FINAL TECHNICAL REPORT
November 1, 1999, through October 31,
2000
Project Title:
DEVELOPMENT OF NEW AND IMPROVED SORBENTS FOR THE COBRA
PROCESS
ICCI Project Number: 99-1/1.2A-2
Principal Investigator: Rachid B. Slimane, Gas Technology Institute
Other
Investigators: Javad
Abbasian and Perin A. Cengiz, Illinois Institute of
Technology
Project Manager: Ken K. Ho, ICCI
The overall objective of this program was to continue
further development of new and improved novel copper-based sorbents for removal
of SO2 and NOx from flue gas, in support of the Copper
Oxide Bed Regenerable Adsorber (COBRA)
process. The targeted areas of sorbent improvement included higher
reactivity and effective sulfur capacity, higher crush strength, and higher
catalytic activity for NOx reduction, which can lead to improvement
in process control and economic utilization of the
sorbent.
To achieve this objective, a total of twenty one
(21) alumina support materials and forty one (41) new sorbents were formulated
using various preparation techniques. The crush strengths of these sorbents
were determined. Thirteen (13) new sorbents were evaluated for their
SO2 sorption capacities in apacked bed
reactor. The regenerability
of seven (7) sorbents that exhibited similar or higher sorption capacities
than the baseline sorbent, were determined over three (3) sulfation/regeneration
cycles. Based on the results of these tests, the sorbent designated as S43-175
was selected as the “best” formulation for durability
studies. A test series consisting
of 20 sulfation/regeneration cycles was conducted with this sorbent in the
packed-bed reactor. The catalytic activities of the baseline sorbent as well
as several new sorbents for NOx removal were also determined.
Four (4) sorbent formulations exhibited improvement in sulfur capacity
and in crush strength compared to the baseline
sorbent. The “best”
sorbent formulation (S43-175) has 3.5 times sulfur capacity and 1.5 times
crush strength compared to the baseline (Alcoa)
sorbent. The reactivity of the S43-175 sorbent after 20 cycles
is about twice that of the baseline sorbent. The crush strength of the S43-175
sorbent does not appear to be adversely affected by the long-term durability
test.
The new sorbents have higher catalytic activities than that of the baseline sorbent. The extent of NOx removal with the new sorbents in the sulfated form exceeds 99% compared to 88.5% for the baseline sorbent. The extent of NOx removal does not appear to be affected by the long-term durability test.
EXECUTIVE SUMMARY
During coal combustion, sulfur in coal is released in the form of sulfur dioxide (SO2) in the flue gas and a small fraction of nitrogen in the form of NO2 and NO, commonly referred to as NOx. The SO2 and NOx emissions are very damaging to the environment because they combine with moisture to form acids which then fall as acid rain. To protect the environment, legislation was introduced requiring electric utilities to adopt available technology for removal of pollutant gases and particulates from coal combustion flue gases so that the increased use of coal is done in an environmentally acceptable manner.
The threat from acid rain is a greater concern in Illinois where over 90% of the high sulfur coal mined is consumed by electric utilities that are based on pulverized coal combustion, while only a very small fraction of the coal-based power plants in Illinois is currently equipped with Flue Gas Desulfurization (FGD) processes.
The development of the Copper Oxide Bed Regenerable Absorber (COBRA) process, which is based on moving-bed crossflow reactor design for the combined removal of SO2, NOx, and particulates, has been pursued in conjunction with the use of Illinois coal. This process has been demonstrated at the nominal half-megawatt scale at the Illinois Coal Development Park in Carterville, Illinois.
The overall levelized cost of the COBRA process is very sensitive to the sorbent price. Any improvement in sorbent performance would reduce capital costs for process equipment as well as for the sorbent itself. Therefore, given the tremendous effect of the sorbent related costs on the overall process cost, it is necessary to conduct a carefully designed systematic study to improve sorbent performance to significantly lower the overall cost of the COBRA process.
Development of improved sorbents for the COBRA process has been pursued in an earlier DCCA/OCDM/ICCI funded project (ICCI project No. 98-1/1.1C-2). In this project, the baseline sorbent (produced by Alcoa) was evaluated in packed-bed experiments. Parametric studies were carried out to determine the effects of operating parameters on the performance of the sorbent. Long term durability of the baseline sorbent was assessed. Physical and chemical properties of the baseline sorbent were also determined. A number of new sorbent formulations were prepared using a modified sol-gel technique. These sol-gel sorbents have significantly higher crush strength and similar surface area compared to the baseline sorbent. However, the reactivities of these sorbents, in the pellet form, were lower than that of the baseline sorbent. To improve the performance of the new sorbents, the chemical composition and the preparation technique were modified. The results indicate that, although the reactivity of the sorbents improved by these modifications, the crush strength of the sorbents significantly decreased. Therefore, additional work was needed to optimize the sorbent composition and preparation technique to further improve the performance of the sorbent.
The overall objective of this program was to continue
further development of new and improved novel copper-based sorbents for removal
of SO2 and NOx from flue gas, in support of the Copper
Oxide Bed Regenerable Adsorber (COBRA)
process. The targeted areas of sorbent improvement included higher
reactivity and effective sulfur capacity, higher attrition resistance, which
can lead to improvement in process control and economic utilization of the
sorbent.
To achieve this objective, a total of twenty one (21) alumina support materials
were formulated using various preparation
techniques. To improve the porosity and
pore size distributions of the sorbents/alumina supports, a number of boehmite
sols were produced using ammonium hydroxide as a hydrolysis catalyst. Various
acidic pH levels were tested during gelation of the sols when producing alumina
pellets to investigate the effect of pH on the macroporosity of the
material. To lower sorbent cost,
the aluminum tri-secondary butoxide (i.e., ALTSB,
Al(OC4H9)3) that was used as the raw material
precursor for production of alumina sol was replaced by the less expensive
aluminum isopropoxide (i.e., ALISOP,
Al(OC3H7)3).
The sorbent preparation technique was simplified by reducing the number
of steps required for producing the sol, resulting in the reduction of the
sol production time from 15 hours to 1-4 hours.
The results indicate that, the alumina pellets produced using the simplified technique with shorter preparation time (i.e., 1-4 hr) have higher crush strengths compared to the alumina from a 15-hr sol. The analysis of these materials indicates that the alumina produced by the simplified technique have comparable physical characteristics to those produced earlier.
A total of forty one (41) new sorbents were formulated
using the lower cost materials and/or simplified preparation
techniques. The crush strengths
of these sorbents were determined for comparison with the baseline sorbent
produced by Alcoa.
Thirteen (13) new sorbents were evaluated for their
SO2 sorption capacities in the packed bed reactor. The regenerability
of seven (7) sorbents that exhibited similar or higher sorption capacities
than the baseline sorbent were determined over three (3) cycles. Based on
the results of these tests, the sorbent designated as S43-175 was selected
as the “best” formulation for durability
studies. A test series consisting
of 20 sulfation/regeneration cycles was conducted with this sorbent in the
packed-bed reactor. The catalytic activities of the baseline sorbent as well
as several new sorbents for NOx removal were also determined.
Among the sorbents developed in this project, four
(4) sorbent formulations exhibited more than 23% improvement in sulfur capacity
and more than 45% improvement in crush strength compared to the baseline
sorbent. The best result was
obtained with the sorbent designated as S43-175, which has 3.5 times sulfur
capacity and 1.5 times crush strength compared to the Alcoa
sorbent. The sorbent designated as 167-WI, which is produced by
wet impregnation of a sol-gel alumina, has 25% higher sulfur capacity and
7 times higher crush strength than the baseline
sorbent. This formulation was selected as a “second best”
sorbent.
A “life-cycle” test consisting of 20
sulfation/regeneration cycles was conducted with the sorbent designated as
S43-175 in the packed-bed
reactor. The results of
this series of tests indicate that the effective sulfur capacity of the S43-175
sorbent after 20 cycles is still about twice that of the baseline sorbent,
while the rates of decrease in the sulfur capacities for the two sorbents
are similar. These results indicate
that the rate of “fresh sorbent make-up” needed to continuously
maintain the desired level of desulfurization with S43-175 is significantly
lower than that of the baseline sorbent.
The results also indicate that the crush strength of the S43-175 sorbent
is not adversely affected by the long-term durability
test.
The catalytic activity of the baseline sorbent
for removal of NOx from flue gas was determined in a series of
tests conducted at different operating
conditions. The parameters studied
included the effect of bed materials as well as gas composition on the extent
of NOx removal from the gas
stream. The baseline tests for
this series were conducted at
400°C
and 2000 hr-1 space velocity using a gas mixture containing
500 ppmv of NOx and 500 ppmv of
NH3. The results
of these tests indicate that the baseline sorbent is capable of removing
69.0% of the NOx in the regenerated state, while in the sulfated
state the NOx removal will increase to
88.5%. The extent of
NOx removal in the absence of NH3 is essentially zero
confirming that NOx removal is accomplished through reduction
by ammonia. The regenerated sorbent in the reduced form exhibits slightly
lower catalytic activity than the oxidized sorbent (64% compared to 69%).
The results also indicate that the extent of the extent of NOx
removal is linearly related to the NH3/NOx ratio in
the flue gas.
Comparison of the catalytic activities of the baseline and new sorbents developed in this program indicates that the new sorbents have higher catalytic activities than that of the baseline sorbent. The extent of NOx removal with the new sorbents in the sulfated form exceeds 99% compared to 88.5% for the baseline sorbent. The catalytic activity of the S43-175 after 20 cycles was also determined. The extent of NOx removal was 99.3% indicating that the long-term durability test did not affect the catalytic activity of this sorbent.