D.1 General Anormal I-V curve has a smooth shape with three distinct parts: * a “horizontal leg” (slightly sloping down); © a “downward leg” (approaching vertical); * a bend or *knee’ in the curve between these two regions. In a normal curve, these three parts are smooth and continuous. The slopes and the shape of the knee depend ‘on cell technology and manufacture. Crystalline silicon cells have sharper knees; thin film modules usually have rounder gradual knees. ‘A number of factors can influence the shape of an I-V curve. Figure D.1 illustrates the main types of deviation that may be present. These shape variations may be present individually or in combination j (Ungp Veo) “I Peco steht , ‘ A Fill factor Rounder knee Current \ Low voltage Normal cove — Voltage NOTE The numbers 1 to 6 indicate curve shape variations that are described in D.2 to 0.7. Figure D.1 ~ I-V curve shapes ‘Small deviations between the measured and predicted I-V curves are to be expected, given the normal uncertainties associated with the measurement of irradiance, temperature and voltage. Small variations between PV modules, even of a given manufacturer and model, will also have an effect. Shading and soiling will also impact the shape of the curve. When deviations are seen, a check should first be made to ensure that difference in shape between the measured curve and that predicted is not due to measurement errors, instrument set-up problems or due to an incorrectly entered module / string data.D.2 Variation 1 - Steps or notches in curve Steps or notches in the I-V curve are indications of mismatch between different areas of the array or module under test. The deviation in the curve indicates that bypass diodes are activating and some current is being bypassed around the internal cell string protected by the diode (string unable to pass the same current of other strings). This can be due to a number of factors including the following * Array or module is partially shaded * Array or module is partially soiled or otherwise obscured (snow, etc.) * Damaged PV cell / module. * Shorted circuited bypass diode. NOTE Partial shading of even just one cell in @ module can cause the associated bypass diode to turn on and ‘cause a notch in the curve, D.3_ Variation 2 - Low current ‘A number of factors can be responsible for a variation between the expected current and the measured current. These are summarized below. Array causes: * Uniform soiling. * Strip shade (modules in portrait orientation). * Dirt dam (modules in portrait orientation). * PV modules are degraded. NOTE Strip shade and girt dam effects have an effect similar to uniform soiling, because they reduce the current of all cell groups by approximately the same amount Modelling causes: + PV module data incorrectly entered © Number of parallel strings incorrectly entered Measurement causes: + Irradiance sensor calibration or measurement problem + Irradiance sensor not mounted in the plane of the array * Irradiance changed between irradiance and |-V curve measurements. + Albedo effects cause irradiance sensor to record overly high irradiance. + Irradiance is too low or sun is too close to the horizon NOTE While the variation shown on the diagram above Is a current lower than expected, itis also possible to find that the measured value is above that predicted by the model I-V curve,D.4 Variation 3 - Low voltage Potential causes for a variation in voltage include the following, Array causes: Conducting or shorted bypass diodes. Wrong number of modules in PV string Potential Induced Degradation (PID). Significant and uniform shading to whole cell / module / string. Modelling causes: * PV module data incorrectly entered. * Number of modules in string incorrectly entered. Measurement causes: + PV cell temperature different to measured value. As cell temperature affects the voltage from the PV module, a disparity between the actual cell temperature and that measured (or assumed) by the I-V curve tracer will cause this shape defect. In such cases a check of the cell temperature measurement method should be instigated before proceeding (e.g. checking a temperature sensor is still attached to the module). A group of strings measured in close succession will often exhibit slightly different amounts of deviation compared with the predictions of the PV model. This is to be expected given that the temperature is usually sensed at a single module and the temperature profile of the array is non-uniform and varying with time. However, if a single string shows substantially more deviation than the others, this is an issue, particularly if the deviation corresponds to module VoclN where the modules have NV bypass diodes NOTE While the variation shown on the diagram above is a voltage lower than expected, iti alse possible to find that measured value is above that predicted by the model I-V curve, D.5 Variation 4 - Rounder knee Rounding of the knee of the I-V curve can be a manifestation of the aging process. Before concluding that this is the case, check the slopes of the horizontal and vertical legs of the I-V curve. If they have changed, it can produce a visually similar effect in the shape of the knee. D.6 Variation 5 - Shallower slope in vertical The slope of the latter portion of the I-V curve between the maximum power point (Ymmpp) and Voc is influenced by the series resistance to the circuit under test. An increased resistance will reduce the steepness of the slope in this portion of the curve.Potential causes of increased series resistance include: + PV wiring damage or faults (or cables insufficiently sized). ‘+ Faults at module or array interconnects (poor connections). + Increased module series resistance. When testing arrays with long cable runs, the resistance of these cables will influence the curve shape and can have an impact on the curve as described here. If this is suspected, either the model can be adjusted to allow for these cables; or the test can be repeated closer to the array (bypassing the long cables). Where this error is noticed on a curve, special attention should be paid to the quality of the wiring and interconnections within the solar circuit. This error can indicate a significant wiring fault or subsequent damage or corrosion affecting the array circuit. Increased module series resistance can be due to high resistance faults within cell interconnects or within the module junction box — due to degradation, corrosion or manufacturing error. ‘An IR scan, as described in the Category 2 test sequence, can be a useful tool to identify high resistance faults. D.7 Variation 6 - Steeper slope in horizontal leg A variation in slope in the upper portion of the I-V curve is likely due to: © Shunt paths in PV cells. ‘+ Module /,. mismatch, ‘+ Tapered shade or soiling (e.g. dirt dams). Shunt current is any current that bypasses the solar cell — usually due to localized defects in either cell or cell interconnects. Shunt currents can lead to localized hot spots which may also be identified through IR testing Differences in /,, between modules in a string can be due to manufacturing discrepancies. If the mismatch is small and randomly distributed across the string, steps or notches may not be Present. While more significant shading will cause steps or notches in the I-V curve, minor shade on some modules in a string or some tapered shade patterns can cause this effect.