Climate change tech (not just renewables)

[PhaseShifter]

Visions of trying to order a cubic kilometre of concrete from my local builders yard, 2.4tonnes / cubic metre, so, that might need a few trucks to deliver it....

Now imagine the amount of water you could hold back building dams out of that concrete.

Using concrete merely for the gravitational potential by lifting its weight will never come close to using concrete pipes and dams to control water flow.

 

[Auguste_Fivaz]

The New York Times runs a rather long essay on the early days of collecting and working up how to exploit methane "eating" microbes from mud volcanoes in the ocean. (Free link) Tricky to work with, some die when exposed to oxygen, others stop working when their environment changes but some are able to tolerate the trip to the lab. The scientists working on this think there is a good potential to use these critters in the next few decades to help with mitigation efforts, removing methane at the source, dairies, dumps, extraction sites, gas storage and pipelines, for example.

That’s why Dr. Henriksen, Dr. Tierney and a genomic scientist, Krista Ryon, have been traveling to some of Earth’s most extreme environments: They want to know whether microbes in these places might be different enough — or just plain weird enough — to help reduce the harms that fossil-fuel use and industrial-age farming have done to the planet.
The three scientists and their collaborators have collected samples from hot springs around Colorado. They have dived to volcanic seeps off Sicily, Japan and Papua New Guinea. To organize their globe-trotting, they founded a nonprofit organization, the Two Frontiers Project, that is funded by Seed Health, a maker of probiotics, and other donors.

So far, Two Frontiers has mostly searched for bacteria that eat carbon dioxide, the main greenhouse gas that is warming Earth. One strain the team found off Sicily has proved so efficient that three gallons of it could theoretically absorb and lock away as much carbon dioxide as a tree, though researchers are still working out how to cultivate and deploy it at scale. Now, Two Frontiers is widening its hunt to devourers of methane, which has far greater heat-trapping power than carbon dioxide, though it stays in the atmosphere for less time.

 

[w00key]

Grid folks are justifiably quite conservative

Indeed. The Spanish blackout started with a renewable plant messing up the local section. It shouldn't have spread, but buggy inverters are a thing and it's much harder to get right than "spinny thing goes round and round" and any reaction is plain physics.

 

[Tom the Melaniephile]

Indeed. The Spanish blackout started with a renewable plant messing up the local section. It shouldn't have spread, but buggy inverters are a thing and it's much harder to get right than "spinny thing goes round and round" and any reaction is plain physics.

Setting aside that the renewables in Spain were (and I believe still are) not allowed to perform these services.

 

[bjn]

Indeed. The Spanish blackout started with a renewable plant messing up the local section. It shouldn't have spread, but buggy inverters are a thing and it's much harder to get right than "spinny thing goes round and round" and any reaction is plain physics.

The immediate cause was a solar plant screwing up, but grids are meant to cope with screwups, so it was a system failure. As I understand it there were meant to be thermal generators on the grid to provide stability and they failed to do that and there aren’t many grid forming inverters in Spain anyway. The full report is yet to come out, so I’m speculating,

 

[tucu]

The immediate cause was a solar plant screwing up, but grids are meant to cope with screwups, so it was a system failure. As I understand it there were meant to be thermal generators on the grid to provide stability and they failed to do that and there aren’t many grid forming inverters in Spain anyway. The full report is yet to come out, so I’m speculating,

The sequence of events got a lot more detail in the Factual Report published recently

(that is the first page of the sequence - there are 3 more pages)

 

[w00key]

The sequence of events got a lot more detail in the Factual Report published recently

Not really. The list you posted begins with a bunch of wind and home PV disconnecting. They don't randomly do that - something is causing severe overvoltage on the grid. Same for the general overvoltage condition on the grid, something / someone fucked up and moved it right to the edge of collapse.

The root cause rumored, not mentioned as it is the content for the followup, is

The Spanish grid operator, Red Eléctrica de España (REE), stated that a "malfunction" at a PV plant in the province of Badajoz triggered a forced frequency oscillation that led to the blackout.

Media reports identified the specific plant blamed by REE as the 500 MW Núñez de Balboa photovoltaic facility, owned by Iberdrola, located in Badajoz.

The oscillation, which had a dominant frequency of 0.6 Hz, was initially described as a local, forced oscillation potentially originating at a photovoltaic power plant on the Iberian Peninsula.

This oscillation was followed by corrective actions from system operators, which, although dampening the oscillation, had the side effect of increasing voltage levels.

 

[demultiplexer]

But the point is that that isn't explanatory. A disconnecting or malfunctioning plant, especially something relatively small (<100MVA), should not cause either these local oscillations or some cascade that triggers the collapse of the majority of the peninsula's transmission grid.

For reference: about a decade ago, we had a hard short fault on Flevoland causing a cycle-by-cycle peak-to-through power oscillation in excess of 500MVA. The amount of current that causes not just one circuit, but all circuits on a 150kV overhead transmission line to start smoking and sagging 3-5 meters (almost to the ground in some places). Literally zero power disruption outside of the faulted area was affected, no oscillations, no nothing. And this wasn't some perfectly resistive, well-behaved short either - it arced and cause very high transients in the transmission lines.

This is why I and other say that it's impossible to say there is a root cause. There are contributing factors, all of which together caused a unique failure mode that was not foreseen or designed for. The malfunctioning plant may be the inciting incident, but it must have had a bunch of other design flaws, miscalculations and other insufficiencies in the transmission grid to propagate this way.

 

[tucu]

Not really. The list you posted begins with a bunch of wind and home PV disconnecting. They don't randomly do that - something is causing severe overvoltage on the grid. Same for the general overvoltage condition on the grid, something / someone fucked up and moved it right to the edge of collapse.

The root cause rumored, not mentioned as it is the content for the followup, is


View attachment 119808

There are >10 pages on the 12:03-12:08 oscillations on the ENTSO-E report. These are the last few paragraphs of that section

Nonetheless, based on PMU data, it is possible to understand that the dominant content of energy is associated
with the reactive power and voltage. In addition, large power fluctuations were detected in the SCADA data of generation in the area of the nodes of Almaraz and Puebla de Guzman. Figure 2-58 shows fluctuations of active power with an amplitude of around 200 MW and reactive power with an amplitude of around 180 Mvar, occurring between 12:03 and 12:08.

All of these elements will be analysed during the investigation to clarify whether any local malfunction could generate a forced oscillatory phenomenon. However, at present, no generator has reported significant malfunctions of reactive power regulation, and thus additional investigations will be conducted in the next analysis phase to confirm the hypothesis of forced oscillation.

In light of all the factors listed in the previous paragraph and the currently available data, the Expert Panel cannot conclude at this stage whether the oscillations were forced or not

edit: the "SYSTEM AND MARKET CONDITIONS BEFORE THE INCIDENT" main section is almost 75 pages

 

[w00key]

Eh, of course it is "just" a contributing factor. If they had scheduled more hydro, gas, whatever spinning generator to the grid, it wouldn't be a problem.

They chaged the HDVC line settings and killed a big bunch of reactive power capacity. Clearly the TSO thought this was fine, and it wasn't.

The report has a million findings and no analysis. My point is that pointing to it is rather useless, it's a 264 pages worth of data and not a lot of thought, that is explicitly reserved for "next steps".

 

[demultiplexer]

But you are drawing conclusions! Even though none have been presented yet. I just don't see what you're saying as credible if there's no analysis presented yet. Especially because the whole firmness argument is used so much in anti-renewables circles and has been so goalpost-shifted through the years that it's not even funny anymore. Like, we already live with less than 10% of the inertia in MJ/GW and MW/GW in the TenneT area that we had in 2000 and all we got was a way, way more robust grid. We're well below any value of firmness that was guaranteed to cause chaos by detractors back then.

Why? Because transmission line design engineers know what they're doing.

 

[w00key]

The factual report is used here as a heavy weapon to club away earlier analysis. Look, here are 200+ pages, anything previously written is clearly wrong.


When it's not the same thing. One was an analysis of the data, possibly wrong, the other is a book full of facts. It's possible B contradicts A but we will have to wait a few months for the final report, and I bet all market parties are also digging through the data to bend it in their advantage.


The renewable lobby in Spain already issued statements that the factual report finds that they are not the cause. Market operators (demand side and traditional power plants) replied with no comment until we have time to go through it.


I expect a bitch fight in the coming weeks as they all issue their own statements.

 

[tucu]

Well, the additional info throws some smoking guns into a swimming pool.

Marking the Núñez de Balboa PV plant as the starting point dilutes given everything that was going on at the same time.

On the inertia argument, the utilities blaming the TSO for only scheduling 10 thermal plants (3 nuclear and 7 CCGT) can be questioned given the additional contribution on the load side provided by the 4.6 GW of pumped hydro that was active (plus the 3GW of hydro generation)

 

[MilleniX]

Well, the additional info throws some smoking guns into a swimming pool.

Marking the Núñez de Balboa PV plant as the starting point dilutes given everything that was going on at the same time.

On the inertia argument, the utilities blaming the TSO for only scheduling 10 thermal plants (3 nuclear and 7 CCGT) can be questioned given the additional contribution on the load side provided by the 4.6 GW of pumped hydro that was active (plus the 3GW of hydro generation)

If the pumped hydro was operating as a consumer and not a generator, do those pumps provide the same kind of reactive power and hence stabilization that the generator turbines would if they were running? I don't see why we can necessarily take that for granted.

 

[tucu]

If the pumped hydro was operating as a consumer and not a generator, do those pumps provide the same kind of reactive power and hence stabilization that the generator turbines would if they were running? I don't see why we can necessarily take that for granted.

While the report doesn't provide a detailed calculation it does mention that the contribution from loads is included in their inertia calculations; it also mentions that the number of plants contributing to inertia was ~45 (the 10 thermal, plus hydro plants >30MW). The inertia values that they published for 12:30 on they day of the blackout are:

A previous report in May from a third party had arrived at a value of 1.3s in the Spanish side. At that point the Spanish TSO responded that it seemed that the report was netting the contribution from ~3GW of pumped hydro and ~3GW of hydro that were active at the time.

edit: worth clarifying that the 4.6 GW of pumped hydro I mentioned earlier are totals from Spain and Portugal at the time of the blackout.

 

[Tom the Melaniephile]

Eh, of course it is "just" a contributing factor. If they had scheduled more hydro, gas, whatever spinning generator to the grid, it wouldn't be a problem.

Or if they had scheduled renewables with grid-forming inverters to perform the same function.

https://docs.nrel.gov/docs/fy24osti/89269.pdf

 

[w00key]

Or if they had scheduled renewables with grid-forming inverters to perform the same function.

https://docs.nrel.gov/docs/fy24osti/89269.pdf

Yeah, now they are looking into this, and classic fixes like synchronous condensors. So it won't happen again, probably.


Oh for that Badajoz plant that was rumored to be patient zero, forcing reconfiguration of the HV grid, the factual report has a chart on page 60. For now, unclear / inconclusive if this was a cause or effect, forced power due to wonky inverters or something else on the grid making it flip out.

With this chart now finally public, I get why some were quick to point at it and say, here's the troublemaker. It sure looks sus, that's a big power spike worth 50% of it's nameplate capacity, going from 30-80% power level (1 pu, "per unit", =100%).

 

[Shavano]

There's an article here from MIT about new startups entering a for profit market in solar geoengineering.

https://www.technologyreview.com/20...ineering-is-bad-for-science-and-public-trust/

Not surprisingly, they're critical of the idea of private for profit companies taking action to cool the atmosphere by releasing chemicals into the atmosphere.

 

[dzid]

New research from the University of Helsinki¹ describes a promising method of capturing CO₂. It uses Direct Air Capture (DAC), improving on previous efforts that were inefficient and impractical due to the high humidity of ambient air and low concentrations of CO₂.

An absorbent derived from a deep eutectic solvent (DES) achieved an absorbent capacity of 0.154 g_CO2/g_DES from ambient air with humidity > 90%. The absorbent does not react with other atmospheric components, and the process is reversible, with the CO₂ able to be released and recycled.

From the abstract:

High stability, low toxicity, very high capture capacity, recyclability, and moisture insensitivity characterize this outstanding sorbent, making it suitable for DAC. This absorbent desorbs CO₂ at just 70 °C within 30 min, effectively addressing one of the most challenging steps in DAC. According to the literature survey, this unique DES outperforms all previously reported liquid absorbents for DAC in terms of recyclability, efficiency, and practicality.

¹ Zahra Eshaghi Gorji et al, Direct Air Capture: Recyclability and Exceptional CO2 Uptake Using a Superbase, Environmental Science & Technology (2025).

 

[Rb87]

So it’s good at pulling CO2 under humid conditions, yes, but it loses 25% of its effectiveness after 50 cycles, and 50% after 100? Seems like they need to fix that, or else they’ll have to make more gigatons of the DES than the carbon they hope to gather.

 

[dzid]

So it’s good at pulling CO2 under humid conditions, yes, but it loses 25% of its effectiveness after 50 cycles, and 50% after 100? Seems like they need to fix that, or else they’ll have to make more gigatons of the DES than the carbon they hope to gather.

I agree, they'll need to improve on that if they can. I don't think it's an insignificant development, though.

 

[Megalodon]

So it’s good at pulling CO2 under humid conditions, yes, but it loses 25% of its effectiveness after 50 cycles, and 50% after 100? Seems like they need to fix that, or else they’ll have to make more gigatons of the DES than the carbon they hope to gather.

Don't think this is necessarily a deal breaker, depending on the specifics. There doesn't appear to be anything in the solvent other than carbon, nitrogen, oxygen, and hydrogen. Depending on how it's synthesized and how you might be able to recondition the solvent after it's been used a few times, and how much carbon it can carry per cycle, I don't think it's a dealbreaker? Maybe you only need to re-distill it to remove contaminants. And this work may lead to further advances.

 

[w00key]

How are we here with direct climate engineering?

Recently I had a silly idea I wanted to test - what if we generated more rain and clouds, artificially, without chemical seeding? It only works in certain areas though. This is an atmospheric condition called inversion, when hot air cannot rise due to external influence so you get extremely humid air trapped near a coastal region, yet it never rains.

If we can poke a hole in it, with the right condition, you can release all the "Convective available potential energy" or CAPE or storm power and create a crazy cloud out of thin air.


This condition exists over 5 area - Red Sea, Persian/Arabian coast, Baja California, Skeleton Coast (Namibia), Atacama (Chili).

If you manage to send a vortex contained thermal plume up to 2-4km, you get rain downwind. Send it up higher and you get ice crystals > 10 km and trade winds will carry it away at 150+ kph leading to slight cooling in a huge area, reflective clouds do that.


The question is, how. One brainstorm session led to another and now it seems like vortices are the key. They hold their shape, even dust devils are hundreds of meters high reaching the bottom of an inversion, and you just need ~10 m/s vertical velocity to get through, to cooler air, condense and release massive latent heat, multiple times the sensible heat as you're dumping 20g/kg of water vapor into a very cold place.

1D sim based on equations says yes. CFD with incompressible solver (Boussinesq) says you can do it with tens of kW of induced spin, create a decent looking vortex and pretty big vertical velocity.

Final boss is compressible solver, initialize with desert heat rising vs Hadley Cell sinking, let them duke it out and see if you get a stable inversion in CFD. Then add spin.


Climate change on a GoFundMe budget (sub $1M). Idk, sounds too good to be true but the idea behind it is using a small nudge to turn a massive lever. Ground heating, the capture zone gets MW scale heating for free, you just provide little bit of airflow so random thermal plumes disappear and get focused into one big one, and twist it into a needle. That twist can be done in tons of ways.

Curved walls or poles with fabric in between is the classic setup to turn sea breeze into rotation. Fans arranged in circles can directly inject tangential rotation. The 2026 way would be placing a bunch of IoT fans on a stick (pole) on a hex grid and create a square kilometer phased fan array, software defined tornado, to control width / height to switch between rain and cooling (max height) mode, and run experiments; recent cfd results says multiple vortices merging > 1 big ass vortex, now you can verify it by programming the fans into 1 vs 3 vs 9 small cells.

 

[wco81]

What happens with unintended consequences like maybe flooding down wind which causes billions in liability?

 

[MilleniX]

What happens with unintended consequences like maybe flooding down wind which causes billions in liability?

Move fast and break things, of course. Grow faster than the courts can get around to finding you liable. Etc

 

[demultiplexer]

How are we here with direct climate engineering?

Recently I had a silly idea I wanted to test - what if we generated more rain and clouds, artificially, without chemical seeding? It only works in certain areas though. This is an atmospheric condition called inversion, when hot air cannot rise due to external influence so you get extremely humid air trapped near a coastal region, yet it never rains.

If we can poke a hole in it, with the right condition, you can release all the "Convective available potential energy" or CAPE or storm power and create a crazy cloud out of thin air.

Unfortunately CAPE content is so extremely low (a few kJ/kg over a volume of maybe 1-10 km3) that you're simply not adding much by inducing some rotor on the local atmosphere.

Generally speaking, geoengineering through cloud seeding is proposing to continuously cloud seed an area in the order of the continent of Australia to be effective on any kind of useful timescale. I feel like in discussions about the subject this is commonly forgotten. There are no kilowatts in geoengineering, it's all a massive undertaking!

 

[w00key]

Unfortunately CAPE content is so extremely low (a few kJ/kg over a volume of maybe 1-10 km3) that you're simply not adding much by inducing some rotor on the local atmosphere.

ECMWF shows CAPE of 1000-2500 during the whole week until Friday over the red sea. Last week also had some good days. And it's winter, not storm season. I think if you pull data from ERA5 for a year you can find a few spots where it would work.

Example random spot with good CAPE coverage in the forecast and "easy" to unlock:

It's ~26°C here, but Tcon, temperature where free convection would rise and rise is 29°C, so inversion capped this energy. Once in a while a chaotic thermal plume will breach it by being >= 29°C but if you have a few m/s vertical velocity and or extra heat you can break through to lifting condensation level (LCL) of 756m and unlock all that latent heat energy.


OpenFOAM says if you heat ground evenly like in a desert, those little thermals won't do much. Add a dark patch with crushed basalt, and you will organize the thermals (it get hotter faster, lower albedo means consistently more heat, air rise, pulls in hot air from surrounding, deleting random thermals in surrounding area) but it loses energy while rising to entrainment so instead of a punch, it's a puff, might work on the best days, won't work on most. Make a vortex, and entraiment drops dramatically with incrased radius or wind speed. Meaning core stays hotter longer, dT vs surrounding air is higher during rise => bigger final velocity, and hits like a punch.

Large dust devils have a size of 10m wide x 1 km tall. That would be enough to reach LCL and poke a hole. The challenge is to engineer a persistent vortex engine - one that adds a few m/s speed due to initial dT providing buoyancy until it hits the inversion. If Tcon - actual temperature is only a few degrees, all the CAPE, moisture is free energy.


Note that if you try this idea not at the inversion spots, it won't work. The Hadley Cell is rising hot air at equator crashing down at ~30°N. The hot and dry air fucks up the natural 6.5°C/km temperature drop so hot air at ground level, as it loses heat when rising, gets colder than air already up there, an infinite, never ending stream from the equator - Tcon > ground temperature = it's hot and humid, but never a cloud in the sky. This creates a pretty unique situation, and 5 of these exist (afaik).

So idk what they did in Australia but it's not this.

 

[w00key]

What happens with unintended consequences like maybe flooding down wind which causes billions in liability?

The winds are 150 kph towards Saudi inland and then Indian ocean. If you "overshoot" and punch too hard, moisture rising to 10km, it stays there for days as tiny ice crystals. So you will be cooling a patch from middle east to Asia, not a concentrated "fuck this area in particular" situation.

The jet stream is why flying east takes like 3 hours less.


As for flooding risks - natural inversion breakdown already cause severe floods. Example, https://en.wikipedia.org/wiki/2022–2023_Saudi_Arabia_floods. Better tap the gas daily than let it build up and release it all at once.

This of course will have to be managed carefully. You want a few mm sprinkle, not flood the people downwind. The key variable to control is vertical speed, that determines what it will be, clouds that only disappear few thousand km away and so smudged by then, it doesn't matter, or local rainfall few km away. Last one needs careful management.


The idea of a fan powered vortex engine is that you can turn it off, and it won't wander unlike natural ones. A passive vortex engine turning the sea breeze (west to east, every day) into a vortex using curved walls has no controllability. This one you can turn off, if on a pan tilt base, even point clockwise and manually destroy it.

 

[demultiplexer]

So idk what they did in Australia but it's not this.

I think you're misreading me - in order to have anything worthwhile on a regional or global scale, your combined seeding event needs to be the size of something like Australia. The total amount of excess insolation to be countered is immense, it's like 0.6% of total global insolation you need to compensate continuously to compensate the current rate of global warming, and that is just the warming effects. Add to that that a cloud seeding event only has its maximum effect at mid-day and isn't useful (and actively harmful locally) at dawn/dusk or night, plus it's only as good if you're making clouds where none would be otherwise, and you need incredibly large areas of active cloud seeding to make a dent in the climate.

It's doable, don't get me wrong, but trying to limit it to very small geographic areas with very specific atmospheric conditions during limited times is not really worth considering. What makes cloud seeding a 'bad' type of geoengineering is that the areas above which it would be super effective are also the areas that are very hard to seed clouds in at will.

 

[w00key]

The total amount of excess insolation to be countered is immense, it's like 0.6% of total global insolation you need to compensate continuously to compensate the current rate of global warming, and that is just the warming effects.

Ah I misread you, yes.


But well of course. No one is thinking that just a few fans on a stick will fix everything right? That's totally unrealistic.

But "the effect is too low so why bother", this line of thinking is dangerous, none of what we do will be a big enough on its own, and if it does, you don't want to, get it wrong and you trigger the next ice age. Like nuking the sahara or whatever do trigger massive amount of dust in upper atmosphere.


So what's wrong with a few tens or hundreds km2 of cloud cover providing local cooling? 30-50% reduced albedo is a few to 10 degrees local cooling. At 40% reflected, you cool 400MW per km2. Not going to fix global warming on its own, certainly, but if you generate a km wide plume for a few hours and let the jetstream smear it out over hundreds of km, it's GW scale.

Consistent few mm of rain with an weak inversion is possible, says the forecasts. Moderate CAPE is sticking around for next 14 days, you actually don't want that too high, the Jeddah epic floods killing people were ~1000J/kg.


"Test plan" right now is just CFDing a bit more. Currently using general purpose solvers, OpenFOAM, it understands temperature, pressure, velocity etc. Need to move to the MicroHH climate model to add moisture and latent heat.

It's just a side project, no one is actually going to do this, I think, but it would be funny to see if a few pan-tilt kW scale fans can generate a vortex on demand with any aspect ratio you want in OpenFOAM and MicroHH, and make it rain. That's an experiment you can run with 100k worth of stuff, Saudi spend way frikken more on desalination.