-------
in addition to operating the towers at the highest cycles of con-
centration (10-15). Colstrip presently keeps the relative satu-
ration of CaSCU.2H20 below 1.10 by lime softening of the makeup
water.
Softening is a technique used to reduce the calcium
concentration in an aqueous stream. In lime softening, CaO(s)
is added to the liquid stream to increase the pH of the
solution and precipitate CaCO 3/s\ . From Equation 2.6 it can
be seen that a reduction in the calcium level reduces the rela-
tive saturation of CaSCU-2H20. Furthermore, softening reduces
the relative saturation of CaC03, and thus the acid requirements
to prevent CaCO3 scale. Lime-soda ash softening is similar but
non-carbonate hardness and silica removal are also accomplished.
Softening is the least expensive treatment that can
be used to control CaSOi»'2H20 scale. Other methods that could
be used, such as brine concentration and reverse osmosis, have
much higher operating and capital costs. If the gypsum scale
potential is not controlled with treatment, the cooling tower
blowdown may be excessive. This may require expensive tail-end
treatment of the ultimate effluent stream to achieve zero dis-
charge .
Softening is a more expensive treatment method when
compared to acid treatment because it requires more expensive
equipment and greater maintenance. For this reason, the size
of the softened stream can have a major impact on the cost of
treatment. A smaller stream to be treated requires lower cap-
ital investment for the flocculator and other softening equip-
ment. Higher concentrations of calcium in the stream to be
treated require smaller flow rates through the softener to re-
move the same amount of calcium. The stream with the highest
concentration of calcium in most cooling towers is the recir-
culating water. Often a small slipstream of the recirculating
water can be softened and remove as much or more calcium than
by pretreating a much larger makeup stream.
Simulations were performed to determine the amount of
calcium that must be removed to keep CaSOi^HaO subsaturated.
In order to account for deviations from equilibrium it was as-
sumed that a softener could effectively reduce the calcium con-
centration to 50 mg/Jt. The results of two simulations perfor-
med to compare slipstream treatment and pretreatment for the
Colstrip cooling system are presented in Table 2-3. In the case
of makeup softening, a treatment rate of 175 £/sec was required
-24-
-------
TABLE 2-3. EFFECTS OF SLIPSTREAM SOFTENING ON
TREATMENT RATE AT COLSTRIP
Cycles of Concentration
Makeup Water Rate, I /sec
(GPM)
Treatment Rate, a/sec
(GPM)
Calcium Removal Rate, kg CaCOs/day
(Ib/day)
Slowdown Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COT)
Sulfate (as SO^)
Nitrate (as NO 3)
pH
CaSO^HaO Relative Saturation*
Makeup
Softening
13.5
175.
(2770)
175.
(2770)
820.
(1800)
534.
143.
540.
227.
6.5
2640.
18.7
7.2
.93
Slipstream
Softening
20.0
170.
(2700)
9.0
(142)
1040.
(2290)
587.
212.
1450.
964.
25.5
3930.
33.6
7.6
1.02
*The critical value, above which scale potential exists is
1.3-1.4 for CaSO^-2H20.
-25-
-------
to achieve 13.5 cycles of concentration. With slipstream^soft-
ening, however, a treatment rate of only 9 £/sec was sufficient
to allow operation at 20 cycles of concentration. The calcium
removal rate increased from 820 kg CaC03/day to 1040 kg/day to
allow for the increase in cycles of concentration, but the treat-
ment rate decreased due to the large increase in calcium concen-
tration in the stream being softened.
An increase in required calcium removal necessitates
a larger slipstream. From Equation 2.6, increases in the con-
centration of either calcium or sulfate raise the relative sat-
uration of CaSO^-ZHzO and thus increase the calcium removal rate
required for scale-free operation. Figures 2-5 and 2-6 show^how
the required slipstream rate increases as a function of calcium
and sulfate concentration in the makeup water when the towers
are operated at 20 cycles of concentration (based on sampled
Col strip makeup water) .
It should be noted here that these curves apply only
to the makeup water composition considered. Variations in other
species' concentrations may cause significant changes due to
chemical complexing. This curve is valid for the compositions
considered and is presented to show trends in the system.
2.4 Recycle/Reuse Alternatives in Cooling Towers
In any cooling system employing cooling towers the
cooling water is recirculated. The degree of recirculation is
measured by the cycles of concentration. Increasing the cycles
of concentration in a cooling tower will reduce the makeup and
the blowdown rates but will increase the potential for scale
formation. Treatment methods can be employed which will reduce
the scale potential of cooling towers which operate at high cy-
cles of concentration.
Two of the plants studied, Bowen and Montour, presently
operate their towers at less than three cycles of concentration.
Operating in this mode the towers do not need any treatment but
do require very large blowdown streams. Simulations of these
systems have shown that both plants could increase the cycles of
concentration without the risk of CaCOj scale if sulfuric acid
were used to control the pH of the recirculating water. These
results show that the blowdown can be reduced by almost an order
of magnitude at increased cycles of concentration using only pH
control to inhibit scale formation.
-26-
-------
250-r
200
5
£L
O
OC
Ut
S
Ul
IT
ui
IT
CO
Q.
150
100
50
30
40 50 60 70
CALCIUM CONCENTRATION IN MAKEUP WATER. MG/L
80
90
Figure 2-5
Slipstream rate as a function of makeup calcium
concentration at Colstrip.
-------
250
200
2
Q.
(9
Z
UJ
5
160
00
100
m
a.
It
m
SO
100 150 200 250 300 350
SULFATE CONCENTRATION IN THE MAKEUP WATER. MG/L AS SO|
400
Figure 2-6. Slipstream rate as a function of makeup sulfate
concentration at Colstrip.
-------
The cooling towers at Colstrip presently employ pH
control and are supplied with softened makeup water. The towers
are operated at 10-15 cycles of concentration without scale for-
mation. Simulations of this system have shown that even higher
cycles of concentration could be attained if a slipstream from
the recirculating water were softened rather than the makeup.
This can be accomplished at reduced capital costs because the
size of the treated stream is smaller if the slipstream is sof-
tened.
The cycles of concentration that can be obtained at
Comanche's cooling towers is limited by the potential for silica
scale. Comanche's towers presently operate supersaturated with
respect to Si02 but no significant scale has been noticed in the
condensers. Without conclusive kinetic data with respect to
Si02 formation, simulations of the cooling towers are not suffi-
cient to determine what cycles of concentration Comanche can
operate at safely.
In order to minimize makeup water requirements and
blowdown rates, cooling towers should be operated at the high-
est cycles of concentration which will not produce scale or
corrosion problems in the condensers. The relative saturation
of many scale forming species can be reduced by acid addition
for pH control. Softening of either the makeup or a slipstream
will reduce the relative saturation of CaS04-2H20 which cannot
be controlled with acid treatment. Utilization of these treat-
ment methods may allow many existing cooling towers to increase
their cycles of concentration safely.
Corrosion problems may be encountered in cooling
towers operating at high cycles of concentration if the chlo-
ride levels are excessive. High chloride levels in the tower
makeup may therefore limit the degree of recycle. Removal of
chlorides is very expensive, requiring sophisticated treatment
such as reverse osmosis or brine concentration.
-29-
-------
3.0 ASH SLUICING SYSTEMS
In addition to cooling towers, a major water consumer
encountered at the power plants studied in this program is ash^
sluicing. Of the five plants considered, two employed wet sluic-
ing for fly ash disposal and all five plants used wet sluicing
for bottom ash disposal. Georgia Power Company's Bowen Plant and
Pennsylvania Power and Light Company's Montour Plant both use
cooling tower blowdown to sluice fly ash on a once-through basis
for disposal in ash ponds. The ash pond overflow is discharged
in both cases. All of the plants sluice bottom ash on a once-
through basis except Four Corners and Colstrip, which have recir-
culating bottom ash disposal systems.
This section first presents a typical ash sluicing
flow scheme with discussions of the particular systems studied.
The process description is followed by discussions of the vari-
ous operating parameters and their effects on the design and
operation of ash sluicing systems. The final portion of this
section discusses how recycle/reuse options in ash sluicing sys-
tems may be incorporated in overall plant water systems to min-
imize water requirements and discharges.
3.1 Process Description
In coal-fired boilers, two types of ash residue are
created by combustion of the coal. Fly ash is that portion of
the ash which is carried out of the boiler with the combustion
gases and bottom ash is the ash remaining in the boiler which
collects at the bottom of the boiler. The relative amounts of
fly ash and bottom ash produced depend on the type of furnace
in which the coal is fired. Burning pulverized coal in a dry-
ash furnace will generally result in about 8070 of the ash being
entrained with the flue gas as fly ash. Cyclone furnaces, how-
ever, retain 70-80% of the coal ash as bottom ash leaving only
20-30% as fly ash (BA-465). The fly ash made up 60-80% of the
total ash at the five plants studied, as would be expected from
dry ash furnaces.
Bottom ash is typically removed from the boiler by
periodic washing and subsequent sluicing of the ash to a pond
for disposal. Fly ash must be continuously removed from the
flue gases to prevent the discharge of large amounts of partic-
ulate matter into the atmosphere through the stack. Available
methods for removing the fly ash from the flue gas include elec-
trostatic precipitators, mechanical collectors, fabric filters,
and wet scrubbers. Fly ash collected by wet scrubbers is disposed
-30-
-------
of in a slurry form. Arizona Public Service's Four Corners
Plant (Units 1-3) and Montana Power Co.'s Colstrip Plant employ
wet scrubbing for fly ash collection. The Four Corners fly ash
scrubbing system is open loop whereas the Colstrip S02/particulate
scrubbing system is closed loop. These systems are discussed in
Section 4.0 and Appendices F and J. The remaining three plants
(Comanche, Bowen, and Montour) and Units 4 and 5 of the Four
Corners Plant collect the fly ash with electrostatic precipi-
tators.
Once the fly ash has been collected, it must be trans-
ported to a suitable disposal site. This may be accomplished by
slurrying the ash with water and pumping it to a pond or by
trucking the ash to the disposal site in a dry form. Fly ash
collected by the precipitators at Comanche is trucked away in
a dry state. At Bowen and Montour, the collected fly ash is
slurried to ash ponds (at about 5% solids at Montour and 7%
solids at Bowen). Figure 3-1 represents the type of fly ash and
bottom ash sluicing operations at Montour and Bowen. The ash is
sluiced to the disposal pond where it settles to 40-50% solids.
The excess water is discharged as ash pond overflow. Detailed
discussions of the Bowen and Montour water systems are presented
in Appendices G and I, respectively.
In areas where water is scarce or regulations prohibit
discharging ash pond overflow, a portion or all of the excess
pond water may be recirculated to sluice ash. Figure 3-2 repre-
sents a recirculating ash sluicing system. In the case where all
of the excess pond water is recycled, the makeup requirements are
determined by the pond evaporation rate and the sludge solids
concentration. Recirculating ash sluice systems with blowdown
streams are used at both Colstrip and Four Corners for bottom ash
disposal. The blowdown at Colstrip is the overflow from the
bottom ash pond to the scrubber ponds. This type of ash handling
is not typically used for fly ash disposal since the amount of
leachable species (mostly calcium, magnesium, sodium, and sulfate)
in fly ash is generally much greater than bottom ash.
A recirculating system is much more susceptible to
scaling since dissolved solids in the makeup are concentrated
in addition to the species leached from the ash. This study
concentrates on the feasibility of using recirculating sluicing
systems for fly ash disposal. The effects of operating param-
eters and fly ash reactivity on the scaling tendency in the
system are quantified in the following sections.
-31-
-------
ASH
EVAPORATION
SLUICE fc
WATER
1
SOLID/LIQUID
MIXING
/
POND
* OVERFLOW
SLUDGE
Figure 3-1. Typical once-through ash sluicing flow scheme
ASH
EVAPORATION
MAKEUP
SLUICE
WATER
*- SLOWDOWN
POND RECYCLE
Figure 3-2. Recirculating ash sluicing flow scheme.
-32-
-------
3.2 Process Variables
Two methodologies were used to investigate the effects
of operating parameters on the scale potential in recirculating
ash sluicing systems. The parameters investigated include ash re-
activity, makeup water quality, C02 transfer with the atmosphere,
and degree of recycle.
First, parametric studies were performed using the ash
sluicing computer model. This model calculates the composition
of the ash slurry pond, and recycle liquors and determines the
scale potential in the system. All flow rates and the amount of
soluble species in the ash are the major inputs to the model.
First, an overall material balance determines the composition of
the pond liquor assuming solid-liquid equilibrium for all species
except gypsum. The remaining stream compositions are then calcu-
lated based on the amount of leachable species in the ash. No
precipitation is allowed in the slurry liquor in order to deter-
mine the potential for scale formation by CaSOtt-2H20, CaC03, and/
or Mg(OH)2 solids. Since most of the calculations are either
material balances or equilibrium predictions, the validity of the
model in predicting scale potential depends on the accuracy of
the ash reactivity input to the model. The ash reactivity used
as input to the model in this study was obtained from leaching
studies using deionized water. Using the ash reactivity deter-
mined from the beaker studies is conservative from a standpoint
of scale-free operation because fly ash tends to be more reac-
tive in deionized water than in water with a higher level of dis-
solved solids. In actual operation the water used to slurry the
ash has a high enough concentration of dissolved solids to depress
the reactivity of the ash.
There are several assumptions which are inherent in
performing simulations with the ash sluicing simulation. These
include:
1) Solid-liquid equilibrium is achieved
in the ash pond, with the exception
of CaSOi»-2H20 which is allowed to re-
main supersaturated.
2) Ash dissolution is essentially complete
before the slurry reaches the pond, and
supersaturation of all species is allowed
in the slurry line.
3) All solids precipitation occurs in
reaction vessels or the pond.
-33-
-------
The long residence time in an ash pond is sufficient
for most of the species to reach solid-liquid equilibrium. Since
CaSCs*2H20 supersaturation has been observed in scrubber ponds,
the model did not allow gypsum to precipitate in the pond. The
model is also designed to handle varying degrees of C02 transfer
in the pond. In many cases two simulations were performed, one
with no C02 transfer and one with C02 equilibrium achieved be-
tween the pond liquor and the atmosphere. These two cases serve
as a boundary on the actual amount of C02 transfer occurring.
In actual ponds the degree of C02 transfer is generally between
equilibrium and no transfer. This computer model is discussed
in greater detail in Appendix E.
The second methodology used to study ash sluicing in-
volved bench-scale operations. These operations were performed
with ash samples collected at the plants and makeup water similar
to that measured at these plants. The bench-scale model used a
mix tank to combine the sluice water and the ash which was trans-
ported to a settling tank which simulated the pond. The pond
liquor was then recirculated to the mix tank where it was added
to fresh ash and makeup water.
The results of the leaching studies for Montour and
Colstrip ash are presented in detail in Appendix K. The bench-
scale studies for Montour and Colstrip are discussed in detail
in Appendix D. The results of the bench-scale and leaching
studies for Bowen, Comanche, and Four Corners ash are presented
in the final report for EPA Contract No. 68-02-1319, Ash Char-
acterization Studies, which was performed in support of this
program and isincluded as Appendix L.
3.2.1 Ash Reactivity
Leaching studies were performed to determine the maxi-
mum amount of soluble species in the coal ashes from each of the
five power plants studied. In these studies, a small amount of
ash was mixed with deionized water in a beaker and the pH was
periodically adjusted with HC1 to maintain a constant value.
When no further alkalinity was leached (pH remained constant
without acid addition) the liquor was analyzed for calcium,
magnesium, sodium, and sulfate. The leachable amount of each
species as a fraction of the ash was then calculated.
Table 3-1 presents a summary of the results of the
leaching studies performed for the five ashes. These results
show that calcium and sulfate are the major species leached from
-34-
-------
TABLE 3-1. ASH REACTIVITY DETERMINED FROM LEACHING STUDIES*
i
u>
Species ,
wt. %
PH
Ca
Mg
Na
so.
Montour Bowen Four Corners
6.0 8.1 6.0 8.5 10.4 3.0 6.0 8.5
.32 .28 1.3 .76 .56 .83 .71 .62
/\O rt O ^» ^ «« _« - .
.04 .04 .12 .10 .07 .02 .02 .02
.76 .79 1.3 1.2 1.0 .15 .04 .07
Coraanche
6.0
2.9
.11
.04
1.2
8.5
2.2
.05
.03
.83
Co Is trip
4.0 6.0 8.0
5.1 3.7 3.2
.66 .21 .06
.55 .60 .57
*Reported as wt. % of ash.
-------
each of the ashes. Magnesium is present in substantial quantity
in the Colstrip ash only. The amount of calcium leached decreases
with increasing pH for all cases but the sulfate remains rela-^
tively constant except for the Bowen and Comanche ashes where it
also decreases with increasing pH. The leachable calcium and
sulfate as a function of pH for each of the ashes are plotted in
Figures 3-3 and 3-4.
The ash reactivity will have a major impact on the
feasibility of recirculating ash sluicing systems with respect
to gypsum scale formation since significant quantities of both
calcium and sulfate may be leached from the ash.
3.2.2 Makeup Water Quality
The composition of the makeup water used in a recir-
culating ash sluice system will influence the composition of the
sluice water and therefore the scaling tendency of the system.
Both the bench-scale experiments and the computer model results
showed that poorer quality makeup water (greater amounts of cal-
cium and sulfate) increased the gypsum scale potential in the
fly ash slurry liquor. Table 3-2 presents results from bench-
scale experiments with the Bowen ash and from computer model
calculations for the Montour ash. The first two columns repre-
sent two closed-loop bench-scale experiments where only the
makeup water quality was varied. The first experiment involved
poor quality makeup water and a gypsum relative saturation of
1.68, well above the critical value for scale formation, was
observed. When a much better quality makeup water was used
(lower calcium and sulfate concentrations) the gypsum relative
saturation encountered was 1.35, which is significantly lower.
The last two columns in Table 3-2 show the results
from two computer simulations of a closed-loop ash sluicing
system using the Montour ash. The first simulation involved 8-
cycle cooling tower blowdown as makeup water whereas the second
simulation used river water as makeup. Again, the gypsum rela-
tive saturation decreased (2.8 versus 3.2) when makeup water
with less calcium and sulfate was used.
For systems which have gypsum relative saturations
greater than 1.3-1.4 (the critical range for scale formation)
one treatment option is to use soda ash softening of a portion
of the pond recycle liquor. The magnitude of the treatment will
depend primarily on the ash reactivity but the makeup water
quality will also affect the amount of softening required
-36-
-------
4-
ra
U
»
O CO
> I
< U.
2-
r
-0
Ny COLSTRIP
COMANCHE
^ BOWEN
Q FOUR CORNERS
© MONTOUR
-0
PH
Figure 3-3. Reactive calcium in fly ashes as a function of pH.
-------
LO
OO
1.0
o
> w
? -5
A BOWEN
<> COMANCHE
© MONTOUR
V COLSTRIP
0 FOUR CORNERS
n
V
0
pH
Figure 3-4. Reactive sulfate in fly ashes as a function of pH.
-------
TABLE 3-2.
EFFECTS OF MAKEUP WATER QUALITY ON
RECIRCULATING ASH SLUICE SYSTEM SCALING
POTENTIAL
Bench Scale Experiments Computer Model Results
with Bowen Ash with Montour Ash*
MAKEUP WATER, mg/£
Calcium
Magnesium
Sodium
Chloride
as
Carbonate (as CO 3)
Sulfate (as S0~)
Nitrate (as N0~)
205.
66.
570.
50.
2.4
1940.
19.
30.
8.5
23.
8.5
100.
10.0
227.
44.
65.
176.
30.
555.
44.
28.
5.5
8.
22.
6.
6.8
5.5
FLY ASH SLURRY, mg/A
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SOO
Nitrate (as
920.
2.4
350.
130.
6.0
2500.
__
800.
1.2
330.
34.
6.6
2100.
__
1690.
120.
67.
182.
4.
4350.
46.
1500.
77.
8.4
23.
1.4
3840.
5.7
Relative Saturation** 1.68
1.35
3.2
2.8
10.6
11.1
10.1
10.2
*No solids precipitation was allowed; calculations were made based on leaching
results which represent a worst case situation (i.e., maximum leachable spe-
cies). Bench-scale experiments gave gypsum relative saturations of 1.1-1.2
for the Montour ash with a reaction tank to allow solids to form.
**Critical value, above which scale potential exists, is 1.3-1.4 for CaSOit'2H20.
-39-
-------
Table 3-3 presents the results of three computer simulations of
closed-loop sluicing operations at Montour using slipstream treat-
ment. As the water quality becomes poorer (cycles in cooling
tower increase) the amount of pond recycle water which must be
treated increases from 2TL using river water as makeup to 30/0 and
36% for 8- and 20-cycle cooling tower blowdown, respectively.
3.2.3 Carbon Dioxide Transfer
In a recirculating ash sluice system, there is poten-
tial for C02 transfer between the atmosphere and the pond water.
The degree of C02 transfer will affect the pH of the pond return
water and thus the fly ash slurry composition. Since both CaCOs
and Mg(OH)2 relative saturations are pH dependent (tend to pre-
cipitate at higher pH), the degree of C02 transfer which occurs
will have a significant effect on the scaling tendency in the
system with respect to these two species.
Table 3-4 presents the results from two of the bench-
scale experiments where the only variable between runs was the
transfer of C02 in the pond. For the run with no C02 transfer,
no CaC03 precipitation was noted in the reaction tank even though
the relative saturation was 7.8, which is above the critical value
of 2.5 (see Appendix C). This may be due to the low carbonate
levels in the system (2-5 mg/Jl) . However, when C02 was bubbled
through the pond liquor, the relative saturation increased to
31.8 and about 2.7 mmole/min of CaC03 precipitated in the reac-
tion tank. The gypsum relative saturation decreased slightly
(from 0.5 to 0.4) due most likely to the precipitation of calcium
as calcium carbonate. In every case where the C02 bubbler was
used in the bench-scale experiments, CaC03 precipitation was
noted in the mix tank (ranging from 0.1 to 2.7 mmole/min), indi-
cating that CO2 transfer between the pond and the atmosphere
could cause CaC03 scaling problems with recirculating ash sluice
systems.
Table 3-5 presents the results from two of the computer
model calculations where C02 transfer effects were studied. As
with the bench-scale experiments, the transfer of C02 in the
pond increased the CaC03 scaling potential in the system. A sig-
nificant increase in the amount of CaC03 solids precipitated was
noted along with an increase in relative saturation. The calcu-
lated amount of CaC03 solids formed increased from 0.03 gmole/sec
for the case with no C02 transfer to 1.85 gmole/sec for the case
with C02 equilibrium in the pond. The relative saturation of
CaC03 in the fly ash slurry increased from 29.3 to 89.1.
-40-
-------
TABLE 3-3. EFFECTS OF MAKEUP WATER QUALITY ON
TREATMENT REQUIRED FOR CLOSED-LOOP
ASH SLUICING OF MONTOUR ASH*
Makeup Source
MAKEUP WATER COMPOSITION, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SOO
Nitrate (as NOl)
SLIPSTREAM TREATMENT RATE, 5,/sec
% OF POND RECYCLE TREATED
CALCIUM REMOVAL RATE, gmole/sec
FLY ASH SLURRY LIQUOR, mg/A
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COl)
Sulfate (as SOO
Nitrate (as NO^)
PH
RELATIVE SATURATIONS**
CaCOa
CaSO^ZHaO
River 8-Cycle Cooling
Water Tower Slowdown
28.
5.5
8.0
22.
6.0
68.
5.5
38.5
27.
0.88
500.
77.
1150.
22.8
1.4
3840.
5.7
10.3
0.97
1.06
227.
44.
65.
176.
30.
555.
44.
43.7
30.
1.06
500.
120.
1420.
180.
4.2
4350.
45.4
10.1
2.4
1.07
20-Cycle Cooling
Tower Slowdown
567.
110.
161.
440.
23.
1430.
110.
50.8
36.
1.36
510.
135.
1940.
455.
3.3
5060.
114.
10.3
1.7
1.08
*Based on ash reactivity determined from beaker, leaching studies (worst case):
no C02 transfer allowed in the pond.
&&
Critical values, above which scale potential exists, are 1.3-1.4 for
CaSO^'2H20 and about 2.5 for CaCOs (see Appendix C)
-41-
-------
TABLE 3-4. EFFECTS OF C02 TRANSFER IN POND ON FLY ASH SLURRY
SCALING TENDENCY (BENCH-SCALE RESULTS)
(Colstrip Ash)
No CO2 Transfer C02 Bubbled in Pond
POND LIQUOR, mg/2,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SOl)
Nitrate (as NO 3)
pH
Relative Saturations
CaCOs
Mg(OH)2
CaS04«2H20
FLY ASH SLURRY, mg/S,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SO^)
Nitrate (as NOl)
PH
Relative Saturations **
1240.
0.*
27.6
39.1
2.4
595.
14.3
12.6
3.7
0.5
1000.
0.*
27.6
39.1
4.8
691.
14.3
12.6
481.
0.*
52.9
31.2
654.
566.
12.4
7.6
12.6
0.4
441.
0.*
25.3
29.1
21.0
614.
14.3
11.7
Mg(OH)2
CaC03 PRECIPITATION RATE
ACROSS MIX TANK, mmole/min
7.8
0.5
0.0
31.8
0.4
2.7
*Magnesium levels were not detectable due to the high pH's in the system
(Mg(OH)a precipitation removed virtually all of the liquid phase magnesium
from the system or prevented the dissolution of magnesium from the ash).
**Critical values, above which scale potential exists, are 1.3-1.4 for
CaSOn'2H20, about 2.5 for CaCOs, and about 3.4 for Mg(OH)2 (see Appendix C)
-42-
-------
TABLE 3-5.
EFFECTS OF CO2 TRANSFER IN POND ON FLY ASH SLURRY
SCALING TENDENCY (COMPUTER MODEL RESULTS)
CaC03 SOLIDS FORMED IN
SYSTEM, gmole/sec
(Bowen Ash)
No CO2 Transfer
.032
CO2 Equilibrium
with Air in Pond
POND LIQUOR, mg/2,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as COa)
Sulfate (as SO^)
Nitrate (as NO? )
pH
Relative Saturations
CaCOa
Mg(OH)2
CaSOit'21120
FLY ASH SLURRY, mg/£
Calcium
Magnes ium
Sodium
Chloride
Carbonate (as COJ)
Sulfate (as S0=)
Nitrate (as NO?)
PH
Relative Saturations**
CaC03
Mg(OH)2***
CaSO^-ZHaO
1170.
.01
187.
33.
0.7
1290.
67.
12.5
1.0
1.0
1.0
1410.
11.
187.
33.
9.0
1850.
67.
11.7
29.3
1460.
1.3
570.
28.
188.
33.
33.
1740.
66.
7.9
1.0
9.3 x 10~s
1.0
1050.
28.
187.
33.
29.
2130.
67.
11.3
89-1
1200.
1.3
1.85
*Equilibrium partial pressure of CO 2 in pond specified to be equal to that
in the atmosphere (3.3 x 10"1* atm) .
**Critical values, above which scale potential exists, are 1.3-1.4 for
CaSO^'2H20, about 2.5 for CaC03 , and about 3.4 for Mg(OH)2 (see Appendix C)
***Solid precipitation was not allowed. Kinetic studies (Appendix C) indicate
that at these relative saturations, the magnesium will precipitate, result-
ing in very low magnesium concentrations in the liquid phase.
-43-
-------
These cases represent the two extremes of what may
actually happen in a plant situation. The samples taken at the
plants studied in this program indicate that some C02 transfer
occurs in the pond but complete equilibrium is not always _?
achieved. At Bowen the partial pressure of C02 was 2 x 10 atm
and at Montour the partial pressure was slightly above the equili-
brium value of 3.3 x 10~4 atm. The degree of C02 transfer at a
specific location will have a significant effect on the CaCOa
relative saturation in the pond recycle and fly ash slurry
liquors.
3.2.4 Degree of Recycle
The previous discussions have considered completely
closed-loop ash sluicing operations and the effects of various
operating parameters on these systems. However, in some cases
completely closed-loop operation may not be desired and treatment
steps might be eliminated. For example, if ash pond overflow is
.used as makeup to a scrubbing system, the ash sluicing network
would be only partially closed-loop since a blowdown stream to a
scrubber is used. Table 3-6 presents two computer simulation
cases for a sluicing system using Montour ash. In the first
case (column one) closed-loop operation is employed so that all
of the pond overflow liquor is recycled to the system. A slip-
stream is taken from the pond recycle to remove calcium and con-
trol gypsum scale potential in the fly ash slurry. A slipstream
treatment rate of about 77 5,/sec was required.
In the second case, gypsum scale potential is control-
led by taking a blowdown stream from the system to keep the sul-
fate level low enough in the circulating liquor to prevent gypsum
supersaturation. A blowdown rate of approximately 51 &/sec was
required to maintain a gypsum relative saturation near 1.0 in
the fly ash slurry liquor. This corresponds to about 1870 of the
pond recycle liquor.
3.3 Recycle/Reuse Alternatives in Ash Sluicing Systems
Of the ash sluicing systems encountered, recirculation
of the ash pond water was only practiced with bottom ash which
is in general much less reactive. Both of the plants which
sluice fly ash do so on a once-through basis. These studies
revealed that a recirculating fly ash system may be employed
but treatment may be necessary to prevent gypsum scale formation.
The amount of leachable species in the ash and makeup water
quality are the major parameters which determine the level of
treatment necessary.
-44-
-------
TABLE 3-6.
EFFECTS OF DEGREE OF RECYCLE IN AN
ASH SLUICING SYSTEM USING MONTOUR ASH*
Fly Ash Rate, kg/hr
(Ib/hr)
Pond Recycle Rate to Fly Ash Sluice, A/ sec
(GPM)
Sluice Water Makeup Rate, 2,/sec
(GPM)
Pond Overflow Rate, &/sec
(GPM)
Slipstream Treatment Rate,** A/sec
(GPM)
Fly Ash Slurry, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CDs)
Sulfate (as SO^)
Nitrate (as NO 3)
pH
- . ***
Relative Saturations
CaCOs
CaSO^-ZHaO
Closed-Loop
Operation
62,200
(137,000)
284
(4,500)
37.8
(600)
0.0
(0.0)
77.0
(1,220)
500
77
1,150
22.8
1.4
3,840.
5.7
10.3
0.97
1.06
Slowdown Taken
from System
62,200
(137,000)
284
(4,500)
44.2
(1,400)
51.1
(800)
0.0
(0.0)
624
37
8.2
22.4
2.2
1,600.
5.6
10.3
3.3
1.0
*Based on ash reactivity from ash leaching experiments (worst case); no C02
transfer in the pond.
**Based on treatment to 50 mg/& Ca to account for inefficiencies.
***Critical values, above which scale potential exists, are 1.3-1.4 for
and about 2.5 for CaC03 (see Appendix C)
-45-
-------
Determining the quantity of teachable species in
different ashes is a difficult task in that the amount leached
depends on liquor pH and composition. Bench-scale experiments
revealed that solutions containing more dissolved species leach
less of the ash than solutions with lower dissolved solids con-
centration. The leaching studies showed the variability in ash
compositions and the pH dependency of the soluble species in
fly ash. The results of the leaching studies represent the
maximum levels of soluble species in each ash studied and were
therefore used to represent "worst case" operation. Pilot or
additional bench-scale studies are recommended to determine more
accurately the solubility characteristics of a particular ash
before a recirculating fly ash sluice system is implemented.
Carbon dioxide mass transfer between the pond liquor
and the atmosphere will affect the CaC03 and Mg(OH)2 scale poten-
tial in a fly ash sluice system. Increased C02 sorption in the
pond will increase CaCOs scale potential and decrease Mg(OH)2
scale potential. Since these scales are pH dependent (higher pH
means more scale potential) problems may be encountered with
very alkaline ashes. However, these scales are typically softer
than gypsum scale and may be eroded by the ash. Installing
reaction vessels prior to the fly ash slurry line may allow
precipitation of these solids in a controlled fashion so that
scale formation in the slurry line may be minimized. Again,
pilot or bench-scale studies are recommended to more accurately
quantify CaCOs and/or Mg(OH)2 scale potential for a particular
ash before a recirculating system is implemented.
-46-
-------
4.0 SO2/PARTICULATE SCRUBBING SYSTEMS
The third type of major water consumer encountered at
the plants studied in this program is combined particulate and
SO2 scrubbing. Of the five plants investigated, two had scrubbing
systems. The Four Corners Plant of Arizona Public Service has ven-
turi particulate scrubbers on three of the five generating units
although some SO2 removal is also obtained. The Colstrip Plant of
Montana Power Co. has combined SO2 and particulate scrubbers on
both of the generating units. Venturi collectors for particulate
removal are followed by spray chambers for SO2 removal.
This section of the report first presents a typical
scrubbing process flow scheme including discussions of differences
between the typical system and actual systems studied. The pro-
cess description is followed by discussions of the various oper-
ating parameters and their effects on the design and operation of
scrubbing systems. The final portion of this section discusses
the potential for scrubbing in various recycle/reuse alternatives
at power plants. Since most scrubbing systems are designed for
zero discharge, they may possibly be used as receptors for the
final water effluent in a cascaded water system to achieve zero
discharge or reduced blowdown for an entire plant water network.
4.1 Process Description
A typical scrubbing system may be divided into three
maj or operations:
1) gas cleaning,
2) solids precipitation, and
3) solids concentration.
The combination of these operations is shown in the simplified
scrubbing system in Figure 4-1. Gas cleaning is accomplished^in
the scrubber vessel, solids precipitation occurs in the reaction
tank, and solids concentration is achieved in the solid-liquid
separator which may be a clarifier, filter, pond, or any combina-
tion of the three. This section presents a discussion of each of
these major operations observed in scrubbing systems and how the
two specific systems studied deviate from the typical flow scheme
in Figure 4-1. More detailed discussions of the scrubbing systems
at Four Corners and Colstrip may be found in Appendices F and J,
respectively.
-47-
-------
MAKEUP
WATER
FLUE
GAS
ALKALI
STACK
GAS
DEMISTER
SCRUBBER
REACTION
TANK
SOLID/LIQUID
SEPARATION
WASTE
Figure 4-1. Typical scrubbing system flow scheme
-48-
-------
4.1.1 Gas Cleaning
In both particulate and combined SO 2/particulate scrub-
bing systems, the boiler flue gas is contacted with a recircula-
ted slurry in the scrubber vessel where the particulates (fly
ash) and/or S02 are absorbed. Venturi scrubbers are used for
particulate removal at both Four Corners and Colstrip. However,
at Colstrip where the scrubbers are designed for combined par-
ticulate and SO2 removal, the venturi sections are followed by
spray sections where most of the SO2 is absorbed.
Spray towers provide more liquid-gas contact area and
longer residence times than Venturis and are therefore more effi-
cient at removing SO2 from the flue gas. At Four Corners only
about 30% of the SO2 in the flue gas is removed in the Venturis,
but at Colstrip 74% of the scrubber inlet SO 2 is removed by the
combination of venturi and spray sections.
To prevent excess carryover of the scrubbing liquor
with the clean gas, mist eliminators are used in the scrubbers.
The mist eliminators are generally washed with fresh makeup water
which falls into the scrubber after being sprayed over the demis-
ters. At Colstrip, a wash tray collects some of the wash water
so that it may be reused along with the makeup water as demister
wash.
4.1.2 Solids Precipitation
The second major operation in scrubbing systems is
solids precipitation. In a lime or limestone based closed-loop
scrubbing system, the S02 which is absorbed must be removed from
the system by precipitation of calcium sulfate and calcium sul-
fite. The required rate of precipitation is determined by the
rate of absorption of S02 from the flue gas. At both Four Cor-
ners and Colstrip, greater than 90% of the sulfite formed in the
liquid phase by S02 sorption is oxidized to sulfate, making the
precipitation of gypsum (CaSO,, «2H20) the controlling rate of
solids formation.
It should be noted that revision of system operation
could alter the sulfite oxidation rate significantly. Systems
with lower than approximately 15% oxidation can be designed for
the so-called "subsaturated gypsum mode" whereby the calcium sul-
fate coprecipitates with calcium sulfite. The results presented
here assume high (>90%) sulfite oxidations.
-49-
-------
The precipitation of the absorbed sulfur occurs in a
reaction tank where adequate time is allowed for crystal^growth
of recirculated calcium sulfate and calcium sulfite particles.
If the residence time of the reaction tank is too small, then the
concentrations of sulfate and sulfite will increase until nuclea-
tion occurs, resulting in scale formation. The scale will most
likely form in the scrubber where sulfur concentrations are
highest.
At Four Corners, where scaling has been noted, the re-
action time for solids precipitation is very small (about one
minute), whereas at Colstrip the reaction tank residence time is
about eight minutes. No scaling has been reported for the Col-
strip scrubbing system. The effects of reaction tank volume on
system operation will be quantified in Section 4.2 which discus-
ses process variables.
The sorbed fly ash solids and precipitated sulfur sol-
ids must be removed from the recirculating slurry at a rate suf-
ficient to control the solids concentration in the circulating
slurry at the desired level. At Colstrip, the blowdown is taken
at a rate sufficient to keep the circulating solids concentra-
tion at about 12%. At Four Corners, the circulating solids con-
centration is about 27o. The low value of 2% is used at Four Cor-
ners to minimize erosion problems. Higher solids concentrations
are more abrasive but require smaller reaction tanks due to the
increased number of precipitation sites. The blowdown from the
recirculating slurry is pumped to the solid/liquid separation
portion of the system.
4.1.3 Solid/Liquid Separation
In order to minimize water requirements for a scrubbing
system, solid/liquid separation is employed to recover a portion
of the water used to slurry the waste solids (ash, CaSCU^HzO,
CaS03'%H20). Higher percent solids also allow for easier disposal
of the sludge as landfill. This separated water can be recycled
to the scrubbing system to eliminate any aqueous discharge and
reduce makeup water requirements. The final sludge water content
and the evaporation occurring in the scrubber determine the makeup
water requirements for the system.
Typical solid/liquid separation techniques include clar-
ification, filtration, and ponding or combinations of these tech-
niques. At Four Corners the 270 solids slurry blowdown is clari-
fied to about 107o solids (typical operation) and pumped to a pond.
The clarifier overflow is returned to the scrubbing system. In
-50-
-------
the ash pond, the sludge settles to about 50% water and the ex-
cess water (not evaporated or occluded with sludge) is discharged
rather than recirculated.
In the Colstrip scrubbing system, the 12% solids stream
is diluted to about 6% solids with pond water and pumped to the
scrubber ponds where the solids (ash and CaSO^»2H20, predominant-
ly) settle to about 50% water. The remaining water which is not
evaporated is recycled to the scrubbing system.
4.2 Process Variables
The effects of important process variables concerning
scale control were investigated as well as the effects on makeup
water requirements for the two scrubbing systems studied using
the computer models discussed in Appendix E. The parameters
affecting the system water makeup requirements are S02 removal
.rate, ash removal rate, and pond recycle rate. Other parameters
considered include liquid-to-gas ratio, circulating slurry solids
concentration, reaction tank volume, and makeup water quality.
The following sections contain quantitative evaluations of the
effects of these parameters on the operation of particulate and
S02 scrubbing systems. Detailed analyses of the Four Corners and
Colstrip scrubbing operations are presented in Appendices F and
J, respectively.
The scrubbing models used in this study calculate all
stream compositions and flow rates in the system using precipi-
tation rate kinetics for CaSCU^HaO and CaS03*%H20, which are
the solids formed in lime/limestone scrubbing systems, and vari-
ous input parameters. These parameters characterize the operating
conditions for a particular scrubbing system and include flue gas
flow and composition, fly ash rate and composition, makeup water
composition, lime addition rate, tank volumes, scrubber feed flow
rate and percent suspended solids, percent oxidation in the system,
and percent solids in the sludge.
Iterative calculations are performed around the scrub-
bing loop through the scrubber vessel and the scrubber recycle
tank (if applicable) until relative saturations and stream com-
positions satisfy the rate equations. The calculations are per-
formed for ancillary equipment such as the reheat and fan require-
ments, and to determine additional stream compositions. Makeup
water requirements are calculated by an overall system balance.
-51-
-------
Several assumptions are inherent in performing scrub-
ber simulations with the model outlined above. These are enum-
berated below:
1) The scrubber exit gas is saturated with
respect to water.
2) Equilibrium exists between C02 in the
stack gas and liquor in the scrubber
bottoms.
3) The scrubber bottoms and stack gas tem-
peratures are the adiabatic saturation
temperature of the flue gas.
4) All oxidation was assumed to occur in
the scrubber.
5) All solids precipitation occurs in reac-
tion vessels.
4.2.1 SO2 Removal Rate
The effect of SO2 removal rate on scrubber makeup water
requirements may be determined by examining the results of two of
the Four Corners process simulations. In the high solids opera-
tions case where the circulating solids concentration is 9%, the
clarifier underflow is 30% and the final slurry discharged is ~LTL,
only 307> 862 removal is specified. 50% SOz removal was specified
for Alternative Two. This alternative mode of operation is with
10% solids in the circulating liquor and 3070 solids in the final
slurry discharged. Table 4-1 presents the characteristics of the
solid waste for each of these cases as well as the operating
conditions and makeup water requirements.
Direct comparison of the makeup water requirements for
these two cases as a function of SC>2 removal cannot be made since
the waste suspended solids concentrations are different. Table
4-1 shows an adjusted makeup water rate based on the solids flow
in the waste and a solids concentration of 30% in the waste for
both cases. The difference in adjusted makeup water requirements
is only 2.1 5,/sec or about 370 of the total makeup for a change in
SO2 removal from 30% to 50%.
This effect is small due in part to the ash comprising
more than 90% of the solids in the waste. Although the amount of
sulfate solids changed considerably, the overall impact is small,
since the sulfate solids represent less than 10% of the total
-52-
-------
TABLE 4-1. EFFECTS OF S02 REMOVAL RATE ON FOUR CORNERS
SCRUBBER MAKEUP REQUIREMENTS
High Solids
Operations
Alternative Two
S02 in Flue Gas, ppm
Ash in Flue Gas, g/sec
(Ib/min)
SO2 Removal, %
Particulate Removal, %
Final Slurry to Ash Pond
Wt. % Solids
Solids Composition, Wt. %
CaSOi»-2H20
Inert (Ash)
Solids Flow, g/sec
(Ib/min)
Makeup Water Requirements, i/sec
(GPM)
Evaporation
Occluded with Solids
Total
Makeup Water Rate Adjusted to 30%
Solids in Waste, £/sec (GPM)
Evaporation
Occluded with Solids
Total
640.
16,200.
(2140)
30.
99.7
17.3
30.0 (475)
79.0 (1250)
109.0 (1735)
30.0 (475)
38.6 (612)
68.6 (1087)
640.
16,200.
(2140)
50.
99.7
30.
3.9
96.1
16,600.
(2200)
7.9
92.1
17,500.
(2300)
30.0 (475)
40.7 (645)
70.7 (1120)
30.0 (475)
40.7 (645)
70.7 (1120)
-53-
-------
solids. The total amount of solids changed only by about 5/0. As
the amount of sulfur removed in relationship to the ash removed
increases, the effects of S02 removal on makeup water requirements
will become more pronounced.
The effect of changes in S02 removal rate on makeup
water rate will be dampened by the fact that water evaporation
in the scrubber frequently represents a major portion of the
makeup water requirement. As the evaporation becomes a larger
part of the makeup water, changes in S02 removal rate will have
a decreasing effect on the overall scrubber makeup rate. At
Four Corners, for the high solids operation case (17% solids in
final waste), the evaporation is about 30% of the total makeup.
However, at Colstrip the evaporation is about 80% of the total
makeup because the final sludge solids concentration is about
50%, and only a small amount of water is lost with the solids.
4.2.2 Ash Removal Rate
The effect of the ash removal rate on makeup water re-
quirements for combined SOa and particulate scrubbing is more
pronounced than that of the S02 removal rate in the situations
studied. The scrubbing situations encountered were for low sul-
fur applications. In the cases studied, most of the solid waste
was ash as opposed to calcium sulfate or calcium sulfite.
As the ash removal rate is decreased, by burning a coal
of lower ash content or by decreasing load for example, the makeup
water requirements for the scrubbing system are decreased since
less water is lost from the system by occlusion with the solids.
By an overall mass balance around the scrubbing system, the makeup
water rate may be calculated by Equation 4.1:
- Xn)
~(1 " Xs) ^
where
M = makeup water requirement, £/sec
E = evaporation (scrubber + pond), £/sec
A = ash removal rate, g/sec
p, = density of water, g/Jl
J_i
XD= weight fraction of ash which is soluble
-54-
-------
XA = wei§ht fraction of solids in waste which is ash
Xg = weight fraction solids in waste stream
The makeup water requirement is the sum of two compo-
nents. One is the evaporation in the scrubber and the other is
the water occluded with the solids in the waste stream. The sec-
ond term in Equation 4.1 represents the water lost with the solid
waste and is directly proportional to the ash removal rate, A.
Table 4-2 presents data from two scrubbing simulations for the
Four Corners Plant and two for the Colstrip Plant which illustrate
the effects of flue gas ash content on scrubbing system makeup
water requirements. The two cases for Four Corners represent a
reduction in flue gas ash content of 60%, from 16,200 g/sec to
6,480 g/sec. The water lost through solids occlusion was reduced
from 17.5 £/sec to 8.0 £/sec resulting in an overall makeup water
reduction of about 20% from 50.7 a/sec to 41.2 £/sec.
Lowering the ash removal rate did not produce a propor-
tional reduction in water lost with the solids since the fraction
of ash in the solids also decreased. From Equation 4.1, lowering
the fraction of ash in the solids, X., will increase the makeup
water requirements. A directly proportional decrease of occluded
water would have been to 7.0 5,/sec insteam of the 8.0 £/sec calcu-
lated. This difference is due to X. changing as well as A, the
ash removal rate, in Equation 4.1.
The two simulated cases for the Colstrip scrubbing
system produced similar results in that a 30% reduction in ash
removal rate caused only a 14% reduction in occluded water.
Again, this is due to the change in the solids composition. The
weight fraction of ash in the solids decreased from 0.568 to 0.474
for these two cases. Overall makeup water requirements were re-
duced by only about 2%, from 58.2 H/sec to 57.1 5,/sec, since most
of the makeup water requirements are needed to replace evaporated
water rather than occluded water.
The ash removal rate will also affect the scaling ten-
dencies in the system. A decrease in ash removal rate will in-
crease the amount of calcium sulfite and calcium sulfate solids
being recirculated. This provides more crystal sites for preci-
pitation and therefore decreases the tendency of the system to
form scale. For example, comparison of the reaction tank liquors
for the two Colstrip simulations shown in Table 4-3 shows a de-
crease in calcium sulfate relative saturation from 1.41 to 1.37
when the amount of ash removed is decreased from 5500 g/sec to
3800 g/sec.
-55-
-------
TABLE 4-2. EFFECTS OF ASH REMOVAL RATE ON SCRUBBER MAKEUP REQUIREMENTS
Ln
I
Ash in Flue Gas, g/sec
(Ib/min)
Particulate Removal, %
Total Fraction of Ash Dissolving
SOa in Flue Gas, ppm
SO 2 Removal, %
Solid Waste
Wt. % solids
Solid Composition, wt. %
CaS03'%H20
CaS(V2H20
Inert (Ash)
Solids Flow, g/sec
(Ib/min)
Makeup Water Requirement, A/sec (GPM)
Evaporation
Solids Occlusion
Total
Four
Alternative*
Three
16,200.
(2140)
99.7
.011
640.
50.
50.
8.5
91.5
17,500.
(2300)
33.2 (525)
17.5 (278)
50.7 (803)
Corners
Alternative**
Four
6480.
(860)
99.7
.011
640.
50.
50.
18.8
81.2
8000.
(1060)
33.2 (525)
8.0 (127)
41.2 (652)
Colstrip
Design
Conditions
5500.
(730)
99.6
.193
790.
74.
50.
4.1
39.0
56.8
7740.
(1020)
50.5 (800)
7.7 (122)
58.2 (922)
Low Ash
Conditions
3800.
(500)
99.6
.193
790.
74.
50.
5.2
47.4
47.4
6560.
(870)
50.5 (800)
6.6 (105)
57.1 (905)
*Repiped scrubber system with adequate reaction tank volume to prevent scale and recycle of ash pond
overflow.
**Same as Alternative Three except ash content of flue gas lowered by 60%.
-------
TABLE 4-3.
EFFECT OF ASH REMOVAL RATE ON COLSTRIP
SCRUBBING SYSTEM SCALING TENDENCY
Design
Conditions
Low Ash
Conditions
Ash in Flue Gas ,
/sec
Ib/min)
Scrubber Recycle Slurry
pH
Liquor Composition, mg/£
Calcium
Magnesium
Sodium
Chloride
Carbonates (as COT)
Sulfates (as SO"^)
Sulfites (as SOl)
Nitrates (as NO 3)
CaSOi^HaO Relative Saturation*
Solid Composition, wt. 70
CaSO^«2H20
Inert (Ash)
CaS03'%H20
5,500.
(730)
5.0
733.
5,285.
444.
117.
153.
21,000.
3,560.
9.6
1.41
38.9
57.3
3.8
3,800.
(500)
5.2
715.
4,267.
448.
132.
156.
17,800.
2,135.
10.9
1.37
47.2
47.8
5.0
*The critical value, above which scale potential exists, is
1.3-1.4 for CaSO^'2H20.
-57-
-------
4.2.3 Ash Pond Recycle Rate
Although most throwaway scrubbing systems are designed
for closed-loop operation (no aqueous discharges) some systems
may operate with a blowdown from time to time. This section des-
cribes the impact on the scrubbing system water balance of oper-
ation with or without recycle of ash pond overflow. The Four
Corners scrubbing system simulations chosen to quantify the ef-
fects of ash pond recycle are summarized in Table 4-4. Detailed
operating data and simulation results may be found in Appendix F.
The first column of Table 4-4 (Alternative Two) rep-
resents operation of the system with a solid waste suspended
solids concentration of 30%. No pond water is recycled to the
system in this case. Alternative Three represents identical
conditions except the ash pond water is recycled to the reaction
tank after the waste solids have settled to 50%, water. The sol-
ids composition was changed slightly due to changes in the liquid
composition in the system, resulting from the recycle of the pond
water. The major influence on the makeup water rate, however,
was the change in solids concentration in the final waste from
307o in Alternative Two to 50% for Alternative Three. The occlu-
ded water decreased from 40.7 I/sec to 17.5 £/sec to reduce the
overall makeup requirements from 70.7 5,/sec to 50.7 £/sec. The
evaporation component of the makeup water requirements increased
slightly in Alternative Three due to the evaporation occurring
in the ash pond.
Recycle of the ash pond liquor, in addition to decreas-
ing makeup water requirements, raises the dissolved solids level
in the scrubbing system. The effect on the system scaling ten-
dency is illustrated by the two simulation cases presented in
Table 4-5. The calcium sulfate relative saturation of the scrub-
ber liquor remains unchanged at 1.07 for both cases although the
TDS increases from 4,000 to 5,200 mg-/A. Therefore, with adequate
reaction time, recycle of the ash pond overflow to the reaction
vessel will not cause scaling problems. The total dissolved
solids concentration increases, however, from about 4,000 mg/& to
about 5,200 mg/£.
4.2.4 Slurry Solids Concentration
The scrubber makeup requirements are not affected by the
recirculating slurry solids content but a study of water recycle/
reuse opportunities in scrubbing systems necessitates an under-
standing of scaling and the parameters which may be used to control
scaling potential. The recirculating solids concentration will
have a significant impact on the scale potential in the system
-58-
-------
TABLE 4-4
EFFECT OF POND RECYCLE ON FOUR CORNERS
SCRUBBER MAKEUP REQUIREMENTS
Ash in Flue Gas, g/sec
(Ib/mln)
Particulate Removal, %
Total Fraction of Ash Dissolving
SOa in Flue Gas , ppm
SO 2 Removal, %
Solid Waste
Wt. % Solids
Solid Composition, Wt. %
CaSCK'ZHaO
Inert (Ash)
Solid Flow, g/sec
(Ib/min)
Pond Recycle Rate, 5,/sec
(GPM)
Makeup Water Requirement , £ /sec
(GPM)
Evaporation
Solids Occlusion
Total
Alternative Two*
16,200.
(2140)
99.7
.011
640.
50.
30.
7.9
92.1
17,500.
(2300)
0.0
(0)
30.0
(475)
40.7
(645)
70.7
(1120)
Alternative Three**
16,200.
(2140)
99.7
.011
640.
50.
50.
8.5
91.5
17,500.
(2300)
20.0
(320)
33.2
(525)
17.5
(278)
50.7
(803)
*Repiped scrubbing system with adequate reaction time to prevent scale, no
ash pond recycle.
**Same as Alternative Two except all ash pond overflow is recycled.
-59-
-------
TABLE 4-5. EFFECT OF ASH POND RECYCLE ON FOUR CORNERS
SCRUBBING SYSTEM SCALING TENDENCY
Alternative Two* Alternative Three**
Pond Recycle Rate, &/sec °-° 2°'°
(GPM) (0) (320)
Solid Waste Solids Concentration,
wt. % 30. 50.
Scrubber Recycle Slurry
PH
Liquor Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonates (as CO 3)
Sulfate (as SOO
Sulfite (as SOl)
Nitrate (as NO^)
TDS
CaSOit'ZHaO Relative Saturation***
Solids Composition, wt. %
CaSO^'ZHzO
Inert (Ash)
6.9
670.
84.
450.
235.
116.
2400.
33.
16.
4000.
1.07
7.9
92.1
7.0
640.
145.
790.
390.
105.
3100.
35.
27.
5200.
1.07
8.1
91.9
*Repiped system with adequate reaction time to prevent scale and no pond
recycle.
**Same as Alternative Two except ash pond overflow is recycled.
***The critical value, above which scale potential exists, is 1.3-1.4 for
CaSO<*'2H20.
-60-
-------
because the solids provide precipitation sites for CaS03-%H20 and
CaSCK-ZHzO. A decrease in the number of sites slows precipitation
rates and therefore maintains more calcium and sulfate in solution,
raising the calcium sulfate relative saturation. This effect is
illustrated by the two simulations of the Colstrip scrubbing sys-
tem shown in Table 4-6. As the slurry solids concentration was
reduced from 12.670 to 7.6%, the calcium sulfate relative satura-
tion increased to 1.73, which is considerably above the critical
range for scale formation of 1.3-1.4.
The recirculating slurry solids concentration is there-
fore an important parameter to control as far as scale prevention
is concerned. A tradeoff does exist, though, in that slurries
with higher solids concentrations are more abrasive.
4.2.5 Reaction Tank Size
In a lime or limestone based S02 scrubbing system, the
control of CaSO^2H20 and CaS03*%H20 scale potential is an extre-
mely important consideration. If scaling conditions are realized
for significant amounts of time in any part of the system, chemi-
cal scale will probably be deposited on equipment which will even-
tually necessitate system shutdown for cleaning. Pure phase cal-
cium sulfite or calcium sulfate kinetics can be described by the
expression:
R = KafCV (R.S. - 1) (4.2)
where
R = precipitation rate
K = temperature dependent constant
a = crystal interfacial area per mass of
precipitating solid
f = weight fraction of the precipitating
species in the solid phase
C = total solids concentration in the slurry
V = reaction tank volume
R.S. = relative saturation of the precipitating
species.
-61-
-------
TABLE 4-6. EFFECT OF SLURRY SOLIDS CONCENTRATION ON COLSTRIP
SCRUBBER SCALE POTENTIAL
Scrubber Recycle Slurry
Suspended Solids, wt. %
pH
Liquor Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonates (as GOT)
Sulfate (as SO")
Sulfite (as S07)
Nitrate (as NOl)
CaSO^-ZHzO Relative Saturation*
Solid Composition, wt. %
CaSOi*'2H20
Inert (Ash)
CaS03-%H20
Design
Conditions
12.6
4.98
733.
5,285.
444.
117.
153.
21,000.
3,560.
9.6
1.41
38.9
57.3
3.8
Low
Solids
7.6
5.09
897.
5,590.
469.
124.
167.
22,300.
3,930.
10.2
1.73
38.7
57.9
3.4
''The critical value, above which scale potential exists is
1.3-1.4 for CaS04*2H20.
-62-
-------
The system precipitation rates are set by the S02 removal rate
and oxidation rate. As can be seen from Equation 4.2, increases
in the weight fraction of gypsum in the solid phasei the total
solids concentration in the circulating slurry, or the reaction
tank volume should decrease the gypsum relative saturation in the
reaction tank for a specific precipitation rate. Table 4-7 illus-
trates the effect of reaction tank volume on relative saturation
for the Four Corners scrubbing system.
The only difference between Cases One and Two of Alter-
native Three is the reaction tank volume. This alternative in-
volved recycle of the ash pond overflow in a repiped system with
adequate reaction tank volume for scale control. As the tank size
was decreased from 37,500 m3 to 21,200 m3, the gypsum relative
saturation in the tank increased from 1.07 to 1.13. The decrease
in reaction time kept more calcium and sulfate in the liquid phase
and less in the solid phase, resulting in the increase in relative
saturation. The change in solid phase composition was very small
and did not have a significant impact on the results. The change
in relative saturation is only about 5% different from what
would be expected by direct ratio using Equation 4.2 (neglecting
changes in solid composition).
These results are based on the assumption that 98.67<> of
the sorbed S02 is oxidized. Four Corners scrubber tests performed
by Arizona Public Service since this study was made indicate that
the amount of sorbed S02 that is oxidized is a function of the pH
of the scrubbing liquor. Lime addition in their venturi scrubbers
at Four Corners raised the pH and reduced the oxidation to a level
that permits operation in the "subsaturated gypsum mode". They
found that the reaction tank volume required with reduced oxida-
tion is significantly less than that predicted using the 98.6%
oxidation assumption in the simulations.
4.2.6 Liquid-to-Gas Ratio
The effects of liquid-to-gas ratio (L/G)on the revised
Four Corners scrubbing system operation are shown in Table 4-8.
The simulation results shown in the first column (Alternative Two,
Case One) represent operation at the present L/G at Four Corners
(4.7 £/m3 & STP) but with adequate reaction time for scale con-
trol. The calculated pH for the scrubber bottoms stream is 2.9.
This low value could cause corrosion problems_if the system oper-
ated in this manner for extended periods of time.
-63-
-------
TABLE 4-7.
EFFECT OF REACTION TANK VOLUME ON FOUR CORNERS
SCRUBBING SYSTEM SCALING POTENTIAL
Alternative Three
Case One*
Alternative Three
Case Two**
Combined Reaction Tank Volume***, m
(ft3)
Scrubber Recycle Slurry
Suspended Solids, wt. %
pH
Liquor Composition, mg/5.
Calcium
Magnesium
Sodium
Chloride
Carbonate (as C0"i)
Sulfate (as SOlj)
Sulfite (as SOI)
Nitrate (as NOl)
CaSOi^-ZHzO Relative Saturation****
Solid Composition, wt. %
37,500.
1.32 x 106
10.0
7-0
640.
145.
790.
390.
104.
3,110.
35.
27.
1.07
21,200.
7.5 x 101
10.0
7.0
670.
140.
760.
360.
93.
3,120.
40.
25.
1.13
Flue
S02
S02
S02
CaSOit^HaO
Inert (Ash)
Gas Ash Flow, g/sec
(Ib/min)
in Flue Gas, ppm
Removal Rate, %
Oxidation, %
8.0
92.0
16,200.
(2140)
640.
50.
98.6
7.9
92.1
16,200.
(2140)
640.
50.
98.6
*Repiped Scrubbing System, more than adequate reaction tank volume for scale
control.
**Same as Case 1 but reduced reaction tank volume.
***Combined volume of all six proposed reaction tank vessels (one per scrubbing
train).
****The critical value, above which scale potential exists, is 1 3-1 4 for
-64-
-------
TABLE 4-8. EFFECTS OF LIQUID-TO-GAS RATIO ON
FOUR CORNERS SCRUBBING OPERATION
Liquid- to-Gas Ratio, £/m3 @ STP
(gal/ 1000 ACF)
Scrubber Effluent Slurry
Suspended Solids, wt. %
PH
Liquor Composition, mg/5,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as C0"i)
Sulfate (as SO=O
Sulfite (as S0=)
Nitrate (as NO!)
CaSOif«2H20 Relative Saturation ***
Solid Composition, wt. %
CaSO^*2H20
Inert (Ash)
Flue Gas Ash Flow, g/sec
(Ib/min)
S02 in Flue Gas, ppm
S02 Removal, %
S02 Oxidation, %
Alternative Two
Case One*
4.7
18.7
10.5
2.9
700.
85.
450.
240.
124.
2,750.
37.
16.
1.16
7.4
92.6
16,200.
(2140)
640.
50.
98.6
Alternative Two
Case Two**
10.0
39.8
10.2
3.9
690.
85.
450.
240.
120.
2,580.
35.
16.
1.14
7.6
92.4
16,200.
(2140)
640.
50.
98.6
*Alternative Two involves repiping scrubbing system, adding reaction tank
volume to control scale. Ash pond overflow is not recycled.
**Same as Case One except L/G increased.
***The critical value, above which scale potential exists, is 1.3-1.4 for
CaSCV2H20.
-65-
-------
The second column in Table 4-8 presents the results
from a simulation with a higher L/G of 10.0 £/m3 @ STP. The
calculated pH of the scrubber effluent stream was 3.9 as opposed
to the value of 2.9 for the previous case. This value is still
not ideal from the standpoint of corrosion control but is some-
what better than the 2.9 pH scrubber liquor. The reason for the
rise in pH from Case One to Case Two is that less SOa is absorbed
per volume of scrubbing liquor so that the change in liquid phase
sulfur concentration across the scrubber is less. This results
in less of a pH drop across the scrubber. In both cases the
scrubber feed pH was 6.9.
A small change in gypsum relative saturation in the
scrubber effluent occurred with the increase in L/G (dropped
from 1.16 to 1.14 when L/G increased from 4.7 £/m3 @ STP to
10.0 2,/m3 @ STP). This again is due to the smaller sorption
rate of S02 per volume of scrubber liquor and therefore a smal-
ler change in relative saturation across the scrubber.
4.2.7 Makeup Water Quality
The effects of makeup water quality on scrubbing system
operation and design were determined by using the Colstrip scrub-
bing system data and varying the makeup water composition input
to the simulation model. Table 4-9 presents the results from
three simulations where the makeup water composition was varied.
The first case represents the design conditions with makeup water
being softened river water. The other two cases represent using
untreated river water and cooling tower blowdown as makeup water.
The results from these simulations indicate that the
makeup water has very little effect on the recirculating slurry
scaling tendency. The relative saturation of gypsum did not
change appreciably between cases. The liquor composition did
change considerably. The total dissolved solids concentrations
ranged from 31,000 mg/£ for the base case to 46,000 mg/£ for the
case using cooling tower blowdown as makeup.
The chloride content of the scrubbing liquor increased
from 120 mg/5, for the base case to 1,560 mg/£ for the case where
cooling tower blowdown is used. The corrosive properties of this
liquor will depend on the materials of construction for the scrub-
ber. The feasibility of use of cooling tower blowdown depends
therefore in part on the chloride content of the initial makeup
water to the plant. Higher chloride levels may cause corrosion
problems in some scrubbing systems.
-66-
-------
TABLE 4-9. EFFECTS OF MAKEUP WATER QUALITY ON COLSTRIP
SCRUBBING SYSTEM SCALING POTENTIAL
Makeup Water Source
Makeup Water Composition, mg/S,
Calcium
Magnesium
Sodium
Chloride
Carbonate (as CO 3)
Sulfate (as SO^)
Nitrate (as NO^)
Scrubber Recycle Slurry
Suspended Solids, wt . %
pH
Liquor Composition, mg/&
Calcium
Magnesium
Sodium
Chloride
Carbonate (as C03)
Sulfate (as SO^)
Sulfite (as SOj)
Nitrate (as NOl)
IDS
CaSOit-2H20 Relative Saturation**
Solid Composition, wt. %
CaSCVZHzO
Inert (Ash)
CaS03-%H20
Case A
Softened
River Water
39.9
10.7
40.3
17.0
6.0
188.
1.4
12.6
5.0
730.
5,290.
440.
120.
150.
21,000.
3,560.
10.0
31,000.
1.41
38.9
57.3
3.8
Case B
Raw
River Water
57.9
10.7
40.3
48.7
17.3
188.
1.7
12.6
5.0
740.
5,280.
440.
330.
150.
20,700.
3,550.
12.0
31,000.
1.41
38.9
57.2
3.9
Case C
Cooling Tower*
Slowdown
534.
143.
540. -
227.
6.5
2,640.
18.7
12.6
5.0
690.
6,190.
3,870.
1,560.
150.
29,400.
3,700.
130.
46,000.
1.42
39.9
56.3
3.8
*Use of this water as demister wash may not be feasible depending on the
amount of S02 sorbed in the demister.
**The critical value, above which scale potential exists, is 1.3-1.4 for
-67-
-------
Since scrubbing systems generally use makeup water as
demister wash, the use of cooling tower blowdown may be limited to
other makeup requirements such as pump seal water. The feasibil-
ity of using cooling tower blowdown as demister wash will depend
on the amount of S02 sorbed in the demisters. Since cooling tower
blowdown may be nearly saturated with respect to CaSOit-2H20 when
maximum cycles of concentration is achieved in the tower system,
only a small amount of SO2 sorbed in the demisters may cause
scaling problems. Colstrip pilot studies indicated that cooling
tower blowdown could not be used as demister wash in that system.
One possible solution to this problem is to dilute the
cooling tower blowdown with a subsaturated stream such as river
water so that when S02 is absorbed, the critical relative satura-
tion of CaSOi»'2H20 will not be exceeded in the demisters. Pilot
studies to determine the required dilution under various operating
conditions are needed before a system change to an existing scrub-
bine process could be made.
4.3 Recycle/Reuse Alternatives in S02/Particulate Scrubbing
Systems
Makeup water is required in S02/particulate scrubbing
systems to replace water lost through evaporation and through
occlusion with the waste solids. Evaporative losses include evap-
oration in the scrubber, from process vessels, and in the pond
system if the pond overflow is recycled to the scrubbers. These
losses are fixed by the flue gas flow, temperature, and water con-
tent and by the exposed area of process vessels and the pond system.
The amount of water lost through solids occlusion depends
on the amount of solids produced and the weight fraction solids of
the waste. Scrubbing systems may be designed to be closed loop,
i.e., no aqueous discharges. The Four Corners study (Appendix F)
showed that recycle of the ash pond overflow liquor had no appre-
ciable effect on the recirculating slurry scaling potential. In
this case, where a pond is used for disposal, the final sludge
composition will be approximately 50% solids (1 kg of water lost
per kg of solids).
The amount of solids to be disposed of will depend on
the ash removal rate and the S02 removal rate. The relative amount
of ash and S02 will determine which removal rate has the most ef-
fect on the amount of solids produced. At Four Corners, where
about 30 kg of ash are removed per kg of S02, the S02 removal rate
has very little effect on the scrubber makeup requirements.
-68-
-------
Since scrubbers can be operated in a closed-loop fash-
ion, there is little opportunity for recycle/reuse but they can
be used as a sink for water streams which might normally be dis-
charged from a power plant such as cooling tower blowdown. The
use of various qualities of makeup water did not affect the
scaling potential of the scrubber recycle liquor in the Colstrip
scrubbing simulations (Appendix J). This indicates that the qual-
ity of makeup water not used as demister wash (for instance, pump
seal water) does not have a major impact on scrubber scaling ten-
dency. The composition of this water will, however, affect the
chloride content of the recirculating liquor. Corrosion problems
could possibly occur if the chloride content of the makeup water
is too high.
The use of water which is saturated or near-saturated
with respect to gypsum (such as cooling tower blowdown) as demis-
ter wash is not recommended since the addition of small amounts
of S02 in the demister may cause scale formation. Combining
cooling tower blowdown with fresh makeup water and using this mix
as demister wash may be feasible in some situations. A careful
process analysis and pilot studies are suggested for determining
the quality of water required for demister wash in a particular
scrubbing system.
-69-
-------
5.0 COMBINED SYSTEMS
The object of this study was to evaluate methods to
minimize the aqueous discharges and makeup requirements_for
coal-fired steam-electric generating stations. In Sections 2.0,
3.0, and 4.0, the operating conditions of the major water con-
sumers at power plants were discussed. In these sections,
methods to reduce the water consumption of the individual pro-
cesses were presented. It is important to use these methods in
an efficient combination to produce an overall system which will
reduce the discharge requirements of the entire plant at a
reasonable cost. This section discusses alternative operating
conditions and the associated costs for the water systems at
the Arizona Public Service Four Corners Station, the Public Ser-
vice of Colorado Comanche Station, the Georgia Power Co. Plant
Bowen, the Pennsylvania Power and Light Montour Station, and
the Montana Power Co. Colstrip Plant.
Detailed discussions of the water systems at these
plants as well as the recycle/reuse alternatives studied for
each are presented in Appendices F through J. It should be
noted here that this analysis was performed to study general
water recycle/reuse alternatives. Actual implementation of
any of these alternatives would require a more extensive inves-
tigation of process variability. More water quality data would
be required along with additional studies to fully characterize
the ash.
5. 1 APS Four Corners
The major water consumer other than the cooling pond
(Morgan Lake) at Four Corners is the particulate/S02 scrubbing
system. Cooling is accomplished by recirculating water from
Morgan Lake. Bottom ash is sluiced in a recirculating fashion
with Morgan Lake providing the sluice water. All of the recy-
cle/reuse alternatives for Four Corners concern themselves ex-
clusively with the scrubbing system since scaling problems have
been encountered and an effective water management scheme neces-
sitates scale free operation of the scrubbing system. Since
Morgan Lake is already part of two recirculating systems (bottom
ash, cooling) no additional recycle/reuse alternatives were in-
vestigated. To do so would involve considering the effects on
heat dissipation from the pond and is beyond the scope of this
project.
-70-
-------
Four alternatives were investigated for the parti-
culate scrubbing system at Four Corners. Table 5-1 presents
a summary of these four alternatives compared to existing op-
erations .
The results of the first alternative simulation indi-
cates that the present system tankage capacity is not sufficient
to allow ample gypsum precipitation to prevent scaling. These
results are based on the assumption that 98.6% of the sorbed S02
is oxidized to sulfate in the scrubbers. In cases where the
oxidation is less than 98.6%, less tankage capacity will be re-
quired to prevent scale. Since gypsum precipitation is the con-
trolling factor at these levels of oxidation, lower oxidation
rates will lower the amount of gypsum to be precipitated and
therefore require a smaller reaction tank. In all four alter-
natives oxidation was assumed to remain at the level measured
at the plant. Studies conducted at Four Corners after this
work was completed showed that addition of lime to the venturi
recirculation tank lowered the sulfite oxidation to a level
where the scrubbers operate in the subsaturated mode with re-
spect to gypsum. The hold tank volume required to prevent
scale in this mode is much smaller than the volumes predicted
in this study where high oxidation was assumed.
In the second alternative, a tank capacity of 37,500
cubic meters (1.33 x 106 cubic feet) was simulated. Gypsum
relative saturations were reduced to levels below the critical
level required for the on-set of scaling. Two cases were
studied with different scrubber liquid-to-gas ratios (L/G).
The existing L/G of 4.7 £/m3 (§STP (35.2 gal/1000 scf) gave a
scrubber bottoms pH of 2.9 and an L/G of 10.0 £/m3 @STP (74.8 gal/
1000 scf) gave a pH of 3.9 (assuming 50% S02 removal), indicating
that higher L/G's are desirable for corrosion control
The third alternative simulation, recycling the ash
pond overflow to the scrubbing system, indicated that the pond
overflow has no major impact on the gypsum relative saturations
in the system but reduces the water makeup requirements from
70.7 Jt/sec (1122 GPM) for Alternative 2 to about 50.8 £/sec
(807 GPM). Also, a simulation with ash pond overflow recycle
using a reaction tank volume of 21,200 nr (7.5 x 105 ft )
showed that a more reasonable reaction tank volume can be uti-
lized. This simulation showed a gypsum relative saturation of
1.19 in the scrubber effluent slurry.
-71-
-------
TABLE 5-1. SUMMARY OF RECYCLE/REUSE OPTIONS AT FOUR CORNERS1
Existing
Condition
Case 1 Case 2
Weight Percent Solids
in Thickener Bottoms 10 30
Hold Tank Volume. 0 0
m3 (ftj)
Liquid to Gas Ratio, 4.7 4.7
I/in' @ STP (gal/scf) (35.2) (35.2)
% Recycle from the
Ash Pond 0 0
S02 Removal, % 30 30
Oxidation, 7. 98.6 98.6
Paniculate Removal
prior to scrubber, % None None
Scrubber Makeup Rate, 223 70.7
e/sec (CPU) (3540) (1730)
Costs :
Capital, 1976 $
Operating, 1976 $ 2
(mils/kWh)
Alternative
Two
Case 1
30
37,500
(1.33 x 10') (1.
4.7
(35.2)
0
50
98.6
None
70.7
(1120)
3,334,000 4
628,000 1
(.128)
Alternative
Three
Case 2
30
37,500
33 x 10')
10.0
(74.8)
0
50
98.6
None
70.7
(1120)
,275.000
,101,000
(.225)
Case 1
30
37,500
(1.33 x 106)
10.0
(74.8)
28
50
98.6
None
50.8
(805)
4,328,000
1,109,000
(.226)
Case 2
30
21,200
(0.75 x 106)
10.0
(74.8)
28
50
98.6
None
50.8
(805)
3,317.000
958,000
(.195)
Allurnat ive
Four
Case 1
30
8900
(0.31 x 10b)
10.0
(74.8)
0
50
98.6
60
41.0
(650)
3,385,000
968.000
(.198)
'Rough cost estimates were made to compare technically feasible options and do not include a "difficulty to retrofit" factor.
2Includes capital cost amortization of 15% per year.
-------
The fourth alternative shows that reaction tank vol-
ume may be decreased further by removing a portion of the fly
ash by dry methods prior to the scrubbing system. A volume of
8900 m3 (3.14 x 105 ft3) was used to obtain a gypsum relative
saturation of 1.19 in the scrubber effluent (60% of fly ash re-
moved prior to scrubber) . Water makeup requirements were also
reduced to 41.0 2,/sec (650 GPM) .
The rough cost estimates of the technically feasible
options (Alternatives 2-4) indicate that three to four million
dollars would be required to upgrade the particulate scrubbing
system so that scale potential is eliminated and water require-
ments reduced. The least expensive alternative was Alternative
3, case two (recycle ash pond overflow, reduced reaction tank
volume), followed closely by the fourth option. Alternative 2,
cases one and two, and Alternative 3, case one showed similar
installed costs. Energy consumption did not vary radically
among alternatives, although Alternative 2, case one (increased
tank volume, low L/G) indicated a lower energy requirement was
necessary and therefore less operating costs. The operating
costs shown in Table 5-1 include capital amortization at 15%
per year.
5.2 PSC Comanche
The major water consumers at Comanche are the cooling
towers and the bottom ash handling systems. There is no S02
scrubbing and the fly ash is presently trucked off site. As
part of the study of the water system at Comanche wet fly ash
sluicing was also considered.
Table 5-2 presents a summary of the three alternatives
which were examined for Comanche. The first one involved using
cooling system blowdown from the towers operating at five cycles
of concentration to sluice fly ash and bottom ash on a once-
through basis. The effects of C02 mass transfer in the ash pond
and the sluice tank were examined for this system. No gypsum^
scale potential was identified in any of the once-through sluic-
ing cases, but potential scaling of CaC03 and Mg(OH)2 was pre-
sent. Although gypsum relative saturation was less than the
critical value (1.3-1.4) variations in ash or makeup water qual-
ity may cause gypsum scale. The calculated relative saturation
for gypsum was 1.24. This alternative will result in an ash
pond overflow of about 32.7 I/sec (518 GPM) for each unit as
compared to the existing configuration pond overflow rate of
about 78 £/sec (1230 GPM) per unit.
-73-
-------
TABLE 5-2. SUMMARY OF WATER RECYCLE/REUSE OPTIONS AT COMANCHE1
I
-^J
-F-
Cooling Tower Makeup Source
Cycles of Concentration in
Cooling Towers
Cooling System Treatment
Fly Ash Disposal Method
Type. % solids
Bottom Ash Disposal Method
Type, % solids
Recycle in Fly Ash
System, %
Recycle in Bottom Ash
System, %
Treatment in Ash Systems
Plant Makeup Requirements
I/sec (GPM)
Plant Discharge
I/sec (GPM)
Costs
Capital Investment, 1976 $
Operating Expenditures, 1976 §/yr 3
(mils/kW-hr)
Additional Cost to Treat Pond
Overflow for Zero Discharge
Capital , 1976 $
Operating, 1976 $/yr 3
(mils/kW-hr)
Total Cost for Zero Discharge
Capital, 1976 $
Operating, 1976 $/yr 3
(mils/kU-hr)
Existing Alternative
Conditions One
Softened River Water Softened River Water
5.0 5.0
Alternative
Two
Softened River Water
7.6
(Sulfuric acid and zinc polyphosphate used for all com
Dry Wet, 10%
Wet, 1% Wet, 47. a
0
0 0
None None
590 (9350) 520 (8250)
156 (2470) 65.4 (1040)
342.000
90.000
(0.02)
8,280,000
2,136,000
(0.43)
8,622,000
2,226,000
(0.45)
Wet, 10%
Wet, 4% *
10%
100%
Brine Concentration
of Makeup (50%)
455 (7210)
28.8 (460)
3,662,000
863,000
(0.18)
3,706,000
944,000
(0.19)
7,368,000
1,807,000
(0.37)
Alternative
Three
Softened River Watur
8.4
litions)
Dry
Wet. 11
100%
None
450 (7120)
30.2 (480)
222.000
38,000
(0. 008)
3,883,000
989,000
(0.20)
4.105,000
1,027,000
(0.21)
'Rough cost estimates were made to compare technically feasible options and do not include a "difficulty to retrofit" factor.
2Sluicing bottom ash at 4% solids may not be feasible due to hydraulic limitations of equipment.
'Includes capital cost amortization at 15% per year.
-------
The second alternative involves using cooling system
blowdown at 7.6 cycles of concentration as makeup to a recircu-
lating ash sluice system. The makeup water was used to sluice
fly ash, and recycled ash pond water was used to sluice bottom
ash. Only 1070 of the fly ash sluice water was recycled from
the pond. Gypsum relative saturations in the fly ash sluice
line were calculated to be 1.54 - 1.74 depending on the level
of C0? transfer in the pond. This range exceeds the critical
relative saturation range for scaling of CaS04'2H20 of 1.3 -
1.4. Therefore, some form of treatment would be required such
as brine concentration of a portion of the tower blowdown.
Lime treatment of the blowdown for calcium removal was found to
be insufficient for scale prevention due to the sulfate concen-
trations in the system. Desupersaturation of gypsum in the ash
pond will also not prevent scaling since only a small portion
of the ash pond liquor is recycled to the fly ash system. This
alternative will produce an ash pond overflow of about 14.4
2,/sec (230 GPM) for each unit.
The third alternative is to continue to dispose of fly
ash in a dry form and sluice the bottom ash on a recirculating
basis at 1% solids using cooling tower blowdown and pond re-
cycle with the towers operating at 8.4 cycles of concentration.
This will provide 16.0 £/sec (260 GPM) of cooling tower blow-
down per unit and will not alter the boiler refractory cooling
requirements. For this alternative about 15.1 2,/sec (240 GPM)
of ash pond overflow per unit is obtained. This water may be
discharged or recycled to the boiler and cooling tower makeup
systems after appropriate treatment.
Rough cost estimates were make for the once-through
sluice system and the recirculating system (Alternatives 1 and
2) using cooling tower blowdown to sluice fly ash with 507o of
the blowdown treated by brine concentration. Operating the
cooling system at 5 cycles of concentration and sluicing the
fly ash and bottom ash on a once-through basis is the less ex-
pensive alternative ($342,000 for capital cost and about
$90,000/yr operating cost, including capital amortization at
157o per year) . The third alternative is the least expensive
of the three. Capital costs are about $222,000 and operating
costs are about $38,000/yr, including capital amortization at
15% per year.
In order to reduce the ash pond overflow to 14.4
£/sec (229 GPM) for each unit by operating the cooling systems
at 7.6 cycles of concentration with the cooling system blowdown
as sluicing makeup, the entire plant ash sluice system will
-75-
-------
require an initial capital cost of about $3.7 million and an
operating cost of about $863,000/yr, including capital amorti-
zation at 1570 per year. These costs do not include the possi-
ble necessity of silica removal.
If zero discharge of ash pond overflow is desired,
the once-through system becomes more expensive due to the
greater amount of ash pond overflow to be treated. A soften-
ing/reverse osmosis/brine concentration system to eliminate
ash pond overflow would require an additional operating cost
of approximately $2,136,000/yr. The total overall costs would
be about $8,622,000 for capital costs and $2,226,000/yr for
operating costs (including capital cost amortization).
The additional costs for obtaining zero discharge
with the recirculating system would be about $3.7 million for
capital costs and $944,000/yr for operating costs, giving total
overall costs of about $7.4 million for capital costs and $1.8
million/yr for operating costs including capital amortization
at 15% per year.
The costs associated with achieving zero discharge
with dry fly ash disposal (third alternative) are about $4.1
million for capital costs and $l,027,000/yr for operating costs.
Since fly ash is currently disposed of in a dry fashion at
Comanche, no additional costs for the fly ash system are needed.
This makes the third alternative the least expensive for
achieving zero discharge. These costs include brine concen-
tration, additional piping, additional pumping costs, and capi-
tal cost amortization at 15% per year.
5.3 GPC Bowen
The major water consumers at Bowen are the cooling
towers and the ash handling systems. Bowen does not employ any
S02 or particulate scrubbing.
Table 5-3 presents a summary of the technically fea-
sible operations and the relative costs of each of these alter-
natives. Two process alternatives were studied for the ash
sluicing system at Bowen. The first case involved using cool-
ing tower Slowdown from the towers operating at 5.7 cycles of
concentration to sluice both bottom and fly ash on a once-
through basis at about 10 wt. % solids (Alternative 1 in Table
-76-
-------
TABLE 5-3. SUMMARY OF TECHNICALLY FEASIBLE OPTIONS AT BOWEN1
Cooling Tower Makeup Source
Cycles of Concentration In
Towers
Cooling System Treatment
Acid Addition Kate, kg/day2
(Ib/day)
Ash Sluice Makeup Source
1 Z Recycle in Fly Ash
~^j System
"^ % Recycle in Bottom Ash
System
Ash System Treatment
Plant Makeup Requirements,
e/sec (CPM)
Plant Discharge Rate,
e/sec (CPM)
Costs
Capital, 1976 $
Operating, 1976 $/yr J
(mtls/kw-hr)
Existing Condition
Makeup Pond ,
Service Water
1.7
None
0 (0)
Cooling Tower Slowdown
0
0
None
3250 (51,500)
1600 (25,000)
--
Alternative One
Makeup Pond ,
Service Water
5.7
U2SOi,
481 (1060)
Cooling Tower Slowdown
0
0
None
1880 (29,800)
255 (4050)
100.000
52,900
(.002)
Alternative Two
Makeup Pond ,
Service Water
15.
HjSOi,
608 (1340)
Cooling Tower Slowdown
60
100
Recycle Softening
1670 (26,400)
41 (650)
1,223,000
402,000
(.018)
Alternative Three
Makeup Pond, Service Water,
Brine Concentrator Uistlllatt
15.
H2SO,
608 (1340)
Cooling Tower Slowdown
60
100
Recycle Softening, Brine
Concentration of Pond
Overflow
1630 (25, BOO)
0 (0)
6,380,000
1,735,000
(.078)
1 Kouyh cost estimates were made to compare technically feasible options and Uo not Include a "difficulty to retrofit
2 As KHIX II2SO,,.
Includes capital cost amortization at 15% per year.
-------
5-3). The effects of C02 mass transfer in the ash pond and sluice
tank on the system operation were investigated. No gypsum scale
potential was identified in any of the cases with once-through ash
sluicing.
The second alternative for the ash sluicing system
involved using cooling tower blowdown from the towers operating at
15.0 cycles of concentration as makeup water to a recirculating
ash sluice system (Alternative 2 in Table 5-3). A blowdown of
41 Jl/sec (650 GPM) is taken from the ash pond for this alternative.
If the pond recycle water remains supersaturated with respect to
gypsum, scaling will occur in this system. However, this situation
may be remedied by chemical treatment. Sodium carbonate softening
of approximately 80% of the pond recycle water will maintain a
gypsum relative saturation of about 1.0 in the slurry line and
prevent calcium sulfate scaling. The calcium carbonate sludge
produced in the softening step may be disposed of in the ash pond.
Zero discharge from the cooling and ash sluicing sys-
tems (Alternative 3 in Table 5-3) may be achieved by installing
a softening/reverse osmosis/brine concentration unit to treat
the above ash pond overflow (41 2,/sec or 650 GPM) and recycling
approximately 50% of the clean water as boiler makeup and the
remainder as cooling tower makeup.
Potential scaling of CaS03 is present in all cases
studied, both once-through and recirculating. However, the fly
ash slurry line possibly can be kept free of plugging by the
addition of a fly ash slurry reaction tank and by frequent
flushing with a water stream of pH 6-7. Pilot or bench-scale
testing is recommended to determine accurately the size of
reaction tank and frequency and quantity of acid washing re-
quired or if other measures are necessary.
The rough cost estimates presented for the alterna-
tives in Table 5-3 indicate that reducing the ash pond over-
flow to 225 £/sec (4050 GPM) by running the cooling towers at
5.7 cycles of concentration and sluicing the ash on a once-
through basis using cooling tower blowdown is the less expen-
sive option (about $100,000 capital cost with about $53,000/yr
operating costs) . This option necessitates acid treatment in
the towers. The operating expenses include capital amortiza-
tion at 15% per year.
Reducing the ash pond overflow to about 41 s,/sec (650
GPM) by operating the cooling towers at 15.0 cycles of concen-
-78-
-------
tration (with acid treatment) and using the tower blowdown as
makeup to a recalculating ash sluice system (with Na2C03 soft
ening of 80% of the pond recycle) has an initial capital cost
of about $1,223,000 and operating costs including capital cost
amortization (15% per year) of about $402,000/yr. The use of
a softening/reverse osmosis/brine concentrator unit to elimi-
nate the ash pond overflow discharge (recycle to boiler and
cooling tower makeup) for this alternative would require a capi-
tal investment of about $6.38 million total. The additional
operating costs would be about $1,222,000/yr, giving a total of
approximately $1,735,000/yr.
5.4 PP&L Montour
The water system at Montour is similar to the one at
Bowen. Once-through wet sluicing of both fly and bottom ash is
employed and there is no equipment to handle S02. The cooling
towers at Bowen and at Montour are natural draft.
Table 5-4 presents a summary of the technically feasi-
ble options for the Montour water system as compared to existing
operations and the relative costs of each of these alternatives.
Four process alternatives were studied for the water systems at
Montour. All alternatives sluiced bottom ash and fly ash at 5
wt. % solids. Mill rejects were sluiced at 0.5% solids. In all
cases ash pond liquor was recycled to the ash sluicing opera-
tion. In one case, Alternative 4, a blowdown was taken from the
system to prevent CaSO^HaO scale. The other three alterna-
tives did not discharge any liquid streams and controlled
CaSOIt-2H20 scale with softening of a portion of the pond re-
cycle water.
The first alternative assumes that the cooling tower
drift from one tower was equal to the design value of 32 £/sec
(500 GPM). Enough blowdown was drawn from the cooling towers
to serve as makeup to the re circulating sluicing operation.
Under this situation the cooling tower should be operating near
8 cycles of concentration.
The second alternative is identical to Alternative_1
except that the cooling tower drift was assumed to be negligi-
ble. This assumption increased the cycles of concentration
from 8 to 20 even though the blowdown rate was not changed.
This had the effect of requiring more softening for the pond
recycle stream because of the poorer quality of makeup water to
the ash sluicing operations.
-79-
-------
TABLE 5-4. SUMMARY OF TECHNICALLY FEASIBLE OPTIONS AT MONTOUR1
Cycles of Concentration
In Cooling Towers
Existing Condition
Alternative 1
1.5 - 2.0
Alternative 2
20
Alternative 3
20
Alternative 4
20
Assumed Drift Rate
in Cooling Tuwers
ll/sec (Cl'M)
62 (1,000)
62 (1,000)
(650)
40 (650)
Ulowdown from Coaling
Towers
4/sec (Ul'M)
725 (11,500)
48 (760)
40 (650)
Z Kecycle in Fly Asli
CO
o
1
Sluicing System 0
Sluice System Makeup Cooling Tower
Source Slowdown
Total Makeup Water Rate,
I/ace (CPU) 1,500 (24,000)
Ultimate Effluent Kate,
£/sec (CPM) 500 (7,900)
Treatment Required None
Costs
Capital, 1976 $
Operating, 1976 $/yr k
(mils/kW-hr)
89.
Cooling Tower
Slowdown
1,000 (16,000)
0
IliSU., (Cooling Tower) '
Ma2COs (Pond Recycle) a
640,000
173,000
(0.016)
89.
Cooling Tower
Blowdown
950 (15,000)
0
HjSOi, (Cooling Tower)2
HaH'Oa (Pond Recycle) s
668,000
187,000
(0.018)
89. 73.
River Water River Water
985 (15,600) 1,035 (16,400)
0 50 (800)
II 2 SO,, (Cooling Tower)2 II2S(\ (Cooling Tower) 2
NazCOj (Pond Recycle)3
622,000 485,000
169,000 103,000
(0.016) (0.010)
'liuugli (;ost uslJniaLes were made to compare technically feasible options and do not include a "difficulty to retrofit" factor.
2Sulfuric acid treatment for CaCOj scale control.
3Na2CO3 softening for Ca removal.
4 Includes capital cosl amortization at 15% per year.
-------
The other two alternatives assume that the cooling tow-
ers can be operated at zero blowdown. This requires that the drift
be at least 65.4 of the design value. Under these alternatives the
makeup water to the ash sluicing operation is obtained directly
from the Susquehanna River or the plant makeup pond. Alternative
3 employed softening and attained zero discharge similar to the
two previous alternatives. Alternative 4 controlled the CaS(K-2H20
scaling potential by the use of a blowdown stream of about 50 Jl/sec
(800 GPM) from the ash pond (both units).
Potential scaling of CaC03 is present in all four cases.
However, the fly ash slurry line possibly can be kept free of
plugging by the addition of a fly ash slurry reaction tank and/or
by flushing with acidic water. Pilot or bench scale testing is
recommended to determine accurately the size of reaction tank and
frequency and quantity of acid washing required or if other mea-
sures are necessary.
Rough cost estimates for the different alternatives are
also presented in Table 5-4. Alternative 4 is the least expensive
due to the fact that no softening was required. The other three
vary mostly in the degree of softening that was required for the
recirculating ash sluicing system. It should be emphasized that
Alternatives 1 and 2 differ only in the assumption concerning
the drift rate in the cooling towers. If more information could
be obtained about the actual drift rate, a more reliable cost
estimate could be made.
5.5 MPC CoIstrip
The major water systems at the 350 Mw/unit Colstrip
plant are the cooling tower and combined SOz/particualte scrub-^
bing systems. Colstrip is designed for and is achieving zero dis-
charge through brine concentration of the cooling tower blowdown
and a disposal pond for the scrubber sludge.
Table 5-5 presents a summary of the two combined system
alternatives for the Colstrip water system as compared to existing
operating and the relative costs of each alternative. All_flows
reported in Table 5-5 refer to those produced from both units.
The first alternative does not involve any changes in
operation of the cooling towers but uses cooling tower blowdown
and untreated river water as scrubber makeup as opposed to soft-
ened river water and brine concentrator distillate as is pre-
sently done. A capital cost of $159,000 is reported for piping
-81-
-------
TABLE 5-5. SUMMARY OF WATER RECYCLE/REUSE OPTIONS AT COLSTRIP1
i
oo
Cooling Tower Makeup
Source
Cycles of Concentration
in Cooling Towers
Cooling System Treatment
Treatment Rate,
«7sec (GPM)
Cooling Tower Slowdown
Rate, ll sec
Scrubber Makeup Source
Plant Makeup Rate
fc/sec (GPM)
Plant Discharge Rate
1/se.c (GPM)
Costs: l
Capital, 1976 $
Operating, 1976 $/yr. 2
(mils/kwh)
Existing
Conditions
Alternative
One
Alternative
Two
Softened River Water
13.5
Makeup Softening
423 (6710)
23.6 (376)
Softened River Water,
Brine Concentrator
Distillate
423 (6710)
0.
Softened River Water
13.5
Makeup Softening
397 (6300)
23.6 (376)
Untreated River Water
20
Slip-stream softening
18 (284)
14.6 (230)
Cooling Tower Blowdown, Cooling Tower Blowdown,
Untreated River Water Untreated River Water
423 (6710)
0.
159,000
-237,000
(-.046)
423 (6710)
0.
275,000
-217,000
(-.044)
]R@ugh cost estimates were made to compare technically feasible options and do not include
a, "difficulty to retrofit" factor.
2Includes capital cost cimortization at 15% per year.
-------
modifications and new pumps. However, a net operating savings
is shown due to a large savings in brine concentrator operation
because of the reduced flow. Only enough cooling tower blow-
down is sent to the brine concentrator to provide the boiler
makeup requirements (only one brine concentrator needed) .
Alternative 2 includes using slipstream treatment in
the cooling tower system in addition to the system changes of
Alternative 1. The towers are operated at 20 cycles of concen-
tration resulting in decreased blowdown. Again only enough
cooling tower blowdown to provide the boiler makeup is sent to
the brine concentrator. A higher capital cost is reported due
to the conversion to slipstream treatment in the cooling sys-
tem. The increased capital charges result in a lower operating
expense savings for this alternative. The savings in brine con-
centrator operating costs represents the major savings of both
of these alternatives.
Although the Colstrip plant is achieving zero dis-
charge, more effective cascading of the water streams in the
plant may be achieved which results in a decrease in operating
costs from the existing level. The capital and operating costs
reported in Table 5-5 do not include any savings which could
have been realized if the Colstrip water system had been de-
signed for the most effective cascading of aqueous streams. A
savings in capital investment could have been achieved by de-
signing the cooling towers for slipstream treatment and by
using only one 150 GPM capacity brine concentrator as opposed
to the two 200 GPM capacity units presently used. It should be
noted that MFC considered sidestream softening but did not im-
plement it since it was not believed to be reliable technology.
The capital savings associated with purchasing one 150 GPM
brine concentrator versus two 200 GPM units totals about $1.9
million based on $7,750/GPM (LE-239) .
5.6 Recycle/Reuse Alternatives in Combined Systems
Minimization of makeup requirements and aqueous dis-
charges from coal-fired steam-electric generating stations re-
quires efficient cascading of the aqueous streams within the
whole plant. The best arrangement for any plant is the water
use scheme which attains the required environmental standards
at the least possible cost. The optimum balance of these con-
flicting requirements is not always obvious. Therefore, several
alternatives were presented for each power plant. Although
-83-
-------
conditions varied from plant to plant the alternative water sys-
tems presented were similar in their overall approach to water
utilization.
The cooling towers require the largest amount of water
at most plants. They also need water of reasonably high quality
to insure that scale will not form in the condenser. The makeup
requirements to the cooling towers are usually supplied directly
from the water source for the plant, the highest quality large
supply of water. The towers concentrate the dissolved solids in
this water because of the evaporation that occurs in the tower.
A blowdown from the tower is usually required to control the
level of dissolved solids and prevent scale formation. The size
of the blowdown stream can be reduced if acid treatment and/or
softening are used to inhibit scale formation. The size and
quality of this blowdown stream makes it a good candidate for
makeup to other plant water consumers such as ash sluicing or
scrubbing operations.
All of the alternatives studied for the four power
plants employing cooling towers have the cooling towers serving
as the major recipient of the plant makeup water. Smaller uses
of the plant makeup include feed to the demineralizers, for
boiler makeup, and general service water. The remaining plant
water needs are supplied by the cooling tower blowdown.
Two of the plants studied, Bowen and Montour, employ
wet fly and bottom ash sluicing. All of the makeup to these
systems is supplied from the cooling towers. The rate of
cooling tower blowdown at Montour is determined by the water
requirement for the once-through sluicing system. Excess blow-
down is discharged at Bowen. In order to meet the once-through
ash sluicing water demand, the blowdown from the towers is much
larger than necessary to control condenser scale. The rate at
which water is delivered to the ash ponds in once-through sys-
tems is greater than the rate at which it is lost from the ponds
through evaporation and occlusion with the pond sludge. This
excess water from once-through sluicing usually is discharged.
Recirculating ash sluicing systems were investigated
as a potential recycle/reuse scheme. The systems were designed
to use much less cooling tower blowdown by sluicing the ash
with a mixture of pond water and cooling tower blowdown. In
some of the alternatives studied for Montour, the size of the
cooling tower blowdown was equal to the losses that occur in
-84-
-------
the pond, thereby attaining zero discharge. At the increased
cycles of concentration realized because of the reduced cooline
tower blowdown, acid treatment was needed to prevent CaC03 scale
in the condensers . Softening of the pond recycle stream was
needed to control CaS0lt-2H20 scale in the ash slurry line. At
Bowen softening of 80% of the pond recycle stream was required
to control gypsum scale in a re circulating ash system with 60%
of the fly ash sluice water recycled. To attain zero discharge
with this system would require an expensive treatment step,
such as brine concentration, for the pond overflow. The quality
of the cooling tower blowdown and the reactivity of the fly ash
at Comanche was such that a recirculating fly ash sluicing sys-
tem would probably scale the slurry lines at 10% recycle without
softening. Softening of the recycle stream at 10% recycle
would not be effective since the recycle represents only a small
portion of the total sluice water. It was recommended that
Comanche continue to dispose of their fly ash by dry methods.
The design of the Colstrip water system and alterna-
tives suggested in this study are similar to the other proposed
systems. The cooling tower makeup is the major portion of the
water entering the plant and the blowdown from the towers serves
as makeup to other water consumers. Some of the water is lost
in the S02 scrubbers via evaporation and occlusion with the
scrubber sludge and some of the water is lost in the recircula-
ting bottom ash system.
The present operation of the Colstrip plant is at zero
discharge. The alternatives proposed for the water system sug-
gest ways to attain zero discharge at reduced cost. The major
savings occurs because the alternatives do not use as much brine
concentration as is presently used. Only enough water to supply
boiler makeup is sent to the brine concentrators.
In summary, the three major water consumers with in-
creased recycle/reuse opportunities at coal-fired power plants
are cooling towers, ash sluicing, and S02/particulate scrubbing.
Cooling ponds may also be considered as major water consumers
but the study of thermal dispersion in ponds was beyond the
scope of this project. An investigation of increased recycle
in cooling ponds would require that the effects on the pond heat
dissipation be considered. The cooling towers demand the high-
est quality water and produce a fairly large blowdown stream.
A desirable method to achieve zero discharge is to limit the
cooling tower blowdown to a level no greater than the total
makeup requirements for the other water consumers. The ash
-85-
-------
sluicing and S02/particulate scrubbing systems serve as good
water sinks because of the water losses occurring through evap-
oration and occlusion with sludges in these systems. Cooling
tower blowdown can be used to sluice ash either in a once-
through or recirculating system or as makeup water to a scrub-
bing system (excluding demister wash).
-86-
-------
GENERALIZED IMPLEMENTATION PLANS
1.0 INTRODUCTION
Limited availability and rising costs of water along
with reduced discharge requirements have placed an ever increas-
ing importance on water recycle/reuse at coal-fired power plants.
The first part of this project involved investigating water recycle/
reuse alternatives of five typical power plants. This section
of the final report presents generalized implementation plans
for the types of recycle/reuse possibilities identified in the
plant studies.
Three types of major water systems were identified as
candidates for increased recycle/reuse options at coal-fired
power plants. These are cooling tower, ash sluicing, and S02/
particulate scrubbing systems. At each of the four plants studied
with cooling towers, increased recycle was possible in the cooling
tower system through the treatment options of sulfuric acid addi-
tion for pH control and/or softening for control of gypsum scale
potential. For each of the two plants studied with once-through
wet fly ash sluicing, a closed-loop recirculating system may be
used with softening to control gypsum scale potential. For scrub-
bing systems, the recycle/reuse options identified include:
1) using a normally discharged stream such as cooling tower blow-
down as scrubber makeup in a cascaded water system, and 2) con-
verting from open-loop to closed-loop operation.
The generalized implementation plans for each of these
types of recycle/reuse options may be divided into four phases:
Phase I - System Characterization
Phase II - Alternative Evaluation
Phase III - Pilot-scale Studies
Phase IV - Full-scale Operation
The first phase, system characterization, is necessary to estab-
lish a data base for determining existing operating characteris-
tics and identifying the types of recycle/reuse options possible.
This includes collecting all available design and operating "J"*-
mation and a sampling program to supplement existing data. The
data base resulting from this phase of the implementation plan
can be used to evaluate recycle/reuse options for both extreme
-87-
-------
and average operating conditions in the second phase. The most
important operating variables and their effects on system opera-
tion are presented in the characterization phase descriptions.
The plant studies performed in this project used a
computer model to evaluate recycle/reuse opportunities. The
results of the plant studies showed that computer models are
effective in identifying water recycle/reuse options in cooling
tower, ash sluicing, and S02/particulate scrubbing systems. The
second phase of an implementation plan involves an analysis of
alternatives to evaluate the feasibility of those alternatives
under various operating conditions. Computer simulation results
can be used to design necessary treatment step to prevent scale
formation, and provide a basis for designing any pilot-scale
equipment required for testing particular options. The evalua-
tion discussions in this document include evaluation criteria
and a methodology for using a computer model to perform the
evaluations.
Pilot-scale studies are listed as the third phase of
a generalized implementation plant. These studies may be required
to determine parameters which may not be easily measured or cal-
culated from the characterization or simulation phases. For ash
sluicing systems, these parameters include: 1) ash reactivity,
2) the effectiveness of including an ash reaction tank prior to
the sluice line to reduce scale potential, and 3) the amount of
C02 transfer occurring in the pond. For scrubbing operations,
pilot-scale studies may be required to determine the operability
of a revised demister wash operation. Pilot studies are not
required for cooling towers since they exist as recirculating
systems and may be changed gradually. Care should be taken when
implementing a cooling tower system modification. A sampling
program to monitor cooling tower operation will minimize the
risk of scaling during implementation.
The final phase of an implementation plan is to make
full-scale modifications. The pilot-scale test results may be
used to design the full-scale modifications. In the case of
cooling towers, the full-scale modifications may be based on the
results of the evaluation phase of the implementation plan. The
following sections discuss each of the phases of an implementa-
tion plan as they apply to cooling tower, ash sluicing, and S02/
particulate scrubbing systems.
-88-
-------
2.0 COOLING TOWERS
Wet cooling towers are generally the largest water
consumers at coal-fired electric power plants. The towers dis-
perse the waste heat from the condensers by contacting a recir-
culating water stream with the atmosphere. Water losses occur
from the tower in three ways: evaporation, drift, and blowdown
The sum of these three streams is equal to the makeup requirements
for the tower. Evaporation accounts for the majority of the heat
lost from the tower. Drift occurs because fine droplets of water
are carried off as a mist due to the intimate contact of the air
and water. Most cooling towers have mist eliminators designed to
minimize drift. The blowdown is a purge stream which is used to
control the level of dissolved solids which will build up in the
recirculating water. Figure 2-1 shows the general flow scheme
for a cooling tower system including the tower and condenser.
The evaporation and drift rates are set by the design
of the tower, the ambient conditions, and the cooling load. The
blowdown rate is maintained at a level sufficient to prevent
scale formation in the condenser. However, as shown by the plant
studies portion of this project, in many cooling towers the size
of the blowdown is much larger than required to prevent scale
formation. Three of the four cooling tower systems investigated
could operate at increased recycle through the use of acid addi-
tion to control pH. By better defining the limits of scale for-
mation in recirculating cooling systems, the blowdown can be
reduced and thus the net water consumption and discharge from
these systems can be reduced.
This section presents a methodology for defining the
limits of safe cooling tower operation and implementing a demon-
stration plan. This plan will allow increased recycle and
decreased makeup and blowdown requirements. Only three of the
four implementation phases discussed in the introduction apply
to cooling towers. These are:
Phase I - Characterize Cooling Tower Operation
Phase II - Evaluate Alternatives
Phase III - Design and Implement Full-scale
Modifications
-89-
-------
EVAPORATION
DRIFT
MAKEUP
COOLING
TOWER
CONDENSER
SLOWDOWN
Figure 2-1. General cooling tower system flow scheme.
-90-
-------
The first phase involves collecting the data base
which is required to identify reasonable system modifications
The second phase identifies the degree of recycle that can be
attained and the amount of treatment that will be required A
detailed design of the modified cooling tower system'can be made
along with a detailed procedure for implementation of the modifi-
cations from the results of the process calculations performed
in the second phase. These calculations will allow specification
of stream flows, temperatures, compositions, and treatment equip-
ment sizes. Operation of a pilot system is not required for
cooling towers because cooling towers already exist as recircu-
lating systems which can be changed in a gradual manner. The
cost of building a pilot system would not be justified even though
there is a possibility that if the towers are operated incorrectly
the main condenser could scale. Care should be taken to safely
modify the operation of full scale cooling towers. A comprehen-
sive sampling program will allow the cooling tower operation to
be monitored so that the risk of scaling is minimized.
The plant studies conducted in this program involved
the first two phases of an implementation plan. Design and op-
erating data concerning four cooling tower systems was collected
and used to evaluate operation both under existing and alterna-
tive conditions. However, only limited operating data was avail-
able concerning these systems.
Implementing the alternatives studied at the existing
plants could require some additional studies to better define
the variations in operating parameters. These studies would in-
volve a sampling program to define tower operation and evaluation
of the system under extreme conditions. The plant studies per-
formed were based on results from one set of grab samples, exis-
ting makeup water quality data, and design information for tower
operating conditions (air flow, circulating water flow, clima-
tological data, etc.).
The following discussions are written to apply to any
cooling tower facility. All phases of an implementation plan are
presented although the characterization and evaluation phases
have been addressed in the plant studies. Since virtually all
cooling tower systems are unique with respect to operating condi-
tions, a detailed characterization and evaluation phase should be
conducted before implementing increased recycle at a particular
plant site. The purpose of the plant studies was to identify and
evaluate the types of recycle/reuse alternatives achievable at
coal-fired power plants. The purpose of this document is to out-
line a procedure for implementing the types of alternatives iden-
tified in the plant studies.
-91-
-------
2.1 Phase I: Cooling Tower Characterization
In the characterization phase the basic data which des-
cribes cooling tower operation is assembled. These data should
include information about both average and extreme tower operat-
ing conditions. This section discusses the important data for
characterizing tower operation. First, the necessary data is
discussed and then a sampling plan to supplement available data
is presented. If sufficient data is available, a sampling pro-
gram may not be required before implementing an increased
recycle/reuse option.
2.1.1 Identification of Process Variables
Table 2-1 represents a data collection sheet which may
be used to assemble cooling tower system operating conditions.
This table includes data blanks for acid treatment, softening,
and other chemical treatments for cooling tower operations. If
the treatment methods are not used, the data concerning them
will not be relevant.
The most important stream to obtain long range data
about is the makeup water. The makeup water quality determines
the quality of all of the other streams in the cooling system.
Small changes in the concentrations of dissolved solids in the
makeup water are magnified in the blowdown because of the con-
centrating effect of the cooling tower. Any long range varia-
tions in the cooling water quality can be traced to changes in
the makeup water. Seasonal variations in makeup water quality
may require variations in treatment levels required to prevent
scale formation in the condenser. In order to insure that the
tower modifications do not cause scaling, the variations in
makeup water should be identified.
The calcium, magnesium, sodium, chloride, sulfate, ni-
trate, carbonate, silica, and TDS concentrations as well as pH
are the most important in defining the makeup water quality. Cal-
cium carbonate, calcium sulfate, and silica are the primary scale
forming species in cooling towers. The magnesium, sodium, nitrate,
and chloride concentrations are important since they affect scale
formation through chemical complexing or effects on activity.
High chloride concentrations may cause corrosion problems. Phos-
phate scales may also form if the phosphate level is high. The
results of the plant studies indicated that phosphate levels in
natural bodies of water are not high enough to cause any scaling
problems. Phosphate levels should be checked to insure that
the makeup water does not have an unusually high level. If high
-92-
-------
TABLE 2-1. GENERAL DATA SHEET FOR COOLING TOWERS
Tower Parameters
A. Drift Rate
B. Approach
C. Cycles of Concentration
D. Circulating Water Temperature Change
II. Ambient Air
A. Dry Bulb Temperature °F
B. Wet Bulb Temperature _ °F
C. Flow Through Tower _ ACFM
III. Cooling System Makeup Water
A. Flow _ GPM
B. Composition
pH
total calcium as Ca^ mg/£
total magnesium as Mg"^ mg/1
total sodium as Na mg/Jl
total chloride as Cl~ mg/5,
total nitrate as NO3 mg/Jt
total carbonate as C0~^ mg/£
total sulfur as SO^ mg/£
total silica as Si02 mg/£
total dissolved solids mg/H
IV. Cooling System Slowdown
A. Flow GPM
B. Composition
pH
total calcium as Ca"1"^ mg/£
_l i
total magnesium as Mg mg/£
-93-
-------
TABLE 2-1. GENERAL DATA SHEET FOR COOLING TOWERS (Continued)
total sodium as Na
total chloride as Cl~
total nitrate as NOl
total carbonate as CO3
total sulfur as SO^
total silica as Si02
total dissolved solids
mg/A
mg/£
mg/£
mg/£
mg/A
mg/£
V.
Condenser
A. Flow Rate of Cooling Water
B. Temperature of Exit Water
C. Condensing Steam Temperature
GPM
°F
VI.
Acid Addition
A. Flow
B. Wt. % K2S
Ib/day
7
/o
VII,
VIII
Softening
A. Additive Type (Lime, Soda Ash, C02)
B. Amount of Chemical Addition
C. Capacity of Treatment Equipment
D. Size of Treated Stream
E. Size of Waste Stream
F. 70 Solids in Waste Stream
Chemical Treatment (Scaling, Corrosion
Inhibitors)
A. Type
B. Amount
Ib/hr
GPM
GPM
GPM
7=
-94-
-------
levels (>01 mg/£) are detected, analysis for phosphate should
be included with the other analyses.
Variations in both the individual species concentrations
and the total level of dissolved solids can be expected. The full
range of possibilities should be identified. This will"insure
that the treatments which are used in the cooling tower operation
will be designed for all of the water qualities which the tower
will be expected to operate with.
For characterizing cooling tower operation it is impor-
tant that climatological data be collected. The wet bulb and dry
bulb temperatures are the most important climatological data that
must be obtained for cooling tower systems. These data are the
determining factors affecting the air flow rate required to meet
the heat dissipation demand on the tower.
The wet and dry bulb temperatures affect the evapora-
tive capacity of the air flowing through the tower. Variations
in these temperatures will cause variations in the evaporation
rate and therefore will affect the makeup requirements of the
system. Increased evaporation rates will cause a higher con-
centration of dissolved solids in the circulating water if blow-
down and drift are constant. Equation 2.1 shows the relation-
ship between evaporation, drift, and blowdown rates and the level
of concentration occurring in the tower.
B + D + E
(2.1)
^ B + D
where
C = cycles of concentration
B = blowdown rate
D = drift rate
E = evaporation rate
From this equation it can be seen that changes in the evapora-
tion rate will directly affect the cycles of concentration if
the blowdown is held constant.
Expected variations in load placed on the cooling towers
should also be identified. Changes in the load should directly
-95-
-------
affect the evaporation rate in the tower because most of the
cooling which occurs is due to evaporation (^90%). The steam
condensing temperature is an important parameter in character-
izing a cooling tower system. The highest temperature in the
system is in the condenser. This is the most likely point of
CaC03 or CaSO<+ scale formation since these species are less
soluble at higher temperatures. The condenser temperature should
therefore be used when evaluating tower scaling potential.
The data collection discussed so far is concerned with
data representing a reasonably long period of time. This data is
very useful to characterize the operating conditions of the tower
under a variety of natural and man-made variations. Of necessity,
this data must be collected over long periods of time and some
of it may not be available. To supplement this data, direct
sampling of the cooling tower's major streams may be necessary.
2.1.2 Sampling Program
In cases where tower operating data is incomplete, a
sampling program should be conducted at the plant for about two
weeks. During this period it is recommended that about 3-5 sets
of grab samples per day be taken. All of the samples should be
taken at approximately the same time to produce a series of "snap
shots" of the system operation which can be used to identify
deviations from steady-state.
The data which are required include much of the informa-
tion listed in Table 2-1. The most important samples are of the
cooling tower makeup and blowdown. If flow data is available, it
should be noted, but the costs of magnetic, turbine, or orifice
metering devices for large streams are too expensive to use for a
short sampling period. The wet bulb and dry bulb temperatures
and the inlet and outlet condenser temperatures should be recor-
ded at the time that the sampling takes place.
The grab samples of cooling tower makeup and blowdown
which are taken should be analyzed using the procedures which are
reported in the EPA's "Manual of Methods for Chemical Analysis of
Water and Wastes". Important parameters include pH, temperature,
total dissolved solids, calcium, magnesium, sodium, carbonate,
sulfate, nitrate, chloride, and silica.
If any treatment is used, such as acid for scale control
or chlorine for biological control, it should be monitored
-96-
-------
continuously. It is more important that average values of these
very small addition streams are known than the instantaneous
rate at the time when the grab sample is taken.
If softening is used for the cooling tower, separate
grab samples_of the inlet and outlet streams from the softeners
will be required. Sample analyses of the inlet and outlet stream
of the softener are required in the same manner that the cooling
tower makeup and blowdown are. The outlet stream from the soft-
ener should also be analyzed for suspended solids to determine
if the flow rate is too large for the clarifier and sedimentation
system. If a substantial level of suspended solids is noticed,
then the softener is being overloaded.
Some of the data which is useful to characterize a
cooling tower cannot be easily measured. In this case design
information or values calculated using indirect measurements must
be used. In general, the water recirculating rate, the air flow
rate, and the drift are not well known, especially in natural
draft towers. Estimates of the water flow rate can be made using
data about the inlet and outlet condenser temperatures and the
electric load on the generator to estimate the waste heat load.
The air flow rate can be estimated using data concerning the
load, the ambient conditions and the approach on the tower. The
drift is not easily estimated and can only be measured
approximately.
To characterize the operation of a cooling tower all
sources of information should be consulted. The main sources are
historical data kept by the power plant and government agencies,
and sampled data which can be obtained over a short period of
time. Any of the required data that cannot be found in these
two sources can be supplemented by design information and in-
direct calculations.
2.2 Phase II; Evaluation of Operating Alternatives
The second phase of a cooling tower implementation
plan is to formulate and evaluate various modes of cooling tower
operation. Formulation of alternative operating conditions will
depend on how the cooling tower fits into the overall plant
recycle/reuse scheme. The plant studies conducted in this pro-
gram showed that cooling tower blowdown can be cascaded to other
power plant water systems such as ash sluicing and S02/particulate
scrubbing.
-.97-
-------
In a plant designed for minimum liquid discharge, the
cooling tower blowdown rate is set by the water requirements of
the systems which use cooling tower blowdown as makeup. Since
evaporation and drift are determined by tower design, the blow-
down sets the cycles of concentration required in the tower (see
Equation 2.1). The types of water systems used at a particular
plant site and therefore the quantity of water required by these
systems determine the cycles of concentration required in the
cooling tower.
To determine the feasibility of operating the cooling
tower system at the cycles of concentration required by a parti-
cular recycle/reuse option, both potential scale formation and
the level of treatment required should be investigated. The
plant studies portion of this project showed that a computer pro-
cess simulation package is a very useful tool with which to study
cooling tower systems. The simulations can be used to identify
the operating characteristics of the tower under increased recycle
conditions. The results of simulations can also be used to iden-
tify the treatment methods required to operate with increased
recycle and the degree of treatment necessary. The following
sections describe the types of calculations required to evaluate
cooling tower operation and a methodology for using a computer
model to perform the evaluations.
2.2.1 Evaluation Criteria
The cooling tower evaluation should involve mass and
energy balances to calculate the flows? compositions, and tem-
peratures of all of the process streams. The calculations should
include predictions of the inorganic aqueous equilibria phenomena
that occurs in cooling towers, including solid-liquid and gas-
liquid equilibria. The solid-liquid equilibria include all of
the important scale forming species: CaCOs, Mg(OH)2, CaSO^-ZHzO,
Si02, and other silicate scales. The gas-liquid equilibria
predicitions are necessary to calculate the evaporation rate and
the degree of C02 transfer that occurs in the tower. These cal-
culations allow cooling tower operation to be evaluated from both
the standpoints of scale potential and treatment requirements.
2.2.1.1 Scale Potential
The most important factor which determines what cycles
of concentration can be reached in a cooling system is the poten-
tial for scale formation in the condenser. If scale does form in
the condenser, the heat transfer can be reduced significantly.
This will raise the steam condensing temperature, increase the
-93-
-------
back pressure in the turbine and reduce the efficiency of the
power generating cycle.
In the plant studies portion of this study the con-
cept of relative saturation was used to measure the scaling
potential of aqueous solutions. For calcium carbonate, a common
scale, the relative saturation is defined in Equation 2.2.
R.S.
(CaC03)
K
(CaC03)
(2.2)
(2.3)
K
(CaC03)
where
R.S.
(CaC03)
K
(CaC03)
c.-" =
relative saturation of calcium
carbonate
solubility product of calcium
carbonate
activity of the calcium ion
M- ++ = moality of the calcium ion
Y
Ca
aC03
M
col
Yco7
activity coefficient of the
calcium ion
activity of the carbonate ion
molality of the carbonate ion
activity coefficient of the
carbonate ion
The activities must be used instead of the molalities of the
ions in solutions to take into account the non-idealities which
exist in aqueous solutions. Equation 2.3 defines, relative satu
ration in terms of ion molalities and activity cont
activity coefficients account for deviations
In the plant studies portion of this project, these
-99-
-------
were shown to become larger as the total dissolved solids of
the solution increased.
When the relative saturation of a species is below 1.0,
no potential for scale formation exists. When it is greater than
1.0, a potential for solids formation exists but scale will not
necessarily form. When the relative saturation of a species is
above its "critical value", nucleation will occur and scale for-
mation is more likely. The critical value for CaSOit-ZHzO is
1.3-1.4 and in this study, the critical values of CaC03 and
Mg(OH)2 were found to be about 2.5 and 3.4, respectively.
The most common scale forming species in cooling towers
is calcium carbonate. The relative saturation of CaCOs is depen-
dent on the concentration of the calcium and carbonate ions, as
shown in Equation 2.3. The calcium concentration depends on the
cycles of concentration. The carbonate concentration depends on
the pH and the liquid-gas equilibrium between gaseous carbon
dioxide and dissolved carbonate. It is necessary that the C02
transfer which occurs between the air and the water in the tower
is accounted for to accurately predict the relative saturation of
CaCOa. This calculation is especially important in cases where
there is no pH control because the pH of the circulating water
is affected by the amount of total carbonate species in solution.
Figure 2-2 is a sample plot of the relative saturation
of CaC03 in the cooling tower blowdown as a function of the
cycles of concentration. This data was obtained using the computer
model of cooling towers described in the plant studies portion of
the project. This plot shows that the relative saturation of
CaCOs increases in a non-linear fashion, and the curve becomes
steeper as the cycles increase. The relative saturation increases
at a faster rate at higher cycles of concentration because the
pH rises dramatically as a function of cycles, as can be seen in
Figure 2-3. The increased pH causes less C02 to be desorbed in
the towers as well as causing a shift of bicarbonate ions to
carbonate ions. The plots shown in Figures 2-2 and 2-3 apply to
cooling tower systems where the blowdown is adjusted to keep the
cycles of concentration at a low level to prevent scale formation.
However, the discussions concerning relative saturation and C02
transfer also apply to systems where treatment is used.
2.2.1.2 Treatment Alternatives
In many cases blowdown alone is not sufficient to con-
trol CaC03 scale and pH control may be instituted. Sulfuric
-100-
-------
1.
.75
s
O
_J
CD
lil
n
O
0.50
O
CE
3
V)
.25
Ul
LU
OC
1.0
2.0
3.0
4.0
5.0
[CYCLES OF CONCENTRATION]
Figure 2-2. CaCO3 scale potential as a function of
cycles of concentration in cooling towers
without treatment (example).
-101-
-------
7.9
7.8
Z
§7.7
S
o
01
I
i-
u.
O
7.6
7.5
1.0
2.0 3.0 4-°
[CYCLES OF CONCENTRATION]
5.0
Figure 2-3.
Cooling tower blowdown pH as a function of
cycles of concentration in towers without
treatment (example).
-102-
-------
acid treatment is used to control the pH of the recirculatine
water because it is the least expensive of most commercially
available acids. The gas-liquid-solid equilibria calculations
identify the pH required to control the relative saturation of
CaC03 to a specified level as well as the amount of sulfuric
acid needed to maintain this pH. Mg(OH)2 and some silicate
scales can also be controlled with acid addition, but CaC03 is
usually the limiting scale.
CaSOif*2H20 is another scale that can be encountered
in cooling systems. Since its solid-liquid equilibrium is not
pH dependent, acid addition will not reduce its relative satura-
tion. The addition of sulfuric acid increases the sulfate ion
concentration and actually increases the relative saturation of
CaSCK*2H20, although in most cases this effect is insignificant.
When problems are encountered with gypsum scale, softening is
required to reduce the calcium concentration, either by pre-
treatment of the makeup water or slipstream treatment of the
recirculating water. The evaluation of a particular option
should involve determining the chemical requirements for pre-
treatment and/or slipstream treatment.
The size of the stream required to be softened must be
calculated. This is especially important because it determines
the size of the softening equipment which will be required. It
is imperative that the softening equipment be able to handle the
largest stream that might be necessary to soften. If the equip-
ment capacity is exceeded, the flocculation and sedimentation
times are reduced and CaC03 solids may carry over into the
cooling water. The results of this carryover could be a heavily
scaled condenser. The maximum stream size should be as small
as reasonable though, because the capital cost of the softening
equipment is relatively high and is a direct function of the
size of the softened stream.
2.2.2 Model Application
A fundamental computer model such as the one used in
the plant studies is a very useful tool in performing the cal-
culations discussed above concerning evaluation of cooling tower
operation. The cooling tower model should initially be verified
for the specific cooling system being studied using the data
collected in the characterization phase. This can best be done
by inputting data collected during the sample periods and com-
paring the simulated results to measured values. If the data
base is extensive enough, the model assumptions can be altered
to improve its reliability. In cooling tower systems parameters
-103-
-------
such as the air flow rate and the degree of C02 transfer are not
always known to a high degree of accuracy. These parameters_may
be adjusted based on extensive operating data to more effectively
model the actual tower operation.
The verified model can then be used to simulate alter-
native operating conditions. The simulations should be performed
for the full variety of conditions which the tower can be expected
to operate under. The analysis of the available historical data
is very useful in the identification of seasonal as well as non-
seasonal extremes in weather and water quality. It is very
important that the worst case operation is simulated, so that the
cooling tower modifications, based upon these simulations, will
be able to handle all expected conditions.
The model should not only be used to simulate cooling
tower operations which account for long range variations in oper-
ating conditions, but also for short-term upsets. The effects of
drastic changes in load, steam condensing temperature, drift,
blowdown, and climate should be identified with simulations.
These simulations identify the kind of changes in treatment rates
that the modified system should be able to handle. It is very
important that upsets in system operation do not cause the rela-
tive saturation of any scale forming species to rise above its
critical value, and scale the condenser.
Several simulations should be performed to verify and
"fine tune" the model. Alternative operating conditions may
require 10-15 simulations of varying ambient and operating con-
ditions. Three to five additional simulations should be performed
to estimate the effects of short-term upsets.
The results of these simulations identify the operating
conditions which can be expected with increased recycle. They
specify the temperatures, flows and compositions of the important
streams under various operating conditions. These results estab-
lish a sound basis upon which to design an acid treatment system,
if necessary, which can operate effectively under all foreseeable
situations. If softening is required, the simulation results
identify the maximum flow through the softener and the maximum
chemical addition rate as well as the quality of the outlet
stream from the softener. The results of these simulations should
establish a sound basis upon which to design system modifications
which will increase the cycles of concentration in the cooling
tower but will not reduce the reliability of the system.
-104-
-------
2.3 Phase III: Cooling Tower Modifications
After the optimum system operation is identified from
the results of the evaluations, a complete design with the new
operating conditions can be prepared. This design should include
specifications of the flows, temperatures, and compositions of
the main process streams as well as all of the new equipment and
instruments necessary to monitor the cooling system.
A plan for the necessary changes in operating conditions
from the initial system to the modified system can be developed
with the new design. The plan will help insure safe operation
and minimize disruption to the main operating objectives of the
plant, the production of electricity.
This section discusses the kinds of modifications that
can be made on a cooling tower to increase the cycles of concen-
tration, and presents examples of how the operating changes could
be implemented.
2.3.1 Equipment
In most cooling systems additional equipment will be
required to significantly increase the cycles of concentration.
This new equipment can include piping, pumps, controllers, feed-
ers, a clarifier, or a thickener. This subsection discusses the
equipment necessary to increase recycle in cooling towers.
Additional piping and pumping capacity may be needed to
reroute the plant makeup to replace the water which is not supplied
by cooling tower blowdown because of the increased recycle in the
towers. This will be a temporary situation required only during
implementation of the modifications when the cooling tower blow-
down is being reduced. If pH control is added to the system, a
pump, a storage tank, and a pH controller will be required. The
maximum capacity of the metering pump should be larger than the
flow rate calculated in the evaluations of worst case conditions.
If slipstream or makeup softening is added to the system, a
chemical feeder with variable feed rates, a flocculator, and a
sedimentation tank will be necessary. The residence time in the
flocculator should be at least 40-60 minutes. Sedimentation re-
quires a 2-4 hour residence time. A slurry pump and piping will
also be required to pump the softener sludge to the ash pond. The
cost of equipment for modifying cooling tower operation will be a
strong function of the size of the system and the levels of treat-
ment necessary.
-105-
-------
Any new treatments added to the system require addi-
tional monitoring. The acid addition can be checked with pH
probes to insure that complete mixing of the acid and cooling
water occurs. If only partial mixing occurs before the cooling
water enters the condenser, corrosion and scale problems might
occur. The inlet and outlet streams from the softener should be
monitored continuously for conductivity, pH, and turbidity. If
a significant increase in turbidity is noticed, the flow rate may
be too large for the flocculator and carryover is occurring, also
increasing the danger of CaC03 scale. A common procedure for
softening operations is to perform "jar tests" to determine what
new operating conditions are necessary for safe operation. How-
ever, if increased turbidity is seen, the blowdown from the tower
should be increased to avoid any scale formation.
If slipstream treatment is used, the flow through the
softener is an important operating variable. The total amount
of calcium and magnesium removed in the softening step is
directly dependent on the flow and concentration of the calcium
and magnesium ions. To effectively control the total amount of
hardness removed by the slipstream softener a flow meter should
be installed to monitor the flow on a continuous basis.
2.3.2 Implementation
Implementation of cooling tower modifications which
involve increased recycle should be performed in a stepwise man-
ner to insure safe operation. Each step should take about a week
of monitoring, to insure that the tower is operating as expected.
Samples should be taken of the makeup and the blowdown at least
twice a day and analyzed for calcium, magnesium, sodium, sulfate,
chloride, and carbonate. The TDS can be measured and correlated
to conductivity measurements which will give a continuous read-
out on the quality of the makeup and blowdown streams. Changes
in the conductivity of these streams implies that changes in
the water quality have occurred and the towers are not operating
as expected.
The manner in which increased recycle is implemented
can be best illustrated with examples. There are four basic
modes of cooling tower operation:
1) Operation at low cycles of concentration
so that no treatment is required.
2) Operation at higher cycles of concentra-
tion with sulfuric acid addition.
-106-
-------
3) Operation at higher cycles of concentration
with sulfuric acid addition and softening
of the makeup water.
4) Operation at higher cycles of concentration
with sulfuric acid addition and softening of
a slipstream from the recirculating water.
In the plant studies, two plants which operate their
cooling towers in the first mode were identified. These plants
operate their natural draft towers at less than three cycles of
concentration. These plants have a very high quality makeup to
their towers as do most plants which operate without treating
their cooling water for scale prevention. Preliminary studies
of these cooling tower operations indicate that by changing to
the second mode of operation, the towers can operate safely at
cycles of concentration as high as fifteen.
To add the capability of pH control to a cooling system
only requires a pump, a storage tank, and a pH controller. After
this additional equipment is installed, the implementation proce-
dure can be initiated. This implementation procedure should take
about five to ten weeks to complete the change from operation at
low cycles of concentration to a higher level.
Each step which increases the cycles of concentration
by reducing the blowdown should be operated for at least a week.
During this period, the pH, temperature, and conductivity of the
makeup and blowdown will be monitored continuously. The plant
operator should be instructed as to what the maximum acceptable
levels in these readings are and to increase the blowdown if any-
one exceeds the maximum. It is recommended that complete chem-
ical analyses of the makeup and blowdown streams be performed at
least twice a day. The results of these analyses can be used to
revise the maximum acceptable levels of the continuous read vari-
ables, if necessary. The maximum temperature, pH, and conduc-
tivity pertain to a condition very close to the maximum accept-
able scale potential.
The relative saturation of CaSO^HaO cannot be con-
trolled by acid addition because it is not pH sensitive. If a
cooling tower is operating near the critical scaling value of
CaSO^HaO, softening is required to increase the degree of
recycle. Softening can be used to reduce the calcium level in
the makeup or a slipstream from the recirculating water. These
levels of operation represent the third and fourth modes listed
above.
-IQ7-
-------
If softening is being added to an existing system,
testing operation of the softener is recommended before the
softened stream is fed into the cooling tower system. Complete
chemical analyses of the effluent will insure that sufficient
calcium is being removed. The pH should be low enough so that
CaC03 or another pH sensitive solid will not form where the make-
up is added to the cooling water. The tubidity of the effluent
can be measured to insure that complete sedimentation is occurring
in the thickener. Jar tests performed on the influent will deter-
mine the optimum lime addition rate. Jar testing involves adding
various dosage rates to samples of the softener influent. The
optimum dosage rate is the lowest rate which still produces a
precipitate that settles readily.
After safe operation of the softening equipment is
assured, operation of the cooling tower with softening can be
attempted. The tower can initially be run at the normal level
of recycle with the softening equipment in operation. Complete
monitoring of the temperature, pH and conductivity of the make-
up, blowdown, and softener effluent is required to track the
system operation. The plant operator should be instructed as to
the maximum acceptable levels of each of these variables. Com-
plete analyses of these streams at least twice a day are recom-
mended. Daily jar tests are necessary to determine the optimum
level of lime addition. Operation at each step should be carried
out at least a week before the level of recycle is increased
further.
Softening the makeup stream will control
scale and is easier to control that slipstream treatment, since
the softener is not in the cooling loop itself. The results of
Phase I of this study have shown that slipstream treatment is
more efficient and can allow a greater degree of recycle in
cooling towers with smaller capital costs for equipment. If a
plant wishes to change its cooling system from pretreatment to
side-stream treatment, this could be accomplished by repiping
the system.
In the case of existing softening equipment, reliability
has been proven for softening the makeup stream. Therefore, ini-
tial tests of the repiped system will be short. The system
should be monitored continuously, as discussed earlier for soften-
ing in general, and the recycle should be increased in a stepwise
manner. Because the quality of the slipstream tends to fluct-
uate more than the makeup water, more frequent jar tests are
recommended. As with the other system modifications complete
change from initial operation to final operation should take
about five to ten weeks.
-108-
-------
In summary, there are four basic modes of cooling tower
operation. As the treatment level becomes more complex from acid
addition to slipstream softening, the degree of recycle achievable
also increases. To change modes and increase the cycles of con-
centration in the cooling tower will usually require about five
to ten weeks after all new equipment is installed. This time is
required to insure safe operation of the cooling towers at all
times and to educate the operating personnel.
-109-
-------
3.0 ASH SLUICING
The ash sluicing system is a major water consumer at
many coal-fired power plants. Most plants sluice the bottom ash
on an intermittent basis from the boiler to a pond where the ash
settles. Fly ash is usually removed from the flue gas by an el-
ectrostatic precipitator, a mechanical collector, or a fabric
filter. The fly ash collected by these means is often sluiced
from the collection mechanism to an ash pond for final disposal.
The bottom ash and fly ash can be sluiced to a common pond or
separate ponds.
Bottom ash is usually sluiced in a slurry of 1-5% solids
and fly ash is sluiced at 5-10% solids. Losses occur via evapora-
tion from the surface of the pond, occlusion with the ash in the
bottom of the pond (-50% solids), and seepage through the pond
liner. The remaining sluice water must then be discharged or
used somewhere else in the plant.
Because ash pond overflow is usually high in dissolved
solids it is not suitable for many other uses in the plant. One
possible consumer of ash pond overflow is a scrubbing system.
However, the makeup requirements for a scrubbing system are
usually much smaller than the ash pond overflow from a once-
through sluicing operation. Therefore, possible recycle/reuse
alternatives involving ash sluicing operations necessarily inc-
lude the possibility of closed-loop or partially closed-loop
systems. The plant studies portion of this project investigated
both closed-loop ash sluicing and partially closed-loop opera-
tions. In a closed-loop system, all of the ash pond overflow
is recirculated to sluice ash. The only water losses from this
type of system are occlusion with the ash in the pond and evap-
oration. In a partially closed-loop system, a portion of the
ash pond overflow is recycled and the remainder is either casca-
ded to a water consumer such as a scrubbing system, or it is dis-
charged.
Figure 3-1 shows a general recirculating ash sluicing
system. Some recirculating bottom ash sluicing systems presently
operate without significant problems. Two of the plants studied
in this project operate recirculating bottom ash sluicing systems
without scale formation in the slurry lines. Bottom ash is essen-
tially non-reactive and does not serve as a major source of dis-
solved solids to the slurry liquor. Fly ash is generally much
more reactive and a significant amount of dissolved species can be
leached from the ash when it is slurried with water. Most fly ash
sluicing systems are once-through operations, with the pond over-
flow discharged.
-110-
-------
ASH
i
EVAPORATION
MAKEUP
SLUICE
WATER
SOLID/LIQUID
MIXING
f
POND
: f SLOWDOWN
i
I
i
T
SLUDGE
POND RECYCLE
Figure 3-1. Re circulating ash sluicing flow scheme.
This section presents a methodology for implementing
a recirculating ash sluicing system. This is broken into four
phases:
Phase I
Phase II
Phase III
Phase IV
System Characterization
Alternative Evaluation
Pilot-scale Studies
Full-scale Operation
The first phase involves collecting a data base which
is required to identify reasonable system modifications. The
second phase identifies the degree of recycle attainable without
treatment to prevent scale formation. If a greater level of re-
circulation is required, the second phase also identifies the
treatment requirements to achieve that level.
The evaluation of alternatives in Phase II will depend
heavily on the accuracy of the ash reactivity data collected in
Phase I. The results of the plant studies portion of this pro-
gram showed that ash reactivity is a strong function of both the
pH and dissolved solids content of the sluice water. The trans-
fer of C02 between the atmosphere and the ash pond will affect
the pond pH and therefore affect the ash reactivity. The_rela-
tionship between ash reactivity and sluice water composition is
-111-
-------
a very complex one and deserves pilot studies to better evaluate
recirculating ash sluice systems. The third phase in the imple-
mentation plan is to conduct pilot studies to better define the
ash reactivity in a recirculating system and to evaluate treat-
ment options for preventing scale formation.
The final phase of the implementation plan for recir-
culating ash sluicing is to design and startup the full-scale
system.
3.1 Phase I: Ash System Characterization
In the characterization phase, the basic data necessary
to evaluate a recirculating ash sluicing system is assembled.
These data include information about both average and extreme
operating conditions. This section discusses the important data
for characterizing an ash sluicing operation. First, the impor-
tant process variables are presented and discussed. Next, a sam-
pling plan to supplement existing data is presented. Finally,
bench-scale studies to further characterize the particular ash
to be sluiced are outlined.
3.1.1 Identification of Process Variables
The major process variables which affect the feasibility
of implementing a recirculating ash sluicing system are listed
below:
1) Reactivity of the ash,
2) Rate of production of ash,
3) Surface area of the ash pond(s),
4) Local climatological data,
5) Sluice water flow and composition,
6) Degree of C02 transfer between the
pond liquor and the atmosphere.
The most important parameter is the ash reactivity.
The ash reactivity is defined as the amount of soluble species in
the ash. The results of the plant studies portion of this pro-
gram showed that the major species leached from the fly ashes
-112-
-------
studied are calcium, magnesium, sulfate, and sodium When the
ash is contacted with water, these species dissolve and may cause
calcium carbonate or calcium sulfate scale formation.
The collection of accurate data concerning ash reacti-
vity is therefore extremely important when evaluating the feasi-
bility of a recirculating ash sluice system. The plant studies
conducted earlier in this program showed that the ash reactivity
varies considerably from plant to plant and that the reactivity
of a particular ash is a function of the sluice water composition.
The best way to collect accurate ash reactivity data is through
sampling and experimental studies. These methodologies are pre-
sented in the following sections.
The rates that fly and bottom ash are produced deter-
mine the amount of water required as makeup to a sluice system.
In a recirculating system, the total makeup required is the sum
of the water lost through occlusion with the ash in the pond and
through evaporation. The rate of ash production and the percent
solids the ash will settle to determine the amount of water lost
through occlusion.
The pond surface area and local climatological condi-
tions determine the rate of water lost from the pond through
evaporation. The makeup rate required is therefore partially
determined by the pond area and climate.
As 'previously stated, the sluice water flow and compo-
sition affects the ash reactivity. The makeup water will, to
some extent, determine the type and levels of dissolved salts
present in the recirculating system.
The last parameter listed, C02 transfer, will affect
the pH and carbonate concentration of the pond recycle liquor,
and thereby affect the ash reactivity. Characterizing the amount
of CO2 transfer occurring is a very difficult task. Not only
does the C02 transfer depend on the pond pH but it also depends
on the amount of surface agitation resulting from wind. The
plant studies portion of this project investigated the two ex-
tremes of CO2 transfer: no transfer and complete equilibrium
with the atmosphere. The actual level will be between these _
two extremes. Pilot studies to evaluate recirculating ash sluice
systems may provide data which can be used to better predict the
level of C02 transfer in full-scale systems.
-113-
-------
3.1.2 Sampling
Sampling of an existing operation is one method to
determine ash reactivity. By sampling the sluice water and the
ash slurry entering the pond, the amount of soluble species in
the ash may be calculated. However, this calculation is depen-
dent on the assumption that no precipitation occurs in the sluice
line. To determine if a potential for precipitation exists,
analyses of the ash slurry liquor can be used to predict scale
potential. A sampling program can also provide information con-
cerning variations in ash reactivity. A one to two week sampling
program should provide sufficient data to evaluate an existing
ash sluice system with respect to ash reactivity.
Three to five samples from each slurry being sluiced
to the pond(s) should be taken each day. It is recommended that
the inlet and outlet streams be sampled simultaneously, if possi-
ble, to obtain a series of "snapshots" of the system operation.
These samples can be analyzed using procedures outlined in the
EPA's "Manual of Methods for Chemical Analysis of Water and
Wastes" for pH, temperature, total dissolved solids, total
suspended solids, calcium, magnesium, sodium, carbonate, sul-
fate, nitrate, chloride, and silica.
If flow data is available, it should be noted, but the
costs of magnetic, turbine, or orifice metering devices for large
streams are too expensive to install for a short sampling period.
Ash flow rates data can be combined with the analysis for sus-
pended solids in the slurries to calculate the fly ash and bottom
ash slurry flow rates.
Samples of the ash which are produced during this sam-
pling period should be obtained and labeled. These samples can
then be studied under laboratory conditions and the results can
be compared to the results measured under actual operating condi-
tions. This comparison can improve the usefulness of the labora-
tory findings.
3.1.3 Laboratory Studies
In many systems the ash may serve as the largest source
of dissolved species in the sluice water. This is especially true
when a recirculating fly ash system is operated where the pond
water is continually brought in contact with fresh ash and only a
small amount of "cleaner" makeup water. Depending on the level
of ash reactivity, dissolved species concentrations may increase
-114-
-------
to the point where CaCO 3> CaSO,, Mg(OH)2> or other scales form
in the sluice line and restrict the flow of the slurry A well
designed system will not operate for any significant period of
time in_the scaling mode. In order to design a recirculatine
ash sluicing system that will not scale under normal operating
conditions it is important to predict the reactivity of the ash.
Samples of both the bottom and fly ash which are pro-
duced at the plant should be obtained if recirculating systems
for both are being considered. In most cases the reactivity of
the fly ash is much higher than the bottom ash. Samples of both
should be studied initially because the reactivity of ash is
affected by many parameters. The reactivity of ash is known to
be a function of the mine it is from, the furnace it is burned
in, and the conditions at the time of combustion. Therefore, it
is important that multiple samples of the ash from all of the
coals burned at the plant are obtained so that a reasonable
range of the ashes produced at the plant are studied. The reac-
tivity of a given ash sample is also a function of the pH and
composition of the sluice water. This requires multiple tests
to obtain a reasonable estimate of how the reactivity will change
with leachate composition.
Two types of laboratory studies were used to determine
ash reactivity in the plant studies portion of this program.
Beaker leaching studies with deionized water were performed at
various pH's to determine the maximum level of soluble species
in the ash. Recirculating bench-scale studies were also conduc-
ted. The results of these two types of experiments indicated
that the amount of species dissolving from ash is much smaller
in a recirculating system than the level determined from the
beaker leaching studies. The bench-scale results should more
closely reproduce actual conditions and therefore should be
used to characterize ash reactivity.
A flow schematic of the bench-scale apparatus used
in the plant studies is shown in Figure 3-2. This equipment
may be operated with varying degrees of recycle by adjusting
the blowdown rate. A mixing tank was used to contact the ash
with pond recycle and makeup water. The ash slurry then flows
from the tank to the settling pond where the ash settles. The
pond recycle and makeup water were transported with peristaltic
pumps and the ash was added via a worm screw feeder.
-115-
-------
ASH STORAGE HOPPER
MAKEUP
WATER
WORM SCREW FEEDER
©
8-
MIXING TANK
-tXh
SLOWDOWN
SETTLING POND
OVERFLOW
- SAMPLE POINTS
Figure 3-2. Bench-scale model of ash sluicing facilities
-------
By sampling the makeup water, pond recycle water and
the ash slurry, and monitoring the flows of ash, makeup water
blowdown and pond recycle, material balances around the mixing
tank may be made to determine ash reactivity. Analyses of the
aqueous samples taken should include calcium, magnesium, sodium
chloride, carbonate, sulfate, nitrate, and silica Variations '
in mixing tank residence time, blowdown rate, recycle rate and
makeup water quality may be used to evaluate ash reactivity and
potential scaling problems under different operating conditions.
In order to facilitate the completion of bench-scale
experiments within a reasonable time period, the size of the
equipment should be considered. With the equipment used in the
plant studies, about 30 hours were required for the system to
reach steady-state. The pond liquor residence time was about 6
hours. Therefore, about five residence times were required for
steady-state operation. The equipment must be sized before ash
samples are obtained so that an adequate amount of ash may be
available for the system to reach steady-state.
The number of experiments required for a particular
ash will depend on the levels of variations in operating para-
meters identified as being likely and the chosen equipment size.
3.2 Phase II: Evaluation of Operating Alternatives
The second phase of a recirculating ash sluicing imple-
mentation plan is to formulate and evaluate various modes of
operation. Formulation of alternatives will depend on how the
ash sluice system fits into the overall plant recycle/reuse scheme.
If the ash sluicing operation is the final water con-
sumer in a minimum discharge cascaded plant water system, then
all of the pond overflow must be recycled. This mode represents
completely closed-loop operation. If some of the pond overflow
is required as makeup to another water consumer such as a scrub-
bing system, then only a portion of the pond overflow is recycled.
This mode represents a partially closed-loop operation.
To determine the feasibility of operating a closed-loop
or partially closed-loop ash sluicing system, both potential _
scale formation and the level of treatment required should be in-
vestigated. The plant studies portion of this project showed that
a computer process simulation package is a useful tool to study ash
sluicing operations. A computer model can be used to evaluate
-117-
-------
scale potential and treatment alternatives based on the charac-
terizing data collected in the first phase. It should be noted
here that the evaluation of alternatives will depend heavily on
the ash reactivity data collected and the amount of C02 transfer
which is assumed. The following sections describe the types of
calculations required to evaluate recirculating ash sluicing sys-
tems and a methodology for using a computer model to perform the
evaluations.
3.2.1 Evaluation Criteria
The recirculating ash sluice system evaluation should
include mass and energy balances to calculate the flows, composi-
tions, and temperatures of all of the process streams. The evalu-
ations should involve predictions of solid-liquid and gas-liquid
equilibria for the species occurring in the system. Important
solid-liquid equilibria include CaCOs, Mg(OH)2, CaSOi>'2H20, Si02,
and other silicate scales. Gas-liquid equilibria prediction is
necessary to account for C02 transfer between the ash pond and
the atmosphere.
Any examination of a recirculating ash sluice system
will require that certain assumptions are made. Steady-state
operation must be assumed so that an answer can be obtained with
a reasonable level of computational effort. Dissolution and pre-
cipitation may be assumed to be instantaneous or calculated rates
which are consistent with measured data can be used. The evalu-
ation should include predictions of system operation under vary-
ing conditions of C02 transfer between the pond and the atmos-
phere. Plant measurements taken in Phase I of this project
showed that some C02 transfer occurs but complete gas-liquid
equilibrium between the ash pond and the atmosphere is not
achieved. The evaluation must assume a reasonable percent solids
in the settled sludge in order to predict the loss of water from
the system due to occlusion with the solids.
No real system operates under steady-state conditions
for a significant period of time, but a steady-state evaluation
under worst case conditions is very usefull for design of a sys-
tem which can operate safely under all conditions. Instantaneous
dissolution and precipitation is a worst case assumption and may
cause overdesign. Therefore, rate data should be used when avail-
able although this is not very easily obtained for many species.
Depending on the reactivity of the ash and the composition of the
pond liquor, C02 transfer can increase or reduce the scaling po-
tential of the slurry in the sluice line. The sludge that nor-
mally settles in ash ponds is in the range of 4070 to 607, solids
-118-
-------
although it is not normally homogeneous. The calculations and
assumptions outlined above allow a proposed recirculatlnS ash
sluicing system to^be evaluated with respect to scale pSLnlia
and treatment requirements. *^*^ Fu^encia
3.2.1.1 Scale Potential
£ The most imPortant factor which determines the feasi-
bility^of implementing a recirculating ash sluicing system is the
potential for scale formation in the ash slurry pipeline. If
significant scale deposits are formed in the pipeline, plugging
and subsequent interruption of the production of power may occur.
In the plant studies portion of this study, the concept
of relative saturation was used to predict scaling potential.
When the relative saturation of a species is below 1.0, no po-
tential for scale formation exists. When it is greater than 1.0,
a potential for solids formation exists but scale will not neces-
sarily form. When the relative saturation of a species is above
a "critical value", nucleation and scale formation is likely.
The critical value for CaSQk'2R20 is 1.3-1.4. In the plant
studies portion of this project the critical values of CaC03 and
Mg(OH)2 were experimentally determined to be about 2.5 and 3.4,
respectively. A detailed definition of relative saturation was
given in Section 2.2.1.1
Although the critical value of CaC03 was experimentally
determined to be about 2.5, samples taken of existing once-through
ash sluicing operations showed CaC03 relative saturations as high
as 38.8 without scaling problems. This may be due to erosion by
the ash slurry or the ash particles may be providing precipitation
sites for CaC03. It should be emphasized here that if the evalu-
ation of an alternative predicts a CaC03 relative saturation
greater than 2.5, a potential for scale formation exists but
operational problems resulting from heavy scale deposits may not
occur. Pilot studies to better define CaC03 scaling problems in
recirculating ash sluice systems are recommended before full-
scale operations are implemented.
3.2.1.2 Treatment Alternatives
In an ash sluicing system where excess scaling potential
is identified, treatment may be required to reduce,^e^"^ po-
tential. The ash sluicing evaluation should consider treatment
technologies and the effects of these technologies on the system
operation. Potential treatment technologies for ash sluicing sys-
tems include:
-119-
-------
1) A reaction tank with sufficient
residence time to allow the major
portion of the reactive part of
the ash to dissolve and to allow
for controlled precipitation of
any supersaturated species that
might otherwise scale in the
slurry line.
2) Softening to reduce the levels of
calcium and magnesium in the water
being recirculated to the sluicing
operation.
3) Treatment to reduce the level of
dissolved solids in the water being
recirculated for sluicing the ash
(brine concentration, reverse osmosis,
etc.) .
The first treatment that should be investigated is
the installation of a reaction tank at the point where the ash
and the water are contacted. A reaction tank in the system will
allow a large portion of the soluble ions in the ash to dissolve
leaving the remaining ash essentially inert. The tank also allows
any supersaturated species to precipitate in a controlled fashion
similar to the way CaSOit-2H20 and CaS03«%H20 precipitate in a
reaction tank in lime/limestone S02 scrubbing systems. The re-
sults of bench-scale ash studies may indicate the rate of leach-
ing from the ash. Precipitation rates of solids can be obtained
from the literature or laboratory studies. The residence time
of the reaction tank should be large enough to allow sufficient
ash dissolution and solids precipitation to prevent scale forma-
tion in the sluice lines. Pilot operations are recommended for
sizing a full-scale reaction tank.
Other possible treatments that will reduce scaling po-
tential in the slurry are softening, brine concentration, reverse
osmosis, etc. These treatments can best be employed on the water
which is recirculated from the ash pond. Softening, the less ex-
pensive of the two options, can reduce the calcium and magnesium
levels in the recirculating water. Other treatments can be used
to remove all types of dissolved solids from a portion of the
recirculating water, or from the pond overflow.
In many cases softening can be used to effectively re-
duce the level of calcium and magnesium to levels where the
-120-
-------
addition of the reactive species from thp
scaling situation in the slurry line Ca
zs i^^ <
^ ^-FI-^- bu B^c?;£:Lea ror Pilot or xu^-scaie implementa-
tion. Softening should be more expensive to build and operate
than a reaction tank but significantly less expensive than other
more elaborate treatments.
Brine concentration can be used as a treatment option
for recirculating ash sluicing systems. A discharge from the
pond can be treated and the clean water sent to the boiler de-
mineralizer and/or some other plant water consumer. The major
effect on the ash sluicing system from this treatment option is
that more makeup water is required to operate the system. Since
the makeup water is lower in dissolved solids than the pond water,
the scale potential in the slurry line is reduced.
3.2.2 Model Application
A computer model such as the one used in the plant stu-
dies is a very useful tool in performing the evaluation of a re-
circulating ash sluice system. However, since ash dissolution
kinetics are not well understood at this time, the results of the
computer calculations are based on instantaneous dissolution. The
evaluations of ash sluicing systems are therefore heavily depen-
dent on the ash reactivity data input. A computer model can be
used to perform multiple calculations which demonstrate the ef-
fects of variations in the input data. The two factors which
are the most difficult to quantify accurately are ash reactivity
and C02 transfer in the pond. For each alternative evaluated,
multiple simulations should be performed to determine the effects
of variations in both ash reactivity and the amount of C02 trans-
fer.
Bottom ash is usually much less reactive than fly ash.
For this reason, most recirculating bottom ash systems will not
be prone to scale formation. If the bottom ash system is going
to be operated completely independent of the fly ash system a
few simulations will usually be sufficient to show that the bottom
ash can be sluiced in a recirculating system without scale. If
the bottom ash is unusually reactive, more extensive study o±
recirculating systems may be required, in which case steps similar
to those proposed for recirculating fly ash systems would be
appropriate.
-121-
-------
The key stream in an ash sluicing system is the slurry
which is transported to the ash pond. If scaling problems are
encountered, they most likely will occur in the slurry where a
significant increase in total dissolved solids is usually exper-
ienced because of the dissolution of the reactive portion of the
ash. When simulating ash sluicing systems, it is useful to assume
that all of the reactive solids identified in the leaching studies
enter the liquid phase instantaneously. This assumption allows
the worst possible scaling potential to be identified. If the
relative saturations of all of the potential scaling solids are
less than the critical values, this indicates that the system
can be safely operated under the simulated conditions.
If scaling is indicated, treatment of the slurry or re-
circulating water may be required. The results of computer simu-
lations may be used to size the necessary treatment equipment
for scale prevention. However, pilot studies are recommended to
better define treatment requirements before full-scale implemen-
tation.
3.3 Phase III: Ash Sluicing Pilot Studies
For ash sluicing systems with essentially unreactive
ash, pilot studies may not be justified. For recirculating sys-
tems sluicing only bottom ash, the results of evaluations may in-
dicate that the scale potential of the slurry is small, in most
cases. These recirculating systems can be designed for full scale
operation as only a pond return line and additional pumps will be
required.
For ash sluicing systems where scaling in the slurry
line is a concern because of a combination of reactive ash and
poor water quality, pilot studies would be very important, and
could produce significant savings in the full-scale design. The
pilot system should include the treatment options identified in
the optimum system which resulted from the evaluations. The
pilot system should be designed at a reduced scale which is di-
mensionally consistent with the expected full-scale design.
Figure 3-3 is a diagram of a general pilot-scale facil-
ity. The system employs a reaction tank, a softener, and a blow-
down capability necessary to implement the three major forms of
treatment discussed previously. The clarifier serves the same
purpose as the ash pond in a full-scale system. The clarifier
separates the slurry into a sludge and a solids-free liquid stream
which can be recirculated via the return line to slurry additional
-122-
-------
FLOW
METER
MAKEUP WATER
ASH
OJ
REACTION TANK
RETURN LINE
CHEMICALS
FLOW
.METER
SOFTENER
SLUDGE -«CT-K
2±3
SLURRY LINE
XFLOW
METER
SLUDGE
SLOWDOWN
Figure 3-3. General pilot scale ash sluicing facility.
-------
ash. A clarifier has a much smaller residence time than an ash
pond, and thus allows the total system to attain steady-state in
a reasonably short time. This allows a larger number of experi-
ments to be performed with the clarifier as opposed to a system
with a large settling pond. If a clarifier cannot achieve the
solids concentration expected in a pond, a vacuum filter may be
added to the system.
A pilot system such as this can best be used on-site at
a power plant. The plant can serve as a source of ash and the
ash pond can be used as a receptacle for the sludge from the
clarifier and the softener. The system should be built near the
ash pond where there generally is enough space for such a facili-
ty. This will allow for short lines from the facility to the
pond and should keep the generally crowded area near the plant
itself free. If this kind of location is chosen a dry method
must be used to transport the ash from the source, near the plant,
to the pilot facility.
Once the pilot facility is constructed on-site, a series
of experiments should be performed with the ash from the plant un-
der various operating conditions. Each experiment should be per-
formed for a long enough period to attain steady-state operation.
The amount of time required will vary with the design and flow
rates chosen. Important variables which should be studied
include the makeup water quality, the residence time in the reac-
tion tank, the degree of recycle, the weight percent solids in
the slurry, and the size of the softened stream. Data may also
be taken to better define C02 transfer in the system.
The makeup water should be similar to the water that is
expected to be used in full-scale operation. If a number of make-
up water sources are being considered, experiments with the full
range of makeup water qualities should be performed. In general,
the water used for these experiments should have a TDS level and
pH near the levels expected in the makeup water for full-scale
operation.
The residence time in the reaction tank is a very impor-
tant parameter to study in the pilot studies. The residence time
of a reaction tank can be changed by varying the level of the
slurry in the tank, and, thereby, the effective volume, or by
varying the flow through the tank. Thus, with one piece of equip-
ment, a wide range of residence times can be studied. The tank
must be large enough to attain the maximum desired residence time
at the minimum flow rate that will still produce turbulent flow
-124-
-------
in the slurry line^ All smaller residence times can be attained
by increasing the flow rate or reducing the level in the reaction
tank.
For a given ash flow rate the makeup water requirement
is determined by the water lost from the system via evaporation
occlusion with solids in the pond, and blowdown from the system!
The sum of the makeup water and the recycled water is determined
by the weight percent solids desired in the slurry. Experiments
should be performed which vary the percent solids in the slurry
and the blowdown from the system. These both directly affect the
percent recycle that exists in the system, and, therefore, the
quality of the water entering the reaction tank. The weight per-
cent solids in an ash sluicing system can usually vary from 0.5%
to 5% solids for bottom ash and 1% to 15% solids for fly ash.
The degree of recycle can be varied from zero, in a once-through
system, to greater than 80% for systems that operate with mini-
mum discharge.
The softening equipment should be designed to operate
on a slipstream from the recycle line. This is the most efficient
location for this operation because the highest concentration of
calcium and magnesium will exist in this stream. For the pilot
design the softener should be large enough to handle all of the
largest recycle streams that the system can be expected to oper-
ate with. In this way the level of softening can be varied from
0 to 100% of the recycle stream for all conditions studied.
Preliminary experiments should be performed under con-
ditions similar to those simulated with the computer model. Sam-
ples of the ash slurry, clarifier underflow, clarifier overflow,
softener effluent and makeup water should be taken periodically
until steady-state is achieved. Analyses of these samples for
suspended solids, calcium, magnesium, sodium, chloride, nitrate,
carbonate, sulfate, silica, pH, and TDS will allow the system to
be evaluated with respect to scale potential. Material balances
may be performed around the reaction tank to calculate ash reac-
tivity. These analyses will also allow evaluation of the soft-
ening equipment performance. These preliminary results can be
compared to the simulated results and used to design further ex-
periments under conditions closer to optimum operation. These
preliminary experiments should only be performed long enough to
attain steady-state operation and obtain samples of the key
streams.
-125-
-------
Experiments under conditions that have been identified
as desirable from the preliminary experiments should be performed
for longer periods of time to insure that scaling will not be a
problem. These experiments can be used to identify optimum opera-
ting conditions. They will identify the minimum reaction tank
volume, the minimum softening requirement and the maximum percent
recycle that can be attained with the ash and makeup water used
at the plant. The total task should take about six months to
perform with 3-6 people working on the system. Costs associated
with the pilot studies will depend on the number of experiments
to be performed as identified from the evaluation phase and pre-
liminary sluicing operations.
The results of these studies will establish a strong
base upon which to design a full-scale system that will operate
with scale problems, minimize the discharge from the ash sluic-
ing system and be as economical as possible to build and operate.
The next subsection is devoted to a discussion of full-scale im-
plementation of a recirculating ash sluicing system at an existing
facility.
3.4 Phase IV: Full-scale Implementation
The final result of the pilot studies should be to
identify a recirculating ash sluicing system design which can
be operated at the power plant. The system should be able to
operate for long periods of time without scale formation in the
slurry lines reducing performance. The design should include
any monitoring equipment which may be necessary to minimize
equipment failure and operating costs. The system should be
designed to facilitate regular inspections at times when the
power plant is undergoing routine maintenance.
This section discusses the kinds of equipment that
may be used to build a recirculating ash sluicing system and
discusses an implementation procedure.
3.4.1 Equipment
For some recirculating ash sluicing systems which
sluice very reactive ash, many pieces of equipment may be
necessary to safely operate the system. For others where the
ash is essentially unreactive, only pumps and piping may be
required. The former case will most likely exist in fly ash
systems while the latter may exist for some bottom ash systems.
-126-
-------
The least expensive piece of equipment that may be
used to reduce the scaling potential of ash slurries is I reac-
tion tank. The reaction tank should be large enough to allow
the major portion of the leachable ions in the ash to enter the
solution^and allow controlled precipitation of any supersatura-
ted species that may be formed due to the dissolution. The
slurry can then be safely sluiced to the ash pond without danger
of scale formation. The pilot studies should determine the mini-
mum volume that will be required to obtain the necessary resi-
dence time.
In^some systems the ash may be so reactive that con-
trolled _ precipitation in the reaction tank may not be sufficient
to eliminate scale and softening of the pond recycle may be
necessary. The softener will reduce the levels of calcium and
magnesium in the recycle water and reduce the relative satura-
tion of CaSCU'2H20, CaCO 3 and Mg(OH) 2. The size of the softened
stream should be determined in the pilot studies. The residence
time in the flocculator should be at least 40-60 minutes and
sedimentation requires 2-4 hours. The sludge produced by soft-
ening can be discharged into the ash pond.
A third treatment option that may be identified as a
result of the pilot studies could be brine concentration. The
level of dissolved solids in the slurry can be controlled with
a blowdown stream. This stream can be treated with a brine con-
centrator which can supply the treated stream to other plant
water consumers such as the boiler or cooling towers. The size
of the blowdown should be determined from pilot studies which
will, in turn, determine the size of the brine concentrator re-
quired.
3.4.2 Implementation
Once the system is built, full scale implementation
should not present a problem. The major portion of the sluice
water in a recirculating system is usually the pond recycle.
The ash pond at most power plants has a very large residence
time, on the order of one to four weeks. Thus, the pond water
and, thus, the water quality of the sluice water will change
very slowly toward its new "steady-state" value. Thus, the
dampening effect of the pond should keep the system from exper-
iencing drastic changes in operation. The system should be mon-
itored very closely through the first month of operation to in-
sure no drastic upsets occur which will cause scale formation.
-127-
-------
4.0 SOa/PARTICULATE SCRUBBING
Previous plant studies under this contract showed that
combined S02/particulate scrubbing systems may be operated in a
closed-loop mode to minimize makeup water requirements and elim-
inate aqueous discharges. These studies also identified the
opportunity for reusing water discharged from other major water
systems as makeup water to a combined S02/particulate scrubbing
system. To minimize plant water requirements and discharges,
the overall plant water system may be cascaded with the scrubb-
ing system being the final step. Scrubbing systems typically do
not require the good water quality that a cooling tower does.
Cooling tower blowdown is therefore one candidate for use as
scrubber makeup in a cascaded plant water system. For plants
with wet ash handling, ash pond overflow may be another candidate.
The plant studies identified two potential problem
areas associated with using cooling tower blowdown or ash pond
overflow. First, if the scrubber makeup water is near satura-
tion with respect to gypsum, scaling problems may be encountered
when this water is used in the demisters. The amount of SOz
sorbed in the demister and the amount of CaCOs solids entrained
in the scrubber exit gas as well as the demister wash rate
determine if scaling is likely.
The second problem area identified involves the dis-
solved solids concentration of the makeup water. The evapora-
tion from the scrubbing system causes the makeup water to be
concentrated. If the TDS level of the makeup is high, excessive
TDS levels (>20 wt. 70) in the scrubbing liquor may result.
Excessive TDS levels can increase the energy requirements for
pumping and decrease the mass transfer in the scrubber. If the
alternate makeup source is high in chlorides, an excessive
chloride concentration may result in the scrubbing liquor which
may cause ocrrosion problems.
These recycle/reuse opportunities were identified for
"throwaway" scrubbing systems where a waste sludge is produced
and disposed of in a pond system. Some scrubbing systems are
used for S02 removal only. These systems may be "throwaway" or
"regenerative". Regenerative systems recover the sorbed S02 as
S02(2,) or H2S0lf(aq) and recycle the alkali as opposed to dis-
psoing of the sorbed sulfur as calcium sulfite or calcium sulfate
sludge. There is opportunity for recycle/reuse in regnerative
systems as well as throwaway systems. In most cases regenera-
tive systems require prescrubbers to prevent contaminant buildup
in the system, and recycle/reuse options may be used in these
prescrubber sections.
-123-
-------
In regenerative S02 scrubbing systems very little water
is required as makeup to the main scrubbing loop. This is becaule
no sludge is produced in which water is occluded and little or no
water is evaporated in the scrubber. The gas entering the main
scrubber in most regenerative systems is near saturation with
respect to water since a prescrubber or quench is normally used
to remove particulates and chlorides from the flue gas upstream
of_the main_scrubbers. Since very little water is lost in the
main scrubbing loop, the concentrations of particulates and
chlorides will build to intolerable levels if they are not
removed by a prescrubber.
The procedure to test and implement the use of cas-
caded water in these prescrubber systems is the same as for
throwaway systems and is discussed in this section. Emphasis
will be placed on throwaway processes, however, since these
types of systems were studied in detail in the previous studies
conducted under this contract.
The four phases to implement a water recycle/reuse
shceme for the major water systems of a coal-fired power plant
were discussed in Section 1.0. These are:
Phase I - System Characterization
Phase II - Alternative Evaluation
Phase III - Pilot-scale Studies
Phase IV - Full-scale Operation
The first phase involves gathering design data to
characterize the scrubbing system. It also includes a sampling
program to supplement the design data. Since scrubbing systems
are subject to operating variations from design, the sampling
will help identify the effects of these variations. Sampling is
also recommended for determining demister operating variables.
The second phase involves evaluating the feasibility
of various alternatives. The design and operating data^collected
in the first phase can be used as the basis for performing pro-
cess calculations to evaluate the feasibility of changing from
open-loop to closed-loop or of changing makeup sources These
evaluations involve determining demister scale potential and
scrubbing liquor composition. Calculations should accurately
predict TDS levels in the system but pilot studies in addition
to calculations are recommended to evaluate demister operation.
-129-
-------
The third phase of implementation is to conduct pilot
studies to evaluate alternative feasibility. Pilot studies will
allow an evaluation of scale potential in the demister. Since
the factors which affect demister scale potential (S02 removal,
carryover, wash rate) are not easily measured, pilot studies
should be conducted to assess the affects of process modifications
The fourth phase is to use the pilot study results to design and
start-up the full-scale modifications.
The plant studies conducted in this program involved
the first two phases of an implementation plan. Design and oper-
ating data for two combined S02/particulate scrubbing systems was
collected and used to evaluate operation under existing and alter-
native conditions. Implementing the alternatives studied at the
existing plants would require additional studies to evaluate de-
mister operation and to better define variations in operating
parameters.
The following discussions are written to apply to a
general throwaway scrubbing system. All phases of an implementa-
tion plan are presented although the characterization and evalua-
tion phases were addressed in the plant studies. Since^virtually
all scrubbing systems are unique with respect to operating con-
ditions, a detailed characterization and evaluation phase should
be conducted before implementing an alternative at a particular
site. The purpose of the plant studies was to identify and evalu-
ate the types of recycle/reuse alternatives achievable at coal-
fired power plants. The purpose of this document is to outline
a procedure for implementing the types of alternatives identified
in the plant studies.
4.1 Phase I: System Characterization
The first phase in implementing a recycle/reuse scheme
for a particular scrubbing system is to characterize the system
operation. This includes collecting all design and operating
data available and supplementing this data by collecting samples
of the important process streams. This section first discusses
the design and operating data required to characterize the system
and then presents a sampling program to obtain additional data.
4.1.1 Identification of Process Variables
The design and operating data necessary to characterize
a scrubbing system can be divided into three areas. These areas
represent the three major operations in a typical S02/particulate
scrubbing operation:
-130-
-------
1) gas scrubbing and demisting,
2) solids precipitation, and
3) solids concentration.
Figure 4-1 is a simplified block diagram which shows how these
three operations are combined. The gas is contacted with a
recalculating slurry in the scrubber section where the particu-
lates^and S02 are removed. Demisters are generally used to
minimize entrainment of the scrubbing liquor in the clean gas.
The absorbed S02 is precipitated as solid calcium sulfite and/or
calcium sulfate in the reaction tank. Concentration of the pre-
cipitated solids and the sorbed ash is achieved in the solid/
liquid separation section. This concentration is made by a
clarifier, filter, pond or a combination of the three. This
section discusses the design data for each of these portions of
an S02/particulate scrubbing systems that is required to char-
acterize the system. All the data requirements discussed in
this section should be easily obtained from the design specifi-
cations submitted to the utility by the scrubber manufacturer.
4.1.1.1 Scrubber and Demister Data
Table 4-1 is a data sheet listing the most important
design data needed to characterize the scrubber and demister
operations of a particular S02/ash scrubbing system. The first
items listed in Table 4-1 for the scrubbers are the flue gas
flow, temperature, and composition. The flue gas characteristics
affect the amount of S02 and ash removed and the amount of water
evaporated in the scrubbers.
The makeup water required for a scrubbing system is
determined by the evaporation rate and the water lost by occlu-
sion with the solid waste (ash, CaS03'%H20, CaSCU2H20). The
amount of solid waste generated, and, therefore, the occluded
water lost, decreases with decreases in the amount of S02 and
ash removed in the scrubbers. As the flue gas flow, S02 content,
or ash content is decreased and the sludge percent solids
increased, the makeup water required by the scrubbing system
will decrease.
The design flue gas characteristics for various plant
loads are therefore very important in determing the applica-
bility of a particular water recycle/reuse option involving the
scrubbing system. The makeup water requirement of the scrubbing
-131-
-------
MAKEUP
WATER
FLUE
GAS
ALKALI
STACK
GAS
DEMISTER
SCRUBBER
REACTION
TANK
SOLID/JJQUID
SEPARATION
WASTE
Figure 4-1. Typical scrubbing system flow scheme
-132-
-------
TABLE 4-1. SCRUBBER AND DEMISTER DATA
Scrubbers
A. Design Flue Gas
Flow (wet gas excluding ash)
Temperature
Composition
N2
02
CO 2
S02
HC1
H20
Ash loading into scrubbers
B. Fly Ash Removal Stage
Liquid-to-gas ratio
Design Removal
C. SO2 Removal Stage
Scrubber Type (spray tower,
marble bed, etc.)
Liquid-to-gas ratio
Design Removal
Ib/hr
°F
vol %
gr/scfd
gal/1000 acf
(outlet)
7
/o
gal/1000 acf
(outlet)
7
/o
II
Demisters
A. Once-through or recirculating loop
if recirculating:
makeup rate
blowdown rate
tank volume
B. Wash Rate
C. Wash water source (fresh makeup,
clarifier overflow, etc.)
D. Rate of scrubbing liquor entrained
with the gas entering the demister
GPM
GPM
gal
GPM
GPM
-133-
-------
system will determine the mode of operation of the water system
supplying the scrubber makeup in a cascaded system. This is
especially important in cases where zero discharge is desirable.
For example, if cooling tower blowdown is used exclusively as
scrubber makeup, the flue gas characteristics determine the
tower blowdown rate and thus the cycles of concentration in the
towers.
The design parameters for the fly ash and S02 removal
stages affect the amounts of ash and precipitated solids dis-
posed of and therefore the water lost by solids occlusion.
Although evaporation accounts for most of the water lost in
closed-loop scrubbing systems, the effects of ash and SOz
removal efficiency should not be ignored. A more detailed dis-
cussion of the parameters affecting scrubber makeup water
requirements may be found in the section concerning the plant
studies conducted in this project.
The design parameters in Table 4-1 for characterizing
the demister section of a scrubbing system include the system
type (once-through or recirculating), the wash rate, the wash
water quality, and the carryover rate. Makeup water is normally
used wholly or in part as demister wash or demister wash makeup.
The feasibility of using cascaded water for scrubber makeup
water in the demister will depend on the demister characteristics
as well as the cascaded water quality. The amount of SC>2 sorbed
and the degree of CaCOs dissolution that occurs in the demister
can be large enough to cause significant scaling, especially if
the wash water is already high in dissolved calcium and sulfate.
CaC03 solids are introduced into the demister loop by scrubbing
liquor entrainment in the gas entering the demister.
Both once-through and recirculating demister wash
operations may be used depending on the wash water requirements
and the total scrubbing system makeup water requirement. If
more water is required for demister wash than for the system
makeup, a recirculating system must be used. A recirculating
system uses a catch tray to collect most of the demister wash
before it falls through the scrubber, whereas a once-through
demister wash system allows the demister wash water to fall
through the scrubber. A recirculating system will be more
susceptible to scaling problems and therefore will require pilot
studies to determine if a cascaded water may be used as demister
wash. A possible solution is to use a combination of cascaded
water and fresh makeup water as demister wash.
-134-
-------
4.1.1.2 Reaction Tank Data
Table>4-2 presents the design data which characterize
the solids precipitation section of a lime or limestone based
SC5Ur o2S o£SHem- ,A reaction tank is used to control CaS03«%H20
and CaSC\.2H20 scale potential by allowing these solids to preci-
pitate on recirculated seed crystals. The makeup alkali and some
makeup water are added to the reaction tank. The alkali dissolves
and combines with the sorbed sulfur to form calcium sulfite and/or
calcium sulfate solids.
The additive type, composition, and addition rate will
affect the liquid phase concentrations of calcium and magnesium
in the system and therefore affect the calcium sulfite and cal-
cium sulfate relative saturations encountered. The design para-
meters for the additive are therefore important in characterizing
the solids precipitation section of a scrubbing system.
The type of water used to slurry the additive as well
as the additive type will determine the scale potential in the
alkali addition system. The use of waters saturated or near
saturation with respect to CaC03 or 03304-21120 may result in
scale formation of one or both of these two species when some of
the additive dissolves. This is especially important for lime
systems where additive dissolution is fast.
The reaction tank volume and slurry solids content are
included in Table 4-2 and affect the precipitation rate. In-
creases in slurry solids content and the reaction tank volume for
a given precipitation rate lowers the relative saturation of the
precipitating species. The required precipitation rates are de-
termined by the amount of S02 sorbed in the scrubber and the oxi-
dation occurring in the system. Higher oxidation causes the re-
quired precipitation rate of CaSO^HjO to increase but lowers
the required CaS03'%H20 precipitation rate.
The final item in Table 4-2 is the rate and composition
of makeup water added to the reaction tank. The quality of the
makeup water will affect the dissolved solids content of the li-
quor in the scrubbing system but was shown to have little effect
on scale potential in the system in the plant studies performed
in this project. The chloride content of the makeup water is im-
portant since this affects the chloride concentrations in the sys-
tem and may cause corrosion problems if the level is too high.
The chloride content of a cascaded water used as scrubber makeup
is therefore an important parameter to consider in a water recycle/
reuse scheme.
-135-
-------
TABLE 4-2. REACTION TANK DATA
I. Alkali Additive
A. Type (lime, limestone)
B. Composition
Lime
CaO or Ca(OH)2 wt. %
MgO or Mg(OH)2
Inerts
Limestone
CaCO 3 wt. %
MgCO 3
Inerts
C. Addition rate Ib/hr
D. Water Source for slurrying additive
(makeup, clarifier overflow, etc.)
II. Tank
A. Volume gal
B. Slurry suspended solids content wt. %
C. Makeup water added to tank
Flow GPM
Composition (design basis)
Calcium mg/£
Magnesium
Sodium
Chloride
Sulfate (as SOl)
Nitrate (as
-136-
-------
4.1.1.3 Solids Concentration Data
^Table 4-3 presents the design information necessary to
characterize the solids concentration section of a scrubbing
system. This data is divided into two parts. The first part
concerns the clarifier and filter which may or may not be pre-
sent in every system. In some cases, the blowdown from the
scrubbing loop is pumped directly to the pond or diluted with
pond liquor and pumped to the pond. In these cases, additional
data concerning mixing tank volume and solids concentrations of
all streams is required to characterize the system. The second
portion of the data in Table 4-3 is for the pond system where
the waste solids (ash, CaS03'%H20, CaS(V2H20) are deposited.
The solids concentration section of a scrubbing sys-
tem determines whether the overall system is closed or open-loop.
In a closed-loop system all water not occluded with the settled
sludge or evaporated in the pond is recycled to the scrubbing
system. The design flow rates and suspended solids concentra-
tions of all of the streams in the solids concentration section
of a scrubbing system are necessary to fully characterize the
system. The effects on the operation of this section of con-
verting from open-loop to closed-loop system design should be
evaluated for this type of recycle/reuse opportunity. Changes
in oxidation resulting from process modifications may have a
significant effect on solid/liquid separation equipment. Calcium
sulfite crystals are platelets and do not tend to settle as
easily as calcium sulfate solids. Careful consideration of
these effects should be made before a recycle/reuse alternative
is implemented.
4.1.2 Sampling Plan
Although design data provides a great deal of informa-
tion concerning system characterization, a sampling program is
necessary to supplement this data. Actual operation will
deviate from design condidtions with changes in coal composi-
tion, water composition, additive composition, and load. Since
these four parameters are continuously fluctuating, a sampling
program to characterize system operation under actual conditions
is necessary. Also, demister operation may be more fully char-
acterized with a sampling program. Material balances around the
demister loop may be performed using_ sample data to determine
S02 removal and carryover in the demister.
Two general types of measurements should be made in a
characterization sampling program. These are analytical and
-137-
-------
'o
TABLE 4-3. SOLIDS CONCENTRATION DATA
I. Clarifier/Filter (if applicable)
A. Feed rate GPM
B. Bottoms suspended solids wt. %
C. Overflow suspended solids ppm
II. Pond System
A. Number of Ponds
B. Surface area of each pond
acres
C. Settled sludge (bottom of pond)
solids concentration wt. 70
D. Pond overflow rate GPM
E. Pond overflow suspended solids ppm
F. Pond overflow recycled to
scrubbing system GPM
G. Pond overflow discharged GPM
-138-
-------
process measurements. Analytical measurements can be used to
characterize the chemistry of the scrubbing system by determining
solids precipitation rates, S02 removal rate, S02 oxidation and
scaling potentials. Process measurements can be used to charac-
terize the system water balance for different operating condi-
tions by determining flow rates for key process streams and
levels for major process tanks.
4.1.2.1 Analytical Measurements
Chemical analysis and sampling schedules can be
divided into two categories, each serving a different function.
These categories are system characterization measurements and
line-out measurements. System characterization samples should
be taken after sufficient time is allowed for the system to
reach steady-state operation. Steady-state operation is deter-
mined from the line-out measurements. When the system has
reached steady-state, the stream compositions will not change
appreciably between sample periods.
The characterization sample points and analyses for a
typical scrubbing system with a recirculating demister wash sys-
tem as shown in Figure 4-2 are presented in Table 4-4. The
analyses to be performed on these samples are divided into three
categories: gas analyses, liquid analyses, and solids analyses.
Gas analyses on the flue gas and stack gas for S02 , C02 , 02 ,
H20, and HC1 will determine the S02 removal rate, HCl removal
rate, and the evaporation rate in the scrubbers.
Liquid analyses are shown in Table 4-4 for the remain-
ing sample points. It should be noted that some deviation from
the sampling plan shown in Table 4-4 should be expected for a
particular scrubbing system. If a system uses a once-through
demister wash with makeup water, sample points 10 and 11 may be
eliminated. Also, if a clarifier is not used, sample points 7
and 8 may be eliminated.
Not all of the liquid analyses listed in Table 4-4
necessarily need to be made for each sampling period. Some vari-
ables, such as the magnesium, sodium, nitrate, and chloride
liquor concentrations, should not change significantly under
normal operation. Changes in makeup water quality, alkali addi-
tive, coal composition, or load are the major factors influencing
the magnesium, sodium, nitrate, and chloride levels. Whenever
step changes in these variables are made, complete liquor analy-
ses should be made until the system is at steady-state (line-out
measurements).
-139-
-------
-p-
o
i
-STACK GAS
DEMISTER
WASH
TANK
- .^y
SCRUBBER FEED(S)
MAKEUP
WATER
TO POND. CLARIFIER,
OR REACTION TANK
SCRUBBER EFFLUENT(S)
CLARIFIER
UNDERFLOW
"-DISCHARGE
SLUDGE
Figure 4-2. Sample points for scrubbing system characterization.
-------
TABLE 4-4, SCRUBBING SYSTEM CHARACTERIZATION MEASUREMENTS
SAMPLE POINTS AND ANALYSES
Flue Gas
Stack Gas
Scrubber Feed(s)
Scrubber Effluent(s)
Alkali Slurry
Makeup Water
CJarifler Underflow"
Clarifler Overflow"1
Pond Return
Demister Wash 5
Catch Tray Overflow6
Sample Gas Analyses
S02 02 O2 |I20 MCI
No .
1 X X X X X
2 X XX
3
4
5
6
7
8
9
10
11
Liquid Analyses
Ca
X
X
X
X
X
X
X
X
X
MB'
X
[X
X
X
[X
[X
IX
X
[X
Na1
X
X
X
X
X
X
X
X
X
Cl1
X
XI2
X
X
X)2
X]2
X]2
X
X]3
C02
X
X
X
X
X
X
X
X
X
N03'
X
[X]2
X
X
IX]2
[X]2
[X]2
X
[X]3
S02
X
X
X
X
X
X
X
X
X
Total
s
X
X
X
'X
X
X
X
X
X
Solids Analyses
%
Solids Ca Mg 2 2 '
X X X X X X
X [XXX X XJ2
X XXX
X [X X X X X]2
X
X
X X X X X X
X X X X X X
These analyses should be performed periodically sln^e they should not change rapidly.
2These analyses can be deleted if differences between these points and sample point 3 are negligible.
3These analyses can be deleted in differences between this point and sample point 10 are negligible.
'"'Ihese sample points may be eliminated for systems which dn not use clarlfiers or thickeners.
srhese sample points may be eliminated for systems which use once-through demlster wash with makeup water.
-------
Solids analyses are shown in Table 4-4 for the scrubber
feed, scrubber effluent, alkali slurry, clarifier overflow,
demister wash, and catch tray overflow streams. In addition,
total suspended solids analyses are shown for the clarifier
overflow and pond return streams.
To characterize a scrubbing system, a sampling program
should last from 1-2 weeks for a given set of operating condi-
tions to allow complete system line-out. The actual time
required will depend on the stability of the operating variables
and the process design. Larger reaction tank residence times
will require a longer period of time for the system to reach
steady-state. Line-out samples to determine if the system is
at steady-state should be taken twice daily.
The cost for conducting such a sampling program will
vary directly with the length of the sampling program and the
number of sampling points and analyses to be made. Meaningful
cost estimates for a characterization sampling program can only
be made for a specific system after a detailed program is out-
lined. The addition of sampling ports and/or laboratory capa-
bilities at a specific site should be considered as well as the
manpower requirements and the number of samples and analyses to
be made.
4.1.2.2 Process Measurements
In order to properly characterize the performance of
a scrubbing system, certain process measurements must be gathered
in conjunction with the chemical analyses discussed in the pre-
vious section. These process measurements include such vari-
ables as liquor flow rate, flue gas flow rates, various tank
levels, and important stream pH's and temperatures. Table 4-5
indicates which process measurements are required for system
characterization purposes.
Several of the desired flow rates may be monitored on
a regular basis by plant personnel. These may include scrubber
feed, clarifier feed, and alkali additive. Before starting a
sampling program all existing flow meters should be calibrated
by using variations in tank levels with time. Periodic calibra-
tions of flow meters should be made throughout a sampling
program.
-142-
-------
TABLE 4-5. PROCESS DATA REQUIREMENTS
Stream Name
Gas Streams
Flue Gas
Stack Gas
Liquid Streams
Scrubber Feed
Scrubber Bottoms
Clarifier Feed
Clarifier Overflow
Reaction Tank
Effluent
Pond Return
Alkali Slurry
Makeup Water
Demist er Wash
Catch Tray Overflow
Demister Wash Loop
Slowdown
Tank Level Data
Vessel
Additive Tank
Reaction Tank
Demister Loop Tank
Surge Tanks
Flow Rate
X
-
X
-
X
X
X
X
X
X
X
-
X
Level
X
X
X
X
Pressure pH
X
X
X
X
X
X
X
X
X
X
X
X
X
Temperature
X
X
X
X
X
X
X
X
X
X
X
X
X
-143-
-------
Important streams whose flows are not already moni-
tored should be included in a comprehensive characterization pro-
gram. The flue gas rate can be measured with a pitot tube
placed in the scrubber gas feed duct. Necessary correlations of
flow rate versus pitot reading can be generated by performing
periodic pitot traverses of the gas duct. This measurement
should be made as close to the scrubber inlet as possible to
minimize errors due to leakage.
Slurry stream flow rates may be monitored by magnetic
flow meters. Additional flow meters should be obtained for
slurry streams whose flows are not already monitored in the
scrubbing loop (scrubber feed, clarifier feed, reaction tank
effluent, demister wash loop blowdown) . The flow meters can be
used for the pilot studies discussed in Section 4.3 as well as
for the system characterization phase.
Water makeup streams (to pump seals or demister wash
tank) should be equipped with rotameters. The rotameters (both
existing and new) should be calibrated before starting the
characterization sampling. Intermittent flows such as periodic
washing of the catch tray may be characterized by noting the
frequency of washing.
The flows around the clarifer and pond are not as
critical as the streams previously discussed for purposes of
characterizing scrubbing system operation. Solids balances and
pump characteristics can be used to estimate clarifier under-
flow, clarifier overflow, and pond return rates. The combined
flow of these three streams is normally small in comparison to
other streams entering the reaction tank. For this reason, any
errors resulting from imprecise flow measurement of these
streams will have a negligible effect on material balances per-
formed around the reaction tank.
Temperature and pH measurements are listed for all of
the liquid streams in Table 4-5. This should be done routinely
as a part of the sampling procedure. Stream temperatures within
the scrubbing loop should remain relatively constant. Some heat
losses occur through pipe and open vessels, however. Initially
all stream temperatures should be recorded. As the sampling
program proceeds, it may be possible to omit some temperature
measurements if only small differences exist.
-144-
-------
_ Tank levels should be recorded routinely. In combina-
tion with water makeup rates, the slurry levels of the larser
vessels will determine the magnitude of the system's water re-
quirements. Increases or decreases in tank slurry levels will
affect the calculation of water evaporation and addition rates
through material balances.
4.2 Phase II: Alternative Evaluation
The second phase of a scrubbing implementation plan
is to formulate and evaluate various modes of operation. For-
mulation of alternative operating conditions will depend on how
the scrubbing system fits into the overall plant recycle/reuse
scheme. ^The plant studies conducted in this program showed that
alternatives include converting from open-loop to closed-loop
operation and changing makeup water source. Alternative makeup
water sources may include cooling tower blowdown or ash pond
overflow.
To determine the feasibility of implementing a recycle/
reuse alternative, both potential scale formation in the demister
and the TDS level (including chlorides) of the scrubbing liquor
should be investigated.
The plant studies portion of this project showed that
an effective tool for performing alternative evaluations is a
process simulation computer model package. A process simulation
package can also be used to evaluate the consistency of the data
collected in the system characterization phase of an implementa-
tion plan. This section first discusses the calculations required
to evaluate scrubber alternatives and then presents a methodology
for using a model to perform the evaluations.
4.2.1 Evaluation Criteria
The scrubbing evaluation should involve mass and energy
balances to calculate flows, compositions, and temperatures of
all process streams. An overall material and energy balance in-
cludes a humidification calculation to determine the amount of
water evaporated by the flue gas, and a prediction of solid-liquid
equilibrium for the solid waste. The important solid-liquid equi-
libria include CaC03, CaS0^2H20, and CaS03'%H20. Of the remain-
23ESE-S: SH£S-£S 5;
-145-
-------
overall balance and the suspended solids concentrations of all
the streams. Calculations around the scrubbing vessel include
the prediction of gas-liquid equilibria between the C02 and S02
in the gas and liquid streams in the scrubber. A common assump-
tion for CO2 is that C02 is transferred between the scrubbing
liquor and the gas so that the liquor equilibrium partial pres-
sure of C02 is equal to the C02 partial pressure in the gas. A
scrubbing process evaluation should include this equilibrium so
that the amount of C02 entering or leaving the system through
the scrubber may be determined.
The evaluation should also include prediction of the
equilibrium partial pressure of S02 above the scrubbing liquor.
Calculation of this parameter will allow an evaluation of the
ability of a given scrubbing liquor to remove a specified amount
of S02 from the flue gas. Any changes in liquor composition re-
sulting from the implementation of a recycle/reuse scheme may be
evaluated in terms of S02 removal capability by comparing liquor
equilibrium S02 partial pressure to the outlet gas S02 partial
pressure. The prediction will allow an estimate of the S02 re-
moved in the demister to be evaluated. The partial pressure of
S02 above the demister wash should be less than the partial pres-
sure of S02 in the stack gas.
The calculations outlined above will allow a particular
scrubbing recycle/reuse option to be evaluated in terms of scale
potential and dissolved solids content of the scrubbing liquor.
The following sections present discussions of the prediction of
scale potential and TDS levels in a typical scrubbing system.
4.2.1.1 Scale Potential
The most important factor which determines the feasi-
bility of implementing a scrubbing recycle/reuse alternative is
the potential for scale formation in the demister. If significant
scale deposits form, the system must be shut down for cleaning.
In the plant studies portion of this study, the concept
of relative saturation was used to predict scaling potential.
When the relative saturation of a species is below 1.0, no poten-
tial for solids precipitation exists. When the relative satura-
tion is greater than 1.0, the species is supersaturated and scale
may form. In the presence of seed crystals, the relative satura-
tion may exceed 1.0 without scale formation. However, as relative
saturation increases above 1.0, a critical value is reached where
nucleation occurs. When this happens, scaling may occur since
conditions favorable to the creation of new crystal nuclei also
tend to produce scale.
-146"-
-------
Past experience has shown that calcium sulfate dihy-
drate will form chemical scale when its relative saturation
reaches 1.3-1.4. A scaling limit also exists for calcium sul-
fite hemidydrate, but it is not as well defined as the limit
for gypsum. A value of at least three to four times the sulfite
saturation level (R.S. = 1.0) is necessary to initiate sulfite
scaling. Some systems have been reported to operate at six to
seven times the sulfite saturation value without scaling. A
detailed definition of relative saturation was given in Section
2.2.1.1.
If a recirculating demister wash system is used, the
evaluation of potential scale formation in the demister should
include a prediction of solid precipitation and dissolution rates
in the demister wash tank. The concept of relative saturation
may be used to calculate rates as follows:
R = KafCV (R.S. - 1)
where
. R = precipitation or dissolution rate
K = a temperature dependent constant
a = crystal interfacial area
f = weight fraction of the considered
species in the solid phase
C = total suspended solids concentration
V = tank volume
R.S. = relative saturation of considered
species
It is important that the scrubbing process evaluation
include rate calculations and determination of the relative sat-
urations of important species in the system. The effects of pro-
cess modifications must be evaluated in terms of these parameters
to insure that implementation of a water recycle/reuse option
will not result in scale formation in the demister.
-147-
-------
4.2.1.2 Dissolved Solids Concentration
If the scrubber makeup water is high in dissolved
solids, excessive TDS levels in the scrubbing system may be en-
countered. The water evaporated in the scrubbing system causes
the dissolved solids in the makeup water to be concentrated in
the scrubbing loop.
Excessive TDS levels in the scrubber liquor can in-
crease the energy requirements for pumping the liquor and can
decrease the mass transfer characteristics in the scrubber. As
the TDS level increases, the viscosity of the scrubbing liquor
will increase. For example, the viscosity of a 25 wt. % NaCl
solution is about twice that of pure water while a 5 wt. °L solu-
tion has a viscosity only about 10% higher than pure water. As
the viscosity of the scrubbing liquor increases, the energy re-
quirements for pumping the liquor will increase.
The TDS level will also affect the surface tension of
the liquor. Excessive TDS levels may cause larger droplet sizes
in a spray tower. This may reduce the mass transfer characteris-
tics of the droplets and adversely affect scrubber performance.
The level of chloride present in the makeup water may
also limit the use of an alternate makeup source in a scrubbing
system. If the chloride level in the scrubbing liquor is too
high, corrosion problems may be encountered.
The material and energy balances discussed previously
will allow prediction of the TDS and chloride levels expected
under alternate operating conditions. The limitations discussed
here should be considered in evaluating the feasibility of con-
verting to closed-loop operation or of changing makeup water
source.
4.2.2 Model Application
As part of a recycle/reuse implementation plan, a pro-
cess model may be used for two purposes:
1) check consistency of data obtained
in characterization phase, and
2) evaluate and optimize alternative
operating conditions.
-148-
-------
The data collected both from design information and the sampling
program can be used as inputs to the model to calculate makeup
water requirements, precipitation rates, and scaling potentials
in the system. The values determined by the model may then be
compared to the measured values obtained from the sampling pro-
gram to check the data consistency and the validity of the pro-
cess model. In addition to comparing stream compositions and
pH's, the calculated equilibrium S02 partial pressure above the
scrubber effluent stream should be compared to the measured S02
partial pressure in the stack gas. If the measured stack gas S02
concentration does not exceed the calculated S02 equilibrium
partial pressure, a gas or liquid phase anlaytical error is in-
dicated. In this case, the analytical data should be re-examined.
Once the data consistency checks have been made and the
process model has been verified, simulations of the scrubbing sys-
tem under alternative operating conditions should be made. Alter-
native simulations should be performed for a variety of conditions
under which the scrubbing system is expected to operate. These
should include different loads and different coal compositions as
well as variations in makeup water quality. The design and opera-
ting data collected in the characterization phase of the implemen-
tation plan should be used to identify the magnitude of variations
expected for the scrubbing system.
Simulations for both typical operation in the alterna-
tive mode and for worst case operation should be made. In the
case where cooling tower blowdown or a combination of fresh makeup
water and cooling tower blowdown is being considered as scrubber
makeup water, several simulations should be performed to optimize
the system configuration or ratio of fresh water to cooling tower
blowdown. This is particularly important for scrubbing operations
with recirculating demister wash systems. The level of S02
removal oxidation, and carryover in the demister section should
be investigated with simulations to determine the long-term
effects of these parameters on system operation.
In addition to investigating long range variations in
operating conditions, simulations to determine the effects of
short-tera upsets should be made. Short-term upsets may include
changes in S02 removal, oxidation, flue gas temperature, and
makeup water Quality. For example, in a cascaded water system
where cooling'tower blowdown is used as scrubber makeup, any up-
sets which occur in the cooling system will cause a change in
the scrubber makeup water quality.
-149-
-------
The results of these simulations should identify opera-
ting conditions which can be expected under a recycle/reuse alter-
native. Stream temperatures, flows, and compositions will be
specified by the model and should allow evaluation of the feasi-
bility of the recycle/reuse option with respect to system scaling
potential and TDS levels. The simulation results may be used as
a basis for designing and process modifications or for designing a
detailed test program for further investigations on a pilot or
prototype scale. The number of simulations required and there-
fore the cost of performing the simulations will depend upon the
type of modifications considered and the stability of the parti-
cular system under consideration. If the scrubbing system is
subject to wide variations in operating conditions, more simula-
tions should be performed to fully evaluate the effects of insti-
tuting a water recycle/resue option. Meaningful cost estimates
can only be made for a particular system undergoing a specific
recycle/reuse modification(s).
4.3 Phase III: Pilot-Scale Studies
After the characterization and simulation phases of an
implementation plan have been completed, pilot-scale testing
should be performed to further analyze the effects of the water
recycle/reuse scheme on scrubber operation. Two types of scrub-
bing system modifications have been identified as parts of a plant
water recycle/reuse scheme. These are converting from open-loop
to closed-loop operation and using a cascaded water stream as
scrubber makeup. This section discusses each of these two types
of scrubber modification separately.
4.3.1 Converting to Closed-Loop Operation
One method identified for minimizing the water require-
ments of a SOa/particulate scrubbing system is to operate the
system in a closed-loop mode. In a closed-loop system all the
excess water leaving with the solid waste is recycled to the
scrubbers. The only water lost with the solids will be that
occluded with the sludge. If a pond is used, the sludge will
be about 50% water. If a vacuum filter is used for solids con-
centration, the sludge may be as low as 3578 water for systems
with high oxidation (either forced by air sparging or naturally
occurring).
Since scrubbing systems are not designed with a sepa- '
rate clarifier, filter, or pond for each module, the isolation
of an individual module is not feasible for testing the effects
of converting from open-loop to closed-loop operation. Either a
pilot scale or one-module prototype installation is required to
-150-
-------
test closed-loop operation. Additional piping will be required
to return the pond overflow for systems using a pond for solid/
liquid separation. For scrubbing systems with high oxidation
a vacuum filter may be installed to effectively dewater the
sludge to about 65% solids. Piping would then be necessary to
return the filter supernatant to the system. Converting to
closed-loop operation will lower the system makeup water require-
ments and, therefore, affect the demister water balance. If a
once-through demister is used, additional water to supplement
the decreased makeup can be taken from the clarifier overflow.
If the evaluation phase showed this option not to be feasible,
a recirculating demister system may be installed. This would
involve adding a catch tray to the scrubber internals, a demister
wash tank, and additional piping and pumps. The cost of the new
equipment will depend on the size of the system and the particu-
lar process option chosen (once-through or recirculating demister)
Once the process modifications have been made, the
pilot or prototype system may be started up. As with the char-
acterization sampling scheme, line-out samples should be taken
until the system has reached steady-state. The time required to
reach steady-state will depend on the particular system design.
Larger reaction tanks and the use of a pond will increase the
system residence time and, therefore, require longer line-out
time.
The sampling strategy outlined in Section 4.1.2 also
applies to pilot testing. Closed-loop operation should be tes-
ted under a variety of conditions to insure safe operation with
the modified system. These conditions will have been identified
by the characterization phase as being likely to occur. Changes
may involve variations in flue gas flow, S02 concentration, and
makeup water composition.
As was illustrated in Table 4-4 of the characterization
section eas analyses are recommended for the flue gas entering
the system and the stack gas leaving the scrubbing system. Com-
plete liquor analyses are recommended for the makeup water, clar-
ifier overflow, and pond return Liquid and solid analyses are
recommended for the scrubber feed scrubber effluent alkali
slurry, clarifier underflow, and demister loop slurries^Liquid
analyses should include calcium magnesium sodium chloride,
carbonate, nitrate, sulfite, and total sulfur Solid P^ase
analyses should include calcium, magnesium, carbonate, sulfite,
and sulfate.
-151-
-------
In many cases the soluble species (magnesium, sodium,
chloride, and nitrate) may not change appreciably throughout the
system since system liquid phase residence times are much shorter
than solid phase residence times. If preliminary results indi-
cate this, then some of the liquid phase analyses may be elimi-
nated or performed less frequently. Notes to this effect are
listed at the bottom of Table 4-4.
Operation of the pilot or prototype facilities in a
closed-loop mode over the range of operating conditions to be
expected on a full-scale will establish a strong basis upon which
the full-scale modifications may be based. Effects on the scrub-
bing operation as well as the optimum demister configuration for
closed-loop operation can be assessed with the results from the
pilot or prototype studies.
4.3.2 Changing Makeup Water Source
As discussed previously for converting to closed-loop
operation, isolation of an individual module of a scrubbing sys-
tem may not be practical to study system modifications. Again,
a one-module prototype or pilot-scale installation is ideal for
testing this water recycle/reuse scheme.
Additional piping and a makeup water mixing tank will
be required in order to study the use of a mixture of existing
makeup water and an alternative source (cascaded water) as scrub-
ber makeup. Using a mixture of makeup waters will allow the op-
timum level of use of cascaded water as makeup to be determined.
It is recommended that four or five different ratios of makeup
water sources be used and the scrubbing system chemistry be char-
acterized under each condition. Scaling potential should be
evaluated for the demister wash loop as well as for the main
scrubbing loop of the system.
The sampling strategy outlined for the characterization
phase of implementation may be used in this phase also. Sample
points and analyses were tabulated in Table 4-4 of the character-
ization section.
The results of these pilot or prototype studies will
allow the design of a full-scale modification to use cooling
tower blowdown (or other cascaded water) or a mixture of the
tower blowdown with fresh makeup water. The pilot studies will
identify any process modifications necessary to prevent scaling
in the demisters. They will also identify the optimum mixture
-152-
-------
of makeup waters to allow reliable system operation if the new
makeup water source may not be used exclusively.
4.4 Phase IV: Full-Scale Operations
The final result of the pilot studies will be a system
design which will allow closed-loop operation or substitution of
an alternate makeup water source. Either one of these process
modifications may require changing the demister loop design to
prevent scale formation in the demister. Converting a once-
through demister system to a recirculating system will involve
installing a catch tray under the demister so that most of the
wash water may be recycled. Additional tanks, pumps, and piping
will also be required.
After the equipment modifications have been made, it is
recommended that the system operation be monitored closely until
steady-state is achieved. The same sampling strategy outlined in
the characterization phase should be used to perform a final eval-
uation of the full-scale modified system. The frequency of sam-
pling may be decreased somewhat since the line-out time required
for a full-scale system using a pond will be significantly longer
than a pilot or prototype installation using a vacuum filter.
After reliable operation is established, a regular sampling pro-
gram should be established to monitor system performance. This
program will be especially important when the plant is undergoing
a step change such as changing the type of coal fired or the
system load.
-153-
-------
5.0 SUMMARY
Generalized implementation plans have been presented
for recycle/reuse alternatives identified in the plant studies
conducted in the first part of this program. Recycle/reuse op-
tions were identified for cooling tower, ash sluicing, and SOz/
particulate scrubbing systems at coal-fired power plants.
Options for cooling towers included increased recircu-
lation resulting in lower blowdown and makeup rates. Increased
recirculation in cooling towers may require the addition of pH
control with sulfuric acid or softening (either makeup or slip-
stream) to control gypsum scale potential. The implementation
plan discussed was divided into three phases: 1) system charac-
terization, 2) alternative evaluation, and 3) full-scale modifi-
cations. This plan presented methodologies for identifying
operating variables and using these variables to evaluate scaling
potential and treatment requirements for alternate operating
conditions. The equipment and sampling requirements for imple-
menting full-scale changes in cooling tower operation were also
presented.
Recycle/reuse options identified for ash sluicing
operations include converting to completely or partially closed-
loop operation. These options may require treatment to prevent
scaling. Possible treatments identified include 1) the use of a
reaction tank to allow ash dissolution and solids precipitation
to prevent scaling in the sluice line and 2) the use of softening
to reduce the dissolved calcium in the pond recycle. A plan was
presented whereby a recirculating ash sluice system may be imple-
mented. Four phases were identified 1) ash characterization,
2) alternative evaluation, 3) pilot studies, and 4) full-scale
operation. The first phase discussion presented a methodology
for identifying ash reactivity with bench-scale experiments. The
second phase presented a methodology for evaluating the feasibi-
lity of a recirculating system based on scale potential and treat-
ment requirements. Pilot studies (third phase) were recommended
for ash sluicing to better define operating parameters which are
not easily predicted and to evaluate treatment options. The full-
scale phase discussion included descriptions of equipment required
and important operating variables.
For S02/particulate scrubbing systems, two potential
recycle/reuse options were identified in the plant studies.
These included converting to closed-loop operation and using an
alternate makeup water source such as cooling tower blowdown or
-154-
-------
ash pond overflow. A general implementation plan has been pre-
sented for instituting these types of modifications. As with
ash sluicing, four phases were identified: 1) system character-
ization, 2) alternative evaluation, 3) pilot studies, and 4) full-
scale modifications. The first phase discussion presented the
important operating variables to be considered and a sampling
plan for obtaining operating data. The second phase presented a
methodology for evaluating the feasibility of recycle/reuse op-
tions. The pilot studies (third phase) presented a plan for
testing operation with a recycle/reuse option. Potential prob-
lem areas are demister scale formation and TDS level in the
scrubbers. The full-scale discussion described the equipment
necessary and recommended a sampling strategy to be used to
monitor system performance.
-155-
-------
REFERENCES
BA-465 Babcock and Wilcox Company, Steam, its generation and
use, 37th ed., N.Y., 1955.
CH-335 Chu, Tien-Yung J., Peter A. Krenkel, and Richard J.
Ruane, "Reuse of ash sluicing water in coal-fired
power plants", in Proceedings of the Third National
Conference on Complete Water Reuse: Symbiosis as a
Means of Abatement for Multi-Media Pollution, Cin-
cinnati, June 1976, Lawrence K. Cecil, ed., N.Y.,
AIChE, 1976.
DE-165 Derrick, A. E., "Cooling pond proves to be the econo-
mic choice at Four Corners", Power Eng. 1963 (Nov.),
46.
DI-057 Dickey, Joe Ben, Jr. and Robert E. Gates, Managing
waste heat with the water cooling tower. 2nd ed.
Mission, Kansas, Marley Co., 1973.
DI-170 Dionex Corporation, Analytical ion chromatography,
Models 10 and 14, operation and maintenance manual.
Palo Alto, CA, Jan. 1976.
EL-094 Electrical World Directory of Electric Utilities,
74-75, 93rd Edition. N.Y., McGraw-Hill, 1974.
FE-001 Feitknecht, W. and P. Schindler, "Solubility con-
stants of metal oxides, metal hydroxides, and metal
hydroxide salts in aqueous solution", Pure Applied
Chem. 6, 130-160, 197-99 (1963).
-156-
-------
FE-102 Federal Power Commission, Steam-electric plant air
and water quality control data for the year ended
December 31, 1972. Atlanta, Georgia, 1973.
FU-001 Fuoss, R. M. , "Ionic association, pt. 3, the equili-
brium between ion pairs and free ions", J. ACS 80.
5059-61 (1968).
GH-001 Ghosh, S. R. , S. C. Ghosh and D. Roy, "Nitrogen Oxides
Absorption. Design of Packed Towers", Technology 2
(3), 149-59 (1965); C. A. 65: 3358d.
GJ-001 Gjaldbaek, J. K. , "Untersuchungen uber die loslichkeit
des magnesium-ydroxyds", II. Die loslichkeitsprodukte
und die dissoziationskonstante der magnesium hydroxyde",
Z. Anorg. Chem. 44, 259-88 (1925).
GU-075 Guthrie, K. M. , Process plant estimating evaluation
and control. Solana Beach, CA, Craftsman Book Co.,
1974.
JA-105 Jaulmes, P. and S. Brun, "Solubilisation et precipi-
tation des sels peu solubles d'acides moyens ou
faibles", Trav. Soc. Pharm. Mont. 25(2), 98 (1965).
KL-052 Klein, David H., Melvin D. Smith, and James A. Driy,
"Homogeneous nucleation of magnesium hydroxide",
Talanta 14, 937-40 (1967).
KL-053 Klein, David H. and James A. Driy, "Heterogeneous
and homogeneous nucleation of strontium sulphate",
Talanta 13, 289-95 (1966).
-157-
-------
KL-054 Klein, David H. and Louis Gordon, "Nucleation in
analytical chemistry, II. Nucleation and preci-
pitation of silver chloride from homogeneous
solution", Talanta 13, 177-86 (1959).
LE-201 Leung, Paul and Raymond E. Moore, "Water consumption
study for Navajo Plant", in Proceedings of the Ameri-
can Society of Civil Engineers, Power Division, 97
(P04), 8564 (1971).
LE-239 Lepper, F. R., Jr., Private communication, Resources
Conservation Co., 3 June 1976.
LI-001 Linden, J. Van der, J. Soc. Chem. Ind. 36, 96 (1917).
LI-068 Linke, William F., comp., Seidell's solubilities.
Inorganic and metal-organic compounds, 4th ed. ,
2 vols., Princeton, NJ, D. Van Nostrand, 1958, 1965.
MC-136 McGlamery, G. G., et al., Detailed cost estimates
for advanced effluent desulfurization processes.
Interagency Agreement EPA IAG-134 (D), Pt. A.
Research Triangle Park, NC, Control Systems Lab.,
NERC, 1974.
NA-205 National Coal Association, Steam-electric plant
factors, 24th ed., WAshington, B.C., 1974.
NE-107 Nelson, Guy R., Water recycle/reuse possibilities:
power plant boiler and cooling systems, final report.
EPA 660/2-74-089. Corvallis, Oregon, National Envi-
ronmental Research Center, Thermal Pollution Branch,
September, 1974.
-158-
-------
OT-001 Ottmers, D. M. , Jr., Simulation of NAPCA venturi
system. Technical Note 200-002-4, Austin, Texas,
Radian Corporation, November 1969.
PA-121 Patterson, W. C. , J. L. Leporati, and M. J. Scarpa,
"The capacity of cooling ponds to dissipate heat",
Proc. Amer. Power Conf. 33, 446 (1971).
PA-227 Palmer-Hostik & Assoc., Private communication,
Houston, Texas, 2 August 1976.
PE-161 PEDCo-Environmental Specialists, Inc., Summary
Report-Flue Gas Desulfurization Systems - April
1975. EPA Contract No. 68-02-1321, Task No. 6,
Cincinnati, Ohio, 1975.
PE-R-277 Perry, John H. , Chemical engineers handbook, 5th
edition, New York, McGraw-Hill, 1973.
RE-211 Resources Conservation Co., Feasibility study for
Arizona Public Service Company Four Corners power
station. Renton, Washington, April 13, 1976.
RO-266 Rogers, A. N. and L. Awerbuch, "Reuse or disposal
of salts from power and desalination plant wastes",
in Proceedings of the 3rd National Conference on
Complete Water Reuse: Symbiosis as a Means of
Abatement for Multi-Media Pollution, Cincinnati,
June 1976, Lawrence K. Cecil, ed., N.Y., AIChE, 1976
TH-192 Thompson C. G. and G. A. Mooney, Recovery of lime
and magnesium in potable water treatment. EPA 600/
2-76-285, Montgomery, Alabama, Black, Crow, and
Eisdness, Inc., December 1976.
-159-
-------
UH-007 Uhlig, Herbert H., Corrosion and corrosion control,
an introduction to corrosion science and engineering.
N.Y... Wiley, 1963.
-160-
-------
TECHNICAL REPORT DATA
If lease read Inuntctions on the reverse before completing!
EPA-6QO/7-78-055a
4. TITLE AND SUBTITLE water Recycle/Reuse Alternatives in
Coal-fired Steam-electric Power Plants: Volume I.
Plant Studies and General Implementation Plans
5. REPORT OATS
March 1978
6. PERFORMING ORGANIZATION CODE
. RECIPIENT'S ACCESSION NO
7. AUTHOR(S)
James G. Noblett and Peter G. Christman
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
P.O. Box 9948
Austin, Texas 78766
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-03-2339
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 6/75-2/78
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES jERL-RTP project officer is Frederick A. Roberts, EPA/ERC,
200 S. 35th St. , Corvallis, OR 97330 (503/420-4715).
is. ABSTRACT
repOrt gives results of an investigation of water recycle/treatment/
reuse alternatives in coal-fired power plants. Five power plants from representative
U.S. regions were studied. The major water systems encountered were cooling, ash
sluicing, and SO2/particulate scrubbers. Results were used to provide general im-
plementation plans for the various options identified. Computer models were used
to identify the degree of recirculation achievable in each water system without for-
ming scale. The effects of makeup water quality and various operating parameters
were determined for each water system. Several alternatives for minimizing water
requirements and discharges were studied for each plant, and rough cost estimates
were made for comparison. An implementation plan is presented for each water sys-
tem and is divided into phases, including system characterization, alternative eval-
uation, pilot studies , and full-scale implementation. This volume discusses the
recycle /treatment/reuse opportunities for cooling, ash sluicing, and SO2/particu-
late scrubbing systems as well as combined systems. It also includes the implemen-
tation plans .
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Combustion
Water Treatment Cooling Water
Filtration Scrubbers
Circulation Sulfur Oxides
Electric Power Plants Dust
Coal Mathematical Model
Pollution Control
Stationary Sources
Water Recycle/Reuse
Ash Sluicing
Particulate
13B
07D
10B
21D
21B
13A
07A
07B
11G
12A
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report I
Unclassified
21. NO. OF PAGES
188
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-161-
-------