Power Input Measurements in Stirred Bioreactors at Laboratory Scale

The overall goal of this protocol is to determine the power input in stirred laboratory bioreactors based on torque measurements. The main advantage of this technique is that commercially available sensors can be used for benchtop-scale investigations, which enables the user to perform measurements for scaling up studies. Visual demonstration of this technique is critical because the proper installation of the torque transducer is important for reliable measurements.

Setup the bioreactor elements and control apparatus on a lab bench. For this experiment, there is a two-liter mixing tank and the vessel holder. There is also a specially-designed holder for the agitator motor.

In addition to a mount for the motor, there is a mount for a torque meter. The special design includes an air bearing with an input for pressurized air. Use a computer and bioreactor control unit for data acquisition and control.

Also, provide a liquid with defined density and viscosity such as pure water or sucrose solution. The first step is to mount the agitator motor and the torque transducer. First place the brushless-servo agitator motor onto the top of the holder.

Secure the motor in place with screws, then orient the torque transducer properly and put it in position. Fix it on the holder using four screws. Fix the metal bellow coupling to connect the motor shaft to the drive shaft of the torque transducer.

Once this is done, get the agitator shaft. Install the impellers on the agitator shaft before proceeding. Next, insert the agitator shaft into the air bushing.

Fix the second metal bellow coupling to connect the agitator shaft to the torque transducer's measurement shaft. Place the holder with the installed torque transducer and the motor drive onto the bioreactor headplate. After selecting the appropriate sensor and test conditions the most critical step is the proper installation of the torque transducer and the impeller shaft that is held in place by an air bearing.

Now put a temperature sensor in the bioreactor. Connect the sensor to the A/D converter. Next, connect the motor drive to the control unit.

Prepare the air bearing by connecting tubing for pressurized air. Apply the appropriate air pressure to the bearing. The completed setup is represented in this schematic.

Computer software uses the control unit to set the motor speed and monitor the temperature. The computer collects data from the torque transducer through an analog-to-digital converter. Open the control software on the computer.

The opening screen is the Settings tab page. Go to the COM port dropdown menu and use it to select the correct port. Click the Connect button to initiate communication with the control unit.

On the same tab page, go to the folder symbol next to the Data File Location text field. Click on it, and a final dialogue box will open automatically. Browse the desired folder and type a file name for the data logging file.

Confirm by clicking the Okay button. The file path and name will be displayed in the Data File Location text field. Now go to the Recipes tab and bring that page to the foreground.

Find the fields labeled PhaseDuration and Agitator. Type the desired values for the recipe phase duration in minutes and the corresponding impeller speeds in the text fields. To save the recipe, click Save.

In the file dialogue, select the desired file path. Type the desired file name. Click Okay, the file dialogue will shut down automatically.

Open the data acquisition software to configure it. A green indicator next to a channel indicates it is initialized and active. Click the Live Update button.

One channel displays the torque signal, which should be appropriately zeroed. Select the tab for the Data Acquisition Job page to switch to the Data Acquisition tab page. There, under Sample Rate Groups, select a data acquisition rate of two hertz.

Under Start of Data Recording, select Immediately at Job Start. Next, go to Stop of Data Recording and select Duration. Define a time that is slightly longer than the measurement.

Now navigate to the Data Storage tab page. Go to the File Format dropdown menu and choose the option ASCII Channel Info. Set a file path for the measurement data in the corresponding text field.

To perform the experiment, return to the bioreactor. Use a funnel to add water or the sucrose solution to the vessel. The experiment is ready to be started.

In the data acquisition software, go to the Data Acquisition Job page. There, click Start, now switch to the control unit software to start the agitator. To use a recipe, go to the Recipe tab.

There, select a predefined recipe for the experiment. On the same page, click the Start button. The process will automatically start and the agitator speed is set according to the settings defined in the recipe.

When the experiment is completed, a window in the data acquisition software will open. Save the data by clicking the Save Data Now button. In the control unit software, a message box will appear indicating that the process has been finished.

Click Okay, go to the Recipe tab page and click the Stop button to finish the recipe. For the raw data processing, a Matlab script can be used, which is provided in the Supplement section of this manuscript. After opening the Matlab software, navigate to the working directory where the experimental data has been stored.

Type the command Torque Results in the command window and hit Enter. This starts the Matlab script that guides the user through the data processing. Select the data logging file for the torque sensor data and click Open.

Type the process parameters in the corresponding text fields when asked to do so. The results are shown graphically and the processed data are stored automatically in a new text file inside the working directory. These data show how the torque signal increased with every step increase in rotational speed.

After a transient peak, about one minute elapses before the measurements are quasi-stable. This depended on the rotational speed and the impeller. The dashed lines indicate a 5%confidence interval around the time-averaged values.

This plot compares the predicted and measured power input as a function of Reynolds number for different sucrose concentrations. The solid lines represent a model which assumes the power is proportional to the cube of the Reynolds number. The circles represent data.

The bioreactor was a two-liter working volume glass vessel with a single Rushton turbine impeller configuration. The triangles in this plot are data from a glass two-liter working volume bioreactor. The circles are data from a stainless steel 10-liter working volume bioreactor.

Both have a single Rushton turbine. The power numbers for fully turbulent conditions are equal at both scales. A comparison of data from one and two-liter working volume bioreactors with modified Rushton turbines and segment blade impellers shows that both have similar decreases in power numbers with increasing Reynolds numbers.

The offset between the two may be due to vessel or agitator geometries. After watching this video, you should have a good understanding of how to set up torque measurements to determine the power input in benchtop-scale bioreactors. Following this procedure, measurements can be performed in both single use and conventional systems.

Using automated measurements, the experimental effort can significantly be reduced and many data points can be obtained in a fast, reliable, and reproducible manner.