United States
                      Environmental Protection
                      Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park NC 27711
                     Research and Development
EPA/600/S7-86/029 Jan. 1 987
&EPA          Project  Summary

                     Analysis  of  Utility  Control
                     Strategies   Using  the
                     LIMB   Technology
                     T. E. Emmel and B. A. Laseke
                       The report gives results of a study to
                      evaluate the impact of proposed acid rain
                      legislation on the potential application of
                      limestone injection  multistage  burner
                      (LIMB) technology to achieve sulfur diox-
                      ide (S02) and nitrogen oxide (NOX) reduc-
                      tions at coal-fired utility power plants.
                       The study found that proposed acid rain
                      legislation, which mandates the retrofit of
                      high efficiency control technologies such
                      as flue gas desurfurization (FGD) or which
                      requires national  S02/NOX reduction
                      levels greater than 10 million tons per year,
                      would significantly reduce the application
                      of LIMB. For regulatory strategies which
                      do not mandate the use of FGD and which
                      require emission reductions of 8  to 10
                      million tons per year, the potential LIMB
                      application  ranges from  15,000 to
                      100,000 MW of coal-fired boiler capacity
                      in the 31 eastern state acid rain region.
                       This Project Summary was developed
                      by  EPA's Air  and  Energy  Engineering
                      Research Laboratory, Research Triangle
                      Park, NC, to announce key findings of the
                      research project that is fully documented
                      in a separate report of the same title (see
                      Project  Report ordering information at
                      back).

                      Introduction
                       A number of bills have been proposed
                      by Congress that would require reductions
                      of acid  rain precursor emissions. These
                      congressional bills would require different
                      mixes of emission control technologies to
                      achieve  S02 and NOX reductions at coal-
                      fired utility power plants. The objective of
                      this research program was to evaluate the
                      impact of proposed acid rain legislation on
                      the potential application of LIMB tech-
                      nology incorporating recent LIMB research
                      and development findings.
  A number of regulatory strategies and
emission reduction targets were developed
by reviewing acid rain legislation proposed
in the 97th and 98th congressional ses-
sions. For each regulatory strategy devel-
oped, the control technology mix of LIMB,
FGD,  and coal  switching required to
achieve the selected emission reduction
level was determined. Next, the maximum
number  of boilers to which LIMB tech-
nology could be applied was determined
by examining technical and regulatory
constraints and  emission  reduction tar-
gets. The cost effectiveness of each regu-
latory case and  control technology mix
was estimated to evaluate the cost of each
control technology mix.

Regulatory Case Development
  The primary differences in the congres-
sional bills are a result of the level of SO2
reductions that is required at each plant
due to plant/boiler specific emission limits
or due to requiring high overall SO2 reduc-
tion levels. All of the bills use 1980 as the
base year for which emission reduction
levels apply. Differences in the method of
calculating excess emissions, implemen-
tation years, financing methods, and state
reduction allocation and implementation
were not considered important for the pur-
poses of this study. The following three
legislative/regulatory cases were analyzed:
Regulatory Case
S02 Reductions,
  million tons/yr
Boiler Performance       10
  Standard
Regional Reduction Levels 10
Regional Reduction Levels   8

-------
  For this study, a regulatory strategy was
developed based on bills which base re-
ductions on boiler and/or state reduction
performance standards (S 1709, HR 4816,
HR 3400). These bills require that existing
boilers must comply with New Source Per-
formance Standards (1971 or 1979) if their
emissions are greater  than a specified
amount of S02  per million Btu of fuel.
These bills also require state wide reduc-
tions. Because these bills require very high
levels of S02  reduction  at  individual
plants/boilers, the use of wet FGD will be
required  at most affected  plants. These
legislative cases are entitled "Boiler Per-
formance" cases, and FGD is applied to
boilers at the largest emitting utility power
plants.

  The other major type of bill  introduced
in Congress (HR 4829,  S 3041) allocates
state level emission reductions generally
based on the portion of emissions from
facilities with emission rates greater than
1.2 Ib S02 per  million Btu fuel input.
These bills allow the states to determine
how the allocated emission reductions for
that state are to be achieved and in some
cases allow trading of emission reduc-
tions. Because these bills  provide much
greater flexibility in how emission reduc-
tions are achieved on a plant/boiler basis,
they do not require the use of certain types
of SO2/NOX control  technologies. Study
cases based  on this type of legislative
scenario are  entitled  "Regional Reduc-
tion" cases.

  The other major difference between bills
that would impact the mix of control tech-
nologies used by utilities is the amount of
emission reduction required because, as
the SO2  reduction target increases, the
average emission reduction needed to be
achieved at  each  coal-fired  boiler  in-
creases.  For this study three  SO2 emis-
sion reduction levels were evaluated: 8 and
10 million tons per year, consistent with
the different levels proposed by the con-
gressional bills reviewed; and 12 million
tons per year,  a sensitivity case to evaluate
the impact that this level  of reduction
would have on the control technology mix
needed  to achieve this high  level of
reduction.
  The  Congressional  bills  differ in the
amount  of credit given for NOX reduc-
tions. For this study half credit was given
for NOX  reductions; e.g., 1.0 ton of NOX
removed equals 0.5 ton of SO2 reduction.
Thus, for this study, a NOX  credit  was
included for  low  NOX   combustion
modification  assumed  to be made with
furnace sorbent  injection.
Region and Boiler Specific
Data Base
  A major part of the study was develop-
ment of a boiler specific data base and
boiler specific control costs for LIMB, FGD,
and coal switching. Developing  an ac-
curate data base for all coal-fired boilers
in the 31 eastern states was not feasible.
However, an accurate data base was easily
developed for the top  100 SO2 emitting
coal-fired utility power plants. These top
100 plants accounted for over 72% of
total U.S. utility power plant S02 emis-
sions in 1980. Results of the applicability
study for the top 100 plants were then ex-
trapolated to the boilers in the 31 eastern
state  region. S02  emission reduction
targets used for each regulatory case, bas-
ed on  allocating  72%  of the emission
reduction target to the top 100 coal-fired
boiler population, are
            tion cases. Figures 3 and 4 summarize the
            results of the 8,10,  and 12 million ton per
            year SO2 reduction cases.
            10 Million Ton Per Year SO3
           Reduction Cases
             Figure 1 summarizes the results of the
            10  million  ton per year S02  reduction
           cases. Two cases were run for the Boiler
           Performance Standard strategy to provide
           an upper and lower bound on the amount
           of LIMB which would be used to achieve
           the  desired  S02 reductions. In both
           cases, FGD was applied to the boilers in
           the top  50 SO2 emitting power  plants
           with post 1965 service year achieving over
           5.5 million tons per year  of SO2  reduc-
           tion. In the first case LIMB  was applied to
           the remaining  boilers which were consid-
           ered technically  applicable  (post  1960
Regulatory Strategy
S02 Emission Reduction
 From Top 100 Plants,
   106 tons per year
 Total Required
 S02 Reduction,
106 tons per year
Boiler Performance Standard
Regional Reduction
Regional Reduction
Regional Reduction
7.2
7.2
5.8
8.6
10
10
8
12
Control Technology
Performance/Cost
  Three coal-fired boiler S02 reduction
technologies were examined: (1) limestone
FGD with 90% SO2 control; (2) LIMB
with 50-60% SO2 control and 50% NOX
control; and (3) switching to 2.5 Ib S02
per million Btu eastern bituminous coal.
  Boiler specific costs for FGD and LIMB
were provided,  using the IAPCS-2 com-
puter model. Table 1 summarizes the cost/
performance assumptions used to make
the computer runs.
  The cost of coal switching was based
on  a coal cost differential of $1.00 per
million Btu above the current higher sulfur
coal. Although boiler specific costs for
high and low sulfur coals were available,
due to the current soft market,  several
plants are actually obtaining  low sulfur
coal at prices below high sulfur coal. This
is not anticipated if many plants were
required  to  switch  coals because  the
added demand for low sulfur coal would
drive up its  price relative to high sulfur
coals.

Discussion of Results
  Figures 1 and 2 summarize the results
of the 10 million ton per year S02 reduc-
           wall/tangential  fired boilers with sulfur
           emissions between 1.2 and 6.0 Ib/million
           Btu). This case results in 69,000 MW of
           FGD application, 13,000 MW of LIMB ap-
           plication and 3,000 MW of coal switching.
           For the second Boiler Performance Stan-
           dard case, coal switching (MAX CS) was
           applied before  LIMB  resulting in 8,400
           MW of  coal  switching.  Because  coal
           switching can be achieved on the 1950's
           boiler to meet the required emission reduc-
           tion target, no  LIMB was applied.
             Three different cases were  run for the
           10 million ton per year regional allocation
           scenario. The first two cases provided an
           upper and lower bound on the amount of
           LIMB which would be used versus  coal
           switching. The other case looks at the im-
           pact of high performance (HP) LIMB (60%
           S02 reduction). For the maximum (MAX)
           LIMB case, LIMB was applied first to the
           applicable boilers resulting in half of the
           boiler population (71,000 MW) being con-
           trolled with the LIMB technology, 15,000
           MW of FGD,  and  11,000  MW of  coal
           switching. For the second 10 million ton
           reduction case,  coal switching was maxi-
           mized (MAX CS) by applying it first to all
           the 1950's boilers. This reduces LIMB ap-
           plication to  65,000 MW and increases

-------
Table 1.   Performance and Cost Parameters Used to Estimate FGD and LIMB
          Annual/zed Costs and Emission Reductions
                           LIMB Performance Parameters
5O% LIMB Cases
                                                                 60% LIMB Cases
50% SO2 Reduction
50% NOX Reduction
Calcitic Hydrate
2.5:1 Ca/S Ratio
7OO°F Quench Rate
             60% SO2 Reduction
             50% NOX Reduction
             Calcitic Hydrate
             3:1 Ca/S Ratio
             700°F Quench Rate
ESP upgrade and SO3 conditioning for control of additional paniculate matter.

                            FGD Performance Parameters
                     90% SO2 Reduction and No NOX Reduction
                     Limestone Slurry Sorbent
                     No Spare Absorbers
                     Number of Absorber Towers Based on Boiler Size:
                        Boiler Size, MW            No. of Towers
<100
100-250
250-500
500-750
>750
1
2
3
4
5
                                General Cost Bases
                          EPRI Cost Premises Used
                          Costs are in 1995 Dollars
                          Equipment Book Life of 15 Years
                          FGD Retrofit Difficulty Factor: 1.2 Times New Plant Cost
 150,000
 1OO.OOO
I
fla
  50.000
                    .•.v.v.v.v Coal Switching Capacity
                    xxxxxxx: FGD Capacity
                    SSB^B LIMB Capacity
                    """" Retired FGD Capacity
                Man LIMB     Max CS    Max LIMB    Max CS
                HP LIMB
                  Boiler Performance
                    Standard Cases
Regional Reduction
   Level Cases
coal switching to 25,000 MW of applica-
tion. For the third 10 million ton per year
reduction  case, high performance (HP
LIMB) LIMB was applied, followed by FGD
and coal switching as in the MAX LIMB
case. This case decreases the penetration
of FDG  due to the greater S02 reduction
achieved by high performance (60%) LIMB
technology.
  Figure 2 summarizes the cost results in
the five 10 million ton per year S02 re-
duction cases. The  boiler performance
standard cases have the highest annual
control cost of $13-$14 billion per year due
to the large number of boilers which must
apply FGD. The regional annual costs of
the regional  reduction  level  cases  are
significantly lower and range from $9.9 to
$11.7 billion per year.
8,  10, and 12 Million Ton Per Year
Cases
  Figure 3 presents the results analyzing
the impact of various emission reduction
scenarios on the application of LIMB. The
10  million ton per year S02 reduction
case is the  same  as for the Max LIMB
regional allocation  case discussed above.
For this case, 71,000 MW of LIMB was ap-
plied to achieve the emission reduction
target.
  For the 8 million  ton per year reduction
case, coal switching to the 1950's boilers
was applied first (lowest unit cost), follow-
ed by LIMB and FGD to achieve the emis-
sion reduction target. This results in boiler
application  of  71,000  MW  of LIMB,
25,000 MW of coal switching, and 3,200
MW of FGD,
  For the 12 million ton per year emission
reduction  case, the application of LIMB
cannot be  maximized  if the emission
reduction target is to be achieved. For this
case, LIMB application was reduced by in-
creasing the use of FGD and allowing all
boilers where FGD  and LIMB were not ap-
plied to switch coal. This results in the
following boiler applications: 38,000 MW
of LIMB, 50,000 MW of FGD, and 25,000
MW of coal  switching.
  Figure 4 presents the annual cost for the
three cases.  The annual costs and unit
costs increase significantly as the emis-
sion reduction  levels increase over 10
million  tons  per year:
Figure  1.    Boiler application results for 10 million ton per year ofSOa reductions.

-------
Annual Reduction,
106 tons per year
8
10
12
Increase Emission
Reduction, %
25
50
Increase
in Cost, %
23
73
Average Unit
Cost, $/ton
1381
1397
1678
These cost increases are due to the signif-
icantly increased application of FGD need-
ed to obtain the very high overall average
emission reductions per boiler/plant.

31 Eastern State Region
  To estimate the potential LIMB applica-
bility for all of the coal-fired boilers in the
31 eastern state region, the number of
boilers in that region that fit the LIMB and
FGD technical applicability was determin-
ed from the 31 eastern state utility boiler
data base. The  amount of capacity for
which LIMB was applicable was 103,000
MW. The amount of FGD capacity for this
boiler population was 108,000 MW.
  The average unit cost of applying FGD
to the applicable boilers not in the top 100
plants is significantly greater due to the
smaller boiler sizes and lower coal sulfur
contents. This  means that LIMB tech-
nology would be favored over FGD, and
the LIMB applicability potential for the 10
million ton per year S02 reduction strat-
egy not mandating the use of FGD could
be as high as  100,000  MW  of  boiler
capacity.

Conclusions
  This study indicates that up to 100,000
MW  of boiler capacity of LIMB application
is possible depending on the type of acid
rain legislation adopted and the amount of
coal  switching that is economically and
politically practical. Currently proposed
legislative strategies requiring SO2 reduc-
tions of 8-10 million tons per  year  will
maximize the application of LIMB because
it is anticipated to be more cost effective
than  FGD. Control strategies  requiring
SO2  reductions  greater than 10  million
tons per year will decrease the application
of LIMB,  because the average level of
S02  control required at each boiler would
exceed that available with a broad applica-
tion of LIMB. Legislative strategies which
would  require   high  levels  of control
(>60%) at each boiler would also  reduce
the application of LIMB unless combined
with fuel substitution.
 7 5,000
  10,000-
,3 5,000
                                                            Coal Switching Cost
                                                     VYxYxW FGD Cost
                                                            LIMB Cost
Figure 2.


 150.000
      Boiler Performance
       Standard Cases
Level/zed annual cost of control (1995 $).
                                               Regional Reduction
                                                  Level Cases
100,000-
 o

I
 is
"550,000
CO
                               Coal Switching Capacity
                               FGD Capacity
                               LIMB Capacity
                               Retired FGD Capacity
                                                            8x10* tons/yr      10x10* tons/yr      12x10* tons/yr


                                          Figure 3.    Boiler application results for 8, 10, and 12 million ton per year cases.

-------
 J 5,000
  10,000-
<0

B>
  5,000-
                         •.•.v.v.v.v Coal Switching Cost
                         WMW FGO Cost
                         S^BB LIMB Cost
Figure 4.
                   8 x JOS tons/yr      10 x 10* tons/yr     12 x 10s tons/yr
Levelled annual cost of control for 8, 10, and 12 million ton per year cases
(1995 $).
    T. Emmel is with Radian Corp.. Research Triangle Park, NC 27709; andB. Laseke
     is with PEI Associates, Inc., Cincinnati, OH 45246.
   Norman Kaplan is the EPA Project Officer (see below).
    The complete report,  entitled "Analysis of Utility Control Strategies Using the
     LIMB Technology," (Order No.  PB 87-WO 574/AS; Cost: $9.95, 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:
           Air and Energy Engineering Research Laboratory
           U.S. Environmental Protection Agency
           Research Triangle Park, NC 27711

-------
  • Indirect
      heated    —Unisulf on
                  Unishale B gases
  • Integral
      combustor —DBA + Stretford on
                  Lurgi gases
                —Unisulf on
                  Unishale C gases

  Major equipment costs were taken
from  EPA Pollution Control Technical
Manuals (PCTMs). ASSP equipment
was sized and costs factored from in-
house data and PCTMs. Costs were fac-
tored to first quarter 1985.
  Results of the cost study showed
changes in incremental capital and op-
erating costs for ASSP relative  to con-
ventional  processing (see Table 1).
  These cost comparisons show that
the best potential for application of
ASSP are processes that already have a
spent shale combustor integrated  into
the retorting  process (e.g., Lurgi, Uni-
shale C, Chevron STB, and Tosco HSP).
Capital and operating cost savings for
Unishale C and Lurgi are primarily a re-
sult of deleting the Unisulf and Stretford
plants.
  Economics for the indirect and direct
heated retorts are good to marginal.
Factors which will affect the economics
are:
  • How effectively combustor heat
    can be utilized  (simple  steam
    raising is the least desirable).
  • The value of steam.
  • The use of fast or circulating fluid
    beds to reduce investment in com-
    bustor equipment.

Phase II Pilot Plant Testing
  Pilot plant tests were performed  in a
bubbling fluid bed combustor of the
type which is integrated into the retort
process. A total of 44 individual tests
were  performed.  Variables evaluated
were  combustor  temperature, solids
residence time, gas residence time, oxy-
gen concentration, inlet gas sulfur con-
centration, staged combustion, and raw
shale injection. Over the entire range of
conditions tested, emissions of primary
pollutants were:

Table  1.    Cost Comparison For ASSP
             Component
               Range
             S02
             NOX
             CO
             Trace Hydrocarbon
               1-38 ppmv
               80-670 ppmv
               0.05-1.80 vol%
               51-8465 ppmv
             Key findings of the tests were:
               • S02  emissions were easily con-
                 trolled to low levels at virtually all
                 conditions tested, probably as a re-
                 sult of the high Ca/S ratios used.
               • NOX emissions were primarily sen-
                 sitive to oxygen concentration, as
                 were S02 emissions to a lesser ex-
                 tent  (Figure 2). Reasonably good
                 NOX control could be obtained with
                 flue  gas oxygen  concentrations
                 below about 3 vol %. The lowest
                 NOX concentrations were seen at 02
                 levels approaching zero but at the
                 expense of higher CO and trace hy-
                 drocarbon emissions.
               • CO and trace  hydrocarbon emis-
                 sions were primarily sensitive to
                 flue gas oxygen concentration (Fig-
                 ure 3). Good control of both could
                 be obtained  at O2 levels above
                 about 2 vol %.
               Emissions of NOX move in a direction
             opposite to S02, CO, and trace hydro-
             carbon emissions. Thus, operating con-
             ditions that minimize all four represent
             a compromise. One test was run which
             produced nearly optimum results.

               Conditions for this test were:
             Bed Temperature       664°C
             Solids Residence
               Time                  9.4 min
             Gas Residence Time      0.9 sec
             Gas Supply
               Velocity             134.1  cm/sec
             Flue Gas 02              2.6 vol %
             Ca/S Mole Ratio         10.3
             Raw Shale/Spent
               Shale Ratio             1:36
               At these conditions the following re-
             sults were obtained:
             SO2                    11 ppmv
             NOX
             CO
             Trace Hydrocarbon
             Combustion Efficiency
                 160 ppmv
                   0.27 vol %
                 388 ppmv
                  89%
Retort Type
   Direct Heated
   Case A, Case B
  Indirect
  Heated
Integral Combustor
Retorting Process
ASSP Incremental
  Cap. Cost, $W6
ASSP Incremental
  Annual Oper. Cost, $W6/yr  +10.83
 MIS/Unishale C

-71.2     -63.2

         +12.07
 Unishale B    Lurgi    Unishale C

+90.2        -13.0    -32.1

-19.21        -2.29    -1.56
                     During selected tests, both combus-
                   tor flue gas and retort gas were sampled
                   and analyzed for  selected trace ele-
                   ments: mercury,  cadmium, arsenic,
                   lead, beryllium, and fluorine. During
                   these tests, solids streams were also an-
                   alyzed for trace elements in an attempt
                   to determine where trace elements go.
                   One run was performed where a spike
                   solution of mercury and cadmium was
                   added to the combustor.
                     Results of the trace element tests indi-
                   cated some relative trends with regard
                   to emissions but, because of the brevity
                   of the sampling, no hard  conclusions
                   can be reached which  would allow ex-
                   trapolation of results to long-term
                   steady-state operations. Some of the
                   key observations were:
                     • Lead,  beryllium and fluorine were
                       found to have low volatility; i.e., of
                       the amounts present in raw shale,
                       only very small percentages were
                       volatilized to the gas streams.
                     • Arsenic was found in significant
                       concentrations in  the retort gas
                       (100-400 ppmv),  although the
                       amount of arsenic found repre-
                       sented less than 15% of that in the
                       raw shale.
                     • So little mercury was present in the
                       raw shale that mercury emissions
                       could not be characterized with high
                       accuracy. Mercury emissions were
                       very low except during the spike in-
                       dicating that mercury, if present in
                       higher concentrations in the raw
                       shale, could possibly  pose emis-
                       sions problems.
                     •  Although significant amounts of
                       cadmium was found in the gases at
                       higher retort and combustor tem-
                       peratures, emissions represented
                       less than 10% of cadmium present
                       in raw shale.

                     There is some evidence that mercury
                   and cadmium introduced to  the  com-
                   bustor during the spike test condensed
                   within the retort equipment and
                   revolatilized over time. However, be-
                   cause of the limited number of samples
                   taken,  it would not be prudent to draw
                   any conclusions. Longer term steady-
                   state operations would have to be stud-
                   ied to determine the fate of mercury and
                   cadmium with more certainty.

-------
I
I
S
Q:
40

35.


30-


25 -


20-


15-


10-

 5 .


 0
                                                                          700
                                                                         -600
                                                                         -500
                                                                         -400
-300
                                                                         -200
                                                                         - 100
I
d
VI

-------
United States
Environmental Protection
Agency
Official Business
Penalty for Private Use $300

EPA/600/S7-86/029
Center for Environmental Research
Information
Cincinnati OH 45268

-------