PV string voltage at low temperature
A 4.7 kWp rooftop PV array on a warehouse in Cork is proposed as a single string of 14 modules. Each module has Voc 41.2 V at STC, Vmp 34.5 V, Imp 9.8 A and a Voc temperature coefficient of -0.29%/°C. The transformerless inverter has a maximum DC input voltage of 600 V, an MPPT range of 150–550 V and a maximum input current of 12 A per MPPT. The recorded minimum ambient temperature at the site is -8°C, and the designer expects cell temperature to be about -5°C at that ambient under low irradiance. The DC cable run is 35 m of 6 mm² double-insulated solar cable from the array to the inverter. The array is mounted on a metal standing-seam roof, and the inverter is installed in a garage. The owner wants to know whether the 14-module string can be connected directly and what DC-side protection and isolation are needed.
1. Calculate the string open-circuit voltage at -5°C and decide whether the 14-module string is acceptable for the 600 V inverter.
Read the worked answer
The temperature coefficient is negative, so Voc rises as temperature falls. The temperature difference from 25°C is -30 K. The change is (-0.29%/°C) × (-30 K) = +8.7%. Module Voc at -5°C = 41.2 × 1.087 = 44.78 V. For 14 modules: 14 × 44.78 = 626.9 V. That exceeds the inverter maximum of 600 V, so the 14-module string is not acceptable, even though its STC voltage is 14 × 41.2 = 576.8 V. The maximum number of modules is the largest integer where n × 44.78 ≤ 600, which is 13 (13 × 44.78 = 582.1 V). A 13-module string leaves only about 18 V margin, so I would also check module tolerance and the inverter's absolute maximum; using 12 modules may be more prudent. The MPPT upper limit of 550 V is also relevant at low temperature.
2. Explain why the inverter's 600 V limit is not just a nominal rating and what could happen if it is exceeded.
Read the worked answer
The 600 V figure is the maximum continuous DC input voltage the input capacitors, semiconductors and insulation are designed for. Exceeding it can cause avalanche breakdown in the switching devices, overvoltage of the DC-link capacitors, and failure of the input protection. The inverter may trip on overvoltage, but repeated excursions can degrade components and void the warranty. There is also a safety issue: the DC side remains live whenever the array is illuminated, so a failure can produce a DC arc that is difficult to extinguish. The design must use the coldest expected cell temperature, not the STC rating, and must allow for module voltage tolerance. The AC RCD will not protect against DC-side overvoltage.
3. The inverter is transformerless. What are the implications for RCD selection, DC isolation and array earthing?
Read the worked answer
A transformerless inverter can inject DC residual current into the AC side under a fault, so the RCD type must be selected from the manufacturer's declaration and the National Rules. A Type B RCD may be required where DC residual current above the threshold is possible; some inverters have integrated DC fault current detection that permits a Type A device, but this must be confirmed, not assumed. On the DC side, isolation must be provided at the inverter and preferably at the array, with clearly labelled DC switches rated for PV. The array frame and mounting system should be bonded to the main earthing terminal as required, but the PV live conductors are not earthed; one polarity must not be bonded. DC insulation testing needs a PV-rated tester and the manufacturer's procedure, and string fuses or a fuse-less design must be checked for reverse-current protection.