20 sats \ 3 replies \ @faithandcredit 15 May 2022
3000mw seems like alot? i thought the total bitcoin network only was about 7000mw power consumption. Unless the 3000mw includes cooling and what not, and i wonder if in the end thats as big an expense in terms of electricity as the miners themselves or maybe its a typo and they are turning off 300mw. still a lot for a single facility. I dont understand :)
reply
5 sats \ 1 reply \ @BTCMiner OP 15 May 2022
https://ccaf.io/cbeci/index
Using CBECI numbers, 7GW is about the theoretical lower bound (i.e., everyone using the most efficient hardware available). The upper bound (i.e., everyone using S9s) is about 34GW. Current best estimate by them is 16.6GW at this moment. It's really hard to know as shipments of new, modern ASICs are increasing (see total hashrate), and nobody knows (yet) how much of the old equipment is going to be finding new homes, which can take months between when powered down at one site, refurbished and tested, then bought and sold on secondary markets, and eventually powered back up at its new home (possibly in another country / continent).
reply
20 sats \ 0 replies \ @faithandcredit 15 May 2022
Thanks this comment helped me put things in perspective because i only used -most effecient miner- in my calculation.
reply
0 sats \ 0 replies \ @BTCMiner OP 15 May 2022
That was a really badly worded Tweet.
The ERCOT grid lost 3,000MW. Riot's (brand new) facility has the capacity of 700MW. They are not yet at capacity, I believe, but their website apparently shows their capacity utilization plan:
https://d1io3yog0oux5.cloudfront.net/riotblockchain/files/pages/bitcoin-mining/whinstone-u-s/hosting_mw.png
[Larger image]Source: https://www.riotblockchain.com/bitcoin-mining/whinstone-u-s
They did not say if the are curtailing for only their own self-mining (probably ~200 MW now), or if they are curtailing for hosted (customer-owned) equipment as well (probably close to another ~200 MW). Presumably the customers would have the option as this is a voluntary act in response to ERCOT's request, and not contractually required nor an order by the gov't.
reply
1 sat \ 0 replies \ @BTCMiner OP 14 May 2022
The exact level of electricity being curtailed by Riot Blockchain wasn't shared.
Also, the facility hosts equipment owned by Riot Blockchain as well as equipment for clients, so it is unclear if Riot is describing curtailment for only the equipment they own, or for all mining occurring at their facility.
reply
15 sats \ 2 replies \ @BTCMiner OP 14 May 2022
Here's an article on the event that triggered this:
ERCOT asks Texans to conserve power after electricity plants go offline
https://www.dallasnews.com/news/2022/05/13/ercot-asks-texans-to-conserve-power-after-electricity-plants-go-offline/
https://archive.ph/kMZta <-- An archive of the article, with no paywall / no subscription required
reply
21 sats \ 1 reply \ @itsrealfake 15 May 2022
It's not clear to me how "hot weather" causes power plants to go "off-line"
reply
1 sat \ 0 replies \ @BTCMiner OP 15 May 2022
They run fine in a lot hotter weather in the summer so it wasn't the heat alone that did it.
The ERCOT grid in Texas is pretty much isolated -- there is limited transmission capacity to import power from (or export power to) neighboring grids.
https://upload.wikimedia.org/wikipedia/commons/thumb/f/f4/NERC-map-en.svg/800px-NERC-map-en.svg.png
[Larger image]Source: https://en.wikipedia.org/wiki/Texas_Interconnection
Thus a failure may cascade into many failures and that will really affect reliability. They overbuilt on wind, for example, but if the wind isn't blowing when they need it the rest of the generation capacity may not be able to make up for the difference. I have no idea if lack of wind was a contributor this time but it has been a common denominator in previous ERCOT grid reliability incidents.
reply