magnetic data enhancement techniques We show the reasons for the failure of an earlier dam on the proposed site We show why the proposal to build a new dam at the same location is not viable 654 31