Quantifying future production from photovoltaic systems of all sizes and into all applications is important for LCOE modeling and by extension, tender and PPA bidding. For these models to be useful uncertainty, that is risk of poor performance, must be appropriately weighted. Underweighting the effect of potential poor performance due to low quality installation practices, low quality components, and unexpected weather pattern shifts can lead to underperforming installations that fail to meet economic goals. As tender and PPA bidding are in part based on production expectations, care must be taken to appropriately quantify the risk(s) of poorer than expected performance.

There are solid and observable reasons humans often fail to foresee negative outcomes and outright disaster. Humans are hardwired to highly weigh optimistic outcomes, and prone to choose data or assumptions that back up their biases.  In sum, we cannot imagine a disaster outside of our experience and thus will, most of the time, discount the probability of it. Optimism bias, the tendency to over weigh the potential of positive outcomes, and under weigh the possibility of negative outcomes is difficult to overcome, even for those trained to look out for it. Precommitment to a positive outcome can outweigh data that might indicate the need to consider a less positive outcome, in this case, poorer than expected performance.   Biases aside, as photovoltaic installations become a larger part of global electricity generation, forecasting the risk of poorer than expected production is crucial to the industry’s maturity and continued health.