CPN JON
Good news for you. Ontario has a deal where you 'buy' a solar power
system (either flat panels on your roof or a tracking system in your property,
a little more in your face but more efficient it's a choice)
They are guaranteeing you something like 50 cents a kilowatt.
That's nuts, I haven't figured out their rationale other than trying to
be green. Anyway something to look into. After it pays for itself (I
think 10 years?) it's starts paying YOU a nice monthly check.
I just looked it up and they are paying 64.2 cents a kilowatt. It's saying
lots of farmers have applied to help 'diversify' their 'crops'. I wish I lived
there on an acreage. I'd have ground based systems and be out there
dusting the snow off :) (it should actually fall since the panels would
be at quite an angle in the winter to follow the sun)
GOOD LUCK!

I never worry about the power being used by my WCG research which includes CEP2 .
I live in a cold climate in Millarville, Alberta, Canada and we have atrocious electrical power production mostly coal based while we use relatively clean natural gas to get oil out of the tar sands. Talking about wasting good fuel and energy.
I run the house at 15 degrees celsius and three rooms even cooler and have my 4 systems in one room where I spend most of my free time, nice and toasty because of the heat from my computers.
I am retired and work (have fun) a good chunk of the day trying to keep up with what happens in science by reading books and articles on the internet. It is an impossible task that can never be completed but all the fun is in the reading, thinking about what I read and reading some more. In other words it is the trip that provides all the satisfaction and I never worry about reaching the destination wherever that may be.
As to the energy for my computers and science research it is almost never wasted. We have to heat the house almost 10 months of the year and the waste heat from the 2 PCs and 2 Macs just help to keep the house warm. Otherwise the 65% efficient natural gas furnace including an enormous electrical fan would come on just a little sooner. One could then argue that the furnace takes more energy, to provide the same warmth that the waste heat of the computers provides. I would not be surprised if the furnace is more expensive.
The only difference is the price of the two forms of energy, gas versus electrical, and that difference is really down to peanuts.
One could still argue about the 2 summer months but then I am usually gone for a few weeks on vacation which always costs way more energy and I shut my systems down when I am gone.
I run the systems at 100 % cpu consumption 24/7 and shut down everything that is not needed such as screens printers etc.
In short there is really nothing to worry about at all. So, go ahead and run your computers and help do the science research.
If you live in a hot climate and have to run a cooling system place the PCs in the coolest place in the house is all I can suggest. I have no experience of living in a hot climate so I best shut up.

I'm a licensed electrician, so I'm definitely not trying to talk you out of PV, but unless your walls and ceilings are insulated to R40+ and R60+ respectively, with a pressure test showing less than 25 square inches of effective leakage area, the initial investment would be better-spent (saving more fuel than the same $$ in PV panels would replace) meeting those criteria, first. Just lowering the effective leakage area from the typical US home's 165 square inches, down to 25 square inches, should save you 30 million heating/cooling BTU per year, and the extra insulation (over the DoE's R13 and R30 recommended minimums) reduces the heating/cooling load another 10 million BTU (40 MBTU is close to 12,000 KW-hours, or over 42,000 megajoules). Of course, a house that tight means putting in a heat exchanger for mechanical ventilation.

Solar hot water is more cost-effective than PV, too, especially if the SHW is used to supplement space heating, as well... the natural gas not used for water/space heating could make electricity cheaper, instead.

Anyway... PV panels are down to $3/watt now, but count on another $3/watt for inverter+installation. So, $30,000 for a professionally designed/installed 5-kilowatt system is not out of line. That's cheaper than the $8/watt estimate I saw recently for building a new nuclear plant... not enough to make up the 20% versus 99% generation uptime, but the cost per watt installed doesn't include the 50%+ voltage drop losses incurred during hundreds of miles of transmission from utility plants, either (voltage drop waste is included in the "Rejected Energy" gray 'pipes' in these charts: https://flowcharts.llnl.gov/ ). I believe the current tax credits would take about $9,000 off that 5KW system, by the way... still, even if you spent $20,000 on insulation and sealing, saving 1000 KW-hours per month beats the 750 KW-hours/month that 5KW PV system would average getting 5-hours/day of sun through the year.

I'm not shilling for HomePower magazine, but I do think it's a good resource for DIY'ers...
e.g. http://homepower.com/basics/started/ and
http://homepower.com/basics/solar/
are the first 2 links under The Basics drop down menu there.

If this post is too far off topic, go ahead and delete it.

You guys really have an amazingly positive attitude towards science and the work of the World Community Grid! We really appreciate this. We are particularly impressed by all the retirees who dedicate their golden years to learn more about science - you are the best! Thanks for crunching CEP2.

Best wishes from

Your Harvard CEP team