Solar Chimneys: Viable Energy Solution or a Lot of Hot Air? | Hacker News Reader

Solar Chimneys: Viable Energy Solution or a Lot of Hot Air?

hackaday.com

35 pointslxm59 comments

> to generate meaningful power, they require huge tracts of land and incredibly tall towers

Preferably towers that won't burn down, fall over, then sink into the swamp.

The first two or three prototypes might do that but the fourth one... will surely stay up!

But would vanish in 24 hours after becoming online. Government would think about scrapping the whole project but suddenly some foreign nation, which wrought a terrible war in the past, would chime in with a proposoal to co-fund the installation, but with one clause: they would name the commander.

The year is 20XX, the place: Solar Chimney 5

I see what you did...

>FATHER: Listen, lad. I built this kingdom up from nothing. When I started here, all there was was swamp. Other kings said I was daft to build a castle on a swamp, but I built it all the same, just to show 'em. It sank into the swamp. So, I built a second one. That sank into the swamp. So, I built a third one. That burned down, fell over, then sank into the swamp, but the fourth one... stayed up! And that's what you're gonna get, lad: the strongest castle in these islands.

At the time the Manzares installation was built the concept seemed very promising. I read about it as a child in an overview of different approaches to solar power generation and to me, the chimney stood out as far more practical than the various mirror setups that were tried at the time (and that remained iconic "future energy" through 90ies and early aughts, e.g. in computer games). But photovoltaics have gotten so much cheaper, all that is left for mirrors, chimneys and the like is nostalgia. Like past futures with flying cars and landline phones.

The chimney idea did have a bit of a revival through the discovery of thermal mass as a way to flatten the day/night cycle, but that's far too short term and undispatchable to really make much of a difference. The gap behind photovoltaic is just too big I think and of we ever have hydrolysers on the scale we need for everything else, adding some more capacity will again win from scale.

These days, even the bad idea of integrating some battery capacity directly on module level would seem more promising to me. So many very good reasons not to do it, biggest one is very different expected service life, but if you do stuff at numbers like photovoltaic modules, scale effects are difficult to predict.

What I want, as a solar future tech fantasy, is installation/service robotics. Central premise: the modules need permanent scaffolding anyways. Focus on that for where innovation happens and make a pair of scaffolding system and machine that are designed for each other so that the scaffolding can serve the machine as as a controlled environment for travel (and power!), and the machine (or family of different machines) can do diagnostics, clean or even replace modules. Bonus points of course if the machine(s) can also lay out more scaffolding, but I'd consider that beyond mvp.

As an interim step, I'd just like to see an installation system that was designed to minimize the now largest cost of a solar setup: labor.

True, but for that alone I'd assume that there's no lack of trying. Innovation through dozens (hundreds?) of suppliers each trying to sell to installation companies.

In old-google-speak those are certainly not moonshot projects, but I'm sue that this organic innovation of many hands, eyes and minds has already achieved quite a bit. In the other hand that kind of innovation is prone to getting stuck in local optima and one of the reasons I believe that the "scaffolding as robot rails" approach might be worthwhile is that it would be somewhat moonshotty, and that moonshottyness could very well knock the state of the art out of local optima if it happens to be stuck.

There is a YC startup doing robots for production of solar plants IIRC. Can't remember it's name but they are often on here trying to hire mechanical engineers to build the robots. Not sure if they are just the initial manufacture, or also ongoing maintenance.

Capex might put weight the opex if you are just using robots to clean the panels after construction?

Charge Robotics, search box did its thing.

https://news.ycombinator.com/item?id=30780455

An interesting read, looks like they have the attitude necessary to innovate without distracting moonshot goals.

An alternative concept is the reverse of this: a very tall shell like a cooling tower, where water (possibly sea water) is injected at the top. Warm dry air is cooled, becomes denser, and flows out the bottom through turbines.

This does not require a large area set aside for heating; it exploits the natural heating of air over any arid area. In a place like Los Angeles it would also serve to scrub pollutants from the air. There is a net increase in atmospheric humidity, so it would increase rainfall downwind of the site, possibly beneficial in desert regions.

These concepts do not scale down well, extracting more energy from the air as the tower becomes higher.

>the solar updraft tower also offers some potential for energy storage

The pleasing part I'd hoped to read.

I'm not intimate with the physics (in general) but am aware that overheating of traditional solar panels can be an issue affecting their efficiency. There are some hybrid variations of PVs (PVTs) that look to remove heat from the panels to increase their efficiency and then use the heat for hot water requirements. Maybe there's an interesting hybrid solution in that where the panels are doing their regular generation at the time of use, and the heat can be used at night time.

My understanding is “hybrid” solar panels, solar trackers, etc. have all basically become uneconomical with the massive drop in solar panel prices per watt. For any additional energy you could extract with these schemes, you can get more energy per dollar just by adding more panels. I suppose in cases where you are space limited something like this might still make sense.

Sounds about right otherwise PVTs would be far more mainstream. I think the point in that is that the energy requirement for hot water is a fraction of the demand for electricity. But perhaps with a system like this, the heat can be more useful when used as a battery and the sun isn't shining.

As ever the LCOE wins, and whatever time of day the energy can be used at.

I think about this a lot and my conclusion is that is too narrow of view of what is happening.

I think it makes sense to use Solar PV as a main source of energy. I also think it makes sense to use other forms of solar energy for a sense of redundency. I have an existing PV array that is nicely covering energy usage. I worry that I'm overly reliant on complex electronics and want to add Solar Thermal as a redundancy.

The solar thermal is actually a whole separate system with the same dependency on sunlight but the human made bits can all be different.

You're worried that PV is complex so you want to add a thermal cycle? That doesn't begin to make sense. Thermal power plants are insanely complicated. PV is just semiconductors.

Not a solar thermal electric power generator, but solar thermal storage. It's just a hunk of matter, usually a water tank.

It'll be cheaper if you charge your thermal store using a PV driven heat pump, rather than try to collect the sunlight as heat directly.

Might be cheaper but I think you failed to understand the need for a redundant heating system. That is isn't wholly based on the complexity of PV.

Secondly there is nothing preventing me from leveraging the cost savings of PV heat generation when available. I just want a parallel heating system that isn't reliant on PV solar generation.

I thought I made that clear.

I also would be able to use the system during the shoulder months where the efficiency of heat generation isn't an extreme need.

Passive solar is simpler the Photo Voltaic solar and I believe there is a rich middle ground in balancing the 2.

Why not a PV heating system and, if you want redundancy, another PV heating system? What failure mode are you imagining that leaves a solar thermal system working but takes out both PV systems?

I see a stronger redundancy by using disparate overall approaches of technology. This is an overall philosophy. I will try to describe the overall weakness I see using the same approach for a redundant system.

A double PV system I see as too homogeneous and thus susceptible to the same risks. A PV system with a solar thermal system for redundant heating I see as more diverse and thus resilient.

I try to design to a 100 year expectation. We all know that PV systems need replacement on a schedule. We also all experience electronic digital systems that may meet are needs but are no longer supported.

I see this currently as a complexity that we accept because the tech is so darn convent.

Solar thermal can take many forms from very passive to a combination of almost passive with the exception of fluid moving systems. The fluid moving systems components are not nearly as specialized as the PV components and thus the overall system is more malleable. This malleability provides for greater variation in replacement parts and short term hacks to get through system failures.

Obviously we disagree. I truly don't understand how highly specified electrical system could be simpler then pumped water and solar gain.

Looks like someone can’t handle variations in opinions and can’t articulate their own ideas.

That seems entirely dependent on available space.

Tracking needs more space, so that works against them. If you are limited by space you would go for fixed tilt.

Or even just 100% coverage of the ground with flat PV panels.

https://www.erthos.com/

Utility scale PV does often use 1-axis trackers, particularly in the US southwest.

I wouldn't ever see limited space as a hypothetical, considering how human history is often just seen as a long sequence of territorial wars.

Territorial wars for productive land. The thing about solar is that it is often best located in land that is useless for other human purposes, and is often vast, like deserts.

And also for water, which is closely related. PV doesn't (or doesn't have to) use any water.

There are entire continents without much in land so unproductive that no-one bothers taking ownership.

When storage is the primary goal, something like the Crescent Dunes Solar Energy Project[0] might make more sense. Basically, bunch of mirrors shining onto a central tower where it heats up molten salt, which is then used to produce steam and run a generator. The key advantage here is that you can just store the hot molten salt in a tank and turn it into steam some time later.

The thing is, PV has gotten so cheap that it might just turn out to be more economical to simply collect the sun with PV and heat the salt with essentially a huge resistor so the energy can be stored for later use.

[0]: https://en.wikipedia.org/wiki/Crescent_Dunes_Solar_Energy_Pr...

This technology is obsolete and uncompetitive now.

True, even with a resistor. And with a chain of heat pumps you'd get the same heat or more, and free cool on the other end of the cascade. Good sites for mirror-based solar tend to be places where coldness is quite valuable.

> The engineering team behind the idea, Schlaich Bergermann and Partner, noted that solar updraft towers really only make sense at these massive scales. Smaller installations aren’t cost competitive with photovoltaic solar panels, but larger ones can be. Large facilities make enough power to offset the huge construction costs,

I think that until there is some revolution in construction logistics, skill, or management techniques, every large construction project is going to have a nearly unacceptable chance of being a boondoggle.

All of construction involves too much time of people hurrying up and waiting for crucial materials/equipment to arrive, waiting for the prior step to finish that can only be finished by particular people highly skilled in some techniques, or similar.

Every retrospective I have seen on construction productivity or on particular projects will cite these problems. Contrast this with manufacturing lines, which experience the same problems during ramp up, but can be solved and stay solved since tomorrow is another day just like today. With construction...

Then again, I'm no expert, just a highly interested amateur, do take my opinions with a grain of salt. But I would definitely make a bet on manufactured tech over constructed tech, in any area , when it comes to cost.

If your critique applies to constructing a large array of scaffolding and glass panels, why doesn't it apply to constructing a large array of scaffolding and photovoltaic panels and wiring and power electronics?

That is an excellent question that gets to the heart of my critique: what roadblocks are there during construction which means that labor gets deadlocked and waiting for the completion of something else, what learning can be iteratively applied, and is the project decomposable or monolithic?

A solar tower loses on all these fronts. First, the staging of the tower has many different components: leveling of the terrain to accommodate a large structure with airflow (perhaps not needed for the heat collector part, but perhaps?), foundations for the heat collectors, construction of the heat collectors to a sufficient pressure standard that the heated air doesn't leak before hitting the turbine, etc. Then the tower itself is a massive construction project with a large and critical foundation that must be built to exacting standards, then the tower itself, etc.

Utility scale solar has the panel structures, inverters, potentially a battery island, all of which can be built independently. Individual workers can work in a highly parallel manner, learning efficiency that can be translated into their next 1000 iterations of the same task. Further, due to the decomposability of the project, you can build a chunk, learn, and do the next chunk in an entirely new way if there's some innovation (maybe a new mounting process comes to market half way through the build). No specialized labor has to wait for the other part to finish their section, to get more of a specialized part that ran out (since all the parts are standardized and mass produced). There's minimal dependency between the different labor skill sets. And where there is a dependency, between the panel structures and the panel mounting, the needs of the structure are minimal. People are even testing dumping panels directly on to earth, without any structure, these days. If any part of panels are damaged, it has nearly zero impact on the rest of the structure. Whereas with a solar tower, any leak in a section affects a much larger structure.

And the decomposability of solar PV means that it can start operating before the entire thing is built. If money runs out, the fraction that gets built can still generate electricity, until the money can be found to complete the project. You can repower part of the project, add to it easily, scale it in any way you want. It's the difference between a cloud data center and a mainframe.

Thanks, that explains a lot. I hadn't really considered that the giant greenhouse had to be nearly airtight, rather than just directing air towards the chimney, and yes it can't be built piecewise and get a useful smaller version.

> There are no dirty panels to clean, for example.

Sure, there aren’t dirty solar panels to clean, because instead there’s the dirty glass (or plastic?) roof of the structure to clean. And access seems much harder, since a PV farm can have access roads between rows of panels, whereas a giant greenhouse is rather more constrained in layout.

One assumes the greenhouse collectors can also be used for food production - if so, this becomes more space efficient than PV by a long way, potentially.

Like agrivoltaics for out-of-season crops.

Probably not. Plants have a relatively small "sweet spot", and greenhouses try quite hard to prevent overheating. The glass panes can be raised to let hot air escape, and they'll even have blinds.

25C (77F) is about ideal, with 20-30C (68-86F) being somewhat acceptable before seeing a huge reduction in crop yield. Higher temperatures will also significantly increase water consumption.

Temperature in a solar updraft tower is about 20-30C above ambient. That pretty much immediately rules out dual use in all but the coldest environments.

Thanks for the reply! I was thinking that the greenhouses would be arranged in a circle, so there'd be a sweet spot in a ring at the very least. But it does sound rather.. difficult. I suppose that sweet spot would also not be in a fixed place, either. Pity.

How fast would the air over the crops move, when generating the claimed 200MW of power? A breeze? Hurricane speed?

Could a fan in a normal chimney generate usable amounts of energy just from the normal suction? Skip the heating the air at the bottom bit?

Define 'usable' and find out how much updraft you would get without heating. Hint: not that much.

That reminds me a bit of what's going in at Lake Nyos, Cameroon.

Basically, there are two "layers" of water in the lake, and due to the lakebed giving off carbon dioxide the lower layer will become increasingly saturated. When the two layers get mixed due to something like a minor earthquake, huge quantities of carbon dioxide get released - resulting in a massive disaster.

The solution here was to essentially just stick a pipe all the way down, so water could flow up and safely release the gas.The interesting part is that this is a self-sustaining process after an initial pumped startup.

Something similar might be possible in other lakes, perhaps driven by a temperature difference rather than gas saturation. I doubt it'd generate enough energy to make it economically viable, though.

Reminds me of OTEC: https://en.wikipedia.org/wiki/Ocean_thermal_energy_conversio...

Interesting, I assume there is some fundamental calculation that takes the radius of the greenhouse and chimney height in order to estimate the energy output.

Yup, and id bet that it doesnt make more than a similarly-sized field of solar panels.

There's a photonic limit [0]

I think the general idea here is that there's a lot of thermal energy that does not get used in a cost-effective way. IIRC the majority of 'efficiency' wrt energy hitting a panel comes out as heat, and here it's a question of if that heat can cost effectively be stored/used.

https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limi...

The Shockley-Quiesser limit only applies to single junction PV cells. PV will likely eventually go with stacked cells of different bandgaps.

My understanding of this proposition is looking at harvesting the thermal energy, which is beyond PV. The chimney is all about the thermal output.

So why did you post a link to something that's explicitly about PV?

Because it's part of the answer to the poster's question which is considered sound science.

Also that on a large scale the thermal energy of a solar array isn't generally accounted for.

> "there's a lot of thermal energy that does not get used in a cost-effective way"

Reminds me of a discussion in the comments on Terence Eden's blog[1], where a commenter argues that photovoltaic panels on home roofs are not the best idea, and solar-thermal panels are better; an edited summary:

"solar thermal, 4 square meters of roof space will generate 1500kWh per year; to achieve the same with PhotoVoltaics [PV] you would need 16 square meters. Solar thermal would roughly address water heating for showers for a family of 4 for using 4msq of panels. If that doesnt represent a useful amount of energy from a small amount of roof space I don't know what does! And with a lower environmental cost, lower CO2 emission, than PV for that energy."

i.e. it's sad that government subsidies are making PV cheaper than solar-thermal, even though it makes worse use of roof space, with worse CO2 profile.

[1] https://shkspr.mobi/blog/2013/12/free-money-from-the-sky/ - comments by 'ben'

But hot water isn't as useful as electricity.

In summer, when solar is plentiful, radiators aren't run, and showers tend to be cool. In winter, when there's higher demand, there's less solar.

Electricity is fungible. It can become heat, or light, or compute, or just about anything. Hot water isn't that versatile.

The right way to use it is not a greenhouse but a pool of seawater, maybe an existing small bay.

The hot air then dissolves a lot of water vapor, brings it upwards while cooling down and rotating turbines, and releases as fresh water.

Now you have a desalination plant that produces energy, not consumes it.

how many of these before they start impacting local weather

TL;DR Solar chimneys aren’t nearly as promising as modern solar panels as far as energy generation goes. They require tons of money and space in order to operate efficiently.

Reply on news.ycombinator.com