It is nice to see our views so closely aligned. My comments are
mainly additive.
On Sep 26, 2011, at 6:39 AM, Jed Rothwell wrote:
Horace Heffner <hheff...@mtaonline.net> wrote:
First, let me say, despite the casual, inaccurate, and one data
point nature of the method shown, it is far better than any
calorimetry applied in Rossi public demos.
I agree it is better for steam. It is the only reasonable way to
measure a mixture of steam and hot water.
.
I think running the steam and water through a condensing heat
exchanger works very well, provided *all* the flow and temperature
variables are recorded very frequently - more frequently than a
bucket test would allow, in any practical sense. The principle
expense I would expect is in accurate digitally interfaced flow
meters. It is always good to have an independent method to confirm
results and to provide confidence in control run calibrations. The
new Rossi device appears to be intrinsically dynamic - never reaching
true equilibrium. It may also have feedback instabilities. Frequent
data points are thus necessary, or a an accurate means provided to
smooth and integrate energy measurements.
I suggested that they combine this method with other methods
because this only produces one data point per test. It only works
when power is stable. So you need another method to be sure that it
is stable.
The bucket test might be a useful common sense backup check. This
could only be accurate if the device could reach equilibrium, which
the new device can not, as evidenced by the time varying duty cycle.
Also, Rossi apparently operates the device, by varying power input,
on occasion.
The steam hose appears larger than Rossi's.
Maybe, but it is difficult to judge how much steam is coming out
just by appearance.
.
Difficult to quantify accurately, but not difficult to judge. The 5
kW steam plume clearly has a much larger diameter, much greater
length, has a higher velocity, and does not require a black
background to observe.
The second temperature measurement was all over the place, due to
the inadequate stirring method. The reading varied from 29.9°C to
31.1°C.
Yup, stirring is the problem. You need to stir vigorously with a
stick or something like a Dremel tool with a paint mixer attached.
(Like a giant eggbeater.)
.
Yes, and people often do not appreciate how a small error in
temperature measurement results in a much much larger errors in delta
T measurements. The small error above indicates a possible 25% error
in the delta T measurement, and thus energy estimate.
When doing boil off calorimetry I made a small stirring device,
designed for 1 liter or less containers, that surprised me. I
attached a thin solid glass rod to a tiny cheap 1.5 V DC motor
included in a kit for use with a small solar cell. It operated well
on much less than 1.5 V. The glass rod was attached directly to the
armature using a thin rubber tubing. The rod was amazingly effective
at stirring, took up very little room in an already cramped cell
(the motor was suspended above the cell), and could be driven with
less than 1 W of power. I used a variable voltage DC power supply.
This could easily be scaled up. One neat thing about it was the rod
could be located on an angle and achieve vertical stirring as well as
cylindrical stirring. The stuff in the cell helped by creating eddy
currents.
No estimate of heat loss through the bucket was made. This means
more heat was produced than measured. It would obviously be better
to insulate the bucket.
I doubt that is a problem. It is easy to find out whether it is a
problem or not. You leave the thermometer in the bucket for 5
minutes, keep stirring, and see how much the temperature falls
every minute. I have often done this. With a bucket of this size
and water at that temperature I do not think the temperature will
fall significantly in a minute or two.
.
I think the temperature in such a bucket falls, or at least can fall,
significantly, considering a delta T measurement is being made. The
more accurate the delta T the longer the test and the bigger the
delta T, but then the more error due to heat loss unless the bucket
is insulated. Also, there is not just one calorimetry constant at
higher temperatures. There is a calorimetry function by temperature
(vs constant) due to nonlinear losses due to evaporation and
radiation. The thermal decline curve has to be measured across the
full range of temperatures that are to be observed. I know from
experience that this thermal decline curve problem is greatly reduced
by good insulation and a reduction of the surface area exposed to
evaporation. Also, the thermal decline is affected by room
temperature, so that adds yet another variable to monitor. It is
interesting that in some tests the participants wore winter clothing,
implying the environment was cold. Other tests were made in summer,
with temperatures above 30°C. If an E-cat actually puts out 10 kW of
heat I would think the place would warm up over time, and thus affect
any steam-bucket measurements.
The scale readings were very blurry in the video, but still not
consistent with the text proportionally. It appears at 0.33 in the
Mario video (see Mario0_33bucket.jpg in separate email) that the
tare was adjusted for (zeroed out on the scale), and the major
divisions are 2kg, and the next lower level division are 1 kg,
since it is stated 10 kg of water was initially in the bucket.
I do not know if the weight scale was zeroed. I recommend a digital
weight scale for this kind of thing.
Yes, indeed. Especially for videoing.
Note that this technique captures all enthalpy including the
kinetic energy of the moving steam.
- Jed
Yes, and this technique accommodates a range of pressures and
temperatures of the steam, as does a condensing heat exchanger. It
is of course necessary to properly account for the added volume of
water and the less than boiling temperature of the isoperibolic mass
in the bucket test. I provided sample calculations for doing this.
Similarly, in a heat exchanger test it is important, especially in a
dynamic situation, to monitor not only the in and out temperature
and flows of the secondary circuit, but also the output flow and
temperature of the primary circuit, in order to determine the
thermal power not extracted by the heat exchanger. The output water
flow rate will not necessarily be accommodated as an input flow rate,
and the output steam/water temperature and mass flow will vary, and
thus will also the resulting steam production in the E-cat. Given
the large thermal mass of the new E-cat the dynamics are complicated,
potentially inherently unstable if the hot water is fed back, and
even dangerous. Determination of inherent instability would require
developing the differential equations describing the circuit, or
perhaps just doing some control tests. The fact that the nuclear
effects are supposedly a function of temperature makes the situation
even more complicated and difficult to control. Then there is the
issue of the requirement for low temperature water to quench runaway
reactions? This control issue might be avoided entirely by dumping
the hot water output and feeding cold water to the E-cat as usual,
and measuring the flow and temperature of water exiting the primary
flow circuit.
Best regards,
Horace Heffner
http://www.mtaonline.net/~hheffner/
