I think you need to look up the definition of feasible. It does not include "that thing that we have never successfully deployed even at the scale of a pocket square", nor "spend billions of dollars to get something we can scoop up off a dry lake in Bolivia".
> if we'd kept funding at the anticipated rate in the '60s, we might have working fusion already
This canard gets passed around in every thread on the topic. It is wrong. Period.
The prediction in question was made in the 1970s when everyone was *absolutely sure* that the latest generation of designs would reach break even. These included, among others, MFTR and TFTR. The prediction, then, was that by developing production versions of these designs one would have a commercial machine, and you could make reasonable predictions on how much those would cost to develop.
Except when they turned on TFTR, it didn't work. Instead, it exposed another group of previously unexpected problems. Some of these also applied to MFTR, which was then shut down shortly after (physical) completion. At that point it was clear that neither was a pattern for a commercial machine, and no one had any idea how much it would cost to fix it, let alone develop a commercial model from those fixes.
ITER represents those fixes. The answer, then, is that it costs about 25 times as much as TFTR to get a machine that produces the results TFTR was supposed to - and that's assuming ITER doesn't fail. Applying the same scaling factors to ITER that they applied to TFTR means the production machines will cost on the order of hundreds of billions to develop.
Meanwhile, in the last few decades the PV researchers, working on perhaps 1/100th the amount of investment, decreased the cost of PV about 200 times. If installed in sunny locations, like Nevada, it produces power for about 4 cents, making it (inflation adjusted) one of the cheapest forms of power in history. When faced with this reality, fusion supporters have retreated to various arguments about why you can't possibly use these sources for X and Y, in spite of people already using them for X and Y all over the world.
No they aren't. All modern pathways look promising until they tried to scale them up. And both of these are subject to the non-equilibrium braking radiation problem, which appears insurmountable.
> but underfunded
The problems with fusion cannot be solved by throwing money at that, that makes the problem *worse*. The problem is not actually physics, its economics, and we already know fusion is not competitive.
> It also could end up much worse, with either shadowing, module mismatch
Well the nice thing about utility scale is that you have some control over this. Shadowing? Call in the bulldozers and level the ground. Module mismatch? Demand the supplier stack them on the pallets in matched groups.
You and I don't have the same advantages, the pallet you get will be +/- ~3% and you get to install on whatever square you have. That said, I did get lucky - my 2 story garage gets shadowed only about 2% of the year, about an hour a day in late winter.
> Interesting blog, BTW.
Thanks! Not a lot of posts this year though, got a new kid.
> Something will need to sit idle or at least underutlized to provide peak load
Yeah, but you make those plants be ones that are OK doing that. Nuclear power is not OK throttling (well, some are).
It's no coincidence that Ontario built the western world's largest coal plants while building out their nuclear fleet.
And it's also no coincidence they are replacing all the coal plants with gas peakers while retiring the nuclear fleet and slowly replacing it with wind.
> The performance ratio of such a power plant could be around 85%, with cable losses, inverter losses and automated cleaning
A VERY minor nit-pick - modern utility scale systems are closer to 90% because they have far lower line losses and inverters with ~98% efficiency. Apple's system (in WV? I can't recall) was specced at 92%
I realize that has no real effect on the bottom line, but I did think it was worth putting out modern numbers.
Maintenance costs are part of OPEX. OPEX on PV is the lowest of any major power source, by far. No moving parts, your main component is a sheet of glass (they are SILICON cells) in an aluminum frame, and the power conversion doesn't even have a transformer any more. And no fuel, of course.
Do you really think it would be more expensive than, say, a nuclear plant? Nuclear fuel is very cheap, but you still have lots (and LOTS) of moving parts to contend with. And cheap fuel isn't as cheap as no fuel at all.
679/648 = $1.05/Wp. That's actually slightly higher than new installs in the US and Europe, but I suspect that's simply because construction started some time ago.
PV is now the cheapest form of power in CAPEX terms, ever. Which is why ~60 GWp is going in this year.
No. This is just wrong. Completely. You need to make your brain unlearn this.
Here is a toy model you can use on your journey... Think of the photon as a cheshire cat. You cannot see the cat, if you try to perceive it completely, it will vanish. You can, however, ask it questions. If you ask it "what is your gizifa", it will say "10". Or in this case, you can ask "what is your momentum", and it might say "5".
Asking certain questions will upset the cat and cause it to change all the other values just to piss you off. So if you ask it what its momentum is, the answer you *might* have got for its gizifa will now change. These values also change on their own over time. So even if you know that its location is 2,7 now, when you ask it again later you will get a different answer. No, this is not *because you asked* (another common misconception), this is inherent to the way the cat works. Some of these values are conserved (like electric charge), others are not (like location) and others are linked together (like momentum and location).
Photons are not particles. There are no particles. "Particles" is the term we use when we refer to these things when you keep asking them what their location is. If you do that, they will give you nice answers like 2,7 and then 3,7 and then 4.7, and you'll go "oh, this thing is travelling along positive X, and it's a point, so it must be a particle!". But the problem is that if you ask it different questions, like its position and the location, then any semblance of particle-like behaviour will vanish. You were fooling yourself, ITS NOT A PARTICLE. Neither is an electron or a proton, or anything else. They're just quanta. It's all quanta.
> It has some pressure when it shines on an object
This is also incorrect.
Newton thought momentum had something to do with mass because he only had large objects to work with. Shotputs have a lot of momentum, and so do planets. But in the "real world" of quantum, momentum is just a number. It's a number like any other, like energy. It's not related to mass. You ask a quanta a question and it will give you an answer. If you ask a photon its mass it will say zero. And if you ask it its momentum it will say 5. These questions are orthogonal, they don't have anything to do with each other.
So why does it LOOK like momentum has something to do with mass? Because the momentum of any one quanta is tiny, so in order to be measurable at macro scales, you need a WHOLE LOT OF QUANTA. It's very easy to make a ball of protons and electrons, because they attract each other. So you put a bunch together and call it a shotput and notice that it has a lot of momentum. But the fact that it has a lot of momentum isn't because it has a lot of mass - it has a lot of mass AND momentum because it *has a lot of quanta*.
It is much harder to make a big ball of photons, they don't attract each other. If you did such a thing, you'd find it had just as much momentum as a ball of matter, but still has no mass.
> and given off an infinite amount of mass/energy
No. Energy is also a measurement of the same sort, it's just a number you can ask for. It has nothing to do with "speed". Some of these values you ask for are more interesting than others because they are concerned, but that's not due to quanta, that's due to the shape of the universe.
> t is uncommon for a lithium-ion battery to withstand more than 1,500 charges before it fails
Bologna. My iPhone 5S is over three years old and still has ~65% charge at the end of a day of use.
> can store a large amount of energy
The paper is behind a paywall, but thanks to Sci-Hub I could read it. It focuses entirely on power density, not energy density, but does have some comparative information in Chart F. According to that, the best-case scenario for this device is about 0.07 Wh/cm^3. A modern li-po is about 0.5, around ten times the energy capacity of this device. That is actually less than some other designs, which have approached li-po but only on microscopic scales.
So don't hold your breath, this is not a device that you will see any time soon.
> wealthiest company on the planet > But holy shit is your comment re: worrying overblown
Profitability appears to have little to do with lasting success. Perhaps the opposite. In spite of being the most profitable company in the computer space in the early 1990s, the entire market imploded almost overnight. I don't relish watching history repeat.
> Probably because the AirPort Express hasn't been updated since 2012 and the > AirPort Time Capsule hasn't been updated since 2013
Indeed, which is perhaps a catch-22. I would suggest the new MBP demonstrates that. More broadly, Apple's attention only to margin is precisely what almost killed them in the 1990s, and its very VERY worrying to see history repeat itself.
I got mine in 2012, and at the time there was nothing like it - a pretty fast ac router with automated backups, printer bridging, simple setup (on the Mac, admittedly) for a few hundred bucks. But then I made the list of things it didn't do:
1) in spite of being relatively fast in general, it was built on the same "premium router" platform as other designs, but it was slower than them 2) while it claimed to offer PC file sharing, the software essentially did not work and could not be used in production 3) it handled printer sharing through port mapping, but did not do the same for scanners. all this needed was a *&^%$% TWAIN driver 4) it has four gigE ports, but exposes only three for some reason 5) a single USB port is not enough if you want to add a backup drive or second printer. the guts had four ports, so why not expose them?
The good news is, like all Apple products, it's pretty bulletproof, so I won't need to replace mine any time soon.
I hate to break it to you, but I've read texts from the Pax Romana era describing how to do this. People have been replacing labor with capital basically forever.
"There was never a job opening for a drone pilot until there was something to fly":rolleyes:
There was something to fly before drones, airplanes. And they had one or more crew per aircraft. In comparison, drones have multiple aircraft per "pilot", and that pilot has far less to do.
" Tesla's Solar Roof Will Cost Less Than a Traditional Roof"
Traditional *expensive* roofs.
Which is fine, don't get me wrong, but if you're expecting this to compete with a $3000 re-roof using asphalt, not going to happen. If you do have a home with something more expensive, then the issue there is that they do tend to last longer and won't be a target for replacement as often.
> smack in the middle of the cold war and nuclear exchange always looming
Ahhh, the good old days.
Want to shit yourself? Read the MX basing report. You can find it online. One option they considered was putting the MX on trains, and when things got nasty, sending the trains out into the rail network. The review board has a calculation that shows the Sovs could bomb every square inch of the US rail network and still have lots of warheads left over to kill everyone.
Sure, get someone else to pay for their education so they can take all your ideas away. That sounds like a formula for rather rapid irrelevance.
Or, get someone else to pay for their education so all the smartest people in their country come and stay in yours. That sounds like a very good idea to me.
Which is fine, because if this does happen they'll all come to Canada instead. I'm still thanking Shrub for being the best president ever when he implemented the snowflake law and instantly created a $1 billion stem cell economy in Toronto. Go Trump!
Slashdot Mirror
> how in the world did they download the fix
I just typed in an IP I know was unused, and 4.2.2.4 for DNS. Downloading now.
And I'm in Canada.
> The only currently feasible ... is solar sails
I think you need to look up the definition of feasible. It does not include "that thing that we have never successfully deployed even at the scale of a pocket square", nor "spend billions of dollars to get something we can scoop up off a dry lake in Bolivia".
> colleague gave me recently
You work in the comic book industry?
> if we'd kept funding at the anticipated rate in the '60s, we might have working fusion already
This canard gets passed around in every thread on the topic. It is wrong. Period.
The prediction in question was made in the 1970s when everyone was *absolutely sure* that the latest generation of designs would reach break even. These included, among others, MFTR and TFTR. The prediction, then, was that by developing production versions of these designs one would have a commercial machine, and you could make reasonable predictions on how much those would cost to develop.
Except when they turned on TFTR, it didn't work. Instead, it exposed another group of previously unexpected problems. Some of these also applied to MFTR, which was then shut down shortly after (physical) completion. At that point it was clear that neither was a pattern for a commercial machine, and no one had any idea how much it would cost to fix it, let alone develop a commercial model from those fixes.
ITER represents those fixes. The answer, then, is that it costs about 25 times as much as TFTR to get a machine that produces the results TFTR was supposed to - and that's assuming ITER doesn't fail. Applying the same scaling factors to ITER that they applied to TFTR means the production machines will cost on the order of hundreds of billions to develop.
Meanwhile, in the last few decades the PV researchers, working on perhaps 1/100th the amount of investment, decreased the cost of PV about 200 times. If installed in sunny locations, like Nevada, it produces power for about 4 cents, making it (inflation adjusted) one of the cheapest forms of power in history. When faced with this reality, fusion supporters have retreated to various arguments about why you can't possibly use these sources for X and Y, in spite of people already using them for X and Y all over the world.
> Focus fusion and Polywell are also promising
No they aren't. All modern pathways look promising until they tried to scale them up. And both of these are subject to the non-equilibrium braking radiation problem, which appears insurmountable.
> but underfunded
The problems with fusion cannot be solved by throwing money at that, that makes the problem *worse*. The problem is not actually physics, its economics, and we already know fusion is not competitive.
> By my calculation, the ~1g of lithium in a common phone
Li6. Not useful.
> It also could end up much worse, with either shadowing, module mismatch
Well the nice thing about utility scale is that you have some control over this. Shadowing? Call in the bulldozers and level the ground. Module mismatch? Demand the supplier stack them on the pallets in matched groups.
You and I don't have the same advantages, the pallet you get will be +/- ~3% and you get to install on whatever square you have. That said, I did get lucky - my 2 story garage gets shadowed only about 2% of the year, about an hour a day in late winter.
> Interesting blog, BTW.
Thanks! Not a lot of posts this year though, got a new kid.
Oh yeah sure, like the kWorld driver on ODroid...
> Solar Panels can actually be MORE efficient on cloudy days
They produce less. Every time. I have panels running since 2010, with daily statistics. Drop in power is basically linear with cloud cover.
So don't talk about all this theoretical BS before you have your own panels up. Which you should.
> Something will need to sit idle or at least underutlized to provide peak load
Yeah, but you make those plants be ones that are OK doing that. Nuclear power is not OK throttling (well, some are).
It's no coincidence that Ontario built the western world's largest coal plants while building out their nuclear fleet.
And it's also no coincidence they are replacing all the coal plants with gas peakers while retiring the nuclear fleet and slowly replacing it with wind.
> The performance ratio of such a power plant could be around 85%, with cable losses, inverter losses and automated cleaning
A VERY minor nit-pick - modern utility scale systems are closer to 90% because they have far lower line losses and inverters with ~98% efficiency. Apple's system (in WV? I can't recall) was specced at 92%
I realize that has no real effect on the bottom line, but I did think it was worth putting out modern numbers.
> Yes, if you ignore all maintenance costs
Maintenance costs are part of OPEX. OPEX on PV is the lowest of any major power source, by far. No moving parts, your main component is a sheet of glass (they are SILICON cells) in an aluminum frame, and the power conversion doesn't even have a transformer any more. And no fuel, of course.
Do you really think it would be more expensive than, say, a nuclear plant? Nuclear fuel is very cheap, but you still have lots (and LOTS) of moving parts to contend with. And cheap fuel isn't as cheap as no fuel at all.
679/648 = $1.05/Wp. That's actually slightly higher than new installs in the US and Europe, but I suspect that's simply because construction started some time ago.
PV is now the cheapest form of power in CAPEX terms, ever. Which is why ~60 GWp is going in this year.
> a photon is a particle, that travels in a wave
No. This is just wrong. Completely. You need to make your brain unlearn this.
Here is a toy model you can use on your journey... Think of the photon as a cheshire cat. You cannot see the cat, if you try to perceive it completely, it will vanish. You can, however, ask it questions. If you ask it "what is your gizifa", it will say "10". Or in this case, you can ask "what is your momentum", and it might say "5".
Asking certain questions will upset the cat and cause it to change all the other values just to piss you off. So if you ask it what its momentum is, the answer you *might* have got for its gizifa will now change. These values also change on their own over time. So even if you know that its location is 2,7 now, when you ask it again later you will get a different answer. No, this is not *because you asked* (another common misconception), this is inherent to the way the cat works. Some of these values are conserved (like electric charge), others are not (like location) and others are linked together (like momentum and location).
Photons are not particles. There are no particles. "Particles" is the term we use when we refer to these things when you keep asking them what their location is. If you do that, they will give you nice answers like 2,7 and then 3,7 and then 4.7, and you'll go "oh, this thing is travelling along positive X, and it's a point, so it must be a particle!". But the problem is that if you ask it different questions, like its position and the location, then any semblance of particle-like behaviour will vanish. You were fooling yourself, ITS NOT A PARTICLE. Neither is an electron or a proton, or anything else. They're just quanta. It's all quanta.
> It has some pressure when it shines on an object
This is also incorrect.
Newton thought momentum had something to do with mass because he only had large objects to work with. Shotputs have a lot of momentum, and so do planets. But in the "real world" of quantum, momentum is just a number. It's a number like any other, like energy. It's not related to mass. You ask a quanta a question and it will give you an answer. If you ask a photon its mass it will say zero. And if you ask it its momentum it will say 5. These questions are orthogonal, they don't have anything to do with each other.
So why does it LOOK like momentum has something to do with mass? Because the momentum of any one quanta is tiny, so in order to be measurable at macro scales, you need a WHOLE LOT OF QUANTA. It's very easy to make a ball of protons and electrons, because they attract each other. So you put a bunch together and call it a shotput and notice that it has a lot of momentum. But the fact that it has a lot of momentum isn't because it has a lot of mass - it has a lot of mass AND momentum because it *has a lot of quanta*.
It is much harder to make a big ball of photons, they don't attract each other. If you did such a thing, you'd find it had just as much momentum as a ball of matter, but still has no mass.
> and given off an infinite amount of mass/energy
No. Energy is also a measurement of the same sort, it's just a number you can ask for. It has nothing to do with "speed". Some of these values you ask for are more interesting than others because they are concerned, but that's not due to quanta, that's due to the shape of the universe.
> t is uncommon for a lithium-ion battery to withstand more than 1,500 charges before it fails
Bologna. My iPhone 5S is over three years old and still has ~65% charge at the end of a day of use.
> can store a large amount of energy
The paper is behind a paywall, but thanks to Sci-Hub I could read it. It focuses entirely on power density, not energy density, but does have some comparative information in Chart F. According to that, the best-case scenario for this device is about 0.07 Wh/cm^3. A modern li-po is about 0.5, around ten times the energy capacity of this device. That is actually less than some other designs, which have approached li-po but only on microscopic scales.
So don't hold your breath, this is not a device that you will see any time soon.
> wealthiest company on the planet
> But holy shit is your comment re: worrying overblown
Profitability appears to have little to do with lasting success. Perhaps the opposite. In spite of being the most profitable company in the computer space in the early 1990s, the entire market imploded almost overnight. I don't relish watching history repeat.
> Probably because the AirPort Express hasn't been updated since 2012 and the
> AirPort Time Capsule hasn't been updated since 2013
Indeed, which is perhaps a catch-22. I would suggest the new MBP demonstrates that. More broadly, Apple's attention only to margin is precisely what almost killed them in the 1990s, and its very VERY worrying to see history repeat itself.
I got mine in 2012, and at the time there was nothing like it - a pretty fast ac router with automated backups, printer bridging, simple setup (on the Mac, admittedly) for a few hundred bucks. But then I made the list of things it didn't do:
1) in spite of being relatively fast in general, it was built on the same "premium router" platform as other designs, but it was slower than them
2) while it claimed to offer PC file sharing, the software essentially did not work and could not be used in production
3) it handled printer sharing through port mapping, but did not do the same for scanners. all this needed was a *&^%$% TWAIN driver
4) it has four gigE ports, but exposes only three for some reason
5) a single USB port is not enough if you want to add a backup drive or second printer. the guts had four ports, so why not expose them?
The good news is, like all Apple products, it's pretty bulletproof, so I won't need to replace mine any time soon.
Circa 2010 (IIRC) the Airport was the best selling router on the planet. Now it's not selling.
> replace these humans who are always bitching
I hate to break it to you, but I've read texts from the Pax Romana era describing how to do this. People have been replacing labor with capital basically forever.
"There was never a job opening for a drone pilot until there was something to fly" :rolleyes:
There was something to fly before drones, airplanes. And they had one or more crew per aircraft. In comparison, drones have multiple aircraft per "pilot", and that pilot has far less to do.
" Tesla's Solar Roof Will Cost Less Than a Traditional Roof"
Traditional *expensive* roofs.
Which is fine, don't get me wrong, but if you're expecting this to compete with a $3000 re-roof using asphalt, not going to happen. If you do have a home with something more expensive, then the issue there is that they do tend to last longer and won't be a target for replacement as often.
> smack in the middle of the cold war and nuclear exchange always looming
Ahhh, the good old days.
Want to shit yourself? Read the MX basing report. You can find it online. One option they considered was putting the MX on trains, and when things got nasty, sending the trains out into the rail network. The review board has a calculation that shows the Sovs could bomb every square inch of the US rail network and still have lots of warheads left over to kill everyone.
What the hell were people thinking?
Two words: blue laser.
> Sorry, but send them back
Sure, get someone else to pay for their education so they can take all your ideas away. That sounds like a formula for rather rapid irrelevance.
Or, get someone else to pay for their education so all the smartest people in their country come and stay in yours. That sounds like a very good idea to me.
Which is fine, because if this does happen they'll all come to Canada instead. I'm still thanking Shrub for being the best president ever when he implemented the snowflake law and instantly created a $1 billion stem cell economy in Toronto. Go Trump!
> I'm genuinely curious
Anyone who starts with this statement almost certainly isn't.
It's part of the quiver that also contains "some of my best friends are".