NCD980602163_19950629_Warren County PCB Landfill_SERB C_Evaluation of PCB Bioremediation in NC Landfill-OCR

JUL 14 '35 02:33PM MWIP/ORNL *6155767865
LOCKHEED MARTIN ENERGY SYSTEMS, INC.
June 29, t 995
Sharron Rogers
State of North Carolina
Department of Environment, Health and Natural Resources
P.O. Box 27687
Raleigh, Norrh Carolina27611-7687
POST OFACE BOX 2009
OAK RIDGE. TDINessee 37831
Evaluation of Polychlorinated Biphenyl (PCB) Bioremediatioo in North Carolina Landfill
Enclosed is a short report evaluating the microbial transfonnation of PCBs in the landfill and the
possibility of bioremediation as an in siru treannent option. The results are quite encouraging
depending on your treatment objectives and goals.
Please contact me after you have had a chance to review the report.
K. Thomas Klassen
Remediation Technology Group
KTK:ccn
Attachment 1) Evaluation of PCB Remediation .Possibilities in North Carolina Landfill
c w/att: A. A. Barnes
J. W. Barton
M. E. Reeves
S. M. Robinson
ncpcbcval .!tr
J.UL 14 '95 02:34PM MWIP/ORNL t6155767865
Direct Assistance Project No. MTAC950516003 Pl: K. Thomas Klasson, ORNL, [phone]
EVALUATION OF PCB REMEDIATION POSSIBILITIES IN NORTH CAROLINA LANDFILL
BACKGROUND
The State of North Carolina Department of Environment, Health and Natural Resources Division (DEHNR) of
Solid Waste Management manages a landfill of soils contaminated with polychlori.nated biphenyls (PCBs).
Background infonnation show that the major PCB contaminant is Aroclor 1260. It bas been $peculated that the
PCBs in the capped landfill may have undergone bacterial transfonnation and the possibility for bioremediation has
been considered.
OBJECTIVE
Oak Ridge National Laboratory (ORNL) was asked to evaluate PCBs in samples taken ftom the land.fill for
bacterial transformation of the PCB and to predict the extent at which PCB bioremediation may be effective. The
project was arranged by the Direct Assistant Program of Lockheed Martin Energy Systems, Inc.
SAMPLES
Three 1-L glass jars of soil contaminated with PCBs were shipped to ORNL on May 25, 1995 ftom DEHNR.
Four samples were collected from each jar and the subsamples were separately extracted with acetone/hexane (l :5)
and the extract was combined with an internal, standard (octachloronaphtalene) and injected into a gas
chromatograph (GC) equipped with an electron capture detector (ECD). The resulting GC chromatograms were
evaluated by comparing peak sizes to a known standard consisting of an Aroclor mixture (1242, 1254, and 1260).
The extracted soils were dried (80°C for 24 hrs) and PCB concentrations were calculated as concentrations in dried
soil.
ANALYTICAL RESULTS
The average concentration of PCBs in the samples was 382 mg/kg (standard deviation 56 mg/kg), and the
average number of chlorines per biphenyl was S.33 {standard deviation 0.10). The GC chromatogram for a typical
sample is displayed in Fig. 1. As references, GC chromatograms for Aroclor 1242, 1254, and 1260 are shown in
Figs. 2-4.
EV ALUATJON OF ANAEROBIC DECHLORINATION lN SAMPLES
Based on historic data, the contaminated soil should contain mainly Aroclor 1260. This is indeed the case
(compare GC chromatograms in Figs. l and 4); however, a large number of smaller peaks in the DEHNR sample
appearing at GC retention times shorter that 23 rnin arc lacking in the Aroclor 1260 sample (Fig. 4). Also, the
average number of chlorines per biphcnyl (S.33) is approximately 20% lower in the DEHNR sample than in the
Aroclor 1260 standard. This lends itself to speculation that dechlorination may have occurred; however, the
'majority of these peaks are most likely from another Aroclor (e.g., 1242 or 1254).
This conclusion is further supported by the eval1.1ation of the number of PCB chlorines at the ortho, meta, and
para positions in the sample relative to Aroclor mixtures. A variety of data collected from various sources is shown
in Fig. 5. It appears that the ratio between the number of chlorines at different positions in the DEHNR sample is
consistent with an Aroclor mixture. The mixture of Aroclor 1242, 1254, and 1260 that best represents the DEHNR
sample is 12 wt% 1242, 27 wt°/o 1254, and 61 wt¾ 1260. This conclusion is based on a non-linear regression
analysis of the components in peaks as analyttd by GC/ECD. A graph depicting the correlation between this
Aroclor mixture and the DEHNR sample may be seen in Fig. 6. Points below the line correspond to peaks
containing congeners at diminished levels in the DEHNR sample compared to the Aroclor mixture; similarly, points
above the line correspond to peaks containing congeners at elevated levels in the DEHNR sample compared to the
Aroclor mixture. Four po.ints, corresponding to peaks 31, 32, 48, 53, and 61, are quite noticeable at some distance
from the straight line. A closer look reveals even more relative differences between the DEHNR sample and the
Aroclor mixture as summarized below together with a possible dechlorination pattern.
Congeners at elevated levels: 25-3, 25-25, 24-25. 24-24, 34-4, 23~24, 24-34, and 236-24.
Congeners at diminished levels: ~31, ~5~-25, 24~:i:. 236-34, ~~3~236 .
......_, Underline indlcares attack position I
JUL 14 '95 02=35PM MWIP/ORNL *6155767865
Direct Assistance Project No. MTAC950516003 Pl: K. Thomas Klasson, ORNL, [phone]
The same conclusion was drawn by researchers at General Electric (1). The dechlorination pattern is consistent
with removal of very limited doubly-flanked meta, singly-flanked para, and singly-flanked meta chlorines.
Stat. of NC DEHNR (Sample 3.3)
13 17 21 25 29
lnl-1
Standtr'CI
I
33
Fig. 1. Typical GC chromatogram of DEHNR sample. (Values
on ordinate correspond to GC retention time.)
Aroclor 1242
1r11pumy
j
13 17 21 15 29
Fig. 2. Aroclor 1242 standard GC chromatogram.
Aroclor 1254
13 17 21 25 29
Fig. 3. Arotlor 1254 standard GC chromatogram.
lllletnal
Sta11dard
I
33
llllemal
Slandard
I
33
~UL 14 '95 02:35PM MWIP/ORNL t6155767865
Direct Assistance Project No. MT AC950516003 Pl: K. Thomas Klasson, ORNL, [phone]
Aroclor 1280
13 17 21 25 29 33
Fig. 4. Aroclor 1260 standard GC chromatogram,
Average No, of
par .. m,~
Clllolln11
• M, Iii
4 ■ "''· (3) .ORNL
•DEHNR
1
A1"1Df 1ll&O
An,c;lot 1254
• • :--. OEHMU.~
Al'OClOI' 1142"a • An>clor 1018 •
ArDcllar1~1
0+---------~-----.----
0 2
Awr&gl No. of ol1ho ChlotlrlH
Fig. 5. Average number of PCB chlorines in different positions of
various samples.
peak l2
z ••• . -,
• pHk41
• peak31 / .
• PMkfl
• • f1Hkli1
Fig. 6 .. Comparison of DEHNR sample with Aroclor mixture
consisting or 12 wt% 1242, 27 wt¾ 1254, and 61 wt¾ 1260.
JUL 14 '95 02:35PM MWIP/ORNL ♦6155767865
Direct Assistance Project No. MTAC950516003 PI: K. Thomas Klasson, ORNL, [phone]
CONSTRUCTION OF A TOOL FOR PREDICTION OF PCB BlOREMEDIATION
To aid in evaluation of bioremcdiation possibilities for the DEHNR landfill, a computer spreadsheet was
developed based on the susceptibility of PCB congeners to undergo bacterial transformations (4,S). Several
anaerobic and aerobic attacks were considered and transfonnation of each PCB congener was evaluated based on
the position of chlorines on the biphenyl rings and the type of bacterial attack. Also, to evaluate bioremed.iation as a
remediation alternative, half.life expectancy in humans together with dioxin-like structures of each PCB congener
were also incorporated into the computer model .(6, 7). Based on GC analysis of DEHNR samples and the
assumption that organisms capable of anaerobic removal of doubly-flanked meta, singly-flanked para, and
singly-flanked mer.a chlorines are present in the landfill, the possibility of in situ bioremediation was evaluated. The
results are shown as they appear on the computer scn:en in Fig. 7.
Spttadsh~I 0.-.,fjQpm.!ll
Tl'liS Ji,t'taGShttt WY. QOvoiopod ~. ,.su~ Of a DOE Oirf'Ct Aislil•r.o•
PIC0,<1mbo)
K.Tllomis~ion
~t l'l!Og,t llladonol L.t>cretor,
P .0. Be• 2009,l;OH, ~k l'lldjl~ TN 371'31 ~°' llli/674-2210
Fig. 7. Computer screen layout for computer prediction of PCB bioremcdiation. The results listed are for
prediction of PCB bioremediation at the DEHNR landfill. (Also part of the computer output, but not shown above,
is a complete listing of individual congener concentrations.)
As noted in Fig. 7, the predicted treatment is quite effective in reducing heavily chlorinated PCB compounds and
producing less chlorinated congeners as indicated by the shift in homolog levels. The majority of congeners
remaining after anaerobic dechlorination are tetracblorobiphenyls. The products were found to have a shorter
half-life expectancy in humans, and dioxin-like compounds were removed by the biotransformation. The total
concentration of PCBs dropped 15% due to the change in avefaie molecular weight of the congener mixture and
does not indicate a reduction of total PCB molecules., only a change in their relative concentrations.
Computer prediction of aerobic attack by 2,3-dioxygenase-containing organisms on the congeners present after
anaerobic transfonnation was also performed. The results indicate that the PCB concentration present after
anaerobic treatment (31 S .3 mg/kg) could be reduced to 182 mg/kg under optimal conditions.
~UL 14 '95 02:37PM MWIP/ORNL ♦6155767865
Direct Assistance Project No. MTAC9S0S16003 Pl: K. Thomas Klasson, ORNL, [phone]
CONCLUSIONS
1. Analyses of soil samples taken from the DEHNR PCB landfill indicate that an Aroclor mixture exists in the
contaminated soil wjtb over 60% of the PCBs pri:sent as Aroclor 1260.
2. GC peak profiles in extracts from soil samples showed a bacterial dechlorination pattern consistent with
removal of very limited doubly-flanked meta, singly-flanked para, and singly-flanked meta chlorines.
3. Under optimal conditions, the dechlorinating bacteria may substantially reduce the fraction of heavily
chlorinated compounds, reduce the half-life expectancy of PCBs in humans, and eliminate the dioxin-like
congeners.
RECOMMENDATIONS
1. Conduct treatability studies to determine the possible amendments needed to promote indigenous organisms to
dechlorinate PCBs present in the landfill.
2. Engineer landfill to accommodate addition of amendments and extraction of samples.
3. Investigate possibility of introducing air into the subsurface to promote aerobic degradation as a sequential step
to anaerobic treatment.
REFERENCES
1. Letter Report from F. J. Mondello (General Electric) to Sharron Rogers (DEHNR), August 26, 1994.
2. J. M. Tiedje et al., Biodegradation, 4, 231-240 (1993).
3. D. E. Schulz et al., Environ. Sci. Technol., 23 (7), 852-859 (1989).
4. W. A. Williams, Environ. Sci. Tech.no/., 28 (4), 630-635 (1994).
S. D. A. Abramowicz, Crit. Rev. Biotechnol., 10, 241-251 (1990).
6. J. F. Brown, Jr., Environ. Sci. Technol., 28 (13), 2295-2305 (1994).
7. S. H. Safe, Crit. Rev. To:cicol., l4 (2), 87-149 (1994).