Money is important. Even if we have a lot of it, there are always multiple ways to spend it : things are always competing for a part of our budget. Is a prestigious mission worth it if we could have made 50 smaller missions instead? Are we doing a one-time spending/benefit, or an investment that we can build upon for the next times?
More generally, money is also the way we take decisions. Not only money, of course, because our choices are also influenced by our ideology (that is sometimes fixed in the rules of the game by politics and law). But most of the time, money is a good mechanism to influence our choices.
During concept studies and when doing general assessments like we are going to do on this blog, we don’t have a precise idea of which equipment from which manufacturer we’ll include in the mission. In fact, we don’t know so much besides early engineering values, like broad requirements, mass, and power budgets. Sometimes, we imagine solutions that do not yet exist – how much would a lunar lander with 20 tons of payload capability cost? What if we make a single one? What if we make 50 of them?
To perform early trade-offs, we can make use of cost models. It only takes minutes instead of weeks but sacrifices accuracy. These models are built from the cost of past space missions. That causes some problems, because there have been quite some disruptions in the space industry recently, like the cost of launch decreasing 5-fold in the last decade. But, keeping that in mind, they’re still a decent tool to make estimations.
Simply put, most cost models have the same structure:
Cost = Dry Mass X × Y
Dry mass is a key factor because the bigger a system is, the more systems will be inside (computers and their software, wires, communications, thermal management systems, power production & storage, …). The X term is sometimes used to break linear scaling: maybe the difference between a 1 and a 2 ton spacecraft is bigger than between a 10 and 11 ton spacecraft. But on the other side, bigger spacecrafts are increasingly complex on interconnections, and require bigger testing facilities, specialized handling equipment… The AMCM model for instance has X < 1 to break linear scaling, while USCM is linear. I personally find linear scaling more logic. Finally, Y is the magic number: it factors everything from complexity to the number of units produced. Intuitively, we know that the Hubble Space Telescope is a more complex system, with a much higher Y factor, than a constellation of mass-produced low-complexity (relatively) Starlink satellites.
Ideally, we want to find analogies to estimate the cost of a space equipment. If you are trying to estimate the cost of an Earth observation satellite, try to look at the cost of the Sentinel program’s satellites. It’s not perfect, but it’s a fair guess to start.
|Program||Company||Dry mass (t)||Cost ($, 2018)||Units||Specific cost ($/ton)|
|Hubble Space Telescope||Lockheed||10.89||2.9B||1||268.9M|
|Sentinel 2||Astrium Germany||1.02||125.0M||1||122.9M|
|Sentinel 1 C & D||TASI||2.20||478.4M||2||108.7M|
|ISS Node 2 & 3||TASI||15.00||510.5M||2||17.0M|
You’ll notice that I specified costs in “2018” dollars. It’s important to account for inflation because projects that happened 30 years ago would look very cheap otherwise. A single $ from 1970 is worth $7.88 in 2018. Note that I also use $ even if I’m from the Euro region: it is convenient because NASA publishes yearly their New Start Inflation Index (NNSI), that is a good number to use as inflation for space projects.
So that’s it, we’ll estimate all costs with the dry mass and a magic number. It’s not very accurate, but most of the time, there is at least an order of magnitude of difference between competing solutions, so that will do.
Most of the time, I will be optimistic compared to the models, because the drop in launch prices will induce a drop in payload manufacturing prices : if missions are cheaper overall, then we can accept more trade-offs. A lot of private actors are also participating to space exploration in the recent years (mostly thanks to NASA’s policy) by searching financing from private actors. This competition coupled with smaller structures tends to reduce the costs of space things compared to what we’ve observed historically.
I hope your learnt something today. Next stop: the Moon ! 🚀