America now has more workers in solar than coal.

[PKRudeBoy]

No, that's wrong.

If you spend $500 on digging holes and $500 on installing solar panels, 1 year later you have a lumpy field and a still working solar panel. It's much better as deficit spending then hole digging.

Hole digging is (almost) pure labour cost and can be ended as soon as the economy has recovered. Solar panels have a large capital cost and to be worthwhile they will be subsidised for decades after any short term economic problems are gone. A better comparison would be $500/year for 3 years vs $1,000/year for 30 years.

Errrr, they've already been "subsidized" by being built by state-financed labor and having the capital cost paid by the same state. They may have to be operated for 30 years in order for the state to see a profit, but that's not the same as being subsidized for 30 years. They have additional capital costs compared to hole-digging, but you also get the solar cell at the end, which can generate electricity. That electricity may be lower-efficiency than other forms of generation, but that makes them worse than the other forms of electricity, not worse than literal holes.

Not necessarily. If you spend $500 on installing solar panels then you're just paying people to put them in. You still have to buy the panel, which to have the same effect on the economy, would have to be built in your country from parts made in and materials extracted from your country. Which, unless you're China probably isn't gonna happen, although at least for the US some of the parts are being exported to China for assembly.

[Terralthra]

EDIT: Perhaps a quantitative example would help.

Plan A: I pay 1,000,000 workers $5/hour each to dig holes and fill them in again. Over 3 years this costs me $30 billion.

Plan B: I pay 1,000,000 workers $5/hour each to produce solar panels. Each hour the solar panel machines also consume $5 of input resources, replacement parts costs, etc. Over 3 years this costs me $60 billion. Over their lifetime the panels produce electricity that is worth $10 billion.

By choosing Plan B I lose $20 billion over having chosen Plan A.

Well, shit, as long as we're allowed to make up numbers, I say Plan C is pay 1,000,000 workers $15/hr (I plan to make sure they actually get above the poverty line) to make solar panels. This costs me $90 billion, plus solar panels which cost me $0.10 of input resources, replacement parts, maintenance, etc. This costs me an additional $100 million. Over their lifetime, the panels produce electricity worth $14 sextillion. Clearly, Plan C wins over both Plans A and B.

Next time you want to "quantify", be prepared to have actual numbers, or it's just an exercise in masturbation in public, and no one here wants to watch you do that.

[energiewende]

1) To what extent could we mitigate this inefficiency by economies of scale?

Very little. In fact even if the panels were free the cost of a grid would still be extremely expensive due to intermittency mitigation.

2) To what extent is it in the state's interest to subsidize solar power anyway, knowing that it is inefficient, so as to avoid the opportunity cost of pollution and global warming down the line?

None. There are cheaper ways to achieve the same thing.

3) To what extent will the cost-inefficiency be affected by the totally predictable increases in energy costs that are likely to occur as the 21st century rolls on?

At the margin, not at all, because fission is also not heavily tied to the price of scarce fuels. This is more of an issue for CCS but unlikely within the lifespan of the next generation of plants.

Well, shit, as long as we're allowed to make up numbers, I say Plan C is pay 1,000,000 workers $15/hr (I plan to make sure they actually get above the poverty line) to make solar panels. This costs me $90 billion, plus solar panels which cost me $0.10 of input resources, replacement parts, maintenance, etc. This costs me an additional $100 million. Over their lifetime, the panels produce electricity worth $14 sextillion. Clearly, Plan C wins over both Plans A and B.

Next time you want to "quantify", be prepared to have actual numbers, or it's just an exercise in masturbation in public, and no one here wants to watch you do that.

You said it was wrong in principle that spending money on capital investment can ever be worse than spending money to produce nothing. I only need to show it is possible for a capital investment to yield a negative return to disprove this.

It is of course possible for capital investments to make a return, but you're implying that it's unknowable whether they will in this case. Your new example only works because you have assumed the value of solar electricity is much higher than the cost of the panels but we know that the opposite is true.

[madd0ct0r]

In the US there are federal subsidies and a patchwork of state subsidies. The federal level subsidy is closest to your description, a 30% "tax rebate" (actually it has nothing to do with the amount of tax due on the transaction) on purchase and installation of solar panels. But this is not enough to make solar viable on its own which is why there are such disparities in uptake of solar power as mentioned in the original report.

so if we look at the federal subsidy, it's an up front capital payment and nothing to do with a 'feed in' system. Conclusion. It's a one off payment that results in a solar panel. You're paying a guy to drill holes (and then bolt the panel on). Let's be honest here. Not that many households are installing solar power 'becuase it's a good investment with a high return.'
The subsidy makes it a less bad investement, but you've got icky human stuff like independence, environmental concern, tech interest or winding up the neighbors at play here.
If you want to look at the patchwork of state level stuff, bring your own magnifying glass. I think you'll also find the large scale 'solar plants' are on different schemes to the average householder, so good luck producing coherent numbers!

speaking of which.

Of the schemes seriously discussed, solar is the least cost efficient means of achieving any of those goals, and the absolute cost is only small so long as grid penetration also remains small.

Is it possible, that as an nurture/incubator scheme, the price paid by the government will be changed as grid penetration grows? Can government policy change over time to reflect new circumstances? Why yes, I believe it can.

[Terralthra]

Well, shit, as long as we're allowed to make up numbers, I say Plan C is pay 1,000,000 workers $15/hr (I plan to make sure they actually get above the poverty line) to make solar panels. This costs me $90 billion, plus solar panels which cost me $0.10 of input resources, replacement parts, maintenance, etc. This costs me an additional $100 million. Over their lifetime, the panels produce electricity worth $14 sextillion. Clearly, Plan C wins over both Plans A and B.

Next time you want to "quantify", be prepared to have actual numbers, or it's just an exercise in masturbation in public, and no one here wants to watch you do that.

You said it was wrong in principle that spending money on capital investment can ever be worse than spending money to produce nothing. I only need to show it is possible for a capital investment to yield a negative return to disprove this.

No, I did not say that it was wrong in principle, I said it was wrong in this case. Feel free to show otherwise. Maybe you can make up a quote of me saying so, like you made up "quantification"?

It is of course possible for capital investments to make a return, but you're implying that it's unknowable whether they will in this case. Your new example only works because you have assumed the value of solar electricity is much higher than the cost of the panels but we know that the opposite is true.

No, I'm implying it's likely they will make a return in this case. Your logic could be applied equally well to the heavily subsidized, government-built system of dams, highways, bridges, libraries, and other infrastructure built during the Great Depression, yet today those form much of the backbone of the transportation infrastructure in the US, and their cost has been paid back many times over.

Unless you provide realistic numbers for wages and input costs, as well as a realistic return output, and show that the capital infrastructure investment is literally not worth the cost, you're just whistling in the dark.

[energiewende]

so if we look at the federal subsidy, it's an up front capital payment and nothing to do with a 'feed in' system. Conclusion. It's a one off payment that results in a solar panel.

It's a one-off payment that doesn't result in a solar panel because it doesn't cover enough of the cost. This is why almost everywhere solar is subsidised on a rolling basis: the costs are too high to be acceptable to governments as a huge lump sum. But this makes no difference to the calculation anyway. Whether the panel is subsidised over 30 years, or the NPV is paid immediately, the sum remains far higher than a project that doesn't require a large, mostly worthless capital investment to be written off.

No, I'm implying it's likely they will make a return in this case. Your logic could be applied equally well to the heavily subsidized, government-built system of dams, highways, bridges, libraries, and other infrastructure built during the Great Depression, yet today those form much of the backbone of the transportation infrastructure in the US, and their cost has been paid back many times over.

Unless you provide realistic numbers for wages and input costs, as well as a realistic return output, and show that the capital infrastructure investment is literally not worth the cost, you're just whistling in the dark.

The relevant numbers are the cost of a solar grid and the value of the electricity it produces. #1 is much larger than #2.

If you just want to spend money on something, anything, you could replace the whole grid with nuclear plants for the same cost as replacing a small fraction of it with solar. Or just demolish fossil fuel plants and rebuild them for that matter. You'd still come out ahead financially.

[Terralthra]

No, I'm implying it's likely they will make a return in this case. Your logic could be applied equally well to the heavily subsidized, government-built system of dams, highways, bridges, libraries, and other infrastructure built during the Great Depression, yet today those form much of the backbone of the transportation infrastructure in the US, and their cost has been paid back many times over.

Unless you provide realistic numbers for wages and input costs, as well as a realistic return output, and show that the capital infrastructure investment is literally not worth the cost, you're just whistling in the dark.

The relevant numbers are the cost of a solar grid and the value of the electricity it produces. #1 is much larger than #2.

This is a) not the first time you've said that and b) not the first time you haven't backed it up with actual numbers.

[madd0ct0r]

so if we look at the federal subsidy, it's an up front capital payment and nothing to do with a 'feed in' system. Conclusion. It's a one off payment that results in a solar panel.

It's a one-off payment that doesn't result in a solar panel because it doesn't cover enough of the cost. This is why almost everywhere solar is subsidised on a rolling basis: the costs are too high to be acceptable to governments as a huge lump sum. But this makes no difference to the calculation anyway. Whether the panel is subsidised over 30 years, or the NPV is paid immediately, the sum remains far higher than a project that doesn't require a large, mostly worthless capital investment to be written off.

To follow Terraltha's lead, can you put real numbers to that? I don't mind if you want to cherry pick a single large state. California, florida or texas being the obvious choices.

[energiewende]

The US has 1.6GW of installed solar capacity. Solar panels cost about $1/W installed. The US solar park is therefore cost about $1.6 billion (ignoring installation, grid extension, intermittency mitigation etc.). With a capacity factor of 14% it produces 1.8 billion kWh of electricity each year. The wholesale price of electricity is 2-4 cents. So the current US solar park would take 25-50 years to pay back just the capital costs of the panels, 1-2x the life of the panels.

This is unfairly generous to solar generators because you pay the $1.6bn up front. If you had made an investment that actually makes money instead you would have a lot more than just recouped your costs after 50 years. If you in index linked funds you should make 3-5% in real terms. Taking lower bound of 3%, you would have needed to have sold the electricity for $7bn to say you broke even and not have your accountant burst out laughing at your gullibility.

Including installation, maintenance and diffuse grid costs isn't amenable to back-of-a-napkin analysis. Let's look at some empirical data. The IEA estimates that wholesale price of electricity generated by solar comes it at best at 15c/kWh - about 5x the best competing technologies. With actual US measured capacity factors, they say more like 30c/kWh. The US Federal government is offering a 30% cost reduction (and not then to the entire supply chain, but ignoring that) when what's needed is more like a 90% reduction.

If you want to include intermittency mitigation... there are no estimates, because no one has an idea how to do it that survives the calculations on back of a napkin stage.

[Simon_Jester]

Where are you getting 1.8 billion kilowatt-hours? I got more like 1.96, but that's a detail I guess.

In any case, the corollary to this is that the solar industry today exists on such a small scale that we shouldn't wonder it's inefficient, even if it doesn't have to be. Most industries are inefficient when they've only existed for a few decades, and most of that solar power capacity in the US has been installed in the past 25-30 years. Heck, 80 to 90% of it has been added in the last five years.

http://en.wikipedia.org/wiki/File:US_So ... uction.png

Now, you can make arguments from theory that solar installation just plain can't get much cheaper, even if we start mass-producing and laying out standardized designs the way France did with nuclear power. And you can argue that for the foreseeable future electricity prices will remain constant, and that therefore that 2-4 cents per kWh will remain constant and make solar unprofitable.

Those are arguments that I think you need to make, for your position to be tenable. The theory behind solar power subsidies is that we are deliberately forcing our way through the expensive prototype and lessons-learned phase of industrial solar power production, in hopes of getting the technology into a reasonably dependable form before it is needed, instead of after.

[energiewende]

Can I be clear what you are arguing: that solar power is in fact cost effective today without massive subsidy, or that the government may be able to make it cost effective in the future? I am willing to discuss the latter but this is talking about the potential return on financial speculation that even the market that did subprime, S&L and the tulip bubble doesn't want to touch with a barge pole. The former is about present reality and is subject to empirical measurement.

If you agree with me that solar power is not presently cost effective without massive subsidy then you do not disagree with the substance of what I have posted.

[Simon_Jester]

If solar power is costing us 10-15 cents per kilowatt hour to produce, then yes, at this time it is not cost-effective without subsidy.

Now, you then go on to point out that the market doesn't want to speculate on solar power becoming profitable. I can think of a lot of reasons for this. The obvious one is that we're talking about a 10, 20, or even 50-year timeframe here. Solar power getting more efficient will require:

-Significant improvements in design of the plants.
-Streamlining production and installation of the solar cells.
-Probable rise in the cost of energy, caused in part by the cost of alternatives to solar such as coal.
-Very probably, breakthroughs in technology that we can hope for but NOT count on (random gibbering by Singular Intellect, the board's resident sunpower-advocate, notwithstanding).

So there's not just risk, but risk that is unlikely to be anywhere near finished paying off until the mid-21st century, possibly not until the late century. This is exactly the kind of thing that governments often invest in, and private companies only sometimes try, because it cannot possibly promise a quick payoff, the way that subprime mortgage securities and investing in a tulip bubble do.

Is it worth it? If the world gets more rational about nuclear power in the next few decades maybe not, though nuclear does have some drawbacks and the accident risk is real if overblown. Unless we go full nuclear, though, it's likely that we'll be grateful 50-100 years down the line for any investment in solar technology (and wind and so on). Because by that time, renewables and much pricier coal will be about the only things left on the energy market, and the cost of electricity is likely to increase until it reaches equilibrium with the cost of renewables.

Anything we do now to reduce the cost of renewables in 2030 and 2040 when we'll have to install them is worth funneling a bit of money into, I think.

[energiewende]

Ok, so we both agree that current solar plants are a waste of money in their own right. Let's see if they are good speculation instead.

If solar power is costing us 10-15 cents per kilowatt hour to produce, then yes, at this time it is not cost-effective without subsidy.

Now, you then go on to point out that the market doesn't want to speculate on solar power becoming profitable. I can think of a lot of reasons for this. The obvious one is that

there is no way to make solar cost effective that will work in a sensible time frame that you can use to look a grandmother in the eye and ask her to give you her pension. Even if you work for Goldman.

we're talking about a 10, 20, or even 50-year timeframe here. Solar power getting more efficient will require:

-Significant improvements in design of the plants.
-Streamlining production and installation of the solar cells.
-Probable rise in the cost of energy, caused in part by the cost of alternatives to solar such as coal.
-Very probably, breakthroughs in technology that we can hope for but NOT count on (random gibbering by Singular Intellect, the board's resident sunpower-advocate, notwithstanding).

So there's not just risk, but risk that is unlikely to be anywhere near finished paying off until the mid-21st century, possibly not until the late century. This is exactly the kind of thing that governments often invest in, and private companies only sometimes try, because it cannot possibly promise a quick payoff, the way that subprime mortgage securities and investing in a tulip bubble do.

Is it worth it? If the world gets more rational about nuclear power in the next few decades maybe not, though nuclear does have some drawbacks and the accident risk is real if overblown. Unless we go full nuclear, though, it's likely that we'll be grateful 50-100 years down the line for any investment in solar technology (and wind and so on). Because by that time, renewables and much pricier coal will be about the only things left on the energy market, and the cost of electricity is likely to increase until it reaches equilibrium with the cost of renewables.

Anything we do now to reduce the cost of renewables in 2030 and 2040 when we'll have to install them is worth funneling a bit of money into, I think.

Yes there is risk. Risk doesn't just mean that an investment may take a long time to pay off, it means it may never pay off or even go to zero. The smart money is saying that solar plants built today will never pay off and investment in production today is not worth it in light of potential future profits if solar becomes dramatically more competitive in the future. Most money on the market is managed by pension funds and they do care a great deal what will happen in 40 years. That's why they invested in mortgages; not because subprime promised a quick pay-off but because they believed CDOs had the ability to turn high risk, low return subprime investments into low risk, low return AAA-rated investments. This belief turned out to be wrong but institutions purchasing CDOs were not out to make a quick buck, quite the opposite, they wanted a steady return over many years.

So why wasn't pension fund money rolling into solar 20 years ago? Firstly, because none of this justifies building millions of panels. It justifies research to some extent but that was paid by the government anyway, for the cost of peanuts compared to subsidising mass production of panels, and as a result panel prices had been dropping for decades before they became "low" enough to make subsidy at all palatable. Second, because no one with money on the line believes that future power grid is going to be based on solar, subsidy farming aside. Even if the panels become very cheap (which they almost certainly will in 30-50 years, though cost of backing material and power electronics probably caps the price to just a bit below grid parity), you need to time shift the energy and doing that on a large enough scale for a low enough price is looking like a fusion-scale problem.

[TimothyC]

If you agree with me that solar power is not presently cost effective without massive subsidy then you do not disagree with the substance of what I have posted.

The one objection I have to this statement is that there are environments (mostly remote locations such as Hawai'i) where it is cost effective without subsidy to install solar power. When I went there this past winter, I talked with the locals and they noted that solar, without subsidy, breaks even or beats 'grid' power on most of the islands. This is independent of the mandate for solar water heating with all new residential construction.

[energiewende]

If you agree with me that solar power is not presently cost effective without massive subsidy then you do not disagree with the substance of what I have posted.

The one objection I have to this statement is that there are environments (mostly remote locations such as Hawai'i) where it is cost effective without subsidy to install solar power. When I went there this past winter, I talked with the locals and they noted that solar, without subsidy, breaks even or beats 'grid' power on most of the islands. This is independent of the mandate for solar water heating with all new residential construction.

That is almost certainly wrong about Hawaii. They probably read some newspaper article that does not account for all the subsidies correctly. If I am wrong I am happy to admit it if I can be shown numbers.

I do grant there are situations solar is legitimately useful, mostly for remote locations. Satellites are the most extreme example, but there are others on earth. However these are a negligible fraction of total electricity generation, and work only because you're accepting higher costs to be able to operate somewhere not attached to the grid.

[Simon_Jester]

Ok, so we both agree that current solar plants are a waste of money in their own right. Let's see if they are good speculation instead.

I don't think "waste of money" is a good description; it's sort of like calling the Stagg Field atomic pile a "waste of money in its own right." Arguably true in the strictest sense, but deeply misleading.

If solar power is costing us 10-15 cents per kilowatt hour to produce, then yes, at this time it is not cost-effective without subsidy.

Now, you then go on to point out that the market doesn't want to speculate on solar power becoming profitable. I can think of a lot of reasons for this. The obvious one is that

there is no way to make solar cost effective that will work in a sensible time frame that you can use to look a grandmother in the eye and ask her to give you her pension. Even if you work for Goldman.

If you're going to make up ends to people's sentences for them, you might as well indulge your solitary vices in private, rather than in public.

If you seriously expect all advisable projects to receive market funding, you are sadly deluded about how markets work. Prototypes and proofs of concept are advisable projects in many cases, but are NOT going to be made profitable. They are not about getting fiscal return on investment in this facility, they are about learning the lessons that can allow a profitable industry to exist in the future.

Nobody does that in the market except the venture capitalists, and even venture capitalists expect the specific firms they invest in to become profitable later.

So why wasn't pension fund money rolling into solar 20 years ago? Firstly, because none of this justifies building millions of panels. It justifies research to some extent but that was paid by the government anyway, for the cost of peanuts compared to subsidising mass production of panels, and as a result panel prices had been dropping for decades before they became "low" enough to make subsidy at all palatable.

Much of what we'd like to do would fall under improvements in plant design and construction technique, I'd think- that does a lot to lower installation costs. But doing that requires people to actually subsidize construction of plants- you can't figure out a better way to build power plants until you have experience building that type of plant.

Second, because no one with money on the line believes that future power grid is going to be based on solar, subsidy farming aside. Even if the panels become very cheap (which they almost certainly will in 30-50 years, though cost of backing material and power electronics probably caps the price to just a bit below grid parity), you need to time shift the energy and doing that on a large enough scale for a low enough price is looking like a fusion-scale problem.

Personally, I'd expect only a limited percentage in solar, most of which would be engaged in providing electricity during peak hours (conveniently, around noon to 6 p.m.). And honestly, I still expect Retired-Simon to be grateful that his electric bill has been reduced by attempts to improve the efficiency and scalability of solar power technology in the 2010s.

If you agree with me that solar power is not presently cost effective without massive subsidy then you do not disagree with the substance of what I have posted.

The one objection I have to this statement is that there are environments (mostly remote locations such as Hawai'i) where it is cost effective without subsidy to install solar power. When I went there this past winter, I talked with the locals and they noted that solar, without subsidy, breaks even or beats 'grid' power on most of the islands. This is independent of the mandate for solar water heating with all new residential construction.

That is almost certainly wrong about Hawaii...

Obvious cost factors: Hawaii gets very strong sunlight year round, being in an equatorial zone with relatively limited levels of cloud coverage. And it's a remote place, so shipping masses of coal there is not very appealing, and artificially increases costs.

[energiewende]

You are willing to give ground when it seems your position is indefensible, which is good, but in this instance I think you have retreated off the field. If you agree that current solar generation is not worth the cost and that in the future it won't be a major grid component then we are hardly disagreeing. I could nitpick many of the specific claims, but why bother?

The bottom line is that if you want to produce electricity as cheaply as possible you should build coal and gas stations. If you want to do that but with near-zero CO2 emission you should build nuclear stations. Solar will be a peripheral component at best, when research scientists have developed panels that don't exist and cannot be built today.

[TimothyC]

That is almost certainly wrong about Hawaii. They probably read some newspaper article that does not account for all the subsidies correctly. If I am wrong I am happy to admit it if I can be shown numbers.

You fail to realize the high cost of electricity in Hawai'i. The average cost varies per island, but it ranges from $0.35 to $0.46 per kilowatt hour (the lowest costs are on O'ahu & Maui, with higher costs on Kauai & Hawai'i, and the highest costs on Molokai & Lanai) for residential customers. [Source: HECO]. With your own numbers at $0.15 per kilowatt hour, Solar is very effective for Hawai'i - even before you add in the non-economic benefits of not burning oil (and the resultant pollution - note that they don't haul in coal because it's not cost effective).

[energiewende]

It seems you're right about the cost of electricity in Hawaii, which in itself may not be surprising since they burn oil unlike pretty much the entire rest of the world. It's hard to understand why, though. Hawaii doesn't produce oil and I don't think that shipping costs would sextuple the price of coal generation. I suspect some NIMBY/governmental issue is at play here.

[Simon_Jester]

You are willing to give ground when it seems your position is indefensible, which is good, but in this instance I think you have retreated off the field. If you agree that current solar generation is not worth the cost and that in the future it won't be a major grid component then we are hardly disagreeing. I could nitpick many of the specific claims, but why bother?

I would call the primary source of extra daytime power over and above base loading to be "a major grid component." And I rather suspect solar will occupy that role unless we come up with something exotic and clever.

The bottom line is that if you want to produce electricity as cheaply as possible you should build coal and gas stations. If you want to do that but with near-zero CO2 emission you should build nuclear stations. Solar will be a peripheral component at best, when research scientists have developed panels that don't exist and cannot be built today.

Between nuclear being inherently rather more expensive than current fossilpower, increased need for electricity to make the liquid fuels we'll be using in vehicles in the 2050-2100 timeframe, and even modest decreases in cost of solar...

Personally I'm imagining things like nuclear, hydroelectric, and geothermal being used for base load, with a rather large solar component that helps handle the peak demand period during daylight hours. Making that solar component cost-effective will therefore matter.

It seems you're right about the cost of electricity in Hawaii, which in itself may not be surprising since they burn oil unlike pretty much the entire rest of the world. It's hard to understand why, though. Hawaii doesn't produce oil and I don't think that shipping costs would sextuple the price of coal generation. I suspect some NIMBY/governmental issue is at play here.

Does "having a world-famous climate with beautiful clear air that attracts tourists from all around the world" count as NIMBYism? They have very strong incentives not to do anything that makes smog/smoke problems worse.

[TimothyC]

It seems you're right about the cost of electricity in Hawaii, which in itself may not be surprising since they burn oil unlike pretty much the entire rest of the world. It's hard to understand why, though. Hawaii doesn't produce oil and I don't think that shipping costs would sextuple the price of coal generation. I suspect some NIMBY/governmental issue is at play here.

There is a NIMBY aspect to some things on the islands (around Hilo no building can be taller than the tallest palm), but some of that grows out of a population that sees the large commercial growth on O'ahu as something to be avoided, and I don't blame them at all. The islands really are spectacularly beautiful. As for why no coal, other reasons would likely relate to the fact that from what I've learned of the history, Coal was never used as a major power source, so there was never the infrastructure established on any of the islands to support the imports of that quantity of solid fuel, while there is a long history of importing petroleum products (look up the Falls of Clyde, still sitting in Honolulu harbor).

[Sea Skimmer]

It seems you're right about the cost of electricity in Hawaii, which in itself may not be surprising since they burn oil unlike pretty much the entire rest of the world. It's hard to understand why, though. Hawaii doesn't produce oil and I don't think that shipping costs would sextuple the price of coal generation. I suspect some NIMBY/governmental issue is at play here.

Given that only Oahu has a civilian port worth speaking of, and it happens to be in the middle of a downtown Honolulu I think a more basic issue is that of how to even unload coal. You simply can not dock large colliers at any of the other islands , and I doubt building an entire port was ever an attractive economic option. Oil power plants meanwhile can burn really shitty grades of crude. NIMBY might be a factor, but unloading is the real issue.

Oil products are all unloaded at offshore buoys, a very elegant and cheap solution which has no equivalent for coal. You can anchor such a bouy just about anywhere and as long as you have a few hours warning tankers can simply disconnect and steam into deep water to ride out storms. Indeed they really wouldn't have to go far at all since the water is so deep around the islands.

[PKRudeBoy]

What about eventually having space solar plants? While it may not be plausible in the immediate future, advances in current solar power might make it possible a few decades down the line.

[D.Turtle]

Personally I'm imagining things like nuclear, hydroelectric, and geothermal being used for base load, with a rather large solar component that helps handle the peak demand period during daylight hours. Making that solar component cost-effective will therefore matter.

I actually expect wind to take up a large part of that. In contrast to solar power, there is always wind blowing somewhere, so that if your grid is large and good enough, wind is quite capable of always providing a certain amount of power.

[PKRudeBoy]

Personally I'm imagining things like nuclear, hydroelectric, and geothermal being used for base load, with a rather large solar component that helps handle the peak demand period during daylight hours. Making that solar component cost-effective will therefore matter.

I actually expect wind to take up a large part of that. In contrast to solar power, there is always wind blowing somewhere, so that if your grid is large and good enough, wind is quite capable of always providing a certain amount of power.

While there may indeed be wind always blowing somewhere, it might not be anywhere close to where you need it. Unless the windmills are evenly spread throughout the globe, theres no guarantee that they will be where you need them.