An Engineering Approach to the Removal of Dissolved Copper from

Highway Runoff

Dan Gunther, P.E.ODOT Geo/Hydro/HazMat Conference

23-Sep-09

Dissolved copper is a huge issue but there are some conflicting statements

and findings regarding its toxicity.

Why are regulators so concerned right now about salmon populations?

• Sick and dying salmon are observed in urban streams all around Puget Sound at rates of 40% to 90%.

• NMFS researcher Dr. Nathan Scholz showed pre-spawn mortality (PSM) is related to toxins in the urban streams.

• Scholz’ recent research has determined the unidentified toxins can be filtered out of urban streams .

Filtered water produced normal Coho embryos

•This is very recent data and Scholz’ team have not yet been able to identify the toxin(s) in the water.

How is ionic copper related to PSM?

• Ionic copper is a major toxin for most marine animals

• Even 2 µg/L adversely affects Coho behavior

• Ionic copper kills the olfactory and lateral line nerves

• Does brake pad copper become ionic copper?

Not all researchers agree on copper toxicity

Sandahl reports: Chum salmon can recover from short-term (4hr) exposures to copper (3-58 µg/L) within 1 day

Not all brake pads contain copper

• Armstrong, Louis, J., (Woodward-Clyde Consultants), “Contribution of Heavy Metals to Storm Water from Automotive Disk Brake Pad Wear”, Santa Clara Valley Non-point Source Pollution Control Program, 10/12/94

• Note that Ford and GM (except one model) have no copper in their brake pads.

Does brake pad copper become ionic copper?

“From the information available, it appears that the brake pad-derived copper is in nontoxic/non-available forms.”1

1 Lee G.F. and Jones-Lee, A., “Regulating Copper in Urban Stormwater Runoff,”

Timely Topics, NorCal SETAC News 11:(No 2) 10-11, June (2000)

What does FHWA say? Does it matter?

From: FHWA Environmental Technology Brief: 2

Is Highway Copper Runoff a Serious Problem? – “Not necessarily.”– “Heavy metals in highway runoff are usually not a toxicity problem.”

How to treat highway runoff:– Detention and retention ponds– Vegetated swales and filter strips– Filtering systems

2 FHWA Environmental Technology Brief, “Is Highway Runoff A Serious Problem?”, Office of Infrastructure R&D, Turner-Fairbark Highway Research Center, 6300 Georgetown Pike, McLean, VA 22101.

• A recent NMFS paper defined ambient concentrations of copper, in surface water as < 3 µg/L*

• Levels of concern (approximate ranges)*– Lethal 19-38 µg/L– Delayed or abnormal migration: 5-25 µg/L– Reduced olfaction and predator avoidance: 0.18-2.4 µg/L

Copper levels of concern

* Hecht et al 2007, An overview of sensory effects on juvenile salmonids * Hecht et al 2007, An overview of sensory effects on juvenile salmonids exposed to dissolved copper: Applying a benchmark concentration approach exposed to dissolved copper: Applying a benchmark concentration approach to evaluate sublethal neurobehavioral toxicity.to evaluate sublethal neurobehavioral toxicity.

What is the EPA Regulated Level for dissolved copper?

1.9 µg/L and 3.1 µg/L are draft chronic and acute regulated copper levels. (Dec. 2003)

“The 4-day average concentration of dissolved copper does not exceed 1.9 µg/L more than once every 3 years on the average (i.e., the CCC) and if:

The 24-hour average dissolved copper concentration does not exceed 3.1 µg/L more than once every 3 years on the average (i.e., the CMC).”

What are the dissolved copper levels around the US?

• East Coast: Chesapeake Bay 6-10µg/L

• West Coast: South San Francisco Bay 3.3 µg/L (dry season) 2.4 µg/L (wet season)

• Copper River in Alaska “frequently exceeds 2.0 µg/L”

• Columbia River - Portland, OR < 2µg/L

Sources of copper in ODOT Region 1(my preliminary calculations and observations)

• What is the estimated dissolved copper concentration from brake pads in Region 1 highway runoff?

• What are other sources of copper entering Portland Metro area watersheds?

Estimated quantity of dissolved copper in highway runoff from brake pads

Assumed number of brake pads per car (front does most of the work) 4Average vehicle miles driven per year 10,000 Average vehicle miles between brake pad changes 40,000 Mass of brake dust per vehicle per year, kg 0.129

Oregon DMV 2004 registered passenger cars 3,153,327 Percent of cars used daily 33% Percent Copper in brake pads 8%

Mass of Copper into Oregon environment from brake pads, kg/yr 10,739

Percentage of Copper from brake pad dust that flows into runoff 30%

Average Oregon rainfall (estimated statewide average), m/yr 0.21Oregon total area, ha 24,862,964 Loading of Copper to environment kg/ha-yr 0.000431

Oregon 2005 State hwy lanemiles 19,048 Oregon 2005 County & City, centerline miles 25,559 Lane + shoulder width 20Oregon roads (state+county+city (not fed. or parks)), sq miles 362 Road area in sq meters 939,115,914 Volume of runoff (liters) 200,000,000,000 Concentration of Copper in runoff µg/L 16

ODOT Region 1 copper sources (includes Washington, Clackamas, & Multnomah Counties)

Copper sources Kg/Year

Agriculture 82,000

Anti-fouling Marine Paint 22,000

Treated Wood 10,000

Brake Pads 7,000

POTW’s 5,000

Drinking Water 2,000Total R1 copper loading 128,000

Which BMP to use?

Question Regarding BMPs

• What issue exists with copper removal in current Best Management Practice (BMP) water quality facilities?– Lower limit of copper removal appears to be 5 µg/L (no

matter what the reported percent removal rate)

• What could engineering solutions look like?– Economical modification of existing systems with

manageable side effects – An add-on device or chemical process that effectively

reduces metals

FHWA Stormwater Best Management Practices (BMP) Database and NCHRP Report 565 are good starting points.

What tools does ODOT have to treat highway runoff for dissolved copper?

• Common BMP types– 100% Infiltration– Wet pond– Media Filter Drain– Dry pond with bottom swale – Vegetated filter strip– Cartridge filter

• These systems were developed for suspended solids removal • Except for infiltration and wet ponds, they have limited ability

to remove dissolved metals at low concentrations

• None of the BMPs can reduce dissolved copper to the currently regulated limit of 1.9/3.1 µg/L

• None of the available BMP’s are capable of reducing dissolved copper to less than 5 µg/L

• Retention Ponds (=wet ponds) are best for metals removal, but they require a lot of right-of-way ($$$)

• Biofilter (=MFD) uses much less area and has fairly good metals removal

Which BMPs are most practical?

NCHRP

FHWA

MFD system key features:

• MFDs use vegetation, soil, and gravel in a coordinated flow-path to remove pollutants

• Plant-filtering with soil-filtering is more effective than plant-filtering alone

• Lowest (median) copper level achieved by WSDOT was 7.1 µg/L for Ecology Embankments (now called media filter drains (MFDs)

• Metal hydroxides and carbonates are relatively insoluble and are likely by-products of systems using dolomite, gypsum and perlite

Proposed MFD enhancement: Add phosphate !

• MFDs would be improved if metals removal was enhanced

• Metal sulfides and metal phosphates, reacting with copper in stormwater, are more chemically inert than hydroxides and carbonates

• Metal sulfides and phosphates are less likely to release metals back to the water once they form

• Fish bone apatite (Apatite II*) sequesters metals by four general, non-mutually exclusive processes:

1. Buffering action near pH 6.5-7.0 causes metals precipitation

2. Phosphate ions promote metal-phosphate precipitation

3. Stimulation of sulfate-reducing bacteria4. Extensive adsorption surface area

* Patent # 6,217,775, PIMS NW, Inc. Dr. Judith Wright

Fish Bone Apatite is one source of phosphate

Ca10(PO4)6(OH)2

F, Cl, Br, CO3, X

CO3, SO4, SiO4, XOy

Pb, Cu, U, Zn, Cd, Th, Cr, Co, Na, Ni, Sr, Rb, Zr, Cs, Au, Ba, Ir, Hg, Se, As, Ta, Fe, and others

• Fish bone apatite is 10 to 100 times better than cow bone apatite and 100 to 10,000 times better than phosphate rock at metal sequestration.

• Fish bone apatite is not contaminated with prions or heavy metals as cow bone and rock phosphates may be.

ODOT Site for Phosphate-metal study

• Study site provided a good composite sample of typical highway runoff

• Box culvert drains approximately 5 miles of I-205

• Accessible location

• Potential regional treatment facility site

Column Filter Tests• One column was apatite + sand

• One column was apatite alone

• Runoff samples were filtered through the columns to reduce dissolved metals

• Runoff samples were filtered through the columns at rates between 0.5 and 14 minutes per 16 oz. (=15 - 0.5 gpm/ft2)

ODOT Study: Results

• Found a workable filtration rate (0.8 gpm/ft2)• Media quantities are reasonable (0.5 -1 ton per 4

acres of impervious surface)• Material cost is reasonable ($750/ton)• Effluent dissolved copper < 5 µg/L !!! (average 2.2 µg/L with minimum of 1.6 µg/L)• Other metals also are significantly reduced

ODOT Study: Preliminary Findings

Proposed MFD modification:Add sulfide !

US26: 185th-Cornell Project

• Initial estimates of space required for bio-infiltration swales indicated a need for retaining walls as well.

• Preliminary cost estimate was $4.4M for swales and walls

• Raise the “V” ditch but provide a high capacity drainage route for runoff

• Make MFD traversable adjacent to pavement edge – use cellular confinement grid to increase gravel stability

• Add pollution MH for fuel spills• Maintenance is essentially

unchanged, MFD media is similar to “sanding rock”

• MFD media life is 10-30 years based on bench scale tests and WSDOT experience

MFD modifications to fit the site

MFD modifications to enhance dissolved copper removal

• MFD media is per WSDOT• Adding phosphate may increase copper removal• Enrich drain rock zone below drain pipe with 5% compost to grow Disulfovibrio

bacteria• Periodic addition of gypsum (CaSO4 ) for Disulfovibrio bacteria• Clay dams along MFD should create anaerobic conditions

Clay Dams

• Clay dams could pool water in a non-turbulent zone below the drain pipe and encourage anaerobic bacterial growth

Similar project using Disulfovibrio for dissolved metals removal

Leviathan Mine in California was owned by Anaconda, is now a superfund site

Successful bio-remediation with sulfate reducing bacteria

“Over an evaluation period of 20 months, from late 2003 to summer 2005, the bioreactor was able to achieve a target-metal removal efficiency of 95 percent. All target metals, except iron, were reduced to concentrations below the EPA interim discharge standards.”

Leviathan Mine Bioreactor

• River rock media, 6”- 20” dimension

• Ethanol for carbon source• Sulfate from gypsum?• Precipitated metal-sulfide is

collected with a cloth filter bag

Results - Influent Copper: 690 µg/L Effluent Copper: 5 µg/L

Thank You

Questions?