Hard to believe. ... Not based on how expensive this would but based on how cheap ground based PV has become. Panels are now hovering around $1/watt, which means that the balance of the system has to be less than $4/watt to compete even if it receives five times the power.
But in any case, if you have aerostat technology, you have access to very steady (and much stronger) winds than you do at the surface. I would think that would be the initial approach since it could produce power close to 24/7 - and uses wind to its advantage, whereas with the pure PV concept, wind is a liability. (Or a combination of both might work well.)
Interesting. At my installation I have gotten in excess of 1000 w/sq m, with an average of around 800 w/sq m at ideal sun angles.
I think their 300W/M^2 is the peak of the noon best monthly average, not of the best day and in the UK, not US SW as I think you are. Also the article is from 2008, and prices have come down for PV cells considerably but the cost of their BOS has not much.
I agree that the wind approach is better, if you want to generate max annual energy (works day AND night) and that high can be much better than tower mounted machine. Here is the MARS (Magenn corp) approach:
Comes with 1 yr warranty, but it isn't cheap.
They admit to a 0.5% loss of He and need to pull down and add more every 6 months - That is high cost in material and labor. My invention version could operate at slightly less than ambient pressure to reduce the He (or H2) loss rate, as used a rigid external frame of cheap (and easy to glue assemble) PVC pipes. {also important, as will be clear soon, PVC withstands harsh UV very well.}
This, my gas bag cage, had a central 20 foot long octagon cylinder with 30 degree pipe elbows adding 10 feet of PVC pipe to the ends,(half of std 20ft sections - nothing wasted in my design) then another set of 30 degee elbow making 10 foot sections at 60 degrees to the main axis and finally 4 of 8 converged 5 feet to the axis and joined in a 4-way pipe union. The other 4 continued on at 60 degree angle from the axis to the wall of a short 12 inch PVC tube. The other end of that tube was held fast to the 4-way unions. Inside that tube were two (possible 3 if needed, but I think not) rotation speed step up cascaded planetary gear sets. The rapidly spinning, on axis, small sun gear of the last drove the shaft of a standard PMA (Permanent Magnet Alternator) used in cars and sold by more than 30 million annually. Every thing used, but the He was cheap and produced in very high volume. The system used a Y shaped top to the tether, like shown for MARS units and that kept the PMA's case from rotating. Lot of finer details, including weigh /lift analysis, AC to optimum frequence AC converter,* cross bracing (also PVC octagons, but with their plains perpendicular to the axis) between 45 degree separated in angle main sections , and cost study omitted here.
* Lamps are even more efficient with mid range audio frequency AC, instead of 60Hz, as internal ionization stays high as AC voltage goes thru V = zero, very quickly.
Also I did not have heavy conductor wires to the ground or their copper cost as I used the power generated in situ to run cheap U shaped UVc discharges lamps to produce OH radical which the CH4 is destroying faster than nature replaces it. - I.e. the economic of my idea has failed with the collapse of carbon off set prices.
With a 20 foot long central cylinder the Magnus lift effect can be huge - hard to model, but perhaps if never stops spinning small pump could replace lost He with air. A ship with 4 vertical spinning cylinders, instead of sails, used Magnus effect to "sail" across the Atlantic years ago, but ship's roll made huge problems with that much angular moment being twisted. Most have never heard of the Magnus effect - it can be very strong lift force. It is why a baseball pitcher can indeed throw a very curved pitch but with cylinder instead of sphere he could hit the third base man!
These UVc lamps are sold by hundred of thousands for sterilization systems - basically are like fluorescent tubes of fused quartz but without the phosphors converting mercury's strong UV lines into visible light so the 50W one can put out 25W as UVc that splits H2O to make the OH radical.
I tell my invention as never planed to make a dime on it, but did want to recover the patent cost in a few years with tiny license fee. It is not a profit maker now with the low cost of carbon credits. - Perhaps if CH4 is killing voters, the government may pick up my idea as to how to kill atmospheric CH4.
Just so you know I had the invention write up done a few months ago, below is some analysis from it for the total volume:
The area of an octagon is (2 + 2 root2) e^2 where e is the length of one of the eight edges. For the largest octagon e = 1.428L or 14.28 feet. The length of the central section of the cage, 2L = 20 feet. Thus the volume of that central section is (2+2 root2)(14.28)(14.28)20 = 19,692.10 cubic feet.
The volume of a right circular cone is (1/3)A H, where A is the area of the base and H is the height of the apex above base. This is also a very good approximation for the volume of a polygon cone with an octagon base with A still the area of the octagon's base and H still the base to apex height.
The second largest volumes of the cage, are the two bottom parts of an 8-sided “cone” with L = 10 long edges inclined 30 degree to the axis. The corners of the largest octagons were 18.66 feet from the axis, so H for the full cone is found from H tan(30) = 18.66 or H = 18.66 / 0.57735 = 32.3201 feet but only the part h from the bottom is real volume in the cage, where h = L cos(30) = 10x0.866 = 8.66 feet. I. e. the real volume truncates in the middle sized octagon, which has area (2+2 root2)(10.4548)(10.4548) = 527.761 square feet. Thus, the volume of a “cone” with this area base and height of 32.32.01 – 8.66 = 23.66 feet must be subtracted from the full volume of the imaginary “cone” base area of 984.61 square feet and height of 32.3201 feet. By the approximation V = (1/3) A H the full volume, largely imaginary, is (1/3)984.61x32.3201 = 10,607.56 cubic feet and the volume not real is (1/3)527.761x23.66 = 4,162.28, so the real volume is: 6,445.28 cubic feet, but there are two such volumes, one on each side of the central 2L long prism. Thus the volume of the three largest central sections is 19,692.10 +2x6,445.28 = 32,582.67 cubic feet.
Text goes on the same way to get the volume of next two sections (and doubles it) to conclude the total is:
35220.63, but call it 34,000 cubic feet of lift gas as the way the balloon is attached to this external cage, to be described later, will make the balloon slightly smaller than the internal volume of the cage.