Gotham JA Miele, Sr, PE

Water conservation cleans Long Island Sound, RL Swanson, DJ Tonjes

Marine vessels serving New York City, W Goyzueta, J Chen, K Byrnes, R Ferro

Line stops avoid bypass in pumping station, F Gallo

Pilot biological nutrient removal, B Bodniewicz, K Mahoney

Enhanced beach protection — 2000, FJ Oliveri, F Loncar, M Ellis

Telemetering in New York, S Rozelman, S Aziz

Job order contracting, MP Quinn, P Schrayer

Operational benefits of celebrating Water Week, RE Adamski, H Einsohn, M Keating, A Lamarche, B Olivieri

CSO signage: expanded notification, S Rozelman, P Lutz, F Loncar

Brooklyn student wins water prize

Executive director's message, P Cerro-Reehil

People and places

Summer 2001 — Vol. 31, No. 2

by Bohdan Bodniewicz, PE and Keith Mahoney, PE

New York City owns and operates fourteen water pollution control plants (WPCPs). Together, they treat more than one billion gallons a day of raw wastewater from all five boroughs. The treated effluent is discharged into the surrounding waters, including Long Island Sound and Jamaica Bay.

Outside view of the PO-55 pilot units

The limited flushing capacity of these waters coupled with excessive nutrient loading promotes prodigious algae growth. As the algae die, they settle to the bottom and decompose. The decomposition of these algae exerts a high oxygen demand, and the result is low dissolved oxygen (DO) concentrations in the bottom waters of western Long Island Sound and Jamaica Bay, a condition known as hypoxia.

The excessive discharge (both point and nonpoint) of nitrogen to the western Long Island Sound has been identified as the single most significant contributor to the hypoxic conditions. Extensive water quality testing and computer modeling of the Sound and East River have been conducted as part of the Long Island Sound Study (LISS). These Nitrogen concerns led to the inclusion of nitrogen control requirements in the New York City SPDES permit.

The SPDES nitrogen control requirements specify several actions that the NYCDEP must implement to diminish the nitrogen discharges. They include:

• Operational modifications to enhance nitrogen removals at WPCPs
• Increased sampling and analysis at the plants
• Preparation of a Nitrogen Control Action Plan
• Design, construction, and operation of both biological and physical-chemical pilot units to evaluate nitrogen removal capabilities of different technologies
• Preparation of a Nitrogen Control Feasibility Plan.

PO-55 pilot facility

NYCDEP consulted Metcalf & Eddy of NY, Inc., to design, construct, and operate a number of nitrogen removal pilot units. The result was the design and construction of a $20 million BNR pilot-testing facility located on the grounds at the 26th Ward WPCP in Brooklyn. The facility was commissioned in September 1998.

The PO-55 pilot facility comprises various BNR processes:

• Suspended growth step-feed BNR pilot units
• Suspended/attached growth biological system
• Biologically activated filters
• Suspended growth nitrification reactor
• Two physical-chemical nitrogen removal units.

The focus of the pilot study has been optimization of the step-feed BNR system, a process that was initially developed by the NYCDEP as part of the Tallman Island WPCP Demonstration Project. The step-feed BNR process proved to be a cost-effective way to remove nitrogen from the wastewater with only minor modifications to the WPCPs.

Step-feed BNR pilot units

Pilot Units 1 through 3 operate in the Step Feed BNR mode. In this mode, the 18-gal/min of primary effluent from the 26th Ward WPCP is sent to each pilot. The primary effluent is then distributed to all four passes of the pilot units as follows:
 
- 10% (A Pass)
- 40% (B Pass)
- 30% (C Pass)
- 20% (D Pass).

Effluent nitrogen data, Sept-Dec 2000

In addition, 800 mL/min of pre-settled centrate is fed to the "A" pass of pilot units 1 and 2, and it accounts for an additional 40% increase in the total nitrogen loading to these pilot units. (Centrate is a side stream from the centrifuges which are used to dewater anaerobically digested sludge before land application or pellitization. The centrate stream contains a high concentration of ammonia and is a significant contributor to the total nitrogen load, particularly in WPCPs that dewater sludge from other treatment facilities.) The first third of each pass is anoxic, and the remainder of the pass is operated at dissolved oxygen (DO) concentrations ranging between 2 and 6 mg/L.

The Solids Retention Time (SRT) of this pilot is maintained between 8 and 12 days depending on seasonal requirements. Supplemental carbon and alkalinity addition are used to optimize the step-feed BNR process; this has resulted in TN concentrations between 4 to 6 mg/L for all three units. See graphs.

Initially, significant problems arose with accurate DO control in all three pilot units and with reliable addition of centrate and methanol. Piping modifications and the fine tuning of existing instrumentation, however, resulted in steady-state operation.

Pilot 2 evaluated a hybrid attached/suspended growth system using fixed media, but this showed no improvements in the nitrification process although it did achieve some simultaneous denitrification. The froth resulting from the BNR process is currently controlled through the addition of a timed spray water header system that breaks up the foam. The foam is then manually removed from the tanks.

Future pilot testing will focus on the following:

• Implementing different automated froth control systems than can be replicated in full-scale facilities
• Evaluating alternative carbon sources
• Further optimizing the step-feed BNR process
• Determining the effects of cold weather on the pilot units' performance
• Determining the effects of storm water flow on the BNR process.

Pilot units 1 thorough 3

Biological filters

The PO-55 Pilot Study also evaluated attached-growth biological filters. These filters were specified in the SPDES Nitrogen Control Requirements and seemed to be the best technology to achieve very high TN removals.

Tertiary filters

Components and results

The nitrification filters consisted of a nitrifying trickling filter and a nitrifying biofilter which were fed final effluent from the 26th Ward WPCP. The average influent NH₃-N concentration to the nitrifying units during the experimental period was 14.9 mg/L, and the effluent ammonia was 6 to 7 mg/L with no alkalinity addition. Both tertiary nitrifying filters provided reliable and relatively maintenance-free operation.

The step-feed BNR process, however, is much more cost-effective for New York City and can achieve almost comparable nitrogen removals. Additional testing will be done on the biofilters, but the primary focus of the pilots will be to optimize the step-feed BNR units to achieve the higher TN removals.

Alternative technologies

In addition to the polishing nitrifying filters, a denitrifying biofilter and an upflow fluidized bed were evaluated on the combined effluent from pilots 1, 2 and 3. The average influent NO_x-N concentration to both biofilters was about 7.7 mg/L, and the average effluent from both units was 5 to 6 mg/L. The poor denitrification performance was directly attributable to equipment breakdown, instrumentation failures, and unreliable methanol feed. Both denitrifying biofilters also had relatively high effluent concentrations of total suspended solids (TSS) of 15 to 20 mg/L.

Biological centrate treatment

The PO-55 pilot study also evaluated the feasibility of removing nitrogen from the dewatering facility's centrate stream. The centrate stream is of particular concern at some plants because it contributes up to 40% of the nitrogen load and only accounts for 1% of the influent flow. Initially, there were concerns that the high ammonia concentrations would be toxic to the autotrophic bacteria. This did not prove to be the case, however.

Pilot 4 was dedicated to biological nitrification of the centrate stream and was operated in a plug-flow mode with a sludge age of about 15 days. The system was able to nitrify about 40% of the influent ammonia before alkalinity became a limiting factor. It could, according to batch tests, achieve basically 100% nitrification with a supplemental alkalinity addition. The pilot developed a white surfactant-like foam which was probably the result of a residual polymer in the centrate stream.

Physical-chemical centrate treatment

The pilot systems focused on removing nitrogen from the centrate stream using both hot air stripping and steam stripping:

1. The centrate was pretreated by a coarse ¼-inch
strainer . . . 
2. followed by settling through an inclined plate
settler . . . 
3. and then additional straining through a 3/64-inch
strainer to prevent media fouling of both stripping units.
4. The centrate was then fed countercurrent to
either the low-pressure steam or hot air.

Hot air stripper

In the hot air stripper, the centrate pH was adjusted with caustic before being preheated in a heat exchanger. The hot air stripper was equipped with random plastic packing. The ammonia that was stripped from solution was sent to a sulfuric acid scrubber in which the ammonia was removed as ammonium sulfate. Operations of the hot air stripper proved to be very chemical-intensive, requiring significant amounts of caustic soda and sulfuric acid.

The steam stripper was equipped with stainless steel structural packing and did not require any pH adjustment as was originally anticipated. The pH of the centrate entering the unit was about 7.5 and leaving the unit was above 9. This was most likely due to the carbonic acid being stripped from solution, thus increasing the pH and converting the ammonium to ammonia.

Steam stripper

The high pH coupled with the high temperatures enabled the steam stripper to remove 33% of the ammonia at a steam-to-liquid ratio of 100:1 and enabled up to 88% ammonia removal at a steam-to-liquid ratio of 300:1. The off-gas from this process was sent to a condenser and recovered as a concentrated aqueous ammonia solution.

Primary sludge fermentation

One of the most expensive aspects of biological nitrogen removal for New York City is the need for supplemental carbon to fully denitrify the NO_x-N produced in the nitrification process. Therefore, it was decided to produce supplemental carbon on site through the fermentation of primary sludge.

Approximately 80 gal/min of degritted primary sludge, with a TSS concentration of less than 1%, was pumped to two 10,000-gal fermenters. This low solids-loading rate was insufficient to produce an adequate concentration of volatile fatty acids (VFAs), however, requiring modifications to the existing system.

Primary sludge fermenters

The units now operate in series:

• The first unit is used to gravity-thicken the
sludge.
• The second reactor is used to ferment the
thickened sludge.

In addition, to hamper the formation of methane, the flow rate to the units was increased to maximize the solids loading and decrease the hydraulic retention time. The process has been producing about 1000 mg/L of VFAs, but this may be insufficient to denitrify all of the NO_x-N in the wastewater.

Future testing will involve enhancing the fermentation process by heating the system along with sending the combined primary and secondary sludge to a dual-phase digestion process to produce VFAs.

Conclusions

• Data from initial testing indicate that step-feed
BNR is technically feasible and would maximize the use of existing secondary treatment infrastructure.
• It can also be implemented at a lower capital
cost relative to other options while having the potential of producing an effluent TN ranging between 5 and 8 mg/L.

Continued research will focus on optimizing the step-feed BNR process, froth control, alternate carbon sources, chemical requirements, and instrumentation evaluation. The findings from these pilot studies will continue to be provided to the design engineers for incorporation into future plant upgrades.
____________
Bohdan Bodniewicz, PE is with Metcalf & Eddy, and Keith Mahoney, PE is with the NYCDEP. Phone 718-595-4966 for Mahoney.