PDA

View Full Version : Green Roof Modelling in the UK with Infoworks CS



paul.mennell
July 26, 2013, 09:21 AM
I am attempting to model flat and sloped green roofs (extensive 100mm thick media) in Infoworks CS. Any ideass anyone? I am interested in other peoples approaches.
I want to do continious simulations over an average year as well as design events, therefore a range of storms and ADWP's. I understand my correct approach for the UK based on SUDS guidance (CIRIA C689) and available literaturewould be to model a maxmum 5mm inital losses, with a recharge rate of 3mm/day for the first day, then 2mm/day for the second day, giving a total recharge back to 5mm over 48hrs. Any thoughts please?
I will updated this topic as I go along and see what other availalbe models have done. I will also post this discussion on the LinkedIN Innovyse group

Robert Dickinson
July 26, 2013, 11:49 AM
Hello Paul,
I did not see the LinkedIn Discussion but maybe these references will help you in your SUDS modeling. Green Roofs have been part of the EPA SWMM 5 modeling software for the last 3 years as part of it Low Impact Development Controls (LID) on Subcatchments. We also have LID controls in InfoSWMM and H2OMAP SWMM from Innovyze. The EPA just released a Nationwide (USA) stormwater calculator with LID's. The image below shows the Green Roof option. If you go this link
http://www.epa.gov/nrmrl/wswrd/wq/models/swc/ there is a good PDF file discussing Green Roofs as part of the LID and Runoff Modeling
Regards,
Bob Dickinson
44

paul.mennell
July 27, 2013, 12:17 PM
Thanks Bob. The EPA National Stormwater Calculator is great. It’s similarly aimed at developers as per the Wallingford Greenfield Runoff and Stormwater Storage online tool http://geoservergisweb2.hrwallingford.co.uk/uksd/surfacewaterstorage.aspx, but simulates a variety of devices and provides results in a sensible manner. In particular for green roofs, input parameters range between, with default in brackets 1-(4)-12 inch media thickness with .2-(10)-20 inch/hr conductivity. The user manual states the calculator uses a 45% porosity/void space ratio. This 45% value appears to attribute porosity with retention ability, based on the "VIRGINIA DCR STORMWATER DESIGN SPECIFICATION No. 5" paper linked to the green roof window in the calculator, suggesting a green roof provides an annual runoff volume reduction of 45% to 60%. The EPA website provides some tools suggesting similar retention value ranges and this ties in with US academic literature papers such as VanWoert et al (2005) measuring 60%. However annual retention is site and device specific. A paper from the UK, from Stovin et Al (2011), measured 50% retention over the study period with a 80mm thick media at Sheffield. Whilst there are other international green roof results published, I don’t know of any other UK green roof papers reporting local green roof performance. The user manual also states the calculator uses the EPA Stormwater Management Model (SWMM version 5), now with functionality to "explicitly model" more LID/SUDS devices including green roofs. SWMM5 downloadable manuals dont provide advice on typical values or ranges or an approach, but the calculator user guide provides information on how it uses the SWMM engine.
The 50% is the reported annual retention of course, and so small >1yr events would be totally retained whilst larger would be less retained as a green roof has finite retention capability, and it may be saturated already from an earlier event, and so not has enough time for evapotranspiration to recharge the detention capability.
There are a few LID/SUDS models you can get to from the EPA LID model page http://water.epa.gov/infrastructure/greeninfrastructure/gi_modelingtools.cfm.
The Virginia Runoff Reduction Method and EPA's Green Long Term Control - EZ Template quote the same 40 - 60/5% retention abilities,
The Center for Neighborhood Technology Green Values National Stormwater Management Calculator uses a Curve Number (CN) of 75 and Rational Method runoff coefficient of 0.75 (the original calculator did anyway, I couldn’t see what the new one uses). I have seen other CN values usefd for green roofs in various published papers.

Robert Dickinson
July 29, 2013, 09:43 AM
Thanks Paul, for posting a discussion on the Innovyze LinkedIn group, I am following the discussion but have not posted anything yet. I did "like" the discussion, however.

As I mentioned we have Low Impact Development and BMP modeling tools for Subcatchments in InfoSWMM and H2OMap SWMM as discussed on the Innovyze blog in this post http://blog.innovyze.com/2012/07/02/lid-and-bmp-modelling-in-infoswmm-and-h20map-swmm/ The possible LID functions on the Subcatchment include: Rain Barrels, Porous Pavement, Bio-Retention Cells, Swales and Infiltration Trench. In the USA, the Philadelphia Water Department has been building many LID's for green roofs so if you search for PWS and LID you will get some suggestions for Green Roof parameters.

Going further afield here is a link to a water quality study of Green Roofs

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0058020

and a blog post on Salon about why Green Roofs are ineffective in NYC
http://www.salon.com/2013/05/18/whats_the_matter_with_new_yorks_green_roofs_partne r/

paul.mennell
July 31, 2013, 01:05 AM
Robert

I have reviewed the PWS and LID and other internet sites with no InfoWorks relavent parameters or approach apart from one research project htat applied 25mm depression storage (absolute I assume) and 20% runoff coefficient. https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&frm=1&source=web&cd=1&ved=0CDYQFjAA&url=https%3A%2F%2Fworkspace.imperial.ac.uk%2Fcivil engineering%2FPublic%2FTechnical%2520papers%2520B% 2F22B-Trioulet-Impact%2520of%2520Urban%2520Creep%2520on%2520the%2 520Hydrology%2520of%2520a%2520Catchment.pdf&ei=mcT4Ue77B9G0hAe2rIDQBw&usg=AFQjCNGn6EUAmIqCBFErYYBYD17qndRNBw&sig2=qjuW2UVFik-S1gUahh0ALQ&bvm=bv.49967636,d.ZG4&cad=rjt

The depression storage/initial losses seems a bit high and dont think the runoff co-efficient can be applied to any storm. The reference given for this approach is not proovided in full and cant be found.
I have found some SWMM approaches documented in literature but dont think the approach is transferable