Without significant feed-in tariffs or other types of government support, photovoltaic energy is not yet competitive with fossil fuel or nuclear power generation. But the technology is on a learning curve and the so called “grid parity” – competitiveness with conventional electricity generation is approaching.

But are we on the right learning curve? How far away from grid parity are we today and how much does it depend on geography or on other specific conditions? And which actions or conditions could speed up the learning path?

Is grid parity a useful parameter, or rather a non-issue? What does “grid parity” really mean? Is it parity with wholesale price or with retail price, with or without taxes?

These and other questions were addressed on a Discussion Webinar on 7 March 2008. The following are a few of the major points arising from that discussion.

How far from grid parity are we today?

How far the current PV cost is still separated from grid parity depends largely on the geographical situation, on the price of grid electricity, and on whether it concerns a cheaper or a more expensive PV system. For a standard roof-top system, grid parity can already be the case in parts of sunny California and South-Western Australia (Perth). In Brussels or London, electricity from such a system is still costing well above the grid electricity price.

Photovoltaic energy is currently on a learning curve. For energy systems, the learning curve generally shows a cost reduction of 15 to 20% for each doubling of the cumulative sales figure. This means that, to halve the cost of PV and reach grid parity, the PV market will need to grow with a factor 8. However, those figures remain predictions and not hard science. Who knows whether the learning curve for PV is following the classic shape, or whether step decreases in costs due to disruptive technology developments are imminent?

The cost for taking a certain technology down on this learning curve until it is competitive is called the “learning investment”. It is carried by government subsidies and the user’s “willingness to pay”.

What does grid parity really mean?

The press is often using the expression “grid parity”, but what does this exactly mean?

• Do you compare with the cost of base load power or with peak power? This choice mainly depends on the correlation between the PV production peak and the demand peak.
• Do you compare with wholesale electricity prices or with retail prices, including a term for transmission? The more the PV production is used locally, the more it should be compared with the retail price.
• Should count with the same tax rates as for electricity from fossil fuel plants? Since the externalities for PV plants are lower than those for fossil fuel plants, it would be rational to count with a lower tax rate for PV systems.

Does PV electricity make less or more use of the grid. If PV electricity is used locally, and reduces peak demand from the grid, it certainly produces a system benefit. A good example could be the coupling of PV electricity to airconditioning in sunny climates, where peak generation of solar electricity naturally coincides with peak demand for cooling. But if the grid is used for electricity storage, and a user becomes a net generator in summer, while being a net consumer in winter, such set-up results in higher network charges.

While traded as a commodity, electricity is actually a highly complex service, with many parameters defining its voltage quality, continuity of supply and quality of service.

But if “grid-parity” is such a volatile concept, does it still make sense using it? In the electricity systems of the future, the concept will almost certainly lose value.

The evolution of the silicon industry

Today, PV growth still largely depends on the silicon industry. It is evident that price reductions in PV modules over the last 10 years happened at least partly thanks to silicon price reductions. In recent years the semiconductor industry started to pay much more focus to solar systems.

Price reductions in silicon and PV systems are sometimes falsely compared with Moore’s law (“the numbers of transistors on an integrated circuit is doubling every 18 months”). Moore’s law applies to scaling in integrated circuits, and even there it is not expected to last forever. It predicts an exponential growth based on increasing miniaturisation of semiconductor components. Price predictions for silicon and PV systems have to take the laws of physics and market conditions into account - fortunately the amount of sunshine on earth does not double every 18 months.

Step improvements are needed

Prices for PV systems are going down, but still cost reductions through learning investments alone might not suffice to reach grid parity in the near future. Step improvements by technological or market breakthroughs will be required as well.

Technological breakthroughs are very likely to happen in the next decades, especially on the domain of thin film PV systems. The Ovionics Triple Junction systems for instance are rapidly improving in efficiency.

Another interesting breakthrough could be the combination of Concentrated Photovoltaic (CPV) systems with thermal Concentrated Solar Power Systems (CSP). Since both systems are harvesting a different part of the electromagnetic spectrum of the sun, there is indeed no technological barrier to combine them and “harvest the same sunrays twice” in an new kind of “Combined Heat and Power” system.

Even if it is not to be combined with CSP, CPV systems could speed up the growth of the PV market, since it reduces the need for PV cells with a factor 400 to 1000 for the same power output.

All this begs the question whether governments should massively invest in the wrong systems that drip-feed electricity into the grid. Shouldn't they rather invest in developing technologies? Of course, it's easier to set-up an off-budget feedin tariff through a levvy on the electricity price than to find the resources on the national budget to develop technology.

A prediction for the future

General predictions today are that standard rooftop PV systems which output correlates with peak demand are to reach grid parity by 2015 in large parts of the U.S., Japan and Southern Europe. In the more temperate part of Europe, this grid parity is expected to happen around 2020. Grid parity for base-load power in this area is only to be expected around 2030.

Note that, even when grid parity for PV systems has been reached and surpassed, countries with large coal mining production, such as Poland, might still prefer to pay the premium for coal fired power plants for socio-economic reasons

Article reprinted from http://www.leonardo-energy.org/drupal/node/2828