3. Best practices for the measurement of 2-amino-2- methyl-1-propanol, piperazine and their degradation products in amine plant emissions (2021)

Baptiste Languille^a, Audun Drageset^b, Tomas Mikoviny^a, Erika Zardin^a, Christophe Benquet^b, Øyvind Ullestad^b, Magnus Aronson^b, Eirik Romslo Kleppe^b, Armin Wisthaler^a*

^aUniversity of Oslo, Department of Chemistry, P.O. Box 1033 Blindern, 0351 Oslo, Norway ^bTechnology Centre Mongstad (TCM), 5954 Mongstad, Norway ^cTotal E&P Norge, Finnestadveien 44, Dusavik, 4029 Stavanger, Norway ^dEquinor ASA, PO Box 8500, 4035 Stavanger, Norway *Corresponding author

Original paper available here (ssrn.com)

Abstract

We herein present the chemical-analytical setup used to measure atmospheric emissions of amines and amine degradation products from an amine-based post-combustion carbon capture plant. The emission measurements were carried out at the Technology Centre Mongstad (TCM) in Norway, in the frame of the ALIGN-CCUS campaign from September 2019 to January 2020, when the amine plant was operated with the CESAR 1 solvent. This advanced solvent is an aqueous solution of 2-amino-2-methyl-1-propanol (AMP) and piperazine (PZ). Four chemical-analytical techniques were deployed for characterizing emission of AMP, PZ and their degradation products: online Fourier Transform Infrared (FTIR) Spectroscopy, online Proton-Transfer-Reaction Time-of-Flight Mass Spectrometry (PTR-TOF-MS), online Proton-Transfer-Reaction Quadrupole Mass Spectrometry (PTR-QMS), as well as manual impinger sampling followed by offline Ion Chromatography Mass Spectrometry (IC-MS) analysis. AMP was detected by all four methods, with the results being in reasonably good agreement. PZ was detected by PTR-TOF-MS, PTR-QMS and IC-MS, but because of the low emission levels (single-digit ppb) the latter two methods suffered from a positive bias (due to an interfering compound) and a large measurement uncertainty, respectively. 17 amine degradation products were only detected by the PTR- ToF-MS analyzer. We present exemplary results from the emission measurements carried out during the ALIGN-CCUS 2019- 2020 campaign and share some of the lessons learned from this exercise.

1. Introduction

Amine-based post-combustion carbon capture (PCCC) is the technologically most mature solution for removing carbon dioxide (CO2) from a flue gas stream. The reduction in CO₂ emissions does, however, come at the expense of amine emissions to the atmosphere [1]. Atmospheric oxidation processes may partially transform the emitted amines into nitrosamines and nitramines [2], which are substances with known carcinogenic or potential carcinogenic properties. Based on a conservative risk analysis, inhalation exposure to the sum of PCCC-derived nitrosamines and nitramines should be kept below an annual average concentration of only 0.3 ng m-3 [3] for the general public. Since it is not possible to monitor such low concentrations in the atmosphere, exposure to nitrosamines and nitramines is calculated taking into account emission rates of amines, as well as the dispersion and atmospheric processing of emitted amines. Amine emission rates need to be measured, but their low concentrations (typically 0.1-10 ppm) and the high humidity levels in the treated flue gas make this a challenging analytical effort.

The Technology Centre Mongstad (TCM) (www.tcmda.com) in Norway is one of the world’s leading facilities for testing and improving CO₂ capture technologies. TCM’s tasks also include the validation and optimization of emission sampling and measurement techniques. TCM has put considerable efforts into the characterization of atmospheric emissions and nowadays routinely monitors amines in the absorber effluent stream using a variety of analytical methods [4]. These include online Fourier Transform Infrared (FTIR) Spectroscopy, online Proton-Transfer-Reaction Time-of-Flight Mass Spectrometry (PTR-TOF-MS) [5], online Proton-Transfer-Reaction Quadrupole Mass Spectrometry (PTR-QMS) [6], as well as manual impinger sampling followed by offline Ion Chromatography Mass Spectrometry (IC-MS) analysis [4]. While amine emission data are routinely reported to the authorities, only few results have hitherto been publicly disclosed because the solvent composition is, in most cases, confidential.

ALIGN-CCUS (Accelerating Low CarboN Industrial Growth through Carbon Capture Utilisation and Storage) is a project financed through the first ERA-NET Co-fund ACT program funded by nine European countries and the European Union Horizon 2020 program. The ALIGN consortium includes 31 partners from industry, research and academia and has considerable involvement of industrial companies and an enterprise organization. The ALIGN- CCUS project aims at accelerating the transition of current industry and power sectors into a future of continued economic activity and low-carbon emissions, in which carbon capture, utilization and storage (CCUS) plays an essential role. For optimizing and reducing costs of PCCC, the ALIGN consortium has implemented test programs at four different pilot plants and testing facilities including TCM, the SINTEF pilot rig at Tiller/Trondheim in Norway, RWE’s Coal Innovation Centre at Niederaussem in Germany, and the Pilot-scale Advanced CO₂ Capture Technology (PACT) facilities in Sheffield in the United Kingdom. TCM’s contribution to the ALIGN-CCUS project is aimed at bridging knowledge gaps as well as reducing HSE, technical and financial risks of technology upscaling.

In the ALIGN-CCUS 2019-2020 campaign at TCM, flue gas from Equinor’s combined cycle gas turbine (CCGT) plant in Mongstad was treated with the aqueous 2-amino-2-methyl-1-propanol (AMP) / piperazine (PZ) (CESAR 1) aqueous solvent. The CESAR 1 solvent was selected due to its lower energy consumption and higher stability as compared to monoethanolamine (MEA). Its promising characteristics have made CESAR 1 the new benchmark IEAGHG amine solvent. Emission control and monitoring as well as solvent consumption were among the study topics of the ALIGN-CCUS 2019-2020 campaign. The main goal of the campaign was to demonstrate that this advanced amine solvent can be used at a large scale and with a real flue gas. Results from the ALIGN-CCUS project are in the open domain, thus offering the opportunity to show, for the first time, qualitative and quantitative results on the measurement of AMP and PZ in the absorber effluent. Emission data will be presented in a companion paper.

In this work, we will present the chemical-analytical methods used for carrying out the emission measurements during the ALIGN-CCUS 2019-2020 campaign and present some of the results along with the lessons learned from this exercise.

2. Methods

2.1 Overview of the campaign and overall system description

The ALIGN-CCUS campaign at TCM was conducted from September 12th 2019 to January 10th 2020. As already stated above, flue gas from a CCGT plant was treated with the CESAR 1 solvent. Figure 1 is a sketch of the analytical set-up used for emission monitoring. Table 1 gives an overview of the analytical methods used for emission monitoring, their time of deployment, the measurement frequency and the compounds measured.

Table 1. Overview of the analytical methods used for emission monitoring, their time of deployment, the measurement frequency and the compounds measured.

Fig. 1. Process flow diagram of emission sampling of absorber outlet. (VP: vaporizer).

2.2 Proton-Transfer-Reaction Mass Spectrometry (PTR-MS)

PTR-MS is a well-established technique for monitoring organic trace gases online (i.e., without sample pre- treatment), at a typical measurement frequency of 1 Hz and down to ppt levels [7]. In PTR-MS, the gas to be analyzed is introduced into a low-pressure reaction cell wherein organic molecules are softly ionized via gas-phase proton transfer reactions with H3O⁺ ions [8]. The protonated analyte molecules are detected in a quadrupole mass spectrometer (QMS) or a time-of-flight mass spectrometer (TOF-MS). [5] have described the use of PTR-MS for monitoring amines in the treated flue gas emanating from an amine-based PCCC plant. For the work presented herein, we used two different PTR-MS instruments, a PTR-TOF-MS analyzer (model PTR-TOF 8000; Ionicon Analytik, Innsbruck, Austria) and a PTR-QMS instrument (model QMS 300; Ionicon Analytik, Innsbruck, Austria). The latter is a cheaper and easier to use instrument version, which was conceived for industrial monitoring purposes. Both instruments were placed in an analyzer house on top of the absorption tower, and connected to the stack via a ~10 m long heated (120 °C) sampling line made of SilcoNert2000®/Sulfinert. The subsampling line of the PTR-QMS instrument included a vaporizer (VP-QMS) for evaporating potential mist particles. The operating parameters of both PTR-MS instruments were as follow: drift tube pressure 2 mbar, drift tube temperature 120 °C, drift tube voltage 500 V. The resulting reduced electric field strength (E/N) was ~150 Td (1 Td = 10-17 V.cm2).

Both PTR-MS instruments were calibrated for AMP and PZ using a HovaCAL® calibration gas generator (model

312-MF; IAS GmbH, Oberursel, Germany). A quantitative aqueous amine solution (AMP, PZ) was prepared by TCM’s analytical laboratory. This solution was evaporated at 180 °C in the HovaPOR evaporator (IAS GmbH, Oberursel, Germany) under a pure nitrogen atmosphere. By increasing the calibration solution flow into the evaporator, the humidity increases in the resulting calibration gas. This allowed us to study the response of the two PTR-MS instruments at different humidity levels.

2.3 Fourier Transform Infrared (FTIR) Spectroscopy

The FTIR instrument (Analect 5000; Schneider Electric) was located in the analyzer house at the bottom of the absorption tower. It was connected to the stack via a 100 m long sampling line made of SilcoNert2000®/Sulfinert. The flue gas was extracted from the stack at a fixed flow rate using a fast flow loop system mimicking isokinetic conditions [9]. The fast flow loop system included a vaporizer (VP-FTIR) for evaporating potential mist particles and a heated filter (HF). The temperatures of these units were adjusted to minimize the thermal degradation of analytes. The FTIR instrument was calibrated using the same HovaCAL® calibration gas generator as described above. Principal component regression (PCR) models were used for extracting CO₂, H₂O, NO, NO₂, SO₂, NH₃, acetaldehyde, formaldehyde, AMP and PZ from the FTIR spectra.

2.4 Impinger sampling followed by Ion Chromatography Mass Spectrometry (IC-MS) analysis

Samples were extracted isokinetically and collected in a standard impinger sampling train including a condensation flask, a filter flask and three absorption flasks. Typical sampling times were 1-2 hours. The samples were analyzed in TCM’s laboratory using a Dionex Integrion HPIC System (model ICS-5000, Thermo Fisher Scientific) which included an IonPac CS19 column and an IonPac CG19 guard column.

3. Results and Discussion

3.1 Measurement of AMP emissions and encountered difficulties

The only compound detected by all four measurement techniques was AMP. The results obtained with the four methods were in good agreement. As shown in Figure 2, the distribution of the measured AMP volume mixing ratios was very similar, with the PTR-TOF-MS, the PTR-QMS and the FTIR instrument measuring median AMP levels of 433, 400 and 388 ppb, respectively. Mean measured AMP levels were 592, 549 and 503 ppb, respectively. Figure 3 shows an exemplary time series of AMP as recorded over a period of ten days. Two impinger samples were collected during this period. The measurements are again in good agreement, the only major discrepancy being the systematically lower levels measured by the FTIR instrument during periods with enhanced emissions. Table 2 shows a systematic comparison for all data collected during twelve impinger sampling periods, with sample collection times ranging from 42 to 136 minutes. The overall agreement was good, even though in some case large discrepancies were observed. It is not clear why the level of agreement varied between the samples. Future work is needed for addressing this issue.

Fig. 2. Distribution of AMP volume mixing ratios as measured in the emitted flue gas by the PTR-TOF-MS, PTR-QMS and FTIR instruments, respectively, over the duration of the whole 2019-2020 ALIGN-CCUS campaign. Boxes represent 25th and 75th percentiles, the black line is the median, and the black dot is the mean. Whiskers represent 5th and 95th percentiles.

Fig. 3. Exemplary time series of AMP volume mixing ratios as measured in the emitted flue gas by the PTR-TOF-MS, PTR-QMS and FTIR instruments, respectively, in the period from October 9 to October 19, 2019. Also included are the two discrete measurement points obtained via impinger sampling followed by IC-MS analysis.

Table 2. AMP mixing ratios as measured by all four measurement techniques during the impinger sampling periods. Standard deviations are given in brackets.

A number of problems and difficulties were observed for all of the measurement methods used during the ALIGN- CCUS campaign. A forthcoming paper will address these issues in more detail and only a brief outline is given here.

The FTIR analyzer was sampling through a 100 m long sampling line. Long sampling lines are known to require long conditioning times for amines, meaning that peak levels were dampened and thus underestimated. In addition, we found that a stainless steel (SS) line originally connected to the FTIR instrument was corroded. The SS line was replaced with a sampling line made of polytetrafluoroethylene (PTFE) during the campaign. AMP degrades when in contact with a hot SS surface. It was thus important to reduce the temperature of the vaporizer (see Figure 1) to a level at which no AMP decomposition is observed. Finally, it should be noted that the AMP concentrations in the treated flue gas were close to the detection limit of the FTIR instrument, which made the measurements with this instrument less accurate.

The response of PTR-MS instrument to amines usually does not depend on the humidity of the sample matrix. This was not the case for AMP. The response of the PTR-QMS and the PTR-TOF-MS instruments varied by a factor of 5 between dry and humid conditions. For obtaining accurate results, it was essential to calibrate both instruments over the full range of humidity levels observed in the flue gas stack. Also, the reduction, processing and analysis of the PTR-TOF-MS data was highly complex and time-consuming and required a full time engagement of an expert scientist.

Impinger sampling followed by ICMS analysis is time consuming and laborious. Only 12 samples were thus collected and analysed over a period of the whole campaign. In addition, the measurement uncertainty was high at low ppb levels. Only AMP (typically >100 ppb) was thus quantitatively reported from the IC-MS analysis.

Finally, it should be noted, that during the ALIGN-CCUS campaign no mist was observed in the treated flue gas. Measurement problems that occurred in the presence of mist will be presented in a separate study.

3.2 Measurement of PZ

PZ was only present at low ppb levels in the treated flue gas, meaning that only the PTR-TOF-MS instrument was able to quantitatively detect it. PZ also exhibited a humidity-dependent response. It was thus necessary to carry out a humidity-dependent calibration of the PTR-TOF-MS instrument. The PTR-QMS instrument also detected an ion signal at m/z 87, which corresponds to protonated PZ. An intercomparison with the PTR-TOF-MS data revealed that a second peak at m/z 87 (which can only be resolved by the PTR-TOF-MS and not by the PTR-QMS) becomes quantitatively relevant at low ppb levels. Since these were typical emission levels during the ALIGN-CCUS 2019- 2020 campaign at TCM, it was not possible to quantitatively measure PZ with the PTR-QMS instrument. Impinger sampling followed by offline Ion Chromatography Mass Spectrometry (IC-MS) analysis suffered from a high measurement uncertainty at low ppb levels, meaning that not quantitative PZ data were reported during the ALIGN- CCUS 2019-2020 campaign.

3.3 Measurement of solvent degradation products

The PTR-TOF-MS analyzer also detected a series of solvent degradation products. The high mass accuracy (typically <10 ppm) of the measurement combined with an isotopic pattern analysis allowed us to identify the elemental composition (i.e., the molecular sum formula) of the decomposition products. Table 3 lists the m/z of the main signals detected in the flue gas, the assigned molecular sum formula and the name of the compound that we assigned (or tentatively assigned) to this signal based on previous work specified in the fourth column. An identification of compounds with a higher degree of confidence warrants complementary analyses by GC-MS or LC- MS.

Table 3. Main *m/z* signals detected by the PTR-TOF-MS instrument in the flue gas that was emitted to the atmosphere when the amine plant was operated with the CESAR 1 solvent. The molecular sum formula was assigned unambiguously; the assignment to a specific chemical substance was based on chemical plausibility and literature data specified in the “References” column. Assignments marked with an asterisk are tentative.

Conclusions

Amine-based PCCC plants emit a variety of organic chemicals into the atmosphere. Treatment of CCGT flue gas with the CESAR 1 solvent generates ppb-to-ppm levels of AMP in the emission stream. Our study has shown that such emissions can be measured with sophisticated chemical-analytical techniques (PTR-TOF-MS; impinger sampling/IC-MS) but also with less demanding methods that are suitable for routine industrial monitoring purposes (PTR-QMS, FT-IR). Special care must be taken to avoid losses (due to adsorption and/or thermal decomposition of AMP) in the inlet system and to calibrate the online analyzers. Only the PTR-ToF-MS instrument was capable of detecting PZ at low ppb levels, which were typical emission levels during the ALIGN-CCUS campaign. The PTR- TOF-MS analyser is also capable of detecting amine degradation products, 17 of which were observed at significant levels in the flue gas after treatment with the CESAR 1 solvent. The unambiguous identification of these degradation products would, however, require complementary analyses using highly specific offline GC/LC-MSn methods.

While it seems unfeasible to make similar chemical-analytical efforts for emission characterization at each amine- based PCCC plant, TCM and its partners provide the know-how and infrastructure to characterize the emission profile of new solvents.

Acknowledgements

ACT ALIGN-CCUS Project No 271501. This project has received funding from RVO (NL), FZJ/PtJ (DE), Gassnova (NO), UEFISCDI (RO), BEIS (UK) and is co-funded by the European Commission under the Horizon 2020 programme ACT, Grant Agreement No 691712 (www.alignccus.eu).

The authors gratefully acknowledge the staff of TCM DA, Gassnova, Equinor, Shell and TotalEnergies for their contribution and work at the TCM DA facility. The authors also gratefully acknowledge Gassnova, Equinor, Shell, and Total as the owners of TCM DA for their financial support and contributions.

References

SEPA Scottish Environment Protection Agency. Review of amine emissions from carbon capture systems. Stirling, UK, pages 1–86, 2015.
Nielsen CJ, Herrmann H, and Weller C. Atmospheric chemistry and environmental impact of the use of amines in carbon capture and storage (ccs). Chem. Soc. Rev., 41 :6684–6704, 2012.
Låg M, Lindeman B, Instanes C, Brunborg G, and Schwarze P. Health effects of amines and derivatives associated with CO₂ capture. 05 2011.
Morken A, Pedersen S, Kleppe ER, Wisthaler A, Vernstad K, Ullestad Ø, Flø NE, Faramarzi L, and Hamborg ES. Degradation and emission results of amine plant operations from MEA testing at the CO₂ technology centre mongstad. Energy Procedia, 114 :1245–1262, 07 2017.
Zhu L, Schade G, and Nielsen CJ. Real-time monitoring of emissions from monoethanolamine-based industrial scale carbon capture facilities. Environmental science & technology, 47, 11 2013.
Zhu L, Mikoviny T, Morken AK, Tan W, and Wisthaler A. A compact and easy-to-use mass spectrometer for online monitoring of amines in the flue gas of a post-combustion carbon capture plant. International Journal of Greenhouse Gas Control, 78 :349 – 353, 2018.
Lindinger W, Hansel A, and Jordan A. On-line monitoring of volatile organic compounds at pptv levels by means of proton-transfer-reaction mass spectrometry (PTR-MS) medical applications, food control and environmental research. International Journal of Mass Spectrometry and Ion Processes, 173(3) :191 – 241, 1998.
Hansel A, Jordan A, Holzinger R, Prazeller P, Vogel W, and Lindinger W. Proton transfer reaction mass spectrometry: on-line trace gas analysis at the ppb level. International Journal of Mass Spectrometry and Ion Processes, 149-150 :609 – 619, 1995. Honour Biography David Smith.
Shah MI, Lombardo G, Fostås B, Benquet C, Morken AK, and de Cazenove T. CO₂ capture from RFCC flue gas with 30w% MEA at Technology Centre Mongstad, process optimization and performance comparison. In 14th Greenhouse Gas Control Technologies Conference Melbourne, pages 21–26, 2018.
Wang T. Degradation of Aqueous 2-Amino-2-methyl-1-propanol for Carbon Dioxide Capture. PhD thesis, 2013.
Cuccia L, Bekhti N, Dugay J, Bontemps D, Louis-Louisy M, Morand T, Bellosta V, and Vial J. Monitoring of the blend 1- methylpiperazine/piperazine/water for post-combustion CO₂ capture. Part 1: Identification and quantification of degradation products. International Journal of Greenhouse Gas Control, 76:215 – 224, 2018.