Estructura de la conducta

1.2.4.1

PV Inverter Configuration

Two types of inverter configuration are employed presently in solar farms. One is called string technology; [32]-[35] where several modules in string configuration feed in to a single large inverter. These large inverters are grouped together to feed the grid. The other is called the micro-inverter, also known as AC module technology, [32], [33] where each individual module has its own inverter and the outputs of all micro-inverters are integrated together to feed the grid.

There are numbers of inverter topologies used for grid connected PV solar farms. There are single step topology for AC modules [36],[37], two stage topology for multiple modules [38],[39], multilevel inverter topology [40],[41], fly-back type inverter [42],[43], fly-back current fed inverter, [44] all of which are summarized in [45]. Different topologies used by several commercial manufacturers are compared and studied in terms of performances in [46]-[50], [35]. A resonant inverter is also discussed in [51]. To construct inverter circuits, manufacturers use Metal-Oxide Semiconductor Field Effect Transistor (MOSFET) [52], Gate Turn Off (GTO) thyristor, and Insulated Gate Bipolar Transistor (IGBT) switches [53]. The present trend is to use IGBT switches [53], [54] because of their low loss and ease of switching [55]. A typical six pulse inverter using the IGBT switches is depicted in Fig. 1.4 [11], [24] which is comprised of 6 IGBT switches with associated snubber circuits for smooth switching operation [56]. The firing pulses to trigger the IGBT switches are generated from the inverter controller.

Figure 1.4 Typical PV solar farm inverter.

1.2.4.2

Inverter Firing Strategy

To generate the firing pulses for the IGBTs in the inverter, various technologies such as Pulse Width Modulation (PWM) [52], [53], Sinusoidal Vector PWM (SVPWM) [24], or Hysteresis control [57] techniques are used. Among all of these techniques, the PWM technique is widely used for high power PV inverter applications [52],[53],[58].

1.2.4.3

Inverter Control schemes

The PV solar farm is required to inject power to the grid at close to unity power factor according to several DG inter-connection standards [59],[60],[61]. In modeling the control system, two control schemes are widely used - Current Sourced Inverter (CSI) and Voltage Sourced Inverter (VSI) or, alternately known as Voltage Sourced Converter (VSC). In a CSI scheme, the inverter input is maintained as a constant current source with the use of DC link inductors, whereas the VSI scheme uses the DC link capacitor to maintain constant voltage source at the inverter input. For CSI, the output of the inverter is the controlled voltage. On the other hand, the output of the VSI control scheme is a controlled current. Due to the lower short circuit current contribution to the grid faults, the VSI with current control strategy is preferred by industries [52],[53],[55]. Therefore, the remainder of the work in this thesis is mainly focused on VSI based inverter control schemes.

Fig.1.5 presents the VSI current controller of a typical PV solar farm inverter which uses d-q current and voltage components as a feedback signal of two current control loops: direct axis current control loop and quadrature axis current control loop [24],[52],[54],[62]-[65]. The three phase voltage (Va, Vb, Vc) and current (Ia, Ib, Ic) measured at PCC (not shown in figure) are converted into d-q components of voltage (Vsdc, Vsqc) and current (Id, Iq) through the Park’s transformation process. In this transformation, it uses the angle obtained from the PCC voltage (Va, Vb, Vc) through a phase locked loop (PLL) oscillator known as the synchronization angle, θ_{. For balanced} three phase signals,

() = cos ; () = cos( −

) ; () = cos( −

) and = ( + )

The Park’s transformation is as follows [108] as given in Appendix – A for a rotating d-q frame at a speed of ω:

()

() =

2 3 .

cos cos +2₃ cos −2₃ − sin − sin +2_{3 − sin −}2₃ " . #

() () () $

where, the quantity, f, represents either the instantaneous voltage or current signals. ω=2πf is the angular frequency and φ is the phase angle of the corresponding components. Therefore, the real power, P and the reactive power, Q in d-q co-ordinate system is given by [108],

% =3_{2 &'}. ( + '. () +,- . =3_{2 &−'}. (+ '. ()

If the direct axis component vd is aligned with the space vector, at steady state, the power expression can be decoupled as follows [108]:

% =3_{2 /'}. (0 +,- . =3_{2 &−'}. ()

The d-q components of voltages vd, vq and current id, iq are constantly rotating with d-q reference axes. Therefore, vd, vq, id and iq are expressed as in Laplace form as Vsdc, Vsqc, Id and Iq, respectively.

In Fig. 1.5, the direct axis current control loop consists of two proportional integral controllers, PI-1 and PI-3. Whereas, the quadrature axis current control loop consists of PI-2 and PI-4. The direct axis current control loop maintains the DC link voltage constant around a reference set by the MPPT algorithm in a converter-less MPPT technique and injects the balance of the power generated from the PV modules to the inverter output. To perform this control function, the DC link voltage, VDC, is compared with the DC voltage reference, VDC_ref, obtained from the MPPT algorithm and is passed through a PI-3 controller to obtain the direct axis current control loop reference, Id_ref. The measured direct axis current signal, Id, is compared with this reference signal, Id_ref, to obtain the direct axis modulation signal, md, through the PI-1 controller. On the other hand, the quadrature axis current control loop is used to maintain the reactive power output of the inverter around a reference value of zero for unity power factor operation[24],[62]. In this case, the reactive power output is compared with the reference value which gives the quadrature axis current control loop reference, Iq_ref, through the PI-4 controller and generates the quadrature axis component of modulation signal, mq, with the comparison of Iq through the PI-2 controller. The direct axis and quadrature axis voltages act as disturbance signals in these control loops and a decoupling factor of ωL is included to decouple these two control loops where, ɷ is the angular frequency in rad/sec and L is the

inductance at the output of the inverter. The detailed mathematical manipulation is given in Appendix-A. However, the output two axis modulation signals md and mq from the controller are converted into three phase signals called modulating signals (ma,mb,mc), as follows:

1 = 1 cos( + 2), 1 = 1 cos( + 2 − 120°), 1 = 1 cos( + 2 + 120°) where, 1 = 61+ 1, 2 = tan9: ;< ;=. > = ?>@A+ >BA C = DEF−G>B >@ 1+ = 1 HIJ( + 2) 1K = 1 HIJ( + 2 − 120°) 1L = 1 HIJ( + 2 + 120°)

Figure 1.5 Generic VSI based PV inverter controller.

In the PWM technique, these ‘modulating signals’ are compared with high frequency triangular signals called ‘carrier signals’ to generate the gating signals for the IGBT devices to inject PV solar farm power to the AC grid at unity power factor and controls the DC link voltage.

1.2.4.4

DC link Capacitor

In a PV solar system the main role of the DC link capacitor, in addition to holding a constant DC voltage, is to maintain the power quality at the DC side which ultimately influences the power quality at the AC side [66]. For smooth operation of the inverter, a comparatively ripple free DC current and voltage is required at the input of the inverter. While using the DC link capacitor in a converter-less MPPT system, the size of the DC link capacitor must be carefully chosen. Otherwise, pulsating power due to extra high

value of capacitor or a power variation due to a very low value of DC link capacitor will be transmitted to the grid [27]. The sizing of DC link capacitors is reported in [52], [65], [67], [68] based on maximum energy storage capacity and allowable minimum DC voltage.