The reload

A Burke-class destroyer leaves a West Coast pier with a fixed number of Mk 41 cells. Those cells are not a software stack. They are canisters that were loaded alongside a crane, in a harbor that had the handling gear, the certified crew, and the time to do the work at something like two cells an hour. Once that ship is west of the date line, the cells it sortied with are the cells it has until it reaches another place that can do the same work. Underway replenishment will give it fuel, food, and parts. It will not, as a matter of practiced routine, drop a live SM-6 or LRASM into an empty cell in a sea state that looks like the Philippine Sea in winter.

That is not a secret. It is the reason Congress, in the FY2027 NDAA markup, asked the Navy for a briefing on at-sea VLS reload instead of assuming the capability already existed as a fleet standard. The Navy has run demonstrations. In October 2024, USS Chosin used a revived Cold War Transferable Reload At-sea Method alongside a dry cargo ship off Southern California and put a canister into a cell while underway — the first such evolution in a generation. In 2025, during Large Scale Exercise work, a crane-equipped support ship transferred canisters into both magazines of a destroyer while the ships were moored. More testing on expeditionary transfer docks is on the 2026 calendar. None of that is the same thing as a Pacific war in which a destroyer empties a forward bank on Tuesday and is a shooter again on Thursday without leaving the theater.

The pierside fact still governs. Reloading VLS the way the fleet actually does it wants specialized infrastructure and days, not hours. The average published rate is about two cells an hour when the pier, the crane, and the crew are all present. A ship that has to leave the fight, steam to a harbor that still has that kit, wait its turn behind other empty combatants, and steam back is not “reloading.” It is exiting.

What the demonstrations actually proved

They proved the mechanics can be made to work in selected conditions. TRAM moves a canister horizontally along rails, tilts it vertical, and lowers it into the cell. The 2025 moored test used a conventional crane on one bank and a frame that handled two canisters on the other. General Dynamics NASSCO has put a concept on the table for an AD(X) destroyer tender, built off the AS(X) submarine-tender design, that would hold station with dynamic positioning and reload several destroyers with only modest changes to the parent hull. The existing submarine tenders — Frank Cable and Emory S. Land — still do weapons handling for boats at a handful of Pacific sites, and they are more than forty-five years old, with replacement money moving slowly enough that a net gain of tenders this decade is not a planning assumption anyone should treat as inventory.

What the demonstrations have not proved, at least in the public record through late summer 2026, is a live round going into a combatant’s cell in open ocean under the weather and threat a Western Pacific fight would impose. An empty canister in calm water is a necessary test. It is not a magazine restored. Sea-state limits for TRAM have not been published as a fleet standard. A method that works alongside a T-AKE off San Diego is still a method that has to survive a missile belt, a damaged crane ship, and a destroyer that cannot loiter.

China does not have this particular problem in the same form. The PLAN fights near its own yards, its own missile dumps, and its own short interior lines. A Type 055 that empties cells can, in principle, reach a Chinese pier on a clock that is not measured in a trans-Pacific transit. The United States is the visiting team. Fuel and food UNREP is a genuine American advantage. VLS UNREP is the advantage that was allowed to lapse after the Gulf War and is now being rebuilt as a demonstration program while the magazine problem from Part 2 is already live.

Geography is the second magazine

The first island chain is long. The list of places that can actually handle the canisters is short. Surface combatants still lean on a small set of Western Pacific ports and on whatever expeditionary kit can be towed into an atoll or a sheltered harbor. Submarines that shoot Tomahawks and heavyweight torpedoes have an even thinner map: Yokosuka, Sasebo, Apra Harbor, and a short list of other handling sites that have been used or named in public, backed by two aging tenders. If those nodes are under fire, or if the crane is a wreck, the magazine ashore does not reach the magazine afloat.

That is why Part 9’s coalition clock is not a sentiment. Japanese and Korean yards, Australian handling, and any allied pier that will take American canisters are part of the reload, not a diplomatic sidebar. A study line in a budget about putting U.S. weapons on allied steel does not create a certified weapons-handling pad. The pad is a notice, a host-nation agreement, a crane, and a crew that has done the evolution more than once.

The wounded-ship problem sits on the same map. A destroyer that survives with damage still needs a dock that can take it. The repair queue in Part 3 already eats hulls the force-structure chart counts as present. After the first exchange that queue is in theater, not in Virginia. A tender that can reload cells but cannot patch a hole, or a dock that can patch a hole but cannot lift a canister, is half of a logistics system pretending to be whole.

What the model does with this

Software belongs here. Magazine accounting, reload sequencing, threat windows for a loitering UNREP pair, and the decision to send a half-empty destroyer home versus keeping it as a sensor are exactly the jobs a model can compress. What the model cannot do is mint a crane ship, qualify a tender crew, or move a live canister through a sea state the test program has not yet signed off.

Treating the next demonstration as a substitute for the next tender is the same Trojan-horse move the series has been naming since Part 1. Opportunity cost dressed up as strategy still applies. Money that buys another rendering of a future UNREP system does not put a canister into a cell on a Tuesday in the Philippine Sea. Money that buys another training cluster on a feeder that was supposed to serve a munitions plant does not either.

The apprentice from Part 12 is on this brief whether the desk writes the name or not. TRAM, a frame reloader, and a tender crane all require people who have done the evolution until it is boring. A yard in the United States that cannot spare those people for a Pacific tender is choosing the wrong ocean.

What this is not

This is not an argument that underway replenishment is a myth or that the 2024 and 2025 tests were theater. They were the first honest work on a gap the fleet had lived with for thirty years. It is also not an argument that every destroyer must become a floating magazine independent of the shore. Some reloads will always want a pier. The question is how many days the ship is a non-shooter while it finds that pier, and whether the pier will still be there.

Allied ports, expeditionary transfer docks, and a purpose-built tender class are the realistic near-term list. Cheap mass from Part 7 changes the mix of what has to go into a Mk 41 cell, which is useful, and does not erase the cells that still have to be filled with the expensive interceptors that keep the ship alive long enough for the cheap mass to fly.

Bottom line

A magazine that cannot be refilled in theater is a one-week navy. A hull that cannot be repaired in theater is a one-exchange navy. The United States still has the better connected-replenishment tradition on earth for everything except the weapon that decides whether the ship is still a combatant. The demonstrations are late and necessary. They are not yet a fleet.

Put the models on the reload window and the threat picture. Put the money on tenders, transfer docks, allied handling pads, and the crews who have already done the lift. Count certified reload points west of the date line the way Part 2 counted cells and Part 3 counted slips.

The war is not won by the side that leaves home with a full launcher. It is won by the side that can fill the launcher again without leaving the fight.

Come back for 14

Even a working reload is only a transfer. The canisters have to come from somewhere that still has them after the first week of firing — a depot, an ammunition ship, a host-nation bunker — that can survive the same raid calendar as the pier.

Part 13 is the lift. Part 14 is the pile behind the lift.

The dump is the next finite input.