Reproducibility of Excess of Power and Evidence of 4He in Palladium Foils Loaded with Deuterium

This is a slide presentation from 2005, authored by M.Apicella(1), H. Branover(2), E. Castagna(3), I. Dardik(4) A. El Boher(2), S.Lesin(2),
G. Mazzitelli(1), M. McKubre(5), F.Sarto(1), C. Sibilia(3), E. Santoro(1), F. Tanzella(5), V. Violante(1), T. Zilov(2)
(1) ENEA Frascati Research Center V.le E. Fermi 45 00044 Frascati (RM) ItalyV. le E. Fermi 45 italy
(2) Energetics, Ltd, Omer Industrial Park 84965 Israel)
(3) La Sapienza University, Via Scarpa, 14 00100 (Roma) Italy
(4) Energetics LLC 7 Fieldview Lane, Califon, NJ 07830 USA
(5) SRI International 333 Ravenswood Ave, Menlo Park CA 94025 USA.

I’ve uploaded the file here. The occasion today is that Kirk Shanahan posted a commentary on this presentation on LENR Forum, in response to my suggestion that he read my 2015 Current Science paper — I had cited this document for a small part of it. Kirk commonly turns every conversation into his favorite topic, this was no exception. Perhaps we will learn something here.

Abd said I should read his paper, so I did. Nothing but recitation of what others say.

Indeed. It’s a review, not something new, and most of what is cited was quite old. However, what I wrote in that paper was considered significant by McKubre, enough to be mentioned in his 2016 ICCF-20 keynote.

In his review of helium-4 and heat correlations in 2015, Abd ul-Rahman Lomax[10] states: “Miles was amply confirmed, and precision has increased. While there are outliers, there is no experimental evidence contradicting the correlation, and only the exact ratio remains in question. In this, we have direct evidence that the effect is real and is nuclear in nature; the mechanism remains a mystery well worth exploration.” For an experimental result of earth shattering importance, first reported publicly in 1991, it took until 2015, 24 years, for this conclusion to be stated with such clarity and conviction. Why? And even now not every researcher in the CMNS world would agree that helium-4 is the primary product, or even a nuclear one!

What was largely new with the paper was the title, “Replicable cold fusion experiment: heat/helium ratio.” This, unlike most LENR experiments, is quantitatively replicable.

He referenced a Powerpoint presentation by a group of authors whose primary CFer is McKubre that details some positive CF experiments ( that I’d like to comment on.

“Primary CFer” is standard Shanahan insult. Because of language like this, I’m not about to take any issues that are legitimately raised here to the CF research community for comment, but if such consultation becomes appropriate, I’ll restate it all, leaving out the load of carp.

Many of the authors are familar names to me. Then again, I actually study the field instead of just throwing darts at it occasionally. Links added.

At the end in the background material there is a slide that actually has a calibration equation on it for their isoperibolic calorimeter. It gives electrolyte temp as a function of input power. The equation is: Telec = -0.1649 * Pin2 + 5.3636*Pin + 24.337, and it has a multiple R2 value of 1, implying it is a very highly precise equation.

The document doesn’t actually say “Telec” but it is probably the electrolyte temperature. It is the average temperature of two PT-100 sensors. The caption does not thoroughly explain what this is. “By electrolysis in LiOD” is vague.” It is presumably a D2O solution of LiOD. But calibration by electrolysis, if that is what they mean, would be problematic. It gets complicated with deuterium evolution, etc. I’d think they would calibrate with a resistor. So this is a question that could be asked of Violante.

That is a curve fit for calibration data, not precisely, only plotted. I’m not thrilled with the claimed R2 = 1, but it may simply mean that the behavior was within measurement precision. This wasn’t a scientific paper, it was a presentation at an APS meeting, by multiple authors working at different institutions. It was not peer reviewed, I expect.

“Their” refers to ENEA Frascati, i.e., Apicella, Santoro, and Violante, and to their Laser-triggered work using isoperibolic calorimetry. that also measured helium.

This can be reversed to predict Pin given the Telec values. I did it by computing Telec for Pin values of 1, 2, 3, 4, 5, and 6W, and then using the Excel fitting routine for a quadratic. I got this equation: Pin = .0022001*Telec2 +.0060493*Telec -2.6978, with an R2 = .99997 (I’m not sure why it didn’t give 1.0, probably round off error). Of course excess power (Pex) is given by Pex = Pout – Pin, and in calibration we set Pout = Pin.

The question is what a small change in calibration constants would do to apparent excess heat. So I started with the McK equation to compute Telec for the Pin’s given above, then changed the linear and quadratic term constants in the reversed equation by +1%, and recomputed the ‘new’ expected Telec.

A small problem. This is not a “McK equation.” This is for ENEA Frascati work. There is much more information on this specific work in Apicella et al, Some Recent Results at ENEA, 2005.

Then I went back to the original McK equation and computed the Pout values for those new Telec’s. At 6W Pin, the shifted equation gives an apparent excess heat of ~78 mW. In my Storms’ reanalysis, I found a +/- 2.5% shift, which translates here to a 195 mW 1 sigma value (for 2.5% shift). Thus the 3-sigma band is +/- 585 mW, which can be rounded up to 600 mW band. Thus theoretically the excess power signal needs to exceed 600 mW to be ‘out of the noise’ if a 1% CCS has occurred.

A signals has no need. We do, as humans. Need for what?

I would certainly not assert that the ENEA results are “major heat,” however, a 1% calibration constant shift would be large. Shanahan’s work and conclusions have never been confirmed. I have not verified his calculations here, but I have no reason to doubt them, they simply are not surprising, in themselves, i.e., Shahanan has been writing this for years.

In careful work, calibrations are not just done before an experiment, they will also be done after. Calibrations do shift, though the magnitude he is proposing seems high. Most electrochemists dismiss Shanahan out of hand, his CCS has been called “random.” However, in fact, he is proposing a systematic shift. If it is systematic it should not be difficult to confirm.

Shanahan’s complaints, though, are not likely to lead to that.

McK, et al have several slides claiming excess power. For example, their 3rd slide shows an excess power peak from a flow calorimeter of ~90 mW in a spike, smoothed I say more like 70 mW. This is approximately the same magnitude as the 1% CCS effect.

This presentation is not “McK, et al.” The lead presenter, by which the paper is cited, is Apicella of ENEA. Slide 3 should be compared with slide 2, which with H2O and 0.1 M LiOH. Notice how the cumulative energy out remains below the cumulative energy in, showing apparent calorimetry capture of 97.5%. There are lots of aspects I don’t like about the presentation. Slide 3 may be mislabeled, at least in the ICCF conference paper this is captioned as isoperibolic.

So … a 1% CCS effect is imagined to appear in the middle of an experiment like this, and then disappear? I do understand the irony: excess appears in the middle of an experiment and then disappears. However, Shanahan’s CCS is an anomaly that has not been confirmed by anyone. He then ignores most of the rest of the evidence.

Slide 6 show]s] “Excess Power at SRI”. They seem to plot an excess power (very noisy) and a smoothed version that apparently uses the right Y-axis based on the figure legend across the top of the graph. Those plots show peak values of ~55 mW (guessing at the units, since they stated ‘Total Power = 214mW’), which is within the 1% CCS 1 sigma.

I don’t know what we are seeing there, the plot is not familiar to me. Many SRI reports do show this kind of noisy power plot. To understand this, I’d want to see the history of this cell and run. It’s simply not there. Whether Slide 6 made for an effective presentation would depend entirely on what was said about it. I’d rather not guess.

Their 8th slide shows more calorimetric results for laser triggered experiments. They plot energy and power on the same graph. Of note is at the start the output power slightly exceeds input power (i.e. positive small excess power signal) but it basically tracks the input power, which is a good indication that the calibration is off or there is something else going on (Storms’ first data set for Pt-Pt F&P cell work showed negative input power feedback due to ground loops). Later on, they get spiky output power when input power is constant. The spikes are about 190 mW peak values (~2.5 times the 1% CCS effect (or just a 2.5% CCS as found in Storms’ results)).

First, see Slide 2. This is a hydrogen control. This is what zero XP looks like, for a few hours. I’d really want to see much more to be thoroughly satisfied, but this is what they have. Hydrogen does not present this messy power output.

Slide 8 does not show more than two days of startup. The scale for this plot was not designed to show what may have been early excess power, but the accumulated energy has deviated early on. The plots of total input energy and total output energy are confusing; cold fusion researchers don’t seem to recognize the communication problem. However, the XP they are concerned with mostly shows up in the last days of this experiment. This is not impressive power, for sure, but it becomes more impressive when compared with the helium measurements.

So my point is that the apparent excess power/energy values shown in these slides could *easily* be a very small CCS. It seems important to me that the reality of these signals need to be determined and not just assumed to be real excess energy.

I don’t assume it’s real excess energy. It is a set of real measurements; unfortunately, we don’t have them. We have calculations done from them. It would be enough for the most important purposes that these measurements are handled consistently. “Very small” is Shanahan’s own view, not some reliable measure. 2.5% CCS hardly seems small to me, for something routinely observed to be much more stable.

Since I am looking at the Apicella, et al, slides – some other points:

In Slide 7, they state some conclusions which I find contradictory. In the first line they say: “(D/Pd > 0.9 in some cases also with less loading) have been observed at ENEA.” Then in the 2nd line they say: “We can conclude that high D loading is a necessary condition for excess of power production during loading of Pd with D.” But if one can get apparent excess power at <D/Pd=0.9, then it is incorrect to conclude that that is a requirement.

Yes, and this indicates poor editing. However, this is the reality: for a very long time, high loading was considered completely necessary. Experiments without that high loading normally did not show excess heat. The apparent contradiction goes away if we think that, say, 85% loading is high, and, historically, it used to be thought that anything above 70% loading was impossible. And that is why the early failed replications stopped loading at 70% or so, and why, now, we can look at those and see the failures as highly predictable.

However, there is recent work by Ed Storms that indicates that once conditions showing XP are set up, the loading may decline and XP may continue, quite a distance below those high numbers. This is unconfirmed, so far, but … it does make sense, given some of the odd events that have been reported, such as the original FP meltdown in about 1984.

Of course they don’t specifically say right there that >0.9 is the requirement, but in Slide 12 they do. These slides were presented in 2005, and today in 2017 the mantra is still “>0.9”. I disagree, it simply takes a little more work to get the effect when the Pd loading is <0.9.

I suggest not confusing initial conditions with what is necessary for a maintained effect.

In slide 10 they show some 4He results for laser-triggered experiments. I note that the indicated background level is ~0.55e16 and the strongest result is 1.05e16, i.e., less than 2X background.

Helium is very energetic. So sue God.

I find that to be ‘working in the noise’, and I require much more replication to be convinced this plot shows anything of value. There’s also no way to evaluate if these signals come from leaks or not.

Leaks would be unlikely to match an otherwise irrelevant and minor excess heat calculation. That is a “way to evaluate.” If Shanahan isn’t convinced, I’m not offended. He has a right to not be convinced. However, how much funding should be allocated to convincing him? What has happened is that the heat/helium evidence was considered by those who fund research to merit replication with increased precision, the classic way to test “pathological science.”

Much weaker evidence was seen by Huizenga in 1994 to be astonishing. So Shanahan’s assessment of that nothing “of value” has been shown is itself without value, worthless, exaggerated polemic, a broken record that he’s stuck playing out, maybe for the rest of his life. Ah, what a dismal prospect!

Slide 11 shows results from the SRI “M4” experiment. I’ve noted elsewhere that I have looked at the calorimetry of this run, and determined that it could well be affected by a CCS. However, there is some *very* fancy data workup going on here, and I require a full explanation of that to be able to evaluate the data’s validity. I asked McK twice for that info and never got it. The He values plotted here never exceed the usual outside air value of 5.22 ppm, and there is no report of what the 4He concentration was in the lab at the time the experiments were run, so we can’t honestly reject leaks once again.

What did Kirk ask? The data in that slide presentation is skimpy. The full report is as shown in my paper. The chart shown is not the original calculation; apparently McKubre found an error in the headspace volume and recalculated. (Which drove Krivit bonkers.) I still consider this work more or less seat of the pants and approximate. I have far higher expectations of Texas Tech and ENEA this time around. I’m hoping they use more of the Miles approach, but Shananan’s ascription of his CCS hypothesis to SRI flow calorimetry is problematic.

If he can convince THH that this is worth taking further, I’ll support that. But, bottom line, the heat/helium data almost totally demolishes Kirk’s argument. It becomes very, very unlikely that bogus heat and bogus helium data would match as well as has been observed, as often as has been observed.

Their 4th slide shows excess power from the Energetics lab that uses the “Superwave” on the input, and they don’t explain the calorimetric method. The excess is about 2.5W on an input of 4W. This clearly needs to be explained further, and the accuracy and precision clearly established, especially when the “Superwave” is being used. Ditto on the 5th slide, which is another Energetics lab results slide showing even greater apparent excess power.

The Superwave work was later replicated by SRI and ENEA, with extensive experimental series; this was published in the ACS LENR Sourcebook, 2008. I haven’t heard much about Superwave lately. The ET work went to the University of Missouri, the SKINR lab there.

None of the scientific approaches, so far, show promise for commercial power generation. Cold fusion at this point is a scientific curiosity. However, sufficient potential is there to justify exactly what both U.S. DoE reviews recommended: further research to address fundamental questions. Originally, it was thought that neutron radiation was important, it is now know that this was a complete red herring. Helium is the identified ash, there really isn’t another candidate with serious support. So measuring helium does become a method of verifying the heat, if that is considered necessary. Because measuring helium is difficult and can be expensive, finding other correlated measures would be of high value for the research.

Aiming at “convincing skeptics” is an obsolete concept that can backfire. Nailing down what is known is of high value, increasing predictability and clarity..

A mind is a terrible thing to waste

Kirk Shanahan is the most-recently published, practically the last standing critic of LENR who has been published in a peer-reviewed journal. His view of himself might be that he has demolished “cold fusioneers,” as he has called researchers and writers, but that they are stubborn and refuse to recognize utter defeat.

Funny how easy it is to imagine that about others and not notice the old saw: that when we have one finger pointing at others, we have four fingers pointing back at ourselves. Any sane skeptic must be aware of this problem, and not rely on self-assessment for conclusions about social position, argumentative success, and the like.

On LENR Forum, kirkshanahan wrote: 

Wow…I thought we had dispensed with Abd’s garbage on this forum. Oh well…one more time…

In this case, I didn’t post my “garbage” to LENR Forum, someone else did, and that’s a sign that someone else saw it as worthy of consideration. I like it this way. Consider it some kind of informal peer review. Zeus46 is anonymous, but so is most genuine peer review in journals.

Shanahan has never figured out how to use the Forum quotation interface, which would allow him to properly attribute the quotation. He makes his attitude clear. Most LENR writers don’t bother with Shanahan any more. I’ve taken him seriously, have agreed that some of the dismissal of Shanahan may be unfair, and, as part of this consideration, I have identified and pointed out errors; yet I have never encountered gratitude for this, only abuse. I continue only for one reason: CMNS needs critique, and it’s not easy to come by, so I encourage it.

Because there are now several posts and pages relating to Shanahan, I’m creating a Shanahan category and will be applying it. This post is a review of his LENR Forum comment. I did not make that comment, Zeus46 did. If Shanahan had difficulty distinguishing Zeus46’s comments from mine, Zeus46 did link here.

ABD quoted me and wrote:

KS wrote: There are 38 references listed. 3 of them refer to the ‘general rejection’ of LENR by mainstream science (they refer to the books by Huizenga, Taubes, and Park).

ABD: The books are references for the statement: “The special condition required to cause the LENR reaction is difficult to create. This difficulty has encouraged general rejection by conventional science [13-15] and has slowed understanding.”

My response: What’s yer point???

Shanahan takes every discussion as a debate, and in a debate, some will never concede fact alleged by the other side. It will either be wrong or “beside the point.” Sometimes I have points to make, other times I simply note, for the reader, fact. What I wrote was simple, verifiable fact, and if there is no point to simple, verifiable fact, then there is no possibility of communication. Consensus can be built from fact.

What, indeed, is the point of Shanahan’s asking “What’s yer point???” ??? Someone seeking straight and clear communication would have written nothing or would have written something like “Yes.” Not what he wrote. This has been going on for years.

ABD quoted me and wrote:

KS wrote: If you look closely at Figure 2, you will see the He/Heat values exceed the theoretical amount in some cases.

ABD: No. In one case, the value is on the theoretical amount, but something must be understood about this data. If what is being calculated is the heat/helium ratio, and if the actual ratio is a constant, experimental error will cause greater deviation from the actual ratio if the produced heat (or helium) are at low values. I have never seen the data presented with careful consideration of error bars as they affect the ratio.

My response: As I put in my original disclaimer, I did this review quickly, and Abd has found a minor error I have made. Let me correct that now.

Does he correct the error? (Yes, below.) “I thank Abd for noticing my error.” Shanahan doesn’t do apologies, not that I’ve ever seen. And then he, has, in the past, gone on to assert that it doesn’t matter, because New Reason He Was Really Right. He doesn’t break that pattern here.

(On LENR Forum, authors may edit their posts, so he could actually fix the error. He could use strike-through to avoid making the comments of others unintelligible. He could point to the correction, etc. As of this writing, the original post has not been touched. When I see an error like that, I immediately address it. In a recent post here, I’d made a huge mistake. When it was pointed out, I immediately unpublished that post, returning it to draft status, responded to the user who had pointed out the error, and created a new post documenting what I’d done. And then I fixed the error, and rewrote the post that had depended on it. Shanahan seems to have no concept of using these fora to develop scientific or social consensus. Will he ever turn around?)

What is funny is that once again, correcting my error places Storms in an even worse light.

Once again, we can see the polemic intent. It is about good light and bad light. Does the light change reality? Bad light on something would be bad interpretation. From my own training, what is highly likely is that Shanahan will careen from one error to another, because the error of others is a matter of certainty to him. He won’t recognize nuances, and what occurs to him as a result of his world-view he will think of as plain and simple evidence or … proof. What does he come up with?

Storms’ Figure 2 is an alternative presentation of the ‘heat/helium correlation’ idea. He plots the number of experiments obtaining a value for the number of He atoms/watt-sec that lies within a specified range versus the mid-range value for that ‘bin’, in a typical histogram approach.


He overlays a Gaussian fit to the data as a curve on the graph. The number of experiments obtaining a He/heat value in the selected range is indicated by a pink box on the plot. Storms also adds a vertical black line on the plot, and labels it “D+D=He”. I observed pink boxes at larger values than the black line.

Here is the plot:

My mistake was to imagine Storms was using the data from his book’s Figure 47, which does show 1 point above the theoretical line and to assume he’d added a couple more (which would be expected based on prior data characteristics). In fact there are several pink boxes at zero values and most are above the black line. Only 1 lies below. So, my mistake, Storms does NOT show any positive values above the theoretical line.

That wasn’t the only mistake, the imagination didn’t fit what was in front of him. Shanahan, as well, knew that this was not the data from the 2007 book, because there were more data points. Yes, he wrote quickly and without caution.

So, I have to ask, what happened to the data point from Figure 47 that was well above the theoretical line? Apparently, without telling anyone, Storms has rejected that datum.

He does tell us in his formally published paper. And I pointed to this. In correlation studies (and that original figure 47 was a correlation study) one will report all data. In attempting to determine a ratio, one may eliminate clear outliers. I discussed all this, and Shanahan starts out responding as if he has never seen any of this. He is reactive and attached to his point of view, which boils down to “I’m right and they are wrong.” Does he go any further than that?

But that radically alters the interpretation of Figure 2. As I noted in other comments, that one datum alters the estimated standard deviation such that the 3 sigma spread encompasses the 0 line as well as going well over the theoretical line. It also swings the average up a bit. If you clip it out, you get a radically different picture, i.e. supposedly ‘all’ data points are now below theoretical (and we (meaning Storms and other CFers) have an ‘explanation’ for that). In my prior comments on Figure 47 from Storms’ book, I discussed why clipping out that high value was an illegitimate thing to do.

Miles reported it. The purpose of Storms’ Figure 47 has been ignored; it appears to me that it was an attempt to show that the ratio settled as the reported energy (or average power for the collection period, similar) increased. As mentioned above, in a correlation study, cherry-picking results is very dangerous. Miles did not do that. He also has zero-heat and zero-helium results (and three outliers of a different kind, experiments where reported heat was significant, but no significant helium was found). All results are part of Mile’s full consideration. Shanahan almost entirely ignores all this.

Storms’ Figure 47, nor his values on the next page, do not consider the 0/0 or 0/energy values. However, that next page does show the “flyer,” and has a note on it: “eliminated from average.”

So of course Storms looks “worse” in this light. The “light” is what Shanahan sees with his eyes closed. He may again excuse his “errors” — if he does admit error here, I suspect he might not — by his having written quickly, just dealing with one paragraph at a time.

So let’s see if he straightens up and flies right:

The functional difference is that including it leads to the conclusion the experiments are too imprecise to use in making the ‘desired’ conclusion. Excluding it means you can use the data to support the LENR idea. But which of these is forcing the data to a predefined conclusion do you think?

What conclusion? And is it “desired” or observed?

Data like this was enough to inspire about $12 million in funding for a project with the first declared purpose being to confirm the heat/helium correlation with greater precision. That’s the only “conclusion” that I care about, long-term. Long ago, within my first year of starting to again look at LENR evidence, I personally concluded that there was much stronger evidence, with a replicable and confirmed experiment behind it, than was commonly being represented — and that includes representation by the CMNS community. There are historical causes for this that I won’t go into here.

It is SOP to exclude an obvious outlier, when calculating a data correspondence, i.e., a ratio, particularly where the outlier has less intrinsic precision than the other values. Whenever this is done, it should properly be reported; it is unfortunately common for LENR reports to only show “positive” results, perhaps because some workers might do dozens of experiments and only see signs of LENR in a few. That is a systemic error in the field that I’ve been working to correct. Some researchers think it is preposterous to report all that “useless junk,” but that is the kind of thinking that has inhibited the acceptance of LENR, allowing vague claims to seem plausible that it’s all “file drawer effect.”

Abd said: “I have never seen the data presented with careful consideration of error bars as they affect the ratio.” – Perhaps, but I have discussed just that before, and now again in summary. Obviously Abd reads what I write, but apparently very selectively (which is typical of people looking to discredit something but not seeking to understand).

And Shanahan’s response here shows how he understands what I write, which is apparently very little. He does not show evidence of my reading “selectively,” yet proceeds to draw conclusions from his own imagination.

He apparently agrees with me, makes the point that he’s said this before (and he may have, I don’t know). I was writing about what was in front of me, his comments, and commenting, mentioning a problem that I know, and if he were interested in the development of consensus, he’d acknowledge the possible agreement. But somehow he converts this to an intention to discredit him.

Rather, my goal is to separate the wheat from the chaff. What is useful about Shanahan’s commentary? As I think I pointed out, few are paying any attention to him any more. The attention he is getting on LENR Forum and here is almost the entire sum of it. As far as we know, he is not submitting critiques of published papers to journals, nor is he writing and submitting original work or reviews. He is more or less, now, confined to complaining about how he has not been accepted, while continuing to display the personality traits that suppress consideration in the real world.

ABD quoted me and wrote:

KS wrote: I have previously commented in this forum on the related Figure from Storms book, which only had 13 numbers on it rather than 17, where I noted that the spread in the data indicates the precision of this measurement is too poor to allow one to make the conclusions Storms does. This hasn’t changed by the addition of 4 points.

From his notice of 17 rather than 13, Shanahan could have realized that this wasn’t the same data. Likewise what Storms writes about “four independent laboratories,” whereas Figure 47 reported from two. What conclusions? I infer several possibilities from this, one of which is that Shanahan is not truly familiar with the evidence. It can be tricky to remember stuff if you believe it is all bogus, it tends to blur into one solid mass of Wrong. (This is an aspect of how belief undermines clear understanding.)

From the Storms paper under review:

This ratio has been measured 17 times by four independent laboratories, the result of which is plotted in Figure 2. This collection shows a range of values with an expected amount of random scatter. Of considerable importance, the average value is equal to about 50% of the value expected to result from d-d fusion. This difference is thought to result because some helium would be retained by the palladium in which the LENR reaction occurred. When efforts were made to remove all the trapped helium from the palladium, the expected value for d-d fusion was obtained [33].

Figure 2 : Summary of 17 measurements of both helium and energy production during the same study [32]. Superimposed on the distribution of values is a fit to the Gaussian error function. The fit is typical of an expected amount of random error being present in the measurements. The value for this ratio resulting from deuterium-deuterium (d-d) fusion is known to be 23.8 MeV for each nucleus of helium made.

Unfortunately, ref 32 is to a Storms paper that does not contain support for the caption. Decent journal editing would have caught this. I have seen the histogram before, but couldn’t find it easily (as I write this, I still haven’t found it); but I was able, without much difficulty, to find the data, given in Storms Current Science paper, which I cited. It is also in his 2014 book.

ABD: Shanahan doesn’t know what he’s looking at. The “Storms book” he is referring to is Storms (2007). Figure 47 in that book is a plot of helium/heat vs excess power, for 13 measurements from two sources: Miles and Bush & Lagowski. The Miles data is more scattered than the Bush data. Miles includes one value with the lowest heat (20 mW). The associated helium measurement generates a helium/heat value that is an obvious outlier.

This newer histogram I think is from data in Storms book (2014), The Explanation of Low Energy Nuclear Reaction. Table 9 (p. 42) is a summary of values. There are 19 values. It looks like Storms has omitted one value (2.4 x 10^11 He/W-sec) as “sonic” (Stringham), one as an outlier (4.4), and maybe one as “gas loading,” (McKubre, Case), then perhaps has added one. Or maybe he left in the Case value (2.0).

My response: “Shanahan doesn’t know what he’s looking at.” – Really? Really??

Really, really, and literally really. Truthfully and on clear evidence. He didn’t know, and has acknowledged that he thought this was from Storms (2007), when it obviously was not. An error. Small or otherwise.

All-too-common interpretive principle: Your errors are fatal, demonstrating ignorance and stupidity and worse, whereas mine are minor, trival, of no consequence, and I was right anyway.

“I think”? Yes, Abd is right, you have to guess at where it comes from.

Well, I did better than guess, but it’s not a certainty, merely very likely, since Storms has published this data at least twice, once in his book (2014), which Shanahan might not have, and once in Current Science in 2015, with the appropriately named Introduction to the main experimental findings of the LENR field — and this was cited in my response.

This was an actual peer-reviewed paper. I know that my own paper’s review in that Special Section of Current Science was real (and even initially hostile!), and also the copy editing was strong. That’s a real (and venerable) multidisciplinary scientific journal. If Shanahan thinks that nonsense is being published there, he could certainly write a response. If they wouldn’t publish it, I would, I assume, working with him to clean it up — or THH could assist, etc. — arrange publication anyway, but I doubt that Shanahan has tried. (We could help him clean it up, and what he submitted to Current Science would be his choice, not ours. I.e., I would advise, with help from anyone Shanahan was willing to allow to see the draft.

As I noted in my initial review, the referencing on this paper stinks. Where the data comes from is actually not specified, so you can’t check it.

That’s correct that it is not specified, but it is possible to check it.

The citation error is one item on the pile of indications that this was a predatory publisher. I’ve seen this happening to more than one older researcher. Takahashi, a genuine scientist, not marginal, published in a predatory publisher’s journal.

However, what’s the topic here? Formally, on LENR Forum, the topic was the paper. So, granted, it’s poorly referenced. What else can we agree upon? Shanahan wrote, however, about the underlying data, and it’s easy to find the substantially identical underlying data. I did not actually research this all the way. The Current Science paper gives references for all the measured values. With only a little work, someone could reproduce the histogram with full references. How important is it?

From my point of view, all this is likely to become relatively obsolete soon. The standing evidence — which Storms does show, as did I in my own paper — was quite enough to justify significant investment in research funding. Shanahan is, too often, focused on being right, whereas the real world is focused on exploring science and especially mysteries with possible major real-world consequences.

How much attention should be given to Shanahan’s CCS and ATER ideas? Basically, unexpected recombination, the major core of this, should be always be considered with the FP Heat Effect, and, where practical, measured (which can include finding upper bounds). That has already been done to some extent (Shanahan seems to mostly ignore this, but he’s welcome to correct me or request confirmation).

Abd makes some interesting guesses about where it comes from, and most importantly, he notes that Storms’ is picking and choosing what to look at. A clear recipe for making the data say what you want it to say, instead of what it actually says.

Again, he could be agreeing. However, I’ve personally gone through the exercise of looking at what data to present in a summary chart. I wanted to present it all, in fact, all the data we have. I came to realize that this was a monumental task, with hosts of data selection problems. Many of the data points are isolated measurements. Then there are variations in experimental technique. I don’t think that Storms selected the data to show based on desired outcome. On the other hand, Storms does not state how he picked what studies to show.

His 2014 book lists 30 helium studies. Many of them provide no clear information about the heat/helium ratio. Many are obviously flawed in different ways. Post-hoc analysis of correlation studies is problematic; it is primarily useful for suggesting further research. Even the Miles work, which is outstanding for this, was not designed in full anticipation of the importance, and was not uniform experimentally. Miles did not set up a full protocol for rigorous correlation study. Close, but not completely. For example, what do you do if some incident creates possible major error in measuring heat? Miles varied the cathode material and created two outliers (that don’t show in the Storms chart). Apparent heat but no apparent helium. Miles later wanted to study this, I think, submitting a proposal to the DoE, which was denied. I suspect that the importance wasn’t established, and investigating Pd-Ce cathodes remains a possible avenue for research. I do not recommend at this point that the Texas Tech/ENEA collaboration complicate the work by trying to explore outliers. Yet. First things first! Keep it as simple as possible, as few variables as possible.

Right now, I’m only considering, and only a little, Shanahan’s response. A deeper study would list all helium studies and set up some selection criteria in an attempt to generate more objective data for a histogram. It might look at the sources for the histogram and compare these studies with the entire body of studies. Until then, my impression is that Storms’ selection criteria were reasonable; particularly if we understand that what is really needed is more precise confirmation, that this does not shut the book, close the case, lead to a final conclusion, and for what purpose would we even think this?

I notice that Shanahan’s critique here is ad hoc and without foundation. He is essentially alleging cherry-picking without showing any evidence for it. The single outlier is acknowledged by Storms in the prior publications. The failure in sourcing is really a journal failure, my opinion; for when a paper is submitted by a scientist in his eighties, I don’t expect perfection. AStorms did not ask me — or anyone, as far as I know — about the wisdom of that submission there. I’ve advised him against spinning his wheels with useless and unfocused repetition of speculations, his “explanations.” He doesn’t like it. So I’ve mostly stopped.

ABD quoted me and wrote:

KS wrote:

This newer histogram I think is from data in Storms book (2014), The Explanation of Low Energy Nuclear Reaction. Table 9 (p. 42) is a summary of values. There are 19 values. It looks like Storms has omitted one value (2.4 x 10^11 He/W-sec) as “sonic” (Stringham), one as an outlier (4.4), and maybe one as “gas loading,” (McKubre, Case), then perhaps has added one. Or maybe he left in the Case value (2.0).

ABD: It’s been confirmed. Maybe Shanahan should actually read my paper. After all, I cited his JEM Letter. It is not a “hand-waving” argument, but, obviously, this cried out for more extensive confirmation with increased precision. And so, I’m happy to say, that work has been funded and is under way. And they will do anodic erosion, I’m told, to test what is apparent from the two studies that did it (McKubre and Apicella et al, see my paper for references). These are the two studies where dissolving the surface of the cathode took the helium level up to the full theoretical value, within experimental error. Two other Apicella (Violante) measurements did not use anodic erosion, and results were at about 60% of the theoretical.

My response: The quote attributed to me is just what Abd wrote immediately above. Cut-and-paste malfunction. If Abd will actually use my quote I might be able to respond.

Apparently Shanahan did not look at my original comment. It’s here, as cited by Zeus46: Reviewing Shanahan reviewing Storms. What is quotation of Shanahan and what is my comment is clear there, I hope. Zeus46 translated the blog format to LF format and incorrectly set up quotations. It was not exactly a cut-and-paste error, but a reformatting error. Shanahan could easily have responded to what was written; after all, he knows what he wrote and then what I wrote, and he could be even more clear if he actually followed Zeus46’s link and read the original.

ABD quoted me and wrote:

KS wrote: Exactly so. So one shouldn’t try to work with these numbers until they are shown to be free of the errors Storms points out, which hasn’t happened.

ABD: Shanahan ignores that correlation can show relationships in noisy data. (This is routine in medicine!) Leakage, quite simply, doesn’t explain the experimental evidence. It could have had an effect on some individual measurements. No, we were not going to wait for “error-free” measurements, but rather how to proceed was obvious: the data shows quite adequate evidence to justify funding further research to confirm these results, and this is a replicable experiment, even if heat, by itself, is not reliable. The variability creates natural experimental controls.

My response: “Shanahan ignores…” No, I don’t. But Abd ignores the point that correlations derived from fictitious data (excess heat is likely not real) are worthless. For the record, I have been using statistics for many years, and Abd has added nothing to my knowledge base.

And I can see here — and, I’m sure, many others who read this can also see — the problem.

First of all, “fictitious data” is not defined. Shanahan is not actually talking about fiction, i.e., made-up, invented data, as distinct from the results of actual measurements (and calculations from measurements). Correlation is how we distinguish random variation from systematic, causally connected variation. What the heat/helium data shows is correlation, which can be quantified. The quantification shows a high probability that the data is not random.

(Storms uses that data to show the kind of variation typical of experimental data which is, by the nature of the work, approximate, not fully precise.)

There is, then, likely, a causal connection. This, in itself, does not show “nuclear,” only that there is likely some common cause.

Shanahan, when he says that the data is “fictitious,” is actually stating, with remarkable lack of sophistication, that because the heat data might be non-nuclear in nature (his own theory), it’s fictitious, not “real.” That’s preposterous. It’s real, that is, there is actually an anomaly, or Shanahan’s entire publication history is bogus. He is simply claiming that the anomaly is not nuclear in nature. Not “real nuclear” heat. But real heat, in some cases, caused by unexpected recombination, or … a real measurement anomaly, systematic, caused by some kind of calibration constant shift, perhaps caused by heat being generated in a place different from expectation or calibration.

This runs into many problems that he glosses over, but one at a time. Cold fusion researchers have studied anomalous heat, it is often called by a neutral name, like the Anomalous Heat Effect. Shanahan agrees there is an AHE. He claims it is due to unexpected recombination or sometimes, perhaps, other causes.

Great, so far. Now, in some studies, there was a search for other results, measurement of tritium, neutrons, transmutations, and other possible correlated conditions, i.e., material, current density, etc., and in particular, and with the most interesting results, helium evolution (generally in the gas phase in electrochemical experiments, but also some other study).

Helium, of course, could, in some experiments, be the result of leakage. That’s been the standard objection for years. However, in some experiments, helium levels rose above ambient. Still, someone might suggest that local helium was high because of nearby experiments releasing helium.

However, would we expect, then, that heat and helium would have strong correlation or weak correlation? If a correlation is proposed, what would be a plausible explanation for it, and how could this be tested? Have those tests already been done? If not, is it possible to suggest that there be tests for this? Is it plausible enough to justify spending research dollars on it?

Shanahan is clearly rejecting the significance of correlation.

“Leakage, quite simply, doesn’t explain the experimental evidence. It could have had an effect on some individual measurements.” – And it certainly does. But in the ATER/CCS proposed mechanism there is a way to get increasing He signals in cells that show apparent excess heat. You all will also note that Abd does not respond to my specification that lab He concentrations need to be reported. Another thing he conveniently ignores.

I’ve made the same suggestion. I don’t ignore this. Once one is arranging many helium measurements, background helium should be routinely measured. However, Shanahan refuses to recognize the infinite regress he is creating. Some local anomalous helium would be very unlikely to correlate with heat. It would contaminate controls as readily as experimental heat-producing cells. Shanahan here is not being specific; he is assuming that increased heat production represents some major difference in cell behavior. In fact, it’s typically only a few degrees C in temperature, and cells with high heat may actually be at a lower temperature, it depends on experimental details.

And then how likely is it that the ratio ends up roughly on the money for deuterium conversion to helium? With reasonable consistency, over many experiments with multiple research groups? The work that it takes to obtain the AHE and the work that it takes to collect precise helium samples is quite different. The sampling with Miles, at least, was done blind. And Miles did measure background helium, and also studied leakage, quantified it.

“the data shows quite adequate evidence” – As I noted, that is true only if you start dropping out data that causes that conclusion to not be true. That’s bad science.

Shanahan is quoting out of context. “Adequate” had a specific referent, which Shanahan ignores. Adequate to justify new and substantial funding to test the hypothesis. What data? What conclusion? Shanahan is struggling with ghosts, cobwebs in his mind. Must be frustrating.

“The variability creates natural experimental controls.” – What? That makes no sense.

No sense to Shanahan, demonstrating that he is lacking in sense. This is really obvious, so obvious that I’m tempted not to explain it unless someone asks. Okay, I’ll say this much, though I’ve said it many, many times.

What happens with FP Heat Effect experiments is that researchers will make a series of cells as identical to each other as reasonably possible. Further, with heat/helium, the same cell is observed for heat and gases are sampled for helium. With different cells, but ostensibly identical, the only clear variation is the amount of heat, so “dead cells” are controls. What is different about a dead cell vs one showing anomalous heat. This is basic science, reducing variables as much as possible.

When Miles reports 33 observations of heat and helium, with 12 showing no heat and no helium, and 21 showing heat (and 18 showing significant helium), that is not 33 different cells, it is a smaller number, with multiple samples of gas taken with heat measured (and averaged) for the gas collection period.

Unfortunately, not all the cells were identical. However, the single-cell results, showing helium varying with average heat in a single collection period, are self-controls of a kind, because the cell is identical. To discuss this further would require very detailed analysis of the Miles work.

That Shanahan doesn’t see the idea shows that he has never deeply considered these reports, which go back to 1991. He looks at them enough to find what he thinks a vulnerability and takes a potshot. It gets old.

If THH here wants to assist looking more deeply at Shanahan’s claims, great, or if anyone else wants to do that, I’ll support it. THH has already started some of this.

ABD quoted me and wrote:

KS wrote: I published a consistent, non-nuclear explanation of apparent excess energy signals, but of course Storms refuses to recognize this.

ABD: Shanahan expects Storms to “recognize” Shanahan’s explanation as “consistent” with the evidence Storms knows well, when Shanahan, with obviously less experience, does not recognize Storms’ opinions, and merely asserts his own as valid?

My response: Read carefully here folks. Abd is pulling a fast one. He implies I ignore Storms’ opinions/conclusions. I don’t, I provide an alternative. I do not assert it is valid, I assert it has the potential to be valid. Like all proposed mechanisms, it must be confirmed experimentally, but that will never happen when the people who can do so refuse to accept it and instead resort to falsified representations of it to justify ignoring it. Abd’s response above is a veiled ‘call to authority’ (“Storms is the authority and Shanahan isn’t, so believe Storms”) which is recognized as an invalid logical technique, often used to intimidate others into silence. It has no inherent truth value.

I have not said “believe Storms,” and on this issue, in particular, I do not depend on Storms for anything (other than I specifically cited in my own paper).

In fact, I encouraged Storms to write in more detail about heat/helium and he actually wrote a paper on it and submitted it to Naturwissenschaften. They came back and requested a general review of cold fusion. I regret that, in fact, because a general review will cover a vast territory whereas cold fusion needs focus on narrow specifics, confirmed results, and especially the clearest and most widely confirmed.

Storms has made errors in his heat/helium publications and I have pointed them out.

My point was that Shanahan appears to expect Storms to recognize his critiques, when Storms has addressed them — at least some of them, and Shanahan has presented a bit of a moving target — years ago and considers the matter resolved. Shanahan uses Storms lack of continued consideration as if it were proof of Storms’ scientific bogosity.

There is a far better approach, that could work to move beyond the limitations that Shanahan experiences, but it seems he is not interested. He prefers to complain about others. And if this isn’t true, he’s quite welcome to demonstrate otherwise. Starting here and now.

At this point I can’t tell if this is Abd or Zeus46 writing, but whoever it is wrote:

“Shanahan’s views are idiosyncratic and isolated, and he has neither undertaken experimental work himself, nor managed to convince any experimentalist to test his ideas. To the electrochemists involved with LENR, his views are preposterous, his mechanism radically unexpected.

I wrote that, and all Shanahan needed to do to identify this would have been to follow the link in Zeus46’s post. He calls it the “full monty.” I.e., the “real deal.”

Yes, I’m sure that response is frustrating. After all, LENR is anomalous, unexpected. However … Shanahan’s explanations are, generally, a pile of alternate assumptions, chosen ad hoc, and his claim is that they have been inadequately considered, but who decides what is adequate and what is not? Shanahan?”

But these paragraphs are nothing but CF fanatic fantasies. There’s nothing in them worth responding to.

“Who decides what is adequate and what is not” is a question, not a fantasy. I then proposed a possible answer: Shanahan. What does Shanahan think? How does he assess this?

I proposed a practical standard: funding decisions. It’s enough if it is funded, not if it is not.

Nowhere in all this does Shanahan point to any “fantasy.”

He is fighting his own ghosts, wasting his own life. It’s quite common, and this has almost nothing to do with cold fusion, itself. It’s a people thing, and that’s my primary interest: people. Not cold fusion, that’s just something that I happened to learn about, for better or worse.

Let’s just remove the outliers

Second of the series of posts I promised on the He/excess heat correlation debate, as noted by Shanahan and Lomax. And this one is a little bit more interesting. Still, I’m going to examine the many issues here one by one, so if you expect a complete summary of the evidence from this post or the ones that follow you will be disappointed.

Lomax here:

[Quoting Shanahan in italics] On the other hand, the energy/helium ratio does not have this problem. The independent errors in the He and power measurements are unlikely to combine and create a consistent value for this ratio unless the helium and energy both resulted from the same nuclear reaction.

Yes. Very unlikely, in fact. On the order of one chance in a million, or more.

As I have noted the value is not consistent, thus the quoted statement is nonsense.

The value is consistent within experimental error.

There is much more of interest in these comments than might first appear.

Continue reading “Let’s just remove the outliers”

Minds open brains not falling out?

First of a sequence of comments on Lomax’s recent blog here on Shanahan’s review of Storms posted in LENR Forum.

Lomax writes:

Ah, Shahanan, obsessed with proof, lost science somewhere back. Science is about evidence, and testing evidence, not proof, and when our personal reactions colour how we weigh evidence, we can find ourselves way out on a limb. I’m interested in evidence supporting funding for research, and it is not necessary that anything be “proven,” but we do look at game theory and probabilities, etc.

I agree with Lomax’s second statement here. Science is exactly about weighing evidence. And I understand the explicitly acknowledged bias: Lomax wants more research in this area. I disagree with the statement that “Shanahan is obsessed with proof”. It would be accurate to say that Shanahan, both implicitly and explicitly, is looking for a much higher standard of evidence than Lomax. There is no proof in science but when evidence reaches an amount that overwhelms prior probabilities we think something is probably true. 99.99% and we call it proof. The numbers are arbitrary – some would set the bar to 99.9999% but this does not matter much because of the exponential way that probabilities combine.

Let us see in detail how this works. Continue reading “Minds open brains not falling out?”

Reviewing Shanahan reviewing Storms

A New Source of Energy using Low-Energy Fusion of Hydrogen

On LENR Forum, Alainco posted an abstract and link to a new Storms article on LENR. Kirk Shanahan promptly reviewed it. This post will study the Shanahan review. It is possible that we will review the article itself more intensely. But first, a little on the journal itself. Continue reading “Reviewing Shanahan reviewing Storms”

If I’m stupid, it’s your fault

See It was an itsy-bitsy teenie weenie yellow polka dot error and Shanahan’s Folly, in Color, for some Shanahan sniffling and shuffling, but today I see Krivit making the usual ass of himself, even more obviously. As described before, Krivit asked Shanahan if he could explain a plot, and this is it:

Red and blue lines are from Krivit, the underlying chart is from this paper copied to NET, copied here as fair use for purposes of critique, as are other brief excerpts.

Ask Krivit notes (and acknowledges), Shanahan wrote a relatively thorough response. It’s one of the best pieces of writing I’ve seen from Shanahan. He does give an explanation for the apparent anomaly, but obviously Krivit doesn’t understand it, so he changed the title of the post from “Kirk Shanahan, Can You Explain This?” to add “(He Couldn’t)”

Krivit was a wanna-be science journalist, but he ended up imagining himself to be expert, and commonly inserts his own judgments as if they are fact. “He couldn’t” obviously has a missing fact, that is, the standard of success in explanation: Krivit himself. If Krivit understands, then it has been explained. If he does not, not, and this could be interesting: obviously, Shanahan failed to communicate the explanation to Krivit (if we assume Krivit is not simply lying, and I do assume that). My headline here is a stupid, disempowering stand, that blames others for my own ignorance, but the empowering stand for a writer is to, in fact, take responsibility for the failure. If you don’t understand what I’m attempting to communicate, that’s my deficiency.

On the other hand, most LENR scientists have stopped talking with Krivit, because he has so often twisted what they write like this.

Krivit presents Shanahan’s “attempted” explanation, so I will quote it here, adding comments and links as may be helfpul. However, Krivit also omitted part of the explanation, believing it irrelevant. Since he doesn’t understand, his assessment of relevance may be defective. Shanahan covers this on LENR Forum. I will restore those paragraphs. I also add Krivit’s comments.

1. First a recap.  The Figure you chose to present is the first figure from F&P’s 1993 paper on their calorimetric method.  It’s overall notable feature is the saw-tooth shape it takes, on a 1-day period.  This is due to the use of an open cell which allows electrolysis gases to escape and thus the liquid level in the electrolysis cell drops.  This changes the electrolyte concentration, which changes the cell resistance, which changes the power deposited via the standard Ohm’s Law relations, V= I*R and P=V*I (which gives P=I^2*R).  On a periodic basis, F&P add makeup D2O to the cell, which reverses the concentration changes thus ‘resetting’ the resistance and voltage related curves.

This appears to be completely correct and accurate. In this case, unlike some Pons and Fleischmann plots, there are no calibration pulses, where a small amount of power is injected through a calibration resistor to test the cell response to “excess power.” We are only seeing, in the sawtooth behavior, the effect of abruptly adding pure D2O.

Krivit: Paragraph 1: I am in agreement with your description of the cell behavior as reflected in the sawtooth pattern. We are both aware that that is a normal condition of electrolyte replenishment. As we both know, the reported anomaly is the overall steady trend of the temperature rise, concurrent with the overall trend of the power decrease.

Voltage, not power, though, in fact, because of the constant current, input voltage will be proportional to power. Krivit calls this an “anomaly,” which simply means something unexplained. It seems that Krivit believes that temperature should vary with power, which it would with a purely resistive heater. This cell isn’t that.

2. Note that Ohm’s Law is for an ‘ideal’ case, and the real world rarely behaves perfectly ideally, especially at the less than 1% level.  So we expect some level of deviation from ideal when we look at the situation closely. However, just looking at the temperature plot we can easily see that the temperature excursions in the Figure change on Day 5.  I estimate the drop on Day 3 was 0.6 degrees, Day 4 was 0.7, Day 5 was 0.4 and Day 6 was 0.3 (although it may be larger if it happened to be cut off).  This indicates some significant change (may have) occurred between the first 2 and second 2 day periods.  It is important to understand the scale we are discussing here.  These deviations represent maximally a (100*0.7/303=) 0.23% change.  This is extremely small and therefore _very_ difficult to pin to a given cause.

Again, this appears accurate. Shanahan is looking at what was presented and noting various characteristics that might possibly be relevant. He is proceeding here as a scientific skeptic would proceed. For a fuller analysis, we’d actually want to see the data itself, and to study the source paper more deeply. What is the temperature precision? The current is constant, so we would expect, absent a chemical anomaly, loss of D2O as deuterium and oxygen gas to be constant, but if there is some level of recombination, that loss would be reduced, and so the replacement addition would be less, assuming it is replaced to restore the same level.

Krivit: Paragraph 2: This is a granular analysis of the daily temperature changes. I do not see any explanation for the anomaly in this paragraph.

It’s related; in any case, Shanahan is approaching this as scientist, when it seems Krivit is expecting polemic. This gets very clear in the next paragraph.

3. I also note that the voltage drops follow a slightly different pattern.  I estimate the drops are 0.1, .04, .04, .02 V. The first drop may be artificially influenced by the fact that it seems to be the very beginning of the recorded data. However, the break noted with the temperatures does not occur in the voltages, instead the break  may be on the next day, but more data would be needed to confirm that.  Thus we are seeing either natural variation or process lags affecting the temporal correlation of the data.

Well, temporal correlation is quite obvious. So far, Shanahan has not come to an explanation for the trend, but he is, again, proceeding as a scientist and a genuine skeptic. (For a pseudoskeptic, it is Verdict first (The explanation! Bogus!) and Trial later (then presented as proof rather than as investigation).

Paragraph 3: This is a granular analysis of the daily voltage changes. I note your use of the unconfident phrase “may be” twice. I do not see any explanation for the anomaly in this paragraph.

Shanahan appropriately uses “may be” to refer to speculations which may or may not be relevant. Krivit is looking for something that no scientist would give him, who is actually practicing science. We do not know the ultimate explanation of what Pons and Fleischmann reported here, so confidence, the kind of certainty Krivit is looking for, would only be a mark of foolishness.

4. I also note that in the last day’s voltage trace there is a ‘glitch’ where the voltage take a dip and changes to a new level with no corresponding change in cell temp.  This is a ‘fact of the data’ which indicates there are things that can affect the voltage but not the temperature, which violates our idea of the ideal Ohmic Law case.  But we expected that because we are dealing with such small changes.

This is very speculative. I don’t like to look at data at the termination, maybe they simply shut off the experiment at that point, and there is, I see, a small voltage rise, close to noise. This tells us less than Shanahn implies. The variation in magnitude of the voltage rise, however, does lead to some reasonable suspicion and wonder as to what is going on. At first glance, it appears correlated with the variation in temperature rise. Both of those would be correlated with the amount of make-up heavy water added to restore level.

Krivit: Paragraph 4: You mention what you call a glitch, in the last day’s voltage trace. It is difficult for me to see what you are referring to, though I do note again, that you are using conditional language when you write that there are things that “can affect” voltage. So this paragraph, as well, does not appear to provide any explanation for the anomaly. Also in this paragraph, you appear to suggest that there are more-ideal cases of Ohm’s law and less-ideal cases. I’m unwilling to consider that Ohm’s law, or any accepted law of science, is situational.

Krivit is flat-out unqualified to write about science. It’s totally obvious here. He is showing that, while he’s been reading reports on cold fusion calorimetry for well over fifteen years, he has not understood them. Krivit has heard it now from Shanahan, actually confirmed by Miles (see below), “Joule heating ” also called “Ohmic heating,” the heating that is the product of current and voltage, is not the only source of heat in an electrolytic cell.

Generally, all “accepted laws of science” are “situational.” We need to understand context to apply them.

To be sure, I also don’t understand what Shanahan was referring to in this paragraph. I don’t see it in the plot. So perhaps Shanahan will explain. (He may comment below, and I’d be happy to give him guest author privileges, as long as it generates value or at least does not cause harm.)

5. Baseline noise is substantially smaller than these numbers, and I can make no comments on anything about it.

Yes. The voltage noise seems to be more than 10 mV. A constant-current power supply (which adjusts voltage to keep the current constant) was apparently set at 400 mA, and those supplies typically have a bandwidth of well in excess of 100 kHz, as I recall. So, assuming precise voltage measurements (which would be normal), there is noise, and I’d want to know how the data was translated to plot points. Bubble noise will cause variations, and these cells are typically bubbling (that is part of the FP approach, to ensure stirring so that temperature is even in the cell). If the data is simply recorded periodically, instead of being smoothed by averaging over an adequate period, it could look noisier than it actually is (bubble noise being reasonably averaged out over a short period). A 10 mV variation in voltage, at the current used, corresponds to 4 mW variation. Fleischmann calorimetry has a reputed precision of 0.1 mW. That uses data from rate of change to compute instantaneous power, rather than waiting for conditions to settle. We are not seeing that here, but we might be seeing the result of it in the reported excess power figures.

Krivit: Paragraph 5: You make a comment here about noise.

What is Krivit’s purpose here? Why did he ask the question? Does he actually want to learn something? I found the comment about noise to be interesting, or at least to raise an issue of interest.

6. Your point in adding the arrows to the Figure seems to be that the voltage is drifting down overall, so power in should be drifting down also (given constant current operation).  Instead the cell temperature seem to be drifting up, perhaps indicating an ‘excess’ or unknown heat source.  F&P report in the Fig. caption that the calculated daily excess heats are 45, 66, 86, and 115 milliwatts.  (I wonder if the latter number is somewhat influenced by the ‘glitch’ or whatever caused it.)  Note that a 45 mW excess heat implies a 0.1125V change (P=V*I, I= constant 0.4A), and we see that the observed voltage changes are too small and in the wrong direction, which would indicate to me that the temperatures are used to compute the supposed excesses.  The derivation of these excess heats requires a calibration equation to be used, and I have commented on some specific flaws of the F&P method and on the fact that it is susceptible to the CCS problem previously.  The F&P methodology lumps _any_ anomaly into the ‘apparent excess heat’ term of the calorimetric equation.  The mistake is to assign _all_ of this term to some LENR.  (This was particularly true for the HAD event claimed in the 1993 paper.)

So Shanahan gives the first explanation, (“excess heat,” or heat of unknown origin). Calculated excess heat is increasing, and with the experimental approach here, excess heat would cause the temperature to rise.

His complaint about assigning all anomalous heat (“apparent excess heat”) to LENR is … off. Basically excess heat means a heat anomaly, and it certainly does not mean “LENR.” That is, absent other evidence, a speculative conclusion, based on circumstantial evidence (unexplained heat). There is no mistake here. Pons and Fleischmann did not call the excess heat LENR and did not mention nuclear reactions.

Shanahan has then, here, identified another possible explanation, his misnamed “CCS” problem. It’s very clear that the name has confused those whom Shanahan might most want to reach: LENR experimentalists. The actual phenomenon that he would be suggesting here is unexpected recombination at the cathode. That is core to Shanahan’s theory as it applies to open cells with this kind of design. It would raise the temperature if it occurs.

LENR researchers claim that the levels of recombination are very low, and a full study of this topic is beyond this relatively brief post. Suffice it to say for now that recombination is a possible explanation, even if it is not proven. (And when we are dealing with anomalies, we cannot reject a hypothesis because it is unexpected. Anomaly means “unexpected.”)

Krivit: Paragraph 6: You analyze the reported daily excess heat measurements as described in the Fleischmann-Pons paper. I was very specific in my question. I challenged you to explain the apparent violation of Ohm’s law. I did not challenge you to explain any reported excess heat measurements or any calorimetry. Readings of cell temperature are not calorimetry, but certainly can be used as part of calorimetry.

Actually, Krivit did not ask that question. He simply asked Shanahan to explain the plot. He thinks a violation of Ohm’s law is apparent. It’s not, for several reasons. For starters, wrong law. Ohm’s law is simply that the current through a conductor is proportional to the voltage across it. The ratio is the conductance, usually expressed by its reciprocal, the resistance.

From the Wikipedia article: “An element (resistor or conductor) that behaves according to Ohm’s law over some operating range is referred to as an ohmic device (or an ohmic resistor) because Ohm’s law and a single value for the resistance suffice to describe the behavior of the device over that range. Ohm’s law holds for circuits containing only resistive elements (no capacitances or inductances) for all forms of driving voltage or current, regardless of whether the driving voltage or current is constant (DC) or time-varying such as AC. At any instant of time Ohm’s law is valid for such circuits.”

An electrolytic cell is not an ohmic device. What is true here is that one might immediately expect that heating in the cell would vary with the input power, but that is only by neglecting other contributions, and what Shanahan is pointing out by pointing out the small levels of the effect is that there are many possible conditions that could affect this.

With his tendentious reaction, Krivit ignores the two answers given in Shanahan’s paragraph, or, more accurately, Shanahan gives a primary answer and then a possible explanation. The primary answer is some anomalous heat. The possible explanation is a recombination anomaly. It is still an anomaly, something unexpected.

7. Using an average cell voltage of 5V and the current of 0.4A as specified in the Figure caption (Pin~=2W), these heats translate to approximately 2.23, 3.3, 4.3, and 7.25% of input.  Miles has reported recombination in his cells on the same order of magnitude.  Thus we would need measures of recombination with accuracy and precision levels on the order of 1% to distinguish if these supposed excess heats are recombination based or not _assuming_ the recombination process does nothing but add heat to the cell.  This may not be true if the recombination is ATER (at-the-electrode-recombination).  As I’ve mentioned in lenr-forum recently, the 6.5% excess reported by Szpak, et al, in 2004 is more likely on the order of 10%, so we need a _much_ better way to measure recombination in order to calculate its contribution to the apparent excess heat.

I think Shanahan may be overestimating the power of his own arguments, from my unverified recollection, but this is simply exploring the recombination hypothesis, which is, in fact, an explanation, and if our concern is possible nuclear heat, then this is a possible non-nuclear explanation for some anomalous heat in some experiments. In quick summary: a non-nuclear artifact, unexpected recombination, and unless recombination is measured, and with some precision, it cannot be ruled out merely because experts say it wouldn’t happen. Data is required. For the future, I hope we look at all this more closely here on

Shanahan has not completely explored this. Generally, at constant current and after the cathode loading reaches equilibrium, there should be constant gas evolution. However, unexpected recombination in an open cell like this, with no recombiner, would lower the amount of gas being released, and therefore the necessary replenishment amount. This is consistent with the decline that can be inferred as an explanation from the voltage jumps. Less added D2O, lower effect.

There would be another effect from salts escaping the cell, entrained in microdroplets, which would cause a long-term trend of increase in voltage, the opposite of what we see.

So the simple explanation here, confirmed by the calorimetry, is that anomalous heat is being released, and then there are two explanations proposed for the anomaly: a LENR anomaly or a recombination anomaly. Shanahan is correct that precise measurement of recombination (which might not happen under all conditions and which, like LENR heat, might be chaotic and not accurately predictable).

Excess nuclear heat will, however, likely be correlated with a nuclear ash (like helium) and excess recombination heat would be correlated with reduction in offgas, so these are testable. It is, again, beyond the scope of this comment to explore that.

Krivit. Paragraph 7: You discuss calorimetry.

Krivit misses that Shanahan discusses ATER, “At The Electrode Recombination,” which is Shanahan’s general theory as applied to this cell. Shanahan points to various possibilities to explain the plot (not the “apparent violation of Ohm’s law,” which was just dumb), but the one that is classic Shanahan is ATER, and, frankly, I see evidence in the plot that he may be correct as to this cell at this time, and no evidence that I’ve noticed so far in the FP article to contradict it.

(Remember, ATER is an anomaly itself, i.e., very much not expected. The mechanism would be oxygen bubbles reaching the cathode, where they would immediately oxidize available deuterium. So when I say that I don’t see anything in the article, I’m being very specific. I am not claiming that this actually happened.)

8. This summarizes what we can get from the Figure.  Let’s consider what else might be going on in addition to electrolysis and electrolyte replenishment.  There are several chemical/physical processes ongoing that are relevant that are often not discussed.  For example:  dissolution of electrode materials and deposition of them elsewhere, entrainment, structural changes in the Pd, isotopic contamination, chemical modification of the electrode surfaces, and probably others I haven’t thought of at this point.

Well, some get rather Rube Goldberg and won’t be considered unless specific evidence pops up.

Krivit: Paragraph 8: You offer random speculations of other activities that might be going on inside the cell.

Indeed he does, though “random” is not necessarily accurate. He was asked to explain a chart, so he is thinking of things that might, under some conditions or others, explain the behavior shown. His answer is directly to the question, but Krivit lives in a fog, steps all over others, impugns the integrity of professional scientists, writes “confident” claims that are utterly bogus, and then concludes that anyone who points this out is a “believer” in something or other nonsense. He needs an editor and psychotherapist. Maybe she’ll come back if he’s really nice. Nah. That almost never happens. Sorry.

But taking responsibility for what one has done, that’s the path to a future worth living into.

9. All except the entrainment issue can result in electrode surface changes which in turn can affect the overvoltage experienced in the cell.  That in turn affects the amount of voltage available to heat the electrolyte.  In other words, I believe the correct, real world equation is Vcell = VOhm + Vtherm + Vover + other.  (You will recall that the F&P calorimetric model only assumes VOhm and Vtherm are important.)  It doesn’t take much change to induce a 0.2-0.5% change in T.  Furthermore most of the significant changing is going to occur in the first few days of cell operation, which is when the Pd electrode is slowly loaded to the high levels typical in an electrochemical setup.  This assumes the observed changes in T come from a change in the electrochemical condition of the cell.  They might just be from changes in the TCs (or thermistors or whatever) from use.

What appears to me, here, is that Shanahan is artificially separating out Vover from the other terms. I have not reviewed this, so I could be off here, rather easily. Shanahan does not explain these terms here, so it is perhaps unsurprising that Krivit doesn’t understand, or if he does, he doesn’t show it.

An obvious departure from Ohm’s law and expected heat from electrolytic power is that some of the power available to the cell, which is the product of total cell voltage and current, ends up as a rate of production of chemical potential energy. The FP paper assumes that gas is being evolved and leaving the cell at a rate that corresponds to the current. It does not consider recombination that I’ve seen.

Krivit: Paragraphs 9-10: You consider entrainment, but you don’t say how this explains the anomaly.

It is a trick question. By definition, an explained anomaly is not an anomaly. Until and unless an explanation, a mechanism, is confirmed through controlled experiment (and with something like this, multiply-confirmed, specifically, not merely generally), a proposals are tentative, and Shanahan’s general position — which I don’t see that he has communicated very effectively — is that there is an anomaly. He merely suggests that it might be non-nuclear. It is still unexpected, and why some prefer to gore the electrochemists rather than the nuclear physicists is a bit of a puzzle to me, except it seems the latter have more money. Feynman thought that the arrogance of physicists was just that, arrogance. Shanahan says that entrainment would be important to ATER, but I don’t see how. Rather, it would be another possible anomaly. Again, perhaps Shanahan will explain this.

10. Entrainment losses would affect the cell by removing the chemicals dissolved in the water.  This results in a concentration change in the electrolyte, which in turn changes the cell resistance.  This doesn’t seem to be much of an issue in this Figure, but it certainly can become important during ATER.

This was, then, off-topic for the question, perhaps. But Shanahan has answered the question, as well as it can be answered, given the known science and status of this work. Excess heat levels as shown here (which is not clear from the plot, by the way) are low enough that we cannot be sure that this is the “Fleischmann-Pons Heat Effect.” The article itself is talking about a much clearer demonstration; the plot is shown as a little piece considered of interest. I call it an “indication.”

The mere miniscule increase in heat over days, vs. a small decrease in voltage, doesn’t show more than that.

[Paragraphs not directly addressing this measurement removed.]

In fact, Shanahan recapped his answer toward the end of what Krivit removed. Obviously, Krivit was not looking for an answer, but, I suspect, to make some kind of point, abusing Shanahan’s good will. Even though he thanks him. Perhaps this is about the Swedish scientist’s comment (see the NET article), which was, ah, not a decent explanation, to say the least. Okay, this is a blog. It was bullshit. I don’t wonder that Krivit wasn’t satisfied. Is there something about the Swedes? (That is not what I’d expect, by the way, I’m just noticing a series of Swedish scientists who have gotten involved with cold fusion who don’t know their fiske from their fysik.

And here are those paragraphs:

I am not an electrochemist so I can be corrected on these points (but not by vacuous hand-waving, only by real data from real studies) but it seems clear to me that the data presented is from a time frame where changes are expected to show up and that the changes observed indicate both correlated effects in T and V as well as uncorrelated ones. All that adds up to the need for replication if one is to draw anything from this type of data, and I note that usually the initial loading period is ignored by most researchers for the same reason I ‘activate’ my Pd samples in my experiments – the initial phases of the research are difficult to control but much easier to control later on when conditions have been stabilized.

To claim the production of excess heat from this data alone is not a reasonable claim. All the processes noted above would allow for slight drifts in the steady state condition due to chemical changes in the electrodes and electrolyte. As I have noted many, many times, a change in steady state means one needs to recalibrate. This is illustrated in Ed Storms’ ICCF8 report on his Pt-Pt work that I used to develop my ATER/CCS proposal by the difference in calibration constants over time. Also, Miles has reported calibration constant variation on the order of 1-2% as well, although it is unclear whether the variation contains systematic character or not (it is expressed as random variation). What is needed (as always) is replication of the effect in such a manner as to demonstrate control over the putative excess heat. To my knowledge, no one has done that yet.

So, those are my quick thoughts on the value of F&P’s Figure 1. Let me wrap this up in a paragraph.

The baseline drift presented in the Figure and interpreted as ‘excess heat’ can easily be interpreted as chemical effects. This is especially true given that the data seems to be from the very first few days of cell operation, where significant changes in the Pd electrode in particular are expected. The magnitudes of the reported excess heats are of the size that might even be attributed to the CF-community-favored electrochemical recombination. It’s not even clear that this drift is not just equipment related. As is usual with reports in this field, more information, and especially more replication, is needed if there is to be any hope of deriving solid conclusions regarding the existence of excess heat from this type of data.”

And then, back to what Krivit quoted:

I readily admit I make mistakes, so if you see one, let me know.  But I believe the preceding to be generically correct.

Kirk Shanahan
Physical Chemist
U.S. Department of Energy, Savannah River National Laboratory

 Krivit responds:

Although you have offered a lot of information, for which I’m grateful, I am unable to locate in your letter any definitive, let alone probable conventional explanation as to why the overall steady trend of increasing heat and decreasing power occurs, violating Ohm’s law, unless there is a source of heat in the cell. The authors of the paper claim that the result provides evidence of a source of heating in the cell. As I understand, you deny that this result provides such evidence.

Shanahan directly answered the question, about as well as it can be answered at this time. He allows “anomalous heat” — which covers the CMNS community common opinion, because this must include the nuclear possibility, then offers an alternate unconventional anomaly, ATER, and then a few miscellaneous minor possibilities.

Krivit is looking for a definitive answer, apparently, and holds on to the idea that the cell may be “violating Ohm’s law,” when it has been explained to him (by two:Shanahan and Miles) that Ohm’s law is inadequate to describe electrolytic cell behavior, because of the chemical shifts. While it may be harmless, much more than Ohm’s law is involved in analyzing electrochemistry. “Ohmic heating” is, as Shanahan pointed out — and as is also well known — is an element of an analysis, not the whole analysis. There is also chemistry and endothermic and exothermic reaction. Generating deuterium and oxygen from heavy water is endothermic. The entry of deuterium into the cathode is exothermic, at least at modest loading. Recombination of oxygen and deuterium is exothermic, whereas release of deuterium from the cathode is endothermic.  Krivit refers to voltage as if it were power, and then as if the heating of the cell would be expected to match this power. Because this cell is constant current, the overall cell input power does vary directly with the voltage. However, only some of this power ends up as heat (and Ohm’s law simply does not cover that).

Actually, Shanahan generally suggests a “source of heating in the cells” (unexpected recombination).  He then presents other explanations as well. If recombination shifts the location of generated heat, this could affect calorimetry, Shahanan calls this Calibration Constant Shift, but that is easily misunderstood, and confused with another phenomenon, shifts in calibration constant from other changes, including thermistor or thermocouple aging (which he mentions). Shanahan did answer the question, albeit mixed with other comments, so Krivit’s “He Couldn’t” was not only rude, but wrong.

Then Krivit answered the paragraphs point-by-point, and I’ve put those comments above.

And then Krivit added, at the end:

This concludes my discussion of this matter with you.

I find this appalling, but it’s what we have come to expect from Krivit, unfortunately. Shanahan wrote a polite attempt to answer Krivit’s question (which did look like a challenge). I’ve experienced Krivit shutting down conversation like that, abruptly, with what, in person, would be socially unacceptable. It’s demanding the “Last Word.”

Krivit also puts up an unfortunate comment from Miles. Miles misunderstands what is happening and thinks, apparently, that the “Ohm’s Law” interpretation belongs to Shanahan, when it was Krivit. Shananan is not a full-blown expert on electrochemistry — like Miles is — but would probably agree with Miles, I certainly don’t see a conflict between them on this issue. And Krivit doesn’t see this, doesn’t understand what is happening right in his own blog, that misunderstanding.

However, one good thing: Krivit’s challenge did move Shanahan to write something decent. I appreciate that. Maybe some good will come out of it. I got to notice the similarity between fysik and fiske, that could be useful.


I intended to give the actual physical law that would appear to be violated, but didn’t. It’s not Ohm’s law, which simply doesn’t apply, the law in question is conservation of energy or the first law of thermodynamics. Hess’s law is related. As to apparent violation, this appears by neglecting the role of gas evolution; unexpected recombination within the cell would cause additional heating. While it is true that this energy comes, ultimately, from input energy, that input energy may be stored in the cell earlier as absorbed deuterium, and this may be later released. The extreme of this would be “heat after death” (HAD), i.e., heat evolved after input power goes to zero, which skeptics have attributed to the “cigarette lighter effect,” see Close.

(And this is not the place to debate HAD, but the cigarette lighter effect as an explanation has some serious problems, notably lack of sufficient oxygen, with flow being, from deuterium release, entirely out of the cell, not allowing oxygen to be sucked back in. This release does increase with temperature, and it is endothermic, overall. It is only net exothermic if recombination occurs.)

(And possible energy storage is why we would be interested to see the full history of cell operation, not just a later period. In the chart in question, we only see data from the third through seventh days, and we do not see data for the initial loading (which should show storage of energy, i.e., endothermy).  The simple-minded Krivit thinking is utterly off-point. Pons and Fleischmann are not standing on this particular result, and show it as a piece of eye candy with a suggestive comment at the beginning of their paper. I do not find, in general, this paper to be particularly convincing without extensive analysis. It is an example of how “simplicity” is subjective. By this time, cold fusion needed an APCO — or lawyers, dealing with public perceptions. Instead, the only professionalism that might have been involved was on the part of the American Physical Society and Robert Park. I would not have suggested that Pons and Fleischmann not publish, but that their publications be reviewed and edited for clear educational argument in the real-world context, not merely scientific accuracy.)

Shanahan’s Folly, in Color

Well, a little color. As covered in It was an itsy-bitsy teenie weenie yellow polka dot error, Kirk Shanahan digitised a chart from page 87 of Storms, The Science of Low Energy Nuclear Reaction, even though the data was on the next page in Table 7. Ah, well, you do what you need to do.

So, today, I loaded the data in to a spreadsheet, and here it is, ODS, and if you need another format, ask. The first plot shows all the data, and looks like the Storms plot, but with a little extra and without the 23.8 MeV/He line; that is equivalent to about 2.6 x 10^11 He atoms/watt-sec.

Continue reading “Shanahan’s Folly, in Color”

It was an itsy-bitsy teenie weenie yellow polka dot error

A comment today pointed out a post by kirkshanahan on LENR-Forum.

zeus46 wrote:

KShanahan. What’s that story about the time you were trying to dispute some ‘cold fusion’ findings by showing a non-correlation between two factors, but ballsed up the analysis, and ended up unknowingly proving it? Or something. Abd used to write about it. Never heard your side of it. Maybe something about a horizontal line on a graph?

In my 2010 J. Env. Monitoring paper, there is a slight error in my discussion
of a specific figure. Abd has tried to use that to discredit everything I write
in a ‘throw the baby out with the bathwater’ style. I replied to him here on
lenr-forum, but in brief… Continue reading “It was an itsy-bitsy teenie weenie yellow polka dot error”