7.2.1.1 Siting and Wind Characteristics
A wind turbine test site has been established directly behind the engineering faculty at Stellenbosch University. The wind conditions in this area are, however, far from ideal. In addition to normal sea- sonal and diurnal variations, the wind is turbulent and unpredictable due to the proximity of numerous buildings and tall trees. Although strong gusts and high wind speeds do occur during certain seasons, the average wind speed at the site is less than 4 m/s. Time constraints did not allow testing during ideal conditions so the results presented here were generally recorded for wind speeds below 8 m/s.
7.2.1.2 Installation of the Turbine, SS-PMG, and GCC
The 7,2 m wind turbine and 15,9 kW SS-PMG were installed atop a tubular tower with an added exten- sion, giving a hub height of approximately 18 m. The extension was added in an attempt to reach wind zones with less turbulence and higher average speed.
134 CHAPTER 7. LABORATORY AND FIELD TESTS
ing the WECS on the tower begins with connecting the SS-PMG and nacelle to the top flange of the tower while at ground level. The spring-tensioned tail vane is then connected to the nacelle (for a description of the yaw control mechanism, see Section 3.1.2).
Finally, the three turbine rotor blades are mounted directly to the front plate of the SS-PMG slip-rotor. The tower must be lifted slightly to allow the final blade to be installed, after which the nose cone can be screwed in place. Once all components are secure, the tower can be lifted into its vertical position by a crane, as shown in Fig. 7.10(a). Bolts secure the tower to the base-plate, which is itself secured to a foundation block.
The fully installed WECS is shown in Fig. 7.10(c), where the nose-cone, blades, SS-PMG, nacelle, and tail vane are all visible. Close inspection will reveal that the tensioning spring for the tail vane has been replaced by steel cable, effectively holding the tail vane perpendicular to the rotor plane. This was after the spring became damaged in the lifting process. The upshot of this arrangement is that no rotor torque reduction can occur under high winds, but none were encountered during the test period so the effect of this change was minimal.
Underground armoured cabling connects the SS-PMG to the GCC, which is installed in a nearby distribution room, shown in Fig. 7.10(b). The GCC employed for the field tests was not equipped with monitoring dials or manual override switches so an external monitoring console that provides these features is also visible. The GCC, in turn, connects directly to the low-voltage electrical distribution board shown in the figure. The resistor cage was carried outside to improve cooling during prolonged testing but could be left in the distribution room during normal operation.
7.2.1.3 Measuring Instruments
In the absence of a shaft-speed resolver, the Norma power analyser was again called upon to measure and log fgen. A Tektronix oscilloscope with differential voltage probes and a clamp-on current probe was
used to measure line voltages and phase current. Phase voltages could no longer be measured because the neutral point of the SS-PMG was not available through the slip-ring coupling in the nacelle—this coupling is necessary because it allows the nacelle to rotate freely without risking cable twist.
An anemometer providing real-time wind speed data was not available during the field tests. As such, it is only possible to attach an average wind speed value to the measurements presented in the sections that follow.
7.2.2
Speed Control
Laboratory investigations indicate that the reduced-gain PI speed controller is functional but does not deliver ideal results under turbulent conditions. To verify these observations, the performance of the thyristor-based speed controller was tested with a real turbine as prime mover under turbulent wind conditions.
Controller gains were set according to the values chosen in Section 7.1.3.1 and, for comparative pur- poses, synchronisation was not enabled. Under this arrangement the turbine and SS-PMG are allowed
7.2. FIELD TESTS 135
(a) Raising the 15 kW SS-PMG WECS and tubular tower on a pivoting base. The extension segment al- lows for a hub height of 18 m
(b) Indoor location of the GCC, where it connects the WECS to the local grid through a distribution board. The monitoring station and resistor cage are also visible.
(c) Close-up view of the tower-top components of the WECS, including the SS-PMG, nacelle and yaw-controlling tail vane
136 CHAPTER 7. LABORATORY AND FIELD TESTS
to run freely until the cut-in rotational speed ( fgen = 0,6 p.u.) is reached, at which point the speed
controller attempts to maintain fgenat 1 p.u. indefinitely.
The PM-rotor speed traces in Fig. 7.11 are for low to moderate wind speeds with significant tur- bulence present. In Fig. 7.11(a) and Fig. 7.11(b) the effectiveness of speed control over relatively long periods is illustrated. Tracking is well centred on fgen =1 p.u. and ∆ ft < 0,02 p.u. for the majority of
both traces. The significant deviations that occur involve fgen falling below the target frequency band,
reflecting the same limitation that is discussed in Section 7.1.3.2: the speed control mechanism can only reduce net torque, not increase it. If wind speed falls below 4 m/s then a reduction in fgenis inevitable.
On the other hand, wind speeds too low to hold the turbine at synchronous speed will not allow the SS-PMG to act as a net exporter of energy. When connected to the grid under such wind conditions, the SS-PMG will ultimately switch over to motoring mode to hold the WECS at speed, which is an undesirable situation. The GCC detects reverse power and disconnects the SS-PMG after a set period, but it may be preferable to avoid synchronisation altogether during marginal wind conditions.
The remainder of the PM-rotor speed plots show tracking over shorter periods. In Fig. 7.11(d), tur- bulent conditions cause rotor speed oscillations between t= 20 s and t= 40 s but control is recovered for the remainder of the recorded period. Fig. 7.11(c), Fig. 7.11(e), and Fig. 7.11(f) exhibit consistent tracking as long as the wind speed remains high enough to support it. Overshoot never exceeds 8 %.
These in-field measurements show that the thyristor-based speed controller is successful at achieving its design objectives under low to moderate wind conditions. It can be expected that fgendeviations will
increase at higher average wind speeds with stronger gusts, but a significant margin of safety exists— overshoot can double compared to the values recorded in Fig. 7.11 without posing a danger to the turbine or SS-PMG.
There is also no sign of the rotor oscillations that occurred during speed control with the VSD-IM as prime mover. As predicted, the wind turbine imparts a more damped response. This leaves room for the speed controller gains to be increased, if necessary, to cope with sites that have higher average wind speeds.