Hacker News Reader

Comment by ars | Hacker News Reader

> One is that two other carbon-free options, solar and wind, are dropping in cost

What's with that myth? They are certainly not "carbon-free" (presumably you mean carbon dioxide free).

Solar and wind use huge amounts of energy to build. The most common way they make money is be spending energy where it is cheap, and selling energy where it is expensive. (But the CO2 is global, as is the pollution.) In essence they are a complicated way of shipping energy.

They have a positive EROEI, but not hugely, just somewhat.

You make a good point on the financial costs of nuclear.

View on HN

"Solar and wind use huge amounts of energy to build"

That's just patently false. And without being too insulting, stupidly so, in a really obvious way.

A 300 watt panel without obstructed view in an average climate is going to get something on the order of magnitude of:

10 hours of sunlight / day 250 sunny days / year 20 years of service

so, 15 million watt-hours, or 15 megawatts hours.

Per wikipedia it takes 400 kilowatt-hours to produce 1 ton of steel in an arc furnace (http://en.wikipedia.org/wiki/Electric_arc_furnace).

That single panel can produce enough energy to produce something like 35 tons of steel over its service life.

Just looking at a state of the art PV panel (http://us.sunpower.com/sites/sunpower/files/media-library/da...) - note, its capacity is 345 watts - it weighs 41 pounds. I am fairly confident that, outside of it being made of pure energy, that the EROEI is pretty decent.

edit: Also, I can't imagine this is a favorable comparison to nuclear, which requires immense amounts of energy-intensive steel and concrete, as well as a fuel that is difficult to mine, process, purify, and separate.

Every one of your numbers are not realistic.

You won't get 10 hours of direct sun, you're lucky to get 6. You are also unlikely to get 20 years out of it - certainly not with zero maintenance (i.e. zero cost). And you are ignoring conversion losses.

Assuming 6 hours and 10 years you get 4.5MW out of it, and let's ignore conversion losses. 4.5MW of electricty at $0.15 per KW/h costs $675.

Installed cost for a 300 watt solar cell? About $500. Not including the inverter.

Cost is a reasonable proxy for energy usage, so your EROEI is about 1.35. Even assuming your figures you have an EROEI of about 4. Less once you include maintenance.

i.e. huge energy costs.

PS. Don't confuse my numbers for financial ROI - if you want that you need to include the time-value of money, and once you include that solar cells loose money even by your measure.

I have solar panels. The A/C conversion efficiency is about 85%. The sunlight doesn't have to be directly incident to generate power, and 10 hours is a perfectly fine estimate, thank you. If anything, it's conservative, like every number I used.

You are pulling the 10 years number right out of your rear end. Many people on old-generation panels are doing 75-80% (above projection) on 25 year old panels. In fact, most panels today are warrantied to 60-80% of nameplate capacity in straight line degradation to 25 years.

Your $.15 power is retail, not wholesale which would be $.02. The $500 installed is also retail, not wholesale. And it is a ROI calculation at that point, not EROEI. I seriously doubt the $500/panel price has any relation whatsoever to the energy requirements to produce a single panel. And I seriously doubt the financial ROE of current panels has any relation to the scaled out EROEI of solar PV infra in the future.

I would spitball the material costs as effective proxies of input energy costs: Plastic, glass: effectively free Copper 1 kg: $7.00 Aluminum 3 kg: $5.00 Silicon 1 kg: $40.00

Now, just randomly cutting my original estimate to 10mWh to humor you and the correct wholesale electricity price of $.02/kWh, the wholesale value of the power is $200. So closer to 300%. I would guess one tenth to one twentieth of the input energy costs of a typical inverter would be less than $10, and even then it's over 250%.

You completely ignore grid transmission losses on your side of the argument, which are 70% or so.

I'm surprised you don't bring up grid level storage, which is solar PVs weakest point, currently.

> I have solar panels.

Did you pay for them yourself, or did you get subsidies?

> I seriously doubt the $500/panel price has any relation whatsoever to the energy requirements to produce a single panel.

It's actually pretty much dead on. Just about every single expense, at the end of the day goes to energy. At the most basic, most fundamental, what else is there to spend money on? Minerals are free - they are just sitting in the ground. The costs is the energy costs of getting them out, not the cost of the mineral itself.

Any costs of making something are the costs of the parts, or the machine, or the transportation, etc. Labor costs at the end of the day go to someone buying those items which fundamentally boil down to energy costs.

> I would spitball the material costs as effective proxies of input energy costs: Plastic, glass: effectively free Copper 1 kg: $7.00 Aluminum 3 kg: $5.00 Silicon 1 kg: $40.00

So where does the rest of the money go? It goes to pay for other energy uses you have not included, that's where.

> the wholesale value of the power is $200.

That's worse for you you know. You are telling me that a panel costing $500 only makes $200 worth of electricity over its entire lifetime?

> You completely ignore grid transmission losses on your side of the argument, which are 70% or so.

That's one of the reasons I used retail prices.

From your reply I think you did not in the slightest understand what I wrote, since your arguments are not helping your case.

AMAZING. That $2000 bottle of Margaux takes over 600 times as much energy to produce as Three Buck Chuck. I knew it tasted better for a reason.

edit: and fine, be a pain. Let's make it apples to apples and make it an ROI comparison. So $500 you can keep the panel, but I keep the retail power cost of $.15/kWh and will also increase retail prices by 4% over a 25 year period. I'm going to use 10mWh again. So 400kWh/yr, or $60 in year 1. In year 1 dollars, that is $1500 over the life of the system. However, using a handy compounding interest calculater, the 4% annual increases in power cost increase the lifetime cost of non-solar power by a factor of 2.5, so that's $3750. So we have a 600% financial ROI. That's unrealistic for solar and I'm ignoring the NPV, but I'm just working with the numbers you handed me.

> That $2000 bottle of Margaux takes over 600 times as much energy to produce as Three Buck Chuck.

Profit is not exempt from my argument. Because at the end of the day profit gets spent, and by paying for the more expensive item you are contributing to energy use once that profit is spent.

Some expensive (in money and energy) item that otherwise would not get purchased has now been purchased because the profit enabled someone to afford it.

> and will also increase retail prices by 4%

That's fine, but I'm not following your math. If you made $60 the first year, and prices went up 4% each year (inflation presumably), your total generation after 25 years is $2,499 worth of power.

Now lets take our $500 and invest it at 6% (which is pretty low), after 25 years you have $2,240.

Or in other words your ROI is virtually nil.

Reply on news.ycombinator.com