United States
 Environmental Protection
 Agency
 Industrial Environmental Research
 Laboratory
 Research Triangle Park NC 27711
 Research and Development
 EPA-600/S7-81 -154  Dec. 1981
Project  Summary
Study  of  Automatic  Control
Systems to Maintain Constant
Percentage SCh  Retention in a
Pressurized  FBC
K. J. Daniel, S. D. Finnigan, and R. M. Reinstrom
  The Clean Air Act Amendments of
1977 indicate  that future emission
standards for SO2 should be based on
a percentage reduction (comparing
sulfur emissions with sulfur feed). In a
pressurized fluidized-bed (PFB) boiler,
sulfur feed (as determined by coal
sulfur content and feed  rate)  and
sulfur removal effectiveness (as deter-
mined by the reactivity and feed rate
of dolomite/limestone sorbent) vary
continually during  PFB plant opera-
tion. The purpose of this study was to
assess the feasibility of using some
type of automatic control system to
maintain a constant percentage sulfur
removal in a PFB system as variations
occurred in key variables, such as coal
sulfur content and sorbent reactivity.
  To conduct this feasibility study, a
transient model of a PFB power plant
was  developed and  validated  for
studying methods of controlling  bed
SO2  absorption characteristics. To
accomplish this, the transient equa-
tion for the population distribution as
a function of size and utilization was
solved. The model uses TGA rate data
for 1337 dolomite as a function of
size, utilization, and temperature. The
kinetic data is integrated over  the
instantaneous population distribution
to determine the instantaneous SO2
absorption rate constant.
  This transient model was used to
assess the potential of  alternative
automatic  control  strategies  for
achieving constant percentage SO2
retention during changes in load, in
coal sulfur  content, and in sorbent
reactivity in a  PFB. Goals  were  to
minimize costs and sorbent require-
ments.  Of  the control strategies
considered, the preferred option iden-
tified in this assessment continuously
monitored the  sulfur content in the
feed coal, and  adjusted the sorbent
feed rate to maintain  a constant
sorbent-to-coal-sulfur feed ratio. This
strategy  does  not  instantaneously
change this sorbent-to-sulfur ratio to
account for short-duration changes in
sorbent reactivity, since the inertia of
the  large mass of partially  spent
sorbent in the bed prevents instanta-
neous  changes  in the sorbent/sulfur
ratio from being effective. However,
long-term variations in reactivity are
accommodated through a feedback
loop which measures retention and
adjusts the sorbent-to-sulfur set point
accordingly. Such a control strategy
minimizes, but does not eliminate, the
need for excess sorbent feed to ensure
that the  PFB does not exceed the
specified percentage SO2 reduction
on a 30-day average.
  A prompt neutron activation tech-
nique appears to be the most promis-
ing  for  the  required  continuous
monitoring  of  coal sulfur  content.
However, such  a system is develop-

-------
 mental, and has not been commercial-
 ly demonstrated.
   This Project Summary was develop-
 ed by EPA's industrial Environmental
 Research  Laboratory.  Research  Tri-
 angle Park, NC. to announce key find-
 ings of the research project that is fully
 documented in a separate report of the
 same title (see Project Report ordering
 information at back).

 Introduction
  The Clean  Air Act  Amendments of
 1977 indicate that  the future emission
standards for S02 should be based on a
percentage reduction. In a conventional
power plant using flue gas desulfunza-
tion  (FGD), percentage  reduction  of
sulfur can,  in principle, easily be moni-
tored and  controlled.  Sulfur  dioxide
(SOi)  concentration  can readily  be
 measured before and after the  FGD unit
to determine the  percentage of  SOz
absorbed by the process. Because no
components  have  inherently  slow
 response, control of the system is rela-
tively straightforward.
  In  the  future,  power plants using a
fluidized  bed will encounter a  different
 situation.  Rather  than  monitor only
 gas concentrations to  determine per-
 centage reduction, coal sulfur content
 and feed rate will need to be monitored.
 Moreover,  the mass of absorbent in a
 bed represents a large inertia that must
 be controlled. Changes in the reactivity
 of the sorbent feed also will affect  the
 control system.
  This system examined the control of
 sulfur capture  in  fluidized beds with
 emphasis on pressurized fluidized beds
 (PFBs). Specifically, the study examined
 how  changes in load and coal sulfur
 content affect sulfur capture  in a PFB
 combustor. In addition, various methods
 of controlling  bed sulfur  absorption
 properties that minimize sorbent usage
 and  maintain a constant percent sulfur
 removal  were examined.
  The study  involved  developing a
 transient model for the PFB power plant,
 validating  the  model  by comparison
 with  experimental data, and using the
 model  to  evaluate  various control
 strategies.  Emphasis was on  phenom-
 enological   modeling  of the  sulfur
 capture processes  in the PFB. However,
 the model also includes representations
 of all major  equipment (e.g., gas  and
 steam turbines) in the PFB/combined-
 cycle power plant.
   The transient rate of sulfur capture is
 determined from the transient distribu-
tion of dolomite particles in the PFB, as a
function of size and  utilization. The
model solves for this transient distribu-
tion by integrating the distribution with
chemical rate constants (also a function
of size  and utilization); the instanta-
neous  rate of  sulfur  adsorption  is
obtained.  The   model was validated
using both transient  and steady-state
data from the experimental PFB Mini-
plant (Reference 1); the comparison of
the model and the data indicates good
agreement.


Results
  Several   interesting  results   were
obtained from the model. The response
of sulfur retention to a change in a Ca/S
ratio shows a strong dependence on the
magnitude  and  direction  of change.
However, the  response is essentially
independent of whether the change is
made in the calcium  feed rate  or the
sulfur feed rate.
  Also,  in spite of rapid changes  in
sulfur input  to the bed, the change in
retention was  very  slow.   Retention
changes slowly because the rate of the
sulfur capture reaction in the bed isfirst
order with respect to SOa concentration
and  also  because the inventory  of
dolomite in the bed represents a large
inertia.
  Four candidate strategies to control
bed  retention   were  identified  and
evaluated. The first was to change the
size distribution of the absorbent feed to
the bed. Because smaller particles have
higher reactivity, the amount of sulfur
absorbed by the bed can be controlled by
changing the  size  distribution  of the
bed. However,   because  of  the large
amount of  mass in the bed, the size
distribution  of the  bed changes very
slowly.  Consequently,  the  response
with this method of control is too slow to
be  practical. Moreover, reducing  the
size distribution  of the feed could cause
additional elutriation which would limit
the maximum excursion that could  be
handled by this method of control. This
type of control would also not be able to
take advantage  of downward  excur-
sions  in  coal  sulfur  content.  This
method of control, thus, had several
disadvantages and no clear advantage.
  The   second  strategy  for  control
studied was to adjust the reactivity of
the  absorbent  fed to the  bed.  This
method  would   also  have  a   slow
response due to the large bed inventory.
In addition, if a highly reactive dolomite
were being used to achieve low steady-
state Ca/S ratio, there would be little
control  margin  to  accommodate
transient  excursions.  Moreover, this
system  could not take advantage  of
downward  excursions of coal sulfur
content.  Because of these shortcom-
ings, this method is not recommended.
  The third strategy investigated con-
sidered blending lowsulfurfuel with the
coal feed to maintain a constant rate of
sulfur feed to the bed. This system will
ensure nearly constant percent sulfur
retention if the reactivity of the dolomite
is constant.  In  addition,  this system
minimizes the  excess dolomite  feed
required. There  are  no technical bar-
riers  to  implementing  this  control
system;  however,  it is  prohibitively
expensive when compared to a system
that uses a sufficient excess of dolomite
feed to ensure that emission limits are
met.
  The fourth strategy investigated is the
preferred method.  This  method
maintained the  dolomite  feed rate  in
proportion to  the rate of sulfur entering
the bed. By exercising the model it was
found that 862  retention  can be con-
trolled within 1-1/2  percentage points
by this method. Control of the dolomite-
to-sulfur ratio in this manner would not,
by  itself, make the  necessary adjust-
ments if sorbent reactivity were to vary;
accordingly, a feedback loop is included
to  measure  retention and adjust the
dolomite-to-sulfur set point as  neces-
sary to  accommodate  long-term
variations  in  reactivity.   Efforts  to
instantaneously adjust the dolomite-to-
sulfur   ratio, to  follow  short-term
changes in sorbent reactivity, would not
be  effective,  since  the inertia  of the
large  inventory of  partially   spent
sorbent  in   the  bed would  prevent
instantaneous changes  in  the
dolomite/sulfur  ratio from having  a
significant impact.
  This strategy reduces excess dolomite
use to a minimum and shows a signifi-
cant potential cost savings  compared to
a system using a constant excess dolo-
mite feed to  ensure compliance with
emission limits. One possible disadvan-
tage of  this  system  is that elutriated
fines may increase  due to  increased
dolomite feed. While available data does
not indicate that this will be a significant
disadvantage, further study  is recom-
mended.
  Because   the  preferred  strategy
requires  measurement  of the  sulfur
content  of the coal  feed to the bed, a
review was conducted of  on-line coal

-------
   sulfur  measurement techniques.  This
   review indicated that,  although many
   techniques are available, the one that
   appears   most  suitable  is  prompt
   neutron activation  analysis of the coal
   feed.  This measurement technique is
   still being developed. The added cost of
   this device was included in the econom-
   ic analysis.
     The methods identified for maintain-
   ing constant percentage sulfur capture
   are applicable for use with atmospheric
   fluidized-bed (AFB) combustors without
   fines recycle, and may  be applicable to
   AFB  systems that incorporate  fines
   recycle. However, recycle of fines is not
   incorporated  in  the  model  and,
   therefore, no conclusion can be drawn
   regarding the effects of fines recycle on
   the results presented in this study.

   Recommendations
     To determine the  economic attrac-
   tiveness   of  the   proposed   control
   method,  the  typical  time-dependent
   sulfur variations in coal feed  must be
   defined. Specifically, the size of excur-
   sion, the  duration  of  excursion,  the
   ramp  rate of change, or frequency
   spectrum must be determined. This will
   aid  in  determining the  margin   in
   dolomite  feed rate that is required  to
   control variations in sulfur emissions.
   More accurate economic trade-offs can
   then be obtained.
     The recommended control  strategy
   requires  that  the  dolomite-to-sulfur
   ratio remain constant. This implies that'
   the dolomite feed be increased and de-
   creased in response to changes in coal
   sulfur content. To reduce this method to
   practice,  experiments should be per-
   formed  to determine  the effect  of
   changing  feed  rate  on attrition  and
   elutriation.
     Moreover,  the   model  should   be
   extended to study advanced AFB con-
   cepts that rely on the recycle of elutri-
   ated fines to  the  bed  to significantly
   improve sulfur capture. The extension
   of the model could aid in the evaluation
   of various effects and options such as
   attrition  rates, recycle  rates,  cyclone
   capture  efficiency,  freeboard  height,
   coal sulfur contents, and load control.

   References
    1.   Hoke, R. C., et al.,  "Mmiplant and
        Bench  Studies  of  Pressurized
        Fluidized-Bed  Coal  Combustion:
        Final  Report,"  Exxon Research
        and Engineering Co. (EPA-600/7-
        80-013, NTIS No. PB 80-188121),
        January 1980.
                                         K. J. Daniel. S. D. Finnigan, and R. M. Reinstrom are with General Electric Co.,
                                           1 River Road.- Schenectady, NY 12345.
                                         D. Bruce Henschel is the EPA Project Officer (see below).
                                         The complete report, entitled "Study of Automatic Control Systems to Maintain
                                           Constant Percentage  SO 2 Retention in a Pressurized FBC," (Order No.
                                           PB 82-110 693; Cost: $15.00, subject to change) will be available only from:

                                                 National Technical Information Service
                                                 5285 Port Royal Road
                                                 Springfield. VA 22161
                                                 Telephone: 703-487-4650
                                         The EPA Project Officer can be contacted at:
                                                 Industrial Environmental Research Laboratory
                                                 U.S. Environmental Protection Agency
                                                 Research Triangle Park, NC 27711
*
U S GOVERNMENT PRINTING OFFICE, 1981 — 559-017/7405

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OM 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
             00003^9

-------