Stronger, Faster, Cheaper: The Future of Wood
Plus: a 50,000 ton forging press, a radio controlled crane hook, and a giant construction tent.
Welcome to Construction Physics, a newsletter about the forces shaping the construction industry.
Stronger, Faster, Cheaper: The Future of Wood
Wood is one of the most popular building materials in the US - of the 1.3 million or so homes built in the US each year, the vast majority of them are constructed with wood.
There’s often an impetus within the industry to move away from wood, and towards materials like cold-formed steel, which are stronger, less combustible, and more uniform. Outside the US, wood construction for single family homes is less common (though it varies from country to country), and wood construction (especially light framed) is often considered cheap, fragile, or unsafe (to the point where cities in the US have banned its use in large projects).
But it seems likely that the future of construction will be one of much more wood use, not less.
By almost any measure, Wood appears to be far better for the environment than other construction materials. Wood production has less embodied energy and less released CO2 than other materials on a unit-weight basis, and on a unit-floor area basis:
Wood is formed from fixing atmospheric carbon, and thus it acts as a carbon sink. A cubic meter of wood absorbs 582 kg of CO2, whereas producing a cubic meter of concrete emits 458 kg. This gives wood at worst a very small carbon footprint (if it’s burned at the end of its life), or even a negative carbon footprint (if it can be recycled).
And wood performs even better than these numbers suggest, because of where that energy comes from. 95% of the energy required for manufacturing lumber products comes from the kiln drying process, which brings the moisture content of wood below 20%. But this process can be fueled by burning the wood detritus left over from sawing the logs down into planks. 65% of a sawmills energy comes from burning this waste wood - an energy source that has no transport, logistics, or refining costs (since it’s right where the wood has been sawn) and is carbon neutral (since it’s just releasing the carbon the wood fixed in the first place).
Where producing more steel or concrete requires enormous excavations (and their environmental effects), wood literally appears out of thin air. Modern tree plantations are managed in a way that allows them to continue to produce wood products. Despite rising (until 2007) timber consumption, both timberland and total forest coverage has increased significantly, particularly over the last 40 years:
The same trend can be seen in Europe - developed countries show net reforestation despite using significant forest resources, due to good forest management, silviculture, and tree improvement. Because of its environmental benefits, countries like Spain or France, where wood is a less common construction material, are encouraging its use. Deforestation is largely a product of burning wood as a fuel source and clearing land for agriculture, not timber harvesting.
Wood being relatively environmentally friendly isn’t new. But what is new is the relative importance of that factor.
The Slow, Steady March of Tree Improvement
Wood is already the least expensive structural material, which is why it’s the material of choice for low-cost construction. But long term, we should expect the economics of it to keep getting better.
Other agricultural products have been cultivated and bred for desirable properties for thousands of years. Modern corn has over 16 times the yield per acre that teosinte, the plant that it evolved from, provides. A 2005 chicken is over 4 times the size of a 1957 chicken:
But breeding trees for desirable traits is a relatively young practice. Tree improvement programs in the US are largely a post WWII phenomenon - prior to that we were still mostly clearcutting forests. And it’s even less time than it appears, because of the length of time it takes a tree to reach useful maturity. Unlike other agricultural products (which have a rotation time of once or twice a year), even a quickly growing tree can take 25-30 years to reach maturity. Breeding is an iterative process, where the successful results of one generation become the seeds of the next generation, and we’re only in the third generation of tree improvement.
The upside of this is that we should expect that most possible tree improvement lies in front of us, despite the great gains that it has already made:
Continued tree breeding, along with genetic engineering and techniques such as clonal forestry (which may alone improve forest production by 50%) should continue to improve tree yields. Southern Pine takes 25-30 years to reach maturity, but Eucalyptus plantations reach maturity in just 6-8 years, suggesting there’s plenty of room for improvement.
So far tree improvement programs have been mostly focused on timber output, and haven’t focused on mechanical properties (in fact, it’s actually come at the expense of mechanical properties - several years ago there was a serious downgrade in southern pine strengths as a result of widespread use of younger trees). But we can imagine there’s quite a bit of potential here, as tree strength traits are highly heritable as any other (though this is complicated by the fact that some traits, like growth rate and strength, seem to trade off against each other, and by the fact that higher strength may reduce things like lightness or workability).
Wood is a highly variable material - design strength values are substantially lower than average strengths to ensure that the design strength is conservative. Even trees bred for increased uniformity, with no change in mechanical properties, would yield substantial benefits.
Much of current wood capacity is limited by defects in the wood - it’s actually possible to increase wood strength 2-3x just by either removing the defects, or distributing them evenly throughout the section. This is what engineered lumber does, suggesting it might be possible to substantially increase strength merely by breeding trees with fewer defects.
It could be argued that the development of higher strength materials haven’t changed the use of concrete and steel substantially. Construction steels are far below the strength of those used in the auto industry, and while high strength concretes are used for supertall high rises, for most buildings lower strengths are used.
But we might expect this to be different with wood. Wood often has to substitute in members of higher strength (engineered lumber, or steel members, or higher strength studs) when normal lumber is insufficient. And strength limitations often force the construction of truly enormous stud packs that could benefit from being reduced.
The Shifting Makeup of Building Construction
Broadly speaking, wood is currently the preferred material for residential construction under 6 stories tall, and concrete and steel are the preferred material for buildings taller than that. It’s hard to do high rise construction with timber (though that’s changing at the margins, it remains a high-cost alternative). As remote working (accelerated by COVID) gains in popularity, it should shift population (and thus construction) away from expensive, dense cities towards smaller, less expensive cities, or to areas more remote from the city center (maybe you decide to live close enough to the office to commute a few days a month, rather than every day). We’re already starting to see this happen. This means, on the margin, fewer highrises and more low rise apartments and single family homes.
Telecommuting also means people are spending more time in their homes, which means people will want, on average, nicer/larger homes (a nicer home becomes a lot more important if you’re spending all day in it). This means more new, large homes, and more custom homes - both of which also favor more wood. Conversely, it means fewer new office buildings, which tend to be steel or concrete.
Many of these trends reinforce each other. As wood gets stronger on average, it becomes easier to get the strength needed for higher grade lumber, making it cheaper, increasing demand. As tree output per acre improves, the size of tree plantations can decrease, making it easier to implement further improvements.
These trends exist in a sea of other trends, some pushing in the other direction. There are no guarantees, but there’s many potentially interesting developments.
 - There’s a strange double standard where European construction is praised for using more mass timber than the US, but our vastly higher use of timber for single family homes is met with skepticism.
 - Though obviously a tree farm is a much less rich ecosystem compared to an old-growth forest.
 - Though some don’t - as people leave dense cities, they become cheaper, making them more attractive on the margin.
Remote Controlled Hooks - Does what it says on the tin - crane hooks that disconnect via remote control (rather than being manually disconnected). Seems like it could significantly cycle time on the jobsite, particularly in areas with a lot of repetitive crane picks (such as precast, though the working load is probably too low for that application).
Site Cover - Giant enclosure you can place over your construction site. If you’re trying to improve construction efficiency, you have two options. You can move more of your operations to a factory, or you can make your jobsite more factory-like. Most flavors of construction innovation try the former, so it’s interesting to see solutions targeting the latter.
Alcoa 50,000 ton forging press - Enormous forging press that was built in 1955, and has forged parts for every current manned US military aircraft, and every airplane built by Airbus and Boeing. Things like this remind me of the extreme linearity in structural engineering - the same forces, concepts, and design solutions continue working from very small scales up through extremely large scales. If you need to mount a small piece of equipment, you might use a ½” bolt. If you need to mount a 16 million pound press, you’ll still use a bolt, it’ll just be 80 times as big.