Environment

3D solar towers offer up to 20 times more power output than traditional flat solar panels

View 2 Images
Two small-scale versions of three-dimensional photovoltaic arrays that were tested by MIT researchers (Photo: Allegra Boverman)
Two small-scale versions of three-dimensional photovoltaic arrays that were tested by MIT researchers (Photo: Allegra Boverman)
Two of the 3D PV arrays tested by MIT researchers that showed a boost in power output ranging from double to more than 20 times that of fixed flat solar panels with the same base area (Photo: Allegra Boverman)

While we’ve looked at the development of solar cell technologies that employ nanoscale 3D structures to trap light and increase the amount of solar energy absorbed, MIT researchers have now used 3D on the macro scale to achieve power output that is up to 20 times greater than traditional fixed flat solar panels with the same base area. The approach developed by the researchers involves extending the solar cells upwards in a three-dimensional tower or cube configuration to enable them to better capture the sun's rays when it is lower on the horizon.

Solar panels placed flat on a rooftop are most effective at harnessing solar energy when the sun is close to directly overhead, but quickly lose their efficiency as the angle of the sun’s rays hitting the panel increases – during the mornings, evenings, in the cooler months and in locations far from the equator. It is exactly in these situations that the researcher’s vertical solar modules provided the biggest boosts in power output.

After exploring a variety of possible 3D configurations using a computer algorithm and testing them under a range of latitudes, seasons and weather with specially developed analytic software, the team built three different individual 3D modules and tested them on the MIT lab building roof for several weeks. The results showed a boost in power output ranging from double to more than 20 times that of fixed flat solar panels with the same base area.

Two of the 3D PV arrays tested by MIT researchers that showed a boost in power output ranging from double to more than 20 times that of fixed flat solar panels with the same base area (Photo: Allegra Boverman)

By going vertical and collecting more sunlight when the sun is closer to the horizon, the team’s 3D modules were able to generate a more uniform output over time. This uniformity extended over the course of each day, the seasons of a year, and even when accounting for blockage from clouds and shadows.

The researchers say this increase in uniformity could overcome one of the biggest hurdles facing solar energy – predictability of electricity supply that currently makes it difficult to integrate solar power sources into the grid.

They add that this uniformity, as well as the much higher energy output for a given area, would help offset the increased cost of the 3D modules, which are higher per the amount of energy generated when compared to conventional flat solar panels.

While the team’s computer modeling showed complex shapes – such as a cube with each face dimpled inward – would offer a 10 to 15 percent improvement in power output when compared to a simpler cube, these would be difficult to manufacture. In their rooftop tests, the team studied both simpler cube modules as well as more complex accordion-like shapes that could be shipped flat for unfolding on site.

This accordion-like tower was the tallest structure the team tested and such a design could be installed in a parking lot to provide a charging station for electric vehicles, according to Jeffrey Grossman, the senior author of the study and the Carl Richard Soderberg Career Development Associate Professor of Power Engineering at MIT.

Grossman and his colleagues believe that with the fall in the cost of solar cells in recent years - to the point where they have become less expensive than their supporting structures and the outlay for the land upon which they are placed - makes it an ideal time to examine the benefits of different solar cell configurations.

“Even 10 years ago, this idea wouldn’t have been economically justified because the modules cost so much,” Grossman says. But now, he adds, “the cost for silicon cells is a fraction of the total cost, a trend that will continue downward in the near future.”

Buoyed by the success of the tests on the individual 3D modules, the team now plans to study a collection of solar towers that will enable them to examine the effects that one tower’s shadow will have on another as the sun moves across the sky over the course of a day.

While the team believes its 3D solar cells could offer big advantages in flat-rooftop installations or urban environments where space is limited, they say they could also be used in larger-scale applications, such as solar farms, once a configuration that minimizes the shading effects between towers has been developed.

The results of the MIT team’s computer modeling and rooftop testing of real modules appear in the journal Energy and Environmental Science.

Source: MIT

  • Facebook
  • Twitter
  • Flipboard
  • LinkedIn
28 comments
Rt1583
Are they equating base area to surface area? I always thought base area was the area occupied by the base of an object, its footprint. If this is the definition of base area, then it is no suprise that they are able to increase the power output since (using the solar tower on the left in the picture) they were able to stack 8 solar panels (say 12"x12") within the footprint of one flat panel (also 12"x12"). It would be interesting to know the power output comparisons of this system, the fixed panel and a motorized sun tracking panel to get a truer idea of its potential.
If they are equating base area to surface area, this would be fantastic since it would seem that a solar systems actual surface area could be increased dramatically relative to its footprint.
So, is base area the same as surface area?
Kumi Alexander
MIT has recently been coming up with solutions to nonexistent problems, from the so-called "flying car" to this worthless monstrosity. The problem with solar cells has never been cost per square mile of land but cost per square foot of solar panel. This not only doesn't solve the real problem, but actually makes things worse by decreasing the efficiency of each solar cell and overcompensating with quantity. Idiotic.
marosini
Mark, I completely agree with you that the REAL problem is always been cost per PV panel surface. And if you consider the energy per PV panel surface this configuration is a clear, huge LOSS, if compared with optimal mounting in unlimited ground footprint! This kind of rack mount could be useful only in very peculiar situations, when you have a small base surface available, and money to waste (in comparison) for the panels. And you can not put two of this towers close, that's for sure. The stability of the output during the day can maybe be interesting for energy management and storage optimization.. But that's not MIT-level kind of an innovation, not at all.
Robert Smithers
Mark and Marosini, you are both absolutely right, the major expense with solar in the past was always cell cost not land area. The fact that things like this are starting to be looked at is because as panel costs fall below 80c/W the balance of costs begins to swap to land area. In utility scale projects land acquisition costs are now around 60% of total project costs. This means that looking forward with cell costs set to fall further strategies to maximise output from minimal land or roof footprint will be more cost effective. It also mean expensive and fallible tracking systems are more cheaply replaced with larger numbers of panels positioned to capture different sun periods is becoming more cost effective. I agree that this is a clumsy looking solution but the significance is in the interest being shown and the implications of it.
Yusuf Khan
This is very stupid. The simple solution for the problem of solar cells not absorbing oblique sun rays would be to motorize the panels to follow the sun and the problem is solved. This just wastes solar panels and money because MIT are too lazy to motorize their PV panels.
Mike.H
I guess that I'm the only one who read the detailed information on this. This option gives you a 100 to 1900 Percent increase in power with a 50 to 300 increase in cell surface. I think that that gives this a bit of promise.
t16460
Where's the data for a avg day? Next thing is these will get longer and maybe start to look like a round tube and suddenly it they've rediscovered a solyndra panel. Can you guess what the installation cost of something like this would be? Great for a military outpost etc but not for a power plant.
Robert Moynihan
OK these are going to be a bitch in high winds
tmtempe
I think it would be easier and very inexpensive to rotate the tower to follow the sun daily and to change the angle of all panels with the seasons than with fixed panels. This could add significantly to the electricity produced per cell.
jeremytp
I'm surprized at how many people are slamming MIT for this research. Granted, you are correct that the article is desceptive because the "efficiency" is defined as power per base area and not panel surface area. However, there are many situations where this kind of technology would be beneficial--especially considering market trends. If the cost of solar panels continues to decrease, it could be cheaper so set the panels up in their configuration rather than spreading out the panels and motorizing them to follow the sun. Also, like they said, this would be the ideal configuration for crowded urban areas. I believe that they would have a good market base to sell these to wealthy urban consumers would are concerned about their enviromental impact (think Beverly Hills).
By the way, Mike.H, I think that you need to analyze the claims in the article again. To me, the picture shows that their design uses either 23 or 32 panels on a base area roughly the size of 2 panels. That would mean a 1050% or 1500% increase in cell surface. Also, I believe that in order to get "20 times" the efficiency (as they claim) they would have to make a tower about 30 panels tall, which would need about 60-120 total panels.