FERMENTATION EXPERIMENT

Monitoring and Control of a Fed-Batch Fermentation

Chemical Engineering Laboratories I, II Senior Lab

Our WebCams and Virtual Instruments use Windows 2000 that can serve only 10 clients at one time. For this reason we must limit access. If you would like a temporary account and password to inspect our teaching methods, please send your request by e-mail to Prof. Bungay

  • Temporary bypass of password for Virtual Instrument
  • WebCT log in
  • General backgound material about fermentation
  • Interim Report Instructions
  •  Student term project on brewing that is helpful
  • Feedback from previous students
  •  Financial support for this experiment
  •  Developer credit 

    • You may use any of this text or the figures in your report, but modify to match your own style. Load from the Public Directory /dept/chem-eng/Biotech-Environ/FermentationLab. The sketch of the apparatus is provided in 3 different graphics formats.

    1. INTRODUCTION

    Fermentation is widely used in the pharmaceutical industry, in the food and beverage industry, and is similar to biological wastewater treatment. This simple, yeast-based, fermentation experiment is used in other chemical engineering departments and gives students an introduction to bioprocessing and the monitoring and control of a small-scale fed-batch bioreactor. It was adapted from an experiment described in Chemical Engineering Education (Teixeira et al. 1992). The experiment addresses basic fermentation operation as well as computer-based data acquisition and control. Fed-batch fermentations are commonly employed in the bioprocess industry in situations encountering either substrate or product inhibition of microbial growth; computer control of feeding schedules is critical to ensure maximum fermenter productivity. The production of ethanol by baker's yeast growing on glucose provides a simple system for a 24-hour period. The objective is to determine the productivity of the fermentation process for a given feeding strategy and to compare with model-based productivity predictions. Computer-based data acquisition and control allow an unlimited range of feeding strategies and for a run that is mostly unattended.

    2. BACKGROUND

    In alcoholic fermentation, using Saccharomyces cerevisiae, the stoichiometry of glucose conversion to ethanol and CO2 is given by Eqn. 1

     C6H12O6 ---------> 2 C2H5OH   +   2 CO2        (1)

     (glucose)                  (ethanol)       (carbon dioxide)

     This equation shows that 0.511 g of ethanol and 0.489 g of CO2 are produced from each gram of glucose reacted. As some of the glucose is used for the production and synthesis of secondary products and cell components, the real stoichiometric yields are 90-95% of these values. Accepting these approximate glucose conversion yields, it is possible to follow the kinetics of a fermentation by measuring the mass of CO2 released. Another important aspect of alcoholic fermentation that employs glucose as substrate and yeast as microorganism is the inhibition of glucose consumption at high glucose concentrations. To avoid this, the fed-batch fermentation is preferred. In this process, fermentation is started batchwise with a low glucose concentration. When all the initial substrate is consumed, a new addition restores the glucose concentration to just below the point of where it produces inhibitory effects. It may be said that, by operating in this way, the fed-batch fermentation is a sequence of batch fermentations of increasing volumes. The apparatus for this experiment (shown in Figure 1) has a magnetically stirred 2-liter flask. Fresh fermentation medium is supplied from another 2-liter flask via a computer controlled peristaltic pump. Both the fermentation and medium flasks are on top-loading balances equipped with serial computer interfaces to enable on-line calculation of mass balances and productivity for the fermentation. The fermentation flask is vented through a meter to provide an independent measure of the CO2 released during ethanol production. A Macintosh computer controls the pump and collects data.
     

    Figure 1. Apparatus


     
     

    3. ABOUT THE EXPERIMENT

    Fed-batch operation can be carried out under different strategies. The baseline experiment is based on total consumption of glucose in each batch. With the experimental setup as described, alternative feeding patterns (namely constant feed of nutrient or stepwise feed of nutrients) can be implemented readily. The algorithm for the control of the whole operation is straightforward. By constantly monitoring the total mass balance (fluids and the amount of CO2 released), it is possible to estimate glucose concentration, Gr, in the medium. The amount of fresh medium, Mf, to be pumped in order to raise the glucose concentration up to a setpoint, Gl, is calculated and the task is automatically implemented by simple on-off control of the pump. The procedure is stopped optionally when a time limit is observed or when the total volume VT set for "added fresh medium" is reached. Our version of the experiment is different from the one mentioned in the paper in a few minor ways such as our use of a mass flow meter on the exhaust gas. One big difference is autoclaving of the medium to insure sterilization. Typical conditions are 121 C for 30 minutes. Although RPI does have an autoclave on the 4th floor of Ricketts Building, it is getting old and temperamental. We will not use it. Before the discovery of microorganisms, some commercial processes such as the production of alcoholic beverages (where the lowering of pH and the accumulation of alcohol by yeast discourage other organisms) had been doing well for centuries. When a goodly number of yeast are inoculated, the process will usually not encounter contamination by other organisms. When contamination is a problem, there should be air breaks in the lines that prevent a direct liquid path for microorganisms to travel. These are also used in the line from the reservoir, but yeast are big and slow and will not grow back into the feed medium in a short time.

     We are going to rely on massive inoculation with Baker's yeast and will circumvent sterilization. Usually, the feed reservoir must be protected from contamination. As medium leaves, air enters. This air must pass through dense fibrous material or a membrane that will filter out microorganisms. The filter itself is sterilized as the reservoir and its connections are autoclaved. Organisms in the fermenter can grow back through the tubing. Yeast are big and slow and might not reach the reservoir in 24 hours. Bacteria however can grow back through the tubing in a few hours. An air break device prevents this. Again, since we are running our experiment for only 24 hours, we will not have to install an air break or a steril filter. Each group will conduct one experiment. You may devise an original experiment of your own. You can earn bonus points if you help the next group to run the experiment. This may be advice before the laboratory session or you may find a few moments to work with them during the session. You only earn points if they include a brief acknowledgment when they write their report.
     

    4. EXPERIMENTAL

    4.1. Prelab

    Before coming to the prelab session, you should read this paper. During the prelab session, become familiar with the apparatus. Follow the lines for the medium from its reservoir to the bioreactor. It is pumped through tubing through the peristaltic pump and into the reactor. Follow the lines for evolved gas. During the run with yeast, CO2 is evolved, and flows through a column filled with desiccant and through the flow meter. Take a look at the software. Pushing the upper right button on the key board boots the computer. Double click on the "Feedback of Balance Reading.vi" icon. The program will start up automatically. Do not actually run the software. Notice the different colors used. Blue generally denotes values that have to be entered by you. Sometimes you have to choose among several options, rather than input values. Check out the different options for the "Feed Back Variable" and "Totalizer mode". Under no circumstances should one taste or drink the fermented products! It is strictly forbidden to remove product from the lab.

     Answer the following questions:

    1. What are the parameters you will have to input to the program?
    2. How will you decide on these values and settings?
    3. Are all chemicals required to run the experiment available in sufficient amounts?
    4. What is the function of the tubing that goes from the reactor to the feed vessel but does not go through the pump?
    5. What type of disk do you need to bring to the lab session?
    6. Is sufficient tubing in stock to replace old tubing?
    7. How will you divide up the work among the group members?
     Write the answers to these questions into your lab journal.

     Discuss the answers with your TA.

    Calibration

    You should not assume that everything works. The T.A.'s are responsible for the apparatus, but you should check at least one feature to verify its accuracy. Please do one of the following:
  • Devise a method for collecting gas after it leaves the totalizer. Correlate the collected volume with the totalizer reading. Click to see a  possible set up.
  • Use the computer to activate the feed pump. Collect water from the pump in a graduated cylinder and check volume versus time to calibrate the pump's rate.
  • You may select another feature and devise a means for testing it.

    • 4.2. Lab Session

    Division of labor is the key to a successful experiment. Make sure everyone has a certain assignment for which he/she is responsible. We will first start up the computer and then the remaining equipment. All data will be displayed on the monitor and saved to floppy disk. You must supply your own disk! There will be no disks available from the TA. Always be aware that the balances are very fragile. You must not overload them, or expose them to shocks or liquids. Follow the step-by-step guide to start up the experiment. All equipment has to be powered down and cleaned before you start.

     1) Start up the computer: Push the upper right button on the key board. Let the machine power up completely before executing the next step.

     2) Start up the software: Double click on the "Feedback of Balance Reading.vi" icon. Let the software start up completely before going to the next step. This is important to start up the serial communication properly.

     3) Power up the remaining equipment: Push the switch on the panel. The pump has to show "P?1" The balances go through a self-test routine after the calibration bar is pressed. The indicator for temperature and gas flow rate will become illuminated. Open the valves upstream and downstream of the gas flow meter.

     4) Install fresh tubing: This will be done by the TA. You must discuss with your TA whether fresh tubing should be installed. You can find a part list for number and length of the required tubing at the experiment. Make sure the tubing is connected exactly as it was before. The connections to the vessels must be horizontal loops. Vertical loops will add weight to the balance reading because the previously fresh tubing relaxes during the course of the experiment. If in doubt, you may copy the layout from the other setup. The pump must have a size 16 tubing.

     5) Prepare the medium: You will need 2 liters of the following medium in Table 1.

    TABLE 1      Medium composition (per liter of medium)

    KH2PO4 5 g
    (NH4)2SO4 2 g
    Carbon source (glucose) 50 g
    MgSO4.7H2O   0.4 g 
    Yeast extract 1 g
    Deionized water to one liter mark

    The amounts mentioned above will be good for 1 liter of medium. Deionized water can be found on the 3rd floor of Ricketts Building. The pH is adjusted to 5.5 with H3PO4. It is good practice to record the amounts actually added into your lab note book. Remove the feed flask from the balance! Make sure the rubber stopper properly seals the feed flask. Place the feed vessel back on the PM4000 balance.

     6) Inoculation: Add 3g of freeze-dried yeast into the reactor. Remove the reactor from the balance! Be sure the rubber stopper properly seals the reaction vessel. Place the reactor back on the PM6100 balance. Center the stirrer bar on the magnetic stirrer unit.

     7) Align the vessels: The vessels have to be in the center of the weighing pan (not the blue spill protection). The tubing leading to the vessels should not be nicked or under stress. Level the balances. A spirit level can be found at the back of the balances. Zero the balances.

     8) Zero the gas flow meter: Pushing the [VIEW] buttom on the totalizer key board lets you select different readings on the display: r = gas evolution rate, P = peak point, u = valley point, no indicator = total gas volume. Pressing the [ENTER] key while the total gas volume is displayed will zero it. Only zeroing the total gas volume will influence the data send to the computer. Do not zero the total gas volume once the program is running. This will cause an error.

     9) Enter the parameters into the program: Use a filename like: MyDisk:MyFile.dat. Do not save data to the hard disk since the device would suffer from constantly being accessed. It is cheaper to replace a floppy disk drive. Suggestion: Do not go above 40% nor below 10% of the glucose concentration in the feed for the setpoint. It is recommended using 0.025 g/ ml as setpoint for glucose. You have only one shot at this. Make sure you set all the parameters correctly. It is also recommended selecting the gas evolution rate for the totalizer reading.

     10) Zero balances: There must be clearance between the balance body and the blue spill protection. Again, check zero of the totalizer for the gas flow meter and the balances. You must not touch anything once the program is running!

     11) Start the program: Click with the mouse on the arrow button in the upper left corner of the window. The running program will initialize the pump and change its display to "P01''. Within 10 minutes, the pump will feed about 300 ml of medium into the reactor.

     12) Start the stirrer: Wait until the pump stopped. Do not exceed 1/2 of the maximum stirrer speed. You are in big trouble, once the stirrer spins out of control. Do not touch the electric connection of the stirrer since this could change the zero point of the balance.

     13) Operation: Observe the reactor for one hour. The pump must come on for 10 minutes after two hours, and add some more medium to the reactor. The flow meter should start recording a gas flow.

     14) Before leaving the lab: Setup a time with your TA to return the next day. The next day

     15) Initial observations: The feed flask should now be empty. The totalizer should show a certain total gas volume.

     16) Record data: Record the elapsed time, the total gas volume, the current gas flow rate, the temperature reading and the readings of the balances in your lab note book.

     17) Stop the program: Click the red stop button on the screen. Print the front panel by pressing the [Apple key] and the [P] simultaneously.

     18) Close the application program: Do not save the program.

     19) Check Data: Double click on the disk icon and verify that the collected data are really on the disk. The data are saved in a Macintosh spreadsheet format. The individual columns are: Column# = time in [min], PM4000 [g], PM6100 [g], Pump [ml/min], Totalizer [ml/min, or ml; as selected], Temperature [C], Go (Glucose conc. at begin of batch), Gt (estimated glucose conc. in reactor), Mco2 (Mass of CO2 expected to evolve during current batch). The installed software allows saving the data to a MS-DOS disk or a Macintosh disk.

     20) Shut down: Clean all equipment and power everything down. Close the valves upstream and downstream of the gas flow meter. It can damage the flow meter if you forget to close the valves! Leave the old tubing in place.
    FLUSH THE TUBING WITH TAP WATER TO REMOVE ORGANISMS AND CRUD.
     
     

    4.3. Interim Meeting

    Bungay has extensive travel committments for the Spring semester of 1999. Instead of meeting with him for the interim report,
    please get your instructions from the web page.
     Click here for instructions for interim.

    Points to consider

     Is this a real fed-batch?

     Compare your experimental setup with the theory of fed-batch fermentation.

     Try to explain your data. You may try to simulate your experiment.

     What is the state of the yeast before inoculation? Is it alive, is it dead?

     What is the difference between the freeze-dried yeast and yeast extract?

     The software will save the following values: Column# = elapsed time [min], PM4000 [g], PM6100 [g], Pump [ml/min], Totalizer [ml/min, or ml; as selected], Temperature [C], Go (Glucose conc. at begin of batch), Gt (estimated glucose conc. in reactor), Mco2 (Mass of CO2 expected to evolve during current batch). Data points are saved at intervals of 1 minute.
     

    FINAL REPORT

    Questions to be answered: Include an explanation of the purpose of the different ingredients. Plot and explain the data. What do they mean. Back the data up through theory and simulation. Consult the literature at the end of the handout. Each page should have the initial of the writer. Do not copy an old FERMENTATION REPORT.

     6. GRADING POLICY

    The scores for presentations are covered in the web page about the interim. In most cases, all members of a team get the same grade. However, a team member who has contributed much better or much worse than the others may get a different grade.

    INCLUDING MATERIAL FROM AN OLD REPORT MAY CAUSE YOUR GRADE TO BE "F".
     
     

    7. LITERATURE ON RESERVE IN THE LIBRARY

    Teixeira, J.A., Sousa, M.L., deAzevedo S.F. and Mota M. "Monitoring and Control of a Fed-Batch Fermenter", Chem. Engr. Education Spring 1992, pp. 94-103.

     Bailey, H.R. and Ollis, D.F. (1986) "Biochemical Engineering Fundamentals". McGraw-Hill, Inc.

     Bungay, H.R. (1992). "Basic Biochemical Engineering". BiLine Associates, Troy NY.

     Video. "The Fermentation Lab". Department of Chemical Engineering. Rensselaer Polytechnic Institute, Troy NY.
     
     

    7. QUESTIONNAIRE

    Setting up this lab required a lot of effort. We will constantly update the experiment as well as the written material. Please answer the following questions and submit them to your professor when you meet to discuss your graded report.

     Rate the following items, 10 points being the maximum, X meaning no opinion:
     

    Quality of the experiment compared to other senior lab experiments (0 to 10)
    Quality of the experiment on an absolute basis -
    Enjoyment of experiment -
    New knowledge gained -
    Clarity of instructions -
    User-friendliness of software -
    Quality of instruction -
    Other sources of information (e.g., reserve literature) -

     Answer the following questions on a separate sheet of paper:

     1) What are positive aspects of the lab handout?

     2) How could the lab handout be improved?

        What would you like to be added to it?

         What should be removed?

        What needs to be corrected?

     3) What are positive aspects of the experimental setup?

     4) What are negative features of the setup?

         Is anything broken?

         How user-friendly is the setup?

         Do you think there is a fundamental bug in the setup?

     5) How could the hardware be improved?

     6) Was your TA competent to answer your questions?

     8) Did you understand how to obtain a good grade?

     9) Please feel free to add anything, negative or positive.