| 초록 |
Monoethanolamine(MEA) has been well-known absorbent in CO2 capture process but in terms of long term operation stability, many problems originated form amine degradation have been raised recently. Thus, the development of online monitoring system is being required. Although FT-IR spectroscopy has many advantages for this system, but it isn't prefereed for quantitative analysis becuase of poor resolution. As a significant breakthorugh, this work presents the quantitative analysis of amine concentration and CO2 loading amount in MEA solutions using PLS regression which is for reducing enormous dimension of raw data. Standard solutions for calibration are prepared in the range of MEA 10 to 50wt% with lean to rich CO2 loading. As a result, predicted MEA concentration and CO2 loading amount are found in agreement with actual data shwoing over 0.998, R2. The presented models are expected to play a key role to estimate process condition for control and it also can be extended to amine mixture solvent with in situ condition as a future work. |
