One aspect of military space warfare that I haven't previously covered is space-based missile defense, an omission that has become important as President Trump has floated a proposal to build just such a system as an "Iron Dome for America". It's worth taking a brief overview of the history of such plans before we get into the weeds of how such a plan would work.
Missile defense is generally divided into three phases: boost, midcourse, and terminal. Boost-phase is actions taken before the missile burns out and/or releases its warheads, while midcourse defense is targeted at warheads before they enter the atmosphere and terminal defense takes place afterwards, although the line between the two is somewhat fuzzy. Obviously, boost-phase defense is the best option if it can be arranged. A single interceptor can take out multiple warheads and the thrusting ICBM is a lot more vulnerable than a warhead. But it is also the most difficult, because there's a narrow window in which to act and people tend to object if you try to put interceptors next to their ICBM silos.
As a result, using a ground-based missile for boost-phase interception is basically impossible, although they work quite well for midcourse and terminal interception. But the appeal of boost-phase interceptions meant that even from the earliest days of the missile age, there was interest in a space-based solution. The earliest serious idea was known as Project BAMBI, and would have involved 500 satellites equipped with large mesh nets which they would deploy in the path of Soviet ICBMs. The upside was that this avoided the Nike-Zeus approach of setting off nuclear warheads over the United States. The downside was that space launch was extremely expensive, and there was no way to avoid the Soviets shooting down the satellites, so the program was shelved, and work focused on missile-based solutions for the next couple of decades.
But in 1972, the US and USSR signed a treaty drastically limiting such systems, and when Ronald Reagan, horrified at the prospect of all-out nuclear war, wanted to build missile defense a decade later, he had no option but to look at space-based solutions. The Strategic Defense Initiative, popularly known as "Star Wars", initially looked at various directed-energy systems, mostly lasers, until it became apparent that all of them were at minimum a couple orders of magnitude away from being practical for missile-defense use.1 The only remaining option was a kinetic system known as "Brilliant Pebbles". Previous systems had involved central "battlestations" that would handle detection and processing tasks, but these were obviously vulnerable to Soviet ASAT systems. Brilliant Pebbles would take advantage of improvements in computers to make each interceptor independent, greatly reducing vulnerability and hopefully cutting cost as well.
This soon ran into problems. Thousands of interceptors would be required to provide even partial coverage, as orbital mechanics meant that each would spend only about 10% of its time over the Soviet Union, and cost estimates swiftly rose from $100,000 per interceptor into the millions. The fall of the Soviet Union, hastened by the need to spend money on projects to counter SDI, just as work was starting on building production hardware meant that it was quickly killed off in favor of programs focused on theater missile defense, which ultimately laid the groundwork for the systems that have performed so well in Ukraine and Israel.
The Executive Order about the new "Iron Dome for America" explicitly references Reagan's SDI, and essentially mandates the construction of a similar system today. While I am entirely in agreement with the goal of building a national missile defense system, very little in the last few decades has moved in favor of a space-based system. The fundamental flaw, that the vast majority of interceptors will be out of position in a crisis, comes from the fundamental laws of physics. The major advantage of boost-phase interception, that a single interceptor can take out several warheads, has been reduced by arms-control treaties. In the late 80s, the Soviets had something like 6,400 RVs on about 1,400 ICBMs, while the numbers for 2022 are 306 missiles and 1,185 warheads. And the US withdrew from the ABM treaty in 2002, removing limits on ground-based interceptors.
The only real advantage that a space-based system has gained since the 80s comes from the massive reduction in launch costs we have seen thanks primarily to SpaceX, and while orbital launch costing an order of magnitude less would indeed help significantly, it only solves a tiny fraction of the problems faced by orbital missile defense. The biggest fundamental issue is that the "brilliant pebble" is basically going to be about as complicated as the ground-launched interceptor, which means that it doesn't make sense to expect the space-based version to be significantly cheaper. Yes, it will probably have greater production volumes, and it can skip the rather annoying and expensive solid booster the ground-launched interceptor needs, but we also need at least 10 of them for every one ground-based interceptor,2 and there's no way that math pencils out. And no, "we'll have someone more efficient build the thing" is not an answer.
But being in space has other, less obvious downsides. A ground-based missile will live in a nice, safe silo, where it doesn't have to deal with temperature swings or radiation or any of the other harshness that comes from the space environment. Nor is it life-limited by thruster fuel, as a space-based version inevitably would be. Even when components start expiring (far later than they would in space), it's easy to pull the missile out and send it back tot he factory for an overhaul. This happens to basically all missiles today. But none of this is possible in space, which means that the entire constellation would need to be relaunched every 5 years or so. Obviously, this could be spread out, but a system that requires paying 20% of the construction cost in maintenance every year is going to get expensive in the long run, raising the possibility that interceptor replenishment falls victim to budget cuts, something that is basically not a concern for ground-based systems.
There's also the fact that the readiness levels of space-based and ground-based missile defense are so wildly different. There are already two ground-based systems that exist today, GMD and Aegis Ashore. Both have been in operation for some years, and expanding their footprint is a matter of building more, requiring little in the way of risky development. The odds of unpleasant surprises that cause significant increases in price are low, and we can in fact expect missile costs to fall as production ramps up. Any numbers for a space-based system are much more speculative, and given the track record of such programs, it seems worth applying a very skeptical eye to them. Aegis Ashore in particular would bring a second advantage, namely that it would also provide air defense to the area it is installed in, a capability that has been entirely lacking since Nike was shut down in the 70s. This is actually another capability called for in the EO, and there appears to be discussion of installing Aegis Ashore on the East Coast to go with installations in Guam and Hawaii, but the space-based portion has gotten all the attention and seems likely to get the lion's share of the funding if the Administration gets its way.3
In conclusion, space-based missile defense has a long and interesting history, but it is a very poor way to provide protection against nuclear attack to the United States. We should instead focus on existing ground-based systems as a faster and surer way to protect against missile attack.
1 Lasers came closer to being used for missile defense in the 2000s, when the US stuffed a couple of 747s full of chemical laser intending to use them to shoot down theater ballistic missiles. But it turns out that the range limitations of lasers make even this pretty dubious, and the program was cancelled. ⇑
2 Note that for Brilliant Pebbles, it was "10% of them will be over the Soviet Union at any given time". The Soviet Union was big, and for someone with more concentrated missile forces, the ratio is going to go way, way up. ⇑
3 I suspect this is due to the influence of Space Force, a problem I predicted back in 2018. ⇑
Comments
Long-time reader, first-time commenter.
I have two quick comments. (1) Moving from ground-based interceptors to space-based kill vehicles lets you move from solid rockets sitting in silos that need to be ready to go at a moment’s notice to reusable space launch vehicles, potentially drastically lowering your overall costs. The kill vehicle is only a small part of those costs. (2) Space-based kill vehicles can intercept throughout the midcourse too, not just boost, so the math on defending against a North Korean launch is not so bad or so simple as saying that North Korea is a lot smaller than the Soviet Union, so the fraction of kill vehicles over North Korea at any one time is a lot smaller. Oversimplifying a bit, any warhead that at some point in its 6000-mile arc through space comes within engagement range of a pebble (a thousand miles or so?) is engageable. That makes each pebble effectively cover a lot more ground, and it justifies the extreme lengths the Brilliant Pebbles folks went to in developing the ClF5/hydrazine propulsion system for those things: small increases in their delta-v mean you need a lot fewer pebbles for your constellation.
I wouldn’t be averse to a large increase in the number of ground-based interceptors, especially if they moved to carrying multiple KVs. But honestly, I’d bet on space-based kill vehicles being more cost effective. Having to replace them every few years doesn’t strike me as all that bad if it encourages you to continually improve the design and allows you to keep the manufacturing line open and operating at scale.
Anon:
Assuming that re-usability actually lowers costs, how much cheaper was the Space Shuttle compared to ELVs again?
But space based kill vehicles are ammunition and therefore expendable so whilst the individual missile may be cheap you'll need a lot of them, far more than you would if ground based plus you have to launch them into space.
Anon:
Unlikely, even if it were the case most of the extra stuff (e.g. sensor platforms) is going to be useful enough that they'll be launched no matter what interceptors you use.
With unlimited resources of course you'd have both a space based system to shoot down missiles right after launch and a ground based system to shoot down any that get past the first layer, but if you don't have unlimited resources you have to prioritize and with current technology space based missile defense does indeed seem to be bad value.
Anon:
Midcourse is harder than boost but should be possible (no stealth in space, warheads likely to be pretty IR bright) but it probably doesn't help enough.
Anon:
Would that even help? More smaller completely independently targeted missiles seem like a better idea.
Anon:
Except if you don't, also if you can't get launch costs down enough.
@Anon
I don’t think this is the case. I will fully grant that modern space launch is cheaper than the motors of the ground-based interceptors. But even if I grant the assertion that we can use this for midcourse, that drops the leverage to maybe 5:1, because as pointed out, the Soviet Union was really big. I do not see the difference in costs making up for that.
I am not sure this is quite true, or at the very least, it implies a rather more complicated system with more complicated pebbles. First, you (probably) lose any leverage from taking the missile out before the MIRVs separate. Second, I suspect this drives up required pebble sophistication quite a bit, and definitely drives up the cost of supporting infrastructure. ICBMs are bright, so you just need fairly simple onboard IR homing. A warhead is much less so, particularly when you’re looking down at it against the background of Earth. It’s not impossible, but it is harder, and you also need a lot more in the way of offboard sensors to cue the pebble.
In an ideal world, I might agree with this, but I have seen enough of military procurement politics to really, really not want to rely on Congress doing the right thing to keep the system intact. Also, this adds considerably to the ground-based system’s leverage once you look over enough years, because overhauls are a lot cheaper than new pebbles.
@Anonymous
That would have been an excellent argument 15 years ago, maybe even 10, but SpaceX as it exists today is reflying boosters by the dozens. Fighting on launch costs is not particularly good ground now.
Yeah, it probably would, enough that there was at least serious discussion of that for the second-generation GBI. You get multiple kill vehicles on a single booster/silo/whatever. I don’t hate the idea, I just would rather concentrate on what we are building right now to actually get things built. (That said, I’d probably want Mk 57 VLS installed at the Aegis Ashore sites so we can go with a bigger missile diameter down the road.)
Thank you for replying. I'm not familiar with Markdown, so please forgive formatting errors.
Most of the problems with the ground-based interceptors have been with the launch vehicles. Fundamentally, it's hard for me to believe that the cost of what is essentially an ICBM is not very large compared to the cost of something, admittedly high tech, that could fit on your desk, and which is a lot less complex with fewer parts. I'd be shocked if the cost ratio was less than 20:1. There's also other parts of the system like sensors, command, and communications, all of which are expensive and more or less common across basing schemes, further decreasing the fraction of overall costs made up by the KVs.
Public descriptions of the pebbles look an awful lot like the GBI KVs with a small boost stage to give them a couple of kps to get to targets that aren't right on their orbit, albeit if those KVs were designed 40 years ago. Look at a good concept diagram of a Brilliant Pebble and one of the GBI KV. They're very similar minus the pebble having a small rocket on the back. Anyway, the KVs are not very complicated (*) - by far the most exquisite thing in there is the telescope.
An ICBM in boost is bright, sure, but it's also accelerating and in atmosphere much of the time. Tbh, I'd always assumed Brilliant Pebbles was primarily going to go after the bus, not the booster and ideally not the RVs after they've MIRVed, so how much leverage you get depends in part on how late the warheads separate from the bus.
(*) One argument that works in favor of ground-basing the interceptors is that the KVs are simple and can in principle be quite small. I believe the telescope is the part that limits how small they can get, but you're talking maybe the size of a toaster. You could launch a lot of those on a booster, particularly if you were willing to make a larger booster. And booster cost doesn't scale much with size. So, while it's practical to consider a fleet of 100 boosters each equipped with 1000 10kg KVs when the technology is better developed, it's probably not practical to consider a 100,000 KV constellation.
Strong disagree here. The task the booster has to perform is basically the same as any SAM, only with more delta-V. You can get a Standard for $2-4 million. Add another million or two for the bigger motors you'll need if you choose to churn out a lot. I'd expect the kill vehicle to be about the same cost, if not a bit higher. Sure, it's smaller, but it's also got all the expensive and difficult bits in it. We could easily have built a suitable booster for missile defense 60 years ago (and in fact did). The same definitely isn't true of the kill vehicle. Yes, obviously military costing is insane, and we don't get GMD boosters in this range, but that's mostly a function of production volume and rushing a system into service before it was really ready.
I expect these to be slightly more expensive for the space-based system, particularly if you try to go for midcourse, but will treat them as a wash for these purposes.
I agree that they're conceptually very similar, and in fact made that point in the OP. I also agree that the most complicated and expensive thing is probably the telescope. But the telescope on something like the GBI KV is designed to look for a relatively warm target against the 4 K background of space. If you want "look down, shoot down" capability against warheads (and not just a very bright ICBM) for an orbital pebble, it's probably going to make the telescope a lot more sophisticated and thus expensive.
Re boost phase, everything I can see suggests that they were actually trying to get the ICBMs before they burned out. Per Astronautix, the total burn time across the three stages of the Minuteman III is about 180 seconds, which some exceedingly rough math suggests means it will have gone about 600 km at burnout. Obviously, this number is off somewhat, but it seems pretty reasonable to assume that the last stage of an ICBM's boost will be in space long enough to allow something like Pebbles to work. (I glanced at Peacekeeper, and the numbers there are going to be broadly similar to Minuteman.) Also, the designer is going to want the boost to occur above the atmosphere as much as possible because drag scales with the square of velocity.
It's a lot more delta-v than a common SAM. The current GBI is a three-stage solid two-thirds the mass of a Minuteman III and with essentially all the same components through busing, just replacing the RV with a KV. The floor on the price of an expendable three-stage solid with perhaps 8 kps of delta-v is pretty high regardless of how much tech you poor into it. Perhaps $50 million? I think the ones they're using cost more like $90 million. SM-3 gets up to near $30 million, and we're talking about something a lot more capable than SM-3.
Whereas throwing an incremental pebble on a Falcon 9 is going to cost a great deal less.
Why is the GBI KV looking up and the pebble looking down? I think they are doing exactly the same job - the bus and RVs are going to spend the vast majority of their flights above the orbital altitude of any space-based KV. It's the same sensor problem and the same technology would be fine. I think that's half the reason people can talk about deploying such a system quickly - the tech is mostly in the bag already and they can think about deploying constellations of a thousand or so $2 million KVs without having to pay $100 million to launch each of them.
Anyway, thanks for your work here. This place is excellent.
First, I suspect the GBI is somewhat overengineered for the job it needs to do. A lot of missile defense systems have beaten their performance estimates by a fair bit. Second, the guidance system on an ICBM is an incredibly expensive piece of kit that would be a lot simpler on an interceptor. That said, we're into the realm where it's hard to get anything concrete on costs. SM-3 Blk IIA does cost around $30 million each, but that's because we're buying 12 a year. If we increased that by an order of magnitude, I don't think the cost per falling 50% is at all unreasonable. Blk IB was about $9 million each in FY23 with 47 procured, rising to $12.5 million in FY 24 with only 27 on order.
Part of my issue is that the pebble is basically still vaporware, which makes me really dubious about cost estimates. The missiles we have today are almost certain to get better if we buy more. Pebbles will get worse.
I fully grant this, and it's the only thing that keeps it even remotely viable. But I just don't think big solids are that expensive. In FY16, the replacement cost for an ICBM motor was about $17 million, which is the only number I can find even in this ballpark. That's going to fall with both smaller missiles and bigger orders.
My bad. I had forgotten how high ICBM RVs typically go. (That said, depressed trajectories are a thing.)
I think the readiness of the tech is overrated. For instance, I know for a fact that we don't have an appropriate thruster, and it would take a couple years to get one designed, built and qualified. And that also starts to push into the next administration, which greatly raises the political risk.
You are most welcome.
Space based Brilliant Pebbles may be more cost-effective than ground based interceptors against ICBMs, but as I understand it they're useless against other nuclear missile threats such as the Russian cruise/theatre ballistic missiles such as Kalibr and Kinzhal. So you're going to need the ground based system anyway, which makes the cost equation very different.
Much of the hardware, the radars and comm links, is going to be the same for both. Using a GBI to shoot down a Kalibr or similar is going to be expensive and overkill, but like using Standards to shoot down a Shahed, still cheaper than not doing so.
A GBI is useless against a Kalibr. Just not what it's for. Same for an SM-3. But there's software in Aegis to control other missiles, and the VLS will take them...
bean:
The first stage is the easiest to reuse, wait until they're reflying upper stages.
bean:
That would require that all the kill vehicles travel is roughly the same direction, a MIRV ICBM can work because the enemy places their targets close enough but if you're shooting at missiles they could be further apart (also in time).
Anon:
You don't need 8 km/s to reach 100 km, though high acceleration would probably be very nice.
bean:
Forcing your enemy to use a lower performance trajectory could be a win, even if not as much as you'd prefer.
I'm not sure what this is supposed to prove. If reusing the upper stage turns out to have unacceptable costs in cost or safety, they can always just go back to not doing that.
Obviously, this is going to be extremely sensitive to a lot of details involving the number of KVs, the delta-V each KV has and the configuration of the incoming missile attack. That said, I think it should be useful in a lot of cases, certainly enough to hopefully justify the cost.
You do need to do quite a bit more than reach 100 km. That said, I believe SM-3 Blk IIA is adequate as an anti-ICBM weapon, and it's not capable of 8 km/s.
Agreed, particularly when it's likely to involve being able to load fewer warheads. That said, I'm not going to speculate on the exact details because I haven't done anything like the necessary math.
You do if you want to defend a large area against ICBMs. You aren't just getting 100 kilometers up: (1) ICBMs are a lot higher than that in midcourse and (2) you're covering a lot of horizontal distance. The strategic ground-based interceptors are not terminal phase point-defense devices.
Regardless of how we here think ground-based interceptors should be designed, they are designed as something the size and performance of an ICBM to intercept targets far from their launch site and to defend a large area. There's a tradeoff between their performance and the number of them you need to have. Higher performance also gives you some chance at shoot-look-shoot.
Anon:
...liquid rockets, and N2H4/ClF5 liquid rockets if I read you correctly, sitting in "lifejackets" in Low Earth Orbit, that need to be ready to go at a moment's notice.
A ground-based system needs a silo, and it needs a high-thrust, high-delta-V solid rocket booster. These are well understood, they are not terribly expensive if you buy them in quantity, and there are essentially COTS models we could probably use if we wanted. The space-based system requires a high-thrust, high-delta-V rocket booster of completely new design, and a space launch system, and a way to keep it healthy for many years in the LEO environment, and paying to replace the whole thing several times over unless the "many years" in the previous sentence is equal to the service life of the ground-based system, and it needs ten times as many interceptors deployed to be able to count on having one in position when you need it.
One of these things is much more expensive than the other, and I'm pretty sure it's not the one where you're complaining about about how it needs a solid rocket booster. Really, they're both going to be using solid rocket boosters. I work with high-delta-V in-space liquid propulsion systems on a regular basis, I know what they cost and how troublesome they are even with much more benign propellants than N2H4/ClF5, and I'm pretty sure that a "Brilliant Pebble" is going to wind up attached to something that looks an awful lot like the upper stages of an SM-3.
Cheap space launch doesn't solve this. Space launch could be free, because basically everything else about the space-based system is more expensive than the ground-based version. Maybe you save a couple million dollars by leaving off the SM-3's Mk 72 first-stage booster, but that's chump change compared to everything else you need.
These numbers are way off. Yes, GBI costs $90 million each. For the entire, bespoke missile produced in very small quantities. That's not the price of the booster. SM-3 for $30 million, ditto - as bean notes, the price dropped below $10 billion when we were willing to order even a few dozen at once, and most of that is going to things other than the booster.
Looking more specifically at boosters, if you go to Lockheed-Martin, or IHI (Japan) or Avio (Italy), $40 million gets you a system that can launch a metric ton or more of payload into LEO, which is an order of magnitude more performance than you should need for a KKV, at an order size of one (1) rocket. Applying some basic scaling laws, a complete ABM-sized booster ordered a hundred at a time should come in at ~$4 million each.
The other parts of the system are more expensive than that, and they don't get cheaper when you put them in space, and you need many times more of them.
The Brilliant Pebbles design planned to use those propellants because they were trying to eke out every bit of performance they could get per gram they had to launch, because launch was expensive, and probably also because it was a cool thing to do and someone gave them permission. I assume today you'd just buy something off the shelf from L3 Harris and accept that you'll get 50 seconds less ISP.
But, for what it's worth, Aerojet was talking about that thruster costing "about $50,000 per stage in production quantities" in the early 90s. Exotic oxidizer aside, it's a small pressure fed hypergolic. It's not the hard part.
On the other hand, I don't think getting something with the kinematic performance of those boosters - we're talking about something in the range of a MMIII - for $4 million is realistic, even in quantity. It's a multi-stage large solid. Those things are complicated, more complicated than they look. If it was realistic, we'd have a lot of successful small launch startups. Sure, you can tone down the propulsion to Standard levels if you're willing to buy a lot more missiles and base them in a lot more locations, but it's not obvious that's good math or politically feasible.
Anyway, I've made my point and don't mean to get in a big argument the first time I comment here. It just looks to me like the economics are in favor of the space-based solution unless booster costs drop a lot; more than I think is plausible.
Talk is cheap. Actual Aerojet thrusters, ones you can actually buy and have delivered, I know what those actually cost. The exact figures are alas proprietary, but your numbers aren't even within an order of magnitude of correct.
I have probably posted this before, but while we are talking liquid rocket fuels, for ease of historical reference -
https://www.sciencemadness.org/library/books/ignition.pdf
Thanks. I'm quoting an Aerojet estimate. It's not my number.
To be fair, the Aerojet estimate is (1) old, and (2) assuming a couple-of-thousand production run, not a lab line. I wouldn't expect it to line up apples-to-apples with the list price for one thruster today. That wouldn't be a sensible comparison.
Anyway, if this is your issue, your problem is with Sam Wiley, not me.
This makes me wonder when someone will decide DIPOLE PYTHON is a pretty good idea and get Elon to make re-usable booster sections that boomerang back onto the launching ship.