Interesting Links for 11-7-2011

[andrewducker]

• The Real Life Social Network - a presentation from last year that sheds light on why G+ is the way it is
  (tags: google socialnetworking)
• Woman sacked for having grey hair.
  (tags: women employment business usa beauty age)
• Converting a video into a photo comic - I'm actually very impressed by this.
  (tags: video html awesome)
• Storing energy as air pressure under water is apparently 50% cheaper than using batteries. Anyone know any more?
  (tags: power electricity pressure water)
• A fantastically funny story of when a small child interjects into a roleplaying game.
  (tags: funny roleplaying)

Comments

[bohemiancoast.livejournal.com]

In my summer quest to hack Japanese, I've discovered the practice of sentence mining, a language learning technique that links SRS systems (flashcards on steroids) with authentic real-life content. One way to do this is to take video for which subtitles exist in two different languages (eg Japanese and English) and then chop it up so that you end up with flashcards that have, eg, a screen shot, audio clip and the Japanese subtitle on the front, and the English subtitle on the back. Which feels a lot like the photo comics you describe above.

The routine that does this is called subs2srs and is reasonably easy to use. It doesn't generate the 'photos of silence' as described in the article but my guess is that it wouldn't be hard to amend it so that it did; and if you looked at the output flashcards in order you'd get the photo story.

The plan is that once I've internalised the entire set of flashcards I'll rearrange them so that the pic + audio are on the front, and Japanese and English subs are on the back, and start again.

[andrewducker]

That sounds like a very good way of exposing yourself to a language, giving yourself a deep immersion in it, with prompts.

Out of interest - did the player on that page work? The way you've worded your comment makes it sound like it didn't.

[bohemiancoast.livejournal.com]

I'm at work! So I didn't try.

[andrewducker]

Now you mention it, that does seem terribly reasonable :->

(It's a slideshow of still images, no sound, if that helps. But at this time of day, it probably doesn't...)

[danieldwilliam.livejournal.com]

I know a little about compressed air storage.

I’m afraid I don’t know anything more about the Hydrostor effort.

A couple of comments made in informed ignorance.

Building things to go in the water is difficult. Maintaining things in the water is difficult. If the water is the sea this is very much more difficult.

1-4 MW of capacity (presumably input and export capacity) is pretty small beer. That’s enough for one or two industrial sized wind turbines.

No vidoes on their site of it working. Not saying it doesn’t work but I’d be more impressed if they showed me.

I also note that doing the same thing under ground appears to cost 1/3 of using a battery at a $1000 per kw of storage capacity which compares well to a new nuclear build.

http://www.physorg.com/news188048601.html

One of the competitions I’m looking forward to over the next decade is the competition between different energy storage and demand management technologies.

[andrewducker]

Yes, I'd also be more impressed if I was convinced it worked.

I'd also like to see some figures on how much capacity wind-farms normally need to store in order to smooth things out.

[danieldwilliam.livejournal.com]

Now that is the billion dollar question.

On a single unit basis a 3GW wind turbine with a capacity factor of 33.3% would need 2.7 GW of storage to become a 0.3 GW baseload plant and / or a 2.7GW peaking plant. I think.

There are lots of variables; the ratio of variable renewables to non-dispatchable plant, baseload demand, peak load demand, the ratio of dispatchable plant on the grid to peak and baseload demand.

It also depends on your tolerance for throwing away free energy and your ability to shift demand about.

Imagine a Grid with 90 GW of wind etc and 80 GW of non-wind, 60GW of which is turn-downable and 20 GW of non-turndownable plant. Peak demand is 80 GW and baseload is 20 GW (not far off how the UK would look if we built all the wind and nukes that are currently technically feasible enough not to cost an insane amount of money). On a great wind day when you’re producing 90 GW during peak period you need to store 30 GW (20 NTD plant plus 10 over demand). If you got the same wind conditions over night you would need to store 90 GW.

The question becomes – how often do we have a perfect windy day? How often do we have a day with above average wind production? How often does that fall conveniently next to high demand periods?

A lot depends on how well interconnected we are. At the moment IIRC we have about 6 GW of interconnection capacity built or planned. With more interconnectors we could physically shift power to other Grids. It’s not inconceivable that we could build a bunch of interconnectors to Norway and use their hydro-schemes as storage (not by pumping water into them but by not running them and storing the water for later). European regulations require us to have interconnections equal to 10% of our installed capacity by some point soon.

With really good interconnection we probably need less storage than you might think as a good wind day displaces dispatachable thermal plant further afield.

Ditto flexible demand (where we store the energy as valuable goods). If you have an aluminium smelter which you can turn on at short notice and take advantage of low prices you can suck up a lot of supply. I’ve also had it suggested to me that data centres distributed around the world might be able to switch on to capture low prices.

[andrewducker]

Switching data centres on/off require you to quickly move data from place to place. Feasible, but possibly quite expensive.

However, for small amounts of data I can see that working pretty well.