Performance based design of LID stormwater systems
Increasing land development is leading to more runoff from rainfall events which strains storm sewer systems and leads to increased flooding. This problem will be exacerbated by the increase in the frequency of severe storms due to climate change. One approach to reducing runoff is through the use of LID technologies such as porous pavements and green roof systems, to retain and dispose of rainfall onsite. However, the hydraulic and hydrologic behavior of these technologies is poorly understood such that it is not easy to use them in performance based stormwater management designs. Our research goal is to address these shortcomings of LID technologies.
Green roof systems
Green roof systems are becoming increasingly popular LID technologies because of their many water quality, thermal, and aesthetic benefits. However, their hydrologic benefits are largely ignored due to a lack of simple hydraulic and hydrologic design and analysis models. The research team is currently examining the behavior of modular green roof systems with the goals of (1) improving their design to maximize rainfall retention and detention capacity and (2) developing simple design and analysis methods so that they can be incorporated into stormwater management plans.
Porous pavement hydraulic behavior
Our research team has undertaken a number of projects to examine the hydraulic behavior of porous pavement systems. Will Martin developed an image analysis technique to measure the porosity of a pavement sample and then examined the relationship between porosity and permeability for porous concrete (Martin et al. 2013, Martin et al. 2014).
Will also examined the reduction of infiltration into the soil that underlies a porous pavement structure due to the blocking effect of the aggregate subbase. Will’s results showed that the infiltration capacity of the soil is reduced, but by a factor that is less than the percentage of the soil in contact with the aggregate (Martin et al. 2015). Tripp West conducted a set of steady-state head loss measurements in porous pavement samples to show that the non-linear term in the porous media flow equations is significant when measuring the behavior of cast-in-place pavements (West et al. 2016).
Patrick Murphy ran a series of experiments to examine the behavior of porous pipe underdrains that are used to speed up draw down in the porous pavement sub-base when soil infiltration capacity is low. Patrick’s experiments showed that for cases where the water surface is above the top of the aggregate, such that the dominant flow direction within the aggregate is vertical, then the hydraulic conductivity of the aggregate plays no significant role and the pipe behaves like an orifice (Murphy et al. 2014). A follow up computational study by Tanjina Afrin has confirmed the experimental results of Patrick and shown that the bulk of the flow through the pipe enters through the pipe side walls in the last few feet of pipe just upstream of the outlet. Tanjina also completed a parametric study to investigate the influence of pipe diameter, pipe wall inlet area, aggregate depth, and the width of the aggregate channel on the pipe discharge. She also improved our understanding the the hydraulics of pipe free overfall’s.
Tripp West investigated the behavior of a porous pavement lane (such as a bike lane) next to an impervious lane (such as a traffic lane). Tripp investigated the distance that the runoff from the impervious lane would flow over the porous pavement before being fully infiltrated. Tripp also ran a series of constant flow rate permeameter tests to characterize the non-linear hydraulic behavior of porous concrete samples. Tripp’s results indicate that standard falling head tests will under estimate the hydraulic conductivity of a sample pavement.
Porous pavement hydrologic characterization
Will Martin developed a new technique for characterizing the hydrologic behavior of undrained porous pavement systems in terms of an effective curve number (ECN). Will developed a method for plotting the ECN as a function of the infiltration capacity of the soil and the effective storage depth of the pavement system. The plots developed vary with local IDF data so we have developed an online application that allows anyone to enter their local 24 hour rainfall depths for a range of return periods and get back an ECN plot for their location (Martin & Kaye 2014). This ECN plot can be used for preliminary sizing of a porous pavement during the early stages of design.
The ECN model was then extended by Will to characterize porous pavement behavior in terms of an initial abstraction and a post abstraction runoff coefficient. This ‘broken line’ approach more accurately reflects the actual rainfall-runoff behavior of porous pavements. Both the initial abstraction and the runoff coefficient are independent of the local rainfall patterns and are, therefore, universally applicable (Martin & Kaye 2015).
Integration of LID into site hydrologic design
We are currently using the results of the research described above to develop methods for integrating LID technologies into site drainage designs. The goals are to (1) establish design methodologies to easily and efficiently integrate these distributed stormwater management practices into standard drainage designs and (2) assess the benefit of this approach at both the land development site scale and at a larger municipal scale. More details on the work can be found in the tabs above or by clicking on the links below:
- Green roof system hydraulics and design
- Hydraulics of porous pavement system
- Design of land developments using LID technologies
- Municipal scale use of LID stormwater technologies
The project is funded by SC Sea Grant.