Scholarly article on topic 'Solid Sorbent CO2 Capture Technology Evaluation and Demonstration at Norcem's Cement Plant in Brevik, Norway'

Solid Sorbent CO2 Capture Technology Evaluation and Demonstration at Norcem's Cement Plant in Brevik, Norway Academic research paper on "Chemical engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"CO2 " / carbon-capture / Norway / cement-industry / field-testing / economic-analyses / solid-sorbents}

Abstract of research paper on Chemical engineering, author of scientific article — Thomas O. Nelson, Luke J.I. Coleman, Paul Mobley, Atish Kataria, Jak Tanthana, et al.

Abstract RTI and Norcem AS have partnered to carry out a pilot-scale CO2 capture technology demonstration in an operating cement plant utilizing an advanced, solid sorbent CO2 capture process. RTI's process technology – currently being developed for coal power plant applications – has the potential to substantially reduce the energy load and capital and operating costs compared to conventional CO2 scrubbing. This work, carried out within the Norcem CO2 Capture Project, has three main objectives: 1) evaluate the technology's economic feasibility for commercial-scale cement plant application, 2) collect sorbent exposure and performance data for the CO2 capture sorbent utilizing simulated and actual cement plant flue gas, and 3) demonstrate, on a pilot-scale, the technology's capacity to achieve effective and continuous removal of CO2 from Norcem's cement plant.

Academic research paper on topic "Solid Sorbent CO2 Capture Technology Evaluation and Demonstration at Norcem's Cement Plant in Brevik, Norway"

(S)

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Energy Procedia 63 (2014) 6504 - 6516

GHGT-12

Solid Sorbent CO2 Capture Technology Evaluation and Demonstration at Norcem's Cement Plant in Brevik, Norway

Thomas O. Nelsona*, Luke J.I. Colemana, Paul Mobleya, Atish Katariaa, Jak Tanthanaa, Markus Lesemanna, and

Liv-Margrethe Bjergeb

aRTIInternational, Post Office Box 12194, Research Triangle Park, NC 27709-2194 bNorcem Brevik AS, Setreveien 2, 3991 Brevik, Norway

Abstract

RTI and Norcem AS have partnered to carry out a pilot-scale CO2 capture technology demonstration in an operating cement plant utilizing an advanced, solid sorbent CO2 capture process. RTI's process technology - currently being developed for coal power plant applications - has the potential to substantially reduce the energy load and capital and operating costs compared to conventional CO2 scrubbing. This work, carried out within the Norcem CO2 Capture Project, has three main objectives: 1) evaluate the technology's economic feasibility for commercial-scale cement plant application, 2) collect sorbent exposure and performance data for the CO2 capture sorbent utilizing simulated and actual cement plant flue gas, and 3) demonstrate, on a pilot-scale, the technology's capacity to achieve effective and continuous removal of CO2 from Norcem's cement plant.

© 2014TheAuthors. Publishedby ElsevierLtd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.Org/licenses/by-nc-nd/3.0/).

Peer-review under responsibility of the Organizing Committee of GHGT-12

Keywords: CO2, carbon-capture; Norway; cement-industry; field-testing; economic-analyses; solid-sorbents

1. Introduction

It is increasingly clear that CO2 capture and sequestration (CCS) must play a critical role in curbing worldwide CO2 emissions to the atmosphere. Since fossil fuel-fired power plants account for about 40% of global energy-related CO2 emissions, the power industry has been the primary target for developing CCS technologies. However, as

* Corresponding author. Tel.: +1-713-942-7864; fax: +1-919-541-8002. E-mail address: tnelson@rti.org

1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/3.0/).

Peer-review under responsibility of the Organizing Committee of GHGT-12

doi: 10.1016/j.egypro.2014.11.686

anthropogenic CO2 emissions continue to be scrutinized and legislation to regulate CO2 advances, or is on the horizon, in many countries, CCS technologies for all major industrial sources of CO2 will most likely be required to meet future CO2 emission targets. The case for CCS in the cement industry is especially strong since, by the very nature of cement production, there is a limit to how much CO2 emissions can be reduced by efficiency gains and/or use of renewable energy. The raw materials (limestone) used in cement manufacturing account for roughly 60% of the CO2 emitted from a cement plant. The other 40% generally comes from the combustion of fossil fuels to obtain the heat needed for limestone decomposition - the chemistry essential to cement manufacturing. Fossil fuel related emissions can be mitigated through increased efficiency and fuel switching, but to avoid CO2 emissions from limestone decomposition, CCS appears to represent the only good option.

To date, there are no cement plants worldwide that utilize CCS technologies to mitigate CO2 emission on a commercial scale. However, Norcem AS - part of HeidelbergCement Group - has taken a leading role in establishing the groundwork needed to deploy commercial CCS technologies within the cement industry. Norcem, in partnership with the European Cement Research Academy (ECRA), has established the first ever CO2 capture test center located at a cement plant - Norcem's plant in Brevik, Norway. This project, the Norcem CO2 Capture Project, has been tasked with the construction of this unique test center as well as testing of various post-combustion CO2 capture technologies at the center. The overall directive of the project is to evaluate both relatively mature CO2 capture technologies as well as technologies in an earlier stage of development. The CO2 capture technology providers are tasked with conducting detailed technical, economic, and environmental assessments in order to determine the commercial viability of their respective technologies within the cement industry, with the prospect of future scale-up and demonstration. The project started in May 2013 and is expected to be complete by March 2017.

As part of Norcem's CO2 Capture Project, RTI and Norcem have partnered to carry out a pilot-scale CO2 capture technology demonstration utilizing an advanced, solid sorbent CO2 capture process. RTI's process technology -currently being developed for coal power plant applications - has the potential to substantially reduce the energy load and capital and operating costs compared to conventional aqueous amine CO2 scrubbing. It is a thermal-swing CO2 capture process utilizing a high CO2 capacity sorbent in a fluidized, moving-bed reactor arrangement. In a separate, parallel project, supported by the U.S. Department of Energy/National Energy Technology Laboratory (DOE/NETL) and Masdar, RTI and Pennsylvania State University (PSU) are working together to develop and improve the novel, fluidizable solid sorbent and process.

The objectives of RTI's work within the Norcem CO2 Capture Project are three-fold: 1) Evaluate the technology's economic feasibility for commercial-scale application within the cement industry, 2) Collect representative gas exposure data for the CO2 capture sorbent and optimize the process for cement plant application, and 3) Demonstrate, on a pilot-scale, the technology's capacity to achieve effective and continuous removal of CO2 from Norcem's cement plant in Brevik, Norway. RTI's work effort is divided in two Phases with Phase I focused on the first two objectives listed above and Phase II focused on the third objective - pilot testing. Phase I work is complete, with the results and conclusions discussed in this paper, while Phase II work has recently started as of August 2014.

2. RTI's solid sorbent CO2 capture technology

RTI's technology is a solid sorbent-based process which selectively removes CO2 from industrial exhaust gas streams through a cyclic, thermal-swing, absorption-desorption process, generating a high-purity CO2 product gas that is "sequestration-ready". The process is being developed as a lower cost alternative to conventional aqueous amine CO2 scrubbing.

RTI's technology consists of sorbent- and process-based components. The sorbent material is based on a CO2-philic poly-amine (polyethyleneimine, PEI) loaded on a high surface area support material (e.g. silica) - first developed at PSU. PEI is particularly well-suited for CO2 capture as it consists of branched polymer chains with a high density of amine groups (primary, secondary and tertiary amines) which selectively absorb CO2. The sorbent captures CO2 via carbamate and bicarbonate chemical reaction pathways. The support material is used to give the sorbent physical strength, allow it to be used in a fluidized-bed process, and to increase the surface area on which CO2 can interact with the PEI reagent.

This sorbent material can feasibly be used in multiple process reactor environments - fixed-bed, fluidized-bed, and transport reactors. Engineering analysis has led us to conclude that the ideal reactor environment for thermal-swing solid sorbent-based CO2 capture is dual fluidized, moving-bed reactors (FMBRs) in which the sorbent is continuously circulated between a CO2 Absorber (for CO2 capture) and a Sorbent Regenerator (for CO2 release and concentration). This circulating, FMBR design concept uniquely meets the technical requirements for solids circulation, heat management, and performance optimization while also addressing issues of scale and commercial availability of process components. A basic block flow diagram of RTI's process is exhibited in Figure 1. The process also utilizes various utilities such as cooling water in the CO2 Absorber and heat exchangers, steam in the Sorbent Regenerator, and electricity for operating the various process equipment.

3. Economic analyses

As part of Norcem's CO2 Capture Project, RTI Fig. 1. Block flow diagram of RTI's advanced solid sorbent CO2 capture was tasked with assessing the technical and process. economic feasibility of the solid sorbent technology

for cement plant applications. Technical viability of RTI's technology was being explored through various field testing activities. Economic viability was evaluated through a preliminary economic analysis of the technology by projecting how the technology will be designed and operated commercially at a cement plant - more specifically, Norcem's cement plant in Brevik, Norway. The overall objective of this work was to provide a preliminary indication of the economic performance of RTI's technology and provide a basis (along with field testing) for determining how suitable RTI's technology is for implementation in modern cement kiln systems. These efforts produced quantitative assessments of the technology's primary economic performance indicators, such as capital cost, operating cost, cost per CO2 captured/avoided, and energy consumption. It also developed qualitative assessments of performance under varying conditions, health risks, environmental risks, safety risks, and potential improvements for RTI's technology for industrial application.

3.1. Scope of economic analyses

The scope of RTI's economic evaluation was defined by Norcem, and their partner Tel-Tek, through a "Benchmark Indicator Report" delivered to all technology providers active under the Norcem CO2 Capture project. Key information supplied in this report includes, 1) boundary specifications for CO2 capture technologies, 2) specifications for uncleaned flue gas from a commercial cement plant, 3) capture rate requirements, 4) waste heat utilization potential at the cement plant, 5) specifications for the CO2 "product" stream, 5) unit prices of utilities, consumables and waste handling, and 6) definition of and specifications for variations that may exist at a commercial cement plant (i.e. case studies). Norcem and Tel-Tek have asked that three different cases for CO2 capture at a commercial cement plant be considered:

• Case 1 (reference case): Full-size cleaning (minimum 85 % CO2 capture), no waste heat available.

• Case 2: Full-size cleaning (minimum 85 % CO2 capture), waste heat available as per the amount specified in the Benchmark Indicator Report.

• Case 3: Reduced-size cleaning, based on a cost-optimal utilization of waste heat available (i.e. the CO2 captured amount is dictated by the waste heat available which is the sole heat source available to meet the thermal energy demand of the process).

Cases 2 and 3 are studies that may offer a glimpse at the unique opportunity for capturing CO2 within a cement facility. It is highly likely that many cement plants worldwide will have waste heat available for use within the capture plant - as is the case with Norcem's Brevik plant. The purpose of including Cases 2 and 3 in these economic analyses is to see if a particular CO2 capture technology can benefit from the quality of waste heat available within the cement plant.

3.2. Design basis

The design basis for this assessment combined the requirements of Norcem and Tel-Tek's Benchmark Indicator Report as well as information and assumptions applicable solely to RTI's technology, i.e. specifications and properties of RTI's CO2 capture sorbent and major components of RTI's CO2 capture process. A summary of the design basis is provided here:

• Specification of the uncleaned flue gas: flow rate (> 300, 000 Nm3/h), component concentrations (e.g. CO2 = 17.8 vol%), pressure (1 bara), and temperature (165°C) of the "uncleaned gas" (i.e. flue gas, exhaust gas) from the Norcem Brevik plant.

• Specifications of system boundaries: 1) Uncleaned flue gas, coming from downstream of the cement plant's baghouse, enters the capture system; 2) Product CO2 gas exits as a high-pressure, high-purity CO2 stream crossing the plant fence; 3) Clean flue gas exits the existing cement plant stack; 4) Any required fuels, electricity, cooling water, and process water are available on-site at given specifications; 5) Waste heat (from the cement plant) may be integrated within the CO2 capture plant (Cases 2 and 3 only); 6) Waste streams generated by the CO2 capture plant will be identified and quantified.

• Specifications of RTI's sorbent: RTI's CO2 capture sorbent is a fluidizable supported-amine sorbent consisting of PEI impregnated on a high-porosity, high surface area, fluidizable support (e.g. silica). The design basis provided the specifications and properties of this sorbent, such as heat of absorption (66,000 kJ/kmol-CO2), heat capacity (1.0 kJ/kg.K), sorbent working CO2 capacity (12 wt%), and many other properties.

• Specifications of RTI's process components: RTI's CO2 capture process operates as a cyclic absorptionregeneration thermal swing process where the solid sorbent is continuously circulated between two fluidized, moving-bed reactors (FMBR) - a CO2 absorber and a sorbent regenerator. In addition to these two reactors, the process contains the following process components for cement plant application: a 1) Direct Contact Cooler, 2) Caustic Scrubber, 3) CO2 Compression and Drying system, 4) Sorbent Conveying system, and a 5) Steam Delivery system. The design basis provided relevant design assumptions, sizing, and other information for the various process components.

3.3. Overview of case studies and economic analysis approach

As mentioned above, Case 1 is considered the "reference case" in which the CO2 capture technology is to treat the entire flue gas stream from Norcem's Brevik plant, removing a minimum of 85% of the CO2 present in that gas stream, without the benefit of waste heat utilization. RTI developed the Aspen process simulation for Case 1 to treat the entire flue gas stream and target the minimum CO2 capture rate of 85%. Thus, RTI's technology was sized to process the entire flue gas flow rate and capture 85% of the CO2 at a run factor of 100%.

Similar to Case 1, Case 2 also encompasses "full-size" cleaning (minimum 85 % CO2 capture), but waste heat is available for use in the capture plant. The waste heat is derived from the hot flue gas stream at the cement plant prior to entering the conditioning towers. A waste heat boiler is used to generate "low-quality" steam (2.4 bara) as RTI's sorbent process uses steam for sorbent regeneration at 110°C. The waste heat boiler generates approximately 35% of the total energy requirement, thus the steam demand from on-site steam generation plant is roughly only 65% of the quantity needed in Case 1. With the exception of the lower steam requirement and the addition of a waste heat boiler to the equipment requirement, the heat and mass balance and equipment specifications for Case 2 are similar to that of Case 1.

Case 3 encompasses "reduced-size" cleaning, based on a cost-optimal utilization of waste heat available at the Norcem Brevik cement plant. Essentially, this case answers the question, how much CO2 can be captured and regenerated at the cement plant if only the waste heat available is used to meet the thermal energy requirements of the capture plant? Preliminary estimates showed that the available waste heat can be used to generate a quantity of steam that can be used to regenerate 30% of CO2 present in the uncleaned flue gas. As a result, Case 3 has been simulated for 30% CO2 capture from the flue gas stream. Capture of 30% of CO2 can be achieved by either treating the entire flue gas flow and capturing 30% or by treating only a fraction of the flue gas flow rate and capturing >30% of CO2. Various sub-cases were explored in order to identify a cost-optimal utilization of the available waste heat. It was concluded that the cost-optimal case is to treat 35% of the flue gas flow and capture 85% of the CO2 present in that slipstream of flue gas.

Each case was simulated using an in-house developed Aspen Plus® process model to generate heat and mass balances, utility demand estimates, equipment lists and specifications, as well as all capital and operational cost indicators.

3.4. Results of technical and economic analyses

A summary of the technical and economic analysis for the three case evaluated is provided in Table 1. The total cost of CO capture for the "full-size" cleaning reference Case 1 is 45.8 €/t-CO2. In Case 2, the cost of CO2 capture decreases to 40.7 €/t-CO2. The major driving factor for this is that in Case 2, 35% of the thermal energy demand is recovered from waste heat available in the cement plant. Considering the capital cost associated with the production of steam using waste heat, the resulting steam cost is 5.8 €/t-steam compared to 17.5 €/t-steam cost for on-site steam generation in Case 1. Finally, in Case 3, the cost of CO2 capture is reduced further to 38.6 €/t-CO2. Initially, it was theorized that the economic improvement in this case would be even greater since all of the steam in Case 3 is available at 5.8 €/t-steam. However, the economic analysis shows that the increased CAPEX per mass of CO2 counteracts some of the steam cost improvement in this case.

Thomas O. Nelson et al. / Energy Procedía 63 (2014) 6504 - 6516 Table 1. Summary of economic analysis results for RTI's CO2 capture technology

Case 1

Case 2

Case 3

Total CAPEX M €

Total OPEX €/y

Steam cost per mass of CO2 avoided [€/t-CO2avoided]

Electricity cost per mass of CO2 avoided [€/t-CO2avoided]

Other variable OPEX per mass of CO2 [€/t-CO2avoided] avoided

Fixed OPEX per mass of CO2 avoided [€/t-CO2avoided]

CAPEX per mass of CO2 avoided [€/t-CO2avoided]

Total cost per mass of CO2 avoided [€/t-CO2avoided]

75.8 33.2 18.6 13.5

8.8 45.8

27.9 12.2 13.5

9.9 40.7

47.9 7.0 0.0

13.5 5.6

3.2 15.9

Electric energy consumption per mass of CO2 [kWh/t-CO2avoided] avoided

Thermal energy consumption per mass of

CO2 avoided

Waste heat utilization per mass of CO2 avoided

[MJ/kgCO2avoided] [MJ/kgCO2avoided]

The total annual thermal energy consumption for the three cases depends on the CO2 capture rate for each case. In Case 1, the entirety of the thermal energy demand is met by on-site steam generation from a natural gas boiler, which is estimated to cost €17.5/t-steam resulting in a capture process steam cost of 18.6 €/t-CO2 avoided. In Case 2 and Case 3, 22 MW thermal energy is recovered in the waste heat boiler, accounting for 35% of the Case 2 steam requirement. This reduces the on-site steam generation demand and lowers the steam cost per mass of CO2 to 12.2 €/t-CO2avoided for Case 2. Since Case 3 uses only the thermal energy recovered in the waste heat boiler, steam cost per mass of CO2 is 0.

Regarding electric power demands in RTI's capture process, the predominant use of electricity is for the compression of the purified CO2 stream for pipeline transfer - accounting for 74% of total electric power consumption. The next highest electricity demand is 16% of the total electricity demand as consumed by the flue gas blower. Through this economic analysis, it is observed that the total electricity consumption is proportional to the amount of CO2 captured when the rate of CO2 capture is consistent. As a result, the "electric energy consumption" per mass of CO2 avoided and "electricity cost per mass of CO2 avoided" is the same for the three cases.

Table 1 also shows the CAPEX per mass of CO2 avoided for the three case studies. Using a capital cost annualization factor of 0.1, the CAPEX per mass of CO2 avoided for Case 1 is estimated to be 8.8 €/t-CO2avoided. In Case 2, with the addition of the waste heat boiler and steam delivery system, the CAPEX increases to 9.9 €/t-CO2avoided. Finally, in Case 3, since only 35% of the flue gas is treated, the reduced scale of the capture process increases the CAPEX per mass of CO2 avoided to 15.9 €/t-CO2avoided. The change in "other variable OPEX costs" across the three cases is due to the change in levelized cost of maintenance materials. A value of 4% of the total installed plant cost, which was provided in the Benchmark Indicator Report, is used for this category.

At this point in time, there are no economic indicators for conventional CO2 capture technology (e.g. amine solvents) applied to the same cement plant, thus a direct assessment of how RTI's technology compares to conventional CO2 capture cannot be made. However, for conventional amines applied to post-combustion coal-fired power plant flue gas, most references suggest that the CO2 avoided cost falls in the range of 40 to 60 €/t-CO2 avoided. From this Norcem Project economic analysis, it is clear that CO2 capture at a typical cement plant may be even more expensive due to > 4x lower flue gas and CO2 flow rate (i.e. less benefit from economy-of-scale) and lack of steam on-site at the

cement plant (i.e. necessitating additional equipment) - thus the CO2 avoided cost range for conventional amines may be even higher. RTI's technology economic indicators fall at the lower end of the noted power application cost range which shows that the technology is economically competitive with both conventional and next generation CO2 capture technologies.

3.5. Next steps for technical and economic analyses

The results of the current economic analysis are based on certain undemonstrated, but achievable, sorbent and process performance assumptions including sorbent working capacity, sorbent makeup rate, operating temperatures, gas and solid residence time, and regeneration CO2 partial pressure. Key performance data will be generated during a pilot-scale demonstration phase of RTI's work efforts and will be used to augment the initial assumptions and update this preliminary economic analysis.

4. Contaminant testing at RTI

Prior to conducting field testing in Norway, it was determined that data on the sorbent performance under simulated cement flue gas conditions and sorbent stability in the presence of contaminants typically present in cement flue gas would be required to help design a lab-scale field test unit. The objectives of this testing, performed entirely in the United States using RTI's existing packed-bed reactor test units, were to compare sorbent performance to simulated coal-fired flue gas, measure sorbent stability in the presence of contaminants, and use performance data to evaluate the need and design of a proper pretreatment system for our technology.

4.1. Experimental: materials, test systems, and procedures

The CO2 capture sorbent used in the lab-scale contaminant testing (as well as other activities in this project) consists of two parts: (i) an amine component and (ii) a porous solid support. The amine (e.g. polyethyleneimine, PEI) acts as the active component for CO2 absorption while the porous solid support (e.g. silica) provides the pore volume and/or surface area on which the amine is supported and/or grafted.

In order to evaluate the CO2 capture performance of RTI's sorbents under simulated cement flue gas conditions, absorption and regeneration experiments were performed using an existing, automated packed-bed reactor (PBR) system developed by RTI. The lab-scale PBR system consists of four main sections: 1) Feed Gas Generation, 2) Packed-Bed Reactor, 3) Gas Switching Valves, and 4) Gas Analysis. The PBR feed gas generation system consists of a bank of electronic gas mass flow controllers and a temperature controlled water saturator. This arrangement allows for the generation of a wide range of feed gas compositions including wet and dry gas mixtures. During testing, the sorbent is contained within the PBR and a thermocouple is situated within the sorbent bed to measure the internal material temperature during absorption and regeneration of CO2. The CO2 concentration in the reactor effluent is continuously monitored using a Horiba VA-3000 CO2 analyzer.

CO2 capture experiments under simulated cement flue gas conditions were carried out using ~2g of and performing 100 cycles of CO2 absorption and regeneration. A "cycle" of CO2 absorption/regeneration consists of 4 steps: (i) CO2 absorption, (ii) system purge, (iii) sorbent regeneration, and (iv) system cooling. Typical simulated flue gas conditions used in absorption testing is provided in Table 2.

The effect of SO2 on the sorbent stability and the sorbent's CO2 loading performance were evaluated by adding 100 ppm of SO2 in the simulated flue gas and performing 100 cycles of absorption/regeneration on a fresh CO2 capture sorbent following the procedure described above.

Table 2. Simulated flue gas conditions used in PBR testing.

Component "Clean" cement flue gas Simulated cement flue gas (with SO2) Simulated coal-derived flue gas

CO2 18% 18% 14.8%

H2O 12% 12% 5.7%

O2 9.2% 9.2% 2.6%

N2 Balance Balance Balance

SO2 0 100 ppm 0

4.2. Results: simulated "clean" cement flue gas testing

Initial CO2 capture experiments were conducted in the PBR system by simulating a "clean" cement flue gas having no contaminants. The primary purpose of these experiments was to evaluate sorbent performance at higher CO2 concentrations than evaluated previously. Since most of RTI's technology development work had been focused on coal-fired power plant application, comparison of the cement-focused PBR studies to results collected on previous coal-focused tests was of particular interest. Figure 2 exhibits CO2 loading data for the solid sorbent over 100 cycles of CO2 absorption/regeneration in the presence of "clean" cement flue gas and overlays a CO2 loading data set obtained from previous testing under simulated coal-fired flue gas. In general, the coal-fired testing exhibited only slightly higher CO2 loading compared to our current result. The sorbent showed initial loading of 9.9 wt% and slowly dropped to about 9.7% toward the end of experiment. In addition, the fairly significant differences in water and O2 concentration also seem to have no beneficial or detrimental impact on the CO2 loading performance.

no SO2 Runl ♦coal-fired flue gas

40 60 80

Cycle Number

Fig. 2. CO2 loading performance comparison with simulated cement flue gas and simulated coal-derived flue gas.

4.3. Results: simulated cement flue gas testing with SO2 and NOx

In order to evaluate the impact of contaminants typically present in cement plant flue gas, sulfur dioxide (SO2) was added to the simulated cement flue gas stream at a level of 100 ppm. This SO2 concentration was selected as an 'average' concentration that may be found in a typical cement plant. Figure 3 shows that the addition of 100 ppm SO2 in the flue gas stream caused the sorbent's CO2 loading to drop by roughly 30% from its initial loading capacity over 100 absorption/regeneration cycles. The "spent" sorbent after SO2 exposure was visually examined and revealed that the portion of the sorbent located at the reactor entrance exhibited a yellow discoloration (compared to the white color

of the fresh sorbent). Discoloration of the sorbent at the entrance of reactor suggests the formation of a stable sulfate salt - the result of a reaction between amine and SO2. It also suggests that the amine has a affinity for SO2 due to its higher acidity compared to CO2.

14 12 _ io

rc ° 6

no SO2 Runi 100 ppm SO2

20 40 60 80 100 120

Cycle Number

Fig. 3. CO2 loading performance comparison of with feed gas containing no SO2 and 100 ppm of SO2.

The PBR testing results clearly show that the presence of SO2 has an adverse effect on the CO2 loading performance of the solid sorbent. If allowed to accumulate on the sorbent, SO2 exposure would result in a higher sorbent make-up rate for the process, and thus a significantly higher operating cost. It is expected that the presence of NO and NO2 in cement flue gas will only increase the rate of sorbent degradation. Thus, it has been concluded that it is imperative to control and minimize these acidic contaminants and limit their interactions with RTI's CO2-absorbing sorbent through pretreatment/guard beds - both in lab-scale testing, pilot testing, and on up to operations at commercial scale.

5. Lab-scale sorbent exposure field testing in Norway

The primary objective of RTI's Phase I efforts within the Norcem CO2 Capture Project was to conduct small-scale field exposure testing of RTI's solid sorbents at Norcem's cement plant utilizing a flue gas slipstream taken from the actual plant exhaust. To this end, RTI designed, fabricated, constructed, delivered, installed, commissioned, and tested a lab-scale system - the Automated Sorbent Test Rig (ASTR) - at Norcem's cement plant. This test unit is capable of housing RTI's sorbent, accepting the cement plant flue gas, collecting a range of sorbent performance data, and is compatible with specific environmental, operational, and processing requirements encountered at Norcem's plant. Specific objectives of this lab-scale sorbent exposure field testing included:

• Evaluating sorbent performance under actual cement flue gas conditions.

• Evaluating sorbent stability and tolerance to various contaminants.

• Determining the efficacy of a pretreatment system for removal of acid gases from actual cement flue gas.

• Determining if sorbent exhibits any critical performance failure due to exposure to real cement flue gas.

• Gaining experience installing and commissioning a research unit in an industrial setting and proving that the system can operate in a stable manner.

• Determining the effectiveness of controlling and operating a research unit remotely - from a different country.

5.1. RTI's Lab-scale Automated Sorbent Test Rig (ASTR)

RTI's ASTR was designed to be a highly versatile system capable of evaluating sorbent materials using real exhaust gas streams under a broad range of process conditions in a highly automated manner. This test unit can handle complex,

multi-condition, multi-stage, multi-cycle experiments without significant hands-on operation. The ASTR system consists of four main sections: 1) Flue Gas Delivery, 2) Pretreament, 3) a Packed-bed Reactor, and 4) Gas Analysis as shown in the block flow diagram exhibited in Figure 4.

The ASTR is capable of continuously extracting real exhaust gas directly from a gas duct and pretreating the gas via compression, temperature control, and strong acid-gas removal (e.g. using guard beds) to continuously generate approximately 3-5 SLPM of conditioned exhaust gas. The conditioned exhaust gas stream is then fed to the ASTR's packed-bed reactor (PBR) containing the CO2 capture sorbent. Continuous exhaust gas extraction and pretreatment is necessary because of the rapid cycling nature of the PBR switching between absorption (using an exhaust gas feed containing CO2) and sorbent regeneration (using a H2O/N2 sweep gas) conditions. The composition of the raw and pretreated (following the guard bed) exhaust gas is monitored using in-line gas analyzers (i.e., monitoring the concentrations of CO2, SO2, NO, NO2).

Fig. 4. Block flow diagram of RTI's ASTR system

The ASTR PBR, a fixed bed reactor, contains a stationary bed of the sorbent and is operated in a continuous cycling mode switching between absorption and regeneration conditions. Under absorption conditions, approximately 1 SLPM of pretreated exhaust gas is fed to the PBR until the sorbent bed is saturated with CO2 as measured by a CO2 analyzer downstream of the PBR. Once the sorbent is saturated with CO2, the PBR switches to regeneration mode. The feed gas is switched to ~ 1 SLPM of humidified N2 and the temperature of the PBR is increased to approximately 120°C to regenerate the sorbent. Depending upon which mode the PBR is operating, the effluent gas will be either treated exhaust gas consisting primarily of O2 and N2 (absorption) - with little to no CO2 present - or regeneration off-gas consisting primarily of CO2 and N2 (regeneration). The PBR continues to switch between absorption and regeneration conditions until the desired number of cycles is achieved, a performance metric is achieved, or a shutdown procedure is initiated.

RTI worked closely with the Norcem team to understand all design and operational restrictions to be encountered at the Norwegian cement plant. Figure 5 shows the ASTR as constructed and installed in Norway.The ASTR was designed as a fully-automated system that requires minimal operator interaction during normal operation. As such, the system requires operational logic to perform experiments in an autonomous and safe manner. Additionally, the ASTR is designed to be operated and controlled remotely. Much of the operation of the ASTR during the various test phases were controlled by RTI staff. RTI is able to operate and control the ASTR (located in Norway) from a desktop computer, or even a smartphone, from the United States.

5.2. ASTR testing results

During the ASTR commissioning, RTI conducted an initial testing and stabilization campaign for the test unit using the delivered flue gas at the plant to confirm the proper function and stability of the system. The flue gas entering the ASTR was at a temperature of about 40°C containing 22 vol% CO2 and very low concentrations of SO2 and H2O. This initial ASTR testing phase consisted of collecting and analyzing 70 absorption/regeneration cycles. Figure 6 exhibits the CO2 loading capacity data of RTI's sorbent during this initial testing period. During the initial testing campaign, RTI's ASTR was able to maintain stable operation over 70+ cycles (with a few deviations due to fluctuations in gas flow and adjustments made for commissioning and shakedown purposes). RTI's ASTR sorbent maintained CO 2 loading performance between 4 and 5 wt% CO2 loading over the 70+ absorption/regeneration cycles - lower than the CO2 loading capacity between 6 to 7.5 wt% observed when testing the sorbent at RTI with "clean" simulated flue gas, but still within a reasonable range to conclude that the ASTR was capable of stable and continuous CO2 capture.

Fig. 5. RTI's ASTR system installed at Norcem's Brevik, Norway cement plant

9.00 8.00 7.00 6.00

^ 5.00 ro 4.00

g 3.00 и

2.00 1.00 0.00

Cycle Number

• CO2 Loading (/) 60 80

Fig. 6. CO2 loading performance of RTI's ASTR sorbent during the initial test campaign at Norcem

Following the commissioning test campaign, the ASTR was operated intermittently between various troubleshooting activities performed to improve the flue gas delivery conditions and to optimize the ASTR operating conditions. Following this troubleshooting and testing period, RTI concluded the Phase I work effort by conducting long-term, uninterrupted ASTR testing to collect a critical amount of CO2 capture data and to evaluate the sorbent performance during stable operation. Figure 7 exhibits the data collected during this final test campaign.

8.00 -

5? •

■M W

Jt 5.00

| 4.00 -

¡3 3.°° -

u 2.00

• CO2 Loading (%)

1.00 0.00

0 20 40 60 80

Cycle Number

Fig. 7. ASTR CO2 capture performance in extended test campaign at Norcem

Extended testing of the ASTR proved that the test unit can achieve stable CO2 loading performance in the range of 6 to 7 wt% loading. More importantly, the performance is stable with no critical failure of the sorbent observed during the testing campaigns. As referenced above, when testing the sorbent in RTI's labs, CO2 loading capacities between 6 to 7.5 wt% were observed, so the sorbent exhibited similar performance when utilizing actual cement flue gas. The sorbent performance at Norcem may be on the lower end of the noted range, but it is theorized that the performance was impacted slightly by low H2O water content in the flue gas - past testing at RTI has shown that H2O in the 3 to 4 vol% range enhances CO2 loading and performance stability over multiple cycles.

In total (across all testing phases), RTI's ASTR collected ~ 300 absorption/regeneration cycles with CO2 capture loading typically in the range of 5 to 7 wt% loading and the sorbent exhibited no critical failure in CO2 capture performance over these test campaigns..

6. Next steps: Field testing of RTI prototype system in Norway

The positive results from the economic analyses, contaminant testing at RTI, and sorbent exposure testing in Norway have collectively justified the continuation of RTI's work efforts into Phase II of the Norcem CO2 Capture Project. Phase II efforts are focused on a larger pilot-scale demonstration of RTI's process at Norcem's cement plant. This will involve the detailed design and engineering of a mobile pilot system and the subsequent execution of procurement, fabrication, construction, and commissioning activities. This pilot system will be installed within Norcem's cement plant, again accepting a slipstream of the plant's flue gas, and operated over extended parametric and long-term performance test campaigns. The overall objective is to demonstrate the long-term thermal, chemical, and physical stability and CO2 capture and sorbent regeneration performance of RTI's novel CO2 capture technology under cement plant flue gas exposure. This testing will allow RTI to collect critical process performance data that can be used in scale-up of the technology and for updating the assumptions and variables used in the economic analyses started in Phase I.

7. Conclusion

Overall, RTI's Phase I work efforts have resulted in furthering the development of an advanced solid sorbent-based CO2 capture technology, understanding the technical and economic advantages and challenges for cement applications, and executing a research project across multiple teams in different countries. A detailed technical and economic assessment was developed with the primary objective to evaluate the technical and economic feasibility of solid sorbent CO2 capture applied within the cement industry. The primary economic indicator calculated from this work -cost of CO2 avoided - was between 38 and 46 €/t-CO2avoided for RTI's technology - indicating the technology to be economically competitive with conventional CO2 capture technologies, which have been shown to have CO2 avoided costs in the range of 40 to 60 €/t-CO2 avoided for coal-fired power plant applications - a potentially lower cost application for CO2 capture given a > 4x higher flue gas and CO2 flow rate (i.e. more benefit from economy-of-scale) and the presence of steam on-site at the cement plant (i.e. eliminating the need for additional equipment, such as a boiler). The economic analysis also indicated that RTI's technology is a good candidate for waste heat utilization -potentially resulting in a 16% savings in CO2 avoided costs for the capture process.

Lab-scale simulated flue gas testing at RTI allowed for an evaluation of sorbent performance in flue gas conditions mimicking those present at Norcem Brevik, Norway cement plant as well as the impact of contaminants typically present in cement plant flue gas. This work demonstrated that CO2 capture performance is reduced significantly in the presence of SO2 and NOx - exhibiting a critical need for a pretreatment system in both small-scale testing at the Norcem plant and in the commercial embodiment of this technology. A flue gas pretreatment system was designed and added for the economic analysis as well as included in the Automated Sorbent Test Rig (ASTR) sent to Norway for real flue gas exposure testing.

The primary objective of RTI's Phase I work efforts was to conduct small-scale field exposure testing of RTI's novel solid sorbents at Norcem's cement plant utilizing a flue gas slipstream taken from the actual plant exhaust. To this end, RTI designed, fabricated, constructed, delivered, installed, commissioned, and tested the ASTR at Norcem's cement plant. Overall, the robust test unit collected ~ 300 absorption/regeneration cycles with CO2 capture loading typically in the range of 5 to 7 wt% loading - and more importantly with the sorbent exhibiting no critical failure in CO2 capture performance over an extended period testing.

As a collective, the positive results and knowledge gained from the Phase I work efforts have justified the continuation into Phase II of the Norcem CO2 Capture Project, in which RTI's scope of work consists of building and testing a much larger pilot system for field testing at Norcem's Brevik, Norway cement plant.

Acknowledgements

RTI would like to acknowledge Gassnova and Norcem AS for providing the funding for RTI's work efforts on the Norcem CO2 Capture Project. Additionally, RTI would like to acknowledge the United States Department of Energy/National Energy Laboratory (DOE/NETL) for on-going funding support to develop our solid sorbent CO2 capture technology from concept to the current level of deployment. RTI's project with the DOE/NETL is funded under a cooperative agreement and is managed by NETL's Carbon Capture Program. Abu Dhabi-based Masdar, a commercial enterprise with a clean energy and sustainable technology focus, is providing co-funding on the DOE-supported project. Lastly, RTI would like to acknowledge technology development partner, Pennsylvania State University, as the originator of the PEI-based sorbents and for their sorbent development and improvement work on the DOE-supported project.