June 20, 2011

Keith Henson describes low-cost access to space and meeting future energy needs

Writer Keith Henson has extensively studied space-based solar power issues. Henson recently published a proposal to reduce the cost of getting payload to orbit by orders of magnitude . In an interview with Sander Olson, Henson describes using skylon rocket planes to release payloads at high altitudes. Concentrated lasers would then heat hydrogen in the payload, providing thrust and propelling the payload into Geosynchronous Transfer Orbit (GTO). Henson believes that this scheme could reduce launch costs to $100 per kilogram, and could make space-based solar power cost effective and feasible.

Keith Henson interview

Question: You have proposed (on Next Big Future) using a laser powered second-stage with Skylon rocket plane to launch payloads into space. How thoroughly have you explored this concept?

I have been researching cost reduction to space for the past five years. For the first quarter of getting into space, the Skylon burns air, achieving remarkably high specific impulse. Specific impulse, alternately, exhaust velocity determine the performance of the engine, and higher is better. Till it runs out of air, Skylon's equivalent exhaust is 10.5 km/s. That's 2 and 1/3 times better than the Shuttle's engines.

Question: What about using a nuclear reactor to increase exhaust velocity?

Heating hydrogen with a multi GW nuclear reactor was tried during the 1960s. It gave very high exhaust velocity but there were reactor mass and radioactivity issues.

Question: And how do lasers fit in?

In the last few years, big solid lasers have become available. It is now seems feasible to have lasers putting out gigawatts of power. Lasers heating hydrogen could result in exhaust velocities as high as 10 kilometers per second, even higher than the nuclear reactors did.

Question: Could ground-based lasers be used to transport power satellites parts directly to geosynchronous orbit (GTO)?

Yes, but to lift a ton put to GEO directly takes a gigawatt laser. Either very large lasers are required, or the payload will be small. Putting up a solar power satellite that weighs 25,000 tons a ton at a time isn't feasible. We need larger payloads.

Question: But you have found a way for a half GW ground-based laser to lift 20-ton payloads to GEO?

Yes. This approach requires putting a reflector at GEO and bouncing laser light off a set of tracking reflector mirrors. This results in a 40 to 1 improvement in payload size per GW of laser. It hasn't been done before, but Jordin Kare convinced me that it will work. The trick is to track and accelerate a 30 ton second stage for 7500 km along the equator. It would take about 500 1-megawatt lasers. In addition, you have to start most of the way to orbit.

Question: How efficient can these lasers be made?

Over 50%. The total power input for all the lasers would be about a gigawatt, which can be provided by power grid. The power grid normally has considerable surplus power available. Cooling the lasers will be necessary, but could be done by refrigerated water.

Question: Your proposal is for a Skylon rocket plane to take a 30-ton satellite to 87% orbital speed, eject the payload, and then land?

Yes, in this sub orbital mode the Skylon goes to an apogee of about 157 kilometers. It ejects the payload at 135 km on the way up then reenters the atmosphere and lands on a runway. The 30 ton second stage consists of 10 tons of hydrogen and 20 tons of structure and payload that reaches GEO. The Skylon lands 10,000 km down range.

Question: How often does the Skylon fly, how many will you need and what will they cost?

Running the expensive lasers at close to full time sets the flight rate. They take 16 minutes to accelerate the payload to geosynchronous transfer orbit and 4 minutes to circularize the payload orbit at GEO.

Three flights per hour delivers 60 tons per hour to GEO. That is 25,000 launches per year, which also gets the cost per flight down. At 72 flights a day and a two day turnaround plus spares it would take perhaps 200 Skylons It sounds like a lot, but a Skylon should not cost any more than a Boeing 777 ($250 M) and there are nearly a thousand of those in service.

Question: How much could the cost be reduced by employing this Skylon/laser combination?

At this scale of operations, the cost could be reduced to $100 per kilogram. We will need to put up half a million tons per year to GEO. The cost needs to be this low for power satellites to be cost effective, that is, less than the cost of electricity from coal. The transport charge would be $50 billion per year, even at those low per kilogram costs. For this cost per kWh, the satellites would be sold for $150 billion, making transport costs a third of the total costs.

Question: How many power satellites would be required to meet all of the earths power needs?

In order to provide for all of the earth's electricity requirements, about 5 terawatts, that is 1,000 5 Gigawatt power satellites. To meet all of the earth's energy needs, including the power required to make liquid fuels, three or four thousand power satellites will be required. Production would go well above the initial target of 100 power satellites per year.

Question: Assuming adequate funding, how long before the first prototype power satellite could be launched?

If someone had a blank check, I think that the first satellite might be constructed in as little as five years. The high capacity transport system is the key, making power satellites is straightforward. If this takes ten years to develop, it might actually be more expensive than if we accomplished the task within five years.

Question: How long before an operational power satellite paid for itself?

If a power company (or group of them) bought a power satellite and the associated rectenna for $1.6 B per GW, and sold the power to the grid for an average of 2 cents per kWh, the sale of power would pay back the purchase price in ten years.

Question: How does this compare to conventional ground-based solar power?

Ground based solar takes at least ten and more likely, twenty times that much investment per GW. Not surprising, the cost they must charge for electricity is at least ten times higher.

Question: How about energy payback time?

It requires 2-4 years for conventional solar cells to repay the energy used to make them. By contrast, a power satellite repays the energy used to make the parts and transport them to GEO within a couple of months. That is at least a 10-fold improvement over ground solar and wind.

Question: Have you collaborated with Reaction Engines LTD, the British company that is developing Skylon?

They have been very helpful. At my request, they analyzed the maximum sub-orbital capacity for Skylon. They are confident that they can release a payload traveling at almost 7 kilometers per second at an altitude of 135 kilometers. There are detailed drawing of the release timing in the Skylon User's Manual.

Question: What stands in the way of Skylon flying? (Besides money.)

The most difficult part of the Skylon is the engines. They are doing something very cleaver to get more energy out of liquid hydrogen than just burning it. It takes 20 kWh per kg to liquefy hydrogen and they are getting part of that energy back by running a helium turbine on the temperature difference of the incoming air (up to 1500 deg C) and the cold hydrogen that flows to the engines.

The impressive part is the 2 GW heat exchanger that heats the helium and cools the air to around –140 deg C. The output of the turbine compresses the cold air to rocket chamber pressure. Compressing air that cold allows them to use lightweight plastic turbine blades

The heat exchanger is made of miles of fine tubing. Reaction Engines has built substantial test sections and they work.

Question: Are you familiar with the Quicklaunch concept, which fires payloads from the sea using a hydrogen cannon? They are confident that $500 per pound costs to LEO is feasible.

That works out to about $2,000 per kilogram to reach GEO. Unfortunately, that's 20 times too expensive for a power satellite. Costs have to come down to $100 per kilogram in order for power satellites to make economic sense.

Question: What is your current assessment of the space elevator concept?

I don't think a space elevator from the earth to GEO is feasible. Even carbon nanotubes may not be strong enough and all the satellites and space junk that orbit the earth will eventually hit and destroy the cable. This is unfortunate, since the energy cost to GEO using a space elevator would be about $1.50 per kilogram. However, there is nothing in orbit around the moon, and Spectra fiber (used for dental floss) is strong enough to build a lunar elevator. Eventually a lunar space elevator, lifting payloads from the lunar surface through L1 will makes sense. Probably after hundreds of power satellites have been built though.

Question: How low can the cost go for power, once the power satellite system is established?

We are looking at producing electricity at half the cost of coal. This will displace coal, then nuclear, and eventually it will displace oil. We can make hydrocarbons by extracting CO2 from the air, which is something that we know how to do. David Keith at the University of Calgary has demonstrated machines that can pull large amounts of CO2 out of the atmosphere at low cost. Electrolytic production of hydrogen is something you can buy today. The Fisher/Tropsch process for making hydrocarbons exists today in billion dollar plants.

Question: Any other energy concepts out there?

There is a concept called StratoSolar, which calls for buoyant solar collection concentrators above most of the atmosphere, and sending the power down in gigantic plastic light pipes. It involves enormous technical challenges because of the wind, but if it works it might be a superior concept to orbiting power satellites. I worked on this for about a year.

There is another possible space transport cost reduction if it becomes feasible to power the first stage with beamed energy, possibly microwaves which are somewhat less expensive to generate than laser power. Ask about it in a few months

Question: What is the biggest game-changer that could be developed within the next few decades?

Advanced nanotechnology and AI are both game changers, and a breakthrough in one will quickly bring about a breakthrough in the other. Like Kurzweil, I see those developments happening in the 2040s. Google "The Clinic Seed" for a fictionalized account.

If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks
blog comments powered by Disqus