Evaluacion Filtrado Fosforo Lake George & Nearby Lakes
Methodology, Evaluation, & Feasibility Study
of Total Phosphorus Removal Management
Measures in Lake George & Nearby Lakes
Daniel L. Reisinger, Mary Brabham, Michael F. Schmidt, Patrick R. Victor, and Larry Schwartz
he St. Johns River is the largest river in process, the St. Johns River Water
Florida, extending from southeast of Management District (SJRWMD, district)
Daniel L. Reisinger, E.I., is a water resources
Orlando to Jacksonville and the has implemented the St. Johns River Algal
intern in the Denver, Colorado, office of the
Atlantic Ocean. The lower St. Johns River Initiative, which will limit the growth and
consulting, engineering, construction, and
Basin has exhibited issues associated with nitrogen fixation from cyanobacteria (blue-
operations firm CDM. Mary Brabham, P.E., is
eutrophication, which has led to the estab- green algae) in the river.
the Middle St. Johns River Basin Program man-
lishment of a Total Maximum Daily Load Algal initiative planning has determined
ager for the St. Johns River Water
(TMDL) for Total Nitrogen (TN) and Total that the TP load needs to be reduced by at
Management District in Altamonte Springs.
Phosphorus (TP). Independent of the TMDL least 84 metric tons per year (MT/Yr) (93
Michael Schmidt, P.E., BCEE, is a vice presi-
dent with CDM in the firm’s Jacksonville office.
Larry Schwartz, Ph.D., P.W.S., is a CDM vice
president in the firm’s Orlando office. Patrick
R. Victor, P.E., is a CDM principal in the
Jacksonville office. This article was presented
as a technical paper at the 2008 Florida
Water Resources Conference in May.
tons per year) from current levels in the St.
Johns River Basin (SJRWMD 2006) through
multiple projects, including TP removal in
Lake George, which is the largest lake in the
river. To evaluate the feasibility of TP removal
in Lake George, the district conducted a com-
prehensive review of potentially viable and
cost-effective Best Management Practices
(BMPs), also referred to as management
measures, for reducing TP concentrations in
The comprehensive review was the first
step in designing a potentially very large
regional surface water quality facility treating
50 to 75 million gallons per day (MGD), or
78 to 116 cubic feet per second (cfs). This
article focuses on the results of this review
and the components that were key to its suc-
cessful completion. Each section presents a
key component in the review of TP removal
in Lake George and nearby lakes.
& TP Removal Goals
Lake George is Florida’s second-largest
lake, with an area of approximately 73 square
miles. It is part of the main stem of the St.
Johns River north of Astor, as shown in
42 • SEPTEMBER 2008 • FLORIDA WATER RESOURCES JOURNAL
Table 1 – Water Quantity and Quality in the St. Johns River at Astor, Florida challenges for the removal of TP. Because of
the large flow and lower concentration, rela-
tively high treatment volumes will probably
be required to remove adequate TP mass. The
relatively low nutrient and solids concentra-
tions are expected to be near or below the
practical limit of many standard technologies
(e.g. wet detention, wetlands treatment), and
infiltration systems on the permeable soils on
the west side of the lake may cause nitrate
impacts in the springs..
TP Removal Goal
Figure 1. Two large lakes, Lake Dexter and the time of this study, 2007. Lake George has potential for very large
Lake Woodruff, are located on a secondary Regular negative flows were recorded at TP removal facilities. The removal goal of 84
channel of the river upstream of Lake George the Astor gage. Since treatment will occur MT/Yr was set for the overall St. Johns River
and were included in the evaluation of some regardless of direction, negative flows were Algal Initiative, but the removal goal of indi-
management measures. included in the possible treatment flow. vidual projects will be decided based on a
The area surrounding Lake George, Lake Summary statistics for the gage as absolute comparison of the potential projects in the
Dexter, and Lake Woodruff is largely in pub- flows are shown in Table 1. initiative. The establishment of a nonbinding
lic ownership or divided into relatively small The nutrient and solids concentrations TP removal goal was considered important to
individual parcels. The district currently co- are also measured at State Road 40 in Astor be able to compare management measures
owns significant conservation lands on the (SJRWMD Stations SJR40 and 20020012). using specific itemized costs.
east shore of Lake George, referred to as the The concentrations were found to be relative- The effect of phosphorus removal on
Lake George Conservation Area and the Nine ly low: less than 0.08 milligrams per liter Lake George concentration was estimated by
Mile Point area, which are shown as shaded (mg/L) of TP, less than 1.4 mg/L of TN, and performing a mass balance at Astor.
areas in Figure 1. The land uses of the area are less than 22 mg/L of Total Suspended Solids. Removing the total St. Johns River Algal
predominately classified as wetlands, silvicul- Average seasonal and annual TP and TN con- Initiative goal of 84 MT/Yr, resulting in an
ture (forested uplands), and agriculture (SJR- centrations at Astor are reported in Table 1. average annual concentration of 0.053 mg/L
WMD 2000). These low concentration levels and the vari- in Lake George, would require at least 900
Lake George is surrounded by sandy ability by season have implications for the MGD (1,395 cfs) to be treated, as shown in
hills between 20 and 80 feet NGVD types of management measures that can be Figure 2. A lesser removal goal of 21 MT/Yr,
(National Geodetic Vertical Datum of 1929). used cost-effectively and for potential effi- resulting in an average annual concentration
Stages in Lake George normally range ciencies of treatment. of 0.072 mg/L in Lake George, would require
between 0 and 0.5 feet NGVD, but will The variability of the water quantity and treatment flow of at least 220 MGD (341 cfs).
increase to two feet NGVD in large flow quality of Lake George provides significant Continued on page 44
events. The lake bottom is relatively uniform
with a mean depth of eight feet and a maxi-
mum depth of 12 feet. The soils in the area
are generally hydrologic group B/D or D
(poor permeability), except on the western
shore where they are almost entirely A soils Figure 2 – Lake George Total Phosphorus Concentration as a Function of TP Removal
(high permeable) (Lewis et al. 1987).
Eleven springs are located in the Lake
George tributary area, with three large
springs, Silver Glen, Juniper, and Salt Springs,
discharging into Lake George on the western
shore (Stewart et al. 2006). Lake George is not
affected by marine salinity, but it is tidally
affected and local sources of salinity have
allowed the lake to support a large number of
marine species. Also shown in Figure 1, the
surrounding area has a large population of
Bald Eagles—a threatened species in Florida
that is currently proposed for delisting.
Summar y of Water Quantity & Quality
Stewart et al. (2006) performed exten-
sive hydraulic modeling of the Lake George
area that provides excellent information for a
period of 1996 through 2004. Measured flow
data was available at the USGS flow gage
located at the State Road 40 Bridge at Astor
(USGS Gage 02236125) from 1994 through
FLORIDA WATER RESOURCES JOURNAL • SEPTEMBER 2008 • 43
Figure 3 – Mechanical Harvester Removing Aquatic
Plants (Texas Aquatic Harvesting Inc. 2007)
Figure 4 – Bathymetry of Lake Dexter where Depths
Less than Three Feet Can Not Be Harvested
Continued from page 43 considered applicable to the configuration 1. Aquatic plant harvesting
A goal of 5 MT/Yr was chosen by the dis- and constraints of the project area were 2. Chemical addition
trict for the study, resulting in an average evaluated: 3. Periphyton flowway treatment
annual concentration of 0.076 mg/L in Lake 1. Aquatic plant harvesting (e.g. hydrilla or
George, which would require between 50 and water hyacinth) Feasibility Study
75 MGD (78 cfs to 116 cfs) of treatment flow. 2. Chemical addition (e.g. ballasted floccula- of Management Measures
This removal rate was chosen because it tion)
would likely require several treatment units 3. Constructed wetland treatment Detailed Planning of Potentia l
to be built, and if a higher goal was later cho- 4. Constructed water hyacinth treatment Management Measure Proces ses
sen, the results could be scaled upward as 5. Diversion of water for reuse Detailed information on the configura-
necessary. 6. Dredging tion, sizing, nutrient removal, and byproduct
7. Fish harvesting disposal were determined in the feasibility
Evaluation of Potential 8. Periphyton flowway treatment (e.g. study. All analyses were performed at a plan-
HydroMentia Algal Turf Scrubber or ning level of detail. The following paragraphs
Management Measures Aquafiber) summarize each management measure and
9. Retention/infiltration key components.
Management Mea sure Review
Each of the nine management measures Aquatic Plant Harvesting—Aquatic
A lake management approach was taken
was reviewed for its benefits, challenges, rela- plant harvesting was considered, using
to determine possible management measures
tive costs, relative time to implement, and ref- mechanical harvesters to remove TP and con-
for phosphorus removal in Lake George.
erences to publications used in order to veri- trol the population of invasive aquatic plants.
Using this approach, short-term and long-
fy the applicability of the management meas- Harvesters use a conveyor belt system to
term measures with offline and in-lake treat-
ure. Approaches that can be integrated for a remove aquatic plants from the water, as
ment were considered.
synergistic benefit were also identified. shown in Figure 3. The feasibility study of
Possible management measures were
determined from lake management, this management measure focused on water
Mana gement Measure Select ion
stormwater control, and water quality hyacinth, which is the main aquatic plant
District and CDM representatives met to
improvement literature, relevant experts, controlled in the Lake George area.
discuss each evaluated management measure
and other implemented projects (both suc- Harvesting was proposed to be the
and to select three measures for further
cessful and unsuccessful). The U.S. main method of control, with herbicidal
detailed evaluation. A feasibility study was to
Environmental Protection Agency (EPA) spraying, which is the current population
be conducted for each selected management
provides comprehensive review of lake and control method, as a secondary measure to
measure, which included the conceptual
reservoir restoration and management tech- assure very low water hyacinth populations
design, TP and TN removal capability, bene-
niques (EPA 1990), which was used as a are maintained. Windrow composting in a
fits, operation, and maintenance require-
starting point. contained site was found to be the most
Because of the uncommon attributes of feasible method to dispose of harvested
The feasibility of each management
Lake George (e.g. large lake, variable flow water hyacinth. The resulting compost
measure was discussed until a consensus was
rates, low inflow concentrations, in-stream, could be given away or sold, depending on
reached on whether or not the measure was
unregulated, and relatively shallow), many demand.
appropriate for further study. The major
management measures were not considered, This management measure is a depar-
decision criteria included its technical feasi-
as discussed with the district, since they ture from existing aquatic plant control in
bility, cost, and the ability to permit the sys-
appeared not applicable or infeasible. The the lakes; therefore, study efforts focused
tem. The three most promising management
following nine lake management measures on determining a feasible methods for har-
measures were determined to be:
Continued on page 46
44 • SEPTEMBER 2008 • FLORIDA WATER RESOURCES JOURNAL
ACTIFLO testing and chemical analysis. At
the time of the testing, the St. Johns River
watershed was in drought condition and the
river exhibited low flow. The results from the
water sample indicated TP concentrations of
0.064 mg/L, slightly below the seasonal aver-
age, and undetectable quantities of ortho-
The bench scale testing conducted by
Kruger Inc. evaluated TP removals associat-
ed with various dosages of cationic and
anionic polymers, ferric sulfate, and ferric
chloride coagulants. The results, presented
in Table 2, show that ACTIFLO ballasted
Figure 5 – ACTIFLO 35 MGD Modular Ballasted Flocculation System (Kruger 2007) flocculation was able to remove at least 85
percent of particulate TP, defined as the dif-
ference between TP and dissolved TP, to
limit of detection for the measurement
Continued from page 44 treatment process in controlled facilities. method. There were undetectable quantities
vesting. Because of the possibility of treating a large of ortho-phosphate in the water, so one-
Large harvesters are limited to operating volume of flow, patented or proprietary sys- third removal of ortho-phosphate was used
in water depths greater than three feet. tems, which generally reduce the footprint of for this study.
Bathymetric data for Lake George, Lake the chemical addition system, were primarily Periphyton Flowway Treatment—
Dexter, and Lake Woodruff indicate that large considered. Periphyton flowways are constructed, biolog-
areas of Lake Dexter and a significant portion The ACTIFLO system that uses ically based treatment systems that pass a
of the near shore area of Lake George and microsand ballast and settling plates to shallow flow of water over a sloped mat of
Lake Woodruff are less than three feet in increase overflow rates (the effective treat- periphyton. Nutrients are removed from the
depth, as shown for Lake Dexter in Figure 4. ment system flow) is shown in Figure 5. water biologically by the periphyton, which
Based on the average standing crop of water Implementation of the ACTIFLO process was are harvested to remove the nutrients from
hyacinth, which was estimated from the evaluated in this study; however, other pro- the system and encourage further growth of
acreage receiving herbicidal treatment prietary treatment systems may be equally the periphyton.
(USACE 2007), no lake would have water effective. These systems are differentiated from
hyacinths at a harvestable depth in an average The actual removal from chemical addi- periphyton wetland systems by their flow
year. tion systems can be estimated only by bench characteristics. Flowways use shallow sheet-
Several methods to remove water and pilot scale tests; therefore, the collection flow and higher velocity, although velocities
hyacinth from areas with less than two feet of and testing of a sample to provide an indica- would be considered low in most treatment
water depth were evaluated. The most prom- tion of possible removal was the key compo- applications. Periphyton flowways can be
ising method proposed is to chop the nent of the chemical addition feasibility conceptualized as hydroponics-style horti-
hyacinth using chopping boats to move the analysis. culture of algae.
chopped material in windrows to deeper A water sample was taken from the inlet Two proprietary periphyton flowway
water for harvesting. Chopping boats are typ- of the St. Johns River to Lake George in April systems were evaluated: Aquafiber and
ically shallow-draft vessels with large pro- 2007 and sent to Kruger Inc. for bench scale Continued on page 48
pellers affixed to the bow to chop the plant
material. The working area will be contained
by turbidity-capturing geotextiles to reduce
effects of the chopping. The benefits of this Table 2 – Bench Scale ACTIFLO Results for Lake George
method were provided by Weedbusters Inc.
and include the following:
Widely used with proven feasibility.
Increases the efficiency of the harvesting
May allow over five times more water
hyacinth to be collected per harvester.
Does not eliminate most submerged
aquatic vegetation or suspend sediment.
The use of chopping boats and har-
vesters was considered feasible and was used
to determine TP removal, TN removal, and
Chemical Addition—Chemical addition
of coagulants to a flow stream and subse-
quent settling of formed floc can improve
water quality significantly. These processes
are used widely in the water and wastewater
46 • SEPTEMBER 2008 • FLORIDA WATER RESOURCES JOURNAL
Continued from page 46 Table 3 – Summary of Management Measure Sizing and Costs
HydroMentia. HydroMentia’s Algal Turf
Scrubber® (ATS) pulses a shallow flow of
water over a bed of periphyton to remove
nutrients, which is shown in Figure 5.
Included in the process are patented phos-
phorus precipitation methods and biomass
Aquafiber has developed patented and
trade secret technologies, including a peri-
phyton flowway system also using shallow
sheet flow. The Aquafiber periphyton
flowway system may include patented ozone
treatment to release nutrients bound in
organic matter, eliminate toxic compounds
found in cyanobacteria, and destroy micro-
invertebrates and their eggs.
HydroMentia and Aquafiber were ent removal. The cost of land was not included, since the
engaged to provide information for this The HydroMentia ATS information was majority of land was in public ownership. All
study. A detailed review of the ATS technolo- considered to be acceptable and was used to costs were based on a 20-year design life and
gy at Lake George was provided for the feasi- represent the periphyton flowway manage- include 20 years of operation and mainte-
bility study by HydroMentia (Stewart E.A. ment measure in the feasibility study. nance.
2007). The report included the facility size,
anticipated removal rates, and an estimated Feasibility Study Discussion of Results
cost per pound of phosphorus removed. Results & Costs Each management measure was evaluat-
Aquafiber provided a short memoran- ed in terms of feasibility and cost to make a
dum documenting the likely TP removal Results & Cost per Pound of TP Removed recommendation for further study and ulti-
from a hybrid system of the periphyton The conceptual specifications and asso- mately, design and implementation.
flowway system and the company’s patent- ciated cost per pound of TP removed for each Aquatic Plant Harvesting was estimated
pending, trade secret technology (Fagan management measure is presented in Table 3. to cost $240 per pound of TP removed. This
2007). Aquafiber did not provide a cost esti- The sizing and TP removal for the feasibility process has a relatively higher level of uncer-
mate for this feasibility study. management measures were based on tainty compared to the other alternatives
The technical review of the document detailed planning using the best available because of a lack of data on the water
provided by HydroMentia found the facility information. Each of the management meas- hyacinth population and disposal site condi-
sizing, TP removal methodology, and costs ures were evaluated based on approximately 5 tions; therefore, a study of the Lake Dexter
to be reasonable. HydroMentia used the best MT/Yr per year of TP removed from Lake and Woodruff water hyacinth populations
available data, results from the S-154 ATS George and the St. Johns River; however, the was suggested. Harvesting water hyacinth in
facility, to determine the periphyton growth exact removal varied by process. Lake George was not suggested, based on the
rate and subsequent TP removal rate; how- Costs in Table 3 are based itemized costs long travel times and low density of water
ever, it was believed that this value may be for each management measure, including hyacinths in the lake.
different in Lake George because of much conceptual probable capital and operational Chemical Addition using ACTIFLO
lower TP levels and environmental condi- cost estimates of site preparation, pumps, could provide high levels of TP removal on a
tions, such as low bio-availability of nutri- and piping. Contingencies, engineering, sur- small footprint at an estimated cost of $350
ents, lower water temperatures, and solar vey, permitting, mobilization, and discount per pound of TP removed. It was recommend-
radiation. Pilot testing was suggested to rates were also included. These costs supple- ed that the ACTIFLO technology be consid-
determine a more accurate estimate of nutri- mented the cost of the removal technology. ered at sites other than Lake George, which
would benefit from the ACTIFLO’s small foot-
print. It was recommended that pilot testing of
both the ACTIFLO system and potential
Figure 6 –
dewatering equipment be performed to better
HydroMentia determine phosphorus removal and operation
ATS S-154 and maintenance costs at a site near existing
Basin Facility in commercial or industrial land uses.
Okeechobee Periphyton Flowway Treatment using
HydroMentia’s ATS provides relatively high
levels of phosphorus removal, and generates
(Stewart, E.A. a marketable byproduct (compost) at an esti-
2007) mated cost of $240 per pound of TP
removed. It was recommended that the
Aquafiber technologies also be considered for
Continued on page 50
48 • SEPTEMBER 2008 • FLORIDA WATER RESOURCES JOURNAL
Continued from page 48 FT Contour (Line). Palatka, FL. St. John’s Herbicidal Spraying of Lake George, Lake
Recommendations for River Water Management District. Dexter, and Lake Woodruff, FL. Electronic
Further Planning & Design http://www.sjrwmd.com/programs/plan_ Communication March 2007.
monitor/gis/docs/metadata/elev_5ft.htm • Camp Dresser & McKee (CDM). 1999.
Three management measures listed in • Lewis et al. 1987. Soil Survey of “Renaissance Jar Testing.” Submitted to Mr.
Table 3 are recommended for further study Okeechobee County, FL. Electronic Report. Jeff Hiscock of Mock Roos & Associates.
through pilot testing: http://soildatamart.nrcs.usda.gov/Manuscr • Stewart, E.A. 2007. “Preliminary
1. The harvest of water hyacinths should be ipts/FL093/0/fl_okeechobee.pdf. Accessed Engineering Assessment for the Application
pilot tested in Lake Dexter. February 2007. of an Algal Turf Scrubber (ATS) for
2. Periphyton flowway treatment should be • Stewart, Joseph B., Sucsey, Peter, and Reduction of Nitrogen and Phosphorus
pilot tested in the vicinity of Lake George. Hendrickson, John. “Meteorological factors Loads into Lake George, Florida.” Prepared
3. ACTIFLO should be pilot tested in a pre- affecting estuarine conditions within Lake by HydroMentia for SJRWMD: Ocala, FL.
dominantly industrial or commercial area. George in the St. Johns River, Florida”. • Fagan, William B. 2007. Aquafiber
Because of the water quality fluctuations in Proceedings of the Seventh International Application to Lake George. Letter to Mr.
Lake George, pilot studies should be performed Conference on Hydroscience and Daniel Reisinger of CDM. June 15, 2007.
for a period of at least one year to evaluate a full Engineering. Philadelphia, PA. September
range of water quality and flow conditions. The 2006. http://hdl.handle.net/1860/732
three management measures are modular and • Hendrickson, J.C. and J. Konwinski. 1998.
complementary; therefore, they can be used Seasonal Nutrient Import-Export Budgets
together if desired for additional removal. for the Lower St. Johns River, Florida. Final
Report for Contract WM598, Florida
References Department of Environmental Protection.
• St. Johns River Water Management District • United States Environmental Protection
(SJRWMD). 2006b. Lake George P Agency (EPA). 1990. “Lake and Reservoir
Removal Project MSJRB St. Johns River Restoration Guidance Manual, Second
Nutrient Discharge Reduction WBS Edition.” Publication # 440490006,
2.3.30.00.08 (part of the SJR Algal Electronic, NTIS # PB93-207926.
Initiative) Charter Document – 11/9/2006. • United States Army Corps of Engineer
• St. Johns River Water Management District (USACE). 2007. Unclassified Acreage,
(SJRWMD). 2000. Elevation: SJRWMD 5 Herbicide Usage, and Man Hours for
50 • SEPTEMBER 2008 • FLORIDA WATER RESOURCES JOURNAL