Grid scale projects in California and net grid demand

[abasile]

I am absolutely sure that I have quoted him accurately regarding the carbon footprint from one person on a flight from LAX to Las Vegas vs 1000 characters in Google search. I recall him saying that each character you type into Google triggers a new response. I know it sounds hard to believe but that is what I heard. I have no basis for questioning the calculations submitted in above posts.

Thank you for the follow up. I trust that you quoted Prof. Wierman accurately. I'm guessing that he hasn't made this claim in peer reviewed papers; otherwise, I'd expect it to have been debunked.

Does the other claim in my post, that "The data centers in total generate more emissions than the entire Airline Industry" seem more credible? I heard the 2% number electricity number many years ago, so 7% now seems very reasonable. 7% of the electric consumption vs the airline industry also seems reasonable to me. Of course the Cloud does many more things than Google Search.

This seems quite plausible to me.

[tbleakne]

I am absolutely sure that I have quoted him accurately regarding the carbon footprint from one person on a flight from LAX to Las Vegas vs 1000 characters in Google search. I recall him saying that each character you type into Google triggers a new response. I know it sounds hard to believe but that is what I heard. I have no basis for questioning the calculations submitted in above posts.

Thank you for the follow up. I trust that you quoted Prof. Wierman accurately. I'm guessing that he hasn't made this claim in peer reviewed papers; otherwise, I'd expect it to have been debunked.

This question is getting old, and and some point we need to move on, but he was very serious that his work in this area was his academic focus. I agree that this claim would have to be backed up better than what I have recollected. We all know that academics do their best to avoid embarrassing errors.

Two thoughts:

1. A 1000 character Google search is very long, much longer than normal, and if in the worse case, each additional character triggered a new response, the total load on the search server farm would grow exponentially. So perhaps a 1000 character search could, in principle, cost many orders of magnitude more that the average search.

2. Prof Wierman's credibility in my view was raised by him reporting in the talk that, in addition to his academic work, he was actively consulting with the actual server farm companies to help them implement his ideas for energy savings. He described how this work gave him access to actual server data.

[tbleakne]

This Sunday, and several days earlier, the LA Times ran a story on the front page that is sure to give encouragement to the nay-sayers of Green Energy.

"Energy goes to waste as state power glut grows" was the Sunday title.


http://www.latimes.com/projects/la-fi-electricity-solar/

They make the point that the amount of power wasted over a year is still small, only a percent or two, but it is growing.

KQED , San Francisco PBS, posted a similar story:

https://ww2.kqed.org/science/2016/04/04/what-will-california-do-with-too-much-solar/

The KQED story had this graph from CAISO fro Mar 27, 2017:



My question is this:

Every clear morning we have the well-known "Duck Curve" where there is a big drop in Net Demand between 7am and 10am as the solar production rises faster that the load. On most days this surge is accommodated by throttling back natural gas production. So why does the graph not show hardly any significant cut back in natural gas production or other generation on this day?

The graph shows that wind declined in mid-day, largely compensating for the mid-day rise in solar.

Nuclear production has be to be steady on a time scale of days because short term cut-backs lead to buildup of intermediate radio-nuclides that poison the reaction. Hydro electric can adjust on time scales of an hour or less, but often the downstream flow has to be maintained. Geothermal is a steady source that is wasted if it is cut back.

My understanding is that natural gas can be adjusted fast, on the order of 15 min. On this date the total maximum production level of 24 GW was about half the normal peak summer load. Perhaps only natural gas plants vary in their adjustability, and all the easily adjusted ones were already off-line.

Perhaps the new battery banks that have been installed to back up the natural gas "peaker" plants so they don't have to run in idle until needed will reduce the minimum natural gas generation.

If anyone has informed answers, please post.

[edatoakrun]

..So why does the graph not show hardly any significant cut back in natural gas production or other generation on this day?...

That graph shows demand on March 27.

It is not very relevant to much higher Summer PEAK demand, often requiring more than twice the MW, after a much steeper daily increase and much higher solar production, resulting in the much more pronounced duck curve, as shown the graph of recent net demand at:

http://www.caiso.com/outlook/SystemStatus.html

It is true that, in a panicked response to the largely fake energy crises created by Enron and other corporations, California has over-built generating capacity.

And yet, there has been virtually no coordinated effort to offer variable BEV charging rates to utilize this excess capacity, and to tune daily and seasonal demand curves to match available generation.

BEV driver to State of CA:

I OWN BATTERIES.

Just tell me when you want me to charge, with discounts on my home power bills, AND my public charge fees...