Catching the Sun
This has been rattling around in my brain pan for a while. And by for a while, that’s probably 15 years with this post as a draft for four years. I’m clearing the mental decks for what is coming next.
The sun is one of the more important elements of our natural environment [Citation Needed]. When it comes to powering our human lives, the sun has all the power we could need. Solar is ultimately the source of all power on Earth, since it is all star-stuff or driven by celestial mechanics (the geo in geothermal only accreted due to the sun’s gravity) but in keeping things more immediate we can split power into segments like Solar, fossil fuels, biofuels, geothermal, hydroelectric, and nuclear. Through combinations of already established technologies we can power our current (pun intended) 1st world lifestyle for many more billions of humans than currently reside on our friendly little rock using any of these mechanisms at sufficient scale (more or less). Today, direct solar power generation is as fast and as cheap of many fossil fuel or biofuel solutions, wind is keeping pace at the front of the cost curve for power generation, geothermal sits behind the pack — both slower to build out and less portable than the fossil fuels and nuclear being much slower to build out and not portable at all as a generating source. Hydroelectric power is cheap and reliable, so long as you only look at the economics of the power generation and not the water source ecosystem, but not particularly portable — as a pumped storage play hydroelectric may have a lasting place as a grid-scale smoothing function rather than primary generation. If scoring by long term consequences, the advantages go to Solar for static (counting wind, as the function of temperature gradients on a global scale) generation and biofuels for portable high intensity power (due to carbon cycle considerations — fossil fuels still have the strongest subsidies for every use case). While battery technology continues to come down in price (because energy density is perfectly reasonable already for all but the most demanding power to weight requirements, e.g. flight and the self-powered lift into orbit). The other major classes are just too slow to deploy either by regulation, or sheer scale required, to fill the need, or non-generalizable across regions and continents. This headline tells the tale: Renewables account for most new US electricity generating capacity in 2021. All the tools we need already exist, we just need to pick from the options and deploy everywhere.
Electrification makes all the math easier, by operating at a base efficiency of twice that of combustion-based power generation — essentially making each unit of electrical power work twice as hard for our needs. This often covers the power-density gap between today’s battery technology and octanes and other chemical fuels. We are at the point where one can watch a Royal Marine perform a ship to ship boarding procedure with a jetpack and wonder if that was JP8 or battery powered (it carries 5.25 gallons of jet fuel, while electrification is more efficient consumption liquid fuels are 20x as energy dense so batteries are still behind JP8 by an order of magnitude in terms of weight). The operative question for most deployments becomes more one of finding the right methods for the specific situation; can we beat photovoltaics between the tropics combined with huge grid or chemical-based transfer mechanisms in order to capture that power and bring it to where it is needed anywhere in the world?
What options do we really have to catch the sun?
Solar motivates several kinds of power generation. Wind is solar conversion through an intermediary of differential heating around the world. Wind can be generated on demand as well, though leveraging existing flows is typically cheaper, more straightforward, and conventional. Wind is more powerful a few hundred feet above the ground, which leads to today’s large turbines (the 248m Haliade-X for instance). Biofuels are another kind of solar derived power created through photosynthesis, with some crops like switchgrass or algae being more efficient than others. As a low-volume distributed power supplement there is biofuel available just about everywhere from cast off as a waste-products from farming and construction, but as a primary a power source, biofuels have some significant drawbacks both in terms of surface volume required to power the cycle of grow-harvest-burn and in repeatability as farming yields have good years and bad years. Direct conversion of solar power through photovoltaics or concentrating solar power are, today, well established and cost competitive if not
preferrable in many areas. These direct conversion plants do suffer one pretty significant drawback — they work best when the sun is out. Wouldn’t it be something if power could be maintained throughout the night?
Well, that’s what solar thermal is for — heat up a sufficient thermal mass and you maintain a useful delta-T throughout the night. An approach to solar power is partway there already: concentrating solar power uses mirrors to heat up a thermal mass — often a molten salt. Why a molten salt? High heat capacity and thermal mass — you can store a lot of heat and maintain a high delta T on a molten salt. And as a high temperature fluid you can pump it around and do the standard generator and heat exchange cycles. So, under the sun things work well with a 1,000C fluid kept hot by many mirrors directing the sun from the nearby area onto the molten salt. Surprisingly, to me at least, these configurations often cool down so much overnight that the molten salt solidifies. I’m not sure why these plants have been built to rely on natural gas burners to bring the salt up to temp in the morning, perhaps NG is just so cheap this made more economic sense than adding another mirror and storing the salt tank underground where it can easily be better insulated and vastly larger (creating a larger high temp reservoir with a smaller bleed out throughout the night — allowing 24/7/365 power generation without gas burners in the mix). This NREL study and summary on gtm point to a number of challenges that are just d’uh level to someone who ‘grew up’ in molten salts, solid oxide fuel cells, and thermoelectrics — you really don’t want to thermally cycle the system because of the differential heating, the effect of differential thermal expansion coefficients in multi-material joints, and the part where everything gets another opportunity to break. Anyway, CSP is an expensive build out, not as clean as you’d hope due to the NG burners, fragile in day to day use due to thermal cycling fatigue, but pretty compact on the ground. If you wanted to get really sneaky park some nuclear waste storage or a geothermal system underneath it to provide the ‘heat’ to keep the system molten through the night — but once you’ve put the thermal mass underground, you have vastly reduced the required heat to keep the molten salt above freezing point until the sun is out again. Seriously… I’m sure it was easier to put the thermal target at the focus point up in the air for alignment-slop (a degree of freedom that allows for the sun to move around in the sky), but just do the math and bounce the light straight down into a hole then close a huge iris, Startgate SG1 style, at night to turn off thermal escape mechanisms — we’ve known how to keep the hot side hot and the cold side cold since the McD-LT.
I’m pretty sure nobody saw this mash-up coming in a section on Concentrated Solar Power. Also, the McD-LT wasn’t the advent of thermal isolation and reliable HVAC, that was probably 3,000 years ago in North Africa — but we’ll get to the windcatchers next.
The counter claim and talking point around renewables is that you still need power when the sun isn’t out and the wind isn’t blowing. That’s true, for what it is worth, but I’m going to show that it isn’t worth very much — we can make the wind blow, reliably, even when the sun isn’t out. This isn’t rocket surgery.
Are there other mechanisms for 24/7/365 solar power generation? Actually, yes. Using a chimney effect one can induce a draft with modest delta T, and since one only needs to cool down slower than the rest of the environment to maintain that delta T this is viable at night (really, without direct solar incidence at all: windcatchers have served as reliable HVAC in desert areas for 3,000 years, but why bother with something that well established?). There has been thought of harnessing the heat-island effect of cities or building out vast collector arrays around a central stack to heat a large volume of air and sweep it up a chimney into the atmosphere — a Solar Updraft Tower. These towers generate power proportional to the height, which establishes the delta T, and the collector area, which establishes the size of the hot reservoir. Doing the math (tower efficiency depends only on the height, the collector just ensures the chimney doesn’t get staved for air) one can get 400MW with a 38sq km collector and a 1.5km tower. That’s extraordinary on both accounts height and area, but not so far out of the realm of possible today to be unworkable, there have been a few trials and a few attempts to construct smaller towers. The Manzanares tower was a test rig built by a German company in Spain in the early 80s. Kids
can make solar updraft towers for school projects, it has apparently become a thing recently. Chimneys aren’t that complicated at the end of the day[Citation Needed]. They aren’t that complicated in the morning either. The exchange for a facility like this is a high land-use cost, up-front lump sum construction financing, and fixed capacity and load profile in exchange for low operating costs for the duration of its service life. Oh, and the non-trivial engineering challenge of the tower itself — at useful scale these things can be really tall, and the
Manzanares tower was damaged in a wind-storm. The company behind the Manzanares Tower (Schlaich Bergermann Solar GmbH) produced a number of scale out options with tower heights from 750m to 1,000m and collector diameters from 3,750m to 7,000m, yielding generating power of 50–200MW. Pushing that up to a 400MW size for easy substitutionality with natural gas generators and we get to the 1.5km tower height.
In this configuration the plant becomes a navigation hazard for air traffic (1.5km tower) and a massive impact on the local ecosystem (raising the temperature dramatically for an area of 38sq km — the CSP plant above is 1/15 the size for 1/3 the power output for instance and a NG plant would fit on a city block). And that’s a lot of land for a power plant of that size, about 100x what you’d devote to a Natural Gas facility of similar size. If the land is cheap and location out of the way a solar updraft tower may make sense for baseline power. If you do something under the collector, for instance a photovoltaic deployment or ‘hot-house’ agriculture, you can double dip on that land — but still that’s a huge chunk of space on a front-loaded capital investment.
There was an Australian company (EnviroMission) a few years ago (ok, starting in 2000) with plans to build a 1.5km tower (operational by 2004), but that never got past website renders (and the revitalized coal industry in Australia created a bit of a hostile environment for environmentalism and led to restarting coal power plants for behind the grid bitcoin mining, which has subsequently become a thing in more parts of the world). The last update on Enviromission from GTM was in 2015, the company posted an update to their website last month about how Covid has increased the need for green energy solutions and things are looking up, but the last tweet is from 2016 and a previous note from 2019 indicates they’ve been unable to meet requirements to relist on the Australian stock exchange. The funding and incentives are backwards all around — leading to struggles everywhere and not just in the land down under.
So, solar updraft towers make a certain sense in the usual configuration, collector below chimney driving wind turbines through the temperature gradient. Some policy and financing possibilities might make funding (im)possible. Can we tweak the ‘normal’ state here and get a significantly better outcome, maybe eliminate one of the major blocking costs? How much can we perturb that configuration and still achieve expected function?
Trolling Elon Musk
There are two promising perturbations I’d like to talk about, the most promising in my mind. First, just lay down a photovoltaic solar plant under the collector. PV arrays already warm their sites to the point of disrupting the local ecosystem, and the addition of a roof over the PV array and steady consistent direction airflow (around 20mph) will reduce the cleaning requirement for the arrays themselves. This one seems like an easy co-generation win, allowing 24.7 baseline as well as daytime production at a much higher rate.
The second is where you have to do something clever, or just outright wrong, with the technology to get the positive outcome. That brings us to the question for Mr. Musk: How steeply can a Boring Company boring machine bore? If instead of standing up a narrow tower in the middle of the array, if you lay it down against a mountain you avoid the navigation hazard and tricky engineering of a freestanding tower — and the chimney still works. This is known, the main problem is that people tend to like to live in the foothills so you still don’t have a good place for a collector. But the chimney draw still considers the relative temperatures at the endpoints, the horizontal path only introduces a drag on the flow (stack or siphon effect, take your pick). That means you can locate the chimney in one place and the collector in another if you have a way of connecting the two. If you could somehow connect a vast collector in the desert to a shaft dropped right down the center of a mountain you can get your 1.5km stack height without any freestanding structure at all!
I’d originally thought of this as a means of establishing self-sustained and cheap to deploy power for Puerto Rico after Maria, where there are sufficiently tall mountains near wide planes and importing power leads to expensive and fragile systems. This topological mix can be found in many places around the world, some near large populations like Barcelona, Denver, even Southern California. The common challenge is that people have developed right up the sides of the hill — you’d need to open a pathway under 5–10 miles of city to get to a reasonably clear area to lay down a vast collector. Tunnels actually do that, and the Boring Company is both good at making holes in the ground and is active in many of those areas. A boring power generation technology for a boring company, is this a match made underground?
Ok, fine, maybe all that works. Where to test/build? Why not just north of Big Bear Lake? From the salt lake in Lucerne valley at 3,500' to the peaks of White Mountain at 7,800' 10 miles to the south you get 1.2km chimney just from the elevation rise and the 38sq km collector basin fits into the existing salt flat, which would put the solar updraft tower in the neighborhood of 400MW in capacity. Various co-generation and day-night thermal storage could be explored as well (PV under the collector for direct generation, salt lakes under the collector for a high capacity thermal store to keep the chimney drawing air through the night and overcast conditions, fields for hothouse farming under the collector for various high value crops, etc. — all valuable experiments to run to find best use of the collector area). Prop up a modest 300m tower at the peak and you get your full 1.5km stack height, plus your prominence remains modest while the mountain mass itself is already a natural navigation hazard air-traffic needs to be aware of — the tower isn’t adding a new hazard to the mix.
Looking at local shaft costs (10' diameter, 200' depth = $1.5M) and Boring Company price per mile the facility construction costs should fall well under the cost to construct a conventional 400MW power plant ($30M for the shaft + $100M tunnel cost + plastic or glass collector << $365M construction cost for similar capacity combined cycle gas turbine facility). It probably makes sense to mix cut and cover with boring for that 10-mile run, or stopping a couple miles short of White Mountain and putting a modest tower on a shorter peak for the same rise. But running the chimney horizontally between the salt flat and the base of the peak provides a lot more site flexibility and keeps the chimney out of sight and protected under the earth. This provides more places you could build these facilities and options to avoid opposition from neighbors — a real consideration for capital projects near population centers.
Anyway, napkin math may vary across different napkins — in particular the shaft diameter is likely linked to airspeed and turbine design and won’t be a nice round number like 10' in an optimal configuration of price vs. performance. My point is that we have incredibly simple solutions to some of our power problems that just don’t fit the conventional funding or construction paradigms. I’m going in another direction now, but this has always struck me as something simple, viable, and just hard to get into real-scale deployment due to the funding requirements and lack of established support infrastructure around construction like this. We aren’t providing a consumption path for an established extractive industry value chain, nor are we patenting a cool new technology to draw the VC interest. Just boring locally sourced energy self-sufficiency. I’m not in a position to do anything with this, so I hope it sparks someone else to think in a new direction and carry on. I’m pretty sure Schlaich Bergermann Solar GmbH does too.
More Solar Updraft Tower reading:
More McD-LT reading:
Big N’ Tasty — Wikipedia
Re-live the commercials: https://mcdonalds.fandom.com/wiki/McDLT