Prem Shankar Jha

Oil, however, is not the only natural resource over which conflicts are likely to break out in the future. Another, even more dangerous, casus belli is water. Disagreements between upper and lower riparian states over their rights to river waters have sparked many wars in the past. Today, with the human population exceeding 6 billion, and per capita water consumption rising rapidly in developing countries, the threat of such conflicts is higher than ever. While solar energy cannot augment the supply of water, it can eliminate the main cause of conflict by making hydro-electric power projects redundant. It can do this by providing peak power far more cheaply, rapidly, and in an eco-friendly way, than either storage or run-of-the-river power plants. Nowhere would the benefits from this be felt more acutely than in the triangle of countries that share the waters of the Brahmaputra (Yarlung-Tsangpo river), China, Bangladesh and India.

The possibility that China might build dams on the Yarlung Tsangpo to divert some of its waters to the arid regions of the north was first mentioned at the first International conference of the Global Infrastructure Fund in Anchorage, Alaska in 1986. Although Chinese officials rubbished the idea as being impossibly expensive to implement, they did not rule out the possibility of constructing dams on the river to generate power. This ambivalence raised understandable alarm in Bangladesh and India but Beijing sought to allay their fears by assuring them that it intends to build run-of the river dams that will redirect, but not stop, the flow of its waters into India and Bangladesh.

These reassurances have not, however, prevented China and India from entering an undeclared race to capture the hydro-electric potential of the Brahmaputra river basin. Chinese writers began to air plans for harnessing the Yarlung–Tsangpo in 2005 but it is possible that India began to formulate its plans after the publication of a book by Li Ling in 2003 titled Tibet’s waters will save China. As the downstream riparian, India is hoping to establish first user rights to stake its claim to an uninterrupted flow of the Brahmaputra’s waters. In International law first user rights start upon the completion of a project, so the number of projects that India has signalled it will take up in the Brahmaputra basin has risen rapidly from 146 announced in a s ten-year Hydro-Electricity Plan unveiled by India’s Central Electricity Authority in 2007, to about 200 today. What is more a scramble has developed to start as many of these, as soon as is possible.

The pace of planning and implementation has also picked up in China. Citing the need to cut down CO2 emissions the 12th five year energy Plan, unveiled in 2012, shifted its emphasis back onto giant hydro-electric projects once more. Chief among these is the exploitation of the hydro-power potential of the Yarlung–Tsangpo. In all, China intends to build 40 dams on the river and its tributaries. Of these 20 dams on the Yarlung–Tsangpo are expected to generate 60,000 MW of power while 20 smaller dams upon its tributaries are expected to generate another 5.000 MW. Eleven of the 20 projects on the Yarlung–Tsangpo are to be located between its source and the Big Bend where the Brahmaputra turns northwards, executes a huge ‘U’ turn and falls from 3,500 metres on the Tibetan plateau to 700 metres, in the undulating hills of Arunachal Pradesh in India. These will generate 20,000 MW of power. The balance, of 40,000 MW will be generated from the Big Bend. The plan, put forward by ex-Premier Li Peng’s family-owned corporation, the Three Gorges Dam Company, is to build a vast tunnel under the ridge that separates the two arms of the Big Bend, and divert 50 billion cubic metres of water a year to the south-eastern slope where it will fall over nine cascading hydropower dams to generate 40,000 MW of peak power. India, for its part, plans to generate 22,000 MW from two large dams on the Brahmaputra in Arunachal Pradesh and 10,000 MW from dams on its tributaries. In all therefore the two countries plan to generate 97,000 MW of power from this tiny region of their respective countries.

Such ambitious but conflicting plans were bound to have a political fallout. Its first indication was an abrupt announcement by the Chinese Ambassador in New Delhi, that China considered the whole of the north-eastern state of Arunachal Pradesh to be a part of Tibet. This was a complete reversal of it’s earlier position, developed in a succession of bilateral negotiations since 1994, that China was prepared to settle for a substantial modification of some parts of the existing temporary boundary, called the Line of Actual Control. What gave the announcement added significance was that it was made on Indian television three days before President Hu Jintao paid a state visit to India. The announcement took the Indian government by surprise and was followed by three years of rising tension along the border. China began to refer to Arunachal as “South Tibet”, and to its principal monastery at Tawang as Tibet’s second most important monastery after Lhasa. It also began to deny visas to Indian offiicals who were serving in Arunachal Pradesh. The tension was not defused till there was a meeting between Premier Wen Jiabao and Prime minister Manmohan Singh designed specifically to prevent its spilling over into military conflict, at Hua Hin, Thailand in October 2009. However, if the hunger for power and water continues to grow, the respite this meeting gave the two countries could prove temporary.

All this took place over plans that were little more than engineers’ dreams. And the dreams have kept growing larger. The Big Bend region is one of the least known area of the world. In all some 360 dams are to be built on slopes with as much as a 70 degree gradient, at the meeting point of three of the youngest and most unstable mountain ranges of the world. But neither China nor India have made even a rudimentary assessment of the impact that tearing down billions of cubic metres of rock and earth to build dams, tunnels and roads, and storing millions, in some cases billions, of cubic metres of water in will have upon the stability of the earths crust in this region.

This neglect is deliberate, for both governments are fully aware that the Himalayas have seen a succession of the most severe earthquakes in recorded history. Four of these, measuring 7.8 to 8.9 on the Richter scale occurred within a span of 53 years between 1897 and 1950. The first and last occurred just 53 years apart in the region immediately south and west of the Big Bend in the Brahmaputra. The 1897 earthquake measured 7.8 on the Richter scale (equivalent to the explosion of 7.6 million tonnes of dynamite, or a medium sized hydrogen bomb) and caused widespread damage and loss of life in what was then called upper Assam. It was caused by the build up of pressure as the Indian (tectonic) plate pressed against the Shillong Plate, a part of the far older Eurasian Plate. The quake occurred when the former shifted 11 to 16 metres as it dived under the latter, over a stretch of 110 kms, along what is known as the Oldham fault. This is one of the largest tectonic shifts recorded so far. Its effects were felt through the entire earth’s crust over this length.

The 1950 earthquake was the severest recorded in the Himalayas. It occurred at Rima, Tibet, not far from the site of the 1897 ‘quake. Measuring 8.7 on the Richter scale it is one of the ten most severe earthquakes in recorded history. Its epicentre also lay on the fault line where the Indian continental plate smashes into the Eurasian plate and dives beneath it. Survivors from the region reported mudslides damming rivers and causing them to rise high when these broke, bringing down sand, mud, trees, and all kinds of debris. Pilots flying over the area reported great changes in topography, caused by enormous land slides, some of which were photographed. An aftershock of this earthquake, a long way to the west of it, was severe enough to register 8.6 on the Richter scale and caused avalanches and floods that destroyed swathes of forest in the Mishmi and Arbor Hills.

Earthquakes in the Himalaya regularly cause landslides that block rivers, causing them to rise till the pressure of the stored water breaks through. The result is a flash flood downstream that causes havoc among the villages and towns that lie in its path. The 1950 earthquake and the 8.6 magnitude aftershock that followed caused avalanches that blocked several of the tributaries of the Brahmaputra. One such dyke in the Dibang valley broke quickly and caused relatively little damage. But another, at Subansiri, broke after water had collected behind it for 8 days and unleashed a 7-metre-high wave that submerged several villages and killed 532 people. In all, the 1950 earthquake killed more than 1500 people in Assam. Geological studies, including the radio carbon dating of sand found on the surface, have uncovered at least one other giant earthquake in the same area that took place in 1548 and two earthquakes in the central region of the Himalayas, that were severe enough to break the earth’s crust all the way to the surface. These occurred in 1255 and 1934.

The 1934 earthquake, which measured 8.1 on the Richter scale and had its epicentre about 10 kms south of Mount Everest, devastated north Bihar and Nepal and killed at least 30,000 people. This occurred before any dams had been built. The dykes that were overwhelmed were of mud and broke in a matter of days. But earthquakes of this magnitude will almost certainly break concrete dams as well. For the Richter scale is a logarithmic scale. An 8.1 magnitude earthquake releases three times as much energy and an 8.7 magnitude releases 23 times as much as a 7.8 magnitude quake. Should any of the proposed dams crack during an earthquake or an ensuing flood, the colossal wave of water, mud and boulders that will be released will kill millions and completely devastate the areas of Tibet, India and Bangladesh that lie in its path. The overwhelming majority of deaths and damage will take place in India and Bangladesh.

India got a foretaste of this some years ago when a flash flood in the Yarlung-Tsangpo, caused by torrential rains and landslides, wiped out an entire island in the Brahmaputra, killing nearly all who lived on it. Chinese hydrologists knew that the flood would occur but did not warn their Indian counterparts. India got another foretaste of what can happen in June 2013 when three days of torrential rains in the valley of the Bhagirathi river, one of the two main tributaries of the Ganges river, caused landslides, that blocked a tributary and destroyed the entire town of Kedarnath when the mud dyke broke, killing between 5,000 and 10,000 people in a matter of hours. Over the previous twenty years the hill slopes overlooking the Bhagirathi valley had been ravaged by the construction of dams and tunnels for the Tehri hydro-electric project, the second largest in India. The rains therefore brought on a catastrophe that many had feared, but hoped would never happen.
The Tehri project has a generating capacity of only 1,000 MW, one fortieth of what the 9 dams at the Big Bend will have, and almost one hundredths of what India and China intend to create in the region. Twenty years from now, when thousands of explosions of dynamite, that have been used to build tunnels for 240 or more power projects, have loosened billions of tonnes of earth and the population has increased tenfold, the entire region will be a calamity waiting to happen.

The Solar alternative

China has become the lowest cost producer of solar photovoltaic panels and heliostats in the world, and India has begun the construction of the largest solar PV power station in the world (4000MWe), as well as the largest Solar thermal power station in Asia (250 MWe). But neither country’s government has grasped the fact that every step they are taking down this road is making their hydro-power development plans redundant For in Spain, a solar thermal power plant that began to supply power to the grid in 2011 has shown that China and India establish 97,000 MWe of generating capacity on less than 10,000 km2 of land. This sounds like a lot of till one realises that it is only one twentieth-fifth of the land area of the Thar desert (that straddles India and Pakistan), and is only a fifth of the land that the Indian government has already reserved within its part of the desert for the construction of solar power plants. It is also barely two-thirds of one per cent of the land in the Gobi desert of China.

The technology is incorporated in Terresol’s 19.9 MW Gemasolar plant at Fuentes de Andalucia. This plant was approved in February 2009 and came on stream in May 2011.It is the first solar plant in the world that is designed to provide power throughout the day in exactly the same way as coal fired plants do today. To do this it stores enough of the sun’s energy, collected from 2650 heliostats in a mixture of potassium and sodium salts to generate power for 15 hours a day using the stored heat alone. Round-the-clock solar power has become possible because this combination of molten salts loses only one per cent of the stored heat in a day. In an average year, therefore, stored heat generates 5475 hours of power while direct sunlight is required for only the remaining 1025 hours, or less than 3 hours a day. Gemasolar is thus generating 129 GWh of power per year, and supplying 110 GWh to the national grid.

It does this from 304,000 sq metres of heliostats spread over 185 hectares of ground. Assuming that the power generation plant takes up another 15 hectares, Gemasolar requires on square km of land on an average for every 10 MWe of generating capacity. Thus the entire 97,000.MWe of generating capacity that China and India are aiming for in Tibet and Arunachal Pradesh can be set up on 9,700 km2 of land.. This sounds like a lot of land but it is less than a quarter of the area of the Thar desert in the Indian state of Rajasthan that the Indian government has already reserved for solar power plants in the second phase of its National Solar Mission. It is also less than one per cent of the land area of the Gobi desert in China.

The Gemasolar plant demonstrates that solar thermal power has many other advantages that make it far more economical than hydro power. First, unlike the Three Gorges dam project that took just under 12 years to complete, the Gemasolar plant came into operation 27 months after it received the go-ahead from the Spanish government. This means that power will be available ten years sooner with solar power plants. In those ten years it will generate additional GDP, and therefore additional savings for investment. Second with power available for 6,500 hours a year every megawatt of installed solar capacity will generate 50 percent more power than the Three Gorges power plants, whose maximum plant availability has been 4,360 hours . It will also be double of the power that India extracts from its 1500 MW Nathpa Jhakri power plant in Himachal Pradesh, in the western Himalayas.

Solar power can meet the need for power at peak load times that hydel power is designed to meet, in as little as two years. Concentrated solar thermal power (CSP) plants can supply power whenever it is demanded. What is more, thanks to the steep fall in the cost of solar panels triggered by China, they can now do so at a substantially lower cost. This has not been immediately apparent because economic appraisals measure only the initial capital cost and the number of hours the plant is able to deliver. They almost never take the cost of delay into account because this is a social cost borne by society and not the investor. Yet this is by far the largest and most important of the three costs .

Comparing the true costs of various energy from various sources

It is not easy to make best practice estimates of the cost of generating power from different sources of energy, specially if they are to be found in different countries. To make this possible, the table below uses the bare capital costs estimated for thermal, nuclear, hydro and solar power in the US, in April 2013, and running costs and Plant load factors found in china, India and Spain .
Comparison of economic cost of alternative power sources. 2013


Thermal     Nuclear      Hydro       Solar
Cap cost/ MW capacity(US$m)           3.246           5.530        2.936      7.3143
Plant Load Factor (hrs/yr)                  64002           78842      28932     6.5004
Cap cost per MWh($)                                  507                701          1014          664
Construction period(yrs) 5                           8                  10                2             55
Net saving foregone 6                               1,350           2,970          4050             0
True Cap cost per MWh ($)7                 1,857            3601            5064         664


  1. Capital cost based upon the technical specifications of the Gemasolar 19.9 MW central tower CSP set up at Fuente de Andalucia in Spain, which came into operation in May 2011.
  2. These PLFs are the actual experience in India. The PLF for nuclear power plants is that achieved in plants that have not experienced difficulties in obtaining uranium.
  3. Based upon price of heliostats prices quoted by Chinese suppliers ($120 per sq.m)and the assumption that these account for half of the total cost of the solar thermal plant. One American supplier is also offering these at $126.
  4. PLF based on 15 hrs of supply per day from stored heat and 3 hours from direct sunlight being delivered by the Gemasolar plant.
  5. Actual construction period at Fuentes de Andalucia was 2yrs 3 months.
  6. This is calculated as the saving out of additional GDP that is foregone during the longer gestation period of the project. In India the GDP in 2011-12 was $1.5 trillion and the saving rate was 36 percent. These ratios have been applied to all the four types of plants
  7. This is row 4 + 6.

In the above table the ‘private’ cost, i.e. the bare capital cost that appears in the investor’s balance sheet, of delivered power (MWh) from different sources is given in Row 3. The ‘social’ cost, i.e. cost to the nation is given in row 6. This is the private cost plus the savings out of GDP foregone through the non-availability of power to the country during the construction period of the project, beyond the earliest date on which the power can be supplied by the technology with the shortest gestation period, i.e solar power. Needless to say, this is only that part of social cost which can be readily calculated in terms of GDP foregone. It does not include less measurable but equally tangible costs, such as employment foregone and damage to the environment.

The table shows that while the ‘private’ capital cost per MWh (excluding operational costs) is lowest for coal-based power plants, it is only 30 percent higher for a solar thermal plant. This difference disappears entirely when we add their operational costs, which are highest for coal based and lowest for solar thermal plants. However it is when we look at social costs that the chasm widens. The social cost per MWh of solar thermal power is one eights of the cost of hydr-power and one third that of coal based power. All of its other benefits – benefits that can save human beings and most other species on the planet from extinction – are additional to the economic benefits calculated above.

Read More