Return to the clickable list of items
151) Experimental errors in excess heat
Ludwik Kowalski (6/30/04)
Department of Mathematical Sciences
Montclair State University, Upper Montclair, NJ, 07043
On May 25, 2004 Richard Eskimos (who prefers to remain anonymous), referring to items #116 to #119 on this web site, asked me (via e-mail): “Has there been any further recent discussion about Shanahan and Storms' papers? I'm still skeptical about cold fusion, but I have problems with Shanahan's criticism. I'd like to know if I'm the only one to give Shanahan's paper a critical look, as opposed to all the negative attention I'm sure any pro-cold fusion paper gets.” In subsequent correspondence Richard elaborated on the above. After reading Richard’s messages I asked him to summarize them in the form of a new item.
What follows is the text he composed for the readers of this web site. I am happy that the form chosen for the item is the same I used in unit #117. In fact, the first half of his essay is taken from unit #117. But then Richard focuses on aspects of the Storms-Shanahan debate that escaped my attention. Richard, who is an engineer in Canada, wrote to me that he is very interested in the cold fusion debate. He emphasized that the essay “isn't intended to address the validity of Storms' finding of excess heat, only the invalidity of Shanahan's criticism of the Storms’ experiment. I'm still very skeptical on platinum.”
In another message he wrote: “Mr. Shanahan's general discussion of systematic and random errors on your web site is certainly factual. He proposes that Mr. Storms has not analyzed the systematic errors adequately: he is probably right, and since no budget is infinite (especially Mr. Storms'), one can always make that claim about any experiment. Also, Mr. Shanahan has the advantage in a discussion about excess energy, since the burden of proof lies with Mr. Storm. However, in his paper, ‘A Possible Calorimetric Error in Heavy Water Electrolysis on Platinum’, Mr. Shanahan states that Mr. Storms has made an error in the experiment. So Mr. Shanahan now has the burden of proof to show exactly what the error is - a general claim of a systematic error because of unaccounted variables is not sufficient.”
As in unit #117, electrochemical nuances are skipped by discussing a setup with two cylinders (magic and not-magic) instead of discussing real electrochemical cells. The nature of what is going inside these cylinders is ignored. Is such an approach appropriate to discuss the Storms and Shanahan controversy? I think it is appropriate because the debate, as far as I understand it, does not focus on electrochemistry. I am glad that Richard found time to elaborate on comments made in e-mail messages. In one message he wrote; "If I were close to retiring, I would definitely be getting more involved.”
Please note that Pin (power input) refers to the rate at which electric energy is used to run a device while Pout (power output) refers to the rate at which heat is generated in that device. Without any additional input of energy the value of Pout can not possibly exceed the Pin. In a common light bulb, for example, Pin can be 60 W while Pout is close to 57 W. The positive difference between Pout and Pin, presumably detected by many scientists, is called excess heat. Dr. Edmund Storms is one of these scientists. Dr. Kirk Shanahan, also a scientist, thinks that claims of discoveries of excess heat result from calibration errors, as described in units #116 to #119. Is he right or is he wrong? The dialog between people below focuses on that question. My comments appear at the end.
= = = = = = = = = = = = = =
On Storms-Shanahan Controversy
Richard Eskimos
1) Student:
My friend, an electrical engineer, discovered an unusual resistor. It behaves normally when the current is low (below 100 mA); the electric power, Pin, and the heat generation rate, Pout, are practically identical, as they should be. But at a larger current the Pout exceeds the Pin. The engineer thinks that he has discovered a new phenomenon; some unknown exothermic process is triggered by the current in the material from which the resistor is made. Is this possible?
2) Teacher:
We do not know everything about nature; new discoveries are likely to be made anywhere and at any time. Your friend may indeed be making an important discovery but I am not willing to accept his claim on faith.
3) Student:
My friend has a calorimeter and other instruments. He said he is willing to make a demonstration. Would you be interested in seeing it?
4) Teacher:
Yes, I would.
5) Engineer (one week later):
Thanks for allowing me to use the laboratory. Everything is ready. Should we start collecting data?
6) Teacher:
I would prefer if you first tell us exactly what you want to do. Where is your “magic resistor”
7) Engineer (reaching in his pocket):
Actually, I have two cylindrical resistors, a magic and a non-magic. They are mechanically identical. The only difference between them is the composition of powders placed in small cavities near the centers of cylinders.
8) Teacher:
What is the nature of these powders? Is it possible that a chemical reaction producing heat takes place in one cylinder and not in another?
9) Engineer:
For the time being I can only say that powders are not consumed. In other words they are not chemical fuels. A qualified chemist verified this by examining powders before and after many experiments.
10) Teacher:
OK, we might ask other chemists to verify this. For the time being we are interested in the small amount of excess heat observed by you. Is it real or not? Small quantities of anything are difficult to measure because of unavoidable experimental errors. Tell us more about your experimental setup.
11) Engineer:
Two insulated wires are soldered to the terminals of my resistors. I will connect each resistor to a d.c. power supply and show that V/I is essentially constant (Ohm’s law), at least up to I=0.3 A. The R will be close to 160 ohms. The input power, Pin, will be calculated as V*I, where V is the measured voltage across the resistor.
12) Teacher:
How accurate are your voltmeter and ammeter?
13) Engineer:
According to manufacturer’s specifications the accuracy of each instrument is 0.5%. This means that the accuracy of Pin will be 1%.
14) Teacher:
And what kind of instruments will be used to measure the heating rate, Pout?
15) Engineer:
For this I have built a flow calorimeter. It is a well isolated container whose capacity is close to one liter. Water enters through the pipe at the bottom (passing through a commercial flowmeter) and exits through the pipe near the top. I will set the flow rate to about 50 cubic centimeters per minute. To measure Pout I will immerse a resistor into the calorimeter. The temperature of entering water, T1, and the temperature of exiting water, T2, will be measured.
16) Teacher:
How accurate are the thermometers and the flowmeter?
17) Engineer:
Temperatures can be measured with an accuracy of plus or minus 0.1oC. The factory-specified accuracy of the flowmeter is 1%. The container is thermally insulated as well as possible; this means that almost all of the heat (98%) is removed from it by water.
18) Teacher:
Your goal is to determine Pout. How will this be accomplished?
19) Engineer:
I will calculate Pout from the following equation: Pout=S*F*(T2-T1), where S is the known specific heat of water (4186 J/kg*C), F is the flow rate (in kg/s) and (T2-T1) is the measured difference of temperatures.
20) Teacher:
That makes sense. Will the (T2-T1) remain constant?
21) Engineer:
At the beginning T2 will be the same as T1. Then T2 will start increasing slowly while T1 will remain constant (room temperature). After about two hours T2 will stop changing and (T2-T1) will remain constant. This will indicate that in each minute heat generated and heat removed (by circulating water) are equal.
22) Teacher:
How accurate will be your determination of Pout?
23) Engineer:
The difference of temperatures will be about 10 oC. Thus the accuracy of the (T2-T1) term will be slightly less than 1%. To be on the safe side I will assume it is 1%. The accuracy at which F will be measured, as I said before, will be 1%. In other words, Pout will be known at the accuracy of 2%.
24) Teacher (addressing students):
Let me make a comment. Our guest is using the word "accuracy;" rather than the word "precision." This is correct. The word “precision” should be used when we are referring to random errors while the word "accuracy" should be used when we are referring to systematic errors. Suppose Pout turns out to be 40 W. Then, knowing that the accuracy is 2%, we would be able to say that the "true value" of Pout can be anywhere between 39.2 and 40.8 watts. Unlike random errors, which may also be present, systematic errors can not be reduced by performing the same experiment many times to obtain the average value. Let me make sure I understand you correctly. You will show that Pout is larger than Pin by the amount that is significantly larger than experimental errors. Is this correct?
25) Engineer:
Not exactly. I would do this if the difference between Pout and Pin were large, for example, larger than 10% of Pin. My approach is different. I will keep the Pin constant and compare the values of Pout for my two resistors. I will do the experiment a number of times with the non-magic resistor, and a number of times with the magic resistor. Then I will compare the results. This means that I am less concerned about the accuracy of my equipment. I am much more concerned about the stability. If my equipment has a systematic error, I want it to have the same systematic error for both resistors. My sensitivity to excess heat now depends on the stability and precision of my equipment (and on how small a change I introduce in switching resistors).
26) Teacher:
I think we should continue discussing your data, with that in mind. How large were the actual values of Pout and Pin for the two resistors?
27) Engineer:
I've kept Pin at 36.4 W for both resistors. Pout for the non-magic resistor turned out to be 35.76W. Pout for the magic resistor turned out to be 36.76W. To me it means that the difference between these two numbers, 1 W, represents the "excess power." It is the rate at which thermal energy is generated inside the calorimeter. I think that this happens through a new nuclear process. What else can it be?
28) Teacher:
I can think of many non-nuclear processes able to generate heat at the rate of 1W. But that is not the issue here. Our goal is to establish reality of excess heat (~ 1 W) and not its origin.
29) Another student, Kirk Shanahan, makes a comment:
I notice that for the experiment with the magic resistor, Pin is 36.4W and Pout is 36.76W. The difference is much less than 1W. In fact, it's 1/3W, less than 1%. I think you have no excess heat, just experimental error.
30) Teacher:
Mr. Shanahan raises a very good point, but before we get to it, I want to know a bit more about the experimental errors. Firstly, about accuracy. Is it possible that the difference of 1 W is nothing but an experimental error due to the limited accuracy of measurements? We know that the true value of Pout can be anywhere between 35.96 and 37.56 W. That is based on your 2% accuracy. Likewise, the true value of Pin can be anywhere between 36.0 and 36.8 W; this is based on your 1% accuracy. Thus, the difference between Pout and Pin, can be as small as -0.84 W or as large as 1.56 W. Do you understand my reservations?
31) Engineer:
Yes. I agree that my equipment could be giving inaccurate values of Pin and Pout. I expected it. That is why I compared a non-magic and magic resistor. If my equipment is reading Pin at 2% too low, I don't mind. I just want to be sure that the reading is 2% too low for both the non-magic and the magic resistor. My first concern was that it might read 2% too low in the morning for the non-magic resistor, and accurate in the afternoon for the magic resistor. So I repeated the run with the non-magic resistor many times, at different times of the day. I also removed and reinstalled the resistor and compared readings. I found that the Pout readings varied by 0.3% over all of these runs. Likewise, the Pin readings varied by no more than 0.1%.
32) Teacher:
From these repeats, you don't know if the readings are all about 2% too low. And as long as that inaccuracy is the same for both resistors, you don't care. Is this your main point?
33) Engineer:
Correct, as long as it is the same for both resistors, I don't need to know how far off I am.
34) Teacher:
So, after these runs, you found that your Pout readings were repeatable to 0.3%. Let's go into the details of the runs. In runs with the non-magic resistor, Pout was 35.7W: on average; an individual reading could have been as low as 35.5 and as high as 35.9W. Similarly, Pin readings could have been anywhere from 36.2 to 36.6W. So, the difference Pout - Pin was from -0.3 to -1.1W. Shouldn't it have averaged out to zero? This sounds like the non-magic resistor is not behaving perfectly.
36) Engineer:
Correct - the equipment is not perfect. I believe the non-magic resistor has Pout = Pin, but the calorimeter is imperfect. The Pout I measure is the power collected by the water from the resistor. Not all the power out of the resistor goes into the water. Some of the power is conducted up the resistor's metal leads into the room. Since it doesn't go into heating the water, I don't measure it. This is a parameter of all calorimeters called the heat collection efficiency. I measure my heat collection efficiency to be 98%. (Pout = 35.67W is 98% of Pin = 36.4W). I have to work within that practical limitation and remove any error it causes if possible.
37) Teacher:
OK, let's look at the magic resistor results. Pout was 36.76W and Pin was 36.4W. With 0.5% error, the range for Pout is then 36.58 to 36.94W. The range for Pin is still from 36.2 to 36.6W. This reminds me of Mr. Shanahan's comment. In this experiment, it could have been possible that Pout and Pin are both 36.6W. Then you would have Pout = Pin and no proof of excess heat.
38) Engineer:
True, based on the assumption that my calorimeter just achieved a heat collection efficiency of 100%. There are two problems with that assumption. One, a 100% efficient calorimeter is difficult, if not impossible, to build. Two, in my non-magic resistor experiments, I measured that my calorimeter is 98% efficient. In order for that assumption to be plausible, you would have to show that changing resistors changed my efficiency. This is not likely to occur; my resistors are essentially identical, as far as the geometry is concerned. I've made the two resistors as mechanically identical as possible and their location in the calorimeter does not change. The electric leads are of the same material and the same gauge. The heat escaping out should be the same. I've removed and replaced the non-magic resistor several times, and the heat collection efficiency did not change.
39) Student Shanahan:
I have another comment. Based on the heat capacity of water and the flow rate, I've calculated a theoretical value for the ratio Pout/Pin. I'm looking at your data for the various runs. I've calculated Pout/Pin for each run. I see that the ratios for the magic resistor runs are clustered near my theoretical value. This leads me to conclude you have no excess heat. I also see that the ratios for the non-magic resistor are clustered at a value 2% lower than my theoretical value. This leads me to conclude that the non-magic resistor experiments have a systematic flaw - I don't know what it is. Possibly all calorimeter experiments have this flaw.
42) Engineer:
Your theoretical value for the ratio Pout/Pin is calculated for equipment that is perfectly accurate and a calorimeter that has 100% heat collection efficiency. A conclusion based on such unrealistic assumption should not be taken more seriously than conclusions based on my experimental data. I did not idealized anything. Is it not obvious?
43) Teacher:
Yes. The theoretical calculation of Pout/Pin does not have a systematic experimental error. You can't compare the experimental ratio to the theoretical ratio on the 0.5% scale, only as a rough estimate of how large your systematic errors are ( which we think is 2%). On the other hand, comparing the non-magic and the magic resistor experimental values to each other allows us to ignore systematic errors which are stable. Thus the 1W difference between the values of Pout seems to be real.
44) Engineer:
I'd like to add that, if my equipment is accurate, the fact that the non-magic resistor measurements are clustered at 2% lower is simply a repetition of my previous statement that the heat collection efficiency is 98%.
45) Teacher:
Since I get the last word, I'd like to say that an experiment which compares the test of the hypothesis against an almost-identical control (or calibration) test has a sensitivity higher than the accuracy limits. Shanahan's suggestion that the magic resistor results should only be compared to a theoretical value rather than to the non-magic resistor isn't correct. The key to making the experiment convincing is to make the difference between the two experiments as small as possible. In this case, the engineer has designed the resistor to be swapped out very easily, which is an advantage. As always, the findings must be confirmed by others. Would you be willing to reveal the compositions of your powders to other scientists?
46) Engineer:
I would, but first I must protect myself with a patent. I do have a family to support; the discovery may lead to practical applications.
47) Teacher:
I understand. This is an important social issue but I do not think that we should debate it here. For the time being I would say that your claim for the discovery of excess heat is credible. Process your patent as quickly as possible and reveal the composition of your “magic powder,” preferably in a reputable scientific journal. This might become an important scientific contribution without leading to immediate practical applications.
P.S. Comments (by Ludwik Kowalski):
1) It is a pity that Richard Eskimos asked me not to show his real name. Will the pending DOE evaluation of cold fusion end the unhealthy situation in which discussing cold fusion is dangerous to young scientists? I hope it will. Declaring that cold fusion is not different from any other area of science, and that additional research in that area is welcome, is the first thing that the new DOE panel should do. In saying this I am assuming that no evidence of fraud, or incompetence (among the major players), will be discovered.
2) Let me compare the proposal presented by the inventor (Engineer) in Unit #117 with the proposal presented by him in Richard’s essay. The goal, in both cases, is to measure x=A-B, where x is the excess power while A and B are two numbers much larger than x. In the unit #117 the value of A (output power) was determined calorimetrically while the value of B (input power) was determined electrically. This approach failed to demonstrate that x is positive because of large uncertainties (systematic errors) in the values of A and B. The approach would not fail if x was several times larger, or if more accurate instruments were used to measure A and B.
In Richard’s story the inventor decided to take a different approach. The value of x is determined not by comparing A and B but by comparing two different values of A. One value, A’, is determined by using the “magic” resistor while another, A”, is determined by using several control resistors. All resistors are geometrically identical; each contains a tiny cavity at the center. In control resistors that cavity contains sand but in the active resistors it contains a sand-looking “magic powder.” This implies that heat collection efficiencies (~98%) are practically the same in all experiments.
In this new context x=A’-A”. A systematic error in measuring A’ and A” can be tolerated, as long as it is practically the same for both A’ and A”. The issue, as stated in voice #25, is no longer the magnitude of the systematic error, the credibility of the discovery depends on the stability of errors. Here is how I would summarize the issue. If the calorimetric error (about 2%, as in unit #117) changes significantly, when the magic resistor is replaced by a control resistor, then Shanahan is right. If, on the other hand, such change is very unlikely. then Storms is right. Is this the crux of the controversy?
3) Let me end with an interesting observation that Richard made when he saw the draft of this document (without my comments at the end). That observation was mostly about unit #148 but will append it here. Richard wrote:
I think the file is ready for posting. Having said that, I do have some philosophical issues with this approach (this is just for discussion purposes - don't let it prevent you from posting the file). This teacher/engineer essay is self-contained, so it re-states the essences of Storms' and Shanahan's papers - that makes it too long. At the same time, there is only room for the high points of each paper - so the essay leaves the field wide open for Shanahan's rebuttals. I don't know how to get around these two issues in a better way, if the essay is intended for a general audience.
In your previous topic, you lamented the fact that scientific papers are much more abstruse than they need to be. It's exactly for these two reasons that it happens: the authors want to cover all the possible future rebuttals, so as not to cause lengthy debates. That makes the paper long and detailed. To compensate for the extra length, they refer to previous papers and the jargon of their field in abbreviated form. That makes reading the paper laborious for anyone who is not familiar with the previous papers.
I think scientific paper writing needs to get into the computer age. A scientist wishing to be accessible could write the "main" paper as they do now, for other scientists. If she wrote it in html, she could incorporate hypetext links. Jargon could be linked to definitions/descriptions; citations to other papers could actually hyperlink to those papers (maybe even the relevant sentences in the paper); calculations (which are often "left for the student") and data files could be linked to, in all their gory detail.
(Comment: Storms' paper originally linked to his data files, and that is how Shanahan analyzed his paper so fully. I found the original data on the Wayback Machine. Without access to the data, Storms' paper is much harder to get a clear picture of.) I don't know why papers aren't written this way, except that most journals are in hardcopy form.