Vorra Deepcut arrived at the forge district workshop two days after Dhug’mhar’s revised survey reached Yohan.

She carried a materials ledger and an expression Zul’jinn recognized, because he wore it himself when a calculation had produced a number he did not want to have produced.

She had come south for the ore survey question, the manganese-titanium concentrations from the eastern range that Khao’khen had flagged for Sakh’arran to verify against existing extraction agreements. That conversation could wait.

She put the ledger on Zul’jinn’s worktable first.

"The substrate question," she said. "You sent it four days ago. I am sorry it took this long."

"You had a reason."

"The third incursion’s northern corridor wall. Triple-density application, same as the eastern wall after the second incursion."

"I authorized it the morning after the engagement, before your message arrived, and I did not connect the two until I was reconciling consumption against your fourteen-day projection three days later."

Zul’jinn opened the ledger to the page she indicated. Two triple-density treatments, eastern wall and northern corridor, both drawing from the same substrate base that fed his bonding compound.

Neither treatment had existed in either of their original projections, because neither incursion had existed when the projections were made.

"Your day-eleven figure," she said. "It assumed the Arch’s consumption stayed at the original baseline rate. It did not."

"The combined draw from both triple-density treatments puts the shared pool’s exhaustion point at five days from now, not eleven."

Five days. Dhug’mhar’s revised survey, which had reached Zul’jinn the previous morning, had already compressed his completion target from fourteen days to seven.

Five days of substrate against a seven-day target was not a gap that careful sequencing could absorb the way the day-eleven gap might have been.

"How much of the original day-eleven gap remains if I stop drawing entirely, right now," he said.

"You would have enough for nine more positions at standard density. You have eleven remaining."

He looked at the ledger, then at the installation arc’s progress sheet, where positions thirteen through twenty-three were listed with their structural criticality notations from his re-survey three days earlier.

Nine positions of standard-density substrate against eleven positions remaining, with two days less time than he had been planning around.

"The kobold ore survey," he said. "The manganese-titanium concentrations. What is the compound chemistry if I substitute manganese-based binder for the void-compound base in the bonding agent?"

Vorra had clearly already thought about this, because she did not pause before answering. "Slower cure. The manganese binder does not achieve full crosslinking at the temperatures your installation sites are running."

"It will reach a bond strength eventually, but the six-hour interval will not be enough to know where it lands."

"Untested."

"Completely untested for this application. The Ironbeard mining operations use manganese binders for permanent seals in stable-temperature tunnel sections."

"Your wall face is exposed, cold, and under a load profile I have never seen the binder used against."

Zul’jinn looked at the eleven remaining positions for a long moment.

Two of them, near the western flanking tower, had been categorized in his re-survey as lowest criticality. The tower’s own structure and the arc’s existing completed installations on either side provided enough redundancy that a position operating below full specification would not compromise the arc’s overall function.

"Position thirteen," he said. "It is in cycle two’s queue already, lowest criticality on the list. I want to run the manganese binder there as a live trial."

He set the note down for a moment before picking it back up.

"Not as a backup plan. As the test. I want to know today, not after the arc is finished, whether this is something I can use."

"If it fails the six-hour check, you have spent a curing cycle on a position you will need to redo with the substrate you are trying to conserve."

"If it does not fail, I have a path to nine standard-density positions covering the seven highest-criticality remaining sites, and manganese covering the rest."

He was already writing as he spoke. "Gorrak does the verification at two hours and again at six. The two-hour check is new, but I need the early signal more than I need him to follow the existing schedule exactly."

Vorra read what he was writing over his shoulder. "If the two-hour reading is low, you will not know yet whether it is slow-curing or failing."

"I will know it is not failing fast, which is information I do not currently have at all."

He finished the note and handed it to her. "Take this back with you. If the trial works, I need to know the manganese supply’s volume and where it is staged, before cycle two’s crews start their day, not during it."

She took the note. "I will also need to revise the Arch’s own projections against whatever you end up consuming here."

"The corridor wall treatment was triple density because it was a new breach point. If your conservation gets the shared pool further than five days, that buys the Arch room too."

"Then we are solving the same problem from two ends," Zul’jinn said. "I will send Gorrak’s two-hour reading to you the moment he has it. Whatever it shows, you should know it before you are halfway back to the Arch."

She left within the hour, taking the eastern road rather than retracing the way she had come, because the eastern road passed within sight of the ore survey’s nearest marked extraction site.

She wanted to look at it herself before she reported back on whether the manganese concentrations were as promising as the survey claimed.

Zul’jinn went to find Gorrak, who was already at the perimeter checking the morning’s first applications.

He gave him the new two-hour protocol with the same flat economy he gave every instruction that mattered.

Gorrak wrote it down without comment, the way he wrote everything down, and went to prepare position thirteen for a compound it had never received before.