Fires of practical interest are often large in scale and involve turbulent behavior. Fire simulation tools are often utilized in an under-resolved prediction to assess fire behavior. Data are scarce for large fires because they are difficult to instrument. A helium plume scenario has been used as a surrogate for much of the fire phenomenology (O'Hern et al., 2005), including buoyancy, mixing, and advection. A clean dataset of this nature makes an excellent platform for assessing model accuracy. We have been participating in a community effort to validate fire simulation tools, and the SIERRA/Fuego code is compared here with the historical dataset. Our predictions span a wide range of length-scales, and comparisons are made to species mass fraction and two velocity components for a number of heights in the core of the plume. We detail our approach to the comparisons, which involves some accommodation for the uncertainty in the inflow boundary condition from the test. We show evolving improvement in simulation accuracy with increasing mesh resolution and benchmark the accuracy through comparisons with the data.