Solar photovoltaic energy: today’s crisis and tomorrow’s technologies

Photo Didier Roux / Member of the French Academy of Sciences, of the French Academy of Technologies, Director of Research & Innovation, Saint-Gobain Group / April 19th, 2013

Could solar power provide some of the needed energy of the future? The much improved availability of natural gas and the crisis that the photovoltaic industrial sector has been experiencing since 2011 serve to make us cautious, viz., not to be over-optimistic. On one hand, we can witness the strategic policies chosen by China and, on the other, the expected advent of new PV cells, could together change the scene. Consequently, we must carefully examine and assess the economics, their dynamics and the supporting technologies.

1. Today’s technologies
ParisTech Review – Today, photovoltaic solar energy is produced thanks mainly to first generation, silicon-based, cells. However, the next generation devices are currently being developed. Could you give our readers a rapid overview of the technologies involved and the issues?

Didier Roux – Silicon-based cells are the mainstay of solar PV technology today, and they still have a future. We can recall the principle that makes a PV cell work; it is simple and has been known since the 19th Century, even though we had to wait for Albert Einstein to really understand the physics of light.

Light is composed of photons, which we can define as “packets of elementary energy”. When they meet matter, the “shock” allows the energy released to be transfered to electrons. This is the so-called photo-electric effect. Usually, the electron finds its way back and the energy brought by the photon dissipates. First generation PV cells modify this process: they capture and forward these electrons, delivering a DC current.

To do so, we use silicon. Why silicon? The simple answer is that it is an abundant element (usually silicon dioxide SiO2), used in a purified form (thanks to the electronics industrial sector) and because it has the required properties. There are several different configurations: mono-crystalline modules which are more expensive but present excellent efficiency ratings per m². Then we have poly-crystalline modules which are a little less efficient and cost far less. Lastly, we can mention amorphous silicon devices i.e., their molecular structure is not that of a crystal. They have the lowest efficiency. Amorphous silicon modules are one example of second generation PV cells, also known as “thin-film” devices.

This thin film cell technology today appears to draw a lot of attention. What exactly does the term ‘thin film’ cover?

You are quite right; the real issue lies in the thin films. There are several players in the field, the First Company, which is American and the French Group Saint-Gobain who have both invested in this technology, which presents some clear advantages.

A word about the technology: fundamentally you place extremely thin layers – we’re talking of several microns only – on a substrate which may be glass, metal or plastic. Industrially assembled solar cells use two types of material: cadmium telluride (CdTe) or 4 component layer devices (copper, indium, gallium and selenium).

You just mentioned ‘second generation’ advantages: we know for example that by using this method, you can make much larger surface modules and, moreover, they can be sliced fairly easily…

The main advantage is that, in terms of the quantity of active matter and the deposit methods used, the industrial processes are less expensive. They are not totally “mature” yet, from the industrialists’ viewpoint, but if we take into account the facilities we are building now, we can observe that progress in this area is very rapid indeed. Using CdTe technology is already proving less expensive than making polycrystalline silicon modules.

Before we look at the marketplace and conditions conducive to second generation development, could you just give us an overview on third and fourth generation technologies, apparently still in their R&D phases?

You are quite right to make this point. We are talking about technologies that are as yet only on the drawing boards, in our laboratories. We can group together some quite different technologies under the heading “third generation”. In some instances, we use organic molecules and/or mineral nanoparticles. The systems we are investigating can absorb light and transport electrons; they are complex and display fragility, viz., they can suffer from radiation damage. But the technologies are attractive inasmuch as, potentially, they will be very inexpensive with efficiency levels around 10% (with so-called Graetzel cells). But they also present a drawback, for which we have not yet found the solution, viz., the natural decay of the organic dyes in the device. The expected operational life span is notably shorter for organic devices, compared with the 20° to 30 year expectancy fro conventional silicon cells.

Multi-junction concentrator cells belong to the category we call “fourth generation” PV devices: the film layers are built up in such as way as to capture solar energy in different parts of the Sun’s spectrum. The main drawback is that today they carry a very high price tag. However, we can also assert that there is no major obstacle to making the fourth generation profitable in the future. We ‘simply’ have to improve, step-by-step, every link in the chain.


But let me underscore one point: this multilayered technology will in fact find itself less in competition with the generations mentioned earlier, than with concentrated solar power devices (CSPs).

Why is this so?

The reason is that the newer technologies require lots of sunlight to operate correctly and even adjunct sun-tracking systems that direct the modules to face the sun as it moves. This is a very different configuration than that used for thin film devices which remain operational even with diffuse lighting or an indirect lighting angle. Multi-junction solar cells like the CSPs I mentioned are really only efficient when there is a very intense solar radiation. In other words, they are perfectly adapted for operations in desert zones. Both technologies will then come into competition with each other.

Only a few years back we saw some large-scale projects for concentrated solar energy installations; apparently this technology has been somewhat abandoned. Is this true?

Yes it is true but not very difficult to understand. If we tend to neglect CSPs today, it is because conventional PV installations have become far more profitable. Siemens, for example, withdraw from the Desertec project. But let‘s not talk about abandoning the project. As was the case for the Font-Romeu Odeillo solar furnace, still operational, Desertec is a pilot scheme and a lot of progress has still to be made. Although concentrated solar energy has been around for several decades now, it has not achieved anything like the maturity of industrial solar PV technologies. On a world scale, CSP only represents at most 1°% of the solar PV capacity installed. In other words, CS still has plenty of room for improvement.

We sometimes read that implementing solar energy facilities in desert zones, far from the consumer outlets, would call for a high level of investment, in terms of grids, distribution… i.e., they carry a steep overhead cost. How do you feel about this?

True again; the very idea of producing power in desert zones to feed Western countries does entail the inconveniency of line losses, given the distances involved. Having said this, we can observe that the countries close to the deserts need more and more electricity as their populations grow rapidly. Concentrated solar power (CSP) or multi-junction PV arrays could lead to a relevant solution to their energy procurement problems. Scientists are working on these issues: the problems of long-distance power transmission are currently being extensively studied and it is already obvious that applicable improvements do exist.

2. Possible future developments for solar energy technologies
Production intermittency is often presented as one of the major obstacles to development of solar PV… Your thoughts here?

Exact. If solar origin electricity increases, it will require a development of electricity storage capacities and this is a one of the major challenges for energy procurement in general. There are some possible solutions – chemical storage via hydrogen for example, or mechanical storage … – but for the time being these solutions are not yet seen as industrially feasible. However, there is no lack of imagination or creativity deployed here. The question is now to identify a profitable approach that can be market readied. This problem has to be addressed and we have no doubt that industrial solutions will develop and come on line.

There is also the question of distribution, with the now well-known “smart” grids. What is you opinion here?

Both elements are connected. “Smart” grids, i.e., they are intelligent, sophisticated and piloted and are necessary as we are led to accepting less and less flexible technologies and, furthermore, disseminated throughout the land ands its installations. They do offer a degree of progress but lead, concomitantly, to a series of questions. Do we, for instance, really need to have large-scale grids that go everywhere? Would it not be more judicious to develop and install grid-independent production, networks and storage for the excess production? Imagine you live in an Indian village, such as a Dhani in Rajasthani, that is, a cluster of homesteads with no connections to the world around them; suppose, moreover, that you have the means to produce the electricity you need and the facilities to store any excess production.

You might reasonably ask yourself – would it be better for me, economically, to connect my home to a grid or should I rather produce “at home” and store the electricity I do not consume? We have a similar situation developing with mobile phones. Some countries, we note, are just not installing any wired networks. Countries that have not as yet invested in large-scale grids prefer now to develop small, local networks rather than large inter-connected networks. “Smart grids”, for these countries may well consist of building and operating local power production/storage models. Today we know how to store electricity for a reasonable cost; there are many places round the world where there would be no interest at all to propose and install grid type networks.

Those opposed to solar energy sources insist that they are still very costly. When, do you feel, solar power will become profitable without being subsidised?

Let me first of all refute a preconceived idea we sometime hear- that PV models consume more energy to build them than they can produce in a normal operational life span (ten or so years). I could have agreed to this assertion in the 1960s, but today what we call “return in energy” (ROE) is attained in about 1 to 3 years after commissioning of the plant, depending of course on the location and on the technologies used.

Now, returning to your question we can compare the competitive edge of solar energy to other forms of electric energy source. But let us be clear: when we talk about electricity, we must not reason in the same manner if we are considering a field of sensors built and operated by an industrialist or positioned on a private roof-top. For the industrialist, what counts is his cost of producing the electricity compared with the cost using other technologies. For a private home-owner, the important factor is the price at which he buys his electricity (including buy-back policies) that matters.

Let us examine the case of private home-owners. We refer to network parity when the cost of production equals the going market price level. This means that the relevance of solar energy is not the same, country by country. In France, for example, the predominance of nuclear power generation allows EDF’s customers to benefit from relatively cheap electricity supplies. But recent drops in the costs of installing and running solar PV installations have changed the deal. I’ll come back to this point. The cost of production for a private customer, if we amortise the initial investment outlay over 20 years, comes to about 0.2€/kWh while the price of electricity in France is 0.1€/kWh. These figures show that network parity has not yet been attained in France. Not so in Italy, for two reasons – 1° electricity is sold at a higher price there (to the consumer) and 2° there is plenty of strong sunlight to hand. The Germans pay about 0.3€/kWh have reached network parity as have certain regions in the USA, in California and in Utah, for example. But parity has not been attained on the East Coast.

The spectacular drop of the cost of PV panels has made solar energy more attractive, but we should not overestimate the consequences of the price drop. In France, the sales price for a panel represents only 20% of the investment needed for a private rooftop installation, and 50% for a power station sized facility. Savings must be sought on the components other than the panels themselves: the panel assembly arrays, the ancillary equipment needed… installing a rooftop installation in Germany costs half as much as it does in France.

Let us now consider the question from the points of view of the electricity producers. The comparison here is made among production costs, even if the latter include items such as the impact on the environment or the cost to be energy independent. And given that investments are needed, it does appear judicious to compare electricity obtained from solar installations with those projected for future nuclear power stations, and not the costs of running today’s nuclear sites, commissioned as they were 40 years ago. Let us examine the French situation again: for a solar power station located in South France, production of electricity would cost 0.1€/kW, compared with 0.08€/kWh expected from an EPR – generation III nuclear power station. There is a difference here but it is not as striking as it was 3 years ago when the difference ratio was 1:4.

3. Market-place dynamics
You mentioned earlier that part of the price drop would stem from a sharp price drop for solar PV panels. Yrs 2011 and 2012 were indeed very difficulty for the PV industrial sector, with several module manufacturers going bankrupt in Europe and the US… How is the sector bearing up today?

The current situation is paradoxical, to say the least. The crisis extends to all PV module assemblers; everyone is losing money, including the Chinese. But the electricity utility operators are buying the modules at very low prices. Numerous projects for solar PV arrays in California or again in Italy, in areas where there is a lot of strong, natural sunlight; the reason is that it is profitable to do so.

Today Asia accounts fro 70% approx. of the world solar panel production. And the Chinese factories involved here are operating at only 50% capacity simply because there are not enough purchasers around. Clearly there is a problem of overcapacity.

In the marketplace that prevails today, were we to apply strict pricing rules, a number of these company would necessarily go bankrupt and into receivership. But the Chinese authorities, fully aware of the situation, firstly no longer demand a reimbursement for government funded investments – thus amounts to a subsidy. Moreover they have launched a huge PV equipment pan, which in the long term will enable them to absorb part, at least of the overcapacity.

How would you personally characterise Chinese strategy here? Do you feel, for instance, that they are waging a war to stymie world competition for their products?

I am not of this opinion. But I do think the Chinese were led into making high investments without anticipating the risk of overcapacity. The interesting point o note is that the Chinese have not decide to close their panel assembly factories and therefore the country will continue massively to produce solar panels ‘Made in China’. The end result will probably be that all other PV panel assemblers in the world will be forced out of the market-place, except maybe that few companies that have demonstrably high quality products, such as SunPower Corp. (US) who will doubtless survive. It is more than likely that by 2020, the Chinese will practically be alone in the solar PV market. In up to that horizon, as demand will gradually catch up with supply offer, panel production will become economically viable. If on the contrary if we see a merging of the Chinese companies accompanied by massive closing of manufacturing and assembly sites, then the price of solar module will rise and the other international actors could become competitive again.

Some companies can hope to survive, inasmuch as they have totally switched their business model. This is the case, for instance, of First Solar (US) a specialist for thin film technologies using cadmium telluride (CdTe). They can turn out the most inexpensive modules but, even then, as prices fall, they became fragile and at risk. The company’s executives then opted for a strategic change of course, moving into the electricity production area. First Solar not only produces but it installs its modules (including in China!); it now produces and sells electricity.

However, it should be noted that First Solar is unique in its category and the sharp drop in prices for first generation solar energy modules did a lot of damage to a market segment that was not yet ready, technologically speaking. Innovative industrial capacity was abandoned and the return on experience has not yet had time to mature. Leaving out a few commercial exceptions, the sector is today far too far behind to catch up – even with drastic savings measures – on the very low sale price for industrial silicon based solar panes “Made in China”.

Potentially, there is still a chance for thin film technologies; they have the means to back it up. Notwithstanding, if we wish to pull through the crisis, we would have to accept losing money for several years – unless of course we introduce radical changes in our business models – before thin film PV panels become economically profitable. This is the equation we have to solve and it can only be done with support from the investors, notably the States themselves. We are not talking about increased efforts and commitments in R&D, but truly about the needed industrial efforts. The challenge is to build larger and larger module assembly factories, which will cost more and more, to make items that you will be selling at a loss for some time to come. This is not an easy challenge to accept. Certain investors have risked a lot here, and they were not obscure personalities: Warren Buffet, to name but one, has just invested in plans to build and assemble PV solar energy module and arrays with SunPower in California.

As you see it, might Chinese PV panel assemblers be interested in 2nd generation arrays?

They are definitely moving in this direction. All Chinese manufacturers today are focussed in silicon devices, but at the same time they have understood that the future lies with the thin film. First Solar launched a 30 MW project this year with an industrial Chinese partner. China is investing in this project. The Chinese manufacturers are going to learn and will acquire the technological licences owned by the competitors who will soon be in dire straits.

So, what about the European PV industrialists’ stance; what is their position today?

Honestly, I do not feel that either the French or the German solar PV manufacturers ca survive if nothing is done. This is a complex ‘game’ where the winners are not necessarily the ones we think of first. Today the Chinese manufacturers are in fact losing money. The buyers (for the modules) are private persons and European and American industrialists. What this amounts to is the assertion that China is subsidising the production of electricity in both Europe and the USA.

Some doubts have been raised about subsidy strategies via the buy-back tariffs practised in Europe. A debate can be held to discuss whether State subsidies are justified if they help a new industrial sector to emerge in Europe; and we can observe that there is a high level risk faced by the German solar PV sector – which is the only European country to have effectively set up such activities. But I underscore that this is only one facet of the story. The German industrialists also developed a machine-tool segment to assemble the panels, and their principal clients here are the Chinese. One of the main reasons behind the fact that the Germans did not insist on setting up trade barriers to Chinese PV panel imports is that they sell the machine-tools to the Chinese. All told, we must recognise here that public subsidies have helped reinforce a strategic industrial sector.

More on paristech review

By the author

  • Solar photovoltaic energy: today’s crisis and tomorrow’s technologieson April 19th, 2013
  • Pingback: google

  • Pingback: Energy | Pearltrees