5 ms·

Now, this is how a good discussion can be had. Not diverting into irrelevant crap but actually asking the commenter to justify the claim --something I will gla

Now, this is how a good discussion can be had. Not diverting into irrelevant crap but actually asking the commenter to justify the claim --something I will gladly do here. NOTE: Breaking this into parts because of post limits. ---------------- Power and Energy ---------------- First things first, we have to be careful not to mix energy with power. We all do it out of convenience, yet they are very different things. The differences also affect system-wide technology and economics. Since I can't assume the audience is entirely technical, here's a reasonable analogy for power and energy: Your garden hose is a low power source of water when compared to a firehose. Anyone understands that a firehose can deliver a lot more water per second than a garden hose. Energy is like the amount of water you need to fill-up a swimming pool. You can fill it up with a garden hose, but it could take days. If you use a high power fire hose you might be able to fill up the same swimming pool in seconds. The water distribution infrastructure currently installed in any city is designes such that every home (or a reasonable number of homes) can use their garden hoses and they will all have acceptable flow rates. If every home in a city was specified to have a fire hose, the size and scale of the water distribution systems would be absolutely massive when compared to what we have today. One way to think of this is that power determines how quickly you can deliver energy. Water flow rate determines how quickly you can fill-up your swimming pool. ---------------- The Solar Panel ---------------- What are the realities of a solar panel? Sure, you buy a 325 W panel. Great. How does that behave in the real world, in the context of a system? There are several variables that affect the output of a solar panel: - Mounting angle - Variation of solar radiation based on time of day - Weather (clouds, rain, etc.) - Dirt - Negative temperature coefficient - Wiring losses - Energy conversion losses - Degradation over time I could write pages on the above. I'll limit myself to a couple of elements that have the most impact on power production. ---------------- The Solar Parabola ---------------- If you look at the output of a fixed solar array, regardless of scale, rooftop or megawatt, it will look like an inverted parabola. Here's a picture of that from my 13 kW array: https://i.imgur.com/Fl8ARJd.png https://i.imgur.com/Fl8ARJd.png This is from March of this year. It should be noted that it never reached 13 kW, it peaked at about 10 kW. This is important, but I'll skip over it for now. A nuclear power plant, over the same period of time, would produce constant power for 24 hours. If we are going to scale it down to this graph, we would get 10 kW every second of the day for the full 24 hours. The first thing we need to calculate, then, is the ratio between full power for the solar period vs. what the equivalent nuclear source would deliver. The way you calculate this is to integrate the curves over the period of interest, say from 8 AM to 8 PM. You are comparing the area of a rectangle to that of the parabola that fits within it. The answer to this is surprisingly simple: The ratio is 2/3. In other words, a solar system --of any scale-- rated at the same peak power as a nuclear power plant, will, at best, produce 2/3 (66%) the energy over the same period of time under --and this is important-- ideal conditions. Just using this number, we can calculate that we need to build a 1.5 GW solar array in order to match the daylight energy output of a 1 GW nuclear power plant. ---------------- Power at Night ---------------- That covers you for 12 hours. What do we do for the other half of the day? Batteries, of course. Well, we need extra energy to pump into the batteries so we can use it at night. The 1.5 GW is used-up. If we keep that math simple, that means we need to double our power production capacity. Now we are up to a 3 GW peak power solar system in order to be able to match the energy output of a 1 GW nuclear power plant. I won't cover the cost, scale and realities of such a massive energy storage installation at this time, just remember this is very much a part of the reality of solar --and a very significant one at that. ---------------- Conversion Efficiency ---------------- What I will talk about is the fact that solar arrays require the power conversion systems in order to move energy in and out of batteries and even out to the grid. Current battery charging technologies are nearly 100% efficient, so I'll ignore the small losses incurred going in and out of a battery and over time (self discharge). In order to get in and out of the buildings full of battery packs you will need to use conversion equipment. This will cost about 20% of the energy you produce. Yes, 20% of what you produce will be converted into heat. At the peak of 3 GW, we would be producing 0.6 GW in heat. That's no joke. That's nearly the peak output of a nuclear facility being used to create heat. I won't get into what this might mean in terms of having to provide for cooling. I don't know how this is handled at such massive scales. I just know that 0.6 GW of heat producing power is a very serious number. OK, this means that our system now needs to be upsized yet again in order to compensate for the 20% conversion loss. 3 GW * (1/0.8) = 3.75 GW That's where we are: Without considering other factors, the solar equivalent of a 1 GW nuclear power plant requires 3.75 GW peak power and a train-load of batteries. Are we done? No, not even close.

robomartin 4y ago
Part 2: ---------------- Unreliable Solar Production ---------------- https://i.imgur.com/SOr30bX.png https://i.imgur.com/SOr30bX.png Take a look at that image. That, again, is from my solar array in March of this year. What happened to the nice smooth parabola? Well, the weather happened, that's what you are looking at. Each one of those horrific dips is a cloud or set of clouds calmly flying by. Yup. Here's the graph for another day around the same time period: https://i.imgur.com/yvTdNX0.png https://i.imgur.com/yvTdNX0.png You can see just how dramatic the power loss can be just because of a few clouds flying by. In one case there's a drop from 7.5 kW to about 2.5 kW. In another a drop from 8 kW to 4 kW. And these drops last HOURS. The net effect is that the peak rating of your solar array is a distant image in the context of real-world solar. My 13 kW array gets taken down to 2.5 kW by a CLOUD. What this means, at a practical level, is that, in order to have the ability to deliver constant power --like a nuclear power plant-- 24/7 your solar system will have to be overbuilt to a larger scale yet. More batteries, lots more, and lots more solar panels. Imagine a city or small town suffering such deep power losses as weather rolls over the one-and-only solar facility. Rain would do the same thing, even worse. How do we even begin to calculate something like this? Well, let's look at an example of day to day energy generation for my system. This is January of this year: https://i.imgur.com/bGuCH2F.png https://i.imgur.com/bGuCH2F.png Here's last month: https://i.imgur.com/8lYKImD.png https://i.imgur.com/8lYKImD.png We have days with less than half the energy output, yes, here in sunny California. I don't even want to imagine what this might look like in other parts of the country/world where they have real weather. Imagine this happening at the scale of a city. In checking the daily output of this system over the last 24 months, I estimate that 7 out of 31 days we are producing at half the rated peak power, if not less. This represents an 11.3% loss of capacity, which we have to compensate for by building a larger system yet. Now we need a 4.2 GW system to match the output of a 1 GW nuclear power plant. Does it end there? Nope. ---------------- Negative Temperature Coefficient and The Seasons ---------------- https://i.imgur.com/EF2L3Hk.png https://i.imgur.com/EF2L3Hk.png That's month-to-month generation for all of last year. You can see that the peak was reached in May, not in the middle of summer as most would think. Why? Solar panels have a characteristic called "Negative Temperature Coefficient". In plain language, it means that they produce less power when they get hot. In the summer, for example. Couple that to variations of solar radiation based on the season and you get the above graph. The difference between the May peak and December low is about a 50% reduction in energy production yet again. This seems to be a common theme, doesn't it? If the design was based on May peak power generation as a constant throughout the year, we now need to compensate for the 22% loss suffered due to seasons and the negative temperature coefficient of the panels. This brings our 4.2 GW array up to 5.5 GW. Again, to match the output of a 1 GW nuclear power plant. It should not be lost in this discussion that this also represents a massive increase in the number of batteries you'll need, heat management as well as the massive amounts of construction materials, labor and land such a system would require. I'll stop here because the rest of the analysis starts to get into deeper technical details an modeling that is hard to convey in this medium. Things like the loss of energy in the very wires used to connect everything together and the statistical failures of a system in the with somewhere in the order of 16 million solar panels (that's roughly what you need for a 5+ GW system. The final number quickly approaches 7 GW as the solar equivalent of a 1 GW nuclear power plant. You also need somewhere over 20,000 or 30,000 acres of land for this installation. No telling what the ecological effect of such a monster might be. The US Department of Energy says that the solar equivalent of a 1 GW nuclear power plant needs 75 times more land. Quoting: “A typical 1,000-megawatt nuclear facility in the United States needs a little more than 1 square mile to operate. NEI says wind farms require 360 times more land area to produce the same amount of electricity and solar photovoltaic plants require 75 times more space.” https://www.energy.gov/sites/prod/files/2019/01/f58/Ultimate%20Fast%20Facts%20Guide-PRINT.pdf https://www.energy.gov/sites/prod/files/2019/01/f58/Ultimate... 1 square mile would be covered with roughly 1.3 million solar panels (each being two square meters). A solar farm requires access isles for installation, cleaning and maintenance of rows of panels. So, out of the 75 square miles the DoE provides as an equivalent, we would have to assume a percentage would have no panels. Here are the results: 33% covered with panels = 30.9375 million panels 50% covered with panels = 46.875 million panels 75% covered with panels = 70.3125 million panels Assuming 325 W panels: 33% case: 10.1 GW 50% case: 15.2 GW 75% case: 22.9 GW And so, the best case presented by the US Department of Energy as the solar equivalent of a 1 GW nuclear facility requires 10.1 GW of peak power generation capacity. This aligns very well with my "at least 7x" calculation. That's my point. It's math and physics. Very basic math and physics at that. And yet most people are living in a delusional cult that looks at solar as this bubble gum and pink unicorns technology that will save the world. Well, based on the science, I beg to differ. Time to actually discuss facts rather than fantasy.

robomartin 4y ago

Part 2: ---------------- Unreliable Solar Production ---------------- https://i.imgur.com/SOr30bX.png https://i.imgur.com/SOr30bX.png Take a look at that image. That, again, is from my solar array in March of this year. What happened to the nice smooth parabola? Well, the weather happened, that's what you are looking at. Each one of those horrific dips is a cloud or set of clouds calmly flying by. Yup. Here's the graph for another day around the same time period: https://i.imgur.com/yvTdNX0.png https://i.imgur.com/yvTdNX0.png You can see just how dramatic the power loss can be just because of a few clouds flying by. In one case there's a drop from 7.5 kW to about 2.5 kW. In another a drop from 8 kW to 4 kW. And these drops last HOURS. The net effect is that the peak rating of your solar array is a distant image in the context of real-world solar. My 13 kW array gets taken down to 2.5 kW by a CLOUD. What this means, at a practical level, is that, in order to have the ability to deliver constant power --like a nuclear power plant-- 24/7 your solar system will have to be overbuilt to a larger scale yet. More batteries, lots more, and lots more solar panels. Imagine a city or small town suffering such deep power losses as weather rolls over the one-and-only solar facility. Rain would do the same thing, even worse. How do we even begin to calculate something like this? Well, let's look at an example of day to day energy generation for my system. This is January of this year: https://i.imgur.com/bGuCH2F.png https://i.imgur.com/bGuCH2F.png Here's last month: https://i.imgur.com/8lYKImD.png https://i.imgur.com/8lYKImD.png We have days with less than half the energy output, yes, here in sunny California. I don't even want to imagine what this might look like in other parts of the country/world where they have real weather. Imagine this happening at the scale of a city. In checking the daily output of this system over the last 24 months, I estimate that 7 out of 31 days we are producing at half the rated peak power, if not less. This represents an 11.3% loss of capacity, which we have to compensate for by building a larger system yet. Now we need a 4.2 GW system to match the output of a 1 GW nuclear power plant. Does it end there? Nope. ---------------- Negative Temperature Coefficient and The Seasons ---------------- https://i.imgur.com/EF2L3Hk.png https://i.imgur.com/EF2L3Hk.png That's month-to-month generation for all of last year. You can see that the peak was reached in May, not in the middle of summer as most would think. Why? Solar panels have a characteristic called "Negative Temperature Coefficient". In plain language, it means that they produce less power when they get hot. In the summer, for example. Couple that to variations of solar radiation based on the season and you get the above graph. The difference between the May peak and December low is about a 50% reduction in energy production yet again. This seems to be a common theme, doesn't it? If the design was based on May peak power generation as a constant throughout the year, we now need to compensate for the 22% loss suffered due to seasons and the negative temperature coefficient of the panels. This brings our 4.2 GW array up to 5.5 GW. Again, to match the output of a 1 GW nuclear power plant. It should not be lost in this discussion that this also represents a massive increase in the number of batteries you'll need, heat management as well as the massive amounts of construction materials, labor and land such a system would require. I'll stop here because the rest of the analysis starts to get into deeper technical details an modeling that is hard to convey in this medium. Things like the loss of energy in the very wires used to connect everything together and the statistical failures of a system in the with somewhere in the order of 16 million solar panels (that's roughly what you need for a 5+ GW system. The final number quickly approaches 7 GW as the solar equivalent of a 1 GW nuclear power plant. You also need somewhere over 20,000 or 30,000 acres of land for this installation. No telling what the ecological effect of such a monster might be. The US Department of Energy says that the solar equivalent of a 1 GW nuclear power plant needs 75 times more land. Quoting: “A typical 1,000-megawatt nuclear facility in the United States needs a little more than 1 square mile to operate. NEI says wind farms require 360 times more land area to produce the same amount of electricity and solar photovoltaic plants require 75 times more space.” https://www.energy.gov/sites/prod/files/2019/01/f58/Ultimate%20Fast%20Facts%20Guide-PRINT.pdf https://www.energy.gov/sites/prod/files/2019/01/f58/Ultimate... 1 square mile would be covered with roughly 1.3 million solar panels (each being two square meters). A solar farm requires access isles for installation, cleaning and maintenance of rows of panels. So, out of the 75 square miles the DoE provides as an equivalent, we would have to assume a percentage would have no panels. Here are the results: 33% covered with panels = 30.9375 million panels 50% covered with panels = 46.875 million panels 75% covered with panels = 70.3125 million panels Assuming 325 W panels: 33% case: 10.1 GW 50% case: 15.2 GW 75% case: 22.9 GW And so, the best case presented by the US Department of Energy as the solar equivalent of a 1 GW nuclear facility requires 10.1 GW of peak power generation capacity. This aligns very well with my "at least 7x" calculation. That's my point. It's math and physics. Very basic math and physics at that. And yet most people are living in a delusional cult that looks at solar as this bubble gum and pink unicorns technology that will save the world. Well, based on the science, I beg to differ. Time to actually discuss facts rather than fantasy.