[SPLIT]Older LENR Experiments were bad, good... in general

Thanks, THH.

Quote from THHuxley

Kirk claims (as I understand):

(1) A condition-variable process at the electrode can cause recombination at the electrode/liquid surface. Whether this happens or no is complex and due to special activation of the metal surface (Recombination Active Environment). He appropriates all of the NAE LENR work: he is agreeing that there is such a complex effect but disagreeing as to what it does.

Yes. I think that his claim may be under-recognized, not understood.

(2) The more recombination at the electrode, the less recombination at the recombiner. This means a larger heat source in the electrolyte and a smaller heat source in the air. This further means the air temperature goes down relative to the electrolyte temperature (a little). This then means that the total heat loss from the cell for given total power released (which is what determines the effective calibation constant) goes down. Hence an apparent over-reading of heat.

You are I think disagreeing with the argument (2) - but it seems pretty water-tight to me. Further Kirk's anecdotal comments (which I'm sure could be checked with Ed) show that for at least one cell this effect is likely to be significant even at relatively small amounts of electrode recombination.[/quote]
Actually, I am looking at, studying, argument 2. The calorimeter in question shows 98% overall capture. To be completely sure about various aspects of this, I'd want to discuss the experiment with Storms. There are details that could be at least marginally important that are not explained in the paper, or not completely explained.

We are toaking about a shift in heat capture from a shift from heat generated in the recombiner (the normal condition) to heat generated by recombination at the cathode, which is a phenomenon not known to occur at the current levels involved. Setting that aside, the idea is as THH stated, and that explanation requires that heat from the electrolyte be more fully transferred to the coolant than heat from the recombiner, in the air above.

The inefficiency of the calorimeter is very likely due mostly to heat loss through the top of the calorimeter enclosure, which is insulated with foam rather than the highly-effective vacuum of a Dewar flask.

The question I have, right off, is how the calirbration was done that determined "98% efficiency +/- 0.5%." There were two calibrations: with Joule heating, i.e., a resistor, and "electroytic," The resistor was an "exposed Pt wire heater." Exposed where? If this is in the electrolyte, could there be electrolytic current between one end of the wire and the other? If an exposed wire were to be used like that, I'd think it would be in the head space. I would prefer to see sealed heating elements, in various locations, to nail the calibration of this. Reading between the lines, I think the wire resistor was in the head space, not in the electrolyte. Then there was electrolytic calibration, using a presumably inactive platinum cathode.

Basically, what we would like to know is the efficiency for heat sources in various locations in the cell.

If my analysis of the cell arrangement is correct, calibration actually rules out what Shanahan is claiming, by measuring the difference between heat in the electrolyte and heat in the head space. However, again, what kind of effect would we expect?

The electrolyte is in intimate contact with the cooling jacket, the inner wall of which is the cell wall. It is not in contact with the lucite cell lid. The recombiner is in the headspace, in contact with the cell wall, the electrolyte below, and the lucite lid. I would expect more heat escape from the recombiner than from heat generated in the electrolyte.

The lucit lid then heats what appears to be an air space inside the Dewar. That air is in contact with the cooling jacked, but also with the foam lid. Some portion of the heat in the cell lid, then, will escape.

Looking at this particular design, then, I come to the position that Shanahan is correct that a shift in heat location position from recombination heat would create an appearance of XE, at some level. The issue becomes "how much."

Storms has:

Quote

Good agreement between the electrolytic- and Joule-based calibrations shows that the location of heat production does not affect the accuracy of the device.

Storms claims that the electrolytic and Joule-based calibrations agree to within 1.2%.

Electrolytic: W = - 0.18317 + 7.1221e-2deltaJ*F
Joule: W = - 0.23893 + 7.2107e-2deltaJ*F

There was a "final electrolytic" calibration done after the experiment:
W = - 0.14405 + 7.0892e-2deltaJ*F

At this point I will notice that the Joule calibration shows watts per degree than the electrolytic calibration, thus experimentally confirming (with one data point) that if calibration is done with recombination at the recombiner, in the head space with the Joule heater -- if I got that right! -- and then this calibration is in use when heat instead appears in the electrolyte, there would be some appearance of XP. This would be a systematic error, we can imagine, affecting all flow calorimeters of similar design.

How much it affects the results is a separate issue. "Impossible" is a common pseudoskeptical claim, when a more neutral approach would be "how much?" Cold fusion was considered impossible, but ... in fact, the predictions were of a very low rate (*very* low), but they then depended on approximations that might not hold under all conditions....

"Impossible" shuts off conversation. So I'll agree that response to Shanahan fell into dysfunctional discussion, such that the points may have been missed, benefiting nobody but allowing some to be happy that they were right.

The input power in a device like this is distributed between Joule heating of the electrolyte and chemical potential energy in the evolved hydrogen and oxygen. The idea of recombination is to return this energy to the cell, leaving only the Joule heating -- and any excess power. Otherwise with open cells, the gases that leave are measured and the lost energy added in. If they are only estimated from electrolytic current, this would be the most vulnerable approach to a recombination artifact. There is also the issue of absorbed deuterium in the cathode to be considered.

The unexpected recombination is from what percentage of the evolved gas? The mechanics of the cell could not allow much. The calibration shift would then be only for a small percentage of the input power. Not the entire input power, by far.

I am still not seeing a major effect.

This experiment was never published under review, it was only a conference paper. I'm not aware of critical response to it. That, unfortunately, is missing for many LENR papers. This was an anomalous result, heat from a platinum cathode and nothing major in the field is standing on it.

Quote from JedRothwell

A variety of blanks tests are done, such as a heater in ordinary water, a heater in heavy water, or Pt in light or heavy water. Only Pd-D produces anomalous heat. (Ed Storms once claimed that Pt-D worked in one instance, but I do not know of other observations of this. If it is replicated it may change the assumptions made in these experiments.)

The experiment being discussed, because Shanahan apparently developed his CCS theory, in part, by studying it, is precisely that single report from Storms.

The fact is, though, that even a palladium cathode will normally not produce heat, and can be used for a "dead cathode" calibration.

Many cold fusion papers tend to leave me with "Huh? What happened? What am I supposed to see in that mass of noise?"

Added. The plot mentioned for SRI P13/P14 is on page 2 of the report given to the DoE in 2004: http://lenr-canr.org/acrobat/Hagelsteinnewphysica.pdf

This is not an impressive level of heat, the peak is about 500 mW. That lasted for 84 hours. If Shanahan wants to explain this with a mysterious recombination effect, he'll need to look at the specific calorimetry. The efficiency of the flow heat capture was 99%. The cell architecture has both the electrolyte and the recombiner relatively far from the coolant outlet, so a positional difference would be suppressed.

Of late, Storms reports having found that active cathodes can be removed from the cell, cleaned with nitric acid, stored, and returned and when loaded immediately show the effect. If that is confirmed, it opens up new vistas.
However, there is a stunning exception: SRI P13/P14. There is a nice plot showing a current excursion, through two otherwise identical cells, one hydrogen, one deuterium, they are electrically in series, so there is the same current through both. The same kind of calorimetry, flow.

The hydrogen cell shows an increase in noise with increased current. The deuterium cell, however, clearly shows calculated XE, significant, more or less tracking the input current (which is a common effect, probably not actually central, recent work from Storms shows what is effectively "heat after death," explained below.)

What that famous plot does not show is that this was the third such current excursion, if I have it right (this is in the full report from EPRI). The first two, no XE. Unfortunately, McKubre didn't plot those "dull boring" results. However, they show that the calorimetry was working, with a "dead cathode" calibration. Hydrogen behavior and deuterium behavior were the same. An anomaly appeared, and strongly. I call it the "chimera of cold fusion." It licked him in the face. After this, McKubre likely harbored few doubts about the reality of the effect, so clearly and immediately seen. Yet it also shows the mystery: what had changed?

Everything was supposedly the same. But ... the material shifts from the stresses of loading and deloading. Something is not being controlled.

The recent Storms work sets up a more or less standard FPHE. However, he has a heater in the cell, used to raise the cell temperature. This is then used to maintain constant temperature, i.e., if there is XE, the heating is backed off. He was seeing XE.

Storms then turned off the electrolysis current. (presumably he increased the heater power to compensate for that known reduction.) XE continued without any electrolysis current. Loading was declining, no reduction in power.

This flies in the face of much expectation.

Higher temperature is generally believed to increase the FPHE. So this experiment, using cathodes extensively conditioned by Storms, would be expected to possibly show XE. However, it was not expected that taking the current density to zero would not cause XE to fall off. As well, when current density is reduced, loading declines. The effect has been well-correlated with loading ratio. Again, reducing the loading did not reduce XE.

I don't think anyone has looked at these experiments this way. I've been suggesting supplemental heat for years, and using "heat replacement" as a check on the calorimetry.

I would have preferred to see the experiment continued. As I recall, when Storms lowered the temperature, the XE then rapidly fell.

It is easy to imagine "some calorimetry artifact" here. But Storms is highly competent.

Highly interesting, unconfirmed. The story of cold fusion. But this approach ought to work with more or less any protocol. It's easy to add, it's just some supplemental heat. Input heating power can be precisely measured.

The result has a theoretical basis. Storms posits that the reaction does not take place in the bulk. His old platinum work is probably an element in his thinking on that. It takes place in structures that form on the surface. What high loading does, in his idea, is to create stress in the palladium, which leads to cracking, which, at a specific size, creates Nuclear Active Environment. But the actual reaction can take place, once the conditions are set up, at much lower loading.

Shanahan posits that "Heat after Death" is from recombination. Old idea. However, there is a problem. When the electrolytic current stops, the cathode deloads, evolving deuterium. There was only a limited amount of oxygen in an open cell, likewise in a closed one. (orphaned oxygen has been vented). So this oxygen could reach the cathode, which is now exposed, butit will be quickly used up, and the evolving deuterium gas, in an open cell, will exclude atmospheric oxygen from entering.

Basically, recombination could explain a small, immediate HAD effect. Not a large or extended one.

Quote from kirkshanahan

"if the heat is shifted away from the water exit," - this is not the relevant parameter. The extent of contact of the water flow tube to the cell surface is the relvant parameter. The thing you should be concerned with is moving the heat out of the low detection efficiency zone into the high detection efficiency zone.

This could be a misleading presentation of the idea. The image presented is of a "low efficiency detection zone" and a "high efficiency detection zone."

In reality, there are differences in efficiency depending on location, apparently, from looking at Storms' data. They are quite small. By they are large ("low" vs. "high") as large, the CCS effect from recombination is then presented as possibly large. Possibly more objectively, there is a small difference in efficiency combined with a low level of recombination at best.

Quote

The vent tube (which you call a water flow exit) in an open cell represents a material pathway for heat to flow out. If the calorimeter design does not atttempt to capture that heat, the calorimeter's overall efficiency would be lower. That probably means a greater CCS effect is possible. See Oriani's paper I have discussed here previously to observe this effect directly.

I was only talking about a flow calorimeter, with a water flow exit. The analysis of an open cell is very, very different. One of the confusions that occurs in discussion this issue is the mixing of many different experiments and experimental conditions. I suggest avoiding that.

Quote from JedRothwell

Abd Ul-Rahman Lomax wrote:

Yes, that is a bad way to do it . . . UNLESS you say the total heat must exceed I*V or it doesn't count. In other words, you assume there is complete recombination within the cell (whether there is or not) and you only count excess above that level. That is a legitimate end-run around this issue.

Sure, but I dislike it. This reminds me of the Rossi "conservative" approach. Instead of making actual measurements of heat that are not focused on evaluation (success or failure) but on accuracy, various possible factors are disregarded. How about measuring things as accurately as possible, and then, if it seems useful, in analysis, pointing out that even if X were entirely Y, results would still be highly significant.

The argument could be sound, and, obviously, if it is an onerous burden to make some measurement, it could be "legitimate" not to measure it. However, it weakens the results. How does one calculate error bars that are sensible?

Quote

In some cases excess heat far exceeds the limits of I*V, making recombination a non-issue. In other cases there is no electrolysis input power for extended periods of heat after death, which again makes the whole issue moot.

Well, Shanahan still tries, by ascribing HAD to recombination, but it's not ... not plausible. However, this analysis should be done on a case-by-case basis. I don't particularly like the Pons and Fleischmann boil-off experiments, it seems just too messy. I like Ed's recent approach of heating the cell to maintain constant (high) temperature, and then turning off the electrolysis current. Not messy at all! And the cathode is still submerged, etc. If there is recombination, it should stop rapidly.

Quote

As I mentioned earlier, along the same lines, suppose you have a crappy flow calorimeter that only recovers 80% of the input power. I would toss it in the trash and start over with a new instrument, but for the sake of argument, suppose you have one. A flow calorimeter gives you an absolute power reading, not dependent on calibrations. (Which is not to say you can skip the calibrations!) Suppose you input 10 W total I*V during a calibration. You find 8 W are coming out in the flow, and 2 are leaking out who-knows-where. In this situation, I would not trust excess heat below 3 W. The total recovered from the calorimeter should be more than input. That's my standard. If the calorimeter recovers only 9 W, that might be 1 W of excess heat, but I would not believe it. That is why you need a high recovery rate.

Yes, or at least it vastly simplifies analysis. I'd still like to see zone calibration. It looks like Ed did a form of that in the platinum experiment. At least that's how I read it, and he explicitly said that the calibrations ruled out a positional variation.

In fact, those results do show a small positional variation. My sense is that looking at quantifying what some unexpected recombination effect could do to results is better than just dismissing it as impossible. It's possible, and it happens, but ... how much?

All the evidence I have seen, so far, is "not enough to have any significant effect on results." But that is rebuttable. As well, it might have some effect on some results, and could be a possible artifact to watch out for.

Overall, Jed, I also trust the calorimetry of experts. If Shanahan had managed to engage skeptical electrochemists and experts in calorimetry, it might be different. Still, as an exercise and for that "unexpected surpfise" that we learn to look for, I want to understand what Kirk is claiming. He also became reactive and so what might be reasonable and sensible, at least ab initio, gets mixed with strenuous argument defending that he's right, and, while I point this out on occasion, you know that I also point it out when cold fusion researchers do the same thing. It's human, but it is not a powerful approach,

A variation on this is "just because you are paranoid doesn't mean that they aren't out to get you." Just because you argue dysfunctionally doesn't mean you are wrong. But ... until we can set aside those defenses, we are not terribly effective in recognizing errors and moving on, and our communication skills can suck.

Quote from Wyttenbach

THHuxley wrote:

Electrolysys (by current) of H2O/D2O is well understood since far more than a century. Any chemical reaction in a liquid is based on an equilibrium, thus recombination is always accounted for.
Did anybody ever ask Kirki, what he claims is going to recombine? H2 and O2 usually shine up at different isolated places. They never meet again..at least not in the fluid!!

Kirk's CCS due to recombination is one of the most plausible systematic artifacts proposed. Yes, my understanding is that the hypothesis is inconsistent with what is known, experimentally, but this kind of a priori dismissal is not useful.

Electrochemistry is famous for complexity. Cold fusion generally is not an equilibrium effect. It is not called an anomaly for nothing. Something is not controlled in the known and confirmed work.

Kirk has proposed an anomaly, some kind of unexpected recombination. Yes, something must recombine and there is no controversy over what it would be, it would be oxygen and hydrogen recombination. Wyttenbach claims that the hydrogen and oxygen "usually shine up in different places," but in most cold fusion experiments, this is some physical separation of the cathode and anode. Sometimes they are quite close. Sometimes there is a cathode and an anode wound around outside it.

Not much oxygen and deuterium are dissolved in the electrolyte, and Shanahan does not propose such dissolution. What he proposes is that bubbles of oxygen may circulate in the electrolyte and reach the cathode. If they do reach the cathode, they could recombine with oxygen there. The same could happen at the anode with hydrogen bubbles.

Kirk asserts some evidence, particularly the "sparkles" that Szpak et al reported way back. We have gone over that evidence. I don't find it convincing at all, but some may.

The recombination issue is not a new one, it's been raised as a possible problem many times. One fact might give pause: recombination occurs as a higher percentage of the evolved gas when the electrolytic current density is low, lower than is normally used in cold fusion experiments. Why?

My sense is that with higher current, an oxygen bubble that approaches the cathode is much more likely to be caught and lifted by the hydrogen bubbles than to actually reach the cathode surface. To recombine, the bubble must merge with hydrogen and then hit the cathode surface, where the catalytic action of palladium would cause recombination.

That it occurs at low current shows that the idea of impossibility is false.

The real question is quantitative.

Actual measurement of evolved gases is done, and shows that recombination is below some level. What level? The problem is that Kirk is alleging an anomaly, somewhat like cold fusion: sometimes it happens and sometimes it doesn't, for unknown reasons.

His position appears to be that one anomaly is just as good as another. His doesn't require revising the laws of physics, though it is an error to think that cold fusion requires revising the laws of physics. It was merely unexpected. ("unknown nuclear reactions" are always unexpected!) As major recombination under the conditions of the FPHE is unexpected. Unexpected by whom? Experts. Experts in what?

And as we start to understand the debate instead of dismissing one side or the other as crazy or wrong, the way beyond it begins to become visible.

THH here, while tending to support Shanahan's ideas, at least as unproven to be false, also agrees with the ultimate resolution: experimental evidence, in work designed to test the hypothesis. Because there is certain work under way to measure the heat/helium ratio, I've asked Kirk what he would predict from that work. I have predictions, which I've given many times.

How is this relevant to the CCS hypothesis? That hypothesis posits that the anomalous heat is due to a non-nuclear process. Heat/helium, if the ratio tightens experimentally, confirms that the heat is from a nuclear process, and the ratio may, it is very possible, tell us much about what it is, though not the actual mechanism, probably. The indications are that helium production, then, will be an independent measure of heat, setting an upper bound on error like that produced by CCS or any other artifact.

Quote from Wyttenbach

Abd Ul-Rahman Lomax wrote:
However, there is a stunning exception: SRI P13/P14. There is a nice plot showing a current excursion, through two otherwise identical cells, one hydrogen, one deuterium, they are electrically in series, so there is the same current through both. The same kind of calorimetry, flow.

This sounds foolish ABD. No two cells have (over a longer period) exactly the same resistance and thus the same voltage. A current excursion would show up in both cells.
Thus conclusion: They worked in parallel.

Knowledge sounds foolish to the ignorant, who analyze it with dysfunctional assumptions.

The cells were in series. SRI uses constant-current power supplies, so they set the current. (I have studied the supply they used in detail to understand claims of power measurement error caused by bubble noise. The bandwidth for the voltage adjustment necessary for constant-current is about 1 MHz, far higher than the frequency range of bubble noise.)

The voltage across the cells will vary, because, Wyttenbach is correct, the resistance will vary, especially because of the difference between deuterium and hydrogen. That information (voltage and current) is used in the calculation of XP, which is what was plotted.

The current was through both cells. But only one showed XP, the deuterium cell. You can see the increase in calorimetric noise in the hydrogen cell.

Given that this is one of the major experiments in the history of cold fusion, having been done by SRI as part of work where SRI was retained by EPRI to investigate cold fusion, to replicate the work done by Pons and Fleischmann and others and to explore the "parameter space," I suggest reading the sources. I pointed to the 2004 DoE review paper, but the EPRI primary source is available. It's a lot of information, huge.

It is years of study of material like that which creates informed commentary. You could start now.

Quote from JedRothwell

Yes. To summarize: it would produce 24,000 times less power and 1,700 times less energy. Numbers matter. Science is supposed to be quantitative. When your "explanation" fails by 3 or 4 orders of magnitude, you need to rethink it.

Not if I believe that these claims are coming from tinfoil-hat pseudoscientists who have somehow managed to delude journal editors into printing their papers, and promoting their nonsense, whereas I am the only true scientist left looking at this soberly and all their claims are wrong. Wrong! Wrong, I tell you! Besides, they stated my theory wrong!

Unfair!

To be fair to Kirk, we would need to look at specific, documented claims, not my restatement of them, and give him an opportunity to say, perhaps, "well, that argument wasn't the brightest bulb in my pack."

Where I would hope to find common ground with Shanahan is in identifying predictions from his theory, and especially ones that would be easy to confirm. Nobody is going to spend $20,000 to confirm his theory. But people might tweak and experiment, make some extra and easy measurement, etc.

It depends on what Kirk comes up with that is much more specific than his usual. He did get down to brass tacks in some of this discussion, so maybe we will see more. Besides, it's educating some here. Including me, Damn! I had to actually look at Storms' math. Okay, it was easy, but I break out in hives. It takes time. I have to be careful, etc.

Discussing these things gives me a much deeper grounding. It creates familiarity, which is useful as long as it does not translate into "I'm right" certainty.

Quote from Longview

Too busy to read all this thread. But, anyone comtemplating quantitatively analyzing this situation should remember that there is another recombination that may be relevant here. That is the recombination of atomic hydrogen to molecular. The delta H is even larger than the burning of molecular hydrogen H + H --> HH -436 kJ/mole v. HH + O --> H2O -286 kJ/mole. Also the atomic recombination is a possible source for HAD phenomena.

Where does the atomic hydrogen come from?

There is some basic chemistry here: the reaction of molecular hydrogen with palladium, which dissociates the molecule, i.e,. the formation of palladium deuteride, is exothermic. It takes energy to break the H2 bond, but this is more than returned by the heat of formation of palladium deuteride. So the reverse reaction, the release of hydrogen -- which will "evaporate" from palladium deuteride -- is endothermic. Yes, the released hydrogen will combine with itself to generate heat, but that is more than compensated by the cooling from the evaporation. Thus atomic recombination cannot be a source of heat for HAD.

There is the "cigarette lighter effect," named after a commercial product that existed at one time, a piece of palladium loaded with hydrogen. When exposed to the air, it would release hydrogen and the palladium surface would catalyze the reaction with oxygen in the air, which would heat the palladium, causing more rapid release. Until it was snuffed out. This is exothermic, obviously. due to the chemical heat from oxidation, which, again, outweighs the endothermy from release.

The problem in HAD wrt the "cigarette lighter effect" would be twofold: the amount of such energy available, and the limited availability of oxygen.