Grid scale projects in California and net grid demand

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
More on offshore wind power for California.

JeremyW said:
Stanford has an excellent paper on some details of off shore site potential in CA: https://web.stanford.edu/group/efmh/jacobson/Articles/I/Offshore/DvorakRenewEn2010.pdf

They suggest a site off the Humboldt Coast which, in summer, essentially looks like a base load power plant. Awesome! They do fail to mention the upstream upgrades necessary after connecting into the Humboldt Bay PG&E system, however. The load and lines in the area cannot support 1500 MW of new generation.

Still, I hope something gets built there in my lifetime. :)
Yes, too bad this Stanford proposal never went anywhere. The wind potential looks very good and the area is one of the few off the California coast shallow enough to avoid floating turbines. The lack of transmission lines into the area you note must be a significant problem. There is a retired nuclear reactor there, so at least there was the possibility of using its substation.

Sunday's LA Times has nice report of plans to build massive wind turbines offshore and possibly onshore, with segmented blades.
http://www.latimes.com/business/technology/la-fi-cutting-edge-windmills-20160313-story.html
The article does not mention the Humbolt area, but near the end it mentions a new proposal from Fall 2015 from Trident Winds for a project off Morro Bay using floating turbines.

http://www.mercurynews.com/science/ci_29086397/californias-first-ocean-wind-farm-works

From several web sites I collected some features and parameters for this proposal. This proposal is still in a state of flux, so all numbers don't match.
1.0 GW wind farm
100 turbines, floating.
34 Km offshore
600 ft high
wind resource 8.5m/sec vs Humbolt Bay 10m/sec
4000 hours per year operation vs 1500 hours per year solar.

10MW per turbine would be very large. Elsewhere a size of 6.5MW was mentioned. Still very large compared to typical 1.5 to 2 MW onshore turbines.
They propose connecting to the grid using the existing substation still in place in Morro Bay for a large natural gas generator that was retired in 2014. This generator was rated at 650 MW.

The farm would not be visible from the beach, but it would be visible from an elevation such as Hearst Castle.
They claim there would be less impact on birds offshore compared to onshore. Not sure they have data to support that.
They expect a long approval period, even 10 years. During that time other large offshore floating wind farms should be built, including one off Scotland.
The same developer had earlier proposed a wind farm off the Oregon coast, but the project cost of $.24/kWh was rejected as too high.

Not quite sure why the array could not be placed farther from land to minimize visible impact of a pretty area. Perhaps 24 km is near the limit of how far AC can be transmitted through the ocean, given the high capacity of sea water. Converting to DC would cost more.
 
Meanwhile, here's a report on the state's decreasing use of coal:
Actual and Expected Energy from Coal - Overview
(I'm having SQL trouble with the link even though it's the CEC website, so you'll need to google it yourself).

I saw some news talking heads claiming that Oregon was the first state to outlaw coal, but California did essentially the same thing a decade ago, by limiting CO2 emissions and prohibiting new coal power contracts or extensions of old ones. I wasn't aware that LADWP was selling their share of Navajo to SRP, but googling,
LADWP board approves sale of coal-fired plant stake to SRP
http://www.publicpower.org/media/daily/ArticleDetail.cfm?ItemNumber=43928

I see that their arrangement with SRP says the supply will end when the sale closes as of July 1st this year, about 3.5 years before LADWP's contract expires. Yay! for us, although it's not as if the plant's going to be shut down.
 
GRA, thanks for your post on a vote in 2015 by Los Angeles DWP to disconnect from coal-generated power imported from Arizona's Navajo plant by mid 2016.

I have also seen reports that a small amount of coal-generated power might still be coming into southern California from southern Nevada, into the Tahoe area from western Nevada, and into northern California from Oregon. The total was a few percent of total California consumption. However, my understanding is that these numbers are declining fairly quickly. Here is a link to a California summary that was updated Dec 2015, but appears to have been generated a year earlier. It says at that point 7% of total was still imported coal power. The Navajo plant shutdown is on this chart, suggesting they are keeping to the schedule.

http://www.energy.ca.gov/renewables...cuments/current_expected_energy_from_coal.pdf
 
tbleakne said:
GRA, thanks for your post on a vote in 2015 by Los Angeles DWP to disconnect from coal-generated power imported from Arizona's Navajo plant by mid 2016.

I have also seen reports that a small amount of coal-generated power might still be coming into southern California from southern Nevada, into the Tahoe area from western Nevada, and into northern California from Oregon. The total was a few percent of total California consumption. However, my understanding is that these numbers are declining fairly quickly. Here is a link to a California summary that was updated Dec 2015, but appears to have been generated a year earlier. It says at that point 7% of total was still imported coal power. The Navajo plant shutdown is on this chart, suggesting they are keeping to the schedule.

http://www.energy.ca.gov/renewables...cuments/current_expected_energy_from_coal.pdf
That's the same report I linked to ;)
 
Two recent reports from NYTimes, first one on lack of progress on building 5 HVDC transmissions to carry wind power from Midwest to population centers in both East, West:

http://www.nytimes.com/2016/03/24/b...rnative-energy-local-stymies-an-industry.html

A 3-to-2 vote by the PUC in MO has stalled one project, Green Belt Express. This is poor; I did not realize the extent of public resistance to such transmission lines. Being DC, there is no radiation. There is no movement or sound like wind mills, only a faint buzz from HV discharge. I assume the lines would have a dedicated corridor so no one would live or work right under a transmission line.

The family whose objections are highlighted talks about the east getting its wind locally, but the wind resource in the midwest is much stronger and steadier than other regions of the country. The Southeast has especially low wind resource. On page 2 of this thread I posted a NREL color map showing these differences.

The second NYTimes report is about the very slow progress of advanced nuclear:

http://www.nytimes.com/2016/03/22/science/nuclear-energy-power-plants-advanced-reactors.html?_r=0

Advanced nuclear reactors is a big subject with many options and tradeoffs. Different designs with different coolants and moderators offer different neutron energy spectra, from thermal spectrum, like conventional pressurized water reactors, to fast spectrum reactors. The faster spectra options can burn spent nuclear fuel from conventional reactors, or breed fissionable fuel from Thorium or U238. These designs generate much less nuclear waste and they should be much safer because they are unpressurized.

Unfortunately the Nuclear Regulatory Commission seems ill-equipped and underfunded to evaluate these designs, with mountains of paperwork. In the last decade, very little public money and only token amounts of private money have been spent on advanced nuclear. Bill Gates has said it is "insane" how little money is being invested in advanced nuclear.

One really big question is whether these advanced designs can be built and deliver power at much lower costs than conventional reactors with typical huge construction over-runs, high operating costs, and frequent down time. Really significant development and prototype operation is required.

The story reports that a team including Southern Company and Terrapower has received $40M of public money to work on their Molten Chloride salt Fast Reactor (MCFR). This is a big switch for Terrapower, which has been funded principally by Gates. Their previous traveling wave concept was liquid sodium cooled. They say they have not abandoned the traveling wave, but liquid sodium can become unstable, so I am glad to see them embracing liquid salt.

A good technical overview of the the MCFR:

http://energyfromthorium.com/2016/01/16/doe-terrapower/
 
As described in earlier posts on this thread, the US has had ZERO offshore wind power. This is about to change.
Construction of a new offshore wind farm near Block Island, off the coast of Rhode Island, is scheduled to be completed this Fall 2016.

The community living on this island has had no grid connection to the mainland, and therefore it has been dependent on expensive diesel power. This makes it a good candidate to be the first to get offshore wind, which is more expensive than onshore wind. A side benefit of the project is that it will pay for a undersea connection from the island to the mainland grid, to allow two-way exchange of power.

map2.jpg


Size: 5 turbines, each 6 MW, 30 MW total

These are very large, 1.5 to 2.5MW turbine size is much more common. Efficiency and production increases with turbine size. Large turbines are difficult to transport and install onshore, but offshore can go larger, with delivery and installation by barge and special ships.

Cost: $ 250M
This is high, $8 per peak watt, vs utility scale solar at near $1 per watt. However, wind measurements at this location predict a utilization factor of %49, which is very impressive. Compare solar, which typically delivers about 1500 kWh per kW per year, which is a utilization of 17%. As they gain more experience with the technology (see below), to cost should drop.

Advanced Technology

Conventional wind turbines use large mechanical gear boxes to convert the relative slow rotation up to speeds compatible with multi-pole 60 Hz generation. These gear boxes are expensive, heavy, inefficient, and they require considerable maintenance. Larger turbines turn slower, so they require even larger gear boxes.

This project is using Direct Drive, which is leading state of the art, eliminates the gear box and its problems. The rotors are permanent magnet, which use scarce Rare Earths elements and the high permanent magnetic fields make assembly quite tricky. Despite the higher initial cost, Direct Drive now has 27% of the offshore market because it offers lower maintenance costs.
Much of the technology, support, and installation labor on this project is coming from Europe. The turbines are being supplied by GE Alstom, which is a large wind power supplier, based in France, that GE purchased last year.

IMG_0471.JPG


Superconducting Direct Drive, which will solve these problems and offer even higher efficiency, is under advanced development. MgB2 is a promising superconductor compound for this application:

http://iopscience.iop.org/article/1...5C74612BFD745A453EA.c4.iopscience.cld.iop.org
 
Great piece in LA Times on how Wyoming wants to tax wind power coming to California:

http://www.latimes.com/nation/la-na-sej-wyoming-wind-tax-snap-story.html

I posted about this long-planned project some months ago on pg 2 of this thread, beneath the NREL map of national wind potential,
http://www.transwestexpress.net/
but the new story has much more juicy political detail.

It is amazing that wind power can be sent 700+ miles from WY to CA for 1$/watt using HVDC. This makes it more than 2 times as expensive in $/watt than local grid-scale solar, but its capacity factor is much larger, several times as many hours per year. WY wind is more consistent than CA wind onshore wind, and even with the additional transmission charge, it is cheaper than offshore wind, especially floating offshore wind.

So both the engineering and economics of this project makes sense, but the politics are unfortunate. It would be best for WY folks to consume their own wind power locally, but they can burn coal cheaper than wind, so they only want to export the wind. They also don't like that wind power generates only a small fraction of as many jobs per kWh as coal. They are putting up none of the money; it is coming from:

"Anschutz Corp., which long has been involved in fossil fuels and owns Staples Center in Los Angeles. It is led by Philip Anschutz, a major donor to Republican candidates and conservative causes."

Right now the WY tax is 1$ per MWh, only .1 cent per kWh, so it does not bring in much revenue. The WY legislators would like more since their coal revenue is declining.
"“Just about every legislator we’ve met with asks us, ‘You tell us how much we can tax you before we put you out of business,’” said Bill Miller, chief executive of the Power Co. of Wyoming, which is planning the wind farm."
Unfortunately Miller made a public statement that they did not need the federal wind tax credit for this project to be competitive, so the WY legislators are saying they want to collect all of the federal credit for themselves.

On the pro side, the project would spend a total of 8 B$, not all in WY. The NREL map shows plenty of great wind power in nearby states, which offer incentives rather than taxes, so Miller is holding firm on rejecting the tax. No new wind farms have been built in WY since the tax was imposed in 2012.
 
Nice lecture relevant to Utility Scale power at Caltech last night, "How Green is the Cloud" by Adam Wierman.

The Data Centers that power the cloud are now consuming 7% of the nations's electricity, up from 2% a few years ago.
The rate of rise Cloud consumption has slowed, but it is still growing 12% per year, vs 2% for total electric demand.

Typing 1,000. characters into Google Search costs the same Carbon footprint as a short airline flight from LAX to Las Vegas. The data centers in total generate more emissions than the entire Airline Industry. I found this very surprising.

Most of the big Cloud players, Apple, Google, Facebook, Amazon, etc have made strides to increase the renewable power they consume, and increase efficiency, and lower cooling costs. However, the millions of servers in a data center are usually running all the time, irrespective of the load. The reason is that booting up a server takes some time, on the order of a minute, and load fluctuations are hard to predict.

Some of the Cloud load is deferrable to times, like the middle of the day, when there is more solar power available. When the load exceeds the available renewable power locally available, the deferrable load can be shifted to other data centers. Moving power clear across the country is not yet practical, but moving the data is. Probabilistic game theory analysis, pioneered by Wierman, shows that it is possible to run a data center with much smaller powered reserve, saving power by keeping more servers off until needed. These algorithmic changes can generate savings equivalent to several mega-watt hours of battery storage per data center.
 
tbleakne said:
The Data Centers that power the cloud are now consuming 7% of the nations's electricity, up from 2% a few years ago.
The rate of rise Cloud consumption has slowed, but it is still growing 12% per year, vs 2% for total electric demand.

Typing 1,000. characters into Google Search costs the same Carbon footprint as a short airline flight from LAX to Las Vegas. The data centers in total generate more emissions than the entire Airline Industry. I found this very surprising.
Wow, these are incredible numbers! I use "the cloud" and do lots of Google searches in the course of my work (it's usually faster than going directly to my reference materials), so that is sobering.

On the other hand, while the cloud is contributing to modest growth in total electric demand, it is primarily replacing other areas of demand. It seems safe to say that, at least for now, cloud electrical load is probably more constant and predictable than the loads it is replacing. That is helpful from the standpoint of grid stability. The same could be said for multi-hour EV charging sessions (though not for DC fast charging).

That the nature of our grid's electrical loads may be trending toward greater stability (not counting residential/distributed solar generation) may buy a bit of time to continue integrating renewable sources without ill effect.

That doesn't diminish the need to seek efficiencies in cloud data centers, of course.
 
tbleakne said:
Typing 1,000. characters into Google Search costs the same Carbon footprint as a short airline flight from LAX to Las Vegas.
And in other news, the earth is flat.
 
SageBrush said:
tbleakne said:
Typing 1,000. characters into Google Search costs the same Carbon footprint as a short airline flight from LAX to Las Vegas.
And in other news, the earth is flat.
I'll admit that I'm also skeptical. Perhaps they are understating the emissions arising from a short flight. But the main point stands - data centers use a lot of juice! I'm not going to stop running my Google searches, though.
 
abasile said:
SageBrush said:
tbleakne said:
Typing 1,000. characters into Google Search costs the same Carbon footprint as a short airline flight from LAX to Las Vegas.
And in other news, the earth is flat.
I'll admit that I'm also skeptical.
Skeptical ? How about a nanogram of a picogram of a femtogram of common sense ?

If you like numbers though, Google said in 2016 it used 0.01% of the world's electricity;
Search is about 10% of it's data servers load;
And it is implementing algorithms to to reduce its electricity use 15%.
https://www.theguardian.com/environment/2016/jul/20/google-ai-cut-data-centre-energy-use-15-per-cent

So on average, Search today (presuming no advances in computing efficiency -- obviously not true) is 0.00085% of the world's electric footprint. I estimate a search query as consuming one watt-hour of energy, although Google's commitment towards powering its data centres with renewable energy drops the actual carbon footprint dramatically. And in fact, Google will be powering 100% of its data centres and operations with alternative energy in 2017 and this excludes the $3.7B invested in alternative energy production.
 
SageBrush said:
So on average, Search today (presuming no advances in computing efficiency -- obviously not true) is 0.00085% of the world's electric footprint. I estimate a search query as consuming one watt-hour of energy, although Google's commitment towards powering its data centres with renewable energy drops the actual carbon footprint dramatically.
I'm willing to give Prof. Wierman the benefit of the doubt based on his credentials. I highly doubt that he was trying to deceive anyone. At the same time, he may have made some very questionable assumptions. I don't know.

That Guardian article also mentions that efficiency at Google has improved 3.5x since 2011, so Dr. Wierman (a CS/Math professor at Caltech) may have been using old information. Further, a single Google search actually involves quite a number of operations besides straight text search. They are searching as you type (multiple searches), running ad servers, and checking/loading shopping data, for instance. Further, there are significant costs to all of the web spidering to gather the data to be searched, so perhaps Prof. Wierman may have been thinking in terms of amortizing those costs across all Google searches (who knows). He may not have been accounting for the use of renewable energy. He may have been lowballing the amount of energy used on short distance flights.

So, while I'm skeptical that 1000 characters' worth of Google searches (or a single search with 1000 characters?) really would generate the same carbon footprint as a LAX-->LAS flight, I'm also very skeptical that searches use only a few watt-hours of energy in total.

Anyway, I'm not terribly concerned about this particular point, as I'm not going to change my online behavior one way or the other. The big picture is that data centers have become a rather significant part of the world's energy usage!
 
abasile said:
SageBrush said:
So on average, Search today (presuming no advances in computing efficiency -- obviously not true) is 0.00085% of the world's electric footprint. I estimate a search query as consuming one watt-hour of energy, although Google's commitment towards powering its data centres with renewable energy drops the actual carbon footprint dramatically.
I'm willing to give Prof. Wierman the benefit of the doubt based on his credentials. I highly doubt that he was trying to deceive anyone. At the same time, he may have made some very questionable assumptions. I don't know.

I agree, the Prof was so misunderstood the entire story is laughable and pathetic.
His lecture is apparently not available on the internet so we can only guess what he was actually saying:

If the flight from LA to Las Vegas is 40 minutes, and
my recollection of CO2 expiration in an adult human of ~ 300 Kg a year is correct, and
he was using old Google data of 0.2 grams CO2 per search, and
his 1000 char search = 100 searches that consume 0.2 grams a piece,

It works out that a Wierman mega-search powered by a dirty grid = the amount of Co2 expired by one person on a 40 minute flight. Not the flight !!! The breathing of one person.

Mostly I am amazed that otherwise intelligent people like yourself can swallow a story line that is off by a magnitude of around 350,000 fold for a simple search.
 
SageBrush said:
Mostly I am amazed that otherwise intelligent people like yourself can swallow a story line that is off by a magnitude of around 350,000 fold for a simple search.
I figured as much. However, this may not be totally crazy. Perhaps 100 kWh could be a lowball estimate for the energy to transport someone from LAX to LAS (an optimistic 70 mpg equivalent per passenger on a full flight, times three gallons equivalent for the distance of 210+ miles, and 33.6 kWh/gallon of gas). If each Google search session involves an average of 10 characters, then we have 100 search sessions, or 1 kWh per search session. If that includes a pro-rated portion of Google's search servers' idle time, the other relevant Google servers, plus all of the data center cooling and maintenance costs/energy, plus a portion of the web spidering and data analysis costs, then maybe 1 kWh/session isn't super crazy. It still seems pretty high, but not by multiple orders of magnitude.
 
abasile said:
However, this may not be totally crazy. Perhaps 100 kWh could be a lowball estimate for the energy to transport someone from LAX to LAS (an optimistic 70 mpg equivalent per passenger on a full flight, times three gallons equivalent for the distance of 210+ miles, and 33.6 kWh/gallon of gas). If each Google search session involves an average of 10 characters, then we have 100 search sessions, or 1 kWh per search session. If that includes a pro-rated portion of Google's search servers' idle time, the other relevant Google servers, plus all of the data center cooling and maintenance costs/energy, plus a portion of the web spidering and data analysis costs, then maybe 1 kWh/session isn't super crazy. It still seems pretty high, but not by multiple orders of magnitude.
Clearly many orders of magnitude off.

Per the NYT Dec 2016 (not known how old the Google data is),

Google search is 10% of services;
Total energy is 3.3 million MWh a year;
Two trillion searches a year.
Works out to about 600 WSeconds a search or around 0.16 Wh a search
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=13&cad=rja&uact=8&ved=0ahUKEwiv0Ieb37bTAhVE3WMKHYDpBVEQFghlMAw&url=https%3A%2F%2Fwww.nytimes.com%2F2016%2F12%2F06%2Ftechnology%2Fgoogle-says-it-will-run-entirely-on-renewable-energy-in-2017.html&usg=AFQjCNGF8sxkBMOfxb0GaSAkPDt7971n0w&sig2=eoHrxWiRrfLDS1hzWFDPSA

You can also start from a google search since Google reports how long it worked out the results. I usually see 0.25 - 0.5 seconds. Sounds like 10-15 CPUs are involved
 
abasile said:
1 kWh per search session. If that includes a pro-rated portion of Google's search servers' idle time, the other relevant Google servers, plus all of the data center cooling and maintenance costs/energy, plus a portion of the web spidering and data analysis costs, then maybe 1 kWh/session isn't super crazy. It still seems pretty high, but not by multiple orders of magnitude.
Aww. come on!

If electricity costs Google 5 cents a kWh,
They serve up over two trillion searches a year,
and Search is ~ 10% of their services,

Just their electric bill would be ONE TRILLION USD a year.
 
SageBrush said:
If electricity costs Google 5 cents a kWh,
They serve up over two trillion searches a year,
and Search is ~ 10% of their services,

Just their electric bill would be ONE TRILLION USD a year.
Wow, I didn't realize Google fields 2T searches/year. That's a lot!

Now, I think it's fair to assume that search is disproportionately CPU and power intensive. Serving up video, and processing videos, would of course also be big. Even if search represents half of Google's energy usage (which I seriously doubt, though I'm guessing it's > 10%), at only $0.05/kWh, that's still $200 billion per year for electricity. Hmm. I don't believe that at all.

Based on 2T searches/year, I'll have to concede your point. The professor may be off by two orders of magnitude! (Three might be a stretch.) Either that, or somehow tbleakne mis-heard or mis-interpreted what he said. That's possible, but I know him to be a sharp guy. As we've speculated, there could be built-in assumptions based on old data (many fewer searches, much less efficient systems, "dumber" algorithms, etc.)
 
abasile said:
Based on 2T searches/year, I'll have to concede your point.
I did a lot of Google searches during this conversation. Lucky for the planet it only cost about 0.2 Wh per search and not 200,000 Wh like a flight from LAX to LAS ;-)

Silly numbers aside, Google powers 100% of its operations with clean energy. And I don't mean REC purchase BS, but real additional generation. That is just phenomenal. The Google environment report they recently published is fascinating reading. They are probably decades ahead of the country in their thinking and actions. No diversion to landfill, re-use, water conservation, huge strides in reducing oil use ... the list is amazing.

Go Google!
 
SageBrush said:
I did a lot of Google searches during this conversation. Lucky for the planet it only cost about 0.2 Wh per search and not 200,000 Wh like a flight from LAX to LAS ;-)

Silly numbers aside, Google powers 100% of its operations with clean energy. And I don't mean REC purchase BS, but real additional generation. That is just phenomenal. The Google environment report they recently published is fascinating reading. They are probably decades ahead of the country in their thinking and actions. No diversion to landfill, re-use, water conservation, huge strides in reducing oil use ... the list is amazing.

Go Google!
I appreciate your efforts to look into this. I'm especially impressed at the high efficiency of Google's search architecture, considering the enormous volume of data that each search can potentially reference. It's an understatement to say that they've really come a long way. And I'm very pleased to know that Google is so focused on environmental stewardship. Truly awesome.

What puzzles me is that Prof. Adam Wierman is currently the chair of the Department of Computing and Mathematical Sciences at Caltech. On his web page (http://users.cms.caltech.edu/~adamw/), the first research interest that he lists is "Algorithms for Sustainable IT". So he ought to know this area quite well! I may want to look more closely at his work to understand more precisely what he has been saying. Perhaps you have already done so.

What bothers me is that this demonstrates the ease with which FUD or "alternative facts" can be spread, quite often unintentionally. Clearly, I'm not immune. I will have to say that I've read more than a few articles in purportedly credible publications, such as IEEE Spectrum, that have seemed pretty dubious when addressing areas that I know relatively well, such as electric vehicles.
 
Back
Top