NCD980602163_19970923_Warren County PCB Landfill_SERB C_Responses to questions on ECO LOGIC Draft Report-OCR

F'.1
SEP 24 -;/Fax Transmittal
ihe infomiation contained in this facsimile message is legally privileged and confidential information interiaed only for ths use of
the individual or entity namea below. If the reader of this message is not the intended recipient, you are hereby notified that any
use, dissemination. distribution er copy of this facsimile is stridly prohibited. If you have received this facsimile ii"\ error, please
Immediately notify us by telephone and return the original rr.essage to us by mail at the address above. Thank you.
Phone:
Fax:
Subject
::
Message:
No. of pages: __ /:....6-=-------
(incuding cover)
From:
I
Company: ELI Eco Logic Inc,
Phone:
(q;q) 7 15-?/?OS Fax: [phone]
~ /4 t44Ju&-d ~ e:tJ db 1cr#( 0.fh
7
(ij/4_{hd ~µ 6«_ -~
ti.JVCL ~ ~ ~ .
kt<~~
SEF' 23 '97 05 : 17PM ELI ECO LOGIC INC.
Comments/questions from Mike Kelly and Patrick Barnes
1) PCB and dioxin removal: Your report indicates problems meeting the treatment level for the
TEQ for Dioxin; have you received any additional tests back from the laboratory, or could you
provide additional information that would demonstrate your ability to meet this dean up goal?
Due to the high detection limits reported by the laboratory for the original analyses of Run 1 and 3
treated soil, ECO LOGIC requested that these samples be reanalysed. Reanalysis was conducted
using a larger sample volume, which resulted in significantly lower detection limits. Tables 1, 2 and
3 provide a summary of the dioxin data for system inputs and outputs for Runs 1, 2 and 3,
respectively, TEQ calculations using half the detection limit for non-detect values, as suggested by
Joel Hirschhorn, were 0.49 and 0.51 ppt for Runs 1 and 3, respectively. Therefore, the values meet
the treatment level for dioxin TEQs of 1 ppt. Problems encountered during Run 2 are addressed
below.
The new data has been included as appropriate throughout the final report.
2) Safety is a key issue for consideration in the full scale operation. Please expand on the safety
issues as it relates to utilization of the gas phase chemical reduction technology during
detoxification. One primary concern is over the use of hydrogen gas and operating under
ambient pressures.
The ECO LOGIC Process uses hydrogen to break down organic contaminants into basic products
such as methane, carbon monoxide and carbon dioxide. These reactions take place in a sealed
system which operates at essentially ambient pressures (within 0.4 psi of atmospheric pressure).
Hydrogen can be explosive only when combined with oxygen or air, and exposed to a spark. ECO
LOGIC has taken numerous measures to ensure that the use of hydrogen does not present a danger
to its on.site workers, and the surrounding environment.
Industrial Use Of Hydrogen
Hydrogen has been used in large quantities in the petroleum refining, chemical, petrochemical and
synthetic fuel industries for decades. Therefore, the use of hydrogen in industry is fairly routine.
The electrical utility indust.ty has also successfully used hydrogen gas for more than forty years, for
such operations as cooling rotor and stator coils in large turbine-generators. Hydrogen is already
the accepted fuel of the aerospace industry, and has been safely handled for years in large quantities.
Although hydrogen has been used in industrial processes for decades, it is a relative "unknown" to
the public. It is therefore important to convey to the public, that there are strict guidelines for the
safe handling and use of hydrogen, and that adequate measures are available and enforced to ensure
the safe use of hydrogen. The ECO LOGIC Process uses hydrogen according to stringent standards
required by both United States and Canadian regulators.
SEP 23 '97 05: 18PM ELI ECO LOGIC INC. F'.3
Page2
Table 1 SUMMARY OF DIOXIN DATA -RUN 1
·Com ouad
2,3,7,8-TCDD <
1,2,3,7,8-PeCDD <
1.2.3,4,i,g-HxCDD <
1,2.3,6.7,8-HxCDD <
: 1.2,3.i.8.9-H..,CDD <
li 1:Z);4:6;7,a-HpCDD
i OCDD i 2,3,7,8-TCDF
I 1.2,3.7,8-PeCDF
12.,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1.2.3,6,7,8-lhCDF
2,3.4.6,7.8-H:xCDF
1,2,3,7,8,9-HxCDF <
1.2,3,4,6,7,8-HpCDF
1,2.3,4,7,8,9-HpCDF
OCDF
,I-1EQ
! Tota.lTCDD <
i Total PcCDD <
: Total HxCDD
I Total HpCDD
15 <
21 <
25 <
19 1 < 22 <
25 1<
3600 :
75 <
31 : <
120 : <
1700,<
300 <
160
19 <
2200 <
960 <
5700 <
350
15 , <
21 : < 25 1 < s20 : <
100 , <
0.3 i <
2.7 ,<
0.2 !<
0.3 , <
0.3 1 <
0.2 <
0.2 <
0.63 <
0.3 <
0.3 <
0.3 <
0.4 ' <
0.51
0.3 , <
0.4 1 <
0.3 <
0.3 ; <
0.2 < ITotalTCDF . ~lPeCDF 900 < OJ < To
1iTo~I HxCDF 3300 0.63 <
1 Total H CDF _. 4900 < 0.3 <
B = Found i'ii'liboratory Mtth00 Bia.nit
PR = Peak is Poorly Resolved -amount is likely overestimated
Product G:u &b~ust Gat
sm3
. < · < 4.6 :
4.2 < ss < i.s ii
11 < 14 < 42 < 7.9 ! 3.1 < 2.0 < 61 I< 111'.
S.2 1< 47 < 7.7 ]
4.s I < 44 < ,.~ .;
4.6 , < 47 < 7.: i
8.41' < 27 < 6,1 ,. I I
3.7 < 3.0 , < 53 < 7.71 3.6 , < 2.9 < 46 < 6.7 ~
5.5 : < 3,7 < 16 < 1.7
4.4 1 < 3.0 < 18 < 1.8 s.s " s.s < 19 < 1.9l1
6.1 < 4,1 I< 25 , < 2.5 1,
4.7 < 4,4 ;< 47 < 1.s ,
7.4 < 6.8 , < .'i.'i ' < 2.1 i
9.1 < 12 ; < 46 < 9.7 :
6.4 . 4.9 63 83
3.61 < 2.6
5.31 < 4.2 I
6,4t< 4.8
I 8,71 < 8.4
3.1 i < 2.0
3.6 ' < 2.9
5.5 :
S.4
-SEP 23 '97 05:18PM ELI ECO LOGIC INC.
Table 2 SUMMARY OF DIOXIN DATA -RUN 2
Treated SoJI
Prc-Ouboll · Post-C.arbotJ. I
Scr""ber W.ate.i Scrubber Wat~
Com uad
2,3,i,8-TCDD < 12.9 3S.8 < 3,1 < 3.6 ' <
1.2.3.7,8-PcCDD 60.4 182 < 6.s' < s.o <
i 1.2.3.4.7,8-HxCDD 82.S 182 < 9.7 < 5.6 <
l,2,3,6,7,8-HxCDD 84 177 1 < 7.9 < 4.4 <
1,2,3,7,8,9-HxCDD !3.6! 196 : < 8.5 < 5.0 <
1.2.3.4.6,7,8-HpCDD 546 ' 166 • < 12.8 < 5.9 1 < ;OCDD 5900 ~ 367 I 234 ,B 6.3 <
2,3,7,B-TCDF 7S.9 , < 18.8 I< 2.41< 2.8 \<
1.2.3,7,8-PcCDF 95 202 '< 41< 3.5 <
2,3,4,7,8-PeCDF 154 i 189 1 < 4,1 i < 3.5 <
1,2.3,4.7,8-HxCDF 1070 ' 206 1< S.2 1 < 3.6 <
1,2,3,6, 7,8-H:!CDF 231 : 213 < 4 ! < 2.7 <
2,3,4,6,7,8-HxCDF 188 228 < S.4 111 4.1 <
' 1.2,3,7,!l,9-HxCDF 81.8 ' 185 < 6.1 1< 4.4 <
1,2,3,4,6_,7,B-HpCDF I 1320 250 < 7.4 < 4.0 <
1,2,3,4,7,8,9-HpCDF 61~ 187 < 10.7 < 5,5 <
,.OCDF 3020 264 < 22.9 < 6.0 <
1 I-1EQ 341.8, 380.7 . 7.2 ~.9
Tota!TCDD < 12.9 38.8 , < 3.1 :< 3,61
· Tots.I PcCDD 60.4. 182 < 6.a l < 5.0 '
Total&CDD 2S0 sss < 8.6: < 5,0 I
TotalHpCDD 1650 : 166 < 12.81 < 5.9
Total TCDF 156 : 34.6 < 2.41 < 2.S
Tota!PcCDF 595 1 189 < 4.1 1< 3.S
Total H.'1:CDF 2230 ; 832 < 5,1 : 4.1
(Total H CDF 3080 ' 437 < 8.7 < 4.6
ll "' .Found in l.Jiboratary MetnoilBfa.rik · ·
PR = Peak is Poorly Resolved -amount is likely overeglimat.ed
P.4
Page3
lab~ust O:is
'dsm3
25.7 < 4.6
42.8 < 6.4
13.4 , < 4.3 1
13.41 < 4.1 [
13.4 < 4.3 ;!
19.6 1 < 3.7 ,
31.8 1 < 3.9 ,
36.7 i < 13.5 i
39.1 : < s:
34.2 < 6.9 :!
6 < 1.9 :I
6.6 < 2.d
7 < 2.3,
!1.9 < 3
18.3: < u l.
22 < 2: I,
22 < 5.5 i
39 7.6 '
·SEP 23 '97 05: 18PM ELI ECO LOGIC INC. P.5
Page4
Table 3 SUMMARY OF DIOXIN DATA -RUN 3
ff-1J '.l'Oft-C.:U a bJputSoiJ TtutcdSoil Scrubber W.:rte1 Scrubber W.ate
C.Oml!_ouad 'l!J. 2,3,7.8-TCDD 1< 1.4 < 0.3 < 1.3 : < 1.1 . < 36.4 < 1,2,3,7,8-PcCDD I< 1.6 < 0.3 < 2.5 i < 2.0 1< 35,0 <
1.2.3.4,7,8-HxCDD 1: 2.1 < 0.3 < 3.1 : < 2.4 '. < <40.6 < 11,2,3,6,7,8-HxCDD 1.9 < 0.2 1 < 2.4 i < 1,9 1 < 37.8 ,<
jl.2.3,7,8,9-HxCDD !< 1.8 < 0.2 < 2.8 i < 2.2 :< 40.6 ! <
1,2,3.4,6,7,8-HpCDD ~.1 1 0.5 < 3.5 · < 2.21 < 28.0! <
1OCDD 539.0 , 7.4 • 14.7 < 3.3 < 103.S j < ;2,3,7,8-TCDP 8,0; o.ss < 1.0 < 0.9 < 42.0 ' <
r 1.2,3.7,8-PeCDF r.. 3.2 < 0.3 < 1.6 , < 1.4 < 47.6 . <
;2,3,4,7,8-PeCDF 8.2 1< 0.3 < 1.6 : < 1.4 < 40.6 < 1,2,3,4,7,8-HxCDF Pit 98.1 < 0.2 < 2.0 \ < 1.6 < 10.2 < :1,2,3,6,7,B-HxC.DF 16,7 : < 0.2 < 1.S !< 1.2 < 6.9 <
!23,4.6,7.S-HxCDF II 7.3 : 0.61 6,2; 4.7 < 7.6 <
1.2,3,7,8.9-HxCDF < 1.8 : < 0.3 < 2.5 < 2.0 < s.o <
1,2,3,4,6,7,8-HpCDF BS.1 · < 0.2 , < 2.7 1< 1.9 < 9.7 <
1.2,3.4,7,8,9-HpCDF 49.2 < 0.3 i < 3.7 , < 2.7 < 11.2 <
OCDF 320: 2.3 '< 4.5 1 < 2.5 ' < 21.0 1 <
I-TEO 21.3 0.Sl 1 3.2 I 2.S: 48
Total TC'.DD < 1,41 < 0.3 ! < 1.3 • 2.1
,Total PeCDD < 1.6 < 0.3 < 2.5 < 2.0 1 1Total HxCDD 2.s :< 0.3 < 2.8 '< 2.1 I
·ToblHpCDD 71.6' 1.0 < 3.5 ;< 2.2
Tota!TCDF 16.3 o.ss < 1.01 < 0.9, TotalPeCDF 57.1 < 0.3 < 1.6 < 1.4
Tota.lHxCDF 195.0 0.61 6.2 4.7
1.Tota!H CDF 225.0 < 0.3 < 3.t '< 2.2
H -Found in Ls.bora.toiy Method Hlan
PR "" Peak is Poorly Resolved -amount is likely O\'ere5timatcd
A-06'/t v~ µ(/ .le ~ ~ ;i ~ fr
'-fol~~~{_
-~~~l
sm3 t
2.61
7.5 I 3.44
3.1
3.4.
5.2 i
7.3 ii
13.5 Ii 6.7 :
5 i j
1.8
u1 :
2.0
2.6
2.0
2.3 i
4., I 6.4 1·
'SEP 23 '97 05=18PM ELI ECO LOGIC rnc. P.6
Page 5
Features of the ECO LOGIC Process to Ensure Safe Use of Hydrogen
The chemical reactions which comprise the ECO LOGIC Process all take place in a sealed reactor
system which is kept within 0.4 pounds per square inch gauge (psig) of atmospheric pressure. The;:
system is monitored t.o ensure that the levels of oxygen remain well below the safe limit: and there
are no significant increases or decreases in pressure. There are several procedures which are carried
out during waste processing, to ensure the safe operation of the ECO LOGIC Process. These
procedures combine to avoid the scen&rios under which hydrogen becomes explosive: mixing with
oxygen, and exposure to a spark.
i) Prior to any hydrogen being introduced into the system, all vessels that • may contain
hydrogen-rich gas are pressure tested to well above normal operating pressure, to ensure
they are leak-proof. This testing includes a final test of the entire system with all vessels
connected.
ii) All vessels which might contain hydrogen-rich gas are electrically grounded through the
main power transformer on site. This ensures that even in the unlikely event that the
hydrogen combines with oxygen and becomes explosive, there is no potential for spark
ignition.
iii) All gasketed pipeline joints that may contain hydrogen-rich gas are connected by conducting
straps or structural conductors, and grounded.
iv) The ECO LOGIC Process operates as a sealed, closed loop system, at nominal atmospheric
pressure (less that 0.4 psig). Therefore, the possibility of the system rupturing due to over-
pressure is e>.1:remely unlikely. Also, the low system opers.ting pressure means that any
small leaks which may occur would release very small amounts of hydrogen -too small to
become explosive.
v) Rigorous procedures are followed for plant operations to ensure that hydrogen-rich gas
never mixes with oxygen or air. For example, all sealed vessels in the system are
completely purged and filled with nitrogen before any hydrogen enters the vessel. The
vessels are monitored and hydrogen is only introduced when the levels of oxygen are well
below the safe limit for a hydrogen-oxygen mixture.
vi) The procedure described in point (v) above is also followed at the end of each waste
processing cycle, when vessels full of hydrogen-rich gas need to be opened. Nitrogen gas
is used in the system as a "buffer" gas between hydrogen and oxygen.
vii) Once a sealed vessel is filled with hydrogen-rich gas, the system is continuously monitored
for oxygen content by process operators, to ensure that any increase of oxygen in the system
l.s immediately detected. Special actions are taken by the system operators which will
correct the condition well before an explosive mixture is created. The special actions are
detailed in a rigorous response procedure for operators that forms part of the Standard
Operating Procedures.
viii) As part of standard system operations, the air around the system is continuously monitored
at numerous strategic locations for explosive conditions due to hydrogen release. Warning
alarms will sound at levels well below an explosive mixture, which gives the system
operators ample time to take the appropriate corrective action.
ix) ~o open flames or smoking are permitted on-site.
---·SEP 23 '97 05: 19PM ELI ECO LOGIC INC. P.7
Page6
The ECO LOGIC Process has been opera.ting safely at various scales for many years. The use of
hydrogen by ECO LOGIC has never posed a safety risk to on-site workers or the surrounding
environment. Commercial-scale operations and further research and development continue to
confirm the ability of the ECO LOGIC Process to operate safely.
The above discussion has been added to the final report in the introductory section.
3) On page 26 of your report, you discuss the ability of the process to handle high moisture in the
soils through utilization of the steam generated during the process. Please discuss what effect
removal of water from the landfill will have on the treatment process. During start up of the
pilot study, it was mentioned that the soil would be better if it were either drier or wetter.
Does the generation of high levels of steam result in accumulated water that must be treated
either on site or shipped off site for disposal?
The ECO LOGIC Process requires water (as steam) in the reactor to aid in heat transfer. Hydrogen
alone is relatively invisible to the infrared radiant heat from the reactor electric heater tubes, but
steam absorbs this heat very well and transfers it convectively to the hydrogen and other gases.
Since it is a chemical reduction process, rather than an incineration process, the presence of water
does not interfere with contaminant destruction. Rather, it aids in supplying some of the hydrogen
necessary for reduction by reacting with methane that is present to form CO and hydrogen.
The water in the waste soil is all sent to the reactor from the TRM during continuous desorption.
The TRM has enough heating capacity to dry very wet soils, sludges and sediments prior to heating
the dried solids to 600°C. The ability to process wet material provides an economic benefit and also
avoids potential environmental emission problems, The high moisture content of the waste feed will
mainly be a concern in selecting the appropriate feed system design to ensure continuous design
throughput. The amount of moisture in the soil is one criteria that influences throughput, as
additional moisture requires more heat input to the system per unit of waste, As a result, the soil
from below the water table will be processed at a slower rate than that from above the water table .
Removal of water from the landfill (in terms of surface water drainage) will provide the process with
a suitable feed material. It is more cost-effective to process the relatively wet material at a lower
rate than to use an additional drying step. The generation of steam does not result in accumulated
water that must be treated either on site or shipped off site for disposal. The water contained in the
soil as moisture is carried as steam from the TRM into the process reactor and recovered in the
scrubber. This water is filtered and carbon treated prior to being stored in tanks for analysis. Once
it has been sho\\-11 to meet disposal criteria, it can be sewered or discharged to open water. This is
generally not a significant cost, since other than a slight salt content, the water is nominally free of
contaminants.
During start up it was mentioned that the soil would be better if it were either drier or wetter. This
comment was specifically in reference to the ease of operation of the treatability-scale soil feed
system, and does not have bearing on the ability of the system to desorb and treat PCBs.
The above discussion has been included in the final report in the discussion of full-scale
applications . ·
SEP 2'.=: '97 05: 19PM ELI ECO LOGIC INC.
Page 7
4) During the initial testing of soils for PCB content, it was demonstrated that the soils tested
were less than 300 ppm. Most soib tested in the past h2ve shown levels of 350-800 ppm.
Although you have given other examples of treatment in the report (PCB oils, dirt, etc.), no
mention is made of actual soil concentrations of PCBs. Can you show that the process works
\\';th higher levels of PCBs?
ECO LOGIC has conducted numerous treatability studies on PCB soil with levels similar to and in
excess of the levels recorded for the Warren County Landfill soil. These studies are summarized
in Table 4 below.
Table 4 TREATABil.JTY..SCALE THERMAL REDUCTION MILL RESULTS
WastePCB Processed Solids PCB
Waste Type Concentration (ppm) (;()ncentration (ppm)
Soil (moist, granulll!', PCB-spiked) 440 0.0039
Soil (moist granular, PCB-spiked) 520 0.0016
Sediment (mudd}', fine. PCB-,piked) 710 0.028
Sediment (muddy, fine, PCB-spiked) 790 0.0097
Sediment (muddy, fine, PCB-spiked) 750 0.065
c;;:,.riiml"'.nl (n,utUc· 1-:--) ,.,oo Nn m 01 n
S) During one of the runs, the temperature was allowed to drop below the 550 degree minimum
operating conditions. This malfunction or mistake would be unacceptable during full scale
operation. Please discuss. What is the overall reliability of the full scale operation?
The pilot-scale system does not have automated process control, and instead relies on the operator
or technician to take a series of measurements manually throughout the course of the run. During
Run 2, an operator error resulted in a drop in the TRM temperature to below the 550 degree
minimum operating condition. Our full-scale operations have a rigorous process control system
which eliminates the possibility of this sort of error occurring, This process control system is
operated by computer and alarms are activated in the event that any of the operating parameters
(including temperature) fall out of range.
Note that even if such a drop in temperature were possible at full-scale, it not have a negative impac.t
on the environment or the safety of site workers and the public. All outputs from the system,
including treated soil, are held and tested prior to their reuse in the system or disposal off-site. If
the soil, or any other system output, was found to contain levels of contaminants higher than the
regulatory criteria, the material would simply be reprocessed in the system until compliance with
criteria ill confirmed.
6) Data was reported for exhaust and product gases. Was there any analysis done on collected
particulate matter?
Gas sampling is performed using a sampling train comprised of a heated probe and particulate filter,
followed by a condenser and resin trap. Particulate is collected on the heated filter: with the vapor-
phase contaminants passing through to the resin. During analysis, both the particulate filter and
resin trap are analysed. Therefore, data reported for exhaust and product gases will include the
contaminants present in any particulate entrained in the gas.
::. .: .
~.'..Jo.....;.~
SEP 2~: '97 05: 20Pl1 ELI ECO LOGIC INC. P.9
Page8
Comments/questions from Joel Hirschhorn ·
1) Table 4: add dioxin TEQ levels in feed, similar to PCB data.
Table 4 has been revised for the final report to include dioxin TEQ levels. As discussed below
(question 4), all TEQ calculations have been made assuming non-detect values are present at half
the dctcc.tion limit .
2) Table 6: data on PCB levels -and perhaps other contaminants -appear to be rounded
numbeTS, compared to Table 4; only actual data should be used. The PCB data in Table 4 does
not correlate exactly with data in Appendix E; it appears that Total PCB + EMPC data were
not used and, ifso, then the report should defend excluding the EMPC data.
All calculations performed used actual numbers provided by the laboratory, with no rounding.
However, for presentation purposes, the numbers have been rounded to 2 significant figures, which
is a commonly used, scientifically defensible practice. In order to maintain consistency throughout
the report, the numbers presented in Table 4 of the report have been rounded to 2 significant figures
for the final report.
Where both a concentration and an EMPC were presented by the laboratory, ECO LOGIC used the
actual concentration, rather than the estimated value, to be consistent with other data presented by
the laboratory. Where only an EMPC was a.va.ila.ble, the EMPC was used as the actual value.
3) All Tables: whenever ND is given as a result place the actual detection limit in parentheses next
to ND. If some data are suspect because of findings in blanks, then that should be noted. For
examplt! it is suspicious that toluene was found at a higher level in Run 2 treated solid than
in original feed.
All tables in the final report have been revised to include the detection limit in parentheses
immediately follo\ving "ND".
4) In the calculation of dioxin TEQs it is not acceptable to assume NDs = 0 if the NDs are high;
a more proper and conservative approach is to use on~half of the NDs.
TEQ values for dioxin and furan data, including the re-analysed data received from the laboratory,
have been recalculated, as provided in Tables 1 through 3 of this response submission. These
recalculations have been performed using one-half the detection limit, as suggested, and are included
in the final report.
5) Section 5: Rhould include some information on the length of time for actual full-scale cleanup
of the site. Should check statements about estimated price of $350-400 per ton and what it
include!; or excludes. Is it really a total, turn-key remediation cost, which statements now
imply?
The cost of processing the Warren County Landfill soil will be accurately detemlined as part of the
Phase II activities. Based on information gained from the pilot-scale study, and current commercial-
scale operations, the per ton processing price is estimated to be $300 to $350. This is based on
processing 100-120 tons per day with a 70-80% availabiUty, with the processing portion of the
·sEP 23 '97 0s:20PM ELI ECO LOGIC INC. P.10
Page9
project thus lasting 90-123 weeks. To allow for ease of comparison with other technologies, this
price does not include costs for activities common to all technologies, such as permitting. public
consultation, excavation, performance testing or disposal of processed solids. These costs would be
negotiated between the prime contractor and the state, with limited cost control by ECO LOGIC.
The costs that ECO LOGIC does have direct control over, and to which this estimate appHes, include
system mobilization and commissioning, all waste p~paration and processing., lab costs for process
outputs analysis, and system demobilization.
The above discussion has been included in the final report in the discussion of economic estimates
for full-scale treatment.
6) Section 4.2.3: would like to see some more detailed explanation of how TRM temperature was
"inadvertently allowed to drop to SOO°C, which is below the minimum operating temperature
of550°C." How would this type of malfunction be prevented in actual full-scale equipment?
Please see response to question 5 submitted by Mike Kelly and Patrick Barnes, provided above.
7) Section 4.2.3: statement about infeasibility of reaching a 1 ppt TEQ should be reexamined,
because it is not correct.
As presented in the draft report, detection limits for dioxins and furans in Runs 1 and 3 were high,
due to the laboratory using a smaller sample volume than was necessary to achieve the required
detection limits. The statement about infeasibility of reaching a 1 ppt TEQ was based on
information provided h.,y the laboratozy (Triangle Labs). They stated that a level of l ppt would be
difficult to obtain, due to the fact that their~ detection limit for 2,3, 7,8-TCDD alone is l ppt.
Despite this concern, reanalysis of the samples yielded much lower detection limits than presented
originally. ECO LOGIC was therefore able to demonstrate that the Process can reduce the TEQ
levels in the soil to below the desirable treatment criteria of 1 ppt. Please see response to Question
l posed by Mike Kelly and Patrick Bames, above, and associated tables.
8) Table 9: would like some explanation for finding of dioxin in Run 3 treated scrubber water and
whether this wouJd be found in full-scale equipment. Tbe explanation on p.25 about
"interference in the analytical technique" is not satisfactory.
The recalculated dioxin and furan data has been provided in Tables l through 3 of this response.
As noted on these tables, levels of dioxin TEQs are lower in all post-carbon treatment scrubber water
samples. Only 2,3,4,6:7,8-hexachlorodibenzofuran was found in the post-carbon scrubber water.
As stated in the draft report, it is the opinion of ECO LOGIC that the presence of this furan congener
is a result of laboratory interference. This opinion is supported by the fact that this congener was
also found in the laboratory blank for Run 2.
9) Would like to see a special sub-section in 4.2 about levels of PCBs and dioxins in all process
waste streams (solid, air, water) and discussion of meeting regulatory requirements and/or
providing treatment of process residuals in company's equipment, or whether offsite
treatment/disposal would be required. Should have some discussion of PCB/dioxin air
emissions.
Outputs from the ECO LOGIC Process are treated solid material (i.e. treated soil), scrubber water
and exhaust gas. Table 5 prO\.-ides a summary of PCB 11.nd dioxin levels in these process outputs.
The Universal Treatment Standards for dioxins and futans, cited in 40 CFR 268.48, lists criteria of
0.001 mg/kg (1000 pg/g) for each of the tetra-chlorinated through hex.achlorinated PCDD/PCDFs.
For Run 1, no tetra-through hexa-chlorinated congeners were detected, with detection limits in the
li,,r dmr -ri-:e
·SEP 23 '97 ~5:21PM ELI ECO LOGIC INC. F'.11
Page JO
area of 0.3 to 0.6 pg/g. For Run 2, where the temperatures in the TRM were not sufficient to fully
desorb the dioxins and furans, the levels of tetra• through hexa-chlorinated congeners were
nevertheless below the Universal Treatment Standard. Levels of tetra-through hexa-chlorinated
congeners in Run 3 were mostly non-detect, with the highest level detected being 0.6 pg/g. Clearly,
all levels are below the Universal Treatment Standards, and Runs 1 and 3 achieved the RFP's TEQ
Target Performance Goal of 1 pg/g (Table S below).
For air outputs, acceptable levels are generally calculated according to risk, rather than an absolute
value that must be met. However, the EPA has recently proposed a Hazardous Waste Combustion
Rule, which states that incinerators must have less than 0.2 ng/dsm3 TEQs in their exhaust gas.
While the ECO LOGIC Process is not an incinerator, this· rule nevertheless can be used as a
yardstick for assessing outputs from the Process. As indicated by the table, TEQ levels in ECO
LOGIC exhaust gas are well below (i.e. approximately 25 times below) 0.2 ng/dsm3. In fact, no
dioxins or furans were detected in either the product gas or the exhaust gas, however, the detection
limits are such that compliance with the RFP's Target Perfonnance Goal of Sxl o-a µg/m3 (0.00005
ng/m3) can not be confinned.
The TSCA office of the EPA. which regulates PCB destruction facilities, does not have a stack level
for PCBs. Their criteria are simply that at least 6-nines destruction and removal efficiency occurs,
and that the levels of PCBs in liquid and solid outputs do not exceed 2 ppm. As discussed in the
draft report and shown in the table below, the ECO LOGIC Process achieved better than 6-nines
DRE, and levels in outputs were well below 2 ppm, in some cases by several orders of magnitude.
Levels of PCBs in system outputs were also well below the Target Performance Goals stated in the
treatability study's Request for Proposal of20 ppb for treated soil and 8x10"" µg/m3 (0.8 ng/m3) for
air emis1,ions.
The levels achieved for treated soil would allow the soil to be either replaced on-site, or landfilled
as clean fill. The quality of the treated water is such that it can be discharged to a municipal sewer
system .
Run 1
PCBs
TEQs
Run 2
PCBs
TEQs
Run3
PCBs
TEQs
Table S SUMMARY OF PCB AND DIOXIN TEQ
DATA FOR PROCESS OUTPUTS
Treated Soil Post-Carbon Scrubber Water
<0.6 ppb <0.008 ppb
< 0.5 ppt < 4.9 ppq
< 0.5 ppb < 0.007 ppb
380 ppt < S.9 ppq
< O.S ppb < 0.006 ppb
< 0.5 ppt < 2.9ppq
Exha"st GOJ
0.17 ng/dsm3
< 0.0083 nwdsm3
0.13 ng/dsm'
< 0.0076 ng/dsm3
0.077 ng/dsm3
< 0.0064 ng/dsm3
The above discussion has been added to the discussion of results in the final report.
· ·SEP 23 '97 05=22PM ELI -ECO LOGIC INC. P.12
Page 11
10) Section 5.5: statement that the process "requires minimal processing of untreated soil prior
to treatment" should be amplified, so that exact processing that may be necessary in full-scale
operation is fully detailed.
The ECO LOGIC Process and TRM soil processing unit do not require extensive pre-processing of
the soil to operate effectively. The soil must be screened to remove oversize material, but no
dewatering is required. Oversize material is defined in terms of the feed mechanism, which would
probably be either a ram feeder or a double intermeshed auger system. For a ram feeder, the
oversize criteria would be material greater than four inches diameter, and for the auger system,
material greater than one inch diameter. Oversize material would either be crushed and processed
in the TRM or treated in the SBV as bulk solids.
The ability to process wet material provides an economic benefit and also avoids potential
environmental emission problems. The amount of moisture in the soil is one criteria that influences
throughput, as additional moisture requires more heat input to the system per unit of waste. As a
result, the soil from below the water table will be processed at a slower rate than that from above
the water table, unless it has dried somewhat after being excavated. It is still more cost-effective
to process at a lower rate than to use a drying step, The water contained in the soil as moisture is
ultimately recovered in the scrubber: filtered, carbon treated, stored in tanks for analysis, and then
sewered or discharged to open water. This is generally not a significant cost relative to the overall
cost of processing.
The above discussion has been included in the final report in the discussion of full-scale
applications.
11) Was PM measured in exhaust gas?
Particulate matter was not measured separately in the exhaust gas. However, as discussed in the
reponse to Mike Kelly and Patrick Ba.mes1 Question 6 above, particulate is collected as part of stack
samplingr and analysed along with the resin to yield a total paniculate and vapor phase contaminant
concentration in the gas stream.
12) The dioxin data. for Run 2, S1 and SS need more discussion. In both cases the levels and
distribution of dioxin congeners is not typical for PCB impurities. But in the treated material,
there is even more dioxins than in the feed material, especially of the most toxic forms. Is it
possible that some reactions have occurred during desorption processing?
During TRM treatment, contaminants in the soil are not only being volatilized for introduction to
the reactor, they are also undergoing gas-phase reduction. A component of this reduction is
dechlorination of the chlorinated compounds, In the case of dioxins and furans, all congeners arc
being dechlorinated. However, there are considerably more higher chlorinated congeners (such as
OCDD and HpCDD) in the input soil. Therefore, during reduction these higher chlorinated
congeners are losing chlorines to result in lower chlorinated congeners.
The conditions experienced during Run 2 are not representative of typical treatment operations, and
therefore the results are anomalous. Because of the low temperature in the TRM during Run 2, full
dechlorination and reduction of the dioxins and furans did not proceed to completion. Therefore,
the pam~rn of dioxins and furans shifted from predominantly higher chlorinated congeners to
increased lower chlorinated congeners during TRM treatment, as illustrated in Figure 1.
Note that the toxicity (TEQ vaJue) of the soil is roughly the same for both untreated and treated
material (i.e. within the same order of magnitude), and so the toxicity is not significantly increasing
overall. Furthermore, as illustrated in Figure 1, the overaH amount of dioxins and furans, with the
-·-SEP 23 '97 0s:22PM ELI ECO LOGIC INC. -p. i:::
Pagel2
exception of the lower chlorinated congeners, decreased considerably over the run. As evidenced
by the excellent post-treatment values for Runs 1 and 3, operation of the TRM at the appropriate
temperature allows for full desorption of the dioxin and furan congeners.
The above discussion has been included in the final report in the discussion of test results.
13) Then: should be more detailed information on exactly how the materials received by the
company were treated and handled prior to becoming designated as the feeds for the three
runs. ·was a composite made from all incoming containers? Did each run correspond to one
particular bucket?
The soil was tr.msported and received in a sealed overpac drum. The drum contained three 5-gallon
buckets of contaminated soil and a vermiculite-type packing material. The individual seals of the
buckets remained intact throughout transport. Each bucket was individually screened, and the
screened material placed jn clean, new buckets. Only two buckets of soil were required for the
treatability study. The soil within each of the new buckets was well mixed prior to grab sampling
for the input soil. Pre-and post soil weights were recorded for each test runs. There was no mixing
of soil between buckets.
The above discussion has been added to the final report in the discussion of process operations.
14) How much oversi:ze material was screened out as a percent of the original soil sample
provided? Wh:it did this material look like?
Due to the physical dimensions of the pilot-scale TRM, material greater than 0.5 inches was
screened out. The weights of input soil and screen rejected soil are provided in the t.able below. The
rejected material consisted of large chunks of material, assumed to be a soil and asphalt mixture, as
well as large stones and unbreakable soHds, At full-scale this material would either be crushed, or
treated in a Sequencing Batch Vaporizer (SBV), which is designed to treat bulk solids. SBV
treatment of rubble, asphalt and concrete mixtures has been performed extensively at ECO LOGIC's
full-scale GMCL facility.
Waite Mtlterial Mi1$J (kg)
Bucket I
Screened Soil 23.72
Rcj cctcd Material 0.99
Bucket 2
Screened Soil 25.31
Rejected Material 1.20
The above discussion has been added to the fmal report in the discussion of process operations.
it;. •. ·--·♦ ••·--:::ff:2:r:-_____ _
·sEP 23 '97 05:23PM ELI ECO LOGIC INC.
Figure 1 DIOXIN AND FORAN PATTERNS IN
UNTREATED AND TREATED SOIL
Dicxin/Furan Pattern in Untreated Soil
Total
Dioxin/Furan Pattern in Treated Soil
6000
5000
4000
g; 3000
2000
1000
0 .;;,. --!11 TCDD ➔xCDI OCDD p
Total
P.14
Page 13
-
SEP 23 '97 05!23PM ELI ECO LOGIC INC. P.15
Page 14
15) The report should present some discussion or its approach to give PCB data in term·s of total
PCBs rather than more typically used method of giving data in terms of Aroclors. At least for
the St data, some discussion of lab data should be given to relate to more typical data on
Aroclors. Should present some information on relative toxicities of different PCBs.
The Request for Proposal states that PCB data for pre-and post-treatment solids must be presented
on a congener-specific and total PCB basis. In order to be consistent, ECO LOGIC had all matrices.
including liquid and gaseous samples, analysed on a congener-specific basis. This approach was
proposed by ECO LOGIC in our March 19 submission, and accepted.
A discussion of relative toxicities of PCBs was not requested in the RFP, and would appear to be
beyond the scope of this treatability study.
TRIANGLE
ENVIRONMENTAL
INC.
Ms. Dollie B. Burwell, Co-Chair
Mr. Henry M. Lancaster, Co-Chair
September 15, 1997
Joint Warren County/State PCB Landfill Working Group
720 Ridgeway Street
Warrenton, NC 27589
Dear Ms. Burwell and Mr. Lancaster:
P.O. Box 41087
Raleigh, NC 27629
919-828-3150
800-849-5115
Fax [phone]
Triangle Environmental, Inc. (Triangle) of Raleigh is leading a team that proposes to
address the Warren County PCB Landfill with appropriate technology to clean the site
and destroy the PCBs located therein using proprietary processes that already have EPA
nationwide permits in place. We believe that we have the most advanced and most cost-
effective technologies for remediating Warren County's PCB Landfill, and we ask that
we have the opportunity to, at our cost, demonstrate this to you and the committee.
We understand that the committee has conducted pilot tests on other technologies and is
anxious to resolve this issue as quickly as possible. Our goal is to offer the committee an
alternative technology that we believe will be in the best interest of the people of Warren
County, without slowing the progress the committee has made. We are requesting a
sample of representative material under the Freedom of Information Act and ask that the
committee review and consider the results from our tests so long as we present them to
you and the committee within 30 days of our receipt of material.
In order to conduct a bench-scale study to demonstrate the effectiveness of the technology
to the committee, we would need a sample of approximately 50 gallons of representative
material. Although the process test run would be completed in 1 to 2 days, confirmation
testing by an independent laboratory would require approximately two weeks. We
currently project that the bench-scale test, including the issuance of a final report, would
be complete within one month of receiving a sample. We will make the necessary
arrangements for pick up and shipment of the sample upon your approval. Our team will
bear all costs associated with the testing and will destroy all PCBs and dispose of the
sample without returning it to the storage facility in Greensboro or the landfill in Warren
County.
RST\PCBL TR2.DOC
Our team will bring the best elements of two significant technologies to the Warren
County site. One is the patented METHEX solvent extraction/soil washing technology
developed here in North Carolina by Environmental Technologies, Unlimited. Second is
the patented PCB destruction technology, the Solvated Electron Technology (SET)
program developed and patented by Commodore Advanced Sciences, Inc. (CASI) of
Albuquerque, New Mexico. Both of these technologies have been thoroughly tested and
verified, and they both hold EPA nationwide permits.
As a brief overview, the solvent extraction process will allow the treatment of the
contents of the landfill to remove the PCBs from the soil and water in the landfill and
concentrate it in the solvent to facilitate the destruction of the PCBs by the SET process.
Triangle believes that this step will shorten the time required to treat all the landfilled
material and reduce the total cost of the operation.
The SET process involves the ambient temperature destruction of essentially all PCBs in
the sample by chemical conversion to remove the chlorine atoms from the PCB molecule
and converting it into common salt. There is no incineration and no potential for
formation of dioxin or similar materials during the reaction. After the destruction process
is complete, the treated material is suitable for return to the landfill or for reuse under
proper conditions.
In summary, we believe that we have assembled a team with the appropriate resources
and technology to meet the goals of the Warren County PCB Committee in a timely and
cost effective manner. We look forward to working with you in this process.
We recognize the importance of this project to the people of Warren County and to the
people of the State of North Carolina.
Sincerely, ·
TRIANGLE ENVIRONMENTAL, INC.
==4;;/1£~
C. Thomas Hendrickson
Chairman
/ra
RST\PCBL TR2.DOC