This is Part Two of a three-part blog series. Here, we describe the use of Tunable Diode Laser Absorption Spectroscopy (TDLAS) for extractive measurements of moisture in the processes of ethylene and propylene production. The principal objective of the work reported here is to characterize the TDLAS-based extractive analyzer with an all-digital protocol for the modulation of the laser drive signal and the demodulation of the detector response.
The Instrument
The AMETEK 5100HD TDLAS is an extractive moisture (water vapor) analyzer that uses tunable diode laser technology to monitor moisture in different process gases including ethylene and propylene. To ensure continuous and representative measurements, and to reduce on-going maintenance requirements, no scrubber is required to account for potential interferences from background gases mentioned previously.
A schematic representation of the instrument is shown in Figure 1. The measurement of water vapor is performed with a DFB laser. The output of the laser is coupled into single-mode fiber, which in turn is connected to fiber-optic splitter. The splitter is used to divide the optical power in an optimized ratio for use in the sample and reference measurements, respectively. Gradient refractive index (GRIN) lenses are used to collimate the output of the single-mode fiber and direct the resulting beam through the sample and reference cells. A multi pass Herriott-type cell was used as a sample cell. The sample cell temperature was controlled with accuracy of +/- 0.1°C and could be set in the range of 60°C – 150°C by setting the temperature of the oven in the sample cell compartment. A reference cell containing water vapor was used to lock the output wavelength of the corresponding laser diodes. The sealed reference cells contained a known concentration of the moisture in nitrogen gas, which had no absorbance in the spectral range of interest.
Figure 1. Experimental setup for TDLAS measurement
The sample and reference cells each contained InGaAs-photodiode detectors, which were connected to a separate input electronics unit. With this configuration it is possible to make simultaneous measurements of unknown samples and known references, which are used to lock the output wavelength for laser. Two nested levels of temperature control are employed to maintain the operation of the DFB lasers at the proper wavelength.
Temperature and pressure in the sample cell are controlled with a thermocouple-based temperature probe and pressure transducer. Automatic correction of the water concentration value is provided by software for the variable temperature and pressure conditions. This correction procedure assumes linear correlation of spectral response (peak amplitude or integral intensity) on temperature and pressure in relatively narrow ranges of temperature and pressure variations.
Testing Methods
Water vapor was added to dried research grade ethylene and propylene gases. A programmable gas mixer was used to blend dried research grade ethylene and propylene with known moisture concentrations to simulate gases of varying composition. The moisture standard used was constructed in-house and consisted of a two-pressure moisture generator and a dilution system. The moisture concentrations produced by the two-pressure system are based only on the thermodynamics of the phase equilibrium established in its main saturation chamber [1-2]. While the two-pressure system is a highly accurate and reliable device, changing the moisture output over a wide concentration range can be a cumbersome and time-consuming process. These difficulties were overcome by the dilution system, which was able to rapidly provide water vapor mixing ratios over a wide dynamic range by diluting the “wet” output from the two-pressure generator with a dry diluent gas. Traceability of the moisture standard was established through monitoring the conditions of the phase equilibrium in the main saturation chamber (i.e., temperature and pressure), and through the controlled dilution stage (i.e., volumetric flow rates).
A calibration was developed for the TDLAS analyzer covering a moisture concentration range of 0-50 ppmv for both ethylene and propylene gases. A multivariate regression was used to establish the calibration model for moisture. Following the calibration, the performance of the analyzer was evaluated. The overall measurement accuracy for the analyte was established, and the corresponding limit of detection (LOD) was determined. LOD values were calculated as the concentration of the analyte required to give a signal equal to three standard deviations of the baseline. According to this definition, LOD characterizes the ability of the analyzer electronics, optics, and algorithm to distinguish the smallest detectable level of the analyte from the background spectra. For evaluation of the LOD, a moisture-free 2F spectrum was used to define the baseline. An estimate for the LOD of 0.4 ppmv was calculated for the measurements of moisture in ethylene and LOD of 0.3 ppmv was calculated for measurements of moisture in propylene.
To learn more about AMETEKs TDLAS solutions for this application, click here.
Read Part One of this series – The Importance of Measuring Water in Ethylene and Propylene Production
References
- Huang, Peter, “Thermodynamic Properties of Moist Air Containing 1000 to 5000 ppmv of Water Vapor”, Proceedings of the RL/NIST Workshop on Moisture Measurement and Control for Microelectronics, National Institute of Standards and Technology, Gaithersburg, Maryland, 1993, pp. 43-51.
- Huang, Peter, “New Equations for Water Vapor Pressure in the Temperature Range -100°C to 100°C For Use With The 1997 NIST/ASME Steam Tables”, International Symposium on Humidity and Moisture, Teddington, England, April, 1998.