Read Part 1 of this post here.
2. Bigger is not better
Like most everything we design, when big things fail, they fail in a big way. Bioretention works best in applications where the drainage area to an individual cell is less than about a quarter acre. Larger bioretention areas are more likely to fail due to erosion because of larger flows, creation of low spots due to variations in the surface, and clogging of the surface layer. This can be avoided by dividing the area into multiple cells with smaller drainage areas.
The largest bioretention area I know of is an example of this. Designed by others, it collects runoff from several acres at a highway rest stop. The measures designed to evenly distribute incoming flow have been overwhelmed by large flow rates, the engineered soils mix layer does not drain quickly enough, and the plantings have died because the roots are sitting in saturated soils.
An example on the other end of the spectrum is the bioretention area in my backyard. It collects runoff from no more than 1,000 square feet of pavement. The engineered soil mix is only 8 inches thick and is a mixture of topsoil and compost that falls far short of the usual specifications, the underlying soils are clays that are not highly infiltrative, and it is grossly undersized per the design manuals. However, it functions beautifully.
3. You need a contractor with knowledge and experience
Bioretention is not a landscape feature nor is it normal site construction. It is an engineered stormwater structure that requires both exactness and a lighter touch. The surface area, elevations, and layer thicknesses on the drawings are very important. If the construction is treated as landscape planting area, these are typically not met, resulting in failure. If the underlying soils and the engineered soils mix are compacted, as is done for base and fill soils elsewhere on the site, the bioretention area will not allow water to infiltrate, resulting in failure. If the engineered soil mix is not protected from sediment during construction, it will clog and result in failure.
A contractor with experience understands the methods needed to avoid these issues, such as the type of excavator bucket, soaking the engineered soil mix to get it to consolidate but not compact, and the critical need for accurate layout. For this reason, I require a pre-installation conference on every project. This approach has worked well and helps me to be able to certify the BMP at the end of the project.