California to Close Diablo Canyon Nuclear Plant

Subtitle: Solar Power to Replace Nuclear

It won't be right away, instead the closures of the two remaining reactors in California will be in 8 and 9 years from now, respectively, in 2024 and 2025.   That news rang across the nuclear camp yesterday, as PG&E, the plant's owner and operator agreed to shut the plant down at those dates and not seek a 20 year extension for the operating license.  Many articles on this are available, one from the Wall Street Journal (see link) does a credible job.   Title: "PG&E to Close California’s Last Nuclear Plant by 2025 - It will be cheaper to shut down Diablo Canyon facility and find replacement power, utility says."

Diablo Canyon Nuclear Plant, image from Google Maps 2016
Arrow indicates twin reactors.   Pacific Ocean to the bottom right.

There are some interesting, biased, pro-nuclear articles too, mostly from those who seemingly cannot believe their beloved nuclear plants are being shut down, instead of being built in greater numbers.   Those articles grind on and on with their favorite themes: jobs lost if nuclear plants close, grid instabilities if nuclear plants are not there to anchor the fragile grid, save-the-planet with carbon dioxide-free power from nuclear plants, and of course the old stand-by, coal and natural gas prices might increase again someday.   What many of the pro-nuclear articles omit is the great capital cost that PG&E would incur to keep the plant running past 2025, and how much money the plant is losing by operating in the present economic conditions. 

Much of the hoopla and angst stems from the pledge by PG&E, one of California's largest utilities, to replace the 2,200 MW of electricity presently provided by Diablo Canyon with a mix of wind, solar, storage, and efficiency improvements - all at no additional cost to the consumers electricity bills.  

Taking the above list of four tentative reasons to keep the nuclear plant online, in order, with jobs first.   The plant employs approximately 1500 people, per PG&E.   Jobs and their loss are also trotted out by other nuclear plant owners across the nation, as those plants are shut down.  The company is to work with various unions to keep some employed to perform the decommissioning (more on that expensive fiasco later), transfer some to other jobs within the company, and perhaps provide severance packages to others. 

Second, the California grid is not at all fragile.  The simple fact is that Diablo Canyon is a drop in the bucket in the California electricity market; only 2,200 MW out of approximately 35,000 MW average production, or approximately 8 percent.   On a high-demand day, when demand reaches 45,000 MW (as it did yesterday), the nuclear contribution is a much smaller portion at only 5 percent approximately.  It is also a fact that another, equal-sized nuclear power plant dropped offline forever in 2012 when the SONGS (San Onofre Nuclear Generating Station) had multiple tube ruptures and spewed radioactive steam into the sunny skies of Southern California.   SONGS' 2,200 MW removal from the grid did not create any blackouts, rolling or otherwise.   The ISO, California's Independent System Operator, simply called for more production from the gas-fired power plants.    Also, in the four years since that time, California has installed at least 7,000 MW of solar power plants.   The grid's frequency stability is assured by the gas-fired power plants, and large hydroelectric plants.  

Third, saving the planet by producing power that is free of carbon dioxide emissions is required only by the false-alarmists who believe that CO2 will overheat the Earth's atmosphere.   CO2 in the atmosphere certainly has not produced any appreciable, nor alarming warming thus far.  

Fourth, the tired ploy of gas shortages that nuclear advocates used in the 60s, 70s, and 80s no longer works.  Natural gas is no longer in short supply with a high price, nor is it likely to ever be in that situation again.  The simple fact is that gas exploration companies know now how to use precision directional drilling (PDD) and hydraulic fracturing to good advantage, producing natural gas in surplus amounts.  The wholesale price is now under $2 per million Btu, due to PDD and hydraulic fracturing.   This is a world-wide practice, not limited to the US.   

So, then, what of the naysayers' claims that substituting wind, solar, and increased efficiency will replace the 2200 MW from Diablo Canyon?    Again, as above, the fact is that California added more than triple in solar MW compared to what was shut down at SONGS.  (7000 vs 2200 at SONGS).   The grid remains stable, blackouts did not occur.    

The wind resources in California are nearly fully exploited, as the state has only three economic locations for wind turbines at Altamont Pass, Tehachapi Pass, and Banning Pass near Palm Springs.   Any future capacity growth would be from retired wind turbine replacements with modern, more efficient turbines.   In addition, state-wide data shows that California's wind power plants have a lower-than-average utilization rate, or annual capacity factor compared to the states in the Great Plains region.  In good years, the wind provides approximately 26 percent capacity factor, and in poor years about 22 percent.   In contrast, the Great Plains states have capacity factors of 35 to 42 percent on an annual basis.  Using rough numbers, 40 for the Great Plains and 25 for California, a wind turbine would produce 60 percent more power in the Great Plains (40/25 = 1.6).   However, the costs to install and operate would be effectively the same.  It makes great sense to build wind turbines in the Great Plains but not in California.  

Increased solar power has some intriguing aspects that will be discussed next.   One major point (allegedly) in the Diablo Canyon shutdown agreement is that PG&E will procure 55 percent of its electricity from renewable sources.   This is 5 percent more than the 50 percent that state law mandates by the year 2030.    As wind power is not likely to increase much, the logical candidate is solar power.  The state has ample sunshine that presently produces approximately 8000 MW at noon (recent data from CAISO).   With a total annual power demand of 300,000 GWh, half by renewables then is 150,000 GWh.  Wind and other non-solar renewables in 2014 produced 34,000 GWh, leaving 116,000 GWh for solar to produce.   With the annual average capacity factor for California utility-scale solar of 26 percent (per EIA and California Energy Commission), the state would then require 51,000 MW of solar installed. 

And there lies the problem.  The solar arrays produce too much power for the grid to absorb on any given sunny day.  51,000 MW of solar output greatly exceeds the typical summer day's peak demand of 35,000 MW.   What, then, to do with all that mandated solar power?    One solution, already proposed and under consideration, is to store at least a portion of the solar energy output as hot oil, or molten salt, to be re-produced as electricity later at night.   Yet another is to increase the pumped storage hydroelectric capacity in the state, and store the energy by pumping water into elevated lakes.   A third solution is to store some of the excess solar energy in grid-scale batteries.   A fourth solution is to store some of the excess solar energy in gravity-based heavy rail storage systems, as the ARES system in Nevada will do when construction is complete.  

Update 1: 6/23/2016 -  More uses for excess electricity include a fifth solution - split water via electrolysis, store the hydrogen for later and produce electricity when needed via fuel cells; and sixth, have a multitude of electric vehicles on smart chargers to charge the batteries with excess power.   --- end update 1

It is an interesting time to live in California.  The last nuclear plant will close in less than a decade.  Solar power plants will be built in great numbers.  The electrical grid will not only survive, it will thrive.   Innovations and economics will, as always, combine to sort out the favored solutions to the various challenges that arise.   

Another article will discuss the expenses of keeping Diablo Canyon online, and why it makes economic sense to shut it down. 

Roger E. Sowell, Esq.
Marina del Rey, California
copyright (c) 2016 by Roger Sowell - all rights reserved

Grid-Scale Energy Storage in Submerged Spheres

Subtitle: Storing the Wind-Energy Makes It Reliable

First, a quote (or paraphrase) from Confucius: “I show a dull man one corner of a room, and he sits in the corner grinning.  I show a wise man one corner of a room, and he shows me the other corners, plus the entire house.”    There are quite a number of people grinning in the corner, based on some of the comments made at WUWT, which started this entire post.

Background

Some low-information commenters on Anthony Watts excellent blog made light (jeered, even) at my suggestion that the recently-announced MIT storage spheres will solve the intermittency problem of wind-energy.   A few, however, asked polite and intelligent questions.  This is my response to those polite folks. 

Pumped Storage Hydroelectric Basics

Now, as to the basics of how PSH (pumped storage hydroelectric) works, and then how the MIT spheres work.   There are numerous PSH sites in the US (more than 22,000 MW at last count by EIA – Energy Information Agency).   One of the largest sites is on the eastern shore of Lake Michigan, the Ludington Plant.   Like all PSH plants, there is an upper reservoir and a lower reservoir.  The lower reservoir is Lake Michigan.  The upper reservoir is man-made, about 360 feet above the lake, and on a sandy cliff-like edge of the lake.  At night, six turbine/pumps run in pump mode by drawing power from the grid to pump water from Lake Michigan into the upper lake.   The next day, the flow is reversed so that water flows from the upper lake through the turbine/pumps into Lake Michigan, this time generating power as needed.    The upper reservoir and land occupy 1,000 acres.  The generators produce 1,872 MW of power at maximum flow.  The penstocks (six of them) or pipes, are 1,300 feet long and 28 feet diameter.  These connect the upper and lower reservoirs.  The plant was built between 1969 and 1973.  As PSH plants go, it is large but has a low elevation change. 

In contrast, the Castaic PSH in Southern California, near Los Angeles, has an elevation change of 1,060 feet and produces 1,247 MW for up to 10 hours in generating mode.   Castaic PSH also draws power from the grid at night to pump water uphill from Castaic Lake into Pyramid Lake, the upper reservoir.  The tunnel connecting the two lakes is 7.2 miles long and is 30 feet in diameter.   Castaic PSH has six pump/turbines and one standard turbine generator.   Because the elevation difference is greater, Castaic has much lower water flow than does Ludington.

These two examples show a low-head and a high-head PSH plant (Ludington is low-head, and Castaic is high-head).  In this context, head is the elevation difference in feet between the upper and lower reservoirs.

MIT Storage Spheres

The MIT spheres on the ocean floor will do exactly the same function: draw power from the grid at night to pump water out of the spheres.  The spheres are not closed as one commenter assumes.  Instead, they are vented by a pipe to the atmosphere.  The sphere acts exactly like the lower reservoir.  It is at atmospheric pressure at all times.   A turbine/pump connected to a motor/generator draws power at night (or whenever the wind blows) and runs in pumping mode to send water out of the sphere into the surrounding ocean.  Air flows from the atmosphere through the vent pipe into the sphere.  The ocean, at that depth, has considerable pressure.  One can estimate the water pressure by dividing the water depth by two.  Thus, 1,000 feet of water will exert approximately 500 pounds of pressure.   (Engineers will know that the exact relationship is 32.2 divided by 14.696, but for estimating purposes, two will suffice.) 

During the day, when peak power is required, seawater is allowed to flow by natural pressure from the ocean through the turbine/pump into the sphere, turning the generator and producing power to the grid.  Water flowing in forces air out of the sphere through the vent line into the atmosphere.   With proper design, about 80 MW will be produced into the grid for each 100 MW consumed from the grid.

As to the servicing and maintenance issues someone asked about, this is trivial.  Proper design will have the entire turbine/pump and generator/motor equipment in the atmospheric pressure zone above the sphere.  Simply put, that building will also be vented to the atmosphere.  Likely, an elevator will convey workers and materials to the submerged sphere, much like in a mine shaft on land.   There is no need to contemplate high-pressure underwater activities.   Purists will say, at this point, yes but what about screens to keep fish and other marine life out of the turbines?   Those screens or similar devices may require periodic cleaning, but that can be done remotely with ROVs.  (remote operated vehicles, think unmanned submarines). 

As to the MIT paper indicating 6 hours of storage, and the naysayers objecting that this is far too little.  It should be pointed out that Castaic PSH has only 10 hours of generating capacity, and about 11 hours for Ludington.   However, these spheres would be storing offshore wind-energy and could require operation for several days.  There are three salient points about PSH generating time: one need only change the generating time by 1) increasing the diameter, 2) adding more spheres, or 3) increasing the head.  Put simply, if the sphere volume is the same and only one sphere is used, one can obtain double the generating time by setting the sphere twice as deep into the water – this increases the head.  Similarly, if one maintains the head constant, one can obtain 8 times the generating time by doubling the sphere’s radius.  Or, one could maintain the head constant and add more spheres of constant radius to obtain the increased generating time.  Note that it is not required to have a turbine/pump with motor/generator on each sphere.   The spheres can be connected one to another by suitable high-pressure pipes.  Very likely, the most economic choice for increased generating time is simply to increase the spheres’ size.   Spheres have a nice property for that, as materials required go up with the square of the radius, but volume increases with the cube of the radius.   One may also excavate out a hollow in the ocean floor and set the larger sphere in place, if water depth is an issue with a larger sphere. 

Now, as to the testing and prototyping as asked about: yes, the MIT publications state the system has been built, has been tested, and measurements taken on an actual sphere. 

Economics 

The economics are much criticized in the comments on WUWT.  It was overlooked, apparently, by the naysayers that MIT stated the cost per sphere will decrease as more are deployed.  This is the economy of mass production.  Henry Ford recognized this with automobiles; it still applies today.  Another cost-reduction will occur as spheres are made larger, this is the economy of scale for unit production.  Yet another cost reduction will occur as spheres are installed along trunk power lines laid on the ocean floor.  It will not be necessary to build the electrical infrastructure again for each sphere. 

Another word about economics: with a suitable number of spheres in place, there will be no need for land-based fossil-fuel power plants to be built in excessive numbers.  Instead of the 1,000 GW currently installed, the US could have only 600 GW installed, and let the spheres do the peak load work.   The savings from not installing 400 GW of on-shore fossil-fuel power is indeed large.  That will offset much of the cost of installing the spheres. 

As to the land-locked cities, spheres can be installed in the larger Great Lakes, with a power grid designed to send power from wind-farms in the Great Plains to those storage systems, then back out the next day.   Even shallow Lake Erie can have storage spheres, they would simply be buried in a suitable hole in the lake bottom.   

Conclusion

This wraps up the MIT sphere grid-scale storage technology.  It works.  It has zero energy cost.  It has very low environmental impact.  It can be constructed now, without waiting for offshore wind-turbines.  It reduces the cost of on-shore generating plants – fewer plants will be required.  Power from the spheres is almost instantaneous and can be at full power in less than 30 minutes.  It quite easily follows the load.  Economy of scale and mass production will decrease the costs.  There is a huge coastline with shallow continental shelf along most of the Eastern seaboard and Gulf of Mexico, so placing numerous spheres is quite possible.   It makes intermittent wind energy very reliable, available on demand.

Roger E. Sowell, Esq.

Marina del Rey, California