Electric Turbopump

[Lunniy Korabl]

Or... I guess just a pump? Is it possible to create an electrically driven turbopump on a rocket, so you get the high isp of a closed cycle without the additional complexity?

I guess the question boils down to whether can you get sufficient energy density out of capacitor banks or lithium polymer batteries to run a pump for a few minutes and still be lighter than the amount of fuel wasted and additional parts required for open/closed cycle turbopumps? Those turbines can be pretty heavy, but I don't have the theoretical background to understand the tradeoffs well enough.

[xenomorph555]

In my opinion the cons would outway the pros, this is most likely why it hasn't been done.

[Lunniy Korabl]

I suppose you're right. A quick google suggests kerosene has an energy density of 43 MJ/kg, while Lithium Thionyl Chloride will do you 2 MJ/kg. Now you'll need oxidiser as well, so with a mixing ratio of 3, the actual energy density of your combined LOX/RP-1 is roughly 32 MJ/kg.

So there's tankage and turbine weights not accounted for, but they'd have to be pretty heavy to make a difference when a kilo of fuel will get you 15 times more energy than a kilo of battery. I suppose the smaller the rocket the more that weight of things like turbines would count compared to fuel.

What is interesting though, is that battery technology is improving with time, whereas kerosene will always have the same properties. In 10 years it's not hard to imagine that it might be cost effective to use electric pumps, especially with stage recovery.

[K^2]

Usually, if your limiting factor is weight and complexity, you forego the pump all together, and use a pressure-fed system. Some liquid fuel SAMs are built that way, because they need to be inexpensive, light, and reliable. By the point where pressure feed isn't sufficient for your rocket, you are really worried about performance, and you have to go with a turbo pump.

[Seret]

I guess the question boils down to whether can you get sufficient energy density out of capacitor banks or lithium polymer batteries to run a pump for a few minutes

Batteries are unlikely to have the required power density, and supercaps won't have the energy density.

[Idobox]

An electric turbopump is just a spinning piece of metal on a vehicle.

That's also what a car, boat or train engine is. Electric trains make sense because they don't carry battery, and electric cars because a kWh of electricity is cheaper than a kWh of oil for tax reasons, but otherwise, chemical energy is vastly superior in terms of cost, weight and performance, and the bigger the engine, the larger the difference.

To get an idea of scale, the turbine power of the Merlin 1B is 1800 kW. To drive that for 10 minutes, you need 1080 MJ, that would be about 500kgs of battery, but only 30-35kg of RP1/LOx.

And batteries can't ever have higher energy density than liquid fuel. The energy is stored in chemical bonds for both case, but in one you release everything as heat, in the other, you try to harness the energy gradient to generate current. The best batteries around, air-metal batteries, are less dense than simply burning metal because of the mass of all the other stuff (electrolyte, electrodes, etc), and Aluminium Lox rockets are actually being studied.

[cantab]

Another factor to bear in mind is that with a turbopump you use the fuel and then toss it overboard. With an electric pump you're carrying the full mass of the batteries (or similar) for the whole time the stage is in use.

[Armchair Rocket Scientist]

Batteries are unlikely to have the required power density, and supercaps won't have the energy density.

Supercaps aren't really necessary. If you're discharging your battery in 3 minutes (the burn duration of a F9 first stage core) you're discharging at 20 C. Most RC vehicle users recommend not drawing more than half of a lithium battery's rated current to maximize its lifespan, but even a reusable launch vehicle probably wouldn't subject its batteries to hundreds of charge/discharge cycles. There are a decent number of batteries rated for 40 C and higher, and a whole bunch rated for 20 C.

The problem is energy density. Idobox said the Merlin 1B needs 1.8 MW for its turbine, so I'm guessing a Merlin 1D would need a little over 3 MW. Let's say 30 MW for an entire first stage. With a burn time of 3 minutes, this means the vehicle would need 5.4 GJ of battery capacity to run its turbopumps, which translates to 1.5 MWH.

I found a 55.5 WH battery on the internet, which weighs 0.349 kg. 1.5 MWH worth of these batteries would weigh 9.4 tons, which is over 50% heavier than all nine Merlin engines, let alone just the turbines and the fuel needed to run them. That is completely impractical.

[Lunniy Korabl]

Well, it's happening. Whole rocket will use just under one MW from batteries: http://www.forbes.com/sites/alexknapp/2015/04/14/rocket-lab-unveils-a-3d-printed-battery-powered-rocket-engine/

Edit: Obligatory hype video...

Edited April 15, 2015 by Lunniy Korabl

[Bill Phil]

Another factor to bear in mind is that with a turbopump you use the fuel and then toss it overboard. With an electric pump you're carrying the full mass of the batteries (or similar) for the whole time the stage is in use.

Not the case with RL-10 (I think...)

[-Velocity-]

I suppose you're right. A quick google suggests kerosene has an energy density of 43 MJ/kg, while Lithium Thionyl Chloride will do you 2 MJ/kg. Now you'll need oxidiser as well, so with a mixing ratio of 3, the actual energy density of your combined LOX/RP-1 is roughly 32 MJ/kg.

So there's tankage and turbine weights not accounted for, but they'd have to be pretty heavy to make a difference when a kilo of fuel will get you 15 times more energy than a kilo of battery. I suppose the smaller the rocket the more that weight of things like turbines would count compared to fuel.

What is interesting though, is that battery technology is improving with time, whereas kerosene will always have the same properties. In 10 years it's not hard to imagine that it might be cost effective to use electric pumps, especially with stage recovery.

Batteries are chemical energy storage systems, so AT BEST, they'll face the same kind of limits in energy density as other forms of chemical energy storage. I'm not a chemical engineer or chemist, but I doubt batteries will ever even approach rocket fuel or high explosives in energy density. Though, some batteries (zinc-air) already "cheat" by taking oxygen from the air to help them generate electricity- but that's cheating, and it only works in a dense oxygen atmosphere.

Also, the rate at which you drain batteries can greatly affect how much total energy you get out of them. The total energy you get out can really plummet FAST as you draw more and more current out of them. For example, you might hook some battery up to 1 ohm resistor, by the time it's dead, you only get like 1/10th of the total energy out of it than if you had hooked it to a 1 k ohm resistor and drained it fully. For pulsed-power applications, you need capacitors, not batteries. And capacitors are much lower energy density than even batteries due to dielectric break down.

Edited April 15, 2015 by |Velocity|

[cantab]

Not the case with RL-10 (I think...)

The RL-10's expander cycle still gets the energy to drive the fuel pumps from burning fuel and oxidizer that is then exhausted. It's just that instead of having a separate pre-burner it's all done in the main combustion chamber. It actually strikes me as a rather clever approach.

[Iskierka]

I believe many turbopumps are electrically-started, as that's a fairly simple way to get the fuel going and you still have umbilicals until T+0, but as many have mentioned, there's a HUGE amount of power required by the pump, far more than batteries are capable of delivering continuously or containing in reasonable mass. When fully matured, batteries will -maybe- match mass performance - the turbine has inefficiencies, and batteries are relatively high efficiency, but batteries are very limited in viable chemistries, and rocketry gets higher energy efficiency than any other chemical generator system, since they're stupidly high pressure and temperature.

At best, once batteries are perfected, they might be roughly equal to the fuel and piping requirements. But that would then ignore that the fuel gets thrown overboard, so unless there's some other super-specialised major advantage in some future rocketry application, it'll pretty much always be that turbopumps are selected for their mass and the fact that really, they're not -that- much more complex. Most of the development cost may be in the turbopump, but it won't come down that much from halving it to a very high performance pump.

[Bill Phil]

The RL-10's expander cycle still gets the energy to drive the fuel pumps from burning fuel and oxidizer that is then exhausted. It's just that instead of having a separate pre-burner it's all done in the main combustion chamber. It actually strikes me as a rather clever approach.

It actually uses the heated up propellant to run a turbine which runs the pumps, and then injects the heated up propellant into the chamber. None is thrown overboard in some designs.

[architeuthis]

Well, it's happening. Whole rocket will use just under one MW from batteries: http://www.forbes.com/sites/alexknapp/2015/04/14/rocket-lab-unveils-a-3d-printed-battery-powered-rocket-engine/

It's a cool project. This type of engine cycle was not even possible until a few years ago, but battery technology has seen some massive improvements this last decade. The first patents for this engine cycle weren't even granted till the early 2000's (to Lockheed Martin), the idea was laughed at prior to then. The thing is this cycle only makes sense for a certain very specific sliver of rockets because for very small liquid fuel rockets with low chamber pressures the weight of the gas regulators and thicker tank walls consequent of a pressure-fed cycle is still less than the weight of a set of multi-stage electric centrifugal pumps and the batteries to run them with. On the other side of the spectrum it is still unreasonable to supply the pump shaft power using batteries, so you have to use turbopump power cycles. For reference a single SSME 3-stage high-pressure fuel pump developed 71,000 hp (53 MW) at about 60% efficiency for about 500 seconds at full thrust. Good luck powering a monster like that with a battery!

Rocketlabs is building a comparatively tiny rocket, but it looks like the numbers worked out for the electric pump-cycle design. Cool!

[cantab]

It actually uses the heated up propellant to run a turbine which runs the pumps, and then injects the heated up propellant into the chamber. None is thrown overboard in some designs.

To clarify, whether it comes out through the main nozzle or not I'm still calling that "tossed overboard". The mass leaves the rocket, unlike a battery that remains on board after its energy is depleted.

[Bill Phil]

To clarify, whether it comes out through the main nozzle or not I'm still calling that "tossed overboard". The mass leaves the rocket, unlike a battery that remains on board after its energy is depleted.

I thought you were referring to the propellant going out a seperate pathway for the high power cycles. Like the gas generator cycle.

Edited April 16, 2015 by Bill Phil