M.C.H. McKubre*, F.L. Tanzella* and V. Violante** * SRI International, Menlo Park, California. * * ENEA Frascati, Rome, Italy.

Transcription

1 The Significance of Replication. What have we learned in 18 years of experiments performed at SRI, about the experimental conditions for success and reasons for failed replication? M.C.H. McKubre*, F.L. Tanzella* and V. Violante** * SRI International, Menlo Park, California. * * ENEA Frascati, Rome, Italy. Presented at the American Physical Society, New Orleans, March 10, Notes by M. C. H. McKubre 1

2 1989 Faced with a Series of Unanswered Questions Q1 Is there unexplained heat? Q2 Is the heat output sensibly correlated with inputs? Q3 Is the heat derived from a nuclear process? Q4 Nuclear ash correlated with the excess heat? Q5 Are their other nuclear effects? Q6 What is the nuclear process? Q7 What is the future? At the beginning one might have posed this set of questions. This would have helped, if more people had done it in a sequential process. There is no point jumping ahead until you are sure of the basis. The particle physicists began with Question 4. Theorists began with Question 6. MITI, with their NHE project, began with Question 7. 2

3 1992 Q1 Unexplained heat source? YES!!! Effect Evidenced on numerous occasions Typical P xs 3-30% (±0.5%) of Total P in (340%) Up to 90 observation of excess power effect Duration several hours to 1 week (>70 at SRI) Sustained, unidirectional heat burst exhibit an integrated energy at least 100 times greater than conceivable energy storage effects Heat production observed for over half the operation time of one cell (C1). Similar heat production observed using 4 different calorimetric methods. YES!!! Emphatically: Bold, with 3 exclamation points. By 1992 I had reached the 99+% conviction level that there was an unexplained, nuclear level, heat source in the D-Pd system. 3

4 1995 Q2 Sensibly correlated with inputs? YES!!! Necessary conditions: Maintain High Average D/Pd Ratio (Loading ) For times >> x D/D (Initiation) At electrolytic i > mA cm -2 (Activation) With an imposed D Flux (Disequilibrium) Heat correlated with: - electrochemical current or current density - D/Pd loading - V ref. surface potential - Pd metallurgy - Laser stimulus For 1mm dia. Pd wire cathodes: P xs = M (x-x ) 2 (i-i ) i D x = D/Pd, x ~0.875, i =50-400mA cm - 2, id =1-10 ma cm - 2, t >200 D/D This heat effect also sensibly correlates with plausible input variables. The bottom function is plotted in Slide 13. The critical thing learned in 1995 was the role of interfacial D flux - which - incidentally - Peter Hagelstein had predicted. 4

5 Hermetic 10- pin Connector Gasket PTFE Top Plate Quartz Liner Electrolyte PTFE Liner Gas Tube Containing Catheter Screws Catalyst RTD Recombination Catalyst in Pt Wire Basket PTFE Spray Separator Cone PTFE Cap Pt Wire Anode SRI Quartz Calorimeter and Degree of Loading (DoL) Cell Pd Cathode Stainless Steel Outer Casing Quartz Anode Cage PTFE Base Cell used in 90% of our degree-of-loading and early calorimetry studies. Important feature: control of impurity sources and distribution. 5

6 Water In Acrylic Toppiece Gas Tube Exit to Gas-handling Manifold Inlet RTD's Water Out Hermetic 16-pin Connector Gasket Water Outlet Containing Venturi Mixing Tube and Outlet RTD's Acrylic Flow Separator SRI Labyrinth (L and M) Calorimeter and Cell Hermetic 10-pin Connector Stainless Steel Dewar Gasket PTFE Plate Catalyst RTD Screws Recombination Catalyst in Pt Wire Basket PTFE Spray Separator Cone Quartz Cell Body PTFE Liner Pd Cathode Brass Heater Support and Fins PTFE Ring Quartz Anode Cage Heater Pt Wire Anode Acrylic flow restrictor PTFE Ring Stainless Steel Outer Casing Locating Pin Stand 99.3% Thermal efficiency. Only the remaining 0.7% needs independent calibration (for high accuracy). Also, only this 0.7% can drift (but it did not). The cell shown in the previous slide goes inside the labyrinth. Two or 4 of these objects are placed in a constant temperature bath (±0.003 K) in a constant temperature room (±1 K - on a good day). 6

7 0.7 P13/14 Simultaneous Series Operation of Light & Heavy Water Cells; Excess Power & Current Density vs. Time I (A/cm^2) Pxs D2O (W) Pxs H2O (W) Heavy water works - light water does not. These two cells were cells operated at the same time, with the same current source (in series), and interrogated with the same measuring instrumentation. 7

8 0.6 P13/14 Simultaneous Series Operation of Light & Heavy Water Cells; Excess Power vs. Current Density P14/D2O Linear P13/H2O Electrochemical Current Density (A/cm 2 ) Same data showing the effect of current density: approximately linear above a non-zero threshold. The scatter is due to departure of the calorimeter from its steady state. At no time have we ever observed a steady state endothermic anomaly. 8

9 C1: Excess Power vs. D/Pd McKubre et al (similar to Kunimatsu et al) ICCF3, Nagoya. 6 C1 Parabolic Atomic ratio (D/Pd) Parabolic (or asymptotic) dependence on average loading above a (rather high) threshold value. 9

10 R/R T1-2,OHF1-3 2 L1, 2, 7, Excess Power vs. Maximum Loading (1) 2 17 (9, 6) 15 L12 T3-4 P4, L11 P3, P5, P6, L4, P16 P8 - P11 C1, P2, L3 AS2 AS1.1 C2, P14, P21 P15 AS1.3 P12, 20 Baranowski P P22 P Max. D/Pd~0.725 P1 1.1 (No heat, Heat) Figure 1 Maximum loading, D/Pd, attained in experiment; determined by R/R. Top left quadrant Max D/Pd < 0.9 => no heat excess Bottom right quadrant Max D/Pd > 0.95 => all heat excess Middle zone 50 : 50 We have done a lot more and I have only one anomalous point which did produce tritium but not measurable heat. 10

11 Q3 Is the heat of nuclear origin? Yes! 100 s to 1000 s of ev s / Pd (D) atom SRI 2076 ev/pd, Energetics >4000 ev/pd Sustained, unidirectional heat burst exhibit an integrated energy at least 10 times greater than the sum of all possible chemical reactions within a closed cell Heat effects are observed with D, but not H, under similar (or more extreme) conditions McKubre et al, Development of Advanced Concepts, EPRI, TR (1994) The heat is too large to be explained by chemistry. It is too big to be storage (and no time to store and no missing endotherms). And it works with D not H. This is (at least) circumstantial evidence that we should be thinking of nuclear effects. 11

12 2000 Q4 Nuclear ash correlated with the excess heat? Q5 Uncorrelated nuclear products? Yes! Compelling Evidence: 4 He closely time and quantity correlated with excess heat 3 H observed in some cases only. Not quantity correlated with excess heat ( ~ 3-4 O.M. down) Isotopics effects possibly at very low level Charged particles:, p + possibly at even lower level Neutrons not observed at SRI (although they can be found using more sensitive detectors at ~10 or more O.M. down from heat) Yes! It is important to mention that SRI was only replicating helium and tritium results obtained much earlier by others. 12

13 M4: Excess Energy - expectation function [Closed, He-leak tight, Mass-Flow Calorimeter, Accuracy ±0.35%] Measured Values Predicted Values Current Burst 1 P xs = M (x - x ) 2 (i - i ) Žx/Žt x =.833, i =.425, r= % Burst Time (hours) Electrochemical Current Density (A cm-2) Blue are data points Green is prediction function from slide 4. r=0.853 is the cross correlation function between blue and green. 73% is the probability that these two curves are linearly correlated. The reason for the drop between Bursts 1 & 2 was primarily due to a (spontaneous) change in the flux of D across the interface. This was for 1 mm dia. Wires. We have checked this function out a lot recently and it seems to work for Vittorio Violante s foils with superwave stimulation (although the current threshold is much lower). 13

14 5 4 M4: Excess Energy - 4 He Correlation [Closed, He-leak tight, Mass-Flow Calorimeter, Accuracy ±0.35%] 62±7% Period of cathode thermal and compositional cycling 69±8% 104±10% Measured Values Helium Helium [ 4 He] (ppmv) 1 Expected from D+D -> 4He + 24 MeV Time (hours) 0 Note: expected values decrease because of withdrawal of samples for analysis with high [4He] being replaced with stock D2 containing 0.34±.007 ppmv 4He. This was done in order to keep the internal pressure above ambient. Our idea was that the missing 40% must have been absorbed (or somehow stuck) very close to the Pd surface - maybe in a junk layer, and that we could get it out by sloshing D back and forth. Given the slope I am not sure the compositional cycling did anything - the 4He may have shown up anyway. 14

15 To Mass Spectrometer Sample Volume Vacuum/Pressure Gauge To Digital Pressure Guage Pressure Transducer Extrel C-50 QMS Thermowell Containing Gas Phase and Solid Phase Thermocouple Sensors Sample Vacuum ppm Helium in Deuterium Calibration Mixtures 4% He 4% D 2 ln Ar Tune-up Mixture Carbon Trap (LN 2 cooled) Nupro 50cc 316SS Sample Flask Solid Insulation Helically Wound Heating Elements Nupro 50cc 316SS Sample Flask Case cell Studies: H 2 and D 2 Gas with Pd/C Catalyst Catalyst 1 Liter Stainless Steel Dewars Vessel 1 Vessel 2 Vessel 1 H2, Vessel 2 D2: 3 Atm. and 200 C. Conflat (Cu) seals - helium leak tested and tight. Calorimetry was: Differential (comparison of T measured in the two beds compared to input heater power), and Gradient assuming a linear gradient between bed - gas - and ambient. This is not ideal calorimetry (hence the uncertainties in the next slide). But the two methods agreed pretty well. 15

16 Gradient Case: Q -Value - Energy vs. 4 He Differential y = 18.36x R 2 = 0.99 y = 18.89x R 2 = 0.95 Gradient Q = 31±13 MeV/atom Differential Q = 32±13 MeV/atom Helium Increase (ppmv/v) Ambient helium = 5.22 ppm. Highest measured in Case experiment = 10.8 (±.01) ppm. The expectation value (24 MeV) is inside the uncertainty - but I am more inclined to believe that we had the same 4He retention issue as in the previous electrochemical result (in this case probably in the C). We simply did not wait long enough. 16

17 Present Q6 What is the nuclear process? Primary product 4 He with ~24 MeV/ 4 He Relevant theory under construction: Hagelstein, Chubb 2, Preparata, etc. 17

18 Future Q7 What is next? Research consortia: e.g. SRI/MIT/NRL/ENEA/Energetics Technical development: > 10 x Heat Out / Power In Positive Temperature Coefficient? Time for Engineering?? 18