Full scale experience with biological process models - Calibration issues

The advent of computer-based modeling tools has greatly improved the ease w ith which biological process modeling can be implemented. However, the cali bration of a site-specific model does not always receive the same level of scrutiny as does model development. Three examples are presented to illustr ate how this can be detrimental to process design: (i) lack of solids mass balance closure can lead to erroneous SRT estimates: (ii) incomplete specif ication of COD partitioning can result in underestimation of plant solids l oading and corresponding under-design of clarification and aeration capacit ies; and (iii) inadequate determination of nitrogen species and COD partiti oning can hinder process design of nutrient removal facilities with stringe nt discharge criteria. Provision of these types of data will greatly facili tate model calibration and improve their utility. (C) 1999 IAWQ Published b y Elsevier Science Ltd. All rights reserved.