United States
Environmental Protection
Agency
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
Research and Development
EPA-600/2-84-193 Feb. 1985
Project Summary
Land Treatment of Petroleum
Refinery Sludges
Leale E. Streebin, James M. Robertson, Herbert M. Schornick, Paul T.
Bowen, Kesavalu M. Bagawandoss, Azar Habibafshar, Thomas G. Sprehe,
Alistaire B. Callender, Charles J. Carpenter, and Vickie G. McFarland
The purpose of this study was to
identify, evaluate and optimize the
factors which influence land treatment
of oily residues. A research site owned
by the University of Oklahoma was
used. A total of 45, 6.1 mx2.7m(6ftx
9 ft), plots were prepared and API
Separator sludge was applied to the
plots at loading rates between 3 and 13
weight percent per year, and loading
frequencies from 1 to 12 times per year.
The soil was analyzed for oil content,
selected heavy metals, selected organic
priority pollutants, pH, nitrate, and
chloride over an 18-month period.
Oxygen levels in the soil atmosphere,
and the emission rate of volatile
hydrocarbons were monitored. A labora-
tory study to identify and quantify
volatile hydrocarbons emited also was
performed. Fractionation analysis of
sludges and recovered oils were done
for saturates, aromatics and polar
compounds and asphaltenes.
Total oil losses were proportional to
the amount of oil applied with mean
losses over the study period equal to 54
percent of the oil applied. Losses of the
saturates fraction were highest followed
by aromatics, polar compounds, and
asphaltenes. Volatile losses as a percen-
tage of the oil applied were relatively
small over the long term, but were
substantial in terms of short-term losses
immediately after application. Biode-
gradation of both total oil and individual
oil fractions followed first-order reaction
kinetics. A composite first-order biode-
gradation rate coefficient of 0.003 day'1
was computed after compensation for
volatilization.
Site monitoring determined that
heavy metals were immobilized and the
organic priority pollutants were degraded
in the zone of incorporation (top 30
cm). Some buildup of metals occurred
over the study period.
Operational considerations such as
sludge loading rates and frequencies,
proper tillage of the zone of incorpora-
tion, prevention of oil percolation and
runoff, and operation of field equip-
ment after sludge application are impor-
tant factors in the design of land treat-
ment facilities.
The full report was submitted in
fulfillment of Cooperative Agreement
No. CR80757810 by the School of
Civil Engineering and Environmental
Science, University of Oklahoma un-
der the sponsorship of the U.S. Environ-
mental Protection Agency. The report
covers a project period from April,
1980 to April, 1983; field and lab work
was completed in June 1983.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental
Research Laboratory, Ada, OK, to
announce key findings of the research
project that are fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
The purpose of this study was to
identify, evaluate and optimize the factors
which influence land treatment of oily
residues. A research site owned by the
University of Oklahoma was used. A total
of 45, 6.1 m x 2.7 m (6 ft x 9 ft), plots were
prepared and API Separator sludge was
applied to the plots at loading rates
between 3 and 13 weight percent per
year, and loading frequencies from 1 to
12 time per year. The soil was analyzed
-------
for oil content, selected heavy metals,
selected organic priority pollutants, pH,
nitrate, and chloride a 18-month period.
Oxygen levels in the soil atmosphere, and
the emission rate of volatile hydrocarbons
were monitored. A laboratory study to
identify and quantify volatile hydrocarbons
emitted also was performed. Fractionation
analysis of sludges and recovered oils
were done for saturates, aromatics and
polar compounds and asphaltenes.
The major objectives of the study are as
follows:
(1) Determine the design criteria for
the land teatment process as it
applies to oily residues. The cri-
teria are loading rates and appli-
cation frequences and tilling fre-
quency.
(2) Study the fate of selected priority
pollutants commonly present in oily
residues.
(3) Assess the atmospheric emissions
from land treatment application of
oily residues.
A sampling program was established
so rates of degradation could be de-
termined. The soil in the zone of incorpor-
ation, top 30.0 cm (11.8 in.) of the
research plots was sampled periodically
for oil content, pH, moisture content, and
nutrients. A 4 x 4 factorial experiment
was proposed with loading rate and
loading frequency as the two variables.
Duplicate combinations of loading rates
and frequencies were established. The
experimental design, including loading
rates and frequencies, were modified as
the study progressed. Limited oil content
monitoring from the unsaturated zone of
heavily loaded plots was performed to
determine the extent of migration of oil
below the zone of incorporation.
The fate of selected organic and inor-
ganic priority pollutants was determined
on two plots with moderate loading rates.
Samples were collected from the top 30.0
cm (11.8 in.) for priority pollutant
analyses. Samples were also taken below
the zone of incorporation to determine if
priority pollutants were migrating.
Atmospheric emissions from land
treatment were also assessed. The
objectives were: 1) to determine the rate
and magnitude of fugitive hydrocarbon
emissions from land treatment of refinery
sludges, 2) to identify the relative effects
of such parameters as sludge loading
rate, temperature, soil moisture content
and relative humidity on the magnitude of
hydrocarbon emissions, 3) to identify and
quantify individual compounds being
emitted to the atmosphere, and 4) to
develop a statistical model to predict the
total volatile emissions rate based on the
above mentioned variables.
Fractionation studies were conducted
on two moderately loaded plots to
investigate the loss kinetics of individual
oil fractions. A fractionation scheme
separated the recovered hydrocarbons
into four fractions: saturates, asphaltenes,
aromatics, and polar compounds.
Results and Discussion
Oil Loss Rates
The oil content of the site soil was
monitored over the study period to
determine oil degradation/loss rates.
Table 1 presents the total losses for the
two-year project period. The oil content
was calculated on a percent dry weight
basis (% dwb). There was good correlation
(R2 = 0.91) between total loss (% dwb) and
total loading (% dwb) during the first year.
The correlation coefficient for the regres-
sion of total second year losses on second
year loadings was 0.81. However, regres-
sion of total losses on the sum of the first
year's antecedent oil content and the
second year's oil loading yielded a
correlation coefficient of 0.94, indicating
that total oil loss was a function of the
sum of antecedent oil plus oil applied in
the second year, and not just that applied
during the second year.
The total loss data for the first year, the
second year, and for the entire study
period indicated that oil losses increased
in proportion to the total oil loadings. An
average of 54% of the total oil applied dis-
appeared during the overall study period
of 657 days with a range from 37.3% to
82.3%. This contrasted with the data
presented in the first year for which the
average percent of applied oil lost was
equal to 39.6% with a range from 9% to
71%.
Rate Constant Evaluation
Rates constants for the disappearance
of oil from the soil were determined for
periods of time immediately following
application and for periods several
months after application. Average first-
order rate coefficients for periods im-
mediately following application was
0.0065 with a standard deviation of
0.0046. Coefficients for time periods
several weeks after application were
0.0046 and 0.0029 with standard devia-
tions of 0.0017 and 0.0010 for the first
and second year periods, respectively.
The fact that coefficients are higher in
periods immediately following application
than during later periods suggests either
some oil fractions are preferentially
degraded, or loss of oil by mecanisms
Table 1. Total Losses During Two
Year Study Period
Plot Percent Percent of Average
dwb Lost Total Lost Percent
dwb Lost
Per Day
(% dwb) (%) (% dwb/day)
1
2
4
5
6
7
8
9
10
11
13
14
15
16
17
18
20
21
22
23
24
25
26
28
29
30
31
32
34
35
36
38
9.22
4.02
2.52
7.89
11.88
3.59
6.15
2.44
7.91
3.96
4.68
1.62
6.88
4.92
5.79
1.39
8.84
15.69
13.86
14.24
3.34
0.88
6.98
13.40
12.51
11.68
6.75
2.64
4.05
13.46
5.23
3.37
46.0
37.3
60.8
53.0
59.3
42.1
63.5
44.5
53.6
46.4
50.9
43.3
46.6
47.5
67.1
49.9
46.6
68.8
61.8
65.7
44.7
39.8
58.0
54.2
55.6
82.3
78.3
53.7
49.2
72.9
41.7
45.1
0.014
0.007
0.004
0.012
0.018
0.005
0.026
0.004
0.012
0006
0.007
0002
0.011
0008
0.010
0.002
0.011
0.024
O.O21
0.023
0.005
0.001
0.011
0.020
0.019
0.018
0.010
0.004
0.006
0.020
0.008
0.005
other than biological, possibly volatilization,
occurs simultaneously with biodegrada-
tion to determine the magnitude of oil
losses via volatilization, a study of volatile
emission rates was performed which is
discussed in the next section. The results
of this study were merged with total
losses to assess the impact on rate
coefficients.
Rate coefficients were calculated
based on total losses and on total losses
minus volatilization which is considered
to be predominately of biological origin.
The removal of volatile losses from the
total losses in the computation of the rate
coefficients had the effect of decreasing
the variance as well as lowering the
mean coefficient from 0.0057 to 0.0033.
It is interesting to note that the 0.0033
value closely approximates the mean
value of 0.0046 and 0.0029 for the first
and second year, respectively, over a
period several weeks after application
and the 0.0057 value correspond the |
-------
value 0.0065 over a short period immedi-
ately following application. Therefore, the
differences in the coefficients determined
from data taken immediately after applica-
tion and those determined weeks after
application can be attributed to the loss of
volatile organics.
Volatile Emissions from Land
Treatment of Petroleum
Residues
The volatile emissions from land
treatment of petroleum sludge were
assessed in this study. Both laboratory and
field studies were used in this assess-
ment. The laboratory tests were used to
measure the volatility of the sludge, and
to estimate weight loss of individual
sludge components as a function of
loading rates, soil temperature, relative
humidity, and soil moisture. Total volatile
losses were also evaluated in the field,
and compared with laboratory results.
The results of the laboratory tests show
that a very sharp rise in the hydrocarbon
concentration in the air appeared during
and immediately following sludge appli-
cation. An abrupt decline from the
maximum concentration through a gradu-
al transition to a lower concentration
followed. The hydrocarbon concentration
in most tests dropped to less than 50
percent of its maximum value within two
hours after application. The laboratory
experiments revealed facts that emission
rates increased with increases in loading
rates, temperature and soil moisture and
decreased with increasing humidity.
The field data showed that higher
loading rates resulted in higher volatile
losses, assuming all other conditions
were constant. It was also found that
there was a variation in the amount of
volatile losses, at a given loading rate,
from one application to another. This
variation could be explained p/imarily on
the basis of volatility of different batches
of sludges. It was found that volatility of
the sludge was a very important factor in
determining emission rates. For this
reason a stripping test was developed in
an attempt to provide a quantitative
measure of relative volatility which could
be related to emission rates.
Two models were developed using
laboratory data relating emission rate in
(g/hr) to loading rate, soil temperature,
soil moisture, relative humidity and time
since application. The first model was
developed for the 10-hour time period
immediately following application, and
the second for greater than 10 hours. The
models are presented below:
Model I - Time < 10 hours
Y = 76.594 + 9.985Xi .769X2 +
8.828X3 - 2.025X4 - 20.645X5
Model II - Time > 10 hours
Y = .184 + .931Xi + .268X2 + 1.879X3
- .371X4- .084X5
Y = emission rate (g/hr)
Xi= percent loading rate
X2 = soil temperature (°F) .
X3 = soil moisture content (%)
X4 = relative humidity (%)
X5 = time since application (hr).
Applying these models to field data,
resulted in a high correlation between the
field results and model predictions. Using
concentrations of hydrocarbons pre-
dicted by the above models in the box
model for calculating equilibrium con-
centrations of air pollutants, the ambient
concentration of hydrocarbons worst
case conditions was found to be below
the Oklahoma Ambient Air Quality
Standards.
Fractionation Studies
Oil recovered from sampling two
moderately loaded plots was fractionated
and the weight fractions determined for
saturates, aromatics, polar compounds
and asphaltenes. Several samples from
the plots were analyzed over the two-year
project period to determine the fate of the
different oil fractions.
The highest total loss during the first
period for both plots occurred for satur-
ates, followed by aromatics, polar com-
pounds and asphaftenes. First period
losses as a percentage of the total applied
were highest for saturates followed by
aromatics, asphaltenes, and polar com-
pounds.
Second period losses were found to
differ substantially from those of the first
period. The most surprising difference
was the decrease in the losses of the
saturates fraction. All fractions with the
exception of polar compounds showed
lower losses during the second period
than the first. The relative magnitude of
the individual fraction losses were
highest for aromatics followed by polar
compounds, asphaltenes and saturates.
Losses of polar compounds increased for
both plots during the second period. The
second period was only approximately
170 days consisting of approximately four
months of relative dormancy during
which time cold weather and saturated
conditions were responsible for low
overall oil losses. The composition of the
sludge applied at the beginning of the
second period was also different than
that applied during the first period. The
weight fraction of saturates was less than
half and all of the other fractions were
from 33 to 295 percent higher than for
sludges applied during the first period.
Anomalous increases in concentration
of asphaltenes (pentane insoluble com-
pounds, saturates, and polar compounds)
were found following the third applica-
tion of sludge. Although these increases
were not expected they can be explained
and have been noted by other researchers.
The time period during which the in-
creases occurred, coincided with cold
weather and saturated soil conditions.
Therefore, anoxic conditions existed with
a possibility of anaerobic decomposition.
During the time period when an
increase in polar compounds was seen,
phenol, 2 nitrophenol and pentachoro-
phenol, as well as benzene, nitrobenzene,
and isophorone, were detected in the soil
matrix. The relatively low apparent losses
of polar compounds may be due to the
production of these compounds as by-
products of the degradation of saturates
and other compounds as has been sug-
gested by several researchers. Thus, the
loss rates as recorded in Table 2 only re-
flect the apparent net losses. Rate co-
efficients were highest for asphaltenes
followed by saturates, polar compounds
and aromatics.
Unsaturated Zone Monitoring
The results of the oil content analysis
show no significant migration belqwthe
zone of incorporation. Analysis of the
unsaturated zone at the end of 406 days
shows that oil content values below 40
cm were similar to the background oil
levels.
Table 2. First-Order Rate Constants for Oil Fractions
Rate Constants (day~1)
Plot
30
30
35
35
Period
7
2
7
2
Asphaltenes
0.0310
O.O160
0.0260
0.0110
Saturates
0.0170
0.0114
0.0140
0.0104
Aromatics
O.OO59
O.OO97
0.0040
O.OO86
Polar
0.0130
0.0055
0.0104
-------
Fate of Metals in Soil
As part of the evaluation of land
treatment, the concentration of metals in
the soil was monitored periodically. The
concentration of selected metals in the
site soil before application of any residues
was determined and compared to the
concentration of the same metals in the
soil at different times during the project.
No significant buildup of metals
occurred during the project period. Zinc
and chromium were present at levels
significantly above background, but the
absolute values were still very low. If the
metal concentrations in the plot with the
highest loading rate are considered, the
useful life of the plot would be limited by
the zinc and cadmium concentrations.
The cadmium concentration in the soil
would reach the critical level in 24 years,
and the zinc concentration in 17 years. No
significant metal migration below the
zone of incorporation occurred.
Modeling and Design of Land
Treatment Systems
Several recommendations relevant to
process modeling and design were made
based on data presented. Although
overall oil losses increase with increasing
loading rates and decreasing loading
frequencies, there is a practical limit
above which operational consideration
such as ability to operate cultivation
equipment and control runoff became
limiting factors.
A maximum hydraulic loading for this
research site based on existing field
conditions was found to be approximately
40 l/m2 (1 gal/ft2). At the oil concentra-
tions of the sludges used in this study (60
- 90 percent) the maximum hydraulic
loading corresponds to approximately 7
percent increase in oil concentration in a
30-cm depth zone of incorporation.
Though higher loadings were in fact
made, operational problems inevitably
resulted.
It must be noted thattheduration of the
study period was not long enough for the
systems to approach equilibrium. There-
fore, the maximum hydraulic loading of
40 l/m2 per application will probably not
be attainable when the oil content
increases as the system approaches
equilibrium. Depending upon the final
equilibrium concentration and the oil
concentration of the sludge, a hydraulic
loading to achieve an oil content per
application of 3 - 4 percent is an achiev-
able goal. Higher concentrations can be
achieved by increasing the number of
applications.
Oily waste land treatment systems
should be designed for equilibrium
conditions'. Equilibrium conditions are
reached when the amount of degradable
material applied is removed (via degrada-
tion and volatilization) in the period prior
to the next application. During and after
equilibration, the possible buildup of re-
fractory organics and inorganics which
may be produced in the process or may be
present in the waste sludge must be
monitored. The buildup of refractory com-
pounds was not found to be a significant
problem in the study; however, equilibri-
um conditions were not achieved.
Though it was shown that the various
oily fractions are removed from the zone
of incorporation at different rates, the
metabolic pathways and biochemical
interrelationships are not sufficiently
understood to warrant the use of a
multiple-substrate process model. Thus,
a pseudo first-order single-substrate
model was developed to predict time to
equilibrium. Application rate and fre-
quency were held constant. If, at equili-
brium, the amount of substrate added, La,
is equal to the amount degraded, (L0 - U),
then the following first-order relation-
ships are valid:
u=u,-u
~ LO \ i - e
And therefore,
La
is an expression for the maximum
equilibrium concentration, where t is the
(constant) time between applications of
La. With Lo known, the number of cycles n
required to reach equilibrium may be
found from the equation:
= La
Z
i=o
Table 3 presents a matrix of equilibrium
values for combinations of loading rate
and loading frequency (LR/LF) which
bracket anticipated practical loading
posibilities. A value of first-order rate
coefficient of 0.003 day"1 was used in the
calculations. Equilibrium was reached
when an increase in maximum concentra-
tion was less than 1 percent of the
previous maximum. Based on the above
assumptions the system would reach
equilibrium in four to five years.
Conclusions and
Recommendations
Conclusions
1 . The project demonstrated that land
treatment is a viable method for
treatment of API separator sludge.
2. Annual loading rates should be
based on projected equilibrium oil
concentrations not exceeding 12
percent oil with an individual applica-
tion maximum of 4 percent oil.
3. Soil should be tilled just preceeding
application and then immediately
following to increase the soil sorption
and holding capacity, respectively.
4. Proper surface slopes are important
to maintain adequate drainage and
control erosion.
5. Rototilling under proper moisture
conditions is important. Tilling under
"wet" conditions resulted in undesir-
able physical changes while tilling
under very dry conditions was not
beneficial.
6. Losses of oil by degradation followed
pseudo first-order kinetics.
7. Variation between sample replicates
and detection-limiting concentrations
hindered monitoring the fate of
priority pollutant present in the
applied waste.
Table 3. Equilibrium Values Assuming K = .003 Day
LR/LF
(%dwb)/(year~')
12/1
12/2
12/4
12/6
12/12
9/1
9/2
9/4
9/6
6/1
6/1
6/4
6/6
La
l%dwb)
12
6
3
2
1
9
4.5
2.25
1.5
6
3
1.5
1
e'Kt
.334
.578
.76 J
.833
.913
.334
.578
.761
.833
.334
.578
.76 J
.833
L
-------
8. Volatile emissions accounted for
about 2/3 of the losses at application,
but only approximately 6 percent of
the total losses over a period of
several months.
Hydrocarbon emissions did not exceed
1979 National Air Quality Standards.
Recommendations
1. Further studies to reinforce the
project's findings should include
optimization of tillage methods under
variable soil moisture conditions and
soil types.
2. The influence of climate variability of
waste constituents in petroleum
refinery sludges, potential for air
pollution, long-term effects of waste
application, closed site revegetation,
and monitoring requirements are
areas needing further research.
3. Full-scale studies to determine
waste generation, waste characteris-
tics, storage, and land requirements
are recommended.
L. E. Streebin, James M. Robertson. H. M. Schornick, P. T. Bowen, K. M.
Bagawandoss. A. Habibafshar, T. G. Sprehe, A. B. Callender, C. J. Carpenter,
andV. G. McFarlandare with The University of Oklahoma, Norman, OK 73019.
Don H. Kampbell is the EPA Project Officer (see below).
The complete report, entitled "Land Treatment of Petroleum Refinery Sludges,"
(Order No. PB 85-148 708/AS; Cost: $26.50, subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, OK 74820
*USGPO: 1985-559-111/10786
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
OCOC329 PS
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STPSiT
CHICAGO IL
------- |