Post-combustion CO2 capture with amino-acid salts

Conventional amine-based solvents used for post-combustion CO2 capture suffer from several common drawbacks.

Ralph H Weiland and Nathan A Hatcher
Optimized Gas Treating, Inc

Viewed : 3417

Article Summary

The most serious is the prohibitive energy used by solvent regeneration. Others include the additional process complexity needed to address solvent volatility, as well as dealing with oxidative and thermal degradation.

 Within the last five years, interest has developed in using the sodium or potassium salt of glycine (NaGly), the simplest primary amino acid, for CO2 capture. Another proposal (Wagner et al.5) is to use the potassium salt of a tertiary amino acid such as dimethylglycine (KDiMGly), promoted with a conventional alkanolamine such as monoethanolamine (MEA).

This paper benchmarks the performance of a 3,000 tonne/day CO2 capture plant using NaGly, MEA-promoted KDiMGly and piperazine-promoted KDiMGly against the standard, 30 wt% MEA. Piperazine-promoted MDEA is also considered. The results are striking — the 
regeneration energy required with piperazine-promoted KDiMGly appears to be about 20% lower that for MEA in an identical plant. Furthermore, solvent rates are lower by about 20%. Combined with a neutralised amino acid’s complete lack of volatility, this finding puts caustic-neutralised amino acids into a class of solvents of potential commercial interest. The paper provides a detailed explanation of how and why a process based on NaGly or piperazine-promoted KDiMGly is likely to perform well.

Post-combustion carbon capture is an applications area within gas treating with its own unique set of difficulties. Very few capture plants of any size have been built, so much of the discussion on carbon capture should correctly use the future tense. Huge gas volumes at essentially atmospheric pressure provide very little driving force for CO2 absorption to take place so contactors will be of very large diameter, and filled with structured packing to minimise pressure drop associated with large gas volumes. Focus to date has been predominantly on thermally regenerable solvents, usually amine-based, although other technologies such as ionic liquids and enzyme solutions (carbonate anhydrases) are receiving serious attention.

The benchmark solvent is 30 wt% corrosion-inhibited MEA, originally developed by The Dow Chemical Company and currently offered by Fluor under the name Econamine FG PlusSM. One of the drawbacks of conventional amines is their high volatility. The contactors in carbon capture are designed to be highly rich-end pinched. In other words, the plants are designed to operate at minimum reboiler duties and minimum circulation rates to keep energy consumption at the lowest possible level. The present obsession with lowering energy needs is quite understandable given that the energy required for an MEA-base carbon capture plant represents roughly one third of the plant’s power generation capability. Energy consumption escalates dramatically as the CO2 recovery requirement is increased. Thus, post-combustion carbon capture plants are intended only to remove the first 85% or so the CO2 in the flue gas. Minimum circulation, minimum regeneration energy, and bulk CO2 removal rich-end pinch the column and conspire to produce a large temperature bulge near the top of any well-designed contactor. The treated gas then will tend to be quite hot, making amine volatility losses very high. Thus, an amine with minimal vapour pressure is desired. Amino acids neutralised with an alkaline metal hydroxide are perfect candidates because they are indeed salts, and have zero vapour pressure. Solvent regeneration consumes the overwhelming majority of the energy needed for CO2 capture using thermally regenerable solvents. A well-designed and optimised capture system for a 300 MW power plant will consume about 100 MW of the power plant output, most of it for solvent regeneration. Thus, anything that can be done to lower regeneration energy has a huge potential benefit to capture plant efficiency. Another challenge peculiar to all post-combustion CO2 capture plants is the unavoidable presence of oxygen caused by the use of excess combustion air in furnaces. Most amines react readily with oxygen and form a host of degradation products, some of which can lead to foaming and corrosion. Amino acids are naturally occurring materials that evolved to function very effectively in an oxygen-rich atmosphere and are quite resistant to oxidation. However, amino acids are almost pH neutral and their amino group is already protonated so they do not react with CO2. In the first part of this paper we discuss the amino acids and how they can be made reactive. In the second part comparisons are made between MEA, NaGly, MEA-promoted KDiMGly and piperazine-promoted KDiMGly. It is also pointed out why promoted MDEA is not a candidate for post-combustion carbon capture.
Base case — 3500 mtpd CO2 capture plant
To compare solvents readily on an equitable basis, a completely non-optimised plant for removing 85% of the CO2 from a 300 MW station was selected as the basis. This amounts to the removal of about 3,000 tonnes/day (mtpd) of CO2 from the combustion gas (at 85% removal efficiency). The raw gas was assumed to contain 13 mol% CO2, 86.8 mol% N2 and 0.2 mol% O2 at 70 mbar and 43°C (110°F). The absorber was simulated with 20 m of Mellapak M250.X structured packing and was sized for 50% of flood. Absorber pressure drop was generally about 20 mbar. The column bottom pressure was assumed to be 70 mbar. The regenerator was simulated with twenty 4-pass trays with feed to tray 3 from the top, and it was sized for 70% jet and downcomer flood.

Solvent compositions of amines were taken at the most commonly-used or conventional values, namely 30 wt% MEA, 45 wt% NaGly, 40 wt% KDiMGly, 45 wt% MDEA, and in the case of promoted amines, a 30:15 wt% mixture of KDiMGly:MEA and promotion with 5 wt% piperazine in the cases of KDiMGly and MDEA. Lean solvent temperature was set at 43°C, the same as the raw gas, and a temperature approach of 5.5°C (10°F) was arbitrarily chosen for the cross exchanger.

Apart from the normal lean-rich cross exchanger, no attempt was made to heat integrate or in any way optimise plant configuration for any amine or blend. No doubt better results could be obtained for any given solvent by paying attention to heat integration, optimising temperature approaches in heat exchangers, more carefully selecting lean amine temperature, etc. so as to minimize OPEX and CAPEX. However, our objective was a simple comparison on a reasonably equitable basis between neutralised amino acids and MEA. Figure 1 shows the base flowsheet.
Amino acids as carbon capture solvents
Salts of amino acids appear first to have been used commercially for acid gas removal around 1935. The Alkazid process developed by IG Farbenindustrie, and later formally by BASF AG after its reëstablishment in 1952, uses the potassium salts of N,N-dimethylaminoacetic acid (also known as the potassium salt of dimethylglycinate) and N-methylalanine for treating refinery, coke-oven, and natural gases. It is noteworthy that refinery and especially coke-oven gases are two of the most severely aggressive treating services from contamination and solvent degradation standpoints. The process appears to have been most commonly applied in Europe, especially Germany, although there are instances of its use elsewhere.

Add your rating:

Current Rating: 2

Your rate:

  • Responsive image Process catalysts
  • Responsive image Catalyst portfolio
  • Responsive image ITW technologies online cleaning
  • Responsive image The sulphur processing expert
  • Responsive image FluegasExact 2700 effective combustion analysis
  • Responsive image OpX® Digital
  • Responsive image Process flares
  • Responsive image Optimize everything plant-wide
  • Responsive image Use your catalyst buying power
  • Responsive image FCC catalysts & additives