Saturday, December 29, 2018

Solar Power is Not as Easy as it Looks

Adding the first increment to the grid is far easier than adding the last, if we want to decarbonize electricity. Review of "Taming the Sun", by Varun Sivaram

Global warming is no longer a future problem, but a now problem, and getting rapidly worse. We need a total societal focus on extricating ourselves from fossil fuels. Putting aside the brain-dead / know-nothing ideology of the current administration, the world is broadly, if grudgingly, onboard with this program. What is lacking are the political will and technical means to get there. California now gets 29% of its electricity (including imports from other states) from renewables, of which 10% is photovoltaic (PV) solar power. The grid operator shows a pleasing daily graph of solar power taking over one-third of electricity demand around mid-day.

A typical day on California's power grid. at mid-day, and fair portion of the state's power comes from solar power (teal). But come sundown, many other plants need to ramp up to provide for peak demand.
 
Varun Sivaram's book is an earnest, somewhat repetitious though well-written and detailed look at why this picture is misleading, and what it will really take to go the rest of the way to decarbonization. Solar power has very bad characteristics for electrical grid power- the grid operator has no control over when it comes in, (it is not dispatchable), and it all tends to come in at the same time of day. While this time (mid-day) is typically one of heavy usage, it is not the peak of usage, which comes during the transition to cooking and evening activities, from 5 to 7 PM. This means that not only does the rest of the grid have to work around solar's intermittency, but the rest of the grid has to constitute a full fleet of power plants for peak needs- solar will not reduce the need for either baseline or peak power capacity.

This is extremely disappointing, and means that adding the first 10% of solar to the grid is relatively easy, but adding more becomes increasingly difficult, and offloads rising expenses to other parts of the system. We do not have the technical means to economically address these issues yet. Solutions come in two basic forms- energy storage, or alternative modes of non-CO2 emitting generation.

Storage technologies by current capacity and capability. Pumping water uphill into reservoirs is the only existing method of storing power in grid-scale amounts over long periods.

Storage is easy to understand. If we could only bottle all that solar electricity somehow, all would be well. Even if we can't save summer power for winter, but save it only for a few days, we could build enough solar generation capacity (at the current cheap and falling prices) to cover our needs at the lowest production time of year, and throw away the excess the rest of the year. This assumes that, over a suitably large geographic area, there will not be so much extended cloud cover that this could not be reasonably planned. But such storage technology simply does not exist yet. The diagram above mentions some of the major candidates. The best known are chemical batteries, like lithium ion. This is how off-grid and home backup systems manage the intermittency of solar power. But these are expensive, which is why it is cheaper to buy power from the local utility than to go off-grid, and also cheaper to build a grid-tied solar system than go off-grid. The most mature grid-scale storage technology is hydropower- pumping water back uphill into a reservoir. This is obviously not available in most places where storage is needed.

Where various storage technologies are in development.

Other methods like flywheels, raising and lowering rocks, etc. are all on the drawing board, but not yet in practical deployment at grid scale, or even demonstrated to be economic at that scale. Making fuels like hydrogen or hydrocarbons from solar energy is another prospect for storage, but again are not currently economical. Hydrogen has been touted as the all-around fuel of the future for many uses, but is so difficult to handle that, again, it is far from currently practical. Getting there will take money and effort. 2050 is when we need the power sector substantially decarbonized, world-wide (if not sooner!). It sounds far off, but it is only about 30 years- a very short time in power technology terms. The scale needed is also gargantuan, so we need these solutions to get off the drawing board as soon as possible- there is no time to waste.

The alternative methods of no-carbon generation are currently wind and nuclear, with CO2 storage (sequestration) from fossil fuel plants as a further option. Carbon sequestration is not a new technology, and is something that would be directly motivated by a carbon tax, though it is also phenomenally wasteful (as are many of our more adventurous methods of producing fossil fuels, like tar sands)- a fair fraction of the energy produced goes right back into compressing and pumping the CO2 back underground. Wind is also getting to be a mature technology, and shares with solar the problem of intermittency, so is not a solution for dispatchable or baseline power. Sivaram does note at length, however, that a helpful technology for both solar and wind is long-distance DC transmission, which would allow rich sources, like the plains states, or the Sahara, to be connected to heavy users.

The dream of the next generation of nuclear power, which has not been demonstrated at grid scale.

That leaves nuclear power as an important element in future power systems. Generation IV nuclear power promises cleaner, proliferation-proof, more efficient, and more sustainable nuclear power. China has several programs in development, as does the US. Again, as with all the other necessary technologies for a fully sustainable grid, these are not mature technologies, and need a great deal of research and development to come to fruition. I will not even delve into fusion power, which is not demonstrated terrestrially in principle, let alone development.

The point of all this, as made at some length by Sivaram, is that the key to getting to a decarbonized future (for electricity, the easiest energy sector to deal with) lies not simply in scaling up the PV present into a glorious future. Rather, it lies in further intensive research and development of a variety of complementary technologies. The next question naturally is: will the private sector get us there, even if there were a carbon tax? The answer is- unlikely. The Silicon Valley model of venture capital is not well-suited to the energy sector, where innovation comes in small increments, the regulatory weather is heavy, and the scale in time and capital to money-making deployment is huge. There needs to be continued, and vastly expanded, government direction of the research, along with much other public policy, to address this crisis.


  • Fed still fighting the last war, or the one before that, or a class war. But good policy it is not.
  • IRS heading towards total impunity.
  • Justice is in peril.
  • What a year...