July 13, 2006
Soy Better Than Corn For Biomass Energy
Soy for biodiesel is better than corn for ethanol but even soy has a very limited role to play as an energy source.
MINNEAPOLIS / ST. PAUL (7/10/2006) -- The first comprehensive analysis of the full life cycles of soybean biodiesel and corn grain ethanol shows that biodiesel has much less of an impact on the environment and a much higher net energy benefit than corn ethanol, but that neither can do much to meet U.S. energy demand.
The study, which was funded in part by the University of Minnesota’s Initiative for Renewable Energy and the Environment, was conducted by researchers in the university’s College of Biological Sciences and College of Food, Agricultural and Natural Resource Sciences. The study will be published online July 12 in the Proceedings of the National Academy of Sciences.
The researchers tracked all the energy used for growing corn and soybeans and converting the crops into biofuels. They also looked at how much fertilizer and pesticide corn and soybeans required and how much greenhouse gases and nitrogen, phosphorus, and pesticide pollutants each released into the environment.
“Quantifying the benefits and costs of biofuels throughout their life cycles allows us not only to make sound choices today but also to identify better biofuels for the future,” said Jason Hill, a postdoctoral researcher in the department of ecology, evolution, and behavior and the department of applied economics and lead author of the study.
The study showed that both corn grain ethanol and soybean biodiesel produce more energy than is needed to grow the crops and convert them into biofuels. This finding refutes other studies claiming that these biofuels require more energy to produce than they provide. The amount of energy each returns differs greatly, however. Soybean biodiesel returns 93 percent more energy than is used to produce it, while corn grain ethanol currently provides only 25 percent more energy.
Still, the researchers caution that neither biofuel can come close to meeting the growing demand for alternatives to petroleum. Dedicating all current U.S. corn and soybean production to biofuels would meet only 12 percent of gasoline demand and 6 percent of diesel demand. Meanwhile, global population growth and increasingly affluent societies will increase demand for corn and soybeans for food.
The authors showed that the environmental impacts of the two biofuels also differ. Soybean biodiesel produces 41 percent less greenhouse gas emissions than diesel fuel whereas corn grain ethanol produces 12 percent less greenhouse gas emissions than gasoline. Soybeans have another environmental advantage over corn because they require much less nitrogen fertilizer and pesticides, which get into groundwater, streams, rivers and oceans. These agricultural chemicals pollute drinking water, and nitrogen decreases biodiversity in global ecosystems. Nitrogen fertilizer, mainly from corn, causes the 'dead zone' in the Gulf of Mexico.
Advances in biotechnology will continue to increase crop yields and increase energy efficiency of agriculture. But total demand for energy will rise quite rapidly. Still, biomass's prospectives will improve when cellulosic technology matures to allow use of all of a plant for production of energy. Yet, even then I expect biomass to play only a minor role in providing energy.
Development of cheaper and higher efficiency photovoltaic materials seems like a better long term prospect than biomass for several reasons. First off, photovoltaics can provide energy all year rather than just during the growing season. Second, photovoltaics will provide energy even during droughts. Third, photovoltaics allow energy to be generated closer to where it gets used. When used to cover a building (e.g. with photovoltaic tiles) photovoltaics generate electricity where it gets used. Fourth, photovoltaics reduce the need for use of additional land to generate more energy. Photovoltaics on buildings and other human structures do not cover more ground than those structures already cover. Fifth, photovoltaics can get placed where few plants will grow (e.g. deserts) and therefore again won't compete as much with wildlife as agriculture does. Sixth, even when plants are growing their efficiency for turning light into chemical energy is lower than what photovoltaics wll achieve in the future.
We should accelerate photovoltaics research and development. Ditto nuclear research. Ditto batteries research (though already plug-in hybrid cars are coming).
In spite of lame government policies with regard to solar power Technology Review reports many signs that solar is starting to take off.
The announcement last month that Palo Alto, CA-based Nanosolar had raised $100 million to finance a new solar-cell factory based on an inexpensive process, similar to that used to print newspapers, and that it will make enough cells to produce 430 megawatts of power annually, is just one sign that new types of solar power are emerging as a viable alternative energy source (see "Large-Scale, Cheap Solar Electricity").
While Nanosolar's new factory capacity, equivalent to one-quarter of the total global solar capacity last year, is unprecedented for a new technology, it's just part of equally impressive overall growth in the solar industry. For the last several years, solar cell production has been doubling every two years, and indicators suggest this will not slow soon, says industry analyst Michael Rogol, managing director of Photon Consulting in Aachen, Germany.
The price of oil just exceeded $78 per barrel. The high cost of oil is doing more to accelerate development of new energy technologies than all government energy policies together. If biomass can make the grade high oil prices mean we'll find out.
This blog does a nice job of covering energy issues. (Energy Pundit?) In a perverse way, we are lucky that political turmoil is artificially inflating the cost of oil, as it gives us a headstart in dealing with the coming peak in world oil production. Take away problems with Iran, Nigeria and this current spat, and oil is at $55. I am more optimistic about biofuels than the proprietor of this blog, because much of the corn we grow actually feeds livestock. I would be happy to see these crops diverted to biofuels, and would not care one whit if the price of beef tripled as a result. Further, there are crops that are much oilier than soy, like rapeseed, that can be grown for biodiesel. The byproduct of rapeseed biodiesel is a high quality livestock feed. Finally, in warm cities like Los Angeles, Miami, and Phoenix, cars can run on straight vegetable oil. This sidesteps the energy intensive transesterification process, greatly improving the energy balance.
When oil hits $100/bbl (next week? tomorrow?), you will see this country marshall our prodigious know-how to secure our energy future. The US has vast coal reserves, billions of barrels of "last resort" oil in ANWR and GOMEX, an endless supply of oil shale ripe for exploitation, abundant arable land, windy coasts, sunny deserts, and proven nuclear technologies. We lack resolve. And leadership.
At what point does solar technology get so cheap that it can affordably purify sea water. When this happens, will the American Southwest bloom? How much carbon would that aborb?
US Land Area: 9.2*10^8 ha
Arable Land in US 18% or 1.7*10^8 ha
Table 822 Cropland:
In Crops: 1.4*10^8 ha
Land in Farms 2002: 3.8*10^8 ha (1953 4.9 10^8 ha)
US Annual fuel consumption: 6.7*10^11 liters
Hippy Dippy Web site:
Crop liters oil/ha
corn (maize) 172
pumpkin seed 534
rapeseed (canola) 1190
oil palm 5950
If we could increase the land in crops by about 20%, we could devote about 3*10^7 ha to oil crops. If we could average 1,000 l/ha*, we would have 3*10^10 of oil. That is about 5% of our current use. In order to run our transportation system on bio-diesel, we must bridge a gap of at least 4 doublings.
Could we bridge the gap? Pushing as hard as we can on automotive technology, we could certainly double fleet efficiency. We might even be able to quadruple it.
1,000 l/ha* is not a stretch. But, of the crops with good oil yields, the only ones on the list above that have any promise and which are not tropical or sub-tropical, are rapeseed and sunflower. However, we have increased corn yields by over 8 times in the modern era. So we might be able to close the gap that way.
*1l of raw oil=~800ml of bio-diesel
Clearly, the algae approach, if it does not interfere with agriculture would be preferable to plowing up that much additional land.
Proposals for solar updraft towers have typically assumed that they would be single use structures: solar to electricity via heat differentials between high altitude air and ground level greenhouse-enclosed air.
Something which would further enhance the value of the solar updraft tower power structure is to use the greenhouse area for algae ponds to add biodiesel, water, fish and salt production.
Doing so brings the proposal from marginally viable to viable, with a net present value, primarily from live fish production, of $3.5 billion, and even allowing for a construction cost double that of the simple solar to electric system.
Let's start with just the value of algae biodiesel:
The greenhouse area required per solar updraft tower of the reference design:
(pi * (5km/2)^2) ? hectares
= 1963.49 hectares
producing peak at peak 200MW via a 1km tall tower.
We now add to this the production of algae biodiesel:
The UNH estimate http://www.unh.edu/p2/biodiesel/article_alge.html for algae biodiesel production is 1 quad per 200,000 hectares. Let's assume only half of the area of the solar updraft tower greenhouse would be available for production at any time (the rest of the area is dedicated to ponds for heat buffering, aquaculture, residential value enhancement and possibly evaporative recovery of high value salts).
That gives us:
(1963.49/2)hectares/tower;200000hectares/quad ? towers/quad
= 203.719 towers/quad
Or about 200 towers per quad of biodiesel.
We can now calculate the biodiesel per tower:
7.2gallon/1e6btu;200tower/quad ? gallon/tower
= 3.5998E+07 gallon/tower
or about 35M gallons of biodiesel per year per tower.
At $2/gallon for wholesale diesel, this yields $70M biodiesel revenue per year.
Now for electrical revenue:
At an average rate of sold production only 1/2 (100MW) of peak capacity (200MW), electrical production per tower per year, is:
100MW;year ? GWh
= 876 GWh
At $30/MWh wholesale:
100MW;year;30$/MWh ? $
= 2.628E+07 $
or about $25M electrical revenue per year.
Interestingly, the biodiesel revenue is nearly 3 times the electrical revenue of a solar updraft tower!
200*200MW or 40GW electrical peak capacity is produced per quad of biodiesel.
Further that same UNH document estimates 19 quads to replace all transportation fuel in the US or 3800 towers, which would also produce 3800*200MW or 760GW or .76TW of electricity.
Current winter capacity in the US is about 1TW:
So this cannot replace the entire US capacity but it can probably support all new growth in demand for the next several decades.
For reference, 3800towers at 1963.49hectares/tower would require:
3800towers;1963.49hectares/tower ? hectares
= 7.46126E+06 hectares
or about 8 million hectares or close to 30,000 square miles -- a figure that cross checks with the UNH figure of 15,000 square miles of optimally productive algae ponds which we are assuming are only half of our land area due to needed additional greenhouse warming area.
An additional advantage of this approach is that the relatively constant wind velocity and direction through the greenhouse disk would allow for
the efficient use of wind for driving the algae raceways.
Now for desalination:
We're going to assume the algae ponds are growing a saline species like CCMP647, and that about half of them are not producing biodiesel. These ponds would be out of algae production but would still be providing water for desalination, a market for the residual salts, live fish, climate control and residential real estate value.
Let's assume that out of the 8 million hectares, half of which is growing algae at reasonable efficiency and therefore providing 4 million hectares of evaporative surface, an additional 2 million hectares are in reasonably efficient production as evaporative ponds, some of which is salt production and some of which is fish production, for a total of 6 million hectares of evaporative surface. Then let's assume the additional difficulty of evaporating from saline cancels that gain out leaving us back at 4 million hectares equivalent fresh water evaporative surface. Using "Open water bodies in the Phoenix area evaporate at about 6.2 acre-feet per year (about two million gallons) per year for each acre of surface area." from:
2*10^6gal/acre;4000000hectares ? gallons
= 1.97684E+13 gallons
or about 20Tgal per year.
Estimating total US demand:
132gal/person/day;300Mperson ? gallon/year
= 1.4454E+13 gallon/year
or about 14Tgal per year.
The entire US requirement for fresh water can be approximately replaced with the desalinated water from the solar updraft towers.
(An objection to this combined use of the solar updraft tower is that the heat of vaporization lost during evaporation will translate into a lower temperature differential between ground and exhaust at the tower head. However, this ignores the recapture of that heat upon condensation -- a phenomenon that drives powerful natural updraft phenomena such as thunderheads.)
Each tower's water output:
1.97684E+13 gallons/3800 ? gallons
= 5.20221E+09 gallons
and at a penny a gallon (remember this is high quality, nearly
1.97684E+13 gallons/3800;.01$/gallon ? $
or about $50 million/year in water revenue. In all likelihood this would be much higher given markets for the distilled water could be found.
Salt is about $25/ton at the mine mouth:
And the ratio of sea water to salt mass is about 65:1 so the revenue
from sea salt is:
25$/ton_salt;65ton/ton_salt;tonm/m^3;5.20221E+09 gallons ?
= 7.57404E+06 $
Or about $8 million/year in salt revenue. Perhaps this can be brought up by arranging radial evaporative ponds to fractionally crystallize higher value salts and accounting for the elimination of a return pipe for waste brine, but from salt value alone it seems barely worth the investment.
Now to live fish:
If the algae is 50% oil, and extraction of the oil isn't total, we can conservatively assume the mass of oil-depleted algae will approximate
the mass of biodiesel:
0.827 g/ml;3.5998E+07 gallon ? tonm
= 124223 tonm
Trophic losses in acquaculture algae grazers are about 1/3:
And the price per kg of live fish at the producer is conservatively $2/kg:
124223 tonm*.67;2$/kg ?
= 1.51009E+08 $
Or about $150 million/year in fish live fish revenue. This is a really big deal! Its equal to the electricity, biodiesel, water and salt production combined!
Totaling up yearly revenues:
$150M for live fish
$ 70M for biodiesel
$ 50M for fresh water
$ 25M for electricity
$ 8M for salt
$303M TOTAL REVENUE
Discount 20% for operation costs ($60M) and the yearly profit available is $240M/year or $20M/month.
A profit stream of $20M/month at 6% interest over 30 years has a net present value of $3.5 billion.
This compares very favorably with the estimated construction cost of the reference tower of $500M to $700M, which, of course, will have to be increased to account for the addition of a condenser to the tower, pond construction, centrifugal algae harvesters, boidiesel equipment, aquaculture equipment and brine transport systems. Indeed, the construction estimate can be doubled without threatening the viability of the system.
PS: The dimensionally solved units expressions above use Unicalc Live from Calchemy:
Interesting question. I expect desalinization costs to fall as technology advances anyway. But add in a big decrease in energy costs and, yes, desal would get much cheaper. The nice thing about using solar for desal is that solar's lack of continuous availability does not pose a problem. Water spends weeks travelling down aqueducts. It doesn't have to get pumped right when you need it. For farming the water can get added to fields whenever the sun shines enough to power the desal operations.
You are in a small minority about meat versus car fuel. The public will wail about high meat costs even more loudly than about high gasoline costs.
The land that is currently used for farming has more water and higher quality soil and climate than the land that is not used for farming. The marginal cost would be higher if more land was pushed into farm production.
You can get about 150-160 bushels of corn per acre of average land now in production. That's about 6200 BTUs per bushel. Well, how many BTUs of sunshine falls on that average corn crop acre per year? I bet it is orders of magnitude more in energy. I'd like to find the answer to that question. It would make a great topic for a post about photovoltaics versus biomass.
Thanks much for the write-up on solar updraft towers.
Some questions and issues:
1) Could such a tower withstand a hurricane? At what cost of damage?
2) I assume these things have a lot of glass on their surfaces. Though would they have glass on their northward faces?
3) Could such a tower withstand a tornado? At what cost of damage?
4) Would they need to be built near oceans for the salt water for the fish? Or could the water be piped affordably some distance inland? I'm thinking the closer they are to the ocean a few problems arise:
- People do not want their ocean views blocked by towers.
- The towers would cast large shadows on expensive ocean-front property.
- Hurricanes hitting land would do more damage to structures closer to the ocean.
5) Could the towers withstand earthquakes? The West Coast of the US gets earthquakes but not hurricanes.
6) Could a Bahama island or other island support such a tower? I'm thinking the Brits have a sizable and not populated island somewhere off the coast of Africa. But the location would preclude the sale of electricity. The idea seems highly location constrained.
7) Could the tower be built in deep water on top of the sort of structures oil rigs use? Probably not viable. Those structures alone can cost a billion dollars can can't support as much as a tower would weigh.
8) What location would create the biggest updraft? Low hot Death Valley perhaps?
How much of the increased efficiency of "soy diesel" over ethanol is due to the diesel engine itself? The data should be broken down into, e.g., kilocalories per dollar as well as hp-hr per dollar, otherwise it is like comparing apples to oranges. (Not that they can't be compared.) Why not compare soy diesel to ethanol in a fuel cell?
Plug in hybrids could cut down drastically on liquid fuel consumption. But if fuel costs drive people to buy them (rather than environmental concerns), don't expect the electricity to be coming from solar.
Solar electricity isn't going to be as cheap as coal electricity for a long time, so from a cost viewpoint, it's almost a red herring (at least without factoring in externalities).
So I take it that you don't think that Nanosolar or any of these other companies has a chance of doing what they're saying? To quote some information about Sterling Energy Systems, the guys planning on building what may wind up as close to 2 gigawatts of solar power out in the Mojave desert:
" They promise a cost of electricity of 0.06$/kWh for large projects. If they manage to do that, this is competitive : roughly the same price as wind power, but production occurs at daytime (more valuable electricity) and is quite reliable (in a desert, weather is quite predictable) "
--- Sumyung Guy
"Well, how many BTUs of sunshine falls on that average corn crop acre per year? I bet it is orders of magnitude more in energy. I'd like to find the answer to that question. It would make a great topic for a post about photovoltaics versus biomass."
NREL says that the annual average solar radiation available (insolation) to a flat plate collector, such as a photovoltaic panel, oriented due south at an angle equal to the latitude of the location, in Iowa, ranges from 4.5 to 5.5 kWh/m^2/day. The highest number in the US is 6.5 to 7 in the desert around the CA, AZ, NV borders. We will use 5 kWh.
[The solar constant is the amount of energy received at the top of the Earth's atmosphere on a surface oriented perpendicular to the Sun’s rays (at the mean distance of the Earth from the Sun). The generally accepted solar constant of 1368 W/m^2 is a satellite measured yearly average. Link]
A Joule is one Watt Second, so 1 kWh is 3,600,000 Joules*. A solar collector in Iowa would collect 18,000,000 J/m^2/day or 6,570,000,000 J/m^2/yr or 6.6*10^9 J/m^2.
One BTU is approximately 1055 J. So the annual solar energy falling in a field in Iowa is about 6.3 10^6 BTU/m^2
1 acre = 4,046.873 m^2, so the acre would recieve 2.5*10^10 BTU. Using your figures above 160bu/acre and 6200 BTU/ Bushel, we get 992,000 BTU of Ethanol per acre. Which is 0.004% of the insolation. I think that is roughly 10.5 minutes of sunshine.
"The land that is currently used for farming has more water and higher quality soil and climate than the land that is not used for farming."
I do not know that this is true. Land has been taken out of production for many reasons, some of which are unrelated to its productivity. Some has been swallowed by urbanization. Some has been idled by the machinations of Federal agriculture policies. At any rate my calculation was not intended to be anything other than a back of the envelope as to whether we have a biofuel future.:-)
Great! Thanks for all that!
To establish a basis for the corn side of those calcs Nebraska gets about 160 bushels of corn per acre.
A growing percentage of corn devoted to ethanol will be offset by advances in biotechnology that will continue the trend of rising corn production, Hutchens said. The state's corn production, 1.3 billion bushels last year, grew from an average of 126 bushels per acre in 2000 to 160 bushels per acre in 2005.
Missouri also gets 160 bushels of corn per acre.
Henggeler said Southeast Missouri farmers are in a period where they have to keep irrigating. And irrigation can improve a crop’s performance tremendously.
“Historically, a corn yield is going to be about 110 bushels per acre and those who irrigate will have 160 bushels per acre, or a 50 bushel difference,” Henggeler said.
For soybeans, dry land yields 30 bushels per acre and irrigated land yields 43 bushels per acre, Henggeler said.
My memory was faulty on the BTUs per bushel. Sorry about that. The number I gave is really the BTU value per pound, not per bushel.
It needs to be emphasized that the energy content of shelled corn is in the range of 8,000 to 8,500 BTU per pound of dry matter, based on bomb calorimeter studies. The term "dry matter" refers to material that is "bone dry." The standard moisture content of shelled corn is 15.5 % moisture on a wet basis. This means that each pound of shelled corn will actually consist of 0.845 pound of dry matter and 0.155 pound of water. Using a median energy content value of 8,250 BTU per pound of dry matter, the energy content of one pound of shelled corn at 15.5% moisture is then 6,971 BTU (8,250 BTU per pound dry matter x 0.845).
So instead of capturing 0.004% of the insolation the amount is 0.004%*(6971/6200)*56 = .25%. Corn biomass still comes out looking bad. Switchback grass converted with cellulosic technology might do a few times better per acre. Not sure if it will reach 1% of total insolation though.
The BTUs per acre for switchback grass converted by cellulosic technology would be a few times higher. But of course there are energy costs as agricultural inputs and in conversion losses.
I, confusingly, wrote: "We're going to assume the algae ponds are growing a saline species like CCMP647, and that about half of them are not producing biodiesel."
I should have written: "We're going to assume half of the greenhouse area is covered by algae ponds growing a saline species like CCMP647, and that the other half of the greenhouse area is covered by ponds not producing biodiesel."
Now to answer RP's questions:
1) I don't think the towers would be constructed to withstand hurricane winds. They are most suitable for desert areas, which are generally not hit by hurricanes.
2) The transparency surface would be the greenhouse disk -- 5km diameter in the reference design. The NREL study concluded that production is limited by maintanence of temperatures conducive to algae growth but that a greenhouse covering for the ponds would render biodiesel so derived marginally economic. The idea here is to cost-share the greenhouse between synergistic uses.
3) A tornado would probably destroy the tower as well as a hurricane.
4) It would be best for them to be near oceans but my guestimate on the cost of ocean water transport is that it would be something like 10% of the cost of construction. The construction cost needs more work since there is a lot proposed here that doesn't come with the $500M-$750M cost estimate of the simple solar updraft tower build just for electric generation.
5) I suspect the towers could be made to withstand the sort of tremors likely to hit the desirable desert areas.
6) An island 5km in diameter could support such a tower and if tropical it would benefit from the warm water surrounding it. Population could surround the greenhouse disk. Each tower can support 200,000 people at a level of consumption similar to the US percapita consumption rate of electricity, fuel, fresh water and fresh fish.
7) I suspect oceanic solar updraft towers would have to be built as floating structures vertically stabilized by extending the tower deep below the surface. This extension would probably require some more conceptual mods using OTEC type technologies to get it to pay for itself.
8) Death Valley would be good for updraft potential but also because land costs are low and it is below sea level so pumping costs to feed it sea water would be lower.
"They promise a cost of electricity of 0.06$/kWh for large projects"
1) With nuclear, "they" promised us that it would be so cheap, it wouldn't be worth metering, so I take what "they" say with a grain of salt, before things get up and running.
2) 0.06 is about what I pay now - my electricity comes almost entirely from coal. Unlike their projection, that includes transmission costs to my house, and is calculated after the energy loss that occurs between the plant and my house. It also includes the electicity company's profit. Checking my "generation charge" on my bill, it's 0.03907/KWH.
3) I live a long way from the Southwest. Somehow I doubt it's going to work as well in DC. :)
Jim Bowery: I like the idea, but I think your tower height is much too low to recover much moisture in desert areas; you need too much cooling to get heated air down to its dew point, and that takes far more delta-H than the proposed tower has. (Albeit not more than Leon O. Billig's concept in "Defeating the son of Andrew" would provide.)
Tom: Nuclear advocates never promised anything "too cheap to meter". That wasn't a promise, that was what one executive said when he was There's a huge difference between wild speculations and promises, both intellectual and moral.
Hmph. Missing sentence fragment should read "... asked to speculate about what might be possible."
Here's the original quote by Lewis Strauss in a bigger textual context:
"It is not too much to expect that our children will enjoy in their homes electrical energy too cheap to meter, will know of great periodic regional famines in the world only as matters of history, will travel effortlessly over the seas and under them and through the air with a minimum of danger and at great speeds, and will experience a lifespan far longer than ours as disease yields and man comes to understand what causes him to age."
Lewis L. Strauss
Speech to the National Association of Science Writers, New York City September 16th, 1954.
I think it is not sensible to hold the entire nuclear power industry responsible for the excessively optimistic hopes of one guy over 50 years ago.
Lots of investors will decide whether to build a new generation of nuclear power plants. We will have to wait to see what they decide. What I find disappointing is that nuclear power is probably not cheap enough to compete without tougher emissions regulations on coal. Therefore effectively the price of electricity must rise either to burn coal more cleanly or to use nuclear power.
I don't think a tornado or a hurricane would destroy the tower itself. It'll probably be made of cheap durable materials and it most likely won't have any features to offer resistance. Like somone said, this is for desert areas where a tornado is a once in a 100 years and a hurricane never hits.
I had an idea that went along with this. There's a patent in the U.S. for a downdraft tower that sprays water at the top of a similar sized tower which cools the air which makes it drop down the tower and out through turbines for power production. The latest incarnation was in Israel, which you can still find references to.
I'd like to know if a dual power tower would be possible. The tower they propose is 400m in diameter for both projects, and I was thinking if you put the down draft tower in the center at 400m and the up draft tower around that the diameter would have to be 501m to get the exact same area. It would seem that the construction costs would be shared by both and a lot of what's needed for both projects they have in common.
So you'd have a double ring, the inner generators would be below ground with vent cold air discharged at the outer edge of the greenhouse area directed inward.
The sun is the power driver both ways. The sun heats up the air under the glass and it goes up, the sprayed water in the inner tower condences because of the warm dry air and goes down. It would be more of a cycle with needed cool air under the glass and needed warm dry air exiting up at the tower.
It's not a perpetual motion hoax or anything, the sun would power this.
The one thing that a downdraft tower needs that strictly speaking an updraft doesn't is water.