Comment by epolanski | Hacker News Reader

Comment by epolanski | Hacker News Reader

Parent Full thread

I have worked as a solar scientist researcher for some time, have coauthored 3 papers in Michael Graetzel[1]'s laboratory and I have some experience in both Graetzel cells and perovskite solutions like the one mentioned in this article.

Long story short: scientific researchers, especially those in well funded laboratories are incentivized by the wrong metrics (in all fields, not unique to solar) and in this field the only metric that matters is efficiency. It doesn't matter if your cell decays by 50% in a day, it reached some great potential in a new way, here's the way for a high impact publication.

Want to work on the other problem though? Lifetime, resiliency? Cheap and affordable and non-toxic materials?

Good luck getting proper funding and exposure. Why? Because scientific papers are a closed mafia, where a set of the most influential scientists (doesn't matter how many scientists the planet has, the moment you start entering a niche the number of people is very low) in their field review the submitted papers. Don't have high efficiency or you're not breaking science? Forget a high-impact journal. You're back at B-tier, C-tier papers, but those won't get you funding and status. Not great for your career.

So what do you do? You play fool and focus your efforts and many many millions of euros to get the next perovskite cell that can reach high 20%s with some twist at least or go for the 30%+ ones.

Just to express how sad and toxic the world of scientific research is (not even mentioning the insane amount of fake data that gets published every day, the politics, etc): instead of being a researcher I now prefer being a web developer writing forms list and tables.

[1] https://en.wikipedia.org/wiki/Michael_Gr%C3%A4tzel

Crystalline silicon cells are already inexpensive, non-toxic, and long-lived. So it makes sense that most researchers are looking for higher efficiency. Back when purified silicon prices spiked around 2007-2008 there were a lot of efforts looking into more affordable (if less efficient) thin film based materials, but those fizzled out as wafered silicon got cheap again and kept widening the efficiency gap over e.g. amorphous silicon.

Since there's also a much bigger solar industry today than in 2007, you also see more practically-minded research coming out of corporate centers (Longi, Oxford PV, GCL System Integration, and others working on perovskite cells and perovskite-silicon tandems). Since they care about shipping real products, they are focused on solving lifetime and durability issues. Of course since this research can give a big commercial edge as single junction silicon reaches its efficiency limits, you'll also only see the really promising work published after it is patented.

Silicon is far from non-toxic to produce (requires insane amount of heat and chemicals to be purified) and dispose.

All these factors are reflected in the price and yet ordinary silicon PV panels are still dominating the market.

Environmental, social and ecological issues are externals and are basically never included in the price.

Could you list some of these chemicals? A quick google search only came up with pretty benign standard industrial chemicals. Disposal is a complete non-issue since the cells are inert and non-toxic (though recycling probably makes more sense than dumping them somewhere).

> So it makes sense that most researchers are looking for higher efficiency.

It makes sense that researchers should be considering metrics other than efficiency as well for new materials or architectures, and there's no reason why this is a uniquely industrial pursuit.

Grandparent's comment resonates - efficiency claims seem to be the only ones generating press releases.

I am not a solar researcher. I'm just an average-joe solar user. I've been doing my own research into "what do I want" for a year now.

One thing I've found frustrating is that solar installers don't seem to understand, or want to share, the fundamentals. They do it one way, the cookie cutter way.

For example, most sites are space-constrained and that leads to concern about efficiency. I'm less space constrained, but that doesn't seem to change their ideas.

Most people have fixed-pitch roofs, so angles are what they are. Being flat, I choose the angle, and direction. There's very little data though on the best angles or the best direction. Are adjustable angles useful? Should I bias the direction to the west to compensate for morning/afternoon consumption patterns? All i get are blank stares.

On the up side, being forced to figure it out is interesting. And over the next few years I can generate actual data, testing various approaches, to see what actually is the most useful approach.

Sure, panel efficiency gets all the glory, but there's a lot more involved once you start building a system, and I think more practical research could be done there.

[1] one interesting point is the cost of the frame versus the cost of the panel. If I just "lie the panels flat" the frame price is negligible. Sure the panels are less productive, but Does spending the frame money on more panels offset that? Is overall production (not effeciency) better or worse? Factoring in summer as well as winter production.

[2] my city allows me to feed power into the grid, giving me about 40% credit. They are effectively a very large, 40% efficient battery, with no capital cost or maintainence cost. Plus they are long term, I can "charge" the battery in summer, use it in winter. I've yet to find an installer though that understands why this is good - they all balk when I say 40%.

FWIW I did my own solar installation, 50 panels in four different orientations (set of 18, two sets of 8 and two more sets of 8). This takes care of spreading the peak around solar noon at the expense of some of that peak. My panels are at 'inefficient angles', but they start generating very early in the morning and they finish really late.

When you look at it from a financial perspective in a net metering setup it is a bit less efficient, but when you look at it as if netmetering no longer exists (which I expect will happen here soon) then it suddenly comes out way ahead. It also has the added advantage that it avoids generating a lot of power when the power price is potentially negative.

My daily surplus in summer months is 80 KWh give or take, but during the winter, even with all this generating capacity I'm still running short. But proper insulation helped to cut down the gas consumption considerably and now energy costs for the whole house (about 2000 sq ft, freestanding) is < 300 euros / month including last winter's gas prices. Next winter should be better still (because I now have 18 more panels).

Total installed capacity is 16 x 265 (older panels) and 34 x 370 (newer, glass-glass panels). Inverter capacity is 4 KW on the high roof and 17 KW on the low one, with about 12 KW of output during the peaks in the summer.

Best day of the year so far made 99.8 KWh, worst days can be a few KWh so then you really need the grid. Average power draw of this house when we're careful is ~300 Watt, but cooking is a pretty big (and usually short) exception to that as is running the water cooker (but that's only a minute or two and during the day doesn't begin to approach the amount of power generated).

One downside of having panels flatter is that they foul up, some 3d printed clips helped with that but it isn't perfect.

Great info thanks. I'm starting with 20 panels, 480w, and a 16kw inverter, so room to add more panels next year if it makes sense. I figure after 2 years of actual data I'd be able to design the perfect setup for this building :)

I've heard of the flat-flat problem of pooling water, so I'd likely be at least 10 degress of angle.

Those little clips make a big difference. They use capillary action to help remove the standing water from the edges and work like a charm.

20x480 is a pretty good system to start with! And yes, insight comes with time, I'm lucky in that I already did a ton of this stuff while in Canada 20 years ago so now I can re-use that knowledge. One of the better things I did here is to create a covered space next to my garage that uses the panels as roofing. Two birds with one stone: very nice covered space and glass-glass panels allow some light to shine through and it helps keep the panels cool because there is plenty of airflow underneath them. Mounting them flush with some polymax in between to make it all watertight was a bit of a job though.

I disagree that science has the wrong incentives. The place for science is to chase new (e.g.) physics, not build something useful.

As a society we have decided that (e.g.) physics should/shall solve our problems with better technology. But that is not what pure academic science is about, or what pure academic scientists care about.

You want to iteratively improve technologies? Give more funding to proper engineers, not physicists/chemists, even if they're sometimes in the 'school of electrical engineering'.

Lifetime, resiliency, non-toxic structure are different dimensions to efficiency and there are likely trade offs between them. Those all can involve what you call “new physics” but parent points out only one gets funding: efficiency. Which parent also points out is suboptimal for society.

Parent claimed only one gets funding, but that is wrong. Industry, capitalist companies work on the other stuff all the time. Parent is likely wrong about researchers not getting funding for that other stuff also, parent likely had trouble getting funding himself due to bad attitude and blames it on other factors.

I like how you decided to trash years of my life and experience and made up a scenario where I'm motivated by envy or failure (and even faulty of bad attitude) behind your keyboard because my experience doesn't fit your views.

If you've been in this field, I'll gladly listen to your experience.

What do you think of MEG/singlet fission cells?

Are you in the Bay Area? If so, we should plan a coffee.

Seconded. It's weird to see your clearly well motivated and insider perspective trashed like that.

FWIW you've helped me understand some of why we keep seeing these efficiency publications that never make it to production. What is impressive is the degree to which cost has reduced as a result of economies of scale and I wonder what the bottom is there. Compared to the price of solar panels even 16 years ago (when I did my previous large installation) it is amazing to be able to get panels at todays' prices for what I perceive to be much better quality panels.

But inverter quality seems to have gone down compared to what I was buying back then, though they too are cheaper.

You are way out of line with this comment, please don't do this because it makes both you and HN look bad.

'Proper engineers' often have no interest in actually understanding the underlying processes, which would make for more than iterative improvements. Industry can fund iterative improvements. Source: Marc Baldo telling me to wait until my paper was published/on the arXiv and then he would 'look at it', while in the meantime trashing simulations and theory in his press releases.

Fair enough. The point of my comment was the incentives are set up in 'science' to research fundamentals (I don't think there are many interesting fundamentals in degradation, but I may be wrong). I am arguing there should be a separate stream of funding/researchers with incentives to work on other types of problems.

Scientists tend to be interested in 'neat cool solutions' or 'neat cool problems' and there should be separate funding for 'important problems with boring solutions' that sits separately to current science funding. I agree leave the small-scale iteration to industry.

A scientist is someone who follows the scientific method for one, engineers do that too. You can be both

There is no agreed upon definition of science or the scientific method. I'm sure we could agree upon a demarcation between scientist and engineer, though, and I was drawing that distinction in a way I think we can all intuit.

I’m not sure that demarcation is easy. As I understand it, the Manhattan Project (for instance) was science Monday, engineering Tuesday, and usually by the same people.

> Give more funding to proper engineers, not physicists/chemists

Materials science departments are often called "Materials Science and Engineering". How do they fall in your categorization?

Without disrespect intended, your words sound like those of someone who has never spent any time in graduate school in a engineering program in physical disciplines. The lines can be blurry.

I get that. Perhaps things are different in different countries, but where I am materials science/eng is drastically underfunded for this reason.

My comment was worded strongly, I meant more that funding could come from different streams fo4 different purposes.

Applied sciences are a thing, you know ...

They are! And they receive fewer funds in my neck of the woods.

Reply on news.ycombinator.com