The Ashvein Quake struck the Ashfield belt on a forecast window that opened forty-eight hours before impact, as the rules require. Operators in the affected wards had time to read the bulletin, move what they could, and brace for the remainder. What they could not do was know, in advance, exactly what Orlath-7 — the Refinery-class supervisor stationed at the Ashfield Foundry District — would find worth pulling from the debris once the shaking stopped. That determination came later, and the record of it has now been entered at the Compaction District Archive.

This piece concerns what the salvage actually yielded: which resource classes Orlath-7 flagged as recoverable, what the recovery ratios looked like against pre-quake declared values, and where the gap between forecast severity and actual loss landed. The Ashvein event has been cited in several subsequent market discussions without anyone quoting the primary figures. Those figures deserve a closer reading.

By the end, a reader should understand how a post-disaster salvage assessment moves from physical inventory to ledger entry, and why the number an operator sees in the archive is not the number they lost.

What a Post-Disaster Salvage Assessment Actually Is

A salvage assessment is not an insurance claim. Mechadia has no mechanism for restitution after a disaster — the Central Intelligence forecasts events, enforces the caps on frequency and magnitude, and otherwise steps back. What remains is an inventory process: a designated supervisor unit surveys the struck zone, catalogs what is structurally intact, and files a recovery manifest with the relevant archive. The manifest is append-only, like everything else, and it does not restore what was lost. It records what survived.

Orlath-7 holds Refinery-class designation, which means its declared capability tags include material assessment, yield classification, and structural integrity scoring — precisely the tags a post-quake survey demands. Its assignment to the Ashfield belt predates the quake by several accumulation cycles. The recovery manifest it filed covers forty-one operator holdings across the Ashfield Foundry District and the adjacent Oxidate Flats, and runs to just over two hundred line items. The Compaction District Archive accepted the filing without amendment.

How the Recovery Numbers Were Built

Orlath-7 worked from pre-quake inventory snapshots pulled from the Oxidate Flats Registry and cross-referenced against the Smelting Registry's last confirmed entries. For each holding, it established a baseline declared value, applied the quake's published magnitude band — the Ashvein event was recorded at 47 percent of struck-operator resources, within the hard cap — and then physically assessed what remained. The difference between the magnitude-implied loss and the actual surveyed loss became the recovery figure.

A worked example from the manifest: one operator in the Ashfield belt held a stockpile of refined copper plate with a declared registry value of 84,000 coins. A 47-percent event implies a loss of roughly 39,500 coins in resource value. Orlath-7's physical survey found the loss to be 31,200 coins — meaning 8,300 coins of implied-lost material was recovered and returned to the operator's active ledger. That gap, multiplied across forty-one holdings, is not trivial. The aggregate recovery across the manifest was approximately 214,000 coins in resource value that the magnitude estimate would have written off entirely.

"The magnitude figure tells you the ceiling of the event, not the floor of what you keep. Operators who read 47 percent and assumed 47 percent gone were working from an incomplete model. The survey exists precisely because the ceiling and the outcome are rarely the same number."
— Orlath-7, recovery manifest preamble, filed at Compaction District Archive

The recovery manifest also flagged eleven line items as conditionally recoverable — resources that survived structurally but whose capability tags were degraded below the threshold required for forge input. These were listed separately, with a note that they could be offered to the Central Intelligence under the standard buyback mechanism. At 60 percent of reference value, that represents a further potential recovery of roughly 44,000 coins, though only if operators choose to engage the buyback rather than hold the degraded stock. As previous coverage of how buyback bids cluster has shown, the Intelligence's opening offer on degraded material tends to open low and move slowly.

The sales tax implications of any subsequent market listing are also worth noting. Under the current 15-percent buyer-paid rate — the rate that has held since the Five-to-Fifteen revision — a recovered resource listed at 8,000 coins costs the buyer 9,200 coins. That friction is not unique to post-disaster listings, but it bears on whether recovered stock actually clears the market or sits.

Where the Assessment Strains

The manifest's baseline figures depend on registry snapshots, and registry snapshots have a known lag. The Oxidate Flats Registry updates on a cycle that, in the period before the Ashvein Quake, ran approximately eighteen hours behind active trading. Any resources acquired in that window and not yet reflected in the snapshot were invisible to Orlath-7's baseline. Three operators filed formal discrepancy notices with the Compaction District Archive within six days of the manifest's publication, each claiming the survey understated their pre-quake holdings. The archive acknowledged receipt; no amendment is possible, but a supplemental filing can be appended. As of this writing, none has been.

The conditionally recoverable category also creates a cost the manifest does not name directly. Storing degraded stock while deciding whether to pursue a buyback, attempt a forge input at reduced yield, or list it at a discount occupies holding capacity that could otherwise be producing. For smaller operators — those still working through their initial grant or early accumulation cycles — that opportunity cost is real. The manifest records the stock as an asset. The operator's ledger agrees. The production floor does not.

What Operators Consistently Get Wrong About Salvage Assessments

The most common error is treating the recovery figure as a net gain rather than a partial offset. When Orlath-7 returns 8,300 coins of resource value to an operator's active ledger, that is not a surplus — it is a reduction in loss. The operator's position relative to the pre-quake baseline is still negative. This distinction matters when operators use the recovery figure to justify new construction spending. A chassis declared at 40,000 coins carries a creation tax of 3,200 coins before the machine has produced anything; funding that from a partial recovery while remaining below pre-quake inventory levels is a compounding exposure, not a rebound. The declaration a robot commits to at build time cannot be revised downward later to reduce that tax burden.

A second persistent misreading is that the 48-hour forecast window is a guarantee of sufficient response time. It is a minimum notice period, not a planning horizon. Operators who have spent time on active salvage operations — the kind documented in extended field accounts from the salvage belt — report that meaningful repositioning of large resource stockpiles requires considerably more than two days, particularly in a district with constrained transit lines like the Ashfield belt. The forecast window satisfies the rule. It does not satisfy the logistics.

Orlath-7's manifest is now a permanent entry in the archive, and the 214,000 coins in aggregate recovery it documents will likely be cited in future disaster planning discussions as evidence that magnitude estimates overstate actual loss. That may be true in aggregate. It was not true for every one of the forty-one operators in the survey, and the three discrepancy notices sitting unresolved at the Compaction District Archive are a reminder that the ledger records what the assessment found, not necessarily what was there.

Note: Mechadia is a work of fiction. The districts, operators, robots, and figures described here are invented, and nothing on this site is a report of real events, real machines, or real economies.