Tobias Jockenhövel & Rüdiger Schneider, Siemens & Helmut Rode, E.ON, Germany
Siemens is developing a process for post-combustion removal of CO2 from power plant flue gases that employs an aqueous solution of an amino acid salt (AAS). The company’s POSTCAP pilot scheme is validating the technology to pave the way for future pilot and full-scale demonstration projects and has already shown the process to have a high capture rate and to be environmentally friendly. POSTCAP has produced results for potential overall plant energy efficiency, the suitability of construction materials, and solvent stability and tolerance to trace compounds such as SOx and NOx. It has also provided insights into potential operating conditions.
POSTCAP is a three-year scheme that Siemens is running jointly with E.ON and involves long-term tests with flue gas. It is a step towards large-scale validation in demonstration plants. It aims to develop Siemens’s second-generation technology for the post-combustion capture of carbon. New steam power plants and combined-cycle power stations can use the technology, which can also be retrofitted in existing plants. Under the process, carbon is captured by an AAS solvent whose low absorption enthalpy and near-zero vapour pressure makes its use economical and environmentally friendly. Siemens has tested AAS capture in its laboratory plant and is testing it in the POSTCAP scheme, which is removing carbon from real flue gas in a slip stream at the pilot plant at the Staudinger Unit 5 power station in Germany, which outputs 510 MW and burns hard coal.
The Federal Ministry of Economics and Technology, BMWi, is funding the project. Staudinger’s owner E.ON Energie is co-financing it. E.ON’s power station is exposing the carbon capture pilot plant to real operating conditions. The pilot plant is well integrated into the power station’s infrastructure, which provides a suitable framework for developing the carbon capture technology. E.ON is also providing operator experience to the POSTCAP project.
This is appropriate since the AAS carbon capture process will have to cope with the varying operational requirements of different commercial flue gas cleaning devices. An important step in developing the AAS technology is the validation of the process. Staudinger’s carbon capture pilot plant, erected by Siemens, works on a flue gas slip stream of 140 Nm³/hour after conventional flue gas cleaning. The main parts of the existing power plant flue gas cleaning system include NOx removal via primary measures, selective catalytic reduction (SCR), electro-filters to remove dust, and a limestone absorber that performs desulphurization.
The POSTCAP scheme collected data sets of operation and solvent parameters to deliver experimental figures that allow the validation of the design of the whole process. The scheme’s long-term investigations with real flue gas are validating these parameters: heat and mass balance; energy demand and capture rate; solvent stability and prevention of crystallization; qualification of material; and environmental impact, to prove low emission levels. Analysis and evaluation of the operating behaviour of the process design under part load and maximal load change gradients, start-up and shutdown continues.
Installation and commissioning
Commissioning of the pilot plant began in August 2009, and its operation will stretch over 16 months. The erection of the POSTCAP pilot plant divides into three parts, as Figure 1 shows.
|Figure 1. Time bar of pilot plant installation and commissioning|
Siemens began planning the design of the process in July 2008. Mechanical completion and commissioning of the pilot plant occurred one year later. All planning permissions were obtained quickly and without restrictions because of the harmless nature of the solvent and the research and development character of the project. An early planning phase initiated procurement of items that had long lead times. Fast-track engineering accelerated the whole of the installation phase.
The main components of the pilot plant are the flue gas cooler, the absorption and desorption column, heat exchangers, desorber top condenser and steam-driven reboiler. POSTCAP extracts flue gas after the flue gas duct. The pilot plant character of the scheme has meant that the CO2 stream and lean flue gas from the CO2 capture plant is fed back into the flue gas duct. The column diameter conforms to DN 200. Its height is about 35 m, including the flue gas cooler, piping and analytical equipment.
A Siemens PCS7 DCS in a container close to the test facility controls the pilot plant’s operation and the recording of all measurements. In a first step, the pilot plant design uses a standard process configuration to allow its results to be compared with those of research carried out elsewhere in the field of post-combustion carbon capture.
By creating an operating schedule for the first operation phase, the project was able to fulfil all the objectives of the pilot plant, as Figure 2 shows. The start-up phase in August 2009 included functional tests with water. These included the start up of the KOH-scrubber, which operates as a flue gas cooler and can simultaneously adjust the SOx load of the flue gas.
|Figure 2. Schedule for the first operation phase of the pilot plant|
Carbon capture using the AAS solvent started in September 2009. Start up and parameter optimization of the plant occurred in the first operation period. Next, the project varied operating parameters, such as solvent pump-around and amounts of steam and processed flue gas, to provide a matrix of process data in different operating conditions. The scheme used these results to validate the existing process model.
This validation allowed reliable prediction of the performance of pilot and full-scale plants under real conditions. The scheme ran a computer simulation with the validated model to obtain an approved operating point, which was adjusted during the long-term operation phase of the plant.
The first long-term test ran in December 2009 and allowed the effects on the solvent of flue gas secondary components such as SOx, O2, and NOx to be investigated. In that month, emission measurements ran in parallel with the long-term test run. The functional tests regarding parameter optimization continued in January 2010.
Tests for the confirmation and selection of construction materials at different process conditions have run in parallel since November 2009. A second long-term test is now running in which controlled amounts of suphur are being input into the process to allow the investigation of the interaction between the solvent and SOx.
Results from the operation of the pilot plant will help confirm the suitability of the capture solvent, for example in terms of its stability and applicability to capture CO2. The solvent has been operating in a laboratory plant for four years. Besides the validation of the simulation model under real power plant conditions, the pilot plant has allowed conclusions to be drawn about the operating behaviour of large-scale capture plants and the selected construction materials.
Functional tests with water had already shown how good the performance of the KOH-scrubber – or flue gas cooler – was. SOx concentration can be safely adjusted to perform related parameter variations to verify the behaviour of the process. The project also determined what the heat losses are at the desorber, results that were of use in the validation of the thermodynamic model. During commissioning and the first test programmes, a three-shift and later a two-shift work system ran to ensure a safe start-up of the pilot plant.
Today, the pilot plant is fully automated and can be operated by remote control. A process that employs an amino-acid salt solution for carbon capture must safely avoid crystallization, but this is well understood due to extensive study of this as a function of temperature and CO2 loading. This knowledge allowed several process arrangements to be made to prevent crystallization. Control of the solvent temperature keeps it well above the point of crystallization.
In the event of an interruption to the process, the pilot plant will switch into standby operation. This will close the flue gas inlet, which will mean no more CO2 can react with the solvent. To prevent fluid stagnancy, the plant will ensure a minimal circulation of solvent.
Also, a well-defined amount of reboiler duty will provide heat to keep the pilot plant at a sufficient temperature. Control of this reboiler duty ensures the minimal heat requirement will compensate for heat losses. In the case of a scale-up, the relative minimal heat demand will be less because the equipment will have a smaller ratio of surface to volume. In addition, the entire amount of solvent present in the plant can be stored in a designated storage tank. In the first test period, damage occurred to a solvent pump. Power plant shutdowns also interrupted the planned operation of the pilot plant as they caused low or no flue gas input. The pilot plant automatic standby mode safely handled these situations.
The entire operation has allowed a general conclusion to be drawn about influencing parameters on pilot plant availability. The scheme investigated planned unavailability, such as during run-up and shutting down procedures on the one hand and zero flue gas supply on the other. No surprising effects were observed.
Heat and mass balance validation
Table 1 shows how in the first operation phase of the pilot plant, in particular for the validation of the model for the thermodynamic process, the scheme adjusted a matrix of different operating points. Figure 3 shows that the plant achieved a capture rate of 90 per cent and above. The CO2 loading of the amino acid salt reached high levels, as expected.
|Figure 3. Capture rate (not controlled, blue) and flue gas input (black) versus time for the test run. The need for dynamic model validation meant the capture rate was not controlled, so oscillated slightly|
The scheme used the measurement data to validate the thermodynamic process model, which predicts the process behaviour over a wide range of operating conditions. The validation involved slightly adjusting the effective mass transfer area of the packing, or the interfacial area factor. Figure 4 compares measured and predicted temperature profiles in the absorber and desorber.
|Figure 4. Comparison of measured and predicted temperature profiles in the columns of the pilot plant|
The validated process model could confirm an energy demand of 2.7 GJ/tonne CO2 for a full-scale capture plant. This would lead to an efficiency drop of only 8.5 percentage points compared with a state-of-the-art hard coal power plant that employs 600 °C technology. This low efficiency penalty includes CO2 compression to 200 bar and all necessary auxiliary power demands, such as for cooling water pumps. The pilot plant features full-scale packing heights and is equipped with the same structured packing as foreseen for a full-scale capture plant. This means that the validated process model can be used to design full-scale capture units for large power plants. There is no need for a scale-up in height, which would influence the behaviour of the process.
Laboratory investigation has shown that the amino acid salt at the POSTCAP pilot plant is stable under thermal stress and different oxygen environments. An estimate of the total loss of solvent per year due to O2 and thermal degradation is less than one per cent of the total solvent hold-up. SOx will react with the solvent to produce sulphite and sulphate, which causes a slight decrease in pH because of the acidic nature of SOx gases.
Nevertheless, SOx does not react with the active substance, the amino acid. There will therefore be no significant loss of solvent, because a reclaimer will selectively remove SOx reaction products. NOx reacts in the same way as SOx, this time producing nitrite and nitrate, which again reduces pH. The reclaimer will remove these too. The amino acid salt has negligible vapour pressure and is stable against the main flue gas impurities. These are two qualities that help make it environmentally friendly. To confirm this, an independent test laboratory is measuring the emissions of the pilot plant. Its results currently show that the capture process produces negligible emissions. For example, it produces no volatile organic compounds and negligible additional ammonia, less than 1 mg/Nm3. Many flue gas impurities are also absorbed by the AAS.
Intensive investigations show that a variety of construction materials are compatible with the solvent. This led to the selection of three types of steel alloy, which had advantages when it came to costs, availability and known resistance behaviour: carbon steel 1.0425, stainless steel 1.4541 and stainless steel 1.4571.
|Pilot plant: mechanical completion in July 2009|
Putting parallel grooves in the samples before installation in the pilot plant allowed the observation of cracks caused by corrosion. These cracks would appear at right angles to the grooved lines after bending of the sample.
The samples spent total residual time of 1370 hours in the ASS-based solvent. All of the samples were resistant under the test conditions. They showed no local corrosion effects, and surface abrasion was insignificant. The conclusion is that stainless steel 1.4571 is a suitable material for this process. Similar qualifications for lower grade materials have provided results that are generally good.
POSTCAP validated the existing AAS process simulation model and showed good accordance between measured and predicted temperature profiles at the absorber and desorber. Solvent stability and low emissions have been experimentally confirmed.
The results indicated that the amino-acid salt is stable against thermal stress and oxygen environments that appear under real operation conditions.