SEISMIC MICROZONATION OF AN URBAN HABITAT

Abstract

The rapid growth of the world’s population over the past few decades has led to a concentration of peoples, buildings and infrastructure in urban areas. Such urban development in the sedimentary plains has increased their vulnerability to earthquakes, due to the presence of soft sediments. The local site effects due to soil characteristics (specifically the shear wave velocities) play an important role in modifying the strong ground motion. For the quantification of such amplification a thorough knowledge of the type of local subsurface soil in a region, is thus essential for a holistic seismic microzonation of the region. Seismic microzonation is a procedure for estimating the total seismic hazard from ground shaking and related phenomena by taking into account the effects of local site conditions. It involves the integration of geologic, seismologic and geotechnical apprehensions for sound land-use planning for earthquake effects so that planners and structural engineers can site and design structures that would be relatively safe from damage during major earthquakes. Such exercise becomes necessary for those areas lying on thick alluvium covers as the strong ground motion is further amplified due to local site effects. Himalayas, one of the seismically active belts of the world, are encompassed by thick alluvium cover from the south where such affects are multiplied due to high seismicity and amplification due to such thick alluvium. Roorkee, an important town in the Indo-Gangetic alluvial plains at the foothills of the Garhwal sub-Himalayan range lying in seismic Zone IV, is vulnerable to strong ground motion on account of high seismicity and amplifications due to local site conditions. It is in this context that the present research work has been carried out to develop methodologies to carry out seismic microzonation of an urban habitat which includes high probabilities of strong ground motion exceedance and local site conditions such as thick alluvium, which include deep soil effects. The main objective of this study has been to carry out seismic microzonation of an urban habitat. A case study has been undertaken for Roorkee which falls in seismic zone IV and is located on thick alluvium of Indo-Gangetic Plains. The study has been carried out in various steps which also define various objectives of the study such as compilation of ii seismicity and geology of the study area and to develop a seismo-tectonic model for the region, predict strong ground motion for various exceedance probabilities, characterise the soil under the study area using various geophysical techniques, estimate the transfer functions and the ground motion at different levels of subterranean soil using 1D wave propagation techniques for the soil classes in the study area, carry out seismic microzonation for the urban habitat viz., Roorkee, and suggest various user specific seismic microzones based on the time period of vibration of various types of buildings in Roorkee municipal limits, and finally recommend improvements in existing seismic microzonation methodologies and their application for seismic microzonation of an urban habitat. The selection of study area was motivated by three factors namely strategic importance, seismicity and local site characteristics. Keeping in view the three factors Roorkee has been chosen as the study area. The area falls in seismic zone IV and is about 20 km from the Himalayan Frontal Fault. It has experienced many damaging earthquakes in the past. Seismotectonically, Roorkee lies on seismogenic feature namely, Delhi-Hardwar ridge which is a linear NE-trending ridge with western boundary defined by Dehradun-Mahendragarh fault. The location of Roorkee is on thick alluvium making it more vulnerable due to amplified motion from seismogenic sources already with high hazard values. The seismic microzonation of Roorkee city has been carried out by combining the existing methodologies and using a novel approach for soil characterisation and response analysis. A two pronged approach has been developed which include the estimation of strong ground motion up to engineering bed rock using GMPEs and the transfer function up to surface using 1-D response analysis (not merely the VS30 methodology). The seismic hazard has been estimated using the probabilistic method for which the seismotectonic model has been developed based on seismicity, geology and prevailing tectonics of the area. The area is divided into seven seismogenic sources and the parameters are estimated to calculate the seismic hazard in Roorkee. The seismic hazard has been assessed by considering the size uncertainties using doubly truncated Gutenberg relationship, whereas for the temporal uncertainties, Poissonian distribution has been used to model earthquake recurrence. It is well known from many earlier studies that the uncertainties in the wave attenuation models contribute significantly to the absolute hazard level and to the total uncertainty in iii the seismic hazard estimates. Based on the available strong motion records, similarity of seismotectonics of the two areas, spectral acceleration requirements, limited parameters availability of independent parameters and the literature review, two attenuation relationships have been selected and used in the present study. The attenuation relationships used in the present study are given by Boore and Atkinson (2008) and Abrahamson and Silva, (1997). The effects of all earthquakes of different sizes, occurring at different locations within different earthquake sources, and having various probabilities of occurrence are integrated into a single uniform seismic-hazard curve that shows the probabilities of exceeding different levels of ground shaking in the city of Roorkee during a specified period of time. The estimation of spectral ground motion has been carried out for three types of soils with shear wave velocities of 1300 m/s (bed rock), 700 m/s (engineering bed rock) and 200 m/s (surface) using GMPEs. The strong motion prediction has been carried out for five return periods (100, 225,475, 2500 and 10000 years) and for ten spectral periods (0.03, 0.05, 0.1, 0.2, 0.3, 0.5, 1, 2, 3, and 4 seconds). Further, the thick alluvium below Roorkee has been explored using various geophysical methodologies (non destructive) for its characterisation considering the soil column below to have three layers over half space. While the upper layers have been characterised using the geophysical techniques, the deeper layers have been characterised using the available data on deep seismic profiling, deep bore logs and the geological sections. Soil characterization of the upper layer has been carried out using (i) Multispectral analysis of surface waves (MASW) using 24 geophones at intervals of 2/4 metres with a 60 kg hammer, (ii) MASW using 10 geophones at intervals of 2/4 metres with a 10 kilogramme hammer and (iii) Micro-tremor analysis using ambient noise. The MASW survey was carried out in about 18 months of time sporadically at 60 sites for quantifying the soil profiles in Roorkee. Further, the MASW with 10 geophones was used along with the microtremor equipment so as to increase the number of points on low frequencies on the phase velocity dispersion curve. This has the advantage of interpreting and computing the shear wave velocities using inverse solutions where the information from MASW and microtremor is used simultaneously, which in turn has increased the depth of information due to low frequencies present in the dispersion curve. The general soil depths below Roorkee has been explored the help of geological information and the bore logs available combining with the MASW and microtremor data to estimate the deep soil information. Based on the information from experimental data, data available from earlier records and iv the geological data from deep seismic sounding (GSI, 2001), the soil in Roorkee has been divided into three categories. The detailed soil characteristics of the three zones of Roorkee have been used for estimating the strong ground motion at various levels and on the surface using 1-D wave propagation technique. The soil information including number of layers, material, thickness (m), Unit Weight (kg/m3), Gmax (kgf/m2) and Vs (m/sec) have been worked out using the above procedure and validated using the empirical relationships and the published work in this area by other workers. The soil data (borehole data) required for the study was obtained from different agencies that are actively involved in conducting soil investigations of the region. This study helped in locating the bedrock and in determination of fundamental resonant frequency at different sites. The soil information in terms of shear wave velocities, densities and the thicknesses of different layers at the sites were used to obtain the ground motion at the surface in terms of spectra and spectrum compatible time histories. The soil characterised for Roorkee (three areas) have been used to estimate the strong ground motion at the surface. De-aggregation was performed to estimate the duration of the time history and spectrum compatible time histories were generated accordingly for their further use by 1-D wave propagation technique. The strong ground motion at the surface was estimated using the 1-D wave propagation analysis, in which the engineering bedrock motion that was estimated using the GMPEs was given as the input to ascertain the strong ground motion at the surface. For this purpose, the frequency domain approach has been used to solve the wave equations for ground motion response in 1-D wave propagation by considering the equivalent linear model (Pro-Shake). The results obtained using 1-D wave propagation analyses have been compared with the strong ground motion prediction using GMPEs. It has been observed that strong ground motion at the surface as obtained from GMPEs is on higher side as compared to the same obtained from 1-D wave propagation analysis. Further, it is observed that the pattern of strong ground motion is governed by the local soil conditions wherein the NW-SE trend of the strong ground motion contours obtained from the GMPEs has been converted into five segmented microzones, when 1-D wave propagation analysis is used. This demonstrates the amplification effect of local soil on strong ground motion, which has to be taken into account during subsequent microzoning. The spectral strong ground motion estimated at the surface has been further subdivided into microzones for various conditions of ground motion exceedance. v Seismic microzonation maps have been finally prepared for Roorkee city for different levels of strong ground motion. Roorkee contains mostly single storied to three storied buildings. Recently, a few multi-storied buildings have also come up. While the buildings in the old city on the western side of Ganges canal are very old non-engineered construction, the recently constructed buildings in the new city on the eastern side of the canal also do not conform to Indian Standard Codes for earthquake resistant design, except for a few high rise buildings that have come up in the last decade. Thus, the bulk of the city is vulnerable to damage and requires immediate retrofitting measures based on the strong ground motion recommended in the present study. Thematic maps have been prepared taking into account the vibration periods of various types of buildings in Roorkee. Further, the design spectra up to 4 seconds of vibration periods have been proposed for the city for different conditions of exceedance of strong ground motion. The exceedance probabilities include the various conditions for earthquake engineering use viz., Maximum Credible Earthquake and Design Basis Earthquake conditions. The local site effects have been found to be dominating in changing the scenario of strong motion at the surface in comparison to the bed rock motion. While the peak ground acceleration determined at bed rock using GMPEs, varies between 0.14g to 0.17 g for 475 years return period with the values increasing from the South-West to North-East, due to the prevailing seismogenic features, the ground motion estimated at the surface using the 1-D wave propagation method varies in a radial pattern, between 0.11g and 0.12g on account of the local soil conditions in the city. The following conclusions have been drawn based on the present study: 1. The various seismic microzonation procedures have been studied and an improved seismic microzonation procedure has been proposed for urban habitats. 2. The seismic hazard assessment for Roorkee has been carried out for various conditions (five return periods and ten spectral periods at surface, engineering bed rock and bed rock). The conspicuous hazard assessed for Roorkee shows that PGA at the surface varies from 0.06-0.07g, 0.09-0.10g, 0.11-0.12 g, 0.18 g - 0.20 g, 0.28-0.31g for 40%, 20%, 10%, 2% and 1% exceedance in 50 years, respectively. 3. The thick alluvium below Roorkee, considered as case study for urban habitat, quantified using geophysical, geological and geotechnical methodologies shows that bed rock is at a depth of 3000m, above which engineering bed rock comes up to a depth of 50 m which is overlaid by soft soil cover having a thickness of about 50 m. vi 4. The comparison of the ground motion prediction using GMPEs having inputs for shear wave velocities versus 1-D response analysis has shown that the latter is inescapable on account of the non linear behaviour of soil and the fact that the GMPEs are empirical in nature. 5. The pattern of microzones is governed by the local site conditions. Further, the GMPEs are giving relatively higher strong ground motion compared to the motion being estimated using 1D analysis. For example the NW-SE trend of strong ground motion values in Roorkee, are different than the results of site specific 1-D analysis, viz., the NW-SE trend has got converted to five segmented microzones, which demonstrate the amplification effect of local soil on strong ground motion. In addition to recommending a better seismic microzonation approach, the main contribution of the present research work is the development of realistic, usable thematic microzone maps, which are the basic inputs for earthquake resistant design of individual structures, and are also the basic documents for land use planning in the city, where the spectral accelerations with definite time periods of recurrence govern land use. Such a product has a direct and critical implication in the strategies being formulated by the local authorities from time to time, for seismic risk assessment and ultimately disaster mitigation and management.