Inverted perovskite solar cell breaks 25% efficiency record | Hacker News Reader

Inverted perovskite solar cell breaks 25% efficiency record

news.northwestern.edu

123 pointsgeox54 comments

Note that at this point in time it is economics, not yield-per-area that dominate any solar cell installation so a new cell doesn't necessarily have to be high yield, it has to be better when you compute total power generated for a given $ invested over the lifetime of the cell. That's pretty much all that matters, everything is just a distraction unless you are severely space constrained a couple of percent more or less isn't going to make a huge difference, and cost savings on existing processes have now compounded to the point that people are more and more looking at the price of inverters and installation as important factors. That's also why you see the push for transformerless inverters (those are bulky and costly so they affect both inverter cost and installation/shipping).

I did a pretty big solar installation over the last two years and I decided to go for quality panels because the glass-glass sandwich panels have a much better economic picture than the glass-plastic ones even though they are more expensive up front.

can you elaborate a bit more on your setup?

6 strings of cells, hooked up to two inverters, one handling two banks on a very high roof, peak production there is 4KW, one inverter handling four strings on covered outdoors space and the garage. The last four are at very shallow angles, so less than ideal but total surface area really makes up for that. We draw about 300W on average and the peak production capacity for the whole setup is close to 15KW (or even slightly higher in full sun in March). Over the last 11 months the system produced 12 MWh on about 4 MHw consumption so net 8 MWh returned to the grid.

34 of the panels are brand new 370W glass-glass panels, the remaining 16 are much older (probably a decade old, maybe more) that still work pretty well rated at 265 W each (but probably closer to about 220 or so).

What's your geography, and what'd it cost up front?

Netherlands.

Panels cost about 210/each for the new ones and the old ones were pretty much free. Inverter: the one upstairs is a new one, 500 Euros ('second chance'), three phase 4KW, the one downstairs is a monster, it's 17KW with up to six strings and cost another 500. The panels on the high roof were installed, about 1000 to do the job and the rest I did myself, add another 500 or so in cables, tubing etc. The power bill went from 8500 / year to 3500 / year, and all of that is the remaining gas usage. We're looking into using any surplus in nov/dec/jan/feb to help heat the house to further offset the gas usage but the economics aren't really there yet as far as I can see.

And even in November we are still offsetting about 75% of our electricity usage, which somewhat surprises me, I'd expected far worse.

thanks!

You're welcome. I've done a bunch of these by now if you are ever in the market for a setup and want to bounce ideas of me then feel free to mail, email is in profile.

> We draw about 300W on average

Dude, teach me. Forget solar panels, if you can explain how you're living on ~1/10th the power of the typical home around here I'd power my house on a hamster wheel.

Lol. Ok: first, you get 25 hamsters...

No, more seriously: we started off by pulling all the breakers and to go 'from scratch', then every time I re-connected a breaker I monitored the increase in power draw. This took a bit of time but after a while we figured out there were a number of really bad consumers: my PC (which had a very high end graphics card in it from a machine learning project a couple of years ago), the NAS (which had 12 bays), an older fridge (though I thought it was quite energy efficient, but this really wasn't the case), a water heater hidden under the kitchen cupboards (that one took forever to track down), some smaller loads but still considerable (a couple of studio monitors, game computer that was on standby when not in use, a large studio mixer, a massive pump for the infloor heat that worked just as well (or even better) with the lowest setting as the one that it was on) and then a large number of really small loads that were always on but rarely needed.

Removing all of those saved us 2/3rds of our power draw, we went from 30 to 35 KWh to around 8 to 10 KWh / day. After that it was relatively easy to get to '0' using solar panels and another batch of them and we were suddenly offsetting our gas usage with surplus electricity (which the power company buys).

No hamsters were harmed in the preparation of this message.

Thanks for this. I'm averaging a kw in a small house and don't understand where it's going. When we put in a hot tub, it only raised the monthly bill about 10% (it's worth that). Gas heat, gas water heater and stove. Most lights are LED at this point. Using laptops, no big graphics machines.

I need to do what you did, though I was thinking about hacking up a clip-on ammeter and some monitoring software to examine the various breaker branches.

> I was thinking about hacking up a clip-on ammeter

The easiest spot is right on the distribution wires, they usually are pretty beefy and that means you don't have to re-hang your meter after every breaker. They're also very well insulated so the chances of shorting something out diminish a lot (but if you do the effects will be far more spectacular ;) ).

In the end I installed a 'Shelly' three phase current meter permanently with the pickups around the mains wires. That gave me very precise logging (and given that the distribution board passes those three wires roughly balanced out across the breakers it is already quite fine grained). It also allows you to spot intermittent consumers and for the money I wished I had installed it earlier during my hunt.

Oh, as for laptop: I ended up buying a second hand W540 which is pretty beefy when you want it to be but normally it sips power (about 29 watts continuous draw, the screen is off because I have two external monitors connected). Oh, and about those monitors: I've reduced the backlight intensity from the default quite a bit and that made a real difference as well (besides the usual power saving and desktop lock settings).

Well, installation prices correlate very well with efficiency. What was always a huge problem for perovskite cells, because they tend to be way less efficient.

They also suffer from lack of raw-material abundance and shorter lifetimes. But it may be possible to fix both of those too. The factory for perovskite cells should require much less initial investment, what is a huge limitation for silicon cells, so it could be revolutionary.

Why I’d installation prices correlate with efficiency?

If you want a 1 GW solar farm you need to pay people per hour which is essentially per panel to physically install them and pay for mounting equipment etc. At 20% efficiency you need 5% more of that labor and mounting equipment than at 21%.

Not just more panels but more wiring to take care of. And more square meters of panel to clean.

And more ground to buy for a given yield.

Ahh I’m with you, i mistook your comment to mean that they were positively correlated with installation prices as opposed to negatively correlated.

At least in my house in Mexico, solar panels were mounted on custom made structures that maximize your house's power generation. That can't be cheap in a high labor cost country. The panel might be cheap but mounting it is expensive.

Grid-scale installations can probably use anything.

However, more efficient cells will probably find new uses where space or weight is very limited.

There are some possibilities that could be opened up at certain efficiency thresholds, though, due to considerations that aren’t purely economic.

For instance, if small panels had high enough efficiency you’d probably see a lot of people who wouldn’t consider roof installations popping them up on their back decks and the like since it’s far less involved to do so. It’d also help work around rules like the one my HOA imposes, where roof mounted solar can’t be street visible.

Yes, when you are severely space constrained it could well make sense. But 25% isn't a big enough difference to enable that. You'd have to do much better than that.

> It’d also help work around rules like the one my HOA imposes, where roof mounted solar can’t be street visible.

Bloody hell, some people just want to watch the world burn.

Wonder if they'd allow a rooftop similar to Tesla's solar rooftop. I totally think HOAs are pretty ridiculous in general.

> It’d also help work around rules like the one my HOA imposes, where roof mounted solar can’t be street visible.

I feel like if we cared about climate change at all there would be federal (global?) laws with severe penalties making rules like the one above highly illegal.

What is disgrace

Definitely a good thing to put on the list of things to complain to legislators about.

For those that don’t have such a list, I recommend you start, otherwise the only opinions they hear are from people paid to capture their attention.

This is really the exception, not the rule. and at least 21 states have laws invalidating it. Mainly the HOAs are there to regulate DIY'ers from having wires hanging down or ladders permanently installed on the house so they can go up there and tinker 3 hours/day 5x a week.

> roof mounted solar can’t be street visible

"Look they have solar panels on their roof. There goes the neighbourhood."

We actually got some comments about that on our house. I asked the neighbors that made the comment if they were willing to put up the difference and I'd be more than happy to take them down. So far no takers ;)

You can be very space constrained… ironically… in space. Well, needs to fit in the rocket.

Yes, but that's not even close to the best that can be done for space applications.

For that there are multi junction cells, with 30% readily attainable (at a cost, but still, you can just order them if you need them).

https://en.wikipedia.org/wiki/Multi-junction_solar_cell

> Note that at this point in time it is economics ... That's pretty much all that matters, everything [else] is just a distraction

What? This doesn't make any sense. How do we invent new things if economics is all that matters? There's a whole process. You prove a thing is possible (literally the thing happening, nothing else), then move onto making the thing reliably reproducible to demonstrate, THEN optimizing for being cheap. It can fail at any point here (and there are more between tbh). This is a really long process. An example might be NASA's TRL scale.

I mean yeah, it is university press and they hype it up, but I'm pretty sure a big reason for that is because comments like this. Being shitty at something is the first step to being kinda sorta good at that thing. Context matters a lot and the context here is: done by researchers, university research, journal publication. This is not Apple coming out and saying they invented something that's going to be on the market in a few months. This is someone at Apple's research lab being like "Hey, that thing we've been researching? I reached a milestone, we shouldn't call it quits yet."

'at this point in time'.

This is now a mature market and mature markets are governed by economics unless you manage to make a massive step forwards. So if that x% improvement comes with an increase in total cost over installed life > x% it's a non-starter.

Solar is a mature market, sure, but that doesn't mean every technology related to solar is mature

That's fair but it does mean that those new technologies - especially if they aim to replace other technologies - will be tested on that parameter. And the difference would have to be large enough for a party to take the risk to go to manufacturing and then for other parties to take the risk to buy the product.

Sort term? No. Long term? Sure. If you're only planning for the next 5 minutes you're not going to make it very far. It's not like these people aren't well aware of all this. That's in fact the motivation behind the research. No one researches things that are useless, only things that seem useless to people who don't know. Again, this is a university researcher, they're kinda just doing their job... I don't know why this is even a contestable thing...

Because that's how the typical arrival of improvements in solar power are usually hailed. I see announcements like these to the tune of one every couple of weeks and they rarely make it to production and if they do they end up as niche rather than mainstream. But some of them do make it. In the meantime, the steady drip of cost reduction due to economies of scale make silicon based solar panels harder and harder to compete with for new technologies, not unlike what we've seen with other semiconductors.

Yeah, again, university. That's typically at least a few years from production if not a full decade. The context matters.

But if we're complaining about the hyping of science, then I'm with you. I haven't been shy about my feelings, as someone finishing my PhD, about academia on here.

I mean, you have to do something in an expensive fashion before you can do it cheaply.

Has this solved every problem? Obviously not. Not even close. Its still an important research milestone.

The problem with a perovskite solar cell is the fragility and life span of the solar cell. If they can stablize the material to last 20+ years it would be entirely worth the cost and development of the solar cells. But right now it is expensive.

They last about 5 years if you're lucky. Although there has been improvements since I last ready, but grain of salt.

I'm fine if they last 5 years, as long as they are a quarter of the price of the ones that last 20, are easily recyclable and have minimal rare earth materials.

They are not at the moment. They are just more efficient. Nirmal solar panels are 15-18%.

Putting it that way, if they are more efficient and cost is the same and they last just as long, then the return on investment will be faster. So it's a win.

As folks are saying, there are already commercially developed panels at 24.7% [1] that should ship within the next year. The existing fleet of Maxeon 3 are apparently going to be 24% this quarter, rather than the 22.7% in the actively shipping parts [2].

Like cache miss ratios, it's hard to remember that 25% is a 10% improvement in efficiency over that 22.7% (and 4% over the 24% claim). But as jacquesm mentions, what matters for system design is either $/watt (installed!), total generation, or whatever other metric. Efficiency might get you more panels on a roof, but so could different panel sizing! One more panel because the sizing works out could easily be 5%.

tl;dr: it's interesting research, but shouldn't affect purchasing decisions.

[1] https://mediaroom.maxeon.com/2023-06-01-Maxeon-Solar-Technol...

[2] https://maxeon.com/us/sites/default/files/2023-05/sp_mst_112...

I was interested until I saw the .edu url. Just another regularly scheduled biweekly battery/renewable/cleantech PR

https://www.science.org/doi/full/10.1126/science.adk1633

Thanks for the full link, but kind of proves the point. We're just months removed from Science publishing room temp superconductivity...

One paper tells other researchers to check it out. Doesn't mean it's:

- reproducible

- scalable

- functional in the real world

That's an unreasonable standard. Things can be interesting without them being a full-blown rolled out product.

Then I'm afraid you don't understand how science works

Did perovskites managed to solve the problem with usage of lead?

Any perovskite article is utterly worthless without some discussion of lifetime/endurance in the wild.

I will believe it when I see it being sold on Amazon and AliExpress. It won't matter if they did 99% efficiency if it never reaches the consumers.

Some people are here for the science, others for the business angle.

This isn’t cold fusion or superconductivity. If you’re interested in knowing the state of the commercial solar industry 5-10 years in the future, these articles are extremely useful. If all you want to do is slap down a credit card you don’t need to read HN.

Reply on news.ycombinator.com