The Case for Flow Duration Modeling

Pacific Water Resources, Inc. (PWR) compared and critically evaluated traditional site-based, single event hydrologic analysis methods versus new methods using continuous hydrologic modeling. Weaknesses of traditional hydrologic modeling methods were identified using a detailed “pilot basin” study. The pilot basin consisted of 235 acres, with nine detention facilities, developed over a ten-year period in the Portland Oregon vicinity. Results indicated that traditional analysis methods failed to achieve their flow control objectives at the watershed scale.

Traditional site-based detention sizing consistently over-sizes flow controls and under-sizes storage volumes providing little or no benefit for smaller, more frequent storms. Also, multiple site-based facilities fail to provide desired flow control at the watershed scale for design storm events, both because of oversized flow controls and “peak stretching” which allows near peak outflows from multiple facilities to overlap.

Continuous hydrologic models, as utilized in Washington State, allow for site-level evaluation of development impacts using a calibrated watershed tool. However, the models in use tend to be regional in nature and not basin or even subbasin specific. PWR’s work in Southwest Washington found that regional parameters used in Washington’s Department of Ecology (DOE) Western Washington Hydrology Model (WWHM) produced poor results when compared to available stream gages that were studied.

Based on this and other experience, PWR developed a cntinuous hydrologic modeling application called the Flow Duration Design Model (FDDM). Continuous
hydrologic modeling permits utilization of a “flow duration standard” rather than a “peak flow standard” for the design of stormwater management facilities for new developments. This site-based, calibrated watershed model differs in several respects from WWHM currently being used in Washington State.

While both use pre-simulated runoff time series obtained from EPA’s HSPF program, FDDM results can be tailored to specific subbasin areas to reflect subtle differences in annual rainfall, elevation, soils or other parameters. FDDM allows jurisdictions to set “target” conditions based on allowable effective imperviousness. The concept of a “target” condition permits the establishment of stormwater management standards based on criteria such as the biological value of downstream areas. Pristine headwater areas can utilize more stringent targets while already impacted areas can use less stringent targets. Target conditions, then, can directly reflect receiving water quality as well as community values and priorities. FDDM also allows varying design thresholds. For example, in this area of Oregon half the 2-year storm is roughly a bank full event, at which shear stresses on channel structure begin the transport of bed materials and erosive forces increase. A jurisdiction’s lower design threshold can be set to half the 2-year event in order to minimize stresses on streams caused by high flow. Similarly, upper thresholds (e.g. the 10-year storm) may be set to reflect downstream flooding hazards or community floodplain policies.Finally, FDDM allows explicit simulation of low impact development (LID) techniques. The goal of LID is to enhance infiltration while reducing stormwater runoff at the site level. Currently the model supports continuous hydrologic simulations of: permeable pavements, green roofs, blue roofs, infiltration facilities, smart cisterns and traditional detention. Many other techniques can be modeled as variations of one of these types. Engineers can use FDDM to assess and compare the cost versus performance of various stormwater management methods.