Huge space communications cost reduction ahead?
While there is general feeling of long term stagnation in the space, we are seeing gradual, steady increase of space based economy in the world GDP. What is more, while government investments are kept more or less steady or stagnating, commercial space revenue steadily increase and have overtaken government space revenue.
Major segment of commercial space economy is based on space communications, primarily geostationary satellites. But we are witnessing a revolution in cost for space based network access.
Classic wide beam satellites (used primarily for video) remain prohibitively expensive for most data applications, since satellite construction and launch costs have not been reduced significantly for GEO satellites. The average monthly cost per Mbps remain well over thousand USD (sample costs have been provided here.
But HTS (High Throughput Satellites) have enabled large reduction in average cost. In the following image I have calculated average cost under assumption of utilization between 50% (for LEO/MEO constellations) and 80% (for GEO satellites), and 3X ROI during designed lifetime of the system.
We can see that even 1st generation of HTS brings order of magnitude cost improvement for data applications. But upcoming constellations (Viasat-3, OneWeb, 03b mPOWER) promise another order of magnitude improvement.
So they have slowed down or even stopped any replacements of classic wide-beam transponders (mostly suitable for video applications), but have gone all in into HTS and MEO. Due to failed OneWeb/Intelsat merger, they have unique position. At the same time, MEO based HTS seems to be a sweet spot with much lower risk and much the same advantages as LEO constellations. So by 2022, their MEO network will be larger than theo GEO network, and majority of the transponders will be narrow beam for data applications. That seems strange, since currently two thirds of their revenue comes from wide-beam video. But it seems to be a shrinking market.
Any new constellation (LeoSat, Viasat MEO, Telesat, SpaceX or any other) will have to provide wholesale monthly prices of less than 10USD per Mbps just to compete with already designed and contracted systems. To get a real foothold, they will have to bring even larger price reductions and provide direct competition to terrestrial networks. Promises of failed early space constellations (Iridium, Globalstar, Teledesic) might finally become reality.
Major segment of commercial space economy is based on space communications, primarily geostationary satellites. But we are witnessing a revolution in cost for space based network access.
Classic wide beam satellites (used primarily for video) remain prohibitively expensive for most data applications, since satellite construction and launch costs have not been reduced significantly for GEO satellites. The average monthly cost per Mbps remain well over thousand USD (sample costs have been provided here.
But HTS (High Throughput Satellites) have enabled large reduction in average cost. In the following image I have calculated average cost under assumption of utilization between 50% (for LEO/MEO constellations) and 80% (for GEO satellites), and 3X ROI during designed lifetime of the system.
We can see that even 1st generation of HTS brings order of magnitude cost improvement for data applications. But upcoming constellations (Viasat-3, OneWeb, 03b mPOWER) promise another order of magnitude improvement.
The latest announced constellation (mPOWER) brings the wholesale monthly cost down to 6.2USD - in the same order of magnitude that is payed for non-dedicated DSL connections. And several times cheaper than MPLS, that has monthly costs in range of 100-300USD per Mbps. That is huge development, since it brings price of satellite data links to the same ballpark as terrestrial counterparts. No wonder that OneWeb founder Greg Wyler reported that the whole capacity of OneWeb initial constellation has already sold before they have launched a single satellite. And that they already plan to triple the capacity.
For example, SES has 51 GEO sats in orbit, and should launch three geosats per year just to keep it at the same level. Its geo fleet has average age of 11 years, and in the last 5 years they have launched seven satellites and have four more scheduled in the next six months. But after that, there is just a single geosat (SES-17) planned for 2020. At the same time, SES has decided to complete original O3b constellation and ordered next generation MEO constellation with much larger capacity that the original. Since they have planned to invest around two billion USD in CAPEX until 2021, we can assume that most of that cost will go into O3b mPOWER constellation.
So they have slowed down or even stopped any replacements of classic wide-beam transponders (mostly suitable for video applications), but have gone all in into HTS and MEO. Due to failed OneWeb/Intelsat merger, they have unique position. At the same time, MEO based HTS seems to be a sweet spot with much lower risk and much the same advantages as LEO constellations. So by 2022, their MEO network will be larger than theo GEO network, and majority of the transponders will be narrow beam for data applications. That seems strange, since currently two thirds of their revenue comes from wide-beam video. But it seems to be a shrinking market.
Any new constellation (LeoSat, Viasat MEO, Telesat, SpaceX or any other) will have to provide wholesale monthly prices of less than 10USD per Mbps just to compete with already designed and contracted systems. To get a real foothold, they will have to bring even larger price reductions and provide direct competition to terrestrial networks. Promises of failed early space constellations (Iridium, Globalstar, Teledesic) might finally become reality.
Comments
Post a Comment