HYDROLOGIC CONSIDERATIONS IN SAFETY OF DAM

Abstract

This dissertation report is based on review of guidelines for the hydrologic evaluation to determine the safety of dams and allied structures. Focus is on hydrological aspects only. Many of the older dams are now characterized by increased hazard potential due to developments in flood plain and increased risk due to structural deterioration or inadequate spillway capacity of the dam. The Government of India has constituted Dam Safety Organization in the Central Water Commission during June 1979 Guidelines issued by Dam Safety Organization are reviewed. There have been 26 major dam failures in India in post independence period. Further, literature review of flood estimates of 62 large dams shows that reassessed design floods are significantly larger than earlier estimates. With occurrence of more severe events in a large sample, earlier estimate of design flood are bound to be revised up ward as illustrated through flood reviews of dams in India. Subjectivity in estimation of PMP and PMF. should be minimized by evolving consensus and coding the criteria, and procedures for estimation of design flood. When various storms are considered for development of Unit Hydrograph for the same catchment a marked variation is observed in the peak as well as the time of occurrence of the peak Therefore average Unit Hydrograph needs to be derived giving higher weightage for the Unit Hydrograph derived from severe storms. Different unit hydrographs should be identified for the various conditions which have major influence on formation and time distribution of the runoff. These unit hydrograph may then be judiciously applied under different conditions. Recent flood hydrographs should be used for derivation of UH. Changes in land use, land cover over the catchment area should be and should be duly considered in the analysis. Errors in forecasting f oods in term of estimation of peak discharge and time to peak due to use of different UH when rainfall is not uniform and when rainfall is assumed to be uniform over the catchment are significant as shown by case study of Baitarni basin(India). Effect of various factors on design flood estimation is analysed through case study of floods at Bhakra dam site on river Sutlej in India DIfferent probability distributions provide significantly different estimates e.g. using 1909-1992 data series of Bhakra dam, 10000 year estimate by EVI is 21, 036.26 m3/s and by PHI is 26,154.53 m3/s (24.33 % higher). Using same probability distribution but different samples from same population also result in significantly different estimates. 10000 year flood estimate using LN II probability distribution are 18,732.75 cumec (1909-92 data series), 15, 064.06. cumec (1909-59 data series) and 24, 588.81 m3/s (1960-92 data series).