Why Single-Event Modeling Doesn’t Work for LIDs presented at Stormcon 2009 Anaheim, CA August 2009 Doug Beyerlein, P.E. Clear Creek Solutions, Inc.

Why Single-Event Why Single-Event Modeling Doesn’t Work Modeling Doesn’t Work

for LIDsfor LIDs

presented at Stormcon 2009 presented at Stormcon 2009Anaheim, CAAnaheim, CAAugust 2009August 2009

Doug Beyerlein, P.E. Doug Beyerlein, P.E.

Clear Creek Solutions, Inc.Clear Creek Solutions, Inc.

Clear Creek Solutions’ Clear Creek Solutions’ Hydrology ExpertiseHydrology Expertise

Clear Creek Solutions, Inc., provides complete range of hydrologic and stormwater modeling services.• Clear Creek specializes in continuous simulation hydrologic modeling.

• We have 30+ years of experience modeling complex hydrologic and stormwater problems.

• We created WWHM3 for Washington State Department of Ecology and are adapting it to model San Francisco Bay Area and San Diego watersheds and facilities.

• We teach WWHM and HSPF workshops.

PresentationPresentation

IntroductionIntroduction

Historical Hydrology Historical Hydrology

Modern HydrologyModern Hydrology

LID DesignLID Design

SummarySummary

Questions & AnswersQuestions & Answers

Historical HydrologyHistorical Hydrology

Slide rule hydrology:Slide rule hydrology: Rational Method (Q = CIA)Rational Method (Q = CIA) SCS methods: TR-55, TR-20, SBUH, HEC-1, SCS methods: TR-55, TR-20, SBUH, HEC-1, etc.etc.

Based on modeling a single-event Based on modeling a single-event

(2-year storm, etc.)(2-year storm, etc.)

Traditionally we have used single-Traditionally we have used single-event design storms to size event design storms to size stormwater and LID facilities based stormwater and LID facilities based on frequency.on frequency.

Design storm assumption: 2-year rainfall Design storm assumption: 2-year rainfall causes 2-year flood.causes 2-year flood.

Historical HydrologyHistorical Hydrology

Single-event design storm Single-event design storm methodology doesn’t work for methodology doesn’t work for stormwater and LID design stormwater and LID design because:because:

Single-event flow frequency standards are based on inappropriate assumptions.

Single-event modeling cannot represent the evapotranspiration and infiltration occurring between storm events.

Single-event inappropriate assumptions:

1. A specific frequency runoff (e.g. 2-yr) will be caused by the same frequency rainfall event.

FACT: The frequency of the runoff event depends on existing soil moisture conditions and the current amount of water in storage in the LID facility.

Single-event inappropriate assumptions:

1. A specific frequency runoff (e.g. 2-yr) will be caused by the same frequency rainfall event.

Seattle 2-Year Runoff (LP3)

Storm Peak (in) 24-Hr Vol (in) Soil Moisture Conditions

Jan-59 0.19 1.53 Wet

Feb-61 0.15 1.55 Wet

Oct-81 0.43 3.55 Very Dry

NOAA Type 1A 0.33 2.00 Average

Single-event inappropriate assumptions:

2. A hypothetical storm shape (e.g. Type 1A) is representative of actual historic rainfall events.

FACT: Actual rainfall events come in many shapes and it is the intensity of the rainfall that is a major factor in how much runoff occurs.

Single-event inappropriate assumptions:

2. A hypothetical storm shape (e.g. Type 1A) is representative of actual historic rainfall events.

Comparison of Seattle Storms for 2-Yr Runoff

0.00

0.10

0.20

0.30

0.40

0.50

0 6 12 18 24

Time (Hours)

Rai

nfa

ll (

in)

Type 1A Jan-59 Feb-61 Oct-81

Single-event inappropriate assumptions:

3. Storm events always start with an average antecedent soil moisture condition.

FACT: Actual rainfall events may occur back-to-back or after a long drought and have totally different antecedent soil moisture conditions.

Single-event inappropriate assumptions:

4. LID facilities are empty at the start of a storm event.

FACT: LID facilities may be partially or completely full of water at the start of a storm event.

Single-event inappropriate assumptions:

4. LID facilities are empty at the start of a storm event.

Full Empty

Storm Events

Single-event modeling cannot represent the evapotranspiration and infiltration occurring between storm events.

The effectiveness of LID facilities depends on what happens between storm events.

Single-event modeling cannot represent the evapotranspiration and infiltration occurring between storm events.

Seattle Monthly Precip and PET

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Month

Pre

cip

an

d P

ET

(in

)

Precip (in)

PET (in)

Precip > ETET > Precip

Single-event modeling cannot represent the evapotranspiration and infiltration occurring between storm events.

San Jose Monthly Precip and PET

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Month

Pre

cip

an

d P

ET

(in

)

Precip (in)

PET (in)

Precip > ET

ET > Precip

Single-event modeling cannot represent the evapotranspiration and infiltration occurring between storm events.

Green Roof Runoff vs Conventional Roof RunoffRunoff small decrease

Large decrease

NOTE: Most LID designs based on NOTE: Most LID designs based on “Runoff Credits” are really based “Runoff Credits” are really based on these flawed single-event on these flawed single-event assumptions.assumptions.

They only hide the problems, They only hide the problems, they don’t fix them and should they don’t fix them and should not be used.not be used.

Modern HydrologyModern Hydrology

Continuous simulation hydrology:Continuous simulation hydrology: Stanford Watershed Model (1966)Stanford Watershed Model (1966) HSPF (1980)HSPF (1980) WWHM (2001)WWHM (2001)

Pseudo-continuous simulation:Pseudo-continuous simulation: SWMMSWMM HEC-HMSHEC-HMS

Modern Hydrology

Continuous simulation hydrology models the entire hydrologic cycle for multiple years.

Stormwater runoff = surface runoff + interflow.

Modern Hydrology

Continuous simulation hydrology explicitly represents hydrologic processes such as interception storage, soil moisture fluctuations, evaporation, transpiration, and the effects of long-term infiltration rather than having to make the assumptions used in single-event hydrology.

Modern Hydrology

Continuous simulation hydrology models the entire hydrologic cycle for multiple years.

Use long-term historic rainfall data to generate long-term simulated runoff, stage, and actual ET data.

For example: 50 years of NWS hourly rainfall data generates approximately 400,000 hours of runoff data.

Statistically analyze the data to compute flow and stage frequencies, durations, and evaluate the effectiveness of LID designs for a wide range of hydrologic conditions.

Modern Hydrology

Flow Frequency (Log Pearson Type III)

Modern Hydrology

Flow Duration (% Time Exceeded)

LID Design

There is nothing magical about LIDs. Water must go somewhere. Water must either:

1. Infiltrate into the soil.

2. Evaporate/transpire into the atmosphere.

3. Runoff.

The majority of the infiltration and ET occurs The majority of the infiltration and ET occurs between storm events, NOT during them.between storm events, NOT during them.

LID Design

The majority of the infiltration and ET occurs The majority of the infiltration and ET occurs between storm events, NOT during between storm events, NOT during them.them.

Infiltration + ET

LID Design

Types of LID facilities used for stormwater mitigation:

Green/vegetated/eco-roofs Rain gardens/bioretention Permeable pavements Impervious surface dispersion

Note: All (except for green roofs) can include infiltration.

Green Roof

Green/vegetated/eco-roof: Relies on ET.

Rain Garden/Bioretention

Rain garden/bioretention/landscape swale

Relies on both ET and infiltration.

Porous Pavement

Relies on ET and infiltration.

Impervious Runoff Dispersion

Dispersion of impervious roof or parking lot runoff on adjacent pervious lawn; relies on ET and infiltration.

Summary

1. What happens between storm events is just as important as what happens during storm events.

2. Infiltration and evapotranspiration are what makes LID facilities effective in reducing stormwater.

Summary

3. Historical hydrology makes inappropriate assumptions.

4. Modern continuous simulation hydrology simulates LID effectiveness over a wide range of hydrologic conditions and is more accurate than the historic single-event methods.

For more information

Contact:

Doug Beyerlein 425.225.5997beyerlein@clearcreeksolutions.com

Joe Brascher 360.943.0304brascher@clearcreeksolutions.com

Or go to: www.clearcreeksolutions.com