.SH
1.1 Installing new plant component models in ESP-r
.PP
It is possible for a user of ESP-r to develop a
new plant component model, in state-space form, and then
to install it within the system. The procedure is as follows
.IP 1.
Firstly the component is represented as one or more
finite volumes between which energy and mass can be exchanged.
.IP 2.
Then by formal energy and mass balance techniques a state-space
equation is established for each finite volume.
Examples of this are given in the principal reference of 6.1.
.IP 3.
The topology of the state-space matrix equation is then entered
to ESP-r's plant components' database along with the
component's describing data as required by the mathematical model.
.IP 4.
A subroutine must now be created according to a specified template
and installed in \fBbps\fR. Its function is to evaluate the
coefficients of the state-space equations, at each simulation time-step,
on the basis of control and component data organised
and supplied by \fBbps\fR).
.PP
While steps 3 and 4 are relatively straightforward, steps 1
and 2 will require a detailed
knowledge of component behaviour and thermophysical
properties (for example capacities and interface heat transfer
coefficients); this is often difficult to find in manufacturer's
catalogues and the like. One possibility is to treat the component
as a single finite volume. The model to result is then a simple
input/output relationship which can embody a model of internal
processes based on manufacturers data. For example, a single node boiler
model can be established in which the heat input is simply
a function of a published part load efficiency curve. In this case the
boiler will have a single capacity term so that a rudimentary account
is taken of boiler capacity.
