One of the questions that frequently arises when discussing green roofs with clients is the feasibility of collecting the runoff from the roof for various uses in and around a building. 

Often this interest is focused on reducing the watering requirement for the roof – after all, in drought prone areas watering your roof seems to be wasteful, right?  Why not install a white roof that you do not have to water?

We will not address the green roof vs. white roof debate in this article, but we will explore the possibility of collecting surplus rainwater for use in watering the roof system when it needs it.  This water could also be used around the landscape, or potentially even in the building.

The latter option, using green roof runoff in buildings, almost always raises eyebrows.  Would you want to be drinking water that just passed through soil? 

We have heard differing opinions on this topic, and were very interested when we found a recent report from the Texas Water Development Board entitled: Effect of Roof Material on Water Quality for Rainwater Harvesting Systems. This report is available for free, and has some applicable findings to our field.

As the title mentions, the study centered on the effect that different roof treatments have on water quality of collected rainwater.  A survey of the most common roofing materials in Texas was completed, and found that the three most common were asphalt-fiberglass shingles, Galvalume® metal panels, and concrete tiles. 

Three 8 foot x 4 foot test roofs were constructed at the Lady Bird Johnson (LBJ) Wildflower Center in Austin to test these three roofing treatments.  Some readers may already know that the LBJ Wildflower Center is home to one of the first green roof research plots in the state, testing various commercial green roofing systems and other cool roof technologies over the course of many a hot, Texas summer.

In addition to monitoring runoff from these three new test roofs, the team also decided to monitor runoff from one of the green roofing modules on site, as well as white roof from the same study.

The team followed best practices for collecting rainwater, fitting all collectors with first flush systems to capture and sequester the dirtiest water before allowing additional water to flow into the two subsequent storage tanks.  Such first flush systems are a standard feature on most rainwater harvesting systems because the initial surge of rainfall across the roof washes any contaminants on the roof into the collection system.

The collected water from each treatment was tested for a number of water quality indicators, which are discussed below.

Results

Conductivity

Conductivity is the ability of a substance to conduct an electric current, and it turns out that extremely pure water is less conductive than water with dissolved salts or other impurities floating around.  This is because these impurities tend to ionize while in solution, and ions (non-neutrally charged particles) allow electric current to flow more freely through a medium.  The green roof used in this study produced the most conductive water samples.

It is hard to draw a finite conclusion from this, except to say that the collected green roof runoff had more impurities in it.  Impurities are not necessarily good or bad, highly conductive water often simply has a more mineral taste to it.  If the Wildflower Center Staff was using city or well water from the Austin area to water this roof, it is possible that minerals from that water were deposited into the green roof soil and are subsequently leeching out into the runoff.

Turbidity / TSS

Turbidity is one measure of how many suspended solids are floating around in a fluid.  These solids, in high enough concentrations, will cause a clear fluid like water to appear cloudy or opaque.  One might expect that water passing through soil would be full of suspended solids; shouldn’t some of that dirt be washing out?

It turns out that the green roof runoff had the lowest turbidity out of any roofing treatment tested:

“The lowest turbidity values were found in rainwater harvested after the first flush from the green roof…an indication that green roofs can effectively filter out particles. It is important to note, however, that all roofs yielded higher turbidity values than the 1 NTU maximum recommended for potable use of harvested rainwater, which is the same as the USEPA’s guideline for filtered surface water.”

The results were similar for Total Suspended Solids (the measure of how much stuff collects on a filter when a sample of water is poured through it), the green roof showed the lowest values. 

Take home point: if you’re going to drink your collected rainwater, you should run it through a sediment filter to remove suspended solids.

Nitrates

Nitrates in water come from a number of sources, and have potential health implications especially for very young children.  While health effects in adults are mild and rate, the presence of nitrates in water usually indicates anthropogenic contamination of some kind in the water.

Because the water samples in this study were collected from roof surfaces, they are less susceptible to the more common vectors of nitrate contamination (fertilizer, human/animal waste).  Almost all of the samples were below the EPA limit, and the green roof had the lowest values of all treatments tested.

Dissolved Organic Carbon (DOC)

DOC tests measure the amount of organic material in a solution.  Organic materials from plants and animals can break down to such small sizes that they can become dissolved in water.  Because DOC is derived from living organisms, it makes sense that the water collected from the green roof had the highest levels of DOC’s.

The big implication here is on the use of chlorine to disinfect collected rain water.  Chlorine reacts with DOC to form all sorts of not-so-good-for-you byproducts.  To avoid this problem all together, collected rainwater from a green roof should be treated with a disinfectant other than chlorine before use.  UV light “filters” and ozone bubblers are proven alternatives to Chlorine.

Coliform

Runoff from the green roof was lowest in both Total Coliform and Fecal Coliform compared to the other treatments, but was present in all treatments after the first flush. Coliform can be eliminated with a UV or ozone disinfectant system, but a green roof goes a long way towards eliminating it before filtration.

Heavy Metals

Roof runoff was also tested for Aluminum (Al), Iron (Fe), Copper (Cu), Zinc (Zn), Lead (Pb), and Chromium (Cr).

“For all rain events, rainwater harvested after the first flush from the green roof consistently showed the lowest concentrations of Al, Fe, Cr, and Cu…the highest Zn concentrations were seen in the harvested rainwater after the first flush from the green and metal roofs; elevated Zn concentrations from the green roof might have been from the solder in the scupper gutter.”

The solder in the gutter was also a potential source for the lead found in the green roof runoff during one rain event.

These are great results, but further study is needed to determine if elevated Zinc and Lead levels were due to contamination outside of the roof of not.  This question highlights the need to be cautious of what materials your harvested rainwater flows across on its way to your tanks, many roofing materials are made from these metals (or alloys containing them).

Considerations

While this study is very informative, it only analyzed a small number of rain events on small test roofs.  Furthermore, the green roof and white roof were almost flat, while the other three treatments were inclined at just over 18 degrees.

The team also found extreme variability when their analysis were scaled up to full sized residences located in different areas of the city, suggesting that geographical location plays a role in determining what ends up in your collection tanks.

Conclusion

So how does this study affect our thoughts about harvesting rainwater from a green roof? 

It would seem that most harvested rainwater can easily be made safe to drink if the levels of heavy metals are not too high.  In many cases, runoff from a green roof enters the filtering/disinfecting process cleaner than water harvested from many popular roofing alternatives – especially with respect to heavy metals.  We should note that water collected from green roofs still benefits from a first flush system.

One should keep in mind that the volume of rainwater collected from a green roof will be less than the volume collected from a less-pervious roofing material.  Many studies suggest that around 80% of the raindrops that fall on a green roof never leave the roof, even on extensive (thin-soiled) systems.

For landscape watering applications, collecting green roof runoff seems to be a no-brainer.  You effectively increase the roofs ability to mitigate stormwater from 80% to nearly 100%, while providing a sustainable, low cost source of irrigation water for the roof during the summer months.  Because less water is captured in such a system compared to a metal or shingle roof harvesting system, less money will be needed to purchase and install a properly sized collection system for a green roof.