Hacker News Reader

Comment by pfdietz | Hacker News Reader

Utility scale PV fields in the US southwest today cost $0.89/W(dc) installed (were you using US residential PV costs instead? Bad boy!) Assuming you were talking of peak power, that's less than $1T for 1 TW. If you were talking about aveage power, the capacity factor there is above 30% now (single axis trackers), so about $3T. How you got $40T from that, god only knows.

You will note that the levelized cost of power from these fields is maybe 1/3 that of your nuclear figures.

> Assuming it takes 10 days to install a solar a 10 kW system, and we manage to install 10,000 solar systems at a time (simultaneously) across the entire nation.

Absurd, since utility-scale systems are orders of magnitude larger than that. There are individual PV fields that have capacity larger than your 10,000 tiny systems.

Your argument from incredulity about manufacturing these systems is not worth anything.

View on HN

I have to point out that you neatly avoided main point in the original post - comparing costs of something that never been done (storing energy at national scale) with something with known price and been done before (powering industrial nation with nuclear)

"Nothing can be done if it hasn't already been done" is not a valid argument. If anything, the large scaleup that would be needed would help, since it would drive batteries way down their experience curves. It's a shame nuclear, unlike renewables and batteries, has not shown good experience effects.

You cant meaningfully compare prices of something that was not ever done with something that is already done and existing.

South Korea would argue with you on the experience part, as well.

> "Nothing can be done if it hasn't already been done" is not a valid argument.

That wasn't their argument. It was that comparing imagined costs to real ones is not valid.

> Utility scale PV fields in the US southwest today cost $0.89/W(dc) installed

That's just not realistic. Not at scale. And, of course, you have to have batteries or the entire thing is pointless. The issue with electric car charging is stressed by power demand. In other words, calculating energy alone does not solve the equation. If you want to charge your electric car in 2 hours vs. a slow 12 hour charge you will need at least six times more power at the socket.

Remember that at higher power (generally meaning, higher current) the power losses increase with the square of the current. Fast charging requires more power than the simple ratio of (time_for_slow_charge/time_for_fast_charge) due to these losses. Super simple hypothetical numbers: If you slow charge at 1 A for ten hours and want to fast charge at 10 A in one hour, your copper losses (energy lost to heating the wires that get the electricity to the plug) are proportional to the square of the current times the resistance of the wiring.

Example:

0 AWG wire can carry 150 A.

It has a resistance of 0.093 ohms per 1000 feet

At 100 A, the loss over a 1000 foot run is: 100 x 100 x 0.093 = 930 W. Almost 1 kW.

This is not meant to be a calculation of actual charging of a vehicle. It's just an illustration of what happens as we face the reality of millions of vehicles fast charging and slow charging on an infrastructure that simply isn't designed to support what we would need to do. I don't even want to think about how much power would be converted into heat on the grid in this all-electric future. Things people just don't talk about when electric cars are put up on a pedestal as examples of a clean "green" future.

Why is it not realistic at scale? What part of it could not be scaled up? There is plenty of space for the PV. There is plenty of space for storage. The US west has effectively unlimited capacity for off river pumped hydro, for example.

The business about cars is weird, as car charging will be needed in a nuclear powered grid too. How is this an argument against renewables? If it's that renewables require more distribution capacity (due to a desire to charge vehicles during the day), then it's also the case that solar can be dispersed to reduce the need for distribution. And if solar + grid level storage provides synthetic baseload more cheaply than nuclear then the distribution argument collapses entirely.

I am not making an argument against renewables. I am saying that the math does not support the idea of making a 100% transition to electric vehicles using just renewables for energy. Nuclear is the only way to get the bulk of what we need without making a mess of things.

You have to consider all the variables that would be affected when you start talking about solar at such massive scales. For example, if I understood you correctly, you are suggesting pumping water to altitude using solar power and then using that potential energy to spin turbines on the way down.

OK, great, sure, Physics 101 problem. Theoretically, yeah, that would work. In practice, you would cause such devastation to the ecosystem doing something like that --again, at scale, not a small lab experiment-- that the entire concept is turned on its head.

The mining and all other things required to make batteries at such massive scales isn't something anyone would want to be around, much else highlight as "green".

BTW, I want electric vehicles. I really do. However, we need to figure out how to do it without hiding a stinking rotten ecological mess somewhere in China so we can feel great about our little "green" electric car on the driveway in Los Angeles.

It's almost like the issue of child labor in China and other places. It's OK so long as our t-shirts are nice quality and cheap? Of course not. Well, the same should apply to the realities of solar and electric vehicles.

Right now my perspective is that we'd be swapping one rotten mess for another. Solar panels and batteries would have to be replaced in massive quantities every 10, 15, 20 years, depending on who you want to believe. At scale, that's billions of, well, everything to trash. Even if we recycle, again, at scale, not a pretty sight.

I don't have all the answers. I probably don't have any answer. I just know that there's a lot of hand-wavy, feel-good, save-the-planet stuff floating around and nobody seems interested in doing the math. It's far easier to convinced the unthinking masses that "green is better" and we need to "save the planet" than to put-up a few slides in a presentation and discuss the reality of having to adopt nuclear power with gusto if we want any of those things to become realities.

In other words, the politicians who have been anti-nuclear can't bring themselves to now be pro-nuclear because that would mean losing votes. Which means we don't do the right thing because politicians want to keep their jobs, not because there are viable alternatives.

Remember that this thread started with an article about China building 150 reactors. Why are they doing that? Because nuclear power, done correctly, is the best option for constant, reliable, consistent and long term energy. If you do that math, including total cost of ownership, this is pretty much indisputable.

Reply on news.ycombinator.com