HPP Gabral Kalam HPP Seismicity Report.Doc
- 1. GABRAL KALAM HYDROPWER PROJECT, SWAT, KPK SEISMIC HAZARD EVALUATION REPORT 1. INTRODUCTION The proposed Gabral Kalam Hydropower Project will utilize the flows of the Gabral River which is a tributary of Swat River in upper Swat. The Gabral valley is at an altitude of 7500 ft., surrounded by Chitral District in the north, Utror valley in the south west, upper Dir district in the west and Mahodand valley in the east. The Project area is in north of Kalam near Osho hill station, which is recognized as the major tourist point in the upper Swat. The site of is about 20 Km by road from Kalam.The location coordinates of weir and Power House (101 MW) of Gabral Kalam Hydropower Project are given below: Weir Latitude 35.505°N Longitude 72.519°E Powerhouse Latitude 35.500°N Longitude 72.577°E The maximum height of weir is less than 10 meters, so according to ICOLD definition, this weir does not fall in the category of Large Dams. So ICOLD guidelines for selection of seismic parameters for large dams (2016) are not applicable for this Project. The Project should therefore be designed as per requirements of building codes or concrete hydraulic structures. The Gabral Kalam Hydropower Project is located in the Kohistan Island Arc physiographic province, a tectonically active region which is sandwiched between the converging Indian and the Eurasian tectonic plates. The Project region has been subjected to damaging earthquakes in the past and therefore it is imperative that a study of tectonic and earthquake history of the region be conducted to determine the seismic hazard to which the proposed Project may be exposed to and to evaluate realistic seismic design parameters for the safe design of the project components. As the Project is located in the collision zone of the Indian and Eurasian plates, therefore the Gabral Kalam Hydropower Project could face a severe earthquake hazard potential. The Geological Survey of Pakistan has placed the Project area in the “Serious Seismic Danger Zone”. In Building Code of Pakistan, Seismic Provisions (2007), the Project area falls in Zone-3. Moreover, within the scenario of the October 08, 2005 earthquake of Pakistan it becomes important to be very cautious regarding the seismic hazard assessment for such an important Project. For the seismic hazard evaluation of Gabral Kalam Hydropower Project, the following methodology was adopted: Study of regional geological and tectonic information collected from available literature and maps. 1
- 2. Compilation of historical and instrumental earthquake data and analysis of the available earthquake record for completeness. Identification and characterization of potential seismic sources in the project region. Evaluation of seismic hazard in accordance with current practices, including: o ER 1110-2-1806 - Earthquake Design and Evaluation for Civil Works Projects o EM 1110-2-6050 – Response Spectra and Seismic Analysis for Concrete Hydraulic structures o Building Code of Pakistan Seismic Provisions (2007) Assessment of soil liquefaction potential. 2. GEOLOGY OF THE REGION 2.1General The geology and geodynamics of the Karakorum-Himalayan region in northern Pakistan are characterized by the interactions of three principal tectonic units: The Eurasian Plate; The Kohistan Sequence; and The Indian Plate. These units have distinctly different lithology and structural settings and are separated by two major branches of the Indus suture (Tahirkheli, et al., 1979; Treloar, et al., 1990; Khan, et al., 1997). Both sutures are marked by the occurrence of a mélange including ultramafic rocks, the southern one also having a wedge of garnet granulites considered to have recrystallized at a depth of more than 40 km. The rocks making up the Kohistan sequence, between the two sutures, are predominantly calc-alkaline plutonics and volcanics with subsidiary volcano sedimentary and sedimentary rocks. Tahirkheli, et al., (1979) have suggested that the Kohistan sequence represents the crust and uppermost mantle of an extended island arc turned on end during the collision of the Indian-Asian Iandmasses. Later studies have shown that the structure of the area is too complex for such a simple interpretation and requires a detailed analysis before final conclusions can be reached about its nature (Coward, et al., 1986). 2.2The Eurasian Plate In the Karakorum area, Gansser (1964) distinguished three tectonic zones: A Karakorum Tethyan zone 2
- 3. A central metamorphic zone with plutonic rocks – Karakorum Batholith; and A southern volcanic schist zone. Of these, the last one is now considered to be a part of the Kohistan sequence occurring to the south of the Main Karakoram Thrust (MKT or the Northern Suture), whilst the first two occur to the north of the MKT. 2.3 The Northern Suture In the section from Hunza to Chalt, there is an almost chaotic arrangement of large lenses, each several kilometers long and several tens of meters wide, of limestone, sandstone, conglomerate and mafic and ultramafic rocks in a matrix of chloritoid slates. The basic rocks with prominent volcanic breccias and greenschists are rich in epidote, chlorite and actinolite. The ultra-basics consist of serpentine, talc-chlorite schists, talc-carbonate schist, calcite- chlorite schists, chromite-chlorite schists, and minor relict harzburgite. Ultra- basic masses are apparently more abundant to the west of Chalt. There are large lenses of quartzite which may have formed in situ or which may be tectonic blocks and limestone intermixed with other sediments. The whole assemblage has the appearance of a major mélange with no simple repetitions, as expected in an imbricate zone. The structures in the high-grade metamorphic rocks contrast with those in the main mélange up-dip as seen from mineral lineations and folds with curvilinear hinges. This tectonic zone is considered to mark the suture between the Kohistan sequence and the Eurasian Plate to the north. There is no evidence of blue- schists, of obducted high-pressure granulites or of an ophiolite, but instead large tectonic lenses of a mélange. 2.4 The Kohistan Island Arc The principal rock units of the Kohistan Island Arc include, from south to north: Jijal Complex; granulite, mafics, and ultramafics; Kamila Amphibolite Complex; mostly norites; Chilas Complex; mafic and ultra-mafic layered complex of gabbros, norites, and dunite intersected by dikes and seams of anorthosite and chromitite; Kohistan Batholith; various calc-alkaline intrusives; and Kohistan Arc Sequence; various meta-sedimentary units and volcanic units typical of an island arc and fore-arc setting. It is important to point out that geologic mapping has shown that the contacts of the major lithologic units in the Kohistan Island Arc area are faulted (e.g. Ghanzafar, et al., 1991), including the southern and northern boundaries of the Chilas Complex. The Kamila Complex is also dissected by numerous shear zones and is bounded to the north by a major shear zone (Kamila Shear Zone). 3
- 4. The Kohistan Island Arc was formed in the mid-Cretaceous and sutured to Asia around 100-85 million years ago. India later collided with the arc after continued subduction beneath the arc complex, now accreted to the active continental margin. After full collision, the arc was tilted, uplifted and dissected, enabling examination of the crustal structure of an immature island arc. Suturing to the Asian active continental margin meant that the arc itself became an active continental margin, and the attendant crustal thickening produce an evolution in magmatism from basaltic to calc-alkaline. This is best observed in the phases of plutonism observed in the gabbro-norite plutons of the Chilas Complex and in the Kohistan Batholith, and also in the surrounding (meta-) volcanics into which these granitic sheets are intruded. The Indus river gorge section through the Kohistan Island Arc reveals an informative section through an island arc from the Main Mantle Thrust (MMT) to which the arc forms the hanging wall, in the south to its footwall position against the Northern Suture. The arc itself is exposed for over 200 km north to south and about 300 km from east to west. The strike of the various tectonic units is approximately east-west; therefore the deepest crustal regions are represented in the southern portions. 2.4.1 Jijal Group A complex of layered mafic and ultramafic intrusions occurs between Patan and Jijal, an area of about 200 km². In the north are garnet-clinopyroxene- plagioclase rocks containing relics of norite, and so it is likely that these are high-pressure metamorphic equivalents of the Chilas complex. The grain size is similar to that of the norite but garnets continued to grow after the deformation and locally grew to cover 8 cm especially in leucocratic veins. Hornblendites may contain hornblende-garnet, garnetite and garnet plagioclase. The overall composition is a high-pressure metamorphic assemblage and the rocks are equivalent to eclogite facies, thus representing the lower parts of the crust. Towards the southern boundary of the complex there is an increase in proportion of clinopyroxenes and hornblendites, until the main ultramafic body is reached, which consists of clinopyroxenites, and dunites which have lenses of layered chromitite up to 5 m thick. It has been concluded that both the granulites and the dunites suffered granulite grade metamorphism at 600 – 700°C and 12 – 14 kb and at 800 – 850°C and 8 – 12 kb respectively (Jan & Howie, 1981). The Jijal Complex is possibly a tectonic fragment of the Chilas Complex that was subducted or downthrusted to a substantial depth against the MMT. 2.4.2 Kamila Amphibolite Belt This is composed primarily of norites; mostly at amphibolite facies (therefore pyroxenes have retrograded to hornblendes). Amphibolite metamorphism is assumed to have occurred during suturing to Asia. The belt also includes banded amphibolites with or without garnet, hornblendites, schists, garnet 4
- 5. gabbros, and anorthosites, diorites, tonalities and granites and thin garnet quartzites and calc-silicate lenses. The proportion of amphibolite is commonly low. The belt is distinctive in that most intrusive rocks are concordant and parallel to the regional trend and have been intensely deformed, many of the coarser leucocratic types becoming augen gneisses. Ghazanfar, et al. (1991) is of the view that these are the oldest exposed unit of the Kohistan sequence and show ophiolitic character. The Kamila belt is dissected by a number of small shear zones and is bounded to the north (adjacent to the Chilas Complex) by a major shear zone, the ‘Kamila Shear Zone’. The belt represents the mid crustal regions of the primitive arc. 2.4.3 Chilas Complex The Chilas Complex is a vast stratiform cumulate body over 300 km long and 8 km thick, dominated by intrusions of calc-alkaline gabbro-norites, which locally show layering (Ghazanfar, et al. 1991). It contains an upward sequence of hypersthene gabbro, major chromite-layered dunite, norite, gabbro, minor troctolite, harzburgite and dunite, and at the top, norite. Particularly impressive are rhythmically-alternating phase-graded cumulate layers up to 0.1 m thick, slump folds, syn-sedimentation faults, and sedimentary breccias. Some layers up to about 0.3 m thick are of almost pure anorthosite. Dykes of pyroxene-hornblende anorthosite cut both homogeneous and layered rocks. The lower dunites are up to 1 km thick and contain 3-m-thick compact chromitite seams. All these rocks show evidence of several phases of deformation. Isoclinal folds in norites have hypersthenes orientated in axial planar fabrics and the penetrative mineral fabric in the norites is parallel to the axial planes of folded pyroxene amphibolite dykes. These relationships suggested a tectonic origin for the main mineral fabrics in the complex. Ghazanfar et al. (1991) have shown both contacts of norite as fault which has led to the formation of norite mylonite in an otherwise very tough dark coloured rock with streaks of white or pale-white colour. 2.4.4 Jaglot Group Occurring north of the Chilas Complex, the Jaglot Group comprises schists intercalated with material of volcanic origin. To the south, the Jaglot schists are intruded by the Chilas gabbro-norites, while to the north the Kohistan Batholith intrudes them. This confusion and overprinting from magmatism means that the Jaglot Group has only recently been defined as a unit (Treloar, et al., 1990). The main lithologies are greenschist facies metabasites, pillows and some volcaniclastic material inter-bedded with pelitic, psammitic and calc- silicate schists, representing clastic and carbonate sedimentary protoliths. 5
- 6. 2.4.5 Kohistan Batholith This consists of a zone of plutonism associated with active continental margin volcanism (i.e. Andean type magmatism). This is a principal unit of the Kohistan magmatic arc and constitutes a 300 km long and up to 60 km broad belt to the west of Nanga Parbat. The Kohistan Batholith is composite and consists of numerous large to small plutons, plugs, dykes and sheets emplaced over a time span of some 75 million years (Kazmi & Jan, 1997). A wide range of rocks has been reported to constitute the batholith: gabbros, hornblendite, diorites, quartz diorite, adamellite, granodiorite, granite, tonalite, pegmatite etc. 2.4.6 Northern Kohistan Arc Sequence This is comprised of various volcanic and metasedimentary Groups – Dir and Chalt Groups. These include Eocene calc-alkaline basaltic-andesitic-rhyolitic lavas and pyroclastic deposits associated with the active continental margin stage of the arc complex. There is a tectonic break between rocks of the northern suture and the volcanic and sediments belonging to the Chalt Group to the south, which make up the northern part of the arc. This group contains meta-greywackes and slates, epidotic grits and tuffs, hornblende-bearing tuffs, chlorite schists, schistose amphibolites, amygdaloidal pillow-bearing basalts and fragments basic volcanics. Further south near Gilgit and Raikhot there are graded psammites and pelites and locally thick piles of deformed pillow lavas, but these occur as screens betweens large plutons of diorite and tonalite. The total thickness of deformed and weakly metamorphosed sediments and volcanics reaches several kilometers but this may involve repetitions by folding and thrusting. The rocks are folded by large upright, tight to isoclinal anticlines and synclines, which plunge east or west. They are cut by thin granitic dykes and by muscovite pegmatites, which are discordant to both cleavage and bedding. 2.5 Indian Plate The bedrock suites south of the Kohistan Island Arc and southern suture zone include those forming the pre-collisional stratigraphy of the Indian Plate plus the syn-and post-tectonic material eroded from the mountain ranges of the Himalayas, Karakoram, Hindukush, and Pamirs. 2.5.1 Salt Range The Salt Range defines the Frontal thrust of the Himalayas, a thin-skinned structure riding on an evaporite decollement. The topographic relief of the Salt Range is produced by blind thrusts and ramp anticlines. 6
- 7. 2.5.2 Molasse Molasse sequences of detrital sediments form the Margalla Hills and the Punjab Plains. All tectonism is thin skinned with numerous southward- propagating thrusts that have produced numerous imbricate zones. 2.5.3 Hazara Sediments The Hazara metasedimentary belt is largely composed of Precambrian to Early Mesozoic sediments. The Precambrian sequence is composed of quartz schist, graphitic schist, marble and gneiss overlain by thick sequence of slate, phyllite and greywacke sandstone. The Precambrian sequence is unconformably overlain by quartzite and argillites. 2.5.4 Mansehra Batholith Imbricated slices of this granitic batholith, intruded into the metamorphic cover, are exposed in the Hazara Syntaxis. It is Cambrian in age. 2.5.5 Metamorphic Cover This consists of late Precambrian – early Cambrian metasediments that have undergone a Palaeozoic low-grade metamorphism, and which are overlain by pre-collisional Mesozoic sediments. These were further metamorphosed and thrusted in the foothill of the MMT synchronous with full collision. 2.5.6 Nanga Parbat Group Rocks of the Nanga Parbat Group represent units belonging to the cratonic Gondwana basement, exposed in the Nanga Parbat – Haramosh Massif syntaxis. The Proterozoic gneisses of the Indian Plate have their northernmost exposure in the Nanga Parbat Syntaxis and represent the lowest structural levels of the Indian Plate observed. They have been mapped and subdivided into three lithostratigraphical groups (Madin et al., 1989). 2.6 Local Geological Setting The local geological setting around project area has been interpreted based on the geological map of Mahodand Quadrangle after Afridi et al. (1999) published by Geological Survey of Pakistan (GSP). Quaternary Deposits Stream Deposits Stream deposits comprise gravels, cobbles and boulders with fine to coarse sand. The deposition is on-going process with the perennial and non- perennial streams. These cover the stream and river beds of active channels. 7
- 8. Alluvial deposits These are old river or stream deposits making terraces along the valley slopes. These deposits comprise gravels, cobbles and boulders embedded in silty sandy matrix. Most of the settlements are situated on these deposits. The top layer of these deposits comprises fine materials and therefore are being used for cultivation. Rock Units Matiltan Granite Matiltan granite comprises white to light grey, medium to coarse grained and porphyritic granite and granodiorite composed of orthoclase, plagioclase, quartz, hornblende and biotite, with xenoliths of quartzite, amphibolite and diorite. Utror Volcanics Utror Volcanics comprise grey, green, maroon red and at places white, fine to medium grained, identified as andesite, dacite, rhyolite with tuffs, agglomerate and pyroclasts. Barawal Banda Quartzite Barawal Banda Quartzite comprises light to dark grey on fresh surface and brownish grey on weathered surface, thin to thick bedded, fine grained quartz which is cherty at places. Barawal Banda Slates/ Phyllites/ Schists This rock units in this formation comprise grey, green and maroon in color, thin bedded, fine to very fine textured, occasionally silty phyllites, schists and slates. Occasional beds of light grey thinly bedded limestone are also present at places. Deshai Diorite Deshai Diorite comprises grey, greenish grey, medium to coarse grained diorite composed of plagioclase, hornblende, biotite with subordinate quartz, hornblended pegmatites and quartz veins. Kalam Quartz Diorite (Associated with meta sediments) It comprises grey, greenish grey, medium to coarse grained quartz diorite composed of plagioclase (andesites), hornblende and biotite. Quartz feldspathic veins and xenoliths of Kalam meta-sediments are present in places 3. REGIONAL TECTONIC FRAMEWORK The geodynamic framework of northern Pakistan is characterized by the collision and coalescence of Eurasian and Indian Continental Plates, which were once separated by the oceanic domains, and creation of the Kohistan island arc in the late Cretaceous. The collisional process started in the late Eocene to early Oligocene with the formation of the Himalayan Ranges and this process still continues. Relative to Eurasia, the Indian plate is still moving 8
- 9. northwards at a rate of about 4 cm/year. The subduction of the Indian plate beneath the Eurasian plate has resulted in folding and thrusting of the upper crustal layers near the collisional boundary. The thrusting has been depicted from north to south in the shape of MKT (Main Karakoram Thrust), MMT (Main Mantle Thrust), MBT (Main Boundary Thrust) and SRT (Salt Range Thrust) the locations of which are shown in Figure- 1. 3.1 Regional Tectonics The geology of northern Pakistan is a superb example of continental collision tectonics. In this area, the three of the world’s greatest mountain ranges converge, the Himalayas, the Karakoram, and the Hindukush. The mountain building process that formed these ranges commenced in Cretaceous time when Indian plate started moving and was carried northward (Scotese et al., 1988). During that time (i.e. Early Cretaceous) Karakoram terrane sutured with eastern Hindukush along the Tirich Mir fault (Zanchi et el., 2000; Hildebrand et al, 2001). Soon after, the intra-oceanic Kohistan arc formed over a subduction zone that dipped beneath the arc, either to the south or to the north (Khan et al. 1993). It is widely accepted that the northward movement of India was concurrent with the accretion to Asia of an intra- oceanic arc system, the Kohistan arc that collided with Asia along the Shyok Suture or MKT. The southern margin of Asia, including the Kohistan arc, then became an Andean type convergent margin, until India collided with Asia. Thrusting of the Kohistan terrane southward over the northern Indian plate margin along the Main Mantle Thrust (MMT) probably took place in Late Cretaceous or Paleocene time and was completed by 55Ma, forming the Indus Suture Zone (Searle et al., 1999). A detailed description of the salient features of the Kohistan magmatic arc and the adjoining Northwestern Himalayan Fold-and-Thrust Belt of the Indian plate is given below. 3.1.1 Kohistan Magmatic Arc Kohistan is an intra-oceanic island arc bounded by the Main Mantle Thrust (MMT) to the south and the Main Karakoram Thrust (MKT) to the north. This E-W oriented arc is wedged between the northern promontory of the Indian crustal plate and the Karakoram block. Gravity data modeling indicates that the MMT and MKT dip northward at 35˚ to 50˚ and that the Kohistan arc terrain is 8 to 10 km thick (Malinconico, 1989). Seismological data suggests that the arc is underlain by the Indian crustal plate (Seeber and Armbuster, 1979, Fineti et al., 1979). The northern and western part of the arc, along MKT, is covered by a sequence of Late Cretaceous to Paleocene volcanic and sedimentary rocks. The central part of the arc terrain is mainly composed of Kohistan Batholith which comprises an early (110-85 Ma) suite of gabbro and diorite, followed by more extensive intrusions of gabbro, diorite and granodiorite (85-40 Ma) which are intruded by much younger dykes and sills of leucogranite (30-26 Ma). 9
- 10. The southern part of Kohistan is comprised of a thick sequence of mafic and ultramafic rocks. These rocks may be divided into three tectono-metamorphic complexes separated by major thrust zones (Figure- 2). The Chilas Complex forms the northern and upper unit. It comprises layered norites and gabbros metamorphosed to granulite facies. It is characterized by a series of south- verging folds. It has been thrusted southwards over the Kamila Amphibolites Complex. The latter consists of amphibolites, meta-gabbro and orthogneisses. This sequence comprises a highly tectonised shear zone. Southward, it is thrusted over the Jijal Complex which forms a tectonic wedge between the Kamila Shear zone and the MMT. The Jijal Complex is largely comprised of garnet-pyroxene-granulites and ultramafic rock (Tahirkheli and Jan, 1979; Coward et al., 1986; Khan et al., 1993; Treloar et al., 1990; Miller et al., 1991). 3.1.2 Northwest Himalayan Fold-and-Thrust Belt The Northwest Himalayan fold-and-thrust belt occupies a 250 km wide and about 560 km long irregularly shaped mountainous region stretching from the Afghan border near Parachinar up to the Kashmir Basin. The Hazara-Kashmir and Nanga Parbat Syntaxes form its eastern margin. It covers all the terrain between the Main Mantle Thrust (MMT) in the north and Salt Range Thrust in the south. This region comprises the mountain ranges of Nanga Parbat, Hazara, Southern Kohistan, Swat, Margalla, Kalachitta, Kohat, Potwar and Salt Range. A major thrust fault, the Panjal-Khairabad Fault divides the NW Himalayan sequence into a deformed southern zone, often referred to as the external or foreland zone and a deformed and metamorphosed northern zone, also known as the hinterland zone (Pivnik & Wells, 1996). The foreland zone comprises the Hazara-Kashmir Syntaxis, Salt Range and Kohat-Potwar fold belt and the Kurram-Cherat-Margalla thrust belt, whereas the hinterland zone comprises the Himalayan crystalline nappe-and-thrust belt. 3.2 Major Tectonic Features The Project site is located in the Kohistan island arc which is sandwiched between the Indian and the Eurasian plates. The major faults of the project region include, from north to south, the Main Karakoram Thrust (MKT), Kohistan Fault, Main Mantle Thrust (MMT), Panjal-Khairabad Thrust, Main Boundary Thrust (MBT) and Salt Range Thrust. The general trend of these faults is predominantly east-west with change in trend due to syntaxial bends. The general description of these major faults is as follows. 3.2.1 Main Karakorum Thrust (MKT) This is the major regional fault representing the suture zone between the two colliding plates. This fault represents the northern boundary of the Kohistan island arc and runs eastward to join Indus suture zone in upper Himalayas and terminate at its junction with Karakoram fault. In the Chitral and Gilgit area, the rocks of Karakoram Batholith are thrusted over the rocks of Kohistan Batholith along MKT. 10
- 11. 3.2.2 Kohistan Fault On the Geological Map of NWFP (2006) published by the Geological Survey of Pakistan, the contact between the Kamila amphibolies and the Satpat ultamafics to the south of Dasu are shown as the Kohistan fault. Along this fault, the rocks of the Kamila complex are thrust over the Satpat complex rocks. This fault runs almost parallel to MMT. 3.2.3 Main Mantle Thrust (MMT) Main Mantle Thrust (MMT) is a northward dipping regional thrust, which separate the Indian Plate from the Kohistan Island Arc. It extends from Khar (Bajaur Agency) in the west to the north of Naran (Kaghan Valley) in the east where it takes a northeast ward bend towards the east of Bunji and gets truncated by Raikot Fault. The thrust inclines steeply near the surface; however, this inclination is believed to decrease considerably with depth likewise as interpreted for other local thrust faults of the region. Structurally the Main Mantle Thrust is characterized by a number of northwest dipping high angle imbricate thrusts, which converge together in the east and being terminated as Raikot fault. A number of other sub-parallel shears associated with MMT and distributed near Chilas and Bunji merge together and join Raikot fault. MMT is almost aligned sub-parallel to the Main Karakoram Thrust in the north and Main Boundary Thrust in the south except the Hazara-Kashmir Syntaxial area, where MMT remains unaffected and continues its journey in the northeast direction to join the Raikot fault. In the east it is abruptly juxtaposed against the Nanga-Parbat-Haramosh Massif, while in the west it meets the Main Karakoram Thrust in Afghanistan. Before joining the Main Karakoram Thrust, it is offset by northwest and northeast trending strike slip faults near Khwaza Khela and Besham. The Patan earthquake of December 28, 1974, having magnitude 6.2, was associated with MMT. The Raikot fault zone and associated structures exhibit remarkable neotectonic features including fault scarps and exposures where Nanga Parbat gneisses overlie Pleistocene tillites. The earthquakes of November 2002 and January 2003 in Astore valley have been attributed to movement in this zone. On the basis of the recorded seismicity and observed neotectonic features both the Main Mantle Thrust and Raikot fault are considered seismically active. 11
- 12. 3.2.4 Panjal-Khairabad Thrust The Panjal-Khairabad Thrust is an important active tectonic feature of regional significance. It runs northwards and parallel to the Main Boundary Thrust (MBT) on the eastern side of Hazara- Kashmir Syntaxis where it is normally called Main Central Thrust (MCT). These faults gradually converge and eventually join about 5 km north of Balakot (Calkin et al. 1975, Bossart et al. 1984 and Greco 1991). In the area west of Hazara-Kashmir Syntaxis, this fault is commonly called Panjal Thrust. A left lateral strike slip fault cuts across both the Panjal Thrust and MBT approximately 6 km south of Balakot, from where onwards the Panjal Thrust continues its independent journey southwards. It is traceable up to Garhi Habibullah from where onward it is concealed beneath Quaternary deposits. The thrust comprises several segments having an aggregate length of about 130 km. To the west this fault passes through the Gandghar range near Haripur and joins the Khairabad fault located on the northern side of the Attock-Cherat range, hence it is sometimes referred as the Panjal-Khairabad fault. The geologic positioning and seismicity associated with the Panjal-Khairabad fault renders it as an active regional tectonic feature capable of generating large earthquakes. 3.2.5 Main Boundary Thrust (MBT) The most significant and active tectonic feature of regional extent is the Main Boundary Thrust (MBT). It is the main frontal thrust of the Himalayan Range, which runs along the Himalayan arc for almost 2500 km from Assam in the east to Kashmir and Parachinar in the west. The MBT along with other associated thrusts forms the sharp conspicuous Hazara-Kashmir Syntaxis. This syntaxial bend is the most dominant tectonic feature of the area as all local major fault systems and geologic structures follow its trend. On the west side of this feature, the MBT initially follows a rather southwest trend and then extends westward reaching Parachinar. Near its surface trace, the MBT dips northward at a steep angle, which becomes sub-horizontal with depth. Islamabad-Rawalpindi area is located at a close distance south of the western limb of the MBT. A number of large to major earthquakes have occurred along the Himalayan Arc east of the Hazara-Kashmir syntaxis during the last two centuries, which places it amongst the most active regions of the world. Much of the seismicity recorded during the last century is attributed to surface and subsurface extensions of the MBT and other associated thrusts. Based on this data, Seeber et al. (1981) have shown that great earthquakes occurring along Himalayan Arc are probably related to slips taking place along this quasi- horizontal detachment surface. 12
- 13. Based on the above, the MBT is considered active having seismic potential sufficient enough to generate large to major earthquakes. 3.2.6 Salt Range Thrust The Salt Range Thrust runs along the southern extremity of the Salt Range between the Jhelum and Indus Rivers. It is marked by thrusting of highly deformed older rocks of the Salt Range over the relatively less deformed Tertiary Sequence of Jhelum Plains (Punjab Plains). Irregular escarpments rise explicitly from the Punjab Plains; however, on the northern side gently dipping strata merge into the Potwar Plateau. The Salt Range Thrust is about 300 km long, having a general trend in an east-northeast direction. It is extensively segmented by northeast and southeast trending minor transverse faults. The fault segments exhibit considerable off-sets at various locations. A significant part of Salt Range Thrust is covered by fanglomerates, while at places near Jalalabad and Kalabagh the thrust trace is clearly visible where Paleozoic rocks overlie the Neogene strata. The Salt Range Thrust terminates in the west against the Kalabagh fault, which is a seismically active tectonic feature of the area. Its eastern termination is near the right bank of the Jhelum River, where it bifurcates and takes a northeast wards bend. In contrast to other parts of the frontal zones in Pakistan, the Salt Range Thrust is marked by a low level of seismic activity which is mainly attributed to the aseismic nature of underlying Cambrian Salt deposits. It has no history of known rupture in moderate to large magnitude earthquakes. However, the entire Salt Range is considered active as indicated by micro-seismic studies and observation of Quaternary deformations in western and central portions of the fault. 3.3 Local Tectonic Features The Project is located in the western part of the Kohistan island arc close to the boundary between the Kohistan Batholith and Utror Volcanics. In the Geological Map of NWFP (2006) published by the Geological Survey of Pakistan (Figure- 3), the contact between the Kohistan Batholith and Utror Volcanics is shown to be a normal contact but some researchers believe that this contact is faulted (Ghazanfar et al, 1991). In the Geological Map of Northern Pakistan edited by Searle & Asif (1995) presented in Figure- 4, this contact is shown to be a faulted one. A regional fault named as Shandur Thrust is marked on GSP Geological Map of Mahudand Quadrangle (Scale 1:50,000) prepared by Afridi et al. (1999) which shows this fault at about 10 km in the northwest of the Project area. This thrust fault has been marked by the Utror Volcanics group of rocks in southeast while by Kalam Quartz Diorite associated with meta sediments in the northwest. The fault is dipping towards the northwest and is directed northeast-southwest ward. It appears that this fault may coincide with the 13
- 14. contact between Kohistan Batholith and Utror Volcanic shown in above referred regional geological maps. 4. EARTHQUAKE RECORD 4.1 General Study of the earthquake record involved several activities: Investigation of the pre-instrumental or historical seismicity Examination of instrumentally recorded earthquake record Interpretative description of the Kashmir earthquake of October 8, 2005 Analysis of the earthquake record Description of interpreted focal mechanisms 4.2 Pre-Instrumental (Historical) Seismicity Before the establishment of seismological observatories, which began at the beginning of 20th century, intensity data collected from the historical records was the only source of earthquake information. Historical Earthquake data is a general account of damage/ loss to life (human & animal) and property. The historical pre-instrument earthquake data has been collected from the description of the earthquakes given in the memoirs or records of travelers, historians and writers. Such earthquakes catalogues have been compiled by Oldham, 1893, Heukroth and Karim, 1970, Ambraseys et al. 1975 and Quittmeyer and Jacob, 1979 and presented in Appendix-A. The historical earthquake data reflects that northern Pakistan as a whole has remained a house of prominent earthquakes. Taxila (25 A.D.) event is probably the most conspicuous one that changed style of building-construction out rightly. An important value of intensity data is that it establishes some understanding of the level of the damage that can be expected to occur in a given region. The catalogue of historical earthquakes for this region is rather sparse and probably highly incomplete. Since the 1700’ s, the historical earthquake data for the northern areas of Pakistan are few and mainly concentrated on the centres of colonial administration. The important tremors for which damage data is available are as follows: Aristobulus of Cassandreia described that the first known historical account of seismicity of northern part of Pakistan in the fourth century B.C. He accompanied Alexander on his expedition to India, who pointed out that the country above the river Jhelum was subjected to earthquakes, which caused the ground to open up so much, that even the river beds were changed (Ambraseys et al., 1975). An important historical earthquake occurring in northern Pakistan was the destructive earthquake of 25 A.D., which ruined the city of Taxila, to which the intensity of IX-X has been assigned (Ambraseys et al., 1975). The effect of this earthquake still can be seen in the excavated remains of Jandial, Sirkap, and Dharmarajika. The building methods 14
- 15. after this earthquake changed, including reduction in the height of buildings, improvements in masonry bracing density, and making the foundations more secure. On March 25, 1869, a large earthquake occurred in the Hindukush region, strongly felt at Kohat, Peshawar, Lahore, and at Khodjend and Tashkent, the shaking lasting 20 seconds; On May 22, 1871, a damaging shock was recorded at Gilgit with many aftershocks. This earthquake was strong enough to be felt as far as Meerut and Agra in India; On January 20, 1902, a large earthquake caused damage in the Chitral area and was felt widely in the Punjab and up to Simla; On July 8, 1909 an earthquake caused destruction in the region of Mankial and Kalam in the Swat valley where Lady Minot’s Hospital was damaged and many houses collapsed, killing 10 people and cattle. Damage area extends to Dir, Karori and Alipurai and was felt in Gilgit, Besham, and to the north up to Tashkent; and The epicentral intensity of all these earthquakes is estimated to be not greater than VIII on the Modified Mercalli (MM) intensity scale. 4.3 Instrumental Seismicity The instrumental recording of earthquakes started in 1904 but very few seismic stations were established in the South Asian region until the 1960’s. However with the installation of high quality seismographs under the World Wide Standard Seismograph Network (WWSSN) established by the U.S. Coast and Geodetic Survey in 1960, the quality of earthquake recording in this region improved and resulted in a better understanding of the seismicity of Pakistan. In Pakistan and most other parts of the world, the seismic record is too short and incomplete to develop a complete sample that is truly representative of the spatial and temporal distribution of shocks over a large period. Nevertheless, all the available information has been gathered for the period covering the last century, which was used to develop a satisfactory and safe assessment of seismic hazard for the Project. For this study, the instrumental record of earthquakes within about 300-km radius of the Project was searched from available earthquake listing obtained mainly from: International Seismological Centre (ISC) England; National Earthquake Information Centre (NEIC) of the U.S. Geological Survey Pakistan Meteorological Department; PAEC Microseismic Network; and 15
- 16. WAPDA, Micro Seismic Monitoring System (MSMS), Tarbela. A composite catalogue of instrumentally recorded earthquakes was prepared by combining these earthquake listings. This is presented in chronological order showing: Origin time; Epicentral location; Depth of focus; Magnitude; and Data source. In preparing this composite catalogue, more weight was given to the data listed in the ISC catalogue because data within this catalogue tends to be more accurate, being calculated with more data than is used in the other listings, and less likely to contain duplicates. Where available, body wave (mb), surface wave (Ms) or local (ML) magnitudes are also indicated. The source catalogues overlap considerably and both automatic and manual procedures that incorporate judgment about source catalogue reliability and priority were used to help eliminate duplicate entries from the combined listing. During the present study, a composite list of seismic events that occurred in the Project region and adjoining areas has been prepared. This composite list includes events within an area between latitudes: 330 to 370 and longitudes: 700 to 750 . This composite earthquake catalogue of Pproject region is presented in Appendix-B. This catalogue comprises 14547 events covering a period from 1904 to 2017. The reporting agencies have given a variety of magnitudes viz: Body-wave magnitude (mb), Surface-wave magnitude (MS), Richter/Local magnitude (ML) or Duration-magnitude (MD) etc. Since attenuation relationships are based on magnitude of given type, a single type must be selected. For data to be used in seismic hazard analysis, all the magnitudes were therefore converted to moment magnitude (MW) by the following equations. Conversion from MS and mb to MW was achieved through latest equation suggested by Scordilis (2006): MW = 0.67 MS + 2.07 for 3.0< MS < 6.1 MW = 0.99 MS + 0.08 for 6.2< MS < 8.2 MW = 0.85 mb + 1.03 for 3.5< mb < 6.2 For ML up to 5.7, the value of ML was taken equal to MW as suggested by Idriss (1985) and supported by operators of local networks in Pakistan. Conversion of ML to MW beyond magnitude 5.7 was done by using the following equations suggested by Ambraseys and Bommer (1990) and Ambraseys and Bilham (2003): 16
- 17. 0.82 (ML) – 0.58 (MS) = 1.20 Log Mo = 19.09 + MS for MS < 6.2 Log Mo = 15.94 + 1.5 MS for MS > 6.2 MW = (2/3) Log (Mo) – 10.73 Where mb is body–wave magnitude, MS is surface-wave magnitude, ML is local magnitude, MW is moment magnitude and Mo is seismic moment. All available types of magnitudes in the catalogue were converted into a uniform magnitude-scale i.e. MW (Moment magnitude) and given in Appendix- B. MW represents area source rather than a point source and the same type of magnitude is mostly being used in the seismic hazard analysis. 4.4 Kashmir Earthquake of October 08, 2005 A powerful earthquake with a magnitude of MW =7.6 struck the northern part of Pakistan on October 08, 2005 and caused widespread damage in Azad Kashmir and adjoining areas of NWFP. The epicenter of this earthquake was located northeast of Muzaffarabad. This earthquake was felt for several minutes in Pakistan, northern India, and Afghanistan. The heaviest damage was recorded in the towns of Balakot, Batal, and Batagram in NWFP and Muzaffarabad, Bagh and Rawalakot in Azad Kashmir where the entire population was effected. Building collapse was also reported in Mansehra, Abbottabad, and Islamabad. Severe cracks were observed in many high-rise buildings in Islamabad. The death toll due to this earthquake exceeded 80,000 people and millions were rendered homeless due to collapse of houses. The earthquake was followed by a series of more than thousand aftershocks, hundreds of them exceeding magnitude 4. This earthquake was caused by the movement due to rupture along a thrust fault named the “Balakot-Muzaffarabad-Bagh fault” which is a northern most branch of the Main Boundary Thrust (MBT) like Riasi Thrust, a main branch of the MBT in Kashmir. Ground ruptures and fresh landslides have been observed along this fault at many places near Muzaffarabad and Balakot. Teleseismic aftershock data and distribution of damage indicates that more than 120 km of this fault between Batagram and Bagh districts ruptured during the major earthquake. The fault plane solution for the main shock given by Harvard Moment Tensor Solution shows a predominant thrust motion and its strike is compatible with the strike of the HFT. 4.5 Analysis of Seismicity The spatial distribution of seismic events recorded in the project region and given in Appendix-B is plotted on Figure - 5. 17
- 18. The distribution of observed seismicity on the seismicity map clearly shows that the project is located in a region of high seismicity. The concentration of seismicity in the northwest of the Project area is from very highly active Hindukush seismic zone where intermediate to deep earthquakes are more predominant. About 80 % of total earthquakes listed in composite catalogue falls in this zone. Another concentration of earthquakes south of the Project is related to seismically active Indus-Kohistan seismic zone and the Hazara- Kashmir Syntaxis where Kashmir earthquake of October 2005 occurred. The Nanga Parbat-Haramosh syntaxis east of the Project area also shows high seismicity. The Jaglot Syncline area northeast of the Project area, where Hamaran and Darel earthquakes occurred, also shows concentration of seismic activity. A number of small to moderated earthquakes are location around the Project area indicating that Project area is also seismically active. The epicenters of three well-studied earthquakes of magnitude 5.9 or above have been recorded in Kohistan island arc east of the Project area (Ambraseys, et al., 1975; Jackson & Yielding, 1983). These earthquakes are: Patan earthquake (28 December 1974); magnitude (Mb) 5.9; 90 km south of the site; close to the surface expression of MMT; Hamran earthquake (3 September 1972); magnitude (Mb) 6.3; 55 km northeast of the site; within the Kohistan Island Arc; and Darel earthquake (12 September 1981); magnitude (Mb) 6.1; 20 km northeast of the site; within the Kohistan Island Arc. The locations of these events are shown in Figure - 6. While the Patan earthquake (28.12.1974) is located close to the surface expression of the MMT, the Hamran earthquake of 3.9.1972 and Darel earthquake of 12.9.1981 occurred within the Kohistan Island Arc, east of the Project site. This shows that active tectonic features are present within or below the Kohistan Island Arc. Both spatial and temporal clustering or concentrations of seismic activity have been observed in the Project region and is distributed over a large area and has not yet been associated with any known tectonic structure in the area. Previously, the cluster of seismicity north of Darel valley has been associated with the Jaglot syncline but recent geological maps have not shown this syncline as fault associated. Importantly, however, the results of new mapping have shown that the boundaries or contacts of the main lithologic units in Kohistan are faulted. In this respect, it is pointed out that the Darel earthquake occurred close to the northern boundary of the Chilas Complex. 4.6 Focal Depth and Mechanism The reported focal depths of earthquakes included in the composite list range from 0 to more than 300 km. In general, the deeper events are related to Hindukush seismic zone whereas other areas have focal depths less than 100 km. In the Kohistan Island Arc, the depths of most of the earthquakes are generally shallower than 70 km and nominal depth of 33 km is mentioned for majority of these events in all the earthquake catalogues, due to the low 18
- 19. resolution in depth calculation in the absence of a proper recording network in this region. It is important to note that majority of the earthquakes in Kohistan island arc area having magnitude 5 to 6 are located up to about 60 km depth while majority of the events with magnitude greater than 6 remained concentrated in the focal depth less than 50 km. The October 08, 2005 earthquake (M=7.6) had focal depth less than 26 km. The available fault plane solutions of earthquakes in this region show predominantly thrust mechanism. Jackson and Yielding (1983) have reanalyzed the phase data of three prominent earthquakes described above. Fault plane solutions for these earthquakes are presented in Figure - 6. The fault plane solution of Kashmir earthquake of October 08, 2005 is also shown on Figure - 6. Fault plane solutions for these earthquakes all show a thrust source mechanism in keeping with the tectonic model described above involving subduction and underthrusting of the Indian Plate beneath the Eurasian Plate. The northeast to north-northeast dipping planes of these fault plane solutions are possibly representing the causative rupture which is in conformity with the observed northward dips of the major thrusts of the region. 5. SEISMOTECTONIC ANALYSIS From the available tectonic and seismic data of the Project region, an understanding about the seismotectonic set up of the Project can be developed. A seismotectonic map of the Project region showing active faults and recorded seismicity is shown in Figure - 7. 5.1 Identification and Description of Seismic Sources The available seismic and tectonic data provides several evidences of the seismic activity along the major faults i.e. Main Mantle Thrust (MMT) and Kohistan Fault passing south of the site and Main Karakoram Thrust (MKT) passing northwest of the Project. Based on this understanding of the seismotectonic setting and faults of the area, the seismogenic features which may significantly influence the seismic hazard for Gabral Kalam Hydropower Project are: Main Karakoram Thrust (MKT), Kohistan Fault, Main Mantle Thrust (MMT), and Shandur Thrust Main Karakorum Thrust (MKT): This is the major regional fault representing the suture zone between the two colliding plates. This fault represents the northern boundary of the Kohistan island arc and runs eastward to join Indus suture zone in upper Himalayas and terminates at its junction with Karakoram fault. In the Chitral and Gilgit area, the rocks of Karakoram Batholith are thrusted over the rocks of Kohistan Batholith along MKT. 19
- 20. Kohistan Fault: On the Geological Map of NWFP (2006) published by the Geological Survey of Pakistan, the contact between the Kamila amphibolies and the Satpat ultamafics to the south of Dasu are shown as the Kohistan fault. Along this fault, the rocks of the Kamila complex are thrust over the Satpat complex rocks. Main Mantle Thrust: Main Mantle Thrust (MMT) is a northward dipping regional thrust, which separates the Indian Plate from the Kohistan Island Arc. It extends from Khar (Bajaur Agency) in the west to the north of Naran (Kaghan Valley) in the east where it takes a northeast ward bend towards the east of Bunji and gets truncated by the Raikot Fault. The thrust inclines steeply near the surface; however, this inclination is believed to decrease considerably with depth likewise as interpreted for other local thrust faults of the region. The MMT is almost aligned sub-parallel to the Main Karakoram Thrust in the north and to the Main Boundary Thrust in the south except in the Hazara- Kashmir Syntaxial area, where the MMT remains unaffected and continues its journey in a northeast direction to join the Raikot fault. In the east it is abruptly juxtaposed against the Nanga-Parbat-Haramosh Massif, while in the west it meets the Main Karakoram Thrust in Afghanistan. Before joining the Main Karakoram Thrust, it is offset by northwest and northeast trending strike slip faults near Khwaza Khela and Besham. The Patan earthquake of December 28, 1974, having magnitude 6.2, is thought to have been associated with movement on the MMT. The Raikot fault zone and associated structures exhibit remarkable neotectonic features including fault scarps and exposures where Nanga Parbat gneisses overlie Pleistocene tillites. The recent earthquakes of November 2002 and January 2003 have been attributed to movement on this tectonic feature. On the basis of the recorded seismicity and observed neotectonic features both the Main Mantle Thrust and Raikot fault are considered seismically active. Shandur Thrust: Shandur Thrust is marked on GSP Geological Map of Mahudand Quadrangle prepared by Afridi et al (1999) which show this fault at about 10 km in the northwest of the project area. This thrust fault has been marked by the Utror Volcanics group of rocks in southeast while by Kalam Quartz Diorite associated with meta sediments in the northwest. The fault is dipping towards the northwest away from project site and is directed northeast-southwest ward. It is assumed that this fault may coincide with the contact between Kohistan Batholith and Utror Volcanic shown in regional geological maps (Figures – 3 and 4). The inclusion of this fault in the hazard analysis would cover the hazard associated with near-site faults, as lot of observed seismicity in this area could be associated with these faults. Towards the east of the site, Hamran and Darel earthquakes also occurred on 20
- 21. undefined faults. Based on observed seismicity around the Project area, this fault is considered active. 6. SEISMIC HAZARD ANALYSIS For seismic hazard evaluation, both probabilistic and deterministic methods were applied. 6.1 Probabilistic Procedure 6.1.1 PSHA Methodology In probabilistic seismic hazard assessment (PSHA), the seismic activity of seismic source (line or area) is specified by a recurrence relationship, defining the cumulative number of events per year versus the magnitude. Distribution of earthquake is assumed to be uniform within the source zone and independent of time. The principle of the analysis, first developed by Cornell (1968) and later refined by various researchers, is to evaluate at the site of interest the probability of exceedance of a ground motion parameter (e.g. acceleration) due to the occurrence of a strong event around the site. This approach combines the probability of exceedance of the earthquake size (recurrence relationship), and probability on the distance from the epicenter to the site. Each seismic source zone is split into elementary zones at a certain distance from the site. Integration is carried out within each zone by summing the effects of the various elementary source zones taking into account the attenuation effect with distance. Total hazard is finally obtained by adding the influence of various sources. The results are expressed in terms of a ground motion parameter associated with return period (return period is the inverse of the annual frequency of exceedance of a given level of ground motion). The seismic hazard model used in the present analysis was developed based on findings of the seismotectonic synthesis. The seismic hazard model relies upon the concept of seismotectonic zones and does not include linear or discrete fault sources. Each seismic source zone is defined as a zone with homogenous seismic and tectonic features, inferred from geological, tectonic and seismic data. These zones are first defined, and then a maximum earthquake and an earthquake recurrence equation are elaborated for each of these seismic source zones. The seismic parameters attached to the various seismic source zones are: a recurrence relationship relating the number of events for a specific period of time to the magnitude; the maximum earthquake giving an upper bound of potential magnitude in the zone; and an attenuation relationship representing the decrease of acceleration with distance. The probabilistic seismic hazard evaluation requires a detailed analysis of distribution of observed seismic data to the seismic sources, determination of 21
- 22. b-value and activity rate of each seismic source and assigning maximum magnitude potential to each seismic source. 6.1.2 Source Modeling – Area Sources For the definition of seismic sources, either line (i.e. fault) or area sources can be used for source modeling. Because of uncertainty in the epicenters location, it is not possible to relate the recorded earthquakes to the fault sources and to develop recurrence relationship for each fault and use them as exponential model. The Project region was therefore divided into five seismic area source zones (area sources) based on their homogeneous tectonic and seismic characteristics, keeping in view the geology, tectonics, seismicity and fault plane solutions of each area source zone. These seismic area source zones in the northern part of Pakistan are shown in Figure - 8. Each of these area sources was assigned a maximum magnitude based on recorded seismicity and potential of the faults within the zone and a minimum magnitude based on threshold magnitude observed in the magnitude- frequency curve for the zone. As the shallow earthquakes are of more concern to seismic hazard, the minimum depth of the earthquakes is taken as 5 km for all area sources except for deep Hindukush zone where minimum depth was taken as 80 km. The source zone parameters used in probabilistic hazard analysis are given in Table-1. Table - 1 Seismic Area Source Zones Parameters for Probabilistic Analysis Zone No. Seismic Area Source Zone No. of Earthquakes above Min. Magnitude Minimum Magnitude Mw Activity Rate /Year b-Value Maximum Magnitude Mw 1 Hindukush 5177 4.1 92.446 1.00 8.0 2 Karakoram 118 4.0 2.107 1.14 7.5 3 Kohistan 545 4.2 9.732 1.07 7.5 4 Eastern Himalayas 309 4.2 5.518 1.24 8.1 5 Western Himalayas 329 4.2 5.875 1.32 7.0 6.1.3 Earthquake Recurrence Model A general equation that describes earthquake recurrence may be expressed as follows: N (m) = f (m, t) (1) Where N (m) is the number of earthquakes with magnitude equal to or greater than m, and t is time period. The simplest form of equation (1) that has been used in most engineering applications is the well known Richter’s law which states that the cumulated 22
- 23. number of earthquakes occurred in a given period of time can be approximated by the relationship Log N(m) = a – b m (2) Equation (2) assumes spatial and temporal independence of all earthquakes, i.e. it has the properties of a Poisson model. Coefficients ‘a’ and ‘b’ can be derived from seismic data related to the source of interest. Coefficient ‘a’ is related to the total number of events occurred in the source zone and depends on its area, while coefficient ‘b’ represents the coefficient of proportionality between log N (m) and the magnitude. The composite catalogue of earthquakes prepared for the Project region provided the necessary database for the computation of b-value for each seismic area source zone. The composite earthquake list contains limited number of earthquakes prior to 1960 and only few of these earthquakes have been assigned magnitude values. Due to installation of WWSSN, the earthquake recording in this region improved and a better and complete recording of earthquake data are available after 1960. A basic assumption of seismic hazard methodology is that earthquake sources are independent. Thus, catalogues that are used to estimate future seismic activity must be free of dependent events such as foreshocks and aftershocks. To the extent possible such events were also eliminated manually, as there are insufficient data to apply rigorous procedures such as that of Gardner and Knopoff (1974) to eliminate foreshocks and aftershocks from the composite earthquake catalogue. The completeness analysis of the overall data for the region showed that earthquake data around magnitude Mw=4.0 is complete after 1960. The converted moment magnitude for the period between 1961 and 2016 was therefore used in the PSHA after excluding the aftershocks. A separate list of earthquakes occurring in each area source zone was prepared through GIS software and magnitude-frequency curves were made for each seismic area source. The b-value for each seismic area source zone was calculated using linear regression through least square method. The minimum magnitude for each area source zone was selected from the magnitude-frequency curve based on completeness checks suggested by Woeffner and Weimer (2005). The b–values, minimum magnitude and the activity rates for the five seismic area source zones used in the probabilistic analysis are shown in Table-1. 6.1.4 Maximum Magnitude To each seismic area source zone, a maximum magnitude potential was assigned based on the maximum observed seismicity in the historical seismic record and enhancing by 0.5 magnitude the maximum observed magnitude in the instrumental seismic record for that area seismic source zone or determining the maximum magnitude of the longest active fault in the area 23
- 24. using Well & Coppersmith equation (1994). The maximum potential magnitude used for each seismic area source zone is given in Table-1. 6.1.5 Attenuation Relationships Because of lack of sufficient strong–motion data covering a larger range of magnitudes and distances, attenuation relationships for the South Asian Region cannot be developed. A number of attenuation equations have been developed from strong motion data collected in other parts of the world. As shallow earthquakes are of more concern for hazard analysis of the Gabral Kalam Hydropower Project, attenuation equations developed for such conditions were considered for use in the hazard analysis for all seismic area sources except deep Hindukush seismic source. For probabilistic hazard analysis, the latest available NGA equations developed under Pacific Earthquake Engineering Research (PEER) Centre, USA by Abrahamson and Silva (2008), Boore & Atkinson (2008), Campbell & Bozorgnia (2008) and Idriss (2008) were used as these equations are valid for tectonically active regions with shallow crustal faulting worldwide. For Hindukush area source, Youngs et al. (1997) attenuation equation applicable for subduction zone was used. The site foundation condition was assumed to be dense soil with Vs30=600 m/sec as insitu shear wave velocity profile of the site is not available. 6.1.6 Results of PSHA The probabilistic seismic hazard analysis was carried out using EZ-FRISK software developed by Risk Engineering Inc. USA. All the parameters defined in Table-1 were incorporated in the model. The mean total hazard curve was obtained by giving equal weighting to all the attenuation equations used. The total hazard curve obtained for the Project area is shown in Figure - 9. This curve shows the annual frequency of exceedance (inverse of return period) of the peak horizontal ground acceleration expected in the Project area. The major contribution to the total hazard is from Kohistan and Hindukush seismic area sources. The results of PSHA are summarized in Table-2. Table-2 Peak Ground Acceleration for Different Return Periods obtained through PSHA Annual Frequency of Exceedance Return Period (years) Peak Ground Acceleration (g)* 0.007 145 0.21 0.002 475 0.30 0.001 975 0.36 0.0004 2500 0.46 0.0001 10000 0.61 * PGA for very dense soil condition (VS30=600 m/sec) 24
- 25. 6.2 Deterministic Procedure In the deterministic procedure, critical seismogenic sources (active or potentially active faults) that represent a threat to the Project are identified and a maximum magnitude is assigned to each of these faults. The capability of the faults is ascertained through observation of historical and instrumental seismic data and geological criteria such as rupture length – magnitude relationship or fault movement – magnitude relationship. The maximum seismic design parameter is then obtained by considering the most severe combination of maximum magnitude and minimum distance to the Project site, independently of the return period. The main tectonic features around the Project site which could be controlling the maximum earthquake hazard are as follows: Main Karakoram Thrust (MKT), Kohistan Fault, Main Mantle Thrust (MMT), and Shandur Thrust Empirical correlations have been developed between maximum potential of a fault and key fault parameters like rupture length, fault area, fault displacement and slip rate. Out of these fault parameters, only fault lengths are known with sufficient accuracy. For the faults around the site, the half rupture length of the faults has been taken for determination of maximum magnitude potential. The maximum earthquake magnitude (in moment magnitude MW) of each of the fault was calculated using Wells & Coppersmith (1994), Nowroozi (1985) and Slemmons et al. (1982) relationships between fault rupture length and magnitude and is given in Table-3 below. Table-3 Critical Faults and Their Maximum Earthquake Potential Tectonic Feature Total Fault Length (Km) Maximum Magnitude Potential (MW) Selected Maximum magnitude MW Wells & Coppersmith (1994) Nowroozi (1985) Slemmons et al. (1982) Main Karakoram Thrust MKT) 200 7.5 7.5 7.6 7.5 Kohistan Fault 150 7.3 7.3 7.4 7.3 Main Mantle Thrust (MMT) 200 7.5 7.5 7.6 7.5 Shandur Thrust 90 7.0 6.9 7.0 7.0 25
- 26. The peak horizontal ground acceleration at the site caused by the earthquake of maximum magnitude occurring at the closest distance to fault was then calculated by using the latest attenuation relationships developed by various researchers from strong motion data from USA and worldwide. As shallow crustal earthquakes are more important for the assessment of seismic hazard to the Project, therefore equations applicable for shallow crustal earthquakes were employed. For the deterministic analysis, the same four NGA equations used for probabilistic analysis were used. The 50-percentile (median) values of the peak horizontal ground acceleration (PGA) were obtained by four attenuation relationships developed for tectonically similar environments are given in Table-4. The NGA equations are preferable over the older equations for the evaluation of seismic hazard in the near field as these are based on a broad spectrum of data recorded in the near field. For all the seismic sources, thrust rupture mechanism have been assumed. The site foundation condition was assumed as dense gravelly soil with shear wave velocity of Vs30=600 m/sec. Table-4 Peak Horizontal Ground Acceleration (PGA) Tectonic Feature Maxi- mum Magni- tude (MW) Closest Distance to Fault (Km) Median Peak Horizontal Acceleration (g) Abrahamson & Silva (2008) Boore & Atkinson (2008) Campbell & Bozorgnia (2008) Idriss (2008) Main Karakoram Thrust (MKT) 7.5 50 0.10 0.13 0.11 0.10 Kohistan Fault 7.3 45 0.10 0.13 0.11 0.10 Main Mantle Thrust (MMT) 7.5 50 0.10 0.13 0.11 0.10 Shandur Thrust 7.0 10 0.40 0.30 0.38 0.35 7. SELECTION OF SEISMIC DESIGN PARAMETERS 7.1 Definitions According to the ER 1110-2-1806 - Earthquake Design and Evaluation for Civil Works Projects, the definitions of design earthquakes are as follows. 7.1.1 Maximum Credible Earthquake (MCE) 26
- 27. The MCE is defined as the largest earthquake that can reasonably be expected to be generated by a specific source on the basis of seismological and geological evidence. Since a project site may be affected by earthquakes generated by various sources, each with its own fault mechanism, maximum earthquake magnitude, and distance from the site, multiple MCE’s may be defined for the site, each with its own characteristic ground-motion parameters and spectral shape. The MCE is evaluated using DSHA methods informed by results from a PSHA. Since different sources may result in differing spectral characteristics, selection of “maximum” ground motion parameters may need to consider different sources and magnitude events to represent the full range of possible maximum loadings e.g., peak ground acceleration from one source may be higher than from another, but reversed for 1s spectral acceleration values. Therefore, both sources may need to be considered in analysis to assess the full range of potential “maximum” loadings. There is no return period for the MCE. 7.1.2 Maximum Design Earthquake (MDE) The MDE is the maximum level of ground motion for which a structure is designed or evaluated. The associated performance requirement is that the project performs without loss of life or catastrophic failure (such as an uncontrolled release of a reservoir) although severe damage or economic loss may be tolerated. For critical features, the MDE is the same as the MCE. For all other features, the minimum MDE is an event with a 10% probability of exceedance in 100 years (average return period of 975 years) assessed using a PSHA informed by the results of a site-specific DSHA. A shorter or longer return period for non-critical features can be justified by the project team based on the Hazard Potential Classification for Civil Works Projects in Appendix B, Table B-1. A Project with a low hazard potential classification may consider return periods less than 975 years, while projects with a significant or high hazard potential classification may consider longer return periods. The MDE can be characterized as a deterministic or probabilistic event. 7.1.3 Operating Basis Earthquake (OBE) The OBE is an earthquake that can reasonably be expected to occur within the service life of the project, typically a 50% probability of exceedance in 100 years (average return period of 145 years) assessed using a PSHA informed by the results of a site-specific DSHA. The associated performance requirement is that the project functions with little or no damage and without interruption of function. The purpose of the OBE is to protect against economic losses from damage or loss of service, therefore, alternative choices of return periods for the OBE may be based on economic considerations. 7.2 Seismic Design Parameters 27
- 28. Design seismic parameters are selected on the basis of the results provided by probabilistic and deterministic approaches, and in compliance with the recommendations of ER 1110-2-1806 - Earthquake Design and Evaluation for Civil Works Projects. 7.2.1 Maximum Design Earthquake (MDE) Acceleration As per ER 1110-2-1806 - Earthquake Design and Evaluation for Civil Works Projects, Hazard Potential Classification for Civil Works Projects in Appendix B, Table B-1, the Project falls in Low Hazard Potential class. As failure of the project would not present a great social hazard, the designer can choose a Maximum Design Earthquake (MDE) acceleration lower than MCE (which is equivalent to 10,000 year return period earthquake). ER 1110-2-1806 - Earthquake Design and Evaluation for Civil Works Projects recommends to adopt 975 year or less return period ground motion for Low Hazard Potential Hydraulic structures. As the Gabral Kalam Hydropower Project is categorized as Low Hazard Potential Hydraulic structure, therefore for all critical structures of the project, the recommended ground motion for MDE is 0.36g (corresponding to a return period of 975 year). For non-critical structures, the recommended ground motion for MDE is 0.30g (corresponding to a return period of 475 year). 7.2.2 Operating Basis Earthquake (OBE) Acceleration The OBE accelerations are selected from the results of the probabilistic analysis which is presented in Figure- 9 in terms of annual frequency of exceedance of different levels of ground motion. The purpose of the OBE design is to protect against economic losses from damage or loss of service for all project structures. The performance requirement is that the Project functions with little or no damage or interruption under OBE conditions. As per definition of OBE given above, OBE accelerations corresponding to 50% probability of exceedance in 100 years (i.e. a return period of 145 years) may be adopted for which PGA value is 0.21g. 7.2.3 Uniform Response Spectra Uniform hazard spectra generated by EZ-FRISK for OBE (145 year return period) and MDE (975 year return period and 475 year return period) are shown in Figure- 10. 28
- 29. 8. CONCLUSIONS AND RECOMMENDATIONS The conclusions and recommendations regarding study of seismotectonic setting of Gabral Kalam Hydropower Project and the resulting seismic design parameters are as follows: a) The Project is located in the Kohistan Island Arc which is sandwiched between Indian and the Eurasian tectonic plates and very active seismically. b) A number of moderate sized earthquakes have been recorded in Kohistan Island Arc during the last 100 years. c) A number of active faults are present around the project site. d) The main seismotectonic features considered critical for the seismic hazard for the project are as follows: Main Karakoram Thrust (MKT), Kohistan Fault, Main Mantle Thrust (MMT), and Shandur Fault e) Both probabilistic and deterministic seismic hazard evaluations were made to determine the expected ground motions at the project site. f) The recommended horizontal Peak Ground Acceleration (PGA) associated with Operating Basis Earthquake (OBE) is 0.21g. g) The recommended horizontal Peak Ground Acceleration (PGA) associated with Maximum Design Earthquake (MDE) is 0.36 g for all critical structures and 0.30g for all other non-critical structures. h) Uniform hazard spectra for OBE and MDEs given for use in the seismic resistant design of the project structures. i) It is recommended that in-situ shear wave velocity profile of the subsoils at weir and powerhouse sites may be obtained for authenticating the assumption of Vs30. j) For safety monitoring purpose Strong Motion Accelerographs may be installed at the weir and Power House site. REFERENCES 29
- 30. 1. International Commission on Large Dams (ICOLD) (2016); Bulletin 148, Guidelines for selecting Seismic Parameters for Large Dams, Paris. 2. Tahirkheli and Jan, (1979); A preliminary geological map of Kohistan and the adjoining areas, N. Pakistan; Geol. Bull. Univ. Peshawar, 11. 3. Treloar et al., (1990); Cooling and uplift histories of the crystalline stack of Indian plate internal zones west of Nanga Parbat, Pakistan Himalaya; Tectonophysics, 180, 323-349. 4. Khan et al., (1997); Geochemical and isotopic constraints on subduction polarity, magma sources, and paleogeography of the Kohistan intra-oceanic arc, northern Pakistan Himalayas; Journal of the Geological Society of London, 154, 935-946. 5. Coward et al., (1986); Collision Tectonics in the NW Himalayas. In Coward, M.P. & Rie, A.C. (eds), Collision Tectonics; Geol. Soc. Lond. Spec. Publ. 19:203-219. 6. Gansser A., (1964); Geology of the Himalayas, Willey Interscience. 7. Ghazanfar M., M. Nawaz Ch. & M. Shahid Hussain (1991); Geology and petrotectonics of southeast Kohistan, Northwest Himalayas, Pakistan; Kashmir Journal of Geology, Vol. 8 & 9. 8. Jan M. Q. and Howie R. A., (1981); The mineralogy and geochemistry of the metamorphosed basis and ultrabasic rocks of the Jijal Complex, Kohistan, NW Pakistan; Jour. Petrol. Vol. 22. 9. Kazmi & Qasim Jan (1997); Geology and Tectonics of Pakistan; Graphic Publishers. 10. Madin, I. P., Lawrence, R. D. and Ur-Rehman, S., (1989); The northwest Nanga Parbat-Haramosh Massif; evidence for crustal uplift at the northwestern corner of the Indian craton, Geological Society of America Special Paper, 232. 11. Scotese et al., (1988); Plate tectonic reconstructions of the Cretaceous and Cenozoic ocean basins; Tectonophysics, 155, 27-48. 12. Zanchi et al., (2000); Mantle exhumation along Tirch Mir fault zone, NW Pakistan: pre-mid-Cretaceous accretion of the Karakoram terrane to the Asian margin; Geological Society of London, Special Publication, 170, 277-293. 13. Hildebrand et al., (2001); An old origin for an active mountain range: geology and geochronology of the eastern Hindu Kush, northwest Pakistan; Geol. Soc. Am. Bull, No. 113, 625-639. 14. Khan et al., (1993); Evolution of the lower arc crust in Kohistan, N. Pakistan: temporal arc magmatism through early, mature and intra-arc rift stages; In: Himalayan Tectonics (P.J. Treloar and M.P. Searle, eds). 15. Searle et al., (1999); The tectonic evolution of the Kohistan-Karakoram collision belt along Karakoram Highway transect, north Pakistan; Tectonics 18, 929-949. 16. Malinconico, (1989); Crustal thickness estimates for the western Himalaya. In Malinconico, L.L. & Lillie, R.J. (eds), Tectonics of the Western Himalayas; Geol. Soc. Amer., Spec. Paper 232: 237-242 17. Seeber and Armbuster, (1979); Seismicity of the Hazara arc in Northern Pakistan: decollement vs. basement faulting. In: 30
- 31. Geodynamics of Pakistan (A. Farah & K.A. DeJong, eds.); Geol. Surv. North Pakistan. J. Metam. Geol. 7, 111-125. 18. Fineti et al., (1979); The Pakistani segment of the DSS-profile Nanga Parbat – Karakul (1974-1975); Boll. Geofis. Teorica ed Applicata, 21: 159-169 19. Miller et al., (1991); Platinum-group element mineralization in the Jijal layered ultramafic-mafic complex; Pakistan Himalaya. Econ. Geol., 86: 1093-1102. 20. Pivnik D. A. & Wells N. A., (1996); The transition from Tethys to the Himalaya as recorded in northwest Pakistan; Bull. Geol. Soc. Amer., Vol. 108. 21. Seeber et al., (1981); Seismicity and continental subduction in the Himalayan arc, in Zagros-Hindukush-Himalayas Geodynamic Evolution; A.G.U. Geodynamic Series, Vol. 3. 22. Oldham, (1893); A catalogue of Indian Earthquakes; Mem. Geol. Survey India, Vol. 19. 23. Heukroth and Karim, (1970); Earthquake history, seismicity and tectonics of the regions of Afghanistan; Seism. Centre, Kabul University. 24. Ambraseys et al., (1975); The Patan Earthquake of 28 December 1974; UNESCO Publication. 25. Quittmeyer and Jacob, (1979); Historical and modern seismicity of Pakistan, Afghanistan, northwestern India and southeastern Iran; Bull. Seismological Society of America (BSSA), Vol. 69, No. 3. 26. E. M. Scordilis, (2006); Empirical global relations converting Ms and mb to moment magnitude, Journal of Seismology. 27. Idriss I. M., (1985); Evaluating seismic risk in engineering practice, Proceedings of the 11th International Conference on Soil Mechanics and Foundation Engineering, San Francisco. 28. Ambraseys, N., Bommer, J., (1990); Uniform magnitude re-evaluation for the strong-motion database of Europe and adjacent areas, European Earthquake Engg, Vol. IV. 29. Ambraseys N., and Bilham R., (2003); Earthquakes in Afghanistan, Seismological Research Letters, Vol. 74 No.2. 30. Jackson, J.A.; Yielding, G., (1983) The Seismicity of Kohistan: Source Parameters of the Hamran (1972.9.3), Darel (1981.9.12) and Patan (1974.12.28) Earthquakes. In Tectonophysics 91: 15-29. 31. Cornell C. A. (1968); Engineering seismic risk analysis, Bull. Seism. Soc. Am.,Vol. 58, No.5 (1968). 32. Gardner J. K. and Knopoff L., (1974); Is the sequence of earthquakes in southern California, with aftershocks removed, Poissonian?, Bulletin Seismological Society of America, Vol. 64, No. 5. 33. Woessner J. and S. Weimer (2005); Assessing the quality of earthquake catalogue: Estimating the magnitude of completeness and its uncertainty, Bulletin Seismological Society of America, Vol. 95 No.2. 34. Wells & Coppersmith (1994); New empirical relationships among magnitude, rupture length, rupture width, rupture area and surface displacement; B.S.S.A., Vol. 84, No.4. 35. Abrahamson, N. and Silva, W., (2008); Summary of the Abrahamson and Silva NGA Ground-motion relations; Earthquake Spectra, Vol. 24, 31
- 32. No.1. 36. Boore, D. M. and Atkinson, G. M., (2008); Ground motion prediction equations for the average horizontal component of PGA, PGV and 5% damped PSA at spectral periods between 0.01s and 10.0 s; Earthquake Spectra, Vol. 24, No.1. 37. Campbell K. W. & Bozorgnia Y., (2008); NGA ground motion model for the geometric mean horizontal component of PGA, PGV, PGD and 5% damped linear Elastic response spectra for periods ranging from 0.01 to 10 seconds; Earthquake Spectra, Vol. 24, No.1. 38. Idriss I. M., (2008); An NGA empirical model for estimating the horizontal spectra values generated by shallow crustal earthquakes, Earthquake Spectra, Vo; 24, No.1. 39. Youngs R. R., Chiou S. J., Silva W. J., Humphrey J. R., (1997); Strong ground motion attenuation relationships for subduction zone earthquakes, Seismological Research Letters, Vol. 68, No. 1. 40. Nowroozi (1985); Empirical relations between magnitudes and fault parameters for earthquakes in Iran; B.S.S.A., Vol. 75, No.5. 41. Slemmons et al., (1982); Evaluation of active faulting and associated hazards, in Studies in geophysics – active tectonics; National Academy Press, Washington, DC. 32