Conference Agenda

The current National Pollutant Discharge Elimination System (NPDES) Municipal Separate Storm Sewer System (MS4) permits in Oregon (Permits) include significant new requirements for stormwater design standards that impact the design feasibility, facility footprint, and effectiveness of stormwater facilities. The Permits require prioritization of low impact development, green infrastructure, and retention in stormwater design standards. For retention, permittees are required to develop a Numeric Stormwater Retention Requirement (NSRR) that retains stormwater onsite and minimizes offsite discharge of pollutants. This presentation will cover methods and data assumptions used in evaluating local rainfall data to identify appropriate design storms for sizing Permit-compliant stormwater retention and water quality facilities.

For one permittee, an NSRR design storm was selected using the annual average runoff-based method to retain 80 percent of annual average runoff. The analysis considered hourly rainfall data from two local gages. The design storm was estimated using two rainfall analysis methods for comparison: a rolling 24-hour method and a storm-event method. The results were sensitive to the analytical methods and the assumptions used as input parameters for those methods such as inter‑event times and period of record from the two local gages. Design storm results differed in size by as much as three times.

For another permittee, after completing a rainfall analysis, a sensitivity analysis was conducted to look at the impact of a proposed NSRR on stormwater facility footprints for scenarios that included a range of infiltration rates, drawdown time limitations, and development types. Facility footprints varied by as much as four times depending on the scenario.

The selection of a design storm for onsite retention requires an understanding of the nuances and significant implications related to different rainfall analysis methods and associated assumptions. It is also important to understand the development characteristics where the retention design storm will be applied. Design storm selection based off a clear understanding of methods and assumptions will best support goals for sizing facilities. This will allow facilities to meet requirements for runoff treatment, address feasibility constraints for implementation, and right-size facilities to effectively manage stormwater runoff and protect water resources.

The questions we ask and answer in the planning and design of stormwater retrofit projects fall into two general categories:

1) What do we need? We usually call this planning.

2) What do we get? We usually call this design.

In stormwater management the answer to what we need is rapidly changing. When many of our storm systems were first built the answer was “we need to get water away as quickly and as cheaply as possible to prevent flooding.” Today, the answer is closer to “we need a multi-use solution to retain water on-site, limit impact to surface and ground water hydroperiods, reduce pollution and erosion to protect and enhance our natural and built environments, provide community assets for place-making and access to greenspace, and prevent flooding.”

When our understanding of what we need is rapidly changing, it is challenging to set goals at the beginning of projects because we don’t know what we can get, especially in highly constrained built-out environments with limited space. The goals must adapt and evolve through the design process. Therefore, the success criteria are less goal oriented, “Did we meet our goals?” and more process oriented “Did our process optimize our solution for functionality and value?”

What we need versus what we get are the two rails of communication that must be advanced together to navigate a successful project with multiple and competing constraints and goals. If one rail gets too far ahead of the other then progress can easily be stalled, sidetracked, or derailed. So how do we tie these rails together to get what we need? We usually call this modeling.

Recent stormwater management planning and design projects will be used as case studies to discuss how modeling practices such as initial conceptualization, identifying constraints and boundary conditions, making explicit simplifying assumptions, and identifying the right approach and level of detail at the right time can be used to determine if we can need less, or if we can get more to achieve successful project outcomes.

Hydrologic and hydraulic models are typically deterministic, making assumptions about input parameters and calculating a single result for each output. However, this simplification may result in cost-increasing overly conservative assumptions. Additionally, this approach is unable to quantify risk presented by low probability events.

For example, ODOT’s guidance for inlet spacing calculations is to use either 30% or 50% as a clogging factor, depending on inlet type and location. Wouldn’t it be nice to know how resilient a collection and conveyance system is if some inlets are more substantially clogged during large storms?

In this talk, Seth will present the basic theory of probabilistic/stochastic approaches to modeling and disucss some applications this is already considered best engineering practice. He will show examples using Monte Carlo simulations with common tools like EPA SWMM and HEC-RAS to create probability curves of the model outputs. The talk will include specific references to open-source Python toolkits that will help others considering creating probabilistic models.

Finally, the talk will address potential uses including risk informed decision making processes, especially with regards to climate change risk, and sensitivity analysis applied to existing deterministic models. It will also discuss some of the limitations of this approach and seek feedback from the audience members who may use similar approaches.