One of the reasons missile defense has always been a controversial idea is that it is impossible to know how well the defense would perform if it is ever used in the context of a nuclear attack. This fundamental uncertainty creates a situation in which you could get away with almost any statement about missile defense. If you are in the business of building up your nuclear forces, you could easily argue that you need more warheads to penetrate defenses. On the other hand, if you are in the missile defense business, you could always say that your system will provide adequate protection against the missile threat of the day. These kind of arguments always worked well during the cold war and are still working today - on both sides of the debate.
The Soviet missile defense program provides a very interesting and important data point in this discussion. We have some idea of what the United States thought it needed to defeat the Soviet system. Thanks to the Katayev archive, we now have the numbers that describe Soviet estimates of the performance of the Moscow ABM system. One document in the archive, reproduced here and translated below, contains a brief description of the program as it stood in 1985. Other documents in the archive provide some useful context.
As the Soviet documents clearly show, the system was never expected to offer anything but a very modest intercept capability - the currently deployed A-135 system was expected to intercept no more than 1-2 ballistic missiles. The document is a bit vague on the definition of a ballistic missile in this case, but it does not seem to mean 1-2 MX-type ICBMs with 10 warheads each - it is more like 1-2 "complex ballistic targets", each of them being a single warhead surrounded by decoys and penetration aids. I guess that given that Moscow was expected to be quite densely targeted some of these "complex ballistic targets" could include a few warheads (here is where nuclear intercept would come handy). The capability provided by the A-135 predecessor, A-35, was even more modest - "a single ballistic missile from some directions." Here "single" probably meant just a single warhead.
The image shows the section of the memo that describes the Soviet missile defense programs that were underway in 1985. Here is the translation (the words and numbers in italics have been written in by hand as it is normally done in secret documents):
The work [on missile defense] has began in the mid-1960s. The TsNPO Vympel of the MRP [Ministry of Radio Industry] has developed the A-35M Moscow ABM system that has been on combat duty since 1979. The system provides a capability to intercept a single ballistic missile from some directions and up to 6 Pershing 2-type missiles from the FRG.
The work on a replacement system, an improved A-135 Moscow ABM system, will be completed in 1987 to provide protection from a strike of 1-2 modern and prospective ICBMs and up to 35 Pershing 2-type intermediate-range missiles. The A-135 system includes a new acquisition and tracking radar, Don-2N (near Pushkino-Sofrino). In accordance with the Decision of the Central Comittee and the Council of Ministers of 15 July 1985, the work has began on further modification of the Moscow ABM system - the A-235 system (intercept of 8-12 complex ballistic targets and up to 40 Pershing-2-type missiles).
The system will be presented for tests in 1995.
At the same time, development is under way of a short-range intercept system S-550 for protection of single highly valuable objects (timeline -1988) and of a 'Sambo' system for protection of ICBM silo launchers (timeline - 1988).
On the next page, not shown here, the memo says:
All development is carried out in compliance with the ABM Treaty, with the exception of the S-550 and 'Sambo' systems, whose deployment would be inconsistent with the Treaty.
Most of the work on the A-135 system was completed in time, but the tests of the system (or, rather, of its test prototype, known as Amur), conducted at the Sary-Shagan from March to October 1987, showed that it needs some serious work. The Ministry of Defense did not accepted the system for service and returned it to the industry. The next series of tests was conducted in 1989 and A-135 was accepted for "pilot joint operations" in 1990 ("joint" here means that industry representatives were operating the system together with the military). It did not begin combat duty until 1995.
The A-235 system was not an entirely new project - the work on system began in 1975. The plan that was approved in 1978 assumed that A-235 will provide defense of Moscow and "the Mosow industrial region" and will be followed by an even larger system - A-1035 - that would extend protection of "key administration and military centers". It appears that the A-235 project has been revived recently. It is unlikely, though, that the design goals - protection of the Moscow region and intercept of 1-2 missiles - have changed significantly.
There is not much information about S-550 and Sambo. As far as I can tell, S-550 was a terminal defense system with a traditional fast interceptor that would protect key command and control facilities as well as Moscow. It was probably based on the work that was done in the S-225 terminal missile defense project, which was closed in 1984 - it appears that the missile was transferred to the A-135 system (a discussion of that transfer on this blog). The 5K17 radar of the system - known as Flat Twin in the U.S. - had been moved to Kamchatka (it has been dismantled some time after 2005). The S-550 project was still active around 1989, but by all indications the work had stopped after 1991.
Sambo has been mentioned in a few documents as an "active defense of missile silos and command facilities." It appears that it was based on the idea of shooting a lot of metal rods at the incoming warhead causing it to detonate prematurely. Something like that has been discussed in the United States as well in the context of providing defense of MX silos.
The 1985 plan called for the Sambo system to be completed in 1987 and for tests too begin in 1989. However, in 1986 this name disappeared from documents and has never been mentioned again. Sambo was apparently replaced by another "active" silo protection system - "Mozyr", although it is not clear if Mozyr used metal rods as well or relied on some other mechanism (in Katayev's notes from the early 1980s, Sambo is mentioned together with another active system, "Aktiv", which was supposed to use explosives to protect the silo). There is some information that the development of Mozyr involved an intercept of a warhead in a flight test conducted in the late 1980s. MilitaryRussia.ru has some interesting photos of what appears to be a Mozyr test facility in Kamchatka.
Now that the Russian government believes it needs to spend some serious money on modernization of its strategic forces, the Moscow ABM system is clearly getting an overhaul. Russia conducts regular tests of short-range interceptors of the system and appears to be working on new version of the interceptor missile. The word is that a new missile will replace the long-range interceptors that were removed from service some time around 2006. There were reports about modernization of the Don-2N radar. Some ongoing programs are hidden behind obscure names, like "Samolyot-M" (here is a very interesting overview of some of these programs prepared by Aleksandr Stukalin).
I should note that although the system is known to rely on nuclear interceptors, all the available evidence suggests that nuclear warheads are removed from interceptor missiles are stored in a (presumably) safe location. There have been reports about conventional intercept capability, but it is unlikely that this can be achieved without a complete redesign of the system. Moreover, the Russian designers and the military are extremely skeptical about a possibility of building a reliable non-nuclear missile defense (in fact, they fully expect that the United States will convert its system to nuclear interceptors at some point having figured out that hit-to-kill doesn't quite work).