Improved reliability of solar photovoltaics

The rapid reduction in the price of silicon solar modules has resulted in a significant shift in the relative contributions to total system cost. The levelized cost of electricity (LCOE) from PV is now dominated by the balance of system components, including inverters, mounting structures and installation costs. As many of these costs are area-dependent, the simplest path to further reducing the LCOE is to increase the power output of the module. This has resulted in a premium on cell efficiency, driving the industry transition to new cell architectures. However, as outlined by Peters et al., the positive impact of increased efficiency is rapidly diminished if the cell degrades at a faster rate, as shown in Figure 1 [1, 2]. This graph clearly shows that the benefits of higher efficiency and lower manufacturing costs can quickly disappear if the degradation rate is increased. For example, the baseline silicon PV panel with an initial efficiency of 20% and an annual degradation rate of 0.6%/year has a similar LCOE as a PV panel with an unrealistically high initial efficiency of 70% and an annual degradation rate of 5% assuming a 25 year lifetime (assuming constant module costs). Similarly, a higher degradation rate requires significantly lower PV module costs to achieve a similar LCOE. Thus, it is crucial to consider possible degradation mechanisms in new, higher-efficiency cell designs.

*Figure 1: Trade-off between degradation and (left) initial efficiency and (right) module costs at a LCOE of 5.2 ct/kWh taken from Peters et al.* [2].

The emergence of passivating contact solar cells in recent years has again raised the bar for production-capable efficiencies, the most promising of which: TOPCon (Tunnel Oxide Passivated Contact) and silicon heterojunctions (SHJ), have enabled initial efficiencies of up to 26.1% [3] and 26.7% [4], respectively, for single-junction devices. These structures rely primarily on the surface passivation capabilities of nanoscale thin film technologies, such as doped amorphous silicon (a-Si) and polysilicon on thin oxides. These solar cells have often been regarded as the last step in single-junction silicon technology development, with the expectation that a silicon solar cell with passivating contacts would provide the foundations for the bottom cell of a silicon-based tandem device (e.g. Si-Perovskite or III-V on Si). Studies have already achieved world records, including a 30% Si-Perovskite device, utilising an n-type SHJ as the bottom collector. However, as shown in Figure 1, there are still significant challenges with the stability of perovskite solar cells that must be addressed before these tandem solar cells will result in lower LCOE.

Figure 2 shows some data from PV Evolution Labs (PVEL) [5] which is a leading reliability and performance testing lab for downstream solar project developers. The data presented in Figure 2 are based on 10 years of testing and the most recent results on potential induced degradation (PID) are indicated in green. It should be noted that all products tested by PVEL are intended for end-use and not development samples, i.e., their data is a good snapshot of the quality of product on the market. Figure 2 shows that 2021, the latest year of testing, was the worst in terms of PID degradation even though all modules were labelled as “PID resistant”.

Figure 2: (top) Schematic illustration PV module power loss for various failure modes as a function of time and (bottom) power loss due to potential induced degradation observed after standard testing of “PID free” modules. Data courtesy of PVEL [5]

Figure 3 shows the three failure modes which are all present in passivating contact solar cells and will be investigated in this project. The left column shows that we have rapid tests for all these failure modes, often orders of magnitude quicker than other tests, while the right column shows that we have mitigation solutions for all of them.

We have shown that surface related degradation (SRD) is a very slow degradation process which becomes more severe for lightly doped surfaces like the ones present in TOPCon contacts [6, 7]. The SRD degradation extent seems to scale with firing temperature [Fig. 2(a)] and is thermally activated [Fig. 2(a-b)], which points towards the involvement of hydrogen. We use that understanding to accelerate its detection, as well as its mitigation.

Potential induced degradation (PID) is making an undesired comeback [5] in PV modules, which is partly related to the move to bifacial solar cells which makes both surfaces susceptible to PID. We have developed a rapid PID test [Fig. 2(c)], that applies salt onto the solar cell surface before stress testing. This allows for the detection of PID-prone cells in less than 30 minutes. In addition, we have shown that the application of atomic layer deposited films can mitigate PID-shunt [Fig. 2(d)] as these layers act as a diffusion barrier (e.g., for sodium).

In our recent work, we have observed very significant degradation of heterojunction solar cells after a damp-heat test [Fig. 5] compared to PERC solar cells. These are damp-heat tests performed on unencapsulated solar cells which significantly accelerated the degradation. We also have patented technology that can prevent many of these failure modes.