Environ. Sci. Technol. 2010, 44, 8930–8935
Applying Ratio Control in a Continuous Granular Reactor to Achieve Full Nitritation under Stable Operating Conditions ´ REZ,* AND A L B E R T B A R T R O L ´I , J U L I O P E ´ N CARRERA JULIA Department of Chemical Engineering, Escola d’Enginyeria, ` Universitat Autonoma de Barcelona, Spain
Received June 9, 2010. Revised manuscript received September 28, 2010. Accepted October 12, 2010.
A ratio control strategy was implemented in a continuous granular airlift reactor to achieve and maintain 100% partial nitrification to nitrite (i.e., full nitritation). The control strategy was designed to maintain a constant ratio between the dissolved oxygen (DO) and the total ammonia nitrogen (TAN) concentrations (DO/TAN concentration ratio) in the reactor bulk liquid. The experimental results demonstrated the feasibility of full nitritation of a high-strength ammonium wastewater with a granular reactor operating in continuous mode, when implementing a suitable control strategy. The effect of the DO/TAN concentration ratio on partial nitrification was fast and reversible, upon switching from complete to partial nitrification, despite the presence of nitrite-oxidizing bacteria (NOB) in the granule. Even at a DO concentration as high as 7 mg of O2 L-1, full nitritation was obtained, decoupling the achievement of partial nitrification in continuous granular reactors from low DO concentrations. Inhibition of NOB by free ammonia was found to contribute poorly to the achievement of partial nitrification. An extremely high volumetric nitrogen loading rate was achieved (6.1 g of N L-1 day-1 at 30 °C), demonstrating that very compact reactors are applicable to nitrogen removal via nitrite.
1. Introduction Nitrogen removal via nitrite has recently gained increasing interest because of its associated economic savings when compared to classical nitrogen removal technologies, in which the ammonium is fully oxidized to nitrate (1). To achieve partial nitrification, several reactor configurations and setups have been tested, but single-reactor highactivity ammonia removal over nitrite (SHARON) (2) is probably the most implemented at full scale (3). In addition, after the discovery of anaerobic ammonia oxidation (anammox) (4) and its later application in a fullscale reactor (5), a two-step treatment consisting of partial nitrification to nitrite plus anammox to autotrophically denitrify to nitrogen gas has become probably the most advanced and economical alternative for nitrogen removal from a high-ammonium-strength stream (3, 5). Anammox has been implemented in several wastewater treatment plants (WWTPs) (5, 6), usually in combination with SHARON, to treat the reject water coming from the dewa* Correspondingauthorphone:+34-935812141;fax:+34-935812013; e-mail:
[email protected]. 8930
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 23, 2010
tering of digested sludge. In fact, recent studies have shown that the actual bottleneck in the overall capacity of N-removal advanced treatment systems is due to the limiting capacity of the first part of the treatment, that is, partial nitrification with the SHARON reactor (5). This limitation is due to the low biomass concentration that can be achieved in the SHARON treatment (7), because SHARON uses a reactor without retention of biomass to achieve and maintain partial nitrification in this type of reactor configuration. Consequently, to improve the capacity of the N-removal via nitrite, it is necessary to develop a robust technology that is able to partially nitrify at higher nitrogen loading rates (NLRs). Recently, a ratio control strategy was successfully applied to biofilm airlift reactors to obtain full nitritation (100% ammonia conversion to nitrite) under stable operating conditions, the so-called automatic control for partial nitrification to nitrite in biofilm reactors (ANFIBIO) (8). In this contribution, we show the results obtained with a granular airlift reactor in which ANFIBIO was used to obtain full nitritation of a high-strength ammonium concentration wastewater. In the study, the main goals were to explore (i) the range of operating conditions in which the ratio control strategy allows full nitritation to be achieved and maintained and (ii) the highest NLR that could be achieved with such a reactor configuration.
2. Experimental Section 2.1. Reactor Description. A biofilm airlift reactor with a working volume of 112 L was used in this study (see Figure S1 in the Supporting Information). The internal diameter of the downcomer was 26.7 cm. The riser had a height of 127 cm and an internal diameter of 18.5 cm. It was positioned at a distance of 10 cm from the bottom of the downcomer. Compressed air was supplied through an air diffuser placed at the bottom of the reactor. The dissolved oxygen (DO) concentration in the bulk liquid was measured by means of an online electrode, and a closed-loop control was implemented to maintain the DO concentration at different setpoint values. The temperature of the reactor was maintained at 30 °C using an electric heating system and a temperature controller. The pH control loop used solid NaHCO3 to maintain the pH at 8.2 and to supply the alkalinity required for nitrification. The total ammonia nitrogen (TAN ) N-NH4+ + N-NH3) concentration in the bulk liquid was determined with an online probe (NH4D sc probe with a Cartrical cartridge, Hach Lange, Du ¨ sseldorf, Germany) during the last 190 days of operation. Except for temperature, all sensors and actuators were connected to a SCADA (supervisory control and data acquisition) system. The biofilm airlift reactor was inoculated with activated sludge from a municipal WWTP (25 L from the main biological reactor of the Manresa municipal WWTP, Barcelona, Spain). The reactor was also loaded with 2.5 kg of activated carbon (AC) to induce granulation (9). The AC particles used had a mean diameter of 1.2 mm and a wet density of 1.25 g mL-1. After the development of the granular biomass (after about 100 days of operation), the added AC particles initially were removed. The main characteristics of the granular biomass at the beginning of the experimental period reported in this study were size, 0.9 mm; granule density, 38 gVS Lparticle-1 (VS: volatile solids); sludge volumetric index ratio at five and thirty minutes, SVI5/SVI30 ratio, 1.03; and settling velocity, 35 m h-1. 2.2. Wastewater. The biofilm airlift reactor was fed with synthetic wastewater with 3.833 g L-1 NH4Cl (1 g of N-NH4+ L-1) and the following compounds (in mg L-1): CH3COONa, 10.1021/es1019405
2010 American Chemical Society
Published on Web 11/04/2010
FIGURE 1. Block diagram of the automatic control for partial nitrification to nitrite in biofilm reactors (ANFIBIO (8)) used to ensure the required oxygen-limiting conditions in the biofilm to obtain and maintain continuous full nitritation under stable operating conditions. The ratio station is applied to the set points; therefore, once a TAN concentration set point ([TAN]SP) and a set-point ratio (RSP) are selected, the resulting DO concentration set point is imposed by the ratio control strategy as [DO]SP ) RSP[TAN]SP. 48.0; CaCl2 · 2H2O, 3.0; KH2PO4, 13.0; NaCl, 9.0; MgCl2 · 7H2O, 6.0; FeSO4 · 7H2O, 0.13; MnSO4 · H2O, 0.1; ZnSO4 · 7H2O, 0.13; CuSO4 · 5H2O, 0.07; and H3BO3, 0.007. This solution was prepared with tap water in a 2 m3 storage tank, the variation in the influent TAN concentration was around 10%, and therefore the TAN concentration was periodically measured off-line. 2.3. Analytical Methods. Regular sampling of the bulk liquid of the pilot plant was carried out to determine the TAN, total nitrite nitrogen (TNN ) N-NO2- + N-HNO2), and nitrate concentrations through off-line analysis. The TAN concentration in off-line liquid samples withdrawn from the bulk liquid was analyzed by means of a continuous flow analyzer. The TNN and nitrate concentrations were measured by ion chromatography (ICS-2000 Integrated Reagent-Free IC System, DIONEX). Volatile solids (VS), total solids (TS), and sludge volumetric index (SVI) were determined according to APHA standards (10). The granular biomass was characterized throughout the whole experimental period (ca. 300 days) in terms of size, aspect, shape, granule density, and settling velocity (see the Supporting Information for further details). 2.4. FISH Analysis. The fluorescence in situ hybridization (FISH) technique coupled with confocal laser scanning microscopy (CLSM) was used to determine the fractions of ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB) in the granules. A detailed description of the applied methodology can be found in the Supporting Information. 2.5. Control Strategy. The control strategy used was the key engineering development for achieving and maintaining stable partial nitrification in the biofilm airlift reactor. The control strategy developed was based on previous research (11, 12), in which oxygen-limiting conditions were found to be the main factor causing partial nitrification to nitrite in biofilm reactors. The control strategy used is an advanced control technique based on a particular application of feed-forward control, called ratio control. The type of ratio station implemented is applied to the set points (see ref 13 for an in-depth description of the advantages of this type of ratio control). Therefore, two different closed loops are implemented: (i) one to maintain the TAN concentration in the bulk liquid (i.e., the reactor effluent, considering a well-mixed liquid phase in the reactor) and (ii) a second one to control the DO concentration in the bulk liquid. To illustrate the control strategy, a conventional block diagram is presented in Figure 1. The variables measured for the two closed loops were the TAN concentration and the DO concentration, whereas the two manipulated variables were the wastewater inflow rate fed to the reactor and the air flow rate, respectively (see Figure 1 for details). With this control configuration, the set
points of both closed control loops ([TAN]SP and [DO]SP) do not depend on either of the measured variables. Both control loops act independently varying the corresponding manipulated variables (wastewater inflow rate and air flow rate, respectively) on demand to keep the corresponding set points. This ratio station is more efficient during transient states than other (conventional) ratio control configurations (13), and this is one of the reasons why we selected this option. The types of controllers implemented were an on-off controller for the TAN concentration control loop and a proportional-integral (PI) controller for the DO control loop. The feedback control loop maintaining the TAN concentration in the bulk liquid allows for a maximization of the treatment capacity at any time, because the loading rate will be as high as possible during the continuous operation of the reactor. With the applied control strategy, the concentrations of the two main substrates of AOB (i.e., ammonium and oxygen) in the bulk liquid are therefore under control. Following the considerations described by Harremo¨es (14), oxygen-limiting conditions in nitrifying biofilm reactors will appear when [DO]
