Passive Phosphorus

Removal Systems

Chad Penn, Josh Payne, Jeff Vitale:

Oklahoma State University

Josh McGrath: University of Maryland

Delia Haak: Illinois River Watershed

Partnership

P FILTER BACKGROUND AND

BENEFITS

Target: Dissolved P

• Soils built up with “legacy” P will continue to release dissolved P for many years

– Conventional best management practices do little for DP

• Objective: construct P removal structures to trap dissolved P in runoff and tile drainage

Need for P filters

• High soil P concentrations contribute to long-term, slow P leak

• This is referred to as a “legacy” P issue

Me

hlic

h-3

Ph

osp

ho

rus (

mg

kg

-1)

Coale, F.J. and R. Kratochvil 2011: Unpublished data

Justification of cost for construction

Current BMPs only

slightly

effective for DP

losses

DP: 100%

biologically

available

DP: sustained

transport

for several

years

$$

P Removal Structure Theory

High P water

PSM layer

Drainage layer Clean water is released

3 Necessary Components

• Effective PSM in

sufficient quantity

• P-rich water must

flow through PSM

• Ability to retain

and replace PSM

Example by-product PSM’s

Acid mine

drainage

treatment

residuals

Bauxite

mining and

production

waste (red

mud)

Steel slag

waste

Drinking

water

treatment

residuals

Fly ash

Waste

recycled

gypsum

Foundry Sand

Selection Process for PSMs

Material Availability

Cost & Transportation

Potential contaminants

Sodium

Soluble salts

Total, acid soluble,

and water soluble heavy metals Sorption characteristics

Physical Properties

Particle size distribution

and bulk density

Hydraulic conductivity

Cartridge Filter?

Portable, easy to

install

Does not work!

Limited amount

of PSM

Confined Bed

• Good for large filter

• Ideal for drainage swales that require high peak flow and non restricted drainage

– Achieved through shallow PSM with large surface area

Confined Bed Example

Overflow weir

3 tons >6.35 mm EAF

slag, treated 150 acres,

25% overall dissolved P

removal after 8 months

Structure has handled

flow rates over 100 gpm

Confined Bed Example

Overflow weir • Structure re-loaded

with >0.5 mm slag

• 33% overall removal

after 18 months

• Still handles high

flow rate

Tile Drain • Similar to bed, but

without confinement

• Allows large amount of material to be used

• Use flow control to build head

• Low cost

• Probably best option for ditches

Box Filter • Easily switch out

material

• Modular design – integrates with flow control

– Agri-Drain

• Small ditches or pond overflow

• Drawback: Small amount of material

Storm Drain Filter

Performance and Lifetime

• It depends:

– Site conditions

– PSM

– Design

0.5

0

5

10

15

20

25

30

0.5 1 510

15

Retention time (min)

P r

em

ov

ed

(g

kg

-1)

P concentration (mg L-1)

10

DESIGN GUIDANCE

General Design Approach

Site hydrology 1. Peak flow rate 2. Annual flow volume 3. Dissolved P level 4. Max footprint

P removal & lifetime 1. Target P removal (%) 2. Target lifetime

+ Model

+

PSM characterization 1. P sorption 2. Safety 3. Physical properties

Design parameters 1. Area 2. Mass of PSM 3. Depth of PSM

Input Output

Model Verification

• Ability to predict design curve is key

– Accuracy of inputs

• Golf course structure

• UMD structures on Eastern Shore

• Pilot scale structure

• Laboratory flow-through experiments

conducted on new materials

Pilot Scale Filter

Example Pond Filter Data

0

10

20

30

40

50

60

70

0 50 100 150 200

Cum

ula

tive

P r

emoved

(m

g k

g-1

)

P added (mg kg-1)

Measured

Predicted

Example Golf Course Structure

P added (mg kg-1)

P r

em

ov

ed

(m

g k

g-1

)

EXAMPLE DESIGN

Design: Poultry farm in OK

Green Creek

flow direction

poultry houses

proposed structure location

Design: Poultry farm in OK

Site Conditions

• Drainage area: 9 acres

– Slope: 6%

• Curve #: 78

– Peak flow rate, 2yr-24hr storm: 16 cfs

– Annual flow volume: 9 acre-ft

• Typical dissolved P: 1 to 2 mg L-1

– Annual dissolved P load: 49 lbs (22 kg)

• Assumed 2 mg P L-1

• Goal: remove 45% of annual P load

Sizing the Structure: design

curve for potential PSMs Measured directly via flow-through sorption tests or predicted based on chemical and physical properties

Unique to each PSM, inflow concentration, and retention time

Structure design: impact of PSM

characteristics

• Need to handle peak flow rate • Need to handle P load

Structure design: impact of PSM

characteristics

• Need to handle peak flow rate • Need to handle P load

PSM Mass (Mg)

Cumulative

year 1

removal (%)

Lifetime

(yrs)

Hydraulic

conductivity

(cm s-1)

Area

(m2)

PSM depth

(cm)

WTR* 7 37 21 0.01 286 2.3

AMDR† 4 50 7 0.009 225 2.2

Fly ash‡ 3 (plus

95% sand) 50 3.6

0.03 (mixed

with 95% sand) 406 13

>6.35 cm slag§ 171 21 1.4 1.0 190 50

Treated > 6.35

cm slag** 36 45 3.5 1.0 40 50

Potential design options to meet

given P removal and 16 cfs flow

Completed Structure

Completed Structure

Completed Structure

• 40 acre tiled drained field to bio-reactor

– Peak flow rate: 80 gpm

– Typical P concentration: 0.2 to 0.5 mg L-1

– Annual P load: 0.5 kg (from 1 mil L yr-1)

– Hydraulic head: 17.89 cm (over 90 ft length)

• Target removal: 3 year lifespan, 25 or 50%

• Maximum area: 400 sq ft

• Material: Sieved Slag

Design: Tile drained field in IN

Design: Tile drained field in IN

Design: Tile drained field in IN

Design: Tile drained field in IN

Design Options: sieved slag

Target lifetime @

25% removal (yr)

Mass (tons)

Depth (in)

Head (in)

Peak flow

(gpm)

Total lifetime

(yr)

Cumulative removal

(%)

Area (sq ft)

1 5.5 6.8 6.8 206 9 10 239

3 16.5 Not enough area

27 10 718

54 10 1436 6 33

Design Options: coated slag

Target lifetime @

50% removal (yr)

Mass (tons)

Depth (in)

Head (in)

Peak flow

(gpm)

Total lifetime

(yr)

Cumulative removal

(%)

Area (sq ft)

1 4.7 6.8 6.8 2181 18 12 148

6 9.5 6.8 6.8 4363 36 12 296

10 15.7 Not enough area 60 12 494

Additional Support

• Design software currently being created

– Provide interactive design guidance based on

user inputs

– Will be made available to NRCS

• NRCS Standard will be completed after

software is completed

• Commercialization may be key to

dissemination

Comparison to other BMPs

• In the short term there is no BMP that can

appreciably reduce soluble P losses where

flow cannot be reduced

– P “mining” with hay crops

or corn to reduce soil P

levels

• Sharpley et al. (2009): only

4.6 mg/kg decrease per year

in Mehlich-3 P with

continuous corn

• Not very fast

Comparison to other BMPs • Treatment wetlands

– Require excessive

retention time (days),

thus requires many

acres of space if high

flow rates are to be

treated

• inefficient

– P is not really removed

from the system

ECONOMICS

Economics Example: Westville

• Metal & custom fabrication: $2677

– ¼” carbon steel

• Slag transportation, sieving, coating: $853

• Earth work for pad & berms: $846

• Paint, seed, & erosion mat: $613

• TOTAL: ~ $5000

• Includes profit from private companies

except for metal painting and installation

• Annual renewal estimated at $1213

Economics Example: Westville

Year $ P removal

(lbs.)

Cumulative P removal cost

($/lb P)

1 4989 22 226.77

2 1213 22 140.95

3 1213 22 112.35

4 1213 22 98.05

5 1213 22 89.46

6 1213 22 83.74

7 1213 22 79.66

Economics

• Economics will vary greatly between

locations

• Westville structure could have been

constructed for less money by using other

materials

• Consider that it costs a waste water

treatment plant 50-200 dollars/lb P

removed

– Easier to remove P at a high P concentration,

point source

www.p-structure.blogspot.com