Burning of liquid pools and wood cribs in large fully developed timber compartment fires

The contribution of exposed mass timber to compartment fire dynamics is often framed by prescribed fuel load density, implicitly assuming fuel-agnostic thermal feedback. This paper interrogates that assumption experimentally using full-scale cross-laminated timber (CLT) compartments with two exposed surfaces (ceiling and side wall) and a movable fuel; either a kerosene pool or a wood crib. High-resolution measurements of heat flux, in-depth timber temperatures, burning rates, opening flows, and gas species demonstrate that fuel chemistry and geometry strongly modulate oxidizer delivery and residence time, thereby governing radiative feedback, CLT burning rates, and external flaming. The pool fire exhibited pronounced radiative enhancement and transient oxidizer starvation near the timber until pool decay. In contrast, the crib burning rate was inhibited, while the CLT burned efficiently. Analysis of the Global Equivalence Ratio (GER) and air bypass ratio revealed significant excess oxygen in the outflow, indicating that entrained air did not permeate the crib but instead oxidized the CLT, leaving unreacted air due to short mixing timescales. Despite unchoked doorway conditions, the crib fire produced bypass ratios and external flaming fractions comparable to the pool fire. The work shows that fuel load and ventilation factors alone are insufficient to describe the mass timber compartment fire dynamics and the CLT performance (e.g. charring). The fuel chemistry, geometry and placement have significant levels of interaction with the compartment geometry. Therefore, the role of the moveable fuel is fundamental, and care must be employed when extrapolating demonstrator experiment results to the fire safety design of mass timber buildings. • Burning rates of pools, wood cribs and exposed timber in large compartment fires • Mass balance in a large timber compartment fire quantified • Entrained air can bypass the fuel leaving unreacted oxygen in the outflow • Exposed timber burning rates are dependent on local flow and oxidizer field • Fuel load density alone is insufficient to characterize timber compartment fires