Lack of outdoor data becomes one of the bottlenecks in preparing perovskite solar cell (PSC) technology for the market. The data scarcity is not only due to the complexity of results interpretation (as it is recorded under constantly changing conditions) but also due to a lack of experimental capabilities for outdoor studies. Recently we have developed a rooftop test field in Berlin suitable for the long-term tracking of research-size PSCs. Apart from weather monitoring, it allows performing maximum power point tracking (MPPT) simultaneously on ~100 individual solar cells. We have also built up competencies in device encapsulation that allowed passing the industrial damp heat test (85 °C; 85% RH) and enabled long-term outdoor experiments [1]. With reliable encapsulation and MPPT tracking capability, we are well set to shine the light on the peculiarities of PSC real-world operation. In this contribution, we share our experience based on more than a year-long outdoor data.

Outdoor data allows estimating the relations between weather conditions and instantaneous cell power output. While for the impact of solar irradiance and spectrum the data is close to our expectations, it is not the case for temperature effect. We observed an increase in device power with temperature both on short (several days) and long (several months) timescales. This contradicts laboratory measurements where negative temperature coefficients were recorded, similar to all conventional PV devices. We attribute such a major artefact in outdoor data to the presence of transient processes. Such processes also strongly affect PSC stability but do not reveal themselves when studied under constant illumination, thus, making this standard indoor stability test an unreliable proxy for predicting PSCs outdoor lifetime. We find light cycling tests [2] to be more suitable for explaining outdoor stability results. Unexpectedly, we found for some PSCs much lower stability under cycled light compared to constant light, which is in contrast with commonly observed recovery in the dark (hence longer lifetime). We also analyze the interplay of seasonally changing weather conditions and slowly occurring degradation processes in the outdoor conditions for 4 different p-i-n PSC architectures in parallel with indoor ageing under constant or cycled light. A spectacular variety in ageing behaviours is observed making it hard to find a single conclusion for all the types of PSCs. More attention has to be paid to the transient processes in the future when evaluating new PSCs (compared to simply reporting 5 minutes of MPP data) in order to forecast their outdoor behaviour.