Principal component analysis (PCA) may reduce the dimensionality of plant m
odels significantly by exposing linear dependences among the variables. Whi
le PCA is a popular tool in detecting faults in complex plants, it offers l
ittle support in its original form for fault isolation. However, by utilizi
ng the equivalence between PCA and parity relations, all the powerful conce
pts of analytical redundancy may be transferred to PCA. Following this path
, it is shown how structured residuals, which have the same isolation prope
rties as analytical redundancy residuals, are obtained by PCA. The existenc
e conditions of such residuals are demonstrated, as well as how disturbance
decoupling is implied in the method. The effect of the presence of control
constraints in the training data is analyzed. Statistical testing methods
for structured PCA residuals are also outlined. The theoretical findings ar
e fully supported by simulation studies performed on the Tennessee Eastman
process.