Last week, the Department of Energy held its annual ARPA-E Summit near Washington DC. One of the keynote speakers, the CEO of DuPont Ellen Kullman, stated that “solar will be globally competitive in 2015”. First Solar is already selling solar power cheaper than coal in sunny areas like New Mexico. This is not a random event after years of improvements in cell efficiency and intense price competition. Does this mean solar is ready for prime time?
Not so fast. Solar power is causing grid engineers and policy makers to scratch their head: it is intermittent and has indirect costs to provide useful energy. Some victims of Storm Sandy realized that first-hand when their solar roofs shut down when the grid blacked-out. What is the point of deploying solar when you cannot use it when it matters the most? Back-up gas generators are still cheaper than deploying solar panels and batteries. The solar sector is at a crossroad.
Last Fall, five solar farm projects waited to get the green light from the California Public Utility Commissions and were postponed until they included some element of energy storage, and addressed other cost concerns. Only two projects out of five were ultimately approved. The awardee, BrightSource, uses molten salt to store heat from solar to generate electricity from steam at a later time. There is a growing realization that California will not meet its 33% Renewable Portfolio Standard target by 2020 without storage. As a result, California is looking at various energy storage technologies under the bill AB 2514, which is scheduled to be implemented in 2015.
The German government has recently announced an incentive to deploy batteries along with photo-voltaics (PV). The subsidy corresponds to $1,080 per kWh of energy storage for PV systems with under 30 kW of peak capacity. This way, utilities see little variation on their distribution network, and the grid can support a higher penetration of solar generation. However, the energy stored at a home, and partially lost due to inefficiencies, does not fully benefit from the feed-in tariffs in Germany. As a result, home-owners would actually lose money in the long run!
The German case is yet another example where the subsidized model breaks down: solar is currently disconnected from the actual energy needs of consumers. What works in a centralized grid does not necessarily make sense for end-users, and vice versa. What is the alternative for energy operators? Natural gas. It pollutes less than coal, and prices are at an historic low in the United States. Utilities already know how to operate gas generators to add capacity on the grid.
The United States are in a particularly interesting situation: vast amounts of untapped reserves are now available thanks to horizontal drilling techniques. This is good news for the economy as cheaper energy will support economic growth in the short-term. Energy veteran and investor T. Boone Pickens has been a vocal advocate for natural gas to restore energy independence of the United States. That is all good in the short-term. But it hides a bigger trend in the long-term. Solar will become clearly the cheaper alternative for consumers, even with natural gas producing electricity at 5 cents per kWh.
As a result, end customers might decide to opt out to be energy independent from an aging grid infrastructure by leveraging renewable energy sources. Community Choice Aggregation programs represent an early instantiation of that trend. Historically, utility rates have grown by an annual average of 2.5% in the last 40 years. In comparison, solar costs have come down by two orders of magnitude during the same period, or an average of 15% per year. That is why solar is becoming increasingly popular among facility owners, as long as it is presented as a service replacement. Power purchase agreements (PPA) represented more than 75% of solar installation last year in the United States.
Now, let's look at a less rosy picture for solar. First, one can argue that utility rates will come down due to natural gas; let's use a decrease of 2.5% instead of the historical increase. Second, let's add the cost of energy storage to solar generation so we can compare "apples to apples": on-demand electricity. The result plotted below is interesting: installed cost of solar (yellow), utility rate (grey), and solar with energy storage (orange) all in dollars per MWh.
The plot assumes reduction in energy storage costs by 15% every 18 months. It is for a sunny area like California. The cost curves for solar in less sunny areas would shift upward, but the overall picture remains the same. Fundamentally, distributed solar costs will come down faster than the utility rates. On-demand solar electricity will hit grid parity only in 2017, and it does not make sense without subsidies today unless a customer pays for more than 25 cents per kWh.
This exercise lays out a very interesting dynamic. Solar with net metering is a no-brainer today but it is not sustainable. It will put increasingly more pressure on the utilities to deal with it. As a result utilities will push back and point to the total cost of operating solar. This is why utilities in California want to maintain a cap of net metering to 5% of peak power. Their point of view is that they are providing the equivalent of "free storage", and should not pay dollar-for-dollar for erratic electricity from solar roofs.
This on-going battle will in turn force solar to find a sustainable business model without subsidies. Several projects are already happening. One project in France and Germany is studying time-shifting as PV's post incentive future. What will be the "killer app"? Time will tell. In any case, the subsididized penetration of solar generation will force incentives on energy storage to end subsidies for both technologies.
The CEO of DuPont is right. Solar will become competitive in a few years, but not without going through a painful transformation from a centralized and subsidized model ("solar 1.0") to a market driven model where customers get the service that they need ("solar 2.0"). Integration of two technologies to deliver a useful service is not unheard off.
The computing industry went through that transformation thirty years ago. Access to computing capacity required a terminal and a line to a mainframe computer. As the cost of data storage and computing came down, the personal computer came to life. However, the real breakthrough for personal computers was desktop publishing when postcript was developed by Adobe's founder, John Warnock. It tied together personal computing and laser printing (picture right).
It is not clear whether industrialized countries will benefit the most from new advances in solar and energy storage. If the United States wants to achieve energy independence from the Middle East with increased oil and gas production, solar could provide a solution for the 2 billion people who do not have reliable electricity. Nobody can put a fence around solar, or tax wind for that matter. Renewable energy integrated with energy storage could open the door to a more organic model.
"Solar 2.0" could become social. For instance, it can provide an answer to a fundamental problem in emerging countries like India: electricity theft (picture left). It is one thing to steal electricity from a central grid in a country where corruption is common place; it is another to steal energy from your neighbor!
This does not mean solar will not continue to play a role in centralized grids in industrialized countries. We can use the analogy with mainframe computers again. The advent of personal computers did not mean their end as a centralized ressource. It forced IT companies to evolve, and provide computing and data storage as a cloud-based service with flexible data centers. To that extent, the IT industry copied the energy industry, and not the other way around.
Until the two paths join again -- distributed solar and solar operated by utilities -- solar is at a critical juncture. Solar 2.0 will branch out from the grid to free up the consumers who are disenfranchised or pay a premium on electricity, while utilities will use savings from natural gas to subsidize the integration of renewable energy and meet government mandates.
For sure, It is not the last time we will hear about natural gas and solar. Just today, Shell released a report that solar and natural gas will become the most important sources in the 21st century.