Thermoelectric generator improvements

[fepps]

Hi All! This is my first post here. I might have an unusual background for a site such as this-- I am a bookseller by profession. But in the 90s I got involved in alternative energy research of the more far-out variety, and eventually went on to work for a couple of companies as a database researcher. All this to say that I'm not a builder but a paper pusher :-)
Over the years my focus has narrowed, and gotten much more conventional. I've spent the last few years reading the patent databases in PV and Thermoelectric generators, looking for technologies that might have been lost or forgotten over the years. I found plenty. In particular, it looks to me like Seebeck generators can be made far more efficient than they are now, through use of some pretty simple means, easily developed. None of it is my invention, just what I gleaned from reading a lot of old patents. Frankly, I'm here to promote the possibility of an efficient TEG, and offer some information that might be useful to anyone who is interested in that possibility. I won't waste anyone's time going through it all, unless there is some interest. Thank you for your time!  Fred

[Bruce S]

Fred;
WELCOME ABOARD!!
You will find there are plenty of people here you are interested in the TEGs and their counterparts.
Please due post some of your findings and don't be too worried about rambling on, we do that on a regular basis 

I have two solid state fridges that I've increased their cooling abilities by merely taking the backs off replaced the single exhaust fan with side mounted squirrel fans to let the hot air out.

Cheers
Bruce

[DamonHD]

Hello!

I have some meaty Seebeck generators that I never deployed.  I was considering going big solar thermal and coverting excess heat to electricity, but in the end went PV and I am doing the reverse for a heat battery!

Rgds

Damon

[SparWeb]

And we do encourage experimentation.  So put the paper aside and pick up that screwdriver!

Whatever your preference, feel free to share ideas.

Another: Welcome to Fieldlines!

[fepps]

Hi Bruce, DamonHD, SparWeb,

Thanks for the warm welcome! I hope you won't regret giving me permission to ramble on about this and that! As to picking up a screwdriver, I've made several starts to it but never get very far. Doing things in the physical world takes a high tolerance for frustration which I never seem to manage.
OK: thermoelectric generators. So much to say. I've surveyed most of the US patent literature on TEGs, and two concepts pop up over and over again in different forms. Neither is in use in current devices, and both show the possibility of great gains in performance.
This is just a broad, general description of the ideas, without references. I can go into the details to whatever degree is wanted.
The first is to create some sort of barrier between the hot and cold side couples that blocks heat flow, but allows electrons to pass freely. This violates the Weidemann-Franz law that says that thermal phonon and electron conduction always go together--one of the main reasons for the low efficiency of TEG. Different patents use capacitive coupling through a dielectric, point contacts, and sparks and arcs, both ionized and vacuum, for this purpose. If the heat flow is blocked then figure of merit Z of the couple increases dramatically. Theoretically, if there were no heat flow at all, you could put a TEG in the wall of a thermos containing hot liquid, and get power until the internal heat dissipated through the walls. This is an exaggeration, but it shows the importance of finding a way to do this. Between the different techniques, capacitive coupling through a dielectric seems the most promising, with the disadvantage that it can't be used with the standard ceramic TEG units because of the way they are put together. One patent simply places a high K dielectric between the ceramic blocks.
The second concept may be more directly usable since it doesn't involve changing the structure of the modules. One of the other major causes of inefficiency in TEG is the high current, low voltage output. This maximizes the I2R losses inside the couple--most of the current produced by heat turns right back into heat. A German inventor solved this issue by using switches to include TEG modules in an AC circuit containing a reactance. The module current doesn't reverse direction of course, but the switches align the module current with the already flowing AC at the zero crossings. Putting a reactance in the circuit and choosing the right operating frequency lowers current and raises voltage through the modules, resulting in increased efficiency. The AC generator is wattless because the voltage rise from the TEG compensates for the voltage drop in the generator.
These two concepts are synergistic in that a capacitive-coupled heat barrier only works if current is changing through the modules, which is the case when the second concept is used. The German inventor does use both concepts with a spark gap for the thermal barrier, and reports very high efficiencies. Interestingly, in his very long patent he also applies the same concept to PV and many other direct energy converters. He states 30% increase in output power for the PV over the MPP. But that's another topic.
I hope you find this of interest!
Fred

[Bruce S]

Fred;
Interesting write-up.
Do you have the ability to let us know the German inventor's name or the paper?
My German is horrible, but getting better. I certainly would be interested in reading his patent or findings.

Many Thanks
Bruce S

[fepps]

Hi Bruce,

Of course! I was going to get into a more detailed outline of the patent later on. I think it probably reads better in German, as the English is a bit tortured. Here's the US patent.

The diagram on pg. 1 and pg. 11-13 of the text are most relevant to the operating principles.
Pg. 14 has an interesting example using a solar panel.

Reinhard Dahlberg, the inventor, was a distinguished pioneer in both solar and Hydrogen power, had more than 80 patents, and once testified before the US Congress on Hydrogen power. He worked at Telefunken, and was later the head of the Satellite solar division at Siemens.

I'll dig up a link to the German patent.

Fred

[fepps]

Hi Bruce,

Attached is the German patent.

Fred

[Bruce S]

Thanks ;
I read the English version. There were some "HUH" parts in there.
I'll read up on the German version to see if those are still in there.
It'll take me a bit, might have to get Sister-in-law to help with translation.

While I'm going over it, I'll keep an eye on the information here too.

Cheers
Bruce S

[fepps]

Hi Bruce,

I'm astonished and gratified you got through the patent so quickly! I quickly recognized the possibilities, but it still took me a couple of months to understand it well enough to explain it to others.

The patent by Esser, attached, shows another way to thermally isolate the TEG elements. It seems much more convenient than Dahlberg's sparks and arcs. I tried to contact Esser but didn't get a response.

It's interesting to note that *even putting a battery* between the two thermo-elements will thermally isolate them (pg. 4, second para.). 

When/if you're interested I have other related material.

Fred

[Bruce S]

Still reading both the German version (my German is worse than I thought) and the Esser PDF.
I have a few things about the assumptions being made in the PDF.
Firstly the claim about making them too much more expensive by using nano-scale diodes is possibly incorrect or was merely based on the early 2000s manufacturing costs. I would state that the use of diodes as a heat flyback barrier isn't correct either since heat isn't necessarily voltage/current dependent.
Using a dielectrics as an insulator however, is interesting.
I'm still reading and may look to see if there's any pier review available on this, either I've not had enough caffeine this morning or something just doesn't read right in the PDF.
I'll be back.
Bruce S
 

[fepps]

Hi Bruce,

I can't find in the patent where he refers to nano-diodes as being too expensive. Certainly if you're talking about the LED examples, they were lab curiosities at that time and are everywhere now. If you're referring to nanomaterials for thermoelectric modules, they've been in development for a long time but none are in commercial production.
There's no mention in any current literature for this patent. You will find references to Dahlberg though--he was well known at the time. There are however some articles that refer to pulsed, resonant, or non-stationary thermoelectric operation. One is attached. This article refers on pg. 6 to the Strachan-Aspden device, which was filmed and demonstrated before experts. I've seen the film and I'm trying to find an article in my vast piles of data now. It used both of the two concepts, consisting of thin metal thermo-elements layered with dielectric plastic sheet in between, driven by AC. It also used PZ stimulation and possibly the Nernst effect. However the inventors didn't fully understand or explain their invention, and Aspden later tied it to his goofy overunity theories, thereby invalidating it.
Fred

[Mary B]

Spark method would require a LOT of shielding making it prohibitively expensive. Regulatory bodies would take a dim view of the RFI generated by it.

[fepps]

Hi Mary,

Yes, I agree. Aside from the RFI, they are finicky and not suitable for a lot of applications. I like the idea of using dielectrics as others have done. Other patents use a very small vacuum gap, or a gap filled with low pressure inert gas.

Fred

[fepps]

Hi Bruce, all,

I've been thinking about what practical experiment could be performed to test out the hypothesis that using AC and switching, along with thermal blocking between the TE elements, will result in higher efficiency.
The first consideration is that standard modules have p and n junctions highly integrated. You can break the little bits of BiTe out of them, but then they are too small to work with. A reasonable sized ingot of BiTe costs about 300 dollars, and you would need one for p and one for n-- and they are almost impossible to work with, tending to fracture along the exact wrong directions. I worked up a couple of designs using metals rather than ceramic but the Seebeck coefficient is just too small.
However CuO is an interesting material for these uses. CuO produced by subjecting a Copper surface to an oxidizing flame has a Seebeck coefficient of 204 uV/K, while BiTe has a coefficient of 160-170 for p and n respectively. I got tipped to CuO from an interesting YouTube video by Nyle Steiner, which you can view by typing in his name and Copper Oxide.
A patent search for 'CuO thermoelectric generator' turned up a fascinating patent with somewhat anomalous characteristics. I've decided that if testing is done, this device should be tested first because it is a game changer, even without fancy electronics to boost voltage, or capacitive layers and such.
The patent is attached. Briefly, the device is a small can of Copper or Tungsten containing CuO created by subjecting Copper wire to a propane torch, with a lead put through the middle of the CuO, and a lead on the can itself. The inventor says that at cherry red temperature (815 C) the output was 4 mA at .2 V, or 80 mW. Now this doesn't seem too significant, but:
1) The can is very small, ideally a 1/4 in. long and 1/8 in. wide. If you do the power density calculations for a device that size and output, you get numbers for a m3 that are too astronomical to print, even if you allow half the space for leads and heat sinks.
2) The whole device can be subjected *to the same temperature*. There is no temperature differential or heat sink. This is thermodynamically suspect right from the start, and implies that something other than the Seebeck effect is operating.
If it is not a pipe dream. I would have left it behind at that point but for the discovery that CuO has such a high Seebeck coefficient. Also, CuO rectifiers are well known, and there are a class of devices that purport to rectify random thermal vibrations with diodes. So I wonder if it is possible that he has done this accidentally on a macro scale? (the other projects use nanodiodes). Even if the claims of the inventor are fallacious, given that CuO is an abundant, easy to make material, why hasn't more been done with it?

Fred

[Bruce S]

This statement "I've been thinking about what practical experiment could be performed to test out the hypothesis that using AC and switching "
TEG's are  DC components adding to heat to one side and keeping the other side cool, will output DC not AC.
Using AC to power a Peltier device will not work correctly either due to the swing nature of AC.

IF you are possibly looking into making the DC into AC via switching then there's a whole new ball of wax.

Bruce S

[joestue]

class of devices that purport to rectify random thermal vibrations with diodes..,


No

[fepps]

Hi Bruce,

Yes, the latter. as the Dalhberg patent suggests. To be clear, he proposes starting with an AC oscillator and then switching the TEG modules so that they are always aligned with the AC current. The effect of this is to allow control of the I/V characteristics of the module that isn't possible otherwise. In general with these DC energy converters--TEG, PV, and other things mentioned in the patent-- you're stuck with whatever these characteristics naturally are. You can transform the voltage outside of the unit, but not within it, where the reduced current can reduce joule losses.

Note that conversion to high voltage is commonly used to reduce losses in power transmission lines. There's nothing unusual about doing this. I take it that you haven't found any basic flaw in the patent yet?

I've come to the conclusion that it's nearly impossible to test BOTH the 'dielectric heat barrier' and 'switched AC' concept at the same time with current modules. However it would be relative straightforward to test the switched AC.

Fred

[fepps]

Hi Joestue,

Most of these projects are patently incredible, but the one attached gives me pause.
It has enough details so it could be replicated. You could see for yourself.

Fred