Genius process design, foolproof operating systems, complex intertwined unit operations, sophisticated pilot plants. All mean nothing if you cannot obtain a sample, analyze and quantify exactly what is in it. Analytical, from method development to sampling procedure, technique, precision, and how to treat the data, is paramount. You must be able to measure it. This is why it is important to bring engineers, chemists, and technicians all together – upfront — to address the analytical needs and hurdles of a project. Often times, unique solutions will have to be sought on how to quantify what is in that sample. MATRIC not only specializes in design and development of new process technology through pilot plant scale, but also has decades of experience in analytical techniques including capturing the sample and method development.
Analytical in Action
Recently, MATRIC worked on a process including water, multiple sulfur components, hydrocarbons, oxygen, and nitrogen, among many other species. It was important for this new process idea to be proven by sampling directly inside a 2000⁰C furnace at a specified point that represented a certain residence time in the reactor. The sample was saturated with water, had available oxygen which allowed further reaction in the sampling system, and it was subject to concentration changes due to equilibrium shifts as temperature was reduced through the sampling system.
It was challenging to construct an apparatus that would allow us to obtain a gas sample in a 2000⁰C furnace and to quench the sample quickly. The sampling system involved ceramic capillaries, pumps, traps, and a quenching mechanism. Furthermore, it was not easy to obtain the sample with a syringe; the syringe draw rate had to be less than the pump/capillary draw rate. If the syringe draw rate were greater, it was possible to back-flow gas from downstream and the sample would not be representative.
Given the sulfur and water components, we had additional obstacles. Water would interfere with the gas chromatograph (GC) analysis and sulfur would condense/freeze out in the lines, plugging them. We needed a trapping mechanism, and phosphorous pentoxide (P2O5) was chosen to trap both water and sulfur. A cartridge was designed, and the method of pulling the sample was modified. Another hardship using P2O5 is that after time the reaction with water formed phosphoric acid, which could absorb other gas components and affect the sample; thus the P2O5 had be replaced frequently. All of this had to be studied to create a method to obtain reproducible representative samples.
Furthermore, the GC analysis itself was complex. With so many components and a wide breadth of concentration ranges (ppm to percent levels), a specialized GC had to be configured with multiple paths, multiple detectors, and special materials of construction to analyze such a sample. Dozens of calibration gases had to be used due to the multitude of ranges plus the non-stability and reactive nature of some of the components. Multiple GC methods had to be developed to analyze different samples at different concentrations.
Then, after quantifying the GC results, the composite had to be corrected for the water and sulfur that were removed. Thus, we now had to devise a way to measure the amount of sulfur and the amount of water and then create a file that would correct the true sample data for the missing components.
In conclusion, the tasks of obtaining a representative reproducible sample cannot be taken lightly. All of the various aspects of obtaining quality data should be considered early in the pilot plant design effort.
— For more information on integrating MATRIC’s analytical capabilities with process design and pilot planting, contact David Statler.