Henri Spanjers

Modelling the production and degradation of soluble microbial products (SMP) in membrane bioreactors (MBR)

MBR biochemical conditions have an effect on membrane fouling and SMP have been attributed to be the main MBR foulant. Thus, predicting the SMP concentration is essential for understanding and controlling MBR fouling. However, existing SMP models are mostly too complex and over-parameterized, resulting in inadequate or absent parameter estimation and validation. This study extends the existing activated sludge model No. 2d (ASM2d) to ASM2dSMP with introduction of only 4 additional SMP-related parameters. Dynamic batch experimental results were used for SMP parameter estimation leading to reasonable parameter confidence intervals. Finally, the ASM2dSMP model was used to predict the impact of operational parameters on SMP concentration. It would found that solid retention time (SRT) is the key parameter controlling the SMP concentration. A lower SRT increased the utilization associated products (UAP) concentration, but decreased the biomass associated products (BAP) concentration and vice versa. A SRT resulting in minimum total SMP concentration can be predicted, and is found to be a relatively low value in the MBR. If MBRs operate under dynamic conditions and biological nutrient removal is required, a moderate SRT condition should be applied.

Direct parameter extraction from respirograms for wastewater and biomass characterization

A method is proposed and described that calculates model parameters and components through a series of numerical transformations on respirometric data. The method is direct because it does not use a model optimisation technique. This direct respirogram evaluation (DRE) method separates the respirogram into individual components, and calculates maximum respiration rate, saturation coefficient and yield factor for the components. An advantage of the DRE method is the simplicity; there are no complications involving modelling, optimisation, and model identifiability. The calculations are model independent, even though the principle implies the Monod saturation term. The most important limitation is that only certain types of respirograms can be identified, that is: individual components must be dominant at a critical stage of their oxidation. The DRE method is tested on simulated respirograms and a number of real respirograms of various characteristics, and the results from the latter are compared to those from a parameter optimisation method. It is concluded that the two methods are not always in agreement, especially when a component is present in a minor fraction.

Biodegradability and denitrification potential of settleable chemical oxygen demand in domestic wastewater.

The effect of settling on mass balance and biodegradation characteristics of domestic wastewater and on denitrification potential was studied primarily using model calibration and evaluation of oxygen uptake rate profiles. Raw domestic wastewater was settled for a period of 30 minutes and a period of 2 hours to assess the effect of primary settling on wastewater characterization and composition. Mass balances in the system were made to evaluate the effect of primary settling on major parameters. Primary settling of the selected raw wastewater for 2 hours resulted in the removal of 32% chemical oxygen demand (COD), 9% total Kjeldahl nitrogen, 9% total phosphorus, and 47% total suspended solids. Respirometric analysis identified COD removed by settling as a new COD fraction, namely settleable slowly biodegradable COD (X(ss)), characterized by a hydrolysis rate of 1.0 day(-1) and a hydrolysis half-saturation coefficient of 0.08. A model simulation to test the fate and availability of suspended (X(s)) and settleable (X(ss)) COD fractions as carbon sources for denitrification showed that both particulate COD components were effectively removed aerobically at sludge ages higher than 1.5 to 2.0 days. Under anoxic conditions, the biodegradation of both COD fractions was reduced, especially below an anoxic sludge retention time of 3.0 days. Consequently, modeling results revealed that the settleable COD removed by primary settling could represent up to approximately 40% of the total denitrification potential of the system, depending on the specific configuration selected for the nitrogen removal process. This way, the results showed the significant effect of primary settling on denitrification, indicating that the settleable COD fraction could contribute an additional carbon source in systems where the denitrification potential associated with the influent becomes rate-limiting for the denitrification efficiency.

Assessment of a two-step nitrification model for activated sludge

The objective of developing empirical models with well determined parameters is to obtain a better understanding of the nitrification of activated sludge. The D-optimal experimental design criterion and nonlinear regression analysis were used to obtain precise parameter estimates. Our experiment consisted of a series of 14 batch runs where activated sludge was spiked with ammonium chloride and sodium nitrite individually and in combination. Specific respiration rate, the ratio of respiration rate to mixed liquor suspended solids concentration, was used as the model response variable. Empirical models for specific respiration rate were specified as piecewise, Monod type functions of ammonium and nitrite concentration. Statistical methods were used to evaluate the effects of nonlinearity and parameter correlation on the models. The effect of experimental design on model specification and on parameter estimation is discussed.

Feedforward control of nitrification by manipulating the aerobic volume in activated sludge plants

A procedure was developed to manipulate the aerobic nitrification volume by means of model based respirometry. The objective was to oxidise all ammonia nitrogen. The required aerobic volume was calculated from the influent flow rate, the influent concentration nitrifiable nitrogen and maximum ammonia oxidation rate which were obtained from respirometric batch-experiments using model calibration and state estimation techniques. Automation of the procedure resulted in the development of the RESCUE (REspirometric activated Sludge and wastewater Characterisation Unit). The procedure was tested for a pilot scale plug-flow activated sludge plant with satisfactory results.

Impact of temperature on feed-flow characteristics and filtration performance of an upflow anaerobic sludge blanket coupled ultrafiltration membrane treating municipal wastewater.

The objective of this study was to assess the operational feasibility of an anaerobic membrane bioreactor (AnMBR), consisting of an upflow anaerobic sludge blanket (UASB) reactor coupled to an ultrafiltration membrane unit, at two operational temperatures (25°C and 15°C) for the treatment of municipal wastewater. The results showed that membrane fouling at 15°C was more severe than that at 25°C. Higher chemical oxygen demand (COD) and soluble microbial products (SMP) concentrations, lower mean particle diameter, and higher turbidity in the UASB effluent at lower temperature aggravated membrane fouling compared to the 25°C operation. However, the overall AnMBR treatment performance was not significantly affected by temperature, which was attributed to the physical membrane barrier. Cake resistance was found responsible for over 40% of the total fouling in both cases. However, an increase was observed in the contribution of pore blocking resistance at 15°C related to the larger amount of fine particles in the UASB effluent compared to 25°C. Based on the overall results, it is concluded that an AnMBR, consisting of a UASB coupled membrane unit, is not found technically feasible for the treatment of municipal wastewater at 15°C, considering the rapid deterioration of the filtration performance.

On-line meter for respiration rate and short-term biochemical oxygen demand in the control of the activated sludge process.

ABSTRACT A meter, based on a novel principle, is described for monitoring of the respiration rate of activated sludge and, as a derivative, the short-term biochemical oxygen demand (BODST). Operational implications and reliability are discussed. A method is presented for measuring four respiration rates of the same sludge under different conditions. These rates are: the endogenous, the instantaneous, the actual and the maximum respiration rate. Theoretical and experimental conditions for the measurement of the different rates are discussed. It is shown that the BODST of both influent and effluent can be calculated from three of the four respiration rates. The method is applied to a completely mixed continuous-flow activated sludge pilot plant. Results from two experimental set-ups are shown and discussed. There is some evidence that the maximum respiration rate is dependent on the loading. It is concluded that the measurements provide relevant information on the activated sludge process and in addition that the influent and effluent BODST can also be calculated. The BODST of the examined wastewater appears to be predominantly caused by ammonium being oxidized by nitrifiers. The measurements can be used to control activated sludge plants with respect to sludge loading, dissolved oxygen concentration and sludge balance. Control schemes are suggested.

Determining short-term biochemical oxygen demand and respiration rate in an aeration tank by using respirometry and estimation

Abstract For good control of the activated sludge process, accurate measurement of the short-term biochemical oxygen demand (BOD st ) in the aeration tank ( S a ) and the actual respiration rate of the activated sludge ( r act ) is required. A previously described measurement technique for these variables requires the measurement of three types of respiration rate of which one rate, the actual respiration rate, involves continuous addition of influent sample to a respiration chamber. The technique is only applicable to a plant with one single aeration tank or to the first compartment of a plug flow reactor. In this paper, a method is proposed for the estimation of S a and r act in a completely mixed aeration tank, no matter whether this is a single reactor or a compartment of a plug flow reactor. In contrast to the other technique, this method does not involve addition of influent sample to the respiration chamber. Instead, the transient respiration rate is measured during two modes of operation which are alternately executed. In one mode, sludge from the aeration tank flows through the respiration chamber, whereas in the other mode, sludge having the endogenous rate is directed through the chamber. The S a is obtained by integrating the mass balance equations for each cycle of two different modes. The measured respiration rate and the calculated BOD st during each cycle of two modes are used to find the kinetic relationship in order to calculate r act . The proposed method has been verified using simulated and experimental data. The estimated S a and r act from experimental data are in agreement with the diurnal variation typical for the wastewater used. The pattern of r act is in agreement with the one obtained by the previously described method. There is, however, a significant difference between the average values of the two methods. A possible explanation for this bias is given. Compared to the previously described method, the proposed method is simpler to use in practice.

Modelling of the dissolved oxygen probe response in the improvement of the performance of a continuous respiration meter.

Abstract In the present technique for the measurement of the respiration rate, the DO-probe is repeatedly subjected to a step change of the oxygen concentration with a time interval of typically 20–30 s. The end value of each response is used to calculate the respiration rate of activated sludge. However, those end values can be inaccurate because of probe dynamics or probe response time. This paper describes an improved method to measure the dissolved oxygen (DO) concentration with one and the same probe at the inlet and outlet of a respiration chamber. The real DO concentration at the end of each response can be estimated by fitting the DO measurements to a first-order response model of the probe. Furthermore, at each step response the time constant of the probe response can be estimated which provides a continuous diagnosis of the probe condition. The proposed method provides a reliable estimate of the real DO concentration and, as a result, the reliability of the calculated respiration rate is improved. The first-order probe response constant is a useful indicator for fouling of the probe membrane.

Comparison of the modeling approach between membrane bioreactor and conventional activated sludge processes.

Activated sludge models (ASM) have been developed and largely applied in conventional activated sludge (CAS) systems. The applicability of ASM to model membrane bioreactors (MBR) and the differences in modeling approaches have not been studied in detail. A laboratory-scale MBR was modeled using ASM2d. It was found that the ASM2d model structure can still be used for MBR modeling. There are significant differences related to ASM modeling. First, a lower maximum specific growth rate for MBR nitrifiers was estimated. Independent experiments demonstrated that this might be attributed to the inhibition effect of soluble microbial products (SMP) at elevated concentration. Second, a greater biomass affinity to oxygen and ammonium was found, which was probably related to smaller MBR sludge flocs. Finally, the membrane throughput during membrane backwashing/relaxation can be normalized and the membrane can be modeled as a continuous flow-through point separator. This simplicity has only a minor effect on ASM simulation results; however, it significantly improved simulation speed.