Mercury that is released as a vapor

Food for thought: What physical property do you believe makes it a vapor? Does that necessary physical property exist outside the bulb?

Now we have Solar thermal generation. This is pretty good tech, but it use a LOT of water.

There's no intrinsic need for that - it just makes the cooling system smaller, cheaper, and more effective (these are heat engines, operating off a temperature differential - the colder the cold side is, the more efficiently they run).

solar PV does not work at night

Concentrating Solar Thermal does (via heat stored in molten salt vats). However it requires specular, not diffuse light (i.e. does not work at all in overcast weather), so it's more suited to desert environments (for example, a large chunk of the southwest US). Efficiencies are also better than with PV (30-40% instead of 10-15%), it's straightforward to scale up, and the technology is well understood (you focus sunlight on a target to make it hot, generate steam with it, and turn a steam turbine with that steam).

something tells me that carbon fiber disks that are carefully stabilized and levitated in a vacuum while spinning incredibly fast...would break into a thousand pieces the second they left containment rather than rolling down the street and through someone's house

Yes, it's basically the same as a bomb with a chemical energy potential equivalent to that of the kinetic energy stored in the flywheel being detonated in the flywheel's containment. If they haven't designed the facility with this in mind, shrapnel from one disintegrated flywheel could hit another flywheel, leading to a chain reaction of disintegrating flywheels that would shut down most or all of the facility. It sounds like they've put the flywheels mostly underground which, with enough spacing, should prevent this.

As for energy levels, if the 1-megawatt-for-15-minutes number in the article is for one flywheel, then that's equivalent to 215kg of TNT. If that's for all 200 flywheels combined, then one flywheel disintegrating is equivalent to 11kg of TNT.

If more than 50% of the voters wanted the patriot act gone...

You need a higher percentage than this. Let's pretend that:

• A representative lives in a semi-contested district where they've, regardless, served for several terms.
• The issue in question is a big one - one that voters may make their next decision on
• The issue isn't one that's aligned with either the dem or rep platform.
• A change on the issue might anger voters who disagree more than it would sell voters who agree.

If only 50% of voters wanted the bill to pass, then about half of the voters who would ordinarily vote for the candidate may vote against them in the next term. While it's often true that the last bullet is just a figment of the representative's imagination, and that half the voters who wouldn't ordinarily vote for the candidate may vote for them now, if a representative has found a platform that gets them elected over and over, they won't want to risk changing it and fracturing their constituent base.

My suspicion is that you're going to wind up reimplementing a good fraction of a CNC machine's functionality, but here's an idea that might save you some time:

Implement the 1.5" increments via a ratchet-like mechanism. Allow the tube to slide down the V of some angle iron placed at a steep (say, 60 degree) angle. The drills are placed halfway down this. In the lower half of the V, there are registration pins (probably bolt heads of the appropriate size, machined to the proper diameter and with some taper for self-centering) every 1.5" for a couple repetitions. The tube is placed into the V, hits the first registration pins, is clamped to the V using electromagnets, and drilled from both directions (you have several options for moving the two drills here, but I kinda like the idea of pushing the drill away from the work with a spring and pushing it towards the work with pneumatic bellows). Once the drilling is done, some electromagnets above the tube pulse for a bit so that the tube pops up and slides down onto the next set of registration pins, and you repeat. The tube falls out of the V when it clears the last set of registration pins. Be generous with the hardware interlocking - at the very least you want to make sure one drill is out before the other drill goes in.

plus one single RPM means that you can really simplify the design of the engine so that a minimal amount of cooling is required.

No, it doesn't mean that at all. This is still a heat engine - it converts a temperature difference (between a hot side and a cold side) into mechanical energy. The smaller that difference, the less heat energy is available to extract and the less efficient the engine can be.

I hear the complaint "teachers will teach to the exam" all the time as an argument against standardized testing. Damn right they will. If this results in a poor education, it means they weren't good exams (e.g., the SAT)

I'd rather have a nation's graduates receive a wide variety of teaching experiences so that they are collectively exposed to millions and millions of different facts and concepts than have them collectively exposed to the same few thousand facts and concepts. That is something that a standardized test by its very nature simply cannot do.

Aircraft can be flown indefinitely and safely as long as they are maintained correctly

Actually, no, they can't, at least not ones made out of aluminum. Aluminum has no fatigue limit, which means that no matter how little you stress it or how much you overengineer it, it will always fail after a certain number of cycles. For this reason, they regularly x-ray large commercial aircraft as they age to look for hidden stress fractures. They want to prevent future accidents like the Comet breakups, Aloha Flight 243, China Airlines Flight 611, numerous firefighting aircraft including a C130, numerous UH-1 Hueys, etc.

Look into the efficiency of a battery sometime. Unless you buy really expensive ones you lose about half of the energy putting it into and getting it back out.

I have looked into it. There aren't many numbers on the Internet - most of the stats on Wikipedia are single-sourced from sites such as powerstream.com, but coulometric efficiency generally varies by cell chemistry type. Lead acid is about 70%, NiCd is about 83%, NiMH is about 66%, Lithium-Ion is 80-90% (and there are some sources that say it's higher than that). For large-scale grid energy storage you might look into Sodium Sulfur, which is reported to have efficiencies of 89-92% and is inexpensive but requires operating temperatures of 300-350C to work. In any case, 50% is very very pessimistic.

I still think that the solar power tower molten salt vat approach to energy storage is the most practical storage option (where hydroelectric pumping isn't available, anyway) in the short term.

Does the angle of the sun come into play here? Is it really the same at 4PM as at noon? Or is 1000 W an average?

A number like 1000W would refer to the peak power output that you'll get from it with the solar cells perpendicular to the sunlight with optimally clear skies. Since the earth receives a maximum of about 1100W/m^2 of solar energy, and ordinary silicon cells are about 12% efficient, you can expect such a system to be a little less than 10 m^2 in size.

4pm is not noon, no. First off, the lower the angle of the sun in the sky, the more atmosphere it goes through, which filters things out somewhat. Second, while you can steer the solar panels so that they are always perpendicular to the sky, most are just fixed to a south-sloped roof and more of their surface area will be going to waste.

If you google for insolation map you can get nifty maps of what areas get how much sunlight. Note that most of these maps are for plain photovoltaic installations, where diffuse light (cloudy skies) is still better than nothing. When you're using solar concentrators (mirrors), those mirrors can't focus anything but specular light (sunny skies). I don't know if any maps have been drawn taking that into account.

In order to do the same job, you would need a massive (read expensive) heat exchanger to do the same job.

And in an environment where you can assume condensation and ice buildup, rendering forced-air circulation useless, that heat exchanger can be drastically simplified - it's just a very long tube. No fins, no fans, just pumps to cover the increased pressure drop, and enough supporting structure to prevent the ice from taking it down.

Not everywhere. For air source heat pumps, the COP* (what you meant when you said efficiency) is about 1.0 at 0 degrees F

I know about COP, I was just keeping it simple for the original poster to allow them to compare to resistive heating.

The useful temperature range of a heat pump is mostly governed by your choice of refrigerant and your cold side and hot side pressures. For example, the freezer in your kitchen is designed specifically to move heat from 0 degrees F air, and it does so efficiently. The theoretical maximum COP of a heat pump with a cold side of -10C (263K) and a hot side of 40C (313K) would be 1/(1-(263/313)), or about 6.26, and it's not uncommon for ordinary consumer appliances to get at least half that. Icing simply limits the rate of heat conduction into the cold side, meaning that you need more surface area to compensate.