Apollo Fusion co-founders Ben Longmier, left, with an Apollo thruster, and Mike Cassidy, with a propellant tank, standing in front of a thruster test chamber. Credit: Apollo Fusion

MOUNTAIN VIEW, California — Satellite electric propulsion startup Apollo Fusion is midway through a 4,000-hour life-cycle test of its Apollo Constellation Engine, ACE, which is aimed at constellations of communications satellites where each satellite weighs between 30 and 1,000 kilograms.

The company said 2017 testing at The Aerospace Corp. confirmed ACE’s overall system performance. Several orders, from unnamed customers, have already arrived even before the life-cycle testing concludes.

Apollo has its own demonstration satellite launching this summer into low Earth orbit. Further down the road, the company says the move to smaller GEO-orbit satellites should play to its strengths, with operators using electric propulsion (EP) to climb from the polar low Earth orbit drop-off point of many launches to final geostationary-orbit position.

As is the case with many tech startups, details of Apollo’s confirmed customers are slathered with non-disclosure agreements, making it difficult to assess where ACE’s first commercial use will be.

But it’s clear the company views 2018 as its coming-out year. Co-founder Ben Longmier discussed the company’s near-term prospects.

You see the market for satellite electric propulsion expanding to smaller satellites?

Yes, people are getting to GTO transfer orbit and then circularizing with EP. You could go all-chemical and pay zillions of dollars. Lately people have been doing GTO as the initial orbit, and then they circularize the orbit with EP.

That’s for larger GEOs. But now there are smaller GEO players coming out and they want to launch a satellite with 100 kilograms of propellant, start in LEO and then boost all the way to GEO.

That way they can get away with launching a 200-kilogram satellite into LEO — those launches happen multiple times a year.

But these are usually launched into polar sun-synchronous orbit.

Yes, but that doesn’t matter. You can do the plane change once you’re up there.

What kind of propellant mass would you need?

It would be a 200-kilogram satellite launched to LEO that would use 100 kilograms of EP propellant to get to GEO. You win so much by using EP. Stationkeeping needs are really small. Over a 15-year satellite life, the stationkeeping is usually around 10% of the entire delta-V.

Where are you on commercial roll-out of your product?

This year we will be delivering to multiple customers and we are working with a half-dozen others now to determine the total number we’ll deploy and what that timeline looks like.

Your production facility here, an expanded garage — does it scale?

In volumes of around 10 thrusters per month we will build ACE ourselves.  For volumes larger than this, we will outsource to a large ISO-9000 certified contract aerospace and defense manufacturer that we have been working with.

Are your early customers demanding to see an in-orbit demonstration before confirming their orders?

No. Most people are actually OK with the lifetime tests from the chambers here, and the qualification of X number of hours — not even the full number of hours, just X number of hours — and then they are ready to buy.

I would say 40% of our customers want to see a full lifetime test and/or an orbital test.

Some people are ready to buy before the 4,000 hours because many constellation players want to be the first to orbit. If they are not the first, they’re not going to make it. Some are comfortable with our extensive thermal-vacuum chamber lifetime testing/test plans and are not requiring flight heritage.  Three customers have offered us potential ride-share flights to demonstrate flight heritage. 

What’s the status of your lifetime test?

It’s ongoing. We need 4,000 hours.

That would take you into July. And you have a launch scheduled for this year?

We will be testing a 6u satellite that will be testing a thruster and the propellant system. The thruster will be fully instrumented and we’ll be boosting. It’s kind of a funny mission. We’ll go into LEO, and fire the thruster, and we’ll use GPS telemetry and boost our orbit, and then boost it again, and we’ll be able to track that by GPS.

And we’ll be able to change our orbit significantly. No other cubesat will have had more delta-V than this one! It will be the highest thrust-to-power satellite ever! That’s going up in late summer.

You have spoken about firm orders. Can you be more specific?

The problem is that everyone’s under NDAs. We have several firm orders and a half-dozen others that are in the pipeline. We can’t be more specific than that, or about the number of units.

How long will you be playing with the satellite to launch this summer?

The whole satellite would last quite a long time but we are going to deorbit to be good citizens. The nominal mission life is two years in LEO. We are carrying enough propellant to do 750 meters per second delta-V. I’d like to go through a good chunk of that. This is a scenario where we can’t fit enough solar panels on the satellite that the thruster really needs, so we’re firing and then waiting and charging the battery, then firing again, and waiting and charging again. So its power limited, but to get through all of the propellant takes a couple of years.

We’re launching our standard thruster, with the same electronics, but with a smaller tank. The tank for this mission is about the size of a hockey puck. It fits right under the thruster. What we would offer a customer is 3.3 kilograms. Competitive flight systems today, right now, are around 20 kilograms.

What does the competitive landscape look like?

We would not consider any of the existing players as competition. We have 3x the performance of anybody. But there are other companies trying to do small thrusters.

Not a company like the startup ThrustMe from France?

Their stuff is fine for 1u or 2u cubesats. The smallest thing that we could fit on is around a 12u — around there. We are targeting the largest number of satellites that are going up and that is in the range of between 30 kilograms and 200 kilograms. But it can go down to 20 kilograms, like a 12u cubesat, and as high as 1,000 kilograms.

An Apollo Fusion thruster during test firing. Credit: Apollo Fusion

We have designed the thruster to go from 200 watts to 800 watts and we can cluster them. We have one customer that wants to cluster six thrusters and fire at 800 watts. They want the highest thruster-to-power satellite ever.

ACE is said to be 15-30 kilograms lighter than the competition. Where does that put its mass?

When you compare our entire system: thruster, PPU, tank, and feed system to all the other competitors on the market for a similar mission we are approximately 15 to 30 kilograms lighter. The mass of our entire system — thruster, PPU, tank, and feed system — is between two and five kilograms ,depending on the exact mission profile.

You make something that is awesome in what it does, and then if you need more power you can stack a few more units around it.

What systems were used for comparison to conclude that your hardware has 3x better impulse per kg and per liter?

We have compared ourselves to every propulsion system on the market that operates across a similar power range i.e. 200 W – 1000 W.

Launch costs are saved because of lower mass using EP. You use a figure of $250K in savings for owner/operator. How is this figure arrived at?

We save money because we’re very small and very light. This number is based on mass savings. It’s $15,000 per kilogram launch cost to space, and ACE saves customers 15-30kg. That makes $250,000 per satellite a very conservative estimate.

Is your product aimed mainly at the LEO constellation market, meaning satellites weighing 500 kg or less?

The majority of orders coming in do orbit in LEO but ACE can support MEO and GEO as well. ACE supports satellites from 20 kg to 1,000 kg.


What are the KPIs that enable a satellite to remain at 300-350 km for 5-7 years with ACE?

Total impulse is large for such a small propulsion system.  Other propulsion systems would be difficult or impossible to fit in small satellites. Flying at low altitudes like 300km requires thrusting against a large drag force on the satellite continuously or nearly continuously. The large total impulse of the ACE system allows stationkeeping for many years at even very low altitudes.

Would you agree that for a startup it’s best not to first go to the government as a customer, but get commercial experience first?

We totally agree. The ideal people to work with are startup like companies. They want to move fast, they are well-capitalized, and they need to beat their competitors.

What are you assuming about the number of 100-kg-satellite-plus constellations that make it to orbit?

We’re compatible with all of them, and we are not dependent on any one of them.