Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 00 (2017) 000–000
www.elsevier.com/locate/procedia

13th International Conference on Greenhouse Gas Control Technologies, GHGT-13, 14-18

November 2016, Lausanne, Switzerland

Chilled Ammonia Process Scale-up and Lessons Learned

Ola Augustsson a, Barath Baburao, Sanjay Dube, Steve Bedell b, Peter Strunz, Michael
Balfe, Olaf Stallmann*c
a
GE Power, Växjö, Sweden
b
GE Power Ltd., Knoxville, TN, USA
c
GE Carbon Capture GmbH, Wiesbaden, Germany

Abstract

The GE Chilled Ammonia Process (CAP) is a post combustion CO2 capture technology that produces a high purity CO2 product
stream that can be utilized in the existing and new markets. The development of the CAP technology began with laboratory
bench-scale experiments to confirm that aqueous ammonia solution effectively absorbs CO2 with sufficiently low vapor phase
ammonia emissions at reduced temperatures. From these results, a technology development program was initiated to scale-up and
validate the process for commercialization.

The approach taken here is illustrative of industrial scale process development and improvement. For the CAP CO 2 capture
technology, the development effort involved an iterative approach as information from the different development stages was
obtained to set environmental & economic targets, develop predictive tools and models for process optimization, and to support
validation efforts at operating facilities. Over the course of the program, the technology was successfully tested on flue gasses
produced from coal, oil, and natural gas combustion, in addition to flue gas produced from refinery applications. Process know-
how and operational experience was gained and together with validated data from bench-scale and pilot plant facilities was
returned to push process design improvement and the development of predictive models. Currently, the CAP design is also
modified and extended into applications involving Urea, Methanol, and Soda Ash Production.

While many lessons learned and process improvement opportunities have been extracted from pilot plant and other test facilities,
pilot plant results and process modeling studies are still unveiling potential for further improvement. Optimization and integration
with the power generation facility occurred in the development phases of several FEED studies for large CCS plants. The CAP

* Corresponding author. Tel.: +49 6134 712 472; fax: +49 6134 712 590.
E-mail address: Olaf.Stallmann@ge.com

1876-6102 © 2017 The Authors. Published by Elsevier Ltd.

Peer-review under responsibility of the organizing committee of GHGT-13.
2 Author name / Energy Procedia 00 (2017) 000–000

design is ready for a demonstration-scale project and now is much improved from the design that was tested originally at
laboratory bench-scale. For example, the CAP scrubbing solution is now operated in a non-solids mode where precipitation is not
a part of the overall operational strategy and the process flow scheme is now modified and improved from original flow schemes
implemented at early pilot facilities to improve performance at reduced cost.

This work summarizes the bench-scale, pilot-scale, and validation facility results and offers insights into the lessons learned and
effort required bringing the technology into commercialization at an industrial scale. The lessons learned from each of the pilot
plants at different sizes are illustrated and the associated impact of the results from each pilot plant in the current CAP product
offering is also discussed. A summary of the important results from CAP test facilities including Stanford Research International
(SRI), WE energies, EONCAP-Karlshamm, AEP Mountaineer, TCM and GE’s pilot facility in Vaxjo, Sweden are presented.
Distinguishing features of the GE CAP are provided drawing comparison to open literature versions of ammonia based CO2
capture processes. Evolution of key performance parameters such as energy demand, product quality, solvent strength, process
flow scheme, etc., at the different plant sizes are also discussed. In addition, the current state of development for extending the
technology into areas where CO2 may be utilized productively are also be addressed in addition to the latest improvement
concepts currently being studied on CAP. Finally, the paper will also summarize the advantages of CAP as compared to
conventional amine based processes.

© 2017 The Authors. Published by Elsevier Ltd.

Peer-review under responsibility of the organizing committee of GHGT-13.

Keywords: GE; CO2; Capture Process; Post Combustion; CAP; Chilled Ammonia Process; Lessons Learned; Status; Scale-up; SRI; WE-Energy;
EONCAP-Karlshamn; AEP Mountaineer; TCM; Technology Center Mongstad; Thermal Energy Demand

1. Introduction

The GE Chilled Ammonia Process (CAP) is a post combustion CO 2 capture technology that produces a high
purity CO2 product stream that can be utilized in the existing and new markets. Primary objective of this CCS
technology is to address GE’s installed base of power plants but also new build and industrial emitters.
The approach taken by GE is illustrative of industrial scale process development and improvement. For the CAP
CO2 capture technology, the development effort involved an iterative approach as information from the different
development stages was obtained to set environmental & economic targets, develop predictive tools and models for
process optimization, and to support validation efforts at operating facilities.
Development of CAP technology began with laboratory bench-scale experiments to confirm that aqueous
ammonia solution effectively absorbs CO2 with sufficiently low vapor phase ammonia emissions at reduced
temperatures. From these results, a technology development program was initiated to scale-up and validate the
process for commercialization. Several pilot and validation facilities using the Chilled Ammonia Process with
increasing capacity per evolution step have been built and tested. The CAP plants treated combustion flue gases
from both power and industrial boilers using several different fuels. In combination, pilot plant and validation
facilities have operated for over 22,000 hours.
Optimization and integration with the power generation facility occurred in the development phases of several
FEED studies for large CCS plants. The CAP design is ready for a demonstration-scale project and now is much
improved from the design that was tested originally in the lab at bench-scale.

2. Nomenclature

CAP Chilled Ammonia Process PVF Product Validation Facility

CCS Carbon Capture and Storage RFCC Refinery Fluidized Catalytic Cracker
CHP Combined Heat and Power (Plant) SAFS Single Absorber Flow Scheme
FEED Front End Engineering and Design SRI Stanford Research International
MEA Methyl-Ethanol-Amine TCM Technology Centre Mongstad, Norway
PCC Post Combustion Capture VP Verification Plant
 Author name / Energy Procedia 00 (2017) 000–000 3

3. Technology

GE’s CAP is a wet-regenerable solvent process using ammonia-carbon dioxide based salts as the solvent. CO2 is
captured in an absorber and then cycled through a regeneration tower to drive off a concentrated stream of CO 2.
Traditionally water washing steps were also employed to recover ammonia.

The reaction mechanism of carbon-dioxide absorption with ammoniated solution has been reviewed in the
scientific community for many years. The following set of reactions, Equation 1 through 6, was used to describe the
chemical equilibrium.

2H 2O  H 3O   OH  (1)


CO2  2H 2O  H 3O   HCO3
(2)
 2
HCO3  H 2O  CO3  H3O (3)

NH 3  H 2O  NH 4  OH  (4)


NH 3  HCO3  H 2O  NH 2COO (5)

NH 4 HCO3 (s)  HCO3  NH 4 (6)

In the CAP absorber system, the reaction of gaseous carbon dioxide into the liquid phase results in reactions with
water and ammonia along with weak acid dissociation to form a temperature and concentration dependent
speciation, which after sufficient time approaches chemical equilibrium. The chilled ammonia process operating
with precipitated solids is designed around the reaction of flue gas CO2 with an aqueous ammonia solution to
precipitate ammonium bicarbonate NH4HCO3. The overall reversible chemical process is shown simplified in
Equations 7 and 8.

CO2  H 2O  ( NH 4 ) 2 CO3  2 NH 4 HCO3  Heat (7)

2 NH 4 HCO3  Heat  H 2O  ( NH 4 ) 2 CO3  CO2 (8)

The capture reaction (Equation 7) is optimal at a temperature between 40ºF and 59ºF (5ºC and 15ºC); to which
the flue gas must be chilled. Regeneration of the capture solution (Equation 8) and recovery of the captured CO2 is
optimal at a higher temperatures and moderate pressure. The regenerated solution, lean in ammonium bicarbonate, is
returned to the absorber system where it is re-used to capture CO2.
The process was originally conceived to operate the absorption at a temperature and ammonium bicarbonate
concentration such that a freely-suspended ammonium bicarbonate precipitate would form. Concentrated,
precipitated ammonium bicarbonate solids could then be used for regeneration. The increased CO2 loading (per unit
water) in the rich solution would reduce the sensible heat load needed for regeneration that is strongly influenced by
the carrier water. Later, pilot operations indicated that the presumed net energy benefit of precipitating solids was
marginal at best and in conjunction with operational stability issues. As a result, current designs foresee operation in
non-solids mode.
More details on the rationale for switching to non-solids mode will be described in subsequent sections of this
article. Nevertheless, still some investigations for solids operation are undertaken in the scientific community, e.g.
by Mazzotti et al. [1], which are also partnered by GE.
4 Author name / Energy Procedia 00 (2017) 000–000

The decision to further explore the potential of the CAP was also driven by its advantages over other solvent
based CO2 removal systems and features that differentiate it from those.

 Lower heat of regeneration (than MEA)

 Use of a low-cost commodity solvent known to utility operators
 Solvent which resists degradation from elevated temperature and exposure to O 2, SOX and NOX
 Use of a higher stripping temperature allows CAP to release CO 2 at a higher pressure leading to lower
downstream compression costs, compared to Amine Systems.
 Higher stripping temperature also increases the window of opportunity to integrate higher grade waste heat
directly, without the problems of degradation and the work-around by degrading the quality of heat.
 A potential reduction in the amount of energy required to capture CO 2, particularly in locations with lower-
temperature cooling water available.

3.1. Process

While the following process description assumes application on a coal-fired utility boiler, the CAP can be
modified, as required, for post-combustion CO2 removal from many different utility and industrial combustion
processes.
A schematic of the CAP is shown in Figure 1 below:

Figure 1: Initial solid mode CAP process

The CO2 capture system has the following main subsystems:

 Author name / Energy Procedia 00 (2017) 000–000 5

1. Flue Gas Cooling.

2. CO2 absorption.
3. High-Pressure Regeneration.

3.1.1. Flue Gas Cooling.

The flue gas exiting the coal power plants flue gas desulfurization plant (FGD) is typically between 120-140°F
(50-60°C). The gas is water saturated and it contains residual contaminants such as SO 2, NOx, HCl, sulfuric acid
mist and filterable and condensable particulate matter (PM). In order to cool the saturated flue gas, both sensible
heat and latent heat for water vapor condensation has to be removed. Direct cooling without heat exchangers using
a combination of cooling towers and mechanical chillers is an efficient and low cost method that results in
condensation of water and the capture of residual emissions from the flue gas. The pH of the water in the flue gas
cooling subsystem is controlled using an alkaline reagent.
The net water balance around the flue gas coolers, with moisture condensing in the direct cooler (DCC1) and
evaporating in the direct contact heater (DCH or DCC2) is balanced considering also a bleed of ammonium sulfate
bi-product solution.
The reduction in volume and mass of the flue gas has the benefit of reducing the size of the downstream
equipment. Likewise, the ID fan is positioned downstream of the cooling subsystem, minimizing its size and power
consumption.

3.1.2. CO2 Absorption

Traditionally the flue gas entering the CO2 absorber is cooled. The gas is relatively dry with less than 1%
moisture and contains very low concentrations of SOx, HCl and PM.
The CO2 absorber is designed to operate with a solution containing a dissolved and suspended mix of ammonium
carbonate and ammonium bicarbonate. The flue gas flows upwards against the falling solution in counter current
flow. Up to 90% of the CO2 from the flue gas can be captured in the absorber. The low concentration of ammonia
in the clean flue gas exiting the absorber system is captured by cold-water wash and returned to the absorber. The
clean flue gas, containing mainly nitrogen, excess oxygen and residual CO 2 flows to the chimney for venting to the
atmosphere.

3.1.3. High-Pressure Regeneration.

After solids separation from the bulk CO2-rich solution from the absorber, the stream routed to the Regenerator
contains mainly ammonium bicarbonate. The CO2 rich slurry is pumped through a heat exchanger, and enters the
high-pressure Regenerator. The pressure required for the CO 2 gas at the plant battery limit is typically 1,500 psi
(100 bar). This represents a compression ratio of around 100, relative to ambient conditions. In contrast, the
proposed process regenerates CO2 at 300 psi (20 bar). This reduces the required compression ratio from 100 to 5,
resulting in a compression train that has fewer stages and consumes less power.
The ammonium bicarbonate in the CO2 rich slurry dissolves as the temperature increases in the heat exchanger
and turns into a clear solution at temperatures above 175°F (80°C). The hot solution is injected into the Regenerator
which is a high-pressure CO2 stripping vessel. Additional heat for stripping the CO2 is provided from a reboiler that
consumes low pressure steam. Bench scale testing has demonstrated that the CO2 gas from the regenerator and
resulting wash system is extremely pure, containing more than 99% CO 2 and extremely low residual concentrations
of ammonia and water.
6 Author name / Energy Procedia 00 (2017) 000–000

3.1.4. Non solid mode process

Over the development phase the CAP process changed operation from solid mode, that is operating the rich
solvent loading beyond the solubility limit of ammonium salts to form slurry, to non-solids mode, where the solvent
CO2 loading is below the crystallization point.

The changes in the process can be seen in Figure 2.

Figure 2: Non-solids flow scheme

The obvious difference to non-solid mode is the omission of the solid handling equipment, like the hydrocyclone
used to concentrate the slurry from the absorber.

4. Industrial scale process development and improvement

GE has established a general approach in the development of new products and processes in the Environmental
Control Systems Sector. This approach formalizes the steps to be taken during the R&D phase.
For any process development, there is certain minimum information that is needed to allow a reliable design and
performance estimation for the commercial product or process. This typically includes:

1. Boundary Conditions (Design Envelope) for the targeted application

2. Thermodynamic base information covering the full range of compositions and operating conditions of the process
3. Calculation models for all relevant chemical reactions
 Author name / Energy Procedia 00 (2017) 000–000 7

4. Calculation methods and/or specification guidelines for equipment

5. Information on process specific aspects not covered by standard estimation methods (Simulation tools, etc.):
5.1. Fouling / scaling behavior of the fluids used.
5.2. Unit operation specific features (e.g. heat leaks or ingresses)
5.3. Operations feedback on process / equipment design and flexibility requirements

Where no other sources of information (such as existing simulators, plants etc.) are available, the following
methods are generally used to obtain necessary data:
 Literature data (satisfy information requirements of items 1-3)
 Experiments at Laboratory or Bench scale (satisfy information requirements of items 1-3)
 Lab pilot plant tests (satisfy information requirements of items 1-4; 5.1; 5.2)
 Field pilot plant tests (satisfy information requirements of items 1-5 all)
 Demonstration plant operation (To confirm underlying assessments at full scale and further detail / optimize
items 1-5 all)

Already existing knowledge of certain aspects of a technology in a company may allow shortcuts in the
development to be taken without increasing the risk position. In such cases the development comprises an adaptation
of an existing process or process modules to new boundary conditions rather than a full-fledged new development.
On the road to building full scale carbon capture facilities for power utilities, increasing capacities of pilot and
demonstration plants are necessary to minimize the risk associated with the knowledge available at the time of their
construction and use. The utility of new concepts is generally required to be proven, first in laboratory tests or at
bench scale, before a further development effort is initiated.
The development plan of the CAP technology as shown in Figure 3 shall serve as an illustrative example of this
approach. The program objectives at the start in 2004 were to provide a commercial product for CO2 capture using
the CAP technology by the end of 2015.

Figure 3: Initial CAP technology development plan

8 Author name / Energy Procedia 00 (2017) 000–000

In order to reach this time-to-market requirement a rather tight schedule for the execution of the different test
phases was defined. This resulted in some overlap in the execution of the different test programs and required
proactive management of the knowledge transfer from one test program to the other.
Several pilot and validation facilities using the Chilled Ammonia Process have been completed to date. The CAP
plants treated combustion flue gases from both power and industrial boilers using several different fuels. In
combination, the pilot plant and validation facilities have operated for over 22,000 hours. A summary of the plants is
provided in Table 1.

Table 1. CAP Pilot and Validation Facility Summary.

Plant Operating Size Fuel
Period
MWel TPY
SRI International 2005-2008 0.25 NA Synthetic gas
Bench-Scale and Pilot Plant
WE Energies Pilot Plant 2008-2009 1.7 13,000 Coal
EONCAP Karlshamn Pilot Plant 2009-2010 1.7 11,000 Heavy oil
GE Laboratory Facility 2009- present 0.25 NA Synthetic gas
AEP Mountaineer Product 2009-2011 20 100,000 Coal
Validation Facility
Technology Centre Mongstad 2012-2014 16 RFCC 80000 Refinery Residue Fluid
Validation Facility CHP 22000 Catalytic Cracker
(RFCC) off-gas and
natural gas combined
heat and power (CHP)

5. The different test facilities and lessons learned

After initial tests to show the principle of CO2 capture using Ammonia-salt based solvents, a small scale bench pilot
plant was used at SRI to further evaluate the chemistry and define the technical hypothesis for further investigations.

5.1. SRI / Växjö facility

The objectives for the SRI facility were focused on proving the
chemistry concept in a CAP batch mode setup. The facility proved that the
CAP process is able to capture CO 2 in solid and non-solid mode. Test
results provided the preliminary mass and energy balances, which were
used to build the WE CAP pilot in Wisconsin.
The large bench scale pilot became operational at SRI in November
2006. Validation activities at SRI were conducted according to the test plan
that was outlined by the field validation team. The primary objectives of
this pilot were:
 Investigating the performance of the chilled ammonia process (CAP) for
capturing CO2 from power plant flue gases;
 Obtaining relevant design data for a pilot scale (1 to 5 MW) system;
 Evaluate absorption solvent regeneration conditions with a batch
regenerator;
Picture 1: SRI Absorber Pilot
 Create an inventory of regenerated solvent for injection in the absorber
system.
 Author name / Energy Procedia 00 (2017) 000–000 9

The construction comprised an 18” diameter large-scale bench absorber system and its auxiliary equipment (gas
and system coolers and water-wash). The batch testing of the absorber system started on November 9, 2006 when
the first test run was performed. Different rates of inlet CO 2 concentration, molarity of the solvent system, absorber
temperature and a wide-range of gas velocities to provide adequate residence time, L/G ratios, and sustained and
varying solvent loading were a few of the parameters that were used to get the data to enable more detailed
evaluations of the process.
Also, over time, many changes have been incorporated in the system to
determine the effect of the following on the CO2 capture efficiency:
 Absorber Packing, Contactor Mechanism, conical/cylindrical shape change of
absorber bottom – To determine the best gas-liquid contacting device for the
required duty.
 Alternative NH3 dosing methods
 Injection of promoters such as Piperazine or MEA.

Most of the runs were basically scouting experiments to understand the effect
various parameters have on the process and to help in preparing a reliable and less-
time consuming strategy for subsequent parametric tests.
Furthermore these tests helped to define the operating conditions for later plant
designs. Particularly the next level pilot design was based on the findings of these
SRI tests.
Picture 2: SRI Regenerator Pilot
The following results from the large bench scale tests were applied to the design
of the WE Energies field pilot:
 A multiple stage, two temperature CO2 absorber design to minimize packing volume to achieve 90 percent
removal, form ammoniumbicarbonate solids and manage ammonia slip to acceptable levels
 Structured packing definition
 Pressure criteria in the regenerator for producing required quality CO 2
 Water wash design
 Process logic and control mechanisms
 Use of lean solution in scrubbing CO2 from flue gas
 Baseline operating envelope for all key process parameters

Due to the small size it was not possible to resolute the behavior of the CAP process with regards to operational
issues. The formation of solids out of the gas phase containing NH3, CO2 and H2O on cold surfaces was detected and
first recommendations for design given. At this time operation in solids mode, having precipitated Ammonium salts
in the solution, was considered also for the next phase of the development.
After the test program at SRI was completed, the test plant was moved to GE’s Växjö laboratories in Sweden and
modified to reflect the latest findings also from the other pilot plants. For example, the operation in non-solid mode
was brought to the drawing board. Prior to the TCM Mongstad Verification Plant engineering a new idea for a
refinement of the non-solids operating mode, the single absorber flow scheme (SAFS), was evaluated and refined in
the Växjö Pilot.
10 Author name / Energy Procedia 00 (2017) 000–000

5.2. WE Energies

GE, EPRI and WE Energies announced in 2007 the development of a CO2 capture, pilot system to be installed
and operated at the WE Energies Pleasant Prairie Power Plant (P4), located in Pleasant Prairie, Wisconsin, US.
Picture 3 shows a picture of the P4 pilot. The P4 power plant was retrofitted with new wet FGD systems for control
of SO2 emissions and the retrofit included the construction of a new chimney.

As the technology developer, GE designed, constructed

and operated the pilot system. The pilot system captured
CO2 emissions from a slipstream of less than one percent of
the Unit 2 design gas flow. At 100% capacity, the pilot
system was designed to capture up to 15,000 tons of CO2
per year.

The field pilot at WE Energies was designed as a ‘proof

of concept’ facility with considerable operating flexibility to
test the different unit operations. The four key criteria,
necessary to validate the technical feasibility of the chilled
ammonia process include the following:
 CO2 removal efficiency (90%); Picture 3: WE Energy CAP pilot, Pleasant Prairie Power Plant

 Low ammonia slip from DCC2 overhead;

 High CO2 quality (with low ammonia slip and low moisture content);
 Low system pressure drop.

Additional objectives of the project included:

 Demonstrate full system operation on actual flue gas, including but not limited to: flue gas cooling using heat
recovery/exchange and chilling, removal of residual pollutants, CO 2 absorption and regeneration;
 Evaluate energy consumption relative to calculated values and to other CO 2 capture technologies;
 Operate the system long-term to identify O&M issues and establish system reliability;
 Conduct field tests to gather operating data from the system and develop objective, third-party techno-economic
analyses to refine current estimates for the performance and lifetime costs of a commercial system;

The pilot logged over 7,000 operating hours and, from January 2009 till October 2009, the pilot reliably operated
24 hours per day, 7 days per week. There had been a total of nine outages, described as follows:
 Two forced outages of the power plant (unrelated to the field pilot);
 One planned outage to provide the pilot operations and validation teams a break over the Christmas and New
Year holiday;
 Three planned outages to support additional modifications to the pilot plant; and
 Three forced outages to: a) to perform maintenance on the mechanical chiller; b) inspect and troubleshoot a
malfunctioning electric heater for the ammonia stripper; and c) replace a gasket on the regenerator column.

The experience in operating the field pilot was invaluable, as the operations and validation teams refined start-up
and shutdown procedures and troubleshooted issues with process operation.
The WE Energies pilot was a first of a kind pilot designed for continuous operation. During the initial months of
operation, the validation team identified a number of issues that required design modifications. As these
 Author name / Energy Procedia 00 (2017) 000–000 11

modifications were implemented, pilot performance steadily improved to the point that stable absorber operation at
100% of design flue gas flow was established in April 2009. From this point, the pilot demonstrated the ability to
meet the key performance objectives, defined earlier. [6]
In 2009, parametric tests were performed on the various pilot sub-systems, including: the ammonia stripper, the
CO2 regenerator, the water-wash column and the CO2 absorber column. As parametric testing was performed,
operating parameters were intentionally adjusted to generate empirical data to be used to predict performance as a
function of these parameters. While this testing provided a deeper understanding of the process, the unit did not
always operate at optimized conditions. WE Energies operation subsequently showed that the energy benefit for
regeneration was offset by increased chilling in absorption.
It must be said that the problems encountered during operation of the pilot, mainly with the formation of solids
outside the absorber were simply categorized as initial operational problems. As a result these problems were not
analyzed into a depth that would have revealed that also some technological issues were involved as root cause.
After review of the development program history this subject is one of the major lessons learned from these early
days.

5.3. EONCAP Karlshamn

The EONCAP pilot was in operation from 2009-04-17 to

2010-01-13, which are ~9⅔ months. During this time the
operation was handled by EON via the local operator KKAB,
to target both a lean operational budget and fulfill KKAB’s
demand that the Pilot be operated accordingly to their standards
& routines. However, test planning, validation and management
was organized by GE.
Initially EONCAP’s operation was focused on energy flow
and efficiency, during the autumn it received requests from the
WE pilot to validate their data and the planned validation
program was aborted and work redirected to what became
called the WE-tests.
From the start of operation, EONCAP focused first on Picture 4: EONCAP Karlshamn Pilot Facility
operational issues that had been experienced in the past, getting
the water balance in control & avoiding issues with plugging. It
did well to avoid plugging as the plant was completely housed, and implemented strategies were successful for
achieving a stable water inventory. With increased operational stability, the issue of ionic buildups and its handling
strategies were started.
During the complete operational time of the CAP, only one truckload of process fluids were shipped out and this
seems to have been caused by a mal operated manual DI-water valve. For the rest of the operational period all
operational wastewater, spills and any rain that entered the containment were handled internally.
For the ionic buildups and the resulting problems with low pH due to ammonium bisulfate that formed at
elevated temperatures, EONCAP started to plan for an appendix stripper to control the ionic balance in the system.
This system was not installed before EONCAP was mothballed.
During the operation of the pilot it became clear that operation in solids mode sets extraordinary requirements
towards process monitoring, control and operation to prevent solid formation in areas, such as accumulations in the
absorber packing, where it would result in transients and eventually process upsets. Furthermore, the formation of
Ammonium carbamate and its effect on materials of construction had been underestimated. In the end increased
corrosion and equipment tightness was affected. The Karlshamn pilot was dismantled after the test program.
Therefore further measures were developed to address these challenges in the almost parallel executed project for
the Mountaineer PVF facility and any other subsequent CAP installation. As deposition of solids was experienced in
piping elements that were normally not in operation, tracing at adequate temperatures was foreseen to ensure
decomposition and avoid precipitation of ammonium salts.
12 Author name / Energy Procedia 00 (2017) 000–000

Aerosols became early on an issue that launched a special program at EONCAP, since this Pilot was the only one
which at this time had both adjustable flue gases and a dedicated stack where aerosols could observed. The aerosol
program included documentation, sampling procedure development, and parametric testing campaigns. Several
strategies and designs to handle aerosols, and circumvent potential plugging from process fluids were started. Since
this program grew and evolved it was not completed within the EONCAP’s operational period.

5.4. AEP Mountaineer Product Validation Facility

A CO2 capture and storage (CCS) pilot plant was constructed at American Electric Power’s (AEP) 1300 MW el
Mountaineer station in New Haven, West Virginia, employing GE Power’s Chilled Ammonia Process (CAP). The
CAP Product Validation Facility (PVF) is approximately a 12-fold scale-up from the 1.7-MW research and
development pilot tested during 2008-9 at WE Energies’ P4 in Wisconsin. The AEP unit was designed to provide
about 110,000 tons CO2/year (100,000 metric tons CO2/year) for injection into geological strata under the
Mountaineer station.

Approximately 1.5% of the full load flue gas flow

leaving the Mountaineer station wet scrubber
(corresponding to approximately 20 MW el) is extracted
and sent to the PVF for CO2 capture and compression.
The product CO2 is then provided to an on-site
injection, storage, and monitoring program. Treated
flue gas, less the captured CO2, is returned to the
Mountaineer station stack. The Mountaineer wet
scrubber, stack, flue gas supply and return ductwork
and the PVF plant are shown in Picture 5.

The captured CO2 was injected into two different

geologic formations via two wells located within the
plant boundary: Rose Run at ~7800 ft. (2380 m) and
Copper Ridge at ~8200 ft. (2500 m). The Copper
Ridge formation performance has exceeded
Picture 5: AEP Mountaineer PVF
expectations, accepting the CO2 at relatively low
injection pressures (~76 bar) with little increase in the
formation pressure (~4 bar increase). The Rose Run formation was initially more resistant to injection than was
expected, but its performance has improved over time, with injection pressures around 76 bar and formation
pressure increases around 13 bar. Three deep monitoring wells were drilled and equipped to monitor CO 2
containment, track carbon storage footprint, and measure downhole properties. [6]

The post-combustion-capture (PCC) retrofit demonstration using GE’s CAP operated for 7900 hours from
September 2009 to May 2011 on the 20-MW equivalent slipstream. During the demonstration, CO 2 capture reached
the designed potential of up to 100,000 metric tons/year. Having been designed for 75% CO2 capture efficiency, the
CAP achieved capture efficiency between 75%–90% at purity of more than 99.9% during the demonstration. Over
50,000 metric tons of CO2 were captured and 37,000 metric tons were injected into permanent storage.

The goal of the capture part of this project was to be able to better judge the adequacy of the design and
performance objectives for the chilled ammonia process through the information obtained on:
 Emissions of all media and pollutants and consumables
 Energy demand, in what form and possibilities for thermal integration into a given power plant
 Trade-offs between emission reductions and energy/reagent consumption
 Finally, plant operability and reliability.
 Author name / Energy Procedia 00 (2017) 000–000 13

Figure 4: Flow diagram of AEP Mountaineer PVF facility

The following key features/modifications were incorporated into the plant to accomplish these goals:
 The PVF design was adjusted to operate in non-solids mode (Based on the experiences from WE Energy
and EONCAP Karlshamn pilots).
 A two absorber system to capture 75% of the flue gas CO2 achieving the total objective of 100,000
tonnes/yr was installed.
 The Water Wash was reduced from three recirculated beds in series at WE Energies to two beds. The top
bed is once through and serves as a polishing stage. The lower bed is a re-circulated system where the
primary amount of NH3 is absorbed
 The refrigeration system includes direct gas cooling with Refrigerant and free cooling during the months
with lower ambient temperatures. Two R-410a refrigeration systems are provided, Low Temperature for
NH3 capture and High Temperature to remove heat of reaction. This configuration improves refrigeration
system efficiency.
 CO2 exits the top of the regenerator at 21 barg with less than 50 ppmv NH3. Moisture in this stream is
reduced to less than 600 ppmv in a CO2 chiller prior to compression. Dry CO2 product is pressured up to
100 barg in a reciprocating compressor.

The impact on energy penalty in non-solids operation mode was found to be minor and partly to be compensated
by proper process operation and heat integration system design. As Ammonia slip into both flue gas and CO2
product was properly addressed through process design, other routes for Ammonia losses came into focus.
Ammonium bisulphate forms in the CAP solution as heat stable salt solvent withdrawn from regenerator to balance
accumulation of impurities SOx slipping DCC and being captured in the Absorber. The continuous accumulation of
this heat stable salt becomes considerable in the amount of ammonia bound, and requires continuous purging.
Therefore another stripper, the appendix stripper, was developed. The use of the appendix stripper was realized in
14 Author name / Energy Procedia 00 (2017) 000–000

this pilot plant allowing reliable main stripper operation and further decreased ammonia emissions and reduced
losses into the ammonium bi-sulfate by-product solvent draw.

Furthermore, conclusions from material tests conducted at the before mentioned pilots found way into the PVF
design increasing reliability and availability.

5.5. CAP Plant at Mongstad TCM

GE has installed its CAP system at another CO2 capture validation

facility, Technology Centre Mongstad (TCM). TCM, which is owned by
Gassnova, Statoil, Shell and Sasol, was the world’s largest facility for
testing of Carbon Capture technologies at its start. The center is located
next to the Mongstad Refinery on the west coast of Norway. The
commissioning of the CAP CO2 Capture validation plant (VP)
commenced in early 2012 with initial start-up in October, 2012. The CO2
capture plant completed its operation and test campaign in August, 2014
with over 6000 hours of operation. A picture of the CAP facility at TCM
is shown in Picture 6.
The unique location of TCM, next to the refinery, provided interesting
opportunities in terms of gases the availability of gas sources to be treated
by the installed CO2 capture plants. The CAP plant at TCM was designed
to treat both refinery off-gas from a cracker operation as well as the
exhaust from a gas turbine based combined heat and power plant.

The plant, which initially was designed in 2008, captured over

39,500 tons of CO2 and achieved capture efficiencies of 85-87% at design
conditions. [8]
The initial test period foreseen covered 13 months, with an extension Picture 6: TCM Mongstad CAP Pilot
option for additional 6 months, based on the test program developed by
GE.

The test program focus was on:

 Process optimization, including: energy efficiency, ammonia consumption and low emissions
 Steady state and transient operations
 Impact of flue gas impurities.

The goal of this pilot plant was to establish an energy efficient process and mitigate risks when implementing a
full-scale Carbon Capture Unit. This unit furthermore was able to show performance of the CAP technology for
industrial applications and flue gases from gas firing.
The TCM Verification Plant design was based on preliminary results obtained from the large bench scale pilot
that was operated at SRI International and later re-located to Växjö. This design included the solids operating
concept that the WE Energies operation subsequently showed that the energy benefit for regeneration was offset by
increased chilling in absorption. Residual elements of this design carried an energy penalty for the TCM
Verification Plant, which were reduced but not eliminated with the modifications implemented.
The TCM Plant was also designed to consider two flue gas sources RFCC flue gas from the refinery with
relatively high concentrations of CO2 and CHP gas with relatively low concentrations of CO2. An emphasis on
obtaining the capture rate and capacity for the RFCC case provided oversized refrigeration and regeneration systems
which reduced the efficiency of the CHP operation. However, by providing the unit with the ability to operate
 Author name / Energy Procedia 00 (2017) 000–000 15

significantly above design margins, the need for operator intervention was minimized. An optimization program
identified the areas where such margins are not needed and reduce them where possible.
During initial operation solid formation was discovered in stagnant piping systems and safety devices which were
relics of an obsolete solids-operation mode. Based on better understanding of the mechanism of solid formation in
different solution and vapor locations with the CAP solution, a solids formation mitigation plan was implemented
successfully to control the temperature in stagnant pipe sections. The required countermeasures were taken and
incorporated into the design standard.
During operation of the Mongstad CAP plant the unit demonstrated to be very robust and forgiving of upsets
including:
 Unit trips at the refinery
 Loss of electrical power
 Control system communication failure
 Flue gas fan trips
 Wide variations in flue gas composition from the RFCC flue gas

The CAP plant normally recovered from significant operating upsets within a matter of one to a few hours. In the
case of substantial RFCC flue gas composition variations, the CAP pilot proved it could continue to operate with no
significant aerosol based environmental emissions and capture CO2 efficiently.
Finally, there was the finding from the AEP MTN PVF facility during which the simulation model was validated
and optimization studies were undertaken leading to an improved absorber design, this improved SAFS design, was
finally validated at the Växjö laboratories in a dedicated test campaign as previously discussed.

6. Test Results

In the following chapters key results that have been retrieved from the pilot operation shall be presented. These
results led a good portion of the decision making in the program execution.

6.1. Results from SRI Large Bench Scale implemented In the WE Energies Design

One of the main questions to be answered already in the initial phase was the behavior of the Ammonia and its
potential losses into the flue gas.
Therefore a parametric test at different operating
3.0
temperatures varying the loading of the solvent was
executed. Here the loading is the molar ratio of CO2 to 2.5
Ammonia species in the solution
The test results in Figure 5 show that with decreasing 2.0
loading, which is also associated with increasing free
NH3(g), vol%

Ammonia in solution, emissions into the vapor phase are 1.5

increasing. A single absorption system will thus not be

1.0
able to meet the required emission limits while having a
sufficient high capture rate within feasible limits. Based 0.5
on these findings a dual absorber system was planned for
the next bigger pilot plant size. 0.0
Further tests included the CO2 capture efficiency in 1.0 0.83 0.71 0.63 0.56 0.5
Loading
dependence of the loading, residence time, packing
height and some other parameters which in the end gave Figure 5: Ammonia slip in relation to solvent loading
16 Author name / Energy Procedia 00 (2017) 000–000

a good starting point for the planning of the next phase in the CAP development program – the field pilot test
program.
The regenerator tests were done in a batch and semi-continuous column system. Objective was to study the
temperature dependency of CO2 and Ammonia release during regeneration. Therefore the test program did not
include the measurement of the required regeneration energy demand. With this in mind one of the immediate areas
of focus for the development team at the time (2007-2008) was to understand and validate the CAP solution/slurry
heat capacities and heat of reaction aspects of the process. Academic collaboration with several groups were
engaged to investigate details of speciation and energy, leading to an early internal energy target for total heating
requirements of 1.9 GJ/ t CO2.
After installation of the system in Växjö also regenerator and stripper energy demand were measured. The system
required about 7 GJ/tCO2.

6.2. Results from WE Energies Design

Despite the operational experiences gained, as

described earlier, the test campaigns conducted also gave
insight into various process parameters effects.
One of the most important results was the capture
efficiency in comparison to the design value.
The WE Energy pilot plant was designed to capture
about 90% of the incoming CO2 in solids mode. During
transformation to non-solids mode this design value
dropped to 80% because of limitations in some of the
existing equipment. As can be seen in Figure 6 this latter
value could be confirmed during the test campaigns. It
was found that the capture efficiency is not so much
depending on the recirculation rate of the solvent but on
the loading of the solution.
Based on these results the simulation model could be Figure 6: CO2 capture efficiency and absorber B recycle variation
validated and its predictive capabilities improved.

Another question to be answered was the behavior of the Ammonia and its losses into the flue gas based on the
previous test program at SRI.
The design anticipated an Ammonia slip in the flue gas at the battery limit of less than 10 ppmv which could be
proven during plant operation. As shown in Figure 7 the Ammonia concentration in the flue gas leaving the DCC2
could be kept between 5 ppmv to 8 ppmv.
The thermal energy demand for the regenerator and
the stripper during solid mode operation of the pilot plant
with flue gas from a coal fired source was determined to
be at 2.5 GJ/tCO2, having an average CO2 capture
efficiency of 82%. This was substantially higher than the
expected 1.9 GJ/tCO2 based on internal theoretical
evaluations.
Based on test results and application of an updated
simulation model, an evaluation of a CAP system for a
reference 800 MW nominal power plant was made. Due
to various limiting factors during scale-up multiple
process trains were considered. Some of the more
obvious limiting factors identified at that time are listed Figure 7: Ammonia emissions via flue gas
below:
 Author name / Energy Procedia 00 (2017) 000–000 17

 Absorber vessel Diameters – There exist practical limitations on the largest diameter mass transfer devices that
can be cost effectively fabricated; this can be limited by: gas flow distribution issues, structural requirements
with spanning the vessels to support the internal mass transfer devices and other issues.
 Chiller Sizes – Larger chillers are custom designed and assembled; the largest chiller that is currently
manufactured is roughly 40,000 tons.
 Pumps and heat exchangers – There might exist size limitations with some of the high pressure slurry pumps and
heat exchangers
 Scale-Up methodology – Program management would ideally scale up the maximum train size on a gradual basis
to minimize the risk exposure on commercial projects. This would suggest initial commercial projects would be
limited to smaller train sizes.

Based upon the above, it was expected that the maximum size of a CO2 train for initial commercial projects
would be limited to around 200-400 MW in size. The various issues mentioned above were addressed after the pilot
operation and gradually the maximum size of a train rose to 400 MW and higher. The development for Wet Flue
Gas Desulphurization systems was a precedent for this approach.
Later process design work and system analysis eventual progressed the capacity of a single train to be suitable for
an 800 MW coal power plant. Nevertheless some of the equipment still required parallel installation.

6.3. Results from EONCAP Karlshamn

EONCAP Karlshamn pilot plant was designed to

remove at least 90% of the CO2 from the flue gas. The 100
Overall CO2 Capture Efficiency (%)

limited raw data was considered to plot CO2 capture 95

efficiency as function of flue gas flow rate. The pilot 90
plant demonstrated that at steady state conditions, the
85 Design: 90%
average overall CO2 efficiency is between 81 and 96% Removal
depending on the operating variables established during 80
test operation. As a result of achieving CO2 removal 75
efficiencies, EONCAP Karlshamn project demonstrated
70
the fundamental viability of the carbon capture 4000 5000 6000 7000 8000
technology in real world conditions such as the inevitable Flue Gas Flow Rate (m3/hr)
starts and stops of a large power plant, changes in
temperature and humidity and the environmental hurdles
Figure 8: Overall CO2 Capture Efficiency
that go along with using any chemical process
EONCAP Karlshamn operation confirmed that the
ammonia slip from water wash system and direct contact cooler system is controllable. Ammonia emissions from
the direct contact cooler (DCC2) are a strong function of pH, temperature and L/G ratio. At Karlshamn, the gas
analyzer system was not able to produce reliable results because of fog formation in the system. Due to the difficulty
of measuring this parameter, no reproducible emission measurement out of the DCC2 system was possible.
In the Chilled Ammonia Process, there was no sign of solvent degradation after regenerating CO 2 loaded rich
solvent at high temperatures and operating more than 2,000 hours. The testing at different CAP plants confirmed
that the typical gas contaminants (SOX, HCL, HF and NOX) do not degrade the CAP solvent.
The CAP solvent showed higher CO2 absorption capacity, high degradation resistance and lower corrosion rate
than MEA. As the solvent capacity establishes the solvent circulation rate, it has a major impact on the absorber
size, piping system, pumps and size of the regenerator system.
The thermal energy demand for the regenerator and the stripper during operation of the pilot plant in non-solids
mode and processing flue gas from an oil fired source was determined to be at 3.9 GJ/tCO2, having an average CO2
capture efficiency of 89%.
18 Author name / Energy Procedia 00 (2017) 000–000

6.4. Results from AEP Mountaineer

American Electric Power (AEP) and GE Power

successfully operated the CCS validation project at
AEP’s Mountaineer Plant in New Haven, WV. The
project, the world’s first facility to both capture and store
carbon dioxide (CO2) from a coal-fired power plant,
represents a successful scale-up of ten times the size of
previous field pilots (WE Energies Pleasant Prairie,
EONCAP Karlshamn).
All the achievements listed below were confirmed
during steady-state operation of the CCS validation plant.
The Availability was measured during 30 days in March
of 2011. The correction of the energy penalty numbers
Figure 9: CO2 capture rate at AEP Mountaineer PVF.
are based on project specific abnormalities (includes
agreed upon adjustments for heat exchanger
modifications, expected temperature approach) and were
reviewed and approved by the owner, American Electric
Power, AEP.
The first phase of the CO2 capture project, which
began capturing CO2 in September, 2009 and started
storing it in October, 2009, underwent mechanical and
equipment modifications in the Fall of 2010 that
contributed to the success of the project. The plant was
operated until June 30th, 2011 with a focus on
maximizing CO2 injection.
The plant was originally designed to capture 75% of
the CO2 in the flue gas. The test results in Figure 9 show
Figure 10: CO2 Product Quality
that this value was continuously reached at design flue
gas feed flow of about 185,000 lbs/h.
Also it could be shown (see Figure 10) that the product purity can be maintained reliably to fulfill even the
strictest requirements for enhanced oil recovery applications. At the same time the Ammonia slip into the CO 2
product is kept well below 20 ppmv so those unwanted losses are also minimized.

In optimization studies done during this project it was found that there is little energy advantage between
optimized CAP systems for solids versus non-solids operation. However, running a design based on solids operation
in non-solids mode significantly impacts energy consumption. For example the optimal internal heat recovery
configuration (i.e. rich/lean heat exchanger network) is different between a “solids” design and a “no-solids” design.
Therefore corrected values were determined in order to give representative figures for key performance indicators
like specific thermal energy consumption. Table 1 is giving an overview on the key performance indicators that
were accomplished. Based on the test results the potential for a CAP system, featuring a design optimized to the
Mountaineer operating scheme, was evaluated. The related expected energy consumption for such plant is listed as
“corrected” value.

An economic study was performed using the updated design basis for CAP from the work done during the
Mountaineer project. The study scaled up the technology to model its application and integration on a 750-MW net
USC PC base plant with 1100°F (593°C) steam conditions and 90% CO 2 capture. Based on the study, the efficiency
loss when applying CAP was calculated to be 9.5 percentage points and the increase in LCOE compared to the base
plant was nearly doubled (the LCOE increased by 59 $/MWh). Although, through continuous improvements and
technology developments the LCOE is projected to be significantly reduced in the future.
 Author name / Energy Procedia 00 (2017) 000–000 19

Table 1: PVF performance

Parameter Unit Value

CO2 purity % >99.9
Availability % 97
CO2 captured tons 51’173
CO2 stored tons 37’404
Capture rate (CR) % 65-85(designed for 75)
Specific Thermal Energy Consumption GJ/tCO2 2.8 (measured) at 75% CR
2.2 (corrected)
Ammonia Emission ppmv <10

If the cooling requirements can be reduced for chilling the ammonia, as is the case with the SAFS and also if
lower-temperature cooling water is available, the increase in LCOE can be as low as 50$/MWh.

6.5. Results from TCM Verification Plant, Mongstad

Despite all these changes in concepts along the way, the TCM Verification Plant delivered the performance for
more than 1900 hours. Two different flue gases were tested during this period. One flue gas was generated by a CHP
plant using natural gas feed. The other flue gas was derived from the refinery fluidized catalytic cracking
regenerator (RFCC). Major difference between these two flue gases is in the CO2 concentration; the CHP flue gas
contained about 3-5% of CO2, whereas the RFCC flue gas had CO2 concentrations of 13- 15.1%. It shall be
mentioned that TCM was GE’s first installation processing CO2 from industrial sources like the RFCC flue gas.
Overall, the TCM test campaign on CAP presented excellent opportunities to demonstrate the effectiveness of
this carbon capture technology and also some of the areas of improvement. The major accomplishment from this
campaign is the demonstration of the process effectiveness on different flue gas sources. During the several months
of CAP operation, the process was very effective in capturing CO 2 efficiently from both CHP as well as RFCC flue
gas. This was accomplished with no reduction in solvent quality or with no emission issues after initial start-up. The
NH3 emission from the flue gas stack as well as CO2 product was well within the environmental limits.

6.5.1. CHP gas testing

During the CHP campaign, 85% CO2 removal was
consistently achieved with the plant operating at 90%
design gas flow rate. The plant data was compared with
model predictions with some deviations observed and
reported.
Particularly mentioned should be the operating period
after week 8 of the CHP flue gas tests, where despite
significant adverse events, the unit was able to maintain
design flow rates and CO2 capture rates above 80%.
Taken together, these achievements showed the inherent
robustness and flexibility of the CAP technology in
meeting operating challenges and adjusting to changing
operating boundary conditions.
The Residual Flue Gas stream is the primary source
for NH3 discharge to the atmosphere. The level of Figure 11: TCM CO2 capture efficiency
ammonia present in the residual flue gas returned to the
atmosphere is an indicator of the effectiveness of the flue gas systems downstream of the absorbers. The process
objective for ammonia in this stream was 2 ppmv. Once the system settled down after the initial weeks of operation,
the readings showed variance in the 2 to 4 ppmv range.
20 Author name / Energy Procedia 00 (2017) 000–000

Different external events caused spikes in the

ammonia levels that can be noted from the charts. More
general fluctuations in the ammonia concentration in the
residual flue gas leaving the DCC2 correspond with
change-outs of the sulphuric acid totes.
During the CHP test program Ramboll, an analytical
company, executed an isokinetic sampling campaign. In
their report amines, nitrosamines, total N-Nitrosamine
(TONO) compounds, aldehydes, and ammonia
measurement results were given. The organic compounds
were all below the detectable limits for the instruments.
Ammonia was detected at 0.5 ppmv, which is less than
the 2 ppmv design objective.
Ammonia is also discharged with the CO2 product. Figure 12: TCM NH3 emissions via flue gas
The level of ammonia in the CO2 Product is an indicator
of the effectiveness of the regenerator. During the period where data was available, the process objective of 10 ppmv
was generally met. Occasional excursions were attributed to transient operating disruptions.

For CAP solution, ammonia is diluted from its delivered concentration to 17 wt%. The literature suggests that
aqueous ammonia systems below 20% concentration require minimal special handling or permitting. As CO 2 is
captured, free ammonia forms ions including ammonium, carbamate, carbonate, and bicarbonate to reduce its
concentration to less than 2 wt%.
Different CAP Unit Operations have different operating conditions. Absorbers operate at 1.1 bara and at
temperatures below 35°C. The Regenerator operates at 19.5 barg and 150°C. With the modifications in place, the
NH3 Stripper operates between 1.0 barg / 125°C and 5.0 barg / 165°C. At these conditions, the ionic solution is not
flammable.
These process conditions have been consistently demonstrated during the TCM Verification Plant operations,
without any significant degradation to the solvent in terms of capture efficiency. A higher than expected ammonia
loss from the final absorber resulted in increased sulfuric acid usage. This is expected at the partial pressure of CO2
operated at the top of the absorber column. This issue is the focus of an optimization program currently underway.
The thermal energy demand for the regenerator and the stripper during operation of the pilot plant in non-solids
mode and processing flue gas from a gas fired source was determined to be at 3.0 GJ/tCO2, having an average CO2
capture efficiency of 87.4%.

6.5.2. RFCC gas testing

From April to July, 2014, the CAP Unit at the TCM
operated a test campaign on RFCC flue gas and
demonstrated its ability to meet the design CO2 capture
efficiency of 85%. The influence of different process
parameters in the Regenerator, and Ammonia
Wash/Stripper systems was documented through a
parametric test series that focused on minimizing steam
demand and sulfuric acid usage as the primary dependent
variables.

The test series illuminated how each of the parameters

affected the primary objectives. At a high stripper
pressure, the final week of operation was sustained at Figure 13: RFCC Capture Efficiency
85% CO2 capture and a steam demand of 2.6 GJ/tCO2
 Author name / Energy Procedia 00 (2017) 000–000 21

removed, without correction for heat losses. During this period, a complete set of validation data was collected for
analysis. The data was reconciled using GE’s CAP simulation model where the measured specific heat consumption
of 2.6 GJ/tCO2 compared well compares to a simulated value of 2.3 GJ/t CO2 that considered with variation attributed
to deficiencies in heat recovery from the lean solution.

Table 2: RFCC test campaign parameters

Parameter UOM Value

Flue Gas Flow Rate Sm3/hr 24800 to 38500
CO2 Concentration in Flue Gas Vol % 13.0 to 15.1
Stripper Pressure barg 0.9 to 4.9
Regenerator Pressure barg 19
CO2 purity % >99.9
Capture rate (CR) % 83-89(designed for 75)
Specific Thermal Energy Consumption GJ/tCO2 2.6 (measured) at 85% CR
Ammonia Emission ppmv <10

The operating data for all key parameters showed the general capability of the technology for conforming to the
design basis. Taken together, these findings support the design capabilities of GE, which incorporates the lessons
learned from the TCM Verification Plant that operated in situ under Mongstad conditions with a local operations
team and a design team familiar with the tenets of site protocols and guidelines.

After the simulation tools were validated against the results of the TCM tests a study was performed using the
updated design basis for CAP from the work done during the Mongstad project. The study scaled up the technology
to model its application and integration on an 800-MW net USC PC base plant with 600°C steam conditions and
90% CO2 capture. The flow scheme featured non-solids operation. Based on the study the specific heat energy
consumption was determined to be as low as 2.2 GJ/tCO2, when cooling water at adequate temperatures is available.

7. Conclusion

The development of the CAP technology began with laboratory bench-scale experiments to confirm that aqueous
ammonia solution effectively absorbs CO2 with sufficiently low vapor phase ammonia emissions at reduced
temperatures. From these results, a technology development program was initiated to scale-up and validate the
process for commercialization.

The approach outlined in this article is illustrative of industrial scale process development and improvement. The
iterations made in the development effort, as information from the different development stages was obtained to set
environmental & economic targets, develop predictive tools and models for process optimization, and to support
validation efforts at operating facilities, were described.

The CAP technology was applied to capture CO2 from different flue gas sources: SRI/Vaxjo Bench Pilot
(synthetic gas containing CO2), WE Energies industrial scale CAP facility (Coal), EONCAP Karlshamn industrial
scale CAP CO2 capture facility (High Sulfur Oil), AEP Mountaineer PVF (Coal), TCM Mongstad (Gas, Industrial
Sources).

In the SRI pilot plant the basic process concepts of CO 2 capture using Ammonium carbonate, first in solid-mode,
later, after relocation to Växjö, also for non-solid mode were proven. Areas for undesired solid formation out of the
gas phase were detected. The WE Energy pilot was the first installation on real flue gas showing the capabilities of
the process with capture rates of up to 89% while having a low Ammonia slip. Solid-mode operation showed to be
challenging and induced the switch to non-solid-mode. With the EONCAP Karlshamn pilot, processing heavy oil
22 Author name / Energy Procedia 00 (2017) 000–000

flue gas with high SOx and NOx concentrations, the non-solid-mode operation was proven and valuable information
about materials of construction were gained. The AEP Mountaineer pilot was the first plant in the world to feature
the whole post CCS chain and furthermore design improvements, derived from Karlshamn and WE Energy
operation, were realized and showed effective. TCM Mongstad finally proved that the CAP can be operated under
feasible boundary conditions with a competitive performance also for flue gas sources other than utility boilers.

The evolution of the thermal energy demand can be seen from Figure 14. Associated with the improvement of the
thermal energy demand over the evolution of the CAP there was also an increase in capture efficiency. In the
beginning the plants were able to achieve about 75% in non-solids mode whereas the TCM Mongstad plant achieved
up to 90% CO2 capture.

There were improvements

10 20
Energy Demand [GJ/tCO2]
identified during the development 9 18

CO2 Production [t/h]

program that will be introduced in 8 16
the design for any future 7 > 20 t/h 14
demonstration project. This is only 6 12
possible because the fundamental 5 10
design tools and methods have 4 8
3 6
been validated and allow effective
2 4
implementation of new features. 1 2
Reference Plant calculations have 0 0

(RFCC, non-solid mode)

(Coal, non-solid mode)

(Coal, non-solid mode)

(Gas, non-solid mode)

(Coal, solid mode)
(Synth, non-solid mode)

(Heavy Oil, non-solid mode)

Växjö Pilot Regenerator

shown that with a dedicated CAP

AEP Mountaineer

TCM Mongstad
system operating in non-solids
WE Energy

EONCAP Karlshamn

Latest Scheme
TCM Mongstad
mode the specific heat
consumption for the post-
combustion capture of a coal fired
power plant is as low as
2.2 GJ/tCO2.

In comparison, conventional
amine-based solvents introduce a
variety of practical problems: Figure 14: Thermal Energy Demand and CO2 Production Capacity Evolution
costly amine solutions, a high rate
of corrosion of the process
equipment and a high rate of amine degradation in the presence of oxygen. In general oxidative degradation mostly
occurs at short times and low temperature with contact in the absorber, but at a rather low rate and at longer and
higher temperature in the stable regenerator. The degraded solvent has to be replaced with make-up and this can be a
significant cost in the other CO2 capture processes like the amine-based ones. The degradation characteristics and
environmental impact of the CAP solvent were also investigated. Oxidative degradation does not occur due to the
presence of oxygen in the flue gas. The formation of heat stable salts leading to a loss of the solvent occurs with the
co absorption of residual SOx entering the absorber with the CO2 rich flue gas.

Several Front End Engineering & Design (FEED) efforts for larger-scale demonstrations of the CAP process
have been carried out, with significant design efforts conducted for these projects, including the 1.5 million tons per
year AEP Mountaineer II project and another 1 million tonne per year CCS project in Alberta, Canada.

Currently, GE is extending CAP technology into other applications and industries involving Urea, Methanol,
Petroleum Refineries, and Soda Ash Production where product yield is further increased when flue gas CO 2 is
reused further downstream in the process. By integrating CAP technology into these processes, there is an
opportunity for improving production efficiency in a cost effective manner.
 Author name / Energy Procedia 00 (2017) 000–000 23

References

[1] Gazzani M, Sutter D, Mazzotti M. “Improving the Efficiency of a Chilled Ammonia CO2 Capture Plant Through Solid Formation: A
Thermodynamic Analysis” Energy Procedia; 2014:Vol. 63. p. 1084-1090
[2] Kozak, Fred, et al. "Chilled ammonia process for CO 2 capture." Energy Procedia 1.1, 2009: 1419-1426.
[3] Sherrick, Brian, et al. "CCS with Alstom’s Chilled Ammonia Process at AEP’s Mountaineer plant." proceedings of MEGA conference,
2008: Baltimore, Maryland
[4] Black, J., et al. "CCS project: Chilled Ammonia process at the AEP Mountaineer Plant." proceedings of COAL-GEN, 2010: Pittsburgh, PA
[5] Bollinger, Rob, et al. "CCS Project with Alstom’s Chilled Ammonia Process at AEP’s Mountaineer Plant." ALSTOM Power Systems, 2010:
[6] V. Telikapalli F. Kozak et al “CCS with the Alstom Chilled Ammonia Process Development Program -Field Pilot Results” Energy Procedia,
2011: Vol. 4.p. 273–281
[7] S. Jönsson, V. Telikapalli” Chilled Ammonia Process installed at the Technology Center Mongstad” GHGT-11 (11th International
Conference on Greenhouse Gas Control Technologies), 18-22 November 2012: Kyoto
[8] Muraskin, D., et al. “Operating Experience at Technology Centre Mongstad (TCM) with Alstom’s Chilled Ammonia Process CO 2 Capture
Technology” in 14th Annual CCUS Conference, 2015: Pittspurgh, PA.
[9] Lombardo, Gerard, Ritesh Agarwal, and Jalal Askander "Chilled Ammonia Process at Technology Center Mongstad–First Results." Energy,
2014: Vol. 51.p. 31-39.
[10] Ennenbach, F. and B. Baburao “Lessons learned TCM – CAP” in Milestone Mongstad, 2014: Bergen, Norway.
[11] Baburao, B., et al. “Chilled Ammonia Process Operation and Results from Pilot Plant at Technology Centre Mongstad” in 8th Trondheim
Conference on CO2 Capture, Transport and Storage TCCS-8, 2015: Trondheim, Norway.