While flooding the shorelines of many lakes and rivers and keeping scores of people cooped up indoors, this uncharacteristically wet summer has also put a damper on the cash flows of solar asset owners across Ontario. In what Environment Canada has dubbed the “Year of the Big Wet”, rain and clouds across the province have led to global horizontal irradiance (GHI) for the region tracking well below average, as seen in Figure 1. GHI is the strongest factor in the eventual energy output of a solar PV asset, and as such is directly correlated to revenue from a site. Solar PV asset owners that have relied on long-term weather averages alone to make revenue predictions have had a bumpy ride this year, but it’s so far smooth sailing for those that properly accounted for solar resource variability when planning and financing their projects.
First, a primer on the use of exceedance probabilities in production modelling. An exceedance probability is the probability that an event will meet or exceed a given value. In the field of solar PV asset performance modeling, commonly referenced exceedance probabilities are P50 and P90. The values associated with these probabilities are the energy production levels that will be exceeded 50% and 90% of the time respectively. P50 and P90 analyses are typically performed on an annual basis. Think of a P50 as the amount of revenue a project will have made by the end of its lifetime. Some years will be bad, some years will be great, but they will all average out eventually. A P90 will help hedge the risk of bad years negatively impacting your operations.
The area to the right of the P90 value in Figure 2 represents 90% of the GHI values one would expect to observe in a given year.
In the context of GHI, a P50 and a P90 value would be the GHI that would be exceeded 50% and 90% of the time. A P90 value will always be lower than a P50 value. While the concept between the two values is the same, their determinations require different approaches.
At the foundation of any solar PV asset production forecast, is a meteorological dataset. P50 values for GHI are readily available as they are sourced from the most prevalent meteorological datasets, such as a Typical Meteorological Year (TMY). A TMY weather file uses decades worth of data to produce a representative single year’s worth of weather data, which is in essence a P50. A P50 value is appropriate when predicting the energy output of a project over a sufficiently long timespan, where the year to year weather variability cancels out.
In situations requiring predictable cash flows, relying on a P50 alone is a risky proposition. For example, to adequately service debt, or to fund development of new assets. A P50 would overpredict revenue during low GHI periods, resulting in tighter operating margins. In locations where there is a large deviation from average GHI (think areas with frequent storms and high atmospheric moisture content) the production could be drastically different from the average. For these situations, a P90 analysis is the right tool for the job.
A P90 value is created by understanding the year-to-year variance in a collection of annual weather files. This usually requires access to several years worth of data, to ensure that an accurate variance and average are established. Once the variance of the dataset is known, a statistical analysis is performed yielding a P90 value.
The Canadian Weather Energy and Engineering Dataset (CWEEDS) is an example of a dataset used for P90 analyses. The dataset includes 492 Canadian locations with at least 10 years of data.
ARI has assessed the GHI data seen in Figure 1 to determine where the past year of solar irradiance in Ottawa sits relative to the P90 threshold, see Figure 3.
By noting that the 2016/2017 average GHI exceeds the P90 GHI we can determine that the low irradiance this summer is within the P90 variance of the CWEEDS data for Ottawa. It is worth mentioning that this average falls below the P50 by 4.6%. This means that a revenue stream approximated with a P90 evaluation should be in good shape for the past year of operation. This longer view of the past year of operation helps give perspective on poor Q2 2017 GHI, which has been 13% below P50 for Ottawa and similar across the province.
Performing a P90, or even a P95 evaluation, is frequently recommended and is often a requirement of take-out financers throughout the industry. Apricity Renewables has a strong foundation when it comes to translating weather variability into actionable and bankable information. Combined with a thorough understanding of other modelling variabilities, we aim to give asset owners a clearer picture of the financial viability of their systems over the entire project lifetime. Contact us here to find out how we can add a new level of confidence to your next solar project.
DISCLAIMER: All data and information provided on this blog is for informational purposes only. Apricity Renewables Inc. makes no representations as to accuracy, completeness, currentness, suitability, or validity of any information on this blog and will not be liable for any errors, omissions, or delays in this information or any losses, injuries, or damages arising from its display or use. All information is provided on an as-is basis. Blogs posts do not necessarily reflect the views or opinions of Apricity Renewables Inc. and should not be construed as such.