Re: Space Solar Power - new CalTech design details

174 views
Skip to first unread message

Darel Preble

unread,
Apr 11, 2016, 10:18:34 PM4/11/16
to M.V. "Coyote" Smith, Peter Garretson, sbs...@googlegroups.com, Paul Jaffe

        SpaceX nails first-stage rocket return landing-at-sea/
 

        Ariane 6 rocket designers-say-theyll-match-or-beat-todays-spacex-prices-on-per-kilogram-basis/
 
        More on CalTech's SSP design ...

"Space solar power"
Daily update April 11, 2016


Inhabitat
SPPI's 2500 orbital solar “flying carpets” could power the planet
One need not venture to the Cave of Wonders to discover the magic of flying carpets. The Space Solar Power Initiative (SSPI), a collaboration between Caltech and global security company Northrup Grumman, has proposed the development of solar paneled “flying carpets,” each nearly the size of a football field, that would orbit in sync while gathering energy. This interstellar solar energy would then be beamed down to the planet to provide clean power across the globe.


When the proposed 2,500 solar carpets are installed in orbit, they will cover an area of 3.5 square miles – though only an inch thick. The solar carpet constellation would be especially useful to provide power to impoverished, rural areas. In places where electrical infrastructure is nonexistent, “it’s easier and more economical to deploy a wireless network there,” says Caltech researcher Ali Hajimiri. To avoid occupying already scarce space, the energy receivers could be built upon already existing buildings or even attached to chicken wire over farmland.

Related: Japan can now beam solar energy from space

The energy would be safely beamed to Earth in the form of microwaves, which would then be transformed into electricity on planet. “The energy density you are transmitting is no more than what you get by standing outside in the sun or using your cell phone,” says Hajimiri. “It can’t induce chemical change, it can just generate slight heating.” Solar panels constantly re-positioned in space where needed would also be more effective harvesters of solar energy than terrestrial plants. “You look at the seasons, day-night cycle and all of that versus having it in space at geostationary orbit (an orbit where a satellite appears to hover over one spot on Earth’s surface), and there is an advantage in space, factor of 9,” says Caltech researcher Sergio Pellegrino.

The Caltech team acknowledges the challenges ahead of them, but are confident that their pursuit is a worthy and achievable goal. “We’re talking about building a new industry, to be sure,” says Caltech researcher Harry Atwater. “But it’s not a pipe dream.”

Via Phys.org

Images via Phys.org and Wikimedia





Ian Cash

unread,
Apr 13, 2016, 7:00:17 AM4/13/16
to power-satellite-economics

Forward to group

---------- Original Message ----------
From: Ian Cash <i...@sicadesign.co.uk>
To: Darel Preble <darel....@comcast.net>
Cc: "M.V. Coyote Smith" <coyot...@gmail.com>, sbs...@googlegroups.com, Paul Jaffe <paul....@nrl.navy.mil>, Peter Garretson <budo...@gmail.com>
Date: 13 April 2016 at 11:40
Subject: Re: Space Solar Power - new CalTech design details

Darel,

This is similar in many respects to the highly modular (down to half-wavelength scale) HESPeruS concept I've researched, where the whole system is essentially a 2D structure.

What puzzles me is this:

"In orbit, a solar panel can bathe in cloud-free, high-noon sunshine 24/7 with the potential of harnessing energy wavelengths that the atmosphere absorbs before they can reach ground-based panels.

“You look at the seasons, day-night cycle and all of that versus having it in space at geostationary orbit (an orbit where a satellite appears to hover over one spot on Earth’s surface), and there is an advantage in space, factor of 9,” Pellegrino said."

How do the tiles remain both sun-pointing 24 hours per day (without self-shadowing or cosine loss) and Earth pointing (within the limits of phased array beam steering) from a geostationary orbit? No rotating reflectors are shown.

The HESPeruS concept achieves this by a shallow tiered arrangement of the solar collector (the whole structure rotates once-per-year to remain sun pointing) which introduces a 90 degree angle with the phased array bore-sight. But the microwave beam angle constraints are only satisfied from an inclined, highly elliptical (Molniya) orbit serving latitudes above 45 degrees.

Can anyone shed any light on this?

Ian

 

--
You received this message because you are subscribed to the Google Groups "Power Satellite Economics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to power-satellite-ec...@googlegroups.com.
To post to this group, send email to power-satell...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/power-satellite-economics/570C5AF2.8050601%40comcast.net.
For more options, visit https://groups.google.com/d/optout.


 


 

Tim Cash

unread,
Apr 13, 2016, 11:11:42 AM4/13/16
to Ian Cash, power-satellite-economics
If it were my money, and some day it will be, I would go for passive gravity gradient stabilization as much as possible.  It is easy to deploy and just plain works, every time.
Second choice, active stabilization backing up passive GG stabilization, in concert with each other.
The keep It as simple as possible principle applies to SSP more so than most other systems ever conceived.  If there is a way to reduce cost, reduce risk, passively, then that decision should be a "no-brainer".


For more options, visit https://groups.google.com/d/optout.



--
Regards,

Tim Cash
8 German Street
Annapolis, MD 21401
cash...@gmail.com
Mobile 410 507-6824
http://www.linkedin.com/in/timcash/

Ian Cash

unread,
Apr 13, 2016, 12:01:04 PM4/13/16
to Tim Cash, power-satellite-economics

Tim,

Having 2,500 free flying modules is one aspect of the SSPI design which is not in common with HESPeruS, which instead uses a rigid fractal geometry while still remaining essentially flat.

I agree GG is a simpler stabilisation method, but is not compatible with an essentially flat Earth-facing structure* - which happens to be far simpler from an in-space construction point of view, compared to the complex 3D structure (but GG-stabilised!) SPS-Alpha, for example.

I've found the SSPI patent with free pdf download:- http://www.freepatentsonline.com/y2016/0056321.html

The patent covers almost every embodiment conceivable, but gives no insight as to how the cosine loss is solved from a geostationary orbit - my guess is that it isn't solved; instead the patent also covers every conceivable orbit!

Here is a link to an IET article on HESPeruS:- http://eandt.theiet.org/magazine/2014/10/space-based-solar-power.cfm

Regards

Ian

* (nor a highly elliptical orbit)

Ian Cash

unread,
Apr 13, 2016, 4:02:20 PM4/13/16
to power-satellite-economics

Forward to group

---------- Original Message ----------
From: Ian Cash <i...@sicadesign.co.uk>
To: Tim Cash <cash...@gmail.com>
Date: 13 April 2016 at 21:00
Subject: Re: Re: Space Solar Power - new CalTech design details

Hi Tim

It's just geometry, but a fundamental issue:

If it's a flat plate facing Earth, that's fine for the microwave phased array antenna side throughout the geostationary orbit.

At 12 noon, the reverse side (with the pv) directly faces the sun - great!

At 8am and 4pm, the panel is angled at 60 degrees to the sun, so intercepts cos(60) = 0.5 power - hence the "cosine loss".

At 6pm through to 6am, it's edge-on or facing away from the sun and will receive zero power.

You could re-arrange this so that the pv always faces the sun, but then at 6am / 6pm the microwave beam is emitted from the edge of the panel - which would give very poor focussing ability in the east-west direction (the effective transmitter aperture is much reduced). In any case, phased arrays (i.e. electronically steered beam) have a limited steering angle away from the perpendicular "bore-sight" angle - approaching these limits causes grating effects where significant beam power is transmitted away from the target.

All GEO/GSO satellites get round this issue by having the solar pv "paddles" on a north-south axis, which rotate to track the sun. Early SP satellite concepts scaled up this idea, but passing GW power levels over rotating joints then becomes a serious engineering problem and a single point of failure.

Later, "sandwich panel" designs are basically back to the flat plate (microwave side to Earth, pv on reverse), but with huge (concentrating) mirrors tracking the sun (rotating on a north-south axis), reflecting power onto the pv. There is still the single-point failure issue, but no power bottle-neck. JAXA have proposed free-flying reflectors, but these would require continuous station keeping thrusters to maintain their position above/below the orbital plane.

Mankins' SPS-Alpha is a highly modular design where the reflectors are split into multiple smaller heliostats, eliminating the single points of failure, but resulting in a complex (but beautiful) 3D structure (see Google images).

For SSPI, like HESperuS, the entire functionality is essentially embodied within a single 10cm tile or 25mm element. The difference is that the Hesperus bore-sight is 90 degrees to the sun-facing pv, allowing the phased array to steer the beam (using retrodirective targeting) to any location above 45 degrees latitude, around the orbit apogee.

The pics show

1) A computed beam intensity pattern (and polar plot) for a very small section of the phased array over the full steering angle range. Note the grating pattern side-lobes, which diminish as the array size grows.

2) How a 25mm scale element is integrated into non-shadowing tiers, a 5 metre scale module, and a sub-scale satellite.

3) How the Molniya orbit enables a solid-state SPS to achieve 75% utilisation for northern latitude rectenna sites.

The simplified 2D structure without km-scale struts allows the mass to be reduced. The SSPI info suggests a 3x3km SPS, intercepting 12GW of sunlight (~2GW electrical, 1GW to grid) would mass under 1000 tonnes - one-tenth that of traditional designs.

That's a BIG deal, and could make the NG/CalTech concept economically viable - providing they actually have a solution for generating power between 6pm and 6am...


Regards

Ian


On 13 April 2016 at 18:53 Tim Cash <cash...@gmail.com> wrote:

I have no direct knowledge of what you speak to, however, given some modest budget, ingenious grad students that wish to secure a jab after graduation, and pay their bills, why should we assume this problem cannot be solved?  If the SSP design were somehow geared to always face the earth (Earth-centric) in some clever, low cost manner, would that not be used to solve the cosine loss problem?  I do not claim to be such a person, I simply do not see this as a show stopper.
The Greenies that do not wish the evil solar power satellite people to fry birds, kill babies,
or wreck their Internet communications and 8K movie streams are the real threat here.
You are correct to point out the loss.  It must be addressed.
Phased_Array_-50_to_50.jpg
Multi-scale_Diagram.jpg
Molniya Orbit.jpg

Keith Henson

unread,
Apr 13, 2016, 5:23:52 PM4/13/16
to power-satellite-economics
On Wed, Apr 13, 2016 at 1:02 PM, Ian Cash <i...@sicadesign.co.uk> wrote:

It's just geometry, but a fundamental issue:

Indeed.  If a power satellite has to stare at the sun with one part and at the earth with the other, something has to twist once a day.

The power joint dates back to early Boeing studies.  A later version set the PV and microwave generators in the same plane as GEO.  Sunlight was reflected to the top and microwaves came out of the bottom and bounced off a rotating reflector to the earth.

Most recent designs twist the light path.  Sunflower just focuses it into a central ring boiler, the rectangular version uses 6 mirrors that follow the sun and focus light into a set of boilers on the leading and trailing edges of the radiator planes.  The recent PPS design uses 40% efficient PV and 60 25 MW radiator tubes in a 1600 m square frame.

I don't understand how the proposed free flying design does it.  However, I tend to be rather skeptical about the kg/kW they are proposing.

Keith



 
x_toearth_comp_00154.jpg
TPS_13.jpg

KENNAN DANDAR

unread,
Apr 13, 2016, 6:19:38 PM4/13/16
to Keith Henson, power-satellite-economics

IT IS ALL SPELLED OUT IN CURRENT BOEING COMMERCIAL!  :)

--

You received this message because you are subscribed to the Google Groups "Power Satellite Economics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to power-satellite-ec...@googlegroups.com.
To post to this group, send email to power-satell...@googlegroups.com.

john strickland

unread,
Apr 13, 2016, 7:20:51 PM4/13/16
to KENNAN DANDAR, Keith Henson, power-satellite-economics
Yes, I did see a commercial today that mentioned space solar power, and I think it was Boeing's!
John

Keith Lofstrom

unread,
Apr 14, 2016, 2:33:28 AM4/14/16
to power-satellite-economics
I looked briefly at:

http://phys.org/news/2016-04-orbiting-carpets-world.html

Which contains this sentence:

"The fleet of 2,500 formation-flying spacecraft in
each solar-energy system would comprise 900 million
functionally independent tiles. The power generated
by each tile would be transmitted to Earth by the
same tile."

I haven't yet found a detailed technical paper about
this idea. CalTech's Harry Atwater is a solid state
materials researcher. Ali Hajimiri is a Caltech EE
working on high frequency integrated circuits. Sergio
Pellegrino designs lightweight deployable structures
at JPL (which is operated by CalTech for NASA).


They may lack some basic knowledge about orbital mechanics,
gravitational gradients, and formation flying. They may
not fully understand the effects of light pressure on
orbital stability. I've written about these for Server
Sky (http://server-sky.com), and made some conference
presentations, but the basic concepts (some of them from
19th century orbital mechanics) have not made it past
the gatekeepers of this space solar power community.

What are commonly called Hill's equations, or more
accurately the Clohessy-Wiltshire equations:
https://en.wikipedia.org/wiki/Clohessy-Wiltshire_equations
describe "compact" formations of orbiting objects.

Assume a formation center on the equatorial plane (easier
to describe, this works for any reference plane):

Objects following the same orbital path at the same speed
will follow each other in that path forever (with some
tiny second-order corrections from solar/lunar tides,
the lumpy shape of the rotating earth, etc.). One
dimensional formations following the same orbit sorta-
kinda stay together over days and weeks, diverge over
years and decades. Small correction thrusts, even tiny
light sail thrusts, can correct the errors.

1) Equatorial orbit formations can be static in the
east-west direction.

Objects to the north or south of that plane are in different
orbital planes, also intersecting the center of the earth;
hence, what is "north" will cross the equatorial plane
perhaps a quarter orbit later, and be south (by an equal
amount) a half orbit later. The details are also dependent
on north-south velocity. The bottom line is that it is
impossible (without a source of thrust proportional to
north-south distance) for an object to "hover" above or
below the equatorial plane, all the way around its orbit
All orbiting objects are either on an arbitrary plane,
or cross that plane twice per orbit.

2) Equatorial orbit formations *cannot* be static in
the north-south direction. They oscillate north and
south sinusoidally with the same period as the orbit.

But wait, there's more. What about radial distance, up
and down relative to the center of the earth? The math
is slightly more complicated. Ignoring the second order
stuff, orbital period is proportional to the 3/2 power of
the semi-major axis, which is half the distance between
apogee and perigee (also the radius of a circular orbit).
So a 10,000 kilometer circular orbit has a period of
roughly 10,000 seconds, while a 40,000 kilometer orbit
has a period of roughly 80,000 seconds. A 1% increase
in radius increases the period about 1.5%.

Imagine two objects in equatorial circular orbits, one
orbit with a meter larger radius than the other. If
the objects start out side by side, after one orbit the
object in the slightly higher orbit will be 9.42 meters
(3 pi) behind the object below it.

3) Equatorial orbit formations *cannot* be static in
the up-down radial direction. They diverge.

However, an ever-so-slightly elliptical equatorial orbit,
with a perigee one meter below the circle and an apogee
one meter above the circle, has the same semi-major axis,
and will have the same period. An object in that
slightly elliptical orbit will seem to follow an ellipse
drawn around an object in the circular orbit - if the
radial (up-down) distance change is 2 meters, then the
east-west distance of the "relative" ellipse is 4 meters,
and it will complete that ellipse in one orbital period.

4) Equatorial orbit formations can "orbit" an equatorial
central orbit in a small 2:1 ellipse with the same
period as the large circular orbit.

This is an approximation; very large deviations distort
the elliptical path around the central point into a kidney
bean shape, and are also strongly modified by Sun, Moon,
Jupiter, Earth shape, residual atmospheric drag, and
general relativity. Fortunately, for small deviations,
these effects can be corrected with very small thrusts;
again, light pressure is sufficient.

5) formations can /rotate/ around the central orbit
line, objects following tilted ellipses around it.
East/west up/down are coupled, north/south can be
synchronized to that, and ellipses can be "nested"
within each other, and "stacked" in the east/west
direction. What they cannot do is maintain a
stable planar relationship relative to the earth or
the sun. They will move, and eclipse each other,
and block each other's view of the earth.

You can look at some simulations of formations here:

http://server-sky.com/DisplacementAcceleration

BTW, the formations illustrated are not suitable for
space solar power transmission; they are intended to
focus 70 GHz, 1 Gbps point-to-point communication
signals at small antennas within the main focus, while
minimizing interference to other antennas nearby.
Large sparse arrays offer more opportunities to place
nulls over nearby receivers. Communication tolerates
huge inefficiencies, >100 dB, while economical grid
power transmission requires < 6 dB losses.

Bottom line: the pretty pictures of "carpets" of closely
packed tiles illustrated in the phys.org article are
incompatible with known orbital mechanics. The material
physics sounds fascinating, though, and the Caltech
authors mentioned in the article wrote some interesting
papers, so I'm glad Darel shared this with us.



Keith (L.)

--
Keith Lofstrom kei...@keithl.com

Roger Arnold

unread,
Apr 14, 2016, 4:48:42 AM4/14/16
to Keith Lofstrom, power-satellite-economics
Good explication, Keith. FWIW, I concur with everything you wrote.

- Roger


--
Keith Lofstrom          kei...@keithl.com

--
You received this message because you are subscribed to the Google Groups "Power Satellite Economics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to power-satellite-ec...@googlegroups.com.
To post to this group, send email to power-satell...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages