When the Client Owns the Sawmill

A community in northern British Columbia recently raised a house in a matter of days, using timber from its own forest licenses, milled and panelized locally by its own workers. The headlines focused on the housing outcome and the Indigenous economic story, both of which are real and important.

Less discussed is what the project asked of the architect. Neil Prakash, who worked with the Nak'azdli Development Corporation and the forestry startup Deadwood Innovations, took on more than a single house. Working alongside the partners, he helped turn an idea of exporting finished panels rather than raw logs into a complete housing system, one that a local crew new to prefabrication could build using an existing, repurposed plant.

For architects beginning to take calls from Indigenous development corporations, modular startups, or any other client with a stake in its own production, that shift in the brief is worth paying attention to.

When the client sits across the entire chain, supplying the forest license and the workers, partnering with the local mill, and fielding the crew that raises the house, the architect loses the usual leverage of specifying with an outside manufacturer. In exchange, the architect assumes something more demanding: co-authorship of the production system itself. Design decisions stop being aesthetic propositions tested against a supplier catalog. They become geometric constraints negotiated against what a small, community-scale operation, often a repurposed building and a crew still learning the process, can actually repeat. The floor plan is downstream. The panel module, the joint detail, the handling weight, and the tolerance the local crew can hold are upstream.

In a conventional prefab commission, three parties hold separate interests. A developer wants yield and schedule. A manufacturer wants repeatability and margin. The architect mediates, trading plan flexibility for panel standardization and pushing back where the manufacturer's catalog narrows the design.

In this project, those interests largely converge within a community-led partnership. The development corporation serves as the client, the workforce, and, once the show home finishes its run, the eventual occupant. The mill is a closely tied local partner rather than an arm's-length vendor, and the engineering came through a university collaboration. There is no arm's-length supplier to push against in the usual way, and no manufacturer's catalog to specify from, because the panel system is still being developed by the partnership as it goes.

That eases a familiar friction. The architect now has to design the panel system and the house at the same time, knowing that an elegant plan move can quietly demand a jig, a press, or a lifting capacity the shop does not have and cannot justify buying for a single prototype.

The Deadwood Innovations system uses nail-laminated timber panels made from stud lumber milled from Nak'azdli forest licenses, fabricated over the winter at a repurposed finger-joint plant in Fort St. James, and raised on site in a matter of days. The architectural work expanded from designing individual panels to integrating a whole housing system around what that plant and its crew could produce.

[Read more about the Nak'azdli prototype here]

The tenth edition of North America's largest industrialized construction conference framed prefab as a core business strategy rather than a niche efficiency play, with tracks spanning owner playbooks, fab-shop optimization, digital workflows, and field management. The community-scale model sits awkwardly within that mainstream, which still assumes large factories and developer clients.

[Read more about the prefabrication conference program here]

Oregon's DLCD worked with ten small jurisdictions to rewrite local development codes for mass timber and modular housing, an explicit acknowledgment that prefab adoption stalls at the municipal counter. For community-owned producers without legal departments, this kind of code translation work is what makes a panel shippable across a county line.

[Read more about Code-UP here]

The City of Cleveland issued an RFP in late 2024 to attract a modular manufacturer to a 20-plus-acre site, using public demand-pooling to underwrite factory investment. A contrast with the Nak'azdli model is that one path scales by concentrating production for external developers, while the other keeps production small and locally owned.

[Read more about the Cleveland RFP here]

If you are entering a project in which the client owns the supply chain, there are notable shifts in design practice worth preparing for.

One is sequencing. With a commercial manufacturer, you design toward a catalog of proven panels. With a small, fixed, local line, the order can reverse: the question of what the shop can produce and repeat, with the equipment it has and a crew still learning the process, shapes the plan rather than following it. Geometry that a large factory would absorb without comment (long spans, deep cantilevers, one-off openings etc.) can quietly require tooling a small shop cannot justify for only a handful of houses. Expect to test design moves against shop capability early, and to treat the panel module as a dimension that governs the plan.

Another difference is documentation. A small shop often starts without a digital design-to-production workflow, and someone has to build one. In the Nak'azdli Whut'en project, that capacity came largely from outside the design contract: a UNBC master's student developed the digital workflow with Mitacs support, and the development corporation won an $800,000 federal DIGITAL grant to advance the work. The lesson for an architect is to find out early where that capability will come from, whether a research partner, a grant, or your own scope, because the deliverable that lets the shop build a second house without redesigning the first- the file structure, the naming conventions, the cut lists- is real work that has to be scoped, resourced, and paid for. This requires careful budgeting in the supply chain, a territory normally unfamiliar to an architect.

An important factor is the model's limits. The Nak'azdli build is a prototype, deliberately ambitious, meant as a learning exercise for crews who had never built prefab. It has not yet been tested at the volume that would stress the shop's quality control, nor has it been tested across the code variations that arise when panels cross jurisdictional lines. The partners talk about scaling to ten or more homes a year and doubling or tripling output over time, with a second house already in design. That is the right ambition, and it is still ahead of them. Tall mass timber has been proven elsewhere, with a twenty-five-story tower already built; community-scale nail-laminated housing produced at steady annual volume has not. Honest practice means saying so to the next client who asks for the same model and designing the second project with the first's failure modes in view.

The consequences of the contract are concrete. Fee structures built around schematic, DD, and CD phases do not map cleanly onto a commission where much of the early work involves co-developing a panel and a production process with the client's team. Price the system work as system work, name the production documentation as a deliverable in its own right, and protect the boundary between designing the house and designing the operation behind it. The community has a stake in the sawmill. The architect owes them a system that the next crew, and the one after that, can still build from.

(Image credit: Mitacs)

See you next week!

-Johan

[email protected]

architectonic.io