(631a) Comparison of Processes for Separation and Sequestration of CO2 from the Combustion of Coal

    
CO₂is a major ?greenhouse? gas and combustion of
carbonaceous fuels is one of the primary sources of its addition to the
atmosphere.  The sequestration of CO₂ from combustion or other
sources is not currently prescribed by law and has no direct economic
advantages but a recent study by the National Academy of Engineering has listed
it among the principal technological challenges of the 21^st century.
This presentation examines the technical feasibility and economic costs of several
processes for the recovery, separation, and sequestration of CO₂ from
the combustion of coal. Our analysis is based on bituminous coal, which,
despite widespread opposition, is generally predicted to be a major source of
energy for the generation of electricity in the next several decades. The
Illinois Basin, which underlies Illinois, western Indiana, and western Kentucky
was arbitrarily chosen for this analysis because (1) it has both large and
accessible holdings of bituminous coal, (2) includes a number of geological
formations suitable for sequestration, and (3) produces 255 million metric tons
of CO₂ per year (11.7% of total US emissions), 92% of which
are from coal-fired power plants. (The other major sources include petroleum
refineries, cement plants, and miscellaneous heavy industries.)

    
The possible venues for sequestration include algae, the ocean depths, depleted
oil and gas reservoirs, unmineable coal seams, and deep saline formations.
Algae are most productive users of CO₂in the
plant world, but the potential capacity and lifetime of this form of
sequestration is difficult to estimate at the present. The algae can also be
burned, which produces thermal energy with no net production of CO₂.
However, before it can be burned the algae must be dried, which is very
energy-consumptive. Because of a lack of the necessary technical information or
industrial experience this alternative was not examined in detail. The choice
of the Illinois Basin as a site eliminated the possibility of sequestration in
the ocean. The oil and gas reservoirs, the coal seams, and the saline
formations have an estimated capacity of 282 million metric tons, 282 million
metric tons, and 4.14 billion metric tons of CO₂, respectively. Insertion
of CO₂ in the oil and gas reservoirs may enhance their
recovery and in coal seams may replace methane that can be recovered, but the fractional
offsets of the cost of sequestration appear to be negligibly small. In saline
formations, the CO₂ dissolves in part and may react in part to
form carbonates, but the latter do not appear to have any economic potential.
If all the electricity produced in the Illinois Basin were produced by the
combustion of coal, and if the corresponding production of CO₂
was all sequestered in the oil and gas reserves there they would be saturated
in 1.12 years. The coal seams have an almost equal capacity, while the saline
formations have a capacity of 16.2 years. This total of only 18.4 years is
discouraging at first sight, but it is the extreme case and provides a window
of opportunity in time to seek other solutions. Because of its greater
capacity, attention was focused on sequestration in the saline formations.

    
For the quantitative calculations of the combustion of coal and of the
separation and sequestration of CO₂, ?Old Ben 26? coal and a
conventional furnace and turbine capable of producing 500MW of electricity at
an overall efficiency of 36% and operated at 65% capacity on an annual basis, were
chosen as representative. The corresponding annual requirement of coal and
production of CO₂ were calculated to be 1.0 and 2.47 million metric
tons, respectively. These quantities amount to approximately 1% of those for
the whole basin. The flue gas produced by the combustion of this low-sulfur
bituminous coal contains a significant amount of ash, CO, and N₂,
and traces of NO_x and SO₂. The ash is
postulated to be removed by electrostatic precipitation, and the NO_x and
SO₂ catalytically. However, these costs were not assigned to
that of sequestration because they are required in its absence. The
concentration of CO was based on the use of 25% excess air. Gaseous N₂is not ordinarily considered a pollutant but it is in terms of the
sequestration of CO₂. It can be almost totally avoided in the
stack gas by burning the coal with oxygen rather than air, but at the price of
a whole set of problems of its own, including the high cost of oxygen and the
necessity of a massive recirculation of flue gas, or of some other expediency,
to reduce the flame temperature. This alternative was not examined in detail.
Therefore, the separation of CO₂ from the flue gas was considered to
be a necessary part of the process of sequestration.

    
Three processes for separation of the CO₂ from the resulting
?clean? flue gas were examined, namely absorption in an aqueous solution of
monoethylamine (MEA), adsorption by and reaction with calcium oxide  (CaO)
, and adsorption by and reaction with magnesium oxide(MgO). Because the
use of aqueous MEA as an absorbent for various chemicals is a well-developed
technology, it was adopted as the base case. The use of CaO as an
adsorbent and reactant is well developed but in the case of CO₂ it
has a serious shortcoming, namely the temperature of 1175K required for the decomposition
of the calcium carbonate (Ca₂CO₃) that is formed. Magnesium
oxide (MgO) was considered as a possible alternative to CaO because
the decomposition of magnesium carbonate (Mg₂CO₃) occurs
at the much lower temperature of 700K, but  the physical handling of powdered MgO
was a real concern. The discovery of a DOE report by Breault and Reasbeck that
suggested the impregnation of MgO on solid granules of alumina (Al₂O₃) resulted in its reconsideration, and our detailed calculations revealed
the resulting process to be more economical than those utilizing MEA or CaO.
These calculations were based on the recovery of 90% of the CO₂  in
245 kg/s of cleansed stack gas being fed from a pipeline at 311K and slightly
above 1bar.

    
The MEA system is conventional except for its size, consisting of 12 identical
18-tray columns, 4.27m in diameter and 15.85m high. On the other hand the MgO
system has two unique characteristics: the  preparation of the impregnated tablets
and the utilization of six identical packed-bed adsorber/reactor/desorbers, 5.0m
in diameter, 6.3m high, and operated in parallel in batch cycles at different
stages. To produce the combined MgO /Al₂O₃
granules,  alumina tablets 5mm in diameter are soaked for up
to 3 days in an aqueous solution of 20-80 gm/l  of magnesium hydroxide. The
water of hydration is then driven off by heating, resulting in a mass ratio of MgO/Al₂O₃
of 1.2. The total requirement for MgO supported on alumina in the six adsorbers
was estimated to be 990Mg with an annual replacement of 10% due to attrition.  Clean
stack gas at 384K and 1.5bar is compressed to 14.6bar before entering a packed
bed in the first step of the cycle. The adsorption and reaction are very
exothermic and recirculation and cooling of a portion of the exiting gas by
heat exchange water is necessary to avoid an excessive temperature. The
recirculation to feed ratio is 12/5. The stream exits the adsorber/reactor at
475K with only 10% of the CO₂ remaining, 90% having been
converted to M₂CO₃. The desorption is
very endothermic and the same of rate of recycling, in this case with heating
by exchange with flue gas is necessary to maintain a sufficient temperature.

    
Before sequestration the CO₂ must be in the supercritical
phase, which requires a pressure of 73.3bars at 311K. The cost of this step was
based on three-stage centrifugal compression to 95.2bar with intercooling and
over a distance of 167 km in a 10-in (0.254-m) pipe. The use of holding tanks in
the event of a temporary shutdown of the compressors was examined but concluded
to be impractical.

    
The estimated costs of separation and sequestration of CO₂using the MEA and MgO/Al₂O₃processes
for the indicated conditions (a 500MW combustor/generator in the Illinois Basin) are 

    
Based on the preceding analyses, it is concluded that separation using MgO
is superior to that using MEA in every respect except operational experience.
The fixed and variable costs are both much less and the plant equipment is more
compact. Even for this process, the overall costs for capture, separation and
sequestration are huge and no easy means of their reduction is apparent. Because
it produces no income,  this process of sequestration can be justified only for
environmental reasons or legal requirements.