The world’s largest constellation operators are plotting considerably bigger second-generation satellites, joining a trend toward more sizable spacecraft in low Earth orbit (LEO).
SpaceX operates by far the largest constellation by satellite count and mass after launching more than 4,000 Starlinks in four years. Most of the Starlinks deployed to date weighed between 200 and 300 kilograms ready for launch — already close to the 500-kilogram cut-off point for being considered a small satellite under some definitions (others set the upper limit at 1,200 kilograms).
Starlink V2 satellites coming in at around 2,000 kilograms each would easily eclipse these, pushing SpaceX even further ahead of the historic title holder for the world’s largest constellation: Earth observation operator Planet with hundreds of satellites that are mostly in the region of 4 kilograms.
SpaceX started deploying bigger V2 Mini Starlinks in February while the company waits on its heavy-lift Starship rocket to be able to launch full-scale Starlink V2s to improve broadband services. SpaceX has not released specifics about the V2 Mini but regulatory filings suggest they are around 800 kilograms.
OneWeb has 634 broadband satellites in LEO and has said its second generation will be bulkier than the 150-kilogram spacecraft currently in its fleet, which recently became the world’s second-biggest by number, at around 500 kilograms each.
Constellation operators will always prefer larger satellites for their second generation, according to Euroconsult principal adviser Maxime Puteaux, versus going smaller with the miniaturization of technology.
Even if they have higher upfront costs, larger satellites promise better performance in payload size and design life, which should make it easier to cover this expense over the network’s lifespan.
The main historical benefit of smaller satellites has been cheaper launch costs, Puteaux said, but this is becoming less relevant as launch prices come down across the board.
All the bells and whistles
Satellite operators seek to find the right balance between their costs and the useful capacity they can provide, whether through megabits per second (Mbps) of broadband or kilometer of imagery.
Compared with high throughput satellite platforms in geostationary orbit (GEO-HTS), LEO systems typically have smaller coverage areas per satellite.
This drives LEO operators to adopt flexible payload technologies to realize the latency advantage their systems have over GEO by being closer to the ground. These technologies include inter-satellite links and steerable or shapeable beams to maximize sellable capacity.
“There has been an exponential progression in hardware capacity and throughput for HTS systems,” Puteaux added, with Starlink and OneWeb outpacing most GEO-HTS solutions in terms of Mbps per kilogram.
By leveraging the benefits of miniaturization and economies of scale, these companies hope to decrease their cost per Mbps while increasing throughput per kilogram of satellite mass.
Notably, Starlink’s first generation has reached congestion in specific beams, calling for more throughput and larger spacecraft that would also help make the most of Starship’s launch capacity and diameter.
It remains challenging for LEO constellations to recoup investments over their operational lifetimes at current market prices for satellite connectivity, noted Emile de Rijk, founder and CEO of GEO satellite manufacturing startup Swissto12.
“The unit economics of broadband LEO constellations is currently significantly higher than those of broadband GEO,” he said.
To address this, LEO operators are geared toward a technical trade where more powerful and flexible payloads are put onto the next generation of broadband LEO constellations, enabling them to push out more throughput.
Optical inter-satellite also help operators optimize ground segment costs, de Rijk added, because fewer ground stations are needed when information can be beamed from one satellite to the next.
Easier the second time
The cost per Mbps is typically lower as you go to larger and larger satellites, said Armand Musey, founder of advisory firm Summit Ridge Group, but it does not make sense to launch more capacity than can be sold.
Given the positive market reception for Starlink, he said it is not surprising to see the company move toward more massive satellites to expand the market further.
Musey said satellites, like most things, are easier the second time around. Second constellations benefit from greater technical manufacturing experience, improved regulatory understanding, and an existing market with distribution in place.
He estimates manufacturing alone is a good 10-20% cheaper than the first time around, and is likely also faster and less risky.
Supported by SpaceX Falcon 9 launches, Starlink’s deployment got a headstart on OneWeb, and its success in securing subscribers also helps pave the way for others to bet on larger spacecraft.
Following multi-billion dollar contracts announced last year with multiple heavy-lift launch providers, analysts anticipate Amazon’s Project Kuiper satellites to come in above 500 kilograms once operational models start launching late this year.
Companies with deep pockets are less incentivized to reduce satellite mass to save costs.
Smaller constellations going bigger
Euroconsult expects the total number of small satellite launches will peak in the next few years and start falling near the end of the decade.
Despite ongoing technology miniaturization, Puteaux expects the quest for performance and the addition of sensors, thrusters, and hardware will result in steady growth in average small satellite mass — even among much smaller constellations.
Companies that started out specializing in small satellites, including LeoStella, Terran Orbital, and NanoAvionics, have also been gradually moving toward larger spacecraft to accommodate this demand.
LeoStella unveiled its largest satellite yet Aug. 6, called LS300, which can reach 500 kilograms when including its payload.
Falling launch costs is the main driver pushing LeoStella and other LEO manufacturing specialists toward heavier satellites, its CEO Tim Kienberger said.
With few rideshare opportunities available in the past, he said “a 500-kilogram mass to orbit used to be almost a dedicated launch,” costing anywhere from $100 million to $200 million.
Making it cost-effective to launch larger LEO satellites with more functionalities also opens up new business models for the space industry.
“Now we’re able to package more in that 500 kilograms,” he said, “which traditionally, even five years ago, would have cost three times as much — probably four times as much, and had less capability on that satellite.”
The 500-kilogram sweet spot?
Terran Orbital started off over a decade ago with 1U cubesats weighing less than 1.33 kilograms and has gradually expanded to larger spacecraft ever since.
The biggest satellite the manufacturer has under construction is 800 kilograms, Terran Orbital CEO Marc Bell said, but he sees the LEO market settling at a sweet spot at around 500 kilograms.
“It seems right now that with 500 kilograms you’re getting the biggest bang for the buck,” he said.
LeoStella agrees, for the time being. Kienberger said rocket providers could one day significantly reduce launch costs for even heavier payloads to change the economics.
“Five years ago, I would have said 300 kilograms is the sweet spot,” Kienberger said.
Euroconsult also does not expect to see a size sweet spot settle any time soon.
“I’m not convinced that 500 Kg fits all needs,” Puteaux added.
While LEO satellites are getting bigger, spacecraft on the opposite end of the spectrum in GEO have been getting smaller with the emergence of manufacturers specializing in mini-GEO offerings.
Swissto12 recently sold three dishwasher-sized satellites to Intelsat, a GEO operator more akin to ordering school bus-sized giants with greater throughput.
It’s difficult to compare trends in the size of broadband LEO satellites with those for GEO because the two are fundamentally uncorrelated, Swissto12’s de Rijk said.
He said the rationale for taking a hit on capacity by going smaller and lighter in GEO is to greatly reduce costs — by between three to 10 times in Swissto12’s case, according de Rijk— as well as reducing time to market.
This enables GEO operators and telecom providers to quickly spin up new regional services, he added, increment capacity on existing markets, and add gap-filling services that would not be economically viable using larger satellites.
At the same time, small GEO satellites give smaller nation-states a more cost-effective option for their own dedicated satellite to provide secure and sovereign connectivity services.
Although de Rijk sees a future for both GEO and non-GEOs, he pointed to how Starlink and other LEO broadband providers have yet to prove profitability with their current generation.
Improving LEO broadband economics takes time, and it remains to be seen whether all these pieces can come together the second time around.
This article originally appeared in the August 2023 issue of SpaceNews magazine.