The millrace elevation work was the kind of project that looked simple from a distance and revealed its complexity only when you were standing in it.
From a distance, it was: raise the millrace channel so that water enters the wheel from above instead of below. Straightforward. A matter of grade.
Standing in the channel, with twelve people, water flowing past your boots, and a head-pressure calculation that needed to be accurate to within two centimeters or the efficiency gain was lost — it was a different proposition.
The millrace was 110 meters long. Junho had walked it with Mara the evening before and marked the work in three sections. Section one: the intake dam, where the diversion from the creek needed to be raised fifty-four centimeters. This was Hendry's work — stonework in running water, the hardest part. Section two: the channel body, a gradual grade adjustment running from the raised intake to the wheel housing. This was bulk earthwork, the kind Mara's twelve people were suited for. Section three: the wheel housing approach, where the elevated water needed to be directed onto the wheel crown with a wooden flume — the final delivery mechanism, Calder's domain.
Three sections. Three crews. Running simultaneously, because the timeline did not allow sequential work.
Junho spent the first morning moving between them.
* * *
Hendry's intake section was the one he checked first and most frequently.
The existing diversion dam was a modest affair — a low stone weir across the creek at the millrace intake, just enough to divert a portion of the flow into the channel. It had been built when the original mill was constructed, done well enough that it had survived two years of neglect without significant degradation.
Raising it fifty-four centimeters meant adding courses of stone on top of the existing weir, which meant working in the creek, which meant wet, cold, slow, careful work.
Hendry had brought his grandson — a boy of fourteen named Ott, who was large for his age and apparently immune to cold water — and between the two of them they had established a working pattern before Junho arrived. Ott in the creek, passing stones up. Hendry on the weir crest, fitting each stone with the particular attention of a mason who had internalized the principle that speed without precision was just expensive failure.
Clk. Clk. Clk.
Stone finding its seat.
Junho crouched at the bank and watched a few courses go in. The mortar Hendry was using was a lime mix — the barony's own limestone, burned in a field kiln two days ago — and it was going in correctly: not too thick, not too thin, each course pressed and checked for level with a straight-edge before the next one started.
'How's the footing?' Junho asked.
'Solid,' Hendry said, without looking up. 'The original base courses are good stone, well-bedded. I'm building on them, not around them.'
'Water pressure on the raised face will be higher than the original design.'
'I know.' Hendry fitted a stone, checked it, added a wedge sliver. 'I'm stepping the upstream face. More mass at the base. It'll hold.'
He anticipated the hydraulic load increase. I didn't tell him to step the face, he just did it because it's the right thing to do.
Junho moved on to section two.
* * *
Section two was organized chaos of the productive variety.
Twelve people in a 110-meter channel, regrading the bed to a half-percent slope, was not quiet work. It was shovels and rakes and a continuous debate about where exactly the grade markers were and whether a given section was high or low.
Junho had set grade stakes the previous evening — sharpened poles driven into the channel bed at ten-meter intervals, each one marked with a notch at the correct elevation for the new grade. The workers' job was to cut or fill the channel bed to match the notch on the nearest stake, using material from the high sections to build up the low ones.
The principle was simple. The execution required constant attention because people had a natural tendency to average rather than match — to make things approximately level rather than precisely graded — and a drainage channel graded approximately was a drainage channel that didn't drain properly.
Wyll had understood this immediately and taken on the role of grade-checker without being asked, moving up and down the work section with a straight-edge and a critical eye, calling corrections.
'Too high here. Half a hand. Move it to the low section at thirty meters.'
'This section's right. Move on.'
'Brin, you've been filling that spot for ten minutes. It's done. Stop patting it.'
Brin, the fifty-three-year-old thatcher who came to everything as a civic principle, accepted this with the equanimity of an old man who had stopped needing to be right.
Junho spent thirty minutes in section two, checking Wyll's checks, finding three places where the grade was off enough to matter and two where it was fine and Wyll had called it correctly. He made corrections, explained why they were corrections rather than preferences, and moved on.
* * *
Section three was Calder, alone, building a flume.
The flume was the elevated wooden channel that would carry water from the end of the raised millrace to the crown of the wheel — the final delivery mechanism that turned fifty-four centimeters of head pressure into a wheel turning at two and a half times its previous speed.
It was not complicated engineering. A rectangular wooden trough, sealed with pitch, mounted on a timber framework over the wheel housing. The critical dimensions were the outlet position — it needed to deliver water to the wheel at the correct angle for maximum rotational force — and the trough dimensions, which controlled flow rate.
Calder had the framework up already. He was working on the trough itself, planing the interior faces of the side boards smooth so the water flowed without turbulence, fitting the base boards with the tongue-and-groove joinery that would minimize leakage.
Shhhk— shhhk— shhhk—
A hand plane working with the grain. The sound of a craftsman in their element.
Junho stood and watched for a moment without interrupting. There was a specific quality to watching Calder work that he'd come to recognize over the weeks — the way the carpenter's hands knew things his conscious mind didn't have to deliberate about. The angle of the plane adjusted automatically. The pressure varied without calculation. Things that would require explicit decision-making from a less experienced hand happened below the level of thought.
That's ten thousand hours of practice, at minimum. More, probably.
He's twenty-six and he has ten thousand hours of wood sense.
'The outlet angle,' Junho said.
Calder didn't look up. 'Twenty-two degrees from vertical. You said the optimal impact angle for maximum torque on the wheel paddles was—'
'Twenty to twenty-five. Twenty-two is fine.'
'I split the difference.' Calder ran the plane down the board again, checking the surface with his palm. 'The pitch delivery. I've been thinking about waterproofing the trough interior.'
'Pine pitch.'
'We don't have pine pitch. We have pine resin — different thing. Less flexible, cracks under thermal cycling.' He looked up now, finally. 'Beeswax and tallow mixture works better for indoor joinery. For a water-carrying channel it's—' he considered, '—adequate for a season. Better than pine resin.'
'And better than nothing, which is our alternative.'
'Yes. I'll mix it this evening. The trough needs to cure overnight before we run water through it anyway.'
He's already solved the problem I was going to raise.
I should stop being surprised by this.
'How long until the flume is ready?' Junho asked.
'Three days for the trough and framework. One day for the pitch treatment and cure. Day five, ready to connect to the raised millrace.'
'And the millrace grade work will be done in four days at current pace,' Junho said. 'Hendry's intake is the constraint — he needs five days for the stonework to set properly before we run full flow.'
'So day six for first water on the new configuration.'
'Day six,' Junho confirmed.
Calder returned to his plane. The shavings curled and fell.
* * *
On the third day of the millrace work, the visitor arrived.
Junho was in section two, working through a grade discrepancy near the eighty-meter mark, when Pell appeared at the channel edge with the particular expression he wore when something had happened that was outside his operational scope.
'There's a man at the farmhouse,' Pell said. 'He says his name is Harwell. He rode up from Crestfall this morning.'
Harwell. Sooner than I expected.
I sent him a message four days ago asking to meet within the week. He came in three days, which means either his timeline is urgent or Colwick's is.
'Tell him I'll be an hour,' Junho said. He was in the middle of the grade correction and leaving it half-done would mean losing the reference points. 'Offer him food and water. He's been riding.'
Pell disappeared.
Junho spent the hour finishing the correction, checking the grade against three stakes, making one further small adjustment, and confirming with Wyll that the section looked right. Then he extracted himself from the channel, removed the mud from his boots with the thoroughness that decency required, and went to the farmhouse.
Harwell was sitting at the table eating bread and cold meat with the unselfconscious appetite of a man who had been on a horse since dawn. He looked up when Junho came in and did not apologize for eating, which Junho took as a reasonable sign.
'Lord Ashmore.' He pushed the plate slightly aside. 'Thank you for agreeing to meet.'
'You came faster than I expected,' Junho said. He sat across from him.
'Lord Colwick's project schedule moved forward,' Harwell said. 'The eastern granary expansion — the master builder revised his timber intake schedule. He needs structural oak in significant volume within four months.'
'Four months from now, or four months from when the original schedule said?'
'From now. The original schedule was six months out. It's been compressed.'
Someone pushed the construction schedule forward. Which means Colwick wants the building done faster, which means the demand is real and the timeline is genuine.
'Tell me the spec,' Junho said.
Harwell produced a folded document from his coat and spread it on the table. It was a timber specification — Junho could read the format immediately even in this world's script, because timber specs were timber specs and the structure of them was universal. Species. Grade. Dimensions. Volume. Delivery schedule.
He read it.
Ping—!
―――――――――――――――――――――――――――
[ ENGINEER'S EYE — COLWICK TIMBER SPECIFICATION ]
Lord Colwick — Harren Granary Expansion
Project: Eastern granary complex, 3-building expansion
Timber requirement:
Primary structural beams: 60 pieces, min. 65cm diameter, 8m+ length
Secondary structural: 120 pieces, 40–55cm diameter, 4–6m length
Species: Oak only (structural grade or above)
Volume: approx. 180–220 cubic meters total
Delivery: 3 staged lots over 4 months
Price offered: 3.2 gold/m³ (primary beams), 2.6 gold/m³ (secondary)
Assessment:
Primary beams: NORTHEAST RIDGE ONLY (current mill cannot process)
Secondary structural: Current mill CAN process (sub-55cm diameter)
Revenue potential: 180–220m³ @ blended ~2.9 gold/m³ = 522–638 gold
⚠ Primary beam delivery requires mill upgrade completion first.
Current mill max diameter: 55cm. Ridge trees avg: 74cm.
Upgrade completion: Day 50 (6 days from now)
First delivery possible: Day 65–70
―――――――――――――――――――――――――――
Junho set the specification down.
522 to 638 gold. From one contract.
Combined with the Brek deliveries already scheduled, that covers the year-one payment. Comfortably.
If I can deliver.
The primary beams need the upgraded mill. I have six days until the upgrade is ready. The question is whether Harwell's timeline can accommodate a sixty-five to seventy day first delivery window.
'The primary structural beams,' Junho said. 'Sixty-five centimeter minimum diameter. You know the ridge stand — Calder's survey will have reached you through Brek.'
'I know about the ridge stand,' Harwell said. 'That's why I'm here.'
'I can't process those trees on the current mill. The saw frame maximum is fifty-five centimeters. I'm upgrading the mill now — new overshot wheel, wider frame. It'll be operational in six days.'
Harwell looked at him steadily. 'Six days.'
'Six days. After that I can process the ridge oak. First delivery in — call it three weeks from upgrade completion, to allow time for felling, extraction path development, and processing.' Junho calculated. 'Day seventy from today, approximately.'
'The secondary structural — the forty to fifty-five centimeter pieces. Those I can take from your current operations?'
'Yes. The second forest section has suitable oak in that range. I can begin pulling secondary structural pieces immediately and stage them for first delivery while the primary beam processing waits for the upgrade.'
Harwell looked at the specification document. He was, Junho had come to understand from their brief road conversation, a man who thought in logistics rather than in value — his first question was always about delivery, not price.
'Lord Colwick's master builder,' Harwell said slowly, 'wants the primary beams first. He needs to set the structural frame before anything else goes up. Secondary pieces come after.'
Junho absorbed this.
The construction sequence requires the large primary beams before the secondary structural. Which means delivering secondary first doesn't help Colwick — it just means his builder has to wait anyway, with secondary material stacking up on site.
So the constraint isn't my mill upgrade. It's whether Colwick will wait for day seventy.
'Can Colwick's schedule accommodate a day-seventy first delivery of primary beams?' Junho asked.
'It can accommodate a day-eighty delivery. Beyond that the construction window closes before winter and the project pushes to next year.' Harwell folded the specification document. 'Day seventy is inside the window.'
Day seventy. I have the room.
'Then I can commit to the specification,' Junho said. 'Day seventy for first delivery of primary structural beams, subject to the mill upgrade completing on schedule. Day fifty-five for secondary structural, first lot.' He paused. 'The price.'
'3.2 gold per cubic meter for primary, 2.6 for secondary,' Harwell said. 'As in the specification.'
3.2 gold per cubic meter for old-growth primary structural is a good price. It's below the specialist market ceiling I identified — the system put that at 4.0 to 5.5 — but Harwell knows the market too and he's come in at a respectable number. I could push for 3.5 and probably get 3.3 or 3.4.
But Colwick's project is active and his timeline is tight. Pushing price now trades relationship for margin. The relationship is worth more over the long term.
...On the other hand, 3.5 to 3.4 on 60 primary beams is real money. This is not the time for sentimentality about relationships.
'3.5 on the primary beams,' Junho said. 'The secondary at 2.6 as offered.'
Harwell looked at him. The familiar assessment look — recalibrating, calculating.
'3.35,' Harwell said.
'3.4,' Junho said.
'Done,' Harwell said. He put the specification document back on the table between them. 'I'll have a formal contract written in Crestfall. You can review and sign it when your second delivery goes through.'
'I'll have a rider to Brek's office in three days confirming timeline.'
They shook hands across the farmhouse table in the way of two people who had concluded a business arrangement and were both satisfied with the conclusion, which was the best possible outcome of any negotiation.
Harwell stood. He looked around the farmhouse — the working documents on the table, the production tally parchments pinned to the wall, Pell's neat administrative notes alongside Junho's engineering sketches. Then he looked at Junho with an expression that was not quite the assessing look but something adjacent to it.
'I've been contracting timber for Lord Colwick for eight years,' he said. 'I've dealt with every supplier in the Northern March.'
'And?' Junho said.
'And I've never been to a supplier's home that looked like this.' He gestured at the wall of documents. 'Most lords have a steward who keeps notes in a ledger. This looks like a shipyard office.'
A shipyard office. That's actually a reasonable comparison.
'It's a working operation,' Junho said. 'It needs to be managed like one.'
Harwell looked at him for a moment longer. Then he put his coat on. 'Day fifty-five for secondary. Day seventy for primary. I'll have the contract ready.'
He left.
Junho looked at the wall of documents. He had not thought about it as looking like anything in particular — it was just the information he needed, organized so he could find it.
A shipyard office.
I'll take it.
* * *
Six days later, on day fifty.
The millrace was ready. Hendry's mortar had cured for five days and he'd checked it twice and said it would hold. The channel grade had been confirmed by Junho walking it end to end with a level — a proper water level, a clay pot filled to the brim and carried carefully, which was the most accurate tool available and was, Junho had to admit, more elegant than it sounded. The flume was built, pitched, cured, and mounted over the wheel housing with the outlet angle precisely at Calder's twenty-two degrees from vertical.
The new wheel was not installed yet.
This was the part that had been worrying Junho since day forty-four.
The overshot wheel was larger than the undershot — 210 centimeters diameter versus the original 130 — which meant a larger, heavier structure that required a wider axle channel through the wheel housing and different bearing geometry. Hendry had widened the housing channel over the past week, stone by stone, a careful process that required removing existing masonry without destabilizing the housing structure. He'd done it correctly — Junho had checked — but it had taken most of the upgrade period.
The wheel itself Calder had built in sections, brought to the site on a flat sledge, and was now in the process of assembling on the foundation beside the housing. Hub, spokes, rim sections, float boards. Larger than the first wheel in every dimension, the elm hub the size of a barrel end, the oak spokes longer than a tall man.
'Today?' Junho asked, arriving at the site.
'This afternoon,' Calder said. He was fitting the last float boards, pegging them at the rim with the iron clips Gorvan had supplied — a new batch, sized for the larger wheel. 'I need to balance it first.'
'Balance it.'
'A wheel this size, if the mass isn't evenly distributed, it'll run rough and the bearing wear will be uneven.' He tapped one of the rim sections. 'I've been shaving material from the heavy side. Two more passes and it'll be right.'
Junho had not thought about wheel balancing. He should have thought about wheel balancing. It was obvious in retrospect — a two-hundred-kilogram rotating assembly with uneven mass distribution would vibrate, and vibration was the enemy of every component in the mechanism.
He thought of it. I didn't. This keeps happening.
At some point I should stop treating it as a surprise and just accept that Calder knows things I don't and that this is a feature, not a bug.
'Take the time you need,' Junho said.
Calder took two hours. When he was done he had Junho spin the assembled wheel on a temporary axle — just lifting one side and letting it rotate freely — and they watched it turn.
It slowed. Stopped. Turned slightly back. Stopped again. Evenly.
'That's balanced,' Calder said.
The wheel installation was the same process as the first time — shear legs, rope lift, a team of people managing the weight as it came vertical. Larger wheel meant larger forces, but the same rig geometry worked, scaled up. Wyll anchored one guy rope with the focused attention of someone who had done this before and knew which ropes mattered.
Grrk— Grrk—
The larger wheel rising.
Clonk.
The axle seating in the widened housing.
A different sound from the first time — heavier, more authoritative. The sound of something with more mass finding its place.
Hendry packed the bearing seats. They waited two hours for the mortar to reach working strength. Junho spent those two hours on the wide-frame saw mechanism — the crank geometry needed adjustment for the larger wheel's different rotational speed, and the new saw frame, twice the width of the old one, needed its guide rails confirmed in alignment before water was added to the equation.
Ping—!
―――――――――――――――――――――――――――
[ ENGINEER'S EYE — PRE-RUN ASSESSMENT ]
Mill upgrade — Day 50 pre-commissioning check
✓ Millrace grade: 0.49% — within spec
✓ Flume installation: Outlet at 22° from vertical — correct
✓ New wheel: 210cm diameter, balanced, bearing seats packed
✓ Axle: Widened housing confirmed, collar seated
✓ Wide-frame saw: Guide rails aligned, blade tensioned
✓ Crank geometry: Adjusted for new wheel RPM
Theoretical performance vs. current mill:
Power output: 2.6x (60–75% water energy captured vs. 25–35%)
Blade speed: 2.3x (est. 160 strokes/min vs. 70)
Max log diameter: 110cm (vs. 55cm current)
Throughput estimate: 22–28 logs/day (vs. 10–12 current)
Risk: Elevated water power = elevated mechanical stress.
Monitor pitman arm and crank fittings closely in first week.
―――――――――――――――――――――――――――
22 to 28 logs per day. More than double the previous throughput.
If those numbers hold under actual load, the Colwick primary beams are achievable well ahead of day seventy.
Don't celebrate projected numbers. Wait for actual ones.
The group had assembled again — not as large as the original test run, but substantial. Calder and Hendry. Wyll and his brother. Mara, who had come from the east field where she'd been supervising the last stages of surface leveling. Pell, who always came to significant moments with the expression of a man cataloguing evidence for a private assessment he would complete later.
Gorvan was there.
Junho noticed him standing at the edge of the group — the big blacksmith on a cart horse, having apparently ridden up from Crestfall without sending word. He met Junho's eyes and nodded once. He had come to see whether the mechanism he'd made components for could handle the upgraded load.
He keeps showing up. Every significant test. He's invested in this working.
I'll remember that.
'Same as the first time,' Junho said. 'Everyone back from the mechanism. Calder, you're on the gate. Half flow first.'
He went inside.
He stood where he could see the whole kinematic chain. Larger wheel. Larger crank. Stiffer pitman arm — Calder had rebuilt it in denser hardwood for the higher load. Wide-frame saw on its guide rails, the new blade — a wider version from Gorvan, delivered three days ago — tensioned and waiting.
'Open it,' he called.
Krrk—
Water into the flume.
Down the elevated channel. Over the lip of the flume outlet. Onto the wheel.
The first float boards caught it and the wheel — the large wheel, the two-hundred-and-ten-centimeter wheel built from elm and oak and Calder's ten thousand hours of wood sense — moved.
Whmm—
Deeper than before. The sound of more mass turning. More energy.
The crank came around faster. The pitman arm moved with the authority of a mechanism that had more power behind it than its predecessor. The wide saw frame descended on its guide rails—
SKKRR—
The sound was different. Louder. More force behind the stroke.
Whmm— SKKRR— Whmm— SKKRR—
Junho walked around the mechanism. Everything moving. Nothing failing. The crank fittings — Gorvan's work — were taking the load without visible distress. The guide rails were holding the wide frame precisely in plane. The axle was turning smoothly in its bearing collars.
'Full flow,' he called.
Krrk—
The gate opened. The flume filled. The wheel accelerated.
The mechanism reached operating speed and the sound of it filled the mill interior with something that was not quite music but had a similar quality — purposeful, consistent, the rhythm of a machine doing exactly what it was built to do.
WHMM— SKKRR-SKKRR-SKKRR-SKKRR—
160 strokes per minute. Junho counted them. 158. 162. 159. The average was holding.
'Log,' he said.
Wyll and his brother carried in the largest log they had staged — a pine from the second forest section, sixty centimeters diameter, which the old mill could not have processed without hand-splitting. They set it on the new wide bed, aligned it against the stop pins.
Calder took the advance lever.
'Go,' Junho said.
SKKRR— SKKRRRK—
The blade meeting the end grain of a sixty-centimeter pine log.
The wheel didn't slow. The mechanism didn't strain. The load went in and the power absorbed it and the blade kept cycling and the log kept feeding and the cut—
WHMM— SKKRRRK-SKKRRRK-SKKRRRK—
The cut advanced. Steadily. Faster than Junho had ever seen a frame saw cut. The sawdust poured from the kerf in a continuous pale stream, piling on the floor below the bed.
Two minutes and forty seconds.
The log fell apart.
THUMP.
Larger than the first time. A heavier sound.
Junho looked at the cut face. Flat. Clean. The slight waviness of a frame saw present but minimal — the higher blade speed had reduced it, the faster stroke leaving less time for deviation. The dimensional consistency was better than the original mill's output.
He walked outside.
The group.
Gorvan had moved forward from the edge. He was standing closer than before, looking at the mill building with the expression of a craftsman hearing his work perform at a level he'd made possible.
Mara was not folding her arms. She hadn't folded them in some time, Junho realized. He couldn't remember the last time.
'That's different,' Wyll said, from behind him. Meaning the sound. The upgraded mill sounded different from the outside — deeper, more deliberate, the exhaust of more power going through a mechanism.
'Yes,' Junho said.
Ping—!
―――――――――――――――――――――――――――
[ QUEST MILESTONE ACHIEVED ]
Ashmore Mill — UPGRADED
Day 50. Overshot configuration operational.
Wide-frame saw: 110cm maximum diameter — confirmed
First test cut: 60cm pine — 2 min 40 sec
Projected throughput: 24 logs/day (first week estimate)
Cut quality: Grade A (improved from B+ on original mill)
REWARD:
[Blueprint: Overshot Watermill (Improved)] — UNLOCKED
[Territory Status: RECOVERING → STABLE]
+250 EXP
New capability: Northeast ridge old-growth oak — NOW PROCESSABLE
Colwick contract: Primary beam delivery window — CONFIRMED
―――――――――――――――――――――――――――
Territory status: Stable.
Not recovering. Stable.
From Critical to Distressed to Recovering to Stable. In fifty days.
That's not a small thing.
Don't stop moving.
He looked at the group one more time.
Then he looked at Gorvan. The blacksmith caught his eye.
'How are the crank fittings?' Gorvan asked. Not from the edge of the group anymore. He'd walked forward, somewhere during the test run, without apparently deciding to.
'Running clean,' Junho said. 'No distress at full load.'
Gorvan nodded. The nod of a craftsman whose work had been tested and had held.
'You'll need a second blade,' he said. 'Wider frame, faster cycle — the current blade will dull faster under load. I can have a second one ready in ten days.'
'Make it,' Junho said.
'Price is the same as the first.' A pause. 'Less the fifteen percent I'm waiving because I want to see the ridge oak come through this mill.'
He's waiving fifteen percent because he wants to see the ridge oak.
There are people investing in this outcome for reasons that are not purely commercial.
That's worth something I don't have a word for yet.
'Thank you, Gorvan,' Junho said.
The blacksmith waved it off, the gesture of a man uncomfortable with acknowledgment, and went to get his horse.
The mill ran on behind them all, the upgraded mechanism turning in the afternoon light, the water falling from the flume onto the wheel crown in a clean white arc and the wheel taking it and turning it into work.
WHMM— SKKRR— WHMM— SKKRR—
The sound of a stable operation.
The sound of something that would still be here tomorrow.
—
[ End of Chapter 11 ]
~ To be continued ~
Wanna read up to chapter 20. All of them are live and free to read on my Patreon.
Join Patreon: patreon.com/MASKO73