The last three posts on this blog have all dealt with the feasibility of reaching very high levels (80% or more) renewable energy in our electricity grids. Specifically, these posts were responses to articles dismissing the possibility doing this quickly, whose arguments I did not feel were well supported by facts and existing research. I will also note that the authors tended to miss or ignore a plethora of real-world developments and even whole technologies.
However, these articles have brought up what to me is a very interesting question: what a world where the vast majority of electricity comes from renewables will look like.
This is something I tend to not write much about; I prefer to cover what is known and actually happening than to proscribe future scenarios. Roadmaps to a better future are inherently useful; however they are inevitably proven wrong in the details, because the real world never plays out the way we think it will.
100% renewable energy is technically feasible
First off, I must reiterate that it is technically possible to run global power systems on 100% wind and/or solar. This has been proven by the island nation of Tokelau in the South Pacific, which began getting 100% of its electricity from the sun in October 2012, joining Iceland in the club of 100% renewable nations.
Incidentally, a number of other islands in the Atlantic, South Pacific, Indian Ocean and Caribbean are all moving to 100% renewable electricity – mostly through wind and solar, though some have hydroelectric plants.
Like Tokelau, most of these islands will rely on massive battery banks to provide electricity when the sun is not out and when wind power is low. While this makes sense for small islands that have limited resources and pay high prices for imported fossil fuels, the rest of the world has less expensive options.
In my last post, I detailed a lot of those options to store and/or utilize energy without batteries in order to reach very high penetrations of wind and solar – including CSP with thermal energy storage, pumped hydro, electricity-to-heat, and electric vehicles with time-of-day charging.
Myriad cat-skinning techniques
There are many ways in which these resources can be deployed, and many studies which suggest how various regions could move to 100% renewable electricity, or even all energy. One of the most prominent is the study by Stanford Professor Mark Jacobson and his team which deals with the United States.
Another is a 2011 study by Ecofys and World Wildlife Fund, which looks at the global picture.
One of my personal favorites is the Zero Carbon Britain project by the Centre for Alternative Technology, a roadmap for moving Great Britain to zero net carbon emissions, in which renewable energy features prominently.
I’ve discussed my points of concern about the Jacobson report in a previous post; namely that I think it makes the technical challenge more difficult than it needs to be (and more expensive) by excluding both biomass and battery storage. Also, I find its cost projections, which are based a rather full accounting of externalized health costs and climate impacts, to be appropriate; however I don’t think we can disregard costs under our current economic system which does not count these externalities. Overall, I consider this a very important report, if a little idealistic.
I like the Zero Carbon Britain project for a number of reasons, including that it looks not only at energy but also agriculture, land use and the built environment. I will also argue that its section on electricity gives what I think is far too much of a role to offshore wind and too little to solar photovoltaics (PV). It appears that even for 2013 when the report was published the authors were a little behind on PV cost trends.
A few weeks ago a friend referred me to another 100% renewable electricity report; this time it is designed for Australia. In this study, three professors at the University of New South Wales (UNSW) look at potential mixes of renewable energy for the nation, and come up with the following ratios as the least expensive way to ensure electricity system reliability with 100% renewables.
Wind 46% (I assume land-based)
Concentrated solar power (CSP, also known as solar thermal electric) with thermal storage (TES) 22%
Solar PV 20%
Biofuelled gas turbines 6%
Existing hydro 6%
While it is not entirely comparable to the Jacobson report (note that Jacobson looks at all energy, while UNSW looks only at electricity), I am partial to this mix because I think that it would come closer to achieving practical ratios with a minimum of energy storage. In fact, the authors include no battery storage. Under this system, Solar PV can meet the large majority of daytime demand, and CSP w/TES can be dispatched to fill “valleys” in the evening and overnight when wind is not as productive.
There will still be technical tweaks in applying this model to the real world. For instance, there is battery storage capacity being deployed on many grids to meet multiple needs, and these batteries can be used to help with some of the more difficult periods, particularly right after sundown when a fast ramp-rate will be needed.
Of course, a lot of excess daytime PV generation could be absorbed by electric vehicles (EVs), particularly if time-of-day pricing is implemented – but that gets into a more full energy model than just electricity. And it would also require a dramatic increase in the rate of adoption of EVs.
Geography and grids
I am particularly fond of the UNSW study because I suspect that this renewable energy mix would not only be ideal for Australia, but that a similar mix would be appropriate for the Western Interconnection – one of the three major grids in the United States. Specifically, in the West there are even higher levels of hydroelectricity, as well as suitability for CSP – which can be built from Wyoming to Eastern Oregon south to the Mexican border.
If you take out the hydro, and substitute that 6% with other resources including biomass/biogas, it would work well for the Texas grid, as West Texas and the Panhandle are both great locations for CSP.
In the Eastern Interconnection, Australia’s mix will not work at all. There is no place north of Florida and East of the Mississippi suitable for CSP, so that will pretty much have to be eliminated. The Eastern grid faces larger problems in terms of resource availability for wind and solar. While newer classes of wind turbines are already helping wind to be deployed in a more areas, a major limitation is that the U.S. South has less potential for land-based wind.
The key resource for the Eastern United States may end up being offshore wind. One major advantage of offshore wind is that the output is more steady, meaning fewer “valleys” to fill with flexible power. But even with this advantage, the technical task of reaching 100% renewable energy will be more challenging and the cost will be higher in the Eastern United States compared to the nation’s two other grids. Solutions may come in the form of needing to make greater use of electricity-to-heat, EV charging and/or stationary battery storage, given the limited availability of hydro.
A note on policies
Existing policies will not allow us to reach these 100% renewable visions. For example, right now PV is much cheaper than CSP with storage, and as a result is much more widely deployed. If we want to meet very high portions of demand with renewable energy, and particularly to do so in the most cost-effective way, we need to make sure that policies reflect the values that different assets provide. This may include redesigning markets, and/or incentives for particular technologies.
There are a few existing policies which point to the kind of proactive planning which will be necessary. California’s AB2514 is a good start towards a grid based on renewable energy; feed-in tariffs on the German model, which set favorable prices for certain technologies, are another approach. A noteworthy development on this model was LIPA’s (now PSEG’s) feed-in tariff which included a locational benefit to generation in certain areas to help defer new transmission.
On false models
Geography matters for renewable energy deployment. There is no one-size-fits-all model, as different resources are available in different locations. If you want to know how to best deploy various renewables while maintaining reliability, you need to do system modelling, as Jacobson and the professors at UNSW have done. Overly simplistic models and clumsy rules, such as the ones pushed by Breakthrough Institute, are not a valid substitute.
And while the UNSW model looks promising from the perspective of today’s technologies and costs, we cannot know with certainty what the ideal mix will be in 2020, let alone 2030. As battery costs are falling rapidly, stationary batteries may be a more practical solution to many needs by the time we actually have wind and solar production in excess of demand on any grid, particularly as they can supply multiple services to the grid and to customers who install them.
We don’t know exactly what the future will look like. Many experienced writers, analysts and even grid operators have made assumptions which have been proven wrong, both about limits to cost declines and the ability of grids to absorb higher levels of wind and PV.
Don’t handicap the future. And remember that there is more than one way to get to where we need to be.