"Control structure design for complete chemical plants"


Response to reviewer A:

1. It is correct that the literature review is rather brief, because I have
referred to the paper by Larsson and Skogestad (2000). Since that paper
was published in a somewhat "obsscure" I have now added a web address where
the reader may get the paper.

2. The discussion about degrees of freedom has been reduced as suggested
by the reviewer.

3. The discussion on the snowball effect has been removed as suggested by the reviewer.

4. I agree that a more thorough example would be good, but this will have
to wait for later work.


Response to reviewer B:

page 4, Table 6:  Some comments on the effect of nonlinearity have been added,
as suggested by the reviewer.

page 6: It seems the reviewer is not quite consistent. First, he seems to
indicate that a single big multivariable controller may be used, but then
he goes on to state that stabikizing control of the plant needs to be 
implemented in the regulatory control layer. I think we agree, and in the
paper I discuss how to design the regulatory control layer. Some comment and
definition of complexity has been added to make this discussion more rigorous.

page 13: I do not agree that level control shouldbe part of the supervisory
control system. The reviewer is probably talking about control of levels
in surge and buffer tanks, where I agree with him. In any case, the supervisory
control layer may generally have access to the setpoints for the levels 
controlled in the regulatory layer.

page 22: I think I agree with the reviewe here. The issues about
reliability and transparent are towards the bottom of my list of
negative points for MPC. I have added a "may be" to meet the objections
of the reviewer.