Assessing biofiltration repeatability: statistical comparison of two identical toluene removal systems.

Biofiltration of volatile organic compounds is still considered an emerging technology. Its reliability remains questionable as no data is available regarding process intrinsic repeatability. Herein, two identically operated toluene biofiltration systems are comprehensively compared, during long-term operation (129 days). Globally, reactors responded very similarly, even during transient conditions, with, for example, strong biological activities from the first days of operation, and comparable periods of lower removal efficiency (81.2%) after exposure to high inlet loads (140 g m(-3) h(-1)). Regarding steady states, very similar maximum elimination capacities up to 99 g m(-3) h(-1) were attained. Estimation of the process repeatability, with the paired samples Student's t-test, indicated no statistically significant difference between elimination capacities. Repeatability was also established for several descriptors of the process such as the carbon dioxide and biomass production, the pH and organic content of the leachates, and the moisture content of the packing material. While some parameters, such as the pH, presented a remarkably low divergence between biofilters (coefficient of variability of 1.4%), others, such as the organic content of the leachates, presented higher variability (30.6%) due to an uneven biomass lixiviation associated with stochastic hydrodynamics and biomass repartitions. Regarding process efficiency, it was established that less than 10% of fluctuation is to be expected between the elimination capacities of identical biofilter set-ups. A further statistical comparison between the first halves of the biofilter columns indicated very similar coefficients of variability, confirming the repeatability of the process, for different biofilter lengths.