That’s a great question!

Our team’s central objective is to help make green roofs more feasible for implementing on a large scale in dense urban areas. With our collective research we have been developing methods for monitoring, modeling and designing green roof systems that could be adapted to different urban settings and needs, while simultaneously addressing issues that can inhibit city wide implementation such as cost, weight and hydrologic performance.

Since our research is being applied in real-­­time to help address an existing and ever growing problem, a comprehensive team across many disciplines allows us to dynamically respond to each others research needs and inputs. This helps to scale this up as a practical technology and allows for data driven design. For example, the development of the vine trellised canopy system will benefit from low cost monitoring methods as well as the alternative substrates research, while the alternative substrates will require input from the understanding of structural restrictions and monitoring metrics.

By approaching our overall research goal from various disciplines, we are able to develop unique and more comprehensive methods for realizing green roof potential.

Hi Peter, your question about possible damage due to wind is also on the minds of a lot of building owners, thank you for bringing this up!

The 7 full-scale green roofs we monitor in New York City were hit by hurricane Irene in 2011, followed by hurricane Sandy in 2012, and several additional extreme events with large gust speeds since their construction. In each of these cases we have not observed any noticeable damage to the green roof or the rooftop structure itself. A number of installers/manufacturers have now started providing wind tunnel testing certificates to identify the uplift resistance of their systems. In addition, building owners who may still have concerns, generally those in heavily trafficked urban areas who worry about wind carrying substrate or other materials off the roof, can also opt to have a geo-synthetic netting material installed over the system.

As far as the vine-canopy design, some vine types will not necessary be able to withstand harsh rooftop conditions, including elevated wind speeds which would hinder their ability to grip and climb. We hope to address this through informed design, where Rob is currently testing 8 different vine species in duplicate that show promise. Additionally, he will be testing ways to make the substrate and supporting framework more resilient – exploring the effect of sub irrigated planters, cover crops, and subterranean ecologies.

Having the opportunity to refer to the large body of work in Germany has been instrumental for our research team and many other groups in North America. For instance, the concept of quadratic relationships between annual rainfall and retention identified by Mentens et. al. 2006 (a review of 18 German language publications), served as the basis of a model Dan and I have constructed for event-based analyses. Similarly, I have been using the 2008 German FLL guidelines, which indicate acceptable ranges of substrate properties, as a means to help identify viable alternative substrates.

While these publications and many other international studies help guide our work, we also recognize that green roof behavior varies quite a bit depending on regional climate, construction method and plant type. Therefore, observations and recommendations for green roofs in Germany (or elsewhere) may not necessarily hold true in North America or New York City. Sometimes this makes direct comparisons difficult, but we do try to build up from the large body of already existing knowledge.

Hi Mary! Connecting the dots between observed green roof behavior and the potential mitigation of runoff problems at the city-wide scale is an incredibly important topic, and one which we have dedicated a considerable amount of time to, so we really appreciate this question!

A couple years ago our group worked on estimating the total area that could accommodate green roofs based on the rooftop density and typical building typologies of neighborhoods throughout New York City (NYC). We found that roughly 20% of NYC rooftop space could be retrofit, corresponding to 8% of the total NYC area, which is about 63 square kilometers (details provided in Geo-Strata 12(2) p.30-32,34). While performance varies between different green roofs, multiple observation methods we use suggest that annual retention in the range of 40-60% is common for current NYC climate conditions. Given that NYCs average annual rainfall depth is 1.3 meters (NOAA Central Park, NYC 1971-2010), the city-wide potential rainfall capture may be on the order of 30-50 billion liters. This storage would be roughly double the estimated 20 billion liters (2008 NYC Stormwater Management Plan) of combined sewer overflow (CSO) pollution in NYC!

Although these calculations are useful in understanding the magnitude of problem and a solution, it is important to note that the mitigation of CSO is dependent on peak runoff rates, for which volume reduction has implications, but are mainly influenced by rainfall characteristics and properties of individual sewersheds. Considering that relationships between temporal rainfall characteristics and green roof hydrographs are still not well understood, and the fact that up-to-date sewer maps in NYC are not widely available (under lock and key in some cases or non-existent in others), the direct assessment of impacts on CSO pollution is incredibly difficult. We are still in the process of using the 2+ years of data from our field monitoring program (which now includes 7 full-scale green roofs throughout NYC), including Raha’s work on models which use low-cost moisture probe readings, to better address this topic. More specifically, a large effort is being made to couple intensity-duration-frequency rainfall curves and event-based green roofs models to understand potential peak flow mitigation for various design storms.

Generally, it is accepted that even the full implementation of green roofs will not eliminate CSO events in NYC, since the maximum rainfall capture of individual events is limited by the capacity of the substrate layer. That being said, events as little as 3 mm in one hour have been shown to generate CSO pollution in the most critical NYC sewersheds (areas in Brooklyn, NYC), this is an event size which green roofs often retain completely. Therefore, reducing or eliminating the occurrence of CSOs during small events (0-12 mm), during which CSO pollution is also highly concentrated compared to extreme events, as well as reducing CSO volume during large events, is likely achievable through green roof technology.

Thank you for the question.

It’s a bit confusing as the loading may not change (in fact it may increase), but the resulting stresses decrease. This model of retrofit is aimed at buildings with reserve capacity in columns, but controlling structural strength in the spans – which is common as many building are constructed with the possibility of adding on floors later.

The red in the video illustration represents the maximum stress in the spans due to the moment. The redistribution illustration is nearly accurate with the redistribution shown, though unscaled. Theoretically, if the substrate were applied as point loads over the columns, the added stress in the spans would be zero.

We are still working on determining the necessary footprint for the substrate; for instance, how much soil the vines need to be properly rooted. Once we have this, an accurate percentage of stress reduction will be possible, but the decrease in the spans is expected to be drastic.

Karen, thank you for the kind words!

With regards to the support of plant growth as shown in the video, the substrates have three main functions (1) provide a rooting structure, (2) store nutrients and water for plant use, and (3) maintain adequate aeration and drainage within the rooting zone. On a broader level, substrates help facilitate many of the environmental benefits provided by green roofs. For example, studies have shown that green roof potential for retaining rainfall, improving runoff quality, creating urban habitats, mitigating urban heat island, reducing ambient CO2 levels, and removing airborne pollutants are all dependent on substrate properties themselves (e.g. sorption) or substrate dominated conditions (e.g. vegetation abundance/health). Given their multi-functionality, the most challenging aspect of substrate design (and luckily what I find most interesting!) is figuring out how to optimize properties for sustaining plant life and providing the environmental benefits mentioned above, while also maintaining constructability by minimizing cost and density.

As far as plant dependence on substrate type, the green roof studies I have read suggest that substrate depth and regional climate are the dominant factors for the selection/success of certain plant types. However, I do think it is likely that, based on chemical composition and other properties, different substrates could favor some plant species. In my plant growth trials I only looked at one Sedum species (Sedum acre) which I noticed to be most prominent on full-scale roofs we study in NYC. This is definitely a topic where there is a lot to be explored.

The cost of green roof substrates on a per area basis will vary by market region, manufacturer, and the selected depth. For a roof with 100 mm substrate depth, I’ve seen prices range from $15-25 per square meter. Depending on which plants and geo-synthetics are selected, substrate can account for around a 35-55% of the material costs per area and about 6-10% of total installation price. If alternative materials could offset the cost of expanded minerals by just 25%, this could save $4-6k on a typical 1000 square meter installation, or $22-35k for larger projects such as the US Postal Service green roof in New York City, for example.

Unfortunately, there is no easy over-arching answer for this one. Depending on what kind of structural system you have, determining your reserve capacity could be simple or more involved.

For example if you have prefabricated steel joists (as is common in post WWII construction) you can take photos and measurements and send it into the Steel Joist Institute, who will return a total capacity.

Custom construction may require modeling or direct calculation of strength.

I see, I appreciate your quick response! Thanks and great work on the video

Kathy, thanks for your question! The green roofs we are researching are referred to as “extensive” green roofs, since their substrate depth is less than 100 mm. These types of roofs are designed to be lightweight and require very little maintenance in order to promote retrofit on existing buildings. However, as you point out, reducing the substrate layer depth limits its ability to hold moisture.

To account for this, green roof designers chose a special variety of plants, which can survive harsh rooftop conditions, called “Sedums”. These plants are succulents which (1) require very little substrate for rooting , (2) store a significant amount water from natural precipitation in their leaves, and (3) have a waxy exterior coating to inhibit water loss. Studies have shown that these plant types can survive an incredible 88 days without water and, in some climates, only need water once every 28 days to exhibit growth.

The roofs we monitor are generally not irrigated and have been healthy thus far. In the winter, the sedums do become dormant and lose biomass; however, they regain biomass each growing season. Interestingly enough, studies have also shown that when green roofs are irrigated “weaker” plants survive and propagate, inhabiting areas that may otherwise have been available to more hardy species. This actually makes the roof more susceptible to plant loss during an exceptionally harsh season or should irrigation cease (perhaps due to equipment failure). Generally, it is recommended that irrigation be used during plant establishment and in extreme drought periods.

There is another type of green roof, referred to as “intensive”, which includes roof top gardens and parks that consist of larger plants (which may differ from Sedums) and deeper soils. These types of roofs typically require routine maintenance, including irrigation, but we are researching how to make soil more resilient in terms of infiltration, absorbance, and dry out time. By testing the effect of cover crops, soil amendments, subterranean ecology, and a variety of sub irrigating systems we hope to make a great environment for plants.

In terms of your potted plants – you should check out sub irrigate planters!

This is very cool, and I hope the practice becomes more common. I am curious if you attempt to use native species of plants, or species that are found in the local habitats? For example, I am sure a green roof in Utah would need to look very different from a green roof in New Hampshire. Is the origin of the species utilized for the roof taken into account?