- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
BG generator, Q and ou - 6/27/99
In earlier posts a method of generating Brown's gas was described
that focused on the value of doing AC electrolysis using insulated
plates, i.e. capacitive coupling, to a slurry comprised of metal,
semiconductor, and insulator particles in an electrolyte. The main
purpose of this post is to look at the value of this approach in
creating a high Q tank oscillator circuit.
One objective this approach is to make most of the current flow in
the cell due to a capacitive transmission, as opposed to resisitive
transmission through the electrolyte, between metal particles. For
this reason, and to limit conductivity of the electrolyte, the slurry
would be expected to benefit from ceramic particles with typically
high dielectric constants, i.e. in the range of 80 to 1200. Small
ceramic particles should also have the useful effect of often
separating conductive particles by small fixed distances. Ceramic
also offers high dielectric strength, on the order of 600 - 1250
volts/mil, and some have high abrasion resistance, thus may be a good
choice for plate insulators or cell sides as well.
There are mixed demands on the conductivity of the electrolyte. One
demand is to keep it low between conductive fragments in the slurry,
in order to reduce waste heat, and to achieve a high Q for the tank
circuit. The other demand is to keep it high across the total cell,
so as to keep most of the current out of the electrolyte and in the
capacitive coupling mode travelling through the insulator granules.
Curent carried capacitively by water, due to the its high dielectric
constant, is somewht inefficient due to the water molecule dipoles
rotating. The insulating granules in the slurry help meet both goals
by increasing the path length through the electrolyte, and by greatly
reducing the electrolyte cross sectional area.
Here is a simplistic tank circuit:
I1
--------------- V1
| |
| -------------
| | |
| | C1
AC L1 | I2
| | |
| | I3 R1
| | |
| -------------
| |
--------------- Ground
AC - AC source
L1 - Inductor
C1 - The cell capacitance
R1 - The net cell resistance
I1 - Input current (rms)
I2 - Cell current (rms)
I3 - Inductor current (rms)
V1 - supply voltage = cell voltage
Xl - Reactance of L1
Xc - Reactance of C2
When the operating frequency is at the resonant frequency fo the tank
circuit L1, C1, R1, the net impedence of the tank circuit is at
maximum is maximum to the AC source, thus the current through the
cell I2 is at a maximum with respect to the input current I1. In fact,
I2 = Q * I1 = I1
Where Q is given by:
Q = Xl/R1
and is a measure of the sharpness of the resonance peak. Since
values of Q over 100 are not uncommon in ordinary resonance circuits,
this is fascinating, and hints at ou behaviour all by itself.
If Xl is fixed, the best way to improve Q is to decrease R1. Looking
at only the tank circuit itself:
-------------
| |
| C1
L1 | I2
| |
| I3 R1
| |
-------------
this is a series resonance. In a series resonance the net impedence
of the inductor and capacitor disappears, leaving the total circuit
impedence Xt:
Xt = R1
For this reason, the more current carried within the cell
capacitively, the lower the net R1. The above is a great
simplification of the cell, which is a lattice of capacitances and
resistances, however, the more current carried capacitively, the
lower the net resistance, and the higher the cell's apparent
capacitance, thus the more the tank net impedence R1 disappears at
resonance, and the higher Q will be.
The higher Q is, the more current I2 that goes through the cell for a
given stimulating current I1. Note that V1 is both the input voltage
and the voltage across the cell, so power follows the current in
proportion.
It remains to balance the conductor concentration in the slurry so as
to get the largest number of "equivalent plates" Ne, and the smallest
possible amount of "equivalent plate separation" Se, and to match the
operating voltage so that
V1 ~= Ne * 2 volts
While the tank current is way up compared to the stimulating current,
and the voltage is the same, thus the apparent power applied is very
high compared to the input power, not all of it is being applied to
generating gas bubbles. The conductive particle concentration has to
be high in order to assure the current flows through many (Nreuse)
conductive particles, on average.
If Q and Nreuse can be made large enough, it is tempting to think the
process may have a COP over 1. This is because electrolysis is
dependent upon the amount of current through the cells conductors,
not on the power applied to the cell. However, a small flaw in this
is that a minimum voltage must be achieved in order to do the
electrolysis, so voltage does play an important role. Still, if Q
can be made large enough, the cell will go over unity.
If this mode of electrolysis approaches a high efficiency, and CF can
be made to occur in the cell and produce excess heat, then such heat
is utilized productively in the electrolysis process to make more
gas, especially if the cell is operating without an (equivalent)
overvoltage. An obvious approach to try would be to use a potassium
carbonate electrolyte with a nickel particle slurry.
A construction option worthy of consideration might be to combine the
degassing pump and cell into a single unit. The cell would then
consist of a flat insulated metal plate on the bottom, with HV AC
applied, a pump body made of an insulating material, containing both
the slurry and a rotor made of dielectric material, with no slurry
exit, a grounded circular plate on top, and a sealed hole through
the top plate containing the rotating shaft, the inside of which is
hollowed out, with lateral holes, to accomodate the gass flow out of
the cell. This is an "all in one" cell.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
More on a Brown's gas generator design - 6/26/99
Earlier I suggested the ideas of using rotating wiper blades or bubble
scrubbing particles in the electrolyte, possibly in combination with
rotating electrolysis plates to improve DC electrolysis. These concepts
were carried over into the AC world of Brown's gas (BG) generators in
the
form of a design incorporating an electrolyte slurry to which the
current
is capacitlvely coupled. The cell is driven capacitively through the
cell
sides, or through an insulating covering over the plates, eliminating
the
need for any exposed metal cell plates. The slurry then provides the
additional function of bubble former, transporter, and scrubber.
This approach has the natural benefit of permitting the cell to act as a
capacitive element in a resonant LC tank circuit. The advantages of
this
were discussed in prior posts. However, the principle advantage of
operating in resonance is that the amount of current (through the BG
generator) and apparent power in the tank circuit in relation to the
stimulating circuit, is maximized when operated in resonance. In
effect,
the impedence of the inductive and capacitive parts of the tank
circuit are
eliminated, leaving only the natural resistance of the electrolyte.
This
natural resistance is greatly diminished by virtue of the fact the
electrolyte is a slurry, with only nominal distance between "equivalent
plate" separations.
Of further interest, is the fact that the water itself has a dielectric
constant. The importance of this has been minimized by the fact that
it is
irrelevent to DC cell operation. However, it is very important to the
operation of a BG generator at high frequency in that, the effect of the
capacitive transmission of energy though the cell is large. This
tranmission mode, being in effect in parallel to the the resistive mode
while in the electrolyte, provides an alternate channel to any point
in the
electrolye for projecting energy through the electrolyte. This fact
further
highlights the possible importance of looking at use of some high
dielectric constant particles in the electrolyte slurry. The higher the
frequency the less the impedence of the water, and the more efficient
the
BG cell, to an extent. Further, if the generator operates in a tank
circuit, the current reducing effect of the capacitive and inductive
elements on impedence disappears.
The limit of efficiency is reached due to the heating effect of the
rotating water molecule dipoles. This is another reason that
insulating
dielectric particles in the slurry may improve efficiency, by reducing
dipole rotation in the electrolyte, while also reducing resistive
current
flow and providing a bubble formation site. My experiments with long
distance high voltage currents through electolytes showed bubble
formation
of the dielectric sides of the vessel, but not in the solution itself.
Providing a vast surface area for bubble formation, along with an
efficient
means of scrubbing those bubbles from the solution, is the basis of this
design. The use of high dielectric constant insulator particles in the
slurry, with or without the metal or semiconductor particles, has a
sound
basis for improving the electrolysis efficiency. Also of interest is
the
possible choice of carbon powder as a semiconductor for the slurry.
It is
lightweight, cheap, and readily available at pet and aquarium stores.
To avoid the need to search for some of the prior material, it is
included
below.
Prior post - A Brown's gas generator design, 6/17/99
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
The main purpose of this Brown's gas (BG) generator design is to
efficiently convert either mechanical action or high voltage high
frequency
AC current into Brown's gas, a stoichiometric mixture of hydrogen and
oxygen obtained from the hydrolysis of water. This design is
original, to
my knowledge, and also untested.
The primary strategy of this design is to use an electrolyte slurry
which
contains a large amount of fine metal particles, say stainless steel,
maintained in suspension by agitation, by pumping the electrolyte
through
the electrolysis system. These particles act like plates in a multi-
stage
BG generator, generating H2 on one side and O2 on the other.
The gas can be removed by running the particles through a degassifier
which
can work in a variety of ways. The degassifier and pump make a natural
combination. The turbulence and centrifugal force of a vortex in a
centrifugal pump can be used to scrub the particles of gas bubbles.
Barriers or vanes can be installed in the vortex to increase turbulence.
The gas is taken out of the pump at the center of rotation. Fluid is
introduced to the pump at a mid radius, and the blades and fluid takeoff
are to the periphery of the pump. The pump exit is diminished in
size over
normal pumps in order to increase the number of fluid rotations
within the
pump. The pump also serves as a backfire preventer, so might be
mounted in
close proximity to a torch, motor or fuel cell.
A very high cell voltage is required, thus lowering the current for a
given
design power. However, the same current goes through lots of particles,
equivalent to a multi plate cell with many plates, so the energy per
mole
of hydrogen should work out equivalently to a low voltage
electrolyser with
very small plate separation. The slurry should be comprised of a large
percentage of bubble forming particles. This generator should be
exceedingly compact for the gas generated, due to the incredible
equivalent
number of plates and the very minimal equivalent plate separation. The
electrolyte resistance per equivalent plate is thus greatly reduced,
converting less of the current into waste heat and thus applying more
directly to gas generation.
Note that, by using a high frequency, the cell is driven capacitively
through the cell sides, or through an insulating covering over the
plates,
eliminating the need for any exposed metal cell plates, and thus
eliminating cell plate corrosion. Doing this also eliminates
excessive arc
formation should a string of particles form a long short. It also
avoids
the formation of long plasma paths, and thus also reduces plasma
erosion,
and cavitation erosion, of the plate insulators, while avoiding
unproductive current surges.
The proposed design works by pumping the electrolyte slurry between two
insulated plates. A high voltage AC gradient is imposed through the
slurry. This can be done by using static metal plates on the outside of
the cell, or insulated plates within the cell, with high voltage AC
applied. Note that one of the plates can be at ground potential. A
third
insulated plate or blade in the electrolyte might be used as a
stirrer/wiper to maintain slurry dispersion and eliminate bubble
sticking.
A novel approach to applying the AC gradient across the cell is to
make one
of the plates an armature, a series of (flat) spokes mounted on an axle,
oriented so as to rotate in very close proximity to one of the flat
(outer)
sides of the cell, which is made of insulating material. The slurry in
this case is contained in a thin container which is circularly
shaped, so
as to maintain the slurry flow along the perimeter of the circle of
rotation of the armature, and between the plates. The plate opposing
the
armature, on the other flat side of the cell, can simply be a round and
grounded metal plate the same diameter as the armature. The armature
surface can be insulated, thus reducing the required thickness of the
insulating cell wall, while simultaneously eliminating corona loss
from the
armature blades. The armature is charged with high voltage DC, thus
should
require no current and thus very little energy to maintain it
electrically.
The electrolysis energy comes strictly from the armature motion.
Maintenance and construction is simplified and made less costly, by
eliminating the need to cut stainless plates, and the need to take
the cell
apart to replace plates, and by reducing plate corrosion. Using filters
and sieves, it should be possible to maintain the particle quality, even
while in operation. One frequent maintenance problem might be
erosion of
the the pump blades, but this might be reduced by choice of blade
material.
This device might be used for investigating various materials for use as
bubble forming particles in the slurry, including semiconductor
material,
and even insulating materials, and mixtures of the above. Also of
interest
for investigation are slurry particle concentrations, operating
voltages,
operating frequency, and waveforms.
Is this all just wishful thinking, something worthwhile to tinker
with, or
a good approach?
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
At 3:18 PM 6/18/99, Michael T Huffman wrote:
Hi Horace,
Sorry about not getting back to you sooner, but somebody nuked my
machine
again, AND changed the password on my server. I just now got back
online:(
I'd like to try this idea. It's completely over my head as to the
capacitative motor idea. Wouldn't this be considered an inductive
coupling,
rather than a capacitance. Like I said it's over my head.
It's capacitive for sure. There are two capacitors (e, g, the
insulating
sides) in series with the electrolyte. The moving plate coming and
going,
making its charge come and go, is the same as a current coming and going
charging a static plate. The electrolyte carries the current between
the
innermost surfaces of the two insulators. It takes a high RPM and a
very
high voltage to get much current through the electrolyte. I like the
idea
mainly because it strikes me as a novel way to convert mechanical
energy to
chemical energy. It may not have much utility. It is not an essential
part of the design. In fact, I would probably that version one unless I
was going to try to close the loop. I would probably build a capacitive
generator separately for experimental purposes initially. I like the
idea
of using a simple spark gap oscillator for a quick and dirty approach.
I think the interesting thing would be experimenting with the metal
slurrys.
Horace Heffner
http://www.mtaonline.net/~hheffner/
