Fire safety is the first question engineers ask about tall wood buildings, and the answers increasingly come from full-scale tests rather than lab-scale assumptions. The American Wood Council and its research partners completed five full-scale mass timber fire tests in a multi-story apartment building, testing exposed and protected cross-laminated timber under realistic fire scenarios. The results feed directly into the sprinkler systems, fire alarms, and passive fire protection requirements that decide whether mass timber gets approved for taller structures.
How Mass Timber Responds to Fire
Wood burns, but thick engineered timber burns predictably. When a fire contacts a mass timber surface, a layer of char forms and insulates the wood below it, slowing heat transfer and preserving structural capacity. That char layer is the reason engineers talk about mass timber fire resistance in terms of section loss rather than collapse.
Charring and Self-Extinguishment
The test program’s most striking result came from exposed timber. Once apartment furnishings and contents were consumed, the exposed cross-laminated timber essentially self-extinguished because the protective char surface blocked further combustion. Engineers specifying mass timber systems in mixed-use building construction treat this behavior as a design input: char depth over a given fire duration, and the residual section that continues to carry load.
Char forms at a predictable rate for a given timber species and density, which lets engineers calculate how much section remains after a design fire. In cross-laminated timber, each layer’s grain orientation affects how the char layer behaves, and the adhesive bond between layers must hold so the char does not fall away and expose fresh wood.
The Five Test Scenarios
The International Code Council Ad-hoc Committee on Tall Wood Buildings defined five fire scenarios, each run in two one-bedroom apartments built from mass timber. The scenarios varied exposed and unexposed cross-laminated timber and included open doors between living and sleeping areas:
- Fully protected structure with gypsum wall board, large furnishings fire
- About 30 percent of the CLT ceiling exposed in living room and bedroom
- Parallel CLT walls exposed, one in each room
- All mass timber surfaces exposed with a normally operating sprinkler
- All surfaces exposed, fire allowed to grow before sprinkler water arrives
Each apartment included open doors between living and sleeping areas, a configuration that tests how a fire spreads through an actual floor plan rather than an idealized test cell. Furnishings and contents provided realistic fuel loads, so the tests measured real fire behavior instead of worst-case design curves.
What the Test Results Showed
Results across the five scenarios were consistent. Protected timber stayed protected, exposed timber self-extinguished once contents burned out, and sprinklers controlled the fires they faced. Real projects are already pressing these findings into code debates, including a hybrid mass timber building project that advanced Seattle building code.
Protected vs. Exposed Timber
In Test 1, the structure fully protected with gypsum wall board was subjected to a large furnishings and contents fire. The test was terminated after three hours without significant charring on the protected wood surfaces. Encapsulation, the practice of covering timber with gypsum, remains the simplest way to keep mass timber out of the fire entirely.
Results at a Glance
The table below summarizes the five scenarios and their outcomes.
| Test | Exposure | Duration | Outcome |
|---|---|---|---|
| Test 1 | Fully gypsum protected | 3 hours | No significant charring |
| Test 2 | 30% CLT ceiling exposed | 4 hours | Self-extinguished after contents consumed |
| Test 3 | Parallel CLT walls exposed | 4 hours | Self-extinguished with char layer |
| Test 4 | All surfaces exposed, sprinkler active | Contained | Single sprinkler controlled fire |
| Test 5 | All surfaces exposed, delayed water | 23 min growth | Sprinkler quickly controlled fire |
Test 2 ran for four hours, longer than Test 1, specifically to determine whether exposed CLT would contribute significantly to the fire. It did not: once furnishings were consumed, the exposed ceiling self-extinguished. Test 3 repeated the finding with exposed walls.
The results give code officials something they rarely have: repeatable, full-scale data from identical compartments. Test 2 and Test 3 produced the same self-extinguishment behavior with different exposed surfaces, which strengthens the case that the result is a property of the material system rather than a quirk of one layout.
Sprinkler Performance Under Real Conditions
Two tests isolated the sprinkler question. With all mass timber surfaces exposed, Test 4 demonstrated that a single sprinkler under normal operating conditions easily contained the fire. Test 5 pushed harder: the fire was allowed to grow in the compartment for 23 minutes before water was supplied, and the sprinklers still brought it under control quickly.
Suppression With a Single Sprinkler
The single-sprinkler result matters for design because it confirms that suppression does not depend on the entire system activating. Reliability comes from the whole chain, from fire pump systems design, installation, and commissioning to head placement and water supply.
Delayed Water Delivery
The 23-minute growth period in Test 5 simulates the realistic failure mode of slow detection or delayed response. Even after significant fire growth, the sprinkler system controlled the fire, which supports the case that suppression plus charring provides redundant protection in mass timber compartments.
For designers, the practical takeaway is that fire protection in mass timber buildings combines passive and active measures. Encapsulation keeps timber out of the fire, charring handles the exposure that does occur, and sprinklers control the compartment before either is pushed to its limit. No single measure carries the whole load.
Water supply planning for a mass timber building follows the same rules as any sprinklered structure, with the added expectation that the system performs through the full burn-out phase of exposed timber. Flow, pressure, and reserve capacity are sized against that longer demand window.
From Test Data to Building Code
The tests were designed to inform code change recommendations from the Ad-hoc Committee. The committee chair said the results will continue to be studied and will help shape those recommendations later in the year, describing the tests as part of the extensive research data used to validate the performance of tall wood buildings.
The Ad-hoc Committee Process
The International Code Council Ad-hoc Committee on Tall Wood Buildings reviews research from multiple programs, and this full-scale test series is one of the largest. Code change recommendations based on the results feed into the next edition of model codes, which means the tests influence regulatory language rather than just academic literature.
The committee drew its five scenarios from real building configurations, including open-plan apartments and the interior doors that compartmentation codes care about. Running the same scenarios in two apartments provided a check on repeatability, so a single favorable result could not carry the committee’s recommendation.
From Test Data to Code Language
Mass timber spans a family of products, from sawn lumber and glulam to cross-laminated timber and heavy timber construction, and code language must treat each product’s char behavior appropriately. Full-scale data gives code writers measured values for char rates, encapsulation requirements, and sprinkler interactions instead of conservative guesses.
The Research Partnership Behind the Tests
The tests were funded in partnership with the U.S. Forest Service’s Forest Products Laboratory and the American Wood Council, and conducted at the U.S. Bureau of Alcohol, Tobacco, Firearms and Explosives Fire Research Laboratory. The facility’s scale allowed real apartment compartments with realistic fuel loads, which is exactly what code committees need.
Who Funded and Conducted the Tests
Public and industry funding kept the program independent enough for code use. A General Technical Report, FPL-GTR-247, on the fire tests was scheduled for release from the Forest Products Laboratory, giving engineers and code officials a citable source for the findings.
Follow-up work will compare the char depths measured in these tests against the predictive models used in structural fire design. Where models and measurements agree, engineers gain confidence; where they diverge, the models get revised before they reach a code requirement.
Why Full-Scale Testing Matters
Full-scale tests answer questions that component testing cannot:
- How char forms on real ceiling and wall assemblies
- How open doors between rooms affect fire spread
- How sprinklers perform in actual apartment geometry
- How exposed timber behaves after contents burn out
- How mass timber compares with advanced materials such as fiber-reinforced polymers and smart materials
Five tests do not close the research program, but they add hard data to the evidence base for cross-laminated timber in tall buildings. Char behavior, sprinkler performance, and encapsulation all have measured answers now, and those answers flow into code language that builders and engineers can rely on. The next tall wood building gets designed with better data than the last one, and that is how a material earns its place in the code.