http://cgi.ebay.com/Thermal-Vacuum-Spac ... 53e4fa6605
A vacuum chamber for the 100MW Polywell demonstrator?
A vacuum chamber for the 100MW Polywell demonstrator?
Someone is selling a beautiful vacuum chamber. I wonder if it's over priced or if the people at Polywell should be bidding.
http://cgi.ebay.com/Thermal-Vacuum-Spac ... 53e4fa6605
http://cgi.ebay.com/Thermal-Vacuum-Spac ... 53e4fa6605
Its probably overkill, and not quite the right shape.
When I designed the chamber WB6 ran in, we arranged to have it built by a company that build large space simulation chambers. They had one under construction that was so large they could have misplaced ours by putting an end bell of the other one over it. The big chamber was a swiss cheeze structure, mounting dozens of liquid helium cryopumps where most chambers would just have walls. Those things are designed for 1e-12 torr, about 4-5 orders of magnitude below what we needed.
Big space chambers are really expensive to transport. The one mentioned above was about 24 ft in diameter. To get it down a road you had to move in the middle of the night, with a couple of cranes. You used the two cranes to lift the chamber over each bridge, as there was no way to pass under. You almost had to deliver to some place near major water as getting one inland was prohibitive.
The industry that uses them need to test $150 million and up satellites, usually a series of them, so they don't mind the price.
When I designed the chamber WB6 ran in, we arranged to have it built by a company that build large space simulation chambers. They had one under construction that was so large they could have misplaced ours by putting an end bell of the other one over it. The big chamber was a swiss cheeze structure, mounting dozens of liquid helium cryopumps where most chambers would just have walls. Those things are designed for 1e-12 torr, about 4-5 orders of magnitude below what we needed.
Big space chambers are really expensive to transport. The one mentioned above was about 24 ft in diameter. To get it down a road you had to move in the middle of the night, with a couple of cranes. You used the two cranes to lift the chamber over each bridge, as there was no way to pass under. You almost had to deliver to some place near major water as getting one inland was prohibitive.
The industry that uses them need to test $150 million and up satellites, usually a series of them, so they don't mind the price.
chris - the difference of 1 order of magnitude (your numbers vs mine) is the difference between success and failure.chrismb wrote:I'm glad you are expressing some experience of how much pumping is required on a large chamber - and not forgetting that a reactor polywell 'core' will be pumping out billions more litres/s of helium to get MW power that are a pumping burden on top of just 'chamber' pumps.
Engineering is the art of making what you want from what you can get at a profit.
viewtopic.php?p=18484#18484
Can anyone [yet] provide me a link to billion litre/s pumps?I think this flux of 3E21 alpha particles would require a pumping speed of 120 billion litres per second to maintain a 1E-9 torr vacuum. That's about 50,000 olympic sized swimming pools of volume - per second. Those will sure be some cool vacuum pumps!!.... Where [on earth] does one buy billion litre/sec vacuum pumps?? Whoever makes them is gonna have a lot of business because we'll need over a hundred of them for each 500MW BFR!
My number is one order of magnitude less which I calculate to be possible.chrismb wrote:viewtopic.php?p=18484#18484
Can anyone [yet] provide me a link to billion litre/s pumps?I think this flux of 3E21 alpha particles would require a pumping speed of 120 billion litres per second to maintain a 1E-9 torr vacuum. That's about 50,000 olympic sized swimming pools of volume - per second. Those will sure be some cool vacuum pumps!!.... Where [on earth] does one buy billion litre/sec vacuum pumps?? Whoever makes them is gonna have a lot of business because we'll need over a hundred of them for each 500MW BFR!
Engineering is the art of making what you want from what you can get at a profit.
I don't think our figures vary, excepting that I was looking at a 500MWe unit and assuming operating pressure was E-9. I think we were still ummm'ing and arrr'ing over what the actual operating pressure would be. Whichever way it swings, I think we agree billion l/s is needed to begin to cover it - whether more or less it is a needed starting point for pumping a 500MWe device.
ITER/toks, at MW power, work by cooling the gases at the divertor. This brings the local pressure around the divertor to oom over the rest of the chamber, and it is at that point that the gases are sucked out, from around the divertor.
This is actually a really neat feature of large toks - the gases that travel around the SOL to the divertor tend to be the heavier stuff, 4He, which is what you wanna suck out.
But I still see similar problems - hydrogen (1H) will begin to contaminate the system, and the lower the A, the harder it is to pump them. Toks have not even got into an operating regime where they can see if a build up of reaction products will contaminate. If ITER does get to the point of emitting 500MW of neutrons for minutes at a time, then maybe they'll start to see a population of decayed neutrons reentering the reaction volume. I think, though, that this will be one of the least of the probles when running long pulses, because the magneto-sonic disruptions from the 4He product will be worse than the issues of trying to pump it. But... just my speculations really, as per polywell.
ITER/toks, at MW power, work by cooling the gases at the divertor. This brings the local pressure around the divertor to oom over the rest of the chamber, and it is at that point that the gases are sucked out, from around the divertor.
This is actually a really neat feature of large toks - the gases that travel around the SOL to the divertor tend to be the heavier stuff, 4He, which is what you wanna suck out.
But I still see similar problems - hydrogen (1H) will begin to contaminate the system, and the lower the A, the harder it is to pump them. Toks have not even got into an operating regime where they can see if a build up of reaction products will contaminate. If ITER does get to the point of emitting 500MW of neutrons for minutes at a time, then maybe they'll start to see a population of decayed neutrons reentering the reaction volume. I think, though, that this will be one of the least of the probles when running long pulses, because the magneto-sonic disruptions from the 4He product will be worse than the issues of trying to pump it. But... just my speculations really, as per polywell.
chris,
Bussard was a pretty canny designer. I think he chose 100 MWth for a reason. All my calculations (heat load on the grids for one - before we found out that with a high enough field it would be "minimal") bear that out. When it was assumed that the alpha load was based on area intercept the heat load was right at the edge between conventional and exotic cooling - 1 MW per m^2 IIRC.
Also Tom Ligon (who should know) says the operational vacuum rqmts are on the order of 1E-6 to 1E-7 torr. Start up was on the order of 1E-8 to 1E-9.
So roughly 1/10th the power out and 2 orders of magnitude less of vacuum takes it from a billion to a million. Doable.
Of course those are just estimates. Reality may be better or worse.
I really can't imagine the US Navy missing a fundamental show stopper like that.
Also note that 50 MW is a very good number per shaft for a ship. A destroyer with 2 shafts requires two Polywells. A carrier requires 4 Polywells + what ever is needed for catapults and aux. eqpt. So 100 MWth Polywells are just what the Navy would order.
The other gain is that with direct conversion you don't need turbines and long shafts. Just wires. Which also means cross connecting reactors is a lot easier. Just reroute the current. It also means more flexibility in ship layout.
As has been noted here before - thermal plant accounts for 80% of power plant cost. The savings might not be quite as great because you do need conversion gear - 2 MV DC to 13 KV variable frequency AC. Plus a motor. So there is that. Still. The total plant should come in at around 50% of what a conventional nuke plant costs. The fuel is for all practical purposes is "free".
Also if you would go back Bussard suggested retrofitting old power plants with as many 100 MWth Polywells as was required to get full plant output when it comes to electrical power plants.
Bussard was a pretty canny designer. I think he chose 100 MWth for a reason. All my calculations (heat load on the grids for one - before we found out that with a high enough field it would be "minimal") bear that out. When it was assumed that the alpha load was based on area intercept the heat load was right at the edge between conventional and exotic cooling - 1 MW per m^2 IIRC.
Also Tom Ligon (who should know) says the operational vacuum rqmts are on the order of 1E-6 to 1E-7 torr. Start up was on the order of 1E-8 to 1E-9.
So roughly 1/10th the power out and 2 orders of magnitude less of vacuum takes it from a billion to a million. Doable.
Of course those are just estimates. Reality may be better or worse.
I really can't imagine the US Navy missing a fundamental show stopper like that.
Also note that 50 MW is a very good number per shaft for a ship. A destroyer with 2 shafts requires two Polywells. A carrier requires 4 Polywells + what ever is needed for catapults and aux. eqpt. So 100 MWth Polywells are just what the Navy would order.
The other gain is that with direct conversion you don't need turbines and long shafts. Just wires. Which also means cross connecting reactors is a lot easier. Just reroute the current. It also means more flexibility in ship layout.
As has been noted here before - thermal plant accounts for 80% of power plant cost. The savings might not be quite as great because you do need conversion gear - 2 MV DC to 13 KV variable frequency AC. Plus a motor. So there is that. Still. The total plant should come in at around 50% of what a conventional nuke plant costs. The fuel is for all practical purposes is "free".
Also if you would go back Bussard suggested retrofitting old power plants with as many 100 MWth Polywells as was required to get full plant output when it comes to electrical power plants.
Engineering is the art of making what you want from what you can get at a profit.
I'm happy to go with these figures. They make more sense. Still, 100 million litres/s Blimey!MSimon wrote:chris,
Bussard was a pretty canny designer. I think he chose 100 MWth for a reason. All my calculations (heat load on the grids for one - before we found out that with a high enough field it would be "minimal") bear that out. When it was assumed that the alpha load was based on area intercept the heat load was right at the edge between conventional and exotic cooling - 1 MW per m^2 IIRC.
Also Tom Ligon (who should know) says the operational vacuum rqmts are on the order of 1E-6 to 1E-7 torr. Start up was on the order of 1E-8 to 1E-9.
There is a paper;
http://iopscience.iop.org/1742-6596/114 ... 012013.pdf
which seems to indicate that the pumps for the main vessel are 300m^3/s [0.3Ml/s], though it also mentions cryopumps attached to the neutral beam injectors (which presumably have a big emissions of gas due to the main fraction that is not ionised and accelerated into the beam that needs to go back into the ion source?? maybe??). These NBI pumps are 3Ml/s. They must be enormous! And we're saying a 100MW polywell (30MWe?) needs 30 of them.
I would think that the bulk of the alphas could be directed into a higher pressure extraction zone. With clever geometry and perhaps by using steering magnets, the vacuum issue could be improved. Brute force is the simplest conceptually, but why not use some creative thoughts.
Best,
mike
Best,
mike
Counting the days to commercial fusion. It is not that long now.
Chris,
My worry has always been the contaminant gas, mostly hydrogen, evolved from the metal surfaces. The early machines could not keep up with it. Hopefully that dropped some orders of magnitude with the later configurations (WB6 and 7 form factors).
If the fuel is diluted, the reaction rate will drop percipitiously, although pulsed operation might still be possible. But pulses between pumpdown would seriously degrade margin, possibly resulting in pulses of net power but overall a useless powerplant.
The ultimate pump is to operate in deep space, but even there you have some amount of structure around the reactor (how else you gonna recover power from it). Thus, even in a near perfect vacuum, there are conductance limits.
All else considered, if strong source of p-B11 alphas can be coaxed out of it, they'll pretty much pump themselves out of the center of the device. The accelerated fuel will also tend to knock anything slow out of the active region (not without corresponding losses, of course). I do worry about what happens after all this stuff hits the walls and generates more stuff. Some of the collector schemes (venitian blinds, etc) may offer some opportunity to aggressively pump near the walls at higher pressures.
But yes, I've always considered the pumping problem to be a potential killer of this scheme as a CW powerplant.
My worry has always been the contaminant gas, mostly hydrogen, evolved from the metal surfaces. The early machines could not keep up with it. Hopefully that dropped some orders of magnitude with the later configurations (WB6 and 7 form factors).
If the fuel is diluted, the reaction rate will drop percipitiously, although pulsed operation might still be possible. But pulses between pumpdown would seriously degrade margin, possibly resulting in pulses of net power but overall a useless powerplant.
The ultimate pump is to operate in deep space, but even there you have some amount of structure around the reactor (how else you gonna recover power from it). Thus, even in a near perfect vacuum, there are conductance limits.
All else considered, if strong source of p-B11 alphas can be coaxed out of it, they'll pretty much pump themselves out of the center of the device. The accelerated fuel will also tend to knock anything slow out of the active region (not without corresponding losses, of course). I do worry about what happens after all this stuff hits the walls and generates more stuff. Some of the collector schemes (venitian blinds, etc) may offer some opportunity to aggressively pump near the walls at higher pressures.
But yes, I've always considered the pumping problem to be a potential killer of this scheme as a CW powerplant.