←back to thread

217 points belter | 3 comments | | HN request time: 0s | source
Show context
proee ◴[] No.41839783[source]
Cost of the Europa Clipper program is around $4-5 billion. Can anyone in the industry shed some light on why these programs are so expensive?
replies(9): >>41839826 #>>41839865 #>>41840107 #>>41840219 #>>41840290 #>>41840594 #>>41842034 #>>41844652 #>>41849478 #
hydrogen7800 ◴[] No.41839865[source]
I worked on one component of the spacecraft which was a derivative of something we've built may times before. However, the test program was entirely unique to Europa Clipper, and most of the cost was in this bespoke testing. The use of a "heritage" component served mostly to lower risk; it did not save much cost overall.
replies(3): >>41840072 #>>41840104 #>>41841076 #
mandevil ◴[] No.41840104[source]
And can introduce unusual failure cases for these bespoke missions. Mars Observer was lost in flight to Mars three decades ago, probably because of the inappropriate reuse of a satellite rocket engine. (1) The space environment out around Jupiter is really quite different from the environment that the JWST is facing around E-S L2 or what the Parker Solar Probe is facing right near the Sun. Even if the component is spec'd to handle the environment, you need to have actual educated humans (read: expensive labor) determine what those conditions will be, and then verify that the part will meet it, and that's where the money goes- to pay all of those humans.

If you built even 15 Europa Clippers the cost per-item would come down enormously (because all of those people's work could be re-used), but since the 1970's NASA has not had the budget for multiple probes per missions. So every mission is bespoke, and has to be done again completely from scratch.

1: The engine was normally used for circularizing the orbit of a geosynch comm satellite, so within a few hours of flight. For doing a Mars Insertion burn it needed to sit fueled for months in outer space, which was not appropriately tested, and probably the fuel tank exploded in flight because of that.

replies(2): >>41840341 #>>41841102 #
1. quotemstr ◴[] No.41840341[source]
Yeah. I'm surprised that we don't have a standard deep space probe bus by now in at least serial production, at least for orbiters if not landers, rovers, and such.

Each mission has unique requirements, but since payload mass costs are coming down, ISTM it should be possible to create a standard buss that meets most requirements most of the rime, even if it's heavier than a bespoke effort for any one mission.

replies(2): >>41841085 #>>41841708 #
2. accrual ◴[] No.41841085[source]
There are somewhat standard busses. Though you're correct, many use custom busses for their specific missions.

https://en.wikipedia.org/wiki/Comparison_of_satellite_buses

For example, the SSL 1300 apparently has hosted 118 satellites so far:

https://en.wikipedia.org/wiki/SSL_1300

Though maybe the distinction between "satellite" and "spacecraft" bears importance here.

3. mandevil ◴[] No.41841708[source]
For things in Earth orbit (LEO all the way up to GEO) we do have some fairly worked out buses that can keep commercial missions done on budget within the fairly well developed parameters of commercial up to GEO. Have, honestly, since the 1980's. Multiple companies from multiple countries have demonstrated this, there is a competitive market for commsats and earth observing satellites (the only two markets where business cases really close). If you are doing science from LEO you can probably re-use a lot of components from those markets.

For exploration missions and anything in deep space (basically, beyond Lunar Orbit) people have kicked around ideas for common buses, there have been plenty of proposals, but no one seems to have enough value in them to be the third or fourth user of one- everyone has found it better to start from scratch than use someone else's bus design. It is possible if there was a sustained, focused effort at one kind of project, say, something where Mars orbiter launches were guaranteed every 26 months for more than a decade, that the investment in a common bus might pay off. But as long as we are bouncing between Mars, Jupiter, Pluto/KBO's, E-S L2, and inside Mercury's orbit, it just isn't actually reusable.

Just as one point, until the past few years everything in the outer planets had to be RTG powered, which requires a totally different design than something solar. It was only with Juno (and now the Europa Clipper) that solar has been demonstrated for outer planets at all, and it is still not exactly a design you'd have off the shelf, nor would the power design you'd want for outer planets solar be at all similar to the design you'd want for inner planets solar. The same is true for comms, for thermal management, for rad-hardening, etc.