As someone who has worked in the nuclear power industry for more than 15 years, let me input my own pedantry:

A RTG is clearly nuclear power. The heat produced is due to a release of binding energy associated with the strong nuclear force.

A RTG is clearly NOT a nuclear reactor. A reactor is a device designed to allow things to react, and to react them in a controlled fashon. The decay of PU-239 cannot be controlled, and no miracle of geometry known to man can effect the rate of decay. A reactor implies control or regulation of the process, but in this case no control is possible.

Also, alpha decay is technically fission, but no one calls it that because it is very common and would provide no useful disctinction when you are trying to describe the process of creating a self sustaining nuclear reaction. If you walk up to a nuclear physicist and start talking about fission when you mean an alpha decay, most will regard you an an uneducated imbicle. When those in the nuclear physics world say something is fissionable, that means it can break into other things besides an alpha (AKA a helium-4 nucleus) and the resultant daughter. When those in the nuclear physics world want to describe something that spits out an alpha, they say that it alpha decays. There are many radionuclides out there that will undergo alpha decay, but which are not classified as fissionable.

This is false. Power balancing and power factor are both vector sums, and since different customers will have different phase imbalances and power factors, those will be averaged out at generation. Harmonics are generally additive, just like real load, except that most of those harmonics will be absorbed by the transformers and signal traps between the load and the generation point. As such, some of the cleanest power you will find on the grid is at the point of generation.

This is complete gobbeldy-gook. These words in this combination mean effectively nothing. Unpacking what you are trying to say would be an effort in futility.

The losses you described are lost in the form of heat, sound, or vibration. They are lost in the loads, along the power lines, and in the transformers, and yes, in the generator, though the majority of the losses will happen closest to where the transient deviation from clean power occurs. This is almost entirely at the load side, but some other noise induced by switching (opening and closing breakers on the transmission and distribution side) or lightning strikes is also possible.

The thing that makes me really sad about this whole situation, is that your technology sounds really cool. It sounds incredibly geeky, and in exactly the kind of way I like. I even think that there is a lot of potential with the technology. My problem is that the words you are using to describe it are nebulous. "Electricity loss" is imprecise. There are dosens of different kinds of electrical losses. It sounds like your technology (assuming it works as well as described) does a really good job of dealing with a lot of those losses, and that there would be a large number of other benefits besides just energy efficiency that would make this really marketable. In the time since reading the article, I've also visited one of the electrical engineering forums where you attempted to defend this technology. I'm incredibly disappointed that you aren't, or don't seem to be capable of, speaking in the sorts of technical terms that those with a science or engineering background need to use to properly assess your technology. I would love to know more about what you are doing, because it's exactly the sort of thing I geek out on, but you don't seem to have a sufficient grasp of the language necessary to explain it in detail.

The author was kind enough to respond to me, and I have responded in turn. Your reply ( I assume you are from 3DFS ) doesn't materially change my response.
Here is the exchange:

That is the claim that is drawing the most fire, but it is an actual claim 3DFS makes, not a mistake, so it's worth understanding.
They are, believe it or not, aware that there are carnot-related losses in electricity generation. Their claim is that the 66% losses DOE claims under that banner are actually a combination of carnot and electricity losses -- that the lost quads DOE puts under there are actually spread out over the grid. They know they cannot eliminate carnot losses. But (they say) they can eliminate 98% of electrical losses. And they say that when all electrical losses are prevented, the actual carnot losses will be revealed as considerably smaller than DOE estimates.

Again: the claim is that carnot losses are exaggerated in the sankey diagram, not that they don't exist, not that they can be eliminated.
That claim may or may not hold up, but it's not the ridiculous claim that so many people on Twitter are busy rebutting.

Hope that helps clarify!

--- And my reply below---

Thanks for the reply, but I'm not buying it (and I know you're just the messenger, so don't take this personally).

While the measuring devices are "analog" in this system, the DOE estimates for conversion losses can literally be taken from the data I indicated earlier (1675 MWt to 570MWe for my plant, or roughly 34% efficient). It's incredibly easy data to gather, and it would be really hard to exaggerate. The only place where there are electrical losses in this system (between the mechanical input from the turbine and the measuring devices on the generator output) is the electrical generator and some buswork. This measurement is before it goes out on the grid, so there is literally no way that this number can be "spread out over the grid". I can also assure you that most of the electrical losses in the generator come from the fact that it is pushing out 20,000 amps. If those I^2*R losses don't account for more than 95% of the electrical losses I would be incredulous - and there is nothing a system like theirs could do to significantly reduce those losses - certainly not by 98%. I would be really impressed if they could pull off 5%... It's not like they're turning the copper in the generator into a superconductor.

Taken from another direction, my plant has a maximum theoretical Carnot Efficiency of about 46% in the summer. Laws of thermodynamics prevent anything higher than that. That 12% difference between theoretical maximum and actual has to account for heated water being discharged constantly to maintain chemistry parameters, friction losses in piping, windage and end-tip losses in the turbine, bearing losses for the turbine and generator, imperfect heat transfer between heat exchangers, thermal losses from steam and feedwater piping, steam leaks, throttling losses, and other ideal vs real turbine losses (made much more significant due to a lack of superheating for the high pressure turbines). That is all in addition to any electrical losses.

There isn't much left to get back from the generator, and no amount of phase balancing and harmonics corrections can fix anything before the generator, especially after you consider the hundreds of tons of rotating mass that would buffer those systems from any harmonics. 3DFS might be doing something really cool, but it sure as heck isn't reducing the conversion losses (as a percentage) on the DOE graph. Then again, my entire argument assumes that the DOE graph is what it says it is. Maybe the DOE is just incompetent at gathering and compiling data. I'll hold out hope that this is merely miscommunication, but what you are describing sounds more like marketing than actual technical data.

Thanks for listening to my rant. :)

Here is an E-mail I just sent to the author:

I just read your article on 3DFS technology. I have been in power generation for 17 years, and I think I can explain the problem you encountered with the professor of electrical engineering who hung up on you. My bullshit detector was going off pretty hard due to a single claim in the article, which you repeated a few times. After reading the full article, I think the technology is entirely possible, and even plausible. The problem is that the conversion losses in the DOE graph are not the same conversion losses that 3DFS is referring to. The DOE conversion losses are looking at the thermal efficiency of the generation being turned in to electricity. For example, a single unit at my nuclear plant operates at about 1675 MW thermal output, but the electric generator only puts out about 570 MW of electricity, of which about another 30 MW is used by the station to power pumps, fans, and other loads. While having a grid full of 3DFS equipment hooked up might increase that efficiency, it would be miniscule. 3DFS’ technology really effects everything from “net generation of electricity 13.70” to the right. The conversion losses 3DFS would be improving is almost entirely on the load side. I cannot rightly say that it will make an average wall-wart AC to DC converter go from 30% efficient to 95% efficient without a much more in-depth understanding, but there is much more room for those kind of gains within the laws of physics than the conversion losses in the DOE graph. Those are limited by carnot efficiency, which applies to all heat engines, like the rankine cycle or brayton cycle. If you correct that deficiency, I suspect that you will get much less negative feedback on the article. I hope this helps!

I agree entirely. I would love to see them try something like this at a US plant. Since 9-11 nuke plants have gotten REALLY serious about security in the US. (thanks in large part to mandates from the NRC) This also makes a nuclear plant about the safest place to be in a zombie apocalypse.

You are wrong on the forever part. The amount of heat produced drops off fairly rapidly. By four days out, you don't need to pump water past it... it just needs to sit in water. By 10 years out, it can be stored in shielded casks cooled by air. You wouldn't want to store spent fuel in the open air due to the possible spread of contamination. Also, that water provides a significant amount of shielding from radiation. Even at ten years out you would get a ridiculous amount of exposure standing next to unshielded spent fuel.

I totally want to know where this hotel is so I can stay there (more accurately be denied the ability to stay there) and sue them for all they're worth. Marital status is a protected class by federal law. I would totally PWN them in court. Seriously, hook me up with an address.

My favorite server names ever. My best friend and I used them to VPN our LAN's together