Design Models for Treatment Wetland Systems at Low Phosphorus Concentrations: DMSTA

Design Models for Treatment Wetland Systems at Low Phosphorus Concentrations: DMSTA

William W. Walker

Robert H. Kadlec

The goal of the Everglades Protection Project (EPP) is to remove phosphorus from runoff waters before it enters the Everglades Protection Area (EPA). Treatment wetlands are the surviving option, out of dozens that have been tested. Those wetlands are characterized by varying proportions of three classes of communities: emergent plants, submerged plants, and periphyton (algae). The acronyms SAV (submergent aquatic vegetation) and PSTA (periphyton stormwater treatment area) have been commonly used to represent the submersed and periphyton communities, respectively. All three assemblages are capable of phosphorus removal.

The STA design model (Walker, 1995) embodied a first order, areal net uptake of phosphorus, and was constrained to be applied to long term average wetland performance. A modified version of the first order, areal model, containing such a lower limit, was proposed (Kadlec and Knight, 1996). That modification was subsequently adopted for design of STA3/4. For long term average wetland performance:

(1)

The STA design model effectively used C* ≈ 2 ppb, a result driven by rainfall P. The later k-C* version allowed C* as a calibration parameter, with resulting values in the 4 - 20 ppb range.

The goal of this work was to develop and calibrate the simplest, highly aggregated model that could mimic the major features of event driven behavior of treatment wetlands in the runoff environment. A simple extension of the STA and k-C* design models is the addition of P storage in the biota of the wetland ecosystem (Figure 1). During periods of high phosphorus availability, that storage will increase, and in periods of P-famine, the storage will decrease. At all times, the ecosystem produces the residual sediments containing unavailable P, that characterized the WCA2A calibration data of the STA design model (Walker, 1995). This extension has been named the Dynamic STA Design Model (DMSTA). It is a unsteady state model that removes phosphorus to permanent burial in proportion to the amount of labile P in storage, and is patterned after a similar model calibrated for other marshes (Kadlec, 1997). The labile pool is presumed to be drawn down by a return flux, or bleed-back. Temporal variations in the water budget (flows, rainfall, evapotranspiration, seepage, storage) are also simulated. The underlying philosophy of DMSTA is to maximize simplicity,
while retaining features absolutely necessary for explaining P removal.

Figure 1. Concepts of the DMSTA structure.

C=P concentration in water, mg P/m3

J=P flux, mg P/m2•yr

L=P loading, mg P/m2•yr

M=water column P storage, mg P/ m2

Q=water outflow, 106 m3/yr

M=water column P storage, mg P/ m2

S=biomass labile P storage, mg P/ m2

V=water volume, 106 m3

The model presumes P removal from the water to labile biomass storage that is proportional to the water concentration and the storage size.

(2)

This process transfers P from the water column to temporary (labile) storage.

A second principal premise of this model is a P return from labile storage to the water that is proportional to the square of the amount of stored labile P. This recycle rate is:

(3)

A third premise is that labile stored P is partially converted to biologically unavailable P in soils and sediments:

(4)

Three rate constants K1, K2, and K3 are adjustable parameters. The combination:

(5)

is the lowest steady state (long-term average) concentration that can be attained by a specific calibration wetland community. Further details are given in Walker and Kadlec (2002).

Calibrations of DMSTA have been accomplished for three genera of treatment wetlands receiving low phosphorus waters: emergent, submergent and periphyton dominated systems. Calibration prototypes were selected based upon vegetation characteristics, size, and dataset duration.Central tendency parameters determined from single-platform calibrations have been determined for the three wetland types, and for combinations.

DMSTA has been used to investigate alternatives for future improvements to phosphorus removal from waters entering the EPA.

References

W. W. Walker and R.H. Kadlec, 2002. Dynamic Model for Stormwater Treatment Areas, website:

Kadlec, R. H., 1997. “An Autobiotic Wetland Phosphorus Model,” Ecological Engineering, Vol. 8, No. 2, pp. 145-172.

Walker, W. W., 1995. “Design basis for Everglades stormwater treatment areas,” Water Resources Bulletin, Vol. 31, No. 4, pp. 671-685.

Kadlec, R.H. and R.L. Knight, 1996. Treatment Wetlands. Lewis Publishers, Boca Raton, FL.

Separate Figure: