Time filter

Source Type

Los Angeles, CA, United States

Kadlec R.H.,Chelsea Management
Water (Switzerland) | Year: 2016

This paper reviews aspects of the performance of large (>40 ha) constructed treatment wetlands intended for phosphorus control. Thirty-seven such wetlands have been built and have good data records, with a median size of 754 ha. All are successfully removing phosphorus from a variety of waters. Period of record median concentration reductions were 71%, load reductions 0.77 gP·m-2· year-1, and rate coefficients 12.5 m· year-1. Large wetlands have a narrower performance spectrum than the larger group of all sizes. Some systems display startup trends, ranging to several years, likely resulting from antecedent soil and vegetation conditions. There are internal longitudinal gradients in concentration, which vary with lateral position and flow conditions. Accretion in inlet zones may require attention. Concentrations are reduced to plateau values, in the range of about 10-50 mgP·m-3. Vegetation type has an effect upon performance measures, and its presence facilitates performance. Trends in the performance measures over the history of individual systems display only small changes, with both increases and decreases occurring. Such trends remove little of the variance in behavior. Seasonality is typically weak for steady flow systems, and most variability appears to be stochastic. Stormwater systems display differences between wet and dry season behavior, which appear to be flow-driven. Several models of system performance have been developed, both steady and dynamic. © 2016 by the authors.

Kadlec R.H.,Chelsea Management
Ecological Engineering | Year: 2010

The dynamics nitrate retention and export were studied at the Des Plaines River wetland demonstration site. Seven wetlands received pulses of river water in discrete pumping events. Twenty-eight wetland events were monitored over 4 years for all hydrologic variables, pumping, rain, storage change, and outflow. Nitrate was measured at high frequency for the ouflows, and at lower frequency for inflows and interior stations. Most events were isolated in time, with sufficient inter-event spacing to allow complete equilibration before the subsequent event. Pumping was selected to provide up to 45 displacements of the wetland water volume. River water averaged 2.3 mg/L of nitrate nitrogen, and wetland effluent averaged 0.9 mg/L. The average mass removal of nitrate was 67% over all events, with a range from 17% to 100%. A calibrated dynamic water mass balance was developed as the framework for interpreting results. Internal hydraulics were characterized by tanks-in-series (TIS) models calibrated to tracer studies. Residence time distributions were describable by three TIS (three wetlands) and five TIS (four wetlands). Dynamic nitrate mass balances were used, in conjunction with a first order areal uptake model, to model the time sequence of NO3N concentrations and flows. Parameter estimation, based on NO3N mass flow fitting, produced rate constants that best described the series of events the wetlands. Rate constants were much higher for the events than for previous steady state performance for the wetlands (k20 = 107 vs. 37 m/yr). Rate coefficients increased at higher water temperatures, with a modified Arrhenius temperature factor of 1.090. Performance for N removal was found to be partially due to displacement of antecedent treated water, and partially due to treatment occurring during the event, and partially due to treatment after the event. Carbon availability was estimated not to limit denitrification, except possibly at the highest nitrate loadings. © 2009 Elsevier B.V. All rights reserved.

Kadlec R.H.,Chelsea Management
Critical Reviews in Environmental Science and Technology | Year: 2012

Large numbers of free water surface treatment wetlands are in use for nitrate reduction. Target applications are field runoff, river and stream improvement, and enhancement of wastewater treatment plants. In total, an extensive database now exists, in many publications and operating reports.Microcosms and mesocosms are not included here because of the lack of transferability to design. A first-order areal model is appropriate, to be implemented with appropriate temperature, hydraulic efficiency, and flow pattern. Annual average rate constants at 20°C have a median of 25 m/year. Performance is better at higher water temperatures, with a modified Arrhenius temperature factor of 1.106. Measured values of the tanks-in-series (TIS) parameter average N = 4.4 TIS. Higher rate coefficients are associated with emergent soft tissue vegetation, and lower efficiencies with submergent vegetation, unvegetated open water, and forested wetlands. Carbon availability can limit denitrification at high nitrate loadings; however, wetlands produce carbon in sufficient quantities to support typical municipal and agricultural loads. Design may be for load reduction or concentration reduction, with the latter requiring larger wetlands. Significant ancillary benefits of ecological diversity and wildlife habitat are certain to accompany the project. A small negative greenhouse gas penalty, which accrues to all new wetlands, is not an important factor. Economic issues may include land cost and pumping cost. Constructed marshes are an ecologically and economically attractive method for reducing nitrate levels in surface waters. Copyright © Taylor & Francis Group, LLC.

Kadlec R.H.,Chelsea Management | Pries J.,CH2M HILL | Lee K.,Municipality of Brighton
Ecological Engineering | Year: 2012

The Town of Brighton, Ontario implemented a 6.2ha marsh in 2000, for the purpose of improving water quality before discharge to receiving waters. The wetlands have successfully operated in this moderately cold climate for over ten years. Phosphorus removal of 2.3gP/(m2yr) was achieved, with an annual areal rate coefficient of 9.2m/yr. The removal is strongly seasonal, with the greatest reductions occurring in spring. The total nitrogen loading was dominated by ammonia (208gN/(m2yr)), with smaller amounts of organic and oxidized nitrogen. Ammonia was reduced to 173gN/(m2yr). Implied areal rate constants were high for mineralization of organic nitrogen (29m/yr) and denitrification (101m/yr), but low for nitrification (4m/yr). CBOD5 was reduced from 5.4 to 3.2mg/L, and TSS was reduced from 13.2 to 7.2mg/L, both with slightly higher values during late winter. The wetland was not effective in reducing pathogens, with Escherichia coli at 167cfu/100ml entering, and 132cfu/100ml leaving. Vegetation was sparse, likely due to muskrats and deep water. Macro-invertebrate diversity was lower than for regional wetlands. Bird use was very high, and birding was a popular human activity. The wetland has been designated as provincially significant. © 2012 Elsevier B.V.

Kadlec R.H.,Chelsea Management | Zmarthie L.A.,Spicer Group
Ecological Engineering | Year: 2010

A wetland system has operated seasonally at Saginaw Township, MI, USA, for ten years. The system consists of extraction, aeration, settling, intermittent vertical sand filtration, a surface flow wetland treatment with recycle, and discharge to the Tittibawassee River. The 0.85 ha cattail wetland treats the full leachate flow, with a total system detention time of 180 days. The high recycle rate creates a lesser wetland detention time of 60 days. Ammonia is the principal contaminant of concern, because it occurs at high concentrations, typically 300-500 mg/L. Ammonia mass reduction averaged 99.5% for the last nine years, with a 95% mass removal in the startup year. Metals were not present in all samples, with modest reductions in those always present (zinc 16%, arsenic 29%, barium 78%, chromium 67%). Volatile organic compounds were removed to below detection, excepting BTEX, which occurred in only 2% of the outflow samples. Base neutral organics, PCBs and pesticides were also removed to below detection, excepting phthalates with an outlet detection frequency of 29%. No pesticides or PCBs were detected in the system outflow. The ammonia removal rate coefficients for the wetland (12 m/yr) was at the 55th percentile of the distribution for other surface flow wetlands. The vertical filter was likely oxygen limited, and functioned with an apparent oxygen utilization of 30 gO/(m2 d). © 2010 Elsevier B.V. All rights reserved.

Discover hidden collaborations