On 10/20/07, David Ernst <david.ernst@davidernst.net> wrote:
> Apologies to everyone who had hoped that this conversation had gone
> away. I've been itching to reply all week, but somehow I've gotten
> less good at spending hours of personal time every day on my
> computer. :) Maybe this is good, but I'm not sure.
Yeah, I think it's good to get away regularly.
> You make an excellent point, I was naively assuming that power plant
> output was roughly equivalent, and that's clearly not the case.
> However, you're arguing it too far in the other direction (see
> below)...
>
> One of my points is that additional power plants, per se, is not
> necessarily all that bad. Indiana has at least 15 power plants, and
> even if they aren't as high-power as many others, they exist here
> without nagatively impacting quality of life all that much (I only
> know where a couple of them are) even though they are about mostly
> using the dirtiest power generation technology still in operation in
> the western world.
I don't believe "without negatively impacting quality of life all that
much" is accuate; see below.
> Put another way, one of my main points is this: a lot of
> environmentalists take an attitude of "every bit of development is
> bad", I look at it more as "we want a very good cost/benefit ratio for
> any development that we do." Even if my numbers were off, I still
> like the cost/benefit of computing.
I look at it this way: I have already grown up in a world
significantly more polluted than that of my parents. At the current
environmental cost of the energy we're using (assuming that it doesn't
go up or down), my grand-children will inherit a world significantly
more polluted than mine, and I'll be surprised if my
great-great-grandchildren (should I have any) know what a clear blue
sky looks like, or what it feels like to swim in the ocean.
For a look at how long it will take for the damage we've already done
through carbon emissions to stabilize (much less return to previous
levels), assuming optimistic figures that we'll start cutting back on
yearly emission growth today, and be actually reducing our yearly
emissions by halfway through this century:
<http://www.mongabay.com/images/2006/graphs/ipcc-lag.jpg>
I don't believe "every bit of development is bad", though I think I'm
probably an environmentalist because it's painful for me to consider
how quickly our world's wild spaces are being destroyed, taking the
away from future generations for negligible economic gain today.
I firmly believe in cost/benefit analysis. I just don't agree that
using our current situation as an acceptable baseline is a sane place
to start. Our current situation is unacceptable, and future
development must reverse the ratio. We must get more benefit for the
cost, so we can bring the cost back to an acceptable baseline.
> I'm quite sure you meant to say "5GW" or equivalently "5,000MW" in
> that sentence. Simple mistake, but three orders of magnitude really
> changes the equations. :)
That's true.
> and added up all of the power ratings of all the plants. The total
> was 25,885 MW, or about 1.75 Itaipus. Using the "25 Itaipus to double
> the world's computers" formula, this means that the COAL FUELED plants
> in Indiana and Ohio provide about 7% of energy required to run the
> world's computers. So, even knowing that this only includes coal fuel
> plants, my previous assertion was about an order of magnitude off. I
> stand corrected. But, 7% is definitely a dent in my book.
I'll concede that's a dent, but how sustainable is that in terms of
how long will that fuel be available in economically viable
prices?...without considering the environmental impact of adding yet
more carbon to our atmosphere.
> On Sat, Oct 13, 2007 at 12:10:36AM -0400, "Simón A. Ruiz" wrote:
> >Unless it turns out that computers are in fact feeding power into the
> >electrical grid, there's really no way to argue against coming up with
> >strategies to deliver computing with less power.
>
> Here's my basic argument: I think it's fair to assume that cars use at
> least 20x as much power in general use than computers do, and
> typically generate more pollution per unit energy. So, add up cars
> and computers and I'm seeing a pie chart showing that 95%+ of the
> environmental impact coming from cars. Then you're saying "let's make
> computers more efficient to benefit the earth" and what I'm saying
> back is "ok, but if we somehow got computers to run without
> electricity, we'd make at best a 5% improvement. I think our energy
> is better spent working on a 6% gain in fuel efficiency for
> automobiles." And, when something like a 50% gain in fuel efficiency
> can be had for simply driving a small car vs. a big one, it's hard for
> me to get too excited about computers.
>
> And then, there's refrigerators and air conditioners...
Well, my first reaction is it's not an either-or choice. Every bit of
development needs to take a good hard look at how to improve their
cost/benefit ratio drastically.
I'm already not driving an SUV, switching to CFLs, minimizing the A/C
and heating I use. It's up to the car geeks and lighting geeks and
HVAC geeks to get excited about revolutionizing the power consumption
of those areas—I don't know enough about internal combustion engines,
myself, to be excited about the next big breakthrough in energy
efficiency (and I'm cynical that the car companies and the oil
companies even care about improving efficiency until the price of gas
doubles again).
I'm a computer geek, on a computer geek mailing list, so that's what I
get excited about and write about.
I agree that improving computer power efficiency is far from the
biggest piece of the puzzle that really turns our environmental rape
around, though I don't think it'll be an insignificant piece.
I understand that when surrounded by SUVs, central air, and Boeing
747s it's hard to get excited about reducing the power consumption of
your computer at home, but I work with computer labs, and if we
consider your estimate correct, then each of our 20 station computer
labs is the equivalent of one automobile and we have the equivalent of
7 or more automobiles running.
But anyways, the main thing that excites me about this is the
possibility of bringing computing to a third-world situation like my
home country (or inner-city USA) in a cheap, sustainable way. If you
can get 10 times as many computers on the same power circuit, costing
the same amount per month on the power bill, you're ahead ten times.
If they have no moving parts, then the failure rates drop dramatically
and maintenance costs plummet.
On the other hand, though not mutually exclusive by any means, if your
power availability is lower, because you're so far from civilization
that you can't connect up to a power-grid (or your local power-grid
sucks, as ours does in my home country) and you produce all your own
power, it's ten times more convenient to use a computer which uses ten
times less energy.
Incidentally, my guess is you've forgotten to include planes, trains,
boats(ships), energy plants (though I guess you do mention
refrigerators and air conditioners, if that's what you meant), all
kinds of manufacturing plants and factories, heating all our
buildings, etc. in your pie graph if you think cars are 95%+ of our
environmental impact.
<http://www.mongabay.com/images/2006/graphs/end-use_sector_2004.jpg>
> Don't forget wind!! :)
I was going to say that the cost/benefit of mounting a wind generator
in southern Indiana is not really all that viable, consider the
up-front and ongoing costs compared to the average wind speed, but
then I saw a new kind of wind generator on the Make blog that will be
significantly cheaper to install and maintain:
<http://www.makezine.com/blog/archive/2007/10/thirdworld_wind_power.html>
So, yeah, wind can become a supplement even in southern Indiana,
especially in the winter months when solar power is less abundant
(thought that happens to be when wind generators need most
maintenance).
And there's Stirling engines...
> And, I doubt that the typical third world
> person is going to be more excited about a computer than they would be
> about a car or an air conditioner as their quality of life improves.
I am a third world person, and my perception is different than yours.
In my tropical third-world country, air-conditioning is not a way of
life like it is here in temperate Indiana, and neither is everyone
owning a car.
Would they rather have a car or air-conditioning? Well, a lot of them
have a chance to and many of those decide it's not worth it. My dad
has a car, and he rarely uses it. We'd rather take taxis or buses the
vast majority of the time. My brother is a taxi driver. We also have a
room with an air-conditioner, but it's only really on when my fiancée
and I are visiting.
But we have three computers at my dad's house. Every one of my
siblings has at least one computer at home (and my nephew has an
XBOX). A good chunk of my extended family has computers in their home.
And these are on and being used regularly (when the power isn't going
out). I am in regular correspondence with my family, and am kept
abreast of news, thanks to computers and the Internet.
The number of institutes in my city alone that are dedicated to
teaching people computer skills so they can get better-paying jobs is
staggering.
The quality of life improvement of having your own car, or having air
conditioning is really not all that great, IMHO, compared to the
quality of life improvement of having a computer with Internet access,
especially when you do a cost/benefit analysis.
> David
Simón
_______________________________________________
BLUG mailing list
BLUG@linuxfan.com
http://mailman.cs.indiana.edu/mailman/listinfo/blug
No comments:
Post a Comment