PhD thesis defense of Jorne Carolus
Oct 16, 2019 - 15.00 uur
Agoralaan Gebouw D
Lokaal auditorium H2
dr. ing. Jorne CAROLUS
Jorne Carolus invites you to the public defense of his doctoral thesis entitled:
"A study on potential-induced degradation: from conventional to emerging photovoltaic technologies".
Promoter is Prof. Dr. Michaël Daenen. Copromoter is Prof. dr. Ward De Ceuninck.
Photovoltaic (PV) energy production is the fastest growing renewable energy source with approximately 100 GW installed in 2018, which brings the cumulative total to over 500 GW. This is the result of an ever-decreasing levelized cost of electricity (LCOE) for PV, mainly attributed to: i) gaining efficiencies, ii) lowering PV module production costs, and iii) improving lifetime. Especially lifetime of PV systems, and thus reliability, has been gaining importance recently. Different types of degradation are studied to assess the power generation during the lifetime of a module. Potential-induced degradation (PID) of PV modules is an important degradation mechanism caused by a high system voltage that drives ion drift towards the solar cell. These ions affect the proper functioning of the silicon solar cells and have been shown to induce rapid and significant performance losses of up to 50% at module level within one year in the field.
Within this doctoral research, we firstly obtained a better view on the presence of PID of the shunting type (PID-s) in the field. This was achieved when a PID test campaign including 49 different PV modules originating from the field was performed. After 96 hours of PID stress, 78% of the PV modules showed a degradation in maximum power output of over 5% and only one-third of the tested PV modules showed a performance loss of less than 20%. In addition, all PV modules underwent a PID recovery test by reversing the polarity of the high voltage source while residing the same environmental conditions. The modules showed an almost-full recoverability as long as the performance loss was under 85%. Once this threshold was exceeded, an irreversible behaviour was observed, which underlines the importance of detecting and curing PID-s before the point of no return.
Secondly, we investigated the impact of PID-s on the different electrical parameters of the 49 different PV modules. Notably, the influence of PID-s on the different parameters of a PV module all show the same trend of degradation. The results indicated that PID-s causes a decrease in shunt resistance, influencing the fill factor in an early stage. When the modules degrade even further (PID-s levels over 40%), the open circuit voltage will be decreased, followed by a decrease in short circuit current. Next to the fill factor, it is possible to recognize PID-s in an early stage by a decrease in voltage and current at maximum power point. Next, we presented the impact of the stress voltage on the behaviour of PID-s. Three different stress voltages (-200 V, -600 V and -1000 V) were applied and intermediate characterisation measurements were conducted. The results clearly indicate that the degradation behaviour in time proceeds according to an S-shaped curve. Interestingly, the stabilized degradation level seems to be the same for all stress voltages and tends to approach 100% degradation when the foil-method is applied. On the other hand, the maximal degradation rate shows a quadratic relationship with the stressing voltage. Furthermore, we have shown that a full recovery is possible when a lower curing voltage than the stress voltage is applied. This is promising for the residential scale PV installations since the maximal curing voltages are often limited due to inverter warranty.
The previous findings were focussing on PID-s in standard n+/p crystalline silicon solar cells, the dominating technology in todays PV market. However, following the recent shift towards mass production of advanced cell technologies (i.e. PERC and PERT), bifacial PV modules are eventually becoming a commercial reality and enter the mainstream. Therefore, we investigated PID stressing methods for bifacial solar cells in a glass/glass packaging. From our results n-PERT solar cells in a glass/glass packaging under bifacial PID stress suffer from a PID mechanism which is only evolving at the front side of the solar cell. However, when the modules underwent monofacial PID stress testing from the rear side only, a similar but slower behaviour was observed, i.e. the PID mechanism evolves at the front side of the solar cell. The explanation of such a faulty interpretation of bifacial solar cells in a glass/glass packaging undergoing monofacial PID stress can be found in the test setup: it is the result of an unintended electric field arising between the grounded inside of the climatic chamber and the solar cell at a negative potential (in our case -1000 V). This work might be interesting when the PID test standard IEC62804 gets updated for bifacial PV modules. Therefore, we included three possible measures which can be considered when monofacial PID stress tests of bifacial PV modules in a glass/glass packaging are conducted: (i) shorting the cell and the non-stressed glass cover side (shielding), (ii) using a floating high voltage source, (iii) replacing the glass cover by a transparent backsheet/frontsheet.
Next to PID testing bifacial n-PERT solar cells, we investigated the physical origin of bifacial PID in bifacial mono crystalline silicon p-PERC solar cells. Our investigations show that bifacial p-PERC solar cells suffer from a combination of both PID-s, occurring at the front/emitter side of the solar cell, as well as PID of the polarization type (PID-p), occurring at the rear side of the solar cell. Since PID-s is affecting the shunt resistance, both the front and the rear side illumination measurements of the solar cell are degrading according to the same trend. PID-p on the other hand has a limited effect on the front side illumination measurements. From these results, it can be stated that the glass/glass packaging and the lack of blanket metallization at the rear side renders such module types more sensitive to PID. Additionally, it has been shown that both failure modes can be easily distinguished by IV and EQE measurements. In the IV characteristic PID-s is witnessed by a loss in fill factor due to a decrease in shunt resistance while PID-p is indicated by a decrease in the short circuit current and the open circuit voltage while the fill factor stays quasi unchanged. Front-side EQE response also shows a specific signature for both degradation mechanisms. PID-s expresses itself by a slight decrease in the short wavelength region (300-400 nm). Whereas PID-p can be recognised by a rather significant drop in the long-wavelength region (800-1200 nm) in front side EQE measurements.
Finally, we reported results of PID tests on perovskite solar cells for the very first time. Metal halide perovskite solar cells have become a major competitor in the run to lower the LCOE of photovoltaic (PV) systems. Commercialization of this new technology mainly depends on the long-term stability of such devices, for which potential-induced degradation (PID) may represent a factor of detrimental impact. The solar cells are found to be extremely susceptible to PID: 18 hours of high voltage stress yielded a performance degradation of up to 95%, which mainly resulted from a decrease in short circuit current. Our results also uncover near full PID recoverability and pave the way towards further research into its mechanisms, kinetics, and mitigation.