Kenneth G. Crowther and Dr. Larry Baxter, Chemical Engineering

In the midst of industrial development we sometimes fail to foresee problems that may arise from current, seemingly economical, methods of production. One oversight in many industries ranging from Power to Caprolactam production is nitrogen oxide (NOX) emissions that damage our bodies and limit visibility. NOX are highly reactive molecules that are responsible for the formation of ground level ozone, acid aerosols, acid rain, and photochemical smog; all which pose a threat to the health of humans and other living things. This problem, however, is finally noticed. Government legislation is going into effect that is reducing the amounts of NOX emissions. The search is now on for ways to more economically and comprehensively reduce these emissions.

One method of NOX reduction is called selective catalytic reduction (SCR). This is essentially a method by which a catalyst is used to effectively reduce NOX into its diatomic elemental forms N2 and O2. However, this process requires a large capital investment and maintains expensive operating costs because of the little that is known about mechanism by which catalysts work and the ‘trial and error’ method by which companies select their catalyst. A project is underway at Brigham Young University to develop better understanding of catalyst to enable companies to choose correct catalyst, extend the life of catalyst, find more effective and cheaper methods to manufacture and reactivate effective catalyst.

To accomplish these tasks must initially extract various characterizing information from many catalysts, such as overall activation/deactivation time, activation energy, and surface activity during reaction. In order to characterize catalyst we have designed and built a lab that allows multiple unattended, safe runs in four parallel reactors for long periods of time. Each of the four simultaneous runs can accomplish different tasks. The software we developed will supervise and control the reactor system including measured quantities of reactants, reaction temperature and pressures, gas component concentration analysis, and data logging.

Here is a brief process description: Various cylinders of compressed gases with precise known concentrations are measured and mixed through a series of mass flow controllers (MFCs), and optionally through a bubbler for the addition of H2O, for the creation of a gas streams that will facilitate learning about reaction activity. A total of four gas streams can be created varying in both mass flow rate and concentration; these mixing parameters are specified on our machine interface through a series of programmed calculations. These four mixed gas streams are sent through their appropriate constant temperature plug flow reactors where the catalyst is located. The four streams that leave the reactors along with four feed streams that bypass the reactors are sent to a 10 channel motor-actuated valve, which selects one stream and for the analytical system the remaining streams from the reactor exit are sent to the vent while the remaining feed streams are plugged.

The analytical system consists of three main devices. The first analyzes NH3 and SO2 in an online, real-time method. These results are logged as the gas proceeds to the gas chromatograph and NOX analyzer. The NOX analyzer measures the concentration of NO, NO2, and total NOX. The gas chromatograph analyzes the concentrations of the remaining gases.

National Instruments (NI) FieldPoint (FP), a digital I/O for distributed network control, collects the signals from all reactor system. We specify data point acquisition to improve network traffic within the FP network. The FP network is connected to the computer via serial cable and all control programming is done in NI’s LabVIEW (LV) programming environment. Data transfer over a serial cable is sufficient because of the slow nature of all changes over the course of 36+ hour runs. FP is the data collection tool of choice because of its cost and reliability compared to other options. It is also very user-friendly and easy to configure. (However, I embarrassingly still found ways to confuse myself in the process of finally configuring correctly!)

The first big obstacle in completing the construction of the lab was the correct wiring of all control devices to allow for both manual and automatic/supervisory control. Although possibly not a big obstacle for some, for the less-experienced chemical engineer this was certainly a challenge accompanied by a great feeling of success upon completion of the wiring and the receipt of valid signals. The wiring and constructed control panel render a very professional and robust look to the reactor system.

The greatest enabling device of the lab is its computer control software. I programmed the human machine interface to automate the setting of reactor system runs and data logging. LV provides the most suitable environment because it’s highly compatible with FP and other NI hardware and is simple to read and learn. It is an object oriented programming environment where objects and classes are graphical images. Object properties and methods are wired to the object to define execution. Execution order is determined by the order in which the objects are then wired to each other. Because of its graphical nature it is ideal for a system where operating philosophy and knowledgeable operators change on a regular basis.

When the reactor system interface is run the initial screen is an image of the process flow. Customized images are used to represent various system controls and quickly communicate to the user required actions for desired responses. Buttons are available to run other programs that automate the setting of flow rates based on known cylinder gas concentrations, specified catalyst properties, and the type of test through the reactor. Once these items are specified and confirmed the computer sends a signal via serial cable to the FP terminal module, which in turn distributes the signal to its appropriate control. The MFCs, solenoid valves, and temperatures are set in an optimized order that minimizes oscillation in the flow system.

Although the LV code is simpler to understand than most, a 12 page document was created with all code specifications and logic of the reactor system interface. This document enables those who modify or operate the lab to navigate their way through the code in its original form.

Learning about technology implementation and the importance of project management is invaluable education. Moreover, I feel successful from having seen many completion stages of a reactor system that adds great long-term value to industry and our economy.