ÐÓ°ÉÔ­´´

Few criticisms of the British government have been as damning as those
that have emerged over the past two months during inquiries into the way
more than £230 million of development aid is being spent on building
a hydroelectric dam at Pergau in Malaysia. ‘Unequivocally a bad buy’ and
‘an abuse of the aid system’ is how Tim Lankester, until recently permanent
secretary at the Overseas Development Administration, described the project
to the Public Accounts Committee of the House of Commons.

The ODA, which is supposed to decide how Britain’s aid budget is spent,
opposed the project but it was overruled. Between 1989 and 1991, the combined
might of the Foreign Office and the Department of Trade and Industry (DTI),
backed by two prime ministers, Margaret Thatcher and then John Major, endorsed
the project as a way of opening up Malaysia to all sorts of British exports.
Judging by the latest revelations, both the technical and the economic viability
of the Pergau Dam as a local development aid project were irrelevant to
the final decision.

Such tactics are not new, however. New ÐÓ°ÉÔ­´´ has unravelled how
a similar project in Sri Lanka, the Samanalawewa Dam, also won British aid
funds despite objections on technical and economic grounds from ODA officials.
As with Pergau, ministers focused on the scheme’s commercial advantages
to British firms rather than on the benefits to local people. In this case,
British engineering firms were appointed to design and help build the project,
and were then expected to win more work around the world from the experience
gained in Sri Lanka.

But that ministerial optimism now seems ill-founded. Furthermore, major
technical problems tarnish the whole saga. Nearly two years after the project
was substantially completed, the reservoir behind the Samanalawewa Dam cannot
be filled because its bed is leaking badly. The escaping water surfaces
in a hillside just downstream of the dam’s right abutment in the form of
a waterfall that has flowed continually since October 1992.

When the abutment first leaked, newspapers warned of a national emergency
as tonnes of soil and muddy water cascaded down the hillside. But within
two weeks the leak was a clear, controlled flow of water, says David Chisnall,
an engineering geologist with Sir Alexander Gibb & Partners, the project’s
British designer. He monitors the discharge and the stability of the hillside
daily. ‘It is under control and totally safe now,’ he says. The hillside
itself has been strengthened with wire mesh, sprayed concrete and long steel
bolts, which have been driven into the rock to keep the surface in place.
A £30 million rescue plan is planned for later this year.

New ÐÓ°ÉÔ­´´ has also discovered that another dam in Sri Lanka, designed
and built by the same British firms, has produced much less electricity
than expected ever since it was opened by Thatcher in 1985. The Victoria
Dam had attracted Britain’s biggest aid contribution until Malaysia’s Pergau
project, and politicians subsequently saw it as providing a shop window
for Britain’s engineering expertise.

In March last year, the minister of state Mark Lennox-Boyd faced questions
in Parliament about the commercial return from Britain’s aid towards building
the Victoria and Samanalawewa dams. He insisted that ‘Britain has enhanced
its reputation in the power construction sector in Sri Lanka’. But that
is not the view among engineering and geotechnical specialists that New
ÐÓ°ÉÔ­´´ has spoken to in Sri Lanka.

Enormous blunder

One local engineer alleged during an official inquiry in Sri Lanka that
British engineers had made an ‘enormous blunder’ over Samanalawewa. A local
professor of geography who lost land when the Victoria Dam was impounded
goes even further. In a letter to New ÐÓ°ÉÔ­´´, L. W. Mediwake writes:
‘Little did the British tax payer know that their aid money has done more
harm than good and created more ill-will than good-will.’

The idea of building the Samanalawewa Dam to generate hydroelectricity
in the hills about 160 kilometres southeast of Colombo originally came from
American engineers in the 1960s. It was developed by Russians in the 1970s,
who conducted geological studies of the area. But nothing came of these
proposals, partly because Sri Lanka decided to look West rather than East
for support to develop its economy.

In the early 1980s, the Sri Lankan government said it would like British
firms, already building the nearby Victoria Dam, to take on detailed design
of Samanalawewa and to get involved in its construction. Norman Tebbit,
trade secretary at the time, took up the offer in January 1985. He proposed
to the ODA that British aid funds should be used to help fund the dam. The
aid, he suggested, should come from the ‘Aid and Trade’ budget, which is
jointly administered by the DTI and the ODA. The purpose of the budget is
to support ‘overseas aid projects of development value which are of particular
industrial or commercial importance to the UK’. As usual, the aid would
be ‘tied’. Not only would the recipient have to repay the loan, albeit at
favourable rates, the money would also have to be spent on goods and services
from companies in the donor’s country.

The British government’s share of the total expected bill of more than
£200 million was a relatively modest £14.4 million, later rising
to £19 million. On top of this, the government approved loans from
the Commonwealth Development Corporation, a government-controlled lending
agency, and banks, backed by export credit guarantees. In all, Britain’s
investment in the project amounted to some £100 million.

In return for the loans and credits, Gibb got the design work, and Balfour
Beatty, one of Britain’s largest construction firms, secured contracts to
lay roads, drive a tunnel and build the power station. But with most of
the money for the project coming from Japan, the main bulk of the construction
work, including the dam building, went to Japanese firms. A firm of Japanese
consulting engineers, Nippon Koei, was also appointed to lead the team of
supervising engineers on site.

From the start, officials at the ODA had doubts about Samanalawewa.
As at Pergau, they believed that Sri Lanka would be better off spending
its aid money on other things. According to a report on British overseas
aid published in 1992 by the National Audit Office, the parliamentary watchdog
on public expenditure, ‘the ODA’s strategy at the time was to give priority
to maintenance and rehabilitation of existing infrastructure in Sri Lanka,
rather than new projects’.

Risky strategy

ODA officials also doubted the value to British trade. According to
the audit office, the ODA shared the Treasury’s assessment that ‘the commercial
case for Aid and Trade support was rather weak’, because ‘recent completion
of the Victoria Dam gave the contractor (Balfour Beatty) an outstanding
reference for securing further business’. Balfour Beatty was already ‘the
second largest user of Aid and Trade monies’. According to ODA officials
today, Tebbit went ahead and backed the Aid and Trade loan because it was
part of a strategy to build Britain’s expertise and reputation as dam builders.

They say that this was a risky strategy in a competitive international
market where most bidders – from Sweden, Austria and Japan, for instance
– had a large home market for dams within which to develop their expertise.
But, though it was draining British aid coffers, it did provide a measure
of success for the companies involved. Before the Victoria Dam, no British
contractor had built such a large dam for more than a decade; afterwards,
Balfour Beatty received government help in securing several contracts. First
with the Swedes on Indonesia’s Mrica Dam, then with the Japanese on the
Samanalawewa Dam, and finally as the main contractor on the Pergau project
in Malaysia. At Victoria and Samanalawewa, Gibb did the design work.

The Samanalawewa scheme aimed to generate electricity by exploiting
the 300-metre difference in elevation between two easterly-flowing rivers,
the Walawe Ganga and the Katupal Oya, as they came within 6 kilometres of
each other in the southern Sri Lankan hills, about 10 kilometres east of
Balangoda. The idea was to build an embankment of rocks with an impermeable
clay core, 100 metres high and more than 500 metres long, across the upper
river, the Walawe Ganga, and a power station beside the lower one. Upstream
of this rockfill dam, a steeply inclined tunnel from the reservoir would
then feed water at speed to the turbines in the power station, from where
the water would be discharged into the Katupal Oya.

Though the proposal sounded straightforward, the scheme got a taste
of the geological problems to come soon after work began in 1987. The site
is in limestone country. Limestone rocks often contain hidden caverns and
underground water channels that make dam construction difficult. And so
it proved. Within months of Balfour Beatty starting work on the tunnel between
the reservoir and the power station, the British contractor hit a local
water table. The tunnel roof collapsed and, for a while, the entire flow
of the upper river gushed uncontrollably through the tunnel at a rate of
up to 300 litres per second. The accident dried out streams, paddy fields
and wells in villages over several kilometres and for several months. It
was solved only by lining the tunnel with concrete – an extremely expensive
business.

More problems followed. When engineers began to fill the reservoir for
the first time in 1991, water poured from the right bank of the dam’s downstream
abutment. Though the first impoundment was only ever meant to be a ‘test
run’, according to Paul Back, Gibb’s chief technical director, the scale
of the leaks surprised him. On site, one of the company’s senior engineers
is said to have thrown up his hands in horror.

The team of engineering firms in charge of supervising work on site
was led by Nippon Koei of Japan, and also included Electrowatt of Switzerland
and a local agency of the Sri Lankan government, the Central Engineering
Consultancy Bureau. According to Back, the supervising engineers failed
to make a detailed site investigation of the dam’s right abutment in 1986
as he had expected them to do. Earlier surveys had indicated that the right
abutment was likely to be geotechnically troublesome, says Back, but he
had hoped to be able to seal any leaky joints with grouting. He concedes,
however, that the job proved to be far bigger and more expensive than he
had expected.

Bursting at the seams

Engineers eventually excavated and lined a tunnel 1600 metres long and
3.5 metres in diameter into the right abutment of the dam, from where they
inserted 15 000 tonnes of cement grouting through 50 kilometres of holes,
drilled in places more than 100 metres deep into the limestone beneath
the dam. To help pay the bills, Japan gave Sri Lanka a further ‘soft loan’
of $27 million.

But it didn’t work. A second impoundment of the Samanalawewa Dam in
October 1992 also failed. As water in the reservoir passed the level at
which power could be generated, recalls Back, a leak of 7000 litres a second
– enough to fill an Olympic swimming pool in less than a minute – sprung
from high on the right flank about 500 metres downstream of the dam. The
pressure of water in the reservoir was literally bursting the hillside at
the dam’s seams. Within a few hours, estimates a report by Gibb’s engineers,
the water had brought down more than 20 000 cubic metres of rock and soil
into the river. Newspapers in Sri Lanka proclaimed an emergency: ‘Engineers
warn of major earthslip’, read the front page headline of The Island. Local
people fled their homes, only returning after site engineers had lowered
the water level in the reservoir, and the leak had slowed down. Chisnall
says the flow has now stabilised at around 2500 litres a second.

The project’s supervising engineers are tight-lipped about the catastrophe.
Both Nippon Koei and Electrowatt refuse to comment; Electrowatt refers inquiries
to Nippon Koei, the team leader, and Nippon Koei says it is bound by a confidentiality
agreement with the project client, the Ceylon Electricity Board. A. N. S.
Kulasinghe, chairman of the Central Engineering Consultancy Bureau, claims
that he has not been allowed to discuss the matter in public since he spoke
to the local press after the second impoundment failed at the end of 1992.
In the subsequent newspaper account, Kulasinghe is reported as saying that
‘proper investigations had not been done before the authorities decided
to go ahead with the project. Had the investigations been done in full .
. . they would never have built the dam on this site.’ He does acknowledge,
however, that his views have not changed since this report appeared. The
Ceylon Electricity Board is even more circumspect. ‘People try to claim
things after the event,’ says S. Ganesharajah, the board’s project director.
‘On the basis of the available information and the final design carried
out by the Russians on which Gibb based its design, there has been no negligence.’

Nevertheless, after the second failure of the Samanalawewa Dam, the
government set up a one-man commission of inquiry under a local High Court
judge, J. F. A. Suza. He reported to President D. B. Wijetunge in October
1993, but the report has not been published and, according to senior engineers
in Colombo, it is never likely to be. Meanwhile in Japan, environmentalists
criticised their dam builders in newspaper articles. But, despite the involvement
of British companies, a British design and British aid, the issue was not
taken up by their counterparts in Britain.

What went wrong on the project? According to Back: ‘The grouting couldn’t
get deep enough. There were open fractures below 120 metres, the limit of
the grouting curtain. So the bottom of the curtain is left hanging and water
is getting underneath.’ It was, he says, impossible to have ascertained
this before grouting work began.

But one local geologist claims he did foresee the leak and warned that
grouting would not plug it. Piyadasa Vithanage, professor of geology and
former chancellor at the University of Peradeniya, has 30 years’ experience
of advising on dam projects in Sri Lanka. He says he was called in by Nippon
Koei to advise on the Samanalawewa Dam before the firm was concerned about
leaks. He told New ÐÓ°ÉÔ­´´: ‘I did four months’ work on Samanalawewa.
I examined the site, the mapping and the survey work for the Japanese contractors.
I told the Japanese that there was a fracture (beneath the dam site) that
could cause leakage. You cannot stop leakage by grouting. This was one and
a half years before the leakage started.’ Chisnall recalls events differently,
however. He says Gibb was the first to discover a fault in the rock and,
as far as he knows, Vithanage had not alerted anyone to this before the
leak occurred.

Cutting costs

Back in the mid-1980s when Gibb was brought in to design the Samanalawewa
scheme in detail, Back decided to move the site of the right bank of the
dam from that first proposed by the Russians. He says the move was only
‘a few tens of metres’, and amounted to ‘fine-tuning’. The aim was to cut
costs, he says.

This fine-tuning might or might not have made a difference to the leaks:
it is impossible to say. But it has fuelled anger in Sri Lanka. M. S. M.
de Silva, a leading Sri Lankan engineering consultant, told the Suza inquiry
that ‘the British consulting engineers aligned the dam to get the shortest
distance between the two abutments, which was an enormous blunder’.

What next at Samanalawewa? The dam is currently filled to a level about
30 metres from its crest; and Back says it should not be raised higher until
the repairs have been completed. But the design of the scheme places the
intake of the power tunnel, which feeds water to the turbines in the power
station, high up on the side of the reservoir. This means that water must
be within 36 metres of the crest before power can be generated. Thus, there
is very little water in store – and storage is essential to ensure power
can be generated outside the rainy season when river flows dwindle.

The reservoir has produced power during 1993, an exceptionally rainy
one in the region. But this is unlikely to last and, later this year, engineers
from Britain, Japan and Sri Lanka want to begin a rescue effort for the
dam. According to Ganesharajah at the Ceylon Electricity Board, ‘we have
practically identified the seepage area now. It is under the (old) riverbed
on the right bank.’ Back says this is where two fault lines cross what is
now the floor of the reservoir.

The ODA sent a team of experts to the site after the second impoundment
failed. They proposed that the leak in the bottom of the reservoir should
be plugged with a ‘wet blanket’ of clay, dropped from a barge onto the reservoir
bed, like a patch over a hole. ‘We (The Ceylon Electricity Board) will dump
one million cubic metres of clay, which will take two years,’ says Ganesharajah.
The blanket will be 10 metres thick and cover 10 hectares. The cost, which
both the Sri Lankans and the British hope the Japanese government will pay
through yet another loan, is put by Ganesharajah at about £30 million.

Marking time

According to ODA officials, some Japanese, alarmed by bad press coverage
over the debacle, wanted the reservoir to be emptied and its entire floor,
stretching for more than 10 kilometres, sealed with sand and gravel. This
would be much more expensive. But Ganesharajah says all parties are now
agreed on the ‘wet blanket’ plan.

Vithanage, whom Back dismisses as a doom monger, is not so sure, however.
He says; ‘They can’t seal the leak. Samanalawewa is a write-off. Whatever
they have told you, they are just marking time. Samanalawewa will become
an archaeological site.’

Not so bright either are the prospects for the Victoria Dam, once the
jewel in the crown of British engineers’ involvement in Sri Lanka. In only
one of its eight years of operation has the dam produced the electricity
expected. The dam was designed to provide 870 gigawatt-hours of electricity
per year, equivalent to about 30 per cent of Sri Lanka’s annual demand.
But in replies to parliamentary questions last year, the ODA rev-ealed that
the dam had failed to reach 70 per cent of this target in each of the previous
six years. In two years, 1987 and 1991, output was less than 500 gigawatt-hours.
In its report, Overseas aid: water and the environment published by HMSO
in mid-1992, the NAO reported that ‘firm power output is expected to be
40 per cent below the original estimate’.

The Victoria Dam, about 80 kilometres north of the Samanalawewa project
and just east of Kandy, was the largest element in a major push for economic
growth in 1977. The aim was to turn Sri Lanka into ‘the new Singapore’,
and the centrepiece was to be the Mahaweli Development Programme. The idea
was to harness the country’s major river, the Mahaweli Ganga, for irrigation
and, especially, hydroelectric power.

In the first half of the 1980s, the Mahaweli programme accounted for
44 per cent of Sri Lanka’s development spending, most of it financed by
foreign aid. The Victoria and Samanalawewa dams were Britain’s main contribution,
amounting to 62 per cent of British aid to Sri Lanka between 1976 and 1989.
Nearly 60 per cent of the cost of the Victoria Dam was met by British aid
on condition that British firms received the main contracts. This amounted
to £117 million, or more than £2 for every person in Britain.

The Victoria Dam is known affectionately at the ODA as one of the ‘jumbo
projects’ set up by the former minister Judith Hart in the late 1970s. The
concrete arch dam is 130 metres high and 500 metres long. Its construction
was dogged by difficult conditions and attracted heroic stories in the British
civil engineering press. So what went wrong?

The main problem, say Back and the ODA, is diminished flow in the Mahaweli
Ganga resulting from poor rainfall in the river’s upland catchments. Sri
Lanka had two major droughts, coinciding with the two years of lowest power.
The reservoir was low in 1987 and almost empty for much of 1991 and 1992,
when the remains of the town of Teldeniya were exposed on the reservoir
bed.

While the history of the Samanalawewa Dam looks like an object lesson
in the risks of overoptimism in geological assessments of potential dam
sites, so the Victoria Dam appears to show what can happen when hydrological
estimates turn out to be bullish.

New ÐÓ°ÉÔ­´´ has spoken to the two hydrologists who, in the late 1970s,
carried out the critical assessments for the ODA and Gibb of how much water
would flow down the Mahaweli Ganga to the Victoria Dam. John Sutcliffe and
Ben Piper, then both at the British government’s Hydraulics Research Station
near Oxford, say that there were some 25 years of data about the pattern
of river flow to go on. During that time, they say, there was nothing like
the run of dry years seen since the dam was built.

Back says that Sri Lanka has simply been unlucky. ‘Two monsoons in a
row failed. But you don’t build for the extremes. The dam is doing fine.
The only constraint is the water.’ Sutcliffe says that ‘by Third World standards
the hydrological data we had there was good’, though officials at the ODA
now privately admit that this is not saying much. Dam projects round the
world, they agree, have been bedevilled by overoptimistic forecasts of how
much water was available.

Leaking Victoria

Local specialists claim that matters are made worse by increasing leaks
either from beneath the Victoria Dam or from elsewhere in the river’s catchment,
perhaps upstream from the Kotmale or Polgolla reservoirs, says Vithanage.
In 1983, the dam site was grouted extensively to stop leaks through an
underground fault shortly after the dam was impounded. Daily pumping from
a specially built drain, or discharge gallery as it is formally known, now
copes with the problem, says Back.

Far from being the making of his country, Mediwake says the Victoria
Dam has been a calamity. It flooded 28 square kilometres of the productive
and densely populated Dumbara Valley, where fertile soils and moderate upland
climate grew export crops such as cocoa, coffee, coconuts and rubber. It
was also the market garden for Kandy, Sri Lanka’s second city. Some 30 000
people lost their land when the reservoir was filled. Mediwake lost 90 per
cent of his estate.

Much of this loss can be attributed directly to British advice. Before
Sri Lanka turned to the West for financial aid, its engineers had wanted
to construct the Victoria Dam to provide irrigation for up to 40 000 hectares
of new farmland, and only secondarily to generate electricity. But after
1977 the priorities changed, and British engineers and aid officials advised
Sri Lankan ministers, against the wishes of some local engineers, to maximise
the dam’s hydroelectric potential. They recommended building it five metres
higher than originally intended so that it could store more water and thus
provide more power during the dry seasons. This, one ODA official admitted
recently, meant ‘we were flooding a large area for a small volume gain.’
But he insisted: ‘At the time, the advantages of the extra height were very
clear. We decided we could go for something better. Now it looks overoptimistic.’

Perhaps the final nail in the coffin of Britain’s funding of dam projects
in the developing world will be the investigation into the Pergau project
in Malaysia. The current inquiry by the Foreign Affairs Select Committee
of the House of Commons, which began this month, follows a hearing by the
Public Accounts Committee in January, which in turn followed a report on
the Pergau Dam by the National Audit Office. Construction of the dam by
Balfour Beatty involves committing an unprecedented sum of money from the
Aid and Trade budget ( £234 million) to a project described as ‘unequivocally
a bad buy’.

Aid for the dam, Lankester told the committee, was agreed between Thatcher
and the Malaysian prime minister Mahathir Muhammad in March 1989. But, two
weeks later, Balfour Beatty submitted a revised price estimate for the job
that had leapt 25 per cent to £397 million. According to the review
by the audit office, published last year, ‘this took account of adverse
geological factors not recognised in earlier assessments.’ Once again, it
appears that overoptimism had landed aid officials in trouble. ‘At that
price we felt very unhappy indeed,’ Lankester told a hearing of the accounts
committee.

The Foreign Secretary Douglas Hurd told Lankester that there were ‘broader
commercial relations with Malaysia’ to be considered. And Malaysia’s former
foreign minister, Rais Yatim, said last month that the aid was ‘undeniably
linked’ with weapons purchases worth up to £1.3 billion, including
28 Hawk fighters. Such a link between aid and arms sales, if established,
would be illegal in Britain, believe committee members.

Work on the five-year project began in July 1991, just as the first
impoundment of the Samanalawewa Dam was underway. According to the National
Audit Office, it was only after work began at the Pergau Dam that it emerged
that extra steel lining would be required in the power tunnel because of
unexpectedly poor rock conditions discovered underground – a familiar story.

The Pergau affair, riddled with controversy because of its link to arms
sales, is a shabby end to British dam-building abroad, but at least it looks
like the end. The ODA appears now to have won its Whitehall battle against
funding large dams. No dams are mentioned in its most recent annual report.
Its emphasis now is on environmental projects, on funding population control
programmes, on women’s issues, on fighting poverty and improving living
conditions in large Third World cities. Back says ‘they’re not interested
in funding dams these days. It’s not flavour of the month any more.’ Perhaps
it’s just as well.

* * *

Britain’s civil engineering firms had a tarnished image overseas long
before Malaysia’s Pergau project and Sri Lanka’s Victoria and Samanalawewa
schemes. One of the more notorious episodes in their recent history began
in the early 1970s, continued throughout the 1980s and looks set to roll
on. It, too, involved the now familiar mix of hydroelectric power, politics
and poor geological conditions.

The tale concerns the doomed contract to build a power station on the
north bank of the Zambezi River, beside the Kariba Dam in southern Africa,
on the border between Zambia and Rhodesia, now Zimbabwe. Mitchell Construction
was the original contractor, and Sir Alexander Gibb & Partners, later
involved in the Sri Lankan dams, was the designer.

Mitchell, which was a medium-sized British contractor at the time, won
the job in April 1971. But the project quickly wrecked the company when
unexpectedly weak rock delayed excavation work underground, and Mitchell’s
claims for compensation were turned down. Within two years, Mitchell had
abandoned the site and gone into receivership, claiming it had been misled
by a geological report, prepared by Gibb, that guaranteed ideal ground conditions.

The project was eventually completed by a Yugoslavian contractor for
a sum, according to David Morrell, Mitchell’s chairman, that was more than
three times as much as Mitchell’s original bid, and substantially more than
his company had sought in compensation. For its part, Gibb said that the
new contractor, Energoprojekt, finished the job more quickly than Mitchell
had claimed was possible. Subsequent claims and counterclaims peppered the
British construction industry for almost two decades.

For years, the consensus within the industry was that Mitchell had overstretched
itself, that the company was on its last legs anyway and that it had bid
too low for the Kariba work in a final, desperate attempt to secure its
commercial salvation.

Determined to give his side of the story a full airing, Morrell eventually
produced one of the most damning tales of the British construction industry
ever seen. What makes his book so intriguing is that Morrell had rubbed
shoulders with the industry’s most prominent figures but had refused, to
his ultimate cost he reckons, to join the ‘inner circle’.

In Indictment: power and politics in the construction industry, published
by Faber & Faber in 1987, Morrell condenses a huge volume of contract
documents, eye-witness accounts and contemporaneous notes into an exhaustive
treatise on the British industry in general and the Kariba project in particular.
In retrospect, he writes, the job was a front to supply hydroelectric power
to the supposedly outcast and illegal Rhodesian regime of the 1970s. And
when he started to complain about the way his company was being treated,
he says, all the main parties, fearing exposure of the political deception,
closed ranks.

Reviews of the book by the construction industry press were mixed, with
some trying to dismiss Morrell as obsessive and his story as extravagant.
But when one industry newspaper was more precise in its criticism, suggesting
that Mitchell’s demise was the result of commercial mismanagement, Morrell
sued. In 1990, in one of the British construction industry’s biggest libel
settlements, he was paid damages of £100 000, plus his legal costs
of more than £200 000. Morrell, currently on holiday in Australia,
says he is preparing another ‘expose’ linked to the Kariba farrago.

Additional research by Sue Carpenter in Sri Lanka and Peter Hadfield
in Tokyo. Fred Pearce is the author of The Dammed, published in 1992 by
The Bodley Head. Price £14.95.