By Pamela Wright, Trans Terra and Manuel Moreau, Eaglebrook, Inc.
After a risk assessment of municipal water supplies, the US Environmental Protection Agency (EPA) determined that New York City (NYC) was at risk and that a water filtration plant had to be built. This assessment was the outcome of regulations announced by EPA June 29, 1989, generally referred to as the Surface Water Treatment Rule (SWRT). These regulations specify criteria for determining filtration as a water treatment technique for public water systems.
The NYC watershed is approximately 2,000 square miles, and supplies NYC with up to 1.4 billion (US) gallons of water per day. The cost of such a plant was estimated to be (US) $6,000,000,000. Capital cost and financing fees were prohibitive, so an alternative was pursued. The end-result was an agreement termed Filtration Avoidance Determination (FAD). The FAD was extensive and included reservoir modeling, surface water evaluations, farm programs, purchase of land in the watershed, and the removal of Cryptosporidium, Giardia, and phosphorus from all wastewater plants in the NYC watershed.
Wastewater treatment plant upgrades
There are approximately 100 wastewater treatment plants (WWTPs) that discharge into tributaries that lead to NYC reservoirs. It was determined that the effluent from these WWTPs had to meet microfiltration quality. Phosphorus was also a concern since it could cause eutrophication of the reservoir.
Pilot studies in 1993 to determine which process train would be most effective in the removal of Cryptosporidium, Giardia, and phosphorus concluded that a sand filter followed by microfiltration or equivalent would be the process train.
In 1995, a pilot project was performed in Delhi, New York to determine if phosphorus levels of 0.10 mg/L could be obtained with filters of standard manufacture. Membranes used in the microfiltration units could be fouled by metal salts. The time between backwash cycles would subsequently shorten, thereby producing more backwash water. No manufacturer of standard filters would guarantee such low limits, so an innovative process was attempted in the pilot.
The Delhi pilot placed two Parkson DynaSandTM continuous backwash upflow filters in series. The first filter was two metres in depth and used coarse sand. The second filter was one metre in depth and used fine sand. A coagulant was added before the first stage filter to precipitate out the soluble phosphorus. Various coagulants were tried, but PACl (poly-aluminum chloride) was the most successful in the project. On average, the process reduced the phosphorus limit entering the filter at 2.33 mg/L to 0.05 mg/L. This was accomplished in conjunction with average TSS (Total Suspended Solids) loadings of 42.5 mg/L.
Later, a second pilot project was performed in Stamford, New York in 1996 to further evaluate the new technology's ability to remove protozoa. The test began in October during a warm Fall season. Two weeks into the pilot, cold weather set in. The process began to experience problems with turbidity. More PACl was added with little, if any, improvement. A Parkson representative with experience in water applications in Québec, suggested trying a coagulant that he found to work better in cold waters. This coagulant was PASS® (Poly-aluminum-silicate-sulfate), manufactured by Handy Chemical (now Eaglebrook, Inc.).
There were initial difficulties with the Poly-aluminum-silicate-sulfate, but all were attributed to the lack of familiarity with the product. PASS does not form a visible settling floc; it hydrolyzes immediately in water and is more effective in removing phosphorus with lower dosing rates. Turbidities in the Stamford wastewater soon began to return to the 0.15 NTU average experienced prior to the drop in water temperature.
A particle counter was used to optimize the dosing rate (see graph). Once optimized, the dosage required little adjustment. PASS was not as sensitive to alkalinity, pH, temperature or fluctuations in TSS and phosphorus. This characteristic was important since the water quality requirements of the NYC watershed required the wastewater operator to become (in essence) a water plant operator. The fewer the adjustments the better.
Pilot Process
In 1997, the DualSandTM process was used on a pilot project on water from an upland reservoir in Stamford, New York. The Village of Stamford needed to reactivate an old reservoir to supplement its wells. The process obtained 4.5 to 4.9 log removals of Cryptosporidium and Giardia in spike challenges performed by Clancy Environmental. The New York State Department of Health approved the technology for potable water treatment and PASS was the coagulant of choice.
Later in 1997, the DualSand system was introduced into a competitive pilot project with microfiltration. The EPA and NYC Department of Environmental Protection worked together to develop a new, more rigorous, spike challenge protocol. Twelve spike challenges would have to be performed to determine if DualSand filters were equivalent to microfiltration in the removal of protozoa. Both technologies obtained seven log removal during these spike challenges. The microfiltration unit was found to have a ruptured membrane, so the first three spike-challenge results for the microfiltration were not used in the comparison.
The DualSand filtration process was approved by the US EPA as an equivalent to microfiltration in the treatment of wastewater in the NYC watershed. In addition to exceptional protozoan removal, the pilot test was found to remove phosphorus to a level of 0.014 mg/L. The ability to remove phosphorus as well as protozoan parasites allowed the DualSand filtration process to be the entire treatment train. This pilot test used both PASS and PACl, with PASS emerging as the better performer.
The capital cost savings from not having to place a sand filter in front of the DualSand system, which is required with microfitration to remove phosphorus, is significant. In addition, the energy costs of this system are approximately eight to ten times lower than microfiltration.
The three pilot tests performed in the NYC watershed have documented the DualSand technology's ability to remove phosphorus and protozoa to exceptional levels. In a recent potable water pilot in Keeseville, NY, the DualSand process, using PASS, produced the following TOC removals:
Date DualSand Inf. DualSand
(Raw Water) Filtrate
10/5/99 4.9 mg/L 2.4 mg/L
10/8/99 4.7 mg/L 2.5 mg/L
10/14/99 3.3 mg/L 1.9 mg/L
10/14/99 3.3 mg/L 1.5 mg/L
10/14/99 3.6 mg/L <1.0 mg/L
The economics of using alum, PACl, or PASS are not immediately obvious. One needs to evaluate sludge production, the amount of turbidity and phosphorus to be removed, pH adjustment and actual chemical costs. To evaluate properly the cost difference between PACl and PASS, a price per poundcomparison is not valid. In warm weather the PASS and PACl dosing rates were about the same. In cold water the PASS dosing rate was about one third that of PACl. As PASS is more expensive than PACl, economy comes from lower dosing rates when treating cold water.