Integrating Renewable Electricity on the Grid
By the American Physical Society Panel on Public Affairs
The United States has ample renewable energy resources. Land-based wind, the most readily available for development, totals more than 8000 GW of potential capacity. The capacity of concentrating solar power is nearly 7000 GW in seven southwestern states.
To date, 30 states plus the District of Columbia have established Renewable Portfolio Standards (RPS) to require a minimum share of electrical generation produced by renewable sources. In addition to state policies, federal policymakers have put forward proposals to establish a national RPS which would make the need for technological developments more urgent.
However, developing renewable resources presents a new set of technological challenges not previously faced by the grid: (1) the location of renewable resources far from population sources, and (2) the variability of renewable generation. Although small penetrations of renewable generation on the grid can be smoothly integrated, accommodating more than approximately 30% electricity generation will require new approaches to extending and operating the grid.
The report’s specific recommendation follow:
Energy Storage
The U.S. Department of Energy (DOE) should:
- Develop an overall strategy for energy storage in grid-level applications that provides guidance to regulators to recognize the value that energy storage brings to both transmission and generation services on the grid;
- Conduct a review of the technological potential for a range of battery chemistries, including those it supported during the 1980s and 1990s, with a view toward possible applications to grid energy and storage; and
- Increase its research and development in basic electrochemistry to identify materials and electrochemical mechanisms that have the highest potential use in grid-level energy storage devices.
Long-Distance Transmission
DOE should:
- Extend the Office of Electricity program on High Temperature Superconductivity for 10 years, with a focus on direct current superconducting cables for long-distance transmission of renewable electricity from source to market;
- Accelerate research and development on wide band gap power electronics for controlling power flow on the grid, including alternating to direct current conversion options and development of semiconductor based circuit breakers operating at 200 kilovolts and 50 kilo amperes.
Business Case
The Federal Energy Regulatory Commission and the North American Electric Reliability Corporation should:
- Develop an integrated business case that captures the full value of renewable generation and electricity storage in the context of transmission and distribution; and
- Adopt a uniform integrated business case as their official evaluation and regulatory structure, in concert with the state Public Utility Commissions.
Forecasting
The National Oceanic and Atmospheric Administration, the National Weather Service, the National Center for Atmospheric Research and private vendors should:
- Improve the accuracy of weather and wind forecasts on time scales from hours to days.
Forecast providers, wind plant operators and regulatory agencies should:
- Develop uniform standards for preparing and delivering wind and power generation forecasts.
Wind plant operators and regulatory agencies should:
- Develop operating procedures to respond to power generation forecasts.
- Develop criteria for contingencies, the response to up-and-down-ramps in generation and the response to large weather disturbances.
- Develop response other than maintaining conventional reserve, including electricity storage and transmission to distant load centers.
Full report available here.
The concern about integrating renewables into our grid is a 40-year-old non-issue regularly re-floated by utilities (and friends) to help dissuade those who are arguing for a more distributed, more redundant, more secure, more reliable grid. In the 1970’s people like Bent Sorensen in Sweden had already proven that 20% renewables could be readily integrated, that crew got to 30% a few years ago and my own calculations are closer to 40% now.
It should be our problem.
The point is this, if we waiver in the mission of decentralizing, decombusting, and de-nuking this hopelessly poisonous electricity grid of ours, we will never even get to 12%. I’d like to see what happens when the grid gets to 50% renewables. What do you think? The power will go out more often?
This first iteration of this argument started with a concern for “what happens if I am in the shower when the wind stops blowing?” There are actually a lot of good answers to that question. First, “are you using electricity to heat your hot water then? Wow, awesome use of coal!” Second, “nothing. Renewables have absolutely nothing to do with peak energy – it’s base load, silly.”
Of course utility companies use every argument against doing anything that they don’t have to, by charter. Like serious DSM. My favorite is arguing against distributed inverted power connections because “the linesmen won’t know which way the danger is coming from.”
This strikes me as a case of asking the wrong questions, and so arriving at the wrong answers.
The assumption of this paper is that we must focus on utility-scale renewables. On that assumption, the APS paper is a good piece of work, and their recommendations are appropriate.
However, just as we must approach the general problem of securing future energy supply holistically, thinking not in terms of “silver bullets” but “silver BBs,” so it is with renewables.
I detailed my perspective on this in my June 2009 post, <a href=”http://www.getreallist.com/se
/>ven-paths-to-our-energy-future.html “>Seven Paths To Our Energy Future</a>, but to summarize: Utility-scale projects face a number of large hurdles that small scale, highly distributed generation (like rooftop PV) do not, including securing sufficiently large sites; environmental reviews; transmission lines; utility interconnection support; extremely high up-front capital costs; long lead times; NIMBYism; the need for storage; and so on.
By contrast, rooftop PV and micro-turbines (both wind and hydro) can be deployed much more quickly, without environmental review, using the existing grid infrastructure, with distributed (and usually private) capital, and without the need for storage. Accordingly, it is no mystery why rooftop solar has been getting deployed at several times the rate of utility-scale solar for many years now. Look at the deployments – rooftop PV still outpaces utility-scale PV two-to-one: <img src=”http://c1.cleantechnica.com/fi
/>les/2010/01/solar_increase.jpg”> Source: <a href=”http://cleantechnica.com/2010
/>/01/10/rooftop-solar-installations- growing-faster-than-utility-scale-s olar/”>”Rooftop Solar Installations Growing Faster than Utility-Scale Solar”</a> <img src=”http://www.getreallist.com/wp-
/>content/uploads/2011/01/US_Grid-tie d_PV_Installations_04-10.jpg”> Source: <A href=”http://irecusa.org/wp-content
/>/uploads/2010/10/Sherwood-IREC-Oct2 010.pdf”>Interstate Renewable Energy Council – US Solar Market Data</a>
If we focused on maximizing rooftop PV first, it would deliver several advantages: 1. It would allow a much faster uptake of renewable capacity 2. It would deliver real-time shading data (from cloud cover, etc.) that grid operators could use to calibrate the output of fossil-fueled plants, without relying on weather forecasting 3. It would require little-to-no enhancement of the grid or transmission capacity 4. It would use only the existing footprint of the built environment, and not require BLM lands, environmental review, etc. 5. It would offer a built-in and distributed opportunity to deploy storage, using off-the-shelf batteries and other technologies appropriate to the size of the various installations 6. It would reduce the size of the problems facing utility scale projects, because less of it would be needed – including land, transmission lines, grid capacity, transience, and storage.
Once rooftop capacity has been maximized (which could easily be in the 30% of total generation, or higher, range), it would be relatively straightforward to buttress those millions of small installations with storage capacity. Once that is accomplished, the challenges of deploying utility-scale renewables, and long-distance transmission capacity, should be much easier to tackle.
I have looked over the American Physical Society’s (APS) report on integrating renewable electricity on the grid and have a few comments.
(1) GRID STABILITY: The issue of grid stability and the use of wind power seems incomplete to me. It is taken as a “given” that wind power can penetrate the present grid up to some high percentage, perhaps 20 o 30% of total supply, before grid stability issues might become important. I’m not convinced of this generalization. In some locations it might be true and less so in other locations.
There already have been grid stability issues at much smaller wind power percentages of national electric energy transmission capacity. The APS authors refer to an event that happened in Texas (ERCOT area) on February 25, 2008 where the local grid was protected by quickly dumping large electric loads when wind production dropped off rapidly. The utility had secured contracts to drop these large industrial users in emergencies in exchange for a financial benefit to them. This grid challenge was a comparatively local event and the degree that wind power might adequately increase its national contribution could well be determined by the sum of many local situations. I believe that to minimize wind power grid stability issues one has to make localized stability analyses that account for the locations and sizes of all sources of electricity, the capacities of their transmission lines, all electric loads on the system and how the system reacts to major changes such as rapidly changing wind power production. Additionally, dumping major industrial users of electricity because of the variability of wind power doesn’t come across to me as a good way to have an overall efficient energy system.
The 20% figure about wind power penetration, such as that in the title of the American Wind Energy’s Association’s report “20% Wind by 2030”, may be influenced by the wind contribution in Denmark, the world leader in percentage wind contribution, which has a 20% wind power component. However, Denmark benefits from its access to neighboring countries including France’s huge nuclear base load generation and Sweden’s and Norway’s quick-response storage hydro systems. What would be useful would be an APS follow-up study to see if claims of existing high wind power penetration all relate to situations where there is significant energy storage like large hydropower facilities.
(2) WIND POWER PREDICTIONS: Greater accuracy in predicting wind energy is desirable from a planning point of view and specialists are working on this. These predictive improvements may mostly benefit utility planning in the one hour to one day range. However, I am not aware of any wind energy prediction expert who would have capabilities in the second-to-minute time range. Yet significant changes in the wind can occur in this short time range.
(3)TRANSMISSION LINES: The APS correctly points out that stringing power lines from wind farms over great distances to their electric loads is a challenge. It will not be easy to get construction permits to run power lines through multiple public and private lands and with many governmental jurisdictions, each of which may have different rules and permitting processes. Although the wind farms themselves do not occupy too much land, the transmission lines would cut a large and long and likely restricted swath across the landscape. Limiting access to existing hiking, hunting, logging, and ski trails almost certainly will be opposed. Complicating this is the fact that there may not be any financial benefit for those people who have such lines run through their property. Besides these financial/political challenges there are technical/economic ones as well. Onshore wind turbines typically have capacity factors in the 30-45% range. This means that the dedicated transmission lines that are connected to them would also only be used 30-45% of the time. Such low capacity factors for the transmission lines is an expensive way to operate.
(4) IMPROVING RELIABILITY: Figure 5 of the APS report shows the result of an analysis which shows that higher overall reliability can be achieved when there is a larger area (more wind farms) to draw from. What is absent from this analysis is that the cost of this wind power would also increase. A larger number of wind turbines spread out over a larger area may mean that there is a larger economic penalty whenever more wind turbines are idle or when so much wind power is produced that it can’t be handled by the transmission system or may exceed that moment’s demand for electricity. Figure 5 seems to be incomplete without an accompanying analysis of cost per kilowatt-hour electric, based on actual data, as a function of distance ( the size of the area that one draws wind power from). Perhaps a simple analogy is applicable. If I have a car that is 35% reliable but I need my transportation to be 99% available, then this one car won’t be sufficient. If I bought a second, third, or fourth car, each 35% reliable, I would begin to approach my 99% transportation availability requirement. However, this would be very expensive. On some days I would have more cars than I need to meet that day’s transportation needs, yet I would need many cars to achieve high availability. Would it not be better to have a single car that is itself 99% reliable?
(5) PROTECTING THE ELECTRIC GRID: APS recommends the development of semiconductor-based circuit breakers that operate at 200kv and 50kA that would operate with a microsecond response time. My concern here is the potential impact of malfunctions of these circuit breakers. During a rapid challenge to the grid, such as quickly decreasing wind power, how many such circuit breakers have to open with a microsecond response time? What if one circuit breaker takes more time to respond or fails to respond? Does this mean that the reliability of each circuit breaker have to be something like 99.9999% ? Conversely, does the inadvertent opening of a single rapid circuit breaker cause some kind of a power failure?
(6) Figure 3 of the APS report. I thought that the output of wind turbines varies as the cube of the wind speed. This figure is much closer to the square of the wind speed. This figure doesn’t quite vary as the square of the wind speed since they state that a 20% error in wind speed results in a 41% error in the output. I would calculate 44% [ (1.20)(1.20) =1.44]. If I am right about wind power varying as the cube of the wind speed then a 20% error in wind speed would be a 73% error in the power output. Either way, squares or cubes, the lesson here is that small variations in wind speed have large variations in output power. If the wind speed dropped from 10 mph to 8 mph the power output would drop by 44% to 73%. These are very large swings in power output for small changes in wind speed that are rather common. So even under rather normal conditions the wind power output is rather “jittery”.
(7) ENERGY STORAGE: One way to do away with many of the wind power challenges is to have energy storage, i.e., you use energy storage as a buffer between the wind farm and the grid. This eliminates the grid stability issue since the grid and the wind farm are not connected directly. With energy storage the output from the wind farm can be more like that from a base load plant. This means the transmission grid would be sized to handle a steady amount of wind power, thereby operating far more efficiently. With energy storage there is no need for microsecond response time for circuit breakers and there is little need to worry about wind power supply and customer demand mismatches. Personally, I think that in order for wind power to ever be a major energy contributor, it must be hooked up to some form of energy storage.
I would eliminate pumped hydro energy storage except where dams already exist. That mass of water that would have to be pumped back up every day would be huge at a 20-30% contribution from wind power. Please note that the APS calculated pumped storage, as large as it was, assumed a storage elevation difference of 500 meters (1680 feet). Where does one find elevation differences like that? Where does one find unused caverns for compressed air storage near wind farms or does one have to run a transmission line over to a cavern to run the compressors there? How many kilowatt-hours can be stored in all the available caverns in the U.S., at a reasonable cost? Because energy storage using chemical bonds rather than “mechanical means”, like pumped storage, often offers much higher energy densities the National Renewable Energy Laboratory is sponsoring research in storing wind power electricity as hydrogen, created by electrolysis of water. When electricity is needed some of this hydrogen is reconverted back into electricity. Such hydrogen storage facilities would be located right at or near the wind farm, thereby eliminating the need to run a separate transmission line to the storage location. Others are investigating using liquid fuels, such as methanol, for energy storage. The methanol would be made from the electrolysis of seawater which contains considerably more CO2 than air. The APS energy storage write-up needs to be expanded to consider more ways to accomplish energy storage.
(8) The APS analysis of the relationship between energy storage and concentrated thermal solar systems omitted an important consideration. It appears that the APS only considered the diurnal variations in available solar energy. However, there are also seasonal variations. Insolation data collected at Desert Rock, Nevada show that the amount of solar energy that these desert solar systems would collect in the winter is about half of that during peak summer months. This affects the economics of the transmission network because the limited output in winter months, a problem similar to the wind power’s low capacity factor/ transmission grid economics issue. Further, the APS also apparently did not consider the fact that there is essentially no water in the very deserts that these solar thermal machines are to operate. Therefore, in order to operate, they would have to use air as the ultimate heat sink. But air temperatures in the desert during summer months during the day are high, making an air ultimate heat sink inefficient. Connecting such desert machines to the grid may not become a large problem if poor economics limits the number of such renewable power plants.
(9) The APS did not attempt to factor in cost estimates of the systems it looked at. It is entirely possible that a rather different result would have been produced by the APS if it screened out various designs when calculated costs that are too high.
Best, Herschel
I don’t want to dump on wind energy. I hope it succeeds wildly, because we are going to need every bit of energy we can find. However, I do want to point out that biomass energy has the considerable advantage of being both the energy collector and the storage system all at once. Plant leaves harvest the solar energy and that energy is then stored in the chemical bonds of plant biomass.
Herschel correctly points out the problems with the intermittent nature of wind energy. Like him, I think wind energy will need to be coupled with some sort of energy storage system if it is ever going to make a very large contribution.
Best, Bruce
In the real world where people build lots of wind turbines, people buy gas turbines and operate them at low power to have the capability to provide power quickly when the wind slows—an expensive option with lots of greenhouse gas emissions.
As a side note, I will be giving a talk at an ARPA-E workshop at the end of the month on thermal energy storage for peak power using nuclear energy. Lots of options—non developed. For those with an interest in some of the storage options, see attached starting on page 13. The earlier part of the talk goes into storage requirements.
There is an interesting and portable storage medium – liquid fuel. We could use the “free” electrical energy off-peak (nuclear and coal) to capture atmospheric CO2 to generate syngas and any liquid fuel from it. I don’t think that the technology is that far away (there are already patents) and the electric utilities may be very interested in a test. There could be “free” dollars for them, CO2 credits and reduced dependency on oil.
Yossie