NCD980602163_20030317_Warren County PCB Landfill_SERB C_Test Run on Contaminated Clay-OCR

Report on 50 ton Test Detoxification of Clay Cap Material March 17, 2003
Report on 50 ton Test Detoxification of Clay Cap Material
Warren County PCB Landfill Detoxification Project
On March 6, 2003 an 50 ton test was conducted to make an initial determination of the
feasibility of treating clay cap material with the existing thermal treatment system.
Obiectives
The "Plan for Testing Detoxification of Clay Cap Material", which appears as
Attachment 1 to this report, described the intent of the test as follows:
The intent of the initial 50-ton test is to determine:
1. Gross handling characteristics, i.e. Does it clog screener, feed system, cyclone,
discharge chutes or pug mill?
2. Carryover of dust into liquid/gas treatment system, i.e. Is the carryover greater
and does it overload that system?
3. Is desorber gas exit temperature of 550°F (50° F greater than normal operation at
this site) adequate to treat the clay cap material to less than the TSCA High
Occupancy Area Treatment level of <1.0 ppm PCBs?
Chronology
A chronology of events during the test appears as Attachment 2 to this report.
Observations and Results
Significant observations and analytical results included the following:
1. No significant problems were encountered when screening the clay prior to the
test.
2. Feed soil moisture values of 25.26 to 28.30% were at the high end of the range of
what had been anticipated, but are viewed by Shaw as likely to be representative
of what would be encountered if large scale treatment of clay were to be
performed.
3. There was a very brief period at the start of the test run when it appeared that the
feed augur might have difficulty handling the clay, but this was quickly resolved
and did not recur.
4. As a result of the higher moistures, it was not possible to maintain both the target
feed rate of 13 tons/hour and the scrubber outlet gas temperature of 550°F.
Despite attempts to maintain a constant gas temperature it ranged widely during
the first hour of the test before dropping below the target temperature. After that
point target temperature was only recovered by reducing the feed rate to 10-10. 5
tons/hour, and even so there were further brief periods when the gas temperature
dropped below target.
Report on 50 ton Test Detoxification of Clay Cap Material March 17, 2003
5. Clay feed was stopped when the screened stockpile was exhausted. Although the
33.44 ton total amount of clay fed was less than the 50 ton target Shaw views the
test duration and quantity as sufficient to achieve the objectives of the test.
6. The analysis of the composite sample of soil treated during the test was 330 ug/kg
PCBs, reported as Arochlor 1260. This was below the target of 1.0 ppm (1 ,000
ug/kg) which had been established for the test. The moisture content of this
sample was 18.3%.
7. In addition to the composite sample of treated soil, four grab samples were taken
at one hour intervals for separate evaluation by Shaw and MSR. The results were
highly variable, with the highest being 350% of the low and 146% of the average
of the 4 results.
8. Attachment 3 to this report shows the amount ofresiduals generated during the
test and compared it to what has been experienced on this site when treating
"normal" feed material from the landfill excavation. The test value of 6. 7 4 % is
approximately 3. 6 times the usual value of 1. 87%.
9. During an inspection conducted after the conclusion of the test it was noted that
the condensate water was turning brown, where it is usually almost clear.
Discussion
This discussion relates each of the observations and analytical results to the
corresponding objective of the test.
Gross handling characteristics, i.e. Does it clog screener, feed system, cyclone,
discharge chutes or pug mill?
Observations land 3 indicate that the clay can be handled without the need for additional
solids handling equipment or extensive modification to the solids handling portions of
thermal treatment plant.
Observation 2, of an average feed moisture of over 25%, is of considerable concern,
especially when taken in conjunction with observation 6, that the treated soil moisture
was only 18 .3%. This indicates that over 7% of the feed would accumulate as water in the
thermal treatment system. While the system had adequate capacity to do so during this
brief test it could not continue to do so in larger scale operations, and the excess water
would eventually have to be discharged.
Normally all of the feed moisture can be used for product cooling and discharged either
as part of the treated soil or as the steam exhausted from the product cooler scrubber
stack. However, it does not appear that these outlets would suffice for feed with the high
moisture levels observed during this test without making the treated soil so wet that it
would be difficult or impossible to handle and backfill.
At present there is no other means for onsite treatment of the excess water, the
stormwater treatment system operated by Shaw is neither designed nor permitted to
handle this waste stream. Given the difficulties that we have experienced in meeting the
Report on 50 ton Test Detoxification of Clay Cap Material March 17, 2003
treatment goals for stormwater, almost entirely due to the passage of clay fines through
the system, it is unlikely that the existing system could be readily modified to handle this
new waste stream, even if the associated permitting considerations would allow this to be
done in a reasonable time period.
Off site treatment of the excess water is an option if the concerns of local citizens can be
addressed. Assuming that 7% of the weight of the clay fed will be discharged from the
thermal system as excess moisture, this results in 140 lbs of water per ton of clay fed, or
approximately 16.5 gallons per ton. If 5,000 gallon tank trucks were used for
transportation to offsite disposal then one truckload would be required for approximately
each 300 tons of clay treated. As of this report, Shaw has made no attempt to obtain
disposal facility approval or pricing for off site transportation or disposal of water.
Carryover of dust into liquid/gas treatment system, i.e. Is the carryover greater and
does it overload that system?
Observation 8 is that there is clearly greater carryover of solids, which was expected due
to the higher percentage of fine particulate in the clay feed. The increase was not enough
to overload the system during this brief test, during which no centrifuge or filter cake was
recycled through the system as is usually done. The increase in carryover from 1.87% to
6.74% would reduce the quantity of treated soil per pass through the system from 98.13%
to 93 .26%. The ratio of these figures is 1.0522, and indicates, very roughly, that the
contract price would have to be increased by over 5% to compensate under the present
system for measurement and payment of treated soils. This calculation does not take into
account the high moisture content of the residuals, which would tend to further increase
the resulting cost.
Further testing would be required to determine the impact of recycling the residuals
through the thermal system. At a minimum, this would be expected cause some further
increase in the percentage of solids carried over and attendant costs. An alternative would
be to ship the residuals to an offsite facility for further treatment, but their high PCB and
moisture contents would also tend to make this expensive. As of this report, Shaw has
made no attempt to obtain disposal facility approval or pricing for offsite transportation
or disposal of these residual solids.
Observation 9 is of concern in that it indicates that fine clay particles in the carryover are
not being completely removed and so are remaining in the condensate water that is used
for product soil cooling. Given Shaw's experience with clay fines in the much larger
stormwater treatment system this result is not surprising. However, previous analysis of
residual solids shows relatively high levels of PCB contamination, as would be expected.
While the condensate water was not tested after this test, it is probably reasonable to
conclude that the portion of the residual particulates that are carrying over into the
condensate water also contain relatively high levels of PCBs. This test probably did not
continue long enough for the condensate water to become sufficiently fouled that it
would have had a significant impact on the level of PB Cs in the treated soil, but this
Report on 50 ton Test Detoxification of Clay Cap Material March 17, 2003
could certainly occur over a longer period of time, thus increasing the risk of failed
batches.
Is desorber gas exit temperature of 550°F (50° F greater than normal operation at
this site) adequate to treat the clay cap material to less than the TSCA High
Occupancy Area Treatment level of <1.0 ppm PCBs?
Observation 6 indicates that treatment to the target level is possible. The instability of the
gas temperature during this brief test makes it difficult to determine if 550°F is adequate,
but it does appear likely that something in the area of 550°F to 600°F is probably
sufficient if it can be maintained consistently.
Observation 7 is of great concern to Shaw, as the wide range of results over such a short
period indicates a high potential for having batches fail during large scale operations. The
fact that the higher results could not be consistently correlated with periods of lower gas
temperature raises the possibility that even achieving more system stable operation may
not completely eliminate the variations in output. This may have been experienced to
some extent even in operation on normal soil, as there have been batch failures that do
not appear to be associated with any abnormal operating conditions, but the limited data
from this test appears to indicate that this phenomenon may occur much more frequently
during treatment of clay.
Determination of what the risk of batch failure is will be a major factor in determining a
final unit price for treatment of clay. As an example, even assuming only a 20% failure
rate would mean that Shaw/MSR would only be paid for 80% of the material processed,
even though the cost of treating a failed batch is essentially equal to the cost of treating a
passing one. To compensate for this, the unit price, which will already have to be
increased due to other factors, would then have to be multiplied by 1/0.8 = 1.25, for an
increase of 25 %. Assuming a potential failure rate as high as 50% would result in
doubling the already increased price. An alternative would be for the state to assume the
risk by agreeing to pay for all clay processed, whether or not it met the treatment
standard.
Observation 4 is the one with the largest certain financial impact. The test target
production rate of 13 tons/hours was already well below the bid rate for normal soil, due
to the expectation that the high feed moisture would prevent being able to achieve normal
production rates. This was the largest factor that resulted in the higher than normal price
of $154.25 quoted by Shaw for conducting the test. As noted in the chronology and
observations "it was not possible to maintain both the target feed rate of 13 tons/hour and
the scrubber outlet gas temperature of 550°F ..... target temperature was only recovered
by reducing the feed rate to 10-10. 5 tons/hour, and even so there were further brief
periods when the gas temperature dropped below target". Coupling this with observation
8 regarding the wide fluctuations in treated soil quality, Shaw believes it would not be
prudent at this time to rely on being able to maintain a consistent feed rate in excess of 9-
10 tons/hour. The average of9.5 tons/hour is only half of the bid feed rate of 19
Report on 50 ton Test Detoxification of Clay Cap Material March 17, 2003
tons/hour, and by itself would justify an approximate doubling of the existing contract
unit rate.
Conclusions
As noted directly above, the anticipated 50% reduction in feed rate, due primarily to the
greatly increased moisture content, would by itself justify an approximate doubling of the
existing contract unit rate of $88.40/ton. While various details of the actual cost structure
could tend to lower this figure somewhat, these would very likely be much more than
offset by the cost increases associated with the increase in percent solids carryover
(minimum 5%), residuals recycling (not yet quantified), the need to further treat or
dispose of excess condensate water (not yet quantified), and risk of batch failures (not yet
quantified, but likely in the range of 20-50% ).
Given the factors cited above, the potential range of prices can be calculated by the
formula:
Price= Base x M x ((SC x RR)+ CW) x BF, where
Base= contract base price of $88.40/ton
M = increase due to moisture impact on production rate
SC = increase due to increase in solids carryover
RR = increase due to recycling residuals, this compounds factor SC
CW = increase due to treatment or disposal of excess condensate water
BF = increase due to risk of batch failures, this compounds all other factors
Insert likely minimum and maximum values for each factor to obtain price range
Minimum
Price = $88.40/ton x 2 x (1.05 x 1.0) + 0) x 1.25 (assume 20% batch failures)
= $88.40/ton x 2.625
= $232.05/ton
Maximum
Price = $88.40/ton x 2 x (1. 1 x 1. 1) + .05) x 2.0 (assume 50% batch failures)
= $88.40/ton x 5.04
= $445.36/ton
The unit prices actually charged by Shaw could in some cased be lowered by the state
assuming the risk on items such as % solids recycled and % batch failures. However, the
overall impact to the state would remain substantially the same for a given level of risk.
While the state could benefit if they assumed the risk and the assumptions and figures
used above then proved to be overly conservative, the cost impact to the state could also
increase if those figures were found to be overly optimistic.
Report on 50 ton Test Detoxification of Clay Cap Material March 17, 2003
Based on the price range calculated above, then for the estimated minimum quantity of
2,000 tons of clay to be treated, the resulting cost to the state would be $464,100 to
$890,720. For a potential quantity ofup to 6,000 tons the costs would be $1 ,392,300 to
$2,672,160.
Please note that the costs above are only those associated with Shaw's contract, they do
not include any costs for the state or its oversight contractor associated with the extended
duration of the project. At 9.5 tons/hour, 85% unit availability, and scheduling the unit as
operating 6 days/week, the conservative calculation of extended duration for the
minimum of 2,000 tons and making no allowance for increase in solids carryover, solids
recycling or batch failures, is as shown below:
Calendar days = 2,000 tons/ (24 hrs/day x 9.5 tons/hr x 0.85 x 6/7) = 2,000 /166.11
= 12.04 days
The same calculation is performed below for four scenarios corresponding to the four
used for Shaw pricing impacts.
2,000 tons, 20% batch failures= 15.05 calendar days
2,000 tons, 50% batch failures= 24.08 calendar days
6,000 tons, 20% batch failures = 45 .15 calendar days
6,000 tons, 50% batch failures= 72.24 calendar days
Questions for Shaw regarding this report should be adressed to Gary Duke at (609) 588-
6373 , e-mail [e-mail].
Dukefile c:\mydoc\nclf\Clay Test #1 -Report .. doc
Plan for Testing Detoxification of Clay Cap Material
Warren County PCB Landfill Detoxification Project
Background
Change order #7 authorized treatment ofup to 200 tons of clay cap material from the
"Mt. Tilley" stockpile at a significant increased unit price cost. The purpose of the
change order is to determine if the clay cap material can be successfully treated in the
existing detoxification equipment and to determine a unit price for the treatment of larger
quantities of the clay cap material. The work will also serve to identify any handling or
processing difficulties that could preclude the large-scale treatment of the clay cap
material in the existing detoxification equipment onsite. The plan developed uses a two-
step approach in an effort to make the needed determinations in a cost efficient manner.
The evaluation will consist of two tests. In the first test, approximately 50 tons of clay
cap material will be treated and last approximately four hours. This is the amount of
operating time that is required for each run during a TSCA performance demonstration
test. It should also be sufficient to identify obvious problems that would preclude large-
scale treatment in the detoxification system. If the results of the first test support further
investigation, the second test will be conducted in the same manner treating up to the
remainder of the 200 tons authorized. The second test will allow MSR to obtain
additional process information to refine the unit price that would be used if a change
order were written to treat larger quantities of the "Mt. Tilley" stockpile.
Objectives
The intent of the initial 50-ton test is to determine:
1. Gross handling characteristics, i.e. Does it clog screener, feed system, cyclone,
discharge chutes or pug mill?
2. Carryover of dust into liquid/gas treatment system, i.e. Is the carryover greater
and does it overload that system?
3. Is desorber gas exit temperature of gas system at 550°F (50° F greater than
normal operation at this site) adequate to treat the clay cap material to less than
the TSCA High Occupancy Area Treatment level of <1.0 ppm PCBs?
During test 1, items 1 and 2 will be qualitative determinations while item 3 will be based
on the analysis of a single composite sample of the treated material processed during the
test. Based on outcome of test, Shaw/MSR will determine if it is practical to attempt
large-scale treatment in the existing detoxification equipment and will establish a likely
range of budgetary pricing for doing so.
Based on the outcome of the test and the pricing proposed by Shaw/MSR, DENR will
determine whether to conduct the second test with the remainder of amount authorized.
Page 1
,. Plan for Testing Detoxification of Clay Cap Material March 1, 2003
The longer test would be required for Shaw/MSR to obtain sufficient quantitative data on
recycle rate, risk of batch failure, etc. to develop a firm price to treat larger volumes of
clay cap material from the "Mt. Tilley" stockpile.
Procedure for Test 1
1. Shaw will excavate 50 tons of material from the "Mt. Tilley" stockpile at the 30'
x 30' grid location designated MTS . This area contains the highest PCB
concentration ( 43 ppm) found to date in the stockpile. Excavate from the ground
surface to a depth of one foot in the 30' x 30' grid.
2. Shaw will transport the material to pad 1, screen it and stockpile the screened
material under the pole barn. Test material will be isolated from other feed
materials already under the pole barn.
3. Immediately prior to the test, the treatment plant should be in normal operation
with the normal feed material. This is the preferred condition, however, the test
can be conducted as the first activity on restart after a shutdown as long as the
feed hopper was empty when shutdown. This may be necessary if the system had
to be shutdown shortly before the scheduled test start.
4. Run the feed hopper (the one the loader dumps into) empty of normal feed
material, then shut off the hopper conveyor. Load the feed hopper part way with
material from the screened feed test pile. Only clay from the test pile is to be fed
during the test. No other soil, centrifuge cake or filter cake should be added or
mixed.
5. Continue to run the desorber feed augers until the amount of normal feed material
remaining in auger hopper is close to the minimum required to maintain the
desorber air seal. Record the weigh belt reading and then start running clay from
feed hopper at a nominal feed rate of 13 tons/hour. Switch the discharge conveyor
to an empty bin at this time and leave it there at least until all the clay material has
been processed and discharged. At the option ofMSR, the conveyor may be left
there for up to 2 hours longer to ensure that all clay has "flushed" through the
system and cannot contaminate a batch of normal material.
6. When the clay feed starts, take the first feed sample from where feed hopper
conveyor discharges onto scale conveyor. Continue taking feed samples in this
manner once per operating hour for the duration of test. All feed samples will be
grab samples and will be analyzed for moisture only.
7. Adjust burners and other operating parameters to achieve a minimum desorber
gas exit temperature of 550°F. The feed rate may have to be reduced in order to
reach this operating temperature. Higher moisture content requires additional
Page 2
Plan for Testing Detoxification of Clay Cap Material March 1, 2003
energy to evaporate water from the material. Reducing the feed rate increases the
time of the material in the desorber and thus the amount of heat transferred to the
material.
8. Note the time when the desorher gas exit temperature and other conditions
stabilize and operations appear to have lined out. Attempt to maintain these
conditions for the duration of the test. Carefully note start and stop times of any
periods when the desorber gas exit temperature drops below 550°F. Continue the
test until the entire stockpile of screened clay is used. This will require
approximately four hours. Record the weigh belt reading after clay feed is
completed to determine the total tonnage of clay used to conduct the test.
9. Treated soil samples will be taken in the same manner as for normal treatment
operations. The first treated soil samples from the test run will be taken one hour
after the start of feeding clay or 20 minutes after the operating conditions are
established and stabilized, whichever occurs first. Two-500 mL grab samples will
be collected. Two additional treated soil samples will be taken at I-hour intervals
for the duration of the test. One of the samples will be used by MSR for process
evaluation. The other sample will be used to prepare a single composite sample
for the test run. At the end of the test, the composite sample will be analyzed for
PCBs to determine whether the test treatment goal of <1.0 ppm has been met.
10. The disposition of the entire pile of treated material from the test run will be
based on the PCB analysis of the single composite sample.
Page 3
Residuals Production during the 50 Ton Clay Test
The residuals record during the clay test conducted on 3/6/03 shows that 2.49 tons of
centrifuge sludge and filter cake were generated. A total of 46.8 tons of feed material was
processed, of which 33.44 tons was clay and 12.64 tons was normal feed material run
after the clay to flush out the kiln.
The 2.49 tons of test residuals equal 5.41% of the total material processed.
The residuals as a percent of kiln feed material for the 5 batches preceding the test run are
shown in the table below
Batch Number % Residuals
93 1.98
92 2.23
91 1.02
90 2.30
89 1.83
AVERAGE% 1.87
The 2.49 tons of test residuals equal 5 .41 % of the total material processed, an increase
over the 5 batch average by a factor of 2. 9.
Assuming that the 12.64 tons of normal feed run during the test contributed the average
of 1.87% residuals, then it contributed .236 of the 2.49 tons. If so, this means that the
33 .44 tons of clay fed during the test contributed the remaining 2.254 tons.
This would result in a residuals generation rate for the clay alone of 6.74%, an
increase over the 5 batch average by a factor of 3.6.
None of the residuals were added back to the clay feed during the test run, as would
normally be done during full scale processing. Since these residuals have a very high
moisture content (>50%) this would cause the actual processing rate to be lower than that
demonstrated during the test, and therefore will further increase the cost of clay
processmg.
Dukefile C:\nclt\Clay test #1 -Residuals Production-Duke
Warren County PCB Landfill
Shaw E&I project #827062
Chronology of 50 ton test detoxification of clay cap material conducted on March 6, 2003
notes by Gary Duke, Shaw E&I
Belt scale Feed Scrubber Feed
totalizer rate outlet gas soil
Time Descriotion of Event tons tons/hr temp-deg F. % H2O
Note: target conditions for the test were 13 tph with minimum scrubber temp of 550 F,
with a target treatment limit of a maximum of 1,000 ug/kg (1.0 ppm) PCBs
Note: immediately prior to clay test the unit was operating normally on its usual feedstock
of contaminated soil from the landfill at a feed rate of approximately 16 tons/hour
10:20 Feed hopper run empty of normal feed , feed belt shut off
10:24 Start loading clay into feed hopper with feed belt shut off
10:25 Take "start of test" belt scale tonnage reading 35,488.42
10:28 Start feed belt to feed clay to system 13-14 601
10:30 Moved discharge conveyor to bin 1 for test
10:45 Took first feed moisture sample, ID# WCLF-SU-Clay1 25.98%
10:48 Check that gas temp and other conditions appear to have stabilized 13-14 627
11:00 Took first treated soil sample, appearance had started to change form normal to a more 13-14 683
reddish color and appeared to have more sand and/or very fine gravel in it
11:30 Gas temp had dropped after 11 :00, got as low as 520 F, dropped feed rate to 10-10.5 tph 10-10.5 530
11:45 Took second feed moisture sample, ID# WCLF-SU-Clay2 28.30%
12:00 Took second treated soil sample 10-10.5 585
12:43 Loaded last of test clay feedstock into the feed hopper
12:45 Took third (final) feed moisture sample, ID# WCLF-SU-Clay3 25.26%
13:00 Gas temp noted as having dropped to 550 F minimum target value, still dropping 10 550
13:07 Took third treated soil sample 10-10.5 530
13:30 Feed hopper empty of clay, belts run clear of feed, take "end of test" tonnage reading 35,521 .86 0
Total amount of clay run during test 33.44
Note: total run was less than 50 ton target but used all of screened clay stockpile
13:35 Loaded feed hopper with normal feedstock and resumed normal feed
14:00 Took fourth (final) treated soil sample, at this time it was noted that the appearance of the 555
treated soil began to change back to its normal appearance instead of the redder color
observed during the test, so this sample may not have been as indicative of clay feedstock
14:00-30 Left discharge conveyor over bin 1 to attempt to allow clay feed to clear out of the kiln
14:30 Moved discharge conveyor from bin 1 35,534.50
Total amount of material run into bin 1 46.08
Amount of non-clay feedstock run into bin 1 (total minus clay fed) 12.64
The 4 treated soil samples were composited and sent for analysis as sample
ID #WCLF-ST-01-094. Results were 330ug/kg PCB as 1260 and 18.3% moisture.
NORTH CAROLINA DEPARTMENT OF ENVIRONMENT AND NATURAL RESOURCES Division of Waste Management ~~~ ~ Ljoo /306 / 'fOO 730 X --qsg fth ~ ~33offh