Page 108 - Handbook of Electrical Engineering

AUTOMATIC VOLTAGE REGULATION 89

to the busbars, falls to near zero when the short circuit exists. The AVR may loose its supply during
this period or perform in an unpredictable manner. The excitation of the generator may collapse,
which is not desirable.

The pilot exciter method is highly reliable and has a fully predictable performance. A small
alternator is mounted on the same shaft, and often within the same frame, as the main exciter. It
receives its excitation from a shaft mounted permanent magnet rotor system. Hence its level of
excitation is constant and dependable. The AC output from the pilot exciter is rectified and smoothed
by components within the AVR cubicle. It can be seen that this method is completely independent of
the conditions existing in the main generator. This is the method usually specified in the oil industry.

4.2 IEEE STANDARD AVR MODELS

In order to standardise the modelling of AVR systems for computer analysis the IEEE, see Refer-
ence 1, has derived a set of block diagrams for the purpose. The model described above is called the
Type 2 and is the most frequently used. If a slip-ring connected main exciter is used then a Type 1
is appropriate.
In Figure 4.1 the block representing the generator shows a function G g . This function is a
complicated combination of the dynamic variables and time constants within the generator equations.
However, in the steady state the numerical value of G g as a gain term varies from 1.0 at no-load
where V is equal to V fd , to typically 0.365 at full-load and rated power factor. This variation G g
needs to be taken into account when the value of the AVR gain G a is established to give an overall
voltage regulation of 0.5% and zero reactive drop.

The saturation function for the main exciter is approximated by a simple exponential function
Bx
of the form, y = Ae ,where x is the output voltage V fd of the exciter and y is the error e e leaving
the summing junction.
The constant A is usually a small number typically in the range 0.07 and 0.1 per unit so that
when the generator is at or near no load the exciter is either not saturated or is only just beginning to
become saturated. The constant B takes account of the extent of saturation that occurs as the exciter
field voltage is increased. It has a typical value in the range of 0.4 to 0.6 per unit.

Figures 4.2 and 4.3 show the open-circuit curves for a wider range of values for A and B in
order to show more clearly the effect that they have on the shape of the curve.
The two constants A and B can be found from data given by the manufacturer for the exciter
open-circuit voltage V fd and the excitation voltage (or current) V a . The data are usually given in
graphical form as actual quantities, i.e. volts and amps. These should first be converted into their
equivalent per unit form by dividing by their values that correspond to the no-load condition of
the main generator. When this conversion is made unit output voltage of the exciter produces unit
terminal voltage at the main generator.
Since there are two unknown constants their solution will require two equations. Hence any
two pairs of data points can be used from the open-circuit voltage curve of the exciter. Using the
notation in Figure 4.2 or 4.3 let these pairs of points be,

V a1 with V fd1 and V a2 with V fd2