University of Houston • University of Houston-Clear Lake • ISSO Annual Report Y2002—pp. 9-12

Vemuri Balakotaiah (UH), Dmitri Lastochkin (UH), Eugene Ungar (NASA-JSC), Brian J. Motil (NASA-Glenn), and Cesar E. Meza, Jr. (UH)

Abstract
Biological processes are currently being investigated for use in spacecraft wastewater treatment. In a biological process, microorganisms are used to degrade organic and inorganic contaminants to carbon dioxide, water, and other metabolic products. One main step in the process is nitrification in which ammonium ions in the wastewater stream are converted to nitrate ions. The reactor used in the process is a small diameter coil (3 mm) of a length about 100 m in which air and wastewater flow co-currently (slug or Taylor flow regime). Aerobic microbes flow on the tube walls. Recent experiments have shown that the bioreactor experienced occasional shedding events where biomass was dislodged from the tube wall.

This study investigates two-phase flow instabilities in tubular bioreactors with the ultimate goal of improving the bioreactor performance and scaling it up from normal gravity to the microgravity environment. Flow experiments used fluids of different viscosity and surface tension (1) to characterize the critical Reynolds numbers at which the slug-annular transition occurs and (2) to investigate the stability of the bioreactor in the annular regime. Near the slug-annular transition point, researchers observed the phenomenon of pressure drop jump and hysteresis in the pressure drop versus flow rate curve.

BIOLOGICAL PROCESSES ARE CURRENTLY UNDER STUDY FOR use in spacecraft wastewater treatment by the NASA-Johnson Space Center (JSC).¹ In a biological wastewater processor, microorganisms are used to degrade organic and inorganic contaminants to carbon dioxide, water, and other metabolic products. One important step in this process is nitrification in which ammonium ions in the wastewater stream are converted to nitrate ions. Traditional reactor designs for nitrification that include continuously stirred tank reactors and trickling filters are unsuitable for use on spacecraft because of their reliance on gravity for aeration. A tubular reactor for aerobic nitrification in a microgravity environment is currently being explored at JSC to allow use of microbes for wastewater treatment on spacecraft. The experimental set-up is shown in Fig. 1.

The tubular reactor consists of a 3.2 mm ID tube, 305 m long in which air and wastewater flow co-currently. Aerobic microbes grow on the tube walls. Because of the small tube diameter and the high surface tension of wastewater, the air/wastewater flow is expected to be gravity independent. If this is indeed the case, fluid flow and biological performance will be identical on earth (normal gravity) and in-flight (microgravity) environments.

Two-types of flow patterns may occur in the nitrifier tube (Fig. 2). When the gas flow rate is high (or the tube length is small), annular flow pattern is obtained. In Fig. 2, the liquid moves as an annular film while the gas flows in the core. Oxygen and nutrients are transferred from the air to the flowing water film and then to the biomass growing on the wall. The second type of flow pattern is the slug flow (also called Taylor flow) in which the air moves as long bubbles of diameter slightly less than that of the tube and the bubbles are separated by liquid slugs. For the nitrifier to operate efficiently, shear imposed by the two-phase air and water flow must be large enough to keep the microbe layer thin, yet must be low enough such that the microbes are not removed from the wall. In addition, characteristics of the two-phase flow must be such that oxygen and wastewater nutrients are effectively delivered to the microbes.

Figure 2. Schematic diagram of the two-phase flow patterns observed in the bioreactor used by JSC for wastewater treatment

The experimental tubular nitrifier at JSC experienced occasional "shedding" events where biomass (microbes) was dislodged from the tube wall. Rapid pressure fluctuations induced flow fluctuations, stripping pieces of biomass from the tube. Shedding also occurred because of inadequate shear. In this situation, a thick layer of biomass accumulated over time, bulging into the flow stream. This mass eventually restricted the flow of air and wastewater, which then caused a pressure increase upstream of the clog. Eventually, the flow broke through, causing a violent expulsion of fluid that stripped biomass from the walls. Shedding at higher flow rates was less extensive and less severe than at the lower flow rates.

A possible explanation for this phenomenon is that the higher shear resulted in a thinner and more robust biofilm. These preliminary experiments led to the conclusion that an optimal shear exists that is low enough such that the biomass can attach to the tube wall, but high enough to maintain a thin, rugged biomass. Researchers also found that the biomass was not concentrated not homogeneously along the tube length, which concentration can be explained by the existence of a pressure gradient along the tube and hence a variation of wall shear stress.

Experimental Studies
The analog of the NASA Johnson Space Center reactor was built in the Two-Phase Flow Laboratory at the University of Houston. The main difference between the two reactors is that the UH reactor was built to study hydrodynamics; no biomass was used in this study. The reactor has two lines: the air line and the liquid line. The air supply was measured by calibrated flowmeters while the liquid flow rate was determined from the weight of liquid passing through the reactor in some chosen time interval. Air and liquid supplies were mixed in the entrance region of the test duct by coaxial injection. A 50-ft long silicone tube with 1/8 in. internal diameter was used in the experiments. The cross sectional flow area of the tube was checked as the volume per unit of length, using water and other fluids as test fluids.

Researchers measured pressure drop over the entire length of the tube by calibrated pressure transducers. Transducers were connected to the line by special connectors so as not to change the cross-sectional area of the tubes. In addition to water, glycerin water solutions of different concentrations were used in this study.

We performed the experiments with more viscous solutions for two reasons: first, dimensional analysis shows that the flow of water in the smallest tube is equivalent to the flow of a more viscous solution in the tube of larger diameter. Since it is not clear at the present time which tube diameter is needed to obtain completely gravity independent flow, we can model two-phase flow of water in small tubes by the flow of viscous solutions, which is easy from the experimental point of view. Secondly, the slug annular transition proceeds clearly and more sharply in viscous solutions than in water, so we can successfully investigate the general physics of the process.

We show in Fig. 3 the pressure as a function of superficial gas Reynolds number at a fixed superficial liquid Reynolds number. (Since the outer pressure in our experiments was always ambient, we refer to what is commonly called pressure drop as pressure at the slug-annular transition.) Note that this plot can be readily represented in terms of corresponding flow rates, which connect to Reynolds numbers through gas and liquid viscosities. As can be seen from Fig. 3, after an initial increase of pressure with the increase of gas flow rate, the pressure drops and flow rate increases, sometimes an order of magnitude without further increase of pressure in the gas line. We should emphasize here that, in reality, the primary parameter in each gas line is pressure; the increase in gas flow rate was provided by increase in air pressure.

The explanation of this phenomenon is simple. At the initial stage of two-phase flow, when one works with sufficient liquid flow rate and low gas flow rate, the flow itself represents a slug (sometimes referred to as a plug) flow where gas cannot pass through a channel because of liquid slugs (plugs). Further increase in the gas flow rate leads to an increase of gas fraction and agglomeration of gas bubbles in the core of a two-phase flow. Finally, at some gas flow rate, the open channel forms in the core, and resistance to gas transition decreases. In other words, the gas "sees" the open channel and, as a result, the gas flow rate jumps up without a change in gas pressure. Hence, slug-annular transition in long channels is accompanied by a jump in the pressure drop at the transition point. Similar phenomena are presented on Fig. 4 where viscosity of water-glycerin solution was 67.7 cp.

One can see from Fig. 4 that the slug-annular transition for a higher viscosity liquid is sharper than for a lower viscosity liquid and the transition point is more clearly determined. Obviously, in such cases, we cannot use any of the existing models based on short channel experiments because all of them predict a monotonic increase in pressure with an increase of the gas and liquid flow rates. In reality, pressure is not a monotonic function of flow rates because of the slug-annular transition.

The same phenomenon can be observed for water, as shown in Fig. 5. The difference between water and more viscous water-glycerin solutions is that the magnitude of the jump in pressure drop at slug-annular transition point is not as large for water. We believe that the magnitude of this jump also depends on the channel length; the jump becomes clearer in sufficiently long channels (as used in the bioreactor). For example, in the Bao et al.¹ experiments researchers measured pressure drop in 150 mm intervals while the slug-annular transition takes place along all the length of the channel. More than that, this phenomenon is complicated by the fact that different parts of the channel have different "resistivity" to slug-annular transition. Indeed, typically for a long channel, pressure drop along the channel is enough to take the gas fraction change into consideration because of pressure decreases along the channel. That explains why the thickness of liquid slugs at the end of a channel is lower than at its beginning. Hence, an initial channel length has greater "resistivity" to transition than an end-length of a channel.

Acknowledgments
This work was supported by UH-NASA Johnson Space Center Post-Doctoral Aerospace Fellowship program.

References
¹Z. Y. Bao et al. "Estimation of Void Fraction and Pressure Drop for Two-Phase Flow in Fine Passages," Trans. Inst. Chem. Eng. 72A (1994): 625-32.

Publications
Lastochkin, D. and V. Balakotaiah. "Fluid Mechanics in a Long Channel of Tubular Reactor Designed for Use in Space," International J. Multiphase Flow. (In preparation.)
Lastochkin, D., E. K. Dao, and V. Balakotaiah. "Inertial Effects and Wave Suppression on Vertically Falling Films," Phy. Fluids. (In review.)
Motil, B. J., V. Balakotaiah, and Y. Kamotani. "Gas-Liquid Two-Phase Flow through Packed-Beds in Microgravity," AlChE Journal 49 (2003): 557-65.

Presentations
Balakotaiah, V. "Experimental and Modeling Studies on Wave Suppression on Vertically Falling Films," AlChE Meeting, Nov. 5, 2002.
Balakotaiah, V. and B. J. Motil. "Fundamental Studies Gas-Liquid Two-Phase Flows through Packed-Beds in Microgravity," 6th Microgravity Fluid Physics Conf., Cleveland, OH, Aug. 2002.
Motil, B. J., V. Balakotaiah, and Y. Kamotani. "Gas-Liquid Two-Phase Flow through Packed-Beds in Microgravity," paper #279j, AlChE Annual Meeting, Indianapolis, IN, Nov. 2002.

Funding and proposals
"Fundamental Studies on Gas-Liquid Two-Phase Flows through Packed-Beds in Microgravity," Co-PI: M. J. McCready, University of Notre Dame and B. J. Motil, NASA-Glenn. NASA-Glenn Research Center, 2002-2006, $405,000.
"Modeling and Experimental Studies on Gas-Liquid Hydrodynamics and Mass Transfer in Bioreactors for Wastewater Treatment in Low Gravity," NASA Fluid Physics NRA, $400,000, four years (pending).
"Modeling and Experimental Studies on Wave Suppression and Occlusion in Annular Gas-Liquid Two-Phase Flows," DOE Basic Energy Sciences, $341,540, three years (pending).
"Pressure Drop and Mass Transfer Studies in Gas-Liquid Two-Phase Slug Flow under Microgravity Conditions," NASA GSRP grant for Mr. Cesar Meza, $72,000, three years (pending).

PDF (K)
Table of Contents

Institute for Space Systems Operations - Y2002 Annual Report
Copyright © 2003