Powdered activated carbon (PAC) for organic micro-pollutant (OMP) removal can be applied effectively on wastewater treatment plant (WWTP) effluents by using re-circulation schemes, accumulating the PAC in the system. This technique is complex because several factors are unknown: (i) the PAC concentration in the system, (ii) specific and average contact times of PAC particles, and (iii) PAC particle loadings with target compounds/competing water constituents. Thus, performance projections (e.g. in the lab) are very challenging. We sampled large-scale PAC plants with PAC sludge re-circulation on eight different WWTPs. The PAC plant-induced OMP removals were notably different, even when considering PAC concentrations in proportion to background organic sum parameters. The variability is likely caused by differing PAC products, varying water composition, differently effective plant/re-circulation operation, and variable biodegradation. Plant PAC samples and parts of the PAC plant influent samples were used in laboratory tests, applying multiples (0.5, 1, 2, 4) of the respective large-scale "fresh" PAC doses, and several fixed contact times (0.5, 1, 2, 4, 48 h). The aim was to empirically identify suitable combinations of lab PAC dose (as multiples of the plant PAC dose) and contact time, which represent the PAC plant performances in removing OMPs (for specific OMPs at single locations, and for averages of different OMPs at all locations). E.g., for five well adsorbing, little biodegradable OMPs, plant performances can be projected by using a lab PAC dose of twice the respective full-scale PAC dose and 4 h lab contact time (standard deviation of 13 %-points).