Mizuno's bucket of water

Quote from THHuxleynew

Now: you state that KS CCS does not apply to McKubre, Storms, Fleischmann, Miles because they all have impeccable error bounds, and that KS has never had any issue with them.

The KS CCS does not apply to anyone. It is a fantasy. If it existed and applied anywhere, it would have been discovered 150 years ago.

But perhaps I misread what you wrote. I thought you were saying that in your opinion: "Many people just don't bother to do proper error bounds, and have some hand-wavy justification that errors << results. That usually is OK, though never good practice." Were you ascribing that view to Shanahan? Or is that what you think? If he says that, he's wrong, and if you agree, you are wrong too.

You dismiss these experiments, so I thought that was your reason.

Quote from THHuxleynew

Surely that is the opposite of what he is saying (whetehr he is right or wrong). I'm now totally confused. When KS says that these experiments do not include CCS he is exactly saying that one necessary element in the error bounding is missing.

Who cares what he is saying? It is obviously wrong. He says it applies universally. If that were right, thousands of experiments and industrial processes would fail catastrophically. A problem that can cause a 5 W error in one cold fusion experiment, and a 100 W error in another, is drastic. It would cause a megawatt error in an industrial process. Shanahan claims that he can disprove all cold fusion experiments with his theory. Since they range in power from 250 mW up to 100 W, that means his CCS can cause drastic errors.

If the CCS only applies to people with "one necessary element in error bounding missing" then it does not apply to the top tier of cold fusion experiments. There is nothing wrong with their error analysis. They were done by experts and they went through extensive, grueling peer-review, lasting years in some cases. If there was a problem with some conventional aspect of science, such as error analysis, someone would have caught it.

Perhaps the claim is that the CCS is universal, and might apply to any calorimety, but it only applies to active Pd-D experiments when high loading and the other McKubre equation parameters are met, and not calibrations. Because . . . because . . . causality does not work? High loading magically reaches out and causes the CCS? Because in the whole history of calorimetry since 1780, it just happens that no other group of people were as incompetent as the cold fusion researchers. Because starting in 1989, a whole group of world-class experts in calorimetry suddenly forgot how to do it.

I leave it to you and Shanahan to come up with a just-so story to justify this nonsense.

Jed, I respect your knowledge and ability to argue logically when it is used. In this case I have made substantive, relevant points, which you do not engage with.

But perhaps I misread what you wrote. I thought you were saying that in your opinion: "Many people just don't bother to do proper error bounds, and have some hand-wavy justification that errors << results. That usually is OK, though never good practice." Were you ascribing that view to Shanahan? Or is that what you think? If he says that, he's wrong, and if you agree, you are wrong too.

Shanahan was my reference for people more generally (he stated this, contrary to my naive expectation). However we have with Mizuno a good example locally. There is no calculated error bound for the calorimeter in the paper, in spite of much work to investigate some aspects of the errors.

You dismiss these experiments, so I thought that was your reason.

That is weird. I am more respectful of Mizuno's experiments, wanting to work out errors, than anyone else here. That is a strange use of the word dismiss. I am it seems more positive about the classic D/Pd experiments than otehrs here - wanting google to put sustained effort into investigating them, getting all possible advice from the community.

Who cares what he is saying? It is obviously wrong.

That is not a scientific argument. I have laid out why while it might not apply, the reasons posted by Alan for its non-application are incorrect. No-one has been able to gainsay my argument, because it is correct. We cannot tell how many CF experiments have results that would be explained by KS argument (in fact it could be nonce, with the semi-anecdotal exception of the results he was given privately and analysed). Therefore anyone who feels those early experiments are important in informing current views of LENR should care about them, and do the work to determine how much they affect each important experiment.

He says it applies universally. If that were right, thousands of experiments and industrial processes would fail catastrophically. A problem that can cause a 5 W error in one cold fusion experiment, and a 100 W error in another, is drastic. It would cause a megawatt error in an industrial process. Shanahan claims that he can disprove all cold fusion experiments with his theory. Since they range in power from 250 mW up to 100 W, that means his CCS can cause drastic errors.

Jed, you are repeating what you said earlier, about textbooks, in a different guise. Any competent calorimeter doing an error bound analysis will incorporate Shanahan's work. It is the maths you need. And in the few cases where it is highly significant, with the unusual conditions in CF electrolysis experiments or something equivalent, it is indeed important. They will not call it CCS, but it will be there in the error analysis. However Shanahan's beef is that the electrolysis CF experiments were not given a complete error bound analysis (because there was no term in there for errors introduced by variation in heat loss between calibration and active systems due to changed heat distribution).

I have said this, precisely and clearly, several times now. You do not engage with it, but dismiss it on spurious grounds "I don't need to consider this because if it were true 100 years of textbooks would be wrong" etc. That is pathoskepticism directed to a small wrinkle in calorimetric analysis, if you want to use the word. I'd rather not label it - just ask you to engage with this specific point which you have been avoiding.

If the CCS only applies to people with "one necessary element in error bounding missing" then it does not apply to the top tier of cold fusion experiments. There is nothing wrong with their error analysis. They were done by experts and they went through extensive, grueling peer-review, lasting years in some cases. If there was a problem with some conventional aspect of science, such as error analysis, someone would have caught it.

Someone did catch it, and published, Shanahan.

Are you telling me that every canard that has been published and peer reviewed is correct? Are you telling me that new ideas (in this case something that is not normally relevant to error bounding, but is in these cases) are not often overlooked by everyone in a field, convinced that "100 years of textbooks cannot be wrong" until some brave soul points it out? Anyway, how about taking just one error analysis that you claim is complete - i will bet you that it is incomplete wrt CCS, probably because the assumption is made that conditions cannot change without a quantitative estimate of how small a change would significantly effect the results.

if we go through this with your best shot one of two things will happen:

(1) You will understand that this analysis, not always done, is important, because your heroes did it.

or

(2) I or Shanhan will explain why it has not been done.

Perhaps the claim is that the CCS is universal, and might apply to any calorimety, but it only applies to active Pd-D experiments when high loading and the other McKubre equation parameters are met, and not calibrations. Because . . . because . . . causality does not work? High loading magically reaches out and causes the CCS?

Glad you ask this, it has been answered many times. In the case of CF electrolysis the reason would be that the necessary active electrode environment that allows ATER is specific to exact conditions, materials, loading, etc. Just as elusive as a claimed LENR NAE.

Over 400 posts of passionate debate about the significance of an anecdote. I guess this is what people mean when they talk about advancing science.

Quote from interested observer

Over 400 posts of passionate debate about the significance of an anecdote.

It was not an anecdote. It was a carefully observed, well documented example of heat after death. Heat after death was later observed hundreds of times. With so many replications, there is no doubt this event was real. The method of measuring this enthalpy is crude but extremely reliable and convincing.

Quote from THHuxleynew

Shanahan was my reference for people more generally (he stated this, contrary to my naive expectation). However we have with Mizuno a good example locally. There is no calculated error bound for the calorimeter in the paper, in spite of much work to investigate some aspects of the errors.

I calculated it from the spreadsheet, and posted a graph showing the source of my calculation. The noise is about 0.1 deg C which translates into 2 W. Take it or leave it. But you are evading the issue. "People more generally" means nothing. The top tier of cold fusion papers covers this in great detail. Despite that, you and Shanahan claim that every single study in this field is invalid. You have no reason. You point to Mizuno to show that Storms and McKubre are wrong. That's pathological skepticism.

Quote from THHuxleynew

Glad you ask this, it has been answered many times. In the case of CF electrolysis the reason would be that the necessary active electrode environment that allows ATER is specific to exact conditions, materials, loading, etc. Just as elusive as a claimed LENR NAE.

And as I pointed out, excess heat is observed when there is no electrolysis, so obviously it cannot be an artifact of measurements made during electrolysis.

Also, the control parameters that cause excess heat are well defined, and they cannot be linked to Shanahan's crackpot hypothesis.

Quote from JedRothwell

It was not an anecdote. It was a carefully observed, well documented example of heat after death. Heat after death was later observed hundreds of times. With so many replications, there is no doubt this event was real. The method of measuring this enthalpy is crude but extremely reliable and convincing.

Calling it an anecdote is not questioning whether it was a real event. It is distinguishing it from a scientific experiment. If you prefer, I will say that there have been 400+ posts about an observed event. And, no doubt, there will be many more as the interested parties continue to dissect it.

Quote from JedRothwell

I calculated it from the spreadsheet, and posted a graph showing the source of my calculation. The noise is about 0.1 deg C which translates into 2 W. Take it or leave it. But you are evading the issue. "People more generally" means nothing. The top tier of cold fusion papers covers this in great detail. Despite that, you and Shanahan claim that every single study in this field is invalid. You have no reason. You point to Mizuno to show that Storms and McKubre are wrong. That's pathological skepticism.

Jed, you need to pay attention to the details here. Neither Shanahan nor I claim every study is invalid. We claim that every study to which CCS might apply (that is most of them, but not HAD ones) needs an error analysis including the affect of changes in cell conditions between active and control runs, without which we cannot know if it is valid, and that this error analysis needs to take into account the fact that system condition change errors apply to all the input power, and so are magnified relative to the excess power. Maybe that is much smaller than the claimed excess. Shanahan says (and you could easily contradict it) that this has not been done. Marwhan et al don't do it, and incorrectly claim it is not necessary to do so.

Shame.

CF experiments in which Pin >> (Pout - Pin) exhibit particular subtle calorimetric challenges (apologies - I read this from a CF discussion somewhere - I'd hope you would agree this is recognised).

Storms reviews calorimeters, and error sources, here, but alas does not explicitly address how to mathematically bound errors. For example he says that calibration constant shift due to changed heat distribution in mass flow calorimetry is small but does and dismisses this, but does not do the quantitative work (from his example) to check whether, in fact, it might be significant. That is Shanahan's point - don't dismiss errors, even if you think they are small, before you determine mathematically what is their effect.

I point to Mizuno as an example I (we all) remember. If you give me one specific documented experiment from Storms or McKubre and highlight the error analysis I will show you why that too is deficient, or show it to you as an example of correct errors bounding including CCS.

Reiterating that I am pathologically skeptic does not help this communication.

If you are others show that this correction is negligible I'd have sympathy with your position. But you don't, you argue incorrectly that it does not exist.

I'll lead you through the maths if you like (v simple algebra), except I suspect you would pay no attention unless it was in a textbook.

Your error bound for Mizuno apparatus above includes only the noise and not the errors? Do you know what "calculating error bounds" means?

We must hope that is not typical of LENR error bounding.

Quote from interested observer

Calling it an anecdote is not questioning whether it was a real event. It is distinguishing it from a scientific experiment.

This distinction is meaningless. In any case, this was the last phase of a scientific experiment, conducted by two PdD scientists, at a National University, wearing white lab coats. How much more scientific can you get?

Your objection is puerile. You will next say it was an anecdote because it was not repeated. It was repeated, hundreds of times, by Fleischmann and Pons. Also, many experiments and observations are not repeated, such as the Top Quark finding. Many cannot be repeated, such as super nova and earthquakes.

Quote from interested observer

If you prefer, I will say that there have been 400+ posts about an observed event.

Observed, measured, graphed, logged, and shortly thereafter described in Japan's most prestigious general circulation magazine, although not in a scientific journal.

Quote from THHuxleynew

Jed, you need to pay attention to the details here. Neither Shanahan nor I claim every study is invalid.

Okay, list one study that you do not dismiss.

Quote from THHuxleynew

We claim that every study to which CCS might apply (that is most of them, but not HAD ones) needs an error analysis including the affect of changes in cell conditions between active and control runs.

The CCS is a figment of Shanahan's imagination. There is not a shred of evidence for it. It only happens magically with Pd-D that meets McKubre's conditions. The alternative explanation, that calorimetry works and this is real heat -- is far more likely.

If heat after death experiments do not need this theory, and if they produce real heat, then the heat is real. That ends the debate. It is not plausible that the heat is caused by a CCS error for a week before the boil off, then for some other cause during the boil-off (such as your magical invisible drops of water that defy the conservation of energy), and then with a third cause in heat after death. That violates Occam's razor, to say the least.

Of course you will say the heat after death results can be explained by some other theory. Don't bother! We know that is what you think. We don't need to hear any more magic invisible droplette hypotheses.

Quote from THHuxleynew

We claim that every study to which CCS might apply

Unfortunately neither THHnew or Shanahan have ever verified their claim with experiment..

The THHnew-Shanahan CCS hypothesis born as a primitive conception in an obscure journal in 2001

will forever remain theoretically pure,

unsullied by citation by analytical chemists or other over 18 years

or by experimental verification..in all that time

even though it has broad application in most of analytical chemistry... according to KS

“Your objection is puerile.”

What objection? Your kneejerk attack is utterly predictable, however.

Quote from interested observer

“Your objection is puerile.”

What objection?

Let me rephrase. Saying this can be distinguished "from a scientific experiment" and it is a mere anecdote is a puerile dismissal. An anecdote is not an event witnessed, measured, graphed and described in detail by two professional scientists. Heat after death is a replicated phonomenon. You have no basis to dismiss it, and no reason to doubt Mizuno's account.

Quote from interested observer

Your kneejerk attack is utterly predictable

This is an anecdote,,,

The statistics on the mean and variability of Jed's kneejerks are not referenced.

Dear all: mods + replicators.

I don't want to go flogging a dead horse here. If there is anyone out there doing experiments and needing to evaluate small excess heat findings (so that Pin >> (Pout - Pin)) then understanding how to do an error analysis and why change in conditions in the cell here gets magnified is vital. Please PM me or like this post - I'll start a new thread explaining how to do error bounds and exactly why a small error caused by change in cell conditions gets amplified when Pin >> (Pout - Pin). For those mathematically minded it is pretty obvious.

If people are only in it for R20 class results, where calorimetry does not matter, this is irrelevant.

In the case of R19, the results here are on the borderline of not clearly being higher than possible errors. For my own satisfaction I'll maybe do an error bound analysis, with individual error values taken from conclusions from the Mizuno thread (which others may agree with or change).

THH

Quote from THHuxleynew

I'll maybe do an error bound analysis

What happened to the THHnew error bounds on Mizuno's anemometry? 64%,, 20% ... 10 % Did they get thrown out of the window?

Quote from THHuxleynew

I don't want to go flogging a dead horse here.

I take that as your non-answer to this exchange:

You: "Neither Shanahan nor I claim every study is invalid."

Me: "Okay, list one study that you do not dismiss."

You are going to bug out and refuse to address this, as I knew you would. You have no actual reasons to dismiss any major study, so you sometimes pretend you don't dismiss them, but more often you evade the issue.

Quote from THHuxleynew

If there is anyone out there doing experiments and needing to evaluate small excess heat findings (so that Pin >> (Pout - Pin)) then understanding how to do an error analysis . . .

Then I would suggest that person read a legitimate textbook on calorimetry, rather than wade through crackpot theories that have never been tested, without evidence, that make "predictions" of events indistinguishable from real excess heat.

Quote from THHuxleynew

In the case of R19, the results here are on the borderline of not clearly being higher than possible errors.

Oh right. Sure. In a calorimeter that has been calibrated dozens of times, that never fails to show a heat balance of zero during calibrations, and that can clearly measure 2 W or better, results ranging from 38 to 108 W are "not clearly higher than possible errors." Yeah. Errors which never show up in the calibrations, by some fantastic coincidence. Yes, 38 W or 108 W is close to 2 W, if you squint and play "let's pretend that numbers don't mean anything."

Okay, list one study that you do not dismiss.

I do not dismiss any study - but where I can't see error bounds smaller than the results I don't consider results show anomalous excess heat.

For example (keeping things simple) I have not seen this bounding done for R19 - though I think we know enough to do it. Your 2W error bound is wrong, as I will explain on another thread if people would be interested. It is complex because there are two separate bounds for absolute and relative to control results.