<%BANNER%>

Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2012-12-31.

Permanent Link: http://ufdc.ufl.edu/UFE0042299/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2012-12-31.
Physical Description: Book
Language: english
Creator: Fanara, Tracy
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Tracy Fanara.
Thesis: Thesis (M.E.)--University of Florida, 2010.
Local: Adviser: Sansalone, John.
Electronic Access: INACCESSIBLE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042299:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042299/00001

Material Information

Title: Record for a UF thesis. Title & abstract won't display until thesis is accessible after 2012-12-31.
Physical Description: Book
Language: english
Creator: Fanara, Tracy
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Statement of Responsibility: by Tracy Fanara.
Thesis: Thesis (M.E.)--University of Florida, 2010.
Local: Adviser: Sansalone, John.
Electronic Access: INACCESSIBLE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042299:00001


This item has the following downloads:


Full Text

PAGE 1

1 INTRA EVENT TRANSPORT OF PHOSPHORUS IN RUNOFF FROM A PAVED URBAN SURFACE By TRACY ANN FANARA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DE GREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Tracy Ann Fanara

PAGE 3

3 To my parents and grandmother whom I have always looked up to, you have been great influences in my life. To Leah for teaching me to take risks, to Natalie for bein g an example of patience and balance, to Amber for your motivating dedication and work ethic to the field of Environmental Engineering, to Ryan for your inspiring and contagious confidence in me, and to Jill, Joey, and Caryn for your unconditional support.

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my family for all their support. I would also like to thank the past and present professors at the University of Florida who have inspired and motivated me for the past ten years, namely Dr. Mazyck, Dr. Chadik, Dr. Delfino, Dr. Bitton, Dr. Boyer, Dr. Martinez and Dr. Koopman. I would especially like to thank Dr. Sansalone for his shared knowledge and committed counsel.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ........... 4 LIST OF TABLES ................................ ................................ ................................ ...................... 6 LIST OF FIGURES ................................ ................................ ................................ .................... 7 ABSTRACT ................................ ................................ ................................ ............................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ............. 11 2 TRANSPORT ACROSS A NON PERMEABLE MATRIX: FIRST FLUSH ...................... 15 Summary ................................ ................................ ................................ ............................ 15 Background ................................ ................................ ................................ ........................ 15 Methodology ................................ ................................ ................................ ...................... 17 Discussion ................................ ................................ ................................ .......................... 20 Resul ts ................................ ................................ ................................ ............................... 22 3 PHOSPHORUS PARTITIONING AND RAINFALL RUNOFF TREATMENT BY ALUMINUM OXIDE COATED MEDIA ................................ ................................ .......... 36 Summary ................................ ................................ ................................ ............................ 36 Background ................................ ................................ ................................ ........................ 37 Methodology ................................ ................................ ................................ ...................... 39 Discussion ................................ ................................ ................................ .......................... 42 Results ................................ ................................ ................................ ............................... 44 4 CONCLUSION ................................ ................................ ................................ .................. 68 LIST OF REFERENCES ................................ ................................ ................................ .......... 70 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ..... 73

PAGE 6

6 LIST OF TABLES Table page 2 1 Determination of First Flush for Rainfall Runoff Events in Gainesville, Florida ............ 24 2 2 Relationship Between Rate Constant and Previous Dry Hours for Rainfall Runoff ........ 25 3 1 Event Hydrology and Partitioning Relationships for Influent and Effluent* Rainfall R unoff Events In Gainesville, Florida ................................ ................................ ............ 48 3 2 Summary of Event Mean Concentrations and Range of Influent Fractions ..................... 49 3 3 Summary of Eve nt Mean Concentrations and Range of Effluent Fractions ..................... 50 3 4 Summary of Event Mean Value of Dissolved Fraction (f d ) and Partition Coefficient (k d ) of Phosphorus for Influent and Effluent Runoff ................................ ....................... 51 3 5 Removal Efficiencies of Phosphorus Fractions ................................ .............................. 52

PAGE 7

7 LIST OF FIGURES Figure page 2 1 Reitz Union Parking Lot (http:// maps.google.com/) ................................ ....................... 26 2 2 Site Setup on West Side of Reitz Union Parking Lot (Not to Scale) ............................... 26 2 4 Mass vs. Flow Limited Determination f or Rainfall Runoff Event on March 20, 2008 .... 28 2 5 Mass vs. Flow Limited Determination for Rainfall Runoff Event on May 16, 2008 ....... 28 2 6 Mass vs. Flow Limited Determination for Rainfall Runoff Event on June 3, 2008 ......... 29 2 7 Mass vs. Flow Limited Determination for Rainfall Runoff Event on June 10, 2008 ....... 29 2 8 Mass vs. Flow Limited Determination for Rainfall Runoff Event on June 21, 2008 ....... 30 2 9 Mass vs. Flow Limited Determination for Rainfall Runoff Event on July 8 2008 .......... 30 2 10 Mass vs. Flow Limited Determination for Rainfall Runoff Event on July 15, 2008 ........ 31 2 11 Mass vs. Flow Limited Determinat ion for Rainfall Runoff Event on July 29, 2008 ........ 31 2 12 Mass vs. Flow Limited Determination for Rainfall Runoff Event on August 8, 2008 ..... 32 2 13 Mass vs. Flow Limited Determination for Rainfall Runoff Event on August 12, 2008 ... 32 2 14 Mass vs. Flow Limited Determination for Rainfall Runoff Event on August 19, 2008 ... 33 2 15 Mass vs. Flow Limited Determination for Rainfall Runoff Event on Sept. 10, 2008 ...... 33 2 16 Mass vs. Flow Limited Determination for Rainfall Runof f Event on Sept. 20, 2008 ...... 34 2 17 Mass vs. Flow Limited Determination for Rainfall Runoff Event on Oct. 8, 2008 .......... 34 2 18 Mass vs. Flow Limited Determination for Rainfall Runoff Event on Oct. 23, 2008 ........ 35 2 19 ................................ ........................ 35 3 1 Infl uent and Effluent Relationships between TDP and PBP for March 20, 2008 ............. 53 3 2 Influent and Effluent Relationships between TDP and PBP for May 16, 2008 ................ 53 3 3 Influent and Effluent Relationships between TDP and PBP for June 3, 2008 .................. 54 3 4 Influent and Effluent Relationships between TDP and PBP for June 10, 2008 ................ 54 3 5 Influent and Effluent Relationships between TDP and PBP for June 21, 2008: ............... 55

PAGE 8

8 3 6 Influent and Effluent Relationships between TDP and PBP for Ju ly 8, 2008 .................. 55 3 7 Influent and Effluent Relationships between TDP and PBP for July 15, 2008 ................ 56 3 8 Influent and Effluent Relationships between TDP and PBP for July 29, 2008 ................ 56 3 9 Influent and Effluent Relationships between TDP and PBP for August 8, 2008 ............. 57 3 10 Influ ent and Effluent Relationships between TDP and PBP for August 12, 2008 ............ 57 3 11 Influent and Effluent Relationships between TDP and PBP for August 18, 2008 ............ 58 3 12 Influent and Effluent Relationships between TDP and PBP for Sept. 10, 2008 ............... 58 3 13 Influent and Effluent Relationships between TDP and PBP for Sept. 20, 2008 ............... 59 3 14 Influent and Effluent Relationships between TDP and PBP for Oct. 8, 2008 .................. 59 3 15 Influent and Effluent Relationships between TDP and P BP for Oct. 23, 2008 ................ 60 3 16 Influent and Effluent Total Phosphorus Removal for the Event of March 20, 2008 ........ 60 3 17 Influent and Efflu ent Total Phosphorus Removal for the Event of May 16, 2008 ........... 61 3 18 Influent and Effluent Total Phosphorus Removal for the Event of June 3, 2008 ............. 61 3 19 Influent and Effluent Total Phosphorus Removal for the Event of June 10, 2008 ........... 62 3 20 Influent and Effluent Total Phosphorus Removal for the Event of June 21, 2008 ........... 62 3 21 Influent and Effluent Total Phosphorus Removal for the Event of July 8, 2008 .............. 63 3 22 Influent and Effluent Total Phosphorus Removal for the Event of July 15, 2008 ............ 63 3 23 Influent and Effluent Total Phosphorus Removal for the Event of July 29, 2008 ............ 64 3 24 Influent and Eff luent Total Phosphorus Removal for the Event of August 8, 2008 ......... 64 3 25 Influent and Effluent Total Phosphorus Removal for the Event of August 12, 2008 ....... 65 3 26 Influent and Effluent Total Phosphorus Removal for the Event of August 19, 2008 ....... 65 3 27 Influent and Effluent Total Phosphorus Removal for the Event of Sept. 10, 2008 .......... 66 3 28 Influent and Effluent Total Phosphorus Removal for the Event of Sept. 20, 2008 .......... 66 3 29 Influent and Effluent Total Phosphorus Rem oval for the Event of Oct. 8, 2008 .............. 67 3 30 Influent and Effluent Total Phosphorus Removal for the Event of Oct. 23, 2008 ............ 67

PAGE 9

9 Abstract of Thesis Presented to the Gr aduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering INTRA EVENT TRANSPORT OF PHOSPHORUS IN RUNOFF FROM A PAVED URBAN SURFACE By Tracy Fanara December 2010 Chair: John J. Sansalone Major: Environmental Engineering Sciences Stormwater is the largest non point source of pollution with respe ct to natural aquatic systems; therefore it is understandable that q uality control of stormwater has recently gain ed recognition In response to research, more educated regulation along with low impact development techniques (LID), is currently being developed and implemented To determine the most effective treatment technique, the chemistry and hydrology of an area must be des cribed. Phosphorus is a problematic nutrient in Florida, which has been determined to have adsorptive properties, and therefore a favorable constituent to control and reduce prior to reducing water quality, and adversely impacting aquatic ecosystems. In this study, rainfall runoff hydrology in relation to mass versus flow limited rainfall events with respect to phosphorus, phosphorus partitioning and relationships, and the utilization of al uminum oxide coated media (AOCM) to bind and reduce phosphorus con tent of rainfall runoff, were researched by monitoring and testing 15 storm events before and after exposure to aluminum oxide coated material. All 15 events that were analyzed showed that total dissolv ed phosphorus was a flow limited constituent, and par ticulate bound phosphorus was found to be flow limited in 11 of 15 events, the remaining fractions. The events that were found to be mass limited did not have a relationship with t he

PAGE 10

10 amount of time between storm events, or previous dry hours, by obtaining a correlation coefficient of 0.112. The dissolved fraction f d was found to be an average of 0.31 prior to treatment by AOCM and 0.21 following treatment, the de crease due to adsorption of the dissolved fraction by AOCM There was a strong linear relationship found between total dissolved phosphorus and parti culate bound phosphorus following exposure to the AOCM treatment system; the same linear relationship was found for 10 of 15 event influent data (prior to treatment) The average removal efficiency (RE) of total dissolved phosphorus (TDP) was 72.8% t he average RE for sediment bound phosphorus was found to be 89.8%, the average RE for settleable bound phosph orus was 81.8%, the removal efficiency for suspended bound phosphorus was 44.6%, and total phosphorus (TP) RE by AOCM treatment was 73%.

PAGE 11

11 CHAPTER 1 INTRODUCTION Development and expansion of civilization vastly decrease environmental abstractions (evapo ration, infiltration, plant and soil interaction), resulting in increased rainfall runoff, which contains pollutants and toxins from both biogenic and anthropogenic activities. Urban environs are currently designed for rapid and efficient transport of sto rmwater, which results in high pollutant loading and significant changes to the rainfall runoff relationship for urban watersheds (Sansalone and Christina 2004.) Although most organic substances can be environmentally and biologically degraded, the resulti ng increased nutrient loading of biogenic material, fertilizers, and pesticides, has distinct adverse effects on the environment. For example, nutrient pollutant loads (phosphorus and nitrogen) are responsible for algal blooms which may culminate in eutro phication of a water body or biologic community. Current regulations regarding sustainable development and best management practices (BMPs) to maintain pre development hydrology include swales, ponds, and wetlands for treatment, redistribution, and quant ity control. The intent of these techniques is to reduce adverse impacts on biological and ecological systems, and to achieve the historical priority of preventing downstream flooding. Further sustainable technologies, such as green roofs (VanWoert, et. Al, 2005), filter media (Davis, Hsieh, 2005; Liu, et.al. 2005), permeable pavement (Brattebo, Booth, 2003), and bioretention ponds (McCuen, Okunola, 2002) have been studied for effectiveness and applicability. The application, integration, and implementa tion of sustainable development, in the form of filter media and pervious pavement techniques, have not become a regular practice in stormwater design due to limited accessibility, minimal public knowledge, initial expense, and lack of regulations in place to mandate these practices or

PAGE 12

12 discharge requirements. Regulatory agencies in the state of Florida are in the process of developing stormwater rules to implement these sustainable development practices. The environmental focus of most newly constructed o r retrofitted drainage systems is pollutants such as hydrocarbons, metals, and nutrients, all of which may bind to particulate matter, also a pollutant of concern. Because phosphorus has adsorptive properties, which allow chemical uptake to be successful, sustainable practices and research would benefit from focusing on removal of this nutrient from stormwater. is applicable; therefore only an initial percentage of ra infall runoff volume (which varies with regulatory agency) is required to be treated by a best management practice ( BMP examples are a lake, pond, swale, and wetland). Sustainable technologies are intended to effectively remove the peak pollutant load, c be effective if a pollutant at a given site is subject to hydrology, therefore, flow limited. This characteristics with respect to phosphorus loading of a subject site in Gainesville, Florida. This evaluation was performed graphically and by categorical analysis to determine whether a storm is flow or mass limited. The results describe phosp horus loading and how the rainfall runoff at this site, and similar sites, should be treated. The last part of this study is to evaluate a common assumption that the amount of time a site remains dry (previous dry hours, PDH) is related to the probability that an event is mass or flow limited. This was done graphically and statistically to events.

PAGE 13

13 Partitioning of the phosphorus content in rainfall runoff is a criti cal phenomenon for assessing treatability, and determining the appropriate treatment technique for a specific drainage basin. The second chapter of this study first focuses on partitioning of rainfall runoff phosphorus, followed by an evaluation of filter media performance used to treat these fractions. The first goal is to analyze the fractions of phosphorus separately and determine the treatment technique that will be most effective in this area. Dissolved phosphorus is removed from rainfall runoff via chemical adsorption, even though adsorption may still play a role in particulate phosphorus removal, filtering sediment ( through a #200 sieve (American Society for Testing and Materials, 2002; Sansalone, Kim, 200 7)), settleable (those fractions 25 m < material < 75 m, found by passing a one liter sample through a #200 sieve before allowing the settleable material to settle in an Imhoff cone for one hour (Sansalone, Kim, 2007)), and suspended (those fractions 1 m < material < 25 m (Sansalone, Kim, 2007)) particles will ensure that the phosphorus associated with the particles is removed from a permeable matrix. Determining relationships between these fractions may assist in derivation of new methodology for es timating partitioning of phosphorus in a given drainage basin. Aluminum oxide coated cementitious media (AOCM) is a clay based oxide media prepared from phosphorus free clayey soil and aluminum oxide. AOCM was chosen for the next goal of this study; to an alyze phosphorus partitioning pre and post treatment, and removal of phosphorus by adsorption and media filtration. This material was chosen for the Gainesville, Florida location on account of the high potential for anion/nutrient adsorption and experimen tal success with phosphorus uptake (Erickson et al. 2007, Sansalone et al. 2004). The adsorption of phosphate on AOCM has been found to be dependent on pH and maximum removal efficiencies

PAGE 14

14 have been obtained at slightly acidic conditions (Wu, T. et al. 200 8.) The point of zero charge (pzc) of aluminum oxide is approximately 9 and the pH of rainfall and runoff in Florida (4.5 and 7.5, respectively) is higher than that of northern states (3.8 and 5.8, respectively); which makes AOCM the more favorable choice when compared to other filter media material, such as manganese oxide coated media, which has a point of zero charge (pzc) below the typical pH range of 6 8 (Liu et al. 2005). The site of interest is a University of Florida parking lot in Gainesville, F lorida as shown in Figure 1. Parking lot rainfall runoff was diverted through the AOCM based filtration system; influent and effluent runoff samples were analyzed to determine pre and post filter media interaction, respectively. Phosphorus fractions for both influent and effluent samples were compared in this study to evaluate AOCM removal efficiency of dissolved and particulate bound phosphorus fractions. Filter media can be used in stormwater design, by directing rainfall runoff through a filter media cartridge, to reduce pollutants prior to outfall. Filter media can also theoretically be used in swales to assist in treatment; more studies must be done for practical application. If AOCM proves to be an effective method of treatment, it can be accepted as a desired design technique for best management practices and low impact development (LID) methods, such as sedimentation tanks and AOCM base d cementitous pervious pavement

PAGE 15

15 CHAPTER 2 TRANSPORT ACROSS A N ON PERMEABLE MATRIX: FIRST FLUSH Summary The runoff events is many times assumed in regulation. Assuming the majority of the pollutant load is delivered in the first portion of the storm may result in treatment regulations for only the first po rtion of runoff. Therefore, if pollutant loading for an event is truly flow limited, thus having a direct relationship with event hydrology, desired results may be falsely assumed, rather than achieved. In this study, rainfall runoff events are analyzed to determine if a mass limited (first flush) or flow limited predominance exists, with respect to phosphorus in a typical northern Florida parking lot. Linear regression results show ed that all 15 storm events demonstrated total dissolved phosphorus was fl ow limited. Over 70% of the 15 events showed a flow limited result for particulate bound phosphorus, therefore 30% of the events were found to be mass limited, with respect to particulate bound phosphorus. This study showed the occurrence of mass or flow limited events was unrelated to the antecedent dry weather period (previous dry hours, PDH), by obtaining a Pearson correlation coefficient of 0.122 of first flush and PDH. Background Governing agencies most commonly regulate land development and civil design by setting quantifiable limits, such as pre development verses post development runoff from a site of interest, downstream pond stage, or ensuring that an initial set volume of runoff has a desired detention time within a constructed pond. For exam ple, in Sarasota County Florida, regulation states that a down stream pond stage cannot increase more than 0.01inch due to development ( www.co.sarasota.fl.us ) Southwest Florida Water Management District calls for the first 1inch of runoff be treated for a wet detention basin ( www.swfwmd.state.fl.us ) where the South Florida

PAGE 16

16 Water Management District states, for wet detention, the first 1 inch of runoff be treated or 2.5 inches multiplied by the percentage of impervious surfaces, whichever number is greater must be treated ( my.sfwmd.gov ) San Diego California regulations state that the first 0.75 inches of rainfall be detained ( www.caltrans.ca.gov ) ; and Italy regulates a mere 0.2 inches of rainfall runoff must be treated before outfall to a natural water bo dy (Gnecco, 2008) However the the first portion of runoff may not result in effective treatment for all storm events and/or all rainfall runoff constituents. A flow limited event is characterized by a relationship between mass delivery and hydrology, when mass loading is proportionate to the runoff hydrograph characterized by the delivery of a disproportionately large pollutant load within the first portion of the runoff hydrograph (Sansalone and Buchberger, 1997); if this phenomena most often exists, treatment of the initial fraction of runoff would then result in treatment of a substantia l portion of the pollutant load. However it has been found in a 2007 research study by Flint and Davis that first flush was only found to occur 22% of the time for total phosphorus. It was found by Sansalone and Buchberger in 1997 that approximately 60% of the normalized TSS (total suspended solid) load was removed from the impervious surface in the first 50% of the storm with the remaining 40% of TSS asymptotically approaching unity as the storm progressed to its conclusion (Sansalone and Christina, 2003 ). Previous research has shown that w ith respect to the physical conditions, a first flush from uniform rainfall intensity has the highest probability of occurrence on small, sloped watersheds or source areas without complex drainage patterns, particularl y if a high proportion of the watershed is impervious ( Sansalone and Cristina, 2004), since the site of interest meets these criteria, this research will test this theory.

PAGE 17

17 It is a common assumption that the amount of time between rainfall runoff events, or previous dry hours (PDH), increases the occurrence of first flush events. The theory behind this assumption is: as time between events increases, biogenic and anthropogenic material accumulates in parking lots, sidewalks, and other impervious surfaces. This accumulation is theorized as more susceptible to mobilization by initial runoff; therefore the pollutant load would be more likely delivered within the first portion of a storm, as opposed to pollutant load having a direct relationship with hydrology. This theory also assumes that a higher fraction of pollutant load is associated with particles, rather than the dissolved fraction of runoff. Previous research has found that variations in constituent loading is dependent upon PDH (Irish, L.B.Jr. et.al, 1998), which supports a theory contrary to the hypothesis that PDH has no affect the occurrence based on rainfall frequency for a certain area. Methodology Aluminum oxide coated material (AOCM) is a clay based oxide media, and was selected based on bench scale testing due its superior efficiency with respect to mass transfer behavior through adsorpti on. This media was prepared from phosphorus free clayey soil and aluminum oxide; the wet clayey soil (above optimum water content) was expanded with a blowing agent and fired to 1000C for 8 hours into bricks, then the cooled bricks were crushed to the me dia substrate, then crushed into media in the size range of 0.85 9.50 mm, coated with 0.5 1 M aluminum salt, reheated to dryness and washed (Sansalone and Ma, 2009). The purpose of this media is to remove particulate matter and adsorb metals, dissolved ph osphorus, and other ions such as chloride, and sulfur through ion exchange. Desorption of phosphorus from this media was found to be negligible, indicating chemical adsorption; hence a strong bond between

PAGE 18

18 phosphorus and AOCM affinity sites (Sansalone and Ma, 2009.) Using in situ testing of this adsorptive material can give an accurate assessment of the media capabilities. The site of interest is Rietz Union parking lot at the University of Florida in Gainesville, Florida as shown in Figure 2 1. The draina ge basin was determined to be approximately .11 acres (450 m 2 ) with a slope of approximately 3% E W and 1.5% N. The catchment area is subject to slightly deviate from the su rveyed catchment area due to rainfall intensity and wind direction. The rainfall r unoff to the storm sewer on the west side of the parking lot was diverted from the sewer system via 6 inch PVC pipe, into a system consisting of a one inch Parshall Flume with a Shuttle Ultra sonic sensor manufactured by MJK, and a drop box. The Shuttle U ltrasonic sensor was calibrated and placed 30 inches above the invert of the Parshall flume. The unit utilizes measurement of water depth to indirectly calculate the flow of rainfall runoff. The Parshall flume, manufactured by Warminster Fiberglass, rece ived water from the 6 inch PVC pipe at a 6% slope located at the storm drain inlet. The flume was calibrated and mounted on wood support so that the flume slope was at 0% as to not affect runoff depth within the flume as shown in Figure 2 3. A number of w eather models were used to predict rainfall, intensity, and duration, including Wunderground (www.wunderground.com), Accuweather (www.accuweather.com), and National Weather Service (NWS) (www.nws.noaa.gov). However due to the sporadic nature of central Fl orida rainfall, many set ups and site mobilizations were executed with no rainfall, or less than 0.1 inches of rainfall, which was set as a minimum for sampling. An antecedent dry weather minimum of three days was also set to ensure more accurate estimati ons of pollutant loadings; this is because the standard assumption for water quality modeling is that pollutants

PAGE 19

19 build up on surfaces throughout the watershed between the rainfall events at a constant rate (Davis and McCuen, 2005), overtime, pollutant depo sition will increase, thus giving a more descriptive data set. Time of rainfall start was recorded as well as the first flow through the Parshall flume. Samples were collected at the outle t of the Parshall flume, 2, 0.5 L bottles and duplicates, 2, 1 L b ottles and dupl icates, and 1, 4 L bottle and a duplicate was taken per sample. Sample administering was dependent on storm duration and intensity as seen by the weather and radar models on the internet, sample increment shortening at the peak of the storm removed and the original storm design was re established. Samples were brought into the lab, where a number of tests were done. Sediment settle able, and suspended mat erial were removed from samples to test for dissolved phosphorus, using a Hach Spectrophotometer 5600. Dissolved phosphorus content was measured by first filtering samples through a 45 m filter, then administering a Hach orthophosphate test. Orthophosph ate reacts with molybdate in an acid medium to produce a mixed phosphate/molybdate complex. Ascorbic acid then reduces the complex, giving an intense molybdenum blue color. Test results are measured at 880 nm by a Hach DR/5000 spectrophotometer. Particul ate bound phosphorus content was analyzed by first separating the suspended, settleable and sediment fractions from each sample. The sediment fraction ( were filtered by a #200 (75 m) stainless steel sieve to retrieve the sediment fract ion (American Society for Testing and Materials, 2002; Sansalone, Kim, 2007) which was then dried and weighed to determine the mass of sediment per sample. To determine the settleable fraction (25 m < material < 75 m) of each sample, sediment free samp les were poured into an Imhoff cone and were allowed to settle for 60 minutes (Sansalone, Kim, 2007), the settled material was then

PAGE 20

20 dried and weighed to determine the mass of the settleable fraction. To separate the suspended fraction (1 m < material < 25 m (Sansalone, Kim, 2007)) from the dissolved fraction, the supernatant (sample (sediment + settleable)) was then filtered through a nominal pore size of 1 m filter, which was dried and weighed to determine the suspended portion. The sediment, settl eable and suspended fractions were then acid digested to convert organic and condensed inorganic forms prior to analysis; therefore releasing phosphates as orthophosphate in order to test the samples using the same methodology as described above for the di ssolved fraction. The acid digestion process uses sulfuric and persulfate acids along with heat to promote conditions for hydrolysis of the condensed inorganic forms. The digested samples are then neutralized to a pH above 2.14. After a dilution factor o f 10, the test procedure stated above for dissolved phosphorus is then administered to determine the fraction of particulate bound phosphorus. A Discussion In determining event first flush charac teristics based on coupled phenomena of water chemistry and hydrology by physical methods in an urban source area watershed, one must identify and characterize flow and mass limited transport behavior of phosphorus loadings. The first flush concept occurs during mass limited events, meaning that the pollutant load is in direct response to particulate matter composition. Flow limited pollutant loads directly correlate with hydrology, thus the hydrologic peak occurring at the same time as the peak of total phosphorus loading. Each rainfall runoff event may fit into either of these two categories influenced by the criteria being analyzed. Partitioning of each constituent, the amount of previous/antecedent dry hours (PDH), rainfall runoff volume, and rainfal l intensity, may all influence whether a storm is flow or mass limited for a specific constituent.

PAGE 21

21 A simple way to determine first flush is mathematically defined by the following equations (Sansalone and Buchberger 1997; and Sansalone and Christina 2003 ): M (t) > V (t) ( 2 1) M (t) = ( 2 2) V (t) = ( 2 3) Where M(t) and V(t) is the normalized mass and normalized volume respectively as functions of time; Q(t) and C(t) are the average flow rate and average concentration during time interval k
PAGE 22

22 Where X and Y are sample means of the rate constant, K, and PDH. A correlation constant near or at 1, represents a strong relationship, and a correlation constant near or at 0, represents a non correlating relationship. Results By plotting cumulative vo lume by cumulative total dissolved phosphorus (TDP) and cumulative particulate bound phosphorus (PBP), respectively, the mass versus flow limited relationship could be seen. By performing a linear regression, this relationship is statistically proven (see Figures 2 4 through 2 18) All storm events display a linear relationship between cumulative total dissolved phosphorus (TDP) loading and storm duration graphically and by linear regression. Therefore all 15 storm events were flow limited with respect to TDP, and the pollutant load was delivered In respect to particulate bound phosphorus (PBP), 11 of 15 events showed the same linear relationship between cumulative volume and PB P, however, four events did show a logarithmic relationship and are therefore considered mass limited, since the pollutant loading was not directly dependent on the hydrology of these events, and mass loading was concentrated in the beginning of the event, resulting in a change in slope when graphically demonstrated. Table 2 1 and Figures 2 4 through 2 18 show these results Though this chapter does not focus on event mean concentrations (EMC) for each event, these data may be found in Table 3 2. In concl usion, for this site, and sites with similar characteristics such as low mass loading, regulations to treat solely the first 1 inch of runoff would not be the most effective treatment method. Although PBP is the predominant mass fraction at this site, and mass limited events were recorded, the majority of the events show that pollutant loading is delivered throughout the duration of an event, and is flow limited, therefore, the peak loading will be at the

PAGE 23

23 peak of the storm, which could be well after the fi rst inch of runoff. Low impact development (LID) practices such as filter media or pervious concrete would be of benefit as opposed depending on a lake with a treatment volume based on one inch of runoff. The first order rate equation constant was determ ined for each storm by power law, and was then plotted against PDH. A relationship between K and PDH is not apparent in the graph. To further solidify a conclusion, a correlation constant of 0.12 was determined for the data, which is well below acceptabl e for correlation to be made, and very close to 0, meaning no relationship at all. In conclusion, PDH should not be used as an indicator of first flush occurrence for an area or a storm, and therefore should not be a source for regulation. Table 2 2 and Figure 2 19 show the non relationship between PDH and the occurrence of first flush.

PAGE 24

24 Table 2 1. Determination of First Flush for Rainfall Runoff Events in Gainesville, Florida Rainfall runoff event, 2008 TDP Part. Bound P PDH V runoff Q inf max Q inf avg Event # (hr) (min) (hr) (L) (L/s) (L/s) 1 20 Mar 2:05:00 F F 230 2066 3.4 0.88 2 16 May 17:20:00 F F 960 2212 8.75 0.9 3 3 Jun 15:30:00 F F 432 308 0.75 0.11 4 10 Jun 14:08:00 F M 168 8002 10.96 1.61 5 21 Jun 12:15:00 F M 144 1137 3.98 0.34 6 8 Ju l 16:46:00 F F 141 31268 13.17 2.14 7 15 Jul 13:16:00 F F 66 22438 13.07 3.6 8 29 Jul 11:40:40 F F 309 1411 3.69 0.55 9 8 Aug 11:06:00 F F 191 474 2.16 0.14 10 12 Aug 14:40:00 F F 93 3866 3.92 0.62 11 19 Aug 12:53:00 F F 117 1461 6.78 0.81 12 10 Sep 16:27:00 F M 264 1569 1.96 0.47 13 20 Sep 13:58:00 F M 234 503 1.09 0.24 14 8 Oct 13:57:00 F F 431 1557 2.75 0.14 15 23 Oct 15:14:00 F F 334 978 1.46 0.29 Median 230 1557 3.69 0.55 Mean 274 5283 5.19 0.86 SD 221 9112 4.30 0.95

PAGE 25

25 Tab le 2 2 Relationship Between Rate Constant and Previous Dry Hours for Rainfall Runoff Rainfall runoff events Time of Event K PDH (rate constant min 1 ) (hours) 20 Mar 2:05:00 0.651 230 16 May 17:20:00 1.711 960 3 Jun 15:30:00 0.968 432 10 Jun 14:08: 00 0.806 168 21 Jun 12:15:00 0.896 144 8 Jul 16:46:00 0.349 141 15 Jul 13:16:00 0.479 66 29 Jul 11:40:40 0.630 309 8 Aug 11:06:00 0.312 191 12 Aug 14:40:00 0.535 93 19 Aug 12:53:00 3.504 117 10 Sep 16:27:00 0.416 264 20 Sep 13:58:00 0.633 234 8 O ct 13:57:00 0.185 431 23 Oct 15:14:00 0.315 334 Median 0.6 230.0 Mean 0.8 274.3 SD 0.8 228.4 Correlation Coefficient 0.12

PAGE 26

26 Figure 2 1. Reitz Union Parking Lot ( http://maps.google.com/ ) Figure 2 2. Site Setup on West Side of Reitz Union Parking Lot (Not to Scale)

PAGE 27

27 Figure 2 3. Site Set Up at Reitz Union Parking Lot, Gainesville, Florida

PAGE 28

28 F igure 2 4 Mass vs. Flow Limited Determinat ion for Rainfall Runoff Event on March 20, 2008 Figure 2 5 Mass vs. Flow Limited Determination for Rainfall Runoff Event on May 16, 2008

PAGE 29

29 Figure 2 6 Mass vs. Flow Limited Determin ation for Rainfall Runoff Event on June 3, 2008 Figure 2 7 Mass vs. Flow Limited Determination for Rainfall Runoff Event on June 10, 2008

PAGE 30

30 Figure 2 8 Mass vs. Flow Limited Determi nation for Rainfall Runoff Event on June 21, 2008 Figure 2 9 Mass vs. Flow Limited Determination for Rainfall Runoff Event on July 8, 2008

PAGE 31

31 Figure 2 10 Mass vs. Flow Limited Deter mination for Rainfall Runoff Event on July 15, 2008 Figure 2 11 Mass vs. Flow Limited Determination for Rainfall Runoff Event on July 29, 2008

PAGE 32

32 Figure 2 12 Mass vs. Flow Limited D etermination for Rainfall Runoff Event on August 8, 2008 Figure 2 13 Mass vs. Flow Limited Determination for Rainfall Runoff Event on August 12, 2008

PAGE 33

33 Figure 2 14 Mass vs. Flow Li mited Determination for Rainfall Runoff Event on August 19, 2008 Figure 2 15. Mass vs. Flow Limited Determination for Rainfall Runoff Event on Sept. 10, 2008

PAGE 34

34 Figure 2 16. Mass vs. F low Limited Determination for Rainfall Runoff Event on Sept. 20, 2008 Figure 2 17. Mass vs. Flow Limited Determination for Rainfall Runoff Event on Oct. 8, 2008

PAGE 35

35 Figure 2 18 Mass vs Flow Limited Determination for Rainfall Runoff Event on Oct. 23, 2008 Figure 2 19.

PAGE 36

36 CHAPTER 3 PHOSPHORUS PARTITION ING AND RAINFALL RUNOFF TREATMENT BY ALUMINUM OXI DE COATED MEDIA Summary Low impact development (LID) in stormwater design is becoming increasingly utilized as urbanization increases and the adverse impacts of these anthropogenic activities are being recognized and recorded. To determine the most effec tive treatment technique, the chemistry and hydrology of an area must be described. The adsorptive properties of phosphorus, known as a problematic nutrient in Florida, causes it to be a favorable constituent to manage, lessening adverse affects on water quality and aquatic ecosystems and aluminum oxide coated media (AOCM) has been proven an effective adsorbent for phosphorus Phosphorus partitioning and knowing the relationships associated with partitioning are essential to determine effective treatment techniques for a specified area. Fifteen storm events were analyzed for partitioning between dissolved phosphorus and sediment, settleable, and suspended bound phosphorus fractions, for influent samples (prior to treatment), and effluent samples (after e xposure to AOCM). All 15 storm events displayed a linear relationship between total dissolved and particulate bound phosphorus after AOCM treatment and ten of 15 events displayed the same linear relationship with respect to the influen t samples. The two largest volume storms show approximate/similar slopes for influent and effluent samples in the relationship between TDP and PBP. The adsorption model was found to be effective with removing dissolved phosphorus in a linear isotherm and the filtration prop erties of the AOCM cartridge was found to be successful in removing PBP mass. The dissolved fraction f d was found to be an average of 0.31, or 31% for influent and 0.21 or 21%, for effluent. The AOCM average removal efficiency (RE) of total dissolved p hosphorus (TDP) was found to be 72.8% and 76.8% when the outlier was removed from the calculation.

PAGE 37

37 The average RE for sediment bound phosphorus was found to be 89.8%, the average RE for settleable bound phosphorus was 81.8%, the removal efficiency for sus pended bound phosphorus was 44.6%, and total phosphorus (TP) was 73%. Background Low impact development (LID) has become a prominent focus of land development, as society has taken environmental sensitivity more seriously, and the economy has taken a dow nturn. The reduced number of perspective home buyers would presumably purchase a builders are working towards LEED (Leadership in Energy and Environmental Design) ce rtification to make their product more marketable. Use of LID alternatives in stormwater has increased substantially in recent years, and range from bioretention, grassed swales to green roofs, pervious pavement, and filtration media; to reduce the level of toxicity before out falling to a lake, river, wetland, or other waterbody. LID development hydrology in an undisturbed condition, and to reduce the impact on soils, vege tation, and aquatic systems effected by new development (Dietz, 2007). Filter media is an attractive LID option, due to application versatility; some examples include use in tanks, ponds, and pervious concrete. If a media has adsorptive properties, it c an be used to adsorb ions as well as filter larger particles containing bound pollutants. However, in the past, phosphorus leaching of media has been problematic due to soil medias used, and fertilization of vegetation used in LID practices (Dietz, 2007). Aluminum oxide coated media (AOCM) is the filter media chosen for this analysis, due to research results suggesting insignificant leaching or desorption of phosphorus from AOCM for neutral pH (Sansalone and Ma, 2009.) AOCM is a favorable material for th e affinity of phosphorus to the adsorptive

PAGE 38

38 properties of AOCM through ion exchange; oxygen atoms of the phosphate ions displace the surface hydroxyl groups, and water molecules associated with aluminum oxide forming a chemical bond (Sansalone and Ma, 2009. ) In this chapter, aluminum oxide coated media (AOCM) will be utilized to determine how this material affects partitioning, and how effectively this LID technique can bind and remove phosphorus from rainfall runoff. The ability to treat stormwater is str ongly influenced by partitioning, therefore it is critical to analyze the partitioning of phosphorus within a watershed to determine the treatability of the rainfall runoff for that area (Sansalone, 2007). The first objective of this chapter is to identif y partitioning of phosphorus between dissolved and particulate bound phase, within each of the 15 rainfall runoff events that were analyzed at a typical parking lot in Gainesville, Florida (See Figure 1 for site map). Operationally, the dissolved portion of a given sample is that which passes through a 0.45 m filter, and thus, the associated phosphorus content is termed as he phosphorus concentration associated with sediment ( settleable (25 m < material < 75 m), or suspended particles (1 m < material < 25 m) (Sansalone, Kim, 2007.) Suspended particles are of concern due to the high surface area, and therefore high potential for pollutants to adsorb to th ese particles. Partitioning can indicate which physicochemical mechanisms will be most effective for immobilization of dissolved and particulate bound mass to water bodies resulting in adverse effects, such as eutrophication (Sansalone J.J. and Buchberger G. 1997). For example, if a particular best management practice (BMP) is efficient in removing only the sediment fraction of the particulate matter, it may not produce desired results of removing phosphorus content for areas high in suspended, settleabl e, or dissolved phosphorus fractions.

PAGE 39

39 Rainfall runoff samples from 15 storms, both influent (before AOCM exposure) and effluent (after AOCM treatment), were analyzed. The relationship between these fractions may also be used to develop predictive tools in the form of general equations or models for efficient, effective, and cost effective, removal of phosphorus from rainfall runoff. The last goal of this chapter is the analysis of AOCM adsorptive filter media for phosphorus removal from rainfall runoff. Influent samples of rainfall runoff and effluent samples following treatment by AOCM were taken from 15 storm events, to obtain a dataset describing the adsorptive behavior of phosphorus to AOCM. Positive results will suggest that AOCM may be used in LID practices to reduce phosphorus concentrations in rainfall runoff, reducing the occurrence of eutrophication and adverse effects to biological and ecological systems of rainfall runoff receiving waters. Methodology Aluminum oxide coated material (AOCM) is a clay based oxide media, and as in Chapter 1, this material was selected based on bench scale testing due its superior efficiency with respect to mass transfer behavior. This media was prepared from phosphorus free clayey soil and aluminum oxide; the we t clayey soil (above optimum water content) was expanded with a blowing agent and fired to 1000C for 8 hours into bricks, then the cooled bricks were crushed to the media substrate, a stable porous material (Sansalone, 2009). This media has proved to rem ove particulate matter and adsorb metals, dissolved phosphorus, and other ions (Sansalone, 2009). As in Chapter 1, the site of interest is Rietz Union parking lot at the University of Florida in Gainesville, Florida. The catchment basin was determined t o be approximately .11 acres (450 m 2 ) with a slope of approximately 3% E W and 1.5% N. The catchment area is subject to

PAGE 40

40 slightly deviate from the surveyed catchment area due to on rainfall intensity and wind direction. See Figure 2 1 for site map. The ra infall runoff draining to the storm sewer on the west side of the parking lot was diverted from the sewer system via 6 inch PVC pipe at a 6% slope, into a one inch Parshall Flume manufactured by Warminster Fiberglass. The flume was calibrated and mounted on wood support so that the flume slope was at 0% as to not affect runoff depth within the flume. A Shuttle Ultra sonic sensor manufactured by MJK, was calibrated and placed 30 inches above the invert of the Parshall flume to measure water depth; depth re adings are then used to calculate flow. A drop box was placed at the flume outlet where samples were taken. See Figures 2 2 and 2 3 for a schematic and set up photos, respectively. A number of weather models were used to predict rainfall, intensity, and duration, including Wunderground (www.wunderground.com), Accuweather (www.accuweather.com), and National Weather Service (NWS) (www.nws.noaa.gov.) Rainfall runoff events producing more than 0.1 inches of rainfall were analyzed. This minimum was set to en sure adequate sample volume to obtain substantial data. An antecedent dry weather minimum of three days was set for accurate estimations of pollutant loadings. An antecedent dry weather minimum of three days was also set to ensure more accurate estimation s of pollutant loadings; this is because the standard assumption for water quality modeling is that pollutants build up on surfaces throughout the watershed between the rainfall events at a constant rate (Davis and McCuen, 2005), overtime, pollutant deposi tion will increase, thus giving a more descriptive data set. Time of rainfall start was recorded as well as the first flow through the Parshall flume. Samples were collected at the outlet of the Parshall flume, 2, 0.5 L bottles and duplicates, 2, 1 L bot tles and duplicates, and 1, 4 L bottle and a duplicate was taken per sample. Sample timing was

PAGE 41

41 dependent on storm duration and intensity as seen by the weather and radar models, sample increment was shortened at the peak of the storm. A 1.0 foot diamete r pipe, extending from the drop box at a 20% slope, delivers rainfall runoff to the sorptive filter and monitoring system, approximately 15 feet from the drop box. The sorptive filter system consists of a cylindrical PVC tank, 22 inches in diameter and 36 inches high, and a concentric cartridge within the cylindrical tank, filled with AOCM media. The system was sized accordingly to treat runoff from the contributing area. The inlet is located 8 inches from the bottom of the tank; after rainfall runoff pa sses through the treatment cartridge, it outfalls from a 3 inch pipe at the top of the tank, where the effluent samples are administered. Two 0.5 L bottles and the respective duplicates, two 1 L bottles and duplicates, along with 1, 5 gallon bucket and a duplicate are filled with effluent sample. Samples were brought into the lab, where a number of tests were done. Dissolved phosphorus content was measured by first filtering sample through a 45 m filter, then administering a Hach orthophosphate test. Or thophosphate reacts with molybdate in an acid medium to produce a mixed phosphate/molybdate complex. Ascorbic acid then reduces the complex, giving an intense molybdenum blue color. Test results are measured at 880 nm by a Hach DR/5000 spectrophotometer. Particulate bound phosphorus content was analyzed by first separating the sediment (material > 75 m), settleable (25 m < material < 75 m), and suspended (1 m < material < 25 m), fractions from each sample. The samples were filtered by a #200 (75 m ) stainless steel sieve to retrieve the sediment fraction, which was then dried and weighed to determine the mass of sediment per sample. To determine the settleable fraction of each sample, sediment free samples were poured into an Imhoff cone and were a llowed to settle for 60 minutes, the settled

PAGE 42

42 material was then dried and weighed to determine the mass of the settleable fraction. The supernatant (sample (sediment + settleable)) was then filtered through a nominal pore size of 1 m filter, which was d ried and weighed to determine the suspended fraction. The sediment, settleable and suspended fractions were then acid digested to convert organic and condensed inorganic phosphorus into a form testable by a spectrophotometer; in other words, releasing pho sphates as orthophosphate in order to test the samples using the same methodology as described above for the dissolved fraction. The acid digestion process uses sulfuric and persulfate acids along with heat to promote conditions for hydrolysis of the cond ensed inorganic forms. The digested samples are then neutralized to a pH above 2.14. After a dilution factor of 10, the test procedure stated above for dissolved phosphorus is then administered to determine the fraction of particulate bound phosphorus. A at this site. Discussion The sum of the dissolved and particulate bound phosphorus, therefore, total phosphorus, concentration can be expressed as: (3 1) This equation states that the total concentration of phosphorus, c T is the sum of the dissolved phosphorus concentration, c d and the concentration of particulate bound phosphorus, c p The ratio of the dissolved and solid phase phosphorus concentrations is termed as the partitioning coefficient ( K d ) in the equation (Glenn and Sansalone, 2002): = c p / c d (3 2) The partitioning coefficient is the ratio of phosphorus fraction concentrations.

PAGE 43

43 To obtain the percentag e of dissolved and particulate bound fractions, the following equations are utilized (Glenn and Sansalone, 2002): (3 3) And, (3 4) Where P is the particulate bound phosphorus mass D, is the dissolved phosphorus mass, and and are the dissolved fraction and particulate bound fraction, respectively. These equations describe partitioning of the rainfall runoff samples at this site, which is potentially representative of parking lots within the state of Florida. For the first goal of this study, influent and effluent cumulative mass of dissolved phosphorus was plotted against the cumulative mass of particulate bound phosphorus to display the relationship, if any, bet ween the fractions of phosphorus throughout influent and effluent samples. A linear relationship will describe a direct relationship between dissolved and particulate bound fractions of phosphorus. This information can be very helpful in prediction model s. Linear regression will verify a linear relationship. Rainfall runoff samples were taken prior to and after filtration to determine the event mean concentration (EMC) of influent verses effluent. EMC values are used to characterize concentrations that represent a flow average concentration for the event (Sansalone et al. 2005), because concentrations of phosphorus can vary substantially during an event, an EMC can characterize the average concentration of a constituent during an entire event. EMC (3 5)

PAGE 44

44 In this equation, M is the total mass of phosphorus over entire event duration; V is the total event flow volume; is the flow weighted average concentration f or the event; c(t) is the time variable phosphorus fraction; q(t) is the time variable flow; t r is the duration of the event and t is time. The second goal of this study is to analyze the effectiveness of AOCM to adsorb and filter phosphorus from parking l ot rainfall runoff. While effluent concentration and mass are the appropriate metric for filter effluent, mass removal can also be quantified as removal efficiency (RE) for any constituent on an event basis using the inflow and outflow loads (Kim and Sans alone, 2008). The following modified equation is the basic equation for calculating the percent removal rate of nutrients or PM. RE (3 6) In the above equation, RE(%) is the percent removal of phosphorus due to treatment by AOCM, V i inf and V j eff are the volume of influent flow and effluent flow during sampling time i and j ; C i inf and C j eff are the mean concentration associated with tim e i and j; and n and m are the total number of influent and effluent measurements taken during event, respectively. The RE of each fraction was also analyzed to further describe the effectiveness of AOCM filter media. Results Influent and effluent total p hosphorus fractions were plotted to show the relationship between total dissolved phosphorus (TDP) and particulate bound phosphorus (PBP) see Figures 3 16 through 3 29 All 15 event effluent plots showed a linear relationship with a linear fit R 2 value o f above 0.95, verified by linear regression. Ten of 15 events displayed the same linear relationship with respect to the influent samples, the remaining five show either a logarithmic

PAGE 45

45 relationship between TDP and PBP cumulative mass, or possibly two linear relationships, where the slope of the linear curve is changed lessoned at a point due to event hydrology or chemistry within the filter cartridge, however, this theory should be studied for verification. I was surprised that at the results suggesting that post filtration, there is a linear relationship, even in events where the influent TDP and PBP relationship was not linear. The most interesting finding being that the volumetrically largest storms (larger by almost a factor of 10), taking place on July 8 th 2008 and July 15 th 2008, show approximate/similar slopes for influent and effluent samples in the relationship between TDP and PBP. This may be due to large flow loadings and less retention time for changes in the chemistry which may affect the slop e of these graphs. Knowing this relationship for high volume storms, may allow effluent phosphorus partitioning to be predicted. The three storms following the July 8 th 2008 and July 15 th 2008 showed, but to a lesser degree, consistency within influent and effluent slope in the relationship between cumulative TDP and PBP mass. Further research should be done on these relationships and the relationship of rainfall runoff volume and partitioning slope. The plots show that the adsorption model is effectiv e with removing dissolved phosphorus in a linear isotherm. This is shown by a strong shift to the left (towards zero) on the x axis (TDP) from influent to effluent. All 15 events also show a reduction in PBP mass from influent to effluent samples, by a r espective shift on the y axis towards zero. Table 3 1 displays the rainfall runoff hydrology characteristics for all 15 events, and Figures 3 1 through 3 15 show the plots and relationships between TDP and PBP mass. As the adsorption sites of the AOCM mat erial are filled, the adsorption efficiency slightly decreases. This is displayed in Figure 3 4, the f d values for the effluent begin to increase in later storms.

PAGE 46

46 In conclusion, TDP is directly related to PBP content following treatment in rainfall runoff events studied at this site. This relationship may be used in prediction models and best management practice decisions based on site characteristics. The result of approximate/similar slopes for influent and effluent samples in the relationship between TDP and PBP for large volume rainfall runoff events may eventually be used to describe influent and effluent partitioning for locations or events experiencing high runoff volume. Further research should be performed on these findings to attempt to develop a universal equation to be used in stormwater quality design. Another conclusion can be drawn that adsorption and particulate filtration by the AOCM cartridge were effective and the relationship between TDP and PBP remained linear throughout all 15 stor m events, and therefore, a change in one will cause a proportional change in the other. To further the success of this best management technique, removal of the suspended fraction should be improved, for this fraction was not removed as effectively as the dissolved, sediment and settleable fractions. A suggestion would be to decrease media size to more effectively remove these particles. Removal efficiencies of phosphorus, in response to aluminum oxide coated material (AOCM) exposure, were calculated for all fractions. The plots of these data are shown in Figures 3 16 through 3 2 9. The average removal efficiency (RE) of total dissolved phosphorus (TDP) was 72.8%, and 76.8% when the outlier was removed from the calculation. The outlier for TDP was the Aug ust 19 th storm which had substantially lower removal efficiency than all 14 other storms (Table 3 5), there was no exceptional hydrologic characteristic of this storm comparatively to other storms and all other fractions had average removal efficiencies f or that storm; however, a linear relationship between dissolved and particulate bound fractions, prior to

PAGE 47

47 and after treatment, was maintained. The average RE for sediment bound phosphorus was found to be 89.8%, the average RE for settleable bound phosphor us was 81.8%, the removal efficiency for suspended bound phosphorus was 44.6%, and total phosphorus (TP) was 73%, see Table 3 5 for removal efficiencies. Tables 3 2 and 3 3 show the event mean concentration (EMC) for each storm, influent and effluent, an d the removal efficiencies for each fraction, and the mean, median and standard deviation for each. Overall, the filter media was found to be effective in phosphorus removal. The removal efficiency for the suspended fraction was lower than the other parts due to the small particle size which is difficult to remove by filtration clogging, and still possibly too large to have effective adsorption of these suspended particles. Decreasing the size of the filter media may increase the removal of the suspended fraction of phosphorus; however research must be done to ensure an effective size with respect to hydraulic pressure and suspended fraction removal. It was also found that the majority of the phosphorus content was particulate bound, with an average f d of 0.31 (31% of phosphorus in the dissolved form) for influent and 0.21 (21% of phosphorus in the dissolved form) for effluent, which also concludes success in the adsorption properties of the filter media by removing TDP, this is demonstrated in Table 3 4

PAGE 48

48 Table 3 1. Event Hydrology and Partitioning Relationships for Influent and Effluent* Rainfall Runoff Events In Gainesville, Florida Rainfall runoff event, 2008 Influent Effluent PDH V runoff Q inf max Q inf avg Event # Graphical Relationship between Tot al Dissolved P and Particulate bound P (hr) (L) (L/s) (L/s) 1 20 Mar 2:05:00 Linear Linear 230 2066 3.4 0.88 2 16 May 17:20:00 Linear Linear 960 2212 8.75 0.9 3 3 Jun 15:30:00 Linear Linear 432 308 0.75 0.11 4 10 Jun 14:08:00 Non Linear Linear 168 8002 10.96 1.61 5 21 Jun 12:15:00 Non Linear Linear 144 1137 3.98 0.34 6 8 Jul 16:46:00 Linear Linear 141 31268 13.17 2.14 7 15 Jul 13:16:00 Linear Linear 66 22438 13.07 3.6 8 29 Jul 11:40:40 Linear Linear 309 1411 3.69 0.55 9 8 Aug 11:06:00 Non L inear Linear 191 474 2.16 0.14 10 12 Aug 14:40:00 Linear Linear 93 3866 3.92 0.62 11 19 Aug 12:53:00 Linear Linear 117 1461 6.78 0.81 12 10 Sep 16:27:00 Linear Linear 264 1569 1.96 0.47 13 20 Sep 13:58:00 Non Linear Linear 234 503 1.09 0.24 14 8 Oct 13:57:00 Linear Linear 431 1557 2.75 0.14 15 23 Oct 15:14:00 Non Linear Linear 334 978 1.46 0.29 Median 230 1557 3.69 0.55 Mean 274 5284 5.19 0.86 SD 221 9112 4.3 0.95 *Effluent Samples taken following treatment by aluminum oxide coated media (AOCM)

PAGE 49

49 Table 3 2. Summary of Event Mean Concentrations and Range of Influent Fractions Rainfall runoff events TDP [g/L] Sediment TP [g/L] Settleable TP [g/L] Suspended TP [g/L] TP [g/L] 20 March 789 1860 533 497 3673 (1428 510) (7778 171) (1782 122) (1203 151) (12191 966) 16 May 517 1693 1371 794 4375 (1523 302) (2999 742) (2994 115) (2223 267) (6988 1964) 03 June 1352 2418 1534 1187 6491 (1767 734) (7564 139) (5764 52) (2083 722) (17178 16 75) 10 June 942 2008 940 1143 5032 (4488 270) (5440 134) (5772 69) (6039 310) (17224 872) 21 June 245 828 465 455 1994 (396 216) (1281 147) (750 68) (574 235) (2814 795) 08 July 222 598 67 1319 2206 (553 144) (3423 121) (11 71 409) (2053 473) (7200 1646) 15 July 476 296 193 802 1768 (1822 216) (841 36) (523 80) (1973 232) (4514 685) 29 July 2664 1188 1003 1341 6196 (4896 1071) (2868 59) (2787 182) (2726 477) (11743 2594) 08 August 579 1510 29 1089 3206 (939 483) (4562 65) (283 0) (1908 687) (7563 1495) 12 August 1023 1499 729 494 3744 (4533 397) (4840 126) (3776 250) (1231 164) (10738 1242) 19 August 410 1231 1271 1367 4279 (986 338) (2480 747) (1717 143) (3917 322) (8866 1820) 10 September 754 822 558 1094 3228 (4394 523) (2709 125) (1249 303) (3964 374) (10215 1292) 20 September 623 519 557 625 2324 (955 472) (2029 135) (1459 104) (1279 146) (5721 995) 08 October 996 392 677 790 2855 (4417 567) (598 58) (1696 53) (1495 414) (7751 1244) 23 October 468 520 254 788 2030 (1078 191) (1392 36) (975 81) (1609 259) (5012 664) Median 623 1188 558 802 3228 Mean 804 1159 679 919 3560 SD 601 656 466 321 1489

PAGE 50

50 Ta ble 3 3. Summary of Event Mean Concentrations and Range of Effluent Fractions Rainfall runoff events TDP [g/L] Sediment TP [g/L] Settleable TP [g/L] Suspended TP [g/L] TP [g/L] 20 March 144 186 429 305 1063 (202 111) (270 80) (1379 0) (373 225) (2135 424) 16 May 135 429 774 851 2189 (461 49) (935 88) (1838 0) (1298 200) (4155 200) 03 June 67 44 136 489 737 (87 50) (93 21) (265 9) (607 364) (947 480) 10 June 227 85 266 325 903 (393 54) (294 52) (701 63) (4 62 191) (1597 402) 21 June 126 89 126 292 633 (147 89) (139 35) (193 21) (377 205) (788 363) 08 July 146 102 0 1258 1505 (221 98) (180 35) (0 0) (1731 647) (2006 895) 15 July 187 59 90 682 1017 (301 51) (138 11) (271 55) (1008 289) (1665 605) 29 July 497 97 279 1199 2072 (733 274) (226 35) (578 104) (1908 458) (3118 1012) 08 August 105 93 4 776 977 (146 36) (145 7) (21 0) (1148 594) (1413 810) 12 August 357 187 124 427 1096 (654 137) (505 2) (625 18) (857 153) (1981 416) 19 August 360 99 244 890 1594 (539 245) (437 7) (485 63) (1740 330) (2677 763) 10 September 299 76 32 443 850 (424 113) (147 40) (410 0) (1093 141) (1904 455) 20 September 255 89 104 429 878 (424 174) (157 20) (250 0) (577 199) (1234 469) 08 October 208 92 134 705 1139 (514 110) (224 32) (289 0) (1245 145) (2147 490) 23 October 87 106 15 463 671 (349 26) (321 41) (170 0) (773 266) (1155 406) Media n 187 93 126 489 1017 Mean 213 122 184 635 1155 SD 121 93 202 309 478

PAGE 51

51 Table 3 4. Summary of Event Mean Value of Dissolved Fraction (f d ) and Partition Coefficient (k d ) of Phosphorus for Influent and Effluent Runoff Rainfall runoff event f d k d ( ) ( L/Kg) Influent Effluent Influent Effluent 20 March 0.43 0.25 11945 112430 16 May 0.21 0.13 18303 176615 03 June 0.38 0.11 11038 220225 10 June 0.32 0.23 13834 97984 21 June 0.25 0.23 43146 136479 8 July 0.10 0.10 137208 417623 15 July 0.24 0.20 57524 308904 29 July 0.49 0.23 9021 141626 8 August 0.27 0.10 10986 623604 12 August 0.36 0.30 14808 137661 19 August 0.18 0.2 7 20963 72798 10 September 0.36 0.33 11978 73812 20 September 0.38 0.29 11593 45576 8 October 0.44 0.21 10138 84926 23 October 0.27 0.15 25417 222488 Median 0.32 0.23 13834 137661 Mean 0.31 0.21 27193 191517 SD 0.11 0.08 33362 155297

PAGE 52

52 Table 3 5. Removal Efficiencies of Phosphorus Fractions Rainfall runoff events TDP Sediment TP Settleable TP Suspended TP TP Removal Efficiency Removal Efficiency Removal Efficiency Removal Efficiency Removal Efficiency (%) (%) (%) (%) (%) 20 Mar 89.1 94.1 52.3 63.6 82.8 16 May 83.5 83.9 64.2 32.0 68.3 3 Jun 96.6 98.8 94.0 72.2 92.3 10 Jun 83.7 97.1 80.8 80.7 87.8 21 Jun 48.2 89.2 72.7 35.5 68.1 8 Jul 74.8 93.5 100.0 63.5 73.8 15 Jul 87.2 93.5 84.7 72 .2 81.2 29 Jul 82.9 92.5 74.5 18.1 69.4 8 Aug 82.7 94.2 87.8 32.2 71.0 12 Aug 64.5 87.3 82.6 12.0 70.2 19 Aug 17.6 92.5 82.0 38.9 65.1 10 Sep 63.1 91.4 94.7 62.2 75.4 20 Sep 57.0 82.1 80.4 28.1 60.4 8 Oct 81.6 79.4 82.5 21.1 64.8 23 Oct 80.1 78.3 9 3.5 37.3 64.7 Median 81.6 92.5 82.5 37.3 70.2 Mean 72.8 89.8 81.8 44.6 73.0 SD 20.1 6.3 12.4 22.2 9.2

PAGE 53

53 Figure 3 1. Influent and Effluent Relationships between TDP and PBP for March 20, 2008 Figure 3 2. Influent and Effluent Relationships between TDP and PBP for May 16, 2008

PAGE 54

54 Figure 3 3. Influent and Effluent Relationships between TDP and PBP for June 3, 2008 Figure 3 4. Influent and Effluent Relationships between TDP and PBP for June 10, 2008 :

PAGE 55

55 Figure 3 5. Influent and Effluent Relationships between TDP and PBP for June 21, 2008: Figure 3 6. Influent and Effluent Relationships between TDP and PBP for July 8, 2008

PAGE 56

56 Figure 3 7. Influent and Effluent Relationships between TDP and PBP for July 15, 2008 Fig ure 3 8. Influent and Effluent Relationships between TDP and PBP for July 29, 2008

PAGE 57

57 Figure 3 9. Influent and Effluent Relationships between TDP and PBP for August 8, 2008 Figure 3 1 0. Influent and Effluent Relationships between TDP and PBP for August 12, 2008

PAGE 58

58 Figure 3 11. Influent and Effluent Relationships between TDP and PBP for August 18, 2008 Figure 3 12. Influent and Effluent Relationships between TDP and PBP for Sept. 10, 2008

PAGE 59

59 Figure 3 13. Influent and Effluent Relationships between TDP and PBP for Sept. 20, 2008 Figure 3 14. I nfluent and Effluent Relationships between TDP and PBP for Oct. 8, 2008

PAGE 60

60 Figure 3 15. Influent and Effluent Relationships between TDP and PBP for Oct. 23, 2008 Figure 3 16. Influent and Effluent Total Phosphorus Removal for the Event of March 20, 2008

PAGE 61

61 Figure 3 17. Influent and Effluent Total Phosphorus Removal for the Event of May 16, 2008 Figure 3 18. Inf luent and Effluent Total Phosphorus Removal for the Event of June 3, 2008

PAGE 62

62 Figure 3 19. Influent and Effluent Total Phosphorus Removal for the Event of June 10, 2008 Figure 3 20. Influent and Effluent Total Phosphorus Removal for the Event of June 21, 2008

PAGE 63

63 Figure 3 21. Influent and Effluent Total Phosphorus Removal for the Event of July 8, 2008 Figure 3 22 Influent and Effluent Total Phosphorus Removal for the Event of July 15, 2008

PAGE 64

64 Figure 3 23. Influent and Effluent Total Phosphorus Removal for the Event of July 29, 2008 Figure 3 24. Influent and Effluent Total Phosphorus Removal for the Event of August 8, 2008

PAGE 65

65 Figure 3 25. Influent and Effluent Total Phosphorus Removal for the Event of August 12, 2008 Figure 3 26. Influent and Effluent Total Phosphorus Removal for the Event of August 19, 2008

PAGE 66

66 Figure 3 27. Influent and Effluent Total Phosphorus Removal for the Event of Sept. 10, 2008 Figure 3 28. Influent and Effluent Total Phosphorus Removal for the Event of Sept. 20, 2008

PAGE 67

67 Figure 3 29. Influent and Effluent Total Phosphorus Removal for the Event of Oct. 8, 2008 Figure 3 30. Influent and Effluent Total Phosphorus Removal for the Event of Oct. 23, 2008

PAGE 68

68 CHAPTER 4 CONCLUSION It is historically known that population and urbanization of the United States has continuously increased, and despite the recent ec onomic recession slowing new development, the population still increased by 0.915% in 2008, alone. In most metropolitan areas of Florida and the United States, the degree of imperviousness is greater than 60% of the metropolitan areas (Sansalone et al. 20 08). This increasing urban sprawl directly results in an increase of rainfall runoff quantity and decrease of rainfall runoff quality due to the addition of impervious surfaces. Another adverse impact of urban sprawl we are challenged with is depletion o f surficial ground water coupled with increase of flooding, which is a strong concern in the state of Florida. Aluminum oxide coated media has proven to be successful in phosphorus removal, which increases water quality of rainfall runoff, and if this med ia is used in pervious pavement applications, the result will not be limited to pollutant uptake, but may lessen the occurrence of flooding, and recharge surficial groundwater. As only one method is described in this study, a diverse set of treatment tech niques and best management practices should be evaluated for the unique runoff chemistry of a drainage basin, to attempt preservation of the hydrologic cycle and biologic/ecologi c presence in the environment. The research in this study can be generalized to small urban pavement watersheds with similar weather conditions, therefore the conclusions are not site specific. Regulators are now looking more closely at current regulations, and recognizing that historically used regulations and best management pr actices are not always effective in removing pollutants from rainfall runoff, which was demonstrated further by this study and suggests that first flush phenomena should not always be used as an assumption on which to base regulations.

PAGE 69

69 Partitioning data should be used in determining appropriate and effective LID practices specific to location. Laboratory research to determine particulate bound concentration of runoff at a site can be expensive and time consuming. Knowing the relationship between dissolve d and particulate bound phosphorus would allow for dissolved analyses to be performed, and the particulate bound fraction to be calculated based on this relationship, therefore making the choice of a treatment technique an educated, less expensive and much easier choice. It should also be a determin ed a relationship between site characteristi cs and partitioning coefficient Therefore, this report should be followed up with further research to determine a predictable relationship between dissolved and part iculate bound phosphorus based on site specific criteria. This study supports that use of AOCM filter media may be very effective in reducing phosphorus loading from rainfall runoff, prior to outfall into natural water systems. However, due to the low re moval efficiency of the suspended fraction, partitioning of a drainage basin should be evaluated prior to implementation of this treatment technique. Since AOCM was very successful in removing all fractions, with the exception of suspended phosphorus, a s uggestion is decreasing the media size to increase the removal efficiency of this fraction. To make a substantial impact, research must continue striving towards more economically feasible LID practices and government must place regulation requiring these practices, to protect wildlife, aquatic systems, and groundwater from anthropogenic and organic pollutants associated with rainfall runoff. We are only in the primitive stages of developing the tools needed in the attempt to achieve sustainability and de ter future degradation of aquatic, biologic, and ecologic systems.

PAGE 70

70 LIST OF REFERENCES Barrett, M. E. (2005) Performance Comparison of Structural Stormwater Best Management Practices Water Environment Research ( 77) 78 86. term Stormwater Quantity and Quality Performance Water Research, 2003, 37(18) 4369 4376. ll Runoff Journal of Environmental Engineering, 129(7) 629 636. Caltrans Division of Construction ( 2003) Stormwater Quality Manuals and Handbooks ( http://www.caltrans.ca.gov/ hq/construc/stormwater/CSBMPM_303_Final.pdf ) of Urban Storm Water Runoff. Journal of Environmental Engineering, 131(11) 1521 1531. Springer Science+Business Media Inc. 149 Recommendations for Future Directio Water, Air, and Soil Pollution 351 359. J. Environ. Eng. ASCE., 133(5), 485 497. Asphalt, 16 18. Runoff from an Ultra Journal of Environmental Engineering, 133(6) 616 625. on and Partitioning of Heavy Metals Associated with Journal of Environmental Engineering, 128(2) 167 185 Gnecco I, Sansalone J LG.Lanza (2008) Speciation of Zinc and Copper in Stormwater Pavement Runoff from Airside and Landside Aviation Land Uses. Water and Soil Pollution, 192(1 4) 321 336 Filter Media: Water Research 41, 2513 2524.

PAGE 71

71 Hsieh, C. H.; Davis, A.P; B.A. Needelman (2007) Bioretention Column Studies of Phosphorus Remov al from Urban Stormwater Runoff Water Environment Research 79 ( 2 ) 177 84. for Analyzing Highway Storm Journal of Environmental Engineering, 124(10) 987 993. Journal of Hydraulic Research, 34, 799 813. Lian, C.; Y. Zhuge (2010) Optimum mix design of enhanced permeable concrete An experimental investigation Construction and Building Materials 24(12 ) 2664 2671 Ru noff Metals by Composite Oxide Journal of Environmental Engineering 131(2) 1168 1176. Journal of Environme ntal Engineering 131(8) 1178 1186. Journal of Hydrologic Engineering 7(4) 319 325 unitoperati University of Florida Publication Journal of Water Resources Planning and Management, 172 179. d First Flush of Metals in Urban Roadway Journal of Environmental Engineering 123(2) 135 142. Journal of Environmen tal Engineering, 120(11) 1301 1314. Journal of Environmental Quality, 37, 1883 1893 Sansalone, J., Koran, J., Smithson, J., and S. Bu J. Environ. Eng. 124(5) 427 440. Journal of Irrigation and Drainage Engineering, 134(5) 666 673.

PAGE 72

72 Water Journal of Environmental Engineering 135(9) 737 744. Sarasota County (2009) Sarasota County Site and Development Application Packet, Planning and Devel opment ( http://www.co .sarasota.fl.us/PlanningandDevelopment/LandDevelopment/documents/Co ncurrentCommercialApplPkt.pdf ) S Dissolved Water Chemistry Load Indices in Rainfall runoff from Urban Source Area Journal of Hydrology, 361, 144 158. Siriwardene N.R.; Deletic, A.; T.D. Environmental Science & Technology, 41 (23) 8099 8103. Southwest Florida Water Management District (2010) Environmental Resource Permit Applications within the Southwest Florida Wat er Manag ement District Part B, Basis of ( http://www.swfwmd.state.fl.us/files/database/site_file_sets/17/erp_basis_of_review.pdf ) Southwest Florida Water Mana gement District (2010) Summary of Amendments to Basis of Review for Env ironmental Resource Permit Application Within the Sout h Fl orida Water Management District ( https://my.sfwmd.gov/portal/page/portal/xrepository/sfwmd_repository_pdf/bor_erp_08 _20_2010.pdf ) Green Roof Stormwater Re tention: Effects of Roof Surface, Slope, and Media Depth Journal of Environmental Quality 34(3) 1036 44 Through Redesigning Stormwater Systems: Looking to the Catch ment to Save the Journal of the North American Benthological Society, 24, 690 705. Proceedings of the Water Environment Feder ation, (91 100), 6945 6957.

PAGE 73

73 BIOGRAPHICAL SKETCH Tracy Fanara graduated with a B achelor of S cience from the Environmental Engineering D epartment at the University of Florida in 2003, after transferring from Hobart and William Smith Colleges in 2000. Follow ing an undergraduate education she worked as a project engineer for four years before returning t o the University of Florida to pursue graduate school