Harvesting and Greenroof Methodology

Three BMPs in BMPFast remove pollutants by keeping water from being discharged without being retention basins or wet ponds: Stormwater Harvesting and Rainwater Harvesting use the water they store, usually for irrigation, and a Greenroof holds and evaporates rain on the roof and can reuse its filtrate from a cistern. None of them is read on the retention curves (Retention Methodology) or the wet detention equations (Wet Detention Methodology). This page explains the method each one uses, with a worked example checked against the program and curves calculated by BMPFast itself.

For all three, the removal is a share of the average yearly runoff that is kept from discharge, and the same share of the nitrogen and phosphorus load. So N removal = P removal = the hydraulic capture.

Stormwater Harvesting

A stormwater harvesting system stores runoff, usually in a wet pond, and pumps it out for irrigation or another use. Water that is used is not discharged. How much of the yearly runoff is used depends on three things: how much storage there is, how fast water is used, and the rainfall pattern. These are tied together by Rate-Efficiency-Volume (REV) curves.

The REV curves

REV curves come from long-term simulations of a harvesting pond with historical rainfall (Harper, 1991). Each curve is one harvest efficiency, the percent of the yearly runoff that is used. It gives the use rate (inches per day over the equivalent impervious area) needed for each harvest volume (inches over the same area). In 2009 the curves were fitted with fifth-order polynomials and extended down to 20 % (Wanielista, 2009). The simulations assume:

  • No net groundwater movement into or out of the pond in a year, and rainfall on the pond equal to evaporation.
  • Water is used twice a week, not the day after rain, at the average yearly use rate.
  • The pool is drawn down no more than 1 inch over the catchment, to protect the vegetation.
  • The results are long-term averages; any one year differs.

Source reference

Harper, G.M. (1991). Design Curves for the Reuse of Stormwater. M.S. thesis, University of Central Florida; the same curves in Wanielista, M.P., Yousef, Y.A., Harper, G.M. and Dansereau, L. (1991), Design Curves for the Reuse of Stormwater (link): long-term simulations of a harvesting pond with historical rainfall give, for each zone, curves of the use rate needed for each harvest efficiency at each storage volume (both over the equivalent impervious area); the simulations assume no net groundwater exchange, rain on the pond equal to evaporation, use twice a week and at most 1 in of drawdown.

Wanielista, M.P. (2009). Review of the REV Curves. Memorandum, May 20, 2009, Stormwater Management Academy, University of Central Florida: fits each REV curve with a fifth-order polynomial, adds curves down to 20 %, drops the 95 % curve, and limits the curves to 0.25 to 6 in of storage (1 to 6 in for the 90 % curve); BMPFast builds its table from these polynomials.

BMPFast holds one set of curves for each meteorological zone, for efficiencies of 20, 30, … 90 %. It builds a table from them: storage from 0.25 to 6.0 inches in steps of 0.05, and for each step the use rate of each curve, from the curve’s polynomial:

Use rate (in/day over EIA) = a5 x^5 + a4 x^4 - a3 x^3 + a2 x^2 - a1 x + a0      x = harvest volume (in over EIA)

Each curve’s polynomial is only good up to a limit (its fit limit, between about 1.5 and 6 inches). Past it the curve is held at its value at the limit: more storage buys no more efficiency on that curve.

Inputs

InputUnitsWhere it comes from
Available Harvest Volumeac-ftThe worksheet (Stormwater Harvesting): the storage the system draws from.
Harvest (use) ratein/week over the irrigated areaThe worksheet: the average water use.
Total Area Available for IrrigationacThe worksheet: the area the water is used on.
Watershed AreaacThe catchment’s post-development area without the BMP’s own area (Catchment Characteristics).
Annual runoff coefficient (ROC)–The catchment’s yearly runoff coefficient, for its zone and land use.
Meteorological zone and rainfall-, in/yrGeneral Site Information.

The method

1. Equivalent impervious area   EIA (ac) = Watershed Area (ac) x ROC
2. Volume over EIA x (in) = 12 x Harvest Volume (ac-ft) / EIA
3. Use rate over EIA y (in/day) = Harvest Rate (in/week) x Irrigation Area (ac) / 7 / EIA
4. Harvest efficiency E (%) = REV table at (x, y), between 0 and 90
5. N removal = P removal = hydraulic capture = E
6. Water demand (MG/yr) = Harvest Rate x 52 x 0.325829 x Irrigation Area / 12
Harvested supply (MG/yr) = E x EIA x Rainfall (in) x 0.325829 / 1200
Supplemental water (MG/yr) = demand - supply, when the demand is larger (otherwise 0)

Step 4 reads the table the way the REV curves are read by hand. At the storage x, it takes each curve’s use rate (in a straight line between the two nearest 0.05-inch columns). It then finds the two curves whose rates are on either side of the actual use rate y, and interpolates the efficiency between them in a straight line. A storage under 0.25 inch is read at 0.25, and over 6 inches at 6. The efficiency is never below 0 or above 90 %. When the use rate is below the 20 % curve, BMPFast extends the line between the 20 % and 30 % curves downwards and reports the efficiency under 20 % that it gives. That is beyond the published curves, which stop at 20 % (Wanielista, 2009); the report of both harvesting worksheets says so in a note, and the value should be treated as an estimate. 0.325829 converts acre-feet to million gallons.

Source reference

BMPFast method (Harvesting, StormwaterHarvesting and RainwaterHarvesting classes; below 20 % decided by Ron, 2026-10-13): the efficiency is capped at 90 %, the highest curve the memorandum keeps; past a curve’s fit limit its last value is held out to 6 in; storage below 0.25 in is read at 0.25 in; and below the 20 % curve the line between the 20 % and 30 % curves is extended, beyond the published curves.

FDEP and the water management districts (2010). Stormwater Quality Applicant’s Handbook: Design Requirements for Stormwater Treatment Systems in Florida, March 2010 draft (not adopted), Section 15.3 Equivalent Impervious Area (Equation 15-1) (the draft, not adopted: PDF, 16 MB): the equivalent impervious area is the watershed area times its weighted runoff coefficient; BMPFast uses the catchment’s annual runoff coefficient (ROC) as that coefficient.

Treatment Depth. The report also gives an equivalent Treatment Depth: the retention depth that would give the same efficiency on the retention curves for the catchment’s CN, DCIA and zone (Retention Methodology). It is a way to compare the harvesting system with a retention basin. It does not change the harvesting removal.

Applicant’s Handbook reference

Section 2.0 Definitions and Terms, “Stormwater harvesting”: “capturing stormwater for irrigation or other beneficial use”.

Appendix O, Table 1 (BMP Efficiencies), Stormwater harvesting: removal is the “annual load of stormwater … not discharged because used for irrigation”, “Evaluated on case-by-case basis (to estimate volume of stormwater reuse, use the rate-efficiency-volume [REV] curve methodology used by DEP”, after Wanielista et al. 1991 and Wanielista and Bradner 1992); BMPFast uses REV curves, with the 2009 polynomial fits. Reading efficiencies below the 20 % curve is BMPFast’s own extension.

Worked example

Example Problem 4 of the BMP Trains manual (Zone 5, 61 in of rain): a 5 ac watershed with an ROC of 0.71, 1 ac-ft of harvest volume, 0.8 in/week used on 3 ac.

StepCalculationResult
EIA5 x 0.713.55 ac
x12 x 1 / 3.553.380 in
y0.8 x 3 / 7 / 3.550.0966 in/day
Curve rates at x = 3.38050 % curve 0.0909, 60 % curve 0.1119 in/dayy is between them
E50 + 10 x (0.0966 – 0.0909) / (0.1119 – 0.0909)52.69 %
Demand0.8 x 52 x 0.325829 x 3 / 123.389 MG/yr
Supply52.69 x 3.55 x 61 x 0.325829 / 12003.098 MG/yr
Supplemental water3.389 – 3.0980.290 MG/yr
Equivalent Treatment Depthretention curves at 52.69 %0.672 in

BMPFast gives 52.691 %, 3.389, 3.098 and 0.290 MG/yr and 0.672 in for these inputs.

After a wet pond (BMPs in Series)

A harvesting pond is often a wet detention pond whose water is also used. In a BMPs in Series with Wet Detention followed by Stormwater Harvesting, the two are combined as a treatment train: N = N(pond) + E x (1 – N(pond) / 100), and the same for P. In Example Problem 4 the pond and the harvesting together remove 74.5 % of the nitrogen and 86.5 % of the phosphorus. See BMPs in Series.

Applicant’s Handbook reference

Section 9.5.1 Treatment Train Nutrient Reduction (Equation 9-5): BMPs in series are combined as 1 − [(1 − Eff1) × (1 − Eff2)]; the pond-then-harvesting removal is this equation.

The REV curves for each zone

Each plot shows the table BMPFast calculates with: for each efficiency, the use rate needed at each storage. A design whose storage and use rate fall between two curves gets an efficiency between them.

REV curves, Florida Zone 1: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 1: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 2: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 2: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 3: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 3: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 4: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 4: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 5: use rate needed for each harvest efficiency, by storage volume
REV curves, Florida Zone 5: use rate needed for each harvest efficiency, by storage volume

Rainwater Harvesting

A rainwater harvesting system collects roof runoff in a cistern and uses it, usually for irrigation. BMPFast uses the same REV curves as for stormwater harvesting, but over the roof: the equivalent impervious area is the roof area times a roof factor, the share of the rain that runs off that kind of roof. The worksheet is in square feet and gallons (Rainwater Harvesting).

Roof typeRoof factor
Asphalt, 0.2 % slope0.4175
Asphalt, 2.0 % slope0.5942
Concrete, 0.2 % slope0.9531
Concrete, 2.0 % slope0.9457
Shingles, 5 % slope0.8371
Shingles, 25 % slope0.9587
Shingles, 50 % slope0.9544
Tile roof, 5 % slope0.9627
Tile roof, 25 % slope0.9678
Tile roof, 50 % slope0.9723

Source reference

Wanielista, M.P., Chopra, M., Hardin, M., Kakuturu, S. and Runnenbaum, N. (2011). Evaluating Runoff and Abstraction from Impervious Surfaces as Components Affecting Recharge. University of Central Florida: runoff as a fraction of rainfall for ten roof surfaces and slopes: the Roof Factors in the table.

1. EIA (sf)              = Roof (contributing) area (sf) x Roof factor
2. x (in) = Cistern volume (gal) / 7.4805 gal/cf / EIA (sf) x 12
3. Use (cf/day) = Use rate (in/week) x Irrigated area (sf) / (7 x 12)
y (in/day) = 12 x Use (cf/day) / EIA (sf)
4. E (%) = REV table at (x, y) for the zone, between 0 and 90
5. N removal = P removal = E
6. Water demand (MG/yr) = Use rate x 52 x 0.325829 x Irrigated area (sf) / (12 x 43,560)

Worked example (Zone 5): a 2,000 sf tile roof at 25 % slope, a 1,500 gallon cistern, 0.5 in/week used on 1,000 sf.

StepCalculationResult
EIA2,000 x 0.96781,935.6 sf
x1,500 / 7.4805 / 1,935.6 x 121.243 in
Use0.5 x 1,000 / 845.952 cf/day
y12 x 5.952 / 1,935.60.0369 in/day
Ebetween the 20 % curve (0.0264) and the 30 % curve (0.0433) at x = 1.24326.21 %
Demand0.5 x 52 x 0.325829 x 1,000 / (12 x 43,560)0.0162 MG/yr

BMPFast gives 26.206 % for these inputs. The plot shows the removal as the cistern grows, for three use rates. A small cistern (under about 0.25 inch over the roof, here about 300 gallons) is read as 0.25 inch, so the curves start flat. At a low use rate the efficiency is below the lowest REV curve (20 %), where BMPFast extends the curves beyond the published range (shaded).

Applicant’s Handbook reference

Appendix O, Table 1 (BMP Efficiencies), Stormwater harvesting: the handbook has no separate row for rainwater (roof) harvesting; BMPFast treats a cistern as stormwater harvesting and uses the same REV method over the roof.

Rainwater harvesting removal against cistern size for three use rates (2,000 sf tile roof, 1,000 sf irrigated, Zone 5)
Rainwater harvesting removal against cistern size for three use rates (2,000 sf tile roof, 1,000 sf irrigated, Zone 5)

Greenroof

A greenroof holds rain in its soil and plants, where much of it evaporates. Filtrate that drains through can be stored in a cistern and reused to irrigate the roof. BMPFast gives the removal from curves for 18 Florida rainfall stations, by the cistern storage in inches over the greenroof area (Greenroof).

No cistern (storage 0):   Removal (%) = c
With a cistern: Removal (%) = max(c, a x ln(Cistern storage, in) + b)
N removal = P removal = Removal

a, b and c belong to the selected station; c is the removal with no cistern. The curve a x ln(storage) + b is below c for storage under about 0.11 to 0.16 inch, so the program uses the larger of the two: a small cistern never gives less removal than none. The worksheet accepts a storage of 0, or 0.01 to 5 inches.

Worked example (Jacksonville #4358: a = 11.49, b = 64.23, c = 40): with 0.9206 in of cistern storage, 11.49 x ln(0.9206) + 64.23 = 63.28 %, larger than 40 %, so the removal is 63.3 %. With 0.1 in, the curve gives 37.8 %, so the removal is c = 40 %. With no cistern it is 40 %.

Applicant’s Handbook reference

Appendix O, Green Roof: “The nutrient reduction capability of these systems can be calculated in the same way as Dry Retention System”; BMPFast differs: it uses rainfall-station curves for the greenroof and cistern (Hardin 2006), not the Appendix O dry retention tables.

Source reference

Hardin, M. (2006). The Effectiveness of a Specifically Designed Green Roof Stormwater Treatment System Irrigated with Recycled Stormwater Runoff to Achieve Pollutant Removal and Stormwater Volume Reduction. M.S. thesis, University of Central Florida (link): the station curves of yearly retention against cistern storage, from simulations at 18 Florida stations; BMPFast’s a, b and c for each station.

FDEP and the water management districts (2010). Stormwater Quality Applicant’s Handbook: Design Requirements for Stormwater Treatment Systems in Florida, March 2010 draft (not adopted), Section 12.8 Greenroof Cistern/Harvesting Design Curves and Equations (the draft, not adopted: PDF, 16 MB): gives a curve % retention = a Ln(S) + b for each of the same 18 stations (Belle Glade: 10.70 Ln(S) + 72.36, about 50 % with no cistern), for cistern storage S from 0.5 to 5.0 in; BMPFast uses the same coefficients and accepts 0.01 to 5 in, using the no-cistern value below about 0.11 to 0.16 in.

Greenroof removal against cistern storage for three stations
Greenroof removal against cistern storage for three stations

Sources

  • REV curves: Harper, G.M. (1991), Design Curves for the Reuse of Stormwater, M.S. thesis, University of Central Florida (Link); polynomial fits and the 20-40 % curves: Wanielista, M.P. (2009), Review of the REV Curves, memorandum, Stormwater Management Academy, UCF.
  • Greenroof and cistern: Hardin, M. (2006), The Effectiveness of a Specifically Designed Green Roof Stormwater Treatment System Irrigated with Recycled Stormwater Runoff to Achieve Pollutant Removal and Stormwater Volume Reduction, M.S. thesis, University of Central Florida (Link).
  • The BMPTrains model: BMP Trains 2020 User Manual, Stormwater Management Academy, University of Central Florida (Link).

Sample problems

Worked problems solved in BMPFast, each with every input, the steps with screen captures, the program’s results, a hand check and the project file to download. Captures are for demonstration only. All of them: BMPFast sample problems.