A Plant Is Not Capacity

An inspector general evaluation published in July found that a new artillery projectile plant in Texas, built at a cost of roughly four hundred and seventy million dollars and opened in the spring of 2024, had not produced a single metal part meeting contract specification as of March of this year. The facility was designed to turn out thirty thousand shell bodies a month across three lines. The equipment had been configured for an older projectile and was being adapted to the heavier one the Army actually wanted.

Set that against the programme it belongs to. The Army spent about four and a half billion dollars on expanding artillery ammunition production between early 2022 and last autumn. The delivered rate in March was around thirty-six thousand rounds a month against a target of a hundred thousand that was supposed to have been reached the previous October.

From the Battlefield to the Balance Sheet

The useful distinction here is between capacity built, capacity qualified and units delivered, and they are three different numbers that get reported as one. A building with machines in it is capital expenditure. A line that has demonstrated it can make a part to specification, repeatably, and has been accepted by the customer, is capacity. Only the third number puts anything in a magazine.

Qualification is where these programmes actually fail, and it is the least visible part of the process. It involves first article testing, process validation, metallurgical acceptance and a great deal of iteration on tooling that was bought against a drawing rather than a proven process. None of that shows up in a groundbreaking photograph or a contract announcement, and it routinely takes longer than the construction did.

The specific failure here is instructive because it was a mismatch rather than a technology problem. Equipment suited to one projectile was pressed into service for another with different weight and length. That is a specification error made at the point of purchase, and it is the kind of error that happens when a programme is under pressure to show that money has been committed.

The bottleneck is also narrower than the headline suggests. Loading and assembly capacity is not the constraint. Shell bodies are, and there are only a handful of qualified producers of them, which is why a single plant failing to qualify moves the national rate.

The Dual-Use Reality Check

Metal parts production of this kind is ordinary heavy manufacturing. Forging, flow forming, heat treatment and machining to tolerance are the same operations that make pressure vessels, driveshafts and pipeline fittings, and the equipment comes from the same builders. There is nothing about a shell body that a competent forging shop cannot learn to make.

What it cannot do quickly is get qualified. The certification burden, not the metalworking, is what keeps the supplier list to a handful of names, and that is a policy choice as much as a technical one. It is also where the reform agenda elsewhere in the department, which is trying to widen the supplier base by cutting administrative barriers, runs into a case where the barrier is genuinely about product acceptance. A shell that fails does so violently, and nobody is going to wave through a new forger on paperwork grounds.

The prime contractor in this case put two hundred million dollars of its own money into replacing the equipment during the summer. That is a real signal about who thinks the demand is durable, and it is worth more than any announcement.

The Capital Signal

The capital signal is that in munitions the scarce asset is a qualified line, not a funded one, and the market keeps pricing the second as though it were the first. Contract awards and facility investments are announced continuously. Acceptance of first articles is not announced at all, which means the information that matters is the information nobody publishes.

The structural read is to ask three questions of any capacity story in this sector. Has the line made the part to specification, has the customer accepted it, and what is the delivered monthly rate now rather than the planned rate later. A new complex under construction in Iowa, funded at six hundred and thirty-five million dollars and due at the end of the decade, is the same bet being placed again on a longer timeline. It may well work. It will not change anything before then, and the rate between now and then is the only number that counts.

Signal: Four and a half billion dollars of artillery investment has produced a delivered rate near a third of the goal, because the binding constraint was qualification and nobody buys that with capital.