DEMAND for clean, economical, and renewable energy

Transcription

1 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 24, NO. 1, MARCH A Single-Stage Single-Phase Transformer-Less Doubly Grounded Grid-Connected PV Interface Hiren Patel and Vivek Agarwal, Senior Member, IEEE Abstract A transformer provides galvanic isolation and grounding of the photovoltaic (PV) array in a PV-fed gridconnected inverter. Inclusion of the transformer, however, may increase the cost and/or bulk of the system. To overcome this drawback, a single-phase, single-stage [no extra converter for voltage boost or maximum power point tracking (MPPT)], doubly grounded, transformer-less PV interface, based on the buck boost principle, is presented. The configuration is compact and uses lesser components. Only one (undivided) PV source and one buck boost inductor are used and shared between the two half cycles, which prevents asymmetrical operation and parameter mismatch problems. Total harmonic distortion and dc component of the current supplied to the grid is low, compared to existing topologies and conform to standards like IEEE A brief review of the existing, transformer-less, grid-connected inverter topologies is also included. It is demonstrated that, as compared to the split PV source topology, the proposed configuration is more effective in MPPT and array utilization. Design and analysis of the inverter in discontinuous conduction mode is carried out. Simulation and experimental results are presented. Index Terms Double grounding, IEC 1547, IEC 61727, maximum power point tracking (MPPT), NEC 690, partially shaded conditions, photovoltaic. I. INTRODUCTION DEMAND for clean, economical, and renewable energy has increased consistently. Among a variety of renewable energy sources available, photovoltaic (PV) appears to be a major contender on account of its abundance, easy availability, and pollution-free operation. Increasing interest in sustainable energy production through PV [1], however, demands attention on various issues such as maximum power point tracking (MPPT) [2], [3], personal safety, grid integration, stability and reliability, power quality, power electronic interface of PV with the grid, and operation under various environmental conditions [4] [7]. With increased level of penetration of PV-based systems into the existing grid, these issues are expected to become more critical with time, since they may affect the performance of the other grid-connected systems [8]. This implies that there is a need for a set of rules and regulations to govern the grid-connected PV systems. At present, there are no such globally accepted standard rules and regulations. However, in some countries, it is mandatory Manuscript received February 19, 2008; revised June 8, Current version published February 19, Paper no. TEC H. Patel is with the Department of Electrical Engineering, Indian Institute of Technology Bombay, Mumbai , India, and also with the Sarvajanik College of Engineering and Technology, Surat , India. V. Agarwal is with the Department of Electrical Engineering, Indian Institute of Technology Bombay, Mumbai , India ( agarwal@ ee.iitb.ac.in). Digital Object Identifier /TEC that the grid-connected PV systems must comply with certain standards like IEEE 1547 [9], IEC 61727, and NEC 690 [10]. IEEE 1547 has provisions related to the performance, operation, testing, safety considerations, and maintenance of the grid PV interconnection. IEEE 1547 and IEC impose limitations on the maximum allowable amount of injected dc current into the grid, which is 0.5% 1% of the rated output current of the source. NEC 690 requires that the grid-connected, PV-based, dc ac inverter topologies must be appropriately earthed (grounded). The grounding of a PV system, referred to as system earthing, needs special consideration due to safety reasons [11], and to minimize the effects of lighting and other surges. It refers to an intentional connection to earth of one of the current-carrying conductors in the PV system. In view of this, in certain countries (e.g., USA), it was mandatory to provide earthing to the PV system when its output dc voltage exceeded a certain level, typically 50 V. According to the revised NEC 690 (revised in 2005), PV inverters with ungrounded conductors are now allowed in USA. However, the ungrounded PV inverters have to fulfill a number of additional requirements (e.g., disconnects and overcurrent protection in both the conductors, provision for minimization of the effects due to surges, etc.) [12]. It is important to note that even though ungrounded PV systems are now allowed, system earthing is still advisable. In addition to the PV string s grounding, the neutral conductorofthe1 φ inverter feeding the utility or the neutral of the utility itself is grounded. This necessitates a PV inverter topology that allows double grounding [10]. The inverter topologies consisting of a transformer can easily realize double grounding. In fact, double grounding can also be achieved with the transformer-less topologies employing split PV sources. As discussed later in Section II, most of the transformer-less topologies, which have double grounding feature, employ a split PV source (i.e., a PV source split into two halves PV 1 and PV 2 ). However, they suffer from the following drawbacks. 1) They generate alternate half cycles of the grid current from different halves of the split PV source. If the two halves (PV 1 and PV 2 ) of the PV source operate in unidentical conditions like different solar insolation levels or different shading patterns, it results in a distorted grid current with high total harmonic distortion (THD) and/or high dc component. 2) Grid current may have high THD and/or dc component if the modules used for PV 1 and PV 2 have some mismatch. 3) Only one of the two halves of the PV source is utilized in a given ac half cycle. This leads to more ripple in the voltage across PV 1 and PV 2. As a result, the average output power extracted from the PV decreases /$ IEEE

2 94 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 24, NO. 1, MARCH 2009 Fig. 1. Schematic of the circuit for the proposed system. 4) As explained later in Section V, when operating under nonuniform conditions, inverters with split PV source are less effective in the utilization of the PV array. This paper reviews various transformer-less topologies and proposes a new single-stage, 1 φ, transformer-less, PV-fed grid-connected system. Fig. 1 shows the proposed configuration employing double grounding. The salient features of the proposed topology are as follows. 1) It meets the double grounding requirement. 2) Even under partially shaded conditions, there is no probability of injecting dc component into the grid. 3) It has a single buck boost inductor and uses the same operating principle (as that of a buck boost converter) in both the halves of the ac cycle. As a result, a symmetrical grid current with low THD and dc component is obtained. 4) It uses only one PV string as the input source, which is used in both halves of the ac cycle. Due to this, the ripple in the voltage of the input capacitor C in (Fig. 1) is less, leading to the extraction of more average output power. 5) The proposed topology is more suitable in effective utilization of the PV array. 6) It can be used irrespective of the PV voltage being greater or smaller than the grid voltage amplitude. This feature renders it suitable for places where the environmental conditions vary over a wide range or the array is likely to receive nonuniform solar insolation (e.g., partially shaded conditions). 7) All the features like MPPT control, inversion, and voltage transformation are encompassed into a single stage. 8) It uses fewer components, and is therefore, more compact. II. REVIEW OF EXISTING TOPOLOGIES Over the years, researchers have presented many topologies that are based on the galvanic isolation provided by the transformer [10], [13], [14]. Another advantage of the transformer is that it can provide voltage transformation. Some single-stage inverter configurations based on high-frequency transformer are also reported [10] to overcome the drawbacks such as lower efficiency, higher cost, and bulky size of the configurations employing line frequency transformer. Several researchers have worked with the PV-fed gridconnected, transformer-less dc ac inverter topologies [15] [18], [20] [22] and the issues related to their operation [23], [24]. To offer a compact and economic design, the transformer-less topologies may have to compromise on the double grounding aspect [15], [16]. A few topologies, which do offer double grounding feature, are shown in Fig. 2. Kasa et al. [17] have presented an inverter based on half-bridge buck boost configuration [Fig. 2(a)]. It employs two parallel-connected buck boost converters, each with its own PV string (PV 1 or PV 2 of the split PV source). It can operate over a wide input voltage range, and has low switching and conduction losses. But, as only one PV string is used over a half cycle, ripple in the voltage across decoupling capacitors is very large, leading to a decrease in the net output power obtained from the PV strings. This could be minimized by having a large electrolytic capacitor, but at an increased cost. In addition, in case of unbalanced power generation from the two PV strings (PV 1 and PV 2 ), the current injected into the grid may not be symmetrical (highly distorted with high THD), and may have a dc component that does not conform to IEEE The configuration shown in Fig. 2(b) utilizes generation control circuit (GCC) [18], which comprises two switches (S 1 and S 2 ) and an inductor L. With GCC, MPPT can be applied to both the PV strings independently. The disadvantage with the configuration is that it is inherently a half-bridge, exhibiting buck characteristics. Hence, this topology is applicable only where input voltage is greater than the output voltage. Due to this, a PV string of large number of PV modules would be required, which is more likely to get affected by partial shading conditions. In such a case, the PV characteristics of these strings may have multiple peaks [19], which makes it difficult to track the global peak, unless each PV module is supported by its own GCC. Another transformer-less topology that can provide a solution to the dual grounding problem is a multilevel (three levels) halfbridge diode-clamped inverter (HBDC) shown in Fig. 2(c) [20]. It produces three different voltage levels: 1) V C1 when switches S 1 and S 2 are ON; 2) zero when switches S 2 and S 3 are ON; and 3) V C2 when switches S 3 and S 4 are ON. Here also, as in the topology of Fig. 2(a), each of the PV strings is loaded only for half the ac cycle. This results in the requirement of a large electrolytic capacitor to reduce the ripple in the capacitor voltage. This topology does not provide voltage boosting, and when the two PV strings generate unequal power, it may generate asymmetrical output current. Kusakawa et al. [21] have presented an inverter topology [Fig. 2(d)] based on the buck boost principle for an ac module, with a low power rating of around 50 W. Unlike other configurations, this configuration uses only one PV string. The circuit is operated in the continuous conduction mode (CCM), which requires a large inductor value. Further, the PV is not actually grounded; thus, the topology does not truly provide double grounding. Nevertheless, during alternate half cycles, either the positive or the negative conductor of the PV remains grounded through the switches S 1 and S 4. All the inverter topologies, shown in Fig. 2(a) (d), operate on the same principle in both the ac half cycles. Schekulin [22] has proposed a topology [Fig. 2(e)] derived from the basic Zeta and Ĉuk (fourth-order converters) configurations. This topology uses minimal number of devices. However, during positive

3 PATEL AND AGARWAL: SINGLE-STAGE SINGLE-PHASE TRANSFORMER-LESS DOUBLY GROUNDED GRID-CONNECTED PV INTERFACE 95 Fig. 2. Existing transformer-less, single-stage, 1 φ PV-based inverter topologies with double grounding. (a) (c) Topologies based on split PV-source [17], [18], [20]. (d) (e) Configurations with single PV source [21], [22]. (f) Proposed topology. half-cycle, it transfers the power to the grid on the principle of a buck boost converter, while during negative half-cycle, it uses the boost principle to transfer the power to the grid. This asymmetrical operation not only makes the control scheme complex, but may also result in asymmetrical current and may inject dc component into the grid current. III. CIRCUIT CONFIGURATION AND WORKING The circuit schematic of the proposed 1 φ PV inverter is shown in Fig. 1. Unlike some other topologies, which use two PV strings, the circuit of Fig. 1 employs only one PV string as the energy source. In addition, the negative conductor of the PV source always remains grounded, thereby providing double grounding. The circuit uses five switches (S 1 through S 5 ) and three diodes (D 1 through D 3 ). It uses only one buck boost inductor L. Inductor L f and capacitor C f form the low-pass filter, which allows only the 50-Hz component of the inverter output current to enter into the grid. The various modes in which this circuit operates are shown in Fig. 3. During the negative half cycle of the ac grid voltage, switches S 1,S 2, and S 5 along with the diode D 1 form a buck boost converter. The switches S 3 and S 4 always remain OFF.Fig.3(a) (c) shows modes I through III, respectively, in which the converter operates during the negative half cycle. In the negative half cycle, switches S 2 and S 5 are always kept ON, while the switch S 1 is triggered with the sine-triangle pulsewidth modulation (PWM) technique. When switch S 1 is ON, energy is stored in inductor L 1 [mode I, Fig. 3(a)]. When S 1 turns OFF, the stored energy is transferred to the grid [mode II, Fig. 3(b)]. During the positive half of the ac cycle, switches S 1, S 3 S 5, along with diodes D 2 and D 3, form a buck boost converter. Switch S 2 always remains OFF. The buck boost converter is now operated by controlling the switches S 1 and S 5 using the signal derived from the sine-triangle PWM. Fig. 3(d) (f) shows modes IV through VI, respectively, in which the converter operates during the positive half cycle. During these modes, switches S 3 and S 4 are always ON. When switches S 1 and S 5 are ON, energy is stored in inductor L 1 [mode IV, Fig. 3(d)]. When S 1 and S 5 turn OFF, the stored energy is transferred to the grid [mode IV, Fig. 3(e)]. The amplitude of the reference sinusoidal waveform used for the sine-triangle PWM (mentioned before) is controlled to track the maximum power point (MPP). The controller implements the perturb and observe (P&O) method, and identifies whether to increase or decrease the amplitude of the reference waveform to achieve MPP. It then increases/decreases the amplitude of the sinusoidal template, which is derived from the grid voltage. Unlike the operating modes shown in Fig. 3, where the operation in both the half cycles is based on the buck boost principle, it is possible to operate the proposed configuration even with asymmetric operating principle. During the negative half cycle, the operating modes are similar to those shown in Fig. 3(a) and (b), i.e., employing the buck boost principle. But in the positive half cycle, instead of operating the configuration as a buck boost converter, it is operated as a buck converter for that period of the half cycle where V PV >v g and as a boost converter for the remaining period. This feature of operating the circuit distinctly as a buck and a boost converter for the positive half cycle can reduce stress on the components. This is a desirable yet unacceptable proposition on account of the asymmetry it introduces between the positive and negative half cycles because the topology does not support distinct buck and boost operations during the negative half cycle. IV. DESIGN OF THE COMPONENTS The design of various components used in the proposed inverter is very crucial for generating a sinusoidal grid current,

4 96 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 24, NO. 1, MARCH 2009 Fig. 3. Circuit diagrams for various operating modes of the proposed configuration. (a) (c) Operation during negative half cycle: modes I III. (d) (f) Operation during positive half cycle: modes IV V. Bold lines show the active current paths. (c) and (f) Operation after all the energy stored in the buck boost inductor is transferred to the grid. which is in phase with the grid voltage. The design also decides the discontinuous conduction mode (DCM) or CCM operation. The analysis leading to the design of the various components for the DCM operation is presented next. A. Design of Inductor L 1 The design of the buck boost inductor should ensure DCM operation of the inverter under all conditions of temperature and peak insolation levels. The inductor should also be able to handle the maximum energy corresponding to the peak power rating of the PV array. If DCM is ensured for peak power conditions, the inverter will operate in DCM for all other conditions. The value of the inductance is obtained for the boundary condition (critical conduction mode), i.e., for the operation on the boundary of DCM and CCM. As the current injected into the grid is in phase with the grid voltage (v g = V gm sin ωt, where V gm is the amplitude, ω is the angular frequency, and t is the time), the maximum (instantaneous) power is injected into the grid at the time when the grid voltage is at its peak (i.e., ωt = 90 ). Thus, the design for the inductor should be in accordance with the current, voltage, and power values at that instant. The maximum power obtained from the PV string is given by P PV max = V PV mpp max I PV mpp max (1) where V PV mpp max and I PV mpp max are the output voltage and current of the PV array, when working at MPP under maximum insolation and uniform conditions. Assuming a lossless inverter, the maximum power injected into the grid is P g max = V gm 2 I gm max 2 = V PV mpp max I PV mpp max. (2) Here, I gm max is the amplitude of current injected into the grid under maximum insolation and uniform insolation conditions. If the switching frequency (f s ) of the converter is very high for a given switching period (T s ), the grid voltage (v g ) and the grid current (i g ) can be considered constant. Under these conditions, at an instant ωt = 90, v g = V gm and i g = I gm max. The energy transferred to the grid during this period is During the ON period E max = V gm I gm max T s. (3) T ON = and during the OFF period L criti peak V PV mpp max (4) T OFF = L criti peak (5) V gm where I peak is the peak inductor current and L crit is the inductance value for the critical conduction mode. From (4) and (5) T s = T ON T OFF = L crit I peak [ 1 V gm 1 V PV mpp max ]. (6) Using (6), the energy stored in the inductor during this period is given as follows: L crit Ipeak 2 = T 2 [ ] 2 s 1 1. (7) 2 2L crit V gm V PV mpp max From (3) and (7), T s L crit = 2V gm I gm max [ 1 V gm 1 V PV mpp max ] 2. (8)

5 PATEL AND AGARWAL: SINGLE-STAGE SINGLE-PHASE TRANSFORMER-LESS DOUBLY GROUNDED GRID-CONNECTED PV INTERFACE 97 From (2) and (8), [ ] T s 1 1 L crit =. V PV mpp max I PV mpp max V gm V PV mpp max (9) Thus, the value of L crit (for the critical conduction mode) is obtained from the PV string parameters, the grid voltage, and the switching frequency. To ensure DCM operation, L 1 (Fig. 1) should be smaller than L crit. B. Design of Filter Capacitor C f The value of filter capacitor C f is obtained by equating the energy released by the inductor L 1 and the energy received by the capacitor C f.if V is the ac voltage ripple on C f,at ωt = 90, the energy balance between C f and L 1 is given by the following: L crit Ipeak 2 2 Hence, = C f 2 [(V gm V ) 2 (V gm V ) 2 ]. (10) From (6) and (11), C f = T 2 s 4L crit V gm V C f = I2 peak L crit 4V gm V. (11) [ 1 V gm 1 V PV mpp max ] 2 (12) which gives the value of C f in terms of L 1, grid voltage, PV string parameters, switching frequency, and acceptable highfrequency ac voltage ripple, superimposed on the sinusoidal capacitor voltage. C. Design of Filter Inductor L f The filter inductor can be obtained as follows: L f = 1 1 ω 2 = C f (2πf c ) 2 (13) C f where f c is the cutoff frequency, which is much less than the switching frequency (f s ) at which the switch S 1 is operated. D. Design of Decoupling Capacitor C in The size of C in can be decided as follows [10]: Hence, C in = P PV 4πfV PV mpp max V PV. (14) C in = I PV mpp max (15) 4πf V PV where f is the frequency of the grid voltage and V PV is the amplitude of the ripple voltage across the decoupling capacitor connected across the PV array. V. SIMULATION RESULTS In this section, various cases are simulated to highlight the issues, arising due to partial shading, control scheme used, Fig. 4. Performance of the configuration shown in Fig. 2(a) under nonuniform conditions. MPPT is implemented based on the information sensed from array PV 1. (a) (c) Output power, voltage, and current of arrays PV 1 and PV 2, respectively. (d) THD of grid current. (e) Voltage across filter capacitor. (f) and (g) Inductor current. (h) and (i) Grid current under uniform and nonuniform conditions, respectively. configuration, improper design of the circuit components, etc. Case a discusses the effect of partial shading on the split-pvsource-based configurations. Case b presents the effect of control scheme employing two different operating principles in the two half cycles, while case c shows the effect of improper selection of the buck boost converter on the grid current waveform. A. Results With a Split Source Topology 1) Case a (Operation of a Split PV Source in Partially Shaded Conditions Where MPPT is Used With an Unshaded Array): Fig. 4 shows the response of the circuit shown in Fig. 2(a) [17] when operating under nonuniform conditions. The PV array, PV 1, is receiving an insolation of 0.5 kw/m 2. PV 2 receives 0.5 kw/m 2 until t = 2 s; 0.1 kw/m 2 for the interval s; and 0.3 kw/m 2 after t = 3.7 s. Hill-climbing method is used for MPPT by sensing voltage and current of the array PV 1. Fig. 4(a) (c) shows that the

6 98 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 24, NO. 1, MARCH 2009 maximum power from both the arrays is being tracked until t = 2 s due to uniform conditions on both the arrays. However, as the MPPT is done by sensing the parameters of PV 1,aftert = 2 s, when PV 2 starts receiving lower insolation, an optimum utilization of the array PV 2 is not achieved. Fig. 4(d) shows that the THD of the grid current is less than the permissible limit until t = 2 s. However, after t = 2 s, the THD of the grid current is about 90% for the interval s and is 48% beyond t = 3.7 s. Simulations were also performed with the MPPT implemented by sensing the parameters of PV 2 in place of PV 1, with the other conditions remaining the same (results not explicitly shown due to paucity of space). It was observed that the THD in grid current after the occurrence of sudden shading increased to 85%, but in steady state, it reduced to within 10% 15% range. However, the steady-state power tracked in this case is much smaller than the previous case (325 W), leading to ineffective utilization of the PV arrays, especially PV 1. B. Results With the Proposed Topology 1) Case b (Performance of an Inverter With Asymmetrical Operating Principles in the Two Half Cycles): The results shown in Fig. 5 are for a PV inverter employing different principles of operation in the positive and negative half cycles. The proposed circuit (Fig. 1) is operated as a buck boost converter in the negative half cycle [Fig. 3(a) (c)]. However, in the positive cycle of grid voltage, instead of operating the circuit as a buck boost converter, the circuit is operated as a buck converter when the PV voltage is greater than the grid voltage and as a boost converter when the grid voltage is greater than the PV voltage. Fig. 5(c) shows the asymmetry in the grid current when the circuit is operated with bang-bang control to control the inductor current in CCM. The amplitude of grid current during the positive half cycle is 2.5 A, while that in the negative half cycle, it is about 2.1 A. Hence, the grid current has a dc component of 0.10 A (about 6.44% of the fundamental component) and its THD is about 15% in CCM. Fig. 5(d) (f) shows the results when the same circuit is operated in DCM. The grid current is more distorted as compared to the CCM operation, with a THD of 22% and a dc component of 0.18 A (10.15% of the fundamental component). Fig. 5(f) shows that the distortion is higher during the positive half cycle of grid current. This is on account of the switching from the buck to boost mode and vice versa. 2) Case c (Effect of Buck Boost Inductor Value on the Performance of the Inverter): Fig. 6 shows the results when an inductor of inappropriate value is used with the proposed topology (Fig. 1). Following values are selected for the various components: L 1 = 210 µh, C f = 20 µf, L f = 2.3 µh, C in = 4500 µf. V and V PV are taken as 20 and 2.5 V, respectively, while f c = 750 Hz. Two parallel-connected PV strings are used as the source. Each PV string has six series-connected PV modules. I PV mpp max and V PV mpp max values for this arrangement are approximately 7 A and 95 V, respectively. With f s = 10 khz and V g = 230 V (rms), the critical value for inductor L crit, using (9), is obtained as 203 µh. Hence, the selected value of L 1 (=210 µh) results in CCM operation. Fig. 6(a) Fig. 5. Performance of a circuit operating on different principles in the positive and negative half cycles. (a) (c) With CCM operation. (d) (f) With DCM operation. shows that when the generated PV power is near its full capacity, the system operates in CCM. As a result, grid current gets distorted [Fig. 6(b)] and injects a distorted grid current with a large dc component [Fig. 6(e)]. Fig. 6(c) shows a 100-Hz ripple in the PV current. This results in a ripple in the capacitor (C in ) voltage. The more the current ripple, the more is the ripple in the voltage and less is the average output power obtained from the PV array. The voltage ripple can be minimized by using a higher value decoupling capacitor (C in ). Fig. 7 shows the response of the proposed converter (Fig. 1) while operating in DCM (with L 1 < 203 µh). Fig. 7(b) shows that as there is only one PV array and only one inductor, the problem of asymmetry in the grid current does not arise. The insolation received by the array before t = 2sis1kW/m 2 and 0.5 kw/m 2 after t = 2 s. Fig. 7(d) shows the amplitude of the reference sinusoidal waveform used for the sine-triangular PWM technique. The amplitude reaches the steady-state value and oscillates about this value. The oscillations are a result of intentional perturbations due to the application of P&O method

7 PATEL AND AGARWAL: SINGLE-STAGE SINGLE-PHASE TRANSFORMER-LESS DOUBLY GROUNDED GRID-CONNECTED PV INTERFACE 99 Fig. 6. Results with the proposed topology with improper inductor size. (a) Inductor current. (b) Grid current. (c) PV string s output current. (d) Amplitude of the reference waveform. (e) THD and dc components of grid current. for MPPT. Fig. 7(h) and (i) demonstrates the effectiveness of the topology, and the control scheme to limit the THD and the dc component to a much lower value of 0.02 and A (about 1% of the fundamental component at low power output under the given partially shaded conditions), respectively. Also, as shown in Fig. 7(e) (g), unlike the split PV-source-based topologies, the MPP is tracked continuously. This ensures effective utilization of the PV array. VI. EXPERIMENTAL RESULTS The experimental setup comprises three parallel-connected PV strings, each with six series-connected PV modules. The specifications of the PV module (output power, current and voltage at MPP, short-circuit current, and open-circuit voltage) at an insolation level of 1 kw/m 2 and 25 C temperature are P max = 38 W, I mpp = 2.29 A, V mpp = 16.6 V, I SC = 2.55 A, and V OC = 21.5 V. Partial shading, using the transparent gelatin papers, is applied artificially by shading the two modules in the first PV string and three modules in the second PV string. Fig. 7. Results when appropriate inductor value is used with the proposed configuration. (a) Inductor current. (b) Grid current. (c) Voltage across the filter capacitor. (d) Amplitude of reference wave. (e) (g) Output current, voltage, and power of the PV array. (h) THD of the grid current. (i) DC component of the grid current. A. Results With the Two-Inductor Topology [16] Figs. 8 and 9 show the results obtained with the circuit configuration that has two parallel-connected buck boost converters [16] and does not have a double grounding. To study the effect of mismatch in buck boost inductor values, the circuit is implemented with buck boost inductors of values L 1 = 0.32 mh and L 2 = 0.29 mh. Fig. 8 shows that due to the mismatch, the grid current amplitudes are not same in the positive and negative half cycles (visible in the power waveform). Fig. 9 shows the variation in the potential of the positive and negative bus with respect to the grid neutral. The absolute maximum voltage at the PV bus is the sum of the grid voltage amplitude and the PV bus voltage. This may lead to an electric shock to a person touching the PV array. Fig. 8. Results with an inverter that uses two buck boost inductors [16]. B. Results With the Proposed Topology Fig. 10 shows the performance of the proposed inverter (Fig. 1) when operating at the global peak power point. The operating voltage, current, and power at this point are 92 V, 2.1 A, and W, respectively. The grid voltage is adjusted (using a variac) to about 100 V with a peak of 141 V. The power supplied to the grid is 168 W. Fig. 10 shows a symmetrical,

8 100 IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 24, NO. 1, MARCH 2009 across the filter capacitor and the buck boost inductor current. Fig. 11(b) shows that for a grid current of 0.82 A, the THD in grid current is 4.9% (case i), which conforms to the standard IEEE 519 (THD 5%), while for a grid current of 0.43 A (corresponding to a very low power level), the THD is 9.2% (case ii). Fig. 9. Variation in the potential of PV bus conductors. Fig. 10. Results with the proposed single-inductor buck boost inverter when operating at the global peak. VII. CONCLUSION Single-phase, single-stage, transformer-less grid-connected PV interfaces, capable of resolving the double grounding problem, have been reviewed. Most of the existing transformer-less topologies achieve double grounding by using a split PV source. Such topologies, when operating under nonuniform conditions, face problems such as inefficient array utilization and dc current injection into the grid. Even inverters sourced by a single PV string, but which operate on different principles in the two halves of the ac cycle, inject a significant dc component into the grid current. A compact PV grid interface, which operates with a single PV source and has the capability of double grounding has been proposed, analyzed, designed, and developed. It is observed that the maximum voltage that can develop on the ungrounded conductor is limited to the PV array output voltage, and hence, the topology exhibits a good safety feature. The topology uses only one PV source, a single buck boost inductor, and a decoupling capacitor that are shared in both the half cycles. This eliminates the problems arising out of asymmetrical operation and mismatch in the components. The THD and dc component of the injected grid current are much lower as compared to other topologies. In addition, due to its inherent nature, it can work over a wide input voltage range. The effectiveness of the proposed inverter configuration to operate in nonuniform insolation conditions is demonstrated with the help of simulation and experimental results. REFERENCES Fig. 11. Results with the proposed inverter at low insolation level. Voltage waveforms have been attenuated ten times with the attenuator probe. (a) Current and voltage waveforms for the PV array, grid, and filter components. (b) THD of grid current at two different low solar insolation levels, λ 1 and λ 2 (where λ 1 > λ 2 ): case (i) for λ 1 and case (ii) for λ 2. distortion-free, sinusoidal grid current that is in phase with the grid voltage. Fig. 11 shows the operation of the proposed inverter when the array is receiving low solar insolation. Fig. 11(a) shows the grid-side and PV-side waveforms along with the voltage [1] O. Wasynczuk, Modeling and dynamic performance of a linecommutated photovoltaic inverter, IEEE Trans. Energy Convers., vol. 4, no. 3, pp , Sep [2] T. Esram and P. L. Chapman, Comparison of photovoltaic array maximum power point tracking techniques, IEEE Trans. Energy Convers., vol.22, no. 2, pp , Jun [3] D. Casadei, G. Grandi, and C. Rossi, Single-phase single-stage photovoltaic generation system based on a ripple correlation control maximum power point tracking, IEEE Trans. Energy Convers., vol. 21, no. 2, pp , Jun [4] M. Liserre, R. Teodorescu, and F. Blaabjerg, Stability of grid-connected PV inverters with large grid impedance variation, in Proc. IEEE PESC, 2004, pp [5] J. H. R. Enslin and P. J. M. Heskes, Harmonic interaction between a large number of distributed power inverters and the distribution network, in Proc. IEEE PESC, 2003, vol. 4, pp [6] W. Libo, Z. Zhengming, and L. Jianzheng, A single-stage three-phase grid-connected photovoltaic system with modified MPPT method and reactive power compensation, IEEE Trans. Energy Convers., vol. 22, no. 4, pp , Dec [7] IEEE Recommended Practice for Utility Interface of Photovoltaic (PV) Systems, IEEE Standard 929, [8] A. Woyte, V. V. Thong, R. Belmans, and J. Nijs, Voltage fluctuations on distribution level introduced by photovoltaic systems, IEEE Trans. Energy Convers., vol. 21, no. 1, pp , Mar [9] IEEE Standard for Interconnecting Distributed Resources With Electric Power Systems, IEEE Standard 1547, 2003.

9 PATEL AND AGARWAL: SINGLE-STAGE SINGLE-PHASE TRANSFORMER-LESS DOUBLY GROUNDED GRID-CONNECTED PV INTERFACE 101 [10] S. B. Kjaer, J. K. Pederson, and F. Blaabjerg, A review of single-phase grid connected inverters for photovoltaic modules, IEEE Trans. Ind. Appl., vol. 41, no. 5, pp , Sep [11] W. I. Bower and J. C. Wiles, Analysis of grounded and ungrounded photovoltaic systems, in Proc. IEEE Photovoltaic Spec. Conf., 1994, vol. 1, pp [12] J. C. Wiles, Photovoltaic power systems and the 2005 National Electrical Code: Suggested practices [Online]. Nov. 2008, Available: tdi/photovoltaic/codes-stds/pvnecsugprac.html [13] S. Saha and V. P. Sundarsingh, Novel grid-connected photovoltaic inverter, Proc. Inst. Electr. Eng., vol. 143, pp , Mar [14] B. K. Bose, P. M. Szczesny, and R. L. Steigerwald, Microcomputer control of a residential photovoltaic power conditioning system, IEEE Trans. Ind. Appl., vol. IA-21, no. 5, pp , Sep./Oct [15] W. Chien-Ming, A novel single-stage full-bridge buck boost inverter, IEEE Trans. Power Electron., vol. 19, no. 1, pp , Jan [16] S. Jain and V. Agarwal, A single-stage grid connected inverter topology for solar PV systems with maximum power point tracking, IEEE Trans. Power Electron., vol. 22, no. 5, pp , Sep [17] N. Kasa, T. Iida, and H. Iwamoto, Maximum power point tracking with capacitor identifier for photovoltaic power system, Inst. Electr. Eng. Proc. Electr. Power Appl., vol. 147, no. 6, pp , Nov [18] T. Shimuzu, O. Hashimoto, and G. Kimura, A novel high-performance utility-interactive photovoltaic inverter system, IEEE Trans. Power Electron., vol. 18, no. 2, pp , Mar [19] H. Patel and V. Agarwal, MATLAB based modeling to study the effects of partial shading on PV array characteristics, IEEE Trans. Energy Convers., vol. 23, no. 1, pp , Mar [20] F. Blaabjerg, Z. Chen, and S. B. Kjaer, Power electronics as efficient interface in dispersed power generation systems, IEEE Trans. Power Electron., vol. 19, no. 5, pp , Sep [21] M. Kusakawa, H. Nagayoshi, K. Kamisako, and K. Kurokawa, Further improvement of a transformerless, voltage-boosting inverter for ac modules, Sol. Energy Mater. Sol. Cells, vol. 67, pp , Mar [22] D. Schekulin, Bundesrepublik Deutschland, Deutsches Patent, Patentschrift DE Cl, Mar [23] M. Calais, J. Myrzik, T. Spooner, and V. G. Agelidis, Inverters for singlephase grid connected photovoltaic systems An overview, in Proc. IEEE PESC, Jun , 2002, vol. 4, pp [24] T. Kerekes, R. Teodorescu, and U. Borup, Transformer-less photovoltaic inverters connected to the grid, in Proc. APEC 2007,Feb.,pp Hiren Patel received the B.E. degree in electrical engineering from the S.V. Regional College of Engineering and Technology (now S.V. National Institute of Technology), South Gujarat University, Surat, India, in 1996, and the M.Tech. degree in energy systems in 2003 from the Indian Institute of Technology Bombay (IITB), Mumbai, India, where he is currently working toward the Ph.D. degree in electrical engineering. His current research interests include computeraided simulation techniques, distributed generation, and renewable energy, especially energy extraction from photovoltaic arrays. He is an Assistant Professor at Sarvajanik College of Engineering and Technology, Surat. Mr. Patel is a Life Member of the Indian Society for Technical Education. Vivek Agarwal (S 92 M 93 SM 01) received the Bachelor s degree in physics from St. Stephen s College, Delhi University, Delhi, India, the Master s degree in electrical engineering from the Indian Institute of Science, Bangaluru, India, and the Ph.D. degree in electrical and computer engineering from the University of Victoria, Victoria, BC, Canada. He was a Research Engineer with Statpower Technologies, Burnaby, BC, Canada. In 1995, he joined the Department of Electrical Engineering, Indian Institute of Technology Bombay, Mumbai, India, where he is currently a Professor. His current research interests include power electronics, modeling and simulation of new power converter configurations, intelligent and hybrid control of power electronic systems, power quality issues, electromagnetic interference (EMI)/electromagnetic compatibility (EMC) issues, and conditioning of energy from nonconventional sources. Prof. Agarwal is a Fellow of the Institute of Electronics and Telecommunication Engineers (IETE) and a Life Member of the Indian Society for Technical Education.