The height of inaccuracy

Last year, people in the Narmada valley were ready once again to face the fury of the river. A week passed but nothing happened .... The villagers had a better measuring system, says RAVI KUCHIMANCHI.

IT was July 20, 2000. The Narmada was visibly rising, a few inches every day until it covered the path between Nimghavan and Domkhedi, crossed a tree and threatened to enter the fields. People prepared for another showdown with the rising river. So, imagine my surprise when I read the gauge at Hapeswar after a week and found the water in almost the same place of about 92 metres. A week passed and another. The water neither invaded nor retreated in defeat. I started calculating.

The dam was 88m high, about 300m or 400m wide and the over- flowing river was at 92m. How much water was flowing over? Easy enough - kinetic energy equals potential energy - we were looking at flows like 10,000 cubic metres per second ... We thought: at this rate let us see how many days the Narmada takes to empty itself.

The Narmada carries 23 million acre feet of water every year and we were less than 40 per cent through a lean monsoon season. Thus at best there was 15,000 million cubic metres water in it. Draining at 10,000 cubic metres per second .... the water would get over in just 15 days.

Days passed but the water level did not fall 5 cm. It was the second week of August and I decided to go to the dam-site to check the water level there ... is it really 92m? Clearly if the Narmada was flowing so high it should have all spilled over by now. And so with a flash light, as the light was fading, on a motor bike past the police check points as we approached the river, I told myself "it cannot be 92" - the moment of truth had arrived. It was the time for Satyagraha. There was a government metre stick half sunk in the river and where it crossed the water, it read "89.2m" ... not "92.2m".

All of a sudden another thought occurred to me ... it was as if a shadow were lifted from my eyes ... oh my god the village people were right. They had a better measuring system than the Government itself. There was nearly 3 metres discrepancy between the Narmada at Hapesar (92m) and the dam (89.2m). What is more, we had solid proof. Proof that depended on the way they measure heights compared to the way engineers measured them.

In the past year or two, people told me with certainty that the waters will enter far more into their villages than the land- acquisition that is going on. They knew where the flood waters came in 1970 and 1994. "Even without the dam, the Narmada's waters rose that high. Then with the dam it will be much more than government figures."

Medha Patkar asked me to carry out an independent survey of levels. I began to observe this: if in Sikka, Vestha Bhai told me that in 1970 the flood waters came up to some tree and in Domkhedi if Dedli Behen told me that it came to some stone, both levels actually tallied when we checked with the theodolite to within a few centimetres. I began to gain confidence that the people of the Narmada valley had a very accurate knowledge of the river. At the Hapeswar temple benchmark that said 105.990m, people from non-government organisations working with the Government of Gujarat said, "You are cross-checking the heights but this benchmark here is also the Government's. How do you know that's right?" They went away laughing, having pondered over the futility of challenging the Government.

The daggers were drawn at the following positions: All the engineering knowledge including mine, was to measure heights above mean sea levels starting from benchmarks. The Narmada villagers, however, measure levels of the houses from the river or level of the river from their houses and had internally consistent observations of floods accurate to a few centimetres. If I started from government bench marks I was just going around in circles.

So what should I do?

I went to the Narmada Satyagraha.

Waking up to the Narmada everyday, I must have sub-consciously moved from one system of measurement to another. It suddenly occurred to me that the river had to be flat - horizontal - because of the dam. Like the water in a swimming pool or a lake that has the same level everywhere. The Domkhedi satyagraha was on a 50 km long lake/reservoir created by the dam which was 88m high. Therefore, all I had to do was to measure heights of villages from the river, like the village people do. There was no need for benchmarks; in fact they could now be challenged.

We found that the Hapeswar benchmark was only 13.8m above the Narmada which itself was at 89.2m, as per the metre stick at the dam-site. Which means the benchmark was at 103m, and not 105.990m as marked. Likewise, the lowest house in Jalsindhi was at 98.4m, and not at 101.5m as claimed by NVDA engineers. This explained why waters invaded the Satyagraha at Jalsindhi in 1999.

In Nimad the dam building engineers working with the CWC predicted, using the computer where the waters will reach in the worst case of a once in 100-year flood. The NVDA uses this to identify the project affected.

What do the engineers say about floods? "At most 133.95m in Kukra." But the field office in Rajghat, Kukra, had itself recorded the floods in 1970 at 136.688m, about 3m higher. Moreover, in the last 30 years, the waters had exceeded these times, the once-in-100-year level that the Government has calculated. Thus there are serious errors in Government levels of all the States in the Narmada valley.

What do survey errors mean? Those who will be displaced are not fully counted. Three metre errors everywhere would mean 20,000 people who will be affected are left out. The Narmada tribunal stipulates that all those who will be affected have to be identified and rehabilitated six months before the building of the dam. Even if the surveys were accurate to the centimetre, it would mean 60 people not counted, a one millimetre accuracy will still leave six people out. Such accuracy throughout the Narmada valley is impossible to achieve using survey instruments which have intrinsic technical limitations. This shows the project is too big and internally inconsistent to make tall promises or claims that people opposing it are scientifically backward and those constructing it are technologically superior. In fact the civil engineering faculty from our esteemed universities will agree that the dam is too high and it is impossible to identify every single affected person six months before submergence. Leave alone rehabilitate them. Where does this leave the implementation of the Narmada Tribunal, even in principle?

The facts are even more bleak. The estimate of people who will be submerged considered by the Narmada Tribunal when it made its final decision was off by 500 per cent. Had our engineers spent a few months in the villages consulting people about the floods in 1970 and the geography of the area, certainly there would have been more accurate estimates. The dam would have been designed for a much smaller height. Commenting on the Bargi dam that submerged 162 villages as against an estimated 102 villages, Aravinda once wrote: "Dam builders have not scored high marks in the Math department".

While we found the survey errors, the Government of Gujarat decided to repaint the gauges at Hapeswar temple. However, it continues to have the same errors - only they are more glaring.

Note: Since water is flowing, the backwaters of the Narmada between the dam and Domkhedi deviate slightly from the horizontal - but this is less than 1 or 2 cm as the water is flowing very slowly.

* * *

A Nimad villager informed us recently that the engineers have made a +/-3 m stamp. So the Jalsindhi house level will now be 101.500 "+/- 3m"!

Construction began on March 23, 2001 of 3m humps above the 90m dam.

The Narmada Control Authority, admitting that rehabilitation has not been completed and that there are thousands of families living below 90m with no alternative land in sight, agreed fo fo for the humps since Gujarat's engineers said they will be able to keep the waters at 90m levels even if humps are added. Does the engineers demand make logical sense or the tribal people's demand - which is stop construction and if need be even lower the dam height and keep water to a level where rehabilitation is actually done?

The writer obtained his B.Tech. in Civil Engineering from IIT- Bombay and Ph.D. in Physics from the University of Maryland, U.S..