Tactical Nuclear Weapons Aren’t Cost-Effective Compared to Precision Artillery
Epistemic status: there is probably a US Army analysis that does this much better.
Tactical nuclear weapons (TNWs) are often times portrayed as being significantly more effective than conventional artillery. My research indicates that they are actually less effective than precision weapons of similar cost.
These have an average deviation of 2m with GPS guidance, far lower than their kill radius. I’ve seen people wondering why Russia wasn’t using S-400 anti-air systems to intercept HIMARS strikes, with some reasoning that Ukraine was using them in conjunction with unguided rockets to make them difficult to identify. The real reason is economics. S-400 systems are something like $400 million (!!) per system (8 launchers + radar + control systems + 72 missile package). This comes to something like $5 million per interceptor launched. In contrast, HIMARS are $3.5 million per launcher + reloader system. The procurement cost for GMLRS systems, used by M270 and HIMARS, is around $150,000 per M30 rocket. M30s don’t need decoys, since any interception system is going to be more expensive than it is.
Compare the cost of an M30 to a tactical nuclear warhead. The B61-7 warhead (300-400kt) costs around $28 million to build, and a lot to maintain. It also needs a delivery system, which is going to increase the price even more. Let’s assume the most optimistic kill radius of 8km, which Nuke Map gives as the 100% 3rd degree burn distance for unprotected civilians. The kill-radius against armored vehicles is going to be far less (probably less than 2km for a ground-burst). In reality, troops don’t neatly fit themselves into an 8km circle. They tend to spread themselves into a line. The vast majority of the killing power of a nuclear weapon is wasted against a line of troops. In order to kill a tank with a tactical nuke, it needs to either be destroyed by the air blast (it’s a 60-ton tank, a detrack is probably the best you get), or literally boil off the metal to the point that the crew is exposed (the nuclear fireball for a 400kt groundburst only ha around a 800m radius for a groundburst). Metal is rather good at protecting vehicle crews from heat and radiation, and modern military vehicles are all fitted with air filters specifically for countering fallout during a nuclear war. In addition, warfare in Ukraine is centered around minimizing losses from artillery, which penalizes TNWs more than precision weapons, since an M30 is still cost-effective when targeting individual vehicles and even individual infantrymen. Once tactical nuclear weapons are deployed, troops will also change their formations to minimize their effects even further and infantry will start wearing more heat protection.
By contrast, an M30 scatters 434 submunitions, each of which can penetrate the top armor of Russian IFVs/APC. They can also sometimes kill a tank, depending on hit location and angle. Each also has a kill radius of 4m against infantry. If you wanted to kill a battle line of infantry, you can either use a tactical nuke to clear 16km of it or hit each 80m stretch with a single M30. This comes to 200 M30s (~$30 million) vs 1 tactical nuke (~$30 million). And that’s assuming pretty optimistic conditions for the nuke, with infantry in the open. A more realistic condition with dug-in infantry and armored vehicles would favor precision weapons even more. In addition, precision weapons can target high-value locations like command posts and ammo dumps individually, which raises their relative value even more.
Ultimately, TNWs aren’t going to be very effective against the front lines. As for logistical strikes, roads are made of asphalt, and high heat/airblast don’t do much to them. Railroads are made of steel, which has the same problem. Any effective logistical strikes with TNWs have to be aimed at targets like logistic hubs, railway stations, and air bases, since those are the places where a single large-radius boom can be better than multiple precision hits. But can a single nuclear weapon really do more than an equivalent cost of precision weapons (hundreds) against such a target? I doubt it. If the Russians do end up deploying tactical nuclear weapons, it will be because they’ve run out of precision cruise missiles, not because they’re the most cost-effective option.
There’s also the elephant in the room, which is that the Russians spend something like 1/10th the amount that the US does on maintaining its nuclear arsenal. This is in addition to having more warheads and 1000+ TNWs, which require more maintenance. They haven’t conducted a nuclear test since the ’96 nuclear testing ban. Given the sheer level of corruption in the Russian military, it is entirely possible that any given TNW would not initiate when deployed. A failed nuclear deployment is highly inviting for a NATO nuclear first-strike, since it shows both a willingness to escalate to a full nuclear exchange and a high likelihood of failure for their nuclear weapons. Given all this, the deployment of TNWs isn’t in Russia’s best interest.
In conclusion, 200 M30 rockets cost as much as a single tactical nuke. They are more effective at killing armor and infantry. They are precise enough to target individual vehicles and high-value targets. They are cost-effective for targeting individual infantrymen. If the Russians start deploying tactical nukes, a simple NATO response would be to massively increase their supply of precision artillery to Ukraine, which may well have more effect on the front lines. The current supplies of M30s alone have already done far more damage to the Russian military than a few nuclear weapons ever could. And they’re already being supplied in large numbers, with thousands already in the country (sources are vague on exact numbers, but intercepted Russian calls claim they’re being used to target individual APCs, which suggests the bottleneck is Ukrainian intelligence analysis, not precision-guided weapons available). If NATO gives its entire supply of 50,000+ M30s and thousands of ATACMs, they will probably end up doing more damage than all tactical nuclear weapons the Russians have at their disposal.
Author’s Notes:
I’ve had trouble finding US military analyses of the effects of nuclear weapons on the battlefield. They aren’t showing up on Google Scholar. Can you guys recommend an easy way to search for such documents?
Congratulations you have discovered the “Revolution in Military Affairs” (RMA) of the early 1990s (really the 1980s but the Gulf War was the showcase), which means that you are literal decades ahead of the rest of the analysis on this topic in this community.
For more information, take a look at the ‘AirLand battle’ concept, and another term that might be helpful is ‘precision guided munitions’.
This assumes that you have perfect knowledge of the exact location of the enemies troops. The advantage of TNWs is you don’t need to be (as) precise.
But you don’t have to have perfect knowledge of all of them at the same time. You can wait for clean shots and pick them off one by one. While in the case of a nuke, you need concentrated and exposed targets for it to do its magic. i.e. it depends on the use case. Against a carrier group? Sure. Against mobile soldiers on dirt bikes? Precision artillery seems a lot more cost effective.
Looking at the Ukraine conflict, one thing that is striking is how many drones are being used. And that’s just what is being published. NATO intelligence is also worth a lot, which seems to be used well (i.e. precision strikes on hidden command posts).
The best targets for tactical nukes are bridges. It is very difficult to destroy a bridge with conventional artillery: Antonov’s bridge still stands as well as Crimea bridge. A tactical nuke with 0.1 −1 Kt range will completely destroy a bridge.
Other possible targets are bunkers and large factories.
The Crimea bridge wasn’t hit with any artillery. https://en.wikipedia.org/wiki/Antonivka_Road_Bridge suggest that artillery is capable of temporarily inhibiting the bridge likely at much cheaper prices. Being able to disable the bridge for a month is likely enough for many military applications as the bridge is valuable after you take the territory.
One needs to be near a bridge to use artillery and this still needs to be high precision strikes with expensive guided missiles, may be 100 of them were used agains Antonov’s bridge.
You could apply the same generalized hierarchical submunition principle to cruise missiles.
A typical cruise missile (tomahawk) has a unit cost of around $2M and carries about a half ton of payload over 1000 km. A switchblade-300 drone weights roughly a kilogram and costs about $4,000 per unit, so in theory each cruise missile could carry several hundred small drone submunitions in the same weight and cost budget, or a dozen larger switchblade-600 anti-tank drones. The smaller drone is highly cost effective against light vehicles and soldiers, the larger drone is highly cost effective (more than 10x, perhaps 100x) against tanks and armored vehicles.
So in theory a single strike of perhaps a thousand or even as few as a hundred drone delivery cruise missiles could cripple the entire russian army, by taking out all the logistics and support vehicles, most of the tanks, etc.
A single C-17 transport plane can carry about 75 cruise missiles (rapid dragon program) to the border of enemy territory, but that could take half a day. A future military cargo version of starship (rocket cargo program) could—in theory—deliver 100 cruise missiles (and thus roughly 40,000 switchblade-300 drones) anywhere in the world in about an hour.
At that point you can also just deploy some of these in orbit to further reduce response time. For the hardened bunker targets (like nuke silos) you want tungsten rods rather than drones, but drones are good for most of the ground vehicles, undesirable leadership outside of bunkers, etc.
Now replace the cruise missile with similar sized rocket, likewise for the drones, deploy an orbital fleet of a few ten thousand or so (similar to starlink) and you have a strong capability to intercept icbms as they enter low earth orbit. If starship comes even within a few OOM of its promise to reduce launch costs to $10/kg that would be well below missile construction cost ($10/kg vs $1000/kg). So the total cost of an 10,000 unit icbm shield (with each unit housing dozens to hundreds of cheap submunitions) could be as low as a few tens of billions.