Metrological quantification of CO in biogas using laser absorption spectroscopy and gas chromatography

Abstract

Biogas has a vital role in the future market for renewable energy. When upgraded to biomethane, it can be injected into natural gas grids if the level of certain impurities complies with the specifications in EN16723. For some of these impurities, suitable measurement methods are lacking which hamper the quality control of biomethane to be injected into natural gas networks. Here, we report on the evaluation of three detection methods suitable for carbon monoxide (CO) in biogas and biomethane applications for which EN16723 specifies an upper limit of 0.1% (1000 µmol ⋅ mol−1). Two of these methods are based on laser absorption spectroscopy (LAS) and one on gas chromatography (GC). Both LAS spectrometers employ direct absorption spectroscopy and operate at 4.6 µm, probing a single CO absorption line in the fundamental CO band. One of them, direct tunable diode laser absorption spectroscopy (dTDLAS), is based on a new interband cascade laser specially designed for biogas and biomethane applications, while the other is based on quantum cascade laser absorption spectroscopy (QCLAS). The GC is equipped with two packed columns (HayeSep Q and molecular sieve 5A) and a thermal conductivity detector. CO amount fraction results in biogas matrices derived using these three measurement methods and are compared to amount fraction values of different, gravimetrically prepared reference gas standards of CO in biogas. These were used to validate the measurement capabilities. The measured CO amount fraction results from LAS and GC covered 10–30 000 µmol ⋅ mol−1 (system measurement ranges, LAS: 3–1000 µmol ⋅ mol−1, GC: 500–30 000 µmol ⋅ mol−1) and were in excellent agreement with the gravimetric values of the gas standards. At 400 µmol ⋅ mol−1, the guide to the expression of uncertainty in measurement compliant relative standard uncertainties of our calibration-free dTDLAS and the gas-calibrated QCLAS systems are estimated to be 1.4% versus 0.5%, respectively. The relative standard uncertainty of the GC CO measurements at 5075 µmol ⋅ mol−1 is 1.3%. This work demonstrates that, by means of GC and LAS, relative standard uncertainties of 1.4% and below can be reached for CO measurements in biogas and that cost-optimized calibration-free approaches not requiring frequent use of gas standards have become available.