Space Solar Power with BFR

[alexterrell]

(This might be a topic for elsewhere, but SpaceX might be a game changer for this, to the annoyance of Musk)

Certainly, space solar power has long been dismissed, primarily due to
1. High launch costs (the original idea killer - coupled with falling oil prices)
2. A lack of space capability meaning we can't make the power sats from asteroids or lunar material.

And here is what Musk said, a few year back:

You see, that's the point I'm making. Take any given solar cell: is it better to have on Earth or in orbit? What do you get from being in orbit? You get twice as much sun, best case. But you've got do do a conversion. You've got incoming photons that go to electrons, but you have to do two conversions you don't have to do on Earth. You've got to convert it to photons and then convert those photons back into electrons.

And that double conversion is going to get you back to where you started basically! So why are you sending them to bloody space?

And by the way, electron to photon converters are not free, nor is sending stuff to space. So then it obviously super-doesn't work.

Case closed! You'd think case closed! But no, I guarantee it's gonna come up another ten times! I mean, for the love of god!

Case closed? Now, I know questioning Musk is unwise these days, but could BFR prove him wrong?

For starters, Musk badly gets his facts wrong above (reality distortion fields can be dangerous). A solar panel in space will deliver ten times as much energy per square metre as most fixed panels on Earth (maybe a steerable array in the Atacama desert could get one third of an orbital array's output - but solar PV is almost always fixed - in orientation and latitude).

And ... two big changes:
- Solar panels can be made lighter in space. How light? Thin film technology is talking about better than 1000W/kg (http://www.spacefuture.com/archive/early_comme
rcial_demonstration_of_space_solar_power_using_ultra_lightweight_arrays.shtml).
- BFR is promising an order of magnitude reductions in launch cost (and orders of magnitude increase in supply).

Let's consider a 6 GW array, delivering 4GW to Earth's grid.
Mass of solar system 6,000 tons
Total mass 10,000 tons in LEO, including fuel for electric thrusters to move sections to GEO. (Assuming the microwave transmitter is light enough)

BFR is being discussed as having a launch cost, once fully operational, of below $100/kg. That means the 10,000 ton solar station would have a launch cost of only $1 billion.

4GW on Earth - average power - could be delivered by about 30GW of solar arrays (Low latitude USA - would need 40GW in Europe). Solar cell prices are stabilising but 30GW of capacity might be installable for $30 billion. (The Saudis are talking about 200GW for$200 billion - https://www.bloomberg.com/news/articles/2018-03-28/saudi-arabia-softbank-ink-deal-on-200-billion-solar-project).

And that of course is for highly intermittent power - especially in higher latitudes where solar is not much use in winter, when electricity demand is higher.

The 6GW of solar cells in orbit will also cost. Traditionally, you'd say that they would cost a lot more than 6GW of Earth cells, but since a large part of Earth cell costs is now supporting aluminium and protective, that may not always be the case.

Let's say Space Power Sat at $1 billion launch plus $6 billion structures, plus $3 billion ground array, total $10 billion, compared to $30 billion, plus a huge amount of storage, for the Earth based solution.

Space Solar Power being baseload - it makes sense to compare it to nuclear. Hinkley C is costing $8,000/KW, or $32 billion for 4GW. OK - that one's expensive, but even with a better PWR 4GW will cost $20 billion. (Barakah in UAE is about $25 billion for 5.38GW - but would cost more in the USA or EU).

These space figures are of course optimistic. But there is a lot of leeway to rise, and still it turns out cheaper.  I'm very sceptical of the claims for BFR, but even it costs $1000/kg - that would add $9 billion: Still affordable. 

So I think the original reason why space solar power turned out too expensive will be gone within a decade. Of course, there will be plenty of other reasons why Space Solar Power might not work, but the economics could start to add up. Thanks to SpaceX.

[JBF]

What are your losses both in transmission and conversion?

[alexterrell]

My losses? None

I think the original NASA / Glaser studies assumed a fairly high DC-DC efficiency of about 80%.

This link
https://www.allaboutcircuits.com/news/wireless-power-transmission-of-solar-energy-from-space/ suggests:

The solar arrays attached to a typical satellite generating 1.6GW in space and an average of 1GW on Earth would measure about 5 to 6 square kilometers and use a transmitting antenna array with a diameter of about 1 km.

https://spectrum.ieee.org/green-tech/solar/how-japan-plans-to-build-an-orbital-solar-farm
This talks about >70% for DC to microwaves, and >80% for microwaves to DC.
(Also interesting stuff about Japanese R&D on this area. Their missing ingredient might be cheap launch costs).

Anything above 50% (DC to DC) is not a major issue if BFR is as cheap as Musk hopes.

[Elmar Moelzer]

I don't understand how you get to 10 times more in orbit than on the ground. The solar intensity at 1 AU is 1.36 kW. On the ground it is about 1 kW, depending on where you are, the weather, etc. A space based solar power station would have to be in a geostationary orbit. So the BFS would have to be refueled for getting the panels into a high enough orbit, since its payload to GEO is relatively small.

[NuclearFan]

You should take into account the development costs of building the transmission system, which needs to be about 10,000x longer range than current technology, building such an enormous satellite that can survive debris impact/puncture and point at the sun while keeping the transmitter pointed at Earth, and the massive expanding solar sheets.  Even satellites today are more expensive than their launch costs, several ton communication sats are >$100 million.  Sure, a solar panel farm will be cheaper/kg by far, but also 1000s of times heavier.

[alexterrell]

I don't understand how you get to 10 times more in orbit than on the ground. The solar intensity at 1 AU is 1.36 kW. On the ground it is about 1 kW, depending on where you are, the weather, etc.

Because capacity factors for fixed solar on the ground are between about 0.1 (Germany, England) and 0.2 (Atacama desert).

In GEO, assuming steering, it will be >0.99.

So 5 to 10 times the CF, and 1.36 times the intensity, gives 7 to 14 times the output.

A space based solar power station would have to be in a geostationary orbit. So the BFS would have to be refueled for getting the panels into a high enough orbit, since its payload to GEO is relatively small.

I would assume modules would be assembled in LEO - perhaps 100MW at a time. Then, with massive amounts of electricity available, electric propulsion would be optimum, with the highest possible Isp.

In this respect, the architecture is very much as outlined by Peter Glaser in 1968, but with BFRs instead of Shuttles.

[philw1776]

I don't understand how you get to 10 times more in orbit than on the ground. The solar intensity at 1 AU is 1.36 kW. On the ground it is about 1 kW, depending on where you are, the weather, etc. A space based solar power station would have to be in a geostationary orbit. So the BFS would have to be refueled for getting the panels into a high enough orbit, since its payload to GEO is relatively small.

This cost, and neglecting conversion losses are precisely why Musk eschews space solar.
This thread is moot given Musk's stated objections un-addressed by proponents.  Why in a SpaceX thread?
If space solar ever happens, which I doubt, it will be done via a lunar construction base.  Bezos?
It's an act of faith on my part bolstered by advances in powerful magnets containment et. al. but fusion power in the 2030s will finally kill the Rube Goldberg scheme of space solar and its ground rectennas.

[speedevil]

I don't understand how you get to 10 times more in orbit than on the ground. The solar intensity at 1 AU is 1.36 kW. On the ground it is about 1 kW, depending on where you are, the weather, etc. A space based solar power station would have to be in a geostationary orbit. So the BFS would have to be refueled for getting the panels into a high enough orbit, since its payload to GEO is relatively small.

I assume it's the weather, but 10* is too pessimistic.

Neglecting spectral effects (some of the 1.36kW is not going to be useful) and assuming that's claiming 136W/m^2 average insolation to solar panels. Or put another way, 1191kWh/year.
Multiplying by a possibly optimistic 20% loss, that comes out to 900kWh/m^2/year.


900kWh/m^2 is beaten even in most of Scotland.

In my part of Scotland, fixed panels get 1240kWh/m^2/year, 2 axis trackers 1600.
In more moderate climates (middle of France) it's around 1520kWh/m^2/1990.
And in Arizona/north-Africa like climates, 2110/2960.

In the best climates, with trackers, it's much closer to 3* than 10*.

10* is only gotten in places where people basically don't live.

[LM13]

My losses? None

I think the original NASA / Glaser studies assumed a fairly high DC-DC efficiency of about 80%.

This link
https://www.allaboutcircuits.com/news/wireless-power-transmission-of-solar-energy-from-space/ suggests:

The solar arrays attached to a typical satellite generating 1.6GW in space and an average of 1GW on Earth would measure about 5 to 6 square kilometers and use a transmitting antenna array with a diameter of about 1 km.

https://spectrum.ieee.org/green-tech/solar/how-japan-plans-to-build-an-orbital-solar-farm
This talks about >70% for DC to microwaves, and >80% for microwaves to DC.
(Also interesting stuff about Japanese R&D on this area. Their missing ingredient might be cheap launch costs).

Anything above 50% (DC to DC) is not a major issue if BFR is as cheap as Musk hopes.

I'm not an electrical engineer and so am out of my depth when discussing antenna dimensions, but is there any way to use a smaller antenna?  How does antenna diameter relate to electrical power, and to distance, and the size of the receiving rectenna? 

[Elmar Moelzer]

I assume it's the weather, but 10* is too pessimistic.

Neglecting spectral effects (some of the 1.36kW is not going to be useful) and assuming that's claiming 136W/m^2 average insolation to solar panels. Or put another way, 1191kWh/year.
Multiplying by a possibly optimistic 20% loss, that comes out to 900kWh/m^2/year.


900kWh/m^2 is beaten even in most of Scotland.

In my part of Scotland, fixed panels get 1240kWh/m^2/year, 2 axis trackers 1600.
In more moderate climates (middle of France) it's around 1520kWh/m^2/1990.
And in Arizona/north-Africa like climates, 2110/2960.

In the best climates, with trackers, it's much closer to 3* than 10*.

10* is only gotten in places where people basically don't live.

That is what I thought as well. I have seen similar math exercises and there never was a factor of 10 improvement.

[alexterrell]

This cost, and neglecting conversion losses are precisely why Musk eschews space solar.
This thread is moot given Musk's stated objections un-addressed by proponents.  Why in a SpaceX thread?
If space solar ever happens, which I doubt, it will be done via a lunar construction base.  Bezos?
It's an act of faith on my part bolstered by advances in powerful magnets containment et. al. but fusion power in the 2030s will finally kill the Rube Goldberg scheme of space solar and its ground rectennas.

Why a SpaceX thread? Because BFR will carry cargos not made by SpaceX. It might even carry cargoes that Elon Musk doesn't approve of.

Personally, I was keen on Space Solar Power as a teenager, decades ago. Then I parked it as "something for after a lunar or asteroid base".

But the point is, with BFR, "launch costs" are no longer an objection. Sure, there are plenty of other hurdles. But if it's to be built, I suspect it'll be a Japanese company buying BFR launches.

[alexterrell]

I assume it's the weather, but 10* is too pessimistic.

Neglecting spectral effects (some of the 1.36kW is not going to be useful) and assuming that's claiming 136W/m^2 average insolation to solar panels. Or put another way, 1191kWh/year.
Multiplying by a possibly optimistic 20% loss, that comes out to 900kWh/m^2/year.

900kWh/m^2 is beaten even in most of Scotland.

In my part of Scotland, fixed panels get 1240kWh/m^2/year, 2 axis trackers 1600.
In more moderate climates (middle of France) it's around 1520kWh/m^2/1990.
And in Arizona/north-Africa like climates, 2110/2960.

In the best climates, with trackers, it's much closer to 3* than 10*.

10* is only gotten in places where people basically don't live.

That is what I thought as well. I have seen similar math exercises and there never was a factor of 10 improvement.

In England, you will normally get 1000KWh per year per KW(peak) of capacity.

In GEO, you will get 365 * 24 = 8,760KWh per year per KW(peak) of capacity.

You will also get more KW(peak) per m2 in space. Maybe not 1.36 times, but certainly more than 1. 

So your solar cells will deliver at least 10 times the energy of solar cells in England. In the months of December and January, when the energy is most needed, they'll deliver about 40 to 60 times as much energy. Northern Europe or Japan would be the most likely target markets.

[whitelancer64]

(This might be a topic for elsewhere, but SpaceX might be a game changer for this, to the annoyance of Musk)

Certainly, space solar power has long been dismissed, primarily due to
1. High launch costs (the original idea killer - coupled with falling oil prices)
2. A lack of space capability meaning we can't make the power sats from asteroids or lunar material.

And here is what Musk said, a few year back:

You see, that's the point I'm making. Take any given solar cell: is it better to have on Earth or in orbit? What do you get from being in orbit? You get twice as much sun, best case. But you've got do do a conversion. You've got incoming photons that go to electrons, but you have to do two conversions you don't have to do on Earth. You've got to convert it to photons and then convert those photons back into electrons.

And that double conversion is going to get you back to where you started basically! So why are you sending them to bloody space?

And by the way, electron to photon converters are not free, nor is sending stuff to space. So then it obviously super-doesn't work.

Case closed! You'd think case closed! But no, I guarantee it's gonna come up another ten times! I mean, for the love of god!

Case closed? Now, I know questioning Musk is unwise these days, but could BFR prove him wrong?

*snip*

You've completely ignored Elon Musk's point: conversion and transmission losses. That's the number one problem with beamed space based solar power.

[alexterrell]

You've completely ignored Elon Musk's point: conversion and transmission losses. That's the number one problem with beamed space based solar power.

I specifically covered it. Conversion and transmission losses are much smaller than capacity factor loss. Which is why Space Solar power was originally proposed and studied in the 1960s.

[speedevil]

In England, you will normally get 1000KWh per year per KW(peak) of capacity.

In GEO, you will get 365 * 24 = 8,760KWh per year per KW(peak) of capacity.

You will also get more KW(peak) per m2 in space. Maybe not 1.36 times, but certainly more than 1. 

So your solar cells will deliver at least 10 times the energy of solar cells in England. In the months of December and January, when the energy is most needed, they'll deliver about 40 to 60 times as much energy. Northern Europe or Japan would be the most likely target markets.

More coffee needed - sorry.

[whitelancer64]

You've completely ignored Elon Musk's point: conversion and transmission losses. That's the number one problem with beamed space based solar power.

I specifically covered it. Conversion and transmission losses are much smaller than capacity factor loss. Which is why Space Solar power was originally proposed and studied in the 1960s.

Power loss from conversion and transmission through the atmosphere is something like 60%

That's hardly trivial, and it's the cornerstone of why beamed space-based space power is not going to happen anytime soon.

It works out to be about as effecitve to have solar panels on the ground rather than in space, without spending the billions needed to send them there.

That's the entire point Musk is making - and you do not address it at all.

[aero]

You've completely ignored Elon Musk's point: conversion and transmission losses. That's the number one problem with beamed space based solar power.

I specifically covered it. Conversion and transmission losses are much smaller than capacity factor loss. Which is why Space Solar power was originally proposed and studied in the 1960s.

Power loss from conversion and transmission through the atmosphere is something like 60%

That's hardly trivial, and it's the cornerstone of why beamed space-based space power is not going to happen anytime soon.

It works out to be about as effecitve to have solar panels on the ground rather than in space, without spending the billions needed to send them there.

That's the entire point Musk is making - and you do not address it at all.

How do those calculations work out for Mars high latitude facilities and for equatorial facilities on the Moon?

[whitelancer64]

You've completely ignored Elon Musk's point: conversion and transmission losses. That's the number one problem with beamed space based solar power.

I specifically covered it. Conversion and transmission losses are much smaller than capacity factor loss. Which is why Space Solar power was originally proposed and studied in the 1960s.

Power loss from conversion and transmission through the atmosphere is something like 60%

That's hardly trivial, and it's the cornerstone of why beamed space-based space power is not going to happen anytime soon.

It works out to be about as effecitve to have solar panels on the ground rather than in space, without spending the billions needed to send them there.

That's the entire point Musk is making - and you do not address it at all.

How do those calculations work out for Mars high latitude facilities and for equatorial facilities on the Moon?

Much better, since there's little atmosphere on Mars and none on the Moon. The problem with the Moon is there are no lunar-stationary orbits, they are unstable because that altitude is outside the Moon's hill sphere. So you'd need a series of solar arrays orbiting the Moon that beam power when in sight of the Lunar station. It makes more sense if you have a lot of locations on the Moon that could benefit from these arrays, but then that becomes a cart-horse issue.

[alexterrell]

You've completely ignored Elon Musk's point: conversion and transmission losses. That's the number one problem with beamed space based solar power.

I specifically covered it. Conversion and transmission losses are much smaller than capacity factor loss. Which is why Space Solar power was originally proposed and studied in the 1960s.

Power loss from conversion and transmission through the atmosphere is something like 60%

That's hardly trivial, and it's the cornerstone of why beamed space-based space power is not going to happen anytime soon.

It works out to be about as effecitve to have solar panels on the ground rather than in space, without spending the billions needed to send them there.

That's the entire point Musk is making - and you do not address it at all.

How do those calculations work out for Mars high latitude facilities and for equatorial facilities on the Moon?

In both instances, you have a problem of scale. This only works from about 4GW upwards on Earth.

Mars is better as a Aerostationary orbit is only about half as far as GEO, so you can scale down to 2GW.

The moon is a bit harder as you'd have to go to L1 or L2, which is actually further - I think 60,000km. On the moon, and perhaps Mars, with no/less atmosphere, lasers might be worth looking at. You drop from a DC to DC efficiency of about 60% (for microwaves) to perhaps 20% (for monochromatic lasers and PV cells), but minimum size would be MW rather than GW.

 

[alexterrell]

Power loss from conversion and transmission through the atmosphere is something like 60%

That's hardly trivial, and it's the cornerstone of why beamed space-based space power is not going to happen anytime soon.

It works out to be about as effecitve to have solar panels on the ground rather than in space, without spending the billions needed to send them there.

That's the entire point Musk is making - and you do not address it at all.

The figures I can find suggest efficiency of >70% x >80%, so the loss is about 40%. Musks point is the efficiency loss is about the same as with Earth based solar cells.

The points Musk forgets to mention are capacity factor and intermittency. Developers of solar power like to forget about these factors, but capacity factor in northern Europe is typically 11% over the year and 2.5% in December. Compared to a 60% transmission efficiency?