NEUTRALIZATION OF HIGH FERRIC IRON ACID MINE DRAINAGE
                         By
        Roger C. Wilmoth and Robert B.  Scott
          U. S. Department of the Interior
        Federal Water Duality Administration
          Norton Mine Drainage Field Site
               Norton, West Virginia
    Eaper Presented Before the Third Symposium
          on Coal Mine Drainage Research
                 Mellon Institute
             Pittsburgh, Pennsylvania
                   May 19, 1970

-------

-------
                NEUTRALIZATION OF HIGH FERRIC  IRON ACID MINE DRAINAGE

                         Roger C. Wilmcth and  Robert  B. Scott

                           U.  S.  Department  of the Interior
                         Federal  Water yuality Administration
                           Norton Mine Drainage Field Site
                        Box 60, Norton, West Virginia  26285

      For the past  year and one half,  the Norton Mine Drainage  Field Site  of the  Fed-
 eral Water Quality Administration has been  conducting research on the  feasibility of
 acid mine drainage treatment  using lime, limestone,  and soda ash neutralization.

      The Norton research facility is  located  at Norton,  West Virginia, on the  banks
 of Grassy Runa small,  heavily  polluted stream of which an estimated  90  percent  of
 the  water flow is  from abandoned mines.   In Table 1,  the water quality of Grassy Run
 is presented.   This water was used for most experiments  reported in this  paper.

                                       Table 1
                            Grassy Run Water (Duality  - 1968

   Parameter                       Units        Max.        Min.         Mean

   pH                                            3.1        2.1          2.8
   Specific Conductance            Mmhos/cm       1800        600          1190
   Hot Acidity (as CaCC^)          mg/1         680        137          466
   Calcium (as CaCC^)               mg/1         360          94          211
   Magnesium (as CaCOo)             mg/1         140          90            93
   Hardness (as CaCO-J              mg/1         470        134          304
   Iron  (Total)*                   mg/1         200          10            93
   Aluminum                        mg/1          70          10           31
   Sulfates              /,v        mg/1         980         270          621
   Flow  (Water Year 1968)v  '          cfs         2?8        1.2          7.52

   *Ferric/Ferrous Ratio Greater than 4:1

      The advent of stronger pollution laws  and the shortage  of reliable at-source
 pollution control  techniques makes treatment  a required  abatement method.  Treat-
 ment  by  neutralization appears,  at this  point,  to be the most  feasible  (cost wise)
 of the various  methods.

     Most mine  drainage  neutralization research efforts have been done on ferrous
 iron mine waters.  In contrast, the studies at  Norton have concentrated on water
 with greater than  80 percent ferric iron.

                                  PURPOSE OF STUDY

     The purposes  of this study were:  1. To investigate the effluent water quality
 characteristics  of high  ferric iron acid mine drainage when neutralized to various
 pH's by three types of neutralizing agentslime, limestone, and soda ash,  2.  To
 compare the effectiveness of each of the three neutralizing agents in neutralizing
this acid mine drainage to various pH's,  3- To determine optimum basic operating
 conditions required to neutralize the acid mine drainage in order to produce an ef-
 fluent water of desirable quality,  4.  To acquire relevant operating data at the
 optimum conditions, derived from actual pilot plant operation,  and 5. To enhance the
 efficiencies, if possible, of the limestone reaction.

-------
                      FACILITY LAYOUT AND FLOW DIAGRAM

     A simplified flow diagram of the neutralization test system is presented
in Figure 1.

     Water from Grassy Run was pumped into a 500 gallon fiberglas tank which
was kept constantly full by a continuous overflow.  This tank served as a
constant head source and gravity fed a constant flow of raw water into a neu-
tralization reactor.  Floxv rates through the systen ranged from 1-6 gallons
per minute.  Control of this flow was accomplished by means of a globe valve
(other type valves fouled or allowed the water flow rate to vary excessively).
The reactor xvas a 55 gallon stainless steel tank on which a flash mixer was
mounted.  The rapid mixing was responsible for some short circuiting.  Tracer
tests yielded a mean logarithmic probability efficiency of 34.8 per cent.  In-
to the reactor, a neutralizing agent could be fed in either the dry or slurry
form.  The dry feeder used was a volumetric type disc feeder which performed
reliably and had a feed accuracy of + 2 per cent.  The slurry feed system con-
sisted of a storage tank exactly like the reactor tank with a flash mixer.
Constant agitation was supplied by the mixer to keep the relatively insoluble
neutralizing agent in suspension.  Small precise chemical feed pumps were used
to transfer the slurry from the slurry storage tank into the reactor.  These
pumps were electrical solenoid type, rapid impulse, diaphragm pumps chosen for
their ability to pump the heavy slurry accurately.

     The reactor contained a pH probe for a recording pH meter.  This meter was
used for precise monitoring and control of the system.

     Water level in the reactor was maintained by float switch operated punpji
which transferred the neutralized water from the reactor into a 1250 gallon
upflow settling tank.  Tracer tests of the settling tank yielded a mean loga-
rithmic probability efficiency of 80 per cent.

     Supernatant from the settling tank, which was considered treated water,
overflowed into a sampling line and then was returned to Grassy Run.  The slttdge
from the settling tank could be either periodically or continually withdrawn
and transferred to a holding pond for periodic disposal.

     The water was filtered prior to treatment with a pressurized rapid sand
filter to remove large particle matter which tended to clog the valves and thus
vary the flow.  Filtration would not be necessary in full scale plants or large
pilot plants.

                                 PROCEDURES

     Normal operational procedures during a test run involved the following:

     The pH was the only meaningful parameter that could be continuously and
immediately monitored and was used for operational control.   The raw water
flow was controlled by the globe valve which maintained a constant flow of
feed water into the reactor.  By adjusting the neutralizer feed rate, a con-
stant pH was established in the reactor.  Samples were taken from the reactor
after a few hours operation, xvhen the system had come to equilibrium.  How-
ever, several days operation was required to allow the formation of an active
sludge blanket before the settling tank effluent was sampled.  Once the sludge
blanket was established, normal sampling procedures were begun.  Sonnies were
taken from the inflow (raw Grassy Run water),  treated water (immediately from
the discharge of the reactor), and from the settling tank effluent.  Rates of
the raw water inflow and neutralizing agent feed xvere measured and the tonper-
ature of all three water samoles was recorded.  A satmle XVAS also taken to

-------
oo

-------
measure the sludge settling rate and volume.  This sample of treated water
was taken immediately from the reactor discharge and placed in a 1000 ml grad-
uate.  The height of the precipitated sludge was recorded every fifteen minutes
for a period of one hour.  A few long term settling tests were made with per-
iodic readings for total time lengths of 24 hours and longer.

     If the slurry feed method of introducing the neutralizing agent was used,
the slurry mixture was prepared on a batch basis in a separate tank and then
dumped into the slurry storage tank.  A flash mixer constantly agitated the
slurry to keep the relatively insoluble neutralizing agent in suspension.
Electronic impulse diaphragm pumps, whose flow could be varied electronically,
transferred the slurry into the reactor.  Even with these specialized pumps,
a five per cent (by weight) limestone slurry was the maximum concentration (of
limestone) the pumps were able to transfer.  Although lime slurries were much
easier to keep in solution (as calcium hydroxide is much more soluble than
calcium carbonate), fouling problems were also experienced with it.  Studies
were not performed on soda ash slurries.

     When the dry feed method of neutralization was used, control of the neu-
tralizer feed rate was accomplished by adjusting a variable speed transmission
on the dry feeder.

     During aeration studies in the reactor, a 3/4" PVC plastic pipe 'U1 was
inserted in the reactor tank.  The sides of the fU* fit vertically against the
sides of the reactor tank and the bottom of the *U' laid near the bottom of
the tank.  The bottom of the 'U* was drilled with a number of 1/16" holes
spaced to provide an equal distribution of air throughout the reactor.  An air
compressor supplied a regulator which, in turn, controlled the air flow through
the aerator.  The compressor provided excellent intermittent service, but over-
heated in continuous duty situations.

     When aeration of the settling tank was studied, a rectangular aerator,
made from four lengths of 3/4" PVC piping, was suspended above the sludge blan-
ket.  Holes 1/16" in diameter were drilled in the rectangular aerator and
spaced to provide maximum dispersion.  Sufficient air pressure was used to main-
tain adequate agitation of the supernatant.  Initially,  the rectangular aerator
was placed eight inches above the sludge blanket, but the turbulance disturbed
the blanket.  The aerator was then moved to a height of fifteen inches above the
blanket.

Analytical Procedures

     As pH was found to vary with time following neutralization and as the solu-
bility of metal salts in the water is dependent upon pH, it was critical to
establish a standard time of analysis following neutralization.

     A study of pH change following neutralization was conducted to determine
the optimum time for analysis.  Neutralized samples of acid mine water were
agitated by flash mixers for ten minutes.  Then the samples were allowed to
set undisturbed and the supernatant pH periodically recorded.  Figure 2 indi-
cates the pH change with time after the addition of lime, limestone, and soda
ash to identical samples of acid mine drainage.

     In the case of lime, the reaction was rapid and the pH was stabilized
after one hour, thus indicating that the reaction had gone to completion.
However, where limestone was the neutralizing agent, the pH had not stabilized
even after 96 hours.  In the case of soda ash, the pH change curve closely re-
sembles the limestone curve and the reaction required a comparable time to go
to completion.  Therefore,  for our use, a standard procedure of analyzing

-------
                          CJ3

                          =c
                          uu
                          CD
                          /t-^"
                          UJ
                     CNJ
                     cs
                          i_t_g

                          	!

                          O-
                          r"""*" ,
                            .




                          GO






                          CD
CO

-------
samples just prior to 24 hours after sampling was adopted.  At this time, the
lime reaction had gone to completion and the limestone and soda ash reactions
have had what we felt to be the maximum detention time which could be found in
any operational neutralization plant.  Also, this 24 hour standard conformed
well to the work schedule of the chemistry laboratory.

Chemical Analytical Procedures

     Water - Cold acidity was measured by automatic potentiometric titration
to an end point of pH 8.3.  Hot acidity was done by potentiometrically titra-
ting a boiling sample to an end point of pH 8.3.  Iron, aluminum, calciun,
and magnesium determinations were by atomic absorption spectrophotometry.
Hardness was the summation of calcium and magnesium.  Standard Methods (2)
were used for the alkalinity, total residue (TR), and sulfate determinations.
The supernatant was used for these maTyses rather than tN? total sample.
Sludge alkalinity and sludge total residue were also determined by Standard
Methods procedures.^)

Utilization Efficiency

     In order to compare the different methods of neutralization, the effici-
ency of usage of each chemical was computed.  This Utilization Efficiency was
defined as the ratio of Neutralising Agent Used/Neutralizing Agent Added.
This concept includes not only the alkalinity used to counteract acidity, but
also includes the excess alkalinity imparted to the supernatant (which will
be discharged into a stream), thus it was felt that a benefit was derived
from not only the acidity removed but also the alkalinity added to the water.
The alkalinity precipitated with the sludge was considered a loss.  Therefore.,
the formula for determining efficiency was:

           Utilization Efficiency (Per Cent) = (H + A) x 100/ (N x P)

           Where H = Hot Acidity Removed (as CaCOo)
                 A = Alkalinity of Supernatant (as CaCOo)
                 N = Neutralizing Agent Used (CaCC>3 equivalent)
                 P = Per Cent Purity of Neutralizing Agent

Particle Size

     The importance of particle size in the neutralization reaction has been
emphasized in studies by Ford, Young, and Glenn,(3) which indicated that de-
creasing the particle size of limestone increased both the rate and effic-
iency of the neutralization reaction.  They recommend a particle size of 00
mesh or smaller.  Ford and Young tested thirteen limestones from the Eastern
United States and rated the limestone used during these studies (Germany Vrlley
Limestone) as the most efficient in reference to reactivity with acid irine
drainage.  A spectrochemical analysis of this limestone was made by Bituminous
Coal Research and is presented in Table 2.

     Since particle size is so important, sieve tests were made on the three
neutralizing agents used.  These size determinations were made on a standard
sieve shaker with 50, 100, 200, and 400 mesh screens.  All tests were nade by
shaking the sample for one hour and then weighing the portion of the sanple
retained on each screen.

     Initial results from the sieve analyses indicated a larger particle size
than was advertised by the suppliers.  It was suspected that water had been ab-
sorbed by the neutralizing agents and was causing the material to agglomerate.
Therefore, the tests were rerun after drying the neutralizing agents by the
direct application of heat.  Table 3 presents the results of these tests.  It
is evident that water had been absorbed.

-------
                                  Table 2
            Spectrochemical Analyses of Germany Valley Limestone
            Reported as Per Cent by Weight of Ignited Sample  (900 C)
            BCR Sample No. 2177

Loss on
lonition    Si02   A1203   Fe2C>3    MgO    CaO    Ti02   Na20   K20	MnQ2

  43.0%     1.0%   0.43%   0.15%  1.16%   97.0%   0.04%  0.02%  0.1%    0.03%

Per Cent CaC03 = 96.5%*
X-Ray analysis of the limestone identified only one compound  - CaC03

*This figure was used in limestone utilization efficiency calculations as P
                                  Table 3
                  Lime, Limestone, and Soda Ash Particle  Size



Lime
Screen Size

50 Mesh
100 Mesh
200 Mesh
400 Mesh
% Not Passing 50 Mesh
Before
Drying
92.1%
36.0%
5.6%
1.5%
7.9%

Dried
89.7%
55 . 1%
43.5%
3.8%
1O.3%
Per Cent
Passing
Limestone
Before
Drying
98 . 1%
42.2%
26.6%
4.5%
1.9%

Dried
99.2%
57.7%
34.4%
14.8%
0.8%

Soda
Before
Drying
49.1%
9.2%
0.01%
0%
50.9%

Ash!

Dri^d
49. C%
10.0%
1.5%
O.6%
51.0%
     The limestone and lime used in this study were obtained from Germany Val
ley Limestone Company, Riverton, West Virginia.*  In order  to obtain  the sral
lest particle size commercially available at a reasonable cost, "rock dvst"
was used.  The soda ash was from Diamond Shamrock Corporation, Cleveland.
     In Table 4, the chemical analyses as quoted by the manufacturers are pre-
sented along with the respective costs of the chemical.

                                  Table 4
                       Manufacturers Chemical Analyses

       Hydrated Lime               Limes tone               Soda Ash

  CaO           72.00% Min.   CaO         .   53.0 %     Na20           58%
  MgO           OO.40% Min.   MgO             0.38%     Na2C03       99-10O%
  CaCOs Equiv-                CaC03          98.8 %
     alent        130% Min.   SiO             0.49%

                                     Cost

        $17.00 per ton            $6.OO per ton             $100.00 per tor?
           in bags                   in bags                   in bags

* Mention of commercial products does not iiroly endorsement by the Federal Water
  Pollution Control Administration.

-------
*.-.....*- r."' oration vs. pii Tests

     Tests to indicate the relative concentrations of the chemical constituents
at various pH's were run by the following procedure:  First, a five gallon
sample of water from Grassy Run was taken into the lab and allowed to  reach
ar-ibient temperature.  The water was then divided into one liter portions.  A
separate sample was used for each pH point in the test.  The neutralizing agent
was added to the sample in sufficient quantity to bring the sample to  the de-
sired pH.  A laboratory mixer was used to agitate the sample.  Once the  sample
pH had stabilized, the supernatant was withdrawn and analyzed.  As discussed
earlier, the time required for pH stabilization was dependent on the neutrali-
zing agent used.  A 24-hour maximum limit was placed upon the reaction times,
thus the limestone and soda ash pH readings and supernatant analyses were made
at the 24 hour point.  By running all the concentration tests from one large
water sample, many variables were eliminated and direct comparisons could be
made between the results of each of the three neutralizing agents.

                                  RESULTS

Concentration vs. pH Tests

     Tests were run on identical samples of Grassy Run water to determine the
quality of the supernatant when the sample was neutralized to various pH's.
Lime, limestone, and soda ash were each tested as the neutralizing agents.
The results of the tests are presented in Table 5.

     When lime was used, the acidity was removed between a pH of 4.6 and 5.0.
At this pH, the iron, sulfate, and aluminum had been decreased to values near
minimum levels and the calcium, hardness,  and magnesium had increased as a
result of the lime addition.

     It was not until a pH of 9.2 that the magnesium was reduced to a low level.
The addition of more lime, beyond a pH of 5.9 only served to increase the pH
and decrease the magnesium since there were no further significant increases
in the alkalinity or calcium concentrations.   Thus,  above pH 5.9,  the lime was
being lost to the sludge.  It should be noted that the iron and aluminum were
never reduced below 2 mg/1.  Referring to our definition of utilization effi-
ciency, it can be seen that the greatest efficiency would be obtained between
a pH of 4.6 and 5.0, because at this pH the demand by the acidity had been
satisfied.  When more lime was added,  at first a small amount was used to in-
crease the alkalinity and the rest was lost to the sludge.   Eventually all of
the further addition of lime would be lost to the sludge.

     In these tests, limestone was not able to increase the pH beyond 6.5, be-
cause of the formation of carbonic acid,  which acted as a buffer.   An indication
of the carbonic acid is seen in the cold acidity analysis as the difference be-
tween the hot and cold acidity is mostly due to carbonic acid.   At a pH of 3.9,
the hot acidity was removed but it wasn't until pH 5.5 that the iron was re-
duced to a minimum of 2 mg/1.   Aluminum reached its minimum at pH 4.3.  Great-
est utilization efficiency, as defined in this paper, occurred at pH 3.9.  Above
this pH, some of the limestone is utilized to increase the alkalinity, but the
majority is lost to the sludge.

     Soda ash removed all the acidity at a pH slightly above 4.7.   The carbon
dioxide liberated during the reaction caused the continued presence of cold
acidity.  Although iron results were erratic, it can be concluded that a con-
centration of approximately 3 mg/1 is the minimum value.  Aluminum reached its
minimum at approximately 2 mg/1 at pH 5.4.  Continued increase of soda ash
addition beyond pH 4.7 served primarily to increase the pH,  alkalinity, and
sodium concentration.  Maximum utilization efficiency occurred at pH 4.7.

-------
                     IT,  C
                         O
                     O -H
                     H -P
                       A
                     rd  n
                     H -P
                         c
                         O
                         o
                         r-

                         0
                        u
                                      I  -p
                                     (ti  n
                                     X  C
                                     i-l  -H
                                     <  r-l
                                      i
                                       
                                     3  3
                                     H  C
                                     < -H
                                     -0  ^3
                                     H  -H
                                     O  O
                                     O  <
    Ifl
-I  (0
rt  ci
<->  c
p  -p
H  H
    rt
 i

 2  ^
 Oi 3
 >
 3  -P
O  -H
 C  >
 O  -H
U
                                                         o
                                                           
                                                         CM
                                                                            CONCMCOCMCOCOCMCOCONCMCOCOCO
                                                                                                               rf  CM H ON  t-(  r-l  H

                                                                                                               CM  co co CM  co  co  co
                                                         o
                                                         CO
                                                         CM
                                                         Tf
                                                         in
                    O
                    00
                                                        vO
                                                        0>
                                                        co
                                                        in
                                                        o
                                                        o
                                                        CO
                                                        oc

                                                        cl
                                                                                       HCO
                                                                           HHCMt-lr-(i-(i-  CO  H
                                                                                                               rH H  H  rH H  O  rH
                                                                                                               rH rH  rt  H rH  r-t  r-(
O O  in vD CO  r-l  CM
\O vO  CO H     r-l  iH
CM H
                                                                        oo  N m  co  in \o c^
                                                                        O  t^ N  H  CO vD CM
                                                                        O  i-t CM  m  vO CO rj<
                                                                        r-l  r-1 H  r-(  rH r-( i-H
                                                                                                                                                   co CTI o  CM  m m  CM
                                                                                                                                                   h> c> N  co  co co  a>
                                                                                                                                                   iH r-l rH  iH  H H  r-l
                                                                                                              O  W O  l-(  O  r-l in
                                                                                                              o\  c\ o\  co  o  co co
                                                                                                              CO  ON O  CO  in  r-l CM
                                                                           o o  o  o o  o
                                                                                                                                                  CM
o  o o o  o  o o
vO  O O O  O  O O
r-l  rt H CM  CM  CM CM
                                                                                                              \o o co  eo  o  'n m

                                                                                                              co co M1  <*  m  m o
                                                                                                                                               p
                                                                                                                                               C
                                                                                                         o

                                                                                                          co
                                                                                                        o
                                                                                                        u
                                                                                                         rt
                                                                                                        U

                                                                                                        (ft
                                                                                                         a

                                                                                                        TJ
                                                                                                                                                                                   G)
                                                                                                                                                                                   1-1
                                                                                                                                                                                   a
                                'C

                                 0
                                o

                                D

                                 g
                                                                                                        V)
                                                                                                        o
                                                                                                        C
                                                                                                                                                                                   rt
                                                                                                                                                                                   -p
                                                                                                                                              
                                                                                                                                              H
                                                                                                                                              U)
                                                                                                                                              
-------
o  p.
 0
 3   
 C  V)
H  >

 C  C
 0  O
    rt
in  M
^3  U
,
J  Q)
P  C

"8
    rt
   ffi
                T3
             P -H

             ^
O   >

D  -H
C   >
0  -H
                ac
                 PH
                                 inrfpooinomoooooo
                                 ^ m  H
                          to
                                 vOOjOiHCOCOOOOOONCMoOOvO



                                 HH     HHHHHHrti-IHHH
                                 CO C\!
                                 i-fHHrHrHt-HHfHr-li-HHHrHH
                                 OWOONOOOOOOOOOO
                                 OOOOOOOOOOOOOO
                                 oooooooooooooo
                                                    O
                                H  H H H  r-l H
                                                           r-4 H i-l H  H C\J
                                                                                                                tf)
                                                                                                                p
                                                                                                               H
                                                                                                                3
                                                                                                                cr
                                                                                                                 cn
                                                                                                               O
                                                                                                  U
                                                                                                  U)


                                                                                                  "8
                                                                                                  U)
                                                                                                  U)
                                                                                                  0)

                                                                                                  
                                                                                                  &
                                                                                                               o
                                                                                                               H
                                                                                                               C
                                                                                                               r-l
                                                                                                               0
                                                                                                               U
                                                                                                               c
                                                                                                              o
                                                                                                               V4
                                                                                                               
                                                                                                              H
                                                                                                               (0
                                                                                                               V


                                                                                                              
                                                                                                              H
                                                                                                               O
                                                                                                              U
                                                                                                              H


                                                                                                              3

-------
     <'iK1 oi the pronounced results of these tests was the continued presence
of iron, even at high pH levels.  The presence of iron is in contrast to
published 'ideal' solubility curves, such as those by Hill(4) which indicate
complete ferric iron removal above pH 5.O.  Similar difficulties in iron re-
moval have been reported by Maneval, et al.'S'

     Since the analyses were made on unfiltered samples, it was suspected that
possibly the iron was remaining in suspension rather than in solution.  There-
fore, three 'suspect1 samples were analyzed for iron, then filtered through
a 0.45 micron filter to remove all particle matter and the iron analysis was
repeated.  The results of this test are in Table 6.  Obviously, the iron con-
centration was not affected by filtration, thus the iron does not exist in a
suspended particle state unless it is smaller than 0.45 microns.  The iron may
be involved in an organic complex.  However, no studies have been made to re-
solve this question.

                                Table 6
                         Effects of Filtration
                        Upon Iron Concentration
Suspect
Samole
L-6
L-7
L-8
L-9
PH
7.0
7.3
8.0
10.0
Total Iron
Before Filtering
6.7
10.1
17.2
2.0
Concentration
After Filtering
6.6
10.2
17.1
2.0
         Filter Size = 0.45 microns
Relative Reactivity Study

     The relative reactivities of lime, limestone, and soda ash were compared
by the addition of measured amounts of each neutralizing agent to identical
separate samples of Grassy Run water.  Graphs of the results of these tests
are shown in Figure 3.  These tests were run allowing five minutes reaction
time between incremental addition of the neutralizing agents.

     In the lime reaction, the steep slope between pH 4.O and pH 8.O is indi-
cative of the quantitative sensitivity of the lime reaction in this portion
of the curve.  It is obvious that a very small change in the amount of lime
added at pH 6.0 can easily shift the pH one to two full points.  Thus, in the
lime reaction, accuracy of feed is of vital importance.

     The slope of the limestone curve decreases rapidly at pH's above 6.0 as
the reaction liberates C02 which, in turn, buffers the reaction.  Consequently,
the curve is nearly flat above pH 7.0 and increases in pH above this point are
very difficult to attain.  However, the shallow slope of the curve in the pH
5.0 - 7.0 range would make the limestone reaction an easy one to control in
an operating system as the reaction would be insensitive to small changes in
the neutralizing agent feed rate.

     The soda ash reaction is quite similar to the lime reaction in the shape
of the curve.  However, this sodium carbonate reaction, as in the case of lime-
stone, suffers from buffering due to the formation of carbonic acid (H2CO3).
It is seen that the formation of H2C03 retards the progress of the reaction
in both limestone and soda ash neutralization.  With time, the H2C03 breaks
down to yield carbon dioxide which escapes as a gas, thus the buffering capa-
city decreases and the pH increases.

-------
                                00
                                LU
        CD   -ec
              cc.


              GO
              GO
              <
              ce
              CD
                               CO
                  CO
                           CO

                           UJ
CD
00

ca

S
                  CD
              oo
              cz:
              CD
                               CD
                               C-
                              1 rl^
                               CD
                               CJ5
CNJ

-------
     Keeping in mind that the scale for limestone is different from that of
lime and soda ash in Figure 3, it is clear that lime is the most reactive a-
gent.  It requires roughly 7.4 times as much limestone as lime in this case
to raise the pH to approximately pH 6.5.  One and one-half times as much soda
ash as lime is required to reach pH 6.5.  It should be remembered that these
tests had only five minutes reaction time and any increase in detention time
would narrow the gap between lime and the other two agents.

     Another comparison that can be made is the amount of neutralizing agent
used at the maximum utilization efficiencies developed in the preceding sec-
tion.  As seen in Table 7, four times as much limestone would be required and
five times as much soda ash as lime.

                                 Table 7
                       Maximum Utilization Efficiencies
Neutralizing
Agent
Lime
Limestone
Soda Ash
pH at Which
Maximum Utilization
Efficiency Occurs*
4.6
3.9
4.7
Grams
Material
Used**
0.10
0.40
0.54
          * Data from Table 5
          **Data from Figure 3
Optimization

     Before optimizing the system, it was necessary to set a goal for the ef-
fluent water quality and then to determine the operating conditions which best
achieved this goal.  The initial intent was to produce the best effluent pos-
sible.  Considerations of cost paled this view somewhat to the more realistic
criteria of meeting and/or exceeding the water quality standards of the mining
states which, in themselves, proved quite restricting.  Table 8 presents the
published criteria relating to mining in West Virginia and Pennsylvania.

     The previous studies had shown that the stream standards of West Virginia
could be met with the exception of sulfate and that the more stringent effluent
standards of Pennsylvania could be satisfied.  It also was evident that the
iron, aluminum, and net alkalinity requirements could be met at a pH less thajn
6.O.  Therefore, in order to meet the pH 6.O requirement, the system must be
operated at less than maximum utilization efficiency.  Referring to Figure 3,
it can be seen that the amount of neutralizing agent required for an increase
in pH from the optimum values given in Table 7 is more for limestone and soda
ash than for lime.  Therefore, the limestone and soda ash neutralization ef-
ficiencies decrease at a higher rate than lime as the pH increases.

     Thus our choice of operating conditions were:

     -pH 6.0 to pH 6.5
     -Iron removal below 7.0 ppm
     -Net Alkalinity (Alkalinity exceeds acidity)

-------
                                   Table 8
                           Water Quality Standards
     State of West Virginia

                pH               5.5 Minimum, 8.5 Maximum
                Iron              10 ppra
                Aluminum          30 ppm
                Sulfates         200 ppm

     These are stream standards.  There are no effluent standards in West
     Virginia.

     Commonwealth of Pennsylvania

                pH               6.0 Minimum, 9.0 Maximum
                Net Alkalinity
                Iron             7.0 ppm Maximum

     These are effluent standards.

Pilot Plant Studies

     Actual pilot plant operation supplied the following data.  This data has
been evaluated and mathmatically reduced to eliminate, as much as possible,
the effects of variations due to the day-to-day fluctuation of water quality
in Grassy Run.

                               Limestone

     Because of the delay in the limestone treated water reaching a final pH,
it was necessary for control purposes to establish a lower pH in the reactor
in order to obtain the desired resulting effluent pH.  For example, for the
test reported in Table 9, a pH of 5.8 was maintained in the reactor which re-
sulted in a pH of 6.6 in the plant effluent.  The pH maintained in the reactor
was called the instantaneous in-reactor pH.  This phenomenon causes some dif-
ficulties in operating a limestone treatment system.

     Tables 9 and 10 show the results of treatment of two different concen-
trations of Grassy Run water with "rock dust" limestone.  In these instances,
the limestone was introduced in the slurry form.

     The cost of limestone is 1.7 times as great to treat the water in Table
10 as in Table 9 and the water's acidity is 1.7 times as concentrated.  Simi-
larly, there is no significant difference in utilization efficiency between
the two .

                                 Lime

     Table 11 shows data acquired from neutralization with lime.  Utilization
efficiency of lime approaches 100 per cent and is easily maintained, even at
high pH's.  Unlike limestone, the pH was slightly less in the plant effluent
than in the reactor.

     Supernatant water quality from lime is relatively good.  However, as in
the case of limestone, the water still remains typically high in hardness, cal-
cium, and sulfates.  Removal of iron with lime was better than that recorded
for limestone as supernatant iron concentrations were consistently below 3.O
ppm.  Better iron removal probably is partially attributable to the higher pH's
attained using lime.

-------
                                   Table 9

                Limestone Slurry Neutralization of Grassy Run Water*

  Instantaneous In-Reactor pH = 5.8**
  Raw Water Feed Rate = 6402 ml/min  (1.69 gpm)
  Limestone Slurry Feed Rate = 132 ml/min slurry  (9.1 grams/min limestone)
  Reactor Detention Time     = 33 minutes (theoretical)
  Settling Tank Detention Time = 12 hours (theoretical)


                                     MEAN DATA***
Parameter Units
pH (after 24 hrs.)
(after 48 hrs. )
(after 72 hrs.)
Hot Acidity (as CaC03) mg/1
Cold Acidity (as CaCO3) "
Iron "
Sulfates "
Hardness (as CaCO^) "
Conductivity Mmhos/cm
Magnesium (as CaC03) mg/1
Alkalinity "
Calcium (as CaC03) "
Aluminum "
Temperature of Water
Acid Removed (Hot) "
Acid Removed (Cold) "
Limestone Utilization Efficiency
Feed
Water
2.9
362
373
195
516
626
927 
171
0
455
33
10C



Reactor
Effluent
6.6
6.8
7.1
0
33
7.3
479
1647
841
175
78
1472
1.6
10C
362
34O
32.1%
Settling Tank
Effluent
6.6
6.8
7.0
O
25
7.0
479
1547
857
170
66
1377
0.9
9C
362
348
31.2%
  Lbs/1000 gal Limestone = 11.8 Ibs/lOOO gal Grassy Run Water
  Cost per 1000 gallons  = $0,0354/1000 gallons
  Cost per mg/1 Acid/1000 gallons = 0.010 cents

  Per Cent Sludge by Volume ( 1 hour settling time) = 3.5%
  Per Cent Sludge by Volume (24 hours settling time) = 3.7%

  *A11 analyses made 24 hours after sampling, except as noted for pH
 **ph of water in the reactor during the test run
***Summary of many sets of data collected during the test run

-------
                                 Table 1O

               Limestone Slurry Neutralization of Grassy Run Water*
  Instantaneous In-Reactor pH = 5.7**
  Raw Water Feed Rate = 4800 ml/min (1.27 gpm)
  Limestone Slurry Feed Rate = 223 ml/min slurry (11.4 grams/min limestone)
  Reactor Detention Time = 43 minutes  (theoretical)
  Settling Tank Detention Time = 16 hours (theoretical)          (
                                 MEAN DATA***
Parameter
                          Units
                 Feed
                 Water
Reactor
Ef f luen t
                                         Settling Tank
                                           Effluent
  pH
  Hot Acidity   (as CaCO  ) mg/1
  Cold Acidity  (as
  Iron
  Sulfates
  Hardness   (as
  Conductivity
  Magnesium  (as
  Alkalinity
  Calcium    (as CaC03)
  Aluminum
  Temperature of Water
  Acid Removed  (Hot)
  Acid Removed  (Cold)
  Limestone Utilization Efficiency
Mmhos/cm
  mg/1
      ti
      tt
                  2.7
                  602
                  595
                  145
                  904
                  588
                 1250
                  188
                    0
                  400
                   44
                  5C
    6.7
      O
     77
      5
    663
   1621
   1100
    196
    122
   1425
    3.8
    7C
    6O2
    518
   31.5%
  Lbs/1000 gal. limestone = 19.7 Ibs/lOOO gal. of Grassy Run Water
  Cost per 1000 gallons   = $O.Q591/1000 gallons
  Cost per mg/1 Acid per 1000 gallons = O.O1O  cents
                                                                    6.6
                                                                      0
                                                                     79
                                                                      5
                                                                    723
                                                                   1588
                                                                   1150
                                                                    188
                                                                     90
                                                                   1400
                                                                    3.9
                                                                    7C
                                                                    6O2
                                                                    516
                                                                   30.1%
  Per Cent Sludge by
  Per Cent Sludge by
Volume ( 1 hour  settling time)  5.0%
Volume (24 hours settling time) = 4.5%
  *A11 Analyses made 24 hours after sampling
 **pH of water in the reactor during the test run
***Summary of many sets of data collected during the test run

-------
                                 Table 11

                  Lime Neutralization of Grassy Run Water*
  Instantaneous In-Reactor pH = 7.9**
  Raw Water Feed Rate = 5964 mg/1 (1.58 gpm)
  Lime Feed Rate = 2.69 grams/rain - dry feed
  Reactor Detention Time  35 minutes (theoretical)
  Settling Tank Detention Time = 13 hours (theoretical)

                                 MEAN DATA***

Parameter Units
PH
Hot Acidity (as CaC03) mg/1
Cold Acidity (as CaCOo) "
Iron "
Sulfates "
Hardness (as CaCOg) "
Conductivity Mmhos/cm
Magnesium (as CaC03) mg/1
Alkalinity "
Calcium (as CaC03) "
Aluminum "
Temperature of Water
Acid Removed (Hot) "
Acid Removed (Cold) "
Lime Utilization Efficiency
Feed
Water
2.8
601
612
157
717
582
1450
135
0
447
33
13C



Lbs/lOOO gal lime = 3.75 Ibs/lOOO gal of Grassy
Cost per 1OOO gallons = $0.032/1000
Cost per mg/1 Acid/lOOO gallons = 0
gallons of
.0052 cents
Reactor
Effluent
7.1
2.1
0.4
2.4
709
1261
1241
162
17
1099
O.8
15C
599
612
103%
Run Water
Grassy Run

Settling Tank
Effluent
6.9
1.7
2.6
1.6
693
1220
1245
145
11
1075
0.9
15C
599
609
102%

Water

  Per Cent Sludge by Volume = 6.9% (24 hours after settling time)
  Per Cent Solids in Sludge =1.7%

  *Chemical Analyses Made 24 Hours After Sampling
 **ph of Water in the Reactor during the test run
***Summary of many sets of data taken during a test run

-------
                                   Table 12

                  Soda Ash Neutralization of Grassy Run Water*

  Instantaneous In-Reactor pH = 6.0**
  Raw Water Feed Rate = 11,000 ml/min or 2.91 gpm
  Soda Ash Feed Rate =7.1 grams/min
  Reactor Detention Time = 19 minutes (theoretical)
  Settling Tank Detention Time = 7.2 hours (theoretical)


                                  MEAN DATA***
Parameter
Units
Feed
Water
Reactor
Effluent
Settling Tank
Ef f luen t
                   CaC03 )mg/l
                   CaCO  ) "
                       J  it
                          it
PH
Hot Acidity   (as
Cold Acidity  (as
Iron
Sulfates
Hardness   (as CaCO3)    
Conductivity         Mmhos/cm
Magnesium  (as CaCO3)   mg/1
Alkalinity              "
Calcium    (as CaCO3)    
Aluminum                "
Sodium                  "
Temperature of water
Acid Removed (Hot)
Acid Removed (Cold)
Soda Ash Utilization Efficiency
Lbs/lOOO Gallons Soda Ash
        2.7
        551
        586
        126
        581
        526
       1300
        205
          0
        321
         41
        28
       126C
                          it
                          tt
 6.1
   O
  73
 7.9
 672
 530
1425
 205
 100
 325
 3.5
 555
12C
                      551
                      513
                     106.896
5.367 Ibs/lOOO gallons of Grassy
                                                               Run
 6.5
   0
   0
 2.3
 768
 541
1500
 208
 103
 333
 2.2
 450
12C
 551
 586
107.3%
Water
  Cost per 1000 gallons = $0.268/1000 gallons of Grassy Run Water
  Cost per mg/1 Acid/1000 gallons = 0.0486 cents

  Per Cent Sludge by Volume = 7.6%
  Per Cent Solids           = 1.5%

  *Chemical Analyses Made 24 Hours After Sampling
 **ph of Water in the Reactor during the Test Run
***Summary of Many Sets of Data taken During a Test Run

-------
     However, the very low per cent solids of the lime sludge (1,7 per cent)
will be a major problem when considering sludge disposal.
                     '                                  r
                                  Soda Ash

     Table 12 contains the results of the pilot plant test using sodium car-
bonate as the neutralizer.

     It is seen that soda ash has only one major advantage over lime or lime-
stone 5n the low hardness concentration.  Utilization efficiency, as in the
case of lime,  approaches 1OO per cent.  However, the one major disadvantage of
soda ash treatment is the cost.  Soda ash costs over nine times as much as
lime per mg/1 acid per 10OO gallons treated.  It costs almost four times as
much as limestone.

     Even though the hardness is low using soda ash, the sodium concentration
in Table 12 was increased from 2.8 mg/1 to 450 mg/1.

     A comparison of the relative settling rates between lime, limestone, and
soda ash sludges is presented in Figure 4.  Sludge from the limestone reaction
compacts to the smallest volume and all settling is completed in approximately
one hour.  Lime sludge has a very rapid initial settling rate (due to the rap-
id lime reaction, high pH's are quickly attained), but suffers from poor com-
paction.  It requires almost five times as long for the lime to reach a state
of final compaction as for the limestone.  In the final analysis, however, bet-
ter than 90 per cent of the total settling is completed in all cases in less
than one hour.  The soda ash settling characteristics are virtually identical
to those of the lime.

     A test was made to compare the per cent sludge solids of lime, limestone,
and soda ash sludges.  Using the same water, samples were neutralized to pH
6.5 (lime and soda ash) and pH 6,2 (limestone, allowed to settle 24 hours, and
then the sludge was withdrawn and per cent solids determined.  Table 13 pre-
sents the results of these tests.  Obviously, limestone sludge is far superior
in per cent solids content.

                                  Table 13
                         Per Cent Solids of Sludges*

                                                        Per Cent Solids
       Neutralizer             Supernatant pH              of Sludge

       Lime                        pH 6.5                    1.5%
       Limestone                   pH 6.2                    9.5%
       Soda Ash                    pH 6.5                    0.7%

       *After 24 hours of undisturbed settling

     Lime and limestone settling tests were also run on a Grassy Run water sam-
ple which had been concentrated 3.3 times by a reverse osmosis unit.  The re-
sults of this test on the concentrated water are presented in Figure 5.  Com-
paring the results from the water as opposed to the unconcentrated Grassy Run
water, we find that the general settling trend of lime does not appear to vary
with increased concentrations.  In the case of limestone, however, marked im-
provements in settling characteristics are apparent in the concentrated water
tests.  In both of these instances, as in the case of the unconcentrated water,
better than 9O percent of the settling is completed in one hour.  The sludge
volumes were greater for the concentrated water than for the mine drainage due
to the larger amounts of iron, calcium, acid, and etc. that were available for
precipi tation.

-------
     CD
     OO
o < n
CO
CO
LO
                    CO
                                                             Csl
                                                             OJ
                                                            CD
                                              CD
                                                                  ILU
                                                                  OO
                                                            co    OO
                                                                  CD
                                                            CO
                                                                  UU
                                                                  CD
JO
     1V101  JO
                                              JO

-------
            CO
                     CO
                                       CSI
                                                   CXI
                                                    CO
                                                   CO
                                                    00
                                                   CSI
                                                   CO
                                                   00
                                                   CO






                                                   CO



                                                   CSI
                                                   CD
                                                   CO
                                                  CD
                                                  CO
(N31VM  JO  3WniOA  IViOi JO %)  ]3QniS JO 3WmOA
                                                               C/3
    cc:

    CE5


    J oo
    I Lt-l
    J_. ,;
CD  U.J !H
                                                              00
                                                              a.

                                                              cS

-------
     iiioxe 14 jLiiuj.cai.es Lne per cent rectuc nons or per cent increases j.u  tae
supernatant quality of neutralized water.  Direct comparisons are made between
lime, limestone, and soda ash supernatants.

Efficiency Improvement Studies for Limestone

                                Aeration

     The studies up to this point had shown that limestone has a raw material
cost advantage over lime ($6.0O vs. $17.00) and has a more dense, less volumi-
nous sludge.  On the other hand, the limestone utilization efficiency was only
32 per cent while lime was 100 per cent.  On this basis, it would appear  that
it would require three times as much limestone as lime to treat the same water
and, therefore, the raw material cost would be 1.1 times higher for limestone.
However, as shown in Tables 1O and 11, the cost for limestone was O.010 cents
per 1000 gallons/mg/1 acid treated as compared to 0.0052 cents per 100O gallons/
mg/1 acid treated for lime or 1.9 times more.  There appears to be a conflict
in these results when, in fact, there is not.  The discrepancy is one of defi-
nition.

     Assuming supernatant alkalinity to be beneficial to the environment, uti-
lization efficiency was defined to include it.  Very little increase in super-
natant alkalinity is seen with lime neutralization as opposed to the high
alkalinity of limestone treated product water.  From a cost of treatment stand-
point, this alkalinity could be considered unnecessary and, thus, detrimental.

     If the alkalinity term of the utilization efficiency equation were to be
excluded and the efficiencies recalculated, then the lime efficiency (Table 11)
would be 98.5 per cent and the limestone (Table 9) would be 26.3 per cent.  It
would appear to be necessary to use 98.5 per cent/26.3 per cent = 1.3 times as
much limestone as lime.  Referring to the definition of utilization efficiency,
it is seen that both lime and limestone efficiencies are expressed as CaCO~.
Thus to discuss cost, or actual usage, it is necessary to reconvert the lime
efficiency from CaCO3 back to CaO and thus it is multiplied by 1.35 (ratio of
molecular weights CaC03/CaO).  The resulting value for lime is 133.0 per cent
as compared to 26.3 per cent for limestone or a ratio of 5.1 times as much
limestone is needed as lime.  Therefore, on a cost basis, limestone costs 1.8
times as much as lime and this corresponds with the 1.9 times figure derived
from the cost per rag/1 acid 10OO gallons columns on Tables 1O and 11.

     In order to have limestone compete with lime, it is necessary to improve
the efficiency of the limestone process.  Several methods were initially con-
ceived:  1. Increase the detention time of the water in the reactor, thus
providing time for more limestone to become solubilized,  2. Increasing the
vigorous mixing action in the reactor to create more complete breakdown of the
limestone, as earlier studies indicated the more rapid the mixing,  the better
the results.  3. Return the sludge with its unused limestone back to the re-
actor, and 4. Aerate the water to strip off the carbon dioxide and thereby re-
duce the buffering capacity.

     The test equipment on hand was unsuitable to perform tests on items 2 and
3, so a new and larger test facility has been constructed to study these factors.
Item number 1 could not be studied directly with the present equipment.   How-
ever, studies in comparing dry feed and slurry feed gave some indirect insight.
Aeration was studied as part of the current study.

     First attempts at aeration were performed in the upflow settling tank.
The data is presented in Table 15.  The effluent pH rose immediately from pH
6.7 to pH 7.1 as soon as aeration began.  However, cold acidity barely dropped
(from 29 mg/1 to 25 mg/1) and the overall rise in efficiency was 0.9 per cent.
This insignificant change certainly does not justify the cost of aeration
equipment and the involved maintenance.

-------
si   fa
 rt  H


w   o
             IT)
             K.
             a
T3
 0)


 rt

 Oi
 M

 O

 B
H
         O
         o
o
 0)

I
 a>
                            o
                            o
                            o
                             0)
                             o
                             3
                            -a
                             o>
                            a
                                    00
                        o
                         6)
                         O
                         3
                        a
                         01
                        a
B
CM*


o
0)
in
rt
Oi
14
0
c
H
^
in
rt

C
0>


o
B
M

.
CO


T3
0>
(0
rt
0)
M
O
B
H
NO
*f
0*



0
OJ
o
3
"8
a
m
i-i
o
i-i

o
0)
U)
*
VH
O
H
B
O
in
0
p
0
0^
(A
rt
o>
^4
o
B
H

\J
rt
N
rt
rH
rt
M
P
3
0>
2

5

0)
3
<*
H >>
0) -rt
H H
& rt
$ S

M
01
P
rt
J5

B
rt
U)
rtj
ON

43
U
i-i

0) ON
B
o w
P 0)
W i-l
Q) J3
B rt
rt H

o>
^
^-*







,_.
H
H

0)
i-l
0? <"N
fi rt
rt H
J
S
0

frl
"*



ffi
a

o
4_>

-a
IA
rt
Q)

O
B
H






O
NO

X
a

0
+J

-o
0>
V)
rt
0)
V4
0
M

6
8






o
Q)
0
3
0)
a






6?
^
O^
Q\







"8
y
^
o
a

S
o






o
0>
o
3
o
0>
a






^
m
O^
Q\







8
u

"S
a

R
vO
ON






o
OJ
o
3
T3
0)
a







.
ON
ON







-o
O
3
-o
0)
04

3J
CO
H






8
o
3
o
01
p^






l&
.
CO








0)
o
3
o
0>
a

m
CO
in
CM




o
0>
U)
rt
0)

u
B
H






Bs
vO
0*
O
CM




o
o
U)
rt
Q)
^4
o
B
M

00
"






-a
o>
o

1
K






K
H
^M
H







a
o>
u
3
O
0)


ON.
CO


0
+*

o
0)
U)
rt
o>
M
0
M





H
X.
8

H
i-l

0
P

o
0>
U)
rt
Q)
^4
0
B
H

in
0
rH
CO




T3
0>
V)
rt
0>

0
B
M






Sx
"1
O

W




-a
o>
(R
rt
0)

u
B
M

N co
S 5*
M




o
a
T3 W
0) rt
U 0)
3 M
-o o
0) B
a M






8? tR
CO Tf
 
t^ t^
ON O
H




t3
Q)
T3 W
o> rt
0 0)

i g
a M

    0>
   P
    0>
    B
    rt
    M
    rt
   0-

a
I

^
"O
rt

<
o
H
0

rt
o
rt

<;



B
0
M
H

U)
0)
P
rt
M-t
H
3
(A

U)
U)
2
o

rt
X
p
rt
rt
P
S
-o
B

u
^
rt
rt
i-l
rt
A
H
<



rt
U
H
tt
U

S
3
B
rt
g
P
^
<
B
3
rt
U)
Oi
B
(^
frt
S
                                                                                     rt
                                                                                     o

-------
                                 Table 15
                    Grassy Run - Slurry Feed Limestone
                   Comparison of Data Prior to and After
                        Aeration of Settling Tank
Instantaneous In-Reactor pH = 6.5*

Parameter Units
PH
Hot Acid (as CaC03) ag/1
Cold Acid (as CaC03) "
Iron "
SO, "
4
Hardness (as CaC03) "
Conductivity Mmhos/cm
Alkalinity ng/1
Calcium (as CaC03) 
Aluminum "
Flow Rate (Mean)
Acid Removed (Hot) mg/1
Acid Removed (Cold) "
Limestone Utilization Efficiency

Feed
Water
2.9
375
379
185
542
564
1100
0
394
30
1.48 gpm
56OO ml/min



Before
Aeration
Settling
Tank
6.7
0
29
2.5
506
1375
1100
72
1219
0.7
After
Aeration
Settling
Tank
7.1
0
25
3.1
544
1263
1100
87
1103
0.5
208 ml/min limestone slurry
10.4 grams/min limestone
375
350
24.0%
375
354
24.9%
Chemistry analyses made 24 hours after samples were taken

Mean Data from settling tank represents five samples taken within 4 hours
after aeration was started
*ph of Water in the Reactor during the Test Run.

-------
                                  Table 16
                       Grassy Run - Slurry Feed Limestone
                      Comparison of Data Prior to and After
                              Aeration of Reactor
Instantaneous In-Reactor pH = 6.5*
BEFORE AERATION
Parameter Units

PH
Hot Acid (as CaO^) mg/i
Cold Acid " "
Iron "
so4
Hardness (as CaC03) "
Conductivity Mmhos/cm
Alkalinity mg/1
Calcium (as CaO>3) 
Aluminum "
Feed
Water

2.8
464
470
136
591
439
1200
0
300
30.8
Flow Rate (Mean) 6867 ml/min
1.81 gpm
Acid Removed (Hot) mg/1
Acid Removed (Cold) n


Limestone Utilization Efficiency
Temperature of Water
18C
Inune- 24 Hrs.
diate Later
(1 Hr.)
6.3
0
41
23
566
1184
1000
56
1044
5

7.0
0
16
7.9
530
1168
1000
62
1031
2.2
2O5 ml/min slurry
10.25 grams/min
464
429
34.8%
18C
464
454
35.296
18C
AFTER AERATION
Feed
Water
(Mean)
2.8
454
472
139
486
452
1167
0
310
34
limestone



18C
Imme-
diate
(1 Hr.)
7.0
0
30
20
466
1263
973
41
1113
5.0

454
442
33.2%
18C
24 Hrs.
Later

7.3
0
9
4.3
542
1360
967
45
1200
1.6

454
463
33.4%
18C
*ph of Water in the Reactor during the  Test Run

-------
     Aeration in the reactor was then studied (Table 16).  A nore pronounced
pH change occurred after aeration (from pH 6.3 to pH 7.6), and the cold acid-
ity was reduced 25 per cent.  However, product alkalinity was reduced 34 per
cent, thus resulting in a 1.6 per cent net loss in efficiency.  There were no
other significant changes in overall water quality.

     In summary, aeration of the limestone reaction proved to be of no value
in increasing efficiency.  However,  as it rapidly strips off CC>2, it could be
very useful in drastically reducing detention time before supernatant discharge
(discharge occurs when desired pH is attained) and would be especially appli-
cable at sites where installation or presence of large holding lagoons (for
the purpose of detention time) would not be feasible.

                          Dry Feed vs. Slurry Feed

     Studies were made to determine if the feeding method--either wet or dry-
would affect the reactions of any of the neutralizing agents.

     Equal amounts of each neutralizing agent were added to identical samples
of acid mine water.  The only difference was that one was slurried with dis-
tilled water while the other was dry fed.  The samples were then agitated by
a flash mixer for five minutes each and then allowed to settle.  The analyses
were made on all samples at identical times after mixing.

     The results of these tests are presented in Table 17 and indicate there
is no significant difference in water quality or efficiency (as indicated by
alkalinitythe only variable in the efficiency equation) between dry feeding
or slurry feeding.

     These results also indicate that increasing detention time to enhance lime-
stone efficiencies by increasing solubilization of limestone (as mentioned
earlier) does not appear promising.   The increase in detention time could well
increase the efficiencies due to increased acid-base contact time or other
factors but the solubility of limestone in water does not appear to have much
bearing on the reactions involved.

                                  Table 17
                Comparison of Dry Feeding and Slurry Feeding
Parameter
pH
Conductivity
Hot Acidity(asCaCO3
Cold Acidity "
Calcium (as CaCO )
Magnesium 
Hardness "
Sulfates
Iron
Aluminum
Alkalinity
Sodium
Raw
Water

2.7
1650
) 532
558
318
212
530
868
120
35
0
2
Lime
Dry
6.1
1460
0
69
812
212
1024
747
2.3
1.1
17
2.3
Slurry
6.2
1460
0
64
812
212
1024
723
2.0
1.2
18
2.2
Limes tone
Dry
6.6
1460
0
90
862
206
1068
723
2,1
1.3
50
2.0
Slurry
6.8
1460
0
12
844
206
1050
688
2.2
1.2
50
2.0
Soda
Dry
6.8
1750
O
65
318
214
532
723
4.9
2.7
97
38O
Ash
Slurry
6.8
1750
0
24
300
201
501
699
6.3
2.4
91
320
All units are expressed in rag/1 except conductivity (Mmhos/cm) and pH

-------
                                    SUMMARY

     a) All three neutralizing agents--lime, limestone, and soda ashcan meet
the pH 6.5, net alkalinity, iron below 7.0 ppm criteria with relative ease.

     b) The major differences between soda ash neutralization and lime or
limestone neutralization are:

     Soda ash produces:  1. Low hardness
                         2. Low calicum
                         3. High sodium
                         4. High cost

     All other chemical parameters are coup arable between the three neutral-
izing agents.

     c) Soda ash costs nine times as much as lime and four times as much as
limestone to treat the same water under the test conditions of this study.

     As soda ash is obviously too expensive for other than specialized cases,
the remaining part of the summary will be confined to comparisons between use
of lime and limestone.  The following observations are made, based upon the
results obtained during these studies:

     d) Lime costs approximately half as much as limestone to treat the same
water under the test conditions of this study.  This refers only to the cost
of the chemical itself.

     e) Limestone sludge is more desirable in that it occupies only approxi-
mately two thirds of the volume of lime sludge and has a higher solids content.
It contains a large residual alkalinity which would be beneficial when disposed
into an acid mine environment (although this residual alkalinity is expensive
and of questionable value to the treatment plant).

     f) Lime is a very reactive material and the neutralization reaction goes
to completion in less than one hour.  The limestone reaction requires from 24
to 48 hours to go to completion and therefore requires a long detention time
before discharge.  This detention time can be expensive; however aeration will
reduce the detention time to one comparable to lime.

     g) The limestone reaction is not very sensitive quantitatively so the ac-
curacy with which constituents are fed into the reactor need not be controlled
with the precision required by lime.  The cost of the controlling system in-
volved could be somewhat reduced for this corresponding reduction in neces-
sary accuracy.  Accidental overtreatment is not the pollution problem with
limestone that it would be with lime.

     h) Lime is capable of attaining higher pH's which may be necessary in
some individual cases for the desired water quality characteristics, whereas
limestone efficiencies drop off rapidly at high pH's and correspondingly, pH's
above 7.0 are very difficult to attain.  In order for limestone to be compet-
etive with lime, methods of improving the utilization efficiency must be found.
Sludge return appears to be a promising method.

     The following observations are presented relative to the plant design and
operation:

     i) We found no difference in the efficiency between dry feed methods of
introducing the neutralizing agent or slurry feed methods.  This applies to
lime, limestone, and soda ash.

-------
     j) Dry feeding offers fewer operational difficulties than does a slurry
feeding and costs less, both initially and in cost of maintenance.

     k) PVC (PolyVinylChloride) plastic pipe should be used for all piping,
and pumps should be of the open face impeller type with all wetted parts made
of 316 stainless steel.  This will resist not only acidity but also the ex-
treme abrasiveness of the liquids involved.  Initial cost will be higher for
this type of equipment but maintenance costs are conservatively estimated to
be on the order of 1/16 or less as compared to cast iron or bronze.  Selection
of low RPM pumps (1750-1800 rpm) will further reduce pump wear.

     1) In all cases, almost all the sludge settling is accomplished in one hour.

                                 ACKNOWLEDGEMENTS

     Credit for the work and ideas involved in this report must be shared with
the following people:

     Ronald D. Hill, Chief, Mine Drainage Pollution Control Activities,
     FWPCA, Cincinnati, Ohio

     James L. Kennedy, Project Chemist, Norton Mine Drainage Treatment
     Laboratory, Norton, West Virginia

     Curtis L. Corley, Plant Operator, Norton Mine Drainage Treatment
     Laboratory, Norton, West Virginia

     Alvin W. Irons, Plant Operator, Norton Mine Drainage Treatment
     Laboratory, Norton, West Virginia

     Howard W. Howell, Chemistry Technician, Norton Mine Drainage
     Treatment Laboratory, Norton, West Virginia

                                   REFERENCES

(1) U. S. Geological Survey.  (Telephone conversation - Charleston Office).

(2) American Public Health Association, Inc.; Standard Methods for the Exam-
    ination of Water and Waste Water, Twelfth Edition, 1965, American Public
    Health Association, New York, N. Y.

(3) Ford, C., Young, R., And Glenn, R., Optimization of Development of Im-
    proved Chemical Techniques for the Treatment of Coal Mine Drainage,
    Bituminous Coal Research, Inc., Monroeville, Pa., Federal Water Pollution
    Control Administration Report No. 14010EIZO1-7O, 1969.

(4) Hill, Ronald D., Mine Drainage Treatment, State of'the Art and Research
    Needs.  Federal Water Pollution Control Administration, (December, 1968).

(5) Maneval, Dr., Charmbury, H. B., and Girard, L., "Operation Yellowboy" -
    Design and Economics of a Lime Neutralization Mine brainage Treatment
    Plant.  Preprint No. 67F35, Paper presented at the Annual Meeting of
    American Institute of Mining, Metallurgical and Petroleum Engineers, Los
    Angeles, Feb., 1967.

(6) Stream Water Standards, Water Resources Division, West Virginia Department
    of Natural Resources, Charleston, W. Va.

(7) Clean Stream Law of Pennsylvania, Sanitary Water Board, Commonwealth of
    Pennsylvania, Harrisburg, Pa.

-------
IT-'

-------