Ukrainian President Volodymyr Zelenskiy this week visited troops around Bakhmut to thank them for the sacrifices that the war against Russia is demanding from them. Some thanks are certainly justified. Contrary to my expectations of 40 days ago, Ukrainian forces have successfully (so far at least) held open the jaws of the trap which Russia has been trying to close around their rear from north and south of the deep salient that stretches from Slovyansk in the west to Severodonetsk in the east.

How has that been possible, in the face of Russian superiority in the air and in artillery firepower? On paper Russian forces should have overwhelmed Ukrainian defences weeks ago. Ukraine has few or no tanks and little to no air power (video reportage from the Donbas salient shows no Ukrainian tanks or aircraft in action, and no aircraft). If Ukraine had any tanks or aircraft they would struggle to find fuel. Ukrainian artillery is present, but thinly, and only slowly being supplemented with retired US howitzers (at least the guns that have survived their journey from west Ukrainian airfields to the salient). Artillery ammunition is also in short supply.

In spite of these material disadvantages Ukrainian infantry cling on to the salient front lines, and only last week were they finally pushed out of Severodonetsk itself. Russian advances are measurable in fractions of kilometres. It all feels uncomfortably like WWI trench warfare.

A closer look at the conflict, and the two sides’ agendas, reveals the likely reasons for Russia’s apparently sclerotic advance, and spoiler alert, Russian incompetence is not one of them.

On the material side the key factor is that Ukrainian forces have had many months in which to dig deep, well-constructed and well-positioned trench and bunker systems in front of Russia’s objectives. A well-sited trench can be stormed by infantry but only at great cost in lives. In 1917 and 1918 the trench systems were breached by using a revolutionary new weapon (tanks), new infantry tactics (effectively storm-trooper tactics with small combined-arms infantry units infiltrating into gaps in defensive positions), new artillery tactics (intense barrages falling metres ahead and to the flanks of the advancing infantry) and a willingness to take very high casualties. Mortality for a single trench attack ran at 10% in 1916, falling to 2% by 1918.

Artillery accuracy

Today the presence of effective front-line anti-tank guided missiles (the Javelin, the NLAW and others) makes the use of tanks to breach trench lines too dangerous, while the presence of man-portable air defence systems like Stinger (MANPADs) does the same for close air support. That leaves artillery, which Russia has in abundance (both weapons and ammunition). However, what has passed unremarked is the painful fact that even first-class artillery, well handled with good target data provided by drone flights, is not a precision weapon.

A large artillery piece firing a single shell from, say, 10 km behind the contact line is surprisingly inaccurate. Targeting is not the precise science that we have grown used to seeing in the context of air-launched precision-guided weapons. The trajectory of the shell is ballistic and unguided.

The shell’s flight through the atmosphere from gun barrel to aiming point is about 50% longer than the line-of-sight range, so a gun fired at a target 10,000 metres away gives its shell a 15,000-metre journey through an unpredictable atmosphere. That takes about 25 seconds. During its flight the shell is subject to atmospheric winds from ground level up to its apogee and back to ground level. The air through which it is passing also provides changing levels of friction according to varying temperature, density and water content. None of these are measurable to the gun-layer or to his fire control computer.

Another major variable comes from changes in the muzzle velocity of each shell. A precise fire control solution depends on an exact prediction of muzzle velocity, but velocity too depends on unpredictable inputs, including the temperature of the propellant charge, the age of the charge (propellant oxidises very slowly in storage, gradually losing its power, so old stock provides a different muzzle velocity to new stock) and even the conditions in which the propellant has been stored over time.

A larger variation in muzzle velocity comes from the level of wear in the gun barrel. As the barrel is abraded by use, more propellant gas escapes around the shell in its passage up the barrel and muzzle velocity falls. Old knackered guns (like America’s retired 155mm howitzers) are removed from service partly because their accuracy falls, but a gun starts wearing as soon as it enters service.

Finally, a gun barrel heats up during sustained use and actually starts to droop under its own weight, marginally yes, but enough to change the ballistic trajectory.

When we combine all of these sources of error the result is that a single shell will have an aiming error of about 0.1%-0.5% of its line-of-sight range. Over a 10,000-metre range that means 10-50 metres.

Some of these errors can be corrected by firing test-rounds at the target. The test rounds provide a real-time measurement for the atmospheric errors and gun barrel wear errors, which are then fed into the firing solution. However, that approach demands that the battery has real-time observation of a target beyond its line of sight, and they warn the target that a salvo may be incoming. Finally, the artillery observer must be able to connect a single shell impact with a single gun (or a salvo with a single battery) to be able to join the dots. Good co-ordination is required along with sight of the target, either from a high vantage point or from a drone.

Muzzle velocity errors can be partly corrected by measuring muzzle velocity for each round. Modern guns are equipped to do this (with a small attachment at the end of the muzzle which is visible), but older ones are not – those US 155mm howitzers again.

Using these aids, accuracy for a single shot can be raised to plus or minus a few metres for indirect fire, but this is where ground fortifications come in. Ukrainian army trenches are typically one metre across – just wide enough for two men with equipment to squeeze past each other – so to hit a target inside a trench you need a +/- 0.5m firing accuracy, or a lot of luck. Strongpoints and bunkers are slightly larger, but buried under metres of earth, steel and even concrete. To destroy a strongpoint Russia needs not only a direct hit, but also a direct hit with a large shell (small shells just move the earth around).

So what we see in the drone video of Ukrainian defence lines is hundreds of shell holes from misses, and a very small number of actual hits on those defence lines.

The same logic applies to shellfire aimed at the roads along which Ukrainian supplies and reinforcements are moving into the salient. A road itself is a larger target, but the vehicles moving along it need not only +/- 1.5m accuracy to be hit, but are also fast moving targets. A road can therefore be well within effective fire range but remain open (if dangerous – fire enough shells and you will get lucky), and that is what we are seeing in the Donbas salient – the supply roads are within 15 km of Russian artillery lines and subject to indirect fire, but have remained open.

These combined circumstances can create a situation in which infantry units stand still while batteries on each side try to hit them, and each other. In this conflict Russia has a large material advantage. Ukrainian batteries are deprived of real-time forward observation of hits (because Russian air defences can take down drones as soon as they appear), are using older, more-worn guns and stocks of ammunition, and have less of it to replace accuracy with volume.

What’s taking so long?

So if Russia is winning the artillery exchange, why has it not also taken the salient? The answer to that probably reveals Russia’s agenda. Kyiv and the western media are presenting the Ukraine war as a traditional conflict for territory, in which victory is measured by the number of square kilometres won. On those terms it is possible to present Russia as both losing and incompetent – it occupies little more Ukrainian territory now than it did 40 days ago when Mariupol fell.

Where Moscow has fought hard for ground (for example towards Popasna) the reason is probably in order to bring the main road into the salient that runs north-east from Bakhmut to Lysychansk within the 20-km effective range of Russian artillery.

But Moscow is almost certainly fighting a completely different type of war – one whose target is not territory but men. Evidence for this assertion (and it is an assertion, as President Putin has not shared his agenda with me yet) can be found in the original statement of Russia’s war aims – “de-militarisation” – and in observation of Russia’s actions in the Donbas salient.

Rather than recreate the lethal battles of WWI by throwing infantry against prepared defensive lines, Russian forces appear to be simply using intensive artillery bombardment to play the odds against Ukrainian lives. Men in trenches and strongpoints are almost impossible to kill, but men need to cycle into and out of those lines for rest and resupply, and when they do that they must move through the open. In open country a large shell has a killing radius of some 50 metres – comfortably within the aiming error of heavy guns. And it is here that Ukrainian forces are losing men at a steady and terrible rate.

This style of warfare also lends itself well to the use of Multiple Launch Rocket Systems. These have poor accuracy but compensate with saturation, payload and range. Firing from the Russian rear, where they are invulnerable to counter-battery artillery fire, they can reach deep behind Ukrainian lines to areas where troops are resting or assembling in open and unprotected areas.

The rate at which men are being killed is in dispute. Russia’s Ministry of Defence releases its estimates of the number of Ukrainian men “neutralised” almost daily. It does not define “neutralised”, which might include serious injuries as well as deaths, and of course its information may be faulty (how would it know?). For what it is worth, the daily tally rarely falls below 200 and often rises to 500. This has been the case since the fall of Mariupol three weeks ago. In contrast, Kyiv for the first time reported its own casualties at about 100 dead per day and 500 injured, a ratio consistent with past dead/injured ratios in a conflict with excellent front-line medical care, but almost certainly a significant under-estimate of the dead.

The truth probably lies somewhere in the middle of the two extremes. If the actual number is 200 dead per average day, and 400 seriously injured (with many more lightly injured but returning quickly to duty), then Ukraine’s forces in the Donbas salient will have seen some 4,000 dead and 8,000 hospitalised since the start of the Donbas phase of the war.

At the start of the assault on the Donbas salient Ukraine appears to have had two full brigades at the eastern end of the salient – the 95^th and 81^st brigades. At full strength these would have had some 10,000 men. When we estimate the number of Ukrainians captured in the salient (badly reported, but somewhere in the low thousands), we can see some numerical support for my contention that Russia’s agenda is to kill, wound or capture Ukrainians while losing the absolute minimum of its own men in the process.

The loss rates above appear to correlate with the reported movement of Ukrainian reserve forces from west Ukraine into the salient – one report cites eight brigades (some 30,000 men) moving east from Ukraine’s western districts. Reserve forces (all with past basic or professional infantry training) may stand little chance of success, or even survival, in an offensive action or in fast-moving manoeuvre warfare, but they are adequate to hold a well-prepared line of trenches and strongpoints.

If Moscow’s agenda is the demilitarisation of Ukraine then drawing men into the Donbas salient where they can be killed steadily by artillery appears to me to be a highly effective and low-cost way of serving it.

If that is clear to me, it must also be clear to Kyiv, so why are Ukrainian forces holding on to ground which, in the end, doesn’t matter that much? Some point to political reasons (while the war continues so does the flow of money and material to Kyiv), but there may be a less Machiavellian reason.

As I said earlier, it is very hard to kill a soldier who is positioned in a well-built trench or bunker. However, when that soldier moves into the open the artillery targeting problem I mapped out above disappears. If the Ukrainian army were to withdraw from the salient en masse much of it would have to do so on foot (Ukraine is short of tanks, armoured cars, trucks and fuel). Walking slowly in the open to the west would expose the army to a murderous bombardment. We have been here before.

On February 26 1991 some 10,000 soldiers of the Iraqi army tried to withdraw north from Kuwait along Highway 80, a six-lane motorway. Caught in the open by Coalition forces, a slaughter ensued, so large that no official figures for the number of dead have ever been released. Those who were not killed or wounded fled. If Ukrainian forces decided to withdraw from the Donbas salient a similar experience might await them, only in larger scale.

It is likely that Moscow is working to trigger an amorphous point of collapse in Ukrainian morale with its relentless whinnying down of the latter's forces with relentless shelling, while protecting its own forces from the same pressures by avoiding pointless attempts to gain territory. If the strategy succeeds that territory will fall into Russia’s hands anyway.

The US army estimates that on average when a unit has lost 30% of its men it can suffer from a collapse in morale and cease to be an effective fighting force. When highly motivated, as Ukraine’s soldiers are, that percentage can be higher.

Time is the unknown factor here, because it is impossible to measure the hearts and souls of Ukraine’s forces. We know that at the outset Ukraine had some 60,000 men of the first quality based in and near the Donbas salient. We can estimate that some 4,000 of those are now dead, some 8,000 are too injured to take part in the conflict, and that some thousands have surrendered or been captured. Adding in losses incurred between day one of the war and the fall of Mariupol (unquantified and probably unquantifiable at present), the original force has probably suffered some 30% total losses, and is probably losing 1% of its men per day.

That means a collapse might occur at any time between now and twenty days from now (when losses reach 50%). Replacement of degraded brigades by fresh reserve formations will delay collapse, but not man for man, since reserve forces may be expected to have collapse trigger points at much smaller mortality percentages than regular units.

Ukrainian counter-attacks (like this week’s attempt to retake part of Severodonetsk) actually play into the Russian strategy. An attack requires men to leave the safety of their defences and move in the open on ground which has been perfectly mapped by Russian artillery. As Russian forces move back Ukrainian forces move into what becomes in effect a killing-ground, where they are likely to suffer around 10% mortality in the space of an hour or two of the attack.

Collapse is contagious – once a major unit collapses it is likely to take its neighbours with it. The collapse of the Azov Brigade was not contagious to the main Ukrainian army because it was physically isolated and 150 km away from the Donbas salient, and because it took place gradually over two weeks, through the siege of Azovstal (which may be why that siege was forced on it by Kyiv).

The arrival of western weapons in the salient serves to delay collapse by a small amount. Outgoing fire (whether 122mm shells or MLRS warheads) will marginally destroy some Russian artillery units and ammunition stocks and reassure Ukrainian men in their trenches, but the margin cannot be large enough to change the dynamic. The only game-changer in this context would be a reversal in control of the airspace over the salient, and that has no chance of happening.

And so the agony of the Ukrainian army continues. Russia may succeed in closing the jaws of its trap (which would lead to complete and sudden collapse and mass surrender) or it may just continue to kill a few hundred Ukrainians each day until collapse comes, like a thief, silently and by night.