Prof.T.Shivaji Rao,

Director, Center for Environmental Studies, GITAM University, Visakhapatam-45




On 30-4-1983 Padmabhushan   Dr.K.L.Rao, an International expert on irrigation Engineering ,a member of the Central Water Commission and a Union Minster for Irrigation and Power gave a special Interview to the correspondents of Indian Express and Andhra Prabha daily news papers at Vijayawada that the Polavaram dam design is highly defective and will not work and hence there is no question of transfer of Godavari water into Krishna basin. He gave a detailed reasoning for this purpose .

“Dr.K.L.Rao ruled out the possibility of diverting surplus Godavari waters to the Krishna owing to defective designing of the Polavaram project.  Only 1800ft. spillway was provided into Polavaram project to clear 40 lakh cusecs of flood waters in the Godavari as against 13000ft. long Dowlaiswaram anicut designed by Sir Arthor Cotton.  Even Prakasam barrage was designed to 6,280ft. long though the flood water would not be  more than 12 lakh cusecs.  Dr.Rao  said it was simple Arithmetic to understand that the Polavaram design would not work he said”

During 1984-85 when the author was  working as an expert committee member of the AP government Irrigation Department entrusted with the preparation of the Environmental Impact Assessment report for Polavaram project, he frequently discussed the crucial issues pertaining to the feasibility of this project with some of the Chief Engineers who often interacted with Dr.K.L.Rao on the technical problems of this project.  Dr.K.L.Rao toured several countries to visit the major irrigation projects and made detailed notes about the design ,operation and safety aspects of the projects.  Dr.K.L.Rao it appears used to exhort the engineers to be careful in designing the spillways so that the failure of dams due to under-estimation of peak floods should not result in collapse of the dams as they occurred at Kadam  dam in Godavari basin, Machchu dam in Gujarat and other dams in India,China and USA. 

Dr.K.L.Rao used to narrate that the magnitude of extreme floods generally increase year after year due to growing deforestation and development of agriculture, industries and human habitations in the catchment areas of rivers. For instance in the case of Godavari the peak flood rose from 15 lakh cusecs in 1850’s to 21 lakh cusecs in 1940s to 30 lakh cusecs in 1953 and 35 lakhs in 1986.  Consequently he felt that liberal provisions must be made for spillway to discharge the increasing levels of peak floods also due to sudden releases of flood  waters from dams in the upstream areas during cloud bursts consequent to intense rains and cyclones.

Some engineers make appropriate forecasts of the probability of failure of earthen dam by using simple algebra.  For instance the storage capacity of the dam is assumed as ‘C’ and ‘G’ is the rate at which water flows out of the dam as per the number and size of the flood gates in the dam.  The unutilized capacity of the dam is ‘E’ and the normal stored water is ‘A’ and so C=A+E.   A cyclonic storm produces peak flood ‘F’ at dam site and the duration of the storm flow is ‘D’ .  The rate at which the dam gets filled up = (F-G). Hence the time taken by the flood to fill up the unutilized capacity of the reservoir = E/(F-G).  If this time of filling is greater than the duration of the flood then all the flood water will be safely stored in the reservoir.  But if the duration of the flood is greater than the time of filling of the unutilized capacity of the reservoir than the flood water received into the reservoir will flow over the earth and rockfill dam resulting in erosion and breaching that causes a catastrophic failure.  The flood flow below the dam jumps suddenly from G to F+C/T where ‘T’ is the time it takes for the emptying of the water in the reservoir. 

Thus in the case of the Polavaram dam the following preliminary engineering calculations  show how and when the dam will collapse due to peak flood estimates of 49 lakhs cusecs by CWC and 93 lakhs cusecs estimated  by Shri.T.Hanumantha Rao, Ex.Engineer-in-Chief  of AP State Government)

A =  Polavaram reservsoir storage            =  200 TMC  = 5600 x106m3

G =  Spillway discharge                                  = 1.0 x 105 cumecs (0.1 million cumecs)

E = Unutilised storage of the reservoir   = 235 TMC = 6674 x 106m3

        above FRL

D = Duration of intense storm                    = 48 hours

F = inflow design flood                                  = 49 lakh cusecs = 1.5 x 105cumecs  (as per CWC estimate)           

F-G  = incremental increase in storage    = 1.4 x 105 – 1.0 x 105 cumecs

Hence the time for filling up of the          = 6674 x 106m3    =16,685 seconds = 46 hours.

Unutilized storage                                              0.4 x 106 cumecs           


Hence if the duration of the storm is more than about 46  hours the flood water will flow over the earthen dam resulting in its collapse.  The above calculations are approximate only and they can be improved by taking up elaborate flood routing exercise and computer modeling.

But if the inflow flood, based on the extreme flood of 26.5 lakhs cusecs of October 2009 at Srisailam against its  100-year flood of 9.4 lakh cusecs is taken into consideration, the extreme flood into Polavaram dam project  could be 33 x 26.5/9.4  = 93 lakh cusecs ( 2.7 lakh cumecs as estimated by Sri.T.Hanumantha Rao.

 In such a case, time of filling = 6674 x 106m3  /1.7 x 105 = hours

Hence, if the duration of an intense cyclone, were to excess  10 hours, Polavaram dam is bound to collapse.Risk = (Probability of an accident) X (costs of damage due to the accident)

Role of  Meteorologists and Hydrologists in Design of Big Dams : 

Besides  the Civil Engineers, Meteorologists, Hydrologists and Environmentalists have to play a crucial role in the design of modern major dams as per the new generation of dam design standards of U.S.A.  The Civil Engineers when designing a dam must get the latest data on Probable Maximum Precipitation (PMP)from the Meteorological experts and the Probable Maximum Flood must be determined by the Hydrological experts. The Civil Engineers must determine the capacity of the dam and the spillway discharge rate through  flood gates to ensure not only safety of the dam but also the safety of the people and their properties downstream due to an anticipated maximum credible accident for the one reason or the other.  Flood gates being an expensive component of the construction of a dam, the civil engineers  naturally consider  a trade of between the cost of the dam and the economic benefits and the degree of safety it will provide to the people and their properties.  The dam design must be based on the probability that an intense cyclonic storm will cause the dam to over flow and consequently the floods destroy the dam inspite of the spillway provided for the design flood.  If this probability is 0.01 then the dam is said to safely handle any small flood upto a 100-year flood.  Sometimes the cyclonic weather conditions that produce one severe storm may persist and produce another severe storm in a sequence.  Thus the policy of operation of the dam becomes a critical factor in estimating the probability of a catastrophic failure floods cause catastrophic failure of a dam because the water stored behind the dam will be suddenly added virtually to the natural extreme flood which becomes an extra burden to be contained by the ill prepared flood banks downstream of the dam.  The failure of one dam in a river basin sometimes lead to the failure of other dams downstream and the flood effect will be cumulative to the detriment of the people and their properties in the vulnerable areas.

Historical Failure of Dams:    

An expert Robert Jensen estimated that since 12th century 2000 dams in the world failed including failure of 200 major dams between 1900 and 1980.  On an average 30 dam failures occurred per decade.  According to International Commission on Large Dams (ICOLD) 2.2% of all the failure rate dropped to 0.5%.  According to another expert Blied the failure rate of dams between 1900 and 1969 is 2.4% for small dams and 1.7% for large dams.  It is estimated that the average global risk of any dam failing in any given year is 1 in 10,000.  Experts estimated that 40% of the dam failures due to overtopping while 30% failures are due to foundation failures.  In China 2300 out of 80,000 failed since 1950 taking the failure rate to 4%. 

Cases of catastrophic dam failures in China in 1975:  

When politicians, officials and contractors take up construction of dams on massive scale under the guise of development programmes there is bound to be scope for avoidable failure of dams resulting in large scale deaths of people and economic ruination of the country.  This situation is reflected by the failure of 2 major dams namely Banquio and Shimantan dams in China in 1975.  Banquio dam was built in 1950 on Ru river, 118 m high with a storage capacity of 492 million cubic meters .  An experienced irrigation expert Chen Xing proposed 12 sluice gates but the authorities reduced them from 12 to 5 and so Chen criticized the Government for compromising with the safety aspects and consequently he was transferred from the project in 1961.  But when the problems surfaced in the project Chen was brought back and he was requested to help the Government in solving the problems.  Since the beginning there were some mistakes in the design and construction of this dam including cracks in the dam and sluice gates.  For rectifying these defects Russian experts were invited for improving the dam and this Soviet design was called an “Iron dam, a dam that could not be broken”.  Perhaps this is a reflection of the ego inherently present among engineering experts who are responsible for failure of many dams.  The dam built in 1950 was designed for a 1 in 1000 year flood (300 mm of rainfall per day).  The peak flood was estimated to be one from a storm that would drop 0.53m of rain over a 3 day period.  In August 1975 the engineers estimates on flood misfired because a 1 in 2000 years flood occurred and poured one years rainfall in 24 hours and a rainfall 1060mm was recorded in 24 hours near the typhoon center.  But new records were set at 189.5 mm rainfall per hour, 1060mm rainfall per day, exceeding annual rainfall of 800mm .   It is reported that by some experts that the cyclone dropped a meter of water in 3 days comprising 0.448m due to the first storm on August 5th.  A  second down pour lasted for 16 hours on August 6th and a 3rd down pour lasted for 13 hours on August 7th .  Another dam on Hong river was also constructed and it also collapsed.  The Shimantan dam was built to withstand a 1 in 500 year flood due to a cyclone that drops 0.48m of rain over a 3 day period and its storage capacity was 95 million cubic meters.  By August 8th, the Banquio and Shimantan dam reservoirs began to get filled up to their full capacity because the run-off into these reservoirs exceeded the spillway discharges but at 12.30AM the water in the Shimantan dam reservoir rose 40cm above the crest of the dam and the dam collapsed and the reservoir emptied its 120 billion cubic meters of water within 5 hours.  Half an hour later the Banquio dam was crested when some people attempted to save the embankment from collapsing.  But soon the dam burst and resulted in creation of a wall of water 6m high and 12m wide moving followed by the emptying of 600 million cubic meters of more water.  In total 62 dams broke and the flood covered a million hectares of agriculture in fields in 29 countries and municipalities.  11 million people in the region were effected, 85,000 were died due to dam failures.  The flood water moved at 14m per second and there was no time for giving warnings and the authorities were hampered because the telephone communication was knocked out immediately and the people did not expect any one of these so called “iron dams” to fail.

Plain areas suitable for anicuts and barrages and hilly areas for big dams:   

Once the Banquiao and Shimantan dams were completed in 1950s many many smaller dams were built.  Initially these dams were built rightly in the mountain regions.  But in 1958 Vice Premier Tan Zhenlin arbitrarily ordered that the dam building should be extended also to the plain areas of China.  The Vice Premier also asserted that primacy should be given to water  storages for irrigation.  A noted expert hydrologist , Chen Xing objected to this policy on the basis that it would lead to water logging and alkalization of farm lands due to a high water table produced by the dams in plain areas and deltas.  Not only were the warnings of Chen Xing ignored but political officials changed his design for the largest reservoir on the plains.  Chen Xing, on the basis of his expertise as a hydrologist, recommended twelve sluice gates for the dam but this was reduced to five by critics who said Chen was being too conservative.  There were other projects where the number of sluice gates was also arbitrarily reduced significantly.  Chen Xing was sent to Xinyang.  But Chen Xing criticized the promoters of mass-scale dam construction  and he was  branded as a “right-wing opportunist”.   When serious problems with the water system developed in 1961 a new party official in Henan brought Chen Xing back to help solve the problems. Even Indian experts like Sir Arthor Cotton, Dr.A.N.Khosla and others asserted that while big dams be constructed in upland hilly regions, plains region must be preferred for anicuts and barrages in the Godavari river basin.  Thus Polavaram dam is a violation of the basic engineering principles.