Oct 2020: We’ll see solar & wind farms with capacity much larger than the grid connection, some with batteries…Solar curtailment is a feature not a bug. Solar’s probably the easiest source of generation to curtail quickly, so can really contribute to grid stability. And sometimes curtailment is a price worth paying for using a grid connection more fully. Hydrogen is going to be a thing, but not for transport and probably not for seasonal power storage – exactly. Using renewables when they’re plentiful to make H2, then using the H2 to make steel or fertilizer, looks quite promising.
Rooftop solar mandates on new construction (like California’s) are a good idea. There will never be a better time in the roof’s life to add solar than when workers are already on the roof, now the kit is so cheap. Building-integrated PV products are usually attempts to sell bad solar products for premium prices to gullible aesthetes and architects.
(Of total decarbonization across all sectors) – getting that last 10-30% of carbon out will be hard, and require some expensive solutions, but now we have the tech and knowhow to get to 70-90% at reasonable cost. Solar may be 50% of electricity worldwide. Tell people at parties that Germany’s at 48% renewables! Minigrids, microgrids, and solar home systems with batteries are bringing reliable electricity to the whole world. It’s happening.
Time to make minor updates to my annual “opinions on #solar” thread.
If you like these, you’ll like my 2019 book, Solar Power Finance Without the Jargon. Five stars on Amazon, apparently “entertaining” and writer knows her stuff”. Also available here: tinyurl.com/y6lc3ohl2. Solar Power Finance Without the Jargon is the book I should have read before seeking a job in renewables, from the perspective of having worked in this for 14 years.
Link to 2019 thread. 2018 and 2017 are linked so you can see what I got wrong. tinyurl.com/y56zx5qc 3. To the opinions!
Solar manufacturing is still a terrible business to be in, though 2020 is better than many years. Competition is vicious and the newest factories have the best technology. Older manufacturers carry heavy debt for factories rapidly becoming obsolete.
Crystalline silicon technology is good enough and is still reducing cost significantly, for example diamond wire saws went from new to 100% of the wafer slicing market from 2016 to 2018.
Multicrystalline silicon has lost to mono, and for utility-scale, bifacial is taking over.
Solar is now the cheapest source of bulk electricity in most sunny countries. We don’t need a technology breakthrough, or subsidies, in solar to achieve massive adoption. Mostly, solar developers just need a grid connection and/or permission to sell electricity.
Wind is getting a lot better, too. And wind blows in the winter. So I do think that we need a somewhat diversified energy mix. But I think that five years ago I would have said, oh, yes, it has to include gas. And now I think it has to include wind. There will be more batteries, there will be more ultra high voltagegrids. We will probably have more sophisticated grid balancing, more sophisticated forecasting. We will just have a lot more solar. I tend to think that all the forecasts that we’re coming out with including Bloomberg NEFs forecasts are still far too low. We’ll probably end up throwing away a lot more solar. So when it’s really, really sunny, solar output really peaks. And I think we may find that in order to get a good amount of solar in our mix across the day, we overbuild solar. We put it on every surface and we end up with too much sometimes and just have to shut it off. Not a bad problem to have. We probably have hydrogen, green, hydrogen. We’ll probably use solar and wind at very low prices to electrolytes, hydrogen. We might not use that for power or for transport, but we’ll probably use that for making steel and also for making fertilizers where we currently use hydrogen from steam methane reforming.
Solar is cheap but it still costs money. Auction bids below $25/MWh do not reflect the full cost of generation, e.g. Portugal’s 11 euro/MWh 2020 auction price is a flutter on the long term value of the grid connection. Middle East headlines are opaque internal transfer prices.
Grid connections are now the gold dust for the solar (and wind) project development business. To optimize use, we’ll see solar & wind farms with capacity much larger than the grid connection, some with batteries. For this reason, solar curtailment is a feature not a bug. Solar’s probably the easiest source of generation to curtail quickly, so can really contribute to grid stability. And sometimes curtailment is a price worth paying for using a grid connection more fully.
Countries that have built little solar to date are not losing to high-solar countries; they waited for a bargain on price and can now hold an auction to find out how little a solar developer is willing to build it for. These first auctions are getting bigger.
In 2017, my analysis team covered 42 countries which were significant solar markets doing more than 50MW/year. Now we cover 125 and keep finding ones we’ve missed. It’s a bit of a pain, honestly, but quite exciting.
Good solar companies don’t need to pitch the enviro-social of solar in general. They have a few meaningful differentiators (above-average recycling efforts, renewable energy purchase plans) but mostly they differentiate on executing solar well, i.e., profitably.
I refuse to get excited about perovskites until a perovskite company can disclose a partnership with a named major module manufacturer. (This was my opinion in 2018. It is my opinion still). the module manufacturers are not sitting on their backsides doing nothing. They would love to have something that gave them a significant edge over their competitors and they are not interested. And when you see the companies that employ most of the solar scientists not being interested, you think maybe this is not such a great thing and this still hasn’t really happened by the way. Some companies have come closer, but there’s no perovskite company with a major module manufacturer partnership. My colleague, Dr. Wang has outlined some of the kind of reasons why perhaps kinds of difficulties. So the efficiency has come up really quickly in the lab. But the lifetime of perovskites even the lab is still quite short. Even a couple of years. Getting them stable out to 25 years is a challenge. And also the amounts of materials that are being made at the current efficiencies and lifetimes are really small. So a lot of these records are being set on a cell that is one centimeter by one centimeter. Scaling that up to an actual module is a massive challenge. And our house position as well as my personal opinion is that it will be at least 5 years before anyone is ready to start a factory that does this…you can you can have technologies that can radically reduce the cost at, you know, some fundamental component level. But that doesn’t translate into bankable projects that actually have a decent lifetime and are going to get deployed at scale.
Floating solar is a thing, but it’s not a new tech. It’s solar onna boat.
Agrivoltaics, likewise, is solar on a field. PV only has synergies with *some* agriculture. Competition for light and restricted mechanical access to crops are often problems. Existing Chinese agrivoltaics are largely PV subsidizing bad farming.
Many current household PV systems are designed suboptimally and may not make economic sense or even perform well. Also, most countries will move to paying a pittance for solar exports, so self-consumption rate is becoming the most important financial parameter for rooftop PV.
To financially assess a proposed rooftop solar system, you will need at least a year’s data on hourly electricity consumption to estimate self-consumption. Also, get it built when you have scaffolding up for something else, scaffolding is expensive.
Rooftop solar mandates on new construction (like California’s) are a good idea. There will never be a better time in the roof’s life to add solar than when workers are already on the roof, now the kit is so cheap.
There’s more innovation now in system design than in module tech (though some there, eg multi busbars, bigger wafers). East-west orientations use less land and give flatter output profiles, costing ~1 percentage point of capacity factor, than equator-facing systems.
Very few people who are not solar project financiers understand tax treatment for solar projects (I don’t) and it’s important enough to make most calculated LCOEs irrelevant to auction prices.
On concentrated solar: Solar thermal tower and heliostat designs are still not working well, although parabolic trough solar thermal has brought down costs more than expected. We may even end up using molten salt for multi-day and seasonal storage… but heat it with PV.
Utility-scale batteries are going to be a massive thing everywhere, both co-located with solar and wind to use a grid connection, and just embedded in the grid. However they’re not going to solve seasonal demand-supply mismatches as the utilization would be too low.
Hydrogen is going to be a thing, but not for transport and probably not for seasonal power storage – exactly. Using renewables when they’re plentiful to make H2, then using the H2 to make steel or fertilizer, looks quite promising.
Nuclear is safer than coal and climate change, and better than gas unless the gas plants are running very rarely. Batteries might help with the unfavourable ramping economics of nuclear (you *can* turn nuclear plants up and down, but you really don’t want to).
We’re finally getting serious about net zero carbon, as of 2020. Whew. Getting that last 10-30% of carbon out will be hard, and require some expensive solutions, but now we have the tech and knowhow to get to 70-90% at reasonable cost.
“We” who are finally getting serious about net zero carbon mainly means governments and large companies. I’m pretty sure the general public has little idea how much progress has been made. Tell people at parties that Germany’s at 48% renewables!
It would still really help if rich people would stop pissing away carbon for no reason.
Solar is a major part of decarbonizing electricity, but probably not more than 50% worldwide, so we need something else too.
Transport will go electric but also we really need to sort out agriculture, shipping and aviation somehow for a sustainably habitable planet.
While moving to a circular economy with 100% recycling rates is essential in the long run, it’s not a challenge for PV in particular; few PV panels have been recycled to date only because the vast majority are still in use. It can be done.
Solar plant operation and maintenance in desert environments will prove more challenging than PV project stakeholders currently expect. Also, climate risk from hurricanes, hailstorms, fire and floods is on the rise for solar as for everything else.
Traded electricity wholesale markets are the worst way of deciding how to dispatch energy resources, except for all the others that have been tried.
Many solar project developers complaining their problem is ‘finance’ are being disingenuous. Their problem is, their project is rubbish and they cannot convince anyone otherwise. (This may not just be a solar thing).
Building-integrated PV products are usually attempts to sell bad solar products for premium prices to gullible aesthetes and architects. (This has been my opinion since 2017. My colleagues Yali and Xiaoting plan some BNEF research in December 2020 to see if they agree. Background from https://re-cognition-project.eu/2020/10/13/building-integrated-photovoltaics-bipv/ : Building-integrated photovoltaics (BIPV) are photovoltaic materials that are used to replace conventional building materials in parts of the building envelope such as the roof, skylights or facades.
The lightweight PV module design will be based on the use of composite materials and polymer films to replace the standard glass-glass configuration for the back sheet and the front sheet, respectively. A process for colouring polymeric adhesive film was developed to achieve the target of aesthetical appealing, lightweight BIPV modules. Shape and size of the coloured and lightweight BIPV modules will depend on the specification of the site where BIPV modules will be installed. Integration of these modules in building skin is made simpler and cheaper, through a newly designed mounting frame. The photovoltaic modules are completely integrated into the building, performing, in addition to those strictly related to energy production, also some or all the functions related to elements and subsystems of closure or shielding. The photovoltaic components of the RE – COGNITION project must, preferably, belong to the BIPV category. In addition to ensuring efficient electrical productivity, the following preliminary requirements must be met:
- control and reduction of the operating temperature of the cells;
- water tightness and infiltration resistance;
- air and draught tightness;
- integration or compatibility with thermal insulation elements;
- Ease of installation and maintenance.
Specifically, the research and development activities of the BIPV component can be carried out according to an evolutionary logic with respect to components already developed and tested. Using special metal or PVC profiles is possible to create a stratification laminate-air gap-insulating layer, which replaces the entire roof or façade covering. This will assure the requirements above listed to be addressed.
Job creation from solar will be less than solar advocates forecast. Relatedly, solar cost will be lower than solar advocates currently forecast. Being a lot of work is not intrinsically a good thing for a source of power. The whole “only an overall growing economy can support a growing renewables industry” idea needs to die. Renewables industry growth *is* economic growth, and jobs, and nobody said economic growth needed to be homogenous. And renewables have been resilient to Covid-19 so far.
Off-grid solar is driven more by people wanting TVs than by people wanting their kids to have light to do homework. And that’s absolutely fine.
Minigrids, microgrids, and solar home systems with batteries are bringing reliable electricity to the whole world. It’s happening.
Commercial and residential power pricing structures are going to get more complicated to reflect cheap midday solar. The main role of smart home technology is to take advantage of this. It’s not currently doing it very well.
There is enough land for lots of solar. There are enough golf courses in the U.S. for about 370GW, ffs.
If you’re recording PV capacity and only have room for one figure, record MW(DC). It will tell you more about what the project will produce and what it will cost than MW(AC), which is just the size of the wire. I will die on this hill.
Anyone buying a new internal combustion car now is pretty silly. EVs aren’t the answer to everything – especially congestion of cities – but they do use much less energy and, with flexibility, can support the grid.