Analysis of last year's Surat Flood

SURAT FLOODS: HOW IT WAS AN PREVENTABLE DISASTER

Why was the water not released before allowing reservoir to be full?

PRESS RELEASE August 22, 2006

The unprecedented floods of Surat city and district in South Gujarat in western India earlier this month and the damages of over Rs 21 000 crores (according to Gujarat govt estimates, actual losses would be much higher if all the losses are properly accounted) were completely preventable. Had the authorities released even 3 lakh cusecs of water from the Ukai dam starting from Aug 1, Surat may not have experienced such a disaster. There was sufficient specific information available to take such an action, as is evident from the rainfall figures in Tapi basin talukas in Maharashtra. Due to lack of timely release of water by the Gujarat water resources department, the storage in the Ukai reservoir was allowed to go up beyond the levels it should have, leading to sudden release of unmanageable quantities of water from Aug 7, for prolonged periods. Since this coincided with the days of high tide when the river’s drainage capacity is further reduced, the floods brought catastrophic results.

Rainfall As we can see from the salient features of the Ukai dam (Annexure 1, see www.sandrp.in/new/surat_floods_aug08.pdf), 97.85 % of the catchment area upstream from the dam site is in Maharashtra and Madhya Pradesh over 81.3 % in Maharashtra alone. We have compiled the list of dates when rainfall in the Tapi basin talukas in Maharashtra was more than 25 mm, during June 1, 2006 to August 14, 2006 (Annexure 2, see www.sandrp.in/new/surat_floods_aug08.pdf). We can see that such instances were spread all over July & Aug ‘06.

It is clear from these figures that Ukai was going to get large inflows in the days to come and there was sufficient actionable information available with the Gujarat water resources dept to release water from Ukai Dam in July itself.

Ukai Reservoir Level This was all the more important considering the rapid rise in water level at Ukai dam as shown in the graph below.

Highest pre monsoon, highest July end levels at Ukai When we compared the Ukai reservoir levels just before the monsoon and at the end of July for the last four years (for which we could get the data from Central Water Commission and Central Electricity Authority bulletins) we were surprised to find that the levels at Ukai dam were the highest this year in last four years, both with respect to the level just before the monsoon and also at the end of July, as is clear from the table below.

Year	Lowest level Before monsoon		Level (m) at the end of July
	Level, m	Date
2003	89.91	June 17	94.56
2004	88.02	July 25	88.16
2005	85.37	June 24	94.26
2006	90.71	June 29	97.8

Source: Central Water Commission, Central Electricity Authority, Govt of India, bulletins of various dates

Water release figures from Ukai reservoir While the project authorities have yet to make public the hourly or even daily figures of inflows, outflows, levels and usage of water at Ukai throughout this monsoon, we could gather following figures from the govt statements and media reports.

Date	time	Inflows, cusecs	Outflows, cusecs	Water level, m
Aug 5	0800 am	85 958	26 664	102.20
	0800 pm	48554	23 640	102.26
Aug 6	0800 am	75 087	124 920	102.14
	0800 pm	330 216	254 780	102.57
Aug 7	0800 am	853 679	409 004	103.46
	0800 pm	1 072 680	816 036	104.22
Aug 8	0800 am	1 053 133	844 092	104.97
	0800 pm	961 466	907 316	105.33
Aug 9	0800 am	856 000	850 000	105.34
	0800 pm	711 757	650 000	105.38

It is disturbing to know that even as late as on Aug 6 evening, the water releases from Ukai reservoir were just 254 780 cusecs when the level had already built up to 102.57 m, that is reservoir was 83.6% full at 5.528 BCM live storage. A day earlier the releases were shockingly low at 23640 cusecs. This was when during the eight days preceding this, all the talukas of Nandurbar district had received over 25 mm rainfall at least for 4-5 days including an instance of 260 mm rainfall in a day and seven instances when rainfall was over 100 mm. Surgana taluka in Nashik was also getting similar pattern of rainfall.

WHAT COULD HAVE BEEN DONE? Had the Ukai dam authorities released 300 000 cusecs of water starting from Aug 1, the reservoir level at Ukai would have been reduced by over 3 BCM by the morning of Aug 6 and the level of water at the dam would have been 94.2 m (about 2.34 BCM storage). After this even with all the additional water coming in, continued outflow of 3-4 lakh cusecs would have been sufficient to ensure that reservoir does not get full and there would have been no big flood disaster in Surat district.

Sufficient actionable information available As we saw above, there was sufficient information available with all concerned about the increasing water level at Ukai reservoir and also about the high rainfall in the catchment.

Additional reasons There were additional reasons why such water releases of 300 000 cusecs starting from Aug 1 or earlier were justified:

1. As per Ukai reservoir design, the reservoir is supposed to have a flood cushion of 1.332 BCM. That means that this amount of storage should not be filled till the end of monsoon. Considering the current live storage capacity of 6.615 BCM, water level beyond 102.07 m (5.283 BCM) should not be filled up, in anticipation of floods. Since level at Ukai was already near this level on Aug 1, water releases of 300 00 cusecs should have been started from Aug 1 or earlier to reduce the storage at Ukai.

2. Moreover, this flood cushion was designed assuming that Tapi River could drain 8.5 lakh cusecs water. However, as noted by the report of the Govt of India’s National Commission for Integrated Water Resources Development (Sept 1999, p 131), Tapi river’s drainage capacity has gone down substantially in view of the experience in 1998 floods when even with an outflow of 6.73 lakh cusecs, Surat was badly flooded. The National Commission had in fact recommended a review of all the parameters of the Ukai dam operation. This should have required even greater flood cushion at Ukai, but neither such a review of the dam operation rules was done, nor was even old flood cushion properly used. All this also warranted release of 300 000 cusecs from Aug 1 or earlier.

3. The Central Water Commission Chair has accepted, as was known to Gujarat water resources dept that current water carrying capacity of the Tapi river is just 3.5-4.0 lakh cusecs. But this crucial information did not seem to be part of the equation of water releases from the Ukai dam.

The Gujarat Govt has a lot to explain why they did not start release of 300 000 cusecs from Aug 1 or earlier.

Power Generation at Ukai (Source: Central Electricity Authority (www.cea.nic.in))

Month	Generation, MU	Target, MU	% of target achieved
Jan 2006	45.79	74	61.88
Feb 2006	44.12	73	60.44
March 2006	53.52	66	81.09
April 2006	56.13	60	93.55
May 2006	50.25	71	70.77
June 2006	29.2	61	47.87
July 2006	13.85	33	41.97

One reason why the dam authorities tend to keep high storages at dams like Ukai is for maximization of power generation. So we looked at the power generation figures from Ukai Dam and the figures we found from the Central Electricity Authority equally shocking.

We can see from the above table is that even as the Ukai reservoir had unprecedented water storage and Gujarat was facing power deficit of up to 1500 MW, Ukai was not being used to generate power. And this situation continued right up to July 2006, when the power generation at Ukai was the lowest in 2006 so far. Here it should be remembered that the target figures in the second column above are fixed by the project authorities and do not signify the maximum power that Ukai can generate. In fact the potential of power generation at Ukai is much higher than these figures.

The loss of power generation benefit to economy is thus a big loss for the nation. The second consequence of this was that Ukai had unnecessarily high water levels throughout this period, leading, ultimately to disastrous floods in Surat district in August. Thirdly, this water flowed away without generating any power, which is also a loss of potential benefit. Who will be held answerable for these losses? (For Power generation figures at Ukai in the recent years, see Annexure 3 at: www.sandrp.in/new/surat_floods_aug08.pdf)

AVOIDABLE DISASTER It is clear from the above account that the unprecedented flood disaster that visited Surat district, earlier this month was totally preventable with lower pre-monsoon storages (which could have been achieved by higher power generation at Ukai in pre-monsoon months) and by timely releases of around 300 000 cusecs starting from Aug 1, for which there was sufficient actionable information available to all concerned. The commission of enquiry set up by the Gujarat govt is not likely to inspire much confidence and what is required is an independent, credible investigation of the events leading to the disaster. In the meantime, Gujarat water resources ministry would do well to release all the figures of inflows, outflows and levels at Ukai Dam from June 1 to Aug 15, 2006.

Recommendations What happened in Tapi basin could also be seen to the replicated in a number of other basins this monsoon, including Sabarmati, Mahi, Krishna, Godavari, among others. To reduce occurrence of such events in future, some immediate recommendations can be made:

Þ National Enquiry A Nationwide independent enquiry should be set up on the issue of floods in India this year, particularly those from the sudden releases of dams all over India. With proper use of water in the reservoir before the monsoon and timely releases of water from the dams, many of these disasters could have been avoided or damages substantially reduced.

Þ High Pre Monsoon storages As shown by SANDRP press release on Aug 10(www.sandr.in/new), a number of dams in these basins had high water storages, up to 47%. The investigation should also look into why this was the case, what were the consequences and how they could have been avoided.

Þ Transparency in Dam operations The inflows, outflows, levels, storages and forecasts at all reservoirs should be routinely made public on daily basis.

Þ The reservoir operation rules, the rule curves and disaster management plans should be in public domain for all reservoirs.

Þ Provide basin wise rainfall data, forecasts Met Dept (& others, including state govts) should give basin wise rainfall data and the rainfall forecasts should be done river basin wise and not geographic area wise as is done now.

Þ Failure of CWC The Central Water Commission’s performance about flood forecasting leaves a lot to be desired. It seems to have totally failed to predict the floods, both due to heavy rainfall and due to releases from dams from all over India. For example, even as a number of river basins in Gujarat were experiencing floods due to releases from dams, the flood forecasts at CWC website was completely silent on all these floods.

Þ Review Dam Operation Rules As recommended by the National Flood Commission (1980) and also the National Commission for Integrated Water Resources Development, periodic review of dam operation rules and rule curves should be taken up in a transparent manner and the same should be made public.

Þ UNDP reports inaccurate UNDP routinely puts out flood situation reports, that seems to have no understanding of how the sudden releases of water from dams are leading to floods. It is expected that UNDP reports would give a more accurate picture of the situation.

Þ Citizens Groups Most importantly, Citizens groups around the large dams in their respective areas all over India need to monitor the dam operations and flood situation and make the govt answerable for the avoidable disasters.

Himanshu Thakkar

South Asia Network on Dams, Rivers & People, 86-D, AD block, Shalimar Bagh, Delhi 110 088

(+91 11) 2748 4654/ 5

cwaterp@vsnl.com , www.sandrp.in