MS

LIST OF TABLES

Packed towers are a particular type of fluid-fluid contactor that have been widely used in the process industries mainly for either gas absorption or liquid stripping operations. For example, packed columns are used in pollution abatement to remove one or more toxic gaseous species from stack gases by scrubbing with a suitable scrubbing liquid.

A schematic of a typical packed column is shown in Figure 1. A typical packed tower consists of vertical process vessel equipped with various external ports for introduction and removal of the process fluids, such as gases and liquids or partially miscible liquid phases. The column internals consist of various hardware for distribution, redistribution and separation of the gas and liquid phases or other process fluids, a grid system for supporting the column packing, and possibly special-purpose hardware for control of column temperature. In the countercurrent mode of fluid-fluid contacting, the process gas is introduced at the bottom of column and the process liquid is introduced at the top.

Packed towers are sometimes chosen over tray towers since they can exhibit a lower pressure drop than tray towers and can exhibit higher efficiency when compared to tray columns using one or more mass transfer parameters as the basis, such as height equivalent to a theoretical plate. Also, packed column configurations can provide substantially larger fluid-fluid specific contacting area for interphase heat and mass transport, they can handle a broader range of liquid to gas mass velocities without a deterioration in performance, the packing can be fabricated using a wider range of materials of construction for improved corrosion resistant, and they are more suitable for foaming systems.

For the countercurrent mode of contacting, the liquid flows downward the column and the gas move upward the column through the packing. In the case of the low gas-liquid interaction regime, the liquid phase is the discontinuous phase, and the gas is the continuous phase [34]

There are two types of packings in packed towers, structural and random. The packing provides a surface where the two fluids can come into intimate contact so that local mass and heat transfer processes can occur. Also, the packing is commonly an inert material, such as ceramic or plastic rings, which is explained below in more detail.

Comment by Graduate Studies.: Below 1 inch margin. Move up.

Figure 1: Packed Column [33] Comment by Graduate Studies.: 12 points, black font. Remove italics.

The performance of a packed column depends on various parameters, including the macroscopic and microscopic liquid and gas distribution, pressure drop, liquid holdup, gas-liquid flow regime, extent of liquid-solid contacting, heat and mass transfer rates, and gas-liquid specific interfacial area. One source of energy loss that occurs is largely manifested in the form of pressure drop. Therefore, knowledge of pressure drop and flow regime are crucial in predicting the performance of packed column.

Figure 3: Structured Packings [37] Comment by Graduate Studies.: Center.

Important hydrodynamic parameters that are needed either for packed column design or rating of performance include overall pressure drop, flow regime, flow regime transition, and liquid holdup. In the case of two-phase countercurrent flow, these parameters have been the subject of numerous experimental studies as well as efforts to develop empirical correlations using data that was measured under a variety of conditions. In a number of cases, the scope of experimental variables investigated, such as gas and liquid mass velocities, gas and liquid fluid transport properties and other related physical properties, column diameter and height, column packing material of construction, shape, and size, and flow regime has been very limited. In addition, the experimental systems often employ various manually operated control systems along with very basic instruments for measurement of key response variables, such as the differential pressure drop across the packed section. Hence, a need exists for an experimental facility where a more modern approach to system operation is utilized and the parameter space can be defined by a broader scope of experimental parameters. This would allow students to obtain a better understanding of the physics associated with packed column systems and to acquire response data that covers a broader range of parameter space.

*Data is collected from web-sited of companies that manufacture column packings [38]

This research is focused upon modification of the existing packed column system located in the Unit Operations Laboratory in Kleberg Hall 139 as part of a larger effort aimed at modernization of various experiments used for teaching the unit operations sequence of courses. The proposed configuration will allow pressure drop measurements to be performed using a high frequency response transducer for a user-defined set of gas and liquid flow rates using either a single-phase air or an air-water multiphase system. The resulting pressure drop data will allow investigation of flow regime, flow regime transitions, and other flow phenomena, such as column flooding, thereby providing students with more in-depth knowledge on the hydrodynamics of single-phase and multiphase flow through porous media.

Comment by Graduate Studies.: Remove extra blank page.

LITERATURE REVIEW Comment by Graduate Studies.: Center.

Various transport processes in packed columns depend upon the fluid flow regime, which in turn depends on the gas and liquid mass velocities and packing dimensions. Various flow regimes exist for countercurrent gas-liquid flow such as trickle flow, pulsing flow, spray flow and wavy flow [34]. In the trickle-flow regime or low gas interaction regime, a discontinuous flowing liquid phase trickles over the packing in the form of streams, rivulets and droplets while the continuous flowing gas phase fills the remaining void space that is not occupied by the liquid and packing.

Understanding the hydrodynamics of packed columns is a key aspect associated with prediction of their design and performance. The hydrodynamics in this fluid-fluid contactor can be described through a mechanistic approach rather than a complex fundamental viewpoint. A description of macroscale and microscale phenomena in trickle-bed reactors, which consist of a packed bed of catalyst with cocurrent downflow of gas and liquid, are provided by Dudukovic and Mills.[21] The performance of a trickle-bed reactor can be partly characterized based on the macroscopic hydrodynamic parameters such as pressure drop, flow regime, liquid holdup, and liquid-solid contacting. All these parameters must be quantitatively evaluated to support the evaluation of reactor performance on various scales of operation. To analyze the performance of packed towers, the same parameters must be evaluated to support the solution of the overall mass, specie, and energy balance relationships that described the packed column performance. This research focuses on development of an automated system for measurement of pressure drop and flow regime where measurements are first performed using the existing manual system to identify existing limitations that must be overcome.

A knowledge of pressure drop is important for estimation of feed pressure, pump and compressor capacities, and for predicting parameters such as liquid holdup, and transport coefficients [1, 2, 18, 19]. A large number of correlations and theoretical models can be found in the literature for prediction of two-phase pressure drop through a packed bed. Table 1 summarizes selected pressure drop correlations for multiphase gas-liquid flow through packed beds. A majority of the correlations are based on the Ergun equation [21] as shown below in eq. (1). In many cases, the modified Lockhart-Martinelli two-phase flow parameters were utilized to capture the presence of both gas and liquid phases as well as the effect of fluid physical properties. Comment by Graduate Studies.: Indent. Indent all the new paragraphs.

38

Table 2. Continued Comment by Graduate Studies.: Below 1 inch margin. Move up . Remove parenthesis.

Table 2. Continued

Table 2. Continued

As explained in a later section, pressure drop data was collected using the packed column system located in Kleberg Hall room 139. The experimental data were compared to values predicted from the correlations of Midoux et al. (1997) and also the generalized correlation described by Larachi et al. (1993). These two correlations are summarized below for later reference. Comment by Graduate Studies.: Indent all new paragraphs.

Midoux et al. (1997) [31]

Larachi et al. (1993) [17]

The Excel simulator can be used in one of two alternate ways. In the first method, predictions can be made using existing packings that are already available in data bank. Alternately, predictions can be made by using packing characteristic supplied by the user. The packing properties can be varied in the box highlighted in color, green.

A packed bed or a trickle bed reactor can be operated in various flow regimes that were broadly classified as the low interaction regime and the high interaction regime. A popular flow regime map developed by Charpentier and Favier (24) is shown in Figure 4. It is based on the so-called Baker coordinates in which the dimensionless group Lλφ/G is plotted against G/λ where λ and φ are parameters that are dependent on fluid properties. Each of these parameters assumes a value of unity when the fluids are air and water. Depending on the values for these dimensionless groups, three flow regimes can exist, namely, the trickling flow regime, the pulsing flow regime, and the spray flow regime. Qualitative features of these flow regimes will be discussed later in this section. The qualitative features of each flow regime are illustrated in Figure 5.

Figure 5: Flow map proposed by Charpentier & Favier, 1975 [24] Comment by Graduate Studies.: Remove extra blank spacing.

One significant aspect associated with identification of the flow regime is that the magnitudes of various key reactor parameters, such as the axial and radial mixing characteristics of the gas and liquid phases, pressure drop, particle-scale liquid-solid contacting efficiency, reactor-scale gas and liquid distributions, and heat and mass transfer coefficients, will depend on the particular flow regime where the reactor is operating [1]. However, the transition from one flow regime to another may differ for foaming and non-foaming systems, along with other hydrodynamic parameters [22, 23].

The designation of a foaming system suggests that a frothy mass of fine bubbles forms at liquid surfaces or at the gas-liquid interface in the presence of flowing gas phase. For example, certain hydrocarbons, such as kerosene, desulfurized and non-desulfurized gas oils, have a tendency to foam whereas organic compounds, such as cyclohexane, gasoline, and petroleum ether do not exhibit this behavior [24]. In industrial practice, some hydrocarbon systems may begin to foam depending on fluid properties, such as surface tension and viscosity. Sometimes, anti-foaming agents or so-called defoamers are added to inhibit foam formation [23]. Therefore, some studies in the literature on pressure drop, flow regime, and flow regime transition have utilized non-foaming systems in experiments. Since air and water are the two fluids used in this work, the results apply to non-foaming systems only. The key characteristics of various flow regimes for a non-foaming system are described below.

a. Low interaction regime. This regime (Figure 5a) is also known as the trickle-flow regime and is observed at low gas and liquid flow velocities in which the liquid phase trickles over the packing in the form of streams and thin films with a minimum interaction with the continuous flowing gas phase. This flow regime promotes stable reactor operation and longer residence times for the liquid phase as desired for chemistries in which observed local reaction rate is relatively slow when compared to other local transport processes. However, reactor operation in the trickle-flow regime may not be preferred when higher heat and mass transfer coefficients are required for reactions whose observed rate may be limited by external transport resistances [26].

b. High interaction regime. This regime is characterized by a significant amount of gas-liquid interaction that is induced by an appropriate combination of gas and liquid mass velocities, gas and liquid fluid properties, packing geometry, packing characteristics, bed voidage distribution, temperature and pressure. The high gas interaction regime can be further classified into the pulse flow, dispersed bubble flow, and spray flow regimes. Each of these regimes is briefly described below.

a. Pulse flow regime. This flow regime (Figure 5b) is characterized by high gas and liquid mass velocities that produces a high degree of interaction between the flowing gas and liquid phases and the stagnant solid catalyst. It is sometimes referred to the pulsing flow regime. At sufficiently high liquid flow rates, the catalyst surfaces are substantially wetted, which leads to the formation of bridges of liquid phase between the catalyst particles. These liquid bridges are ruptured and subsequently pushed downward by the gas flowing through the bed as soon as they grow to a sufficiently large size. As a result, alternating liquid and gas rich slugs (or “pulses”) can be observed flowing through the catalyst bed. This leads to an increased pressure drop across the packed bed and generates fluctuations in the liquid holdup. Comment by Graduate Studies.: Align left. Fix. Comment by Graduate Studies.: Fix.

b. Spray flow regime. The spray or mist flow regime (Figure 5c) is achieved at very high gas-to-liquid flow ratios and is characterized by a very high degree of gas-liquid interactions as a result of significant drag at the local gas-liquid interface. The magnitude of the drag is large enough the liquid phase is broken down into tiny drops that are subsequently entrained in the downward flowing continuous gas flow. This regime is not very common in the industry because most reactions conducted in practical applications require higher liquid to gas flow ratios for significant reactant conversion to occur. Since it is very difficult to discern a boundary between the spray flow and trickle-flow regimes by simple visual observations, both are sometimes grouped together and described as the gas continuous flow regime [12, 25].

c. Dispersed bubble flow regime. The dispersed bubble flow regime (Figure 5d) occurs at sufficiently low gas velocities and high liquid-to-gas flow ratios. These ratios cause the gas phase to be broken down into small bubbles that flow through the continuous liquid phase as a dispersed phase. This phenomenon also results in a high level of interaction between the flowing gas and liquid phases and the stagnant catalyst particles. This flow regime is characterized by rapid fluctuations in the column hydrodynamic, such as the pressure drop, liquid holdup, and the gas-liquid specific interfacial area.

Flow regime transition is generally concerned with the various multiphase flow phenomena that occur when changes in various system parameters, such as the gas and liquid mass velocities, physical properties of the gas and liquid phases, and catalyst bed characteristics, results in a transformation in the flow regime from one regime to another. This change either may or may not be readily discernible by normal visual observations. Nevertheless, the ability to predict flow regime transition is important in the analysis and prediction of reactor performance and operation. Other parameters, such as pressure drop, liquid holdup, liquid-solid contacting, and mass transfer coefficients, to name a few, are also dependent on the flow regime. Therefore, significant research has focused on identifying the transition from trickle flow to pulse flow or bubble flow, and on quantifying the associated hydrodynamic parameters. Many authors have proposed flow maps and correlations to identify transition from one regime to another. For example, Figure 1 shows the first flow map proposed by Charpentier & Favier, 1975 that is accepted widely in literature.

There are two different types of liquid holdup in a packed bed since the total liquid hold up is the sum of both the static and dynamic holdup.

a. Static Holdup. Static holdup represents the volume of liquid per volume of packing that remains in the bed after the gas and liquid flows are brought to an abrupt stop and the bed has drained. It depends on packing surface area, roughness of packing surface, contact angle between the packing surface and the liquid and system properties. The static holdup has been correlated using the Eötvos number (Saez et al., 1991).[39]

b. Operating/Dynamic Holdup. The dynamic holdup corresponds to the volume of liquid per volume of packing that drains out of the bed after the gas and liquid flows to the column are abruptly stopped. Liquid holdup is an important parameter in packed column design and operation since it is related to the residence time of the actively flowing liquid phase and also contributes to the effective local thermal capacity of the combined packing and fluid that are present. Dynamic holdup is generally a function of the gas and liquid flow rates, gas and liquid physical properties, and packing characteristics.

Main research objective is to identify the shortcomings associated with the controls, instrumentation, and operation of the existing packed column system in the Unit Operations Laboratory in view of the ABET assessment and perform experiments (1) to determine pressure drop (2) to observe flow regime transitions and (3) to compare the measured pressure drops with those calculated using existing correlations for both single phase gas flow and two-phase gas-liquid flow.

Identify the shortcomings associated with the controls, instrumentation, and operation of the existing packed column system in the Unit Operations Laboratory in view of the ABET assessment.

Perform experiments to determine allowable range of operation for existing system in terms of gas and liquid mass velocities, pressure drop, flooding and flow regime transition. Then, compare the experimental measurements to existing correlations for pressure drop and flow regime to assess the validity of these correlations and to expand the scope of existing methods used for data analysis since this is currently based on simple Ergun equation-like analysis.

Design a modern PLC-based process control and data acquisition system using the existing packed column system as the initial platform that can be expanded in the future. Design and implement a unique Human Machine Interface (HMI) for control of the overall system, various sub-systems, and implementation of experiments and develop recommendations for future work.

The experimental setup currently used in the Unit Operations Laboratory is described in this section. This includes an overview of the experimental objectives of interest for this particular application and the limitations of the current experimental setup shown in the process and instrumentation diagram (P&ID) and as well as a photograph reference.

The primary objectives of this experiment are: (1) to determine overall pressure drops at various gas and liquid rates; (2) to determine the loading point and the flooding point at various liquid flow rates starting with “dry” conditions; (3) to observe flow regime transitions; (4) to compare the measured pressure drops with those calculated using existing correlations for both single phase gas flow and two-phase gas-liquid flow; and (5) to compare experimental data for flow transition in two-phase gas-liquid flow with predictions using existing correlations.

By varying the gas and liquid inlet flow rates, the pressure drop behavior, gas-liquid flow regimes and flow-regime transitions can be studied. Manual controls are used for flow rate settings using basic process control elements such as rotameters, pressure regulators and gauges. A U-tube manometer is utilized for pressure drop measurements. This manometer is not capable of capturing high frequency pressure drop fluctuations. Also, Flooding point at various liquid flow rates could not be determined due to gas bypassing, and this setup is limited to countercurrent mode of gas liquid contacting.

Rotameter for Air Flow

Rotameter for Water Flow

Ballast Tank

City Water Supply

Figure 7: Piping and Instrumentation Diagram of Existing Packed Column System Comment by Graduate Studies.: Figure out of 1 inch margin. Move in .

Raschig Rings

Pump

Rotameter for Water

Rotameter for Air

Air Inlet

Air Outlet

Water Inlet

Water Outlet

Manometer

Figure 5: Lab Scale Packed Column

3-in. ID x 60-in. H

Figure 8. Lab Scale Packed Column

The experiment involves determination of pressure drop (inches of H2O) per foot of packing (∆P/L) for various air and water flow rates. The dimensions of the packed column are 5 ft. long with an inner diameter of 3-in. that is packed with ¼-in. by 3/8-in. Raschig rings. Air is fed from the bottom of column from the compressed air lab header and water from the top of column from the city water supply. Both air and water flow rates are controlled by rotameters on a scale of 0 to 100 with a scaling factor of 0.0418 for air to convert flow in terms of SCFM and a scaling factor of 0.0078 for water to convert in terms of gpm.

a. Single phase operation:

With damp packing and no water flow, turn on the air flow and wait until a steady state is attained. Then note down the manual reading given by manometer, which shows difference in water column in terms of inches of water column (units for pressure drop). This process is repeated by increasing the gas flow rate and note down the respective manometer readings.to determine pressure drop, loading and flooding points.

b. Multi-phase operation:

With damp packing and no water flow, turn on the air flow and wait until steady state is attained, then turn on the water flow. Make sure to wait some time till steady state is attained and then note the manometer readings. This process is repeated for various flow rates of air and water to determine the pressure drop, gas mass velocity, superficial velocity and also study hysteresis effect.

Data was collected for following operating range, analyzed for trends and compared experimental data with predicted correlations.

Gas Flow: 0.418 to 4.18 SCFM

Air compressor is turned on, keeping the water flow at zero. Gas flow rates are increased gradually, and corresponding pressure drop is noted from manometer. Figure 9.1 depicts the behavior of superficial velocity on pressure drop for a single-phase flow.

Also, for single phase flow, experimental pressure drop is not in correlation with Ergun equation for given porosity of system (Ɛ=0.68). But, for a reduced voidage (Ɛ=0.51), most of the experimental data is in correlation with predicted values based on Ergun equation. It is evident that the voidage has been reduced due to local accumulation of fouling material in packing (figure 9.3), because the system is installed 40 years ago and has not been cleaned since then.

Ergun equation used to calculate pressure drop for voidage of 0.68 and sphericity of 0.85

Data was collected for following operating range, analyzed for trends and compared experimental data with predicted correlations.

Gas Flow: 0.418 to 4.18 SCFM

Liquid Flow: 0.078 to 0.312 gpm

For multi-phase operation, gas flow rate is first set to zero and column is allowed to wet for some time. Now, rotameter for gas flow is turned on and manometer reading is noted for a given flow rate of air and water. Liquid flow rate is kept constant and gas flow rate is increased gradually, while noting the manometer readings, and then readings are noted for the same liquid flow rate and decreasing the air flow rate. Figures 11 and 12 shows how pressure drop increases with increase in gas flow rate, but the data is limited for higher flow rates of gas and for cold flow conditions. Graphically, flooding and loading points are determined by plotting pressure drop of column against gas mass velocity. A point where the orientation of graph is changed is located as loading point and the path where it takes sudden peak is located as flooding point.

4772

Figure 12: Experimental Pressure Drop vs Air Mass Velocity

Mode of operation is same as described above for a multi-phase flow. Figures 14(a) and 14(b) shows the comparison of experimental values vs predicted using two correlations by Midoux and Larachi. Because of fouling in packing, porosity of existing system is changed to 0.51 (as observed for single phase operation). Larachi’s excel simulator is used to generate results for various gas and liquid flow rates for porosities 0.68 and 0.51 in multiphase flow. A huge error was observed between experimental and predicted values for given porosity of system, but the system is in correlation with Larachi’s predicted data, generated for porosity of 0.51.

Hysteresis was also exhibited by the system, mainly at higher gas flow rates. This observation is reported by many authors [4,5,32]. Figure 14 (a-b) shows the hysteresis effects observed a various gas and liquid flow rates. At high gas and liquid flow rates, liquid flows down in form of channeled rivulets through relatively dry packing and spreads to other parts of packing that were not in contact with liquid earlier (at low liquid flow rates). Once the liquid flow rate is high enough, a portion of packing gets covered in a thin film of fluid. This film should be broken by gas flow. Therefore, liquid flow is decreased, and the gas encounters higher pressure drop than, when compared to higher liquid flow rates.

Comment by Graduate Studies.: Below 1 inch margin. Move up.

Figure 15 is re-defined plot of pressure drop correlation by Eckert for the existing system. This graph helps in predicting the pressure drop for given gas and liquid flow rates.

Using the same mode of operation for multi-phase flow, flow regime maps developed with experimental data and are compared to the maps developed by Charpentier-Favier [24] and Fukushima-Kusaka[30]. From the plot, one can see that the existing system only operates in trickling and pulsing flow regime. The operating conditions correspond to a very low gas to liquid ratio, quite common in the industry but rarely encountered in literature Operation at higher flow rates in packed column should be studied to know more about flow regime transitions (visually).

To obtain more consistent readings (represents industrial columns), column should be subjected to high interaction regimes. So, there is a need to upgrade the existing system to automation.

Pulsing Flow

Spray Flow

Trickling Flow

Figure 16: Flow Regime Map by Charpentier-Favier Comment by Graduate Studies.: Placement of figures is consistent.

System is controlled manually, pressure drop (∆P) measurements and flow regime transition from low interaction to high interaction are limited by ranges for gas and liquid flow rates. Flooding cannot be approached due to gas bypassing at liquid seal of system (“U” neck installation at bottom). Simple manometer is used for pressure drop measurements and (manual data collection) it cannot capture higher frequency fluctuations which are indicative of flow regime transition. Also, experimental values are noted manually, and data collection has limited the scope of experimental studies. System is also limited to single column size (H, D, H/D), single packing type (Raschig rings) and single mode of gas-liquid contacting (countercurrent).

I. AUTOMATION OF PACKED COLUMN

Figure 18: Piping and Instrumentation Diagram of Automated System

Determination of pressure drop can be carried out in an automated system of packed column for a single-phase gas flow and two-phase liquid-gas flow under different contacting modes. Dimensions of column are 8ft long and 9 inch diameter (derived using excel worksheet of packed bed simulation by F.Larachi and B.P.A. Grandjean, Laval University, Canada). Types of packings to be used in this experiment can be started with spheres, pall rings and raschig rings with three different diameters (to maintain constant D/dp), and further study of pressure drop on fluid properties can be determined by changing the liquid media like ethylene glycol, methocel.

Air is introduced into the system from compressor to a 10-gallon ballast tank (T-1) through a ¾” PVC pipeline. The ballast tank provides a buffer volume between the header and the downstream piping system so that any upstream fluctuations in the header supply pressure, such as those due to intermittent operation of the compressor, are minimized. Both the pressure and temperature of air that enter the ballast tank are measured by a pressure transducer (PT-2) and temperature transmitter (TT-1). A manually operated forward pressure regulator (FPR-1) is located ahead of the ballast tank so air is supplied to the downstream test section at a relatively constant pressure. The exit side of the ballast tank is connected to a supply line that contains a mass flow controller (MFC-1) having an upper limit of 2000 SLPM based on air. The pressure upstream of the mass flow controller is measured with a pressure transducer (PT-3) while the pressure downstream of the mass flow controller is regulated by a manually operated back pressure regulator (BPR-1). The pressure on the exit side of the mass flow controller is measured with another pressure transducer (PT-4). A tee is located downstream of the back pressure regulator with a manually operated ball-valve (V-2) so the gas flow can be diverted to a wet test meter for calibration of the mass flow controller. A thermocouple (TT-2) is in close proximity of the column side port for a final measurement of the air temperature before it is introduced to the column. The air exit is located at the top of the column since the air and water flow in countercurrent flow. A pressure transducer (PT-7) is used to measure the pressure in the air vent line while the temperature is measured using an in-line thermocouple (TT-3).

Water is introduced to the system from the laboratory deionized water supply header. The tank level is monitored using a differential pressure transmitter (DPT-2) which is also used to actuate a level control valve (LCV-1). The water temperature in the tank is monitored using a thermocouple (TT-4). The tank can be drained through a two-way ball valve (BV-10) that is attached to one side of a tee in the tank exit line. The electrical power for the pump can be turned on or off through an 115V solid-state relay that is actuated from the HMI of the PLC. The fluid pressures at both the inlet and exit sides of the pump are measured by pressure transducers PT-9 and PT-10, respectively and fluid temperature at the pump exit is measured by an in-line thermocouple (TT-5). Under normal operation, the water is directed to the top of the column where the temperature is measured before being introduced using an in-line thermocouple (TT-6). The water exits the column through a glass pipe that is connected to the bottom head. The water temperature at the exit of the column is measured using another in-line thermocouple (TT-7).

The system is designed so the gas and liquid flow rates can be varied over wide ranges so the gas and liquid interaction can be varied from the low gas-liquid interaction (trickling flow) regime to the high gas-liquid interaction (pulsing flow or spray flow) regime. A differential pressure transmitter (DPT-1) connected across the packed section of the column provides a means of measuring the overall column pressure drop during process.

Since the system is automated, one can specify up to 15 different air flow rate and water flow rate data pairs at which pressure drop data would be recorded. In addition, each of air-water flow rate settings would be maintained for specified dwell times where the maximum dwell time is 10 minutes. Once all the flow rate settings have been selected, the flow rate sequence would be initiated, and the system would then proceed to record pressure drop vs flow rate data for the user-specified dwell-time. The resulting pressure drop versus flow rate data are then stored in a user-specified file on the computer system hard drive for subsequent data analysis. A standard operating procedure (SOP) for identifying column flooding will require development after the automation system is completed since this has not been done before with the current experimental setup.

Key Features for Automated System

1. Using of high-capacity liquid pump and gas mass flow controllers with greater ranges will expand the scope of differentia pressure (∆P) and flow regime measurements.

2. Installation of Liquid-level control system in column sump will prevent gas bypassing at all gas and liquid flow rates, thereby allowing flooding to be investigated.

3. Digital differential pressure transmitter has a response time that will measure higher frequency pressure fluctuations both in time domain and frequency domain.

4. HMI-PLC system allows user to specify up to 10 gas-liquid settings where data is recorded for a user-specified time and display in real time.

5. System is configured so that columns having various dimensions, column packing, column internals, and modes of gas-liquid contacting can be installed in the future.

HMI is based upon Wonderware InTouch™ System Platform where system is installed with Siemens 505 Series Programmable Logic Controllers (PLC) with Control Technology, Inc. (CTI) input/output modules and supporting technology.

Key Features & Capabilities of this PLC are.

1. Analog Input/Output of Voltage or Current

2. Digital Input/Output for solenoids, relays, etc.

3. Wide range of temperature sensing devices

4. RS-232or RS-485 protocols for instruments

5. Ethernet to PC interface module

6. Other specialty modules and PID control

Figure 20: HMI Process Historian

Statistical Design for Experiment

Packed columns are one of the most capital-intensive components of commercial process technology and student training on various aspects, especially that gained through hands-on training, is well-justified and a useful addition to their education. Manual operation has notable merit and has been integrated into the automated system through proper design of the process control and data acquisition system. Existing packed columns that are commercially available for teaching purposes have limited capabilities and still require significant capital resources.

Significant progress has been made toward the completion of the packed column automation. The overall design automation control system has been finalized. The HMI and its underlying programming parameters for control have been developed with only minor adjustments and improvements remaining. The final programming parameters may not be available until the physical set-up is completed. The automation instruments and equipment needed have been ordered and received

The path forward largely depends on the available budget for the project and the overall installation of equipment, which will be an extremely involved process because of the number of instruments and the fact that support structures will need to be added to the current structure for some of the heavier equipment (namely control valves), adding to the already labor-intensive processes of piping and wiring the system. The two main possible options for moving forward with the installation of equipment and validation of HMI would be to request bids from professional contractors, or to establish a team of engineering students in relevant disciplines (chemical/mechanical/electrical). A combination of these two scenarios may be possible. For instance, the university may be able to find instrumentation/piping/control technicians or engineers working in the Corpus Christi area that would be willing to serve as advisers to an engineering team; then the completion of the project would provide a valuable learning experience to students. Also, the university would benefit by saving labor costs while still having professional oversight.

Upgrading the existing packed column system provides flexibility with desirable features and provides an attractive alternative for academic program improvement. Automation of a pilot-scale packed column for teaching and other training purposes requires significant effort as well as various financial and human resources for successful implementation.

Raschig rings are used in current experiment to determine the pressure drop across column for single and multiphase mixtures. Pressure drop is not only a function of flowrates, phase properties but also voidage and packing/catalyst size. As our main aim is to fill the knowledge gap between low gas-liquid and high gas-liquid interaction regime, development of automated system is resourceful

Use the automated system to measure the pressure drop for different packings using single and multiphase fluid mixtures.

· Using various protocols for different contacting/flowrate modes.

· Implement tracer method for measurement of liquid holdup.

Columns having different diameters and heights with fluids having a wide range of fluid properties (µ, σ, ρ). can be installed to study the effect of column dimensions on pressure drop and flow regime transitions. Various dumped and structured packings that produce bed structures having various voidage distributions and ranges for packing surface to volume ratios can also be used.

Use various liquid and gas flow rates so that the flow regime is varied from low gas-liquid interaction regime to the high gas-liquid interaction regime and vice-versa to study hysteresis effects. Studying the hydrodynamic parameters at higher liquid and air flow rates can also be made for different column dimensions.

Compare the experimental measurements to existing correlations for pressure drop, flow regime and liquid holdup and update the existing Larachi data bank with newer data.

1. Shah, Yatish T. Gas liquid solid reactor design. Vol. 327. McGraw-Hill International Book Company, 1979.

2. Ranade, Vivek V., Raghunath Chaudhari, and Prashant R. Gunjal. Trickle bed reactors: Reactor engineering and applications. Elsevier, 2011.

3. Kan, Kin-Mun, and Paul F. Greenfield. "Multiple hydrodynamic states in cocurrent two-phase downflow through packed beds." Industrial & Engineering Chemistry Process Design and Development 17.4 (1978): 482-485.

4. Levec, Janez, A. E. Saez, and R. G. Carbonell. "The hydrodynamics of trickling flow in packed beds. Part II: Experimental observations." AIChE journal 32.3 (1986): 369-380.

5. Kan, Kin-Mun, and Paul F. Greenfield. "Pressure drop and holdup in two-phase cocurrent trickle flows through beds of small packings." Industrial & Engineering Chemistry Process Design and Development 18.4 (1979): 740-745.

6. Rao, V. Govardhana, and A. A. H. Drinkenburg. "Pressure drop and hydrodynamic properties of pulses in two‐phase gas‐liquid downflow through packed columns." The Canadian Journal of Chemical Engineering 61.2 (1983): 158-167.

7. Ratheesh, S., and A. Kannan. "Holdup and pressure drop studies in structured packings with catalysts." Chemical Engineering Journal 104.1-3 (2004): 45-54.

8. Ellenberger, J., and R. Krishna. "Counter-current operation of structured catalytically packed distillation columns: pressure drop, holdup and mixing." Chemical Engineering Science54.10 (1999): 1339-1345.

9. Larkins, Robert Pruett, R. R. White, and D. W. Jeffrey. "Two‐phase concurrent flow in packed beds." AIChE Journal 7.2 (1961): 231-239.

10. Sai, P. S. T., and Y. B. G. Varma. "Pressure drop in gas‐liquid downflow through packed beds." AIChE journal 33.12 (1987): 2027-2036.

11. Larkins, Robert Pruett, R. R. White, and D. W. Jeffrey. "Two‐phase concurrent flow in packed beds." AIChE Journal 7.2 (1961): 231-239.

12. Sato, Yuji, et al. "Pressure loss and liquid holdup in packed bed reactor with cocurrent gas-liquid down flow." Journal of Chemical Engineering of Japan 6.2 (1973): 147-152.

13. Turpin, Jim L., and R. L. Huntington. "Prediction of pressure drop for two‐phase, two‐component concurrent flow in packed beds." AIChE Journal 13.6 (1967): 1196-1202.

14. Specchia, V., and G. Baldi. "Pressure drop and liquid holdup for two phase concurrent flow in packed beds." Chemical Engineering Science 32.5 (1977): 515-523.

15. Clements, L. D., and P. C. Schmidt. "Two‐phase pressure drop in cocurrent downflow in packed beds: Air‐silicone oil systems." AIChE Journal 26.2 (1980): 314-317.

16. Ellman, M. J., et al. "A new, improved pressure drop correlation for trickle-bed reactors." Tenth International Symposium on Chemical Reaction Engineering. 1988.

17. Larachi, F., et al. "Experimental study of a trickle-bed reactor operating at high pressure: two-phase pressure drop and liquid saturation." Chemical Engineering Science 46.5-6 (1991): 1233-1246.Dudukovic, Milorad P., and P. L. Mills. "Contacting and hydrodynamics in trickle-bed reactors." Encyclopedia of fluid mechanics 3 (1986): 969.

18. Gianetto, A. and V. Specchia, Absorption in packed towers with concurrent downward high-velocity flows: I-interfacial areas. Quaderni Dell'ingegnere Chimico Italiano, 1970. 6: p. 125-133.

19. Ergun, S. and A.A. Orning, Fluid flow through randomly packed columns and fluidized beds. Industrial & Engineering Chemistry, 1949. 41(6): p. 1179-1184

20. Eckert, J. S. "Trays and packings: selecting the proper distillation column packing." Chem. Eng. Prog. 66 (1970): 39-44.

21. Dudukovic, Milorad P., and P. L. Mills. "Contacting and hydrodynamics in trickle-bed reactors." Encyclopedia of fluid mechanics 3 (1986): 969.

22. Chou, T. S., F. L. Worley, and D. Luss. "Transition to pulsed flow in mixed-phase cocurrent downflow through a fixed bed." Industrial & Engineering Chemistry Process Design and Development 16.3 (1977): 424-427.

23. Sodhi, Vijay, and Ajay Bansal. "Analysis of Foaming Flow Instabilities for Dynamic Liquid Saturation in Trickle Bed Reactor." Analysis 356 (2011): 2736.

24. Charpentier, Jean‐Claude, and Michel Favier. "Some liquid holdup experimental data in trickle‐bed reactors for foaming and nonfoaming hydrocarbons." AIChE Journal 21.6 (1975): 1213-1218.

25. Sicardi, S., H. Gerhard, and H. Hoffmann. "Flow regime transition in trickle-bed reactors." The Chemical Engineering Journal 18.3 (1979): 173-182.

26. Ramachandran, P. A., M. P. Duduković, and P. L. Mills. "Recent advances in the analysis and design of trickle-bed reactors." Sadhana 10.1-2 (1987): 269.

27. Honda, Gregory S., et al. "Effects of prewetting on bubbly-and pulsing-flow regime transitions in trickle-bed reactors." Industrial & Engineering Chemistry Research 54.42 (2015): 10253-10259.

28. Larachi, F., et al. "Onset of pulsing in trickle beds: evaluation of current tools and state‐of‐the‐art correlation." The Canadian Journal of Chemical Engineering 77.4 (1999): 751-758.

29. Reinecke, Nicolas, and Dieter Mewes. "Investigation of the two-phase flow in trickle-bed reactors using capacitance tomography." Chemical Engineering Science 52.13 (1997): 2111-2127.

30. FUKUSHIMA, SUSUMU, and KATSUHIKO KUSAKA. "Interfacial area and boundary of hydrodynamic flow region in packed column with cocurrent downward flow." Journal of Chemical Engineering of Japan 10.6 (1977): 461-467.

31. Benkrid, K., S. Rode, and N. Midoux. "Prediction of pressure drop and liquid saturation in trickle-bed reactors operated in high interaction regimes." Chemical Engineering Science52.21-22 (1997): 4021-4032.

32. Honda, Gregory S., et al. "Effects of pre wetting on bubbly-and pulsing-flow regime transitions in trickle-bed reactors." Industrial & Engineering Chemistry Research 54.42 (2015): 10253-10259.

Sample Calculations for Single Phase Gas Flow

Density of air, ρG = 0.075 lbm/ft3

Viscosity of air, µG = 2.419 lbm/ft-hr

Column diameter, d = 0.25 ft

Cross sectional area, A = = 0.049 ft2

Flow rate of air, QG = 1.254 SCFM = 75.24 ft3/hr

Superficial velocity of air, uSG = QG/A = 1532.78 ft/hr

Diameter of particle, dp = 0.014328 ft

Porosity of column, Ɛ = 0.68

Sphericity, φ = 0.51

Substituting above inputs in Ergu’s equation,

Pressure Drop = 0.022 in H2O/ft

Sample Calculations for Charpentier-Favier Flow Regime Map

Flow rate of air, QG = 0.836 SCFM = 50.16 ft3/hr

Flow rate of water. QL =0.078 gpm = 0.6256 ft3/hr

Gas mass velocity, G = (QG )(ρ)/A = 76.6388 lbm/ft2-hr

Liquid mass velocity, L = (QL )(ρ)/A = 795.2927 lbm/ft2-hr

=1

Charpetier constant, λ =

=1

Charpentier constant, ψ =

Sample Calculations for Fukushima-Kusaka Flow Regime Map

Equivalent diameter of particle, deq = 0.27 ft

Viscosity of water, µL = 2.419 lbm/ft-hr

Reynolds number for air = ReG = (deq)(µG)/G = 47.03

Reynolds number for water = ReL = (deq)(µL)/L = 8.88

Table 6.2: Comparison of Experimental vs Predicted Pressure Drop

Table 6.2(a): For Liquid Mass Velocity, L = 1.079 kg/m2-s

Table 6.2(b): For Liquid Mass Velocity, L = 2.158 kg/m2-s

Table 7: Pressure Drop Correlation by Eckert