Open Access System for Information Sharing

Login Library


Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

Biokinetic Evaluation of Multiple Nitrifying Bacteria Growing Simultaneously by Use of 16S rRNA Gene Quantification

Biokinetic Evaluation of Multiple Nitrifying Bacteria Growing Simultaneously by Use of 16S rRNA Gene Quantification
Duong Xuan Nguyen
Date Issued
Nitrification is an important step in biological nitrogen removal process. Along with the development of advanced processes and reactor configuration for the improvement of nitrogen removal efficiency as well as the development of molecular techniques for investigating nitrifying, biokinetics still keeps an important place in these trends to bring the process advancements to application in practice. This is because biokinetic analysis provides relevant information for process design, optimization, upgrade of existing systems, process control, and education as well. Biokinetic analysis of nitrification process is usually performed through the analysis of oxygen (respirometry) and proton (titrimetry) consumptions. These approaches are popular and well studied because those measurements can be rapid and automatic. But, several assumptions must be made when applying these techniques for biokinetics. For example, microbial concentrations are assumed to be invariant during the courses of biochemical reactions. This is certainly not true in any nitrification process. Also, if the oxidation of ammonia, nitrite, and organic matters occur at the same time, it is impossible to measure oxygen consumptions separately for each reaction. As a result, several experiments must be separately performed to analyse biokinetics for those three oxidation reactions. This is also a problem because biokinetic parameter estimates from individual experiments may not describe correctly the growth rate of microorganisms and utilization rate of substrate in one reactor simultaneously. 16S rRNA genes quantification technique such as real-time QPCR has been proved to be a rapid and reliable tool for quantifying targeted and important microorganisms in wastewater systems. These 16S rRNA genes can be used as microbial concentration in biokinetic analysis of the nitrification process because one nitrifying bacterium has one 16S rRNA gene operon copy number per genome. This fact has been found and reported in literature. Compared to traditional biokinetics (respirometry and titrimetry) in nitrification, biokinetics with 16S rRNA genes of nitrifying bacteria has several advantages. First, assumption of unchanged microorganism concentrations is not necessary. Second, multiple biokinetics can be studied with 16S rRNA genes for targeted nitrifying bacteria which cannot be done by traditional biokinetic methods. Models that describe the microbial growth and substrate utilization based on three most popular microbial growth equations (i.e., Monod, Haldane, and Contois) were investigated in this thesis to find the most suitable one. Microbial interaction with Lotka-Volterra equation was also used to study the impacts among ammonia-oxidizing bacteria (AOB). The Akaike information criterion (AIC), which includes both least squares errors and number of parameters, was used as a criterion for choosing the best model. Three batch experiments were run to obtain the data for model selection and biokinetic analysis. Experiment B1 was prepared with artificial ammonia wastewater with 25%v/v seed from an inoculum system fed industrial wastewater
experiment B2 was run with artificial ammonia wastewater with 5%v/v seed from another inoculum fed by artificial ammonia wastewater
and experiment B3 was operated with steel processing wastewater with the same seed as in experiment B1. Five ammonia-oxidizing bacteria (AOB) groups (i.e., Nitrosomonas europaea-lineage, N. mobilis-lineage, N. nitrosa-lineage, N. cryotolerans-lineage, Nitrosospira spp.) and two nitrite-oxidizing bacteria (i.e., Nitrobacter spp. and Nitrospira spp.) were targeted with seven primer and probe sets. However, only two AOB subgroups (N. europaea-lineage and Nitrosospira spp.) and one NOB subgroup (Nitrobacter spp.) were detected. Therefore, the differential rate equations describing microbial growth rates of N. europaea-lineage, Nitrosospira spp., and Nitrobacter spp. and describing substrate (ammonia and nitrite) utilization rates were established. Model with Contois equation was chosen because its AIC was the smallest in all cases (i.e., 16.8 for artificial nitrogen wastewater and 7.4 for steel processing wastewater). Although Monod was the most popular equation for describing specific microbial growth rate, models with Contois equation showed better results. This is probably because Contois takes both microbial and substrate concentrations into consideration in its expression. However, the specific growth rate in Monod type equation depends only on substrate concentrations. This rate, as a result, ignores the microbial role in the biological systems, which might be problematic in Monod equation. Uncertainty analysis of biokinetic estimates were also performed to derive the standard deviation because the models can only be used if the parameter quality is determined. This analysis was implemented through the computation of Fisher information matrix (FIM) and then inverted it to derive the parameter estimation covariance matrix. FIM contains the important information about the sensitivity functions and if any function is linear with one or more functions in FIM, it will cause the singularity of FIM. Thus, the inverse of FIM to obtain covariance matrix is impossible. In most cases, the sensitivity function for Contois biokinetic parameter, KX, was found to be linear with the sensitivity function of maximum specific growth rate, m. The sensitivity function of KX was removed from FIM to avoid the singularity of FIM. As a result, the standard deviation for this parameter for this parameter was not estimated. The biokinetic parameters estimated from the batch experiments (i.e., B2 and B3) were used to model the system at the continuous operation. Experiment C1 was performed with artificial wastewater (influent NH4+-N of 100 mg/L and 600 mg/L, respectively, hydraulic retention time (HRT) of 10 days) whereas experiment C2 was run with steel processing wastewater (influent NH4+-N and SCN- of 500 mg/L, hydraulic retention time of 10 days and 7 days, respectively). The prediction for the case influent NH4+-N of 100 mg/L and HRT of 10 days was close to the experimental data. However, model predictions were not good in three remaining cases. This is probably because the simplified models did not consider all the phenomena in the dynamic conditions such as lag, shock loads, etc…The extended Kalman filter (EKF), then, was used to improve the model prediction which is usually described in process control. In four cases, the predictions with EKF were found to describe adequately the experimental data. The use of 16S rRNA genes was proved to be successful to evaluate biokinetic parameters in both artificial ammonia wastewater and steel processing wastewater. This is different from the traditional biokinetic approaches (i.e., respirometry and titrimetry tests) in the nitrification process. This idea may open a new direction towards biokinetics of nitrifying bacteria (grow simultaneously or separately) in this process.
Article Type
Files in This Item:
There are no files associated with this item.


  • mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Views & Downloads