Small-scale PV Inverters EEA Conference, Wellington. Bill Heffernan, Patrick Chen and Ankur Mishra June 2016

Transcription

1 Small-scale PV Inverters EEA Conference, Wellington Bill Heffernan, Patrick Chen and Ankur Mishra June 2016

2 Introduction Concern exists regarding the effects of many inverters connecting to the LV system Power Quality effects Safety anti-islanding Partially addressed by AS4777.2, Better addressed by AS/NZS4777.2:2015 Additional factors addressed by EEA Guideline Further factors not yet addressed by anybody 2

3 Test brief Set up a test facility in accordance with AS/NZS4777.2:2015 ( the standard ) Test a representative set of inverters according to all sections of the standard Identify any factors not covered by the standard and test these too Record and analyse all data, to identify and examine areas of concern

4 Problems with The Standard Appendix A2 specifies a test voltage of 240V, not 230V The standard jumps between testing at % of rated I, P & S. In Appendix D2 it actually changes from S to I half way through. Only P DC can be directly controlled repeatably Standard Phase and Neutral impedances, defined in Appendix A5, make some specified tests impossible to conduct V nom-max incorrectly stated as 246V in Appendix H (248V)

5 Test Circuit Schematic Lenovo T510 Laptop NetGear D-Link Hub Switch Lenovo T430s Laptop USB USB Vsense Impedance bypass Chroma 62150H PV Simulator PV array DC isolator DC Surge Protector DC Inverter Under Test AC+DC AC Surge Clamp AC Surge Clamp AC AC Manual Grid- Disconnect Switch DC A A A j0.15 Ω Phase impedance Neutral impedance j0.1 Ω Ametek MX45 Grid Simulator Load Vsense CH4 Hioki PW3198 PQ Analyzer CH A. Eberle PQ-box 200 PQ Analyzer 3 Phase Power Distribution Board Three Phase Main Supply Rev Date: 15/06/2016

6 Effect of Source Impedance jωl X Lp R Lp X Ln R Ln L p L n V i V L V s V R R L p = 0.24Ω + j0.15ω L n = 0.16Ω + j0.10ω L loop = 0.40Ω + j0.25ω (Actually 0.4Ω + 800uH) 5kVA inverter supplying 21.7A (i.e. maximum allowable for single phase): V s = 230V V R = 21.7 x 0.4 = 8.68V V L = 21.7 x j0.25 = j5.425v V i = ( ) = V

7 Selection of Inverters Type Code Rated AC Power, VA 1: Micro-inverter A : Micro-inverter B : 2kVA string inverter C : 2kVA string inverter D : 2kVA string inverter E : 2kVA string inverter F : 3kVA string inverter C : 3kVA string inverter G : 3kVA string inverter H single phase inverter models from 7 manufacturers 215VA to 5kVA Chosen by ubiquity on the networks of four NZ lines companies 4: 4-5kVA string inverter F : 4-5kVA string inverter F

8 Tests Performed Basic Power Quality Performance DPF; Current Harmonics; DC Injection; Ripple Susceptibility Demand Response Modes Power Quality Response Modes Volt-Watt; Volt-VAr; Fixed Cosφ ; Fixed VAr; Cosφ (P) Protective Functions Active & Passive Anti-Islanding; Sustained Limits; Underfrequency Transient

9 Harmonics with Pure Sine Supply

10 Harmonics with 4% THD Supply

11 DC Injection Current Magnitude, ma DC Current Limit

12 Ripple Signalling Ripple Frequency, Hz Number of Systems

13 Ripple at 283Hz Harmonic Magnitude, ma Off On 1.75% Ripple Voltage at 283Hz A215, I5 A215, I6 C2000, I5 C2000, I6 F2000, I5 F2000, I6 C3000, I5 C3000, I6 F5000, I5 F5000, I6

14 Ripple at 317Hz Harmonic Magnitude, ma Off On 1.75% Ripple Voltage at 317Hz A215, I6 A215, I7 C2000, I6 C2000, I7 F2000, I6 F2000, I7 C3000, I6 C3000, I7 F5000, I6 F5000, I7

15 Ripple Amplification Current and Voltage harmonic magnitudes, at C3000 inverter terminals, with 1.75% Ripple at 1050Hz. Ripple Voltage magnitude is doubled in this case.

16 Volt-VAr & Volt-Watt Response EEA Guideline recommended response Response achieved by C3000 Inverter

17 Active Anti-Islanding Voltage, V Current, A Voltage Current Unloaded response of weak grid-simulating LC load circuit. Q 102VAr

18 Under-frequency Transient Frequency, Hz Time, s Applied under-frequency transient (from Transpower Companion Guide for Asset Testing)

19 Conclusions None of the inverters yet tested are fully compliant, because DRMs have not been implemented All protective functions are met, BUT default limits are often not correctly set for NZ Some inverters can fully implement the EEA Draft PQ Guideline Voltage Support response, BUT this is not trivial to set up correctly Harmonic performance is marginal with some models, although DC limits are met Ripple signalling can cause significant current responses, which can amplify ripple voltage Only one model tripped out during the system under-frequency transient

20 Final Words We are keen to test new inverter models which claim full compliance with AS/NZS4777.2:2015 Thank you for your attention Any Questions?

21

22 Thank you to the supporters of the EPECentre and the Power Engineering Excellence Trust (PEET)