Process engineering: Wastewater treatment plant goes the distance

Editor's Note: There is a figure that accompanies this article that can be downloanded in PDF format via the "Download Now" button at the bottom of the page.

The Piney Point fertilizer plant in Palmetto, Fla., was shut down in 1999 after its parent company declared bankruptcy. The state of Florida in 2001 started managing the facility, which had made phosphoric acid. A series of ponds at the site held 1.2 billion gal. of acidic, ammonia-laden process water. These ponds were in danger of overflowing, thereby spilling contaminated water into Tampa Bay.

To prevent a disaster, the state organized a team to manage the site and select wastewater treatment and offsite disposal methods to handle the contaminated ponds. They chose a mobile reverse osmosis (RO) system from USFilter, Warrendale, Pa., to treat about 800,000 gpd so the water could be discharged into Tampa Bay without adversely affecting the environment.

Why Piney Point was a problem
During phosphoric acid manufacturing, phosphate rock reacts with sulfuric acid and forms calcium sulfate, or gypsum, as a waste byproduct. This waste gypsum is known as phosphogypsum. Each ton of phosphoric acid produced generates about 4.5 tons of phosphogypsum. The phosphogypsum is mixed with hot acidic process water to create a slurry that is pumped to a disposal area on site. This area is called a phsophogypsum stack, known locally as gyp stacks, whereas the pumps, ditches, surge ponds and other means for collecting and conveying the phosphogypsum and process water are called the phosphogypsum stack system (Figure 1).

When the Piney Point facility was abandoned, its phosphogypsum stack system contained about 1.2 billion gal. of acidic process water, about half of which was stored in ponds on top of the gyp stacks or contained in adjacent above-grade cooling ponds. The process water was contaminated with a variety of pollutants, such as ammonia nitrogen, phosphates, fluoride, radioactive compounds and trace metals, in concentrations that exceed the Florida Department of Environmental Protection (FDEP) water quality standards.

According to FDEP, the main environmental risk posed by Piney Point was a catastrophic failure of the more than 30-year-old gypsum stack dikes. Such a failure would release acidic process water into an estuary along Tampa Bay, adversely affecting sea grasses and marine organisms.

Once rainwater comes into contact with process water or another contaminated part of the plant, it adds to the total volume of wastewater at the site. Due to the size of the Piney Point facility, more than 12 million gal. of process water are collected for every inch of local rainfall. The year the site was shut down, the area experienced heavy rains and tropical storm Gabrielle.

After the area received 19 in. of rain in September 2001, a group of local engineers developed a list of alternatives to quickly reduce the quantity of process water at Piney Point.

The team and FDEP investigated the feasibility of transferring the water offsite. They found, however, that the daily capacity of other fertilizer plants or Tampa wastewater treatment plants was insufficient to handle the volume of process water at Piney Point. Therefore, discharge options had to be developed for both the short-term recovery from Gabrielle and the long-term closure of the site.

For short-term recovery from the storm, the U.S. Environmental Protection Agency (EPA) issued a permit allowing fully treated wastewater to be discharged 100 miles offshore into the Gulf of Mexico. About 248 million gal. were discharged offshore using two 10-million-gal. tankers between July and November 2003. The long-term closure required reliable and sustainable options to remove between 1.0 Mgpd and 1.5 Mgpd from the site during a three-year period.

The promise of success
The FDEP considered RO technology as the most promising treatment method for producing a high-quality effluent with low concentrations of ammonia while avoiding the neutralization and disposal costs associated with the existing two-stage lime process at Piney Point. There were two concerns about using RO, however. The first was the absence of historical data on RO performance under acidic conditions. The second was the necessity of disposing the RO reject by sending it back to the gypsum stack. This would eventually cause constituents in the ponds to build up; this water would have to be treated before the site could be closed.

RO technology had never been used for any length of time to treat process water for direct discharge from a phosphate plant. Earlier trials, some at Piney Point, proved the efficiency of RO, but the long-term impact on operation, maintenance and materials of construction were unknown.

In spite of this, FDEP decided to phase in RO technology at Piney Point using mobile equipment to treat water 24/7. The feed water characterization and effluent quality specifications are listed in Table 1.

Table 1. Discharge effluent limitsConstituent Typical Feed Concentration Contract Specification Calcium, mg/l

551

Magnesium, mg/l

229

Sodium, mg/l

1,290

Potassium, mg/l

196

Aluminum, mg/l

8.4

Barium, mg/l

0.02

Bicarbonates, mg/l

0.78

Sulphates, mg/l

5,200

Chlorides, mg/l

100

Phosphates, mg/l

1,600

<0.5 mg/l as P

Nitrates, mg/l

0.26

Flourides, mg/l

150

2.8

6.0-8.5

Silica, mg/l

200

Iron, mg/l

5.6

Manganese, mg/l

2.9

TDS, mg/l

11,500

Turbidity, NTU

TSS, mg/l

Color, PCU

110

BOD, mg/l

TOC, mg/l

TKN, mg/l

650

NH3-N, mg/l

600

< 1 mg/l

The objectives of Phase I were to demonstrate the feasibility of using RO and to reliably produce an effluent of 90 gpm within the limits shown in Table 1. Phase I was to be completed within three months or after 5.2 million gal. of treated water was produced, whichever came first. Production would then be ramped up to 450 gpm during Phase II based on the knowledge and experience gained during Phase I. Phase III was optional, and was proposed as a low-cost extension.

To meet the fast-track timetable, the team chose a mobile system. Within five weeks of the contract award, the system was started up. Phase I equipment included dual-media roughing filters, followed by multimedia polishing filters and two-pass RO (Figure 2).

The system operates without adjusting the low pH conditions in the feedwater. The RO section has two passes; the first operates at a very low pH (less than 3.0), and the second pass runs at close to neutral pH. The first-pass operates at low pH to prevent scaling due to silica, fluoride, sulfates and phosphates in the feed water, and to improve the rejection of ammonia. The main reason the second-pass runs at close to neutral pH is to improve the rejection of fluorides, silica, phosphate, organics and other weakly ionized compounds. The contractor added mobile ion exchange polishers to ensure consistent effluent quality within the stringent contract specifications. The raw water characteristics versus the system effluent quality during Phase I is shown in Table 2.

Table 2: Mobile system cleans up pond water
Parameter	Units	2002 Process Water Values	System Effluent	Contract Specifications
Color	PCU	70	NA	NA
Fluoride	Mg/L	170	<2	<5
Calcium	MG/l	591	<0.5	NA
Phosphorous as P	Mg/l	1,600	<0.2	<0.5
Ammonia	MG/l	700	<1.0	<0.9
pH	Units	2.85	6.0-8.5	6.0-8.5
Silica	Mg/l	210	<0.5	NA
Sulfate	Mg/l	4,600	<1.0	NA
Conductivity	μs/cm	10,500	<25	<50
Total Nitrogen	Mg/l	730	<1.0	<2.0
TOC	Mg/l	72	<1.0	NA
Tubidity	NTU	15	<1.0	NA

Although the effluent met the contract specifications, fluctuations in the water temperature and the amount and intensity of daylight affected algae concentrations in the pond. Since no one thought algae could survive in the low-pH pond water, this problem was unexpected. A seepage ditch along the edge of the gypsum stack contained less algae since it was not subject to sunlight; it also had lower turbidity. Modifications were made to the feed piping and ditches so the system could take feed water from this ditch.

To reduce the rate at which submicron-sized algae fouled the RO membranes, the 5.0 micron prefilters were replaced with 1.0 micron cartridges. Another filtration system was then placed upstream to decrease the cost of changing these cartridge filters. Continuous microfiltration (CMF) technology from USFilter and ballasted flocculation clarification technology from Krüger Inc., Cary, N.C., were chosen and a three-week pilot study was conducted. The membranes, however, experienced some permanent fouling, even after a clean-in-place procedure was performed. Next, a ballasted flocculation clarifier was tested using hypochlorite at dosing rates between 2.5 mg/l and 4.0 mg/l to kill the algae, and bentonite to adsorb/destabilize the suspended particles. This process chemistry proved to be the solution for removing submicron algae.

Move to close
Steady state operations were achieved after nearly one full year of operation. During this time, two Actiflo clarifiers, also from Krüger, were installed, and the contractor patched, relined and re-bedded the media filters, and added additional RO capacity. From July 2003 to May 2004, the system produced an average of more than 490 gpm; once the improvements were made, the system was often able to process 550 gpm.

As the volume of free water in the ponds decreased during the latter months of Phase III, the RO concentrate began to impact the ponds. The pH, turbidity and algae increased, requiring daily feed water testing so adjustments could be made to the pretreatment upstream of the clarifiers.

In November 2003, the site moved from emergency-response mode to closure mode. Individual ponds or cells would be closed by lining them with a high-density polyethylene (HDPE) liner.

The RO system effluent met the contract specifications and demonstrated reliable performance, while the mobile equipment allowed flexibility for expansion. At Piney Point, RO technology to-date has treated more than 400 million gal. of process water, ultimately discharged to Tampa Bay. The ponds will be lined this year and the remaining free and interstitial water will be treated and disposed of using the RO system and the lime precipitation equipment.

As the lined ponds collect rainwater, the Piney Point site will help solve the Tampa Bay area’s water supply shortage. A site that once threatened disaster for a coastal estuary will be transformed into a reservoir that can hold 1.2 billion gal. of fresh water.

Bill Perpich, Jr., is a Tampa, Fla.-based industrial sales engineer for USFilter. He has 20 years of experience in the water treatment, groundwater remediation and solid/hazardous waste treatment industries. He has authored and presented several papers on groundwater treatment. E-mail him at [email protected].