When it comes to assessing the impact of low-impact development (LID) and green infrastructure (GI) practices, site development teams are frequently asked to determine the percentile of rainfall events that can be managed by the site development solution.
For instance, the LEEDv4.1 Rainwater Management credit asks project teams to retain on-site the runoff from the developed site for, at minimum, the 80th percentile of regional or local rainfall events. The credit further rewards teams that can manage the 85th or 90th percentile of rain events.
Similarly, SITESv2 requires that site development projects retain the precipitation volume from the 60th percentile precipitation event, with an additional credit that awards points for retaining or treating the 80th, 90th, and 95th percentile precipitation event on-site.
Figure: If a site development strategy can manage the 80th percentile storm event for its location, then all the precipitation from 80 percent of the historical storm events on record should be able to be captured on-site. Image courtesy of Daniel Overbey.
So, how is the percentile precipitation event determined? Methodologies may vary and there are various resources available in the marketplace today to help assist project teams with such calculations, but a best practice aligned with LEED and SITES is to conduct the following four steps.
1. Obtain rainfall data for project location.
Document at least 10 years of historical rainfall data (or as much historical data as possible) representative of the project climate conditions based on proximity to site, elevation, region, etc. In general, a 30-year period of rainfall record is preferred for the analysis. Such raw data are collected by most airports.
2. Organize the historical rainfall data appropriately.
Organize the historical rainfall data such that daily totals are arranged. Then, import the data into USGBC's LEEDv4.1 Rainfall Events Calculator or another spreadsheet.
Again, a minimum of ten years of precipitation record data is needed to determine the percentile event for a project. Thirty years or more is recommended. This data is recorded daily (i.e., 12:00am to 11:59pm) for every single day in which precipitation was recorded. In many instances, this will be a LOT of data.
3. Review the records, looking for anomalies, and identify and remove any erroneous or flagged data points.
It is also unnecessary to retain daily data for days in which 0.0 inches (i.e., negligible amounts) were measured.
4. Determine the percentile rainfall amounts.
The LEEDv4.1 Rainfall Events Calculator automatically determines the various percentile rainfall amounts. If using another spreadsheet software, apply a percentile function (or similar) to obtain results.
My firm was part of a team that recently completed a LEED Gold certified library in the International Marketplace district of Indianapolis, Indiana that features a site development strategy capable of managing a 98th percentile storm event. Using data from the nearby Indianapolis-Eagle Creek Airpark weather station, our project team determined that the area averaged 122 days of measurable precipitation every year for a total annual average of 42.4 inches. In terms of assessing our on-site rainwater management strategy, our civil engineer determined the following:
- 80th percentile storm event: 0.80 inches
- 85th percentile storm event: 0.93 inches
- 90th percentile storm event: 1.14 inches
- 95th percentile storm event: 1.52 inches
- 98th percentile storm event: 2.05 inches
- 100th percentile storm event: 9.99 inches
Note the enormous jump between the 98th and 100th percentile storm event. It is quite common in many locations to have occasional flash-flood events. By definition, managing a 100th percentile event would require on-site sequestration that could handle even the most extreme precipitation event. This is not practical in the majority of circumstances.
EPA 841-B-09-001 Technical Guidance on Implementing the Stormwater Runoff Requirements for Federal Projects Under Section 438 of the Energy Independence and Security Act (December 2009) offers a detailed explanation regarding the calculation of percentile storm events using daily rainfall records.