Techno-Economic Analysis of Constant-Flow Woodchip Bioreactors

HighlightsTechno-economic analysis was performed for multiple scales of bioreactors operated under a variety of conditions.The unit cost decreased as the bioreactor size increased.The unit cost increased in bioreactors with longer HRTs and bypass flow due to reduced treatment capacity.One large bioreactor was more cost-effective than multiple smaller bioreactors.Abstract. Woodchip denitrification bioreactors are a relatively new, edge-of-field technology used to reduce nitrate-nitrogen (NO3-N) from subsurface tile drainage. The removal rate of nitrate is influenced by many factors, including temperature, dissolved oxygen, and hydraulic residence time (HRT). The objective of this study was to conduct a techno-economic analysis (TEA) for four scales of woodchip denitrification bioreactors operating at three HRTs (2, 8, and 16 h), designed with bypass flow or with a low probability of bypass flow, to determine the cost to remove 1 kg of NO3-N at each bioreactor scale and at each HRT. Several assumptions were made: the flow rate required to achieve a 2 h HRT on a per m3 basis could be achieved at all scales, the same mass removal of NO3-N was achieved on a per cubic meter basis, and the 2 h HRT did not have any bypass flow at each scale. With these assumptions, the lowest unit cost was observed for the large-scale bioreactor sized to have a low probability of bypass flow at 16 h HRT, with a resulting cost of $0.74 kg-1 NO3-N removed. The highest unit cost was observed for the pilot-scale bioreactor designed with bypass flow to achieve a 16 h HRT at a cost of $60.13 kg-1 NO3-N removed. At longer HRTs with bypass flow, a greater percent removal of nitrate has been observed with a lower mass removal rate. By having a low probability of bypass flow in the design, a higher mass removal and percent removal of nitrate were observed, leading to the above results. Contrasting this trend, the total and annual costs were highest for the large-scale bioreactor and lowest for the pilot-scale bioreactor. However, it was determined that 783%, 280%, and 54% increases in total cost for the pilot-, small-, and medium-scale bioreactors would be incurred to implement the number of bioreactors (66, 24, and 4, respectively) required to treat the same volume of flow as one large bioreactor. These results can be used to inform future design decisions and inform stakeholders of the approximate unit cost of installing a denitrifying woodchip bioreactor over a range of expected field conditions. While a larger bioreactor with a low probability of bypass flow may represent a more cost-effective investment, the potential for unintended, negative byproducts needs to be considered in the design. Keywords: Denitrification, Nitrate, Tile drainage, Water quality, Woodchip bioreactor.

Reproduction of Large-Scale Bioreactor Conditions on Microfluidic Chips

Microbial cells in industrial large-scale bioreactors are exposed to fluctuating conditions, e.g., nutrient concentration, dissolved oxygen, temperature, and pH. These inhomogeneities can influence the cell physiology and metabolism, e.g., decelerate cell growth and product formation. Microfluidic systems offer new opportunities to study such effects in great detail by examining responses to varying environmental conditions at single-cell level. However, the possibility to reproduce large-scale bioreactor conditions in microscale cultivation systems has not yet been systematically investigated. Hence, we apply computational fluid dynamics (CFD) simulations to analyze and compare three commonly used microfluidic single-cell trapping and cultivation devices that are based on (i) mother machines (MM), (ii) monolayer growth chambers (MGC), and (iii) negative dielectrophoresis (nDEP). Several representative time-variant nutrient concentration profiles are applied at the chip entry. Responses to these input signals within the studied microfluidic devices are comparatively evaluated at the positions of the cultivated cells. The results are comprehensively presented in a Bode diagram that illustrates the degree of signal damping depending on the frequency of change in the inlet concentration. As a key finding, the MM can accurately reproduce signal changes that occur within 1 s or slower, which are typical for the environmental conditions observed by single cells in large-scale bioreactors, while faster changes are levelled out. In contrast, the nDEP and MGC are found to level out signal changes occurring within 10 s or faster, which can be critical for the proposed application.