At first sight, Tehri is a wonderful place to build a dam: a narrow
Himalayan gorge at the end of a wide trough-shaped valley that could hold
3.5 cubic kilometres of water. Indian geologists picked it out 30 years
ago and today engineers are about to start work on a dam 260 metres high,
close to Tehri town on the River Bhagirathi, a tributary of the Ganges.
The dam will be the largest in Asia. Behind it, the Bhagirathi valley
will be flooded for 45 kilometres and its tributary, the Bhillunguna, for
35 kilometres. In the plains south of the Himalayas, farmers are looking
forward to extra water from the reservoir to top up their irrigation canals;
the citizens of New Delhi want more water for their taps, and the industrialists
of northern Indian cities such as Kanpur and Allahabad want more hydroelectricity
to power their factories.
Advertisement
The people of Tehri are less enthusiastic about the $2 billion project.
The reservoir will flood their town as well as 23 villages in the valley,
displacing more than 80,000 people. And the 200,000 inhabitants of three
towns downstream would be well advised to move out as well if the dam is
completed. Seismologists believe there is a serious risk of a large earthquake
in the valley within the lifetime of the dam and that it will destroy the
dam and inundate the towns below.
Shivaji Rao from the college of engineering at Andhra University has
been a member of the Indian government’s environmental appraisal committee
for river valley projects. He warns: ‘if the Tehri reservoir emptied in
half an hour, floods may reach Rishikesh in one hour and Haridwar in another
15 minutes.’ A 40-metre wall of water would rush down the narrow valley.
‘Almost all the people in Devaprayag, Rishikesh and Haridwar may be killed
and all towns and villages up to Meerut, some 250 kilometres from Tehri,
may be severely damaged; millions of people will be exposed to epidemics.’
The whole Himalayan region is very active seismically because of the
movement of the Earth’s continental plates. The Indian plate to the south
is pushing its way beneath the Eurasian plate at a rate of about 5 centimetres
a year. there have been nine earthquakes that measured more than 7.5 on
the Richter scale in the Himalayas in the past century, but none in the
Tehri area since 1828. Vinod Gaur, until recently director of India’s National
Geophysical Research Institute and the country’s best known seismologist,
points out that Tehri sits close to the line of a ‘seismic gap’ along the
boundary between the Indian and Eurasian plates. A seismic gap is a zone
along an active plate boundary where there have been relatively few earthquakes
in recent times – so few that stress is building up to the point where a
quake is likely.
The central seismic gap along the Himalayan plate boundary extends 700
kilometres from Kashmir, where an earthquake at Kangra in 1905 measured
8.6 on the Richter scale, to northern Bihar, site of an 8.4 quake in 1934.
Rao says that the time elapsed since the quakes of 1905 and 1934 is ‘long
enough to require ruptures by at least three major quakes of magnitude 8
to release the accumulated strain’. Unfortunately the Tehri dam, which will
be just 15 kilometres from the plate boundary, has been designed to withstand
only the likely consequences of a quake of magnitude 7.2 James Brune, a
former president of the Seismological Society of America and professor of
geophysics at the University of Nevada, wrote recently that the Tehri dam
site is ‘one of the most hazardous in the world from an earthquake point
of view .. Although we cannot predict precisely the time of the future
quake, there is little doubt that it will occur at anytime in the next few
hundred years and has a high probability of occurring during the projected
lifetime of the dam.’
Lesser quakes are a regular feature of the region. Locals in Tehri say
that they feel a tremor every two or three years. Some studies suggest that
these tremors have become more frequent since the early 1970s. Sunil Roy,
who chaired an environmental appraisal of the dam for the government in
the early 1980s, says: ‘What makes the Tehri dam particularly dangerous
is the fact that there are four major faults in the region along which movements
of the Earth’s crust are likely to take place.’
At Tehri itself the river flows along a fault, although geologists are
divided about whether it remains active. There is also the Mahr Fault, a
tear fault where the rocks have slipped in the same way as on a tear in
a piece of paper, exposed about 4 kilometres downstream from the dam. It
may be connected to the main plate boundary and passes about 7.5 kilometres
beneath Tehri. During the two decades that have elapsed since the Geological
Survey of India highlighted the danger, no systematic attempts have been
made to explore further the risks posed by the faults. Gaur complains that
the lack of data has been used to discredit opponents of the dam, but says
that ‘the probability of a major earthquake whose rupture zone may traverse
the dam site is high’.
Calculating the risks
The central problem for earthquake scientists at Tehri lies in converting
measures of the intensity of a quake into an estimate of the likely strains
that would be placed on a surface structure such as a dam. Engineers from
Roorkee University in India, when investigating the Tehri area for the Indian
government, at first calculated that the worst credible quake there would
measure 7.2 on the Richter scale. They further calculated that such a quake
would move the ground at the surface at accelerations as great as 0.446
g (g is the acceleration due to Earth’s gravity).
The dam is designed to resist an acceleration of 0.5 g. But there is
now widespread concern that this is not enough. The row reached a new intensity
in April of last year following the report of a new committee of experts
set up by the government under the chairmanship of the director-general
of the Geological Survey of India, D P Dhoundial, to look at seismological
risks in the area. It concluded after representations from Gaur, its lone
seismologist member, that the worst quake could, as opponents of the dam
had long contended, have a magnitude greater than 8. Yet the committee,
which was dominated by engineers, refused to revise upwards the estimate
of peak ground acceleration, claiming that 0.446 g was still the most that
could be expected.
Within days of the report’s release, however, Gaur had repudiated this
conclusion and asked Brune, the American who first devised an equation for
calculating peak ground acceleration, to review the data. In a detailed
reply forwarded by Gaur to the Indian government in May, Brune wrote that
his formula had been wrongly applied by the dam’s backers. The critical
issue was the extent to which the vibrations from a quake would be weakened,
or attenuated, as they spread through rock. Dhoundial’s committee had assumed
that any major quake would be along the plate boundary, and hence 15 kilometres
or more away from Tehri, and that it would attentuate according to Brune’s
formula. It concluded that a peak ground acceleration above the earthquake
itself (some 2 g) would weaken to less than 0.5 g at Tehri.
But Brune said his formula, developed for a study in the Imperial Valley
in California, was ‘not applicable to the Tehri Dam region’. In California
the shock would weaken swiftly in thick layers of sedimentary rocks. But
in the hard rocks between the plate boundary and Tehri, and especially for
quakes deeper than 1 kilometre, he expected ‘very little attentuation over
most of the path’. Brune pointed to other failings in the Indians’ calculations:
features of the land surface, even in solid rock, could amplify ground motions
caused by an earthquake by a factor of two or more. And faults could branch
from the plate boundary and give rise to a quake much closer than 15 kilometres.
They might also focus the energy from a quake towards the dam. In December,
Brune travelled to the Tehri dam site and afterwards said that modern equipment
had recorded ground accelerations above 1 g at five earthquakes in recent
years and at more than 2 g during a quake in northern Canada in 1985. He
warned: ‘There is no question that accelerations near 1 g might be produced
at Tehri dam. The question is, what is the probability, and what risk is
acceptable? In my opinion this question deserves much more investigation.’
Faced with Brune’s critique, the expert committee stood its ground,
with Gaur now dissenting. The government now appointed Jai Krishna, a former
vice-chancellor of Roorkee University to review the evidence. He gave his
support to the expert committee. According to Nalini Jayal, a former top
civil servant at the Ministry of the Environment and now a leading environmental
campaigner, Krishna had, like his Roorkee colleagues, ‘been fully involved
in supporting the Tehri dam’ and was not a seismologist. He said that Brune’s
case was being ignored ‘because any increase of peak ground acceleration
would imply a total redesigning of the dam at a prohibitively uneconomic
³¦´Ç²õ³Ù’.
As well as damage to the dam in a big quake, smaller tremors pose a
further threat in the narrow Himalayan valleys – landslides. A large landslide
into the reservoir would set up a wave that could breach the dam or pour
over the top. Such an event in Lake Vajont in Italy in 1963 killed 2000
people. Again, Tehri seems a prime candidate for disaster. Twenty years
ago, the Geological Survey of India noted in its project report the remains
of ‘a number of major landslides along the Bhagirathi river .. These slides
are expected to be aggravated during conditions of rapid drawdown of the
reservoir,’ it said, and predicted ‘a few more fresh slides’ near the dam
itself. Since then, the dam has been earmarked as a source principally of
‘peak load’ electricity for the state grid in Uttar Pradesh. This means
that its turbines will be switched on and off frequently, alternately raising
and lowering the water levels in the reservoir – giving precisely the rapid
drawdown identified as dangerous.
Another growing fear is that the weight of water behind the Tehri dam
might itself trigger an earthquake. There is increasing evidence from around
the world that such ‘reservoir induced earthquakes’ do happen. When the
Tehri reservoir is full, its water will weigh 3.5 billion tonnes. Harsh
Gupta, former director of India’s Centre of Earth Studies and now vice-chancellor
of the University of Cochin, points out that 6 out of 20 reservoirs round
the world with heights above 150 metres have witnessed this phenomenon,
often as they filled up. Filling a reservoir at Koyna near Bombay seems
to have led to a shock of magnitude 6.5 in 1967, which broke the dam and
killed 177 people. This reservoir was much smaller than that planned for
Tehri and was in a region of low seismic activity. Gupta says that the fractured
and crushed rocks of the Bhagirathi valley will increase the risk of a reservoir
induced earthquake. The broken rocks will allow the weight of the reservoir
above to force water down into zones where it could trigger earthquakes.
Gaur agrees. He says that ‘the creation of a large reservoir in a region
which may already by critically stressed, might induce rock failure’, resulting
in a breach of the dam.
Many conclude that Tehri is a disaster waiting to happen. It would follow
40 other failures of big dams in India (a failure rate of 9.2 per cent).
Besides the Koyna disaster, the collapse of the Machu dam killed several
hundred people and destroyed the town of Morvi in Gujarat in 1972, and lesser
collapses have occurred at Kadam in 1958, Panchet in 1961, Nanaksagar in
1967, Chikahole in 1972, Dantwala in 1973 and Aran and Hinglow dams in 1978.
But if Tehri collapsed, casualties would be measured not in hundreds but
tens of thousands.
A long battle without a victory
The Tehri dam project was first mooted in 1949 by the Geological Survey
of India, which completed its detailed project report in 1969. But progress
has been slow, first because of a shortage of cash, and second because of
growing opposition from seismologists and the townspeople of Tehri, who
created the Anti-Tehri Dam Struggle Committee in the 1970s. The committee
won its first victory in 1980, when the government, which had given the
project clearance in 1972, asked Roy to review the project. Through the
early 1980s, while Roy’s working group sat, work on site continued with
the construction of diversion canals and the clearing of rocks from the
river bed near the dam site.
Urgency returned in 1986 when a treaty signed during a visit to India
by Mikhail Gorbachev earmarked one billion roubles in Soviet aid and technical
assistance for the project. Reports at the time said that the Soviet Union
wanted to provide aid for India’s nuclear programme, but the two countries
could not agree terms. So Tehri was pulled from the hat as a compromise.
Usha Rai, a journalist for the Times of India, says that ‘the decision was
taken quickly because the Russians were eager to assist India in some field.
It was a political decision, taken at the highest level and disregarding
the advice of the Department of the Environment as well as Sunil Roy.’
But Soviet scientists were soon uneasy about the dam’s ability to withstand
an earthquake. An analysis carried out in 1987 by Alexander Fink, the chief
Soviet project engineer concluded that ‘the design needs a revision because
it does not meet the safety standards applied to similar structures in the
Soviet Union’. He called for the dam’s base to be widened from 1100 metres
to 1500 metres. This he said, ‘will securely keep the dam on its feet and
will not allow it to tumble down’. It would also increase the price by an
estimated 25 per cent.
In 1986, Roy wrote to the government, complaining about its continuing
‘failure to increase the data on seismicity from additional locations, monitoring
of seismic activity of faults ..’ His letter concluded: ‘I have chaired
innumerable committees and groups in India and in other parts of the world.
I have never encountered such an unbending dogmatic approach to all issues
which were not positively framed to ensure continued work on the Tehri Dam,
whatever the cost.’
Few independent observers believe that the Indian government wants a
dispassionate analysis of the risk of the Tehri project. Roy’s environmental
assessment group, in an interim report in 1980, recommended a series of
studies into the environmental impact and seismic risks of the project.
They were essential before the go-ahead could be given, it said. Yet few
of its proposals have been taken up. Even the small network of seismic observatories
and boreholes necessary to check whether the fault in the Bhagirathi river
bed is still active were not established.
Will the dam be finished and the reservoir filled? India has a reputation
for starting but failing to complete larger development projects, and opposition
remains fierce. After a hunger strike at the dam site over Christmas 1989
by the Gandhian environmental campaigner Sunderlal Bahuguna, the government
agreed to a moratorium on construction until new reviews had been completed.
Last September, the government released funds for work on dam safety
and research. Locals say this money appears to have been diverted into construction.
When I visited the town in November 1990, work seemed to be in progress
on a large coffer dam across the valley below Tehri, and water is being
diverted through a tunnel. Coffer dams are built to divert water away from
the river bed to allow construction of the main dam. The construction company
has told townspeople that it will begin the main dam later this year.
Night and day, Tehri town has to withstand dust, arc lights perched
high on the surrounding hills, and a steady stream of trucks rumbling through
the town, across the bridge and onto the dam site, carrying rock to and
fro. To add insult to injury, for several hours in the middle of the night
during my visit, an armed guard, complete with fixed bayonet, blew a whistle
as each truck crossed the bridge. The echoes from the high mountain walls
filled the valley for miles. Locals speak of rock blasting at night, a practice
the construction company denied until presented with a large rock which
catapaulted through the roof of a house next door to the home of the local
police chief early one morning.
The dam’s supporters cleared one more hurdle in November, when the Indian
Supreme Court, after five years’ deliberation, turned down an application
from a Tehri lawyer, Virendra Saklani, that the dam should be stopped because
the government had not adequately investigated the earthquake risks. The
court said that it was not competent to judge whether the right decision
had been reached. Saklani, a veteran freedom fighter from pre-Independence
days and now a frail man in his 70s, has fought the dam for almost 20 years,
extracting from government departments the project reports which first revealed
the extent of the fears of seismologists about the dam’s safety. He spent
the past five years pursuing the argument through the Supreme Court. When
I met him in Tehri in November he had just received the ruling, but refused
to be downhearted. ‘We must now do what I originally wanted to do back in
1978’, he said, ‘and pursue the case from the start again in the local courts.’
Opponents of the Tehri dam argue that it is an unwarranted and dangerous
assault on the environment and people of the western Himalayas. More than
80,000 people are to be up-rooted and 270,000 hectares of fertile land,
representing much of the best land in the district, will be flooded in return
for a few decades of peak load electricity before the dam fills with Himalayan
silt. They argue that more electricity could be obtained more cheaply and
much more quickly from small ‘run of the river’ hydroelectric plants that
do not involve large dams. Such a strategy has been successfully adopted
by the neighbouring state government of Himachal Pradesh. Unlike Tehri’s
power, this electricity would be available for local use as well as for
export to the plains.
Meanwhile, the central government’s auditor-general has cast serious
doubt on the official cost-benefit assessment for the dam. But overshadowing
all the debates stands the ever-present threat of a seismic disaster, a
disaster that would make Bhopal pale into insignificance. Nalini Jayal,
a former top civil servant at the environment ministry, issued a sombre
warning in a letter last November to Maneka Ghandi, the environment minister.
Unless the Tehri dam is redesigned to withstand a major earthquake, he said,
‘there is no doubt whatsoever that the existing design cannot withstand
an earthquake that must come in the forseeable future. The catastrophic
consequences for millions of people living downstream and the wiping out
of Rishikesh and Haridwar are mind-boggling.’
* * *
A historical lesson for the dam builders
West of Tehri, in Pakistan, the River Indus drains much of the western
Himalayas, flowing out to the Arabian Sea. Historically, it has a fearsome
reputation as the Lion river, a barrier to armies and a bringer of flash
floods. But in recent years, the Inus has been seen in more positive terms
as a tempting source of energy in a country with relatively small fossil
fuel resources. One of the world’s largest dams has been built at Tarbela,
where the Indus flows out from the Himalayan foothills into the Vale of
Peshawar, before cutting through the low front ranges to the plains. But
150 years ago, a landslip blocked the Indus, forming a natural dam roughly
the same size as Tehri. The story of what happened when this dam burst has
lessons for the people of Tehri.
The Indus crosses many active faults in its journey through the mountains.
One reaches the surface where the river cuts along the flanks of the mountain
Nanga Parbat (8125 metres high). This fault carries the crystalline rocks
that make up the mountain itself onto gravels laid down by the Indus. Other
parts of this fault change the course of the river. These are clear signs
that the fault has moved at some time in the past, which may be ancient
history in human terms, but are only yesterday in the geological record.
We do not know what the recent pattern of minor earthquakes in the region
has been because no suitable programmes of seismic monitoring have been
carried out. But there is some information. Historical records tell of an
earthquake on the flanks of Nanga Parbat in 1840. The most interesting aspect
of this quake for today’s dam builders is the devastation that followed
six months later in the valley downstream.
The 1840 earthquake triggered a huge landslip on the sides of the Indus
gorge, which blocked the valley, forming a 300-metre-high natural dam. In
the following six months, the Indus built up a lake stretching 60 kilometres
behind the landslip. Then, in June 1841, there was another landslip. Although
this was much smaller than the first one, it delivered enough debris into
the lake for water to surge over the top of the dam. The natural barrier
to the Indus failed, and the lake water was released almost instantaneously.
After travelling more than 300 kilometres down the gorge, the water emerged
from the mountains at the place where the Tarbela dam is now, and continued
out across the Vale of Peshawar.
Fifty kilometres away across the plains, at Attock, the floodwaters
overwhelmed a Sikh army that had taken advantage of the abnormally dry bed
of the Indus to establish a camp. An eyewitness told of the destruction:
‘I looked and perceived that all the dry channels were already filled and
that the river was racing furiously in an absolute wall of mud, for it had
not at all the colour of water .. It was a horrible mess of foul water,
carcases of soldiers, peasants, war-steeds, camels, prostitutes, tents,
mules, asses, trees and household furniture, in short every item of existence
jumbled together in one flood of ruin.’
Apart from giving an insight into the priorities of a 19th century Sikh
army, this account emphasises that there was not a gradual build-up of water,
but a sudden inundation of the camp by a bore estimated to be more then
25 metres high. Many hundreds of the Sikh group were drowned as well as
untold others across the vale. The devastation continued downstream at least
as far as Dera Ishmail Khan, the site of one of modern Pakistan’s nuclear
power installations.
Travellers today on the Karakorum Highway pass the site of the 19th
century dam at Nanga Parbat, on their way north to Gilgit and the Chinese
border. The remnants of the 1840 landslip are still there, just up valley
from traces of swift and strong water flow high above the present level
of the river. Scree and glacial deposits have been washed off bedrock that
is buried elsewhere, and the terraces look as if they have been cut by waves.
It is a place that those in favour of large hydro-electric projects would
do well to visit.
The Great Indus Flood of 1841 offers a dramatic illustration of dam
failure, albeit a natural one. It is also a warning for future hydroelectric
projects on the river. Because of the huge load of sand and mud carried
by the Indus, Tarbela’s reservoir is silting up, with a delta growing into
it at more than 1 kilometre a year.
Over the past few years, the Water and Power Development Authority of
Pakistan have been directed to draw up a list of new dam sites on the Indus
to take up the falling capacity. WAPDA scientists were fully aware of the
1841 flood and recommended a new dam downstream from Tarbela at Kalabagh,
away from the worst earthquake areas.
After years of wrangling, this option was rejected because the reservoir
would have drowned a large area of fertile Punjab farmland and upset the
politically influential and wealthy landowners. WAPDA scientists, caught
between this powerful lobby, the politicians and Pakistan’s growing energy
needs, now recommend a new site at Basha, high in the mountains. The site
is marked out, and World Bank finance is being sought. Once complete, the
Basha reservoir will stretch up to the site of the 1841 dam break. Another
dam site, at Patan, lies on a belt of modern earthquakes. In 1974 a destructive
quake killed many thousands of people in the surrounding hills.
And if this dam goes ahead, Pakistan looks set to build a ladder of
hydroelectric schemes up the river, with those downstream depending on those
higher up being unaffected by earthquakes. Or maybe they will all silt up
before they have a chance to fail. Hydroelectric power on this scale seems
to be neither a renewable nor a particularly safe option for the people
of Pakistan, particularly when they try to tame the Lion River.
Fred Pearce is the author of Green Warriors published by The Bodley
Head earlier this month.
Rob Butler lectures in earth sciences at Leeds University, and works
on faults in the Himalayas and other mountain ranges.