Dust down those orbital power plans
Sylvia Else
politics, and little to do with the environment, decided to throw $Au 10
billion into the bottomless pit that is renewable energy.
Lest it all get turned into yet more solar panels and windfarms, I
invite all comers to submit their plans for orbital power satellites. At
least then we might get some technological advance for our money, even
though I doubt we'd actually see any orbital power.
Sylvia.
Keith Henson
Please let me know more, especially who to talk to. I am hkhenson on
Skype or hkeithhenson at gmail dot com
Keith
Sylvia Else
It'll be a while. The government has made the announcement, and
supposedly has the numbers in parliament, but the legislation won't be
passed until later this year. A single by-election in the meantime could
yet lead to the whole thing unravelling.
Sylvia.
Bohica Bohica
You could make a shit load of parabolic reflectors aimed at the hot
part of a Stirling engine, these things are about 6m wide and produce
about 10Kw
A Spanish comp[any makes them. The main problem is the colour, all
shiny and not a bit og brown or green on them :-)
\
Julian
Peter Fairbrother
That's actually close to what the generating part of an orbital power
sat should be - lots of mirrors feeding sunlight to a Brayton cycle gas
turbine. Forget acres of solar cells, they are too heavy and too
expensive and too fragile.
A Brayton cycle engine in that size range is lighter than a Stirling
engine, no regenerator needed. Not as efficient, but cheaper and lighter
to launch.
-- Peter Fairbrother
Keith Henson
> Bohica Bohica wrote:
> > On Jul 12, 9:17 pm, Sylvia Else <syl...@not.here.invalid> wrote:
> >> The Australian Government has, for reasons that have much to do with
> >> politics, and little to do with the environment, decided to throw $Au
10
> >> billion into the bottomless pit that is renewable energy.
>
> >> Lest it all get turned into yet more solar panels and windfarms, I
> >> invite all comers to submit their plans for orbital power satellites.
At
> >> least then we might get some technological advance for our money, even
> >> though I doubt we'd actually see any orbital power.
>
> >> Sylvia.
>
> > You could make a shit load of parabolic reflectors aimed at the hot
> > part of a Stirling engine, these things are about 6m wide and produce
> > about 10Kw
> > A Spanish comp[any makes them. The main problem is the colour, all
> > shiny and not a bit og brown or green on them :-)
>
> That's actually close to what the generating part of an orbital power
> sat should be - lots of mirrors feeding sunlight to a Brayton cycle gas
> turbine. Forget acres of solar cells, they are too heavy and too
> expensive and too fragile.
And it would be hard to scale up to the number needed for 100 GW/year
of new construction.
> A Brayton cycle engine in that size range is lighter than a Stirling
> engine, no regenerator needed. Not as efficient, but cheaper and lighter
> to launch.
The turbines themselves are around 1/10th of a kg/kW. The
concentrating reflectors, radiators and heat absorbers seem to make up
the bulk of the satellite.
I have offered a spreadsheet before to anyone interested.
It's partly a refutation of an influential paper published
in 1962 and never revisited as far as I can tell.
What I did was very simple. In the radiation spread sheet, the first
column is absolute temperature, column B is deg C. Col C is
radiation per square meter at 0.95, D is at 0.1. E is how many square
meters per kW based on C (both sides radiate). D isn't further used.
Column E is the area to radiate on kW. F is the Carnot efficiency
from 1400 K down to the radiation temperature, G is the 75% of F based
on the typical real turbines. H is the square meters required to
collect one kW out at 100% of Carnot efficiency based on
1.366kW/meter^2. I is how much area it would take to collect sunlight
based on .75 of Carnot efficiency. I is the area it would take to
radiate heat from ideal Carnot, K is the area for real (.75) of
Carnot. L sums the areas for ideal Carnot cycle, M sums the radiator
area plus collector area at the temperature required to get rid of the
heat rejected by a real (75%) Carnot cycle.
This doesn't take into account the reflector (concentrator) loss or
the re-radiation loss from the working fluid heater, but I have
reasons to think both will be small.
Of course I could have expressed area as a function of the sum of the
two areas computed from radiation and Carnot efficiency as a function
of T and solved it analytically by setting the derivative to zero. I
find spreadsheets give me more insight though.
Assuming the radiator and collector mass per square meter is about the
same, then you can see from the graph that the minimum occurs a bit
above 100 deg C, which is far below the 370-650 deg C quoted in an old
paper here:
http://contrails.iit.edu/DigitalCollection/1961/ASDTR61-322article42.pdf
I can't say for sure what the mass per unit area of radiation or
collection are. I need to analyze a canvas tube (like an air
mattress) radiator filled with low pressure gas and air float
charcoal, Buckey balls or BeO. Assuming they are both around a
kg/m^2, a kW should come in around 3.2 kg. Turbines and generators
are around 0.1 kg/kW based on Boeing 777 engines. Transmitters have
been analyzed at less than a kg/kW. So giving room for such parts as
power conductors and the joint to the transmitter, it *might* come in
at 5kg/kW.
If anyone has some spare web space to hang a small xls file, I can
send it to you.
Keith
> -- Peter Fairbrother
Peter Fairbrother
[...]
> Assuming the radiator and collector mass per square meter is about the
> same, then you can see from the graph that the minimum occurs a bit
> above 100 deg C, which is far below the 370-650 deg C quoted in an old
> paper here:
>
> http://contrails.iit.edu/DigitalCollection/1961/ASDTR61-322article42.pdf
I'd use something like 1,000 K as Tl. High efficiency and high rate heat
radiation in space is problematic unless the temp is high. Radiative
heat dispersal is about 100 kW/m^2 for the low temp radiator.
Incident radiation on the collector is 1.1 MW/m^2, the mirror (which
weighs 0.005 kg/m^2 excluding support) concentrates sunlight from 1.33
kW/m^2 to 1.1 MW/m^2, approximately 820 times at 80% efficiency.
Th is 1800 K, Carnot efficiency is 44%, assumed overall efficiency to
local electricity is 29%.
> I can't say for sure what the mass per unit area of radiation or
> collection are. I need to analyze a canvas tube (like an air
> mattress) radiator filled with low pressure gas and air float
> charcoal, Buckey balls or BeO. Assuming they are both around a
> kg/m^2, a kW should come in around 3.2 kg.
I do not understand that. Ignoring the mirror, which I think - actually,
I don't know what you are doing -
In my example design the single sided collector has a mass of 5 kg/m^2,
the double sided radiator 1 kg/m^2.
The gas contact areas are 15 times the collecting or radiating areas.
The coefficients of convective heat transfer are 800 and 80 W/m^2 K (the
gas in the high temperature one is at twelve times the pressure of the
low temperature one). The temperature difference across each is 100 K -
the collector surface is at 1900K, the radiator surface at 900 K.
One m^2 of collector produces 400 kWe at the station, and needs 8 or 10
square meters of radiator, so 15 kg of collectors and radiators are
needed to produce 400 kWe, or 0.0375 kg/kW.
My numbers might be a little hard to achieve, though they are meant to
be only medium-tech at best, so let's be very generous and say 150 grams
per kW. That's still 20 times less.
Turbines and generators
> are around 0.1 kg/kW based on Boeing 777 engines. Transmitters have
> been analyzed at less than a kg/kW. So giving room for such parts as
> power conductors and the joint to the transmitter, it *might* come in
> at 5kg/kW.
>
> If anyone has some spare web space to hang a small xls file, I can
> send it to you.
Yes please. I seem to be missing something in your argument. Will put it
up too.
-- Peter Fairbrother
>
> Keith
>
>> -- Peter Fairbrother
>
Peter Fairbrother
> On 20/07/2011 3:35 AM, Peter Fairbrother wrote:
>> Sylvia Else wrote:
>>
>> The compressor and turbine, I don't see why not. It's only one moving
>> part, gas bearings are well-developed technology and there are no
>> critical rotating seals.
>>
>> It's much simpler than a Stirling engine, and they reckon they can make
>> those work for long periods in space.
>>
>> The generator? Yes, I'd think so too.
>>
>> The heat exchangers might need some work though, probably multiple
>> redundant circuits. leak sealing and gas refills or something. Likewise
>> the mirrors and mirror pointing stuff.
>>
>>
>> But would it need to last 20 years without maintenance? If you are going
>> to build it in the first place, you'd need a good launch capability
>> anyway.
>>
>> And if it's providing a goodly proportion of your energy, you'd want to
>> be able to fix it if it breaks. no matter what the built-in reliability
>> claimed was.
>
> I would assume that there would be enough examples in orbit to provide
> redundancy in the case of failure. After all, even if you can perform
> in-orbit maintenance, it's unlikely you can do so at the kind of short
> notice required for power supply failures that cause blackouts.
>
> If it lasts 20 years on average without intervention, then you can
> probably afford to deorbit it when it breaks, and send up a new one. If
> you're talking about in-orbit repairs you're almost certainly talking
> about manned missions, with all that that entails. One thing the shuttle
> missions have shown us is that getting up there is just the start.
> Fixing things in zero-g while wearing a space suit is not such an easy
> task.
Hmmm - suppose a 1 million square meter (just over 1.1 km across) mirror
shining on a 1000 square meter (35 meters across) collector/Brayton
engine/radiator. It collects a little over a gigawatt of solar energy,
and produces maybe 500MW of electricity. It short-range-beams that to
the downlink satellite, which beams it on to Earth, and that provides
300MW on Earth.
The downlink satellite services a flight of - what, 100 of these
powersats? - giving 30 GW electrical on Earth, equivalent to 20 large
nuclear power stations. Say you have three downlink stations which can
be fed by any of the powersats as required, and 300 powersats, that's
maybe 80 GWe on Earth, or GWeE. :)
The 300 powersats can be deorbited (or perhaps brought to LEO for
refurbishment in a shirt-sleeve atmosphere?) if they fail, so an average
life of twenty years is acceptable. I don't think that would be too hard
to achieve, though they might need a scheduled resupply of maneuvering
fuel. However as they have ample power, ion drives using very little
fuel might be okay.
I've been considering what to make the powersats from, the hard part is
the high temperature solar collector. First, operating fluid. Initial
options are hydrogen, helium, methane, water, neon and argon. Strike
helium for leakyness and neon for lack of availability. Methane, water
and hydrogen are liable to have reactivity problems over 20 years, and
we are left with argon. A little heavy, but in the amounts needed the
extra weight is lost in the noise.
Apart from that, argon is all good. First it's cheap and readily
available. It is monatomic, which is thermodynamically useful, as it
gives better efficiency. And it is almost completely inert chemically,
which means we can use it a high temperatures (with concomitant high
efficiency) without worrying too much about chemical corrosion of the parts.
On to the collector. It has an area of 1,000 square meters, and receives
1 MW/m2 of energy. That's a blackbody temperature of 2050 K. We could
make it really thin, say 0.1 kg/m2 - like a thin sheet of paper - so it
would only weigh 100 kg, but a very thin collector would be leaky, and
it would require many tiny channels. We could make it from
carbon-carbon-carbon-carbon or something, but that would be leaky and
fragile.
I suggest zirconium (or a zircalloy) at 0.8 mm thick, which would mass
about 5 tons. That won't melt even if the gas supply fails. The
collector is made from 2 large sheets of 0.4mm zircalloy, one
corrugated, which are roller-welded in strips, so as to leave channels
for the gas between welds. This provides a very low-leak solution.
The waste heat radiator could be made from something less heat resistant
and lighter, possibly titanium. It would need an area of about 6,000
square meters, but unlike the collector it could be double sided,
meaning 3,000 square meters overall. Mass, about 10 tons.
The engine could be about 10 tons - compared to a jet engine which
weighs 4 tons in aircraft form and produces 60 MWe in marinised form,
that's 15 MW/ton, the spools are simpler and do less work and there are
no combustion chambers, so 30 MW/ton should be quite possible using
available technology.
A bit of handwavium now, I'm out of time:
Add 20 tons for the alternator, 5 tons for the beam transmitter, 10 tons
for fuel, 5 for pipework etc, and 5 for the mirror, and we have a mass
of 70 tons. If the mirror support structure can come in under 30 tons,
that's 100 tons for the powersats.
300 of them is 30,000 tons in GEO. Plus you need the downlinks, say
10,000 tons each, or 60,000 tons in total. For 80GWe.
-- Peter Fairbrother
Keith Henson
Unless you know something I don't, you take a ~50% hit in each
microwave link.
> The downlink satellite services a flight of - what, 100 of these
> powersats? - giving 30 GW electrical on Earth, equivalent to 20 large
> nuclear power stations. Say you have three downlink stations which can
> be fed by any of the powersats as required, and 300 powersats, that's
> maybe 80 GWe on Earth, or GWeE. :)
>
> The 300 powersats can be deorbited (or perhaps brought to LEO for
> refurbishment in a shirt-sleeve atmosphere?)
It's a good idea. Boeing first thought about it in the 70s and even
did some great artwork of assembling power sats in LEO. Then someone
worked out what would happen to them in the months of going from LEO
to GEO on ion engines. It was not nice. Even in those days they got
hit several times with space junk. :-(
> if they fail, so an average
> life of twenty years is acceptable. I don't think that would be too hard
> to achieve, though they might need a scheduled resupply of maneuvering
> fuel. However as they have ample power, ion drives using very little
> fuel might be okay.
>
> I've been considering what to make the powersats from, the hard part is
> the high temperature solar collector. First, operating fluid. Initial
> options are hydrogen, helium, methane, water, neon and argon. Strike
> helium for leakyness and neon for lack of availability. Methane, water
> and hydrogen are liable to have reactivity problems over 20 years, and
> we are left with argon. A little heavy, but in the amounts needed the
> extra weight is lost in the noise.
>
> Apart from that, argon is all good. First it's cheap and readily
> available. It is monatomic, which is thermodynamically useful, as it
> gives better efficiency. And it is almost completely inert chemically,
> which means we can use it a high temperatures (with concomitant high
> efficiency) without worrying too much about chemical corrosion of the par
ts.
Argon is good, but it turns out that supercricial CO2 is even better.
I favor two cycles because the more heat you can convert to electric
power, the less you have to radiate.
That's kind of optimistic. 60,000 t for 80 GW is 750 t per GW. My
numbers (and Dr. Phil Chapman's) come in at around 5,000 tons per GW,
6.7 times as high. On the other hand, Solare's numbers come out a lot
lighter so you are not that far from various estimates.
Keith
> -- Peter Fairbrother
Keith Henson
> Keith Henson wrote:
>
> [...]
>
> > Assuming the radiator and collector mass per square meter is about the
> > same, then you can see from the graph that the minimum occurs a bit
> > above 100 deg C, which is far below the 370-650 deg C quoted in an old
> > paper here:
>
> >http://contrails.iit.edu/DigitalCollection/1961/ASDTR61-322article42.pdf
>
> I'd use something like 1,000 K as Tl. High efficiency and high rate heat
> radiation in space is problematic unless the temp is high. Radiative
> heat dispersal is about 100 kW/m^2 for the low temp radiator.
That's not what the minimum mass calculation show, at least for the
assumption that collector surface and radiator surface have about the
same mass per unit area. I am assuming about a kg/m^2 for both,
taking into account the supporting structure.
What you want is for the sum of mass for the collector and radiator
per kW, and taking into consideration the Carnot efficiency to be at a
minimum.
Here is the graph. http://www.htyp.org/Space_radiator
The minimum came out 130 C with not much penalty between 75 C and 200
C.
Of course, there could be an error in the spread sheet. If you can
find one, please let me know.
Keith
Peter Fairbrother
Keith Henson wrote:
> On Sat, Jul 23, 2011 at 10:41 AM, Peter Fairbrother
>> Ok, the problem is that the spreadsheet ignores the mirror.
>>
>> You have the collection and radiating areas at approximately the
>> same mass per unit area. This is wrong. The collector is very much lighter
>> than the radiator per unit area, at least 20 times lighter and
>> maybe 100 times lighter.
>>
>> You have calculated the minimum total area of collector and
>> radiator, but not the minimum mass.
>>
>>
>> Your collector is collecting at 1.33 kW/m2, and weighing 1 kg/m2,
>> but that's ridiculously heavy. For a start, the collector cannot
>> collect at 1400 K without a concentrating mirror, it's thermodynamically
>> impossible.
>>
>> The mirror is presumably thin aluminium or metallised mylar, and
>> weighs in at about 0.005 kg/m2.
>
> I agree with you *if* you can tell me how to support accurate
> pointing mirrors over km scales without structure. Virtually all of
> the collector mass is structure,
Agreed. I allowed 0.005 kg/m2 for the mirror itself, the same for the
high temperature bits, and 0.15 kg/m2 for structure.
The obvious ways to do this include gas-filled tubes, spinning a round
mirror, and a double very low pressure envelope, but I don't have a
specific design in mind.
Anyway, no matter what the structure is, it isn't going to weigh 1
kg/m2, or anything like that much.
The pointing doesn't have to be that accurate - I have the pointing
ratio [1] at 1 in 60, so a fairly easy pointing accuracy of 1 in 600
would give 90% efficiency.
[1] the distance between mirror and pickup, divided by the width of the
pickup.
>> The high temperature part of the collector can be very much smaller
>> than sunlight collecting area, and thus the overall mass of the
>> collector can be very much less than 1 kg/m2, or the mass of the radiator.
>>
>> I'd use something like 0.025 g/m2 for the collector mass, and 1
>> kg/m2 for the radiator mass.
>
>> This gives a minimum mass at about 720K, see:
>
>> http://www.zenadsl6186.zen.co.uk/minimum_mass.xls
> I *think* I can make a 1kg/m2 self sealing, radiator surface at ~130
> deg C. I don't know how at 450 C. Any ideas? Also are you counting
> the heat transfer fluid?,
Two sheets of thin alloy, about 1 meter by 2. say 0.2mm Ti alloy, that's
1.8 kg/m2, and the radiator is double sided.
High surface area on the insides for good transfer between the fluid and
the metal. The sheets are roughly roller-welded in lines at say 2cm
intervals along the 2m axis. The welds do not have to be leak-free, they
are only there to keep the sheets from moving apart under pressure
(0.4MPa). This gives a relatively low initial leakage.
There are two cutoff valves so that if punctured the section is
isolated. Larger sub-sections of the whole also have cutoffs. The cutoff
valves could be pyro, pyro melting, chemical or other things.
I am ignoring the mass of the transfer fluid, it's a couple of litres of
argon at 0.4MPa, weighing 7 grams per square meter, or 0.7% of the
radiator mass.
>> You may also notice that the total mass is now about 0.07 kg/kW,
>> rather than 1.5 kg/kW.
>
> Actually, my estimate of the total mass was 5 km/kW, but that was
> after taking a 50% transmission loss, so the power at the satellite
> including transmitter and the structure that keeps the antenna flat
> to 1/4 wave is 2.5 kg/kW.
Agreed the transmission loss to Earth is about 50%.
However trying to keep the huge main power Tx antenna flat to 1/4 wave
sounds like a .... bad ... idea.
It makes the electronics a little easier, but the penalty in structure
mass is so huge that it isn't really even worth considering.
-- Peter F
>
> You should perhaps talk to the Solaren people, they are down in that
region.
>
> I hope you are right.
>
> Keith
>
>> -- Peter Fairbrother
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> Keith Henson wrote:
>>> Here you go.
>>>
>>> Keith
>>>
>>> On Sat, Jul 23, 2011 at 6:46 AM, Peter Fairbrother
>>> <zenad...@zen.co.uk> wrote:
>>>> If you could send me ac opy of the spreadsheet please?
>>>>
>>>> -- Peter F
>>>>
>>>>
>>>> Keith Henson wrote:
>>>>> On Jul 21, 10:47 am, Peter Fairbrother
>>>>> <zenadsl6...@zen.co.uk> wrote:
>>>>>> Keith Henson wrote:
>>>>>>
>>>>>> [...]
>>>>>>
>>>>>>> Assuming the radiator and collector mass per square meter
>>>>>>> is
about the
>>>>>>> same, then you can see from the graph that the minimum
>>>>>>> occurs a bit above 100 deg C, which is far below the
>>>>>>> 370-650 deg C quoted in
an old
>>>>>>> paper here:
>>>>>>>
>>>>>>>
http://contrails.iit.edu/DigitalCollection/1961/ASDTR61-322article42.pdf
>>>>>> I'd use something like 1,000 K as Tl. High efficiency and
>>>>>> high rate heat radiation in space is problematic unless the
>>>>>> temp is high. Radiative heat dispersal is about 100 kW/m2
>>>>>> for the low temp radiator.
>>>>> That's not what the minimum mass calculation show, at least
>>>>> for the assumption that collector surface and radiator
>>>>> surface have about the same mass per unit area. I am
>>>>> assuming about a kg/m2 for both, taking into account the
Keith Henson
Gas filled tubes . . . . how do you keep the from being punctured?
Spinning round inflated mirror, same problem. Plus you need to
precess the mirror over a year to keep it pointed at the sun. Now we
are talking bearings.
> Anyway, no matter what the structure is, it isn't going to weigh 1
> kg/m2, or anything like that much.
> The pointing doesn't have to be that accurate - I have the pointing
> ratio [1] at 1 in 60, so a fairly easy pointing accuracy of 1 in 600
> would give 90% efficiency.
>
> [1] the distance between mirror and pickup, divided by the width of the
> pickup.
It's an optics problem.
> >> The high temperature part of the collector can be very much smaller
> >> than sunlight collecting area, and thus the overall mass of the
> >> collector can be very much less than 1 kg/m2, or the mass of the ra
diator.
>
> >> I'd use something like 0.025 g/m2 for the collector mass, and 1
> >> kg/m2 for the radiator mass.
>
> >> This gives a minimum mass at about 720K, see:
>
> >>http://www.zenadsl6186.zen.co.uk/minimum_mass.xls
> > I *think* I can make a 1kg/m2 self sealing, radiator surface at ~130
> > deg C. I don't know how at 450 C. Any ideas? Also are you coun
ting
> > the heat transfer fluid?,
>
> Two sheets of thin alloy, about 1 meter by 2. say 0.2mm Ti alloy, that's
> 1.8 kg/m2, and the radiator is double sided.
I am not sure you grok the scope of a power sat. For 2.45 GHz, the
smallest practical size is 5 GW on the ground, 10 GW into the
transmitter, even for 60% efficient, 16.7 GW sunlight in and 6.7 GW
waste heat. 12-13 square km of reflectors into the heat cavities.
It's worth working out the flow of heat sink fluid.
> High surface area on the insides for good transfer between the fluid and
> the metal. The sheets are roughly roller-welded in lines at say 2cm
> intervals along the 2m axis. The welds do not have to be leak-free, they
> are only there to keep the sheets from moving apart under pressure
> (0.4MPa). This gives a relatively low initial leakage.
>
> There are two cutoff valves so that if punctured the section is
> isolated. Larger sub-sections of the whole also have cutoffs. The cutoff
> valves could be pyro, pyro melting, chemical or other things.
>
> I am ignoring the mass of the transfer fluid, it's a couple of litres of
> argon at 0.4MPa, weighing 7 grams per square meter, or 0.7% of the
> radiator mass.
And how fast does the argon need to be moving to transfer the waste
heat?
Keith