In March 1885, Acharya Jagadish Chandra Bose wrote in his diary: “I have been long thinking whether the vast solar energy that is wasted in the tropical regions can in any way be utilised. Of course trees consume solar energy. But is there no direct way of utilising the radiant energy of the Sun?”

Along with his research on radio waves, he then proceeded to invent a device called Tejometer which could be called a solar energy conversion device.

However this method was not followed until the National Physical Laboratory was set up in Delhi in the 1950s when solar heaters and concentrators were among its first projects. The West was busy exploiting Middle East oil while India had ample reserves of coal. It was not till the energy crisis in 1974 that the focus turned to forms of Renewable Energy.

The Silicon solar cell was invented at Bell laboratories by Chapin, Fuller and Pearson in 1954, just after the transistor had yielded efficiencies of only 4 – 6 per cent. However, as a case of serendipity, it was just in time to power Telstar and other satellites.

With research, efficiencies increased to near 20 per cent, but the cost / watt remained high. Nowadays space vehicles are powered by 3-layered tandem solar cells based on Gallium Arsenide which can reach 43 per cent efficiency.

Situated in the Tropics at the same latitude as the Sahara, India is blessed with abundant solar energy, with an average of 300 sunny days / years and 7 – 9 KWh energy per day.

This is much more than China or Europe. However, India lacked a leader such as Homi Bhabha who led our march to Nuclear Energy. As a result, China has captured the Indian solar photovoltaic market after annihilating the European, in particular the German, Solar PV manufacture.

The Internet shows that Chinese PV modules are available in India at Rs. 25 per watt, while domestic manufacture sells at Rs 92 / watt.

Was it always like this? Till the mid-eighties Indian PV output, led by Central Electronics Ltd (CEL Sahibabad) far exceeded that of China. Attending PVSEC in 1986 in Beijing, I found that Chinese technology was fairly backward and limited to so-called Electric Factories.

In India, solar irrigation was successfully taken up in Tripura and solar water heating systems were promoted by a Gujarat company. With a large internal market, bureaucrats failed to explore export possibilities. The basic material for 90 per cent of solar PV, is, as is well-known, Silicon.

Derived from white sand, quartzite rock or even rice husk, there is no dearth of the source material. On the other hand, India is highly dependent on Australia, Canada and Tajikistan for the import of Uranium for its nuclear reactors.

The tragedy is that India does not manufacture even 1 kilogram of semiconductor-grade Silicon. In 2013 China produced 90,000 tons, Korea 70,000 tons not to mention the USA, Norway, Germany and Taiwan.

China has increased its capacity to 227,500 MT in 2016 growing to 319,000 MT in 2017 and 450,000 MT by 2018. It already has 54 per cent of the Polysilicon market and yet has imported 159,000 MT in 2017.

Single crystal Silicon is at the heart of all IC chips, mobile phones, TV sets and power devices. These are microelectronic products while solar PV is an example of microelectronics requiring 5 – 8 tons of Polysilicon per Megawatt of electricity generation. India has recently increased its projected output from 20 GW to 100 GW. This will require 500,000 tons of Polysilicon.

How can we compete with China? Till the 1990s there was a small plant in Tamil Nadu with a capacity to produce upto 100 tons every year, but the actual output was 20 tons per year.

The technology was outdated but instead of modernising and expanding, this was forced to close down. Without production of the basic raw material, is it possible to compete with China?

In the mid-eighties there was a debate worldwide on alternative low-cost technologies including Cadmium sulfide and Amorphous Silicon (a-Si) solar cells. These did not require energy-intensive crystal growth and could be fabricated as thin films on glass substrates.

What was not realised by its proponents was that not only were the semiconductor properties of a-Si much inferior to those of crystalline Silicon but the cells also degraded under sunlight.

As is usual when large funds become available, three very senior scientists, with scant knowledge and experience of semiconductor technology, jumped into the fray for developing amorphous Silicon and cornered the major part of funds, amounting to crores in those days.

A study group consisting of Dr B K Das of NPL, Prof O P Agnihotri of IIT Delhi, myself, then at IIT Kharagpur produced in 1987 a “ State of Report on Polycrystalline Silicon Technology for Solar Cell Applications” under the aegis of the Department of Non-conventional Energy Sources.

However, all attempts to develop Polycrystalline Silicon technology in the country were abandoned and the limited resources on “Si”. These cells, being flexible, had a small market share till the mid-Nineties, but with their module efficiencies peaking, 8 per cent of these are now defunct the world over.

Large companies in the USA have declared bankruptcy. The wrong choice has crippled the Indian PV programme. C-Si modules have reached over 20 per cent efficiency and command over 88 per cent of the solar PV market and now have a projected life of 25 years. Due to shortsightedness. India now depends entirely on foreign imports.

That `low cost’ goes with low quality was realised too late. For solar parks producing MW of power, the area required is obviously of paramount importance. With c -Silicon the requirement is a few sq miles / MW. From $ 100 per watt in the Seventies, solar PV prices have plummeted to less than $ 0.50 / W with grid parity being achieved by 2020.

Is there another power source that is so modular that can be scaled from watts for mobile charging to gigawatts for industries? Another unique advantage is the short time for installation.

While thermal and nuclear plants take anything from 6 to 10 years to become operational, megawatt level solar plants can be installed in a few months and gigawatt level in a year. The only thin film technology with 8 per cent market share is based on Cadmium Telluride, which is also not manufactured in India.

There are means to augment overall efficiencies and energy output. Bifacial Si cells are available, where the rear side is away from the sun. These can gather light from the surroundings ~ the albedo ~ thus adding 10 per cent to the output.

A mechanical arrangement can track the sun during the day (single axis tracking) which could increase the energy output by 10 to 15 per cent, at an increased system cost and complexity.

Taking this a step further, double -axis tracking could follow the sun from summer to winter and increase by another 5 per cent, since normally PV modules are kept inclined at an angle equal to the location’s latitude.

Concentrators can be used to focus solar radiation using plastic lens on smaller cell areas but this requires double-axis tracking and cooling.

Research is being conducted in the quest for lower cost and higher efficiencies. A new slogan is the search for “earth-abundant materials” but what could be more abundant than Silicon, the second most abundant material in the earth’s crust? Its only disadvantage is that it cannot absorb enough light as a thin film. The latest material of promise is based on organic-inorganic perovskites ~ tri-methyl lead iodide to give it the full name.

Discovered about 10 years back, efficiencies have leapt to 22 per cent for small area cells. Increasing the area to square metres throws up the usual problems of uniformity.

However, these films are sensitive to oxygen and water vapour and so stability over 1000 hours has remained a major problem. In comparison Si cells retain efficiencies over 25 years. Can one identify any car or electronic gadget that has a comparable life-span, with the price per watt plummeting by a factor of 100 in the last three decades?

This not to suggest that solar energy has no drawbacks ~ diurnal and variable nature, low intensity of only 1 KW / m2 at best, the need for batteries for storage purposes.

Thus solar energy must be supplemented by conventional power generation. Nevertheless, the contribution of renewable energy together with wind farms could be increased from the present 11 to 30 per cent over the next decade, reducing the dependence on coal and imported oil. In the USA, renewable energy (including hydroelectric) already accounts for over 30 per cent.

The proponents of nuclear energy cite Einstein’s equation E = mc2 to promote the advantage of Uranium over coal. However, for nuclear, as well as for thermal power stations, energy is first produced as heat which is then converted into electricity.

It has been shown that a 1000 MW nuclear station requires 1 million gallons of water per day. No wonder most nuclear power plants are located near the sea. With a head-start of 70 years, nuclear energy in India has just reached 3 to 5 per cent of the energy basket.

Nuclear energy is sometimes labelled as “clean” and “green” in that it does not produce CO2 directly, but the hazards of Chernobyl and Fukushima speak to the contrary.

The writer is author of Photovoltaic Science & Technology (Cambridge University Press), 2018