Quarterly Progress Report for by ltm66165

VIEWS: 13 PAGES: 20

									         Quarterly Progress Report for
          USEPA Grant S-82874601-1



Evaluate Pilot and Full-Scale Treatment Processes
  to Remove TBT from Industrial Wastewater




                  Submitted to:

                  Ruby Cooper
                 Project Officer
            Water Permits Division
       US Environmental Protection Agency
                Ariel Rios Bldg
            Washington, D.C. 20460



                       by


                 Gary C. Schafran
 Civil and Environmental Engineering Department
             Old Dominion University
             Norfolk, VA 23529-0241




                  May 15, 2003
Summary of Recent Study Efforts

Research efforts in the most recent quarter were directed toward completion of the long-
term study of tributyltin (TBT) removal with a laboratory granular activated carbon
(GAC) column. This study was conducted over a nine-month period and for the last
seven months of the study, was run on a continuous basis; during the first two months the
column was run intermittently. At the conclusion of column operation, the GAC in the
column was extruded and sectioned in 1.5 cm intervals. Portions of each section were
extracted with methanol to determine the organotin content as a function of depth.

         The influence of organic matter on the decomposition of TBT by UV photolysis
was examined in a series of laboratory experiments since organic matter may interfere
directly or through secondary reactions with the decomposition of TBT. Organic matter,
measured as dissolved organic carbon (DOC), concentration, have been observed to vary
in the final effluent of the full-scale treatment plant by approximately a factor of five and
elevated DOC concentrations may inhibit TBT decomposition. The organic matter stock
solution used in this study was generated by grinding and extracting organic matter from
fouling organisms that were collected from a waterway in southeastern Virginia. This
stock solution was generated and provided by Dr. Michael Unger from the Virginia
Institute of Marine Science through a collaborative Sea Grant project.

        Previous study efforts with the full-scale treatment plant revealed that particulate
TBT was often the dominant fraction of TBT in the final effluent and that polymer was
also present in the final effluent. A study to examine interactions between polymer and
TBT was conducted in the most recent quarter to determine whether TBT complexes with
or can be absorbed into polymer molecules. If this interaction occurs, it may enhance the
transport of TBT through the GAC columns and into the final effluent of the full-scale
treatment plant.

        Proposals were received in response to the request for proposals1 and reviewed by
the study technical committee. A system consisting of three, 80-kW UV reactors
proposed by Trojan Technologies, Inc. was selected and will be delivered this summer.




1
 Appendix 1, February 15, 2003 Progress Report for USEPA Grant S-82874601-1, Evaluate Pilot and Full-
Scale Treatment Processes to Remove TBT from Industrial Wastewater.



                                                                                                   2
Long-Term GAC Column Study of Tributyltin Removal

The long-term, laboratory GAC column study has been described in previous progress
reports and the setup and operational conditions will be summarized here. The focus of
the study was to generate a TBT breakthrough curve and develop an understanding of the
adsorption capacity for the conditions under which the column was operated. The study
began in July 2002 with influent solutions made up daily consisting of deionized-distilled
water spiked with TBT-Cl and NaHCO3 to achieve an influent pH value of approximately
7.0. During the study, influent conditions were changed at various points and these are
summarized below (Table 1). Samples were collected at approximately daily intervals
and subsequently analyzed for TBT and other organotin and inorganic tin concentrations.
From the start until bed volume 1,500 the column was operated intermittently (10 to 12
hours per day, five days a week). The volume of water treated under this condition
represents 7% of the total volume treated during this study. After bed volume 1,500 the
column was run in a continuous flow mode and was interrupted only for a short duration
each day while switching over to a new influent solution.



                                   Table 1.
           Long-term, Laboratory GAC Column Operational Conditions

        Bed Volumes                      Influent Composition
          0 to 3,293      TBT = 2 x 106 ng/L;                 pH = 7.0
                                         5
       3,294 to 16,227    TBT = 8.2 x 10 ng/L;                pH = 7.0
      16,228 to 17,303    TBT = 0 ng/L                        pH = 7.0
      17,304 to 18,276    TBT = 0 ng/L                        pH = 8.5
      18,277 to 19,266    TBT = 0 ng/L                        pH = 5.5
      19,267 to 22,630    TBT = 8.2 x 10 5 ng/L               pH = 7.0



        Although TBT was the only tin species spiked into the influent water, dibutyltin
(DBT), monobutyltin (MBT), and inorganic tin were all observed in the column effluent
along with TBT (Figure 1; inorganic tin not shown). The low concentrations of DBT,
MBT, and inorganic tin in the early stages of the study could have resulted from these
constituents being present in the stock TBT solution that passed through the column.
However, the concentration of all tin species rose dramatically over the course of the
study indicating that the other tin species were by-products of TBT decomposition. This
trend clearly indicates that TBT is not just removed by adsorption onto the carbon surface
but that reactions occurring at the surface are causing its degradation. As noted in the
February 2003 progress report, biological activity within the carbon may play an
important role and may facilitate the decomposition of TBT.




                                                                                        3
                                               TBT              DBT                MBT
                       1000000


                           100000
   Organotins ng/l as Sn




                           10000


                            1000


                             100


                              10
                                    0   5000         10000         15000         20000   25000
                                                 Cumulative Bed Volume Treated




             Figure 1. Organotin concentrations in the effluent of the long-term, laboratory GAC
                       column study. Vertical lines indicate the beginning and ending of a period
                       where influent waters were not spiked with TBT.



        For a period of approximately 3,000 bed volumes (x days) the influent to the
GAC column was prepared without the addition of tributyltin and at three different pH
values, pH 7.0, 8.5, and 5.5, in succession (Table 1). During this period the effluent TBT
concentrations were observed to decline as expected but it was clear that
desorption/mobilization from the GAC column was occurring (Figures 2 and 3).
Tributyltin concentrations were elevated throughout this period averaging (± s.d.) 5,350 ±
4,590 ng/L as Sn at pH 7, 933 ± 851 ng/L as Sn at pH 8.5, and 1,849 ± 922 ng/L as Sn at
pH 8.55. Consequently, for a highly saturated column of GAC, when water of lower
TBT concentration is passed through the column, it appears that desorption will occur to
achieve equilibrium between the carbon surface and the aqueous phase.




                                                                                                 4
                      10                                                             1000000

                                  pH       TBT
                                                                                     100000
                       9




                                                                                                      TBT (ng/L as Sn)
                                                                                     10000
Effluent pH




                       8
                                                                                     1000
                       7
                                                                                     100

                       6                                                             10

                       5                                                             1
                            0      5000          10000       15000     20000     25000
                                                 Bed Volumes Treated



                     Figure 2. Effluent TBT concentrations and pH values for the long-term,
                               laboratory GAC column study.




                        9                                                             100000



                        8

                                                                                               TBT (ng/L as Sn)
                                                                                      10000
       Effluent pH




                        7

                                                                                      1000
                        6         pH

                                  TBT

                        5                                                             100
                        15000      16000         17000        18000    19000      20000
                                                 Bed Volumes Treated



                     Figure 3. Effluent TBT concentrations and pH values for the period of zero TBT
                               influent concentrations for the long-term, laboratory GAC column
                               study. Effluent pH values correlate with influent pH values of 7, 8.5,
                               and 5.5.




                                                                                                                         5
        The influent pH variation portion of this study was conducted to determine to
what extent the variation in pH of waters being passed through a GAC column might
affect the mobility of TBT previously removed on granular activated carbon. During this
period, the highest concentrations were observed at pH 7, immediately after beginning
the zero TBT influent concentrations and the lowest concentrations were observed
during the pH 8.5 influent phase. Based on adsorption isotherm data it was expected that
at higher pH, elevated release of TBT might occur resulting in higher effluent TBT
concentrations and that at pH 5.5, TBT release would be minimal due to greater affinity
for the carbon surface. Higher TBT concentrations were observed at pH 5.5 indicating
greater release of TBT from the carbon relative to pH 8.5. This observation contradicts
previous adsorption isotherm observations that related greater adsorption at lower pH
values. A similar trend was observed with DBT, however, MBT and inorganic tin were
both elevated at higher pH values and were lower at pH 5.5 (Figure 4, MBT only). As
noted previously, it appears that there is a microbial (bacteria) influence on the removal
of TBT. It is possible that variations in the release of TBT that were observed in this
effort may not only reflect equilibrium partitioning between the aqueous phase and the
GAC surface but may also reflect uptake/release of TBT and possibly other tin species
from bacteria. The potential role of bacteria as adsorbers of TBT is currently being
evaluated.




            10                                                               1000000
                          pH
                                                                             100000
             9
                          MBT
                                                                             10000


                                                                                      MBT (ng/L as Sn)
             8
    pH




                                                                             1000
             7
                                                                             100

             6
                                                                             10

             5                                                              1
                 0       5000       10000       15000       20000       25000
                                   Bed Volumes Treated


         Figure 4. Effluent MBT concentrations and pH values for the long-term,
                   laboratory GAC column study.




                                                                                                         6
        During the course of the column study, it became apparent that the organotin
species in the effluent were changing in concentration and in relative amounts to each
other (Figure 5 and 6, Table 2). Initially, the ratio of TBT to DBT was relatively high
with an average ratio of 12.1 (n = 27). After switching to continuous column operation
the ratio of TBT to DBT decreased (2.2 over the next 27 samples and 1.2 through the
remainder of the study). This trend is consistent with development of an active
population of microorganisms and increased conversion of TBT to DBT over time with
lower adsorption of DBT relative to TBT. The ratio of DBT to MBT was also observed
to change over the course of the study with the ratio of DBT to MBT increasing. DBT
has been observed to be more favorably adsorbed to GAC than MBT (this study,
unpublished data) so the increasing ratio would not be expected if DBT and MBT were
present at similar concentrations before GAC adsorption. Based upon this trend, it would
appear that biological activity or a GAC-surface mediated reaction preferentially
degraded TBT to DBT at a rate that exceeded the conversion of DBT to MBT over the
longer-term operation of the GAC column.




                                                  20
       Concentration Ratio of Indicated Species




                                                  18              TBT:DBT

                                                  16              DBT:MBT

                                                  14
                                                  12
                                                  10
                                                   8
                                                   6
                                                   4
                                                   2
                                                   0
                                                       0   5000        10000       15000     20000   25000
                                                                       Bed Volumes Treated



           Figure 5. Effluent organotin ratios for the long-term, laboratory GAC column
                     study. Concentration ratio is also equivalent to molar ratio due to
                     species being represented on an “as Sn” basis.




                                                                                                             7
                                                       Table 2.
                                       Average Concentrations of Tin Species in the
                                Small-Scale GAC Column Effluent For Indicated Periods of
                                                     Bed Volumes

                                                    Average Concentration (ng/L)
                                Bed Vol.      TBT     DBT        MBT       Inorganic   Total
                                0-5K             266      94     129       383    871
                                5K-10K          7695    4799    2261     3900 18,655
                                10K-15K       31,319 29,850     8412     8872 78,453
                                15K-20K       22,217 30,945 12,851 19,125 85,138
                                20K-23K       88,279 115,792 19,548      9666 233,285
                                                       Average Percentage
                                0-5K            30%     11%     15%       44%   100%
                                5K-10K          41%     26%     12%       21%   100%
                                10K-15K         40%     38%     11%       11%   100%
                                15K-20K         26%     36%     15%       22%   100%
                                20K-23K         38%     50%      8%        4%   100%




                               120%

                               100%
Tin Species Distribution (%)




                                                                                               Inorganic Sn

                               80%
                                                                                               MBT

                               60%                                                             DBT

                               40%                                                             TBT

                               20%

                                0%
                                       0-5K      5K-10K    10K-15K      15K-20K   20K-23K
                                                          Bed Volumes


Figure 6. Distribution of effluent organotins as a function of period of column
          operation for the long-term, laboratory GAC column study.




                                                                                                              8
         The relationship between TBT and the byproducts of its decomposition are
important to discern to better understand the processes occurring within a GAC column.
Byproducts of TBT decomposition were examined in terms of the relationship with TBT
to better understand how the GAC column affects TBT removal (Figure 7). The
relationships were examined for three different periods: 0-16,227 Bed Volumes (BV),
16,228- 19,226 BV, 19,227 – 22,630 BV. These periods correspond to prior to, during,
and following the period where no TBT was included in the influent. These periods are
differentiated due to potentially different conditions affecting the relationships between
the tin species.

         Dibutyltin concentrations were positively correlated to TBT concentrations (DBT
= 1.08TBT + 788; r2 = 0.76; excludes the period during when influent TBT = 0 ng/L)
suggesting similar influence on the mobility for each of these constituents (Figure 7a).
Prior to the period when no TBT was in the influent the concentrations of the two species
in the effluent were similar. When TBT was removed from the influent, TBT
concentrations declined by a factor of over 300 while DBT concentrations declined by
only a factor of 10. Since neither TBT nor DBT was present in the influent, this trend
indicates preferential loss of DBT relative to TBT from the GAC surface that may be
related to equilibrium partitioning. TBT has a higher affinity than DBT for GAC surface
and if present at similar concentrations on the surface of the GAC it would be expected
that in the absence of either constituent in the influent, that effluent concentrations of
DBT would be higher than TBT. TBT and DBT concentrations both increased to similar
and higher concentrations following reintroduction of TBT to the influent indicating that
the period without TBT introduction did not dramatically influence subsequent
adsorption.

        Monbutyltin was positively correlated to TBT (MBT = 0.15TBT + 2023; r2 =
0.48; excludes the period during when influent TBT = 0 ng/L) indicating that TBT
concentrations were also influencing the formation and/or release of MBT from the GAC
column (Figure 7b). During the period when TBT was not present in the influent MBT
concentrations exhibited considerable scatter with a number of samples at concentrations
elevated above the observed trend prior to and after the TBT-free influent. These
elevated MBT concentrations reflected the pH influence previously described where
MBT exhibited enhanced mobilization and/or formation at higher solution pH values.

        Inorganic tin is generated by complete debutylation of TBT and consequently is
without an organic moiety. Because of this condition, it might be expected that inorganic
tin would behave differently to the organotins undergoing GAC adsorption. For the
column study, inorganic tin was positively correlated to TBT but with greater scatter than
observed with DBT and MBT (Figure 7c). As observed with MBT, inorganic tin
concentrations were substantially increased in the column effluent during the pH 8.5
influent period. Comparing effluent inorganic tin concentrations versus pH for the entire
period of the study revealed an apparent pH influence (Figure 8). The only data point
inconsistent with a trend of increased solubility at higher pH, was the one at pH 9.2
which was the first effluent sample analyzed. A low inorganic tin concentration at this
elevated pH value was likely due to the lack of tin retained on the column at this point.



                                                                                             9
                               1,000,000

                                                                                          a)
                                100,000
DBT (ng/L as Sn)




                                 10,000


                                  1,000


                                    100


                                     10
                                           10   100       1,000       10,000    100,000    1,000,000
                                                           TBT (ng/L as Sn)

                               1,000,000


                                100,000
                                                                                           b)
MBT (ng/L as Sn)




                                  10,000


                                   1,000


                                    100


                                      10
                                           10    100       1,000       10,000   100,000        1,000,000
                                                           TBT (ng/L as Sn)


                               1,000,000


                                                                                           c)
                                 100,000
        Inorganic Tin (ng/L)




                                  10,000



                                   1,000



                                     100
                                           10    100       1,000       10,000   100,000        1,000,000
                                                            TBT (ng/L as Sn)

                                Figure 7. Relationships between TBT and TBT byproducts for the long-term,
                                          laboratory GAC column study. ( 0-16,227 Bed Volumes, -
                                          16,228- 19,226 BV, 19,227 – 22,630 BV)



                                                                                                            10
                            100000



                             80000
    Inorganic Tin (ng/L)




                             60000



                             40000



                             20000



                                0
                                     4      5         6          7         8          9         10
                                                                pH


                           Figure 8. Effluent inorganic tin concentrations as a function of pH for the long-
                                     term, laboratory GAC column study. All data points included.




        The relationship between the adsorption density (adsorbed TBT) and effluent
TBT concentration is an important association to understand and was examined for this
column study. Calculation of the apparent adsorption density was made by conducting a
mass balance around the GAC column and assuming that any TBT or its byproducts not
accounted for was adsorbed on the column. The mass of TBT retained on the column
was calculated based on the following expression calculated each time a sample was
retrieved (calculated on a tin basis):

                           Apparent adsorbed TBT = TBTin – TBTout – DBTout – MBTout – Inorganic tinout

The expression is based on the assumption that any DBT, MBT, or inorganic tin observed
in the effluent originates from TBT in the influent that was converted to these species
while in the column.

        The relationship between apparent adsorbed TBT and effluent TBT
concentrations was observed to follow the trend of a typical adsorption isotherm with
increasing effluent TBT concentrations at higher adsorbed TBT (Figure 9).
Concentrations were generally below 2,000 ng TBT/L up to 50 mg TBT/g GAC while
above this apparent adsorption density, effluent concentrations were observed to increase
dramatically.



                                                                                                           11
                              125
                                                                                                                   a)
                              100
TBT Adsorbed (mg/g GAC)




                              75


                              50                                                                 Bed Volumes
                                                                                                   0 - 16.3K
                              25                                                                   16.3 - 19.3K
                                                                                                   19.3 - 22.6K

                              -
                                       -        50,000     100,000     150,000     200,000    250,000     300,000       350,000

                                                                  Effluent TBT Concentration (ng/L)




                              125

                                                                                                                        b)
                                                Bed Volumes
                              100
    TBT Adsorbed (mg/g GAC)




                                                   0 - 16.3K
                                                   16.3 - 19.3K
                                  75               19.3 - 22.6K



                                  50


                                  25



                              -
                                           10            100            1,000          10,000            100,000          1,000,000

                                                                     Effluent TBT Concentration (ng/L)


Figure 9. Relationship between effluent TBT concentrations and apparent
          adsorption density TBT (calculated). ( before,       during, and •
          after the period of zero influent TBT concentration). Effluent
          concentration plots: (a) rectilinear, (b) log10 .




                                                                                                                                      12
        The different periods previously identified in the long-term column study were
plotted individually in Figure 9 to illustrate an apparent change that occurred to the
relationship between the adsorbed and the effluent TBT concentrations. After the period
when TBT was left out of the influent during which small amounts of TBT and other tin
species were observed to be removed from the column (calculated reduction of 0.9 mg
TBT/g GAC) the TBT concentrations were shifted lower (to the left in Figures 9a and 9b)
relative to extrapolation of the trend observed prior to the period of no influent TBT.
This may reflect some change to adsorbents on the carbon surface (decomposition?) that
provided additional available surface when TBT was reintroduced to the influent.

        Following the completion of the long-term, laboratory GAC column study, the
GAC column was extruded intact and sectioned into six, 1.5-cm depth sections. A
portion of each GAC section (average mass = 0.93 gram) was then extracted by placing
the GAC in 250 mL of 99% methanol and placing on a shaking table for 24 hours. After
24 hours the GAC was allowed to settle and the solution was carefully poured off leaving
the GAC behind. New methanol was then added and the procedure was repeated again.
This procedure was conducted three times for each portion of GAC resulting in three
samples (labeled 24, 48, and 72 hour) for TBT analysis. The tin species were measured
in each solution and the results were used to calculate the amount of each tin species
desorbed from the GAC column. Results illustrated that the majority of the tin species
were removed from the carbon with the first extraction (24 hours) with successively
smaller amounts with each additional extraction. The amount of TBT extracted for each
section as a function of time illustrates the variation for all species (Figure 10).



                    0

                    1

                    2

                    3
       Depth (cm)




                    4

                    5

                    6                                                 24 hrs
                                                                      48 hrs
                    7
                                                                      72 hrs
                    8

                    9
                        0   1   2    3       4       5       6    7            8   9
                                    TBT Desorbed (mg TBT/g GAC)



       Figure 10. TBT desorbed by depth as a function of extraction time for the long-
                  term, laboratory GAC column study. Depth is measured from the top
                  of the GAC column first received the influent passing through the
                  column. Times indicate extraction times.



                                                                                       13
        Extraction of the GAC with methanol allowed characterization of the tin species
that were retained (and extractable) on the GAC column as a function of depth.
Surprisingly, tributyltin was one of the minor tin components on the GAC with MBT and
DBT present on the GAC at greater concentrations than TBT (Figure 11). It is clear from
the trend that the majority of removal occurred in the upper portion of the column with
80% of the tin species present in the top one-third of the column and 94% present in the
top half. The distribution of tin species in the column is consistent with mass transfer
expectations with elevated driving forces (transport from the bulk phase) and adsorption
occurring at the point of highest concentration – at the top of the column. The low
adsorption at the lower portion of the column is consistent with the observation that 85-
95% TBT removal was still occurring at the end of the column study.
         Depth From Top of Carbon Column (cm)




                                                0.75

                                                2.25

                                                3.75

                                                5.25

                                                6.75
                                                           TBT   DBT        MBT   Inorganic Sn
                                                8.25

                                                       0   5           10           15           20

                                                                 mg Sn/g GAC


       Figure 11. Distribution of organotin and inorganic tin concentrations as f unction
                   of depth for the long-term, laboratory GAC column study.



        It is clear from the findings of the long-term laboratory GAC study that TBT was
significantly converted to other tin species and that these tin species were both lost from
the column and retained on the GAC surface. Unknown is the mechanism that is driving
the conversion of TBT. GAC is capable of promoting the conversion of organic
compounds abiotically through reaction with the GAC surface and this may explain the
conversion, particularly in the early phase of the study. However, biological removal
(adsorption) and conversion may also play an important factor, particularly as



                                                                                                      14
performance was observed to significantly improve when conditions favoring microbial
growth and activity (i.e. continuous flow) occurred. Studies are currently being
conducted to evaluate the potential for microbial communities to provide additional
adsorptive capacity.


Laboratory UV Studies

Laboratory studies examining UV-initiated decomposition of TBT have been conducted
to better understand the influence of solution composition and better elucidate the
mechanisms of TBT decomposition. Specific studies are described below.

Effect of DOC on TBT Removal by UV Treatment

       Dissolved organic matter (measured as DOC) is present in all shipyard waters and
at concentrations after treatment with GAC at levels between 1 and 5 mg/L. Since DOC
may directly absorb UV light or react with oxidants formed during UV irradiation, it is
important to understand what how DOC may affect the rate of TBT decomposition.

        Removal of TBT was examined in batch mode using synthetic waters (DI spiked
with TBT) and adjusted to five DOC concentrations. These values covered a range that
would be expected at the end of the treatment process train currently being used. The
influent water was prepared by adding 3.7 ml of TBT stock solution (1.022 mg/L TBT as
Sn) and varying amount of DOC stock solution (357 mg C/L) in 3.79 L of DI water to get
a concentration of 1000 ng/L TBT as Sn. The DOC stock solution was prepared at the
Virginia Institute of Marine Sciences by grinding and extracting aquatic organisms that
would be similar to those that might attach to ship hulls and then acidifying the solution
to prevent biological deterioration. The pH of each prepared water was adjusted to 7
using 0.1N NaOH.

        The results illustrate that DOC lowers the rate of removal of TBT but that the
impact appears to be more substantial at low DOC concentrations and that at increasing
DOC concentration the effect decreases (Figure 12). At zero DOC, the first-order rate
constant for DOC was observed to be 0.23 min-1 while at DOC concentrations between
2.2 and 9.5 mg C/L the rate constant varied from 0.14 to 0.12 min-1, respectively. The
decrease in TBT degradation in the presence of DOC could be due to competitive
absorption of light by DOC (i.e. DOC serves as a light filter). However, it would be
expected that there would be a greater effect of increasing concentrations of DOC than
observed here. Consequently, reduction in the rate of TBT decomposition may be
through a secondary reaction where the available DOC was sufficient at low DOC to
affect TBT decomposition by an alternate, indirect pathway while direct UV photolysis
was primarily unaffected.




                                                                                       15
                   0.25


        k (min )
        -1
                    0.2


                   0.15


                    0.1
                          0   2         4          6         8         10
                                        DOC (ppm)

       Figure 12. First-order rate constants for TBT decomposition as a function of
                  DOC concentration in solution.




Effect of DOC on the Photodecomposition of Hydrogen Peroxide

Hydrogen peroxide has been observed to increase the rate at which TBT is decomposed
and it is planned for use in the full-scale system to be acquired. DOC is known to affect
the UV decomposition of H2O2 by absorbing UV light as well as by acting as a scavenger
of hydroxyl radical which is produced by the decomposition of hydrogen peroxide. In
this manner, DOC can influence the chain reaction of hydrogen peroxide decomposition
and potentially conversion of TBT. The effect of DOC on the photodecomposition of
H2O2 was investigated with the laboratory reactor as described above.

        The DOC source was the same as noted above (VIMS) and three different DOC
concentrations were investigated, 2.5, 5 and 10 mg C/L. Solutions were spiked with 50
mg/L of H2O2 and irradiated for the desired period of time. At the end of the desired
irradiation period, the lamps were turned off and the sample was immediately analyzed
for H2O2 concentration. The results illustrate an effect of DOC on the decomposition of
H2O2 that is similar to that observed for the decomposition of TBT. The rate of
decomposition was affected most significantly between 0 and 2.2 mg C/L of DOC and
much less above this DOC concentration (Figures 13 and 14). The similarity of the effect
of DOC on the decomposition of these constituents suggests that the mechanism by
which DOC interferes with their decomposition is similar.




                                                                                      16
              0.00
                                                                 Blank
                                                                 0 ppm
ln(C/Co)
           -1.00
                                                                 2.5 ppm
                                                                 5 ppm
           -2.00
                                                                 10 ppm

           -3.00
                         0         4          8           12
                                    Time (min)

Figure 13. Natural log of the ratio of hydrogen peroxide remaining at each
           irradiation time and DOC concentration. Note that blank was as 0 mg
           C/L DOC solution placed in the UV reaction vessel but the lamps
           were not turned on.




                   0.3
                                                                 k (H2O2)
                   0.2                                           k (TBT)
       k (min-1)




                   0.1

                    0
                         0   2.5        5       7.5       10
                                   DOC (mg C/L)

Figure 14. Rate constants for H2O2 and TBT decomposition by UV photolysis as
           a function of solution DOC concentration.




                                                                           17
TBT Interactions With Coagulant Aids

Previous study efforts with the full-scale treatment plant revealed that particulate TBT
was often the dominant fraction of TBT in the final effluent and that polymer was also
present in the final effluent. A study to examine interactions between polymer and TBT
was conducted to determine whether TBT complexes with or can be absorbed into
polymer molecules. In this experiment, deionized-distilled waster was spiked with
tributyltin chloride to a concentration of 2,440 ng TBT/L and then separated into six 1-
liter aliquots to which a 1 mg/L polymer dose was applied to each. Cationic, anionic, and
nonionic polymers were all examined as well as a control (no polymer). Following a
period of mixing, an aliquot of each solution was filtered through a 0.1 µm filter and then
TBT analyses were conducted on the filtered and unfiltered samples for each solution.
Previous work with polymers had exhibited that large, organic polymers could be
retained on a 0.1 µm filter. Hence, if TBT was associating with the polymers, it was
expected that the TBT would be captured on the filters and measured at much lower
concentration in the filtrate.

         TBT concentrations in the majority of the unfiltered solutions were relatively
close to the initial concentration suggesting that the borohydride-TBT analytical method
was capable of measuring TBT in the presence of the polymers (Figure 15). The one
exception was the Selfloc-2250, an anionic polymer that was measured at 23% below the
initial TBT concentration. The filtered sample TBT concentrations were all significantly
lower in the polymer containing samples, particularly the cationic and anionic solutions.
This result indicates that TBT does associate with the polymers and that if polymers carry
through a treatment system, which we have previously observed that they do, then they
may be capable transporting associated (complexed, absorbed, or adsorbed?) TBT.

        An interesting finding was made in preparing to conduct the polymer study. Each
polymer was added to DI water and measured for TBT prior to the experiments to prove
that they were TBT free (this was expected). However, TBT was measured in one
polymer (Magnifloc E30) at a concentration of 123 ng/mg polymer. The presence of
TBT was confirmed when replicate solutions were made.




                                                                                        18
                 2500

                 2000
    TBT (ng/L)




                 1500

                 1000

                  500

                    0
                            l




                                                                                            )
                                         42




                                                                         73
                                                       30




                                                                                                       ic)
                           tro




                                                                                          c
                                                                                       ni
                                       E-




                                                                     78




                                                                                                      on
                                                     E-
                        on




                                                                                        o
                                                                                     ni




                                                                                                    ni
                                      ic




                                                   c
                     C




                                                                    ZE




                                                                                   (A
                                                     i




                                                                                                  (A
                                       n




                                                  on
                                    io




                                                                 c




                                                                                                  50
                                                                              11
                                                 i
                                 An




                                                               ni
                                              on




                                                                                                22
                 Unfiltered
                                                               io




                                                                              10
                                              N




                                                            at




                                                                                              c-
                                                                            c-
                                                         C




                                                                                          lf lo
                 Filtered                                                flo
                                                                       na




                                                                                       Se
                                                                     ag
                                                                    M




                 Figure 15. Filtered and unfiltered TBT concentrations measured in the indicated
                            polymer solutions. Polymer concentration was 1 mg/L as product.
                            Control contained no polymer.




Acquisition of a Full-Scale Ultraviolet Treatment System

Proposals were received in response to the request for proposals2 and reviewed by the
study technical committee. A system consisting of three, 80-kW UV reactors proposed
by Trojan Technologies, Inc. was selected and will be delivered in early August.


Future Efforts

The delivery of the full-scale treatment plant will occur in August and will be integrated
with the CASRM treatment plant and subsequently performance tested. The influence of
the operational condition of the CASRM treatment plant in front of the UV system and its
impact on the UV system will be examined in detail.



2
 Appendix 1, February 15, 2003 Progress Report for USEPA Grant S-82874601-1, Evaluate Pilot and Full-
Scale Treatment Processes to Remove TBT from Industrial Wastewater.



                                                                                                             19
        Additional UV experiments are being conducted to characterize the DBT and
MBT rates of decomposition so that an overall model of organotin decomposition by UV
photolysis can be developed. A number of continuous flow GAC column experiments
have been conducted with TBT-containing shipyard waters. TBT, copper, and zinc
removal have been examined and efforts have recently been completed on the studying
the effect of backwashing on the remobilization of these metals. The influence of
sudden changes in pH on metal remobilization from the columns has also been studied.


Acknowledgments

Research results presented in this report have been due to the efforts of Old Dominion
University research assistants Ram Prasad, Prem Bhandari, Khalid Qadwai, Shilpa
Shivakumar, and Steven Roelands. In addition, TBT (and other butyltins) analyses and
identification of other tin compounds in some of the long-term GAC column study
samples have been conducted at the Virginia Institute of Marine Sciences under the
direction of Dr. Michael Unger. The DOC used in the UV experiments was also
developed in Dr. Unger’s laboratory.

       Operational management of the CASRM full-scale treatment plant has been
coordinated by John Soles, Frank Thorn, and Chris Walter of Northrop-Grumman
Newport News Shipbuilding (NG-NNS) and Mike Ewing of Norshipco and Frank
Wheatley of Colonna’s Shipyard.




                                                                                         20

								
To top