Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
"Bun is such a sad word, is it not?" -- Watt, "Waiting for Godot"

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Isothermal bubble pump heat engine

Highly efficient heat engine with no valves and no solid moving parts at all.
  (+2, -1)
(+2, -1)
  [vote for,

[As requested, here are answers to "where the heat comes from, why this is a superior manner in which to harness said energy"]

Heat to operate this device comes from any pair of hot and cold temperature reservoirs. For example, it could operate between the condenser of a steam power plant and the cooling towers, absorbing heat from the condenser and delivering 90% of it to the cooling towers and 10% of it as work to a generator. Or it could operate in places where Stirling engines are used now, with solar collectors or using landfill gas. Or it could operate in a place like Death Valley, taking heat from the ambient and delivering it via a heat pipeline up a nearby mountain. A heat pipeline is just a large gravity-feed heat pipe, say a 60 inch diameter pipe carrying refrigerant vapor up the mountain and a smaller pipe carrying condensed liquid refrigerant back down.

It should be superior to existing Stirling engines because it doesn't have solid moving parts (so is more efficient) and is inherently larger scale. Also it ought to do a better job of keeping the working gas at constant temperature.

How does the heat energy get converted to work? See the paragraph about molecules of gas. Gas molecules collide with slugs of liquid accelerating them and slowing (cooling) the gas molecules. The slow molecules collide with hot tubing walls and get accelerated, raising the temperature, while the atoms in the walls vibrate less, reducing the temperature.

[revised description, of a slightly different machine, and I will be revising it further.] Injecting bubbles into moving fluid May 7th, 2006 by archimerged

The isothermal bubble pump is very simple, but not so flexible. There is no control over the spacing or size of bubbles once they are created when the machine is manufactured. So I propose a machine made of two hyperbolic segments of bubble pump with injectors at the inputs and collectors at the exits. The injectors are under automatic control. One injector feeds cold low-pressure gas bubbles into the downward moving fluid stream at the nearly horizontal top of the cold compression hyperbola. The other feeds hot high-pressure gas bubbles into the upward moving fluid stream in the nearly vertical bottom of the hot expansion hyperbola.

Alternating slugs of fluid and gas flow through the two hyperbolas, but elsewhere in the machine, flows of fluid and gas are separate. Heat is exchanged between high and low pressure fluid using a fluid-only countercurrent heat exchanger. Since there is no gas involved, a U tube easily brings the hot low-pressure fluid down to the level of the high-pressure fluid and returns the cold low-pressure fluid to the original level. Incidentally, inside the heat exchanger the fluids have equal hydrostatic pressure; importantly, the pipes are in direct thermal contact.

Separately, heat is exchanged between high and low pressure gasses in a gas-only countercurrent heat exchanger which can be placed anywhere in any orientation. Typically, it would comprise a series of insulated low pressure tanks of descending temperature, filled with coils of high-pressure pipes flowing in the opposite direction so cold gas enters the coldest low pressure tank, absorbs some heat, flows to the next warmer low pressure tank, absorbs more heat, etc. The volume of these tanks is unimportant so long as very little heat is lost and there is very little net heat flow in or out either end. The amount of heat gained by the high pressure gas should be essentially equal to the heat lost by the low pressure gas. The more tanks provided, the longer the gas stays in the tanks, and the larger the volume (so the relative surface area is lower) the better this goal is achieved. A single heat exchanger can serve many pairs of bubble pump hyperbolas.

This machine does not require a jump-start because the bubble injectors and collectors are inherently irreversible, and it cannot function as a heat pump unless the temperatures of the heat reservoirs are reversed. The liquid always flows down the compressor and up the expander, heat always flows from heat source to heat sink, and the motor always runs forward, but if the heat source is colder than the heat sink, forcing the motor to turn forward will pump heat from the cold source to the hot sink. Forcing the motor to turn backward will just pump fluid around the circuit with no gas flow.

The bubble collectors are simply vertical tubes leading to gas reservoirs. These tubes have such large diameters that the bubbles detatch from the walls of the tube and float freely (and irreversibly) upward. The gas pressures are always high enough to prevent fluid flow upward, so fluid never fills the gas reservoirs or flows into the gas-only tubes above the reservoirs. Hot fluid leaving the nearly horizontal upper end of the hot expansion hyperbola takes a sharp bend downward to the hot input port of the fluid countercurrent heat exchanger, but the bubbles escape upward into a short vertical pipe leading up to a gas reservoir. Cold fluid and bubbles arriving at the bottom of the cold compression hyperbola flow around a sharp bend from nearly vertical to horizontal. After a short horizontal segment, the bubbles escape up into the high pressure gas reservoir while the fluid flows to the cold input port of the fluid countercurrent heat exchanger.

Hot gas from the hot low-pressure reservoir and cold gas from the cold high-pressure reservoir flow slowly through the gas countercurrent heat exchanger and into the cold low-pressure reservoir and hot high-pressure reservoir. From there, the hot high-pressure gas feeds the bubble injector at the nearly vertial bottom of the hot expansion hyperbola, and the cold low-pressure gas feeds the bubble injector at the nearly horizontal top of the cold compression hyperbola.

A positive displacement pump somewhere in the fluid-only circuit converts fluid flow to rotary motion or rotary motion to fluid flow, always in the forward direction (the pump cannot run backward). A motor-generator converts electricity to or from shaft motion.

Thinking of gas at the level of moving molecules, one can imagine how and why the expansion hyperbola provides motive force to the liquid. At the bottom where the gas bubbles are injected, the hydrostatic pressure is so high that the bubbles are just big enough to touch the walls of the tube, and they are not expanding. But the fluid is moving upward in the nearly vertical hyperbolic tube, and the hydrostatic pressure drops as the amount and weight of the fluid above decreases. So the gas expands, doing work on the slug of liquid behind and ahead. Molecules of gas collide with the liquid, accelerating the liquid and decelerating the gas molecules, so the temperature and pressure drop. Molecules of gas also collide with the hot walls of the tube, gaining energy so the temperature and pressure of the gas rise. The faster heat flows into the gas, the more work it can do on the liquid and the faster the liquid moves. Nearer the top of the hyperbola, the slope of the tube decreases, so that a larger increase in volume is permitted without reducing the pressure so much. The goal is isothermal expansion: the pressure-volume product is constant and all of the energy removed from the hot walls of the tube is used to accelerate the moving fluid while the gas stays at constant temperature.

Thus, if no energy were being extracted from the fluid motion, the fluid would continue to accelerate. This is generally true of heat engines — they speed up as long as heat (microscopic disordered kinetic energy) is supplied faster than energy is removed, storing the excess as macroscopic ordered kinetic energy. Here, the moving fluid plays the role of a flywheel.

[original description follows]:

A bubble pump comprises a tube filled with alternating slugs of liquid and gas. The tube must be narrow enough so that surface tension keeps the liquid slugs completely isolated from one another. Preferably, the liquid should not wet the walls of the tube.

The isothermal bubble pump heat engine is an amazing device. It comprises a hot heat source, cold heat sink, many identical cleverly shaped loops of sealed tube containing alternating slugs of a heavy liquid and the working gas threaded through the heat source, heat exchanger, and heat sink, and finally, a means for extracting energy from the motion of the liquid. For example, magnetic particles suspended in the liquid would interact with an electromagnetic field, transferring energy to the field as the expanding gas does work on the slugs of liquid. An applied electromagnetic field is also required to start the flow any time the engine is stopped, and if the engine is operated in reverse, it will pump heat from the cold sink to the hot source.

The slugs of fluid and gas move as rapidly as the heat flow and energy extraction permits, and the amount of energy extracted is expected to be close to the maximum possible given the temperature difference between the heat source and sink, allowing for losses to friction, parasitic heat flow, temperature drops between the sources and the gas, etc.

The bubble pump loops take the shape of the heat engine's PV diagram (with pressure increasing downward): the height of the tubing in the ambient gravitational field is proportional to the pressure of the slugs of gas at a given location, and the volume available for a given slug is greater as the tubing is more horizontal.

It is necessary that the slugs of liquid be larger than the slugs of gas, or else the relationship between height and pressure will not be uniform, but this can be allowed for in the shape of the loops.

The isothermal expansion and compression segments form hyperbolas. The expansion segment will be at higher pressure. The heating and cooling segments must be inside a countercurrent heat exchanger. For isobaric heating and cooling, the segments are horizontal but the gas volume does not necessarily vary as the horizontal position. Vertical segments would not generally cause constant volume heating and cooling, but as expected the pressure would not be constant.

Does this really produce isothermal expansion and compression? Not precisely. A more complicated mechanism with reservoirs and a means for feeding alternating slugs of gas and liquid into the isothermal tubes would operate efficiently under a wider variety of conditions, but with some control over the heat flow, this machine will operate efficiently, and it has an appealing simplicity.

Now to build one…

No doubt someone wants a diagram. I think in words, usually. So a word picture:

Isothermal compression starts in a tube which is almost horizontal and slopes down to the right. The slope increases as the depth increases. The slugs of gas are at maximum volume for their temperature, which is cold because the isothermal compression tube is in thermal contact with the cold heat sink. The gas and liquid is moving at a reasonably constant velocity to the right and down. Farther down, the slugs of gas decrease in volume, because the liquid behind is being accelerated by gravity while the liquid ahead is being decelerated by the higher pressure gas ahead of it. The gas temperature increases and heat flows into the cold heat sink. If the temperature tends to continue rising (because the engine is running too fast for the cold heat sink to carry away heat) then the pressure will increase more and tend to slow down the engine. The hyperbolic shape of the curve will result in compression occuring as fast as possible without increasing the temperature.

Isobaric (constant pressure) heating occurs in a horizontal segment of tube at the maximum depth and maximum pressure of the engine. Ideally, the heat is obtained via countercurrent heat exchange with the isobaric cooling segment, but that segment operates at minimum pressure, at the top of the machine. Gravity feed heat pipes will not accomplish the job: capillary action heat pipes are needed. Countercurrent heat exchange is important because the heat capacity of the liquid may be substantial, and that heat should not be coming from the heat source or ending up in the heat sink, but should stay in the hot portion of the liquid.

After the gas and liquid are heated to the hot temperature, the isothermal expansion step begins. The liquid and gas slugs move suddenly upward and to the left, at a steep slope which decreases smoothly to a nearly horizontal path before connecting to the isobaric cooling unit. This phase produces the force which keeps the liquid and gas moving against friction and the electromagnetic load applied by the external field to the magnetic particles in the liquid, and of course, against the resistance of the compression step. As the weight of the liquid above the gas decreases, the applied pressure decreases and the gas expands. It does work on the slug of liquid ahead of it and behind it. The energy for this work comes from the heat absorbed in keeping the gas at constant temperature.

Does the isothermal bubble pump need a check valve in the circuit? With one, it shouldn't need to get a push start, but then we couldn't say the machine has no valves. Inertia serves a similar function to the check valve. When expanding gas warms back to the isothermal temperature, it applies additional force to the slug in front and the slug behind, decelerating the slug behind and accelerating the slug ahead.

Archimerged, May 06 2006

Archimedes Submerged http://archimerged.wordpress.com/
See also posts on my blog regarding isothermal expansion and compression [Archimerged, May 06 2006]

Renewable Energy Design http://renewableene...le_pump_heat_engine
Isothermal bubble pump heat engine page on Renewable Energy Design wikia. Anyone can edit. [Archimerged, May 07 2006]

info on heat pump with Bubble Pump http://www.me.gatec...andy_phd/index.html
a machine that actually works and its proven [jhomrighaus, May 08 2006]

An existing bubble pump http://www.bubbleac...bble_pump_works.htm
This bubble pump pumps hot water from a solar collector through a heat exchanger using water vapor bubbles. The bubble pump is straight, not curved, so the water makes a fountain instead of flowing smoothly. The description attributes the pumping action to hydrostaic pressure, not to expansion of the gas. Since there isn't any heat entering the gas as it expands, it isn't isothermal and indeed most of the pumping action comes from the hydrostatic pressure. [Archimerged, May 08 2006]

Bubble Pump Design and Performance [pdf] http://www.me.gatec...rgy/SusanThesis.pdf
M.S. Mechanical Engineering Thesis by Susan Jennifer White, Georgia Institute of Technology. 77 pages. Oriented toward pumping refrigerant. Experiments with vertical air-lift pump. No heat flow into the pipe; all lift is due to hydrostatic pressure. No experiments in which the tube is not vertical. Used air and water, not air and mercury. [Archimerged, May 08 2006]

Hydropower plant costs http://hydropower.i...s/plant_costs.shtml
Hydroelectric plant capital cost $1700 to $2400 per kW. Operation and maintenance $0.007 per kWh. [Archimerged, May 09 2006]


       I'd love to comment, but I'm busy. Could you summarize this in 200 words?
sninctown, May 06 2006

       who do you know can summarise this in 2,000 words - prat!
po, May 06 2006

       Summary: Invest a lot of resources in building an efficient heat engine, and it produces enough energy for enough years to pay back the investment, not to mention that the other easy sources of energy have already been done and demand keeps increasing. This heat engine is like a hydro plant, only you don't have to flood millions of acres to collect the energy.
Archimerged, May 07 2006

       //many identical cleverly shaped loops of sealed tube containing alternating slugs of a heavy liquid// ... got to be a +
xenzag, May 07 2006

       So this is a free energy machine?
Galbinus_Caeli, May 07 2006

       the gas and liquid, how are they kept in discreet slugs? Oh surface tension and narrow tubes, ge that should eliminate any frictional losses.   

       Another great theoretical experiment.   

       [galbinus], [archimerged] has a history of idealized physics oddities that are overly complex and seem to produce huge amounts of work for minute inputs.
jhomrighaus, May 07 2006

       [Galbinus_Caeli] It is no more free energy than a windmill, a hydro plant, or any machine that uses heat to produce motion. But yes, you can operate it without fuel.   

       [jhomrighaus] Have you looked at a mercury manometer? Mercury and gas stay nicely separate. If you make a U tube with a series of slugs of mercury and air, and excite the harmonic motion so it flows back and forth, it takes quite a while to decay. That is not a sign of lots of friction.   

       Bubble pumps work.   

       You asked me to get rid of the valves. I did. Now what do you want?   

       This machine is not complex at all.   

       The machine produces an appropriate amount of work for the heat input: for 270K to 300K and around 97% mechanical efficiency, 100 joules of heat in from the heat source, 5 joules of work out, 95 joules of heat out to the heat sink. But unlike a toy Stirling engine, this machine scales up.   

       This time I will edit the description to make improvements. The original text is on the blog and the wikia.
Archimerged, May 07 2006

       so build one and demonstate that it works and will solve the worlds energy problems. Post on a legitimate engineering site and get input from qualified professionals then provide a link so we can all see.
jhomrighaus, May 07 2006

       I posted here because I got useful questions on the earlier incarnations of this idea, which helped lead to this version. I hope another useful discussion will occur. If you are not interested, just don't read it.
Archimerged, May 07 2006

       Can I suggest a re-write? Basic concepts first (e.g. where the heat comes from, why this is a superior manner in which to harness said energy etc.) THEN get into the hardcore details.   

       I know a reasonable amount about the subject you appear to be interested in, however I find your ideas almost unreadable as they are currently written. I think you will get a better response from people if you modify your text so as not to be such an ordeal to read.   

       Thanks in advance.
Texticle, May 07 2006

       See link for lots of information and they built a prototype after haveing calculated and demonstrated performance.
jhomrighaus, May 08 2006

       That device was invented by Albert Einstein and Leó Szilárd in 1926. I got the idea of using a bubble pump from a description of that refrigerator, which I read about a year ago. Their purpose was to build a refrigerator, not to extract work from available temperature differences. They use ammonia, hydrogen, and water, not mercury and air, and there are phase changes involved.   

       [xenzag] The device I describe is certainly not "over unity." It absorbs heat and puts out maybe 5 or 10% of the heat as work and the rest as heat at a lower temperature. It is only unusual in that because it uses only liquid and gas for moving parts, it has very low friction and is able to extract some work from a smaller temperature difference than usual. I have deleted the Zero point energy link as completely irrelevant.
Archimerged, May 08 2006

       Thanks for the re-write.   

       My understanding of bubble pumps is that they are not particularly high pressure machines. They are essentially an improvement on simple thermosiphoning, used where thermosiphoning is insufficient, and where mechanical pumps are unwarranted or undesired e.g. solar hot water.   

       Based on that alone I would be skeptical as to the economics of generating electricity this way. Power available is of course pressure drop across the turbine multiplied by the volumetric flow rate. Consider the sort of flows required to generate anything 'worthwhile'. If that is difficult to visualise, consider torque and rpm of the turbine.   

       Simply saying that the machine can be 'scaled up' neglects significant construction and operational costs that I believe would vastly outweigh the revenue gained from electricity generation.
Texticle, May 08 2006

       Ordinary bubble pumps do not heat the gas to counteract the drop in temperature as the gas expands. Also, they use a vertical tube, and the bubbles have to be able to start the flow from a standstill. And they have to work using the liquid that needs to be pumped. Here we choose the best liquid -- probably mercury -- to eliminate inefficiency and complicated flow patterns. Slugs of fluid move at the same rate as slugs of gas without leaking.   

       The power produced is about 10% of the heat power flowing from heat sink to heat source. Bigger heat collectors and heat sinks, more power. Bigger temperature difference and the 10% figure increases.   

       The compressor hyperbola can carry all of the liquid, and 10% more moles of gas than the expansion hyperbola. In that case, the additional moles of compressed gas are returned to low pressure through a device which extracts work from gas flow.   

       Or, the compressor hyperbola can carry only 91% of the liquid and 100% of the gas the expansion hyperbola carries. Then 9% of the liquid returns to the lower tank through a device which extracts work from fluid flow.   

       Even if this conversion of gas or liquid flow to electricity is only 30% efficient, the device still operates. It would keep running even if the conversion was only 3% efficient. Inefficient conversion to electricity does not stop the machine, because the electricy is not in the closed cycle.   

       Hydropower construction costs can be $2,000 per killowatt. See the link for cost comparisons.   

       The useful lifetime of the mercury is infinite, and this machine would also serve to keep all of the mercury in one place where it stays out of trouble. So you can't include the cost of mercury in the 50 year capital cost.   

       Why would operating costs be much different from hydro plant costs?
Archimerged, May 09 2006


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle