@article{CheenakulaGriebelMontagetal.2023,
  author    = {Cheenakula, Dheeraja and Griebel, Kai and Montag, David and Gr{\"o}mping, Markus},
  title     = {Concept development of a mainstream deammonification and comparison with conventional process in terms of energy, performance and economical construction perspectives},
  series = {Frontiers in Microbiology},
  volume    = {14},
  journal   = {Frontiers in Microbiology},
  number    = {11155235},
  editor    = {Huang, Xiaowu},
  publisher = {Frontiers},
  issn      = {1664-302X},
  doi       = {10.3389/fmicb.2023.1155235},
  pages     = {1 -- 15},
  year      = {2023},
  abstract  = {Deammonification for nitrogen removal in municipal wastewater in temperate and cold climate zones is currently limited to the side stream of municipal wastewater treatment plants (MWWTP). This study developed a conceptual model of a mainstream deammonification plant, designed for 30,000 P.E., considering possible solutions corresponding to the challenging mainstream conditions in Germany. In addition, the energy-saving potential, nitrogen elimination performance and construction-related costs of mainstream deammonification were compared to a conventional plant model, having a single-stage activated sludge process with upstream denitrification. The results revealed that an additional treatment step by combining chemical precipitation and ultra-fine screening is advantageous prior the mainstream deammonification. Hereby chemical oxygen demand (COD) can be reduced by 80\% so that the COD:N ratio can be reduced from 12 to 2.5. Laboratory experiments testing mainstream conditions of temperature (8-20°C), pH (6-9) and COD:N ratio (1-6) showed an achievable volumetric nitrogen removal rate (VNRR) of at least 50 gN/(m3∙d) for various deammonifying sludges from side stream deammonification systems in the state of North Rhine-Westphalia, Germany, where m3 denotes reactor volume. Assuming a retained Norganic content of 0.0035 kgNorg./(P.E.∙d) from the daily loads of N at carbon removal stage and a VNRR of 50 gN/(m3∙d) under mainstream conditions, a resident-specific reactor volume of 0.115 m3/(P.E.) is required for mainstream deammonification. This is in the same order of magnitude as the conventional activated sludge process, i.e., 0.173 m3/(P.E.) for an MWWTP of size class of 4. The conventional plant model yielded a total specific electricity demand of 35 kWh/(P.E.∙a) for the operation of the whole MWWTP and an energy recovery potential of 15.8 kWh/(P.E.∙a) through anaerobic digestion. In contrast, the developed mainstream deammonification model plant would require only a 21.5 kWh/(P.E.∙a) energy demand and result in 24 kWh/(P.E.∙a) energy recovery potential, enabling the mainstream deammonification model plant to be self-sufficient. The retrofitting costs for the implementation of mainstream deammonification in existing conventional MWWTPs are nearly negligible as the existing units like activated sludge reactors, aerators and monitoring technology are reusable. However, the mainstream deammonification must meet the performance requirement of VNRR of about 50 gN/(m3∙d) in this case.},
  language  = {en}
}
@phdthesis{Bung2023,
  author    = {Bung, Daniel Bernhard},
  title     = {Imaging techniques for investigation of free-surface flows in hydraulic laboratories},
  doi       = {10.25926/BUW/0-172},
  pages     = {XXIII, 218 Seiten},
  year      = {2023},
  abstract  = {This thesis aims at the presentation and discussion of well-accepted and new imaging techniques applied to different types of flow in common hydraulic engineering environments. All studies are conducted in laboratory conditions and focus on flow depth and velocity measurements. Investigated flows cover a wide range of complexity, e.g. propagation of waves, dam-break flows, slightly and fully aerated spillway flows as well as highly turbulent hydraulic jumps. Newimagingmethods are compared to different types of sensorswhich are frequently employed in contemporary laboratory studies. This classical instrumentation as well as the general concept of hydraulic modeling is introduced to give an overview on experimental methods. Flow depths are commonly measured by means of ultrasonic sensors, also known as acoustic displacement sensors. These sensors may provide accurate data with high sample rates in case of simple flow conditions, e.g. low-turbulent clear water flows. However, with increasing turbulence, higher uncertainty must be considered. Moreover, ultrasonic sensors can provide point data only, while the relatively large acoustic beam footprint may lead to another source of uncertainty in case of relatively short, highly turbulent surface fluctuations (ripples) or free-surface air-water flows. Analysis of turbulent length and time scales of surface fluctuations from point measurements is also difficult. Imaging techniques with different dimensionality, however, may close this gap. It is shown in this thesis that edge detection methods (known from computer vision) may be used for two-dimensional free-surface extraction (i.e. from images taken through transparant sidewalls in laboratory flumes). Another opportunity in hydraulic laboratory studies comes with the application of stereo vision. Low-cost RGB-D sensors can be used to gather instantaneous, three-dimensional free-surface elevations, even in flows with very high complexity (e.g. aerated hydraulic jumps). It will be shown that the uncertainty of these methods is of similar order as for classical instruments. Particle Image Velocimetry (PIV) is a well-accepted and widespread imaging technique for velocity determination in laboratory conditions. In combination with high-speed cameras, PIV can give time-resolved velocity fields in 2D/3D or even as volumetric flow fields. PIV is based on a cross-correlation technique applied to small subimages of seeded flows. The minimum size of these subimages defines the maximum spatial resolution of resulting velocity fields. A derivative of PIV for aerated flows is also available, i.e. the so-called Bubble Image Velocimetry (BIV). This thesis emphasizes the capacities and limitations of both methods, using relatively simple setups with halogen and LED illuminations. It will be demonstrated that PIV/BIV images may also be processed by means of Optical Flow (OF) techniques. OF is another method originating from the computer vision discipline, based on the assumption of image brightness conservation within a sequence of images. The Horn-Schunck approach, which has been first employed to hydraulic engineering problems in the studies presented herein, yields dense velocity fields, i.e. pixelwise velocity data. As discussed hereinafter, the accuracy of OF competes well with PIV for clear-water flows and even improves results (compared to BIV) for aerated flow conditions. In order to independently benchmark the OF approach, synthetic images with defined turbulence intensitiy are used. Computer vision offers new opportunities that may help to improve the understanding of fluid mechanics and fluid-structure interactions in laboratory investigations. In prototype environments, it can be employed for obstacle detection (e.g. identification of potential fish migration corridors) and recognition (e.g. fish species for monitoring in a fishway) or surface reconstruction (e.g. inspection of hydraulic structures). It can thus be expected that applications to hydraulic engineering problems will develop rapidly in near future. Current methods have not been developed for fluids in motion. Systematic future developments are needed to improve the results in such difficult conditions.},
  language  = {en}
}