We examine an ensemble of 48 simulated clusters to determine the effects of small-scale density fluctuations and large-scale substructure on X-ray measurements of the intracluster medium ( ICM ) mass .
We measure RMS density fluctuations in the ICM which can be characterized by a mean mass-weighted clumping factor C \equiv \langle \rho ^ { 2 } \rangle / \langle \rho \rangle ^ { 2 } between 1.3 and 1.4 within a density contrast of 500 .
These fluctuations arise from the cluster history of accretion shocks and major mergers , and their presence enhances the cluster ’ s luminosity relative to the smooth case .
We expect , therefore , that ICM mass measurements utilizing models which assume uniform density at a given radius carry a bias of order \sqrt { C } \approx 1.16 .
We verify this result by performing ICM mass measurements on X–ray images of the simulations and finding the expected level of bias .

The varied cluster morphologies in our ensemble also allow us to investigate the effects of departures from spherical symmetry on our measurements .
We find that the presence of large-scale substructure does not further bias the resulting gas mass unless it is pronounced enough to produce a second peak in the image of at least 1 % the maximum surface brightness .
We analyze the subset of images with no secondary peaks and find a bias of 9 % and a Gaussian random error of 4 % in the derived mass .