SIMULATION OF JUNE 2013 FLOOD AT TEHRI DAM

Abstract

Tehri Dam is located in River Bhagirathi, around 1.5 km downstream of confluence point between Bhagirathi and Bhilangana in Tehri Garhwal District of the state Uttarakhand. The height of the earth and rock-fill dam is 260.5 m from its deepest foundation level. The main purpose behind the construction of high dam is to collect the large volume of water during the flash floods and use them for power generation, supply drinking water and irrigation in a controlled manner throughout the year. The average water availability at Tehri dam is 8000.0 MCM out of which about 39% comes from melting of glaciers in the Himalayas where as the rest 61% is trapped from the flash floods during the short monsoon season. The design flood discharge adopted for the spillways design of Tehri dam is 15,540.0m3ls. Recently on 171h June 2013, the maximum flood discharge measured at the dam axis was 7535.0 m3/s, which is nearly half of the design discharge. In the present study, June 2013 flood at Tehri dam is analyzed and simulated by different geomorphologic approaches in order to develop a flood estimating and forecasting model for Tehri dam for similar floods in future. For the study purpose the whole catchment is divided into three major parts on the basis of catchment physiography. The Geomorphological IU1-1, Geomorphoclimatic IUH and GIULINash model are used for the simulation of 16th , 17th and 18th June 2013 flood. The resultant flood hydrographs from each model is compared with the observed hydrographs. Similarly the flood volume retained in the Tehri reservoir during the period is compared with the simulated volume from different model. The observed peak flood discharge is 7535.0m3/s and the simulated discharge from GIUH, GcIUH and GIUH-Nash model are 6856.0m3/s, 6980.0m3/s and 6901.5m3/s respectively. The gross inflow volume retained in the Tehri reservoir is 748.9 MCM and the simulated volume by GIUH, GcIUH and GIUH-Nash model work out to be 723.8 MCM, 733.5 MCM and 735.6 MCM respectively. The Nash-Sutcliff efficiency for GIUH, GcIUI-1 and GIUHNash models are 75.8%, 79.1% and 80.5% respectively. The results of GIUH-Nash model interms of model efficiency, percentage error in time to peak and percentage error in observed volume are better than that of GcIUH and GIUH model.