|
QuarkJets posted:I don't think that it's intellectual dishonesty to compare the actual costs of things that you can buy and use right now. I should have been more clear, they're arguing that the cost of solar and wind over coal and natural gas means that solar and wind are inherently inferior technologies that will never make economic sense. The claim is that solar and wind were created purely for environmental reasons rather than being a cost effective source of energy which is pure bullshit. It would be akin to claiming that electric cars are inherently inferior to gasoline powered cars based on the state of the market in 2015, which is another idiotic argument I have seen from the same group of people.
|
#
?
Dec 6, 2015 21:40
|
|
|
|
| # ? May 23, 2024 13:37 |
|
|
Yeah. "Solar can never ever be worthwhile for anything lol" is a dumb argument, but when discussing short term and mid term implementations current cost matters regardless of future targets.
|
|
#
?
Dec 6, 2015 21:56
|
|
I CANNOT EJACULATE WITHOUT SEEING NATIVE AMERICANS BRUTALISED!
|
MaxxBot posted:I should have been more clear, they're arguing that the cost of solar and wind over coal and natural gas means that solar and wind are inherently inferior technologies that will never make economic sense. The claim is that solar and wind were created purely for environmental reasons rather than being a cost effective source of energy which is pure bullshit. Arguing based on projections of anything that's supposed to happen more than about twelve months out is equally ridiculous.
|
|
#
?
Dec 6, 2015 22:20
|
|
|
Yeah, even just pointing to a graph of $/kW for each technology pretty quickly dispells that notion. Even coal has been growing in cost
|
|
#
?
Dec 6, 2015 22:21
|
|
|
Rent-A-Cop posted:Arguing based on projections of anything that's supposed to happen more than about twelve months out is equally ridiculous. It would be dumb trying to predict exactly how cost trends will play out over time but I don't think it's ridiculous to say that solar is most certainly going to be cheaper than coal at some point in sheer $/kW. Just like it would be dumb for me to try to predict exactly how fast the SSD and GPU in my computer will be 4 years from now but I can say with basically 100% certainty that they will be faster than they are now barring some sort of catastrophic world event.
|
|
#
?
Dec 6, 2015 22:29
|
|
|
MaxxBot posted:I don't think it's ridiculous to say that solar is most certainly going to be cheaper than coal at some point in sheer $/kW. Honestly I think that any argument that Thing X will be better than Thing Y "at some point" is unfalsifiable and thus ridiculous on its face.
|
|
#
?
Dec 6, 2015 22:32
|
|
|
You're comparing a hundreds of years old technology that hasn't seen significant cost reductions in some time to a much newer one that has a recent trend of rapid reductions in cost. Sure the trend won't continue forever and no one can predict exactly what will happen but for solar to not drop in cost to the level of fossil fuels would require a really dramatic and sudden stagnation of this trend.
|
|
#
?
Dec 6, 2015 22:45
|
|
|
Rent-A-Cop posted:Why wouldn't that be ridiculous? The cost of solar is entirely in the fabrication and deployment of the panels. Those costs are not likely to go up in the future. Deployment costs will be primarily labor and will get more efficient over time. Fabrication costs have dropped dramatically and will likely continue down. With coal, you've got the cost of mining and transporting, as well as running a power plant to turn it into electricity. Those costs are not likely to go down in the future. The coal industry is mature and significant cost savings are going to be difficult to find. Thus, predicting that solar is going to be cheaper than coal in the future seems almost certain and far from ridiculous on its face.
|
|
#
?
Dec 6, 2015 22:49
|
|
|
MaxxBot posted:It would be dumb trying to predict exactly how cost trends will play out over time but I don't think it's ridiculous to say that solar is most certainly going to be cheaper than coal at some point in sheer $/kW. It certainly will if current trends continue, but there's no reason to assume that they will continue. First, most of the research is public money, and public money could easily dry up before solar and coal ever reach parity. The imminent loss of the federal solar home rebate and the various issues being created by local governments aren't helping, either; more expensive solar systems leads to fewer consumers, leads to less production, leads to more expensive solar systems. This also leads to less private research money
|
|
#
?
Dec 6, 2015 22:53
|
|
|
I'm not optimistic about US government support of solar by any means but it seems like China, India, and even Saudi Arabia are pretty committed to it.
|
|
#
?
Dec 6, 2015 23:03
|
|
|
MaxxBot posted:I'm not optimistic about US government support of solar by any means but it seems like China, India, and even Saudi Arabia are pretty committed to it. "It seems like" and "pretty committed" are two phrases from which you should not derive total certainty. We're still probably decades away from coal and solar reaching cost parity at today's rate of advancement, and a lot can happen in that time. Predicting solar's total dominance on that time scale is not much less idiotic than predicting solar's total failure on that time scale
|
|
#
?
Dec 7, 2015 00:03
|
|
|
So it sounds like the US power grid may be very vulnerable to CME's, small nuke EMPs (assuming they can be delivered at around 70km up, even small stuff like 50kt), and of course sabotaging substations. That doesn't even include potential for hacking SCADA systems and the AURORA attacks. My understanding is that a coordinated sabotage of a handful of key locations can knock the grid out for over a year.
|
|
#
?
Dec 7, 2015 01:12
|
|
|
The terrists have won.
|
|
#
?
Dec 7, 2015 01:40
|
|
|
Three-Phase posted:So it sounds like the US power grid may be very vulnerable to CME's, small nuke EMPs (assuming they can be delivered at around 70km up, even small stuff like 50kt), and of course sabotaging substations. That doesn't even include potential for hacking SCADA systems and the AURORA attacks. It's pretty vulnerable to Godzilla, too.
|
|
#
?
Dec 7, 2015 01:51
|
|
|
^ Heh. blowfish posted:The terrists have won. I hope not. Plus there are other technologies being discussed (EMP mitigation/hardening, microgrids for critical infrastructure) that can help out. But you lose electrical transmission, everything just shuts down. (No power, no clean water, no wastewater treatment, no refrigeration of food...)
|
|
#
?
Dec 7, 2015 01:52
|
|
|
Doesn't part of the problem lie with trying to bring all the power plants back online and having them fight each other? edit: A little google fu tells me, assuming you're talking about the same report, that it's a manpower issue and failure of control equipment. Hmm. Maybe require power companies to have an emergency action plan? There's a lot of small generation plants that should be easier to get online right after a nation wide outage. You could probably map out a plan to bring them online and focus on power to essential facilities in the nearby areas, utilities, communication, and public health. You'd need a supply of parts though to replace all the burned out machines. Don't CME's have some warning? Killer-of-Lawyers fucked around with this message at 09:58 on Dec 7, 2015 |
|
#
?
Dec 7, 2015 09:45
|
|
|
Killer-of-Lawyers posted:Hmm. Maybe require power companies to have an emergency action plan? They do, and they generally work quite well for local or regional outages. If everything is hosed then all bets are off.
|
|
#
?
Dec 7, 2015 15:24
|
|
|
Killer-of-Lawyers posted:Don't CME's have some warning? Not much. A few days on the upper end and ~2 days on the lower end. Not nearly enough warning to take any meaningful precautions. My understanding is that a big CME would still wreak havoc on the grid even if it were all offline, though I could be wrong.
|
|
#
?
Dec 7, 2015 15:26
|
|
|
Well, it would depend if transformers have enough wire in them to blow themselves out with out being attached to the actual transmission lines. If they do then there isn't much you can do. Also you can unplug the control equipment which would protect that at least. Basically, you would have an easier time if you had some warning and shut the grid down before hand then if you were caught unaware. I'd be more interested in hardening the network against CME's then a nuclear detonation in space, since the latter falls under nuclear deterrence and outright warfare. edit: And when I mean shut the grid down I mean physically isolating the transformers and equipment from the wires, not just turning the plants off. Killer-of-Lawyers fucked around with this message at 16:47 on Dec 7, 2015 |
|
#
?
Dec 7, 2015 16:45
|
|
|
Wouldn't "shuting the grid down" against a CME just be shutting down and tripping everything in a specific order? Just with calculated delays between to minimize the damage that will cause. The whole 12 hour warning thing that I hear a lot seems like a much bigger number then what you have to perform the shutdown. If you want a 60 second delay between each small breaker just to allow some modicum of stabilized shutdown but that would still take hours. It's going to take an hour or two just to make the high level determination to begin the shutdown. And more time past that to place critical infrastructure and hospitals in a state where it's safe to disconnect them.
|
|
#
?
Dec 7, 2015 17:07
|
|
|
M_Gargantua posted:Wouldn't "shuting the grid down" against a CME just be shutting down and tripping everything in a specific order? Just with calculated delays between to minimize the damage that will cause. There is the assumption that an EMP would trip protective systems further down the line and might limit the damage, but will leave substations in an unusable state. EMP is also not universal, it might knock out some while not even touch others.
|
|
#
?
Dec 7, 2015 17:20
|
|
|
The problem is that those scenarios are sufficiently unlikely that spending a lot of money to prepare for them isn't cost effective. Modularizing the grid so that parts of can be isolated if necessary is generally going to be a design element going forward when there is little to no additional cost to a project. Spending billions to retrofit the existing system is just not viable. The only significant concern is hacking into control systems, and that's already being dealt with, as it's a problem for everything on the internet everywhere.
|
|
#
?
Dec 7, 2015 17:28
|
|
|
Deteriorata posted:The only significant concern is hacking into control systems, and that's already being dealt with, as it's a problem for everything on the internet everywhere. Most of the solutions in this case have been to isolate control systems on a non-internet connected intranet.
|
|
#
?
Dec 7, 2015 17:41
|
|
|
CommieGIR posted:Most of the solutions in this case have been to isolate control systems on a non-internet connected intranet. Generally the best form of network security.
|
|
#
?
Dec 7, 2015 18:36
|
|
|
OwlFancier posted:Generally the best form of network security. Pretty much. There was some click bait article going around about Chinese hackers infiltrating Nuclear Power Plant control systems, which is both stupid and wrong, since even the few nuclear plants with digitized systems have isolated any sort of networked controls from the internet connected enterprise network.
|
|
#
?
Dec 7, 2015 18:53
|
|
|
CommieGIR posted:Pretty much. There was some click bait article going around about Chinese hackers infiltrating Nuclear Power Plant control systems, which is both stupid and wrong, since even the few nuclear plants with digitized systems have isolated any sort of networked controls from the internet connected enterprise network. I read on a sitcom that if they aren't airgapped they are vulnerable. :atomz:
|
|
#
?
Dec 7, 2015 19:27
|
|
|
Nevvy Z posted:I read on a sitcom that if they aren't airgapped they are vulnerable. :atomz: ....actually, that was a real thing: http://www.wired.com/2015/07/researchers-hack-air-gapped-computer-simple-cell-phone/ Granted, it doesn't allow you to CONTROL the system, but gather information from it.
|
|
#
?
Dec 7, 2015 19:35
|
|
|
Even if someone did hack a nuclear power plant control system somehow, all the poo poo it controls has manual backups anyways.
|
|
#
?
Dec 7, 2015 20:41
|
|
|
CommieGIR posted:....actually, that was a real thing: I'm confused as to how they expect to get the malware onto the otherwise secure system but this whole thing is cool as gently caress.
|
|
#
?
Dec 7, 2015 20:44
|
|
|
-Troika- posted:Even if someone did hack a nuclear power plant control system somehow, all the poo poo it controls has manual backups anyways. Pretty much. Everything has a backup of a backup in modern nuclear plants, and its part of the whole resistance to digitization in the nuclear industry as well. Nevvy Z posted:I'm confused as to how they expect to get the malware onto the otherwise secure system but this whole thing is cool as gently caress. Its more of a snooping tool. There's this one too: http://www.wired.com/2015/03/stealing-data-computers-using-heat/ Where you CAN actually send commands over air gap. But the proof of concept still requires both systems to be infect, which makes it kinda silly, but assumes that you could infect the system before it is air gapped and then hack it over the air gap with the previous infection
|
|
#
?
Dec 7, 2015 20:48
|
|
|
QuarkJets posted:It certainly will if current trends continue, but there's no reason to assume that they will continue. Research money has not been an important factor in driving down the price of solar. The plummeting cost of solar is not some Moore's Law-esque thing where technology is being incrementally developed towards a manifest destiny assured by fundamental physics; in terms of performance, existing PVs are not too far off from what will ever be achievable. Certainly, the orders-of-magnitude drop in price has nothing to do with orders-of-magnitude improvement in the technology over the last few decades. PV technology has not changed substantially over the last 20 years, and the kinds of solar cells being sold (and for which prices have been plummeting), are 1970's technology. The prevailing trends in cost per kWh are purely due to experience curve and economy of scale type effects, which are generic and expected for any industry when production volume increases greatly. There is no fundamental manufacturing or physical reason to think that these trends, which have persisted over 4 decades, 5 orders of magnitude in shipped volume and >2 orders of magnitude in price, will suddenly hit a brick wall. As such, it is very reasonable to assume that a factor of ~4-10 increase in PV production will result in prices being cut in half or by more. PV power, in general,, is currently 2-3 times more expensive than fossil fuels on a LCOE basis. Therefore it is very plausible that PV will reach parity with fossil fuels if there is a ~1 order of magnitude further increase in PV production volumes. The loss of rebates/subsidies is a problem for cheap photovoltaics, then, but not because of anything to do with research.
|
|
#
?
Dec 9, 2015 21:49
|
|
|
Morbus posted:Research money has not been an important factor in driving down the price of solar. I totally agree. I think you'll have trouble getting QuarkJets to agree though. I maintain that a lot of the cost reduction in the cells is mostly due to incremental improvements in the efficiency of solar silicon production (that stuff probably hasn't been heavily researched in a long time). Probably there have been a lot of system-level improvements which have lowered the solar electricity price, too. Most solar cell research in the past 10 years is on junk optical materials and nano-wires, and it is unlikely that that stuff will ever matter economically.
|
|
#
?
Dec 9, 2015 23:34
|
|
|
Morbus posted:Research money has not been an important factor in driving down the price of solar. The plummeting cost of solar is not some Moore's Law-esque thing where technology is being incrementally developed towards a manifest destiny assured by fundamental physics; in terms of performance, existing PVs are not too far off from what will ever be achievable. Certainly, the orders-of-magnitude drop in price has nothing to do with orders-of-magnitude improvement in the technology over the last few decades. PV technology has not changed substantially over the last 20 years, and the kinds of solar cells being sold (and for which prices have been plummeting), are 1970's technology. First, you're focusing on the efficiency of converting photons into electrons, and it's true that research on cutting-edge maximum efficiency panels does not meaningfully impact the end-cost of solar PV. I think that this is short-sighted; there are a million other ways in which research has benefitted the end-cost of residential PV. For instance, we have much more efficient microinverters and much more efficient silicon harvesting techniques than we did 20 years ago. These kinds of technologies (and others) have had a significant impact on the cost of residential PV. Second, a German think-tank seems to believe that the end-cost of large-scale solar is going to come down significantly thanks to advances in module efficiency. I don't think that this is supported by the historical data, so I don't personally put much stock in it, but others seem to. They think that 35% efficiency in large-scale PV deployment is a possibility, but it's probably a pipe dream. Third, your statement that modern silicon panels are "1970s technology" is as accurate as saying that a sports car engine is 19th century technology: technically true, but not in a meaningful way. As a point of reference, most of the commonly-used crystalline silicon solar cells have an efficiency of 25% (e: in ideal conditions, of course). But even cutting-edge Silicon cell research didn't reach this efficiency until the late 1990s (e: also in ideal conditions), and that was a research project, not a commercially-available solar panel. The manufacturing of panels today also does not at all resemble what it took to manufacture silicon panels in the 1970s. quote:The prevailing trends in cost per kWh are purely due to experience curve and economy of scale type effects, which are generic and expected for any industry when production volume increases greatly. There is no fundamental manufacturing or physical reason to think that these trends, which have persisted over 4 decades, 5 orders of magnitude in shipped volume and >2 orders of magnitude in price, will suddenly hit a brick wall. This was actually covered in my post; demand in the US is going to drop significantly once the federal rebates expire. But I'd like you to show some proof that the cost per kWh is "purely due to experience curve and economy of scale type effects". Do you have any? I believe that economies of scale and experience are making a big impact, but it's laughable to suggest that none of the research conducted in the last 40 years has had any cost impact. QuarkJets fucked around with this message at 00:07 on Dec 10, 2015 |
|
#
?
Dec 9, 2015 23:56
|
|
|
QuarkJets posted:First, you're focusing on the efficiency of converting photons into electrons, and it's true that research on cutting-edge maximum efficiency panels does not meaningfully impact the end-cost of solar PV. I think that this is short-sighted; there are a million other ways in which research has benefitted the end-cost of residential PV. For instance, we have much more efficient microinverters and much more efficient silicon harvesting techniques than we did 20 years ago. These kinds of technologies (and others) have had a significant impact on the cost of residential PV. I agree that one of the main cost-reduction drivers for solar electricity were improvements in solar silicon production efficiency and cost. I would be shocked if that were largely research-driven, though. QuarkJets posted:Second, a German think-tank seems to believe that the end-cost of large-scale solar is going to come down significantly thanks to advances in module efficiency. I don't think that this is supported by the historical data, so I don't personally put much stock in it, but others seem to. They think that 35% efficiency in large-scale PV deployment is a possibility, but it's probably a pipe dream. Wow, so they must be predicting that III-V semi-conductor tandem cells that are used on satellites will hit mainstream. 35% efficiency is beyond the limiting efficiency for an unconcentrated single junction cell. Conventional wisdom is that the satellite solar cells necessarily must be high cost solar cells, but I have not really heard a convincing argument as to why they have to be high cost solar cells. QuarkJets posted:As a point of reference, most of the commonly-used crystalline silicon solar cells have an efficiency of 25% (e: in ideal conditions, of course) The very best silicon mono-crystalline solar panels that you can buy are probably almost 25% efficient, but not quite. SunPower is a pretty big silicon cell manufacturer who is committed to manufacturing high efficiency silicon solar cells. They probably still hold the module (not the cell) record. But you are getting a little mixed-up here: most solar cells bought and sold today are not mono-crystalline silicon cells, they are poly-crystalline silicon cells with much worse efficiencies, ~half of 25%. One big reason why the efficiency is worse is because the grain boundaries in the polysilicon are sites for unwanted conversion of excited-state electrons into waste heat. If you read the famous paper on solar cell efficiency written in the early 60's by William Shockley, you'll learn that the state of the art cell in those days was 14% efficient, like these common cells sold today. Also, "Black silicon" referred to in the article is not really anything new. It's certainly not a third class of solar cell. Most silicon cells have engineered textures on them, and the idea of engineering textures on solar cells to aid in light absorption is probably 30-40 years old. The benefit of the texture is that it traps light inside the cell--instead of the light passing through the cell once (or twice if there is a reflector on the back of the cell), the light gets scattered internally inside the cell off of the texture. If you do a good job on the texture, you can use ~50x less of the solar cell material to absorb the same amount of light. Also in the class of solar cell materials, silicon is one of the more weak absorbers of sunlight (being a so-called 'indirect bandgap' semi-conductor), so it's no wonder that that idea caught on with companies. silence_kit fucked around with this message at 05:15 on Dec 10, 2015 |
|
#
?
Dec 10, 2015 04:42
|
|
|
QuarkJets posted:Third, your statement that modern silicon panels are "1970s technology" is as accurate as saying that a sports car engine is 19th century technology: technically true, but not in a meaningful way. As a point of reference, most of the commonly-used crystalline silicon solar cells have an efficiency of 25% (e: in ideal conditions, of course). But even cutting-edge Silicon cell research didn't reach this efficiency until the late 1990s (e: also in ideal conditions), and that was a research project, not a commercially-available solar panel. The manufacturing of panels today also does not at all resemble what it took to manufacture silicon panels in the 1970s. 25% isn't what is available to purchase and install today, the 2010 journal article linked in the article you quoted even says http://www.nature.com/am/journal/v2/n3/full/am201082a.html posted:The highest energy conversion efficiency reported so far for research crystalline silicon PV cells is 25%.  These are research panels, not things you can go out and get installed today. Maximum panel efficiency that is available for purchase right now is 20.4% with SunPower panels. silence_kit posted:I agree that one of the main cost-reduction drivers for solar electricity were improvements in solar silicon production efficiency and cost. I would be shocked if that were largely research-driven, though. The maximum theoretical efficiency of a cell is 34%, something lower than that for a module. Black silicon mentioned in the article is monocrystalline, which is black, as opposed to polycrystalline which is blue. The research cell efficiency has actually been pretty slow in advancing the last 10 or so years, it is still advancing but not at the rate it was in the 90s and first bit of the 00s. Most of the advancement in retail efficiency in the last 10 years has been about refining the manufacturing processes to bring retail panels closer to research panel efficiency. 
|
|
#
?
Dec 10, 2015 07:27
|
|
|
silence_kit posted:I agree that one of the main cost-reduction drivers for solar electricity were improvements in solar silicon production efficiency and cost. I would be shocked if that were largely research-driven, though. I don't think that anyone is arguing that those gains, or any of the gains in the realm of solar power cost reduction, are largely research-driven. The argument is that none of those gains were research-driven, which is plainly false quote:The very best silicon mono-crystalline solar panels that you can buy are probably almost 25% efficient, but not quite. SunPower is a pretty big silicon cell manufacturer who is committed to manufacturing high efficiency silicon solar cells. They probably still hold the module (not the cell) record. I'm not mixed up at all; the 25% figure is just what the article says. I'm aware that typical industrial-grade solar panels are in the range of roughly 15-20%. I edited for clarification QuarkJets fucked around with this message at 09:12 on Dec 10, 2015 |
|
#
?
Dec 10, 2015 08:40
|
|
|
fermun posted:Black silicon mentioned in the article is monocrystalline, which is black, as opposed to polycrystalline which is blue. I've never looked at the absorption spectra for polysilicon, so I might be wrong, but the color is probably a consequence of the low efficiency cells not absorbing sunlight as well as possible and not achieving silicon's max short-circuit current. The black versus blue here isn't really a material property but is a consequence of cell design. Black silicon is a newer term for silicon with an engineered texture on its surface. But almost all silicon cells have textures on them, making all textured cells 'black silicon'. The texture enables you to (theoretically) use 50x less silicon and still absorb the same amount of light.
|
|
#
?
Dec 10, 2015 15:35
|
|
|
silence_kit posted:Wow, so they must be predicting that III-V semi-conductor tandem cells that are used on satellites will hit mainstream. 35% efficiency is beyond the limiting efficiency for an unconcentrated single junction cell. Conventional wisdom is that the satellite solar cells necessarily must be high cost solar cells, but I have not really heard a convincing argument as to why they have to be high cost solar cells. Because launch costs are so high that anywhere you can save mass by using a higher-cost component, it's probably worth saving mass by using a higher-cost component. A 10% increase in solar cell efficiency means you need 10% fewer cells, which means a lot at $10,000/lb to LEO.
|
|
#
?
Dec 10, 2015 15:46
|
|
|
Phanatic posted:Because launch costs are so high that anywhere you can save mass by using a higher-cost component, it's probably worth saving mass by using a higher-cost component. A 10% increase in solar cell efficiency means you need 10% fewer cells, which means a lot at $10,000/lb to LEO. I understand why space people will pay almost anything for a higher efficiency cell. What I don't understand, and no one has ever convinced me of this, is why the III-V cells necessarily must be high cost. I think that most of the high cost is just a consequence of the III-V semi-conductor electronics industry being a boutique electronics industry servicing customers who are willing to pay a lot of money for their products. silence_kit fucked around with this message at 16:51 on Dec 10, 2015 |
|
#
?
Dec 10, 2015 16:42
|
|
|
|
| # ? May 23, 2024 13:37 |
|
|
Perhaps the complexity of their production does not outweigh the increase in efficiency except when transport costs are tens of thousands of dollars per pound?
|
|
#
?
Dec 10, 2015 17:41
|
|