The aim of this paper is to evaluate several multivariable model-based cont
rol algorithms for controlling nitrogen removal in activated sludge process
es, We contend that the control of nutrient removal in activated sludge sys
tems is a multivariable control problem, rather than a multiloop problem. T
he work is a simulation study of a predenitrification activated sludge plan
t. Several model-based controllers are evaluated and compared against a bas
e-case and multiloop controller. We introduce a quantitative measure to aid
in controller evaluation. Model based control algorithms are shown to be a
ble to provide tight control of N removal, thereby offering significant ben
efits in terms of deferred capital expenditure. This is in return for incre
ased operating costs. (C) 1999 Elsevier Science Ltd. All rights reserved.