Earlier this summer, BUILD’s Andrew van Leeuwen traveled to Lyons, Oregon, for a fascinating tour of Freres Engineered Wood’s production line, talking with Kyle Freres, Justin Harries, and Robert Rumplik along the way. They discuss the engineering that sets veneer-based mass timber apart from lumber CLT: fire performance, panel strength, and CNC machinery built for precision. The Freres team shares insights on code barriers, the use of wildfire-damaged logs, and three generations of family history. The future of American mass timber depends not on proving the product works, but on building the domestic capacity and code acceptance to build with it at scale.
How far back does your family’s history go in the lumber business, and growing up, did you always assume you’d return to the family business?
Kyle Freres: I’m third generation. My grandfather started in 1922 with lumber, moved to a facility in Lyons, Oregon around 1942, then switched to veneer in the early ‘50s. I worked elsewhere for a while, but yes, I always knew.
Why the switch from conventional lumber to veneer?
KF: The appeal of veneer is recovery. The best sawmills get around 50% recovery turning round logs into conventional boards; the rest becomes chips and sawdust. In our veneer process we get over 70% recovery, because we’re unwinding the log like a roll of paper towels rather than sawing it apart.
How small a log can you actually peel?
KF: We can peel any block over five inches in diameter.
How many people work at Freres Engineered Wood?
KF: Company-wide, including maintenance, sales, and everyone else, we’re at about 490 people. We have about 50 mechanics/mill-service staff company-wide: roughly 40 millwrights and 10 electricians.
How has the business model evolved over the years?
KF: Until about 2017 we were basically a commodity producer: we harvested logs, peeled veneer, sold most of it, and used some internally to make plywood panels. We still do all of that, but this facility lets us take the higher density-graded veneer and turn it into engineered wood products such as Laminated Veneer Lumber (LVL) beams and columns, rim boards, LVL studs, as well as thicker, large-scale beams. We don’t compete on high-volume commodity LVL, like I-joist flanges, but on the boutique, project-specific side (often making products for specific projects rather than stocking a catalog), we can do specialized things that the big commodity producers can’t.
We also run a co-generation plant, where the offcuts get ground up and burned for fuel. That process also yields a biochar byproduct.
Justin Harries: This biochar byproduct can be used for agricultural applications and for carbon sequestration.
How does the Laminated Veneer Lumber (LVL) product you make relate to Cross Laminated Timber (CLT)?
KF: Cross-Laminated Timber is certified under Performance Rated Glulam (PRG) 320, the official rulebook that says how cross-laminated timber panels have to be made and tested so builders and inspectors can trust they’re strong and safe enough to use in buildings. Anything meeting that standard is technically CLT, whether it’s lumber-based or, like ours, veneer-based. We use the term “mass-ply” partly to differentiate from plywood, since this is a massive, engineered piece of wood, and partly because people expect to hear “CLT” and we want a term that signals it’s veneer-based.
JH: The cross-laminated timber process was invented in Austria and I worked for an Austrian producer before joining Freres. The U.S. market is much newer to this than Europe, which has had a 20-year head start. Once people saw Mass Ply CLT was stronger, could go thinner, came in wider panels, and wasn’t imported, they started asking why architects were still specifying CLT from other countries when a domestic mass-ply alternative existed.
How does mass-ply perform in the fire tests as compared to CLT?
KF: Unlike a lumber-based CLT, which alternates boards at 90 degrees layer by layer, we build the cross-lamination into every one-inch layer. That means if a lumber CLT panel burns through its outer layer in a fire, it loses two structural plies at once, one side and the adjacent orthogonal layer of the panel. Since our layup is homogeneous through the thickness, we get consistent structural performance as the panel chars. We’re also denser (about 37 lb/ft³ versus roughly 32 lb/ft³ for dimensional lumber), so we char slower than lumber, even though our glue content is higher.
How does mass-ply perform for strength as compared to CLT?
KF: Because our plies are so thin, we have a lot of flexibility in how we lay them up. Standard panels alternate cross-bands like traditional lumber CLT; thinner panels might use only one or two cross-bands to improve span rating in the primary direction; and for two-way slab designs, where the panel itself has to act like a beam in both directions, we lay it up to get strength both ways.
Does the process incorporate different types of trees?
KF: We only use Douglas fir because it’s the dominant species here in the Pacific Northwest, and it’s arguably the superior structural softwood.
The Freres mass-ply panels are left exposed on many projects. Is it possible to specify the product in a cabinet-grade finish?
KF: We could attach a surface panel if someone asked, but we don’t usually recommend it: once a structural product goes in early in construction, it’s hard to protect a cabinet-grade finish throughout the rest of the build. What we do offer is a range of architectural appearance grades: knot-free or tight-knot finishes, a traditional board-look finish, and an edge-grain finish that reads almost like bamboo. I wouldn’t call any of it cabinet-grade, but it covers most of what architects are looking for.
Walk me through the process here at the Freres mill, start to finish.
KF: We peel veneer and sort it by density grade, lay up 4×8 panels by density (essentially engineered plywood), then cut a joint into a panel about an inch thick and scarf-joint those together end-to-end until we have a long piece of LVL. We rip those to width, sand them, and use those pieces to build our larger panels. It’s the same idea as lumber CLT finger-jointing boards for length; we’re doing the same thing with LVL laminations instead of 2×6 boards.
JH: Those laminations are ripped down to about 9.5-inch widths so that when they come together the joints are tight, and they’re a genuine structural layer, not a cosmetic veneer.
KF: A robot picks and places the panel sections, glues the joint, and feeds them either straight into production or into inventory. We stagger the seams so no seam ever lines up directly over another: a 4-foot piece next to an 8-foot piece next to a 2-foot piece, alternating, so the panel is never structurally weak along one line.
JH: The joint uses a radio-frequency press, which essentially microwaves the glue line to cure it; it can process four joints at once rather than one at a time. The press was purpose-built for this process.
What glues are used in the process?
KF: Two types. A phenol-formaldehyde resin on the primary glue line, accepted under the California Air Resources Board (CARB) and other formaldehyde-emissions standards; more formaldehyde actually off-gasses from the wood itself than from our resins. Our secondary glue line is a melamine resin with a formic-acid catalyst, a cold-press system. It’s held up well and actually performs better in fire testing than a lot of the resins used in Europe.
What are the mass-ply size limitations?
KF: Panels currently run up to 48 feet 6 inches long out of the press, and we have the ability to extend that to 60 feet in the future if it makes sense to add sections to our press. Practical shipping width tops out around 10–12 feet before wide-load permitting gets complicated. We’re certified in one-inch increments from 2 inches up to 12-plus inches, versus lumber products that are stuck at fixed thicknesses (1.5″ or 3″ board dimensions). That extra flexibility can save a meaningful amount of material. For example, where a 5-ply lumber CLT floor section might run about 7 inches, we can typically hit the same performance in less material.
Sometimes structural engineers specify a camber into a beam, an intentional curve that allows a beam to sit flat once a structural load is imposed on it. Is it possible to build this into a mass-ply panel?
KF: Not really. Mass-ply beams are typically used on edge, so camber usually isn’t needed. We can cut visual curves for something like an exposed roof member if an architect wants that look.
When clients are comparing concrete, steel, and mass timber on cost, what’s the pitch?
JH: The building is lighter, so foundation work (footings, concrete) is significantly less, which is a big cost saving. You also get one specialty erection crew instead of multiple trades sequenced around concrete cure times, and that crew can be roughly 30% smaller. Panels come pre-cut, so you’re flying pieces into place rather than forming and pouring.
KF: A client of ours in Oakland, California estimated about a 30% reduction in overall project schedule versus concrete construction, and that translates directly into savings on both money and financing time. We built a 50,000-square-foot warehouse in six weeks using this approach.
What kind of dimensional tolerance do you work to, and does that ever create friction with other trades?
KF: Our CNC-cut (Computer Numerical Control) panels hold to about a sixteenth of an inch. The friction point is the interface with anything poured on site: concrete isn’t held to that tolerance, so we build in wiggle room, often a quarter to half an inch, to absorb variation in the concrete.
JH: You also have to allow for “building creep” (long-term movement), so panel-to-panel gaps are usually at least a sixteenth of an inch, ideally closer to a quarter inch.
What insulation R-value do you get out of these panels?
JH: About R-1.5 per inch with mass-ply.
KF: That number doesn’t capture the fact that there’s essentially no air infiltration through a solid mass-ply panel, and no thermal bridging the way you’d get through steel or concrete connectors, so real-world performance is better than the R-value alone suggests. For comparison, rigid foam insulation runs roughly R-5 per inch.
JH: That thermal mass and lack of bridging are why we’re seeing traction in refrigerated/climate-controlled logistics and industrial buildings, including early interest from data centers, where temperature control matters.
Are panels being used for elevator or stair shafts instead of poured concrete cores?
JH: Yes. In non-seismic regions we’re seeing real traction using our panels as elevator and stair shafts, since a concrete core is usually the long pole in the schedule due to cure times.
What code or jurisdictional issues would you still like to see resolved?
KF: Fire classification is still the big one. We’re categorized as combustible even though we can meet the required fire performance, so on a lot of high-rise projects, even where we could otherwise do a floor every day or two, jurisdictions require encapsulating the wood in gypsum every five stories. That forces contractors to slow down and cover up a product that’s meant to be left exposed, which adds real cost.
JH: It has gotten better: the latest code cycle is more favorable to exposed mass timber because there’s more test data now and more flexibility with this building type.
How do U.S. building codes compare to Europe’s?
KF: Europe’s market is simply more mature; they’ve been doing this for decades longer, so there’s more acceptance. The U.S. market has stricter, more fragmented code requirements; it ends up being state- or even city-specific, so we’re sometimes negotiating specifics with a building department in San Francisco or Oakland even when a design is already accepted under the international codes.
Do you have staff that work directly with code officials or the International Code Council (ICC)?
KF: Less with the ICC directly: we hold an ICC certification mainly for the LA area, but PRG 320 certification and product reports certified by APA, the Engineered Wood Association, cover most jurisdictions. WoodWorks is the organization we lean on most; they’re good at educating code officials and showing acceptable fire-rated assemblies. In the end it’s the engineer of record running the calculations and stamping the design who satisfies code for a given building.
How involved is Freres with the design of the connections between mass-timber components?
KF: That’s really the engineer’s and fabricator’s scope. We can advise on panel-to-panel connections, but beam-and-column connection design and fabrication for steel hardware usually goes to a third party, and it often comes down to how the architect wants the connection to look.
JH: We do offer design-assist: from a napkin sketch we can model the whole building for a budgetary material takeoff, then share that model with a fabrication partner, who reviews it and prices the connection hardware. Hardware often gets pre-installed at the factory so the job-site crew is really just assembling pre-made pieces; a lot of the connection tolerances, like angle and torque spec, are easier to hit in a shop than in the field.
What is the tallest mass-timber project you’ve manufactured panels for?
KF: The tallest project we’ve done so far is 18 stories of wood over 7 stories of concrete, a 25-story building in Oakland. We’ve got a couple more projects coming later this year at around 10 stories of wood. Most of our recent stock has gone to warehouse/office space and some LA-area residential projects, plus work with a couple of San Diego developers.
What are you developing next with mass-ply, and are there markets you’d like to get into?
KF: Industrial projects are a big focus right now; they’re also a stepping stone into data centers. We have a fairly standardized industrial building design for our product that’s been successful; it’s not the flashiest application compared to a signature building, but it’s a great fit for what we do. The pitch starts with the environmental story, carbon sequestration and renewability, but the bigger misconception is cost: people assume wood construction costs more. We built our own facility specifically to prove we’re cost-competitive with concrete tilt-up or pre-engineered metal buildings.
Do you produce mass-ply for one-off custom residences?
KF: All the time. A lot of it comes through word of mouth for a specific structural need; we’ve done a lot of roof sections for high-end homes in Jackson Hole because of the heavy snow loads there. We’ve also laid an entire floor system for a home in a day with no floor joists, just intermediate supports every six to eight feet.
If I understand correctly, fire-damaged/charred logs can be used for mass-ply.
KF: Yes. The 2020 Labor Day fires here burned over 200,000 acres, and about 5,000 of those acres were on land we own. In 2021, over 70% of the logs we processed were salvaged. The biggest constraint with fire-salvaged logs is time. By roughly the two-year mark, the wood has degraded too much to process: it starts rotting, picks up insect damage, and develops sun-checks. So, there’s a hard deadline built in, and the threat of a single environmental lawsuit can halt a salvage sale entirely, even though we’re required to replant regardless.
Where did this material get used?
KF: These salvaged logs were used at Portland International Airport (PDX) designed by ZGF: the roof structure used a 2-inch panel across a 9-acre span, and roughly 70% of the wood in that roof was salvaged from the 2020 Labor Day wildfires.
JH: Because the panels are so dense, we can design point-supported floor plates on a 16-foot grid, no beams, just columns and a solid plate. That’s exactly how the roof structure works at the oWow project in Oakland, California.
Why would a lawsuit target the salvage of fire-damaged trees?
KF: Some environmental groups oppose any post-fire logging on the theory that further disturbance harms the ecosystem. Our view is that leaving the dead material in place just fuels the next fire, delays getting the land back into productive forest, and costs communities the timber-sale revenue that many rely on.
Before the Northwest Forest Plan in the 1990s, local counties were largely funded by revenue-sharing from federal timber sales. The federal government doesn’t pay property tax on the land it holds, so a share of timber-sale revenue went to counties instead. Federal harvest volume dropped from about 5 billion board feet a year to roughly 500 million, a 90% drop, and those counties lost that funding almost overnight.
That urgency shows up in our own replanting numbers: in a typical year we plant over 300,000 seedlings, but in the year after the Beachie Creek fire tore through Oregon’s Santiam Canyon over Labor Day in 2020, we planted 3 million seedlings.
What are the current challenges for the business?
KF: Mostly market-related rather than product-related. Mass timber demand has been soft the past couple of years because projects are highly sensitive to financing costs, and interest rates rose sharply coming out of COVID.
Imports have been another issue. Coming out of COVID in 2021, a lot of product moved at elevated prices, which encouraged more imports into the U.S. Tariffs came in last year to address that, and we support the idea in principle, since it helps with dumping. But the rollout left a gap: for roughly six months (around March to October), imports were still allowed in ahead of the tariff taking effect, which flooded the supply chain right before the change.
There’s also a structural loophole working against us. Canada subsidizes its wood products industry, and the existing softwood lumber agreement only covers dimension lumber boards, not glulam, CLT, our mass-ply panels, or pallets. That means a Canadian producer can turn raw lumber into a glulam or a pallet and ship it into the U.S. with no tariff protection at all, which puts us at a real disadvantage on the engineered-product side.
Is anyone in Europe doing mass-ply the way you are, and why not pursue the patent there?
KF: Not that we’re aware of. Companies like Stora Enso have looked at it and there’s clearly interest, but our patent only covers North America, Australia, and New Zealand. If a European producer started shipping panels into the U.S., our patent would apply; producing it in Europe for the European market, it wouldn’t. Our patent attorney told us it’s much harder to get comparable protection in Europe and South America. We also got about 90% of the way through a Japanese patent application before it was denied; we think Japan viewed it as a threat to its domestic industry.
Any plans to build a second facility elsewhere in the U.S.?
KF: We’d like to. A comparable plant in the Midwest or East Coast would give us regional market access, since wide panels are hard to ship cross-country but our smaller 4×8 “power panel” format already ships everywhere. Doug fir isn’t available in those regions, so we’re working with a partner to get a Southern Yellow Pine version of the product certified, or we could keep shipping finished West Coast product east, which we already do by rail. I’d like to see it happen before I retire, but you never know.
Any shipping challenges getting product down to the Bay Area or LA, and is truck width or length ever a limiting factor?
KF: By truck it’s manageable. Concrete trucks are heavier, so this product actually means fewer truckloads to a job site, but the panels are large, so you need adequate laydown space or a fast pick-and-place plan to get panels off the truck quickly. Width is the real constraint, not length: eight feet is easy, ten is fine with a flag escort, but once you’re over 12 feet you start needing pilot cars and permitting gets more complex.
Do you ship by boat to places like Hawaii?
KF: We’re actually working on that right now. The obstacle is termite treatment. A dip treatment works but has to happen on a 4-foot basis, which limits how wide a panel you can treat and ship. Lumber-based CLT has a glue-line termite treatment option that only works with thin (roughly eighth-inch) veneer sheets, like ours, so if we can get termite treatment into our glue line, we may be able to treat a full 8-foot panel without dipping it. We’re also working with a dip-treatment company on expanding their tank size to handle an 8-foot panel.
Tell me about the giant CNC machines at the end of the production line.
Robert Rumplik: We have two CNC machines. One, made by the German manufacturer Weinmann, has been in place since the facility was constructed in 2017. The second is an Italian-made SCM machine installed a couple of years ago. It runs two independent spindles on a tandem bridge, so when the same operation, drilling or cutting, is needed on both halves of the split table at once, each spindle handles one side simultaneously, roughly doubling throughput on that work. It carries over 60 tool positions, including four saws: two 800mm blades and two 1150mm blades, each pair run together. The table itself is split into two zones, so we can load and cut a new billet in one zone while finishing the previous one in the other, when billets are short enough to allow it.
What is the level of accuracy on a machine like this?
RR: Accuracy is very tight. We float within a sixteenth to an eighth of an inch across a 48-foot billet. The machine was calibrated once at install and hasn’t needed it since; we run a quality-control check every morning on the first billet of the day, checking all dimensions against the shop drawings.
What do you do if/when the machine goes down?
RR: We have a second CNC as backup; we added it after the first one broke down for four or five weeks during the PDX Airport project. Day-to-day maintenance, like greasing and cleaning rails and spindles, is done in-house. For bigger issues, a technician overseas can log into the machine remotely to help diagnose problems; for major repairs, a technician flies out.
KF: It’s a bit of an unfortunate reality that most of this specialized equipment (the presses, the big CNCs) comes from European manufacturers who’ve been doing this much longer. If we built this plant again today, we could probably source almost everything else domestically, but not the specialty gear.
What does a machine like this cost?
KF: Around $2.4 million, not including shipping and installation.
JH: Nothing else in the plant approaches that. Most other equipment is under $1 million; it’s really the scale of this machine that drives the number.
What’s the best “mistake” you’ve made on this machine?
RR: I was training another operator who cut a file incorrectly, and I was able to edit the cut file live to salvage the piece rather than scrap the billet. The machine has a visualization step where you can watch the cut plan before it runs and catch problems, and if something’s off you can correct the offset in real time, in millimeters, and keep going. It was a great lesson in on-the-fly problem-solving.
What’s the reward for you in doing this work?
RR: I take real pride in this. Driving past PDX Airport, or through Oregon State University or the University of Oregon, and knowing I helped cut the wood in buildings there, that’s a good feeling.
KF: We’re providing shelter, which is a basic need, at a time when the U.S. has real housing shortages, and we’re doing it with a lower-carbon substitute for concrete. I don’t buy the idea that slowing down technological progress is the responsible path here. I think the industries working on these problems are the ones that are actually going to solve them.
Kyle Freres leads Business Development at Freres, where he’s helped guide the family company for over two decades since returning in 2001. A Southern Methodist University graduate in Industrial Engineering and Mathematics and former PricewaterhouseCoopers consultant, he brought that large-scale systems background home to lead projects like the Evergreen BioPower co-generation facility and the Mass Plywood Panel (MPP) plant, and remains focused on the technology reshaping both the mill and the future of construction.
Justin Harries leads Business Development at Freres, bringing more than fifteen years in the building materials and engineered wood industry, including stints at Barn Pros and Montana Timber Products before five years as Business Development Manager USA for Binderholz Group, the Austrian mass-timber producer. Since joining the company in 2024, that firsthand look at Europe’s head start on cross-laminated timber has shaped how he helps architects and developers understand what sets Freres’ mass-ply panels apart.
Robert Rumplik has worked for Freres in a variety of capacities since 2005. Robert worked at the Freres Plywood facility laying up plywood until moving to the Mass Ply CLT facility in 2020 as a CNC Operator. He is an avid participant in CrossFit and a key member of the Mass Ply team.