Gas treating simulation – a holistic perspective part 2: carbon capture
This article is the second in a three-part series examining real-world cases in which the behaviour of a treating plant is somewhat counterintuitive. Part 1 focused on a treating problem experienced in the CO2 removal section of an ammonia plant.
Nathan A Hatcher, R Scott Alvis and Ralph H Weiland
Optimized Gas Treating
Viewed : 2944
There, a seemingly-small departure from an operating condition recommended by the process technology’s supplier resulted in failure to treat adequately by a very large margin. In the present article, we turn our attention to the surprising relationship between reboiler duty and solvent circulation rate in a pilot plant for CO2 capture operating to meet a specified fractional CO2 removal. To understand the behaviour, one must appreciate the difference in operating philosophy between conventional deep CO2 removal and the removal of a limited amount of CO2.
As pointed out in Part 1, developing a good understanding of the behaviour and performance of any gas treating plant requires one to look at the plant in its entirety rather than focusing too soon on some specific detail. The main tool used in the analysis of the present case is mass transfer rate simulation, specifically the commercially available ProTreat simulator, that creates a virtual plant on a computer. The virtual plant is a digital image based on all the real factors that affect performance. The scale is microscopic and quite comprehensive. The model reveals intricate detail and is very satisfying from an engineering science perspective. More pragmatically, it can lead to the solution of difficult troubleshooting exercises, on the one hand, or to the selection of correct operating conditions for a new plant, on the other. In every case, the engineer is left with intimate knowledge of what makes a given plant tick. In this Part 2 we consider a performance map for a CO2 capture plant using MEA.
Post-combustion CO2 capture
Figure 1 shows the pilot plant processing scheme. The plan was to remove 90% of the CO2 from the entering flue gas and the engineers were interested in exploring the extent to which reboiler duty (regeneration energy) depended on solvent circulation rate. The absorber contained 10 m depth of structured packing with a specific surface area of roughly 250 m2/m3. The cross exchanger was to operate with a 10°F temperature approach. When a series of simulations were run over the solvent flow range from 1,800 to 6,000 gpm and the reboiler duty was adjusted to achieve 90% removal, the rather unusual looking curve shown in Figure 2 was obtained. There are two minima, a maximum, and two asymptotes at the far left- and right-hand sides. The engineers conducting the study initially believed the simulator predictions were in gross error. Based upon results from an equilibrium stage, efficiency-based simulator, they expected to see a single minimum in the reboiler duty followed by a steady increase in reboiler energy input. However, as we shall see, this odd and surprising behaviour is quite real.
The reason for this seemingly strange behaviour is that the absorber moves from a rich-end pinch condition in Region A at the far left of Figure 2 (the desired mode of operation for carbon capture), to a lean end pinch in Region C at the far right (the desired mode of operation for deep CO2 removal). The solvent flows in Region C are far higher than needed for only 90% CO2 removal: these flows are much more typical of a conventional MEA column used to treat to low CO2 levels. To make such high flows work economically for CO2 capture, the reboiler duty must be greatly reduced so that the lean solvent loading is very high. At a high solvent rate, a high lean loading is completely consistent with what is needed for the equilibrium partial pressure of CO2 over the lean solvent to have a high enough value that it results in only 90% of the CO2 being removed.
The regions marked A, B, and C in the figure highlight areas where it might be worthwhile to look at temperature profiles for clues1. It is no surprise that the combination of low reboiler duty and low solvent rate is an efficient way to remove 90% of the CO2— indeed, the rich end pinch is purposely produced by low solvent flow and is exactly what is used to limit CO2 absorption. However, why does a higher solvent flow eventually require more reboiler energy to be expended? The answer is that as solvent flow increases, the temperature bulge spreads to much of the interior of the absorption column making the centre region so hot it can do little or no absorbing. Therefore, it leaves more CO2 in the gas, unless, that is, the solvent gets stripped cleaner by more reboiler steam. At point B the bulge temperature is nearly 170°F and only the ends of the column are effective in removing CO2. As the solvent rate goes still higher, beyond the peak at 4,000 gpm, the bulge continues to move down the column and treating becomes increasingly lean-end pinched. The column has entered an operating region more suited to deep CO2 removal than carbon capture, but it is forced to remove only limited CO2 by being grossly under stripped.
Under lean-end-pinched conditions, solvent lean loading controls the outlet CO2 concentration and as the solvent rate increases, higher and higher lean loadings are adequate for treating to remove only 90% of the CO2. Despite the fact that the solvent flow is higher so more reboiler energy will go into heating the rich feed to its bubble point on the stripper’s feed tray, less stripping still requires less energy and the curve falls through another minimum. It should be apparent that to the left of the first minimum a further decrease in solvent rate will neew to be offset with a substantial decrease in lean loading, hence higher reboiler duty. To the right of the second minimum, a higher solvent flow will require a gradually increasing reboiler duty to heat the solvent to the feed tray temperature, the solvent lean loading being fairly constant beyond 6,000 gpm.
Maxima and minima are always caused by the existence of (at least) two factors opposing each other with one factor dominating on one side of the minimum and the other dominating on the opposite side. The factors at play here are (1) lean-end versus rich-end pinch, and (2) solvent net loading capacity versus the solvent flowrate. So there are two pairs of factors. There is more than one minimum and the situation is obviously complex. In part, the complexity is a result of one trying to examine carbon capture (left hand side) and high purity treating (right hand side) at the same time. These two regions and goals are at opposite ends of the treating spectrum and looking at them simultaneously (i.e., holistically) leads one to suspect that the simulator is just plain wrong. However, if these two regions are considered separately from each other, the analysis becomes much simpler, and a lot of the complexity goes away.
Add your rating:
Current Rating: 3