Large Scale Bio-Remediation Projectsinfohouse.p2ric.org/ref/31/30406.pdf · The initial project to be discussed is an abandoned refinery site. It is located in the mid-west, next

LARGE SCALE 610-REMEDIATION PROJECTS John A Christiansen, PE.'

Clayton R. Page, PhD.' Michael D. Thomas, M.S., E.I.T.'

'Environmental Remediation, Inc., Baton Rouge, Louisiana

Bioremediation of hazardous waste and petroleum hydrocarbons and sludges in

soils are still described in scientific literature as an "emerging technology". In fact, today

there are very large scale applications of bioremediation techniques to contaminated

sludges in soils in a variety of regulatory environments. Bioremediation has been

accepted for the Superfund program as evidenced by the number of Records of

Decisions (ROD) in which bioremediation is named. By the end of 1989 there were

twenty five (25) ROD'S which included bioremediation. This amounted to over 15% of

ROD'S written to date. In addition to this acceptance by CERCLA programs,

bioremediation is being used in a number of RCRA underground storage tank and

emergency response sites.

The purpose of our paper is to describe two such large scale bioremediation

applications. The first is the description of the bioremediation of an abandoned refinery

site in the mid-west. This site is being remediated under RCRA using slurry reactor and

land treatment techniques. The volume of waste to be treated is more than 150,000 cubic

yards of sludge and contaminated soil. The second application is the treatment of crude

contaminated beaches in Valdez, Alaska, with emphasis on the development of

specialized microbial cultures. Our discussion in this case will center on the isolation,

adaption and screening of microorganisms for application at this site.

795

The initial project to be discussed is an abandoned refinery site. It is located in the

mid-west, next to a major river. As shown in Figure One, the site consists of three major

surface impoundments which are part of the refinery’s waste water treatment system.

Under the RCRA closure plan, these impoundments were regarded as a single waste

management unit. The surface impoundments, identified as pit, pond and lagoon,

contained different types of petroleum waste. Of these, the pond represented the most

concentrated material. Table One provides characterization information, including a list

of average leveis of contaminants at the site. Of the contaminants, oil and grease was

selected as a tracking parameter during characterization and remediation. Volatile

Organics (VOAs) were of concern during material handling and bioremediation. The

principal volatile compounds of concern were Benzene, Touiene, Ethyl Benzene and

Xylene (BTEX). The limiting parameters for the project are polycyclic aromatic

hydrocarbons PAHs, in particular, carcinogenic PAHs. Under the RCRA closure plan,

these PAHs were divided into carcinogenic and non-carcinogenic categories. The

cleanup standards for these individual constituents are shown in Table Two.

TABLE ONE

Site Characterization

Parameter

Votume, cys STEX, mgi’ig (avg.j PAHs, mg/kg (avg.)

Carcinogenic Total

Oil & Grease, % (avg.)

Pond - Pit

17,000 55,000 288 99-i

640 205 1960 653 19 15

50,000 N/A

235 600 7.5

796

TABLE TWO

RCRA Refinery Site Cleanup Criteria

Parameter

Metals Benzene Benzo (a) pyrene and

Dibenzo (a, h) anthracene Sum of Eight Potentially

Carcinogenic PAHs Total PAHs

Cleanup Level

E.P. Tox Levels 3.2 mg/kg

90 mg/kg

160 mg/kg 300 mg/kg

The bioremedial design of the project involves slurry reactor treatment for the

pumpable sludges and land treatment for the non-pumpable sludges. To achieve this

treatment and remain within the context of RCRA, the existing surface impoundments

would also have to serve as the treatment units. Accordingly, it was decided to store

approximately 50% of the contents of the pond sludges in the pit area, as well as create

a temporary storage area for sludges and soils within the confines of the lagoon. The

remaining lagoon sludges were transferred to this temporary storage area and the lagoon

itself converted into a 6.5 acre land farm and a 1 .O acre storm-water management area.

The units were partitioned as shown in Figure Two.

The basis for this treatment system was extensive laboratory and field treatability

studies conducted during the a 36 month period by several consultants. From these

studies, the ha!f-!ife ya!uns fQr key constitg$nts were idectified as she:"::: i:: Table Three.

Hydraulic retention times for continuous and batch processing were developed. Batch

processing was selected due to variability of the waste, with slurry reactor retention times

of 60 to 135 days identified for the most concentrated constituents. An additional benefit

797

identified during slurry reactor treatment was the destruction of organic solids and the

subsequent volume reduction of sludges. A wastewater treatment system was provided

in order to allow the discharge of free liquid contaminated with degradation metabolites

and salts during the treatment process.

TABLE THREE

Degradation Estimates for Slurry Reactor

Constituents Pond - Pit Lagoon

Oil and Grease, Half-life, Days 69-1 13 20-58 25-65 Total PAHs, Half-life, Days 29-69 1 6-24 NIA Carcinogenic PAHs, Half-life, Days 19-42 44-82 N/A Average Design Hydraulic Retention

Processing, Days 135 90 90 Time Assumed for Sludge

N/A - These sludges were not subjected to bench scale studies

To effect the treatment, a slurry reactor was constructed out of the existing pond.

To maintain a solids concentration of about 10% during slurry treatment, high speed

direct drive mixers were utilized, as shown in Figure Three. These devices used a

polyurethane foam filled float to maintain buoyancy in the reactor. Suction was taken

from the sides of a cone underneath the float and pump through a volute, by high speed

propeller. The pumping action was considered essential to suspend grit, as well as

organic solids. Several of these mixers were placed throughout the slurry reactor, as

shown in Figure Four. !r! order tc! prevent .ncnntm!!ed ccnuring or eresie!? of the pond,

a proprietary mixer frame was designed which acted as a mixer positioning control and

test block. Oxygen transfer was the next requirement. After evaluating the diffused and

mechanical aeration systems, mechanical aeration was selected for several reasons. The

798

primary reason was the recognition that volatilization concerns would be present with

both types of systems. Volatilization of BTEX peaked after the first 48 to 96 hours of

treatment and declined drastically thereafter. With controlled start-up, volatilization would

be effectively controlled. The combination of surface aerators and direct drive mixers

achieved a mixing level of approximately 160 HP per million gallons or about 150% of a

normal activated sludge system. In monitoring during the slurry reactor, treatment was

set so that no detectable carcinogenic constituents would be found outside the exclusion

zone of the treatment unit.

The residuals from slurry reactor treatment, along with non-pumpable sludges,

would be treated in a land treatment unit. The land treatment unit was approximately 150

feet wide by 1,720 feet long. In order to insure adequate removal of water, prevention

of constituent migration and operation under variable loads of sludge, a double liner leak

detection system was used. The first liner consisted of not less than two feet of

compacted soil on which was placed a 200 mil geo-textile drainage weave. On top of this

was placed an 80 mil HDPE liner. On top of the liner was placed more than 24 inches

of porous fill soil containing a drain system. The drain system was designed to remove

the maximum hydraulic load to the land treatment cell within the 24 hour period. During

the construction phase of this project, the lagoon had to be emptied of sludge. After the

construction of the storage cell, approximately 50,000 cubic yards of sludge and 10,000

cubic yards of sludge from the pond were moved in a period of 40 days using hydraulic

dredge and pumping techniques. Free liquid was maintained on top of the lagoon during

this time in order to minimize air admissions. An air emissions control curtain was

constructed around the operating portions of the site to allow admissions monitoring and

799

to isolate areas of high admissions from areas of site operating personnel. During

excavation, operators were often found in Level B personnel protective equipment though

operating environments were reduced to Levels C and D as soon as air monitoring

indicated no threat to human health and environment existed.

Operations consisted of batch treatment of more than 19,000 cubic yards of waste

in the slurry reactor for approximately 60 to 75 days. During this time, monitoring was

conducted for control parameters and target hydrocarbon parameters. Control

parameters consisted of microbial measures, such as Adenosine Triphosphate (ATP) to

estimate total microbial population, and nutrient parameters, such as nitrogen and

phosphorus. In addition, operating parameters such as dissolved oxygen uptake, pH

alkalinity and other activated sludge functions were observed. Target hydrocarbons were

observed monthly in order to accurately access the degradation of target hydrocarbons.

The slurry reactor was de-energized and a sampling crew collected grab samples from

more than 40 sampling locations. These locations were then composited on-site to

provide a target hydrocarbon sample. Mass balances were calculated for both mixed

liquor (suspended organic matter) and settled sludge in order to assess target

hydrocarbon degradation.

During the first year of operation, more than 3.8 million pounds of oil and grease

were loaded into the slurry reactor and processed until at least 67% degradation was

achieved. At this point, the reaction was stopped and the residuals applied to the land

treatment unit. As shown in Table Four, the oil and grease mass balance indicates more

than 74% degradation in the first batch. Subsequently, in the Spring of 1991, the land

treatment unit was sampled and passed the cleanup standards shown in Table Two. A

800

new batch of sludge was loaded into the reactor for treatment. Thus, a large scale

bioremediation unit was used to successfully treat refinery sludges in the batch slurry

reactor followed by land treatment of residuals. More than 19,000 cubic yards of sludge

were treated in the first batch.

TABLE FOUR

Reduction of Mass Oil and Grease Expressed in Pounds

- Date Dav Lbs. of Oil & Grease

7/25/90 0 3,866,650.00 8/22/90 28 1,813,400.00 1 0/3/90 70 1,255,330.00 10/24/90 91 992,727.00

% Reduction

0.00% 53.1 0% 67.53% 74.32%

The second study of bioremediation follows the extensive work carried out by

Exxon, under supervision of the United States Environmental Protection Agency (USEPA)

and the Alaska Department of Environmental Conservation to clean-up the spill from the

Exxon Valdez in March of 1989. During the first 12 months of remediation, Alaska

bioremediation work consisted of several elements, including:

1. Physical and chemical treatments to remove oil from beaches using skimmers, hot water and steam cleaning and chemical surfactants and dispersants.

2. Applications by Emon of oleophillic fertilizers to about 1 10 miles of polluted beaches.

3. Evaluation of the potential of adding microbes to Alaska’s beaches to stimulate biodegradation.

801

This third effort involved testing of the biological product under laboratory

circumstances simulating the marine environment. In order for field evaluation to be

considered, the potential return on the addition of microbes would have to be greater

than that seen from the addition of nutrients and fertilizers alone.

Several types of nutrients had already been evaluated on the polluted beaches.

These included slow release, water soluble fertilizers in both briquette and granular forms,

a water soluble fertilizer which was applied with a sprinkler system, and liquid oleophillic

fertilizers. The oleophillic fertilizers had been specially formulated to keep nutrients and

oil in contact. The initial work done on-site indicated that significant visual changes in oil

removal, improvement in total microbial and specific hydrocarbon degrading bacterial

populations, low concentration of ammonia and phosphate in seawater and some

decrease in oil over time. Based on the available evidence, research concluded that the

addition of fertilizer accelerated the break down of oil by two to four times the natural rate

of biodegradation. The water soluble fertilizer was deemed the most effective when

applied by the sprinkler system. For large-scale applications, the use of oleophillic

fertilizer in slow release granules was determined to be more practical to use. This

became the basis for the second year’s bioremediation efforts.

In order to develop a more comprehensive method of responding to oil spills and

more effective biodegradation products, USEPA appointed National Environmental

Technology Applications Corporation (NETAC) to develop protocols to screen

bioremediation products for enhanced degradation capabilities or weathered crude.

NETAC selected a panel that represented academia, the bioremediation products

industry, and bioremediation research institutes. EPA requested proposals on promising

802

bioremediation products that could be used to clean the contaminated beaches. NETAC

received solicitations from thirty-nine companies. The criteria for non-feasibility were that

the product could not contain the following:

1. Any genetically engineered organisms,

2.

3.

4.

Any known or suspected human or animal carcinogen,

Any known pathogen to humans or indigenous flora and fauna, and

Any chemical listed in 40 CFR 268 Land Disposal Restrictions.

The NETAC panel met in March, 1990 to review the proposals for demonstrated

effectiveness of the products in degrading oil, feasibility of applying the product on a

widespread basis, toxicity, experience and capability. Eleven promising bioremediation

products were identified and recommended to EPA for laboratory testing.

During the first quarter of 1990, tests were conducted to evaluate and screen

biodegradation products. The methods of analyses were developed by the EPA’s Risk

Reduction Engineering Laboratory in Cincinnati, Ohio. The first method used to evaluate

the products was stirred electrolytic respirometers. Studies were conducted using

continuous oxygen-uptake measuring, Boith Sapronat (model D-12) , electrolytic

respirometer. These were temperature controlled through water baths surrounding the

unit and equipped with recorders with digital indication and direct measurement of

oxygen uptake velocity curves. The experimental design consisted of duplicate testing

of each commercial product. Each experimental respirometer flask was charged with:

1.

2.

3.

Weathered crude oil, 250 mg

Commercial product at vendor specified concentrations

Synthetic sea water, 250 ml

803

The procedure to create the liquor used rocks contaminated with weathered

petroleum crude from which crude was extractor using two liters of sea water. The

respirometers were incubated at 15" C to simulate Alaska's conditions in the dark and

continuously stirred by magnetic stirrers. Control flasks were used to represent the

background oxygen uptake of the product in sea water without oil. A non-nutrient control

was used to represent the oxygen uptake of organisms from washed beach material and

sea water on weathered oil without any external source of nutrient addition, Le.,

background levels.

At the same time the respirometry analysis was being conducted, shaker flask

microcosms were set-up to assess the quantitative changes in oil composition by

chromatographic separation of the individual components. Flasks were set up to

correspond to those shown in Table Five. These were 2.0 liter flasks containing 1 .O liter

of sea water with 10,000 ml/l of weathered oil and commercial products, and inorganic

nutrients at ten times the level used in the respirometer Flask. The higher concentration

of weathered oil was used to improve the final sensitivity of the chemical analysis. In

addition to 22 flasks set up, EPA set up 18 supplemental microcosms representing nine

sterile product controls and nine sterile background controls. Sterile background controls

were sterile oil and sea water, but non-sterile product evaluated the effects of competition

from naturally occurring micro-organisms.

804

TABLE 5

Experimental Design for Respirometric Studies

Reaction Weathered Commercial Vessel - Oil Product Seawater TOTAL

TEST FLASKS: P" + + + 2 F1.2 + + 2 - CONTROL FLASKS: - CP, + + 2 CF C,-inoculum

- + 2 + 2 -

- 2 - Ci-no nutrient + + TOTAL 12

PN duplicate commercial products (n=9) F1 I2 chi CF c1, c2

fertilizer (inorganic N + P nutrients) controls (product and fertilizer) inoculum and non-nutrient controls

Analysis from the second set of shakeflasks was conducted by measuring the

alkane and aromatic fractions of oil extracts. The aliphatic fractions were measured using

gas chromatography using a flame ionization detector. The aromatic fractions were

characterized by gas chromatography/mass electrospectometry. In addition to gas

chromatography analysis for oil constituents, nutrients and microbial testing was

conducted. Microbial testing consisted of total micro-organisms and total hydrocarbon

degrader populations. Microbial growth for one product is shown in Figure Six. Growth

on the hydrocarbon is indicated. Results were also quantified by oxygen uptake and

a!kane degradattinn, exyge!? uptake assocI2ted ?.rith one prcduct gr0V"m is she:.::: in

Figure Seven. The most significant results of alkane degradation for day eleven are

shown in Figure Eight. This chart illustrates alkane reduction over nutrient and sterile

controls. Two of the biological products exceeded the alkane degradation seen from

805

fertilizer alone at a significant statistical level (p 2 0.05) on day eleven. The two biological

products tested do show improved alkane degradation for day eleven over the product

used on the Alaskan beaches in 1989, as shown in Figure Nine.

With these results in hand, NETAC prepared to set up a field bioremediation

demonstration using plots for control, nutrient amendments and the two NETAC

recommended microbial products. Tests were conducted from the end of July to late

August in 1990, a period of 27 days. USEPA and NETAC researchers concluded that

while laboratory results were promising, field results on the two microbial products were

inconclusive since no statistically significant enhancement of the rate of biodegradation

occurred in the field. It was noted that tests were conducted on oil that had weathered

and degraded naturally during the 18 month period since the Exxon Valdez spill. The

more easily degraded components of the oil had already disappeared prior to the field

testing.

Conclusions:

The first case study described involved the successful design, construction and

execution of a large scale bioremediation project. For work done in the first phase of

this project it is apparent that bioremediation of a petroleum refinery waste sludge to

reduce carcinogenic and non-carcinogenic chemical constituents and reduce volume was

successful using the combination of the slurry reactor and land treatment methods. In

the second case study, commercial microbial inoculum had a demonstrated afPinity to

degrade crude petroleum constituents over fertilizer in laboratory tests. However, field

results were inconclusive for the demonstration of a full scale process, due primarily to

the limited time frame of the test conditions.

806

REFERENCES

1. Bioremediation of Hazardous Wastes, U.S.E.P.A. - O.S.W.E.R., Technology Innovation Office, August, 1990.

2. Selected Data On Innovative Treatment Technolorries, U.S.E.P.A. - O.R.D., EPA/600/9-90/041 , December 1990.

3. "Bioremediation In The Field", U.S.E.P.A. - O.S.W.E.R., No. 1 , EPA 540/2-90/004, November, 1990.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

"Operating Plan for the Closure of the Sludge Pond, Sludge P. 4, and Wastewater Treatment Lagoon As a Single Waste Management Unit", Submitted to Missouri Department of Natural Resources and U.S.E.P.A. Region VI1 by Amoco Oil Co., July, 1989.

"Sampling and Analyses Program in Connection with RCRA Closure/Lagoon Cleanup", submitted to Missouri Department of Natural Resources by Amoco Oil Co., November, 1988.

"Biological Treatment Demonstration, Biological Treatment of Petroleum Refinery Sludges From the Sugar Creek Facility Tank - Scale Study", Remediation Technologies, Inc., Ft. Collins, Colorado, December, 1988.

Bioremediation For Marine Oil Stills, U.S. Congress, Off ice of Technology Assessment, OTA-BP-0-70, May 1991.

Berkey, E., et al. "Evaluation Process for the Selection of Bioremediation Technologies", Environmental Biotechnolonv for Waste Treatment, Plenum Press, New York, New York, 1991.

Vanosa, A., et al. "Tests to Screen Bioremediation Products for Efficacy Against Weathered Crude Oil Using Respirometer and Shaker Flask Microcosms", U.S.E.P.A. - Risk Reduction Engineering Laboratory, Cincinnati, Ohio, March 23, 1990.

"Special Notice", U.S. Commerce Business Daily, February 12, 1990, No. 1953N.

R.M. Atlas, 1984. Petroleum Microbiolonv, MacMillan Publishing Company, New YOTK, Xew -fork. \ I - .I

M.R. Overcash, et all 1979. Desicln of Land Treatment Svstems for Industrial Wastes - Theow & Practice, Ann Arbor Science, Ann Arbor, Michigan.

U.S. Environmental Protection Agency (EPA), 1986. Test Methods for Evaluatinq Solid Waste: Phvsical/Chemical Methods. SW-846, 3rd Edition.

807

14. APHA, 1985. Standard Methods for the Examination of Water and Wastewater, 15th Edition.

15. EPA, 1083. Methods for Chemical Analvsis of Water and Wastes. EPA-600/r-79- 020.

II

808

PRODUCT G

i o g

I O 8

i o 7

105 0 z

i o 4

i o 3 0 10

Non-Sterile

Sterile Product

Sterile Background

No Oil

20 30

TIME, DAYS

40 50

FIGURE 6

809

-L 0

0 0

O I

tn 0

CUMULATIVE 0 2 UPTAKE, mg/l h)

si: 0 0 0 Q Q 8

0

I

810

PRODUCT G 16000

2000

0

UNSTERILE

STERILE PRODUCT

STERILE BACKGROUND

-0- NO NUTRIENTS

ALKANES FIGURE 8

81 1

1

0 0

t

E W

(D

I

y! m

a,

3 W

a,

n

812

DRAWING NOT TO SCALE

FIGURE 3 SLURRY MIXER

I

815

Top o f dike

Bottom of reactor Water line

. .

i Sludge ; Storage ; Cell

....................................... .

\ L u,,.,.,, - ., .. - - Y - Y - ,. ,. ,. Y - ..

llsolation Curtain i FIGURE 4

SLURRY REACTOR LAYOUT \

c3 031 0

4000

3500

3000

2500

2000

1500

lo00

500

0

0 28

Time (Days) 70 91

REDUCTION OF MASS OF OIL AND GREASE EXPRESSED IN POUNDS

817 FIGURE 5

OIL & GREASE MASS BALANCE BATCH ONE

ESTIMATING ENVIRONMENTAL RESTORATION COSTS I N THE DOE COMPLEX

Marc Zocher Environmental Restoration Technical Support Office

LOS Alamos National Laboratory Los Alamos, New Mexico 87545

Introduction

The United States Department of Energy (DOE) is embarking on one of the largest environmental restoration undertakings of this century. The diversity of locations, site conditions, nature and extent of contaminants, and regulatory environments make this an exceedingly complex, and therefore costly, program. The planned budget for fiscal year 1992 alone exceeds 5.9 billion dollars for DOE environmental restoration and waste management activities'. minimize costs and maximize returns (lower risk to human health and the environment), good cost estimating tools and techniques must be employed.

Traditional Cost Estimating

The DOE has been estimating the costs of construction successfully for years. Typical construction projects within the DOE are laboratory and process buildings, offices, support facilities, and infrastructure (roads, utilities). Most of these projects and project components follow typical construction practices and are built by a contracted work force.

The in-house estimating staff that prepares government estimates follow policies and guidelines that cover the development of the base estimate, the application of contingency, and the use of escalation factors. Many sources of data exist to assist the estimators.

In order to

They include:

0 Commercially available estimating handbooks (e.g. R . S . Means). These books compile actual productivity , equipment, and material costs from many sources across the United States. By developing this large of a database, the actual cost history of the included project components can be used as good reference material.

0 Site specific actual costs. Estimators rely on data from projects completed at their installations for cost elements. past, this data is invaluable for determining site specific cost drivers.

Computer models and estimating tools. general shell to develop internal databases, or specific models that contain their own cost elements to use and mndify.

If similar projects have been completed in the recent

0 These computer software packages are either a

' DOE Five-Year Plan (1990). This figure includes both validated and unvalidated estimates and will be updated in the September, 1991 release of the five-year plan.

818