DEVELOPMENT OF AN OPERATIONAL INFLOW FORECASTING SYSTEM FOR TEHRI DAM

Abstract

Tehri is a multipurpose project, located at 145 km downstream of Gangotri glacier on river Bhagirathi. Tehri project is important for power generation, irrigation and flood control. Tehri reservoir is a major source of water for Rabi crop in Uttarakhand and Uttar Pradesh. It has increased 200,000 ha command area in Uttarakhand and Uttar Pradesh and strengthened the intensity of irrigation in 670,000 ha of the command area. Dam is also a major source of drinking water for 7 million people. Tehri dam has got good flood space and has played a very vital role in moderating the 2010 and 2013 floods of Uttarakhand. The present study has been taken up with the aim of developing an operational inflow forecasting system for Tehri Dam. This in turn required, identification and setting up of real-time hydro-meteorological network in the basin, and development of mathematical flow forecasting models for the sub-catchments contributing to Tehri Reservoir on 6 hourly and daily bases. The thesis has been arranged in seven chapters. The broad objective and need of inflow forecasting system for Tehri dam are explained in Chapter 1. The details of the study area and the physiographic analysis for the sub-catchments are presented in Chapter 2. The identification of the hydro-meteorological sites, earth station, specifications of the instruments, data transmission details and the website development is explained in Chapter 3. The review of literature for the inflow forecasting models has been presented in Chapter 4. The inflow forecasting models have been developed using HEC-HMS, GIUH based Nash models and stochastic models. The details of the inflow forecasting models are presented in Chapter 5. The results are presented in Chapter 6. Finally, Chapter 7 presents the conclusions drawn from the study, limitations and the scope for further improvement. Automatic weather stations have been installed at eleven locations in the catchment of Tehri dam. These stations are operational since June 1, 2016. The data of June 1, 2016, to Dec. 30, 2018, have been used in the study. The following steps have been adopted for the development of mathematical models: i. Development of the GIUH based Nash model for 16 ungauged tributaries; ii. Event-based modelling of Tehri catchment including modelling of snow-fed sub-catchments using HEC-HMS; iv iii. Continuous flow modelling of Tehri catchment using HEC-HMS and iv. Development of stochastic models for 6 hourly forecasts during monsoon season and daily forecast during the non-monsoon season. The conclusions drawn from the study are: 1. The indigenous system installed for the Tehri dam is performing satisfactorily June 1, 2016. This type of system can be developed and applied for the other storage hydropower projects in the country. Transmission of data through GSM/ GPRS is having less administrative formalities. The GSM/GPRS data transmission is cost effective, and its space requirement is much lesser as compared to the VSAT / INSAT communication system. GSM/GPRS based system is working satisfactorily for data transmission. 2. Autoregressive and Autoregressive models with exogenous inputs have performed very well for all the sites of Tehri catchment. For the forecasting of monsoon flows with 6 hours lead time ARX (1,1) model has performed very well with NSE more than 82% at Tehri dam. Reservoir levels were forecasted 78% of the time within the range of + 10 cm accuracy in 6 hourly forecasting during monsoon season. In one- day advance forecasting during the non-monsoon season, 47 % of the forecasts are within the range of + 5 cm accuracy without updating of parameters. With updating of parameters of the model, these models performed far better. During the period 18.11.2018 to 30.12.2018 more than 90% of the forecasts are within the range of + 5 cm accuracy. 3. Stochastic models are easy to use and require fewer data. These models have performed very well in operation inflow forecasting system for Tehri dam. The Operational inflow forecasting system for the Tehri dam is operational now. However, it needs to be further improved by strengthening the hydro-meteorological data network in the system, minimization of manual intervention in the system, strengthening of the data transmission system by introducing VSAT data transmission at some of the stations and integration of automatic gauge of Tehri reservoir with earth station of the inflow forecasting system. The mathematical models developed in the study also need to be further improved to increase the lead time by incorporating forecasted rainfall.