Environmental Engineering Reference
In-Depth Information
chosen as
τ
v
=
0
.
02 s. The simulations were carried out in MATLAB
R
7.4 with Simulink
R
.
The solver used in the simulations is ode23tb with a relative tolerance 10
−
3
and a maximum
step size of 0
2 ms. The synchronverter can feed pre-set real power and reactive power to the
grid (called the set mode) and can automatically change the real power and reactive power fed
to the grid according to the grid frequency and voltage (called the droop mode).
The simulation was started at
t
.
0 to allow the PLL and synchronverter to start up (in
real applications, these two can be started separately). The dynamics in the first half second is
omitted. The circuit breaker was turned on at
t
=
=
1 s; the real power
P
set
=
80 W was applied
at
t
=
2 s and the reactive power
Q
set
=
60 Var was applied at
t
=
3 s. The droop mechanism
was enabled at
t
=
4 s and then the grid voltage decreased by 5% at
t
=
5s.
18.4.1 Under Different Grid Frequencies
The system responses when the grid frequency was 50 Hz are shown in the left column
of Figure 18.5. The synchronverter tracked the grid frequency very well all the time. The
voltage difference between
v
g
before any power was applied was very small and the
synchronisation process was very quick. There was no problem when turning the circuit breaker
on at
t
v
and
1 s; there was not much transient response after this event either. The synchronverter
responded to both the real power command at
t
=
3svery
quickly and settled down in less than 10 cycles without any error. The coupling effect between
the real power and the reactive power was reasonably small. When the droop mechanism
was enabled at
t
=
2 s and the reactive command at
t
=
4 s, there was not much change to the real power output as the frequency
was not changed but the reactive power dropped by about 53 Var, about 50% of the power
rating, because the local terminal voltage
=
v
was about 2
.
5% higher than the nominal value.
When the grid voltage dropped by 5% at
t
4 s, the local terminal voltage dropped to just
below the nominal value. The reactive power output then increased to just above the set
point 60 Var.
The same simulation was repeated but with a grid frequency of 49
=
1% lower
than the nominal value. The system responses are also shown in the left column of Figure 18.5
as well for comparison. The synchronverter followed the grid frequency very well. When the
synchronverter worked at the set mode, i.e. before
t
.
95 Hz, that is 0
.
4 s, the real power and reactive power
all responded to the setpoint exactly, respectively. After the droop mechanism was enabled at
t
=
4 s, the synchronverter increased the real power output by 20 W, that is 20% of the rated
power, corresponding to 0
=
.
1% drop of the frequency. This did not cause much extra change
to the reactive power output, just slight adjustment corresponding to the slight change of the
local voltage
v
.
18.4.2 Under Different Load Conditions
The same simulation with the nominal grid frequency was repeated but with the local load
changed from
R
=
1000
to
R
=
4
at
t
=
1
.
5 s. The system responses are shown in
the right column of Figure 18.5. The local voltage
v
dropped immediately, with some short
big transients, after the load
R
was changed to 4
at
t
=
1
.
5 s and the voltage difference
between
v
and
v
g
increased because of the load current drawn from the grid. There was some
Search WWH ::
Custom Search