Hacker News Reader

The problem with "shut up and calculate" is that sometimes you forget to take things into account. People objecting on grounds of land use elsewhere in this thread have been worried about the size of the infrastructure and mines supporting the manufacture of the solar panels. I have no idea what the ingredients of a solar panel are, or how to calculate W/m^2 after taking area for manufacture into account, but it's plausibly "not quite so many". Then you need to start thinking about batteries for energy storage at night, and so on, and the area increases.

It's important to take all of those things into account, and moreover to consider that we may have differing values and may need to compromise; but, in order to do that, we first need to be able to arrive at a correct understanding of the basic facts of the situation. For example, whether there is or is not enough land for solar power to replace current energy production from coal, other fossil fuels, and nuclear power (there is, with almost three orders of magnitude to spare); whether solar panels do or do not contain rare earth elements (they do not and never have); which mines produce products needed for solar-panel manufacture or nuclear power, and in what quantities, producing what quantity of environmental damage; how much lithium is available for batteries; and so on.

Lacking a correct understanding of even the most basic facts of the situation, we cannot hope to reason about the consequences of possible courses of action.

As you can see above and elsethread, I've spent some time and effort to try to figure out some of the basic facts of the situation. I'm certainly no expert — for example, I only found out this year that solar panels include screen-printed silver paste, which accounts for on the order of 10% of the cost of photovoltaic modules and on the order of 10% of global silver production — and so I'd appreciate some help.

Thank you for expressing this better than I could have.

You're welcome! Overall I've found this thread extremely disappointing.

>or claiming that solar panels are made out of rare earth elements (!!)

I responded to so many messages and I tried my best to always refer to 'solar and wind' in tandem (because without wind as a kind of a complement, solar is completely not feasible), and the one message where I got sloppy and referenced only solar panel with a passing reference to 'rare earth minerals' and that's the only aspect that is picked up. OK. Let's move past this. Let's say neither windmills nor solar panels use any rare earth minerals, just for argument's sake. That's not really true, certainly not for windmills, but let's say that it's true ... because that's not the salient point here.

Re: your calculations. You forgot that Solar and Wind need to be overprovisioned because they not only need to collect enough energy for right now, but also enough to store to bridge the daily, seasonal and inter-annual intermittency. There are no batteries capable of that, but if there were, you have to add that in there somewhere as well because that road leads to more land-use and more ecosystem stress.

>This is 550 thousand km², a circle of 420 km radius (groovy, mang!), about 80% of the area of Texas

That is an insane amount of land - I don't understand why so many are so flippant on this point. After all, you're placing high-tech devices in that area. Devices that need maintenance. Devices that break-down. Devices that need to be decommissioned and landfilled and then replaced. If you're covering an area of 550 thousand km² with panels, there is an equivalent volume that you need to dig up somewhere else, ship and process, and of course you then have to landfill, and you have to do that every few decades forever. Right? Are we OK with that? Is the environment OK with that? After all, this is on top of everything thing else we will be doing, like agriculture and transportation.

And this is for TODAY. The world is projected to grow to around 10-12 billion. At the same time, per capita energy use will continue to increase.

>You should be ashamed of yourselves

This kind of emotional appeal is not conducive to good discourse. I made an argument how I see things today. For all the talk about wind and solar being the energy of the future, I don't actually see where solar and wind are actually powering a modern economy. I see a lot hope and handwaving. I see Germany building pipelines to ship natural from Russia and signing multi-decade contracts with that nation while being held up as a model nation. Why not invest in solar and wind deployment instead, especially since Russia is a geopolitical adversary in many ways?

I don't understand how the intermittency gap is bridged given that there is no battery technology capable of storing excess generation enough to last a minimum of a a few days, but more likely several weeks. And finally, I see a world that while growing in population to around 10-12 billion, will exponentially increase per capita energy requirements at the same time. I look at all that, I shudder at the amount of solar and wind collectors that will be required to manufacture and deploy and the stress it will put on existing ecosystems. I don't understand why so many pretend this is a solved-problem, especially since we have an energy-dense alternative, with no engineering challenges to solve and with decades of experience.

(1/2)

> without wind as a kind of a complement, solar is completely not feasible

This is not correct, and there is no reason to even suspect it might be true.

> the one message where I got sloppy and referenced only solar panel[sic] with a passing reference to 'rare earth minerals'

It wasn't a passing reference; it accounted for the vast majority (99.9% to 99.99%) of the resource and environmental impact you attributed to "a typical solar panel":

> A typical solar panel weighs around 40 pounds. All that material has to be mined from somewhere, then transported, then processed. Some of that material are rare earth minerals that have an outsized mining requirement (like for every 1 pound, you need to turn over 1000 pounds or 10000 pounds of earth). All of that material will need to be landfilled at some point in the future as well, no matter how recyclable the solar and wind collectors are (and today, they aren't recyclable at all).

Your describing it as "a passing reference" reinforces the impression that you have no interest in what might or might not be true, only in finding excuses to attack people. The question of whether making a 40-pound solar panel requires 30 kg of rock or 20 000 kg of rock is absolutely central to your purported primary concern here: the environmental impact of mining and waste management. Yet you dismiss this as "a passing reference"? How is it of no consequence to you whatever if a problem you are concerned about is literally a thousand times less serious than you had thought?

The only correct thing in your paragraph quoted above is that a typical solar panel weighs around 40 pounds (18 kg in non-medieval units), and that materials cannot be incorporated into a panel without being mined, transported, and processed. (And some stuff about wind generators I'm not even going to address.) As you can easily verify, a square meter of 100-μm-thick silicon wafer only accounts for 230 grams of that; the other 98% of the weight is mostly glass, aluminum, silver, and copper, which are extremely recyclable and have been thoroughly recycled since their discovery — centuries at least, millennia in some cases.

It's true that nobody's recycling the silicon at present, but that's not because it's difficult — it's that there isn't enough scrap solar-grade silicon out there to recycle, because the vast majority of the panels ever made are still being used, not scrapped.

> You forgot that Solar[sic] and Wind[sic] need to be overprovisioned

I did no such thing. I referred to PV capacity factors explicitly in about half a dozen places in the comment you are responding to. Depending on how you count, about a third of the text in the comment to which you are responding, saying that it "forgot that solar...need[s] to be overprovisioned" is specifically about capacity factors; that is to say, it is specifically addressing the fact that solar needs to be overprovisioned, and calculating by roughly how much. I could not be further from forgetting that solar needs to be overprovisioned.

As for wind, my comment didn't mention wind at all.

My notes on batteries are in https://dercuano.github.io/notes/lithium-supplies.html.

> also enough to store to bridge the daily, seasonal and inter-annual intermittency

No, that is a bad idea. You may need to store energy to bridge daily intermittency, but for seasonal and interannual intermittency you need to overprovision. (Except in the Arctic and Antarctic, where you should probably just get fuel or electricity shipped to you.) Where do you think the seasonal and interannual intermittency problem is worst, and how bad is it? Then we can calculate by how much extra we need to overprovision to handle it, or how far we need to run transmission lines to avoid it. Please stop this vague gesturing at potential problems and provide specific, verifiable data and careful reasoning, as I have done.

> you're placing high-tech devices in that area. Devices that need maintenance. Devices that break-down. Devices that need to be decommissioned

Solar panels do not need maintenance, although they are more efficient if you dust them occasionally. Some of them do break down, but this is unusual, usually due to design or manufacturing defects or very extreme weather. They generally do not need to be decommissioned. Typically they have a design lifetime of 25 years but continue to produce slightly reduced amounts of power for decades after that. https://www.nrel.gov/state-local-tribal/blog/posts/stat-faqs... https://www.nrel.gov/pv/lifetime.html Silicon solar cells haven't existed for long enough for us to see how long they last naturally; the first ones manufactured 60 years ago are still working.

Inverters and charge controllers do break down and need to be replaced every ten years or so, but there are many fewer of those; the maintenance load is similar to that of any other kind of power plant serving the same load.

(continued...)

>You may need to store energy to bridge daily intermittency,

And the greatest intermittency is nighttime. Happens every day. Chemical batteries are not cost effective even if we had the resources.

But sure, continue to just hand wave that off.

I didn't handwave it off. Nobody had mentioned the cost-effectiveness of chemical batteries in the thread so far. I wrote a 1000-word essay about the issue two years ago, exploring the resource ramifications, which I linked from the above comment, and this other 600-word essay here which goes into more details on battery costs and expected internal rate of return on batteries under different incentive schemes: https://dercuano.github.io/notes/energy-storage-efficiency.h...

(2/2)

> That is an insane amount of land

If any of Texas, Thailand, or Botswana can singlehandedly power the entire planet's marketed energy consumption, which is what that "insane amount of land" adds up to, there's no land shortage for solar.

> If you're covering an area of 550 thousand km² with panels, there is an equivalent volume that you need to dig up somewhere else

550 thousand km² times "40 pounds" per m² is only ten billion tonnes, which is the weight of a 1-kilometer-radius sphere of rock. Botev Peak in Bulgaria, say, or the Matterhorn, although of course in practice you'd want to use ten or twenty billion tonnes of sand and bauxite that don't require climbing mountains to get them. Please do the math yourself before bringing up irrelevant trivialities like this.

> For all the talk about wind and solar being the energy of the future, I don't actually see where solar and wind are actually powering a modern economy.

Well, it's possible you're unaware of the relevant facts, so I'll explain.

Until four years ago, solar energy was more expensive than other sources of energy, so modern economies were built on those other sources of energy. Until about seven years ago, solar was much more expensive. Now, solar energy is cheaper. But most power plants are more than four years old; I think the median age is something like 15 years old, 30 years for nuclear plants, 40 years for both US nuclear plants and US coal plants. Typically planning and building a power plant is a process that takes a few years, too, and utilities and regulators are justifiably cautious about innovations.

So let's look at new installations rather than installed capacity.

For example, China built 38.4 GW of coal capacity last year https://www.reuters.com/article/us-china-coal-idUSKBN2A308U and 71.7 GW of wind capacity and 48.2 GW of solar capacity https://www.reuters.com/article/us-china-energy-climatechang.... Taking into account typical capacity factors of 40% for wind, 25% for solar, and 60% for coal, that adds up to 23 GW average new coal, 29 GW average new wind, and 12 GW average new solar. That last number doubles about once every three years. So I think it's fair to say that, even in the coal-heaviest country in the world, the transition to solar (and, I grudgingly admit, even more toward wind) is quite clear: the great majority of new power plants are running on renewable energy.

You may not be aware of this, but Peabody Energy, the world's largest coal company, went bankrupt in 02016. They've emerged from bankruptcy but their latest annual report says they lost US$211 million in 02019, after turning a profit briefly the year before — but nothing compared to how they lost US$2 billion in 02015.

https://www.sec.gov/ix?doc=/Archives/edgar/data/1064728/0001...

Their latest quarterly report from 02020 has them losing US$1.6 billion in 9 months.

One of the factors in Peabody's problems is that in the Southwest of the USA, where some of their major customers are, continuing to operate already built coal power plants like the San Juan Generating Plant and the now-defunct Navajo Generating Station is no longer economic; solar energy has driven down prices (averaging US$26.58/MWh at the Palo Verde trading hub in 02019 according to this article) well below the coal plants' operating expenses (for example, US$44.90/MWh at the San Juan Generating Plant).

https://pv-magazine-usa.com/2020/05/28/record-low-solar-ppas...

> I see Germany building pipelines to ship natural[sic] from Russia and signing multi-decade contracts with that nation while being held up as a model nation. Why not invest in solar and wind deployment instead,

This is another question I answered in the comment you are purportedly responding to: Germany's solar capacity factor is an abysmal 10%.

Let's unpack what that means. A 280Wp solar panel module in California with a 28.1% capacity factor produces 79 watts average (690 kWh/year, worth about US$28 at a wholesale price of US$40/MWh); the same module in Germany produces about 28 watts (245 kWh/year, worth about US$9.80 at wholesale). But the module costs about US$50 wholesale in both places, or US$300 if you buy it on Amazon. So, absent subsidies, a solar plant is a much worse investment in Germany than in, say, Qatar.

Why is Germany often held up as a model nation? Germany was an early pioneer in a variety of energy reforms (the "Energiewende"), including massive investment in utility-scale solar, rooftop solar, and utility-scale wind, as well as energy-efficiency programs like Passivhaus. Still, in Germany it's still cheaper to run existing coal and gas plants than to build new solar plants, so Germany is no longer a leader in this field; the leaders are now countries like China, India, Qatar, and Chile.

Also, I didn't mention this, but Germany uses about 80 gigawatts in only 357 thousand km² of land, largely due to its high population density of 230 people per km². By contrast, the world has about 50 people per km², using 18 terawatts in 150 million km². So Germany, despite its high efficiency, uses about 0.2 W/m², almost twice as high as the world's 0.12 W/m².

So, in summary, Germany has a higher energy demand, higher competition for land, and a terrible solar resource, so it's one of the last places in the world you'd expect solar to be cost-competitive.

> I look at all that, I shudder at the amount of solar and wind collectors that will be required to manufacture and deploy and the stress it will put on existing ecosystems.

If you were to chop down a forest to build a PV farm, it would put stress on an existing ecosystem, but in a desert it will provide shade and windbreaks, and in arable land it will return cultivated land to being a habitat usable by native plants and animals. And those are typically much cheaper than chopping down forests. (They also don't concrete over the site anymore. Too expensive.) And, as outlined above, the mining and manufacturing are of a trivial scale compared to the deployment. So there is no reason to suspect that the environmental effect of solar energy will be net negative in the next couple of decades, even if compared to a nuclear alternative.

It's likely that in 20–50 years, as solar energy production vastly exceeds current world energy production, that it will become environmentally devastating, as people desperately seek to cover any scrap of land or ocean with increasingly efficient solar panels to harness every last little bit of terrestrial insolation. Unless we have greener ways of making such decisions by then.

> we have an energy-dense alternative, with no engineering challenges to solve and with decades of experience

Now that solar is so cheap, the alternatives are too expensive except for niche uses, and most of the alternatives are also wrecking the climate — nuclear, hydro, and geothermal being the exceptions. Even in the 1970s, before Three Mile Island, nuclear power plants cost over US$1 per average watt, which is about twice what PV plants in favorable locations cost now.

> This kind of emotional appeal is not conducive to good discourse

You know what's conducive to good discourse? Basing your discourse on easily verifiable facts instead of easily falsified nonsense is conducive to good discourse. Rigorous arguments based on specific, easily checkable calculations are conducive to good discourse. Responding point by point to what other people have actually said, instead of baselessly accusing them of "forgetting about" one of the primary topics of their comments, is conducive to good discourse. Trying to inundate careful consideration of relevant points with piles of nonsense you haven't done the most basic consistency checks on, thus challenging others to do your homework for you if they want to disagree — that is not conducive to good discourse.

And that is what you were doing.

I don't know how to appeal to you to step up your game, if not emotionally. It looks to me like you're trying to monkeywrench good discourse and prevent it from happening, and you're immune to appeals to your shame.

>One of the factors in Peabody's problems is that in the Southwest of the USA, where some of their major customers are, continuing to operate already built coal power plants like the San Juan Generating Plant and the now-defunct Navajo Generating Station is no longer economic; solar energy has driven down prices (averaging US$26.58/MWh at the Palo Verde trading hub in 02019 according to this article) well below the coal plants' operating expenses (for example, US$44.90/MWh at the San Juan Generating Plant).

And thank you for providing the context for CA's rolling blackouts.

Drive base load generation off the market for the current darling "renewables", then when the renewables that ARE NOT predictable or consistent do the inevitable and fluctuate, the base load plants are no longer available to fill the gap.

I'd be right there with you celebrating the solar win if solar was able to 100% replace the functionality of those plants. But it can't!

Eliminate the plants but don't replace the functionality and what do you get? Rolling blackouts. A common occurrence in CA. Why visit the 3rd wold when you can just bring it home?!?

I don't think this is the full context for CA's rolling blackouts, although it's true that the Navajo Generating Station did sell to California.

The California rolling blackouts were back in the summer, when solar energy production is highest. If reliance on solar were the whole story, those blackouts would have gotten worse and worse from summer through fall and winter. Instead, they only lasted a total of about three hours over two days.

https://www.latimes.com/environment/story/2020-10-06/califor...

> Officials have consistently said that intermittent power sources such as solar panels and wind turbines didn’t cause the rolling blackouts.

> Energy providers collectively under-scheduled the amount of electricity they expected to need. That allowed power plant operators to sell their juice to customers in other states, resulting in thousands of megawatts being exported even as the Independent System Operator warned that rolling blackouts were imminent.

But "officials" turned out to be wrong. Intermittent power sources did cause the rolling blackouts — though they weren't alone.

https://www.utilitydive.com/news/california-releases-final-r...

http://www.caiso.com/Documents/Final-Root-Cause-Analysis-Mid...

> the three major causal factors contributing to the August outages were related to extreme weather conditions, resource adequacy and planning processes, and market practices. In summary, these factors were the following:

> 1. The climate change-induced extreme heat wave across the western United States resulted in demand for electricity exceeding existing electricity resource adequacy (RA) and planning targets.

> 2. In transitioning to a reliable, clean, and affordable resource mix, resource planning targets have not kept pace to ensure sufficient resources that can be relied upon to meet demand in the early evening hours. This made balancing demand and supply more challenging during the extreme heat wave.

> 3. Some practices in the day-ahead energy market exacerbated the supply challenges under highly stressed conditions.

If I'm reading this report right, the immediately precipitating event was that the Blythe Energy Center, a gas plant, went down, so when solar generation had declined enough at 18:38 on August 14, there wasn't enough reserve power, so they started rolling blackouts on half a million people until 20:38, when solar generation was just about at 0.

Then, the next day, storm clouds and lowered winds reduced the power generation available throughout the peak-usage afternoon hours, until the CAISO scheduling coordinator erroneously ordered the Panoche Energy Center (another gas plant) to ramp down generation. So at 18:28 they started rolling blackouts, which lasted 20 minutes until 18:48.

The Governor appealing to people to conserve electric power, and no more big gas generating stations failing, allowed them to avoid rolling blackouts in the following days.

So it sounds like they need more batteries. Two more hours of batteries would have been sufficient in this case (for about 5% of their 45 GW load, about 2500 MW: 5000 MWh, or 16 TJ in SI units), but ultimately you'll need a few days of batteries, over a million MWh for California, about 3 or 4 PJ. (Pessimistically assuming there's no demand response or other storage.)

At the US$111/kWh cost for lithium-ion batteries cited in https://dercuano.github.io/notes/energy-storage-efficiency.h... this means about they would have needed US$555M in batteries in this case, or probably about a billion dollars for the whole facility, and a million MWh of batteries to move California entirely to solar would of course be US$111B of batteries, which could be invested US$11B per year over the next ten years. This is a significant investment, but hardly unaffordable — PG&E, for example, has about US$17B in yearly revenues, and Southern California Edison US$12M, so even the financial savings from switching to cheaper solar energy are likely enough to pay for storage. Presumably as battery prices come down these numbers will shrink further, and there will probably be substantial demand response due to the residential TOU migration described at greater length below.

Part of the problem is that their planning process was structured around peak demand, which is the hardest thing to handle with baseload-type plants like coal and nuclear; but, in this case they handled peak demand fine, even in the face of higher-than-usual loads! The problem was after the peak-demand hour: demand declined, but solar generation declined even faster, which was a thing they hadn't taken into account in capacity planning!

Another aspect of the problem was that the gas plants they were relying on for peaking couldn't operate at their normal capacity at such high outdoor temperatures, so the very heat wave they were struggling to survive in was also impairing their generation capacity.

With respect to the potential impact of things like the Navajo Generating Station and the San Juan Generating Plant, they California's imports were limited not by available generation capacity but by available transmission capacity:

> The imports category includes both non-resource-specific resources as well as resource- specific imports like those from Hoover Dam and Palo Verde Nuclear Generating Station. Total import bids received in the day-ahead market were between 2,600 MW and 3,400 MW (40-50%) higher than the August shown RA requirements from imports. Despite this robust level of import bids, transmission constraints ultimately limited the amount of physical transfer capability into the CAISO footprint.

So, for the reduction in coal generation in New Mexico or Arizona specifically to be a causative factor in California's blackouts, someone would have had to build more transmission lines from Arizona. But the transition to renewables in general, that was a causative factor.

(However, it's not entirely clear to me that all the transmission lines were operating at capacity — they emphasize the California–Oregon lines.)

One of the big things they're doing in response is moving residential customers to time-of-use rates, which means that they'll be able to save money on your electrical bill by using energy during the day instead of at night. That should make it possible to sell appliances that save money by, for example, freezing ice during the day and using it for cooling in the afternoon and at night.

It's hard to tell through the insane level of CYA bureaucratese in this report but it looks like they're also installing more battery capacity.

Thanks for the great posts.

Have you seen any numbers for the life expectancy of grid connected batteries? I heard A solar project in Hawaii expects to replace the batteries every 5 years with daily cycles. So for California it would be $111 million /yr in battery costs. Manageable, but not a trivial amount to be factoring in to the cost of firming up the capacity of wind/solar.

You're welcome!

The life expectancy is a really important point, and one I hadn't taken into account in my calculations! The thread at https://news.ycombinator.com/item?id=26231021 talks a bit about this. Laptop Li-ion batteries typically last only a few hundred cycles, but presumably you can do better than that if that's what you're optimizing for; Sir Bearington claims 1500 to 2000 discharge cycles, which sounds vaguely plausible and matches the 5 years you're suggesting, but I'd be interested in a deeper dive. Do you know the name of the Hawai‘i project?

I think US$111 billion over 5 years works out to US$22 billion per year, not US$111 million per year, which is probably just past the "manageable but not trivial" level.

I mean, suppose California needs 45 GW of power at the worst moments, as the CAISO report linked above suggests, and you want to provide that with nuclear power at a ballpark 01970s figure of US$1.50/watt (rather than the current contentious figures of around US$7/watt). Further, let's suppose that USA nuclear power's typical 90% capacity factor https://en.wikipedia.org/wiki/Capacity_factor#Nuclear_power_... is not something you can control (for example by scheduling refuelings to not coincide with summer) and that opex is zero. So you need 50 GWe of nameplate capacity, which hypothetically costs US$75B using 01970s practices and wages. If we simply divide by 30 years (rather than using debt financing and calculating an IRR) we get US$2.5 billion per year. At modern USA nuclear construction costs of US$7 per peak electric watt we end up with US$17.5 billion per year, which would still be less than the US$22 billion cost of replacing US$111 billion of batteries every 5 years. Debt financing makes these numbers a little better, but not a lot, and it reduces the cost of short-term projects like 5-year batteries a lot more than it reduces the cost of medium-term projects like 30-year nuclear plants.

(Does that seem backwards? At 5% interest, an infinite-lifetime annuity can be purchased for 21 years of its yearly earnings. If the annuity expires after 30 years, like a nuclear plant, the NPV is reduced to 16 years' worth of earnings, and if it expires after 5 years, like a battery, it's worth 4.5 years of earnings. So if you debt-finance your nuclear plants you end up paying 46% of your money to the bank and only 54% to GE or whoever builds the plants, while if you debt-finance your 5-year-lifetime batteries you pay 91% of your money to Panasonic and only 9% to the bank. So a 5% yearly discount rate drops the value of nuclear-plant joules by about 40% because most of them are so far in the future.)

Right now, batteries and other storage are only used a couple of hours a day at most, rather than the 22 hours or so that add up to my "million MWh" figure, so the battery price is actually an order of magnitude lower. (And we can expect that when we're just paying the cost of recycling old batteries into new batteries, rather than the cost of mining all that lithium, the price will go down.) So, right now, Li-ion batteries are far more economical than nuclear plants, but to bear the base load over days of low sun, either:

· Li-ion batteries will have to get substantially cheaper;

· wind, other forms of generation, or other storage technologies like lead-acid and compressed-air storage would have to be able to supply most of the load, as gas and coal do at present;

· some of the batteries would have to be borrowed from other uses (such as electric vehicles) that pay for most of their cost;

· demand response and efficiency improvements will have to be able to cut demand by a factor of two or three; or

· some combination of these.

For example, you could very plausibly imagine a conjunction of Li-ion batteries getting 25% cheaper than the US$111/kWh (US$31/MJ) price cited above, down to US$83/kWh (US$23/MJ); wind supplying 25% of power demand when the sun is down or clouded; drawing on an additional 30% "spinning reserve" of parked Teslas feeding energy back to the grid; and demand response reducing off-peak power usage by 25%. So our 45 GW of power demand drops to 33.8 GW from demand response; parked Teslas supply 23% of that, and wind supplies another 25% of it; and so the demand on the utility-scale storage plants is only 18 GW. 24 hours at 18 GW is 430 GWh, which costs only US$36 billion rather than the US$111 billion cited above.

We can hope to do a lot better than that, but it seems like an eminently achievable set of improvements.

Transmission loss? How are you getting the power generated from "empty" land to where it's actually needed?

The easiest way is to distribute that Thailand-sized area of solar panels around the globe, close to the point of use. This won't help Germany, which will still need either 2000-km-long transmission lines from Extremoduro or Algeria or something, or a lot more 10%-capacity-factor solar panels than they already have, but for electrical power in most of the world, photovoltaic power tends to decrease transmission losses (which are already included in the world marketed energy consumption numbers) rather than increase them. That's because thermal electric power plants tend to be in the 100MW–10GW range where they're most cost-efficient, while PV generation can be economically scaled down to 0.3 kilowatt at present.

Indeed, if the prices of modules and power electronics keep dropping as they have been, there's a point where it becomes uneconomic for most energy users to pay for offsite generation, transmission, distribution, and billing; around 02013 there was a lot of talk about the risk of a "utility death spiral" as more and more grid users deserted the grid for their own private PV farms, leaving the remaining grid users (steel mills, diamond foundries, Bitcoin mines, and the like) to pay higher and higher prices to maintain the stranded infrastructure. So far this has not materialized and probably will not.

However, of those 18 terawatts, about a third are transport fuels, mostly gasoline for cars and diesel for land shipping, but also passenger trains, bunker fuel for cargo ships, and jet fuel for airliners and air shipping. Although ships can plausibly tow their own PV rafts, the others will probably be the last strongholds of fossil fuels due to the low efficiencies of electrolysis (typically around 60%) and the Fischer–Tropsch process (typically 50%), which adds up to "transmission losses" of around 70%, relative to how they're measured in the IEA's report today. And if you're doing this to power your Antarctic research base's electrical needs, you lose another 60% (total losses: 88%) to Carnot.

The good news there, though, is that the electrical part of the equation gets a "Carnot boost": a kilojoule of coal or gas burned to generate electricity only yields 400 J of electrical energy. So we only need 400 W of PV generation to replace each kilowatt of coal mining for electricity!

Okay, I've answered three of your questions. Now it's your turn. How much are current transmission losses typically, and in maximal and minimal cases? How are they affected by transmission-line distance, and how much does the step-up and step-down equipment cost for different voltages? How does HVDC help? Are there better alternatives to water electrolysis followed by Fischer–Tropsch? How does the global distribution of current human energy use relate to the distribution of solar resource?