PROCEEDINGS OF
THE OPEN FORUM
ON MANAGEMENT OF
PETROLEUM REFINERY
WASTEWATERS
PRESENTED BY
THE ENVIRONMENTAL
PROTECTION AGENCY
THE AMERICAN PETROLEUM
INSTITUTE
THE NATIONAL PETROLEUM
REFINERS ASSOCIATION
THE UNIVERSITY OF TULSA
-------
PROCEEDINGS
OF
OPEN FORUM
ON
MANAGEMENT OF PETROLEUM REFINERY WASTEWATERS
Presented by
The U. S. Environmental Protection Agency
The American Petroleum Institute
The National Petroleum Refiners Assdeiation
The University of Tulsa
Francis S. Manning
Editor and Project Director
Professor of Chemical Engineering
University of Tulsa, Tulsa, Oklahoma
Fred M. Pfeffer
Project Officer
Research Chemist
Robert S. Kerr Environmental Research Laboratories, Ada, Oklahoma
-------
ACKNOWLEDGEMENTS
An Open Forum on Management of Petroleum Refinery Wastewaters was held at the
Camelot Inn, Tulsa, Oklahoma. This four-day (January 26-29, 1976) symposium was
sponsored by the U. S. EPA in the form of Grant No. R803957-01-0 to the University of
Tulsa. Fred M. Pfeffer and Francis S. Manning served as Projector Officer and Prefect
Director respectively. In turn the University of Tulsa contributed partial matching funds;
while the American Petroleum Institute and the National Petroleum Refiners Association
helped defray the costs of publishing the written proceedings of the Open Forum.
The cooperation of the API, the NPRA, and many petroleum industries in suggesting
and engaging speakers and publicizing the Open Forum is gratefully acknowledged.
Success of any project is aided immeasurably by support "from the top". Such
support was provided enthusiastically by Nick Gammelgard and Arne E. Gubrud, API; Herb
Bruck, NPRA; Bill Galegar, EPA, Ada/Okla.; and J. Paschal Twyman, University of Tulsa.
Of course the speakers' contributions cannot be overestimated. Not only did they
present most knowledgeable and timely papers but they also reviewed the resulting discussions.
The Project Director and Project Officer regret that it is impossible to identify all
of the numerous colleagues and friends who contributed so much to this symposium. How-
ever special thanks are due Katie Whisenhunt, Shirley Clymer, and Cathy Whisenhunt for
managing the registration and typing the proceedings; Ed Andrews for tape recording the
proceedings; Nelda Whipple and Dawn Buss for typing the proceedings; and Tom Redmond
for redrafting many figures.
11
-------
CONTENTS
Acknowledgements (ii)
List of Participants (v)
Welcome 1
Keynote Address 3
Session 1 Petroleum Refining Guidelines 17
Allen Cywin "Establishment of Petroleum Effluent Guidelines" 19
Martha Sager "The Role of the Effluent Standards and Water Quality 29
Advisory Committee"
Joe G. Moore, Jr. "The Role of the National Commission of Water 39
Quality (NCWQ)
Robert T, Denbo "Economic Impact of Wastewater Effluent Guidelines 51
for Petroleum Refining"
Session II Permits for the Point-Source Category" 57
Carl J. Schafer "The Environmental Protection Agency's Permit System" 59
Robert F. Silvus "Current Approach to Refinery Permits" 63
William K. Lorenz "Petroleum Industry Experience With The NPDES Permit 69
System "
Harless R. Benthul "The Enforcement of Permits" 77
F. T. Weiss "Measuring the Parameters Specified in Permits" 91
Banquet Proceedings and Address 105
Session III Biological Treatment 113
W. Wesley Eckenfelder "Activated Sludge Treatment of Petroleum 115
Refinery Wastewaters an Overview"
James F. Grutsch "A New Perspective on the Role of the Activated 127
Sludge Process and Ancillary Facilities"
Charles E. Ganze "Case History on Biological Treatment of a Petroleum 153
Refinery Wastewater"
Mohammed A. Zeitoun "Optimization of the Activated Sludge Process 159
Through Automation"
FredM. Pfeffer "The National Petroleum Refinery Wastewater Characterization
Study" 183
Session IV Sludge Management 197
Carl E. Adams, Jr. "Sludge Handling Methodology for Refinery Sludges" 199
Jacoby A. Scher "Processing of Waste Oily Sludges" 233
• • •
in
-------
C. B. Kincannon "Oily Waste Disposal by Soil Cultivation" 257
R. L. Huddleston "The Disposal of Oily Wastes by Land Farming" 273
Session V Activated Carbon Treatment 293
Davis L. Ford "Current State of the Art of Activated Carbon Treatment" 295
C. T. Lawson "Cautions and Limitations on the Applications of Activated 345
Carbon Adsorption to Organic Chemical Wastewaters"
Joyce A. Rizzo "Case History: Use of Powdered Activated Carbon in an 359
Activated Sludge System"
Leon H. Myers "Pilot Plant Activated Carbon Treatment of Petroleum 375
Refinery Wastewater"
M. A. Prosche "Activated Carbon Treatment of Combined Storm and 399
Process Waters"
Session VI Miscellaneous Topics 411
Lial F . Tischler "Inherent Variability in Wastewater Treatment" 413
R. T. Milligan "Reuse of Refinery Wastewater" 433
Sterling L. Burks "Biological Monitoring of Petroleum Refinery Effluents" 445
Ronald G. Gantz "API - Sour Water Stripper Studies" 459
Session VII Future Research 483
Wilson K. Talley "The EPA's Role in Future Research" 485
Arne E. Gubrud "The Industry's Role" 497
Joseph F. Molina "The University's Role in Future Research" 501
Summary 509
IV
-------
PARTICIPANTS
OPEN FORUM
January 26-29, 1976
Arthur J. Abington
Applications Engineer
Mapco
1437 South Boulder
Tulsa, OK 74103
Ray W. Amstutz
Industry Manager
Williams Bros. Waste Control
6600 South Yale
Tulsa, OK 74130
Mike Anderson
Pollution Chemist
Amerada Hess
Box 425
Purvis, MS 39475
Raymond V. Anderson
Independent
217 Gadman Drive
Williamsville, NY 14221
Gary Bartlett
Sales Engineer
Chase Incorporated
P.O. Box 42546
Tulsa, OK 74145
R.N. Beals
Manager, Engr. & Admin. Services
Chemplex Company
P.O. Box 819
Clinton, |A 52732
Harless R. Benthul
Chief, Legal Branch
EPA
1600 Patterson St., Suite 1100
Dallas, TX 75201
Carl E. Adams
President
AWARE, Inc.
P.O. Box 40284
Nashville, TN 37204
Jack W. Anderson
Coordinator, Ref. & Gas Plant Operations
CRA,lnc.
P.O. Box 7305
Kansas City, MO 64116
R.C. Anderson
Superintendent
CRA,|nc.
P.O. Box 311
Scottsbluff, NE 69361
W.L. Banks
EPA
1735 Baltimore Avenue
Kansas City, MO 64108
Glendon W. Bassett
Environmental Control Engineer
Lion Oil Co.
Lion Oil Refinery
El Dorado, AR 71730
Norman L. Benecke
Manager, Environmental Engineer
The Lummus Company
1515 Broad Street
Bloomfield, NJ 71730
John H. Bentz
Senior Engineer
Exxon USA
P.O. Box 3950
Baytown, TX 77520
-------
Robert E. Bergman
Sales Engineer
Johnson Div., UOP, Inc.
Box 111, LA-PA Twoer
Conroe, TX 77301
Edgar H. Bienhoff
Plant Superintendent
CRA, Inc.
P.O. Box 608
Phillipsburg, KS 67661
John C. Blake
Bechtel Corporation
P.O. Box 3965
San Francisco, CA 94119
Edmund D. Blum
Coordinator-Environmental Programs
Union Oil Co. of CA
Union Oil Center, Box 7600
Los Angeles, CA 90017
Desmond H. Bond
Process Engineer
Ford, Bacon & Davis Texas
P.O. Box 38209
Dallas, TX 75238
Richard Brown
Dept. Staff
Mitre Corp.
1820 Dolley Madison Blvd.
McLean, VA 22101
Robert H. Bruggnik
Director of Environmental Control
Clark Oil & Refining Corp.
Box 297
Blue Island, IL 60406
Marion Buercklin
Coordinator Environmental Affairs
Sun Oil Co.
Box 2037
Tulsa, OK 74102
Milton R. Beychok
Consulting Engineer
1 7709 Oak Tree Lane
Irvine, CA 92664
Richard J. Bigda
President
Richard J. Bigda and Associates
5577 South Lewis Avenue
Tulsa, OK 74105
Robert Blick
Systems Manager
Ameron Process Systems Division
1000 South Grand Avenue
Santa Ana, CA 92701
G.C. Blytas
Sr. Staff Research Chemist
Shell Development
14323 Apple Tree
Houston, TX 77024
Clifford W. Bowers
Director, Technical Operations
Metcalf & Eddy Industrial
50 Staniford St.
Boston, MA 02114
Herb Bruch
Technical Director
National Petroleum Refine rs Association
1725 DeSales St. NW
Washington, D.C. 20036
B .B. Buchanan
R &D
Phillips Petroleum
Batlesville, OK 74003
Sterling L. Burks
Assistant Director, Reservior Center
Oklahoma State University
Stillwater, OK 74074
VI
-------
Robert D. Burns
Process Engineer
Great Canadian Oilsands Ltd.
Box 4001
Ft. McMurray, Alta, Canada T9H-E33
Janis Butler
Lab Director
Wilson & Co .
Box 28
Salina, KS 67401
R.D. Cameron
Assist. Air & Water Conservation Coordinator
Texaco Canada Limited
90 Wynfbrd Drive]
Toronto, Ontario, M3C 1 K5, Canada
John Carpenter
Mobil Oil Corporation
635 Elk Street
Buffalo, N.Y. 14210
A.W. Catanach, Manager
Env. Control & Laboratories
Neches Butane Products Co.
P.O. Box 817
Pt. Neches, TX 77651
Frank J. Cebula
Lead Process Engineer
Sun Oil Co.
Box 426
Marcus Hook, PA 19061
J.J. Chavez
Process Technologist
Caltex Petroleum Corp.
380 Madison Avenue
New York, N.Y. 10017
B.E. Clark
Tech. Manager
Williams Bros. Waste Control
6606 South Yale
Tulsa, OK 74136
VII
Lee C. Burton
Environmental Engineer
OK State Health Dept.
NE 10th and Stonewall
OK City, OK 73118
John Byeseda
Graduate Student
University of Tulsa
600 South College
Tulsa, OK 74105
J.N. Cardall
Environmental Staff Engineer
Mobil Oil Corporation
3700 West 190th Street
Torrance, CA 90509
Rosemary Carpenter
Engineer
Ameron Process Systems Division
1000 South Grand Avenue
Santa Ana, CA 92705
Jack L. Caufield
Environmental Engineer
Toscopetro Corp.
P.O. Box 2860
Bakersfield, CA 93303
Samuel Z. Chamberlain
Prod. Eng.
Monsanto
P.O. Box 711
Alvin, TX 77511
Cho K. Ching
Environmental Engineer
U.S. EPA
26 Federal Plaza
New York, N.Y. 10007
Edward L. Clark
Ind. Sales Manager
Union Carbide
2 Greenway Plaza E.
Houston, TX 77046
-------
William R. Clarke
Reject- Eng. Air & Waier Consv.
Texaco, Inc.
P.O. Box 2389
Tulsa, Ok 74101
Larry T. Clere
Environmental Engineer
Sun Oil Co.
P.O. Box 920
Toledo, OH 43693
Harold H. Coffman
Manager of Quality Control
Petrolite Corp. Bareco Division
Barnsdall, OK 74002
R.'G. Coker
Texaco Refinery
902 W. 25, Box 2389
Tulsa, OK 74101
Gary Congram
Proc. Eng. Editor
Oil & Gas Journal
P.O. Box 1260
Tulsa, OK 74127
Leon R. Corpuz, Jr.
Mgr. of Utilities
Pasco, Inc.
P.O. Box 277
Sinclair, WY 82334
Dennis Creamer
Analytical Chemist
Continental Oil Co.
5801 Brighton Blvd.
Commerce City, CO 80022
Ben Crocker
Process Engineer
Williams Bros. Waste Control
6600 South Yale
Tulsa, OK 74136
Dr. L. Davis Clements
Assistant Professor
Texas Tech University
School of Chemical Engineering
Lubbock, TX 79409
Michael J. Coffey
Manufacturing Representative
Dow Chemical Co.
P.O. Box BB
Free port, TX 77541
Bruce W. Cogswell
Process Engineer
Champlin Petroleum Co.
Box 552 -Ref. Office
Enid, OK 73701
H .H. Comstock
Director of Environmental Control
Phillips Petroleum Co.
2029 Firfax Road
Kansas City, KS 66115
Jack B. Cornett
Exec. Vice President
W.R. Holway & Associates
4111 South Darlington
Tulsa, OK 74135
Allan Craig
Partner
Craig & Keithline
1129 East 15th Street
Tulsa, OK 74120
L.W. Cresswell
Conoco
Ponca City, OK 74601
Ernest F. Cross
Manager of Engineering
Brown & Root, Inc.
P.O. Box 3
Houston, TX 77001
VII I
-------
Joseph B. Crowley
Technical Coordinator
Marathon Oil Co.
539 South Main
Findlay, OH 45840
Leonard J. Daniels
Director of Marketing
Air Resources, Inc.
P.O. Box 87
The Woodlands, TX 77373
Mike Davis
Supervisor of Environmental Control
Southwestern Refining Company
P.O. Box 9217
Corpus Christ!, TX 78408
Wi 11 ia m Dea n
Associate Analytical Chemist
Sun Oil Co.
17th and Union Avenue
Tulsa, OK 74102
James F. Dehnert
Pilot Plant Supervisor
Phillips Petroleum Co.
Avon Refinery
Martinez, CA 94553
Robert T. Denbo
Environmental Control Coordinator
Exxon Co.
Baton Rouge, LA 70801
F.A. Devine
Senior Project Engineer
Exxon Research and Engineering G>,
P.O. Box 101
Florham Park, N.J. 70801
John H. Dobson
Technical Services Manager
Champlin Petroleum Co.
Corpus Christie, TX 78408
Allen Cywin (WH 552)
Director, Effluent Guidelines Div.
EPA
401 M St.,S.W. East Tower, Rm 913
Washington, D.C. 20460
Carter B. Davis
Ch. Chemist
OKC Refining Co.
Box 918
Okmulgee, OK 74447
Irv Deaver
W. Div. Mgr.
Ford, Bacon & Davis TX.
7966 E. 41st
Tulsa, OK 74145
Wallace R. Decker
Special Projects Coordinator
Derby Refining Co.
P.O. Box 1030
Wichita, KS 67201
Paschal B. Dejohn
Project Leader
1C I United States
Route 202 & Murphy Road
Wilmington, DE 19897
Albert P. Dennis
Senior Associate Engineer
Mobil Research & Development Co.
P.O. Box 1026
Princeton, NJ 08540
Harold L. Dinsmore
Process Engineer
Williams Bros. Waste Control
6600 S. Yale
Tulsa, OK 74136
V.O. Dodd
Texaco
P.O. Box 2389
Tulsa, OK 74101
IX
-------
Jay R. Dungan
Project Manager
Ford^acon & Davis Texas
P.O. Box 38209
Dallas, TX 75238
M.L. Dunton
Amoco Production Co.
P.O. Box 591
Tulsa, OK 74102
Larry Echelberger
Environmental Coordinator
Marathon Oil Co.
P.O. Box 1191
Texas City, TX 77590
Charles D. Edwards
Product Manager
ARCO Chemical Co.
P.O. Box 370
Sand Springs, OK 74063
Kenneth Eisdorfer
Project Engineer
Coalcon
1 Penn Plaza
New York, N.Y. 10001
S. David Ellison
Manager, Industrial Processes
CH 2 M Hill
1930 Newton Square, Room 201
Reston, VA 22090
R.E. Etter
Environmental Coordinator
Sun Oil Co.,-Tulsa Refinery
P.O. Box 2039
Tulsa, OK 74102
R.T. Evans
Environmental Coordinator
Mobil Oil Co.
Box 546
Augusta, KS 67010
R.D. Dunn
Phillips Petroleum
Bartlesville, OK 74003
John H. Dyer
Process Engineer
Union Texas Petroleum
P.O.Box 2120
Houston, TX 77001
W.Wesley Eckenfelder
Professor
Varderbilt University
Box 6222-Station B
Nashville, TN 37203
Robert Ehrlich
Environmental Coordinator
Hess Oil Virgin Islands Corp.
Engineering Dept. Box 127
St. Croix, US Virgin Islands 00850
J.P. Ellis
Operations Supervisor
Standard Oil Co. of California
324 W. ElSegundo Blvd.
El Segundo, CA 90245
Richard Elton
Project Engineer
Engineering Science
3109 N. Interregional
Austin, TX 78740
Edward D. Evans
Chief Chemist
Skelly Oil Co.
P.O. Box 1 121
El Dorardo, KSK 67043
R.A. Farnham
Chevron Research
Richmond, CA 94802
-------
B.J. Flemming
Project Engineer
Texaco, Inc.
P.O. Box 2389
Tulsa, OK 74101
Davis L. Ford
Regional V.P.
Engineering Science, Inc.
3109 N. Interregional
Austin, TX 78722
Arthur E. Franzen
Supt. Reclamation & Conservation
Amoco Oil Company
Whiting Refinery
Whiting, IN 46394
Milton G. Freiberger
Process Engineer
Williams Bros. Waste Control
6600 South Yale
Tulsa, OK 74136
Neale Fugate
Project Manager
Ford, Bacon & Davis Texas
P.O. Box 38209
Dallas, TX 75238
William C. Galegar
Director
R.S. Kerr Research Laboratories
EPA
Ada, OK 74820
Ronald G. Gantz
Supervising Process Engineer
Continental Oil Co.
P.O. Box 1267
Ponca City, OK 74601
N.E. Garland
Environmental Engineer
Ethyl Corp.
P.O. Box 341
Baton Rouge, LA 70821
Peter J. Foley
Environmental Engineer
Mobil Oil Co.
150 East 42nd Street
New York , N.Y. 10017
Orval L. Fouse
Supervisor, Utilities & Environmental Eng.
Gulf Oil Co.—U.S.
P.O. Box 701
Port Arthur, TX 77640
Allden R. Frederickson
Supervisor Environmental Control
Murphy Oil Corp. Meraux Reg.
P.O. Box 100
Meraux, LA 70075
George Frondorf
Engr. Suprv.
Amerada Hess Corp.
Purvis Refinery, P.O. Box 425
Purvis, MS 39475
R.E. Funk
Environmental Coordinator
Cities Service Co.
P.O. Box 300
Tulsa, OK 74102
P. N. Gammelgard
Senior Vice President, Environmental Affairs
American Petroleum Institute
2101 L St., N.W.
Washington, D.C. 20037
Charles Ganze
Facilities Supervisor
Gulf Coast Waste Disposal Authority
910 Bay Area Blvd.
Houston, TX 77058
John E. Garner
Manager, Environmental Affairs
TOTAL Leonard, Inc.
P.O. Box 231
Alma, Ml 48801
XI
-------
Refer C. Gaskin
Supervisor, Environmental & Quality Control
Kerr-McGee Chemical Corp.
P.O. Box 25861
OK City, OK 73125
William G. Geubelle
Analytical Chemist
Conoco
P.O. Box 1267
Ponca City, OK 74601
Dan H . Gillum
Environmental Engineer
Phillips Petroleum
Box 271
Borger, TX 79007
E.F. Gloyna
Dean of Engineering
University of Texas
Austin, TX 78712
James W. Godlove
Environmental Engineer
Phillips Petroleum
10 Bl Phillips Bldg.
Bartlesville, OK 74004
H,M. Gomaa
Environmental Engineer
Lummus Co. Canada
255 Consumers Road
Willowdale, Ontario, Canada
Fred W. Gowdy
Senior Staff Engineer
Exxon
P.O. Box 551
Baton Rouge, LA 70821
Robert W. Griffin
Manager, Special Projects
NUS Corp.
1910 Cochran Road
Pittsburgh, PA 15220
Rene Gaudet
Environmental Engineer
Tenneco Oil
P.O. Box 1007
Chalmette, LA 70043
G .H. Gaumer
Refinery Representative
Phillips Petroleum
Avon Refinery
Martinex, CA 94553
John A. Ghser
Director Lab and Environmental Affairs
Gulf Oil Co. US
Box 2487
Santa Fe Springs, CA 90670
H.W. Goard
Phillips Petroleum
Barflesville, OK 74003
Robert Goldsworthy
Project Engineer
Standard Oil Co. of CA
324 W. El Segundo Blvd.
El Segundo, CA 90505
W,.J. Gossom
Vice President
Resource Sciences Corp.
6600 South Yale
Tulsa, OK 74136
John H. Gray
Environmental Engineer
Ashland Petroleum Co.
1409 Winchester Ave.
Ashland, KY 41101
Jim Grutsch
Coordinator of Environmental Projects
Standard Oil (Indiana)
200 East Randolph Drive
Chicago, IL 60601
XII
-------
Arne E. Gubrud
Director, Environmental Affairs
American Petroleum Institute
2101 L St., N.W.
Washington, D. C. 20037
John D. Hallett
Senior Engineer
Shell Oil Co.
P.O. Box 2463
Houston, TX 77001
Mike Henebry
Mkt. Mgr. of Petroleum & Chemicals
Environmental Research and Technology
Concord, MA 01742
Thomas Hickey
Process Design Engineer
CE Lummus
1515 Broad Street
Bbomfield, NJ 07003
Guy Hillard
Mgr. of Quality Control
Witco Chemical Corp.
77 N. Kendall Avenue
Bradford, PA 16701
Larry C. Holland
Environmental Coordinator
Cities Service Co.
P.O. Box 300
Tulsa, OK 74102
W.F. Hoost
Process Manager
Pullman - Kellogg
Three Greenway Plaza East
Houston, TX 77046
R.L. Huddleston
R &D
Continental Oil Co.
Ponca City, OK 74601
Jack A. Guthrie
Assistant Refinery Manager
Sun Oil Co.
Box 426
Marcus Hook, PA 19061
Pat Havener
Mechanical Engineer
Navajo Refining Co.
P.O. Box 159
Artesia, NM 88210
S.F. Heneghan
Engineer
Amoco Oil Co.
P.O. Box 182
Wood River, IL 62095
L.E. Hilgers
Lead Process Engineer
Sun Oil Co.—Tulsa Refinery
P.O. Box 2039
Tulsa, OK 74102
C.J. (Pete) Hoffman
Manager, Environmental Services
Champlin Petroleum Co.
P.O. Box 9365
Fort Worth, TX 76107
Harold Holt
Process Engineer
Great Canadian Oil Sands Ltd.
Box 4001
Ft. Me Murray, Alta ., T9H33E3T Canada
Jan F. Horwath
Environmental Coordinator
Bukeye Pipe Line Co.
P.O. Box 368
Emmanus, PA 18049
Darrell W. Hughes
Plant Supt.
OKC Refining Inc.
P.O. Box 918
Okmulgee, OK 74447
XIII
-------
E.E. "Gene" Humes
Design Engineer
Crest Engineering
P.O. Box 1859
Tulsa, OK 74101
M.K. (Don) Hut-ton
Mgr. , Mec & Environ. Eng.
Kerr-McGee Refining Co.
P.O. Box 25861
Oklahoma City, OK 73125
Raymond L. Jackson
General Manager
Environanental Design Eng. Co.
Box 1026
Findlay, OH 45840
John O . Johnson
Utilities Superintendent
Sun Oil Company
Tulsa, Refinery, P.O. Box 2389
Tulsa, OK 74102
Ted Johnson
Staff Assistant
Hunt Oil Co.
P.O. Box 1850
Tuscaloosa, AL 35401
Robert S. Jones
Sr. Environmental Technologist
Continental Oil Co.
P.O. Box 1267
Ponca City, OK 74601
Jack A. Kamps
Mgr. of Engineering
Toscopetro Corp.
P.O. Box 2860
Bakersfield, CA 93303
S.M. Khan
Mgr. Refinery and Utility Sales
The Ducon Co.
147 E Second Street
Minneola, NY 11501
XIV
James (Jim) P. Hutchinson
Regional Administrator
Champlin Petroleum Co.
P.O. Box 9365
Fort Worth, TX 76107
George F. Jackson
Crest Engineering, Inc.
P.O. Box 1859
Tulsa, OK 74135
Jere M. Jo hnson
Supervisor, Environmental Engineering
Exxon Co., U.S.A.
P.O. Box 3950
Bay town, TX 77520
K .R. Johnson
Process Engineer
Sun Oil-Tulsa Refinery
P.O. Box 2039
Tulsa, OK 74102
L.W. Jones
Research Asso .
Amoco Production Co.
Box 591
Tulsa, OK 74102
John R. Kampfhenkel
Lead Refinery Engineer
Suntide Refining Company
P.O. Box 2608
Corpus Christ!, TX 78403
W.G. Kelly
Manager of Environmental Operations
Atlantic Richfield Co.
Box 2679 --T.Z.
Los Angeles, CA 90051
R. Kilpert
Senior Engineering Associate
Exxon Research and Engineering Co.
P.O. Box 101
hlorham Park, N.J. 07932
-------
C.B. Kincannon
Texas Wafer Qualify Board
P.O. Box 1.3246, Capitol Station
Austin, TX 78711
Dr. A..T. Knecht
Supervisor
Atlantic Richfield Co.
400 E. Sibley Bldg.
Harvey, IL 60426
Charles E. Knipping
General Foreman, Wastewater Comtrol
Clark Oil & Refining Corp.
P.O. Box 7
Hartford, IL 62048
R.D. Kuerston
Chem. Eng.
Phillips Petro. Co.
1523 Hampden Rd.
Bartlesville, OK 74003
Gerald Lamb
Process Manager, Dravo Corp.
1777 Borel Place
San Mateo, CA 94404
Juanifa G. Lairmore
Environmental Control Technologist
Standard Oil of California
P.O. Box 1300
Pascagoula, MS 39567
Cyron T. Lawson
Project Scientist
Unbn Carbide
Tech Center
So. Charleston, WV 25303
Eddie Lee
Public Information Officer
U.S. EPA
Rt. 5, Box 411
Ada, OK 74320
Conrad W. Kleinholz
Zoology Graduate Student
Oklahoma State University
Life Science West 03
Stillwater, OK 74074
Eva Knettig
Chem. Engr.
Imperial Oil Enterprise, Ltd.
Samia, Canada
Ray Knutson
Environmental Eng.
Koch Refining Co.
Box 3596
St. Paul , MN 55165
Walter C. Lackemann
Chief Chemist
The Refinery Corp.
5800 Brighton Blvd.
Commerce City, CO 80022
Neale V. Lamb
Engineer
Phillips Petroleum Co.
Bartlesville, OK, 74003
Les Lash
Marketing Specialis!1
Envirofech
Eox 300
Salt Lake, City, UT 84110
Peter Bu Lederman
Director R & D
EPA
401 M St., S.W.
Washington, D.C.
Norman Lehnhardt
Tesoso Petroleum
P.O. Box 17536
San Antonio, TX 78286
xv
-------
Carl Lev!
Pac. Div. Mech. Eng.
Champlin Petroleum
P.O. Box 125
Wilmington, CA 90748
R.I. Loeffler
Chief Chemist
Rock Island Refining Corp.
500 W. 86th St.
Indianapolis, IN 46268
WM. K. Lorenz
Sr. Eng. Engineer
Sun Oil Co.
Box 1135
Marcus Hook , PA 19061
Lester M. McCright
Project Manager
Williams Bros. Waste Control
6600 South Yale
Tulsa, OK 74136
Wayne McLaury
Senior Staff Engineer
Crest Engineering, Inc.
P.O. Box 1859
Tulsa, OK 74101
Joe F- Molina
Cicil Eng. Dept.
University of Texas
ECJ 3.6
Austin, TX 78712
Kris D. Manchanda
Mkt. Res. Supervisor
Union Carbide Corp.
Old Saw Mill River Road
Tarryfown, NY 10591
Francis S. Manning
Professor/ ChE
University of Tulsa
600 S. College
Tulsa, OK 74104
Henry E. Lloyd, Jr.
Technology Services Engineer
Getty Oil Co. (Eastern Operations) Inc,,
Delaware Refinery
Delaware , DE 19706
Douglas G. Lofgren
Environmental Engineer
Phillips Petro.
10 Bl Phillips Bldg.
Bartlesville, OK 74003
W.C. McCarthy
Phillips Petro. Go.
Bartlesville, OK 74003
Michael D. McGee
Process Engineer
Continental Oil Co.
21 0 So.' 9 th
Ponca City, OK 74601
Patrick McMahon
Process Engineer
Brown & Root
Box 3
Houston, TX 77001
Russ C. Mallatt
Standard Oil
200 E. Randolph Drive
Chicago, IL 60601
Jim Mann
Burmah Oil & Gas
909 S. Meridan Suite 603
Ok City, OK 73128
Ralph E. Maple
Manager Refinery Services
Merichem Company
P.O. Box 61529
Houston, TX 77208
XVI
-------
Richard Marcum
Sanitary Engineer
Alaska Department of Environ. Conserve,
Pbuch O
Juneau, AK
Charles D. Martin
Senior Engineer
Crest Engineering
4310 S. Lakewood
Tulsa, OK 74135
F.J. Martin
Supt.
Union Oil of California
Box 237
Nederland, TX 77627
Anil Mathur
Student
2602 E. 10th Street, Apt. 2
Tulsa, OK 74104
Umesh Mathur
Environmental & 208 Project Engineer
Indian Nations Council of Governments
630 W. 7th Street
Tulsa, OK 74104
Krishna Merchant
Tertiarary Management
University of Tulsa, ChE Dept.
600 S. College
Tulsa, OK 74104
WM. R. Miles
Environmental Engineer
Charter Inter Oil Co.
Box 5008
Houston, TX 77012
Otis Miller
Process Engineer
CRA, Inc.
P.O. Box 570
Coffeyville, KS 67337
James A. Maricle
Air & Water Conservation Coordinator
Mobil Oil Corp.
P.O. Box 8
Ferndale, WA 98248
D.P. Martin
Director, Environmental Affairs
Gulf Oil Company, U.S.A.
P.O. Box 2100
Houston, TX 77001
J.L. Marzak
Vice President—General Manager
Metcalf & Eddy Industrial Div.
50 Staniford Street
Boston, MA 02114
Jimmy Mathur
Process Engineer
Brown & Root
4100 Clinton Drive
Houston, TX 77001
R.V. Mattern
Supt. Environment Cons.
Shell Oil/Chem.
P.O. Box 2633
Deer Park, TX 77586
John Migiliavacca
Process Engineer
Brown & Root
P.O. Box 3
Houston, TX 77001
Gilbert Millar
Manager-Ufilities
Brown & Root
4100 Clinton Drive
Houston, TX 77001
Richard B. Miller
Supt., Chemical Operations
Skelly Oil Company
P.O. Box 1121
El Dorado, KS 67042
XVII
-------
Robert Milligan
Pollution Cbntrol Technologist, Mgr.
Bechtel Corp.
P.O. Box 3965
San Francisco, CA 94119
Joe. G . Moore, Jr.
University of Texas, Dallas
P.O. Box 688
Richardson, Texas 75080
Gary W, Munson
Engineer
LaGloria Oil & Gas Co.
P.O. Box 840
Tyler, TX 75701
R.E. Myers
General Marketing Manager
Nalco Chemical Co.
2901 Butterfield Road
Oak Brook, IL 60521
Alee J. Nash
Environmental Engineer
Conoco
P.O. Box 37
Westlake, LA 70669
W.H. Nichols
Chief Project Engineer
Champlin Petroleum Co.
Box 552
Enid, OK 73701
William R. Oberman
District Manager
Envirotech Corp.
9235 Katy Road, Suite 202
Houston, TX 77024
W.P. Olivent
Senior Mechanical Engineer
GulfOil Co. U.S.A.
P.O. Box Drawer G
V,
'enice, LA 70091
XVII I
G .T. Minnick
Senior Environment Engineer
Sun Oil Co.
P. O. Box 2039
Tulsa OK 74102
Ronald F. Morgan
Advanced Scientist
Marathon Oil Co. -Denver Research Center
P.O. Box 269 (7400 S. Broadway)
Littleton, CO 80120
Leon H. Myers
Research Chemist
R.S. Kerr Environmental Research Labs, EPA
Ada, OK 74820
Ted Nairn, Jr.
Coordinator Air & Water Conservation
Cosden Oil & Chemical
Box 1311
Big Spring, TX 79720
John C. Nemeth
Manager-Environmental Sciences Dept,
Law Engineering Testing Co.
2749 Delk Road, SE
Marietta, GA 30062
Howard W. Nickle
Operating Technical Specialist
Sun Oil Co.
Box 426
Marcus Hook , PA 19061
G .F. Oldman
Effluents Adviser
British Petroleum
Britannic House, Moor Lane
London EC2, England
Ken Orr
Project Manager
Great Norfhern Electric Co.
2016 W. 43rd Street
Kansas City, KS 66103
-------
Leonard Orwin
Engineer
Metcalf & Eddy Engineering
Box 10046
Palo Alto, CA 94303
David Parrish
Research Assistant
Oklahoma State University
LSW 402
Stillwater, OK 74074
Howard Patterson
Chemist
Mapco Inc.
11391 E. Tecumseh
Tulsa, OK 74116
Fred Pfeffer
Research Chemist
R.S. Kerr Environmental Research Lab
EPA —P.O. Box 1198
Ada, OK 74820
B. Vail Prather
Staff Specialist
WM. Bros. Waste Control
6600 South Yale
Tulsa, OK 74136
Jean B. Reeder
Technical Editor
L.R. Reeder & Associates
5200 South Yale
Tulsa, OK 74135
George Reid
Professor
University of Oklahoma
202 West Boyd Street, Room 301
Norman, OK 73069
A. Kim Reyburn
Manager-Pollution Control
Crest Engineering, Inc.
P.O. Box 1859
Tulsa, OK 74101
XIX
R.E. Palmer
Senior Process Engineer
The Litwin Corp.
P.O. Box 282
Wichita, KS 67201
Don Patterson
Chief Chemist
Apco Oil Corp.
P.O. Box 857
Arkansas City, KS 67005
Wes Perkins
Coordinator Environmental Affairs
ARCO
Box 1346
Houston, TX 77001
M.L. Fbrter
Project Administrator
Phillips Petro. Co.
14 D 1 Phillips Bldg.
Bartlesville, OK 74004
Marvin A. Prosche, Mgr.
Refinery Technology
Atlantic Richfield
1801 E. Sepulveda BUM.
Carson, CA 90745
Pedro Regnault
University of Tulsa
819 S. Harvard
Tulsa, OK 74112
John C. Reidel
Sr. Process Engineer
Williams Bros. Process Services
6600 South Yale
Tulsa, OK 74136
William R. Reynolds
Mechanical Engineer
Navajo Refining Co.
P.O. Box 159
Artesia, NM 88210
-------
Mrs. Myria M. Rivera
Assoicate Chemist
Yabucoa Sun Oil
Box 186
Yabucoa, Puerto Rico 00767
Gary Roehr
Chemist
Indiana Farm Bureau Coop, Ass'n Inc.
1200 Refinery Road
Mt. Vernon, IN 47620
G.A. Rohlich
Civil Eng. Dept.
ECJ8.6.
University of Texas
Austin, TX 78712
Jim E. Russell
President
Aquaculture Industries, I nc.
2116 West 24 th Avenue
Still water, OK 74074
William D. Rutz
Mgr., Lab Sciences & Chemicals
Mapco
1437 South Boulder
Tulsa, OK 74114
Martha Sager
EPA
(WH 551) Crystal Mall, Bldg. 2
Arlington, VA 20460
J.K . Sargent
Manager Marketing
Dravo Corp.
One Oliver Plaza
Pittsburgh, PA 15222
H.F. Scarsdale
Environmental '.Coordinator
Skelly Oil Co.
P.O. Box 1650
Tulsa, OK 74102
xx
Joyce A. Rizzo
Sun Oil Company
Box 1135
Marcus Hook, PA 19061
Arthur Rogers
Mobil Oil Corp.
635 Elk Street
Buffalo, N.Y. 14210
W.L. Ruggles
Project Administrator
Phillips Petroleum
14 D 1 Phillips Bldg.
Bartlesville, OK 74004
Simon Russo
Student
University of Tulsa
801 North Gary Place
Tulsa, OK 74110
Edward R. Sager
Project Manager
Williams Bros. Waste Control
6600 South Yale
Tulsa, OK 74136
Max Samfield
Chemical Engineering
EPA
Room N-123, Mail Drop 62
Research Triangle Park, NC 27711
James R. Satcher
Assistant Manager Refineries
Southland Oil Co.
P.O. Box 128
Sandersville, MS 39477
Carl Schaffer (En 336)
Acting Director, Permits Division
EPA, Crystal Mall Bldg. 2
1921 Jefferson Davis Highway
Arlington, VA 20460
-------
Jacoby A. Scher
Environmental Engineer
Fluor Engineering and Construction
4620 North Brae wood
Houston, TX
K.F. Schmidt
Amoco Production Co.
Box 591
Tulsa, OK 74102
Edward C. Sebesta
Project Engineer
Brown & Root, Inc.
P.O. Box 3
Houston, TX 77001
George Sexton
Gooridnator-Environmental Controls
Beta Laboratories, Inc.
Somerton Road
Trevose, PA 19047
Robert Silvus
Chief, Industrial Branch
Texas Water Quality Board
P.O. Box 13246, Capitol Station 78711
Ausfin, TX 78711
John Single/
WBWC
1433 W. Loop S #634
Houston, TX 77027
Dave Skamenca
Sales Engineer
Eimco BSP-Suite 202
9234 Katy Freeway
Houston, TX 77024
John H. Smith, III
Sales Engineer
Jacobs
837 South Fair Oaks Avenue
Pasadena, CA 91105
Marc Schillinger
Environmental Engineer
Atlantic Richfield Co.
400 E. Sibley Blvd.
Harvey, IL 60426
J.W. Scrivner
Environmental Engineer
Amoco Oil Co.
P.O. Box 8507
Sugar Creek, MO 64054
Robert P. Selm
Wilson & Co.
Box 28
Salina, KS 67401
E.M. Sheets
Laboratory Section Supervisor
Sun Oil Co.
Box 2039
Tulsa, OK 74102
Gary W. Simms
Chemist
Marion Corp.
Rangeline Road
Mobile, AL
Suzette Sivyer
Environmental Technician
EPA
100 California St.
San Francisco, CA 94111
John Skinner
Chief Chemist
Kerr-McGee Refining Corp.
Wynnewood, OK 73098
Phil J. Smith
Process Engineer
Ford, Bacon & Davis Texas
P.O. Box 38209
Dallas, TX 75238
XXI
-------
S.R. Spears
Texaco Refinery
P.O. Box 2389
Tulsa, OK 74101
David O. Story
Director of Process Engineering
Gulf Oil Corp.
2935 Front Street
Toledo, OH 43697
E.J. Sullivan
Environmental Specialist
Amoco Oil Co.
P.O. Box 182
Wood River, IL 62095
W.L. Swander
Manager, Process Engineering
Sun Oil - Tulsa Refinery
P.O. Box 2039
Tulsa, OK 74102
Kenneth J. Tacchi
Chemist
Tretolite
369 Marshall Avenue
St. Louis, MO 63119
J. Anthony Tall is
Environmental Systems Consultant
NUS Corp.
71 1 Louisiana Street
Houston, TX 77022
Richard Thorstenberg
Engineering Assistant
Continental Oil Co.
P.O. &ox 1267
Ponca City, OK 74601
Jul io Ortiz Torres
Director-Technical Planning
Caribbean Gulf Refining Corp.
G.P.O. Box 1988
Sun Juan , Puerto Rico 00936
Larry W. Stinnett
Senior Process Engineer
Kerr- McGee Corp.
Box 305
Wynnewood, OK 73098
Ronald D. Stover
Assistant Refinery Superintendent
Indiana Farm Bureau Coop Ass'n., Inc.
1200 Ref inery Road
Mt. Vernon, IN 47620
John C. Suplicki
Application Engineer
Ecodyne - IWTD
2720 U.S. 22
Union, NJ 07083
Nicholas D. Sylvester
Professor of Chemical Engineering
University of Tulsa
600 South College
Tulsa, OK 74104
Wilson K. Talley
Assistant Administrator - R & D
EPA
40 M St. S,W.
Washington, D.C. 20460
Gerald Thornhill
EPA
1600 Patterson Street
Suite 1 100
Dallas, TX 75201
Lial F. Tischler
Manager, Austin Office
Engineering-Science, Inc.
3109 N. Interregional
Austin, TX 78722
John W. To we
Engineer
Sun Oil Co.
1600 Walnut
Philadelphia, PA 19103
XXII
-------
Paul C. Tranquil I
Environemental Engineer
The Standard Oil Co. (Ohio)
HSOSouthMetcalf St.
Lima, OH 45804
T.J. Tseng
Environmental Engineer
Ralph M. Parsons Co.
Pasadena, CS 91124
Michael Ray Twitchell
Process Engineer
Charter Internation Oil Co.
Box 5008
Houston, TX 77012
Herbert Uhlig
Manager Petroleum Technology
Stone & Webster Engineering Corp.
1 Pen Plaza
New York, N.Y. 10001
R.E.Van Ingen
Manager, Mfg. Environmental Conservation
Shell Oil Co.
P.O. Box 2463
Houston, TX 77024
W.A. Wadsack
Supervisor, Environmental Control
Cities Service Oil Go.
P.O. box 1562
Lake Charles, LA 70601
Joe D. Walk
Project Director
Standard Oil (Indiana)
200 East Randolph Dr.
Chicago, IL 60601
Gordon J. Wanless
Senior Technologist
BP North America, Inc.
620 Fifth Avenue
New York, N.Y. 10020
xxiii
D.V. Trew
Environmental Coordinator
Cities Service Co.
P.O. Box 300
Tulsa, OK 74102
W.N. Turtle
Process Engineer
Williams Bros. Waste Control
6600 South Yale
Tulsa, OK 74136
J. Paschal Twyman
President
University of Tulsa
600 South College
Tulsa, OK 74104
Russ Vandenberg
Manager- Industrial Sales
Resources Conservation Co.
P.O. Box 936
Renton, WA 98124
Gary R. Van Stone
Adsorption Systems Specialist
Calgon Corp.
4800 West 34th Street
Houston, TX 77043
J.N. Wakefield
Process Engineer
The Litwin Corp.
P.O. Box 282
Wichita, KS 67201
Henry F. Walter
US ERDA
Mail Stop-D-227
Washington, D.C. 20545
David Ward
Manager Engineering and Adminstration
Chemplex Co.
P.O. Box 819
Clinton, IA 53732
-------
William L. Warnement
Manager - Environmental Engineering
Crown Central Petroleum Corp.
P.O. Box 1759
Houston, TX 77001
L. Duvall Webster, Jr.
Professional Engineer
Turner, Mason, & Solomon
1950 Mercantile Dallas Bldg .
Dallas, TX 75201
Fred Weiss
Shell Development
Box 2099
Houston, TX 77001
R.B. Wells
Sales Supervisor
Amoco Chemicals
3815 Dacoma
Houston, TX 77018
Morris Wiley
Coordinator-Environmental Protection Dept.
Texaco, Inc.
P.O. Box 509
Beacon, N.Y. 12509
Edward J. Wolf
Senior Process Engineer
Gulf Oil Co. ~ U.S.
13539 E. foster Road
Santa Fe Springs, CA 90670
Marvin L. Wood
Deputy Director
R.S. Kerr Environmental Research Lab
Box 1198
Ada, OK 74820
Thomas Zale
Process Engineer
Sun Oil Co.
1608 Walnut Street
Philadelphia, PA 19103
James E. Watkins, Jr.
Utilities Chemist
Sun Oil Co.
Tulsa Refinery, 17 th & Union
Tulsa, OK 74102
Irving Wei
Senior Technical Consultant
Process Res. Div.
Environmental Research & Technology
Concord, MA 01742
R.A. Weldon
Research Engineer
Sun OilCo.
Box 1135
Marcus Hook . PA 19061
John White
Marketing Manager
Mapco
1437 South Boulder
Tulsa, OK 74103
Jim Wilkerson
Tank Farm Foreman
Marion Corp.
P.O. Box 526
Theodore, AL 36582
Joe F. Wood
Technical Advisor
Skelly Oil Co.
Box 1 650
Tulsa, OK 74102
Robert L. Wortman
Acting Director, Water Quality Service
OK State Dept. of Health
NE 10th & Stonewall
OK City, OK
Eugenio Zavala
Student
University of Tulsa
2423 East 5th Place, Apt. 33
Tulsa, OK 74104
XXIV
-------
M.A. Zeitoun
Senior Research Specialist
Dow Chemical
Freeport, TX 77541
XXV
-------
XXVI
-------
WELCOME ADDRESS
William C. Galegar, Director
Robert S. Kerr Environmental Research Laboratory
E.P.A., Ada, Oklahoma
Honored Guests, Ladies, and Gentlemen, It is a great pleasure to welcome
you to the first 'Open Forum on Management of Petroleum Refinery Wastewaters. "
My hope is that this will be only the first of a continuing series of technical
information exchanges between the industry, academia, governmental agencies at
all levels, and the interested public. This conference, as you will note, is a
series of sessions dealing with problems and solutions to protect the environment.
In each session the program representatives have attempted to offer a balanced
presentation of view points. The list of speakers is most impressive, and they are
all quite knowledgeable in their fields.
Success or failure of any conference, however, can be linked to two
action words-"participation" and "communication" by attendees. While not being
an active part of any of the committees responsible for the symposium, I was
close enough to experience the enthusiasm of your respresentatives from the
American Petroleum Institute, the National Petroleum Refiners Association, the
University of Tulsa, and the Environmental Protection Agency.
Last evening, upon entering the hotel, the undercurrent of expectation was
very evident. Meeting old and new friends convinced me that you came to
"participate" and "communicate". You have already, by your past and present actions,
indicated the true welcome of this meeting; I have the priviledge of expressing that
feeling. On behalf of the organization sponsoring this symposium, the working
committees who planned it, the speakers for the various sessions, and our host,
Tulsa University, I bid you welcome. May the next three days be as helpful to
you as I expect them to be to me.
BIOGRAPHY
William C. Galegar holds a BS in Chemical
Engineering from Oklahoma State University and an
MS in Chemical Engineering from the University of
Oklahoma. His career has included 1 1 years with
Oklahoma Department of Health and 16 years
with the Environmental Protection Agency and its
predecessor agencies. Mr. Galegar is currently
Director of the EPA's Robert S. Kerr Environmental
Research Laboratory in Ada, Oklahoma •
-------
-------
KEYNOTE ADDRESS
"WASTEWATER CONTROL — WHERE ARE WE, WHERE ARE WE GOING, AND WHY?
Milton R. Beychok
Consulting Engineer, Irvine, California
I am sure that all of us must be impressed by the broad and diverse range of view-
points represented by the speakers assembled for this conference. Very briefly, we have:
— The EPA Divisional Directors for Effluent Guidelines and for Permits
— The Chairman of the EPA's Effluent Standards and Water Quality Information
Advisory Committee
— The Program Director of the National Commission on Water Quality
— A speaker from the Texas Water Quality Board
— The EPA's Assistant Administrator for Research and Development
— Speakers from a number of the major oil refining companies
— Speakers from several well-known engineering and consulting firms
— Eminent university professors
— Key administrators from the EPA's Robert S. Kerr Research Laboratory
Aside from being impressive, these diverse viewpoints give us a unique opportunity to
share perspectives and to explore important policy and technological questions related to
industrial wastewater control. Toward that end, I would like to explore a few key issues
in this 'keynote1 address and one of those is the problem of the communication of
viewpoints between industry and the federal EPA.
EFFECTIVE COMMUNICATION REQUIRES A TWO-WAY STREET
Effective communication works best on a two-way basis. That may be a cliche, but
it is nevertheless true. Mr. Joe Moore of the NCWQ recently stated that (1), "We must
not only listen, but we must also hear what is said." And I would add that we can accom-
plish much more by listening to each other than by shouting at each other.
In the three years since enactment of PL 92-500 (the Federal Water Pollution Control
Act Amendments of 1972), at least 634 lawsuits have been instituted involving implemen-
tation of the Act (2). I submit that there would be far less court litigation between in-
dustry and the EPA if industry could effectively participate in developing effluent and
emission control regulations and if industry viewpoints were truly given serious con-
sideration. I fully realize that some industry viewpoints are non-productive and perhaps
overly opposed to meaningful regulations. But the large majority of industry is willing to
accept equitable regulations based upon realistic data and commercially-proven techno-
logy. The EPA should solicit industry viewpoints very early in their process of developing
regulations, and provide industry a real opportunity to participate in their decision-
making. It is not enough to merely invite formal written comments within a legalistic
30-60 days after publication of a criteria document or proposed promulgations espe-
cially not when those criteria or proposed promulgations represent months or years of work
and many irreversible decision steps already taken by the EPA and their outside contrac-
tors. What industry wants is early involvement, face-to-face dialogues and a meaningful
opportunity to help formulate proposed regulations through all the stages of decision-
making. Formal comments made after the fact, or during public hearings conducted under
-------
the rigid rules of law, are poor substitutes for a meeting of minds around a conference table.
Formal comments and rigid public hearings simply invite adversary debate merely to esta-
blish a legal record what we need instead is the chance to voice, to hear and discuss
valid viewpoints. I submit again that a working partnership between industry and the EPA
would drastically reduce the court litigation involving PL 92-500. Obviously, the same
participatory partnership between the EPA and the State regulatory agencies would be
equally helpful.
It is probably too late to institute any working partnership of industry and the EPA
involving the 1977 and 1983 effluent limitations mandated by PL 92-500. But it is not too
late to learn some lessons from the past three years and apply them to the tasks and
decisions that lie ahead of us. For example, what of PL 92-500's objective regarding the
elimination of the discharge of pollutants (EOD) by 1985 the so-called 'zero discharge1
goal? And what of the yet to be issued toxic pollutant standards called for in Section 307
of PL 92-500? And what of the recent decisions (3) by the Third and Seventh Circuit
Courts of Appeal which made a clear-cut distinction between Section 301 effluent
limitations as being 'minimum' levels of control and Section 304 guidelines as being "factors
or ranges' to be used by the permit issuers in deciding whether or by how much any indivi-
dual permit is to be more stringent than the 'minimum' level Section 301 limitations? Do
these Circuit Court decisions lend more validity to the matrix approach for Section 304
guidelines as advocated by the EPA's ES&WQIAC committee? Certainly, these are ques-
tions which must be of urgent concern to all of industry and there is still an opportunity
to tackle these questions by a broad-based working partnership between industry, the EPA,
and the State regulatory agencies.
WHERE ARE WE NOW?
I think most of us would agree that the tasks to be accomplished in order to achieve
the mandated 1977 and 1983 goals of PL 92-500 are now defined in terms of the EPA's 1977
and 1983 Effluent Limitations and their New Source Performance Standards. Obviously,
resolution of the recent Circuit Court decisions and of the pending backlog of litigation
may very probably add new dimensions to the 1977 and 1983 regulations. But at least we
have a framework upon which to assess where we are now, what we may accomplish by 1983,
and what it may cost.
We are indeed fortunate at this conference that the National Commission on Water
Quality has very recently released its Staff Draft Report. That Commission was created by
Section 315 of PL 92-500 and charged with making a complete study of:
"... all aspects of the total economic, social and environmental effects of
achieving or not achieving the goals set forth for 1983. "
The NCWQ staff draft summarizes PL 92-500 in terms of the Act's primary objective, two
key goals and a number of specific policy statements. The primary objective of the Act is:
"to restore and maintain the chemical, physical and biological Integrity of
the Nation's waters."
-------
The two key goals of the Act are given by the NCWQ as: (4)
— To reach, "wherever attainable," a water quality that "provides for the protection
and propagation of fish, shellfish, and wildlife" and "for recreation in or on the
water" by July 1983.
— To eliminate the discharge of pollutants (EOD) into navigable waters by 1985.
Focusing on the 1983 goal, the NCWQ staff draft states that "much of the nation's waters
are already of sufficient quality to meet the interim (1983) goal" as the result of pollution
control efforts prior to PL 92-500 or as a natural condition (5). In order to be more de-
finitive, the Commission selected certain minimum water quality criteria that would con-
stitute meeting the 1983 goal one of which was a dissolved oxygen level of at least
4 ppm during 'worst case1 conditions of seasonal low flow (5). Based on an assessment of
41 sites, the NCWQ developed the data in Table 1. These data indicate that 7-8% of
the nation's waterways do not presently meet the 'worst case1 dissolved oxygen criteria
for satisfying the 1983 goal. Thus, in effect, we might define 'where we are now' by
saying that 92-93% of our waters already satisfy a key criteria for meeting the 1983 goal
of PL 92-500.
WHERE ARE WE GOING NEXT AND AT WHAT COST?
Clearly, where we are going next is to achieve the 1977 and 1983 goals. Table 1
tells us that achieving those goals will upgrade the 7-8% of our waters which do not pre-
sently meet those goals. But at what cost?
Table 2, developed from data in the NCWQ staff draft, sheds some light on the mag-
nitude of the costs involved. The total capital expenditure required by industry and by
municipally owned treatment systems will be about 233 billion dollars (1975 basis), and
the annual operating costs will be about 16 billion dollars. I must stress that these figures
do not include interest charges for borrowing the capital to finance this vast program, and
that these costs will be spread over a period of at least 10 years. On the basis of about
65 million families in the nation, these costs amount to $3,600 of capital expenditure
plus 245 $/year for each family or about 600 $/year per family for at least the next
10 years.
However, the costs will in fact be paid by each of us in the form of taxes and price
increases for manufactured goods. Since tax burdens and buying power are graduated in
terms of income level, Table 2 estimates that most of us in the middle and upper middle
income brackets might expect to finance the 1983 goals by paying from 50-100 $/month
for at least the next 10 years. But remember that these figures do not include interest
charges of inflation. A simple 7% per annum interest charge for 10 years could easily
increase that monthly payment range to 70-140 $/month per family.
The 1983 benefit will be an upgrading of 7-8% of our nation's waters. The cost
might be 70-140 $/month per family for 10 years. Each of us will probably assess that
balance from a different perspective in deciding whether it is an equitable balance.
-------
WHY GO BEYOND 1983?
As mentioned earlier, the second goal of PL 92-500 is the elimination of the discharge
of pollutants (EOD) by 1985. For simplicity, I will refer to that goal as EOD in my further
remarks. What do we mean by EOD? The NCWQ staff draft uses the following definition:
"The elimination of the discharge of pollutants shall apply to the
removal of those constituents which are added during use of the water.
The resultant discharge must be of equal or lower concentration than
that of the original supply."
I cannot accept that as a satisfactory definition, since it does not define what con-
stitutes a pollutant. Do pollutants include simple inorganic salts such as sodium and magne-
sium chlorides or calcium and magnesium carbonates and sulfates? If so, on what basis
or criteria have these simple salts been defined as being 'pollutants'?
Nor can I accept the fact that the second part of the definition is not consistent with
the first part. If we simply boil some water to produce steam, the resultant boiler blow-
down will have a higher inorganic salt concentration than the original supply despite the
fact that no salts were added during use of the water. Or if we simply evaporate some
water in a closed loop cooling tower, the resultant cooling tower blowdown will have a
higher salt content than the original supply again with no salts being added during use.
Are these salts truly undesirable or toxic pollutants?
I am reminded of the following philosophical parable:
First philosopher: "Why did the ancients consider the world to be flat
rather than round?"
Second philosopher: "I suppose simply because it looked flat."
First philosopher: "Then how would it have looked to them if it had
looked round?"
The point of my parable is that we had better take another look at this definition of EOD.
Dissolved inorganic salts may look like pollutants to some people, but they don't look like
pollutants to me. I would most certainly agree with defining pollutants to include phenols,
cyanides, BOD, heavy metals and so forth but I cannot agree that simple inorganic
salts are pollutants.
If we achieve the 1983 goal, presumably all of our nation's waters, "wherever attain-
able" will have a quality that "provides for the protection and propagation of fish, shell-
fish and wildlife . .. and for recreation in or on the water." Then why go beyond that?
What do we gain from EOD? Table 3 presents the answer to that question as given in the
NCWQ report and the answer would appear to be that the benefits of EOD will be very
minimal. The NCWQ report is very explicit on this point (7):
"the predictable environmental effects (of EOD) would appear to be
minimal, particularly in the absence of adequate control of nonpoint
sources (urban and agricultural runoffs)."
-------
On the question of costs for achieving EOD, the NCWQ report does not quantify
total costs but only estimates unit costs in terms of $/1000 gallons for various innovative
technologies that might possibly be used. However, the report does say that the costs
would be (7):
"prohibitively expensive, and the economic and social effect would be too
severe to be absorbed within the foreseeable future."
The NCWQ report also points out that EOD technologies are considerably more energy
intensive than those required to achieve the 1983 goals (8). Although energy require-
ments for the innovative EOD technologies are estimated in terms of KWH/1000 gallons,
the report does not quantify the total energy requirements. However, I make my own esti-
mate (in Tables 4, 5, and 6) based largely on the data within the NCWQ report. As
shown on Table 4, I have estimated the total energy requirements to achieve EOD in terms
of barrels per day of fuel oil and then expressed the fuel oil requirement as a percent-
age of our current oil imports of some 7 million barrels per day. If we assume a post-1983
industrial effluent flow of only 10% of the current nationwide industrial water intake, and
if we assume we must produce at least a highly concentrated slurry for disposal by landfill
or by land spreading, then the energy requirements to achieve EOD in the industrial sec-
tor will range from about 5-20% of our current oil imports.
Based on the NCWQ's projection of post-1983 municipal effluent flows, Table 5 pre-
sents an estimate of the EOD energy requirements for the municipal sector. Assuming that
half of the municipal effluent can be disposed of as a 2 wt% brine (via ocean disposal,
solar evaporation ponds, injection wells, etc.), and assuming that the other half must be
concentrated to a slurry for land disposal, then the municipal sector energy requirements
to achieve EOD will amount to about 10% of our current oil imports.
On Table 6, the total industrial and municipal energy requirements to achieve the
1977, 1983 and EOD goals are presented. The total will range from 1,430,000 to 2,660,
000 barrels per day of oil, or about 20-38% of our current oil imports. The EOD require-
ments relative to the 1977 and 1983 requirements are very high. One must certainly agree
with the NCWQ's statement that the EOD technologies are considerably more energy inten-
sive than those required for the 1983 goals.
On the basis of costs versus benefits, the case against EOD appears overwhelming:
— The environmental benefits would be very minimal.
— The economic and social costs would be prohibitive and too severe to absorb in the
foreseeable future.
— The energy consumption would defeat any hopes we have of significantly reducing
our dependence on imported oil.
Finally, in my opinion, the removal of simple water-soluble inorganic salts from our
industrial and municipal effluents is a complete absurdity. As an intellectual engineering
exercise, we very probably could remove those salts. But what a waste of the nation's
wealth and genius! And after the salts are removed as a brine ... or a slurry ... or eva-
porated to dryness, how then do we dispose of them? In fact, how do we prevent their
ultimate return to Mother Nature by dissolution in rainfall or by leaching into underground
-------
aquifers? The NCWQ draft would seem to agree with me on this point:
"The dominant consideration for achieving EOD relates to the removal
of soluble salts. Removal of these salts creates large quantities of brine
which must then be disposed of ... at very high costs. "(8)
THE TOXIC POLLUTANT STANDARDS
"Toxic pollutant standards: paralysis or progress?" that was the title of a recent
article in the Journal of the Water Pollution Control Federation (9). The purpose of the
article was to analyze the delays in developing the toxic pollutant standards mandated by
Section 307 of PL 92-500. Let us examine the chronological history of EPA's actions to
date on these standards:
Published a list of 9 toxic pollutants July 1973
Revised their criteria for selecting toxic pollutants,
re-issued original list of 9 pollutants plus 25 other potential
pollutants Sept 1973
Proposed effluent standards for the 9 toxic pollutants Dec 1973
Held public hearings on proposed standards Apr-May 1974
Officially withdrew the proposed standards, and
released a "Draft Advance Notice of Proposed Rule-
making for Toxic Pollutant Standards" in which new
proposed standards were presented on an informa-
tional basis Nov 1974
In the 14 months since releasing that "Draft Advance Notice ...," no further official pro-
posals have been issued. According to the schedule mandated in PL 92-500, the toxic
pollutant list was to be published in Jan. 1973, the final standards were to be promul-
gated by Jan. 1974, and source compliance was to be achieved by Jan. 1975 some
12 months ago.
Much of the delay can be attributed to two factors: (9)
— The lack of adequate scientific information on the effect of pollutants at extremely
low concentrations in water.
-- The counterproductive provisions in PL 92-500 which required virtually impossible
time schedules that were subsequently forced upon the EPA by court orders, and
which required trial-type adjudicatory public hearings.
I can readily sympathize with the EPA on these points. But whatever the past rea-
sons for delay, I would urge again that it is not too late to work with industry and with the
State regulatory agencies in developing a practical set of toxic pollutant standards upon
a firm base of scientific information.
EFFLUENT LIMITATIONS VS. EFFLUENT GUIDELINES
As I said earlier, recent court decisions (3) have made a clear-cut distinction between
PL 92-500's Section 301 effluent limitations and Section 304 effluent guidelines. In the
case of the American Iron and Steel Institute versus the EPA (10), the Third Circuit Court
-------
of Appeals ruled:
"Section 304 requires the EPA to develop effluent guidelines to assist per-
mit issuers in setting individual point source discharge limits. Section 304
is not satisfied unless EPA specifies 'factors' and 'permissible ranges' of
effluent limitations to be taken into account by the permit issuers in setting
individual permit limits. These permissible ranges must be within the
nationwide parameters established by Section 301 effluent limitations."
A few weeks after that ruling, the Seventh Circuit Court of Appeals in ruling upon a
similar case issued a decision (10) which agreed with the Third Circuit's ruling on this
point.
It would appear that these rulings (unless reversed by further appeal) mean that all of
the EPA's effluent limitations promulgated to date are Section 301 nationwide parameter
limitations and now the EPA must establish Section 304 effluent guideline 'ranges' and
'factors' within those nationwide limits. I hope that the Director of the EPA's Effluent
Guidelines Division (who is our next speaker) will address this point because it should
be of vital concern to industry.
It would also appear that these rulings should lead to a re-consideration by the EPA
of the matrix approach for Section 304 guidelines which has long been advocated by the
EPA's ES&WQIAC advisory committee. Mr. Leon Myers of the EPA's Robert S. Kerr
Research Laboratory has been quoted as saying (1 1) that his group plans to study the appli-
cation of the ES&WQIAC matrix approach to the petroleum refining industry. I hope that
we may hear more about these plans from Mr. Myers when he speaks to us later in this
conference.
Personally, I have very little knowledge of the ES&WQIAC matrix approach, and I
have no opinions concerning the merits of that approach. But it was certainly disappoint-
ing that the work of the ES&WQIAC was not discussed or evaluated in depth anywhere with-
in the 832 pages of the NCWQ staff draft report. That is an oversight which should be
corrected in the NCWQ's final report.
THE PETROLEUM REFINING EFFLUENT LIMITATIONS
Leaving the area of broad issues, I would now like to focus upon a specific technical
point which concerns the petroleum refining effluent limitations as applied to treatment
plant design.
First, it would be useful to review how the refining limitations were developed. In
1972, the API and the EPA conducted an extensive wastewater survey of 167 refineries.
150 of those refineries submitted comprehensive data, and 136 were considered acceptable
after screening. Upon eliminating those which used once-through cooling water for more
than 3% of their process cooling, 94 were left in the final data base. That data base is
probably the most realistic information ever collected for the wastewaters from the petro-
leum refining industry. However, the data were based on one-time samples and did not
-------
represent attainable long-term performance. The EPA analyzed ^the data from the 94 re-
fineries by statistical regression techniques to determine how refinery classification, size,
and processing configuration affected the raw waste load. The regression analysis led to
the five refinery subcategories, and the size and process factors, promulgated in the efflu-
ent limitations.
The next step by the EPA was to determine the attainable levels of long-term (annual)
pollutant concentrations in the treated wastewater discharge from each of the five refinery
subcategories, as well as the attainable long-term average wastewater flows. The data
for these determinations were obtained by visiting refineries, reviewing monitoring records,
and from various other sources.
Finally, data were gathered on the day-to-day and month-to-month variations in
ref'nery wastewaters discharges. These data were analyzed by statistical techniques to
arrive at variability factors which related long-term averages to short-term values. Thus,
the attainable annual averages were multiplied by the applicable variability factors to arrive
at the daily or 30-day averages in the final promulgated effluent limitations.
Unfortunately, the attainable long-term averages (flow and concentrations) and the
variability factors are not included in the actual published EPA regulations (40 CFR 419).
Due to the various amendments and revisions that have taken place, the unclear presentation
in the original Development Document, and to the fact that the preambles to the various
Federal Register promulgations are not included in the final Code of Federal Regulations —
it is virtually impossible at this point in time to determine the applicable long-term averages
and variability factors with any degree of certainty. The waste treatment plant designer is
left only with the amended short-term effluent limitations as published in the Code of
Federal Regulations.
However, it is most important that the waste treatment plant be designed to achieve
the more stringent long-term levels if the plant is not to exceed the allowable peak levels
for daily and 30-day averages. For that reason, I have attempted to reconstruct (as best
I could) the applicable variability factors and equivalent annual averages. Tables 7 and 8
present some of those reconstructions for:
1977 limitations for existing sources (30-day averages)
NSPS limitations (30-day averages)
These reconstructions reveal a number of anomalies. These anomalies may be due to my
inability to correctly reconstruct the currently applicable information, or they may be due
to inconsistencies which the EPA should correct. For example, in Table 7 for the 1977
existing source limitations:
—The annual average values were obtained from dividing the promulgated 30-day
averages by the annual average wastewater flow and by the variability factor, and
applying a conversion factor to arrive at ppm.
—We can see a number of trivial anomalies in values, most probably due to incon-
sistent round-off procedures.
--But in the COD values, the anomalies amount to as much as 20% between the
extremes. I am fairly sure that the four COD values, shown enclosed in a box,
were not meant to exhibit a 20% difference.
-------
11
—Looking at the four NhL-N values, shown enclosed in a box, I am at a loss as to
why the more complex refinery subcategories (Lube Oil and Integrated) are expected
to achieve 38% lower NHL-N values than the less complex subcategories.
Table 8 illustrates the same anomalies for the New Source Performance Standards, and here
the four NHL-N values vary by a factor of 2-to-l.
Since the waste treatment plant design should be based on attaining the annual averages,
I would strongly urge the EPA to:
—Publish a clear and explicit listing of the latest applicable values for long-term
flows and concentrations in the format of Tables 7 and 8 — for daily averages as
well as 30-day averages, and for the 1983 limitations as well as 1977 and the NSPS.
—Publish a clear explanation of the COD and NhL-N anomalies that appear to exist.
SUMMARY
In closing, I wish to express my deep appreciation for the opportunity to address this
conference and for your patience with my lengthy address. I can only hope that it will
stimulate some open discussion of the issues we face, and I look forward to listening to and
hearing what the other speakers have to say.
REFERENCES
(1) Environmental Industry Conference, Council on Environmental Quality, Washington,
D. C., December 10, 1975
(2) Staff Draft Report, National Commission on Water Quality, November 1975, page V-6
(3) American Iron and Steel Institute vs. EPA, Third Circuit Court of Appeals, November
1975
American Meat Institute vs. EPA, Seventh Circuit Court of Appeals, December 1975
(4) Staff Draft Report, National Commission on Water Quality, November 1975, pages
1-1 and 1-2
(5) Ibid, pages 1-23 and 1-24
(6) Ibid, page IV-10
(7) Ibid, page 1-74
(8) Ibid, page 76
(9) "Toxic Pollutant Standards: Paralysis or Progress?", JWPCF, December 1975
(10) 8 ERC 1321, Environment Reporter, Decisions No. 10, November 28, 1975
8 ERC 1369, Environment Reporter, Decisions No. 11, December 12, 1975
(11) Environment Reporter, Current Developments, page 1252, November 14, 1975
-------
12
DISCUSSION
Umesh Mathur: I found your comments extremely interesting.
work for the Indian Nations
, ^TGovernments'and we are a designated Section 208 planning agency. All of our
guidelines (for Section 208 long-term regional planning) tell us to assume that industry will
meet the EPA's effluent limitation standards. We can then evaluate the long-term effect
of industrial discharges within our region on that basis.
However, if our evaluations on that basis indicate that long-term water quality goals
will not be attained, what are we to do then? Can industrial NPDES permits call for more
stringent limitations because Section 208 long-range planning indicates that more stringent
controls may be needed?
Is there any real purpose in long-term planning based on receiving water quality when
NPDES permits are issued on technology-based effluent standards?
Milton Beychok: I am not sure that I fully understand your questions. Each individual state
already has the option of setting NPDES effluent limits which are more stringent than the
established EPA guidelines. And the court decisions (which I discussed in regard to effluent
limitations versus guidelines) seem to mandate a range of values to assist the permit issuers
in deciding how stringent to make the individual permit. But I don't know how a Section
208 agency interfaces with the state NPDES permit issuers.
As for technology-based standards versus long-term receiving water quality, I guess
that Congress crossed that bridge when they wrote PL 92-500 and decided upon technology-
based standards.
Umesh Mathur: Perhaps we can have the EPA speakers address this point later on, because
I think it is of great importance.
BIOGRAPHY
Milton R. Beychok is a consulting Chemical
Engineer in the field of environmental technology.
He has a B.S. degree in Chemical Engineering
from Texas A & M University, and he is a regis-
tered professional engineer in California and Texas.
He is a Diplomate of the American Academy of
Environmental Engineers and is a member of the
AlChE, Air Pollution Control Association and the
Water Pollution Control Federation. He has
served on the California Water Quality Control
Board and consulted for the EPA, the National
Science Foundation and the National Commission
on Water Quality. Prior to entering private
practice, he was with Fluor Engineers &
Constructors for 20 years.
-------
TABLE I "WHAT DO WE ACHIEVE BY 1983 COALS?'
13
0.
Q.
Ill
O
0
Ul
>
o
VI
(4 PPM DISSOLVED OXYGEN)
MINIMUM GOAL DURING THE
LOW FLOW SEASON
7-8% OF WATERWAYS DO NOT PRESENTLY
MEET THE 1983 MINIMUM GOAL
I
10%
20%
30%
40%
50%
% OF WATERWAYS WITH DISSOLVED OXYGEN EQUAL TO OR LESS THAN LEVEL SHOWN
Reference: Page 1-24, NCWQ Staff Draft Report, November 1975
Involving 9, 000 river and stream miles, 10, 500 miles of coastline, 6,000 square miles of estuaries, and 6,000
square miles of lakes and reservoirs. The total surface of the watersheds where each of the sites is located
encompasses 20% of the nation's land surface and a population of 80 million people (6) .
TABLE 2 "WHAT WILL 1983 GOALS COST?"
CAPITAL
COST ($)
ALL INDUSTRY (1)
MUNICIPAL SYSTEMS (2)
TOTAL
75,000,000,000
158,000,000,000
233,000,000,000
ANNUAL
OPERATING COSTS ($)
13,900,000,000
2,000,000,000
15,900,000,000
ESTIMATED MONTHLY COST PER FAMILY FOR NEXT 10 YEARS (3)
(IN TAXES AND PRICE INCREASES)
INCOME LEVEL:
29,000 - 37,500 $AEAR
20,700- 29,000 $AEAR
17,700- 20,700 $AEAR
13,100- 17,700 $/YEAR
MONTHLY COST
$80 - 100/MONTH
$50 - 65/MONTH
$40- 50/MONTH
$25- 30/MONTH
(1) Page II - 81, NCWQ Staff Draft (2) Page II - 28, Ibid
(INTEREST COSTS AND
INFLATION ARE NOT
INCLUDED )
(3) Page II I- 273, Ibid
-------
14
BOD REDUCTION - -
TABLE 3 "BEYOND 1983 . . . WHAT DO WE GAIN FROM EOD?"
(By achieving EOD for point sources
after achieving 1977 and 1983 goals)
INSIGNIFICANT COMPARED TO UNCONTROLLED NON-POINT
URBAN RUNOFF AND AGRICULTURAL RUNOFF
TOXIC SUBSTANCES REDUCTION - -
EUTROPHICATION REDUCTION
SUSPENDED SOLIDS REDUCTION
OVERALL BIOLOGICAL IMPACT -
MINOR COMPARED TO UNCONTROLLED URGAN AND
AGRICULTURAL RUNOFFS
VERY LITTLE SINCE NON-POINT SOURCES CURRENTLY ACCOUNT
FOR 80% OF TOTAL NITROGEN AND 50% OF TOTAL PHOSPHORUS
ENTERING OUR WATERWAYS
VERY LITTLE SINCE 90% OF THE LOAD ENTERING OUR WATERWAYS
IS FROM NON-POINT SOURCES
ONLY 6-10% OF WATERS PRESENTLY UNSUITABLE WILL BE UPGRADED.
THUS, 5-10% of 7-8% OR PERHAPS ABOUT 0.4 - 0.8% OF NATION'S
WATERWAYS WOULD BE UPGRADED.
Reference: NCWQ Staff Draft Report ( Nov. 1975) Pages 1-77 and 1-78
TABLE 4 "HOW MUCH ENERGY TO ACHIEVE EOD FOR INDUSTRY?"
GROSS
USE
NET INTAKE (1)
15 x 10 gals/yr
ALL
INDUSTRY
EVAPORATION,
CONSUMPTION,
' LOSSES
EFFLUENT
ca. 10%/of INTAKE
EVAPORATION
TO A SLURRY
50-200
KWH/1000 gals
ENERGY CONSUMED TO ACHIEVE EOD
DISPOSAL
OF SLURRY
EFFLUENT (AFTER 1983)
% OF/INTAKE (1) GALSAR KWH/1000 GALS (2) KWH/YR EQUIV./BPD OIL (3) % OF CURRENT/IMPORTS (4)
10
10
1.5x 10'
1.5 x 10
200
50
3x 10
11
0.75x 10
11
1,371,000
342,750
19.6
4.9
(1) CURRENT INDUSTRIAL WATER NET INTAKE = ca. 15 x 10 GALS/YR (p. 11-53, NCWQ Staff Draft).
EFFLUENT (AFTER 1983) ASSUMED TO BE 10% OF INTAKE.
(2) 50-200 KWH/1000 GALS (60-240 BTU/POUND) TO ACHIEVE A CONCENTRATED SLURRY (p. 11-144, Ibid)
(3) EQUIVALENT BPD OF OIL = (KWHAR) ( 10,000 BTU/KWH)/(365 DAYS/YR) (6 x 106 BTU/BBL OIL)
= (KWHAR) (4.57x 10" )
BPD = BARRELS OF OIL PER DAY, WITH HEATING VALUE OF 6 x 106 BTU PER BARREL
10,000 BTU/KWH = FUEL CONSUMED IN GENERATING ELECTRIC POWER AT ca. 33% EFFICIENCY
(4) CURRENT OIL IMPORTS TAKEN TO BE ca. 7 MILLION BARRELS PER DAY
-------
15
TABLES "HOW MUCH ENERGY TO ACHIEVE EOD FOR MUNICIPAL EFFLUENTS?1
OCEAN DISPOSAL,
SOLAR EVAPORATION,
50% INJECTION WELLS, ETC.
EFFLUENT FLOW 5(Mo-1 0.7x 1Q9 GALS/DAY (1)
(AFTER 1983 GOALS) CONCENTRA CONTAINING 2 WT%
9 TION DISSOLVED SOLIDS _
35x10 GALS/DAY EVAPORATION LANDFILL
CONTAINING 0.04 JŁg _ TO A SLURRY _
WT % DISSOLVED _
10 KWH/1°°° 50-200 KWH/1000 GALS(2)
CONCENTRATION = (35 x 109 GALS/DAY)(365 DAYSAR)(10 KWH/1000 GALS)(4.57x 10"6)
= 584,000 BPD OF OIL
EVAPORATION = (0.35 x 109 GALS/DA Y)(365 DAYSAR)(50 to 200 KWH/1000 GALS)(4.57 x 10"6)
= 29,000 to 117,000 BPD OF OIL
TOTAL ENERGY REQUIRED = 613,000 to 700,000 BPD OF OIL
8.8 % to 10.0 % OF CURRENT IMPORTS
(1) Page 11-155, NCWQ Staff Draft
(2) Page 11-144, IBID
TABLE 6 "HOW MUCH ENERGY (NATIONWIDE) TO ACHIEVE 1977, 1983 AND EOD GOALS?"
% OF OUR
10 BPD OF OIL CURRENT IMPORTS
Manufacturing Industry:
1977 goal (1) 0.17 2.4
1983 goal (2) 0.20 2.9
Steam-Electric Industry:
1983 thermal limitations (3) 0.02-0.13 0.3-1.9
Municipal Effluents:
Treatment, collection, sewer control (4) 0.09 1.3
TOTAL TO ACHIEVE 1983 GOALS 0.48-0.59 6.9-8.5
EOD for industry (5) 0.34-1.37 4.9-19.6
EOD for municipal effluents (5) 0.61 0.70 8.8- 10.0
TOTAL BEYOND 1983 (w/EOD) 1.43 - 2.66 20.6 - 38.1
(1) 376 trillion Btu/year (p. 11-100, NCWQ Staff Draft)
(2) 446 trillion Btu/year (p. 11-100, Ibid)
(3) 45 to 294 trillion Btu/year (p. 11-106, Ibid)
(4) Page 11-41, Ibid
(5) See Tables 4 and 5
-------
16
TABLE 7 "1977 EFFLUENT GUIDELINES FOR PETROLEUM REFINING (1)"
EFFLUENT LIMITATIONS IN POUNDS/1000 BBLS OF CRUDE OIL (2,3)
SUBCATEGORY
TOPPING
CRACKING
PETROCHEMICAL
LUBE OIL
INTEGRATED
BOD-5
4.25
5.5
6.5
9.1
10.2
TSS
3.6
4.4
5.25
8.0
8.4
COD
21.3
38.4
38.4
66
70
OIL
1.3
1.6
2.1
3.0
3.2
EQUIVALENT ANNUAL
TOPPING
CRACKING
PETROCHEMICAL
LUBE OIL
INTEGRATED
VF(4)
TABLE
15.0
15.5
15.3
14.3
15.0
1.7
8 "NEW
10.2
10.1
10.0
10.2
10.2
2.1
79.9
115.2
96.0
110.0
109.4
1.6
4.9
4.8
5.3
5.0
5.0
1.6
PHENOL
0.027
0.036
0.0425
0.065
0.068
AVERAGE p
0.10
0.10
0.10
0.10
0.10
1.7
NH3-N SULFIDE
0.45
3.0
3.8
3.8
3.8
.p.m.
1.8
9.6
10.1
6.8
6.3
1.5
0.024
0.029
0.035
0.053
0.056
Cr
0.071
0.088
0.107
0.160
0.17
CONCENTRATIONS
0.10
0.10
0.10
0.10
0.10
1.4
0.25
0.25
0.25
0.25
0.25
1.7
Cr
0.
0.
0.
0.
0.
(4)
0.
0.
0.
0.
0.
1.
\J 1
0044
0056
0072
on
on
02
02
02
02
02
4
(5)
20.0
25.0
30.0
45.0
48.0
SOURCE PERFORMANCE STANDARDS FOR PETROLEUM REFINING (1)"
EFFLUENT LIMITATIONS IN POUNDS/1000
SUBCATEGORY
TOPPING
CRACKING
PETROCHEMICAL
LUBE OIL
INTEGRATED
TOPPING
CRACKING
PETROCHEMICAL
LUBE OIL
INTEGRATED
VF(4)
BOD-5
2.2
3.1
4.1
6.5
7.8
TSS
1.9
2.5
3.3
5.3
6.3
COD
11.2
21
24
45
54
OIL
0.70
0.93
1.3
2.0
2.4
PHENOL
0.016
0.020
0.027
0.043
0.051
EQUIVALENT ANNUAL AVERAGE p.
14.8
15.6
15.2
15.0
15.1
1.7
10,3
10.2
9.9
9.9
9.9
2.1
80.0
112.5
94.7
110.7
110.9
1.6
5.0
5.0
5.1
4.9
4.9
1.6
0.11
0.10
0.10
0.10
0.10
1.7
NH3
0.45
3.0
3.8
3.8
3.8
p.m .
3.4
17.1
16.0
10.0
8.3
1.5
BBLS OF CRUDE OIL
-H SULFIDE
0.012
0.017
0.022
0.035
0.042
Cr
0.037
0.049
0.068
0.105
0.13
CONCENTRATIONS
0.10
0.10
0.10
0.10
0.10
1.4
0.25
0.25
0.25
0.24
0.25
1.7
(2,3)
Cr +
0.
0.
0.
0.
0.
(4)
0.
0.
0.
0.
0.
1.
0025
0032
0044
0072
0084
02
02
02
02
02
4
(6)
10.5
14.0
19.0
30.5
36.5
NOTES FOR TABLES 7 & 8
(1) AS AMENDED ON MAY 20, 1975
(2) BEFORE APPLYING SIZE AND PROCESS FACTOR MULTIPLIERS. BASED ON 30-DAY AVERAGES.
(3) EXCLUDING ALLOCATIONS FOR RUNOFF AND BALLAST WATER
(4) EQUIVALENT ANNUAL AVERAGE p.p.m.
= (30-DAY AVERAGE, LBS/1000 BBLS CRUDE) (120)
(ANNUAL AVG. WASTEWTR. FLOW, GALS/BBL CRUDE) (VF)
VARIABILITY FACTOR (VF)
= RATIO OF 30-DAY AVERAGE TO ANNUAL (LONG-TERM) AVERAGE
(5) ANNUAL AVERAGE WASTEWATER FLOWS USED AS BASIS FOR 1977 GUIDELINES PER EPA DEVELOPMENT
DOCUMENT, EPA-440/l-74-014-a, APRIL 1974
(6) ANNUAL AVERAGE WASTEWATER FLOWS USED AS BASIS FOR NSPS GUIDELINES PER EPA DEVELOPMENT
DOCUMENT, EPA-440/1-74-014-a, APRIL 1974
-------
SESSION
"PETROLEUM REFINING GUIDELINES"
Chairman
William C. Galegar
Director, Robert S. Kerr Environmental Research Laboratory
U.S.EPA Ada, Oklahoma
Speakers
Allen Cywin
"Establishment of Petroleum Effluent Guidelines"
Martha Sager
"The Role of the Effluent Standards and Water Quality Information
Advisory Committee (ES&WQIAC) in Furthering the Scientific
and Technical Developments Towards Clean Water"
Joe G. Moore, Jr.
"The Role of the National Commission of Water Quality(NCWQ)
Robert T. Denbo
"Economic Impact of Wastewater Effluent Guidelines for
Petroleum Refining"
17
-------
18
-------
"ESTABLISHMENT OF PETROLEUM REFINING EFFLUENT GUIDELINES"
Al len Cywin
Director, Effluent Guidelines Division
E.P.A., Washington, D.C.
Let me start by saying I am always pleased to be back in Oklahoma, having learned
how to fly here in World War II at Norman; and I am always pleased to talk and partici-
pate in conferences with the petroleum industry. Mr. Brinegar, President of Union Oil
was talking to me recently and said "You know you folks in the Environmenta1 Protection
Agency and the environmental movement ought to thank our industry as we got you started
as a result of such incidences as the Tory Canyon and Santa Barbara oil spills. " Well,
of course, there are other industries and there are other reasons and we have had a whole
series of laws in which this nation tried to tackle the problem of water pollution control.
Three years ago the Congress over Presidential veto enacted Public Law 92-500. Now
why did they do so - why did the public demand cleaner water - why did the Congress
intensely debate the issue for two years - why did they pass one of the most comprehensive
pieces of legislation in order to establish new mechanisms and specific goals for water
pollution control departing in significant ways from a whole series of laws passed }ust two,
three, four, five years earlier? Now why does the law state in its very first sentences,
in the very first page, that the objective of this Act is to restore and maintain the chemical,
physical and biological integrity of the nation's waters? In order to achieve this objec-
tive it is hereby declared that consistent with the provisions of this Act, I) it is the
national goal that the discharge of pollutants into navigable waters be eliminated by 1985;
now that is the goal. Is the national goal that wherever obtainable an interim goal of
water quality which provides for the protection and propagation of fish, shell fish and
wild life and provides for recreation in and on the water be achieved by July I, 1983; now
that's a goal also. It is a national policy that the discharge of toxic pollutants and toxic
amounts be prohibited; now that's a policy. In order to accomplish these goals the Act
created a new regulatory mechanism requiring uniform technology based effluent standards.
As you know a level of standards for existing plants cal'ing for Best Practical Control
Technology is to be achieved by July I, 1977 followed by a more stringent requirement
to be achieved by July I, 1983 which, and I am quoting from the Act, "shall require
application of the Best Available Technology Economically Achievable for such category
or class which will result in reasonable further progress toward the national goal of
eliminating the discharge of all pollutants. " The Congress places faith in technology,
practicable, available,and achievable by 1983. Why?
For one reason industrial wastes are the principal point source of controllable water-
borne wastes. In terms of the generally quoted measurement of just BOD,-, fhe gross
wastes of industries are about three times greater than that of the domestic sewered waste.
Moreover the rate of U.S. industrial production is increasing at a rate of over four times
of that of the population growth rate. But BOD,, is just one consideration, industry also
discharges the bulk of the metals, chemicals, toxicants, oils and greases to our waterways.
Recent drinking water surveys illustrate the great amounts of noxious waste discharges that
provide a threat to our drinking -water supplies. Therefore of paramount concern is that
19
-------
20
public health now and in the future is at issue. Also at issue are other major beneficial
uses of water - for agricultural purposes, for recreational purposes and indeed for use by
industry itself. Industry which presently uses three times the volume of water required for
domestic water supply is a major beneficiary of water pollution control.
Now let us look at the mechanisms for achieving industrial water pollution control in
an equitable fashion on a national basis. As 1 have said industry has a stake in meeting
the objectives of the Act since it not only is a water pollution control act but a water
conservation act. The U.S. needs for industry, for agriculture and for domestic use, are
increasing all the time, but as we all know, the volume of water available in the United
States is finite. The 1972 amendments to the Federal Water Pol'ution Control Act, have
as a goal the elimination or reduction of pollutant discharges through effluent limitations
that encourage recycle and reuse and the law is to be implemented through a series of
permits issued by the Federal Government and the States. At this date, we have more than
half the states of the union implementing the program.
The industry effluent limitations are to be technology based; progress, not status quo,
is to be emphasized. Two new terms are coined, best practical control technology avail-
able based on the average of the best practices within an industry or class or category of
industries and best available technology economically achievable, which could be based
on a single application within an industry. Technology transfer is permitted where there
is little or no technology being practiced and, of course, we are interested in reducing
the amount of pollution not just the measurement of pollution so mass units are used wherever
possible for our limitations.
Now for industry, and by the way industry is required to do more than municipalities,
best practical is to be achieved by July I, 1977, best available by July I, 1983 and also
new source performance standards are to be established and the latter is to be based on best
available demonstrated technology or technology transfer. Now these are for point sources
that discharge directly to the waterways.
In addition to that, we are to establish pretreatment regulations and these could be
and usually are somewhat different for existing plants that are tied into municipal systems
and for new sources. By the way we have something like 40,000 industrial point source
discharges and we have something like 300,000 commerical and industrial establishments
on municipalities, but basically ingoing through the program we have had to establish
some five different regulations, and there they are. Now the July I, 1977 regulation, as
I said before is to be based on the average of the best practices, and by the way these
terms are not something we conceived at EPA. In establishing what a law requires one has
to follow the history of the law and it does get complicated in our democracy because
often times as the House works through its side it differs somewhat from the Senate; then
there's a conference committee and a conference report and reports to the Senate and
congressional record. These ideas and these thoughts on the effluent regulations are
therefore based not only on public discussions of the issue, intense discussions and debate
within EPA and within the federal establishment, but also on briefing the Congress on a
very regular routine basis on these ideas and these thoughts throughout the development
arid history of this program. Everything that we have done from the very first definitions
-------
21
has been supplied to fhe Congress. There are a number of factors that we have to rake
into account. By fhe way in establishing these kinds of regulations, no one had ever
done this before and no one had token into account1 many of these factors which became
quite important as fhe country entered into inflation, recession and energy crises. But
right from the beginning we did take into account as the law required the total cost
versus the effluent reduction benefits, the age of the plants, where this was a factor,
the reduction processes, where these were factors, process changes, engineering aspects,
non-water quality environmental impacts, in other words what are the tradeoffs between
transferring waste from one media to another, etc. There are only three places to put
waste - air, land and water.
Now occasionally, as I said before, we do practice technology transfer. There are
industries that are doing very little or nothing, yours is not one of these. The 1977
standard emphasizes end of process treatment but it does include in-process treatment
where this is common practice and in very many industries there are very good housekeep-
ing practices on-going and certainly should be used across the industry. The objective
is to reduce the point source pollution discharge. Mr. Moore will be discussing the
aspects of the Act which go to questioning whether or not this particular date or this
particular goal should be achieved but the 1983 goal in the Act (the only way one could
study this is to establish what the standards are) is quite a difficult one. It establishes
that really you had better take the best available technology and apply this across the
industry. The law which was passed on October 18, 1972 envisioned that there would be
a ten year effort to reach that objective and in-process controls are suggested. Alter-
native water uses, water reuse conservation, by-product recovery, reuse of waste -
water constituents, good housekeeping and quality control are all techniques in use in
your industry and other industries and which can be utilized in many other applications.
The objective in the Act, the ultimate objective is no pollutant discharge. Now in fact
a few of our regulations do recognize that certain plants and certain industries are indeed
achieving no pollutant discharge at the present time but in most cases the regulations do
have limits and these limits are not zero. Now to go through the thought processes we
take a major industry category, and some 28 were published as a minimum in the basic
law, and within that industry we look at like families, or classes or categories (as the law
states) of plants, by type, by size, or by age, or by climate, or by treatability or finished
product and we call these sub-categories. We then establish what the raw waste loads
are for these sub-categories, what the best or the average of the best in-planf controls
are and fhe end of pipe treatment and come out with our standards. Once again there
are five sets of standards. We look at the annual costs, both infernal and external. This
was a departure from previous rulemaking in most other agencies in that we actually took
our costs and tried to establish what the economic consequences of these regulations were
and these were considered and debated within and without the federal establishment before
fhe regulations were promulgated. Now for publicly-owned treatment works, which are
referred to in the law as POTWs, fhe law requires secondary treatment by July I, 1977
and if can require industry prefreatment and it does say in fhe law that industrial waste
which cannot be treated or will interfere with the operation of the municipal treatment
plant or pass through untreated should be prefreated. The law established fhe technology
based limitations as a minimum; if further stated that where wafer quality could not be
achieved by applying fhe technology based limitations something more stringent may be
-------
22
required. It also said that states may make their limitation more stringent than the Federal
limits for whatever reasons occurred to the states. Now let me just explain some of the pro-
cedures we have gone through in order to come up with this. First of all the law itself
expressed the view that for existing plants and at least for those 28 industries ( and the
petroleum refining was one), that we should publish proposed regulations twelve months
after the law was passed from October 18, 1972 to October 18, 1973 and some four or five
months thereafter the new source standards. We started the program building a little bit
from the 1899 permit program. Recall that the Corp of Engineers suddenly discovered and
we discovered about 1970 that the 1899 Rivers & Harbors Act called for a "permit" system
for discharges to waters. For a variety of reasons no more than 20 permits were issued in
that two year period and by that time Public Law 92-500 overtook the effort. Now if you
recall the President vetoed the bill and the Congress overrode the veto on October 18,
1972. Because of that we were inhibited somewhat in building up the resources necessary
to implement the bill as quickly as we would have liked to, but in any event starting with
a secretary and myself about four years ago and still only with a real small handful of
staff, we undertook to hire outside contractors to help us. We can start with the fact that
good contractors were hard to come by in a hurry. With regard to previous comments by
Mr. Beychok, I remember asking, if Fluor wanted to participate and Mr. Beychok said "Oh
no1.", so we had to pick and choose the best we could. We also set up certain other review
mechanisms. The law provides for us to consult with other federal agencies, states and
other interested persons. Now of course the people affected by the regulations are the
most interested persons, but you also have environmental groups and professional societies
and others. Now EPA, in establishing what our program plan would be, where our dead-
lines would be, who our contractors and who our project officers would be, what our
definitions would be, called in many groups, other federal agencies, states, industrial
organizations, environmental professional groups and explained what we were going to do
and solicited their cooperation. Now a lot of people didn't think the whole thing was for
real although, of course, many did. Of course, the cooperation was spotty across
industries and even between companies within industries. Within your industry, an
environmental committee was established headed by Russ Mallett and we had a very good
dialogue with that group throughout the rulemaking process. Now we also had to establish
our effluent standards and water quality information advisory committee and get that set
up and for a little while I was the executive/secretary/director of that committee and
carrying water on both shoulders. We asked that committee to hold public hearings and
Dr. Sager will go into that aspect of their operations. Faced with these deadlines, EPA
set some very stringent timetables on the contractual efforts and we told our contractors
that their very first draft report would be exposed to the public. Now this is highly
unusual, I don't think anyone here would want to expose his first draft effort to the public,
but we did. We did it for a purpose, part of the purpose went with a covering letter that
went to the public. We said, that if you don't like what you see in this document, please
let us know what you would like to see but give us your background, your data, your
rationale — for any changes, we just can't pluck things out of the air. In some cases this
was successful, in other cases not so successful. In addition to that external review;
within EPA, we tried to establish throughout the agency regional office and lab personnel
who might be knowledgeable in a particular industry and we set up what we call working
groups of these people for an internal review. These were technical task forces, and we
ourselves -- (we just didn't wait for comments to come in) we ourselves went over these
materials and in some cases wrote and rewrote these documents several times. We then
-------
23
as an agency came out with a proposed regulaMon. This is the first time EPA's name
appeared on it and a development document which was a manual which explained how
we got from here to there. Everyone of the factors in the Act is a chapter, purposely
written in that guise. We also took all of our background data, file cabinet loads, and
made these available to the public in what we call a "freedom of information center, "
and once again we asked the public to tell us what it is they would like to see in there
and give us their data, their background and their rationale, for any changes.
Now the timetable of 30 days or 60 days whatever they turn out to be were very
short because during this period of time, not only were we faced with the congressional
deadline which we didn't quite make (we made it for a number of regulations but not quite
all) but the National Resource & Defense Council took us to court and the court set us on
specific deadlines. Now what the court said is you will do competent professional work
on this schedule and they told us what the schedule was. We have tried and so far we
have stayed out of contempt, but in addition to that the court also added another 30
industrial categories, because the first 28 or 30 covered about only half the permit
applications. Now in addition to all of that, of course, as I say there are some five
regulations and we have established some well over 500 sub-categories so we have
promulgated and proposed over 2500 regulations in less than four years. As a part of
the regulation writing process !n the United States in fact as a part of our very basic
system of democracy we have three parts of government. We have the executive, the
congressional and the judiciary and too many people lose sight of the fact that the rule-
making procedures under the administrative procedures act and also under this Act,
Public Law 92-500, call for public lawsuits if the public so desires. Now a number of
our regulations were not sued, and a number of the lawsuits that were entered against
us we have long since settled out of court, some for, to me, very trivial reasons. 1 will
give you one. The textile industry sued us and before they did they came in and said we
want to take on a research study to illustrate whether or not the projected numbers which
we have reasonably agreed to for the 1983 standards could really be met through mi xed
media filtration. If we, the textile industry, do that would you look at the results and
we said as public servants we are obligated to do that and the law tells us to look at
this every few years. They said, well our lawyers tell us that we have to have this in
writing for after all you might not be here by the time our results come in so they sued
us and we and they developed a letter. We said if they come up with the results we will
look at it and that was the way that suit was settled. Others were a little more compli-
cated than that and others have gone right to the mat.
Now with respect to the two law suits previously mentioned by Mr. Beychok, I will
give you my interpretation. Neither Milton Beychok nor I are lawyers, and perhaps
this will all end up in the Supreme Court. The Seventh Circuit did not mention ranges
and I think in my view I could not have written a better treatise for the position taken by
EPA, than they did. The Third Circuit, although it did mention range, qualified that
term by saying that we have to give very precise guidance to the permit writers as to how
to apply anything over and above that minimum number. We will be wrestling with this
issue in the months ahead. We have still to hear from several other Circuits. I think of
great significance is the fact that we are all citizens and we have all expressed our views
and indeed the House of Representatives had an oversight hearing on our program a year
-------
24
and a half ago; something like the Church Committee on Intelligence. I would commend
the report resulting from that hearing to your reading.
My summary of it is that the House said to itself; gentlemen this is a very complex law,
people of goodwill are trying to work their way through it and let the process continue.
With that I think I am on time to answer questions.
DISCUSSION
W. C. Galegar: We have talked primarily about history of the guidelines program, what do
you think the future holds in this area?
Allen Cywin: Well immediately ahead if we did nothing else, we still have a number of
industry categories to get out still in the court order deadlines, for example by January 30
(this Friday) we have to get out pulp and paper - phase II and we have a series of others to
develop and these are all in process. Every time we promulgated a new source standard as
law required we also promulgated a new source pretreatment standard. We have a very
complex procedure in existence for pretreatment standards for existing sources and we have
undertaken a program to try to clarify that somewhat better. So we still have a lot to do
in the field of pretreatment. More than that last spring we proposed to the agency (I am
just spinning this off and I know it is being taped so take this in the context of just discus-
sion), a conception whereby 307a, would be used really as a last resort. For toxics in
general, once they are established on a list, should be integrated into 301 and 306 as we
go about reviewing and upgrading those regulations. Conceptually this seems to be taking
hold in the agency and conceptually we are negotiating with NRDC and others on this
issue. Here again we have been taken to court by the environmental organizations to get
on with the job in both pretreatment and in toxics. So basically we still have a number of
regulations to finish. The court suits themselves will give us a lot of work to do for exampl e,
the Eighth Circuit court remanded some of the standards in grain milling and just gave us an
opportunity to go back and look at it again. As public servants, if our review had shown us
to be wrong we would easily have admitted it, but in this case we felt we were correct and
added 2500 more pages to the record to show that the standards were based on sound data.
The lawyers again in responding to that for the other side said all you did was prove you
were right in the first place and therefore you are arbitrary and capricious. So some call
Public Law 92-500 the lawyers employment act and in some cases, the lawyers call it the
engineers employment act. Perhaps it's both, but there is an awful lot of work to do to
finish what we have started.
Frank Manning: Would you care to comment how you selected the parameters that were
Iisted in your guidelines and promulgations, for example BOD, why did you choose that,
instead of COD or TOC?
Allen Cywin: Basically, of course, it started with a look at the industry and the common-
ly regarded prinicpal pollutant parameters. Often times as we developed our regulations
we dropped parameters because we thought it might cost too much to monitor and we knew
that a certain amount of reduction would take place anyway if the technology was employed,
so we tried to keep our parameters to a minimum. In the case of the petroleum refining
-------
25
industry we had been discussing the parameters even before the law was passed with your
technical advisory committee for a long period of time and we are in general agreement
on the parameters that were selected.
Milton Beychok: Allen, I'm not sure whether you have had a chance yet to read the
Commission's staff draft report but what is your personal opinion about the 1985 goals and
the elimination of discharge as they relate to the removal of dissolved inorganic salts?
Allen Cywin: Well let me just say that the agency through testimony submitted by Mr.
Ruckelshaus before the House in 1972, I guess even in 1971, gave its ideas about universal
no-discharge. As I say some plants and some industries are achieving this now. Other
places like storm sewers and for certain dissolved solids, I personally as an engineer,
don't think we are going to achieve that for God knows when if ever in many industries.
This does not mean that the goal is not a good goal you see. This does mean because of
our population growth as the Commission reports and because of the industrial growth,
we have got to get as close to that as is economically and technically feasibly possible.
And for a lot of industries and a lot of plants this means from going from virtually unlimited
discharge to something that is quite within the realm of possibility and is being practiced
today. We have not suggested any technologies that those of us who have been working
in this field for a long time have not worked with for a long time.
Morris Wiley: I believe Mr. Cywin that you stated that industry was discharging about
three times as much waste as municipalities if I understood your statement. The National
Commission on Water Quality indicated that the industrial water intake was a little more
than half of the municipal waste discharges and the information that I would have would
suggest that say in the oil and grease pollutant parameter that municipalities are larger
sources than industry and probably true in BOD as well. What is the basis for saying
that industry, apart from agriculture, discharges more pollutants and where can one find
this information so that it can be examined by industry.
Allen Cywin: The basis of my information is the 1968 Department of Commerce report
(and my three-times statement was in terms of BOD,-)- If you will just think a large pulp
and paper plant introduces in one plant as much BOD,- as the population of a million
people. A cow introduces ten times more BO D^. than a person and so forth. I am inc-
luding feed lots as they are very large industrialoperations. The processing of animals
(you can estimate in terms of petroleum refining) can get 100,000 chickens in one place
you can get 10,000 cattle in one place and that's a large point source of discharge. I
was looking at a film this morning just before I came down here on the Today Show for
just a second and it was about California dairy operations. Everything was automated,
cows came into a line, they got washed down, the machines were automatically applied,
the machines were automatically taken off, they milk a cow in eight minutes - they just
run them through this machine, there, wo-san awful lot of cows in that place and a lot
of residual material.
Anonymous: Where can you find population equivalents for livestock?
-------
26
Allen Cywin: For that one number look at the 1958 census Department of Commerce.
As far as the cattle and what not we have published some papers which give population
equivalents and so has The Economic Commission for Europe of the U.N., but we are
talking about concentrated animal feeding operations. The environmental protection
agency was taken to court once again to include every cow and chicken not every farm,
but we have established population equivalent limits based upon research, you know
minimum limits, size limits and we have also established the fact that these have to be in
concentrated areas.
J. Anthony Tallis: During your presentation you mentioned toxicity analysis is going to
Fe factored into the guidelines in terms of cost. How is this a part of the policy of deter-
mination as to limitations? Maybe I can quote an example:-
as to whether a dam should be built in a certain location is useful in comparison
with alternative source of water supply, water supply being an absolute necessity
for industry and municipalities.
In something like this, what is the break-off point?
Allen Cywin: Let me answer in two ways, first of all the term is cost versus benefits of
effluent reduction. In other words, do you double the cost by taking out 2% more or 5%
more of BOD, or of suspended solids, NUS was one of our contractors and they established
a good series of graphs; for example, the iron and steel industry which showed what the
levels of technology would be, what their costs were and how much tonnage on a model
plant basis which would apply across the industry would delete from discharge to the water
environment. As a part of this analysis, we had a separate series of contractors look at
the economics of the Industry as a whole. Now where are looking at the industry across
the board with reference to what technology can achieve and in some parts of the industry
you will find the smallest plants achieving let's say the best technology. Normally you
will find in economics that the smallest plants are the most marginal and in many cases
you will find in our regulations size cutoffs for example, where there is no or a lesser
regulation because we don't feel that that segment of the industry should be regulated on
a technology basis because the cost or the total economics across that spectrum of plants
is prohibitive. Or you may find some differences by geography as for example in the
Alaskan crab operations,we have a different limitation than the contiguous states. We
have a number of variabilities throughout our regulations on this basis. I don't know the
exact numbers for example and the history behind them but obviously that Mr. Beychok
was showing and calling anomalies but I would guess that the average of the best practices
in those sub-categories created those anomalies, since certain sub-categories especially
more complex you perhaps have more sophisticated operations or designs and therefore are
taking out more pollutants and you should because you have more in there than just BOD,-
and suspended solids. We have not identified the toxics at this stage but the more
sophisticated chemical operations, petrochemical operations, are eventually going to have
to address themselves to this point of toxics.
W. L. Ruggles: What is the future of toxicity in the guidelines?
'. w:n: Amongst our own staff you will find different attitudes and different
-------
27
statements. Toxics obviously are something that can get quite emotional. You see in
Kepone for example, I think most of you have seen enough on TV recently of Kepone,
and the emotions that this has generated. One of the reasons why EPA has not been
able to come out under 307a with some real hard numbers as yet is because of the great
divergence of opinion as to I) what you can achieve and 2) the obvious economic costs
that are involved. You run to two viewpoints, if it's my health that is in danger I don't
care what the cost is, if it's your health I might not care either, but you most certainly
would and I am sure that this kind of issue is going to be debated not only for water
pollution but it is being debated for air pollution and OSHA and other regulations. It's
a very hard thing to cope with because what is toxic to some people in toxic amounts or
to the aquatic life isn't something we can prove quite easily across the board. There is
a point of view that let's play it safe until we find out some more about it. Industry on
the other hand obviously says well prove it and then we will do something about it. You
get these two different points of view which EPA is trying to balance all the time. So
let me say I think we will bring forth a balanced prospective when we finally get the
thing out, I hope so.
-------
28
BIOGRAPHY
Allen Cywin was appointed Science Advisor
for the Environmental Protection Agency's Office of
Water and Hazardous Materials in February 1976.
Prior to this, Mr. Cywin organized and served as
director of the Effluent Guidelines Division in the
Office of Water and Hazardous Materials.
Mr. Cywin also served as Director, Division
of Applied Science and Technology in the Water
Pollution Control Research and Development Office,
Assistant Commissioner for Operations and Engineering
of the Community Facilities Administration, and
Chief of the Saline Water Conversion Demonstration
Plant Program of the Department of the Interior
where he was responsible for directing the design, con-
struction and operation of the Nation's first large saline
water conversion plants.
Mr. Cywin has also worked for the Resarch
Division of the Navt Facilities Engineering Command and
is retired Seabee Reserve Officer.
-------
"THE ROLE OF THE EFFLUENT STANDARDS AND WATER QUALITY INFORMATION
ADVISORY COMMITTEE (ES&WQIAC) IN FURTHERING THE SCIENTIFIC AND
TECHNICAL DEVELOPMENTS TOWARD CLEAN WATER"
Martha Sager
Chairman, ES&WQIAC EPA7 Washington, D.C.
STATUTORY BASIS
The Effluent Standards and Water Quality Information Advisory Committee (ES&WQ-
IAC) is a statutory Committee established under Section 515 of PL 92-500.
"Sec. 515 (a) (1) There is established on Effluent Standards and Establishment
Water Quality Information Advisory Committee, which shall be
composed of a Chairman and eight members who shall be appoint-
ed by the Administrator within sixty days after the date of enact-
ment of this Act.
"(2) All members of the Committee shall be selected from the Membership
scientific community, qualified by education, training, and
experience to provide, assess, and evaluate scientific and
technical information on effluent standards and limitations.
"(3) Members of the Committee shall serve for a term of four Term
years.
MEMBERS AND STAFF SUPPORT
PL 92-500, the "Federal Water Pollution Control Act Amendment", was enacted in
October 1972. The ES&WQIAC members were appointed on February 3, 1973 for four
year terms - which may be renewed. The committee is composed of a chairman and
eight members. The staff consists of an Executive Director with administrative and
secretarial support.
ACTIVITIES
The data bases and information relative to industrial point source effluent discharges
are of such a particular technical nature that from its earliest meeting in March 1973,
members of ES&WQIAC decided to work closely with the technical engineers and project
officers in the Effluent Guidelines of the Office of Water Planning of EPA so that differing
engineering viewpoints might be settled at an early stage in guidelines development to
cut down the Administrators involvement in issues of such technical nature at the time of
promulgation.
This relationship was developed in concert with the director of the Effluent Guide-
lines Division who arranged the first joint meeting with project officers of EPA early in
April, 1973.
The Committee has carried out its statutory functions through public hearings; public
29
-------
30
meetings; industrial workshops; critique and review of documents; special studies; and
agency, industry, congressional; and public interest group liaison.
FIRST YEAR ACTIVITIES, MARCH 2, 1973 - MARCH 1, 1974
In its first year of effort the ES&WQIAC conducted 60 Industrial Workshops, 7 Public
Hearings, and 15 Committee Meetings. It provided detailed comments on an extensive
series of technical documents and proposed effluent limitations related to the initial 27
basic industries specified in PL 92-500, and proposed a new approach for developing effluent
limitations for industrial point source discharges became later known as the Matrix Method.
MATRIX APPROACH
(BNA News Article 1973)
"This new matrix approach for establishing effluent limitations under the Federal Water
Pollution Control was developed by ES&WQIAC and presented to EPA Administrator Russell
E. Train September 26, 1973.
The new method entails a modification of the guidelines based on state-of-the-art waste
water treatment technology. In addition to current subcategorization of industries, considera-
tions which affect an individual industrial operation are included in mathematical models used
to establish as best practicable technology. The current data base available from EPA and
industry serves as a baseline for establishing raw waste load. The baseline is modified as
additional data are generated and as predictive models of treatment technology are improved.
The subcategorization of industries considers four constraints on effluent standards of
performance, these include treatment technology, operational characteristics, economic equity,
and geographic and climatic considerations."
The matrix provides a quantitative, technically sound basis for subcategorization of
industrial categories which is necessary to insure that effluent limitations are consistent with
the requirements of the act and realistically achievable by classes and categories of point
sources. This method for quantitatively determining effluent limitations could be interpreted
as a National Guidelines.
Basic steps for development of an effluent standards model by the matrix method are
outlined, as follows: (1) Develop statistical plots of raw waste load and flow for an industrial
category. Determine if size, age, or type of process has a major effect and if so break the
statistical plots accordingly. Eliminate plants which have extensive water reuse or product
recovery not normally practiced in the industry. Base lineflow/unit production is one stand-
are deviation from the mean. (2) Determine from waste water treatment operating data on
best practicable technology if the waste water characteristics have an effect on process per-
formance under optional design operating conditions. If so, categories should be established
with appropriate effluent qualities. (3) Define operation characteristics of the industry.
Establish effluent qualities for each grouping. (4) Delineate geographical and climatic
effects. Best practical technology in some cases can be defined by area if spray irrigation
or evaporation ponds are applicable. However, in all cases an alternative for discharge
-------
31
using waste water treatment should be considered because land availability or cost could
mitigate against land disposal, waste water characteristics, particularly temperature,
should be considered. (5) For each of the groupings defined in (1), the cost of compli-
ance/unit production should be generated. Best practicable technology is defined and
costs are computed by selecting the large plants with prevalent technology. Processes are
selected and effluent quality defined for each grouping considering a reasonable cost equity.
(6) As additional data is complied, it would be added to the data base for current best
practicable technology. Treatment models would be updated as the state-of-the-art ad-
vances.
Industry support for this quantitative method was strong and was evidenced by several
hundred letters and telegrams to the Administrator of EPA requesting its evaluation by the
Agency.
In December, 1975, the matrix method was discussed by the EPA's R&D division at
Ada, Oklahoma for review and revision, if necessary.
In addition the Committee developed data formats for information acquisition from
industry and received and transmitted more than 320 communications to EPA regarding
establishment of effluent limitations. Under its responsibilities for Section 307, i .e., on
toxic substances the Committee: conducted special hearings; developed extensive lists of
materials proposed for inclusion in the original toxic substances list by EPA; prepared a
comprehensive bibliography of reference material on toxic substances and provided a posi-
tion paper for EPA on toxic substances regulations. (Details are provided in the Commit-
tee's First Annual Report covering the period March 2, 1973 to March 1, 1974).
SECOND YEAR ACTIVITIES, MARCH 2, 1974- MARCH 1, 1975
In the Committee's second year of operation a comprehensive Work Plan was developed
covering 10 key areas of Committee interest in providing, assessing and evaluating scienti-
fic and technical information for the Administrator of EPA, according to its mandates.
This FY'75 Work Program is shown in Table 1.
The Committee held 5 public meetings, and 40 industrial workshops to obtain techni-
cal information for assisting EPA in developing industrial effluent limitations. The Commi-
ttee also provided comments and recommendations on 56 documents and rule-making pack-
ages developed by EPA for establishing industrial effluent limitations.
In April 1974, at the request of the House Committee on Public Works, the Committee
provided comments on "Problems in Implementation of the Federal Water Pollution Control
Act". The summary of this testimony follows:
The ES&WQIAC Members have been concerned with the following critical issues
with regard to the effluent limitations for industrial point source discharges being develop-
ed by EPA under Sections 304 (b), 306, and 307 of PL 92-500; specifically:
(1) EPA's interpretations of definitive terminology and the effects of these interpre-
-------
32
tations on the promulgated limitations, i. e., Exemplary Plants; National Standard;
Guideline; Effluent Limitation; BPT; and, Zero Pollutant Discharge.
(2) The effectiveness of the methodology selected by the agency to develop industrial
point source effluent limitations; i.e., selection of a descriptive and qualitative
contractor's approach rather than a quantitative mathematical methodology.
(3) Specific items which resulted because of the definitions adopted by the agency
amd the application of the methodology selected:
(a) % BOD reduction required of BPT by 1977;
(b) Factors excluded generally in establishing subscategories in many of the
promulgated regulations;
(c) Economic equities largely imbalanced in favor of large "most exemplary
plants";
(d) Lack of quantitative economic analysis of non-water impacts and energy
requirements in the promulgated documents to date;
(e) Lack of economic and scientific substantiation with regard for Section
307 and toxic materials effluent discharges; and,
(f) Lack of cooperation by the Effluent Guidelines Division with the
Committee efforts to provide, assess and evaluate scientific and
technical information for the administrator of EPA according to the
mandate for the ES&WQIAC in Section 515 of PL 92-500.
Special information gathering conferences were planned or conducted to facilitate
exchange of technical information. In November 1974 the Committee sponsored a Federal
Interagency Conference to determine the quantity and quality of technical information
derived by EPA in establishing industrial effluent limitations. As a result of this confer-
ence, recommendations were made to the EPA Administrator for improvement in utilization
of these government resources.
A special task force was constituted in December 1974 to review and provide recom-
mendations to EPA on a newly proposed approach to developing Toxic Substance limitations.
A report "ES&WQIAC Report on Review of Draft Advance Notice of Proposed Rule-Making
on Toxic Substances" containing ES&WQIAC recommendations was subsequently transmitted
to EPA.
THIRD YEAR ACTIVITIES, MARCH 2, 1975- DECEMBER 31, 1975
During the Committee's third year of operation the Committee continued its activities
of meetings, workshops, special task forces and reviews.
On March 6 a Hearing was conducted on "Privately Owned Sewage Treatment Works"
including public interest groups, industry and Federal agency representatives.
A conference on radioactive wastes was conducted on April 3 with participation of
all Federal Agencies with responsibilities in the area.
The Committee continued its detailed review and comments on all EPA documents
-------
33
related to industrial effluent standards development and the associated rulemaking packages
and evaluated 20 of these during this time period.
Public Meetings were conducted addressing industry problems of meeting Best Avail-
able Technology (BAT) specifications for 1983. The schedule for these meetings with
milestones is shown in Table 2. The first meeting on BAT was held on April 24, followed
by workshop/meetings in July, August, and September.
Four special studies were initiated by the Committee to provide additional informa-
tion to EPA during this third year of operation: (1) Development of a Report on the Matrix
Method for Establishing Industrial Effluent Limitations; (2) Development of an Approach
for Establishing Best Available Technology (BAT) Under PL 92-500; (3) Analysis of Litiga-
tion on Implementation of PL 92-500; and (4) Analysis of Toxic Substances Legislation.
Five reports were prepared for publication in calendar year 1975:
(1) Second Annual Report - ES&WQIAC
(2) The Matrix Method - A Methodology to Assist in Establishing Industrial Effluent
Limitations under Public Law 92-500
(3) Summary of Contentions of Industry in Litigation Pending Under Sections 304 (b) and
306 of Public Law 92-500
(4) An Analysis of Current and Proposed Federal Legislation Seeking to Control the Use
of Toxic Materials
(5) An Approach for Establishing Best Available Technology (BAT) Under Public Law
92-500 - With Applications to the Organics, Synthetics and Plastics Industry
A new ES&WQIAC Toxic Substances Task Force was established in November 1975
to assist EPA in implementation of Section 307 of PL 92-500. A proposed rationale for
implementation was developed. A workshop/meeting was also developed and scheduled
for January 15 and 16, 1967 to review the EPA plans and ES&WQIAC proposal for imple-
mentation of toxic substances standards under PL 92-500.
CLOSURE
The ES&WQIAC is pleased with its output during its years of tenure since March
1973. We believe the evaluative techniques, methods of operation and range of inputs
are uniquely different from the typical advisory committee pattern of operation and could
possibly become a template for future operation of scientific and technical advisory
committees in the Federal arena.
-------
34
DISCUSSION
W. L. Ruggles: Will toxic-materials restrictions generally be even more severe than the
present 1977/83 guidelines?
Martha Sager: The toxic limitations conceptually should be more stringent because of their
direct relationship to the physiology of aquatic organisms and because of harmful effects in
minute quantities.
Irvine W. Wei: Please describe the matrix approach.
Martha Sager: It is a little difficult to discuss the matrix method in a brief time. We are
preparing a comphensive report on the "Matrix Method" which will be available shortly.
Briefly the method utilizes a series of matrices in which the first matrix characterizes the
individual inputs and leads to the definition of the standard raw waste load for each
category or sub category. Then these results are input onto a second matrix which yields
process designs for alternative choices as a function of the effluent quality. In turn these
are then applied to the economic matrix which generates a series of costs versus effluent
quality relationships for the alternatives from the second matrix. Through consideration of
cost versus effluent quality policy decisions are next made in establishing the ultimate
effluent standards for each sub category. This procedure is shown in Figure 1.
BIOGRAPHY
Dr. Sager is Chairman of the Effluent
Standards and Water Quality Information
Advisory Committee (ES&WQIAC), a
statutory committee established under
PL 92-500. She is also Professor and
Director of the Environmental Systems
Management Program within the Center
for Technology and Administration,
College of Public Affairs, at the
American University, Washington, D.C.;
Consultant in Environmental Management;
Member, Air Quality Evaluation Panel;
Technical Advisory Board, Department of
Commerce; Panelist, Measures of Air
Quality, National Bureau of Standards;
and a Specialist in Environmental
Engineering and Industrial Micobiology.
-------
35
TABLE 1 "ES&WQIAC COMMITTEE WORK PLAN (FY75)"
It has been proposed that the following tasks and work plans be adopted for ES&WQIAC
for the new fiscal year in order to accomplish the maximum production of end product in
relation to providing, assessing and evaluating scientific and technical information for
transmission to the Administrator of EPA.
I. Document Review
Overall engineering responsibilities for review of documents from Effluent Guidelines
Division.
Don Bloodgood (coordinator) Ramon Guzman
Wes Eckenfelder Lloyd Smith
II. Early Inputs for Current Effluent Limitation Development
Early ES&WQIAC guidance on approach and document preparation for the development
of effluent limitations. This will include workshops and special tasks for at least the
following industries:
Miscellaneous Foods and Beverages Ore Mining and Dressing
Machinery and Mechanical Water Supply
(Others)
All Committee Members (Brossmar - Coordinate, development)
III. Anticipating Inputs for Future Effluent Limitation Development
Ongoing review for revision and updating of effluent limitations. Data input design
and acquisition. Development of a broader base of data sources. Possible Matrix
Workshops.
Glenn Paulson Martha Sager
Wes Eckenfelder Martin Brossman (utilize additional
Blair Bower technical staff support when
available)
IV. Public Hearings for Public Interest Groups
Selected series of meetings involving public interest groups in relation to the possible
economic implications and environmental impact on the public of PL 92-500. This
task will have particular reference to ES&WQIAC responsibilities in order to advise
the Administrator about public reception and response to the progress EPA is making
on PL 92-500 in the area of industrial point source discharge.
Glenn Paulson Martin Brossman
Bob McCall
V- Federal Liaison
Liaison with other federal agencies on positions and problems relating to PL 92-500
and preparation of reports for transmittal to the Administrator.
Martha Sager Bob McCall Martin Brossman
-------
36
VI. Special Legal Tasks
Coverage and preparation of written reports to ES&WQIAC on hearings and legal actions
related to designated industries and toxic materials. Maintenance of liaison with the
General Counsel's Office in EPA and other federal agencies as required.
Blair Bower Rose Mattingley Bob Grieves
VII. Matrix Development
Coordination of development and implementation of matrix method by EPA.
Martha Sager Bob Grieves
Wes Eckenfelder Martin Brossman
VIII. Coordination with National Academy of Science
Coordination of National Academy of Science review of EPA in relation to impact on
ES&WQIAC responsibilities (with special reference to the inclusion of the matrix in
the EPA R&D Program).
Glenn Paulson Lloyd Smith
Martha Sager Martin Brossman
IX. Economic Analysis and Liaison with National Commission on Water Quality
Expansion of ES&WQIAC economic analysis inputs and coordination with NCWQ.
Blair Bower Martha Sager
Wes Eckenfelder Martin Brossman
X. Toxic Effluents
Review of effluent limitations, analysis of legal and technical aspects in liaison with
other federal agencies on Section 307 of PL 92-500.
Lloyd Smith Martha Sager Glenn Paulson
(participation dependent
on current court actions)
-------
37
TABLE 2 "BAT TASK FORCE SCHEDULE"
(Revised July 22, 1975)
(Organ!cs, et al, and Paperboard)
ES&WQIAC MEETING DATE LOCATION MILESTONE
MEETING ACTION
July 18 & 19
(Fri./Sat.)
Aug. 18 & 19
(Mon./Tue.)
Sept. 25 & 26
(Thurs./Fri.)
Oct. 30 (Thurs.)
Univ. of Kentucky Completion of Review Data;
Lexington, KY Data Collection Decide on Methods
of Analysis
Vanderbilt Univ.
Nashville, TN
ES&WQIAC
Wash., D. C.
ES&WQIAC
Wash., D. C.
Completion of Develop Report
Initial Analysis Outline
Completion of
Draft Report
Final Action
on Report
Review Draft
Report, Revise,
Produce Final
Version
Agreement on
Disposition of
Final Report
-------
38
OPERATING DATA STATISTICAL
FROM EXISTING CHARACTERIZATION OF RWL DATA
TREATMENT FACILITIES FROM ALL PLANTS IN INDUSTRY
• •
•^^ ^P*
RWL
PARA METE
FLOW
BOD
COD
SS
LJ
V.
<
$
Q
PROCESS PRODUCT CLIMATIC
R AGE SIZE TECHNOLOGY SPEC EFFECTS
1
VERIFICATION
OF PROCESS
PARAMETERS
<•
OTHER
SU8CATEGORIZATION
SELECTION OF SRWL
IN- PLANT
MODIFICATIONS ""
SELECTION OF
BPT
(BAT) ALTERNATIVES
t
SELECTION OF PROCESS DESIGN PARAMETERS
A.S.
F/U
A.L. T. F.
V
Sl/t Oh UNII S ^ ,
L.D.
A. R.
A.C. OTHER
L
"
DEF NITION OF EFFLUENT QUALITY (Sc )
S
BOD
COD
SS
OTHER
A.S.
A.L. T.F.
L. 0.
A.C. OTHER
'•
COST MODELS
OPERATING
CAPITAL
A.S.
A.L. T. F.
<
L.D.
A.C. OTHER
EFFLUENT QUALITY -COST
RELATIONSHIPS FOR
BPT (BAT) ALTERNATIVES
i
ECONOMIC
IMPACT
S DESIGN
in
LJ
0
O
-------
"THE ROLE OF THE NATIONAL COMMISSION ON WATER QUALITY (NCWQ)"
Joe G. Moore, Jr.
(formerly, Program Director, NCWQ, Washington, D.C.)
Head, Graduate Program in Environmental Science
The University of Texas at Dallas
Sometimes when I see what is going on today I wonder if the "good old days" were
as good as we sometimes want to remember. My first experience with Bill Galegar had
to do with the adoption of the Texas water quality standards. He had the proper federal
view that nobody at the state level knew anything and I had the proper state view that
we weren't going to have these "outlanders" telling us what we ought to do in our state.
In one of the early conversations with him I said, "I want you to know that we don't like
these foreigners coming down here telling us what to do." Bill sort of drew himself up and
got a scowl on his face and said, "I want you to know that my forefathers welcomed your
forefathers to this country when they came;" that did give me pause, I asked, "What do
you mean by that?", and he said,"Well, you know I come from Oklahoma, you know
Oklahoma was settled by the Indians. I happen to belong to a tribe of Indians who have
not yet executed a treaty with the United States government."
How many of you know what the National Commission on Water Quality is? Well,
may be I had better tell you something about this operation, with apologies to those of
you who do know. The National Commission on Water Quality, created by Section 315
of Public Law 92-500 is made up of five members of the Senate, five members of the
U.S. House of Representatives and five Presidential appointees. The Commission was
created in the law with a statutory deadline to make a report to the Congress on or before
October 18, 1975. As usual with government bureaucrats, we were two million dollars
short and six months late, if the Commission manages to make its report in March of 1976.
The Commission members — and by the way, of the Congressional members you will
recognize a large number of names involved in the development of Public Law 92-500 —
are as follows: the presidential appointees, Nelson A. Rockefeller, who was then
governor of the State of New York and now Vice-President, has remained chairman
throughout the. life of the Commission; Dr. Edwin A. Gee, senior vice president of DuPont
in the Wilmington, Delaware office; William R. Gianelli, a consulting engineer from
Pebble Beach, California who is probably better known as the Director of the California
Department of Water Resources, the agency responsible for developing the California
water project; Mr. Raymond Kuduleis, the Director to the Department of Public Utilities
for the City of Cleveland, Ohio and Mr. S. Ladd Davies, Director of the Arkansas
Department of Pollution Control and Ecology; from the Senate, Muskie from Maine,
Randolph from West Virginia, Bentson from Texas, Baker from Tennessee and Buckley from
New York; and from the House of Representatives, Jones from Alabama, Jim Wright from
Texas, Harold R. "Bizz" Johnson from California, Bill Harsha from Ohio, and James C.
Cleveland from New Hampshire. There were three additional Commissioners, John A.
Blatnik, a former Congressman from Minnesota, James R. Grover, a former Congressman
from New York and Carl E. Wright the Chairman of the Arkansas Commission on Pollution
Control and Ecology who died during the summer of 1973 and was replaced by the
Director of that particular agency.
39
-------
40
The Commission's charge in the statute Is a very narrow one, "The Commission shall
make a full and complete investigation and saidy of all of the technological aspects of
achieving and all aspects of the total economic, social and environmental effects of
achieving or not achieving the effluent limitations and goals set forth for 1983 in Section
301(b)(2) of this Act. " That assignment was developed after some controversy in the
conference committee on P.L. 92-500. By the time the Commission was actually operable,
nearly a year after the passage of P.L. 92-500, attitudes even of the congressional members
had changed, and the Commission early decided that in addition to taking a look at the
narrow requirement for 1983, the Commission would also examine the progress being made
under the 1977 requirement to the Act and examine as well the potential impact of the
elimination of the discharge of pollutants — the 1985 goal of the Act. After a great deal
of discussion,the Commission, in February of 1974, published a study design stating what
it intended to do in the time remaining to it.
The Commission appropriated or was authorized an appropriation of $15 million
initially, has secured an additional $2 million authorization and a $2 million appropriation,
so that by the time it concludes its work it will have spent at least $17 million. About
$14 million of that total will have been spent on contract research projects defined by a
professional staff to fulfill the stated study design and represented in some 90 contractor
reports. I have one with me, or at least a part of the one with me on the petroleum
refining industry. This is the contractor's report that went through two drafts and has now
gone into the National Technical Information Service of the Department of Commerce.
This does not include the Appendix which contains the data base; but if you can imagine,
some of them run to four volumes this thick, so you can have some idea what sort of stack
you'd get if you looked at all the reports. For any of you who may be interested in
finding out about any report of the Commission, or getting a list of the reports, or getting
a copy of the draft staff report or its summary, you should write to the National Commission
on Water Quality, P.O. Box 19266, Washington, D.C. 20036, and we will be glad to
send you a publication list. If we don't have a particular report, we can give you some
idea about when it is likely to become available.
The contractor's report on the petroleum refining industry was done by Engineering
Science of Texas, Inc. You have two staff people from that concern later on in your
program talking about various waste treatment technologies, Davis Ford and Lial Tischler.
Rather than take up my time trying to explain to you a technical subject from a person
not versed in technical matters, I would suggest if you have questions you might catch
them while they are here with regard to the actual technical aspects of the Commission's
analysis of petroleum refining. Mr. Beychok, whom you have already heard, was a
consultant to the Commission.
There were some 75 consultants used by the staff and an additional 30 or 40 specific-
ally mentioned in the Commission's authorizing legislation.
The staff itself consisted — the program staff — consisted of some 30-35 professionals
recruited by me and my deputy, Jim Smith, who came to the Commission from the
Conservation Foundation. That professional staff oversaw the actual execution of contracts
-------
41
for Hie various studies and then produced a report almost as lengthy as the one I have
just shown you at some 900-1,000 pages; this report went through three publicly
distributed versions, and probably three private versions that were prepared by the
staff before any were released publicly. The staff draft report consists of several
chapters; I am going to give you the general organization of the document. I have with
me only the first chapter called Issues and Findings, this is the one that contains
questions and answers conforming with the Commission study design. For those who wish
merely a cursory examination of the Commission's work, this is some 82 pages long, and
you can almost pick the subject in which you have an interest and take a look at it in
that particular section. If you are more interested in the detailed subject matter areas I
will give you the chapters so that you will know the contents of the full report.
Chapter II is an analysis of the capabilities and cost of technology for achieving the
effluent limitations required by 1977 and 1983 as well as what remains to be done after
1983, in a technological sense, toward achieving the 1985 goal of the elimination of
discharge of pollutants. This is an assessment of the capabilities of the public and
private sectors to apply the defined effluent limitations for 1977 and 1983 as well as those
more stringent to protect water quality. Chapter III is an analysis of the economic effect
of the cost of applying the necessary technologies on both a micro-economic and macro-
economic scale as well as the social effect of these changes. Chapter IV is a description
of present water quality and environmental water-based and related terrestrial conditions
and projections of anticipated change which may result from implementation of the Act;
this is where Mr. Beychok got some of the information he showed in the slides a moment
ago. Chapter V is an identification of the effects of the Act in eleven selected regions
in the nation. Chapter VI is an assessment of the public and private response as
institutional segments finance, implement, manage and enforce the nation's water
pollution control program. This Chapter concerns itself not only with the inter-govern-
mental institutional relationships but also the response of the various regulated segments
in terms of whether or not they can institutionally achieve the requirements of the Act.
In addition to the more traditional aspects of municipal and industrial technologies,
economic effects and results, there is also discussion of the requirements related to
agriculture, particularly with regard to feed lots, irrigation return flows and agricultural
non-point sources.
Generally speaking, the Commission staff draft concludes that the cost of achieving
the 1977 requirements and the 1983 requirements even in today's economic climate, are
not insurmountable. However, the Commission staff draft inevitably has raised the question
as to whether or not the incremental improvement to be realized from the achievement of
the 1983 requirements, once the 1977 requirements have been met, is worth the probable
cost. And so the question is, really, do the changes in water quality, as a result of the
incremental requirements of 1983, justify the expenditure of the funds to achieve the 1983
requirements. Mr. Beychok did read for you the general conclusion of the staff with
regard to the elimination of the discharge of pollutants.
-------
42
Let- me just give you some figures that relate to the petroleum refining industry in
terms of the total. To reach the 1977 requirements for all refineries presently in place,
the technology assessment staff concluded that the cost would be roughly a billion dollars.
The economic impact analysis indicated that as a result of imposing these requirements
there would be certain plants that would inevitably close or consolidations that would
occur, and therefore the economic impact cost would be roughly $830 million as contrasted
with a billion dollar cost. You must understand that the technology contractors were
asked to price the various technologies for existing facilities. The economic impact
analyses — and by the way, they are contained in separate volumes, so this does not
include the economic impact analysis — the economic impact analysis would generally
show lower figures because of in those industries that may be affected so as to cause
plant closures, the economic impact figures are generally lower than the technology
cost figures. To get that in perspective for the 1977 requirements, the total estimated
economic cost is roughly $37 billion, so petroleum refining would account for roughly
one billion of the $37 billion.
The operation and maintenance costs for those facilities once in place in petroleum
refineries would be $142 million over and above whatever operation and maintenance cost
there is in 1975 for waste water treatment facilities. The table you saw earlier assumed
that the total of all these figures would represent the total O & M. There is a piece
missing, simply because we were not always able to get the actual O & M figures for the
base year. When you see O & M costs in Commission reports, these are incremental over
the base line — which is to say when we talk about 1977 O & M costs, we are talking
about the requirements, the O & M costs attached to the billion dollars of expenditure
that would be required to get to the 1977 requirements.
For 1983, the technology costs — and unfortunately the economic analysis did not
analyze in detail potential closures, primarily because most of them apparently occur in
meeting the 1977 requirements — the cost figures for both technology and economic
impact are the same for the 1983 requirements. The total cost is roughly $1.2 billion for
the refinery industry out of a total industrial cost of $23 billion for industry to achieve
the 1983 requirements. In just comparing these two gross figures and assuming that the
magnitude is reasonably valid, the cost to achieve the 1977 requirements is greater than
the cost to achieve the 1983 requirements.
In addition to these two estimates, the Commission's staff, using a massive macro-
economic model in which the staff does not have unlimited confidence, prepared a
prediction as to the cost of new source performance standards for those plants that would
be constructed between 1975 and 1983. For all new plants, the estimated cost is $20
billion, but the staff believes that that figure is much "softer" than the other two I have
just given you. The O & M figures for the 1983 requirements would be $429 million over
and above the incremental costs of 1977, and for new source performance standards, the
estimate was $290 million for operation and maintenance.
I must emphasize that in any tabulation you see of costs produced either in the staff
draft reports or analyses of the Commission's contractor reports or staff draft reports, it is
inevitable that the cost figures are chosen or selected from a range. We did not attempt
-------
43
to showall the ranges of costs that were selected or produced by the contractors; thus,
you will occasionally see a variety of figures credited to the National Commission on
Water Quality. What this means is that someone has chosen a series of figures from the
ranges either in the contractors reports or supporting documentation and have put together
a table of his own. For example, nowhere in the Commission staff draft work will you see
complete tabulations of total costs to the nation including all three — municipal,
industrial and agricultural — costs, simply because the number of variables or the range
of figures is such that it is very difficult to do. for example, you saw the figure of
$158 billion, for achieving the treatment requirements for publicly owned treatment works.
The range from which that figure is chosen is somewhere between about $65 billion and
$450 billion, so that the $158 billion is a figure that the staff chose as one that appeared
to be reasonable in terms of the requirements of the Act. I must emphasize that for publicly
owned treatment works there are six categories of eligible federal construction grant
assistance. Depending upon which of those categories might be funded, one can account
for a wide range of costs; for example, will municipalities be required to collect, contain
and treat all urban storm water run-off. If they do, and they have to apply secondary
treatment to that volume, you can regard that cost as one that contributed to that extreme
figure of some $450 billion. I caution you that when using cost figures, it is difficult at
best to get a series that may approach what you judge to be realistic, but you will see
figures that vary considerably.
The important thing to remember is that the cost for achieving the publicly owned
treatment works requirements of the Act are considerably higher if all categories are
funded and are considerably higher than those for achieving the industrial requirements.
At the same time it is obvious that industries have moved substantially farther towards
achieving the 1977 requirements by the deadline in the statute than have municipalities,
because the municipal program depends substantially upon the level of federal funding and
whether or not Congress and the Administration are providing the intended level of federal
participation.
Generally speaking, most industries believe that even the costs as I have given them to
you are severely understated. There are those who believe, however, that the costs are
overstated. As you might expect, Al Cywin and his group in EPA argue that we have
substantially overstated the costs and therefore distorted the consideration of Public Law
92-500. There are those who believe, particularly in meeting the 1983 requirements,
industries will strive for ways to minimize the costs through changes in production processes
and that therefore the 1983 costs have been substantially overstated, which is to say that
as the economics of waste water treatment approaches the point at which the costs become
a significant factor in the original investment or changes in investment, then industries
will find ways to adopt or adapt technologies that will be less expensive than those that
have been analyzed by the Commission.
One of the issues that confronts anyone examining this subject is the question of the
variability of the effluent limitations, and you have already heard some discussion of this
topic this morning. The Commission generally concluded that in the petroleum refining
industry EPA's effluent limitations, as presently stated, will preclude 100% compliance
by all industries in which the technology specified may have been installed on a uniform
-------
44
basis. This then raises the question, must the investment be doubled in order to ensure
stand-by equipment for every piece of the wastewater treatment process? Of course, it
is in examining these kinds of questions that disputes about cost arise.
Now without getting into the detail, let me just go to some general conclusions, then
I am going to talk about what is likely to happen based upon Joe Moore's prognostication,
since the Commission has not yet adopted any recommendations and now there may be
serious questions as to whether the Commission itself will be able to agree on what should
be recommended to the Congress. First of all I think it came as something of a surprise to
those of us, at least in the Commission staff, that there is a greater cost for reaching the
1977 requirements and a greater water quality improvement from reaching the 1977 require-
ments than is estimated for the 1983 requirements and water quality changes. Recognizing
that part of the difference may be accounted for in the methodology for making projections,
I think it nevertheless came as a surprise to us that the greater step — according to our
analyses — the greater step is to achieve 1977, and the smaller step is that to incrementally
go to the 1983 requirements provided, of course, you ignore for the moment the new source
performance standards which must be met over the entire period from 1975 until 1983.
You heard mention earlier of the four parts per milljon minimum dissolved oxygen
level. Remember that that particular level is a minimum during sustained low flow con-
ditions. Obviously, if that level of dissolved oxygen persisted in some places in the
country for long periods of time, it is conceivable the biological integrity of that body of
water could be substantially impaired. There has been some argument by environmenta-
lists that we should not have chosen four parts per million but should have chosen six,
which I presume would then suggest that we could compromise on five as a logical measure
for dissolved oxygen. I mention that because there is in the basis for the Commission's
analyses, grounds for differences of view, and so the Commission's results, I am sure, will
undergo considerable discussion in the months ahead.
The Commission has just concluded five public hearings, the last one of which was
held in Washington on last Monday and Tuesday. The general conclusion is that there was
substantial support for the genera! thrust of the staff draft report with a great deal of
quarrel about individual pieces. Each industry has generally argued that its costs have
been understated and therefore the total should rise; environmentalists and public interest
groups have argued that there should be no relaxation in the deadlines in the statute.
You heard some discussion this morning about the elimination of the discharge of
pollutants; the staff early decided to us the "net" definition, that is, there should be no
net increase in the constituents of the water from the intake to the discharge. You must
also appreciate that the staff was only allowed to examine the elimination of discharge
of pollutants after a great^deal of Commission disagreement; finally there was a trade-off.
Those who had objected to looking at anything but 1983 were concerned because of the
delay in the publicly-owned treatment-works construction grant program and they agreed
that the Commission could look at the elimination of the discharge of pollutants provided
the Commission would also look at progress being made toward the 1977 requirement,
particularly for publicly-owned treatment-works. While they did take a look at EOD
(elimination of discharge), the Commission staff's effort was limited when compared with
-------
45
the effort that went into looking at the 1983 requirements. Obviously, in the
elimination of discharge of pol lutants, one of the major concerns is what do you do with
the municipal plants — the publicly-owned treatment-works — and obviously the question
of dissolved inorganic salts becomes critical if EOD is to apply to municipal facilities.
Strange as it may sound, most people who talk about elimination of discharge want to
talk about it just in terms of industrial wastes and not necessarily in terms of the total
waste stream from all point sources. Also, it is obvious from the Commission studies,
that in some cases, not only will the point source controls be obsajred by the contribution
from non-point sources even with the 1983 controls in effect, but also if all dischargers
also went to the elimination of discharge, the results from this step would in some cases be
obscured by the contribution from non-point sources. The real question is one of balance
between the 1977 requirement, the 1983 requirements, the elimination of discharge and
what must be done for non-point sources where they now appear to be the major source of
pollutants to the nation's waters.
Some discussion about the ESQUIAC Committees matrix approach might be appropriate,
I must emphasize something about the way this Commission functioned. The question was
presented early, should the Commission examine some other scheme for achieving national
water quality objectives other than that contained in Public Law 92-500? As you should
gather from the listing of the membership of the Commission itself, a majority was involved
in the formulation and adoption of Public Law 92-500 and therefore they had already con-
cluded that the structure of 92-500 was the preferable one for achieving water quality
goals. For example, there was the question, should the Commission examine the prior
program of water quality standards to achieve water quality objectives? Of course, the
Congress believes it rejected that approach to the national water pollution control program
in the 1972 Act and, furthermore, most of them that speak on the question will say that
they became convinced that water quality standards were not achieving national water
quality objectives. In other words, they tend to believe that water quality standards is an
approach to a national pollution control program has been discredited. I say they believe
that whether or not the arguments upon which they believe it are valid.
Thus in looking at the program, the staff was to some degree restricted in what it
could examine.
Generally speaking, the technology contractors and the economic impact analysis
contractors were asked to look at five levels of pollutant control; (1) BPT, "best practicable
control technology currently available, " generally as defined by the Environmental
Protection Agency; (2) BAT, "best available technology economically achievable,"
generally as defined by EPA; and (3) the " elimination of the discharge of pollutants, " as
used in the definition I gave you earlier. There were two other levels for which we
attempted to get technological costs, one somewhere between BPT and BAT or below BPT
and one between BAT and EOD, so that, theoretically, there should be an analysis of at
least five levels in every contractor's report. As a practical matter, the contractors
achieved various degrees of success in their examinations. Fortunately, in petroleum
refining, because there was a very good data base though, actually six levels were analyzed,
primarily because of the question of the -- as I understand it — of the use of activated
carbon and the possible costs that might be imposed upon the industry for that requirement.
-------
46
There was an attempt to look at various levels of treatment. However, the staff did
generally not look at schemes for either achieving effluent limitations or for achieving
national water quality objectives outside the general structure of Public Law 92-500.
Thus we did analyze the ESQUIAC approach.
With that, let me go to the issues as I see them at the moment. First of all, will
there be recommendations? Despite the fact that the majority of the Commission comes
from those who sponsored the legislation and presumably therefore endorsed section 315,
there is some indication that those who feel the general thrust of the staff draft report is
inconsistent with their present or preconceptions about Public Law 92-500 believe that
perhaps it would be advisable for the Commission to make no recommendations. Without
recommendations they then do not have to contend with any suggestion other than what is
presently contained in 92-500. There is a question even now as to whether or not there
will be recommendations, and it takes interesting forms. We have a great deal of dif-
ficulty finding an agreeable time at which the Commission can meet, because if meetings
can be postponed sufficiently long and the appropriation can be exhausted, then obviously
the Commission cannot come to recommendations within the appropriated funds. The
second question is, will there be recommendations that mean anything? Obviously if you
can't stop the recommendations, then maybe the thing to do is to keep there from being
any meaningful recommendations. The next question is, will there be any action in 1976?
There are those who believe that 92-500 was a product of a political year and the worst
thing that could be done would be to amend it in a political year. Those who may dis-
agree with the general thrust of the staff draft can then argue in 1977 that the Commission
staff's data base was in 1974, or 1973; that a great deal of time has elapsed and therefore
the Commission's work is out of date —another year has passed and so all those facts and
figures don't really mean anything and what we should do is postpone any consideration of
any changes until late in 1977. Think for a moment what that does with the deadline of
July 1, 1977. The argument is being made, for example, that after all, 1983 is seven years
away, and there is plent of time for Congress to act. My reaction is if they are as slow
in considering changes to 92-500 as they were in adopting it in the first place, Congress
should have started last September in order to consider amendments to 92-500 if they are
going to get them out of the congressional process before the 1983 deadline.
The next series of statements are strictly mine and you cannot attribute them to the
Commission. Without casting aspersions, I am going to use the phrase that we have come
to use as a result of our association with the National Academy of Sciences and National
Academy of Engineering Advisory Committee. I will credit the phrase to Dr. "Reds"
Wolman, a professor at Johns Hopkins University. The phrase is, "If I were running the
zoo," this is the way we could express our opinions and get them credited to us individ-
ually. And so, "If I were running the zoo," I would argue that there should be a delay
in the requirement of "best available technology economically achievable" for industrial
and agricultural point source discharges and a delay in "best practicable waste treatment
technology over the life of the works" for publicly-owned treatment-works beyond the
July I, 1983 deadline, (1) provided that in the interim, every effort is made adequately
to control toxics, either substances or pollutants — recognizing that a large number of
toxics are being controlled in permit conditions; (2) provided that there is an aggressive
-------
47
program for the adoption, modificaHon and enforcement of water quality standards
wherever achievement of the 1977 requirements will not achieve adequate levels of water
quality; (3) provided that there are stringent new source performance standards imposed
upon all new plants constructed in the future; and (4) provided that there is an evaluation
of the progress that will have been achieved by substantial compliance with the 1977
requirements before the course of the national water pollution control program is
irrevocably committed to BAT.
1 wish to make clear that that does not include an abandonment of the technology
based effluent limitations in Public Law 92-500. In my view, the achievement of either
a state or a national program requires the "marrying" or the joining of technology-based
standards with individual discharger permits. I would like to mention that that has not
always been the federal view. I was Chairman of the Texas Water Quality Board at the
time we were negotiating with old Federal Water Pollution Control Administration for
water quality standards and I was publicly chastised in a meeting in Houston, Texas in
June of 1967 because of the Texas permit system for individual dischargers. That system
was called by my predecessor, Mr. Jim Quigley, the process by which one grants "permits
to pollute. " I couldn't wait for an opportunity,and eventually caught Jim in an audience
at a meeting with the pulp and paper industry in New York City, to remind him of what
can transpire in eight short years in terms of national viewpoint and suggested that just as
the nation had decided to go with Texas in that case, they might find other ways to go
with Texas in terms of the national water pollution control program. I do think there must
be a technology base and permits.
I do have some misgivings about permits administered from the Federal level, and I
think the experience with 92-500 has demonstrated some of the problems. You see, as a
practical matter, the flexibility that everybody argues for the effluent limitations for
industrial categories very probably exists in the individual permit conditions. Those in
EPA who develop the effluent limitations don't like to admit that their associates in the
permitting process may not necessarily be following the effluent limitations in every detail.
We have collected enough evidence to indicate that the permits do not always agree with
the effluent limitations and may not be uniform across the country.
Now, "If I were running the zoo," there should be case by case exceptions to the
1977 deadlines, not only for publicly-owned treatment-works because of the delay of
Federal construction grant funds, but also for industries as well, because our evidence
indicates that industry will have to spend substantial sums of money or at least there will
have to be substantial cash flow in the next two years if they achieve BPT; therefore there
should be a delay for industry as well as municipalities conditioned upon good faith efforts
to attempt to meet the 1977 requirements. Parenthetically, you should appreciate, from
the discussion already of legal issues, that in some cases, the challenge either to the
effluent limitations or of the permits themselves will not be concluded by July 1, 1977.
In fact, there may be some that will then have judicially-applied schedules for achieving
the BPT requirement which could extend as long as five years beyond the conclusion of the
litigation — which could very well run into the 1983 deadline — just for achieving the
BPT requirement.
-------
48
All right, "If I were running the zoo/' I think fhere should be priorities for the con-
struction grant categories for publicly-owned treatment-works with a concentration of the
funding where it should be. The nation is not likely to produce the kind of money that is
contemplated in our estimates of the costs for publicly-owned treatment-works.
Generally, I believe there should be a retention of the elimination of the discharge
of pollutants as a goal, simply because there are some cases in which this is a technological
possibility. I should emphasize, in my view, that in most of those cases in which EPA has
used elimination of discharge as a part of the effluent limitation, it more often than not is
an elimination of the discharge of water as well as pollutants — which is to say, closed
cycle systems. I see no serious problem with retention of the elimination of discharge as
a goal, retention of the 1983 interim goal — the so called "fishable-swimmable" test, and
even retention of the ultimate goal of "restoring the physical, chemical and biological
integrity of the nation's waters." I do have a great deal of difficulty with the word
"restore," and probably the goal would have more meaning in the administrative sense if
it were changed to read "achieve and maintain" the physical, chemical and biological
integrity of the nation's waters.
There should be flexibility in the regulations and some delegation of authority to state
and local units of government, but that delegation should be conditioned upon an accept-
ance of the general philosophy of Public Law 92-500, adequate state and local resources
to achieve those provisions of the statute and a post audit of performance by the Environ-
mental Protection Agency. For agriculture, there should be some special way of control-
ling non-point sources; there must be some consideration of the effect these have upon the
water quality of the nation.
I would like to close by suggesting that we not overlook that 92-500 has built into it
a series of cycles. The effluent limitations are subject to review — conscious review —
at intervals of five years. The permits are issued for maximum periods of five years except
in the case of new source performance standards where the term is ten years; but there is a
cycle in 92-500, and I think it might be well for EPA to recognize that cycle and make
the necessary corrections or adjustments that should be made in effluent limitations and
permits in the second round. I predict that if there are not relaxations in the deadlines
and some adjustment in some of the rigid regulations issued by EPA, the entire statute
faces the prospect either of becoming administratively impossible — which is to say it
would collapse by virtue of inability to administer it, or — despite what some will say —
I think the Act will face serious challenges in the Congress — if not in the next two years,
certainly the next four. The very structure of the Act could be seriously weakened in any
such challenge.
DISCUSSION
Umesh Mathur: In the business of pursuing permits, I think EPA policy has been
traditionally to encourage the states to take over as much of this function as possible,
and obviously this gives the state the responsibility of defining in some manner the exist-
ing water quality within the boundaries of the state. Now I know that the State of
Oklahoma has declared that all the waters in the state are "water-quality limited" as
-------
49
defined in the Act. The State of Kansas which is just upstream from us on the Arkansas
has declared that every water body in their state is "effluent-quality limited" and I have
not seen any determination as to whether or not either oF these two designations are in
fact valid. And to me what this means is that those of us who are in the planning business
really have no ideas as to what EPA feels is the right attitude to take. Certainly if you are
doing regional long-term planning for facilities as well as non-point sources in industries
and storm water run-off, and you begin on the basis that the whole of the State of Oklahoma
if really filthy and the State of Kansas is extremely pure, then I feel like you get into
the contradictions and will have interstate battles eventually even if all the regulations
came out in the right manner with all the right numbers. Because what one state adds on
to the water leaving its state will be cumulative to what the next state adds on and so when
you get to Louisiana you will never be able to do anything with water quality because of
all the incremental additions that have been made along the way.
Joe Moore: You have hit upon what is in the statute, a difficult dichotomy. The construc-
tion grant program state allocations are based upon needs. You have states in the Southwest
and generally the South that do not reflect substantial needs measured against a secondary
treatment yardstick. The reason you have the declarations that waters are water-quality
limited is to get the needs increased. The State of Texas, by the way, has done the same
thing; most of its waters are classified as water-quality limited. The reason is that if it
didn't do that and set higher-requirements for publicly-owned treatment-works construction
— if you didn't get the needs increased — it wouldn't get a very large share in the con-
struction grant program. As you can gather from the list of names I read, with Bentson in
the senate and Wright in the house, both on Public Works Committees from Texas, the
decrease in Texas construction grant funds wasn't very well received when this first
allocation was made under 92-500. The question of water quality limited waters, I think
has largely been ignored by EPA; no judgment has really been made as to whether or not
the designation is valid. That is the reason I used the words "an aggressive program of
water quality standards with approval, actual approval by EPA." To show you what has
happened, in the President's budget message that just came out this week — if we properly
understand the first language we saw, the construction grant program will hereafter — if
that recommendation is adopted — will hereafter be limited to the achievement of second-
ary treatment only. There will be no advantage to a state declaring its waters "water-
quality limited". You see, I think the "water-quality limited" designation is to get more
construction grant funds, but affects every other activity in which you are engaged.
Actually, "if I were running the zoo," I would recommend that the construction grant
program be terminated on a date certain about ten or fifteen years from now, after which
there would be zero federal funding in order to force, not only the Agency but also the
Congress and the States to sort out their priorities for publicly-owned treatment-works.
You asked an earlier question about section 208; you will find tremendous reliance among
theoreticians on the whole planning structure of Public Law 92-500. Now I have been
involved in government a long time, and planning is the ideal solution to budgeteers. If
you are planning, you don't have to act and so you can plan for ever and you don't have
to spend large sums of money. The problem with the planning provision of Public Law
92-500 is they will be worthless unless there is a linkage between the planner and the
actor. The way the thing is structured at the moment, I don't think the linkage exists.
-------
50
To show you how far-sighted the budgeteers can get, they have now decided that they
will withdraw all federal funding for planning after another year . We will then have
plans that will be out of date within two years, because plans in my view are worthless
unless they are constantly updated. I hope I am proved wrong, but I am afraid that there
is a tremendous amount of dust being stirred up under the guise of 208 that may settle on
everybody and there will be no pollution control when it is over with.
Milton Beychok: You talked about the fact that the report, the staff draft, did not in any
place put the total cost for municipal industry and I agree it didn't. You also talked or
gave the caveats about the ranges, and I agree with that. I would point out that we have
not only the Public Law 92-500 but the Clean Air Act, because essentially the American
public believes in clean water and clean air as it believes in motherhood, and because it
believes someone else will pay for it. Despite all of the caveats, all of the ranges you
might have to attach to it, the greatest service your Committee could do to the American
public in this election year, is to let them know that it might cost them $150 a month.
Joe Moore: I appreciate that comment. One of the statements that was made in the course
of the Washington hearings was that for the first time, there was the cost of a program
stated in terms that average citizens, ordinary folks, should be able to understand; for
example, even in this section, (by the way this chapter is available like this bound in just
an 82 page document and we distributed many more of these than we did the more detailed
report), it actual ly says what the price rises are expected to be. You will always know if
we have understated them, you can expect the price rise is going to be greater than that.
There are some shortages that are mentioned. The industries and geographical areas of the
country that are likely to be most severely impacted, insofar as we were able to analyze
them, are stated; and just to show you the kinds of things you can get into, the analysis
shows that generally industry will be most adversely affected in the northeastern and central
part of the country and that there could very well be a shift of industry to the South and
Southeastern part of the country. In these kinds of things, even with all the caveats, I
agree that there is some value in being able to state things so that people understand them.
-------
"ECONOMIC IMPACT OF WASTEWATER EFFLUENT
GUIDELINES FOR PETROLEUM REFINING"
Robert T. Denbo
Coordinator of Environmental Control, Exxon Company, U.S.A.
Baton Rouge Refinery, Baton Rouge, Louisiana
This paper deals with the economic impact of the guideline requirements to the
petroleum refining industry. The major points to be covered include:
- a summary study on costs of wastewater improvement for which a status report was
presented to an API Division of Refining meeting in the spring of 1975.
- a case study of the experience of a major refinery in investment cost to achieve
1977 guideline requirements.
- insight into perspective of relative impact of discharges in terms of organic loading
from the nation's petroleum refineries compared to discharges from other sources.
- suggestions for corrections to Public Law 92-500, the Federal Water Pollution
Control Act amendments of 1972, will be presented.
The Environmental Protection Agency has issued the wastewater guidelines for Best
Practicable Control Technology Currently Available (BPCTCA) to be achieved by U.S.
petroleum refining by July 1, 1977. Based on a study sponsored by the American Petroleum
Institute and completed in mid-1973, the petroleum refining industry will have invested
approximately $1 billion to meet the requirements of BPCTCA. To arrive at this figure,
the contractor made estimates on approximately 100 refineries and adjusted this investment
to cover the entire 247 petroleum refineries in the country. Information available on the
status of wastewater treatment trains as of 1972 was used and equipment was added to
upgrade the treating sequences to include facilities estimated as being required to meet
BPCTCA. The target treating sequence included facilities shown in Table 1 . It is apparent
that this treating sequence incorporates the latest in today's technology.
This paper will not discuss the merits and demerits of the most recent guidelines for
petroleum refining. This is covered in detail in other places. Suffice it to say they are
tough. Some refineries will have real difficulty in meeting the requirements, even with
the target treating train, if they can meet them at all. Also, the daily variability require-
ments for some parameters are quite restrictive. The assumption of the cost study is only
that the investment provides the industry with the BPCTCA treating sequence and not that
guidelines are met.
By now many refineries are far enough along with their wastewater treatment programs
to provide accurate investment information based on funds already expended or appropriated.
It is of interest to consider the investment being made by the current largest U.S. refinery —
the Baton Rouge Refinery of Exxon, Company, U.S.A. — to achieve BPCTCA. Total
investment at this location since 1967, will amount to approximately $61,000,000. The
breakdown of the investment is shown in Table 2. For this particular refinery the configura-
tion of the sewer system made it less attractive to segregate clean, once-through cooling
water than to eliminate it by conversion to a recirculating system employing cooling towers.
This conversion cost $16,000,000.
51
-------
52
In-plant source corrections to reduce wastewater flow and pollutant load cost
$2,000,000.
Because of high annual rainfall, stormwater segregation and detention are quite
expensive at this location. The cost of these facilities amounts to approximately $7,000,000.
The expansion of the sour water stripping facilities cost $1,500,000.
Site preparation was exceptionally costly, requiring expenditure of $7,000,000.
Because of the shortage of adequate sites, it was necessary to reclaim 15 acres previously
used for storing oily silt removed in past years from the separator system when silty river
water was used for once-through cooling.
The $27,500,000 for the Wastewater Treatment Plant includes — in addition to a two-
train, activated sludge unit with aeration basins to provide 18-hour detention — sewer
segregation, a corrugated plate interceptor, equalization, dissolved air flotation, sludge
dewatering and disposal and the associated utilities required.
The Baton Rouge Refinery came onstream some 66 years ago. These costs illustrate
that solutions to retrofitting problems for an older refinery are expensive and are not of a
nature that is readily anticipated in a generalized study approach such as the API study.
The investment at the Baton Rouge Refinery to meet BPCTCA amounts to about $133,000
per 1,000 barrels of daily crude run, or almost twice the average for the entire petroleum
industry in the API study. This comparison does not contradict the API study, but it does
suggest that the API costs are not too high.
It has been established that the dollar costs for BPCTCA for petroleum refining are
significant. Let's examine what we are getting for this money. This can be done in terms
of investment per unit of cleanup. The $1 billion to go to BPCTCA reduces BOD,, by some
357,000 pounds per day for the entire industry. This amounts to an investment of$2,800
per pound of BOD- removed each day.
At this point I would like to comment briefly on certain economic facts. When
considering the costs of improvement of water quality, let's not forget who pays these costs.
In the cases of municipal facilities, it is clear to us that we do — as taxpayers. It is not
so clear that we also pay the costs to industry since these costs are passed on as higher
product prices to us as consumers. The essential difference is that it is usually easier to
require industry to spend money than it is to increase taxes. Costs as higher prices don't
have to be voted upon. The most important consideration in developing regulations should
be whether the citizen is getting his money's worth.
It now appears that by the end of 1977, the largest portion of the U.S. petroleum
refineries will have installed facilities to meet BPCTCA. If we decide to proceed with the
1983 requirements of Best Available Treatment (BAT) for petroleum refining, we will then
be going after reducing BODj. at a cost per daily pound 3-5 times that of a daily pound of
BOD- reduced by installing facilities to achieve BPT. And we will achieve a reduction of
only about 126,000 pounds per day of BOD-.
-------
53
For perspective on the impact of petroleum refining, let's look at the total of all
other sources - industrial and municipal. On this basis, the BOD_ from total petroleum
refining amounts to about 1% of the total. While it is true that BDDj- may not be the best
measure of pollutant load, it was the only parameter for which estimates of the other
categories were available. One would anticipate that the relative amounts of total
organic pollutants are roughly in the same proportion as
Some of my numbers are rough and can be verified when final data become available
from studies by EPA. The figures are probably good enough to provide perspective. If
all of the petroleum refiners in the U.S. would suddenly disappear and we could get the
required energy from sunlight, the overall impact on the nation's waterways would be
almost indiscernible.
One other comment on the relative impact of petroleum refining. None of the
comparisons have included the impacts of pollutants in runoff from feed lots, agricultural
runoffs or from land disposal operations. It becomes more obvious what constitutes the
greatest potential for cleanup.
SUGGESTIONS FOR MID-COURSE CORRECTIONS TO PUBLIC LAW 92-500
Now for some suggestions for mid-course corrections to Public Law 92-500.
It is generally believed that all responsible citizens agree with the objective of
environmental legislation if it is to protect human health and well-being. It appears
that there is much that is right with Public Law 92-500 in working toward this end. How-
ever, it is not clear that technology-based guidelines uniformly applied are the right
answer for correcting the pollution problems of the nation's waterways. Suggestions are
in Table 3.
The assessment of pollutants in drinking water known to affect human health should be
completed, and then high priority should be set for removing the known problem pollutants.
In order to avoid wasting of our technical expertise, natural resources and money,
what is done after this first step should be reevaluated. For each contemplated regulation
questions like the following should be answered:
- For example, does the impact of a water regulation on land destroy the net benefit
anticipated by the author?
- Does the regulation require consumption of excessive energy?
- Does the manufacture of equipment necessary to implement the regulation cause
excessive pollution somewhere else?
- Do the effects of inflation, balance of payments and possible infringements on other
social goals destroy the desired impact?
- Are we taking control steps in priority order? We should insist on this since we are
paying the bill .
These questions should be adequately answered before proceeding in our control
program. This is a business-like approach. We have the time, the expertise and the
-------
54
commitment to do this and do it correctly. Improvements from BPCTCA will have bought
the time for us. We can't afford to take the chance of wasting the very natural resources
we are trying to conserve.
DISCUSSION
Sheriff Khan: As you know, the refinery industry has generally installed electrostatic
collectors to control the FCC effluents whereas Exxon is the only company which has
installed a wet-gas scrubber. Would you be able to outline the economic and the techno-
logical benefits plus any comment on the legal aspects?
R. T. Denbo: Yes. We are discussing control of FCC particulate emissions to the atmosphere.
Since it is wet-gas scrubbing, then I guess it would qualify as a water topic. On that basis,
I will consider this a just question and we will go ahead with it. Let me mention the reason
for the development in the first place. An engineer who had been an expert in the catalytic
cracking area for some time was not satisfied with the previous experience with electrostatic
precipitators. He felt there must be a better way to achieve the Clean Air Act requirements
that were implemented by the States. Further, some States such as New Jersey required the
removal of so-called condensible particulates in early regulations for cat cracking emissions.
Steel and some other industries had employed wet gas scrubbing successfully for years and
the application was considered for cat cracking. The Baytown and Baton Rouge refineries
of Exxon Company, U.S.A. participated in pilot plant work that indicated that wet-gas
scrubbing had several advantages over electrostatic precipitators since you could remove
SO~ and condensibles in this system. So, consequently an economic comparison was made
thar indicated that investment and an operating cost were about breakeven with electro-
static precipitators. Consequently, the major refineries in Exxon have decided to install
these facilities.
BIOGRAPHY
Robert T. (Bob) Denbo is the coordinator of
environmental control at Exxon Company, U.S.A.
Baton Rouge Refinery. In this capacity, he is
responsible for development of long-range goals
for air and water conservation and solid waste
disposal problems. Mr. Denbo received a B.S.
degree in chemistry from Louisiana State Univer-
sity in 1948, finishing the work for his degree
after spending four years in military service during
World War II. Mr. Denbo has been with Exxon in
Baton Rouge since 1948. He is currently chairman
of an Exxon committee concerned with environ-
mental control for all Exxon refineries. He is
vice-chairman of the API Committee on Refinery
Environmental Control, a member of the Sub-
committee of the API Division of Environmental
Affairs Water Quality Committee and other
industry committees.
-------
55
TABLE 1
efficient in-plant water use practices
segregated storm runoff and detention
sour water stripping
gravity separation
equalization
added removal of suspended oil and solids by either dissolved air flotation,
filtration or coagulation
activated sludge biological treatment
postfiltration
facilities for dewatering oily and biological sludges and sludge disposal
TABLE 2 "INVESTMENT REQUIRED TO ACHIEVE BPCTCA AT THE
EXXON COMPANY, U.S.A., BATON ROUGE REFINERY"
Project
Purpose
Investment
$
River Water Replacement Project Eliminate the use of once-through 16,000,000
cooling water to reduce effluent
rate by about 90%
In-Plant Waste Load Reduction
Storm Water Handling
Sour Water Stripper
Reclamation of Old Oily Silt
Storage Pond by Sand-Lime
Stabilization
Wastewater Treatment Plant
(Activated Sludge)
To further reduce flow and abate 2,000,000
pollution at the source
Segregate, detain and treat 7,000,000
storm water
Remove ammonia and H~S from 1,500,000
effluent
Develop site for Wastewater 7,000,000
Treatment Plant
Provide treatment to achieve 27,500,000
BPCTCA
Total
61,000,000
-------
56
TABLE 3 "SUGGESTED MID-COURSE CORRECTIONS TO PUBLIC LAW 92-500"
(Steps assume BPCTCA is implemented as currently scheduled in mid-1977.)
Complete the assessment of pollutants known to affect human health.
- Determine sources and abate.
At that point provide an adequate period for reassessment of any additional steps.
- For each new, contemplated regulation evaluate:
the impacts of that regulation on other areas of the environment — air;
water; land
energy consumption
use of natural resources to install the control facility
impacts in other segments of the economy
Impose only those regulations that are determined to provide net benefits that justify
the net costs to society, with special consideration for the specified uses and the
assimilative capacity of each waterway.
-------
SESSION II
"PERMITS FOR THE POINT-SOURCE CATEGORY"
Chairman
Jerry T. Thornhill
Regional Representative
Office of Program Integration
U. S. EPA, Region VI, Dallas, Texas
Speakers
Carl J. Schafer
"The Environmental Protection Agency's Permit System
Robert F. Silvus
"Current Approach to Refinery Permits"
William K. Lorenz
"Petroleum Industry Experience With The NPDES Permit System1
Harless R. Benthul
"The Enforcement of Permits"
F. T. Weiss
"Measuring the Parameters Specified in Permits"
57
-------
BIOGRAPHY Jerry T. Thornhill
Jerry T. Thornhill holds a B.S. in Geology from the
University of Texas at Austin. Mr. Thornhill has been
with the EPA in Dallas, Texas since 1969. During this
time he has served as: - Contingency Plan Officer; Chief,
Federal Activities Branch; Chief, Oil & Hazardous
Substances Branch; and Deputy and Acting Director,
Hazardous Materials Control Division - before assuming
his present position of Regional Representative, Office
of Program Integration in 1974. Mr. Thornhill has also
served with the Texas Water Development Board as Water
Quality Specialist and as Assistant Director, Ground
Water Division; the Texas Water Commission; and the
Texas Board of Water Engineers.
58
-------
THE ENVIRONMENTAL PROTECTION AGENCY'S
PERMIT SYSTEM
Carl J. Schafer, Acting Director
Permits Division
Office of Water Enforcement
U.S. Environmental Protection Agency
Washington, D. C.
Created by the Federal Water Pollution Control Act Amendments of 1972 (Public
Law 92-500), the National Pollutant Discharge Elimination System (NPDES) replaces and
improves upon the old permit system under the 1899 Refuse Act. It is part of the compre-
hensive effort set in motion by the 1972 law to prevent, reduce, and eliminate water
pollution.
After listening to all the excellent speakers this morning, I have to tell you they
make up a hard act to follow. I thought, too, of all sorts of clever opening remarks as
I sat in the audience but as I stand here, they've either vanished or have been used by
somebody else. I would just like to say that it is a pleasure to be here, and it is a
special pleasure to renew acquaintances with many of you with whom I have worked in
years past in setting up some of the initial considerations for permit limitations before
there were guidelines. The Permit Program is established under Section 402 of the
Federal Water Pollution Control Act as amended in 1972 and constitutes one cornerstone
of a comprehensive program to insure that the intent of the Act is complied with by all
point source dischargers to the waters of the United States. Under the Act, National
Pollutant Discharge Elimination System (NPDES) permits specify effluent limitations
and schedules of achievement which involved both compliance monitoring requirements
on the part of the discharger and compliance monitoring on the part of the regulatory
agency, be it State or Federal. Obviously, the purpose of all NPDES compliance
monitoring is to assess whether or not the limitations are being met and, if not, whether
enforcement actions are necessary.
The Program is designed for ultimate delegation to each State, and State participation
with and without delegation has been most gratifying. While the States retain primary
responsibility to combat water pollution, they do so within the framework of the Federal
law. Wherever States encounter difficulties in administrating the NPDES, the Environ-
mental Protection Agency may intervene to assist in resolving the problems, be they legal,
technical, or administrative. Currently, a total of 27 States manage their own NPDES
programs, all of which are monitored constantly by Federal EPA Water Enforcement
personnel. The Environmental Protection Agency is a highly decentralized effort in
itself as it delegates authority to its ten regional offices, each of which exercise a high
degree of autonomy.
I'd like to address today the actual substance and content of NPDES permits. By
and large, you can break an NPDES permit down into four general components: (1) the
limitations; (2) the compliance schedule; (3) the self-monitoring and reporting requirements;
and (4) the boiler plate. Limitations, of course, are based upon effluent guidelines, or
59
-------
60
water quality standards in their absence, under Section 402 (a) (1) of the Act. The indi-
vidual determination by EPA for a particular plant regarding what constitutes Best Practicable
Treatment (BPT) is generally written in pounds, depending upon production levels at the time
the limitations are determined. Concentration limits are also specified depending upon
various other requirements. Compliance schedules cite specific steps to achieve compliance
with the Act by July 1, 1977, as well as reports of progress on the achievement of each step
and, in particular, reports of failure to make progress.
Each of the requirements in an NPDES permit, whether it be interim or final, is
legally enforceable. Each permit is for a fixed period of time not to exceed five years and
application for renewal must be made 180 days prior to the expiration date. Most of the
major permits were issued by December 31, 1974, and we therefore expect the next round
of permit issuances to take place during 1979 and 1980.
I would like to give you a little bit of insight into where we stand right now, what
the status of the program is, and what is going on. There has been a great deal of interest
expressed regarding the Seventh Circuit Court decision by Judge Flannery on feedlots and
I don't think we know yet exactly how many new point source discharges this is going to
add to the list, but for the time being at least — that not withstanding we have a total
of 59,048 identified dischargers requiring permits. Permits have been issued for 43,420 of
that number, which is a 74% issuance rate. The breakdown on those figures indicates that
16,000 municipal permits have been issued out of 19,000 dischargers identified and in the
non-municipal field, which includes industrial, agricultural and Federal facilities, we
have issued just under 26,000 permits out of the approximately 40,000 dischargers that have
been identified.
Out of the 59,000 permits to be issued, of which EPA has issued 43,000, you might
be interested in what our adjudicatory hearing experience has been. To date there have
been 1,863 adjudicatory hearings requested. In other words, only 3 percent of all the
permits issued have been contested. Of those adjudicatory hearings requested, 1,105 are
still pending. A look at those pending requests reveals that it is not necessarily the major
discharges that give us the most grief, because by and large we will have devoted more time
and energy and thought to the limitations going into those permits. The minor dischargers,
on the other hand, pose some very difficult problems in peripheral areas which must be
resolved before a reasonable permit can be issued. Such is the case for petroleum bulk
terminals, two or three other sub-categories, some water treatment plants and other classes
of permits which make up a good portion of the 1,105 adjudicatory hearing requests yet to
be resolved. It is also interesting to note that only ten of the 750 adjudicatory hearings
that have been resolved required the use of all the formal adjudicatory proceedings avail-
able under the law. By far, the majority of all adjudicatory hearings requested are settled
prior to trial.
Insofar as the petroleum refining industry is concerned, most — if not all — of the
approximately 250 petroleum refineries in the nation have now been issued NPDES permits.
Of those, only about 20 are currently pending under the adjudicatory hearing request
review process, having been contested on some requirement or condition of the permit.
Typical issues in the adjudication of petroleum refining permits include deep well injection
-------
61
controls/ storm water runoff treatment, supply-water silt disposal and other miscellaneous
items.
Overall, the NPDES program has proven quite workable in the petroleum refining in-
dustry. The basic requirements for Best Practicable Treatment of process water do not seem
to have constituted a major source of disagreement between the industry and EPA. Earlier
statements provided at this Forum give further proof that the industry as a whole is well on
its way to achieving the installation of the equipment necessary to meet Best Practicable
Treatment.
To sum up, NPDES requirements were designed to make sure that discharges of pollu-
tants into the nation's waters are subject to uniform minimum controls and that the restrictions,
terms, and conditions of a permit are substantially the same whether issued by EPA or a State.
In brief, water pollution is a national problem, and the permit system is part of a national
program to combat that problem.
I will be happy to take any questions at this time.
BIOGRAPHY
Mr. Schafer received his degree in Chemical
Engineeeing from Georgia Institute of Technology, and
has extensive experience in industrial production and
development with Shell Chemical Company (Petro-
chemicals, Houston, Texas) and Celanese Figers Company
(cellulose fibers, Cumberland, Maryland). He joined
EPA in January 1972, and is currently Chief of the
Permits Assistance Branch, Office of Water Enforcement,
in Washington, D. C. As such, he is responsible for
quality control,technical assistance and policy coordina-
tion and policy coordination for EPA regions and states
writing National Pollutant Discharge Elimination System
permits for industrial point source discharges.
-------
"CURRENT APPROACH TO REFINERY PERMITS"
Robert F. Silvus
Texas Water Quality Board, Austin, Texas
HISTORY
Much of Texas is water short, and water is a valuable resource which has always
needed protection. The first official recognition in Texas came in I860, when a law was
passed to make water pollution illegal. But, it wasn't until the 1930's that the degree of
allowable pollution was quantified by a State Health Department requirement of secondary
treatment of sewage effluent from cities. From that time until 1967, the State Health
Department was responsible for water pollution control in Texas. The Texas Water
Pollution Control Board functioned (as a part of the Health Department) from 1961 to 1967
and initiated a permit system for both municipalities and industries. The Texas Water
Quality Board was established by the legislature as an independent agency in 1967. It
presently performs the administration of the Texas Water Quality Act and the Solid Waste
Act and has responsibilities under the Subsurface Disposal Act. The Board is composed of
three public members and representatives from four other State agencies with interest in
water.
Most State permits were written 10-15 years ago and revised since then as water
quality problems were recognized, plant expansions occurred, etc. The initial efforts
were more of a cataloging operation to denote the location and approximate quality and
quantity of discharges. State law requires control of effluent volume as well as average
and instantaneous quality. A self-reporting system was established in 1970, and com-
puterized data for each outfall of each discharger is available for review. Traditionally,
the State has relied on treatability studies to set limitations for industrial permits with the
overall goal of "good secondary treatment".
SINCE EPA
There has been quite an impact on the State permit system since the implementation
of the NPDES permit program beginning in 1972. The nationwide guidelines adopted by
EPA provide another and important means to check industrial permits for reasonableness of
effluent requirements and equity among permit holders. The State permit format has been
revised to resemble the NPDES format, and a working agreement established with the EPA
whereby we write most NPDES permits for the EPA for Texas dischargers. Currently, we
consider (1) waste load evaluations and allocations for the various stream segments, (2)
stream standards, (3) treatability studies, (4) self-report data, (5) the existing State per-
mit and (6) guidelines in considering a permit revision. It should be emphasized that
before the guidelines were issued most refiners in Texas had either already built or con-
tracted for wastewater treatment plants which can now meet or exceed guidelines. In
fact, six of the twelve exemplary refineries used to develop the guidelines are located in
Texas. These are Coastal States and Champlin at Corpus Christi, Union Oil at Beaumont,
Marathon at Texas City, Shell at Deer Park (near Houston) and Phillips at Sweeney.
Most treatment systems include oil-water separators, dissolved air flotation units, equali-
zation basins, biological treatment units, clarifiers, and a sludge handling system.
63
-------
64
Having taken care of the major treatment problems during the late 1960's and early
1970's, most Texas refiners can concentrate their efforts on the remaining problems of
storm water handling and hazardous metals control and should be able to meet guidelines
in 1977 without additional major expenditures. However, a few old refineries do not yet
have adequate treatment facilities and will have difficulty meeting even the 1977 guide-
lines. It is almost impossible to carry out a large wastewater handling and treatment
program over a period of years without the continuous commitment of corporate manage-
ment to the program. The State strongly advocates a management policy which
encourages timely investments in concrete and steel rather than in a good staff of
attorneys.
PERMIT PREPARATION
Now that you know how we arrived at our current 1976 state of affairs, you would
probably like to know what is involved in the preparation of a waste discharge permit.
The Texas Water Quality Board has divided the various water bodies in the State into
segments and each segment is regularly monitored. A set of stream standards was adopted
in 1967 and subsequently revised several times. Those stream segments which have
problems are designated "Water Quality Segments", and some special limitations designed
to correct the problem are placed on all dischargers to the segment. In the case of the
Houston Ship Channel, a computer model was developed, and a waste load allocation
expressed in pounds of BOD^ and ammonia-nitrogen was established by the State for each
discharger (including municipalities). The EPA reviewed and concurred with these
limitations. In addition, the State has issued several special statewide orders which affect
many dischargers. Chief among these is the Board Order governing the discharge of
hazardous metals. The latest revision of this order is stringent and may require special
treatment to remove or replace catalysts or cooling water additives which were previously
discharged. When an engineer first begins to draft a permit, he looks at the stream
standards, waste load allocations and special orders to determine their applicability.
A second question for existing dischargers concerns their current performance. The
Texas Water Quality Board has 12 district offices around the State staffed by about 100
employees. By reading their inspection reports and reviewing the self-reporting data from
the permittee, the engineer can answer questions about current performance. What he
does not know is whether refinery operations were normal and whether the refinery was
operating near its capacity. This information along with unit capacities, etc. is usually
obtained by direct contact with the company.
There is an existing State permit for each refinery. Most State permits were written
to include concentration limitations and dry weather flow. The engineer will calculate
the limitations (in pounds) which are established in the existing State permit for later
comparison with guideline calculations. The guideline numbers will be used in the new
permit unless the existing State permit requires something more restrictive.
From the NPDES or Corps of Engineers application, we would attempt to determine
the unit capacities so that size factors and process configuration factors can be calculated.
Frequently, additional information from the company is required. A storm water
-------
65
management plan is normally a very important ingredient in every permit, and a ballast
water treatment system is important in many cases. Storm water runoff can be estimated
from a plot plan of the refinery and historical rainfall data. A better source of data is
historical in-plant concentration and volume measurements on storm sewers. Historical
ballast water receipts can be used to estimate the ballast water contribution to the permit.
Ballast water is being treated successfully in combination with other refinery wastewaters.
Normally, storm water data is the most difficult item to quantify in a permit. Since
storm water runoff is a problem common to all refiners, some further elaboration of the
type of recommendations which we make is in order. First, we recommend that process
and storm water sewers be separate. Second, we recommend that all storage tanks be
diked and that tanks not be drained into the diked area; draining water from tanks into
the firewall area has been common practice at some locations. Third, we recommend that
all process areas be curbed and provided with drains so that contaminated storm water will
not be discharged without treatment. It has been the practice in many plants for
operators to rinse sample bottles and pour the contents down the drains or on the slab and
to drain equipment which is opened for maintenance on the slab. If it rains before this
material is flushed to a treatment unit, the storm water will be contaminated. Our fourth
recommendation is that all contaminated storm water be captured for eventual treatment
and that clean storm water be discharged directly. The permit usually defines clean
storm water in terms of (1) pH, (2) TOC and (3) Oil and Grease content, and this storm
water can be released without treatment. As mentioned earlier, historical volume and
quality measurements provide the best data on the magnitude of the storm water treatment
problem. Impoundment of the contaminated runoff is the most common approach. This
allows the collected water to be treated over a period of time; this reduces the hydraulic
loading of the treatment plant which, in turn, reduces the investment.
Impoundment is not always practical because of the geography of the plant or avail-
ability of land. In such cases, exceptionally good housekeeping practices are necessary,
and the additional investment for an oversized treatment plant may also be necessary. A
final recommendation is that runoff from non-point sources including those areas held for
future development be diverted away from potentially contaminated areas through an
adequate system of dikes and ditches. In many cases, efforts expended in preventing the
contamination of rainfall runoff can greatly reduce the magnitude of the problem.
Another common problem involves the compliance with chromium limitations. Limi-
tations are usually based on the volume of chromium-treated cooling water discharged
and not on the total plant effluent. Since not every refiner has had success in substi-
tuting other water treatment chemicals for chromium, some plan to continue to use
chromium but to treat the cooling water blowdown prior to discharge. There are three
processes under evaluation to achieve removal. These are (1) chemical precipitation and
clarification, (2) ion exchange, and (3) electrolysis. Chemical precipitation with clari-
fication is the most proven process but each has its advantages. Each of the processes
will be installed at some manufacturing plant in Texas before July 1, 1977, so we will
soon learn how well they perform in routine service. It is very difficult for a company to
meet the required chromium levels if the discharges are not confined to a few large
volume streams. For this reason, the engineer who prepares the permit will consider each
plant situation individually.
-------
66
Having reviewed the stream standards, waste load allocation, Board orders, operating
history, NPDES applications and other company input, and the existing State permit, the
engineer will simply compare the calculated limitations using the existing State permit or
allocation with calculations using the guidelines and pick the more stringent limitations
for the July 1, 1977 requirements. Limitations are norrfraUy placed on BOD5, TSS, COD,
oil and grease, phenol, ammonia-nitrogen, sulfides, chromium and pH. Volume limi-
tations and requirements on other hazardous metals, temperatures, and occasionally on
chlorides and sulfates are added. Interim limitations (until July 1, 1977) are usually
based on self-report data, field reports, current State permit limitations and federal
applications. Limitations for each time period on individual samples and on volume are
also placed in the State permit. At this point, it is normal to discuss the proposed permit
with the company to determine the accuracy of assumptions, applicability of particular
guidelines to the refinery, currentness of the data, time required for compliance, and
approved expansions or other major changes to take place during the five year period
covered by the permit. Once all points have been explored, the State permit is set for
public hearing and a similar NPDES permit is sent to the EPA for review. The two permits
then proceed through the various public notice and hearing procedures where they may be
modified as a result of new evidence introduced by the public or other governmental
agency. Finally the permits are adopted by the Texas Water Quality Board and the EPA.
With good communication between EPA and the State during the public hearing procedures,
the permits will contain the same limitations and the same sampling and reporting require-
ments.
These procedures have been completed for most refineries in Texas, and 1977
objectives are set. There remain questions concerning 1983 limitations. It is assumed
that final filtration will be a standard treatment step if the 1983 guidelines remain
unchanged. Additional biological treatment or physical-chemical treatment or some
combination will also be required. The technology is available, and considering the
excellent progress which most refiners have made in the last decade and the knowledge
accumulated during this period, I feel that the major problems in meeting the 1983 guide-
lines will be financial and not technical.
ENFORCEMENT
State enforcement action is initiated primarily by the district offices around the
State. In addition to regular inspections, all dates in permit compliance schedules are
entered into a computerized data system which is regularly checked. It is an understate-
ment to say that it is more difficult to achieve a continuous quality level with biological
treatment units than with other refinery unit operations. Recognizing this fact, the State
has not normally pursued enforcement actions for isolated violations of permits caused by
weather changes, etc. However, where repeated violations occur, enforcement action
will be taken. This frequently involves an enforcement hearing and subsequent recommen-
dations to the Texas Water Quality Board for remedial action. Flagrant violations are
normally referred to the Attorney General for prosecution either directly or following an
enforcement hearing. It remains the intent of the State to enforce State permit require-
ments .
-------
67
SUMMARY
A partial picture of the background and current procedures for the development of
permits for petroleum refineries has been presented. Most State permits were originally
issued 10-15 years ago and revised as necessary. Traditionally, the State of Texas has
relied on treatability studies to set limitations with the overall goal of "good secondary
treatment". Following the establishment of EPA, a working agreement was reached
whereby the State drafts most NpDES permits for Texas dischargers along with a parallel
State permit. Current permits are derived by combining information from waste load
evaluations, self-report data, existing State permits, NPDES applications and guidelines
and company input. Most refiners already have secondary treatment systems which will
meet or exceed 1977 NPDES requirements. In general, storm water handling and
chromium and zinc removal constitute the problems to be solved by 1977.
BIOGRAPHY
Robert F. Silvus is Branch Chief of the Industrial
Permits Branch, Central Operations Division , Texas
Water Quality Board. He received a B. S. degree
in Chemical Engineering and an MBA degree from
the University of Texas at Austin. He is a professional
engineer registered in the State of Texas. He had 15
years of industrial experience in Research and Developmenl
with Ehtyl Corporation and Jefferson Chemical Co., Inc.
before joining the Texas Water Quality Board.
-------
68
-------
PETROLEUM INDUSTRY EXPERIENCE WITH THE NPDES PERMIT SYSTEM
Harold F. Elkin
Coordinator, Environmental Affairs
Sun Oil Company, 1608 Walnut Street, Philadelphia, Pa. 19103
and
William K. Lorenz
Senior Environmental Engineer
Sun Oil Company, 1608 Walnut Street, Philadelphia, Pa. 19103
ABSTRACT
Sun Oil reviews its experiences with refinery NPDES permits. This discussion
emphasizes the special problems encountered with water quality limited streams, effluent
sampling, laboratory analysis, and self-reporting requirements. Recommendations are
made that may simplify negotiation and compliance with future NPDES permits.
INTRODUCTION
Most of the 250 U.S. petroleum refineries have now been issued final National
Pollutant Discharge Elimination System (NPDES) permits. Compliance and enforcement
are now actively underway. Over the last three years, the refining industry has had many
problems with the NPDES permit system and has hopefully learned many important lessons.
We are speaking with you this afternoon primarily on the basis of Sun Oil Company's
experience at our six domestic refineries. Sun has accepted final NPDES permits at all
six refineries. Our permits are based almost entirely on the EPA petroleum refining
effluent guidelines with certain more stringent conditions based on water quality-related
effluent standards.
We are planning to achieve our 1977 permit limitations through the use of conven-
tional treatment technology including sand filtration. At this time, we are not certain
that this technology can achieve the daily 24-hour maximum effluent limitations for all
parameters. Sun's largest refinery at Marcus Hook, Pennsylvania will join a regional
treatment plant in early 1977. We believe this experience is typical of the problems
encountered by most U.S. refiners. Many of these experiences are also relevant to the
marketing and production phases of the oil industry. The purpose of this paper is to dis-
cuss our most significant NPDES problems and to make some recommendations that may
simplify negotiation and compliance with future NPDES permits.
Several of the more vexing NPDES problems occur throughout all the stages of the
refinery NPDES permit process. For clarity, we will discuss these issues on a functional
basis, rather than chronologically. Table 1 lists the three subject headings we feel are
worthy of special discussion at this conference. We have excluded reference to the EPA
petroleum refining effluent guidelines, and will limit our discussion to the basic NPDES
problems which are somewhat independent of any specific set of effluent regulations.
Further, we assume our audience has some experience with the NPDES permit system so
69
-------
70
that we will not dwell on the procedural and legal aspects of negotiating and appealing
NPDES permits.
1. Water Quality Limited Streams
According to the Engineering-Science Report for the National Commission on
Water Quality, 68 refineries, having a cumulative crude capacity of 6,800,000
bbl/day, now discharge to water quality-limited streams (Table 2). Thus, over one-
third of U.S. refining capacity is partially, if not totally, controlled by water
quality standards, rather than the EPA technology-based effluent standards.
The impact of the water quality-related effluent limitations can range anywhere
from Best Practicable Control Technology Currently Available (BPCTCA) limitations
to zero discharge. Individual states are responsible for classifying specific river
segments as effluent quality-limited or water quality-limited on the basis of EPA-
approved state water quality standards.
Fora refinery discharging into a water quality-limited segment, permit
negotiations can be a complex and frustrating experience (Table 3).
First, it is necessary to establish the water quality standards upon which the
effluent limits are based. We have found that these standards are not always
available in specific form in the legally-promulgated state regulations. In some
states, the regulations are narrative with the specification of numerical water
quality standards up to the state agency. In the latter case, it is very important
to determine if the internal state agency decisions have been open to public comment
according to the relevant Administrative Procedures Act.
The states have employed a variety of methods for calculating the total
allowable effluent load that will achieve water quality standards on a particular
stream. These techniques range from sophisticated computer models to mere
educated guesses. No individual method is satisfactory for all circumstances. A
major stumbling block appears to be the unavailability of historical data. The
important fact is that the individual company must analyze the state methods in
sufficient detail to determine if there is an adequate basis for establishing its
effluent standards.
The next step is to allocate to all individual dischargers on the stream segment
a portion of the total allowable segment load determined by the mathematical model
or educated guesswork. This process opens up a complex set of economic equity
problems. Mathematical models are available to set load allocations in a variety
of ways. As you all know, it is very difficult to quantify the relative environmental
impacts of a variety of different effluents to an individual stream segment; thus, it
is almost impossible to objectively determine individual abatement schedules.
Nevertheless, these decisions are being made every day and all dischargers,
municipal and industrial, should play an active role in setting load allocations.
-------
71
Wafer quality-limited river segments will also be classified for the 1983
water qualify standards. We can expect that our next round of NPDES permits
will also include a significant number of refineries discharging to these river
segments.
2. Effluent Sampling and Laboratory Analysis
One of the most significant changes brought by the NPDES system is the
imposition of uniform requirements for measuring wastewater flows and the obtaining
and analyzing of representative effluent samples (Table 4). Prior to PL 92-500,
all Sun refineries had evolved their own monitoring programs based on local and
state regulations. However, some waste streams like tank farm storm runoff were
never measured or sampled before the 1972 Federal Water Act.
As you know, in most cases the interim and final refinery effluent limitations
are based on mass-weight (Ib/day) discharge rates. This requires that the flow
measurements, sampling techniques and analytical methods be of comparable
precision and accuracy. It makes no sense, whatsoever, to monitor a multi-million
dollar treatment facility with trained chemists using sophisticated analytical
instruments if the flow measurements are only within - 20% of the real flow rate.
Inadequate historical flow data is one of the main reasons that our refineries
have had some non-compliance incidences with interim permit conditions. Such
problems stem from both the lack of the quality and quantify of historical flow data.
Hopefully, these problems will not reoccur since we are now establishing flow
measurement systems that should provide an adequate amount of accurate and
precise flow data. For example, our Marcus Hook, Pennsylvania refinery will
join a large municipal-industrial regional treatment system that requires us to
install a flow meter accurate to within _+ 1%.
Prior to PL92-500, our refinery practices were about evenly split between
grab and composite samples. We are now running more composite samples as
required by,NPDES permits. However, we expect to continue faking grab samples
on segregated storm runoff. In developing our specifications on sampling methods
and sample preservation, we have relied on the guidance of the EPA technology
transfer handbook entitled, "Monitoring Industrial Wasfewater." This reference
is quite complete and is an excellent source of information on sampling equipment
and flow measuring devices.
Laboratory testing is rapidly becoming the most critical part of the NPDES
compliance monitoring system.
As an example, we are all spending a lot of money to build sophisticated
treatment systems. We measure our performance by laboratory results. But what
does it all mean if another laboratory analyzes an identical sample and obtains
results five or ten times higher than our own? This is a very significant problem
-------
72
if the second laboratory happens to be an EPA laboratory and the samples are being
analyzed as part of an EPA annual site inspection.
We have only begun to look into the areas of analytical precision, analytical
accuracy, and interlaboratory variability. Our fundamental interest is in producing
reliable information that is legally defensible in administrative and judicial
proceedings. If we really do have a non-compliance situation, we need accurate
data to identify and solve the problem. On the other hand, if we are in compliance
with the effluent limitation, we must have accurate data for use in any governmental
enforcement action. In the long run, we need accurate and precise effluent data so
that the 1983 Best Available Technology Economically Achievable (BATEA) effluent
standards better reflect reality.
By way of example, the State of Texas is now implementing an analytical
quality control program for use in NPDES compliance monitoring. Once this
program is operational, the Sate will back up its effluent sampling results with
information on the precision and accuracy of the test method. If an industrial
discharger wishes to challenge the State analytical results, he will be at a distinct
disadvantage without comparable precision and accuracy data on his own laboratory.
These quality control programs are expensive and time-consuming to operate, but
they may be necessary in order to maintain a balance in the enforcement of NPDES
pe rm i ts.
3. Enforcement of Self-Reporting Requirements
As we all know, the bases for the enforcement of NPDES permits are the many
self-reporting requirements that the permittee must satisfy. Table 5 shows the major
headings of a typical NPDES permit issued in EPA Region VI. The format may vary
throughout the country, but the main subjects are included in all NPDES permits.
Part 1, Section A details the specific effluent limitations and monitoring
requirements, but does not stipulate the form or frequency of any self-monitoring
reports.
Part 1, Section B sets forth the specific timetables for implementing the limit-
ations in Part 1, Section A. These timetables require progress reports from design
of facilities through the attainment of operational levels. We have been successful
in negotiating reasonable compliance schedules, but still recognize that the interim
dates are only educated guesses based on very little prior experience in designing
and building refinery wastewater treatment facilities. As long as the final deadline
is satisfied, we believe that the interim dates should have some flexibility. The EPA
or the states cannot now agree to interim date modifications without formally
modifying the permit including proper public notice. We suggest that some red
tape could be cut by a revision in EPA's regulations to allow them some internal
flexibility in modifying interim compliance dates as long as final deadlines can be
achieved.
-------
73
Part1 1, Section C covers the details of analytical test procedures, reporting
of results and retention of records. Under this section, the report of all
analytical results is prepared for EPA or the states every three months. The
regulatory agencies can seek enforcement actions against the permittee on the
basis of the quarterly report.
The procedures and reports of this section do not reflect the accuracy and
precision of the compliance monitoring process nor the problem of interlaboratory
variability. We know that the EPA effluent guidelines attempt to reflect the
variability of wastewater effluents. However, we are not satisfied with the EPA
methodology. The total variability of a wastewater effluent must be broken down
into its component parts. Basically, we need to differentiate between the
controllable vs. uncontrollable sources of variability. Then, a compliance
monitoring system could be developed that 1) does not try to limit the uncontrollable
variability and 2) does attempt to limit the controllable variability sources in a
probabilistic manner. We are still learning about the variability issue and have
no specific systems to offer at this time. But, we are convinced that the existing
data base is not adequate.
Part II, Section A includes a subheading on non-compliance notification.
Within five days that a permittee becomes aware of exceeding any daily maximum
effluent limitation, the permittee must notify the agency. This notification must
describe the problem and specify what, if any, remedial actions are being taken.
We consider this requirement to be an example of unnecessary red tape. Further,
we have had some difficulties in implementing the five-day requirement, especially
when commercial laboratories are performing the analyses. After all, it is our
legal responsibility to achieve both the daily maximum and monthly average
effluent limitations. Why should we have to take special steps on the daily
maximum limitations and not on the monthly average limitations? Each daily
non-compliance is a separate violation subject to considerable fines and adverse
publicity. We believe that we can manage our wastewaters without the added
burden of this daily maximum non-compliance notification. Our recommendation
is to delete this provision from NPDES permits.
The remaining sections of the NPDES permits do not include any specific
self-reporting requirements that are worthy of specific mention in this paper.
CONCLUSIONS
This concludes our prepared remarks on experiences with petroleum refinery NPDES
permits. To sum up:
I. Over one-third of all U.S. refining capacity discharges to water
quality-limited streams. Special considerations must be taken in
order to negotiate an optimum set of permit conditions. Water
quality-limited discharges will also be an important factor in the
1983 permit negotiations.
-------
74
2. Effluent sampling and laboratory analytical procedures are now becoming
more standardized and more sophisticated. All phases of the compliance
monitoring system must be of comparable precision and accuracy.
Analytical quality control programs and round-robin testing may become
a necessary part of the NPDES permit process.
3. Self-reporting is the primary basis for enforcing NPDES permits. Most of
the existing requirements are necessary, but some improvements can be made
to cut red tape and improve the identification of an actual non-compliance
incident.
We believe that our experiences are typical of other refiners and also relate to
problems that will be encountered by the marketing and production activities of the oil
industry. We have purposely not dwelled on the EPA effluent guideline regulations
believing that these criticisms are being handled in a separate forum.
BIOGRAPHIES
William K. Lorenz is Senior Environmental
Engineer in the Environmental Affairs Department
of Sun Oil Company. Bill holds the following degrees:
Bachelor of Chemical Engineering from Villanova
University, Master of Science in Environmental
Engineering from Drexel University, and expects to
receive a Masters in Business Administration from
Drexel University in 1977.
Harold F. Elkin is Coordinator of Environmental
Affairs for Sun Oil Company in Philadelphia. He
holds the following degrees: B.S. in Chemistry,
University of Pennsylvania; B.S. Engineering,
Drexel University; M.S. Environmental Engineering,
University of North Carolina. He is a professional
engineer registered in Pennsylvania.
Mr. Elkin has served on the Pennsylvania Air
Pollution Commission, advisory panels to EPA, and
numerous industry environmental committees.
-------
75
TABLE 1 SELECTED PROBLEM AREAS IN NPDES PERMITS
1 - Wafer Qualify Limited Segments
11 - Sampling and Laboratory Analysis
111 - Enforcement of Self-Reporting Requirements
TABLE 2 DISCHARGE LOCATION OF REFINERIES IN THE U.S.
Crude Throughput
Type of Effluent Limitation Number of Refineries 1,000 bbIs/Stream Day
Effluent Quality Limited 106 5,729
Water Quality Limited 68 6,954
Municipal System 34 1,720
Zero Discharge 38 615
TOTALS 246 14,918
Source: Engineering Science, "Petroleum Refining Industry Technology and Costs of Wastewater
Control" Table II -- 6, page 11-25
TABLE 3 WATER QUALITY BASED EFFLUENT LIMITATIONS
o Applicable State Water Quality Standards
o "Modeling" Techniques
- Computerized River Model Studies
- Manuel Computation River Model Studies
- Educated Estimates
o Individual Load Allocations
TABLE 4 EFFLUENT SAMPLING AND ANALYSIS
o Lack of Historical Data
o Flow Measurements
o Analytical Quality Control
o Interlaboratory Variability
o Legally Defensible Monitoring Programs
TABLES OUTLINE OF NPDES PERMITS
EXAMPLE: EPA REGION VI
PART I A - Effluent Limitations and Monitoring Requirements
B - Schedule of Compliance
C - Monitoring and Reporting
PART II A - Management Requirements
B - Responsibilities
PART III - OTHER REQUIREMENTS
-------
76
-------
"THE ENFORCEMENT OF PERMITS"
Harless R. Benthul
Chief, Legal Branch, Enforcement Division
U. S. E.P.A., Dallas, Texas
Before I get into the substance of my discussion I would like to extend personal
thanks to Bob Silvus. He did something in his talk that I have been waiting for for some
time, and that is Bob acknowledged that two people, non-lawyers presumably, can look
at the same piece of information or the same data and reach different conclusions. I had
heard since I got a law license that it was unique to attorneys that we could look at the
same piece of English language and forever argue two different conclusions, but I am
grateful to Bob that he has made progress for us, who knows tomorrow we may be con-
sidered people as well. I am going to talk a little bit about enforcement of federally
issued NPDES permits. Federally issued because in the Region where I work, Texas and
the four bordering states, we don't yet have a state which has been formally delegated
NPDES permit authority, although we do work with those states on various cooperative
bases. Nevertheless my discussion will be on enforcement of federally issued permits.
We are now just over three years from the enactment of public law 92-500. Many of
the deadlines of the Act have come and gone. Others are yet before us - you as the
operators of regulated facilities, we as EPA the regulatory agency. Two deadlines are of
specific interest today. They are first, the December 31, 1974 moratorium expiration
date which became the target by which all major permits (and some others) were to have
been issued and second the July 1, 1977 requirement for attainment of best practicable
control technology currently available (BPT).
The December 31, 1974 date saw the issuance of virtually all permits for major dis-
chargers in Region VI and EPA nationally. What is a major industrial discharger is a
matter of evolution. It has included generally large production facilities such as
refineries, chemical plants, certain power plants, metal processors and others, not based
solely on size or diversity of waste, but on the strength or toxicity of waste or effect on
the receiving stream. Those to be added to the major permit list in the future will
probably be based on considerations such as potential harm to the receiving stream,
significance of pollution abatement facilities, extent of hazardous material handled and
the like. Finally, it should be remembered that designation of major permits is in large
part a label for purposes of allocation of our resources.
The number of such industrial permits in Region VI, is now about 340. It should be
noted that the claim of having issued those permits rests in part on the procedural system
which allows uncontested portions of permits to become effective notwithstanding the
fact that an adjudicatory hearing is pending on other portions of the permit. Depending
upon the day of the count, the number of such pending adjudicatory hearings is around
100 in Region VI and about 1100 nationally.
Parts of almost all of those 100 contested permits, all the other 200 plus majors, as
well as many minor industrial and some municipal permits are effective. They have been
77
-------
78
effective for a year or more in many cases. As you are aware all of these industrial
permits require the discharger to attain effluent limits equal to BPT by the statutory date
of July 1, 1977 or other more stringent limitations established pursuant to Section
301(b)(l)C based on water quality standards or other federal or state law. BPT may be
defined by effluent guidelines promulgated pursuant to Section 304 or in the absence of
guidelines on a case by case basis by application of professional engineering judgment.
As is well known to this particular group, many effluent guidelines remain in liti-
gation. Yet if one can detach himself from this hopefully temporary delay in our joint
implementation of the Act, it is clear that if the July 1, 1977 mandate is to be realized,
one thing that must be done is to uphold the credibility of the issued permits.
Parenthetically, I would point out that EPA administrator Russel Train recently in a
case involving Bethlehem Steel, concluded that the July 1, 1977 date is a mandatory
date and permits must require permitees to meet it. This position upheld a previous
decision of the general counsel. That was No. 26 if you want to look at it or have your
attorneys look into it. Enforcement of these permits is an element of maintaining that
credibility and enforcement is well under way.
The principal legal basis for enforcement of NPDES permits is generally Section 309
of the Act. I am not going to bore you by reading that whole page, I will read a
section of it of particular importance. Potentially available also for direct enforcement
of permits is Section 504. Also available for collateral support of a municipal permit is
Section 402(h) by which injunctive relief is available to prohibit additional pollutant
load being introduced where there is an existing violation of that municipal permit.
However for most violations of NPDES permits, Section 309 will be the primary
enforcement tool. Since we are here concerned with federally issued permits the parts
of Section 309 of most immediate interest are Sections 309(a)3, 309(a)4 and 309(b),(c)
and (d). This results from the fact that Sections 309(a)l and (2) are concerned with
state-issued permits and 309(e) deals with municipal permits. I will not in this discussion
deal with enforcement of true municipal permits where there are violations caused by
industrial or other non-domestic waste contributors. There are other classes of permit
holders which claim some attributes of municipalities as defined in the Act, but treat
almost exclusively industrial waste. These applicants have been in our Region and will
continue to be dealt with on the basis that each industrial contributor is a party to the
permit and responsible directly to EPA along with the applicant for performance of the
permit. These permits are not generally addressed here either. Thus we will be dealing
with enforcement of single-party, federally-issued permits for industrial facilities.
Section 309(a)3 reads as follows,
"Whenever on the basis of any information available to him, the administrator
finds that any person is in violation of Section 301, 302, 306, 307 or 308 of
this Act, or is in violation of any permit condition or limitation implementing
any of such sections in a permit issued under Section 402 of this Act by him
or by state, he shall issue an order requiring such person to comply with such
-------
79
section or requirement, or he shall bring a civil action in accordance with sub-
section (b) of the Act. "
A few remarks are in order about word usage in this section. First, the word "any"
appears twice in the first two lines preceding both information and person. This would
appear to be all inclusive and cover self-reported, agency-discovered and citizen-
reported information (information relating to violations). Then note that the language
encompasses both violations of certain enumerated sections of the Act and violations of
permits which implement those sections. Then follows the directive to the administrator.
He sha 11 issue an order requiring compliance or he shall commence a civil action in
accordance with Section 309(b) (emphasis added). I will touch later on this directive
language. The authority to so act has been delegated to regional administrators.
Section 309(a)4 provides for sending a copy of these administrative orders to states; it
provides that service of orders shall be by personal service, that they shall state with
reasonable specificity, the nature of the violation and it also puts an outside limit of 30
days on the time for compliance. This section also creates a right on the part of the
recipient of the order, to a conference as a condition to effectiveness of the order
relating to a violation of Section 308. Section 308 deals with monitoring, reporting,
sampling, inspection and the like. Parenthetically I would say that in orders we have
issued dealing with monitoring requirements of the permit or violation of monitoring
requirements we have been very careful, we think, to be sure to post date the effective-
ness of those orders and to explicitly advise people that they had a right to a conference
before the order became effective.
Section 309(b) authorizes civil actions for what is called "appropriate relief",
including (but not limited presumably to) injunctive relief for any violation for which an
order could have been issued under Section 309(a)3 (that is the section I read just a
minute ago). Section 309(b) also establishes venue, that is where the suit is to be filed
for such proceedings and requires immediate notification of the appropriate state.
Section 309(c)1 contains the criminal sanction for violation of enumerated sub-
stantive provisions of the Act or a permit implementing them. The criteria for Section
309(c)l criminal conduct is "wilful or negligent". Liabilities of $2500 to $25,000 or up
to one year in prison or conceivably both, per day of violation are imposed and the
limits are double for a second offense.
Section 309(c)2 makes criminal a knowing false statement, representation or certifi-
cation of certain documents filed or required to be maintained under the Act. Tampering
with monitoring equipment is a criminal offense. Penalties range up to $10,000 or six
months imprisonment or both.
Section 309(c)3 adds for the purpose of 309(c), any responsible corporate officer to
the general definition of the term 'person' as contained in Section 502(5). 502(5) defines
person generally for purposes of the Act. 309(c)3 adds to that definition for the purposes
of 309(c), the term "responsible corporate officer".
Finally, Section 309(d) creates a civil penalty of up to $10,000 per day of violation
-------
80
of the same sections listed in 309(a)3 or a permit condition implementing them. On its
face this liability, the civil penalty liability, is strict. That is, it applies without the
necessity of proof of fault on the part of the violator. The question of how many
penalties are assessable for multiple violations in one day has been addressed so far by
only one court that I am aware of. That was the Detrex case which indicates a $10,000
penalty per day only may be assessed. That is, not several $10,000 penalties in the
same day for multiple violations and that is based on a comparison of language in Section
309 and Section 311 which is the oil-and-hazardous-material section. One case
addressed the point and that was the result. In my opinion there will be more heard on
that point as to whether more than one $10,000 penalty could be assessed for more than
one violation - for instance, for violation of two different effluent parameters.
Let me talk now about the mandatory aspects of Section 309(a)3 which was again
the language that I read for you out of the statute a few minutes ago. I have referred
previously to Section 309(a)3 as directive in nature. This is because it says that the
Administrator shall, and I emphasize shall, issue an order or commence a civil action
upon the finding that any person is in violation of a permit as well as in violation of the
enumerated sections of the Act. Literally read, there is no discretion except as to
which remedy to exercise - issue an order or commence a civil action.
To date we have had to exercise some discretion because there have been more
violations (of various levels of seriousness I should add) than we have had resources with
which to develop orders or prepare referrals or court actions. Consequently we have
tried to issue orders and commence civil actions in cases where we felt it most important
to do so. We have noticed others by warning letter. Notwithstanding resource limits on
the part of EPA and the States for that matter, the language of Section 309(a)3 remains
what it is.
This language using the term "shall" is pretty clearly mandatory upon the Admin-
istrator once there is a finding of a violation. The requirement of 309(a)3 is qualified
elsewhere in Section 309 only if a state agency has been delegated NPDES authority by
the provisions of Section 309(a)l. Now if you refer to Section 309(a)l, it suggests the
argument that the only circumstances where the Administrator has an alternative, and I
will come back to that in a minute, to issuing an order or bringing a civil action under
309(a)3, is where a state has been delegated NPDES authority. However the EPA may
be free to seek civil or criminal penalties in addition to the order or civil action con-
templated by 309(a)3. In any event if we are operating under 309(c)l, where a state
has authority to issue permits under this law, even then the alternative to issue an order
or commence a civil action appears to me to be limited to a 30 day waiting period after
giving notice to a state of a violation to see if the state does or does not commence
appropriate action. This option is not available where, as here, we are dealing with
federally issued permits. Now the word alternative that I used, was used advisedly be-
cause of language in another section of the Act, 505(a) of the Water Pollution Control
Act, which authorizes citizens suits to compel the administrator to perform any act or
duty which is not discretionary, and I will risk boring you by reading part of section 505.
It reads - "except as provided in sub-section (b) of this section, any citizen may
commence a civil action on his own behalf against any person including the United
-------
81
States and any other governmental instrumentality or agency, to the extent permitted by
the eleventh amendment to the constitution, who is alleged to be in violation of a) an
effluent standard or limitation under this Act or b) an order issued by the Administrator
or state with respect to such standard or limitation, or against the Administrator where
there is alleged a failure of the Administrator to perform any act or duty under this Act
which is not discretionary with the Administrator". Then it goes not to say that federal
district courts will have jurisdiction without regard to the amount in controversy or
citizenship of the parties to enforce such an effluent standard limitation or such an order
or to order the administrator to perform such act or duty as the case may be herein to
apply any appropriate civil penalties under Section 309(d). Well, so what does all that
legalese get for us? Referring back to 309(a)3, which is the part you will recall that
said that on the finding of a violation by any person the administrator shall issue an order
or commence a civil action, a compelling argument can be made that once a finding of
a violation has been made on the basis of any information available by the Administrator
in a pre-NPDES delegation situation, he has no discretion but to issue an order or
commence a civil action per 309(a)3 and that such duty is enforceable by a citizen
through Section 505(a).
This argument is very similar to the result of a case that was decided under the Clean
Air Act called Wisconsin Environmental Decade versus Wisconsin Power and Light et al;
one of the et als being the administrator of the EPA. This was in the U.S. District Court
for the western district of Wisconsin and decided in 1975. You will pass it on to your
attorneys that case is printed in volume 7, Environmental Reporter Cases, page 2022.
Here the court construed part of Section 113(a) of the Clean Air Act an conferring a non-
discretionary duty on the Administrator to issue a notice of ciolation, which duty is
enforceable under citizen suit provisions of the Clean Air Act. Now I am going to bore
you again with some reading, which I think you will find interesting. This is Section
113(a)l of the Clean Air Act.
"Whenever on the basis of any information available to him, the administrator
finds that any person is in violation of any requirement of an applicable
implementation plan, the administrator shall notify the person in violation of
the plan and the state in which the plan applies of such finding. If such
violation extends beyond the 30th day after the date of the administrator's
notification, the administrator may (emphasis added) issue an order requiring
such person to comply with the requirements of the plan or bring a civil action
in accordance with sub-section (b).11
The "may" that I read refers to what can be done after the 30 days expires. "Shall"
was back up earlier - whenever on the basis of any information the administrator finds
that any persons in violation he shall notify the person in violation. It reads very much
in words like 309(a)3 in the water act which we have been reading. The citizen suit
provisions of the Clean Air Act are very very similar to Section 505(a) of the Water Act
which I read to you. The court in the Wisconsin case also indicates that the act of the
Administrator 30 days after the notification is discretionary and that the initial finding
which generates the notice of violation must be a good faith effort based on investigation
of any evidence available. By analogy, and I emphasize by analogy, similar criteria
-------
82
could apply to a founding of a violation in connection with Section 309(a)3 orders or
civil actions. The case is also interesting in the fact that the violation was on alleged
violation of a state implementation plan as compared with an NPDES permit. The court
holds that the EPA may, in determining whether a violation has occurred which requires
that a notice of violation be issued, defer to the state's interpretation of its plan as long
as the state's interpretation is reasonable and consistent with the Act.
Thus there appears to be a compelling conclusion, if not inescapable, on the basis
of this analysis and that case that I have cited, that the 309(a)3 requirement to issue an
order or commence a civil action is non-discretionary once a violation is found to exist
and that that duty is enforceable by citizen suit through Section 505. In checking this
last week, I discovered the existence of one unreported case in a Federal Court in the
northern district of Alabama which reached a contrary result on whether this is a dis-
cretionary or non-discretionary duty. I would caution that the result was reached on a
preliminary motion and that the case is on appeal. It has not turned up in the reporters
yet to my knowledge.
Now this is not to suggest that issuing an order or commencing a civil action such as
an injunction under 309(a)3 is an exclusive remedy. We believe, and we have operated
on the premise that the civil penalty and criminal sanctions of 309(c) and (d) compliment,
in appropriate cases, Section 309(a)3. There has been one case that held that the
309(a)3 remedies are not exclusive. That is a case which was decided, I believe, in the
District Court for the district of Arizona where it was claimed that EPA had issued an
administrative order and therefore was not entitled to collect penalties for violations that
were covered by the administrative order. The judge said that's not right, the two go
together.
Finally, it should be noted in the face of all this grimness that there may in fact be,
and may in law be, a place in the execution of Section 309 for what has traditionally
been called prosecutorial discretion. However, the extent to which this may be
engrafted onto the Act when Congress used the language it did presumably being aware
of the concept of prosecutorial discretion remains to be seen. And. also if there is dis-
cretion it may lie in the area of the finding of a violation. I think this is sort of hinted
at in the rational behind the Wisconsin case that I mentioned. Time and experience will
tell. This doesn't mean that we, EPA, are going to issue administrative orders on trivial
violations and it doesn't mean we are going to refer bad cases to the U.S. attorneys,
but that we see our enforcement responsibilities as rather explicit and rather demanding.
I would like to conclude by making a few observations on some of our experience to date
in enforcement of permits. We have been attempting to enforce these permits in Region
VI, and I think nationally, as long as we have had them issued. During fiscal year 1975,
approximately 65 administrative orders were issued to a variety of dischargers for various
violations. Beginning in fiscal year 1976 or July 1975, a greater emphasis on enforce-
ment of permits has taken place. During the first six months of fiscal year 1976, that is
July to December 1975 inclusive, we have issued 48 administrative orders involving 38
violations of effluent limits, 13 violations of reporting requirements and 6 violations of
schedule of compliance. During the same period there were a total of 4 referrals to
U.S. attorneys for prosecution. The prosecutions have all been for civil penalty
-------
83
collections and one case for injunctive relief involving a municipality. The referrals
involve for the most part, violations of effluent limitations or serious bypassing, that is
running waste around the treatment plant, and we have recommended collection of civil
penalties for each day in violation.
It is worthy of observation that a problem keeps recurring which we are trying to
discourage by making referrals as it comes up in appropriate cases. The pattern is that a
difficulty of some nature will arise within a production facility or perhaps independent of
production, but in any event the facility operator is faced with a choice of whether to
discharge untreated or partially treated waste in violation of his permit or curtail
production. In the instances we have had occasion to carefully examine so far, we have
concluded that a referral was in order. I should say in some of the cases, not all of the
cases. The point we are trying to make is that if there is a choice of violating a permit
and keeping production going, then that choice is in all likelihood going to have a price
attached to it, and that concludes my prepared remarks. I will attempt to answer
questions if anyone has any.
DISCUSSION
Milton Beychok: I have a question and I am not sure whether it should be addressed to
you or one or more of the previous speakers. We have heard a great deal this afternoon
about compliance schedules for industry, violations by industry and enforcement of
sanctions against industry, be they discretionary or non discretionary. I would like to put
the shoe on the other foot. There is one part of PL 92-500 that places what is in my
opinion the mandatory responsibility upon the EPA and it says that before granting a new
source permit the EPA must develop an environmental impact statement. So I have two
questions, how many new source permits has the EPA granted, and how many environ-
mental impact statements have they written?
Harless Benthul: In Region VI, we have granted, I want to say a handful of new source
permits and by that I think and I really honestly don't know, I think the number is on the
order of maybe 10 or perhaps less or perhaps a few more. How many impact statements
have we prepared? I don't believe we have published any impact statements Mr. Beychok.
We have asked for and have received environmental assessments in at least some of those.
In one or two cases that come to mind, we have received an impact statement for review
which has been prepared before the facility owners got to us because they had to get a
federal license or permit earlier in the game. Now if I may, let me add one other point.
The requirement is as I recall not that self executing that automatically-if it's a new
source, there must be an impact statement. As I recall 511(c), it says that only two
things are "major federal actions" under this law, one of which is issuance of new source
permits. Well given that it is a major federal action because it is a new source permit
then there may or may not be a requirement for an impact statement depending on the
various criteria that are contained in NEPA and in the cases that have developed under
NEPA. Does that answer your question?
Milton Beychok: Well I understand that you could have a negative declaration and get
around the EIS and that is what you are really saying that all it required was that NEPA
-------
84
apply. Are you saying then that out of the ten new source permits, if there were no
impact statements written, that in each case a negative declaration was filed. And
simply telling me that you asked for one doesn't answer the question because it is EPA
who must prepare that statement.
Harless Benthul: Your latter statement is correct. We do have responsibility for pre-
paring the impact statement. We have asked for and received the assessments in those
cases. What I was trying to say to you was in at least some of those, and don't hold me
to the ten, that is a matter of memory and I honestly don't know if there were ten or less
or more, but in some of those, by the time the people got to us, there had already been
an impact statement prepared as a result of their dealing with some other agency,
before they got to us, and some other agency had to do something like granting a license
of one kind or another which they felt like required an impact statement.
Milton Beychok: Just one further question, does PL 92-500 spell out any sanctions
against the EPA, should they not follow Section 511 properly?
Harless Benthul: Well we are back to 505; if need be. The sanctions I think are there
in abundance under NEPA, if we take a major federal action significantly affecting the
human environment and don't prepare the 102(2)C statement its "back to go" for us.
Whether 505(a) of the Water Pollution Control Act applies I honestly don't know. I
would have to give some consideration to whether or to what extent preparing an impact
statement is discretionary and again, horseback, off the top of my head, I would say
there is a good likelihood it is not discretionary, that if a court found that we issued a
new source permit to a facility which was a major federal action significantly affecting
the human environment and we hadn't prepared an impact statement, I would say the
judge would say go back and prepare it.
Frank Manning: You have mentioned many times in your talks when a violation occurs.
Is this always cut and dried, doesn't a violation occurring depend upon a measurement
and analysis? Don't you find cases where one analyst might say yes it's over a guideline,
another might say it's under, in sort of a case like this. One, do these kind of cases
occur and if they do occur, what do you do then?
Harless Benthul: The answer is that yes both on the self-reporting data and as a result
of our own inspections we turn up apparent violations which are to use an electronics
term "in the noise level", either because it's such a small percentage over the limit or
because it's within the accuracy of the analytical technique or in another area if a
schedule of compliance date is missed by a week or two weeks or a month, technically
that is a violation. As a practical matter, going back to what 1 said earlier, we try not
to issue administrative orders on relatively trivial violations and we hopefully are not
going to send any bad cases to the U.S. attorney. So again as a practical matter there
is some discretion exercised. Does that answer your question?
Umesh Mathur: I noticed in your presentation that you stayed away from the whole
area of pretreatment, particularly for toxic materials. Obviously the NPDES permit
system applies only to a direct discharger to the navigatable waters. And so if somebody
-------
85
was to say, Well if 1 am going to have to discharge to the river then obviously I am going
to have to get a permit from EPA, so why don't I just discharge into the sanitary sewer
system because there is no regulation that covers that, except EPA's 40CFRI28 which is
really saying nothing, because 40CFRI28 does not specify concentration levels or any of
the mass loadings permissible per unit of production, and so this area has been very grey
and so what I would like to point out here is that the construction grants program under
Title 2 has been running around all over the country requiring municipalities that are
going to get federal funds for building facilities to pass ordinances that will supposedly
control these pollutants. Now the problem of that is that there is no uniformity in these
various and sundry ordinances that have emerged and as a result of that nobody in industry
that is using sanitary sewers for discharging what are called incompatible or toxic
pollutants depending upon what you choose, has any idea as to what those numbers are
going to say eventually when and if EPA comes around to defining pretreatment standards
for these materials. Now the difficulty that I am trying to point out here is very real
because once an industry was given an ordinance which stayed in effect for two years, if
EPA took two years to determine what pretreatment they should require, and this industry
went in and built pollution control facilities to meet both standards that have been
established on a completely arbitrary manner in the face of any definitive guidance from
EPA, then for the municipality to go back to the industry and say look now that EPA has
come out with a new set of numbers you are going to have to shoot for this new set of
numbers; I feel is going to cause a management problem for the municipalities involved.
I was wondering why the construction grant problem cannot be restrained and told not to
require facilities that are going to require federal funds to necessarily have the industrial
waste pretreatment ordinance in the face of the fact that EPA itself is under suit to
produce those pretreatment guidelines and they have failed to do so for I think they are
already two years too late, and they have still not produced and so what is the point of
forcing municipalities to do it.
Harless Benthul: Well thank you for raising a very pertinent point in this whole dis-
cussion"!I can't very reliably answer your question as it relates directly to the con-
struction grant program. Hopefully, I can give you some reassurance with respect to the
rest of it. The pretreatment regulations generally required that as of the end of the three
year period which I think about upon us in the order of a year or so, contributors of
incompatible waste or waste which would pass through or interfere with the operation of
the treatment works, have some pretreatment satisfactory to prevent that happening
implemented. Now you are quite correct that that is not implemented is a general rule or
any specific basis, that is in terms of limitations in either concentrations or flow or pounds
per day from the industrial contributor to the municipal treatment facility. Nevertheless
the regulation is there, it has been on the books for a period long enough that we feel
like it is about ready to be implementing. Coincidentally, we had a crew investigating
a municipality in an industrial facility last week which fits this circumstance, and there
very well may be some enforcement activity resulting from that. It doesn't by the way
concern anybody in this room. It is not a petroleum or refinery operation. The rest of
the reassurance is, I hope, that is something I mentioned earlier in my talk, and that is
that category of industry permit applicants which has some attributes of a municipality
under the Act, that is they are a creature of statute under state law, they have bond-
issuing powers and the like, but treat totally if not largely, industrial waste. We are
-------
86
permitting those applicants on the basis that the contributors of that waste is a party to
that permit and is therefore directly responsible to us for a violation of that permit. My
answer to you is not specific as it relates to the grant program. What it is, is that we
think we have some enforcement power on pretreatment by virtue of the regulation which
you mentioned and by virtue of the reference in Section 309(a)3 to violations of that
portion of the statute that requires pretreatment. So I would have to agree that pretreat-
ment has probably not been implemented on a specific uniform basis to the degree that
you would like or that EPA would like, but we do recognize that there is a problem and
we try to deal with it to some degree at least.
Pete Hoffman: The question I have deals with those permits which have not been given
by the EPA, on applications that were filed say back in the original Corps of Engineers
which I think was July 1971 . If during the period of time in which this permit is pending
conditions change significantly, on that discharge, what is the burden upon the operator
to file a revised permit or revised application on changing conditions on a permit that
has never been issued in the first place and secondly, is it not true what you were saying
that if a citizen wants to they could file suit and force the EPA to fine that company for
not having a permit?
Harless Benthul: I will answer your last question first. Yes it is possible. Again the
citizen has to prove that or has to show, make a prima facie showing that the duty
involved is not discretionary and that we haven't performed it. You are describing a
facility for which there was an application filed under the Refuse Act Permit Program
and which has never been permitted?
Pete Hoffman: Yes
Harless Benthul: I didn't think there were any of those left around Mr. Hoffman.
Alright, you filed the application timely, I assume, and you say there have been some
changed conditions. Have you notified us of those changed conditions? Therefore, if
we were to proceed to issue a permit tomorrow, we would be issuing a permit based on
facts which are not current. Would that be a fair assumption?
Pete Hoffman: Yes, Sir
Harless Benthul: Mr. Hoffman, I would have sworn that we have issued permits to all
your facilities, but obviously we haven't. I would say that, and had very good
experience with you in doing it I would add, if there have been changed circumstances
i think I would make a record of it if nothing else by letter to the Regional Office and
at least by doing so you have put us on notice that the facts are different than we
apparently think they are. Someone else was going to help me and far be it from me to
turn down any help. Whoever was going to Mr. Hoffman I would like a little expansion
on that last thinking.
Bob Funk: We have a similar situation to Mr. Hoffman. Being specific we have a
number of offshore production facilities which have permits filed which have never
received a permit from Region VI and these kind of facilities are subject to change as
-------
87
time goes on due to the nature of the discharge. You mentioned a letter to inform EPA
of what the change is, what level of change would trigger a letter, this is our real
problem.
Harless Benthul; While I doubt you are ready for this, what I am referring to is a major
change, significant change and those are only words, I realize. You are talking about
offshore facilities, you are absolutely right you haven't got any permits for those and
Mr. Hoffman are you talking about minor industrial facilities, not what we would probably
call a major permit? You are right there is a bunch of applications lying around for
minor, what we have in the past called minor discharges, and they have not been issued.
Several have, on the order of 1500-2000 in Region VI total. We will get to you sooner
or later and again, if there has been a change in circumstances, again for the like of a
better phrase, major, significant and I know that is not much help, but those you should
notify us.
I am delighted to hear it and thank you.
Bob Funk: During your discussion you cited a court case limiting penalties for violations
to $10,000 per day regardless of the number of violations and you mentioned that you felt
like there would be further challenges to this interpretation. Can I draw the conclusion
that Region VI would contemplate requesting penalties above $10,000 a day for multiple
violations?
Harless Benthul: Yes, you can draw from that we think we will encounter some cases
where it is appropriate to ask for a penalty, conceivably $10,000 per day for more than
one violation which results in conceivably $20-$30-$40,000 per day violation. I don't
believe we have done that yet. I don't think we have asked for more than $10,000 per
day, but it has been because we haven't run across a case where we think it fits. I do
believe though that there will be cases where it is appropriate even within the language
of the judge in that Detrex case to ask for penalties, multiple penalties, for multiple
violations on the same day.
Morris Wiley: I have a question I think that concerns many of the audience here about
the pretreatment requirement. EPA in publishing the guidelines for municipal secondary
treatment has included fewer parameters than they have in some of the other industrial
guidelines. For example, COD is not included in municipal secondary treatment but it is
included in the refinery guidelines. Now if the refinery or other industrial plant dis-
charged a pollutant to a municipal treatment plant which is required by its guidelines to
treat for implementation of the 77 standards, but is not included in the municipal
monitoring program, what would be the view of the pretreatment requirements which might
be required of that discharger?
Harless Benthul: Sounds like it might be a case where it would be in order to put a COD
limit on the municipal permit.
Morris Wiley: Is that being done at all?
-------
Harless Benthul: It is being done in those, that class of cases I mentioned earlier which
are sort of quasi municipal type permits. I don't know that it is being done in ordinary
municipal permits. We have a case pending in one of the states in this Region, where
that is precisely the issue. We have a court decree from a couple of years back that
imposed a requirement on a municipality and us as parties to the settlement, that there
will be certain pretreatments of metals and influence limits on metals coming into the
treatment plant and it also is a requirement that there be effluent limits on those same
metals which, of course, would be significantly above what you normally see in a
municipal permit as you are aware. And that one is still up in the air, I don't know
where it is going to wind up.
Morris Wiley: Metals would be a good case in point because the refinery effluent
guidelines do place a limitation on chromium and it is well known that this is removed
by either a municipal or industrial primary and secondary treatment and yet I don't
think there is any basis for evaluating what chromium limit municipalities should dis-
charge in a permit.
Harless Benthul: Well I wouldn't place the greatest emphasis on the fact that one set
of guidelines requires controls of COD and chrome and lead and another one might not.
Those as you undoubtedly know, vary from sub-category to sub-category. For instance,
one of the organic chemicals or may be all of them for instance, don't have a BPT
limitation for COD, it doesn't come up until 1983. Well I personally don't know about
that, but nevertheless that is what the regulation says.
Morris Wiley: If the COD limitation were placed on a municipal discharge and they
received discharges from both the refinery and the organic chemicals plant then how
would that be resolved to determine who was responsible for what pretreatment?
Harless Benthul: The total COD load coming from the municipal treatment plant
couldn't be any more than what would be BPT for those two industrial facilities. So how
you translate that back up through the municipal treatment plant, back to the effluent
from the two industrial facilities, I am to many years away from engineering to know if
I ever did and I probably didn't. Sorry.
Bob Wortman; Isn't the intent of the law where you have an industry discharging into
a municipal system the actual intent I would think is that any discharge in the system
should not cause any harm to the sewer or harm to any of the biological treatment pro-
cesses. And yet the effluent from the sewage treatment plants should meet stream
quality standards and I would think if you meet the intent of the law this would be the
essential thing to do regardless of all of the other parameters.
Harless Benthul: Well I go along with you up to the point where you said it would be
sufficient if you met the stream standards, remember we have technology standards which
would apply to the municipality for starters. If you start backing away from placing
requirements on however many industrial dischargers there are to the municipal facility,
if you'd start backing away much from the technological standard that they would other-
wise be required to meet, then you are placing them at a competitive advantage with
-------
89
the people who are discharging and treating their own waste and just that advantage
might or might not be offset by the cost recovery ordinance under the construction grant,
assuming there is one.
Bob Wortman; But I thought that was one of the concerns to encourage more central-
ization of treatment processes and to me there should be some built in advantage to an
industry going to the city, otherwise he is not going to go and you are going to have
more permits, more administrative expenses and everything. In other words there are a
lot of things to consider here.
Harless Benthul: You are right there are and one of them is not placing people at
relative advantages or disadvantages when they should otherwise be similarly situated.
BIOGRAPHY
Harless Benthul holds a B.A. from Texas A&M
University and a Bachelor of Law degree from S.M.U.
in 1964. He has vast experience in the engineering
field, seven years total and has been in the private
law practice four years and three years in the
practice of corporate law. His present title is Chief
of the Enforcement Legal Branch in Environmental
Protection Agency Region VI in Dallas.
-------
90
-------
"MEASURING THE PARAMETERS SPECIFIED IN PERMITS"
F. T. Weiss
Shell Development Company
Houston, Texas
ABSTRACT
Much effort has been expended in recent years to develop and evaluate methods for
the determination of wastewater parameters. It is important to review the validity of the
analytical methods and to determine that they relate to the ecology of the receiving waters.
The detailed procedures of the methods must be carefully standardized and cooperatively
tested on samples representing actual effluents.
INTRODUCTION
In order to establish and enforce environmental regulations it is necessary that there
be a reliable, accurate and meaningful measuring system. The measuring system is com-
posed of a complex interrelation of several factors. First, of course, are the test methods
themselves. These should be chosed to determine those components and properties of the
effluents for which there is a need for measurement. The methods must be carefully
examined to determine the significance of procedural operations and standardized by inter-
laboratory comparisons. Precision of the methods should be determined from interlaboratory
testing and accuracy determined, if possible, from analysis of Standard Reference Materials.
Well established procedures must be set up and followed to provide for proper sampling and
to ensure stability of the samples prior to actual analysis. Analytical chemists should
always be a part of the overall planning process.
A essential consideration in establishing a meaningful measuring system is to make
certain that the analytical data obtained on the effluents relate to the ecology of the
receiving waters. Factors that should be considered in assessing the health of the receiving
waters include processes acting upon the components of the effluent, as it passes into the
receiving water, as evaporation, dilution, chemical reactions, and biological degradation.
ORIGIN OF THE METHODS
The methods specified in the permits represent various stages in the history of
analytical technology, some quite current while others derive from an earlier period of
water analysis methodology. Among the more current techniques are atomic absorption
for determination of metals and specific ion electrodes for determination of individual
ionic species. These powerful new techniques now permit the determination of specific
components with a higher degree of accuracy. Examples of test methods which derive from
an earlier period of technology in water analysis include the determination of oil and grease
and of biochemical oxygen demand. It is interesting to review the history of these older
91
-------
92
methods in regard to the significance of their present applications.
The present methods for determination of oil and grease in water stem from a commi-
ttee review held in Los Angeles by the City Bureau of Engineering some years ago. The
paper covering their work (1) was published in 1941 and used the now familiar extraction
methods for determination of grease in sewage sludges and industrial wastes. The procedure
described initial acidification of the sample, extraction of the grease with a solvent,
evaporation of the solvent, and weighing the residue. The extraction procedure itself,
however, was taken from yet an earlier method published in 1933 by the American Petroleum
Institute (2) for the analysis of oil-field wastewaters.
The history of the test for biochemical oxygen demand (BOD) in water analysis goes
back many more years. The test was recommended in 1912 by the British Royal Commission
on Sewage Disposal (3). Almost fifty years ago, in 1928, it was possible to present a
review paper entitled "Technic and Significance of the Biochemical Oxygen Demand
Determination" (4) and, even then, make the statement that the BOD determination had
been in use for many years. The dilution method, taken from the 1917 issue of "Standard
Methods" (5), is essentially the BOD method in current use. The data reported (4) were
taken from analysis of sewage effluents, the drainage canal of the Chicago Sanitary District,
and of an effluent from a corn products manufacturing plant.
A characteristic of these older methods, and that also of certain others being used as
test procedures, is that the results are empirical. The data obtained are dependent upon
the test variables themselves and do not provide compositional data. In certain of these
methods, the parameter being measured is defined by the methodology itself.
STANDARDIZATION GROUPS
The long history of the analysis of water led inevitable to a confusion of methods and
interpretation of results. To overcome these problems, standardization groups were
organized many years ago. A committee of the American Association for the Advancement
of Science reviewed methods for the analysis of water and published a report (6) in 1889
which covered methods for determination of "free" and "albuminoid" ammonia, nitrites,
nitrates, and oxygen-consuming capacity. Further developments led the American Public
Health Association to appoint a committee who produced in 1905 the first edition of the
now familiar "Standard Methods". Revised and enlarged editions of "Standard Methods"
have been produced over the years under the combined sponsorship of the American Public
Health Association, the American Water Works Association, and the Water Pollution Control
Federation (7).
The American Society for Testing and Materials (ASTM) made a review in 1925 and
established that there was a need for cooperative testing in the sampling and analysis of
industrial waters (8). As a result, ASTM Committee D-19, on water analysis, was
established in 1933 with eleven members and has grown in scope and membership, parti-
cularly in the past ten years. Membership now is over 500 and the number of methods
currently printed are 135 (9). Active groups meet twice yearly to review and introduce
methods covering such topics as biological monitoring, sampling techniques, radiochemical
-------
93
analysis, determination of inorganic and organic components, analysis of brackish water,
and the identification of waterborne oils. An essential part of the ASTM process is inter-
laboratory comparison of results on practical samples in round-robin fashion to determine
precision and reliability of the methods. Such testing goes on during the course of the
year and results are reviewed at the semi-annual meetings.
The U.S. Environmental Protection Agency (EPA) has developed a valuable Manual
covering "Methods for Chemical Analysis of Water and Wastes" (10) now in its second
edition. This set of methods was collected, reviewed and published by the Method Devel-
opment and Quality Assurance Research Laboratory at the National Environmental Research
Center at Cincinnati, Ohio. Assistance was provided by other governmental laboratories,
both state and federel. The Manual is a basic reference for monitoring water and wastes
in compliance with the Federal Water Pollution Control Act. Although other test procedures
may be used, it is stated in the Manual that the methods therein described will be used by
the EPA in determining compliance with applicable water and effluent standards.
European standardization groups are also active in promulgating methods for the
analysis of wastewaters, but their publications do not directly impinge upon the practice in
the United States on measuring wastewater parameters. Exceptions, of course, will be with
the development of new analytical technology or in topics which have a large geographical
impact. For instance, the characterization of waterborne oils has been the subject of
activity not only by the ASTM and others in the United States, but also by the Institute of
Petroleum in England (11).
FEDERAL GUIDELINES FOR ESTABLISHING TEST PROCEDURES FOR THE ANALYSIS OF
POLLUTANTS
The specific approved test procedures to be used for measuring the parameters
specified in permits derive from Part 136, Title 40, Code of Federal Regulations. Section
136.3 identifies the test procedures which must be used in any permit application pursuant
to sections of the Federal Water Pollution Control Act. Updated lists of approved methods
in the Code of Federal Regulations are published from time to time in the Federal Register.
The most recent proposed amendments to the test procedures for aqueous samples were pub-
lished in the Federal Register issue of June 9, 1975 (12).
These proposed amendments covered a great deal of ground since they include not
only correction of minor errors but also allowed new techniques and added a considerable
number of additional parameters for possible measurement. The corresponding proposed
table of approved test procedures from the Federal Register of October 16, 1973 (pages
28759 and 28760) lists 71 parameters for measurement. The 1975 table lists 115 parameters.
Although the text briefly reviews the fact that additional parameters have been added, the
magnitude of the additions is obscured by the way in which the 1975 table was constructed.
In the effort to keep the parameter listings of the 1973 table unchanged, the 44 new para-
meters have been introduced as subheadings. For instance, the new listing of Rhodium is
introduced as 35 (b) under Potassium (35) and Potassium - dissolved (35a). Ruthenium Is
then introduced as 35 (c).
The recent issue of the Federal Register is an example of a way in which test
-------
94
procedures have been introduced without undergoing a thorough standardization process and
interlaboratory review. For example, a number of the new test procedures which have
been introduced are directed to the determination of the dissolved portion of metal com-
ponents, generally using atomic absorption for final measurement. The procedures differ-
entiate between the portion of the metallic constituents which will pass through a 0.45
micron filter and those which will be retained. In nearly all cases the reference is to a
single page of the EPA Methods Manual (10). This Manual does not provide sufficient
procedural details for method standardization. In fact, a note is included in the EPA
Methods Manual in the general method detailing techniques for atomic absorption which
suggests the need for (unspecified) special precautions for stabilization of certain types of
metal containing samples. Large differences can be expected in behavior between metal
components of such diverse properties as sodium, for example, and titanium. Because of
these factors new sets of proceudres should always be subjected to careful and detailed
review and proper standardization and validation before being applied to the permit process.
This should include crosstesting of samples amongst a group of laboratories for intercompo-
sition and development of data on the reliability of the procedures.
Although the current Federal Register lists 115 parameters, in actual practice con-
siderably fewer are used for control of petroleum refinery wastewaters. Table I lists twelve
parameters which are most commonly required with a brief statement of the method recom-
mended.
REQUIREMENTS FOR SATISFACTORY TRACE ANALYSIS
Measurement of certain components specified in the permits is a practical example
of trace analysis which is generally considered to be the determination of a component or
components in a sample vastly diluted by other materials. Many methods have been used
for trace analysis. (13), (14) During 1972 an intercompany project was initiated to develop
techniques for measuring trace concentrations of metals in oil (14). Although their analyti-
cal medium was different from water, their detailed studies of the principles of trace analysis
apply to the determination of trace components in water.
Sampling. It is well acknowledged that analytical results can be no better than the
sampling and preservation techniques employed. It cannot be overstressed that samples
must accurately represent the conditions existing at the point of sampling and that samples
must be properly preserved in that condition until the analysis is made. Although each of
the reference compilations provides guidelines on sampling, in actual practice good sampl-
ing techniques are not always employed. For accurate results in trace analysis, clean
sampling containers of the proper material, caps, thorough sampling procedures, and proper
preservation techniques must be carefully established and rigorously follwed in actual use.
If these points are not properly observed, the results obtained can be meaningless. It
must be emphasized that anything which contacts the sample, whether apparatus, storage,
or shipping containers and closures, the reagents or the atmosphere can serve as a potenH
ial source of elemental contamination or provide a sink for loss of some components, such
as by adsorption.
Storage and Stability. The instability of dilute aqueous solutions has been widely
observed. (10, 14) The best rule is to analyze samples as soon as possible after they have
-------
95
been taken. Changes that take place in a sample can be due to chemical or biological
reactions. Metal cations may precipitate, form complexes or adsorb onto surfaces. Organ-
ics7 including hydrocarbons7may undergo biochemical oxidation. A general rule to
increase the stability of aqueous samples is to acidfy approximately to a pH of 2. HgClj
can be added for bacterial inhibition. Refrigeration at temperatures near freezing or
below can be helpful in preserving a sample which cannot be analyzed immediately.
Standards. The best means to evaluate laboratory performance is the analysis of a
standard sample in which the composition of the determined component is known exactly.
A series of quality control samples is available on request from the Environmental Protection
Agency (15) for the determination of many components including certain trace metals,
ammonia, nitrate, phosphate, pH, and so on. These have been widely used to provide
checks on reagents and laboratory procedures. More exact primary standards are the
Standard Reference Materials being prepared by the U.S. National Bureau of Standards
for trace metals in water. The composition of NBS Standard Reference have been certified
by at least two independent analyses. Suitable Standard Reference Materials are the best
means of evaluating the entire proceudre and to establish its accuracy. In the absence of
a standard sample, "spiked" samples prepared by adding known amounts of a constituent
to an effluent can be used.
SIGNIFICANCE OF THE DATA
It is important to divide the test results into those which provide compositional data
and those which are empirical. In the first category, compositional, are included those
methods for determination of ionic constituents such as dissolved metals, nitrogen compounds
such as ammonia, nitrite or nitrate, and sulfide ion. Since these are all defined compon-
ents and determinable by at least two independent analytical methods, their concentrations
can be determined with some degree of accuracy and the effect of interferences investigated
on the precision and accuracy of the results. Interlaboratory comparisons of methods and
data, preferably, by several techniques can establish the validity, accuracy, and precision
of the data.
On the other hand, those methods which provide empirical results, including the
determinations for oil and grease and biochemical oxygen demand, defy a definition of
accuracy. For the empirical methods the best that can be done is to standardize the
procedural descriptions and then determine precision by interlaboratory round-robin testing
under exactly defined conditions. It is not possible to run any of the empirical tests by
two sets of procedures to determine accuracy since the results are defined by the procedures
themselves.
Relationship to Ecology of Receiving Waters. The prime consideration in establish-
ing effluent standards should be to relate the analytical data of the effluent wastewaters to
the ecology of the receiving waters. The factors which should be considered in the
receiving water are toxicity (both acute and long-term) to the flora and fauna, depletion
of oxygen, odor, and detrimental changes in the esthetics of the receiving waters and the
adjacent land areas. Factors that should be considered in assessing the effects of effluent
composition on the receiving waters include such processes as evaporation, dilution, air-
water interchange, chemical and photochemical reactions, and biological degradation. (16)
-------
96
The relationship between laboratory bioassays and actual results in the field are not necessarily
straightforward. For example, the chemical form in which a trace metal is released or the
form it adopts after release is highly significant in its biological impact (17). There is an
increasing body of evidence indicating that there are natural processes operating to reduce
both the concentration and toxicity of trace metals dissolved in water. In order for a
heavy metal to be toxic it must be in the ionic state. (18,19,20) In most natural waters
much of the free metal ions would probably be rapidly bound to organic substances, nat-
urally present in the water, decreasing the relative percentage of the ionic species There
is indirect evidence that organically chelated heavy metals in aqueous solutions do not have
as great an effect upon organisms as do solutions of the metal salts. (20) This could be due
either to the fact that the organo-metallic complex is too bulky to enter a biological system
or it could be due to the lack of availability of the metal for reaction with enzymes within
the cells.
Precision of Data. From the standpoint of quality control and agreement with
regulatory standards it is important to have an understanding of the significance of precision
of the results obtained by the various methods. Both ASTM Committee D-19 (21) and the
Methods Development and Quality Assurace Research Laboratory of the Environmental
Protection Agency (22,23,24,25) have practices for evaluating analytical methods for
water. After a method has been sufficiently investigated to determine the significant
variables and establish their control, it is possible to organize an interlaboratory, round-
robin testing program. A written copy of the method being investigated plus the necessary
samples for each test are supplied to each laboratory. From the interlaboratory tests, it is
possible to derive terms for method precision, which is usually done by calculating standard
deviations. Standard deviation is the measure of the dispersion of the data around their
average, expressed as the square root of the quantity obtained by summing the squares of the
deviations from the average of the results and dividing by the number of observations minus
one. The standard deviation is generally given as a measure of the precision of methods
which have been subjected to comparative evaluation. The next factor to consider is the
confidence level of the data obtained at two different laboratories. As Youden (26) points
out, it is well known that a reasonable number of measurements must be made to give
confidence in the results. A number of factors enter into establishing confidence limits on
data sets. At a 95 percent confidence level an approximate rule (26) is that, for sets of
4 or 5 determinations, average results between two laboratories should not differ by more
than approximately 3 times the standard deviation. When analytical methods are used at
or near their lower level of sensitivity, there is a greater tendency to error and loss of
reproducibility.
PROCESS-TYPE INSTRUMENTATION
The use of continuous monitors to determine effluent parameters on flowing systems
is particularly valuable, if it can be made effectively. Such continuous measurements
provide for the fully automatic regulation and control of certain parameters and avoid the
delay and error in sample taking and handling. It is always necessary to carefully review
the operation of a continuous monitor under field conditions for a sufficient period of time
to make certain its behavior is satisfactory and that its response is to the parameter to be
measured and not to an artifact in the system.
-------
97
Continuous instruments are available and have been commonly used for many years
for measurement of pH by use of the glass electrode (ASTM D 1293). Commercial process-
type water quality analyzers are currently becoming available for a number of other com-
ponents. These include the use of specific ion electrodes for ammonia and sulfide, and
ultraviolet spectroscopic instruments for phenols. An excellent detailed coverage of such
instruments was recently published by the Lawrence Laboratory of the University of Califor-
nia (27).
EXAMPLES OF CRITICAL ANALYSES
Oil and Grease Determination. A considerable number of methods have been
applied over many years for determination of "oil and grease" or extractable materials in
aqueous samples. Table II provides a summary of the principal features of some twenty
such methods. For each of these, more detail is given in a recent report (28) prepared
for the American Petroleum Institute. The kind of results that may be obtained from use of
this large array of empirical procedures may be quite different. For instance, Methods 4
and 10 both involve initial extraction with Freon 113. Method 4 utilizes a gravimetric
determination following solvent evaporation, which causes also some evaporation of light
hydrocarbons. Method 10 utilizes an intrared spectrophotometric measurement and does
not involve evaporation. Consequently, it can be anticipated that comparative measure-
ments, v/here light hydrocarbon components are present in the sample, can be higher with
Method 10. It is interesting that the Federal Register of June 9, 1975 (12) lists both the
gravimetric and infrared methods as alternative procedures without commenting on differences
that could be anticipated in results.
Another area of concern is the relative solvent power of common solvents for petro-
leum constituents. The components of wastewater are a complex mixture of compounds of
different nature and are not all equally dissolved by different solvents. For instance,
Freon appears to extract less material of an asphaltic nature and a "rag" or undissolved oily
film has sometimes been noticed.
Another significant difference in results can be obtained when an absorbent is used
to separate polar from nonpolar materials. Methods of this type can provide for the deter-
mination of hydrocarbons in effluent waters. A review of Table II shows that there are
several available methods utilizing adsorbents to separate polar materials from hydrocarbons
to allow the more specific determination of the latter. These involve different applications
of adsorbents such as the use of alumina (Method 5), Florisil (Methods 11 and 17) or silica
gel (Method 13). An extensive round-robin test was recently managed by the Robert S. Kerr
Laboratory of the Environmental Protection Agency, Ada, Oklahoma to ascertain the validity
and reproducibility of Method 13. It is generally the experience that the efficiency of
separation of polar organics from aromatic hydrocarbons may depend to a considerable extent
on the polarity of the solvent, the activity of the adsorbent and the adsorbent/sample ratio.
A study of this type has been recently published (29) and indicates that the variables should
be carefully established with such methods for accurage and meaningful results.
Continuous monitors have been considered for determination of small quantities of
oil in effluent waters. The principles and experience with several of these techniques are
indicated under Method No. 21. Most of those methods involve interaction of the com-
ponents of the sample with light of several wavelengths. It is important that the spectral
-------
98
region (that is, ultraviolet, visible, or infrared) relate properly to the components of the
samples. Although these techniques do show promise, several have shown limitations in
field use in that they may be somewhat fragile and some tend to undergo fogging in the
optical path. It is necessary that the optical instruments properly discriminate between oil
and interferences such as suspended solids.
ASTM Committee D-19 is currently conducting interlaboratory comparisons with many
participants of several of the methods of oil and grease analysis. Samples being circulated
are practical effluents and contain both petroleum and non-petroleum constituents.
Chemical Oxygen Demand. The dichromate oxidation method for chemical oxygen
demand is another empirical test with a history of use of some years (30. As is pointed out
in "Standard Methods" (7) it is important to follow a standardized technique since only a
part of the organic matter is oxidized. This is particularly important with nitrogen com-
pounds since the extent of their reaction can be variable, depending upon structure, as can
be seen from Table 111. The data in Table III are taken from references being reviewed for
consideration in the ASTM D-19 Committee studies on the method for Chemical Oxygen
Demand. Because of the way the test is performed, volatile, nonpolar components, such
as benzene and toluene can be partially lost. The limited reaction of many nitrogen com-
pounds depend upon the structure of the compound. For instance, pyridine with one heter-
ocyclic ring is essentially nonreactive while quinoline undergoes extensive reaction, no
doubt due to initial attack on the homocyclic ring.
Biochemical Oxygen Demand (BOD). As pointed out earlier, this empirical method
has a long history of use in the analysis of wastewaters. The test has significance as an index
to the susceptibility of the components of wastewater to biodegradation. The test involves
measurement of the amount of oxygen consumed in a 5 day period after the sample has been
inoculated with sewage bacteria. Despite its long history of utility, the BOD method has
been recognized to have many limitations (31). The limitations include such factors as the
differential rates of biochemical oxidation of various compounds, including the effects of
nitrifying bacteria which convert nitrogen compounds to higher forms. The rates of reaction
may vary markedly with different sources of inocula. "Standard Methods" (7) points out that
the BOD test is of limited value in measuring the actual oxygen demand of surface waters.
The extrapolation of test results to actual stream oxygen demand is highly questionable since
the laboratory environment does not reproduce actual field conditions, particularly as
related to temperature, sunlight, biological population, water movements, and oxygen
concentrations.
ASTM Committee D-19 came out with a similar assessment of the BOD method. The
action of the ASTM was to discontinue without replacement, ASTM D 2329, "Test for Bio-
chemical Exygen Demand of Industrial Water and Wastewaters". (9) The ASTM group con-
cluded that the BOD concept is important in determining water quality but that the laboratory
bottle incubation procedure typified by the usual BOD test is ill defined, producing data
which are limited. The ASTM recommends several alternate tests which have been better
standardized to assess pollution potential. These tests include those for Total Organic
Carbon (ASTM D 2579) (9) and Total Oxygen Demand (ASTM D 3250) (9). The subject is
still under close review by this ASTM group.
-------
99
Cyanide Determinations. The issue which has become of concern in the determina-
tion of cyanide in effluents is the differentiation between simple cyanide and complexed
cyanides. In common practice simple or "free" cyanide is defined as the CN ion or the
HCN molecule. Complex cyanide is taken to mean any molecular composition in which
the CN group is bound to a metal through coordinate bonds. Examples of complex cyanides
are those of iron: Fe (CN), and Fe (CN)7 . The practical distinction between simple
and complex cyanides depends upon the production of HCN under relatively mild conditions.
For example, the procedure of Roberts and Jackson (32) provides for distillation of hydrogen
cyanide, in the presence of zinc acetate, under reduced pressure. Under the conditions
of the test the decomposition of ferrocyanide is completely prevented. The cyanide in the
distillate is determined by colorimetric pyridine-pyrasolone procedure.
There is currently an extensive effort in ASTM Committee D-19 involving inter-
laboratory round-robin tests with many participants examining several of the methods which
have been recommended for cyanide determination. The aim of this testing program is to
establish the validity and precision of the cyanide methods and to establish the significance
of interferences.
CONCLUSIONS
1. The large scope of the procedures for measuring the parameters of aqueous
effluents has given rise to methods whose variables have not all been completely under-
stood or standardized. Results from some of these methods are dependent upon the proced-
ural steps employed.
2. It is important to review and decide on what components in effluent waters should
be measured. The purposes and definitions of the measurements must be tied to real needs
relating to the ecology of the receiving waters.
All methods must be thoroughly examined to determine the significance of the
procedural operations. The final methods must be carefully standardized and cooperative
testing on samples representing actual effluents.
REFERENCES
(1) Pomeroy, R. and Wakeman, C. M. (1941), "Determination of Grease in Sewage,
Sludge, and Industrail Wastes", Ind. Eng. Chemistry, Analytical Edition, 13,
pages 795-801.
(2) American Petroleum Institute (1933), "Disposal of Refinery Wastes", Second Edition,
Section I.
(3) Royal Commission on Sewage Disposal (1912), Eighth Report, I. "Standards and Tests
for Sewage and Sewage Effluents Discharging into Ribers and Streams", Cd 6464,
HMSO, London.
(4) Mohlman, IF. W., Edwards, G. P., and Swope, G.(1928), "Technic and Significance
of the Biochemical Oxygen Demand Determination", Ind. Eng. Chemistry, 20, pages
242-246.
(5) "Standard Methods for Examination of Water and Sewage" (1917), page 71.
-------
100
(6) Caldwell, G. C. (1889), "A Method, In Part, For the Sanitary Examination of Water
and for the Statement of Results, Offered for General Adoption", J. Analytical and
Applied Chemistry, 3, page 398.
(7) "Standard Methods for the Examination of Water and Wastewater" (1971), Thirteenth
Edition, American Public Health Association, New York.
(8) Clarke, F. E. (1972), "The History of Committee D-19 on Water" in ASTM Committee
D-19 Handbook.
(9) Annual Book of ASTM Standards (1975), "Water", Part 31.
(10) "Methods for Chemical Analysis of Water and Wastes" (1974), Second Edition, U. S.
Environmental Protection Agency, National Environmental Research Center, Cincinnati,
Ohio.
(11) Whitham, B. T., Duckworth, D. F., Harvey, A. A. B., Jeffrey, P. G., and Perry,
S. G. (1974), "Marine Pollution by Oil - Characterization of Pollutants, Sampling,
Analysis, and Interpretation", Institute of Petroleum, Great Britain.
(12) Federal Register (1975), Vol. 40, No. Ill, Monday, June 9, pages 24535-24539.
(13) Weiss, F. T. (1970), "Trace Analysis" in "Determination of Organic Compounds:
Methods and Procedures", Wiley-lnterscience, pages 317-334.
(14) Hofstader, R. A., Milner, O. I., and Runnels, J. H. (1976), "Analysis of Petroleum
for Trace Metals", to be published by the American Chemical Society.
(15) U. S. Environmental Protection Agency, Analytical Quality Control (1975), "Newsletter"
Environmental Monitoring and Support Laboratory, Cincinnati, Ohio 45268, October,
page 6.
(16) "Petroleum in the Marine Environment" (1975), National Academy of Sciences,
Washington, D. C.
(17) "Trace Methods" in "Marine Bioassays" (1974), Workshop Proceedings, Marine Techno-
logy Society, 1730 "M" Street N.W., Washington, D. C. 20036, pages 76-93.
(18) Lerman, A. and Childs, C. W. (1973), "Metal-Organic Complexes in Natural Waters:
Control of Distribution by Thermodynamic, Kinetic, and Physcial Factors", in "Trace
Metals and Metal-Organic Interactions in Natural Waters," Edited by Singer, P. C.,
Ann Arbor Publ., Inc., Ann Arbor, Michigan.
(19) Morel, F., McDuff, R. E., and Morgan, J. J. (1973), "Interactions and Chemostasis
in Aquatic Chemical Systems: Role of pH, pE, Solubility, and Complexation", In
"Trace Metals and Metal-Organic Interactions in Natural Waters", Edited by Singer,
P.C., Ann Arbor Publ., Inc., Ann Arbor, Michigan.
(20) Wilson, R. C. H. (1972), "Prediction of Copper Toxicity in Receiving Waters", J.
Fish. Res. Bd. Canada, ^9, pages 1500-1502.
(21) ASTM D 2777-72 (1975), "Standard Recommended Practice for Determination of
Precision of Methods of Committee D-19 on Water", In "Annual Book of ASTM
Standards", Part 31, American Society for Testing and Materials, Philadelphia,
Pennsylvania.
(22) Winter, J. A. and Clements, H. A. (1975), "Interlaboratory Study of the Cold Vapor
Techniques for Total Mercury in Water", Water Quality Parameters, ASTM Special
Technical Publication 573, pages 566-580.
(23) Winter, J. A. and Midgett, M. R. (1969), "FWPCA Method Study I, Mineral and
Physival Analysis, An Evaluation of the FWPCA Analytical Methods for Surface Waters",
Analytical Quality Control Laboratory, Cincinnati, Ohio 45202.
-------
101
(24) Winter, J. A. and Midgett, M. R. (1970), "Method Study 2, Nutrient Analyses,
Manual Methods, An Evaluation of Analytical Methods for Water and Wastewater",
Environmental Protection Agency, Analytical Quality Control Laboratory,
Cincinnati, Ohio 45202.
(25) Winter, J. A. (1971), "Method Study 37 Demand Analyses, An Evaluation of
Analytical Methods for Water and Wastewater", Environmental Protection Agency,
Analytical Quality Control Laboratory, Cincinnati, Ohio 45202.
(26) Youden, W. J. (1951), "Statistical Methods for Chemists", John Wiley & Sons, New
York.
(27) "Instrumental for Environmental Monitoring" (1975), Water, Lawrence Berkeley
Laboratory, University of California, Berkeley, California 94720, Second Update, June.
(28) "Review of Analytical Methods for Determination of Oil and Grease in Produced
Waters from Oil and Gas Extraction Industry Operations" (1975), Prepared for the
American Petroleum Institute Committee on Environmental Conservation - Division
of Production, 300 Corrigan Tower, Dallas, Texas 75201.
(29) Bridie, A. L., Bos, J., and Herzberg, S. (1973), "Interferences of Non-Hydrocarbons
in Oil-in-Water Determination," J. of the Inst. of Petroleum, 59, 570, pp.263-267,
November.
(30) Moore, W. A., Kroner, R. C., and Ruchhoft, C. C. (1949), "Dichromate Reflux
Method for Determination of Oxygen Consumed", Anal. Chem., 21, pages 953-957.
(31) Marske, D. M. and Polkowki, L. B. (1972), "Evaluation of Methods for Estimating
Biochemical Oxygen Demand Parameters", Journal Water Pollution Control Federation,
44, pages 1987-2000.
(32) Roberts, R. F. and Jackson, B. (1971), "The Determination of Small Amounts of
Cyanide in the Presence of Ferrocyanide by Distillation Under Reduced Pressure",
Analyst, 96, pages 209-212.
DISCUSSION
Larry Echelbergen I am wondering what Shell does to ensure the accuracry of its numbers.
What do you do at the individual refineries? Do you spike samples or periodically on a
routine basis, for example, send them to independent laboratories? What do you do or
what do most of your refineries do to ensure accuracy of your results?
Fred Weiss: In the Company for certain types of analyses we have for some years run our
own type round-robining for some product materials. For effluents, which I take your
question to be, because the transport of effluent samples is very difficult, we do run cross
comparisons with commercial laboratories in each of our areas. The data I have seen from
a number of the cross comparisons is generally fairly good. Another thing that we have done,
although I don't know how widely it is done throughout the refinery laboratories, is to get
those samples from the EPA that are essentially standard materials. There are a whole
series which have been made available and we have run those.
-------
102
BIOGRAPHY
Fred Weiss is Senior Staff Research Chemist,
Analytical Department, Shell Development Company,
Houston, Texas. He holds degrees in Chemistry from
fhe University of California at Los Angeles
(B.S., 1938) and from Harvard University (M.A.,
1939 and Ph.D., 1941). He has been involved in
research with the Shell Companies since 1941. His
special interests have been in the detailed analysis
of organic compounds and the application of methods
to trace analysis. In recent years much of his work
has been devoted to environmental measurements.
TABLE 1 PARAMETERS COMMONLY REQUIRED FOR PETROLEUM REFINERY PERMITS
Parameters
Biochemical Oxygen Demand 5 day (BOD,)
Chemical Oxygen Demand (COD)
Oil and Grease
Total suspended solids
Total Organic Carbon (TOC)
Chromium, Total
Chromium (VI)
Cyanide
Phenols
Ammonia
Sulfide
PH
Method
Modified Winkler or probe
Dichromate reflux
Extraction
Glass fiber filtration
Combustion-infrared
Atomic absorption or colorimetric
Extraction and atomic absorption,
colorimetric
Distillation and colorimetric
Colorimetric
Several techniques
Titrimetric or methylene blue
Electrometric
-------
TABLE II
SUMMARY OF ANALYTICAL METHODS FOR DETERMINATION OF OIL & GREASE'
TYPE
u
t-i
H O
> a
<; s
o
VOLUMETRIC
METHODS
,j
S w
| §
5 H
H §
2
SAMPLING
en
11
fe H
M M
O S
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
METHOD DESIGNATION
Chloroform Extractable Matter
Solvent Extraction
Soxhlet Extraction
APHA Oil & Grease
Hydrocarbon and Fatty Matter
Reflux Distillation Method
Petroleum Ether Extraction
Hexane Extractable Material
Flocculation - Extraction Method
Oil and Grease - Freon Extractables
Hydrocarbons in Water or Soil
Combined Methods
Petroleum Hydrocarbons
Volatile and Non-Volatile Oily Material
Ultraviolet Fluorescence
Ultraviolet Absorption
Gas Chromatography
Visible Color Method
Color Comparator
Ultraviolet
Visible
Infrared
Fluorescence
Light Scattering
REFERENCE
ASTM D-1178
ASTM-2778
EPA Manual 00550
APHA - 137
Std. Methods P. 413
ASTM D-1340
a)
ASTM D-1891
API 732-53
EPA - 00560
Concave 1/72
API 0 & G Comm
1975 EPA Study
API-733-58 5. Lit.
Concawe 111-72
Hach Handbook-1973
Champion Chem. Co.
EPA, ASTM, APHA
STATUS
Discontinued
Limited Use
Wide Use
Wide Use
Limited Use
Discontinued
Limited Use
Discontinued
Limited Use
General Use
Limited Use
Proposed
Proposed
Limited Use
Limited Use
Limited Use
European Use
General Use
Very Limited
Wide Use
Spsc if ic
Applications
CONDITIONS
OF EXTRACTION
pH 3-4
pH Variable
pH - 2 Soxhlet
Acidic
Follov Method 4
Acidic
No pH Adj.
pH-4
pH-4
pH-2
Acidic
pH < 7 > 7
Acidic
Acidic
None
Acidic
Acidic
None
None
EXTRACTANT
Chloroform
Any Solvent
Hexane or Freon-113
Freon or Pet. Ether
Freon-Hexane
Benzene, Others
Petroleum Ether
Hexane
Ether
Freon 113-IR
CCL4 (IR)
CCL4 (IR)
CCL4 (IR)
CCL4 (IR)
CCL4-Hexane (UV)
Hexane (UV)
CCL4 or Pentane (GC)
3-4 Solvents
Special Chemical
ADSORBENT
None
None
None
None
Alumina
None
None
None
Fe(OH)3Floc
None
Florisil
None
Silica Gel
None
None
None
Florisil
None
None
COSTS
Moderate
Moderate
Moderate
Moderate
Moderate
Moderate
Expensive
Moderate
Inexpensive
Expensive
Expensive
Expensive
Expensive
Expensive
Expensive
Expensive
Expensive
Inexpensive
Inexpensive
pH Control, Refrigeration, Time-Dependent, Glass Bottles
None
As Received
None
Moderate
Inexpensive
Expensive
Expensive
Moderate
•For more information, see Reference 28.
O
CO
-------
104
TABLES BEHAVIOR OF VARIOUS TYPES OF ORGANIC MATERIAL IN
ASTM METHOD D-1252 FOR CHEMICAL OXYGEN DEMAND
Component Reactivity, Percent of Theoretical
References (a) (1) (2) (3) (4)
Aliphatic Compounds
Acetone 98 96 94
Acetic Acid 98
Dextrose 95
Diethylene Glycol - 70
Ethyl Acetate - 85
Methyl Ethyl Ketone - 90
Aromatic Compounds
Acetophenone 89
Benzeldehyde - 80
Benzene 60 41 -
Benzole Acid 98 - 100 100
Dioctyl Phthalate 83
Ortho Cresol 95-95
Toluene - 45
Nitrogen Compounds
Acrylonitrile 48 - 44
Adenine - - - 59
Aniline 80 - 74
Butyl Amine 57
Pyridine 0 I 2
Quinoline 87
Trimethylamine 1 -
Tryptophane 87
Uric Acid 64
(a) References:
(1) Unpublished data.
(2) Dobbs, R. A., Williams, R. T., Anal. Chem. 35, 1064 (1963)
(3) Buzzell, J. C., Young, R. H. F., Ryckman, 57 W., "Behavior of Organic
Chemicals in the Aquatic Environment: Part II, Dilute Systems," Manufacturing
Chemists Association, April 1968, page 34.
(4) Chudoba, J. and Dalesicky, J., Water Research, V. 7, No. 5, 663, (1973).
-------
BANQUET
Master of Ceremony --Dr. J. Paschal Twyman
President
The University of Tulsa, Tulsa, Oklahoma
Paschal Twyman; "I am Paschal Twyman, and, as President of the University of
Tulsa it is my privilege to welcome all of you to this Open Forum Banquet. As you are
well aware this highly successful symposium on "Management of Petroleum Refinery Waste-
waters" is co-sponsored by the EPA, the API, the NPRA, and T.U. Just before this
banquet I was told that T.U. were the only initials which should be spelled out as Tulsa
University; because the others - E.P.A., A.P.L, and N.P.R.A. - are so well known to
all of us.
"As your M.C. for this banquet it is my pleasure to introduce to
you the people at the head table. Starting at my left, that is your right, they are:"—
Fred Pfeffer, of the R. S. Kerr Environmental Research Laboratory, Ada,
Oklahoma. Fred is the Project Officer for the EPA Grant
sponsoring this Open Forum.
Milton Beychok, Consulting Engineer, Irvine, California. Our distinguished key-
note speaker.
Herb Bruck, Technical Director, N.P.R.A. Herb is representing the NPRA,
who are, of course, one of the sponsors of this symposium.
Nick Gammelgard, Senior Vice President, Environmental Affairs, A.P.I. Nick is our
banquet speaker and is also representing the A.P.I., the other
industrial sponsor of this symposium. "
"Now on the other side starting at my right hand is:" —
Bill Galegar, Director, R. S. Kerr Environmental Research Laboratory, Ada,
Oklahoma. Bill, as you remember, welcomed us this morning.
Nancy Manning, Research, biochemist at St. Francis Hospital in Tulsa. Dr. Nancy
Manning is also Mrs. Frank Manning.
Frank Manning, Professor of Chemical Engineering at Tulsa University. Frank is
Project Director for the EPA Grant financing this Open Forum."
"Frank Manning will now introduce the banquet speaker."
Frank Manning then introduced Mr. Gammelgard. Mr. Gammelgard's address and
biography follow on page 107.
105
-------
106
BIOGRAPHY J. Paschal Twyman
Dr. Twyman holds the following degrees
from The University of Missouri at Kansas City:
B.A. in Sociology; M.A. in Educational
Administration; and a Ph.D. in Education. Dr.
Twyman served as a Professor of Education and
Associate Director of the Research Foundation
at Oklahoma State University and as Director
of Research, Assistant to the Chancellor and
Professor of Education at University of
Missouri at St. Louis/ before joining Tulsa
University in 1967, as Vice President for
Research and Development. Dr. Twyman be-
came President of Tulsa University in 1968.
Dr. Twyman has extensive research
experience and has published extensively on
higher education. Dr. Twyman is also most
active in numerous educational, charitable and
community-service organizations.
-------
BANQUET ADDRESS
"LET'S CHANGE COURSE AND SPEED"
P- N. Gammelgard, Vice President
American Petroleum Institute, Washington, D. C.
I am very glad to have been invited to participate in your Forum here in Tulsa. The
wastewater management subjects that you will be discussing in your sessions this week are
matters of great concern to refiners in this country —and, I have found, in other countries
as well. On a recent trip to the USSR, for example, I found Russian refiners talking about
some of the same things that you are talking about here.
A good many of the wastewater management problems that refiners face in this country
can be traced to the Federal Water Pollution Control Act. That is hardly news. At the
time of the passage of the 1972 Act, Congress was frustrated by the lack of progress that
had been made under earlier versions of the Act. It determined that major legislation,
with teeth in it, was necessary. But it also knew perils and pitfalls would be involved,
as is the case with all legislation of such broad sweep and bite.
In this case, for example, many important water pollution control problems had not
yet even been clearly defined. And certainly, to meet the ambitious goals of the Act,
it would be necessary to develop new control technology, always a tricky business. Back
in 1972, therefore, Congress wisely built into the Act some mechanisms that would facilitate
making any necessary changes that might be indicated in the future.
Among other things, Congress established the National Commission on Water Quality.
As you know, the Commission has just completed hearings on the staff draft of its report
to Congress, which is due in March. As a result, some of the tools needed for making
the kinds of changes envisaged by Congress are now at hand. I am referring to the really
impressive amount of data in the draft report itself and the public hearing record.
This evening, I would like to offer for your consideration a general proposition which
I believe would be helpful to bear in mind as we go about making such changes. Briefly,
it is that our country is best served by the kind of environmental legislation that responds
to clear environmental needs, has been carefully weighed against other pressing national
needs, and promises to yield benefits commensurate with the costs involved.
In light of this proposition, I would like to talk with you, first, about the impact of
the Federal Water Pollution Control Act on the petroleum industry — and on the economy
as a whole, for problems like the ones you are discussing here have reprecussions far beyond
the refinery gate. Second, I would like to make some suggestions for amending the Act
in ways that will enable this country to move ahead toward reasonable water quality goals,
efficiently and effectively.
First, regarding impact, as this audience is well aware, life under the Act is very
difficult for the regulated industries now and, unless some sound changes are made, will
107
-------
108
be a great deal more difficult in the future. As things presently stand, the EPA Administrator
must require what is, in his judgment, the best practicable control technology currently
available to be installed by 1977 and the best available technology economically achievable
by 1983. Beyond that date, the Act sets a national goal of elimination of discharge of pollu-
tants, or EDOP, for 1985. Implicit in these three dates and in the increasingly stringent
requirements for each date is a rapidly rising cost forecast that is of great concern to the
petroleum industry and, indeed, to all affected industries.
That concern is justified. Many of you are familiar with the cost findings of API's
1973 Brown and Root study of the economics of refinery wastewater treatment. The study
concluded that the cumulative capital investment in refineries alone to achieve the three
levels of wastewater treatment required by the Act, adjusted to 1974 dollars, would reach
$2.5 billion in 1977, $4.7 billion in 1983, and $6.7 billion in 1985.
Of even greater concern to us in the petroleum industry is the fact that much more
attention needs to be given to the cost-benefit ratio associated with water pollution control
requirements. The Brown and Root study plotted capital costs against water pollution
reduction — measured in terms of the three increasingly stringent levels of treatment techno-
logy and Biochemical Oxygen Demand, or BOD, discharges. The cost per annual pound of
BOD removed from refinery effluent streams by the first level of control is estimated, again
in 1974 dollars, to be about $10. The cost of application of the second level of technology
for 1983, although it would bring only a small reduction in discharges, would just about
double to some $20 per annual pound. And the cost of elimination of discharge of pollutants
by 1985 would soar to more than $80 per annual pound — about eight times the cost of the
first level of control. That, in our judgment, is completely unreasonable.
The industry expenditures on water quality have already been substantial — some
$2.8 billion between the years 1966 and 1974. The very serious question now arising is
whether, after already making notable progress, the petroleum industry should be required
to continue to commit ever larger amounts of money to achieve even smaller reductions in
water pollution. The anticipated reductions are so small, in fact, that in many, if not
most cases, they could not even be accurately measured in the receiving waters.
In other words, whatever degree of success the petroleum industry and other industries
should attain, at whatever cost, the quality of the receiving waters - - the name of the game,
after all — would remain about the same.
That bleak outlook is not the petroleum industry's alone. A few months ago, the
Environmental Reporter carried a story on several of the regional assessment studies sponsored
by the National Commission on Water Quality. These studies confirm the dismal cost-benefit
picture. With regard to the Commission's study of the densely populated and industrialized
Houston Ship Channel-Galveston Bay area, for example, the Reporter had this to say:
The study concludes that the costs of meeting 1983 and 1985
scenarios would be disproportionate to the benefits of
achieving the goals of the Act. Changes in water quality
and hence recreational opportunities/lumber of aquatic and
terrestrial biota, and aesthetics would be minimal . . .
-------
109
Yellowstone country, on the other hand, is far from being densely populated and
industrialized. Yet here, too, the Commission's investigators reached the same conclusion.
Again according to the Environment Reporter:
The report for the Yellowstone River Basin . . says the high
concentration of nutrients in the river basin appears to be
of natural origin. The river's tributaries contain high
concentrations of sediments and total dissolved solids,
mostly from natural and diffuse nonpoint sources . . . and
they would not be significantly affected by implementation
of the Act.
With so little to be gained in the way of improving actual water quality, even after
heavy expenditures, it would seem wise — indeed mandatory — to assess our position in
light of genuine environmental needs, other national needs, and, especially, costs and
expected benefits.
Everyone I know in the petroleum industry is acutely aware of the need for such
assessment. The primary mission of the industry, of course, is to help supply this country's
energy needs. That is our reason for being. And we know that environmental expenditures
on the order of magnitude required by the Act are bound to have an adverse effect on the
industry's ability to generate the capital necessary to maintain a satisfactory growth pattern
and meet future energy needs.
Capital formation is one of the most important problems — possibly-- facing the
petroleum industry. In the years ahead, if the nation is to achieve reasonable energy
self-sufficiency, the industry's capital needs are going to be staggering. It is therefore
essential that legislators and government environmental planners — at some point along the
ling, and the sooner the better — consider all of the industry's environmental expenditures
together and then relate them to the industry's expected capital requirements for all purposes.
The plain fact is that the amount of capital available to the industry in a given year
is limited. The capital used to meet environmental requirements simply is not available to
meet other pressing requirements, including, particularly, exploration and production —
the bedrock upon which a major part of the nation's energy supply structure is built.
What these environmental costs are likely to mean is becoming clearer and more urgent
by the day. In the same year that the Brown and Root refinery study was completed, API
did what it has long been recommending that legislators and government environmental
planners do. It embarked upon a much broader study covering all environmental costs to
all segments of the industry. Phase I of that study, a literature survey, was published in
1974. Phase II, forecasting environmental costs to the industry for present and future
environmental regulations, is now being finalized.
According to the preliminary findings of Phase II of the study, the cost of meeting
anticipated federal, state, and local environmental regulations between 1975 and 1985
to the petroleum industry — expressed in 1974 dollars — will be $20.1 billion. This figure,
incidentally, does not include the cost of the 1985 goal of EDOP, which, again, we
-------
no
believe to be unreasonable and most probably unattainable. Excluding EDOP, the cost of
meeting water quality regulations is estimated to be about $8 billion, or some 40 percent of
total environmental expenditures.
The magnitude of the additional $20.1 billion environmental investment for the 1975-
1985 period can be put in perspective by comparing it with recent and projected levels of
total capital spending by the petroleum industry. That $20.1 billion, for example, is almost
as much as the $21.3 billion that the U.S. petroleum industry laid out in total capital and
exploratory expenditures during the two-year period 1972-1973. In the next decade, however,
and for as far ahead as we can see, the industry's total capital requirements -- in current
dollars, with a five per cent inflation factor -- are expected to be on the order of $40 billion
a year. The question is: Where is all of that capital going to come from?
The coming capital bind is there for all to see. The pool of funds available to the
nation for pollution control is limited — including, certainly, funds for water pollution
control. We simply do not have the wherewithal to accomplish all that has been promised by
the Federal Water Pollution Control Act within the time frame specified.
I would like to turn now to some suggested amendments to the Act. Like everyone
else who must deal with the provisionsof the Act, we in the petroleum industry have compiled
a long list of such suggestions. This evening, however, I want to mention just three of the
most important ones, designed to provide some of the things now sorely lacking — namely,
achievable deadlines, realistic water quality goals, and ensuing environmental benefits
commensurate with the costs to society:
1 . Extension of the 1977 Deadline. EPA has already concluded that because of lack
of funds, 50 per cent-- some 9000 — of the municipalities that will be serving 60 per cent
of the 1977 population will not be able to comply with the 1977 sewage treatment require-
ments of the Act. The Agency therefore recommends that the 1977 deadline be extended to
about 1983 — for municipalities. That may or may not be the optimal date, but we in the
petroleum industry agree that an extension certainly is indicated. We also feel, however —
and very strongly — that the deadline should be extended for industry, too, especially since
such extension would not prevent EPA or the states from requiring earlier compliance where
possible through the individual permit program.
As you know, a part of the problem is that some refineries and smaller facilities like
terminals and service stations discharge to municipal sewers. Such facilities must rely on
municipal treatment plants for removal of compatible pollutants in order to comply with 1977
requirements. It is worth adding that industry must bear 100 percent of the costs, while
municipalities need bear only 25 per cent of the capiral costs for federally approved projects.
Clearly, to extend the deadline for municipalities and not for industry would be inequitable
in the extreme.
2. Modification of the Requirement for Best Available Technology Economically
Achievable for 1983. Once both municipalities and industry are in compliance with the
first treatment level requirements — at whatever date compliance is achieved — it would
be unwise to continue to require application of tremendously expensive second level tech-
nology everywhere, indiscriminately. The National Commission on Water Quality's studies
-------
Ill
show that water quality is improving faster than expected and that implementing the 1977
requirements of the Act will produce better water quality than was anticipated at the time
of enactment. The Commission's studies also show, as I mentioned earlier, that the benefits
of implementing the second level of technology would not in general be commensurate with
the costs. The limitations and other controls imposed by the Act for second level technology,
therefore, should be applied only in those cases where they are necessary to achieve
receiving water quality and then only when the environmental benefits exceed the costs to
society.
3. Redefinition of the Term "Pollutant." In general, the Act fails to distinguish
between natural and man-made pollutants. The Act also fails to recognize that substances
now termed pollutants are not in all cases harmfn' Salt and suspended solids, for example
— whether natural or resulting from man's activities — often occur without harm to waters
or to the biota in them. What is needed is a definition of "pollutant" which excludes
substances which are discharged in quantities too small to cause damage. It would then be
possible to establish — not the Act's present national goal of elimination of discharge of
pollutants, which just is not going to happen — but a national goal of elimination of dis-
charge of harmful quantities of pollutants, which well could happen.
These amendments, if enacted, would not solve all of the nation's water quality
problems. They would clarify some of the nation's water quality goals and set reasonable
deadlines and procedures for attaining them. And from clear goals and reasonable dead-
lines come the capacity of regulatory agencies to regulate wisely and the capacity of industry
and municipalities to comply at acceptable cost.
This position was championed last week by Bill Ruckelshaus, "Congress has an almost
neurotic fixation for setting deadlines on environmental matters that are unachievable. Then
by promising something we cannot achieve we become incapable of measuring progress. A
delay of one year in meeting the standard that denied reality to begin with makes it look
as if EPA is copping out and all the progress that has been made is ignored. Adjustments to
some of the schedules are necessary for a sane and a reasoned approach to such a vital
national goal. The question should not be whether an environmental law is tougher but
whether it is wise and in the best interests of the whole country. "
John Queries, last December, also espoused these views in a letter to me, "On the
side of disappointment, we are all now wiser but sadder as to the ability and solve national
pollution problems quickly. It is now evident that the work of cleaning up pollution will
require intensive efforts throughout the next ten or fifteen years and that many important
environmental goals will not be reached any sooner than that. Passing a law does not by
itself solve a problem. Manpower, money and in some cases technica, advances must also
be provided and experience leads me to conclude that in a national program tackling complex
problems the achievement of results takes patience as well as persistence. The risk this
creates is that if the intensity of effort weakens along the way some of the goals may never
be achieved. "
The national badly needs to move ahead toward its water quality goals at a more
considered pace, allowing ample time to take into account ciM of its peoples' needs and
-------
112
how best and most economically to meet them. To answer to that need, let's change course
and speed.
BIOGRAPHY
Peter N. Gammelgard is the Vice President of the American
Petroleum Institute responsible for Industry Affairs. He holds
a B.S. degree in Engineering from Northwestern University.
He served as a Commander in the U.S. Navy during World
War II. He has been a Vice President of the Air Pollution
Control Association and past President of the National
Petroleum Refiners Association. He has participated in
environmental conferences around the world and has been
active in the affairs of the World Energy Conference.
-------
SESSION III
"BIOLOGICAL TREATMENT"
Chairman
George W. Reid
Regents Professor of Civil Engineering
University of Oklahoma, Norman, Oklahoma
Speakers
W. Wesley Eckenfelder
"Activated Sludge Treatment of Petroleum Refinery Wastewaters
an Overview"
James F. Grutsch
"A New Perspective On The Role Of The Activated Sludge Process
and Ancillary Facilities"
Charles E. Ganze
"Case History On Biological Treatment Of A Petroleum Refinery
Wastewater"
Mohammed A. Zeitoun
"Optimization Of The Activated Sludge Process Through Automation"
Fred M. Pfeffer
"The National Petroleum Refinery Wastewater Characterization
Study"
113
-------
114
BIOGRAPHY
George W. Reid, born 1917, Indianapolis, Indiana,
attended Shortridge High School, Purdue, Harvard and Johns
Hopkins Universities, studying Civil and Sanitary Engineering.
Employed by the Indiana State Health Department, a
sanitary engineer with USPHS, MCWA and CDC. Has
taught at Purdue, Johns Hopkins, Florida and George Tech.
Currently teaching at the University of Oklahoma as
Regents Professor of Civil Engineering and Environmental
Sciences and is the Director of the Bureau of Water and
Environmental Resources Research. Holds degrees from
Purdue and Harvard Universities.
Has published numerous articles, etc., in the field (170);
is and has been consultant to the USPHS, EPA, Interior, USAF,
WHO (PAHO), AID and others in water and public resources systems. Recently involved
in extensive research on modeling of natural and man-made public systems. Lectures at r
recent Urban Water Systems Conferences and seminars in Colorado, New Mexico, Bogota,
Lima, Europe, etc.
-------
ACTIVATED SLUDGE TREATMENT OF PETROLEUM REFINERY WASTEWATERS
AN OVERVIEW
W. Wesley Eckenfelder, Jr.
Distinguished Professor of Environmental cTnd Water Resources Engineering
VanderbiIt University, Nashville, Tennessee
This discussion will consider aerobic biological treatment in the activated sludge
process. The two basic equations which describe the aerobic biological mechanism are:
(a1) cells (a)
Organics + O2 + N + P > new cells + COo + H2O
(k)
(BOD,COD,TOC)
(s)
+ nondegradable soluble residue (1)
(b') (b)
Cells + O2 > CO2 + H2O + N + P + nondegradable cellular
(Xy) residue (2)
Equation (1) is the synthesis equation and occurs in the presence of degradable
organics and biological cells with oxygen, nitrogen, and phosphorus present. Equation
(2) is the endogenous breakdown of the biological cells, and occurs continuously in an
aerobic environment.
In order to design an aerobic biological process it is necessary to balance equations
(1) and (2). Since petroleum refinery wastewaters contain an unknown mixture of
organics, it is necessary to operate a laboratory or pilot plant program to define the
coefficients to balance these equations. In some cases experience elsewhere may be
employed. The pertinent parameters for design purposes in equation (1) are the coefficient
a1 which is the fraction of organic matter removed in terms of BOD, COD or TOC
oxidized for energy; the coefficient a which is the fraction of organic matter removed
converted to new cells expressed as volatile suspended solids; the rate coefficient k is a
measure of the rate of oxidation of the organics present in the wastewater. The
coefficient b in equation (2) is the fraction per day of the biodegradable volatile
suspended solids oxidized. Since it requires approximately 1 .4 pounds of oxygen to
oxidize one pound of biological cells as volatile suspended solids, the coefficient b1 is
generally approximately 1.4b_. For all practical purposes organic matter removed in the
process is either oxidized for energy or converted to biological cells (except for the
small fraction of organics which result in nondegradable by-products as shown in equation
(1) ). Therefore, on a COD or ultimate BOD basis, the coefficients a +a' = 1 . As
shown in equations (1) and (2), the biomass as volatile suspended solids is denoted by the
symbol X and the organics expressed as BOD, COD or TOC by the symbol s. It should
be noted that the nondegradable soluble residue indicated in equation (1) results from the
accumulation of nondegradable organic by-products from the biological oxidation process.
It has been found that the magnitude of this value usually falls between 1 and 5 percent
of the influent COD. The nondegradable cellular residue as shown in equation (2) will
amount to approximately 23 percent by weight of the volatile
115
-------
116
suspended solids generated in equation (2). It can therefore be concluded that the pertinent
parameters required for bio-oxidation design are those indicated in equations (1) and (2),
namely a1 , a, k, b and b1. Typical values found for petroleum refinery wastewaters are
summarized in Table 1. These reactions as they would occur in a batch oxidation system
are shown in Figure 1 .
Figure 2 illustrates the breakdown of 1 pound of degradable COD in accordance with
equations (1) and (2). Analysis of the wastewater in Figure 2 indicates that the COD/BOD-
ratio is l:0.75;a' and a are 0.5 on a COD basis. It is assumed that 1.4 Ibs CL will oxidize
1 Ib cell mass as VSS.
DESIGN RELATIONSHIPS
As can be deduced from equations (1) and (2) nondegradable volatile suspended solids
will accumulate in the process as auto-oxidation of the cellular mass proceeds.
In order to develop process design relationships it is therefore first necessary to define
the degradable fraction, x, of the volatile suspended solids (biomass) in the aeration process.
The degradable fraction is a function of the sludge age or the length of time the volatile
suspended solids are in the process as shown in Figure 3. Sludge age in the activated sludge
process can be defined by the relationship:
X (Ibs)
Sludge Age =^- ^^ (3)
in which AX is the volatile suspended solids wasted per day from the process. As the sludge
age is increased the nondegradable materials accumulate and the degradable fraction decreases
as one would expect. The relationship shown in Figure 3 represents average conditions for a
petroleum refinery wastewater.
The kinetic relationship employed for the design of a completely mixed activated sludge
process is based upon the removal of most soluble simple organic compounds as a zero order
reaction. In a multi-component wastewater a number of zero order reactions occur con-
currently. The summation of these zero order reactions can be expressed in an exponential
form for a batch reaction or employing equation (4) for a completely mixed continuous flow
reactor.
v
In cases where the influent concentration of the wastewater varies markedly, equation (4)
may be modified to equation (5) which considers variable influent wastewater strength.
S - S
S A (5)
X t ' " V o
v
When COD or TOC are employed as a parameter of organic strength, the nondegradable
portion of the TOC or COD must be substracted from the total as shown in Figure 4. It
-------
117
should be noted that the reaction coefficient k in equations (4) and (5) relates to a specific
wastewater composition. If the wastewater composition varies the reaction rate coefficient
k may also be expected to vary.
It should be noted that the effluent concentration defined by equations (4) and (5) are
soluble concentrations. In order to determine the total BOD in the effluent it is necessary
to add that contributed by the effluent suspended solids. While the BOD contributed by
the effluent suspended solids will be a function of sludge age (increasing sludge age de-
creases the degradable fraction) for most activated sludge processes in the petroleum refining
industry, a value of approximately 0.3 mg BOD/mg SS may be used.
The total oxygen requirements in the biological process are related to the oxygen
consumed to supply energy for synthesis (equation 1) and the oxygen consumed for endo-
genous respiration of the biological cells (equation 2). Total oxygen requirements can
therefore be defined by equation (6) and shown in Figure 5.
09 /day = a ' (S - S ) + b ' x.(X ) (6)
z — o e — v
If nitrification is occurring in the process this oxygen will also be included in equation (6)
and considered in the coefficient a1.
Sludge accumulation from the biological treatment of petroleum refinery wastewaters
can be similarly calculated. Volatile suspended solids accumulation, AX , is equal to that
resulting from synthesis (equation 1) plus the influent volatile suspended solids not degraded
in the process, minus the volatile suspended solids oxidized by endogenous metabolism
(equation 2). This relationship is shown in equation (7) and Figure 6.
AX = a (S - S ) + S. - bx(X ) (7)
v — o e i v
One further consideration in the activated sludge process is that the biological sludge
generated in the process flocculates, settles and compacts in the final clarifier. In most
cases the sludge settling characteristics have been found to relate to the organic loading to
the process (F/M) expressed as pounds of BOD applied per pound of MLVSS.
SQ
F/M = --
In order to insure good settling properties, the F/M should be maintained within an accept-
able range as shown in Figure 7. If the F/M becomes too high, frequently non-settling
filamentous growths will result creating sludge bulking. If the F/M remains too low, auto-
oxidation of the biological floe and resulting floe dispersion causes an increase in effluent
suspended solids and a bulking condition.
The final clarifier is an important consideration in the process design of the activated
sludge process. The solids flux to the clarifier and hence the underflow solids concentrations
attainable are related to the zone settling velocity of the biological sludge mixture. These
relationships are shown in Figure 8.
-------
118
ATTAINABLE EFFLUENT QUALITY
Considering the performance of biological treatment plants the primary factors which
will affect effluent quality are aeration basins, temperature, changes in organic loading
(as reflected in the F/M) and changes in product mix (as reflected in the k value). Where
temperature is critical for effluent quality during winter operation the aeration system should
be selected to minimize heat loss, usually a submerged aeration system. By contrast, high
temperatures during summer operation can be minimized by the use of surface aerators which
maximize heat loss from the aeration basin. Covered basins as employed in the high purity
oxygen process will maintain high temperatures during cold weather operation.
Changes in organic loading can best be dampened by equalization. Since, all other
things being equal, the effluent BOD will be directly proportional to the influent, the
equalization basin can be designed for whatever maximum effluent levels are permissible.
Variations in organic loading can also be compensated by increasing the sludge recycle,
thereby maintaining the F/M constant. The most difficult effect to assess is changes in the
reaction rate k due to changes in wastewater composition. In some cases the plant operating
conditions can be modified for a change in plant production schedule. Equalization will
also help to dampen this variability. While the effects of temperature and organic loading
variation can be calculated from the design relationships, it is not possible in most cases to
calculate the effect of product mix. In this case a statistical approach using operating data
is necessary.
Long-term case histories of a number of activated sludge plants in the petroleum refining
industry have been reported in a study for the National Commission on Water Quality (2) and
are summarized in Table 2. The variability in effluent quality which might be expected for
these parameters are shown in Table 3.
It may be expected, therefore, that as a refinery becomes more complex there will be
a greater variation in wastewater composition and hence a greater variation in k. The net
effect of this considering equations (4) and (5) is to increase the variability in effluent con-
centration, S , on a daily basis. Refineries in the northern part of the United States would
be expected to have a greater monthly variability due to seasonal temperature effects than
those in the south.
LIST OF SYMBOLS
a Ibs VSS produced per Ib BOD, COD or TOC removed
a1 Ibs oxygen required to oxidize one Ib of BOD,COD or TOC
A Area
b Ibs VSS destroyed per day per Ib oxidizable MLVSS
E1 Ibs oxygen per day per Ib oxidizable MLVSS
C. Oxygen concentration in the basin
F/M Food-to-Microorganism Ratio
hr Hour
k Organic removal rate coefficient, days
MLVSS mixed liquor volatile suspended solids
-------
119
N oxygen transfer coefficient, corrected for field conditions
N Oxygen transfer coefficient under standard conditions
OR Clarifier overflow rate (gpd/sq ft)
Q Flow rate
R Recycle flow
S effluent organic concentration
S Influent organic concentration
x° T-
t Time
T Temperature
T Basin temperature
S. Average influent suspended solids
X Volatile suspended solids
a Ratio of oxygen transfer rate of waste to water
REFERENCES
(1) Eckenfelder, W. W., "Principles of Biological Treatment," Vanderbilt University,
Nashville, Tennessee, 1975.
(2) Engineering Science, Inc., "Petroleum Refining Industry Technology Costs of
Wastewater Control," Report to the National Commission on Water Quality, June, 1975.
(3) Engineering Science, Inc., "Activated Sludge Treatability of Petroleum Refinery
Wastewaters," 1975.
BIOGRAPHY
W. Wesley Eckenfelder holds a B.E. from Manhattan College;
a M.S. from Pennsylvania State University; and a M. E. from
New York University. Professor Eckenfelder has served on
the faculty of Manhattan College, the University of Texas at
Austin, the University of Delft in Holland, and is now a
Distinguished Professor of Environmental and Water Resources
Engineering at Vanderbilt University. Wes is also Board
Chairman of AWARE, Inc. and has written over 200 publi-
cations and 11 books on water quality management. Pro-
fessor Eckenfelder has received many awards.
-------
120
DISCUSSION
Bob Worton: Would you care to comment on the effect of po I/electrolytes on keeping sus-
pended solids down in the final effluent when operating with a high degree of nitrification.
Wes Eckenfelder: The polyelectrorytes probably would not help much, since denitrification
occurring in the final clarifier would lead to floating sludge due to the nitrogen gas bubbles
released during the process of denitrification.
Bob Worton: Do you know any research in this area?
Wes Eckenfelder: Specifically on nitrification, no. I think the answer to this problem,
however7~is to remove the sludge from the final clarifier before active denitrification sets
in.
Juanito Chavez: My first question is also on the polyelectrolyte injection. In commercial
application is there any specific point where you would inject the polyelectrolyte in such a
way that you effect good distribution and at the same time not destroy the biological floe.
Wes Eckenfelder: To my knowledge, I know of no full-scale plant that has been using
polyelectrolytes for effluent suspended solids control for long periods. All of the experi-
mental studies that I have been associated with have injected the polymer prior to going
into the final settling tank. In view of the fact that most of these have not been pre-
engineered to optimize the process, much of the data is erratic.
Bill Oberman : Would you care to comment on the use of dissolved air flotation instead
of secondary clarification, where poor settling sludges are involved?
Wes Eckenfelder: Some years ago I was involved in engineering an activated sludge system
employing BAF in lieu of secondary clarification. While a reasonably concentrated sludge
was obtained for re-cycle, the effluent suspended solids from the flotation unit were highly
variable. It is possible that advances in flotation practice may overcome this problem to
date.
Larry Eckelberger; What type of variability factors can be achieved in your opinion?
Wes Eckenfelder: I quote here data that generates out of the Petroleum Industry report to
the National Commission on Water Quality. Monthly variability in BOD is 2.7 times the
average with daily variability in BOD being 5.4 times the average. For effluent suspended
solids, monthly variability is 2.5 and daily variability 3.8. I consider these realistic limits.
John Garner: This question may be a little premature and may be that it will be answered
later, but I have found in raw waste BOD that I require 20 to 25 aeration horsepower and in
practical application I find I need 50-55 horsepower.
Wes Eckenfelder: I have no figures specifically for the petroleum refining industry, but in
the pulp and paper industry, we generally estimate 45 pounds of BOD removed per day per
horsepower as a rule-of-thumb. I see no reason why this would be markedly different in
petroleum.
-------
121
TABLE 1 "BIOLOGICAL TREATMENT COEFFICIENTS FOR PETROLEUM WASTES (3)"
Refinery
Category Size
(EPA) (MBPD)
E
C
D
C
E
B
C
406
60
90
63
427
205
88.4
k*
I/days
BOD COD
4.15 2.74
Sludge Growth Coefficients Oxygen Requirements Coefficients Influent
BOD Basis COD Basis BOD Basis COD Basis mg/1
a b1 a b1 a1 b1 a1 b1 BOD
0.595
3.84
97
86
92
0
79
0.5 0.08
0.5 0.05
7.24
0.5
0.5
0.44
0.26
0.2
0.43
0.06
0.06
0.1
0.03
0.08
0.10
0.47 0.28
0.57
0.06
0.34
0.35
0.46
0.40
0.52
0.1
0.11
0.06
0.08
0.05
0.01
0.14
0.6 0.05
0.88 0.10 -
244
575
396
153
600
345
160
* at 24 C k = organic removal rate at 24 C
1 based on total VSS
Sub-
category
C
C
B
B
B
(lOOOb/S.)
a
60
63
135
63
205
BOD
13
16
18
28
12
COD
174
130
132
183
112
mg/1
TSS Oil
78
30
23
41
15
9.4
9.0
9.9
Phenol
.053
.15
.04
NH3-N
1.1
6.7
8.2
COD
509
980
782
428
170
563
806
TABLE 2 "LONG TERM EFFLUENT CONCENTRATIONS FROM THE ACTIVATED SLUDGE PROCESS (2)"
Capacity
S
0.27
TABLE 3 "VARIABILITY OF EFFLUENT LEVEL IN ACTIVATED SLUDGE TREATMENT (2)"
Daily Min
Max
Ave
Monthly Min
Max
Ave
BOD
3.8
7.6
5.4
2.2
4.6
2.7
COD
1.9
5.0
3.6
1.5
1.9
1.8
TOC
3.5
2.9
TSS
2.4
5.2
3.8
1.5
3.3
2.5
Oil
3.9
5.2
4.7
1.8
2.9
2.2
Phenols
4.5
5.2
4.9
1.4
3.4
2.5
NhL-N
O
9.
10.
10,
2
4.1
3.4
-------
122
o
LJ
O
z
o
o
Nondegradable
Residue
Figure I. SCHEMATIC OF AEROBIC BIOOXIDATION PROCESS
(COD) a'=0.5
(BOD5)a' = 0.75
a = 0.5 (COD)
a = 0.5/1.4 = 0.36(VSS)
0.36 Ib
VSS
NEW CELLS
0.27-K4= 0.39
0.23-0.36 = 0.083
IbVSS RESIDUE
0.39 Ib 0,
USED
Figure 2 .
-------
0.8
0.6
0.4
K n p
<•> w.t
o
o
03
U.
O
o:
u.
0.2 0.4 0.6
_F_ /mg influent BOD
M I mg MLVSS-day
123
0.8
1.0
Figure 3. CORRELATION OF BIODEGRADABLE FRACTION OF
MLVSS, x, WITH ORGANIC LOADING, F/M.
400
300
200
o
V)
100
BOD-
Sample Correlation
For COD or TOC
Shows effect of
variable composition
on organic reaction
rate, K
Non- biodegradable fraction
V
I i l
I
0 10 20 30 40 50 60 70
SOLUBLE EFFLUENT BOD OR COD, Se(mg/l)
Figure 4. DETERMINATION OF ORGANIC REACTION RATE
COEFFICIENT ( EQUATION 5 )
-------
124
1.0
0.2
X
x
BOD-
Somple correlation
for COD or TOC
I
0.2 0.4 0.6 0.8
Sr [ Ib organics removed ^
xXtf Vlb biodegradable VSS-dayJ
1.0
Figure 5. DETERMINATION OF OXYGEN UTILIZATION COEFFICIENTS
(EQUATION 6)
Sample Correlation
indicates influence
influent VSS on
sludge sythesis
coefficient
Sample
Correlation —
for COD or
TOC
.0.4 0.6 0.7
/Ib organics removed \
xXv\lb biodegradable VSS-daJ
Figure 6. DETERMINATION OF SLUDGE PRODUCTION COEFFICIENTS
(EQUATION 7 )
-------
140
^ 120
o>
E
125
o
(O
a
UJ
a
z
UJ
a.
V)
CO
h-
UJ
U.
u,
UJ
100
80
60
40
20
ZSV
±
_L
0.2
0.4 0.6
Ib influent total BOO
0.8
M
4.0
3.0
2.0
1.0
o
o
Ul
I-
UJ
CO
Ul
o
N
1.0
Ib MLVSS-day
Figure 7. CORRELATION OF EFFLUENT SUSPENDED SOLIDS AND ZONE
SETTLING VELOCITY WITH F/M.
"0 0.2 0.4 0.6 0.8 1.0
TOTAL SUSPENDED SOLIDS CONCENTRATION (%)
Figure 8 . BATCH SOLIDS FLUX CURVE FOR CLARIFIER DESIGN
1.4
-------
RAW WASTE
GRAVITY OIL SEPARATION! EQUALIZATION
I
Skimmed Oil
RESIDUAL OIL
SSUSPENDED
SOLIDS RE-
MOVAL
BIOLOGICAL BOD REMOVAL
a FINAL
CLARIFICATION
TREATED
EFFLUENT
CO
o
EQUALIZATION f]~
BASIN
OIL
SKIMMER
(Floating)
| ^ Return
I i Sludge
Dewatering Float Sludges ,
TO
-^
RECOVERY
OILY SLUDGES TO TREATMENT 8 SUBSEQUENT
RECOVERY OR DISPOSAL BY INCINERATION
Conditioning
Chemicals
TO FINAL
DISPOSAL -*
(LANDFILL,
INCINERATION, ETC.)
T Conditioning
GRAVITY THICKENER ^J^ Chemicals
OR AIR FLOTATION
Supernatant
f THICKENER J.^ Chemical!
.OTATION -T^f X I
Inn* \ /
\^S
Allernotive I
AEROBIC
DIGESTION
BASIN
SLUDGE FLOW
-PROCESS WASTE FLOW
DEWATERING THICKENING STABILIZATION
EXCESS BIOLOGICAL SLUDGE HANDLING FACILITIES
FIGURE 9:
SCHEMATIC PROCESS DESIGN FLOWSHEET FOR ACTIVATED SLUDGE SYSTEM
-------
"A NEW PERSPECTIVE ON THE ROLE OF THE ACTIVATED SLUDGE
PROCESS AND ANCILLARY FACILITIES"
J. F. Grutsch
Coordinator of Environmental Projects, Standard Oil (Indiana)
R. C. Mallatt
Manager, Environmental Conservation, Standard Oil (Indiana)
Refiners have found the activated sludge process (ASP) to be a very effective and
broadly applicable process. The versatility of the ASP has impressed many users and
emphasis has been placed on this aspect of the process. However, design and operation
of the ASP have not proceeded in a direction that is optimum for meeting future effluent
quality goals. The ASP is generally used to achieve major reductions in suspended and
colloidal material, as well as dissolved substances. When the refiner is almost wholly
dependenton ASP to achieve end-of-pipe treatment objectives, this operating mode may
achieve the best overall reduction in contaminants. In the future, however, the refiner
will be confronted with very restrictive effluent limits on suspended solids, BOD, COD,
ammonia, and some bioresistant materials, as well as variability limits for each parameter
as conditions for obtaining an NPDES permit. Under these conditions, the ASP becomes
only one of a series of process steps required to meet the objectives. When viewed in
this context, the ASP can be optimized to achieve further control of process variability
and increased removal of soluble contaminants to concentration levels below that achiev-
able by activated carbon. The key to optimizing the ASP in this new mode is to design
and operate the preceding end-of-pipe process facilities for maximum removal of inter-
fering colloidal and suspended contaminants.
APPLICABLE REFINERY TREATMENT TECHNOLOGY
Best practicable control technology currently available (BPCTCA), as defined by
the federal Environmental Protection Agency (EPA) for petroleum refineries, is based on
both in-plant and end-of-pipe control technology. Suggested in-plant controls include:
1. Installation of sour water strippers to reduce the sulfide and ammonia
concentrations entering the treatment plant.
2. Elimination of once-through barometric condenser water by using surface
condensers or recycle systems with oily water cooling towers.
3. Segregation of sewers, so that uncontaminated storm runoff and once-
through cooling waters are not treated normally with the process and
other contaminated waters.
4. Elimination of polluted once-through cooling water by monitoring and
repair of surface condensers, or by use of wet and dry recycle systems.
EPA recognizes also that good housekeeping practices and precautionary measures
127
-------
128
taken during the turnaround of operating units are important considerations that can have a
substantial impact on waste loads discharged to end-of-pipe water treatment facilities.
BPCTA end-of-pipe treatment technology is based upon the existing waste water
treatment processes currently used in the industry. Although various options are possible,
as shown in Figure 1, the most sophisticated and complete end-of-pipe treatment sequences
usually make provision for equalization and storm water diversion; initial oil and solids
removal in API or parallel plate separators; further oil and solids removal in filters, clarifiers,
or dissolved air flotation units; carbonaceous wastes removal by any of several biological
oxidation processes or combinations thereof; and a final clarification or polishing step,
consisting of chemical coagulation or destabilization and either dissolved air flotation or
granular media filters for phase separation. For BATEA, in 1983, the contaminant con-
centrations achievable by an additional activated carbon process element is an indicated
objective.
More specifically, the principal objectives of the end-of-pipe treatment sequence for
process design considerations are as follows:
1. Maximize removal of soluble contaminants contributing to biochemical oxygen
demand (BOD).
2. Maximize ammonia removal.
3. Maximize removal of contaminants contributing to chemical oxygen demand
(COD).
4. Maximize final effluent clarity.
Achieving these primary objectives gives rise to some secondary design objectives:
5. Effect the efficient removal of gross quantities of solids and oil.
6. Provide for hydraulic and chemical equalization of refinery wastes.
7. Minimize the generation of excess biological sludges.
8. Minimize any odor and bulking problems from biological sludges.
9. Minimize IOD (Immediate Oxygen Demand) of water entering secondary
treatment.
10. Minimize sulfides in the effluent.
1 1 . Discharge an effluent containing dissolved oxygen and a good threshold odor
number.
-------
129
SECONDARY TREATMENT IS THE KEY TO OPTIMIZING THE END-OF-PIPE
TREATMENT SEQUENCE
If the overall treatment sequence is to be optimized, the principal objectives
listed above point to secondary treatment as the key end-of-pipe treatment element.
Various other process steps are designed to enhance its performance. Further, in contrast
to prevailing practice, secondary treatment must focus on removal of only soluble con-
taminants. This limited objective for secondary treatment requires the objectives of the
remaining elements in the treatment sequence to be defined, establishes a basis for operat-
ing conditions that maximizes the performance of each treatment element, thereby optimiz-
ing the effectiveness of the overall treatment sequence. Adherence to this design philos-
ophy can yield an effluent quality superior to that achievable with an add-on activated
carbon treatment element.
OPTIMIZING PRETREATMENT PAYS OFF
Optimizing pretreatment not only permits the ASP to be designed and operated in an
especially attractive mode of exceptionally high sludge age as will be described later,
but major savings in energy costs and capital costs can be made. Consider, for example,
the indicated raw waste load in Figures 2-4 to be average data. These quantities are
highly variable in a refinery and may be increased five to ten fold by many situations
including, for example, (1) an occasion of high pH or excessive sulfides which contributes
to increased colloidal matter with high zeta potentials, (2) sulfonate waste from an alky-
lation unit, which peptizes and disperses contaminants, (3) loss of catalyst fines that
remain dispersed in water, (4) loss of residual oils, coke fines, wax, clay, or lube stocks
and additives, which emulsify oil and solids and tend to escape the API separator, (5) loss
of efficiency in the API separator by hydraulic overload, etc. Excessively high raw waste
loads from the API separator are characterized by the fact that they consist primarily of
discontinuous phase materials. Adequately providing for chemical and physical equali-
zation in the intermediate treatment section will provide a more consistently uniform
quality effluent over a very wide range of raw waste loads.
Maintenance of high sludge age in the ASP requires that the feed be free from oily
inerts that accumulate in the activated sludge mass and impart undesirable properties to
it that preclude operation in the high sludge age mode. In large and complex refineries
the oil and grease in the raw waste load can vary over a year's time as shown in Figure 2.
Whereas the 50% probability (median) is only 70 mg/1 oil, about 5% of the data may be
in the 600+mg/l range due to causes previously outlined. Optimized ASP operation at
very high sludge ages has been achieved with excellent sludge properties when all dis-
continuous phase oil and solids are removed. Pretreatment can achieve these oil levels
indicated in Figure 2 when treatment consists of chemical destabilization and filtration.
Removing oil from the process water effluent prior to biological treatment recovers
the oil in a sludge much easier and less costly to handle than if the oil is commingled in
a waste activated sludge. Waste activated sludge is costly enough to dewater but when
commingled with oil, the oil blinds precoat filters, increases significantly the volume of
waste sludge to be dewatered, and makes the final disposition of dewatered oily sludge
more troublesome.
-------
130
Equalizing feed quality in terms of organic loading can also be achieved by the same
pretreatment as shown for BOD/COD data in Figures 3 and 4. Colloidal and suspended
matter contributes significantly to the total BOD/COD;and periods of storm flow which
flushes sludge and silt from sewers, high pH, loss of emulsifiers, coke fines, clay fines, etc.,
typically give a highly variable raw waste load which is frequently underestimated because
of sampling and testing difficulties. Removal of the discontinuous phase contaminants can
have a large impact on reducing the total organic load and the variability in quality of
water entering secondary treatment.
Equalizing the organic loading results in substantial savings in capital and operating
costs, simpler operations, and better effluent quality with less variability in quality of
water leaving the secondary treatment facilities. For example, data from Figures 2, 3, and
4 are as follows:
% Probability Less
Than Indicated Value
Contaminant Treatment After 50 95 98
Oil and Grease, mg/1 Primary 70 600 800
Intermediate 4 10 12
BOD, mg/1 Primary 185 380 400
Intermediate 80 96 105
COD, ma/1 Primary 400 980 1,400
Intermediate 220 310 350
An engineer faced with design of an ASP to operate in the nitrification mode at 2 mg
DO/1 with uniform effluent quality faces an almost insurmountable challenge, both econo-
mically and performance-wise, if he has to deal with the variability of the raw waste load.
Considering that data in the preceding table represent a year of operations, they also reflect
(1) seasonal, (2) weekly, (3) daily, and (4) hourly cycles. The final design of an ASP
based on primary effluent would be such a compromise that year-round operation would be
unsatisfactory, yet the unit would be expensively over-designed. If, for example, a guide-
line of 1 Ib. 0,~/lb. COD is used and the 98% probability COD value is selected as a design
basis, the raw waste would require four times the aeration horsepower the intermediate
treated waste does. Actually, it would require substantially more, because the alpha
(oxygen transfer) characteristics of raw waste are not as good as the intermediate treated
waste.
Another example of the impact of the representative data on the ASP is recognition
that, for the same sludge age, the equilibrium solids level in the ASP relates directly to
organic loading. One critical aspect of clarifier design is the solids flux rate. The
increased solids circulating at the higher organic loading can increase substantially the
clarification capacity required. Thus, once the engineering and operating details of an
ASP unit are addressed, the designer soon recognizes that proper design of the intermediate
treatment section will contribute to lower costs, improved and simpler operation of the
secondary treatment facilities.
-------
131
PROCESS CONTROL OF THE ACTIVATED SLUDGE PROCESS
A key to designing and optimizing the operation of an ASP unit as an integral part
of an end-of-pipe treatment sequence is first, to recognize that it is only a part of the
treatment sequenceiand operation of each part should be optimized in light of what it can
do best, or is required to do by another part to perform optimally; and second, to provide
for use a method of process control that optimizes performance.
Performance of the ASP in terms of BOD/COD removal relates to organic loading;
i.e., the food-to-mocroorganism (F/M) ratio. Eckenfelder (1) has summarized concisely
the effect of organic loadings on ASP performance when treating domestic wastes in the
manner shown in Figure 2. These data and the work of others (2,3) show that better BOD
removals, total oxidation, nitrification, and less excess sludge are achieved with lower
F/M ratios. The consideration which most often limits the ability to produce a high qua-
lity effluent is development of an activated sludge mass with poor settling characteristics
at organic loadings under .2 F/M, where increasing values of SVI are observed. Sludge
volume index (SVI) is a measure of the sludge settling characteristics and is the volume
occupied by 1 gm of sludge after the aerated liquor settles for 30 minutes. While the SVI
test has shortcomings beyond the scope of this discussion, it can be a useful process indi-
cator. On the basis of these data, optimal ASP operating conditions are usually cited
as F/M loadings in the range of 0.2 to 0.5.
Our ASP development work has shown that optimum ASP operating conditions are
achieved by removing essentially all the dispersed colloidal contaminants by chemical
destabilization and filtration prior to ASP treatment. With removal of colloidal matter
in the feed, the SVI of the activated sludge did not increase at sludge ages well in excess
of 35 days, but consistently measured 70-80 with excellent flocculating and settling
properties.
Thus, contrary to comments in the literature that operation at high sludge age
(F/M
-------
132
material that intensifies the zeta potential and enhances the dispersion of the biomass.
Forster's ZP-SVI relationship in Figure 6 shows SVI to be remarkably sensitive to ZP and
it is easily perceived how the highly negative charged particles in refinery primary efflu-
ents cause the SVI to deteriorate.
For refinery effluent waters, the impact of accumulating colloidal solids in the acti-
vated sludge mass must not only be addressed, but dispersed phase oil and oily solids also
must be considered. Oily materials contribute to electrokinetic effects and destabilization
of the oily colloidal system allows the agglomeration and coalescing of the oily matter into
the activated sludge mass; this, in turn, leads to other deleterious effects. For example,
the coalesced oil matter associating with suspended solids can cause so little gravity
differential with water that the associated materials pass through the clarifier to the effluent.
The coalescing of the oily materials also encourages the formation of a stable, voluminous
emulsion mass consisting of oil, water, solids and air. Typically this mass floats on the
surface of the aeration tank and the discontinuous emulsion phase builds rapidly because
its surface properties encourages agglomeration of like materials. The emulsified mass
does not biodegrade significantly. Dispersion, or redispersion, of emulsified material in
the aeration tank typically causes the material to enter the clarifier where it collects on
the surface. If the clarifier is not designed to trap floating material, the material
escapes in the effluent causing a significant degradation in effluent quality.
Well-settling activated sludges have been observed to have a density of about 1.016.
When associated with oily matter, it is misleading to simply calculate the final sludge
density from composition and component density data, because the oily, emulsified mass is
much more buoyant than calculated, due to entrapped air.
ASP PROCESS CONTROL DYNAMICS
The amount of colloidally dispersed inert solids and oil entering the aeration tank
affects the degree of process control available to the operator. Colloidal and suspended
matter in the influent to the ASP at quite low concentration levels can make the process
non-operable at the optimum conditions.
The ASP process control is via the food-to-microorganism (F/M) ratio, or perhaps
more practically the sludge age (SA). The sludge age and F/M ratio are related as follows:
J_ = a ( A F/ A T) - b (Equa. 1)
SA M
where, a = the sludge yield coefficient,
b = the endogenous rate coefficient,
M = the mass (Ibs) of microorganisms in the system,
SA = the sludge age (days), and
A F /AT = the mass (Ibs) of food (BOD or COD) supplied per day.
The F/M process control involves measurement of the BOD or COD load ( A F/A T)
per day and adjustment of the sludge inventory (M) to maintain a desired ratio. The sludge
age method of process control can provide a simple hydraulic means to achieve the same end
-------
133
by sludge wastage. The sludge wastage rate to maintain an indicated sludge age is cal-
culated from the equation:
u/ * /KA^nN i Ł(VA+V)XA - QX ) (Equa. 2)
Wastage (MGD) 1 7_ A c' A e ! ^ '
t x~or y SA3
\. f e "
where, X = sludge concentration in recycle
X = sludge concentration in effluent
XA - MLSS
V . = volume of aeration tank, MMGD
V = volume of clarifier, MMG
c '
Q = feed volume, MMGD
The sludge age method of process control has many advantages which are discussed
in detail by Walker (5).
The accumulation of negatively charged colloids, inerts and oil in the activated
sludge mass is hypothesized to be a principal cause of poor sludge settling properties (SVI)
and dispersed solids, thereby limiting the effective SA or F/M operating range. The
sensitivity of the ASP to influent inert solids can be readily demonstrated. For example,
if undesirable materials such as heavy oils, catalyst fines, clay, coke fines, metallic
sulfides, etc., escape preceding treatment at the rate of 30 mg/1, become flocculated by
activated sludge and accumulate, their effect can be predicted. Assuming for simplicity
of calculation that the inerts can escape only via sludge wastage and not via the clarifier
overflow, the equilibrium concentration levels in the waste and recycle streams and
aeration tank can be estimated using equations (1) and (2).
The quantity of inert suspended materials flocculated by the activated sludge mass
at equilibrium is shown in Table 1 for two cases; one in which the cell yield is 0.4 Ibs/lb
COD removed and the second in which the yield is 0.6. For each case, equilibrium
conditions are calculated for sludge ages from 5 to 45 days. In columns C through J
equilibrium data are shown for only soluble substrate (COD), and in columns K through
S the impact of the inert suspended matter accumulating in the system for a rate equivalent
to 30 mg/1 in the feed.
Not expressly illustrated by the data in Table 1 is the desirable change in biological
population dynamics when operating at the higher SA. At low SA, bacteria predominate
in the activated sludge mass. As SA increases, organisms which are comparatively slow
reproducers accumulate to a greater extent. One effect of operation at high SA is to
allow accumulation in the activated sludge mass of nitrifying and other organisms to
degrade the more "bio-resistant" contaminants present. A second effect is the accumu-
lation of protozoa in the activated sludge mass. In an ASP operating at high SA, the
role of protozoa is chiefly effluent clarification achieved by predation on the dispersed
bacteria; i.e., those tending to carry out in, and thereby contaminate, the clarifier
effluent (6). Bacteria within the floe are relatively immune to predation.
-------
134
The soluble BOD/COD/TOC remaining in the effluent from an ASP relates importantly
to the SA in the treatment of refinery effluents. As described earlier, the COD/BOD ratio
for refinery effluents may be quite high. As shown in Figure 7, the BOD is usually removed
quite readily by the ASP with little attention to process details. However, operating at
comparatively low SA's usually results in an effluent containing substantial remaining con-
tamination measured as COD. Operating the ASP at increasingly higher SA results in
increasing COD reductions, as shown quantitatively in Figure 7. The TOC removal curve is
similar to the COD curve. Dickenson and Giboney (7) in a study of refinery effluent treat-
ment observed that, with sufficiently long periods of acclimation, the organisms in an ASP
can acquire the capacity to metabolize components normally biorefractory. A specific
example of such a material was found to be tertiary butyl alcohol, which took a 5-month
acclimation period before degradation was obtained. Many other such observations are
reported in the literature and have been experienced by the authors. Thus, operating at
high SA will continue to reduce significantly the COD, whereas the observed effect on the
BOD may be slight. Additionally, we have achieved total organic carbon (TOC) levels
below 10 mg/1, superior to that achievable with activated carbon.
BIOLOGICAL REASONS FOR LIMITING THE OBJECTIVE OF BIOLOGICAL TREATMENT
TO REMOVAL OF SOLUBLE CONTAMINANTS
Only soluble organics can be absorbed through the cell membrane of the microorganism.
Above certain minimum "saturation"concentrations (typically perhaps 5-10 mg/1) of degrad-
able organics, the rate of removal is independent of degradable organics concentration, but
directly proportional to the amount of microorganisms present. Thus, the rate of biodegrad-
able organics removal is determined by the rate of metabolism of the microorganisms down to
certain minimum concentrations. Below these minimum "saturation" concentrations, the
rate of biodegradable organics removal is determined by their rate of passage through the
cell membrane. As the concentration gradient across the cell membrane decreases, the
rate of mass transfer decreases, and further biodegradable organics removal under these
conditions involves increasing the reaction time.
Large particles of biodegradable organics are removed by two possible mechanisms:
(1) the microorganisms can evolve a mechanism for engulfing or directly passing large mole-
cules into cells where confined enzymes initiate oxidative biochemical reactions, or (2)
they excrete the necessary enzymes causing the reaction sequence to be initiated outside the
cell membrane. These exoenzymes are produced and controlled by mechanisms not under-
stood, and in the activated sludge process it is not possible to predict the influence of
process variables on the kinetics of colloid and large particle biodegradation . This latter
mechanism involves bacteria, and is believed to be the principle mechanism operative in the
activated sludge process.
STRUCTURE OF THE BACTERIAL CELL
Figures 8 and 9 are schematics of a bacterial cell and bacteria cell walls and mem-
brane which are useful in picturing the extreme complexity of biological oxidation. The
food material must pass through the cell wall and the cell membrane. The outermost surface
of most bacteria is a slimy capsule varying in thickness up to 100,000+ A . The composition
of capsules varies with species and may consist of polymers of glucose or other sugars, amino
-------
135
acids and sugars/Or polypeptides. Capsules generally consist of about 98% water, and
the capsular layer probably serves the microorganism as an osmotic barrier; i.e., a
mechanism for guarding against too rapid an influx or efflux of water (8).
The soft outer layer of the cell wall is 100-150 A° thick and consists of an outer
layer of proteins and polysaccharides and inner layers of lipids and possible lipopoly-
saccharides and proteins. The acidic polysaccharides are partially responsible for the
negative electric charge on the bacteria surface.
The rigid layer is about 50 A thick and is comprised mainly of heteropolymers of
peptidoglycans arranged in a coarse three-dimensional foraminous mesh which provides
rigidity and strength to the cell wall.
These outer layers offer no obstacle to inward passage of molecular oxygen, minerals,
amino acids, glucose, and large organic molecules. Cell wastes can pass outward unob-
structed (9).
The periplasmic space is about 100 A thick and, in some bacteria, contains degrad-
ative enzymes which break down the larger molecules passing through the cell wall into
simple sugars, amino acids and other low molecular weight fragments. These lower mole-
cular weight fragments are transportable across the cell membrane by transport systems.
Membranes consist almost entirely of proteins and lipids (14). The proteins serve as
enzymes and provide the distinctive functional properties. The lipids provide the gross
structural properties and are a permeability barrier.
FUNCTIONS OF THE CELL MEMBRANE
The cell membrane is about 100 A thick and serves three key functions: It acts as
an osmotic barrier and is impermeable to ionized substances and nonionized substances with
molecules larger than glycerol; it provides a housing for active transport systems, and
provides a site for key enzymic reactions involved in energy metabolism. Cell membranes,
therefore, are highly selective in regard to the substances that may pass through them,
inward or outward (10).
Substrates may be transported through cell membranes counter to concentration
gradients. This process is called active transport because it requires added energy. Active
transport is accomplished by enzyme transport systems sometimes called permeases. These
systems are so efficient for concentrating metabolites that the osmotic pressure in the cell
can reach 300 psi (14). Like other enzymes, permeases are specific for the substrate
involved and are inducible.
REACTIONS WITHIN THE CELL
With reference to Figure 8, transport of the substrate through the cell wall and
membrane is only a small part of the over-all biological system. Once inside the cell,
the substrates are acted upon by systems of enzymes called endoenzymes. There are
several thousand different enzymes within the cell of a single bacterium (15). Endo-
-------
136
enzymes undertake two general types of processes inside the cell; synthesis of cell compo-
nents in food reserves, and the release of energy from food stuffs. The energy is either
stored in special molecules like adenosine triphosphate (ATP) inside the cell or is immediate-
ly used for any of the active processes of the cell. Recently it has been discovered that
bacteria can energize the active transport of many substances by the oxidation of various
chemical substances without the transfer of phosphate groups (17).
Endoenzymes do not act individually but as parts of coordinated and sequentially
operating systems. Whatever effects one portion of the intracellular enzyme system has
some effect on all parts. The activity of an enzyme is inhibited by accumulation of the
end-products of the enzyme catalyzed reaction. In a sequentially operating enzyme system
excessive accumulation of a reaction product may inhibit the reaction not only of the
enzyme manufacturing the reaction product, but all prior enzymes in that sequence. This
is an important form of automatic control called feedback inhibition. In the presence of
excessive amounts of end-products, not only is enzyme activity inhibited but the actual
synthesis of the enzymes themselves may be repressed. If a cell normally synthesizing a
certain substance is artifically supplied with that substance from an extraneous source,not
only is activity of the enzyme inhibited, but synthesis of some or all of the enzymes in the
production sequence for that substance is repressed until the enzymes are needed again.
This is called feedback repression. Differentiation is made between inhibition of the action
of the enzymes by their end-products (feedback inhibition) and repression of the synthesis
of the enzymes themselves by the accumulation of end-products (feedback repression) in the
enzyme sequence (Figure 10). In a sequentially operating system the end-product of each
enzyme can be the inducer of the next enzyme in the series and the inhibitor or represser of
the preceding enzyme, thus carrying forward the work of the enzyme factory (18).
In the cell, energy is derived from the enzymic bio-oxidation of foods, and the energy
liberated is transmitted by high energy compounds, notably certain organic phosphates; or
directly as recently discovered. The exact manner in which energy is transferred from low
energy level oxidized foods into high energy foods is not yet fully understood, but it has
been suggested that the participating enzymes are perhaps hundreds in number and must have
a specific orientation to each other in the cell (18).
REACTIONS EXTERNAL TO THE CELL
Suspended and colloidal matter require solubilization by exoenzymes preparatory to
transport into the cell. Many enzymes probably exist free in colloidal suspension in the
fluid matrix of the cell. Some of these are secreted to the exterior of a cell. They are
called exoenzymes. Exoenzymes are mainly digestive in function. By hydrolysis they
decompose complex organic matter, such as proteins, cellulose, and fats, to simple soluble
molecules of amino acids, glucose, glycerol and fatty acids. These relatively small mole-
cules can be transported through the cell membranes of many microorganisms, there to be
utilized as food. Among the lower fungi, algae, and bacteria, digestive enzymes are to
a large extent excreted aimlessly into the surrounding fluid where they may or may not come
into contact with food. They may be wholly dissipated by dilution, convection currents,
or other factors (12).
-------
137
STRUCTURE OF ENZYMES
Enzymes are large complex molecules v/ith molecular weights ranging from about
10,000 to more than 1,000,000 and consisting of two components; a colloidal protein part
called apoenzyme, and a non-protein, noncolloidal, low molecular weight, thermostable
part called coenzyme (organic) or cofactor (inorganic). The two components constitute
the complete active enzyme called holenzyme but commonly referred to by the term
enzyme. Coenzyme molecules generally carry the distinctive reactive portion of an
enzyme. Acting with the apoenzyme, the coenzyme brings about the specific substrate
reaction that is characteristic of that particular enzyme. In many coenzymes the
distinctive reactive portion is a familar vitamin; nicotinic acid or its popular derivative
niacin; otheisare vitamins of the B complex such as thiamine which is vitamin B. or
riboflavin vitamin B^. In fact, all vitamins that function physiologically have been
found to act as a reactive group of one or another coenzyme (13, 19).
ENZYME MECHANISMS
In most enzyme reactions, the specific apoenzyme involved appears first to attach to
the substrate at certain specific sites involving only 10-20 amino acid residues confined
to an area of surfaces perhaps 15-20 A in diameter (19). These sites represent recipro-
cally corresponding physical structure and anionic and cationic groups in the molecule
of the substrate. This preliminary combination appears to place certain bonds in the
substrate under stress. The coenzyme, because of its appropriate molecular structure,
then combines with a part of the substrate. For example, a halogen ion or amino group,
etc. This ion or group is then either passed by the coenzyme to another coenzyme or to
a different substrate molecule, or may be liberated as waste into the surrounding fluid.
The final result depends on the nature of the reaction being catalyzed. The enzyme,
freed of the altered substrate residue, is then ready to combine with another substrate
molecule and repeat the process. If any energy is released by the reaction, it is partly
lost as heat and in part taken up into the cell substance by the formation of energy-
rich compounds that contain certain types of bonds as organic phosphate bonds, thioester
bonds, and some others in certain coenzymes. The energy stored in such energy-rich
compounds is later used by the cell in motility and cell synthesis.
ENZYME STEREOCHEMISTRY
All known enzymes are globular proteins whose active site is either a crevice,
shallow depression or a pit and, stereochemically, conform to the principle of the
"hydrophobic in, hydrophilic out" protein model. All charged, and most uncharged,
polar groups are on the surface and the great majority of the nonpolar hydrophobic groups
are buried in the interior. This has a profound chemical purpose in that the reactant
molecules can be taken out of the water where water molecules keep charged molecules
apart, and into the enzyme which is made up of hydrocarbons with a low dielectric
constant. In the organic solvent of the enzyme, strong electrical forces can be brought
to bear on the reactant causing them to react rapidly (15).
Being colloidal proteins, enzymes are macromolecules that may consist of hundreds
of the 22-24 amino acid residues linearly linked in different arrangements. Trypsin,
-------
138
chymotrypsin and carboxypeptidases, for example, contain 200-300 amino acid residues per
molecule, or between 3,000 and 5,000 atoms. The active site involves perhaps less than
20 of the acid residues and the three dimensional structure of the entire molecule is necessary
for maintaining the stereochemical structure of the active site. Changes in the three
dimensional structure of the protein are typically accompanied by a loss of enzymic activity.
Since each enzyme is different from all others with respect to molecular configuration,
each is unique or specific. Each enzyme can react only with certain particular substrates
that have a corresponding stereochemical structure. Specificity of enzymes varies greatly
in degree (19).
Lysozyme, the first enzyme to have its detailed molecular structure worked out, ribo-
nuclease, and carboxypeptidase A are examples of "small enzymes" of molecular weights - -
14,600, 13,700, and 34,600 respectively. These enzymes are ovid shaped and carbo-
xypeptidase A has dimensions of about 40 X 44 X 52 Angstroms. Most enzymes are far larger
than these (16).
GENETIC INFLUENCE ON ENZYME PRODUCTION
Each species of living cells has a genetically determined natural endowment of certain
functioning enzymes. These are characteristic of and constantly present in all of the cells
in that species of cell. Such inherited enzymes, part of the normal constitution of the cell,
are called constitutive enzymes. In addition, many cells possess genetic determinants or
genes for the synthesis of such enzymes which are genetically repressed. These genetically
repressed enzymes are manufactured only when an inducing substrate or related substance
enters the cell. In the presence of the inducer the represser of the appropriate synthesizing
mechanism is removed and the enzyme specific for the inducer substrate is synthesized.
Such enzymes are said to be inducible. Therefore, although genes determine the full
enzymic potentialities of a cell, environmental factors in the form of inducers, determine
just which of the latent enzymic potentialities of a cell shall appear under any given
circumstances. The cell is evidently not under the necessity of synthesizing all its potential
enzymes but only those needed (11).
ENZYME MEDIATED REACTION EQUILIBRIA
Enzymic reactions are reversible. The accumulation of products affects the action
of any enzyme in either direction, as would be expected in chemical equilibria. Under
any set of constant conditions, the equilibrium point for an enzyme catalyzed reaction is
constant. There is a constant relationship between concentration of enzyme and concentra-
tion of substrate. Up to the point of "saturation", the rate of reaction increases with
increase of ratio of one component to the other. With a constant amount of enzyme,
increase of substrate increases rate of reaction until every molecule of enzyme is fully
saturated with substrate. Further additions of substrate cannot increase the rate of reaction.
Conversely with a fixed amount of substrate, rate of reaction increases with the additions
of enzyme until all molecules of substrate are in contact with enzyme. Further additions
of enzyme do not effect the rate of reaction. In many instances enzyme catalyzed reactions
appear to proceed in only one direction because the equilibrium is never reached. Under
normal conditions in the living cell, enzyme reactions are constantly pushed in this manner
toward the one or the other side of the reactions. Theoretically reversible reactions cannot
-------
139
actually reverse when large differences in energy levels are involved, since resynthesis
cannot be brought about by the same enzymes because they cannot restore the lost energy.
To reverse the reaction requires that work be done by other systems of enzymes that capture
new energy from other sources (11, 20).
ENZYMES SENSITIVITIES AND INHIBITIONS
Enzymes are sensitive to all of the various precipitating and coagulating factors that
effect proteins in general; temperatures above 80 C, excessive concentrations of heavy
metals or of hydrogen or hydroxyl ions, active chemicals like chlorine, and strong alkalis
quickly destroy all types of proteins. In addition to being sensitive to deviations in
temperature, enzymes are very sensitive to variations in pH, osmotic pressures (salinity),
and UV and other radiations. Whatever affects enzymes, of course, affects microorganisms.
The functioning of some enzymes may be inhibited by certain non-metabolizable
substances whose molecular structure is analogous to that of a true, metabolizable substrate.
Such metabolites like inhibitory agents are called metabolite antagonists and because of
their molecular structure, can pre-empt the specific combining site on a particular enzyme
protein to the exclusion of the true metabolite. The enzyme or coenzyme is not necessar-
ily destroyed but it can no longer function.
ENZYME CHEMISTRY
A key to recognizing the important process design responses which must be made in
an integrated end-of-pipe treatment sequence in which the activated sludge process is
an element, is recognition of the complex enzyme systems operative and their requirements.
Exoenzymes excretedaimlessly into the surrounding water by bacteria for hydrolysis of
colloidal and suspended solids into smaller soluble molecules so they can pass through the
cell membranes to be utilized as food, not only may be dispersed and lost in the effluent,
but are exposed to and unprotected from environmental changes in pH, salinity, and other
conditions impacting on the stereochemical conformation or properties of proteins, causing
deterioration in their functioning. Endoenzymes on the other hand, are at least partially
protected from these environmental changes by the bacterial cell wall and membrane.
Thus, using the biological process for removal of only soluble contaminants, as we recom-
mend, is consistent with obtaining a stable system less influenced by transient changes in
environmental conditions. An additional important consideration contributing to capacity
and process stability is that activated sludge units utilizing only soluble contaminants can
be operated at very high sludge ages. This permits much higher biological cell inven-
tories, which in turn increases stability, capacity, and purification while minimizing
generation of excess biological cells.
SUMMARY
1 . A systems optimization of a refinery end-of-pipe treatment sequence points to
the activated sludge process (ASP) as the key element.
2. To achieve maximum water quality, system optimization points to reversing
the historic work horse role of the ASP; i. e., it should be used only for
-------
140
the removal of essentially soluble contaminants.
3. Reversing the role of the ASP yields a dramatic series of beneficial effects:
a. The SVI characteristics of the activated sludge mass are excellent at
very high sludge ages; i.e., over 40 days.
b. Process control is greatly simplified by operating at a very high sludge
age; i.e., greater than 40 days.
c. Using very high sludge age for process controls eliminates the need for
many process control tests.
d. Operating at very high sludge age produces an exemplary effluent
very low in TOC and other contaminants.
e. The cell yield coefficient is remarkably low at very high sludge age.
f. At very high sludge age the population dynamics of the sludge mass
improve.
g. Maximum ASP capacity for purification is achieved by operating at
high sludge ages.
4. Systems optimization wherein the key element is using the ASP for removal
of only soluble contaminants permits clear definition of the roles the other
elements play; i.e., colloidal and suspended matter must be essentially
all removed in pretreatment sections.
5. The technology is available to handle the new requirements made by the
process operations in their changed roles.
6. Current research developments on cell membranes and enzyme systems
support strongly this new role for the ASP.
7. Cell genetics, wherein inducible enzymes or mutant species are required,
support operating the ASP at very high sludge age for maximum purification
capacity.
8. Bacteria are essentially enzyme factories. Enzymes are sensitive to temperature,
pH, excessive concentrations of heavy metals, oxidizing agents, salinity, UV,
and other radiations. Using the ASP for removal of only soluble contaminants
is consistent with obtaining a stable system less influenced by transient changes
in environmental conditions since the endoenzyme systems are at least partially
protected from these environmental changes by the bacterial cell wall and
membrane.
9. This recommended systems optimization saves significant energy and other
operating and capital costs of the end-of-pipe sequence.
10. This recommended systems optimization minimizes the generation of solid
wastes, and the solid wastes generated are amenable to disposition.
-------
141
REFERENCES
(1) Eckenfelder, W. W., Manual of Treatment Processes, V. 1, Chap. 4, p. 7,
Environmental Science Services Corp.
(2) Downing, A. L.7 "Factors to be Considered in the Design of Activated Sludge Plants,"
Advances in Water Quality Improvement, Gloyna & Eckenfelder Editors, Univ. of
Texas Press, p. 190.
(3) Wuhrmann, K., "Research Developments in Regard to Concept and Base Values of the
Activated Sludge System," Advances in Water Quality Improvement, Gloyna &
Eckenfelder Editors, Univ. of Texas Press, p. 143.
(4) Forster, C. F., "The Surface of Activated Sludge Particles in Relation to their
Settling Characteristics," Water Research, V. 2, (1968) p. 767-776.
(5) Walker, L. F., "Hydraulically Controlling Solids Retention Time in the Activated
Sludge Process," Jour. WPCF, V. 43, No. 1, (1971) pgs. 30-39.
(6) Jones, G. L., "Role of Protozoa in Waste Purification Systems," Nature, V. 243,
pgs. 546-547.
(7) Dickenson, R. L., Giboney, J. T., "Stabilization of Refinery Waste Waters with the
Activated Sludge Process: Determination of Design Parameters," Proceedings of the
Industrial Waste Conference, Purdue University, 19, pgs. 294-303.
(8) Frobisher, et al, Fundamentals of Microbiology, W. B. Sounders Company, Philadel-
phia (1974), p. 198.
(9) Ibid, pp. 201,204.
(10) Ibid, p. 204.
(11) Ibid, p. 92.
(12) Ibid, p. 116.
(13) Ibid, p. 89.
(14) Fox, C. F., "The Structure of Cell Membranes, " Scientific American (February, 1972).
(15) Dickerson, R.E., Geis, L, " The Structure and Action of Proteins," Harper and
Row, New York (1969), p. vii.
(16) Ibid, pp. 88,96,97.
(17) Luria, S.E., "Colicins and the Energetics of Cell Membranes, " Scientific American,
(December, 1975).
(18) Lehninger, A. L., "Energy Transformation in the Cell, " Scientific American,
(May, 1960).
(19) Finean, J. B., Engstrom-Finean Biological Untrastructure, Academic Press, New
York and London (1967).
(20) Aiba, S., Humphrey, A. E., Millis, N. F., Biochemical Engineering, Academic
Press, New York, (1965), pp. 38,39.
-------
142
DISCUSSION
Milton Beychok: Jim, what you have really said is that if you filter well ahead of the
activated sludge unit, you can cycle up in sludge age and the process behaves much better
in terms of capacity, effluent quality, and operating simplicity because it's only working
on soluble BOD.
Jim Grutsch: Yes.
Milton Beychok: What do you do in your DAF (dissolved air flotation) or filter unit that is
different than other people so you can achieve this?
Jim Grutsch: We address the water chemistry. Essentially everything in nature has a
negative electrical charge on the surface. Colloidal particles exist because they are
repelling each other and resist agglomeration which would facilitate phase separation. In
a filter, for example, the surface of a granular media filter is also negatively charged. If
we just attempt to filter out negatively charged solids with a granular media filter, a large
fraction of the negatively charged colloids will be repelled by the media and rapidly break
through to the effluent. Similarly with a DAP unit, the flotation air has a negative electri-
cal charge. This is the most important reason why Wes Eckenfelder's DAF unit wouldn't do
a very good job clarifying activated sludge from mixed liquor; the negatively charged fine
particles in activated sludge repel the negatively charged flotation bubble resulting in
poor phase separation efficiency.
The key to making DAF units or filters work effectively, therefore, is to reduce the
electrical charge on the colloidal particles so that they can agglomerate and be trapped in
granular filter media, or separated by attached air bubbles. There are some excellent
chemicals available for this purpose at reasonable cost. Determination of the most cost-
effective chemical treatment is easily done by comparing zeta potential titration curves.
The use of zeta potential titration curves is, unfortunately, in its infancy but this tool will
see greatly increased use \A the coming years.
Anonymous: Do you encounter any problems from oil and grease?
Jim Grutsch: The electrical charge on oil particles is also negative and the previous
discussion on colloids also applies to oil and grease particles. Granular media filters
should be protected from gross slugs of oil, and the equalization basin does this. The
higher the ratio of inert solids-to-oil in the feed to the DAF of filter unit, the lower the
oil in the effluent.
Robert L. Wortman: Would you care to comment on the effect of polyelectrolytes on keeping
suspended solids down in the final effluent when operating with a very high degree of nitri-
fication ?
Jim Grutsch: This question is central to the main thrust of our paper and I would like to
develop the response in some detail so that the importance of optimizing the end-of-pipe
sequence as described is not overlooked.
-------
143
1 . We have no experience using pal/electrolytes in the activated sludge unit
clarifier at high sludge ages because the need doesn't exist when proper prefil-
tration is used. The average amount of total suspended solids in the clarifier
overflow is less than 10 mg/l , and less than 5 mg/1 from a final stage of granular
media filtration (also not using chemicals). This is due principally to the excellent
SVI maintained of about 80, with no "arms and legs" in the supernatent. A wide-
well clarifier is used, but it is typically fully loaded; i.e., overflow rates greater
than 70 gpd/s.f. and solids flux rates greater than 25 Ibs/D/s.f. Actually, the
design capacity of the clarifier in this service has not been determined.
2. "A very high degree of nitrification" specified in Mr. Wortman's question equates
to operation at a high sludge age (or low F/M). With reference to Figure 5 and
Wes Eckenfelder's data, under these operating conditions the SVI of the sludge
increases which presumably gave rise to the question. A reason for the deteri-
oration of the SVI is shown in Figure 6; i.e., the electrical properties of the
sludge are deteriorating.
Electrical properties (ZP) of the sludge deteriorate because of increasing accu-
mulation of electrically charged inerts in the activated sludge mass at increasing
sludge age. This can be visualized better by plotting some of the data in Table 1
as I have done in Figures 11 and 12. Data from Part A of Table 1 is plotted in
Figure 11. The curve for inert solids shows the effect of capturing only 30 mg/l
by the sludge mass. Thus, for this comparatively weak waste, by comparing the
ratio of inert solids to biological solids it is observed that at 20 days sludge age
half the sludge is inerts,and at sludge ages greater than the biological fraction.
This inert fraction has a negative zeta potential, and at some ratio of inerts to
biological solids the capacity of the biopolymers to destabilize the solids (i.e.,
keep the zeta potential of the sludge mass low enough so that fine particles are
not repelled and thereby dispersed) is surpassed and the SVI deteriorates (Figure
6) and "arms and legs" appear in the supernatent. Without the inerts in the
activated sludge mass, excellent SVI's are achieved at very high sludge ages.
Our objective, therefore, is to remove completely suspended matter in the
influent to the activated sludge unit.
Figure 12 shows the mixed liquor solids characteristics for the conditions of
Part B, Table 1.
-------
144
BIOGRAPHY
James F. Grutsch is Coordinator-Environmental
Projects, Standard Oil Company (Indiana). He holds under-
graduate and graduate degrees in chemistry from Indiana
University. Prior to his present assignment with Standard,
Jim served successively as Group Leader for finishing,
blending and reclamation at the Amoco Oil Whiting
refinery, and Coordinator of Waste Disposal for Amoco.
Jim taught undergraduate chemistry for six years at
Indiana.
BIOGRAPHY
Russell C. Mallatt is Manager-Environmental and
Energy Conservation, Standard Oil Company (Indiana).
He holds degrees in chemistry from the University of
Illinois and the University of Rochester. Prior to his
present assignment with Standard, Russ served successively
as Chief Chemist, Chief Process Engineering and
Technical Services Superintendent of the Amoco Oil
refinery at Whiting, Indiana.
-------
145
TABLE 1 "MATERIAL BALANCES AROUND ACTIVATED SLUDGE PLANT FOR VARIOUS OPERATING MODES'
Basis: Flow = 20 MM GPD; Total Aeration and Clarifier Tank Volumes = 10 MMG; Thickened
Sludge From Clarifier = 8000 mg/1; No Solids Loss in Clarifier Overflow; COD = 150 mg/1
A. Basis: a = .4; b = .05; COD Removal = 25000 Ibs/D (150 mg/1)
Sludge Age Equilibrium MLVSS Wastage Rates
Sludge Yield
Case
1
2
3
4
5
6
(days)
5
10
15
25
35
45
mg/1
480
800
1029
1333
1527
1662
Total Ibs
40
66
85
111
127
138
/
/
/
/
/
r
032
720
783
200
375
572
F/M
.625
.375
.292
.225
.196
.181
Ibs/D
8006
6672
5719
4448
3639
3079
gpd @ .
120
100
85
66
54
46
8% Solids
,000
,000
,716
,666
,541
,148
Ibs/lb COD
.32
.267
.229
.178
.145
.123
MGD Clarifier Clarifier Flux,
.8% Solids I bs. Sol ids/Day
1.2
2
2.57
3.33
3.82
4.16
84,868
146,784
193,693
259,365
303,352
334,884
30 mg/1 Inert Suspended Matter in the Incoming Water Hove the Following Effects
mg/1
Inerts in
Wastage, mg/1
5000
6000
7000
9000
11000
13000
MLSS
Inerts
300
600
900
1500
2100
2700
MLSS
Total
780
1400
1929
2833
3627
4239
Inerts/MLVSS
ratio
.625
.750
.875
1.125
1.375
1.625
MGD
Draw (c
1.
3.
4.
7.
9.
10.
Clarifier
i> .8% Solids
95
5
82
08
07
6
Equilibrium
MLSS, Ibs.
65,052
116,760
160,879
236,272
302,492
353,533
Clarifier Flux.
Ibs Solids/Day
142,789
274,386
399,300
639,825
879,344
1,081,810
Wastage Rates
Ibs/D gpd
13,
11,
10,
9,
8,
7,
roio
,676
,725
,451
,643
,856
@ .8% Sol ids
195,000
175,000
160,750
141,650
129,541
117,750
B. Basis: a = .6; rest as above in A
Sludge Age Equilibrium MLVSS
Wastage Rates
Case
7
8
9
10
11
12
(days)
5
10
15
25
35
45
mg/1
720
1200
1543
2000
2291
2492
Total Ibs
60,
100,
128,
166,
191,
207,
048
080
674
800
062
859
F/M
.417
.25
.195
.15
.131
.12
Ibs/D
12010
10008
8578
6672
5459
4619
gpd @ ,
180
150
128
100
81
69
.8% Solids
,000
,000
,567
,000
,820
,230
Ibs/lb COD
.48
.4
.34
.267
.218
.185
Sludge Yield MGD Clarifier Clarifier Flux,
Draw @ .8% Solids Ibs. Solids/Day
1.8
3
3.86
5
5.73
6,23
130,905
230,184
307,045
417,000
491,622
545,145
30 mg/1 Inert Suspended Matter in the Incoming Water Have the Following Effects
Inerts in
Wastage, mg/1
3333
4000
4667
6000
7333
8667
MLSS
Inerts
300
600
900
1500
2100
2700
MLSS
Total
1020
1800
2443
3500
4391
5192
(nerts/MLVSS
ratio
.417
.5
.583
.750
.9167
1.083
MGD Clarifier
Draw® .8% Solids
2.55
4.5
6.11
8.75
11
13
Equilibrium
MLSS, Ibs.
85,068
150,120
203,746
291,900
366,209
433,013
Clarifier Flux.
Ibs Solids/Day
191,828
367,794
531,981
839,213
1,135,249
1,428,942
Wastage Rates
Ibs/D gpd@ .8% Sol ids
17,014 255,000
15,012 225,000
13,583 203,583
11,676 175,000
10,463 156,821
9,623 144,222
-------
146
Optional Refinery Treatment Sequence
FIGURE
Treatment
Pre-or Inplant
Treatment
Primary
Treatment
Objectives Phenolics,S=,NH3, Free Oil and
RSH, F~ Acid Sludge, Suspended Solids
Oil Etc.,Removal a Removal
Water Reuse or
Waste Equalization
Processes Unit Separators
Steam Stripping
Fuel Gas Stripping
Air Oxidation
Neutralization
Surge Ponds
API Separators
CPI.PPI Separators
Sludges
Intermediate
Treatment
Emulsified Oil,
Suspended and
Colloidal Solids
Removal
Chem. Coagulation
SAir Flotation
Chem. Coagulation
S Filtration
pH Control
Immediate Oxygen
Demand Reduction
Equalization of
Wastes
1
Sludges
F
Tertiary
Treatment
Variable
Objectives
Chem.Coagulation
Q Air Flotation
Chem. Coagulation
8 Filtrotion
Activated
Carbon
Sludges
Sludges
FIGURE 2. Probability Plots of Oil a Grease
Data Before and After Inter-
mediate Treatment
o
•s.
a>
(A
o
a>
w
O
03
1000
800
600
400
200
100
80
60
40
20
10
8
6
_ Raw Waste Load from
API Separator
O
o
Waste Load after
Intermediate
Treatment
0.5 I 2 5 10 20 40 60 809095989999.5
%Probability less than Indicated Value
-------
147
FIGURE 3
Probability Plots of BOD Data Before
and After Intermediate Treatment
400
2 300
M
TJ
O
a 200
x
O
"o
O
'i
a*
o
o
'o
100
80
60
50
ORaw Waste Load from API Separator
•Waste Load after Intermediate Treatment
I 1
0.5
1
10 20 30 40 50 60 70 80 90
% Probability less than Indicated Value
95
98 99 99.5
FIGURE 4
Probability Plots of COD Data Before
and After Intermediate Treatment
1500
1000
800
600
400
o
E
a
o>
>>
x
O
"o
u
1
0)
200
100
0.5 I
ORaw Waste Load API Separator
• Waste Load after Intermediate Treatment
O
I I
I I
10 20 30 40 50 60 70 80 90
% Probability less than Indicated Value
95 98 99 99.5
-------
148
ioo h
90
80
70
FIGURE 5
SVI Of Activated Sludge In Unit Treating Refinery
Effluent Pretreated by Filtration
BOD
Removed
( from Eckenfelder)
SVI (from
Eckenfelder)
_L
I
SVI Found with
Prefiltration
li ii
.15 .2 .3 .4 .5 .6 .8 1.0
Loading,Lbs.BOD/D/LB MLVSS
I l I I | I I
35 25 15 10 5 3 2
Sludge Age,Days (a = .6 Ł = .05)
300
200 o>
E
_3
O
100 I
50 «
1.5
FIGURE 6
Relationship of Activated Sludge
SVI with Zeta Potential In-Situ(Forster)
-2lr-
-20
>
E
_o
c
o
a.
a>
N
-19
-18
I
I
I
I
I
I
60 80 100 120 140
SLUDGE VOLUME INDEX
160 180
-------
149
400-
FIGURE
Soluble BOD and COD in ASP
Effluent at Various Sludge Ages
20 30 40
Sludge Age, Days
FIGURE 8
Schematic of a Bacterial Cell, Its Biochemical Activities,
and Exoenzyme Solubilization of Insoluble Substrates
Soluble Substrates
Soluble Substrates
from Actions of
Expenzymes
r^^T
~
^ Permease Enzyme
f'f, ^ Transport System
Cell
Membrane
\Wcell Wall
CO^ and Other
Metabolic By-Products
-------
150
FIGURE 9
Schematic of Layers Observed in Bacteria Cell Walls and
Cytoplasmic Membrane (Approximately to Scale)
• ••••••••*-••**" »•••••••
'«••»• • • • • • •_•• • •«. • • • • • • • .•
',. • • • • • • *. r
*. • • ••••••*. •••.•!«••••«•»•
•«•••••••. i • . • • i • ?. •••••»•!•
• » • «.•» •«. i•.••••«•.•••••"•••
::•:•:•: }c
>d
• Outarmost Layer, Protaim and Polysacch«rid«s
b Lipids
c Possibly Lipopolyuccharidas
d PouibJy Protains
a Fibrous H«taropolym«rs of Paptidoglyeans
f Pariplasmic Spaca Containing Dagradaiion Enzymas
g Outar Layar of Call Mambrana
h Middla Mambrana Layar, Bimolacular Phospholipid Layar
i Innar Layar of Call Mambran* (--25 A°)
A C*puil< Matirul; 200-100,000+ A°
B Soft L«y«r. 10O-1SO A»
C Rigid Liy*r; -SO A°
D Pcriplumic Sp*c«; ~100 A°
E Cytoplwnic Mwnbnrw: -100 A°
Pr Proteins
P, Polyucchwido
En EniynM Trantport Synwm (PtnnnMil
FIGURE 10
Enzymes Work in Sequentially Cooperating Systems
Enzyme Synthesizing
System
Feedback Repression
Feedback Inhibition
-------
151
Figure 11
Character of Mixed Liquor Solids
with Sludge Age
6000
4000
2000
0
Mixed L|quor Suspended Solids (MG/L)
Flow,20 MGD
Retention Time,12Hrs.
a = 0.4 - b = 0.05
COD Rmvd.,25000 Ibs/D
Inert Solids in,
so MG/L ..^^ /Inert Solids
_ . , 0 ...
Total Solids
CM ounuo v ^fff
'"
Biological Solids
6000
4000
2000
0 5 10 15 20 25 30 35 40 45
Sludge Age (days)
Figure 12.
Character of Mixed Liquor Solids
with Sludge Age
6000
4000
2000
0
Mixed Liquor Suspended Solids (MG/L)
Flow,20 MGD ....•
Retention Time,12Hrs. ..•••""
a = 0.6 - b = 0.05 ...••'"
COD Rmvd.,25000 Ibs/D ...<,LTota| QOIJHS
Inert Solids in, ..,•" l1.^1^
so MG/L .••' Bioloical Solids
Inert Solids
6000
4000
2000
0 5 10 15 20 25 30 35 40 45
Sludge Age (Days)
-------
CASE HISTORY ON BIOLOGICAL TREATMENT
OF A
PETROLEUM REFINERY WASTEWATER
Charles E. Ganze, Facilities Supervisor
and
Joe P. Teller, Deputy General Manager
Gulf Coast Waste Disposal Authority
Houston, Texas
Gulf Coast Waste Disposal Authority's American Facility located in Texas City, Texas,
provides intermediate, secondary and tertiary treatment for a 23 million gallon per day
waste stream.
The Research and Development and Process Design work was done by Amoco Oil and
its Parent Corporation, Standard Oil of Indiana. A number of process patents have been
applied for, with several granted and others pending. Williams Brothers Construction has
been licensed to construct similar facilities incorporating the patented processes.
The wastewater treated at the American Facility is generated primarily by Amoco
Oil Company's Texas City Refinery which currently processes approximately 350,000 barrels
of sweet crude per day. In addition to the wastewater generated in the crude refining
processes, sulfuric acid concentrating unit, cokers, and other refinery related process, this
treatment facility treats wastewater generated by Amoco Chemical Company's Texas City
Plant "B", other Amoco Chemical Company processing units located at the Refinery and all
storm water collected in the 500 plus acres inside the battery limits of both Amoco Oil and
Amoco Chemical Company's Plant "B".
To characterize the influent wastewater, parameters listed in the wastewater dis-
charge permits are used. The data presented represent the values actually observed over
7 months operation.
Influent Characteristics
Average Maximum Minimum
BOD, mg/1 183 1,110 36
TOC, mg/1 118 620 30
TSS, mg/1 234 1,908 32
O&G, rng/1 11 110 1
NH.-N, mg/1 39.4 100 9
Phenols, mg/1 6.7 85 1.2
Sulfides, mg/1 0.4 4.0 0.07
PH 11 2
The various waste streams have received primary treatment prior to reaching the
Gulf Coast Facility. In many cases, streams have passed through unit oil separators for
the removal of gross quantities of oil before being discharged to the refinery's collection
system, and then at the end of the collection system, all wastewater is processed through
153
-------
154
API Separators for removing the remaining oils and the bulk of the suspended solids. As is
generally the case with refinery effluents, the waste treatment system receives a waste
stream that varies greatly in both quantity and quality. The process design was based on a
knowledge that the treatment system would experience a highly variable waste load and
that more stringent effluent requirements could be expected in the future.
To describe the waste treatment system, it is broken into three steps: Intermediate,
Secondary, and Tertiary. The intermediate step incorporates the use of physical and chem-
ical processes which are designed to protect the secondary or biological step. It consists
of equalization with pH adjustment and prefiltration. As in many waste treatment facilities
treating industrial wastes, equalization is considered a necessity. It serves several functions:
first, it helps smooth out some of the variations in types of wastes and waste strength; second,
it helps protect the secondary facilities from sudden surges in flow; and third, it helps re-
duce the overall treatment chemical costs by allowing the various wastestreams to mix, and,
in many cases, neutralize each other, especially in the case of pH. The 4-hour equaliza-
tion time provided is considered minimal for adequate protection to the biological system.
Prefiltration in this system is performed by pressure-type multimedia filters. The
design hydraulic loading to the prefilter system is 5 GPM per square foot. Using only
organic polyelectrolytes as filter aids, the prefilters are designed to remove the suspended
solids and oils that remain after primary treatment and removal of much of the colloidal
materials. The logic in keeping an extremely high quality water as a feed to the biologi-
cal step is: first, to reduce the organic loading by removing the suspended solids, colloidal
materials and oils; and second, to reduce the amount of inorganic solids that are put into
the system. When the inorganic solids and oils are kept out of the biological system, the
settling characteristics of the biological sludge is improved; with a better settling sludge,
a higher sludge age can be attained; thus we then start attacking the more difficult to treat
organic compounds and ultimately obtain biological nitrification.
The backwash from the prefilters is clarified with the overflow recycled to the equal-
ization basin and the underflow (sludge) mixed with digested waste-activated sludge and
pumped to a land farm for disposal.
The product water from the prefilters passes into the second step of our treatment
system, the biological system. At this point, two aeration tanks, which operate in parallel,
provide 12 hours retention time at design hydraulic loading. This reasonably long retention
time improves the treatment capabilities for handling shock loads. Aeration is provided by
ten 150-horsepower, floating, low-speed surface aerators. From the aeration tanks, the
mixed liquor passes into the secondary clarifiers. The two 140-foot diameter, 18-foot
side-water-depth-wide feed well clarifiers operate in parallel. The hydraulic loading on
the secondary clarifiers is 600 gallons per square foot per day, which is generally considered
adequate. The sludge removal mechanism is the multiple pick-up, sludge sucker type.
The return sludge is pumped by centrifugal, recessed-impeller type pumps. Waste sludge
is pumped to a gravity thickener where the underflow passes into an aerobic digester, and
the thickener overflow is recycled to the equalization basin.
The overflow from the secondary clarifiers then flows to the third step of our treatment
-------
155
system, the tertiary. The tertiary step consists of polishing filters. These final or post
filters are designed simply to remove the biological solids that escape the secondary
clarifiers. These gravity type, dual media filters operate at a design hydraulic loading
of 4 GPM per square foot with the backwash pumped back to the aeration tanks as the
solids are of biological origin.
Operations began before all construction activities were complete; we have experi-
enced and are continuing to experience, the usual start-up mechanical and design problems.
Among the significant mechanical problems are electrical switch gear malfunctions, an
aerator gear box, backwash sump pumps, a clarifier scum pump motor, and the underdrain
nozzles on the final filters. Design problems that we have noted to date include: no
special provision for operating in sub-freezing weather (freezing sample lines, service
water lines, caustic lines, and wet-leads to instrument-primary devices); and a widely
fluctuating float in the Parshall Flume which is caused by excessive turbulence as a result
of too short an approach to the throat. Even without all equipment and facilities fully
operable, a number of interesting observations have been made. Over a three month
period, when mechanical problems were at a minimum, we observed what we consider
phenomenal reductions for a secondary waste treatment facility.
BOD 92%
TOC 85%
Ammonia-N 88%
These removals were experienced while maintaining the following conditions:
MLSS 3400 rng/1
Sludge Age Approximately 50 days
F/M 0.07 pounds BOD/pound MLSS
with no apparent deterioration in sludge settleability (Average SVI -85).
The average performance for the first seven months of operation was as follows:
REMOVALS
BOD Reduction 88%
TOC Reduction 77%
TSS Reduction 94%
Oil & Grease Reduction 91%
Ammonia-N Reduction 51%
Phenols Reduction 96%
Sulfides Reduction 72%
It should be emphasized that much of the data collected for these calculations do not
reflect a stable operation.
We feel that by completing the remainder of the construction, clearing up all the
initial mechanical problems generally associated with start-up, and by continuing to
improve on our own operational procedures, we can meet the conditions set forth in the
existing permits which extend through June 30, 1979. Beyond this, we foresee even more
-------
156
stringent requirements for BOD, TOC, and Ammonia-Nitrogen. Provided our treatment
capabilities improve as expected, this facility will produce an effluent that will comply
with anticipated future requirements.
DISCUSSION
J. J. Chavez : How much ammonia do you have in your raw influent and in the final efflu-
ent stream and where do you think you get the bulk of ammonia removal in your system?
Charles Ganze: The ammonia is removed, biologically, in the aeration tanks. The average
influent concentration (feed to the aeration system) was about 40 ppm with a maximum of
about 100 ppm or so.
J. J. Chavez: How about the final effluent?
Charles Ganze: The final effluent - it would go into the .05 ppm range, very difficult to
analyze.
Anonymous: Please discuss your choice of filters.
Charles Ganze: Economics is one thing. The pressure type filters are a little more expensive.
You can give them a little bit higher loading. The multimedia does give us some advantages
on handling the colloidal materials and the suspended solids. Now the gravity-type filter
again is used on the effluent, again the only thing that we anticipated getting from that were
biological solids.
G. T. Minnick: How do you dispose of your aerobically-digested sludge?
Charles Ganze: The sludge that we actually waste is a very minimal amount.
As Jim indicated we observe very low sludge yield, we are operating again at 50-day
sludge age at an elevated temperature and we have a low sludge yield. The aerobically
digested sludge is mixed again with this back wash from the prefilters and it is pumped to a
land farm area. Now we call this a land farm and not a land fill. It is actually tilled
into the soil. At this facility we have some 150 acres dedicated to this activity. We just
take a tractor with a disk out there and till it into the soil.
R. A. Farnham: What are the calculated variability factors associated with your process?
Charles Ganze: We don't think we have enough data to try to put together any thing like
that, I only showed you some maximums and minimums that we have observed over this
relatively short period of time, which is seven months, and we haven't tried to put together
any probability type numbers.
Anonymous: What is the loading on your secondary clarifier?
Charles Ganze: 600 gallons per day per square foot, surface loading.
-------
157
Leonard Orwin: At the beginning of your talk you mentioned that there were several
features of this plant for which you had applied for patents, I wonder if you are able to
point out any of those?
Charles Ganze; Standard Oil, since they did the R & D and design work, has applied
for the patents. I'll ask Jim Grutsch if he will come back up here and speak on that for
just a moment.
Jim Grutsch: More than 20 claims are allowed in the patent covering areas of (1) pre-
treatment, (2) the activated sludge process, and (3) the integrated end-of-pipe treatment
sequence. Some of the specifics are as follows: Pretreatment (1) The use of a portion of
the pretreatment basin for reduction of immediate oxygen demand (IOD). (2) The use of
certain catalysts for IOD removal. (3) The addition of colloid destabilizing chemicals follow-
ed by filtration of suspended solids and oil and grease to levels of 10 mg/1 and less.
Activated Sludge Process (1) Two stages of sludge treatment in facilities for one unit; i.e.,
recycle of part of the clarifier return sludge to: (a) the aeration tank rundown line, and/or
(b) a thickener thence to the aeration tank rundown line, and/or (c) a thickener, aerobic
digester, thence to the aeration tank rundown line. (2) Operating at a sludge age greater
than 10 days. (3) Pretreating the activated sludge unit influent to 10 mg/1 or less suspended
solids and oil and grease. (4)Using air or oxygen between stages of the activated sludge unit.
For example, turbulence in the wide well region of the clarifier was visible in the slide.
This turbulence is intentional. Air is educted into the buried rundown line so that the air/
water mixture is under a hydrostatic head to improve oxygen transfer. When the air is re-
leased in the wide well clarifier, it provides flocculating energy. Process. A process
consisting of an equalization section incorporating hydraulic equalization, pH control,
pre-aeration, colloid destabilization, and filtration to 10 mg/1 or less of suspended solids
and oil and grease, followed by an activated unit incorporating two stages of sludge recycle
and interstage aeration, in which the average sludge age is greater than 10 days.
A Tulsa based firm, William Brothers Waste Control, is authorized to use these feat-
ures.
BIOGRAPHY
Charles W. Ganze is Facilities Supervisor at the
Gulf Coast Waste Disposal Authority's 40-Acre and
American Facilities in Texas City, Texas. He has
acquired two degrees from East Texas State College. He
holds a B.S. in General Science/Biology and a M.S. in
Earth Science with a minor in Biology. Charles taught
science in the LaMarque Independent School District
(1967-1968) and the Andrews Independent School
District (1966-1967) and was employed by Union Carbide
Corporation as a Secondary Waste Treatment Research
Worker prior to his employment with the Gulf Coast Waste
Disposal Authority in 1973.
-------
158
BIOGRAPHY
Joe P. Teller is Deputy General Manager in
charge of Operations at the Gulf Coast Waste Dis-
posal Authority in Houston, Texas . He holds a
B.S. Degree in Civil Engineering from Texas Tech-
nological College and a M.S. Degree in Sanitary
Engineering from the Georgia Institute of Tech-
nology. Mr. Teller was employed as Deputy
Director of the Texas Water Quality Board in
Austin before joining the staff of the Gulf Coast
Waste Disposal Authority in 1972.
-------
OPTIMIZATION OF THE ACTIVATED SLUDGE PROCESS
THROUGH AUTOMATION
Mohamed A. Zeitoun
The Dow Chemical Company, Texas Division, Freeport, Texas
As an introduction, I would like to go over our accomplishment's in extending fhe
application of the activated sludge process for the treatment of brine wastewaters.
ACTIVATED SLUDGE TREATMENT OF WASTEWATERS OF HIGH SALT CONTENT
Many petrochemical processes produce organically contaminated wastes of up to
12% NaCl . Glycols and glycerine made by the chlorohydrin process result in a brine
waste of 6-10% NaCl and constitute the major organic waste load at the Texas Division
of The Dow Chemical Company.
Up to 1969, operation of the activated sludge process was believed to be possible
at salinities less than 2% NaCl . Although numerous bacterial species living in very high
salt concentrations were known, utilization of halophilic bacteria in a biological
oxidation process was lacking.
Under an EPA grant (1969-1971) we started with a fill and draw procedure to find
the minimum dilution required. We started with sewage sludge and slimes from ditches
and holding ponds exposed to the wastewater from the glycol production facilities,
and at 2% NaCl concentration. Then we raised the NaCl concentration gradually, over
a period of 6 weeks, to 10% salt. To our surprise, no dilution was required. Then we
operated laboratory continuous units, obtaining 80-88% TOD removal at a residence
time of 12 hours, a good settling sludge, but a turbid effluent. A chemical flocculation
step was required to clarify the effluent and maintain low suspended solids.
We then built a mini-plant on a skid for a flow rate of 0.5 to I . 0 gpm and fed it
continuously from the production facilities. We then learned the necessity of good
equalization. The mini-plant (Figure I) is essentially equalization-aeration-bio-settl ing-
flocculation and a final settler.
At that point in time we believed that activated sludge can be extended to the
treatment of brines up to 12% salt content. The operation was checked in a pilot plant
(150 gpm) using commercial size equipment. Its operation continued for over two years
where several sludge handling methods were tested and it served as a technology center
for the design of several treatment facilities around the world. We spent about $1 million
on this testing facility.
In the Texas Division at Freeport, Texas, we utilized existing holding ponds and
built a mammoth aerated lagoon system using submerged Kenix mixers. It is treating
20,000 gpm in 45 acres of aerated basin at a capital cost of about $5.5 million. Now
we are adding air flotation, settling tanks and pressure filters to remove the suspended
solids from the aerated lagoon effluent at a capital cost of $7 million.
159
-------
160
At Stade near Hamburg, Germany, we builf an activated sludge planf to treat a mixed
effluent of glycol, glycerine, and other minor streams, which were tested in mini-plant
operation. This plant has been in operation for about 2 years. It is designed for 9,500
gpm at a cost of $7 million. Another plant built at Sarnia, Canada, has been in operation
for about 4 months, and is treating 3,000 gpm at a capital cost of $4.5 million.
The Louisiana Division at Plaquamine is building a Unox activated sludge plant at a
cost of $8.5 million for treating 8,500 gpm and is scheduled to start early 1977. A total
of $33 million has been invested in plants treating 60 million gallons per day, (Table I),
of brine wastes all over the world. All this resulted from an initial study sponsored by an
EPA grant costing the taxpayer $150,000. This is the story of the best spent $150,000 by
EPA.
AUTOMATION OF THE ACTIVATED SLUDGE PROCESS
The problems of automated control of activated sludge plants lie in selecting parameters
for control and developing sensors for these parameters. The common practice of monitoring
plant operations with random daily or weekly analysis is inaccurate and clearly inappropriate
for control and optimization. Parameters to control the process must be measured on a con-
tinuous basis or must be obtained rapidly enough on a repetitive basis so that updated data
are available for use within the time constraints of the process.
The objectives of the reported study were to develop on-line control systems that will
(a) add nutrients (N&P) in proportion to the total carbon in the feed, (b) control the sludge
recycle rate by an F/M signal that is measured as the ratio of the total carbon in the feed
to that in the mixed liquor, (c) control the chemical flocculant addition at the end of the
process in proportion to the turbidity of the bio-settler overflow, (d) divert the feed to a
holding pond if a toxicant is detected in the feed by an instrument "Biological Inhibitor
Detector", that measures the oxygen uptake of the biota on repetitive basis.
The existing, skid-mounted mini-plant was renovated and new electrical circuits for
the instrumentation were added. The control panel (Figure 2) included three 2-pen recorders.
The influent flow, set manual'y at the board, and the pH of the influent were both recorded.
The sludge recycle was manually controlled and recorded. The alum flow rate and the
turbidity of the final effluent were both recorded and controlled at the board. The pH of
the final effluent was both recorded and indicated on the board.
Influent pH Control
The influent pH control system (Figure 3) was designed for fast response in a well
mixed system. By coupling the pH feedback loop around the recycle feed pump, the loop
capacity, mixing time and measurement lags were kept to a minimum. The pump also
served as the feed pump to the aeration basin and provided the feed at the proper pH
independent of the volume and pH of the influent in the equalization tank. The pH
control system had a high and low limit switch that closed the feed valve if the pH got
out of control and automatically opened the valve as the control range was regained.
-------
161
Food to Microorganisms (F/M) and Nutrients Addition Controls
The F/M control system is shown schematically in Figure 4, together with the
nutrients addition control system. An Ionics Model 1212 Total Carbon Analyzer was used
to measure the total carbon in the feed and that of a homogenized and diluted mixed
liquor sample from the aeration basin. To achieve this, the sample handling system of
the carbon analyzer was modified to accommodate two sample streams, one a continuous
flow and the other a batch sample. A schematic of the sample handling system is shown
in Figure 5. A continuous sample of the feed passes through a 3 micron Cuno(R) filter,
through a valved rotameter and through a two-way solenoid valve (A). It then passes
through the Analyzer sample injection slide valve to a 3-way solenoid valve (B) and on
to drain. A 1:1 dilution of the mixed liquor sample aids in the homogenizetion step
required to obtain a representative sample. Two timers operate sequentially to first
operate pumps to provide the aliquot and next operate the Virtis(R) 45 homogenizer.
Upon completion of the homogenization time, another sample pump empties the homo-
genizer contents into the stand-pipe above solenoid (B), which allows deaeration of the
sample before analysis. All three solenoid valves are energized at this moment; two-way
solenoid (A) is closed and the drain connections of valves (B) and (C) are closed. The
homogenized sample passing through the stand-pipe above valve (B) flows by gravity
through the sample injection valve and into the stand-pipe above solenoid valve (C)
until the stand-pipe levels approach equilibrium. Flow in this direction flushes the
prior feed sample from the injection valve. Upon injection of the homogenized sample,
all of the solenoid valves are de-energized, restoring the flow of the continuous feed
sample stream through the system. Meanwhile,the batch sample in both stand-pipes is
drained.
The output from the Total Carbon Analyzer is converted to a pneumatic F/M signal
with a 0 to 1 range. The signal is fed to a recorder-controller which adjusts an output
pneumatic signal proportional to the offset between the desired and actual F/M valve.
The output signal controls the percent energized time of a 60-second cycle timer, that
operates a normally closed solenoid valve allowing the flow of the recycled sludge to
waste. The system also includes a high signal selector and regulator that is set to
designate a maximum waste flow rate independent of the F/M controller. This safety
device prevents bacteria washout in case of control system failure
Nutrients Addition Control: Constant nutrients addition, based on flow rate, will
occasionally result in excess or insufficient nutrients addition. Insufficient addition
decreases bacterial viability and decreases the organic carbon removal efficiency.
Excess nutrients addition, besides being economically unsound, can cause downstream
algae blooms and subsequent pollution problems. High nutrients residual may also
prevent the recycle of industrial treated effluents.
The activated sludge mini-plant was operated with varying flow of nutrients to the
aeration basin based upon the measurement of the total carbon in the feed. Ammonia
and phosphorus concentrations, in the biosettler overflow, were analyzed routinely to
determine the optimum C:N:P ratio, for the waste water being treated. Figure 6 shows
the relationship between ammonia consumption and microbial growth. Although the
-------
162
experimental data reflects the ammonia consumption for cellular growth plus the ammonia
absorbed onto the bio-floe, the experimental value of 0.133 compares closely with the
theoretical value of 0.15 according to cell formula Cj-
In order to determine the minimum nutrients requirement, the ratio of nutrients
addition to feed total carbon concentration was progressively lowered until analysis of the
bio-settler overflow showed a very low concentration of N and P. The results shown in
Figure 7 indicate a large adsorptive capacity of the mixed liquor bio-floe for values of
N :C greater than 0.2. The C:N:P ratio of 100:9:0.54 was found to be optimum for the 8%
salt containing wastewater. Using this proportionally the nutrients control system assured
microbial viability and the effluent concentration was maintained below 5 ppm ammonia
and I ppm phosphorus.
F/M Control System: A transient or unsteady state model of the completely mixed
activated sludge process is developed to simulate the dynamics of the F/M control system.
The Monod kinetic equations are incorporated into the substrate and microorganisms
material balance equations around the aeration basin. The substrate degradation is assumed
to occur only in the aeration basin.
The scheme of F/M proportional control is inserted into the model as a function of the
wasted sludge flow rate.
W = A (F/M) + B, O - W - W
v ' ' max.
where W - wasted sludge flow rate
A and B are constants related to a specified gain.
A maximum waste flow rate, W is imposed to prevent complete cell washout in case
the control system should fail. The' control is related to the recycle flow by:
T= R+ W
R = T - [A (F/M) + B]
where T = total biosettler underflow rate
R = recycle flow rate
The dynamics of the unsteady state bio-settler operation are computed in terms of a con-
stant compaction ratio, X /X. There are obvious limitations in using a constant ratio to
describe the unsteady state performance of a biological settler, however it was adequate
for the predicfionof the bacteria concentration in the aeration basin. Figure 8 shows that
the sludge compaction ratio remained relatively constant during a 50% increase and
decrease in feed concentration.
The Monod type kinetic equations were found adequate in simulating the dynamics
of the activated sludge treatment of propylene glycol waste water because of four basic
reasons:
-------
163
I. The Monod model was developed to describe the behavior of pure cultures.
In the fresh water glycol system, 80% of the microorganisms present were of
one species, Pseudomonas. In the brine (6-8% NaCl) glycol system 90%
of the culture was identified as one halophilic bacterium.
2. The propylene glycol represented the only soluble carbon source in the feed.
3. The carbon energy source was rate limiting. Deficient nutrients (N&P) were
added by the nutrients control system.
4. The pH and temperature were regulated for the optimum rate of biological growth
Sludge growth coefficients and substrate removal kinetic constraints were determined for
freshwater glycol feed and saltwater glycol feed. For the freshwater biota, the yield
coefficient Y = 0.21 gm VSS/gm TOD removed, and the decay coefficient k . = 0.06
day , while for the saltwater bacteria, these values were 0.133 and 0.02 respectively.
The substrate utilization rate for the freshwater system is best described by:
dF - k xS]
dT
where k = 5.0 day
and K = 117 mg/l as TOD
For the saltwater bacteria, K is very large compared to S.. (substrate concentration in the
aeration basin), and the rate equation reduces to:
dF = k x $
dT '
where k = 0.0132 (mg/l TOD)'1 day''
Testing of the F/M Control System: The mini-plant test runs were started with the F/M
controller on manual, i.e., constant recycle and sludge waste flow rates were set. The
system was allowed to approach a steady-state conditbn, in terms of reactor microorgan-
ism concentration, substrate removal efficiency, and bio-settler characteristics such as
sludge concentration and sludge blanket height. The system was then put on automatic
control and a step increase in the concentration of the feed was made by adding an
aliquot of industrial grade propylene glycol to the equalization tank.
An automatic Total Carbon analysis was made every six minutes during these runs,
with alternate analysis of feed sample and homogenized mixed liquor sample. To reduce
the volume of data, each F and M concentration plotted in this section is represented as
the average of five total carbon analyses. Waste, recycle and F/M data are also hourly
averaged values.
A blank F/M test in which no control is affected shows the dynamics of cell growth
under continuous flow conditions and is used for comparison for F/M control test system
-------
164
response. Figure 9 shows the mixed liquor total carbon and F/M response to a 47% increase
in loading. The solid lines are the predicted responses using fhe transient state computer
model. A very good fit is obtained for the increase in mixed liquor total carbon. Because
of a non-linear calibration of the total carbon instrument above 50% of scale, data for F/M
signal lie below the computer prediction of F/M, till the mixed liquor total carbon values
reach the same non-linear range as the feed values, then the predicted curve agrees with
the experimental data.
The results of one of the control tests made with freshwater gIycoI feed are presented
in Figures 10 and II. The initial conditions of this test run compare closely with those
used in the blank test (Figure 9). The model prediction of the increase in mixed liquor
total carbon (t = 0 to t = 27.5 hours) fits the data well. The F/M signal data lie above
the predicted F/M curve for the first several hours following the step increase and below
the predicted curve following the step decrease, because an average value of feed con-
centration is used in the simulation, rather than individual data points. A good represen-
tation of recycle and wastage response is obtained with a fast F/M response (Figure 11).
The prediction of the transient biological state after the step decrease in feed concentra-
tion (t = 27.5 hours) is not accurate because of an inadequate representation of the bio-
settler dynamics. However, the final steady state computer prediction ( t =50 hours) fits
the data well.
In order to have a meaningful basis of comparison for the F/M control tests, it is
necessary to have a blank run with identical initial conditions. The initial steady state
must be the same in terms of feed, recycle and waste flow rates, substrate concentration,
reactor and bio-settler underflow microorganisms concentrations, temperature, and
physiological conditions of the biota. While this is difficult to achieve experimentally,
a valid comparison can be made by using the mathematical modeling which predicts the
transient accumulation of microorganisms.
The computer simulated curves in Figures 12 and 13 show the differences in response
characteristics for the freshwater propylene glycol system with F/M control versus the
system with no control. For the case of an 8.5-hour hydraulic residence time with an
initial recycle ratio of 0.55 it takes 17 hours for the controlled system F/M to return to
within 20% of the set point after the step increase in loading (Figure 12). This represents
a 63% reduction in response time. Figure 13 represents the system having a 6.95-hour
hydraulic residence time and an initial recycle ratio of 0.3. The F/M ratio returned
to within 20% of the set point in 10 hours, less than a third of the time required for the
system without control. The larger mass of waste sludge available for recycle allowed a
69% reduction in response time.
One of the most important achievements of this control system is maintaining the
effluent quality. Figure 14 shows the relationship between substrate concentration in the
aeration basin for the blank and the F/M control simulation of Figure 13. The control
enables a fast return to the desired effluent quality.
Figure 15 shows the simulation of an F/M blank test and an F/M control test for the
brine propylene glycol waste water system. Low volume of waste sludge is available for
-------
165
additional recycle, due fo the lower yielding halophilic bacteria, which results in the
rather long response time of 33 hours with F/M control. Manual adjustment of the recycle
and waste flows would be adequate for loading control because of the long response times
involved. On-line F/M control for high salt systems does not seem justifiable unless an
aerated sludge storage or stabilization tank is available to supply additional recycle in
periods of high loading.
Chemical Flocculation Control System
The overflow of the bio-settler is too turbid, especially in case of saline wastewaters,
and needs additional clarification. A chemical flocculation control system was used to
remove the suspended and colloidal matter from both the fresh water and the highly
saline biologically treated effluents.
The schematic in Figure 16 illustrates the flocculation controls and process flows.
The proportional control was based on a turbidity measurement. Aluminum sulfate was
used as the flocculant, and the pH was controlled (proportional-integral) at the level
corresponding to the optimum for alum flocculation. Both feed forward and feed back
modes of control were tested. A Hach Surface Scatter Turbidimeter, No. 2426,
operated better than other types tried because it has no contact between the sample and
the optical surface.
The optimum pH for alum flocculation was determined by a series of jar tests at
various pH values, on samples having the same initial turbidity. The optimum alum
dosage was found to increase above pH 6 and below pH 5. At a pH above 7.0, the
settled floe was more fluffy and had a higher subsidence line.
A stoichiometric relationship was found between the bio-settler overflow turbidity
and the optimum alum dosage, at constant pH (Figure 17). Comparing point 5 with point
6 in this graph illustrates the effect that the physiological condition of the bacteria has
on the alum requirement for optimum flocculation. Data point 5 represents a series of
jar tests on a sample of bio-settler overflow whose initial turbidity was increased from
25 JCU to 44JCU by adding aeration basin mixed liquor. Thus the sample contained
larger bio-colloids of older microorganisms. On the other hand, data point 6, where
the higher optimum alum dosage was required, represents a case of finely dispersed
microorganisms, as the sample was collected during a period of very high growth rate
of microorganisms in the aeration basin.
Because of the lag time of several hours associated with the flocculator and final
settler, control stability was a major problem in feed back control. The feed forward
mode of proportional control resulted in a high quality effluent and stability was not a
problem.
The feed forward control was tested on both the propylene glycol fresh water system
and the glycol-salt waste water. Typical performance is given in Tables 2 and 3
respectively. An average of 98% of the suspended solids were removed by the floe-
-------
166
culaHon step in the fresh water system and 85% were removed in the salt system. The
chemical flocculation step increases process efficiency in terms of TOD removal. For
the fresh water system the overall process efficiency increases from 90.9% TOD removal
to 93.4%. The turbidity of the final fresh water effluent was reduced to 2 to 3 JCU
whereas the bio-settler overflow turbidity was typically 15 to 25 JCU. The bio-settler
overflow turbidity for the saline waste water system was reduced by 85%, from 35 to 80
JCU to 5 to 10 JCU.
Biological Inhibitor Detector
The objective was to develop an upstream sensing device for toxic loads or spills to
an activated sludge process. If the presence of a toxin in the feed to the plant can be
detected rapidly enough on a repetitive basis, the feed could be diverted to a holding
pond until the toxin is identified and eliminated, or bled at a low rate to the treatment
plant.
Oxygen uptake rates, both Warburg and oxygen electrodes, have been used to
measure the effect of toxins and inhibitors on activated sludge. Tests were conducted to
determine an oxygen uptake measurement cycle that would be sensitive to the presence
of a toxic substance in the feed to an activated sludge plant and that could be automated
to rapidly detect the toxic effect on a repetitive basis.
The oxygen uptake rate and the total oxygen consumed by propylene glycol standard
solutions containing various concentrations of toxic substances were measured as 10 ml of
the sample was introduced in 500 ml of activated sludge under controlled conditions of
mixing, aeration, and temperature.
The sensitivity and reproducibility of the total oxygen consumed, as determed from the
area under the oxygen uptake curve, was found superior to that of the oxygen uptake rate,
as determined from the initial slope of the oxygen concentration curve when the aeration
was cut off. The rate of oxygen uptake was more dependent on the concentration of the
viable biomass than the total oxygen consumed. The automation of a measurement cycle
utilizing the area under an oxygen uptake curve was simpler since it did not require cutting
off the air every time a test sample was added.
In order to utilize the area under the oxygen uptake curve to monitor the toxic effects
of a feed to an activated sludge plant, it is necessary to distinguish between the changes
due to the varying concentrations of the substrate in the feed and the presence of toxic
substances. A comparison of the oxygen uptakes of two standard propylene glycol samples
before and after the exposure of the bacteria to a feed sample was the measurement cycle
adopted to measure the toxic effects of the feed. The oxygen uptake of the first standard
established a reference activity of the bacteria sample, which was then exposed to the
feed to the process for long enough time to degrade the substrate in the feed. Then the
same bacteria sample received an equal volume of the standard and its oxygen uptake,
compared to that of the first standard, was expressed as an activity ratio in response to the
feed sample.
-------
167
Figure 18 is a schematic illustration of the measurement cycle (60 minutes), when
the feed contained no toxic substance. The activity ratio would be approximately
equal to 1.0. Figure 19 shows a measurement cycle when the feed contained a toxic
substance resulting in an activity ratio much less than 1.0.
Several models of an instrument based on the automated measurement cycle of three
oxygen uptake curves were built in-house and used extensively in screening the bio-
degradability and toxicityof industrial waste waters, organic and inorganic substances.
The instrument can be regarded as a differential respirometer, and has the added feature
of measuring a BOD value of the feed in less than one hour. The feed BOD value,
calculated from the ratio of the feed oxygen uptake to that of the first standard multiplied
by the BOD of the standard, compares very well with BOD,- for values of 100 to 800 mg/l
(Table 4). 3
PROCESS DESIGN CONSIDERATIONS - CONTROLLED ACTIVATED SLUDGE
The development of on-line control systems for the activated sludge treatment process
opens up many possibilities for flexible and innovative process design.
Figure 20 illustrates a conceptual design of an activated sludge treatment system
including the F/M control system and the biological inhibition detection (BID) system.
Equalization dampens fluctuations in the organic content, in flow, in pH, and in
other physical-chemical characteristics of the feed, thus it is parr of the F/M control
system. The equalization tank is also partof the feed forward inhibition detection
control system. It provides a time delay, together with dilution of a possibly toxic
feed during the biological activity analysis period. A toxic feed would be diverted to
a holding basin by the BID control system. After determination of the chemical nature
of the toxin, it could either be bled back into the system or treated by other means.
Pre-treatment could include chelation of toxic metals and sorption of toxic organics.
The F/M control system, based on flow measurement and organics concentration,
includes an aerated waste stabilization tank which supplies additional microorganisms
for recycle in periods of high loading. The stabilization tank may be located in the
sludge waste line as shown in Figure 20, or it could be located in the sludge recycle
line; depending on which site is more adequate in terms of microorganisms viability and
effects on bio-flocculation.
If the feed sample for F/M control is taken before the equalization tank^ F/M
control will be more anticipatory in nature with very short response time. The control
system could then include a mini-computer which would be programmed to calculate
the hydraulic effect that the equalization will have on the feed flow and composition
and would affect F/M control on this anticipatory basis. Hence, extra sludge recycle
could begin before the higher loadings affect the aerated biomass.
-------
168
Another available control system that needs to be incorporated into the plant design
is an aeration basin oxygenation system that optimizes oxygen allocations and controls the
oxygen supply to the stabilization tank.
Nutrients and pH control would be provided when required. Turbidity control would
be provided for designs that employ flocculation as the tertiary treatment for colloidal and
suspended solids removal.
In conclusion, future plant designs have to incorporate both control systems and
predictive models in order to optimize the operation of the activated sludge process to be
able to meet the more closely regulated effluent limitations. Both design and operation
of sophisticated plants depend upon the availability of reliable sensors and the instrumen-
tation necessary to control the processes.
REFERENCES
Details of the research reported in this paper are given in two reports published by
U.S. Environmental Protection Agency and are for sale by the Superintendent of Documents,
U.S. Government Printing Office, Washington, D.C. 20402.
(I) "Treatment of Wastewater from the Production of Polyhydric Organics, " 12020 EEQ
10/71 Water Pollution Control Research Series (1971), includes 18 references.
(2) "Optimizing a Petrochemical Waste Bio-Oxidation System through Automation,"
EPA-66-/2-75-02I Environmental Protection Technology Series (June 1975), includes
61 references.
ACKNOWLEDGEMENTS
The work reported was performed at the Texas Division of The Dow Chemical Company
by the author as project director, and N. J. Biscan, J. H. Culp, H. C. Behrens, W. D.
Spears, and R. W. Murray. W. F. Mcllhenny was the Project Manager.
These projects were partially supported by a grant from the Water Quality Office of
the Environmental Protection Agency. Appreciation is expressed to the personnel of
Roberts. Kerr Environmental Research Laboratory, Ada, Oklahoma, for their cooperation
and assistance.
DISCUSSION
Morris Wiley : Was this a calcium or a sodium brine?
Mohamed Zeitoun: Mainly sodium chloride.
Morris Wiley: Have you tried treating this in combination with other organic waste?
-------
169
Mohamed Zeitoun: Yes.
Morris Wiley: Is that desirable or undesirable?
Mohamed Zeitoun: That depends on the same factors governing the possible combination
of fresh wastewaters. Our aerated lagoon treats a mixture of organic waste that were
tested in the laboratory and determined to be compatible and treatable. Some organic
wastes were found to be non-toxic to the system, but don't degrade in the presence of
the main substrate which is glycol and glycerine.
Morris Wiley: Do you notice either positive or negative synergism between the waste-
waters, by that I mean you get enhanced activity or decreased activity by mixing the
brine waste and other types of waste?
Mohamed Zeitoun: Do you mean by other type of waste, that which is less saline?
Morris Wiley: Yes.
Mohamed Zeitoun: Micro-biological tests showed that the optimum salinity of this
halophitic bacteria is 6-7% NaCl .
Morris Wiley: Do you have problems with salinity swings causing osmotic pressure
difficulties on the organisms?
Mohamed Zeitoun: Gradual changes of salinities up to 10% and down to 4% have very
slight change of activity. Practically, this is not a problem.
BIOGRAPHY
Mohamed A. Zeitoun is a senior research
specialist with the Dow Chemical Company, Texas
Division, Freeport, Texas. He holds the following
degrees in Chemical Engineering: B.Sc. Cairo
University, Cairo, Egypt (1950), M.S. and Ph.D.
(1958) Texas A&M University. He taught at Cairo
University and Texas A&M. Mohamed has been
conducting research work in desalination and
industrial waste water treatment since 1962 and has
been with Dow since 1966.
-------
170
TABLE! BIOLOGICAL OXIDATION PLANTS TREATING BRINE WASTES
IN THE DOW CHEMICAL COMPANY
Process
Activated Sludge
Pilot Plant
Aerated Lagoon
A e ra t io n
Sludge Handling
Activated Sludge
Activated Sludge
Unox Activated
Sludge
Location
Freeport, Texas
Freeport, Texas
Stade, Germany
Sarnia, Canada
Plaquamine, LA
Design Flow
Rate, gpm
150
20,000
9,500
3,000
?,500
Capital Cost
Million S
1.0
5.5
7.0
7.0
4.5
8.5
Remarks
Incl. 2 years operation
Start-up late ''976
Start-up early 1977
TOTAL
60 MGD
$33.5 million
TABLE 2 ALUM FLOCCULATICN SYSTEM PERFORMANCE*
GLYCOL-FRESH WATER SYSTEM
Effluent Before
Chemical Flocculation
Effluent After
Chemical Flocculation
Day of
Operation
1
2
3
4
5
6
7
8
9
10
MLSS
mg/1
45
22
20
20
11
36
52
86
34
55
TOD Removal
%
88.8
95.5
93.5
89.3
92.0
95.2
84.!
90.7
88.3
91.7
MLSS
mg/1
4
0
0
0
0
1
0
0
0
0
TOD Removal
%
89.7
96.0
95.8
93.0
93.9
95.4
87.7
94.9
94.0
93.9
* proportional control with optimum alum dosage according to Figure 17.
effluent turbidity maintained a t 2 to 3 JCU
-------
171
TABLES ALUM FLOCCULATION SYSTEM PERFORMANCE*
GLYCOL-SALT WASTE WATER SYSTEM
Effluent Before Effluent After
Chemical Flocculation Chemical Flocculation
Day of MLSS TOD Rerroval MLSS TOD Removal
Operation mg/1 % mg/1 %
I 91 76.2 24 91.9
2 164 78.9 17 91.9
3 52 63.2 0 65.9
4 43 66.6 3
5 23 90.3 0 90.8
6 214 75.8 18 96.3
7 66 71.5 o -
8 43 69.0 19
9 52 75.4 21 80.3
10 40 78.4 18 81.0
* alum dosage at optimum (10% greater than doses in Figure 17)
TABLE 4 COMPARISON OF INSTRUMENTAL
BOD AND BOD
BOD,- Instrumental BOD*
mg/1 mg/1
190 197
536 482
588 565
815 839
* Standard propylene glycol of 400 mg/1 BOD,- used for the instrument
calibration .
-------
172
Figure 1
ACTIVATED SLUDGE MINI PLANT.,
Flow Diagram
Influent
Equalization
Acid
Nutrients
i—« 1
Feed
Pump
Aeration Basin
Sludge Recycle
Bio-Settler F1 peculator. Final Settler
Alum
Caustic
Waste Sludge
Figure 2.
ACTIVATED SLUDGE PILOT PLANT
Side View
Acid
Tank
PHV-1
PHT-1
PHE-1
Instrument
Panel
ooo
AIC-2
000
Nutrients
Tank
AT-2
FV-1
P-1
P-2
<3Ł>
i f 1 i
-------
Figure 3.
INFLUENT FLOW AND pH CONTROLS
173
Acid
Addi ti on
To -*
Equalization
From
Equalizati on
To
Aeration
Basin
Ft-1
Figure 4.
ACTIVATED SLUDGE PILOT PLANT
Nutrients Control and F/M Control
F/M Signal
Total Carbon
Analyzer
Dilution &
Homogen izer
/
*•
i o-Settler
P-l
Feed Pump
Sludge Recycle:
SIudge
-------
174
Figure 5.
F/M CONTROL SAMPLING SYSTEM
Solenold-C
1
Feed
Sample
TCA . Ł
Slide [J i 4
Valve \
3 '
Homogeni zer
— t n
T Mixed
(c9* Liquor
' -^ ~r c=* Sample
^ — ) Solenoid-B
^ L__ Dilution
1 (c-% Water
Drain i=a
> 1
/ O-V Solenoid-A
Drain ^
W Rotameter
n
Filter
-a
ja
O-
E
Figure 6.
QJ AMMONIA CONSUMPTION BY MICROORGANISMS'
0,6
0,5
0,4
0,3
I 0,2
•<
0,1
0,0
Slope = 0.133
' 0,0 1,0 2,0 3,0 4,0 5,0
[•Hero organ ism Growth, Ib vss/day
-------
Figure 7
DETERMINATION OF AMMONIA REQUIREMENT
175
50
o o
E
o.
ex
40
o
E
30
20
10
10 .20 .30 .40
Nitrogen to Carbon Ratio in Feed
.50
.60
Feed
Total
Carbon ,
mg/1
Mixed
Liquor
Total
Carbon ,
mg/1
w w w
500
400
900
700
500
Figure 3.
UNSTEADY STATE VARIATION OF THE SLUDGE COMPACTION RATIO
(DATA FROM RUN C)
600 r
SI udge
Compact!on
Ratio,
3,Or
2,0
1,0
0,0
0
10
20
30
50
Time , hours
-------
Figure 9 .
F/M BLANK TEST - BACTERIA GROWTH
AND F/M RESPONSE
(Constant Recycle Flow) Run A
Figure 1C.
F/M CONTROL TEST - BACTERIA GROWTH AND F/M RESPONSE
RUN C
Feed Flow 1938 ml/min
Sludge Recycle Flow 1070 ml/min
Sludge Waste Flow 85 ml/min Feed
Total
Carbon, 5QQ
Feed Flow 1938 ml/min
Initial Recycle Ratio 0.55
600
^ 50°
E- 400
o
300
800
^ 700
*• 600
500
0,8
1 0.7
en
" 0,6
Feed
Mixed L i_q u o r^ °o
oo
00
O Experimental
Computed
00
00
_ o
mg/1
Mixed 90°
Liquor _nn
Total 700
Carbon,
mg/1 500
. O.°ooo°
00
000
O Experimental
— Computed
F/M
Signal' 0,8 —
\ \
\ \
20 30
Time , hours
10 15 20 25 30 35 40 45 50
Time, hours
-------
Figure 11.
F/M CONTROL TEST - RECYCLE AND WASTE RESPONSE
RUN C
Feed
Total
Carbon
mg/1
F/M
Si gnal
1200 r
Recycl e
Waste
FT ow,
m 1 / mi n
000
Po°
O Experimental
—— Computed
Mixed
Liquor
Total
Carbon,
mg/1
F/M
20 30
Time, hours
Figure 12.
RESPONSE CHARACTERISTICS WITH AND
WITHOUT F/M CONTROL - LOW LOADING
Recycle
Fl ow,
ml/mi n
600
500
400
1000
900
800
700
600
0,9
0,8
0,7
0,6
0,5
1200
1100
1000
Control Simulation
Blank Simulation
Feed Flow 1938 ml/min
_Init1al Recycle Ratio 0.55
Initial Loading 1.5 gm TCp/gm TCM«da;
I
I
I
I
10 20 30 40
Time Since Step Change, hours
50
-------
Figure 13.
RESPONSE CHARACTERISTICS WITH AND
WITHOUT F/M CONTROL - HIGH LOADING
Figure 14.
COMPARISON OF AERATION BASIN SUBSTRATE CONCENTRATION
AND WITHOUT F/M CONTROL
CO
WITH
1
reed
Total
Carbon ,
rng/1
Mixed
Li quo r
Total
Carbon ,
m n t 1
mg/ I
F/M
Recycl e
Fl ow ,
ml/min
600
500
400
1000
900
800
\J W \J
700
600
—
— r. j. i r • 1 i •
^~~~~ Control oimulation
Blank Simulation i n
0 9
Feed Flow 2325 ml/min F/M
r~ Initial Recycle Ratio 0.3 Signal no
Initial Loading 2 . 1 gm TCF/gm TCM-day U,«
^^ 0,7
^^^ w | /
Digital Computer Simulation
Feed Flow 2325 ml/min
Initial Recycle
Ratio 0.3
X. """"'—- -^, ___ Blank
^^^"--..^ "~
.S ___ — — — """ F/M Control
- xC— " ""
=-^
/u
1,0
0,9
0,8
0,7
/"\ r"
60
Substrate
— ^ Cone. , _
V Total 50
\ -*^_ Carbon,
N. — — mg/l 40
_ \^_ :=i. — __
1 ^^^^^^ 30
°'6 1 20
900 r- 10 32,5
800
700
J.U
—
| ^^^^ °
~~ X"-**.
^•v ^^
J ^W x,^
1 ^^. "^"^
— 1 ^k ^"^.^^Blank
/ ^X. ~~"~~~^
- 1 X^^^
I ^*^*«^
~~^ F/M Control
—
—
1 1 1 1 |
0 10 20 30 40 50
1 1 I | 1 Time Since Step Change, hours
0 10 20 30 40 50
Time Since Step Change, hours
-------
179
Feed
Total
Carbon,
mg/1
Mixed
Liquor
Total
Carbon,
rog/1
F/M
Figure 15.
RESPONSE CHARACTERISTICS WITH
AND WITHOUT F/M CONTROL
800
700
600
500
WOO
1200
1000
800
0,9
0,8
0.7
0.6
0,5
1200
Recycle
How, 1100
m 1 / m i n
1000
Salt-Propylene Glycol Systen
Control Simulation
Blank Simulation
Feed Flow 1938 ml/min
Initial Recycle Ratio 0.55
Initial Loading 1.5 gm TCp
gm
33
20 40 60 80 100
Time Since Step Change, hours
Figure 16.
FLOCCULATION CONTROLS
Al urn
Bio-Settler '
(50 gallons)
Caustic
Floe-Settler
(50 gallons)
Final
Effluent
>
Chemi cal
Flocculator
(10 gal Ions)
'Waste
SIudge
7)Feed Forward Turbidity Measurement
Feed Back Turbidity Measurement
-------
180
Figure 17.
OPTIMUM ALUM DOSE AS A
FUNCTION OF INITIAL TURBIDITY
0,4r-
OJ t.
CO OJ
O +->
O T-
E \
13 t^
r— 01
L.
J_
O)
Q-
90
80
70
60
o
Z 50
(0
c
m Li(]
o "u
c
o
S 30
CD
X
o
10
20 30 40
Time, minutes
50
60
-------
Figure 19.
BIOLOGICAL INHIBITOR DETECTOR
SCHEMATIC OF MEASUREMENT CYCLE WITH TOXIN IN FEED
181
° 90 r-
80 -
„ 70 -
c
01
Ł 60 n
I 50
C
a) i|Q
o '^
c
o
u
Ł 50
en
X
10
20 30
Time, minutes
Figure 20.
CONTROLLED ACTIVATED SLUDGE TREATMENT
Control Unit
Feed
Control Unit
Final
Effluent
Aerated
Stabi1i zati on
Tank
Recovery
or
Disposal
SI udge
Handli ng
-------
THE NATIONAL PETROLEUM REFINERY WASTEWATER CHARACTERIZATION STUDY
Leon H. Myers, FredM. Pfeffer, and Marvin L. Wood
U.S. Environmental Protection Agency
Robert S. Kerr Environmental Research Laboratory
Ada, Oklahoma 74820
ABSTRACT
The EPA and API jointly conducted an investigation to provide a verifiable data
base under actual operating conditions in the treatment of petroleum refinery wastewater.
This data base consisted of raw waste loadings and activated sludge treatment efficiencies.
Industrial participation in the raw waste load phase accounted for 84 percent of the U.S.
petroleum refining capacity.
INTRODUCTION
In 1965, with the passage of amendments to the Federal Water Pollution Control
Act, the API Committee on Air and Water Conservation instigated a project to secure
representative data on wastewater treatment and control practices in domestic refineries
of the petroleum industry (1). The response to this project represented 93 percent of the
industry's crude processing capacity. From this response, a profile of wastewater effluent
quality was prepared reflecting the various types of operation in 1967.
The Refuse Act Permit Program and the impending legislation which became the
Federal Water Pollution Control Act Amendments of 1972 (2) necessitated that the Envir-
onmental Protection Agency acquire a scientific characterization of various wastewater
streams and a factual determination of the efficiencies of available wastewater treatment
processes so as to ensure protection of the environment from any adverse effects caused by
wastewater discharged under a permit system. In January of 1971, a series of contracts
was let by EPA's Office of Water Programs to assemble available information and to utilize
consultant expertise as a base from which effluent limitations might be prepared. At that
time, the prevailing concept was to determine "standard raw waste loads" coming from
the various unit processes, apply the optimum contaminant removal efficiencies afforded
by available wastewater treatment processes, and then determine allowable effluent char-
acteristics. There were limitations, however, in the usefulness and adequacy of the
available and current technical data. The API offered to cooperate with EPA in con-
ducting a survey of the petroleum refining industry to update the 1967 data. Following
Agency acceptance of the offer, a planning committee was formed consisting of represent-
atives from EPA's Office of Research and Monitoring, EPA's Water Quality and Refuse
Act Permit Section, State water quality offices in Texas and California, the API, and the
consulting field.
The committee agreed that the refining classification scheme used in the 1967 API
survey should be incorporated into the plan of a joint API/EPA study. This classification
assumes that the extent of processing complexity of oil refining would have a parallel
impact on the character and quantity of contaminants entering a given refinery wastewater
183
-------
184
system. The five refinery classifications were and are:
A. Crude topping (Atmospheric and vacuum distillation)
B. Topping and catalytic cracking
C. Topping, cracking, and petrochemicals
D. Integrated (Topping, cracking, and lube oil processing)
E. Integrated plus petrochemicals
Applying this fundamental refinery classification scheme to the nation's refineries, Class A
represents about 5 percent of the domestic crude capacity, while Classes B, C, D, and E
account for 36, 23, 26, and 10 percent, respectively (3).
The survey objectives agreed to by the committee were:
1. To determine the refinery raw waste load;
2. To verify analytical results from representative refineries;
3. To determine the contaminant removal efficiency of selected refinery waste-
water treatment systems; and,
4. To evaluate and update interim reference guides for the petroleum refining
industry.
The results of the raw waste load survey have been reported by the API (4). This presenta-
tion today concerns the second and third objectives: analytical validation and treatment
efficiency. A draft report on the study is under consideration for publication (5).
ANALYTICAL VERIFICATION SURVEY
For the purpose of validating data obtained in the questionnaire portion of the raw
waste load survey, EPA conducted a study of 17 of the 150 refineries which submitted the
questionnaires. Selection of refineries to be monitored by EPA was made by the joint
API/EPA Committee from the list of participating refineries by use of a random number table.
Code numbers for plant identification were used to insure confidentiality. Validation
consisted of a statistical comparison of results obtained when refinery and EPA personnel
conducted thorough monitoring and sample-splitting programs at each of the 17 refinery
locations. Responsibilities during the study were shared. The refineries' personnel
documented flows and operating conditions, collected, composited, split, and preserved
the samples for independent laboratory analyses within 24 hours of sampling under observa-
tion by EPA personnel.
Refinery Selection. Table 1 shows the number of refineries by class and crude cap-
acity which were monitored by EPA. Also included are the questionnaires received from
each refinery class, showing that for purposes of validation, EPA monitored 10 percent of
the respondees in each class. Nominally, a 10 percent random sampling of a total popu-
lation constitutes a statistically sufficient sample, and the results of many polls or surveys
are based on much less data than such a sampling. Thus, the survey proposed by API in
actuality approximated a study of the entire population, i.e., the petroleum refining
industry. This conformity to strict scientific and statistical procedures continued during
survey sampling, analysis, and interpretation of the data.
Analytical Procedures . Selection and implementation of laboratory methods are
critical decisions for any survey. Analytical methods must be reliable; that is, they should
-------
185
have widespread usage among water quality laboratories, compensate for interferences in
samples, and be precise and accurate. In addition, the methods must conform to the
equipment, skills, and analytical time limitations of the laboratory and the survey plan.
The procedure for each method must be clearly described, including references and any
modifications necessary because of interferences.
The joint API/EPA task group selected the 22 parameters for the study as shown in
Table 2. A complete wastewater characterization of any industrial effluent would require
analyses of every known wastewater parameter. The wastewater characteristics evaluated
during this study do not reflect all known parameters but do represent the major constituents
expected in this industry's wastewater.
In addition to these 22 parameters, EPA's sample splits were subjected to further
analyses. A comparison was made of the hexane and freon gravimetric procedures for
"oil and grease"; and additional parameters were measured: color, turbidity, volatile
suspended solids, specific conductance, nitrite nitrogen, and acute fish toxicity. All
participating refinery laboratories were instructed to adhere to the EPA's approved methods
(6), as modified slightly in 1972; copies of these methods were distributed accordingly.
Analytical Quality Control. A host of determinate errors inherent to any survey of
this nature dictated the incorporation of an analytical quality control program. Beginning
with the planning phase, emphasis was placed on such laboratory variables as selection and
sensitivity of instruments, glassware cleaning, and reagents. Control of performance was
accomplished by following the EPA's approved methods of analytical quality control devel-
oped at the Robert S. Kerr Environmental Research Laboratory (RSKERL) in Ada, Oklahoma.
A systematic program of accuracy and precision testing was implemented. Initially, a
sample of API separator effluent was examined by each of the methods in order to ascertain
problem interferences as evidenced by comparison of duplicate results for precision and
spike recovery results for accuracy.
Factors such as the number of samples received and time limitations for completion
of analyses precluded duplicating and spiking each sample. The laboratory quality control
program involved running a blank, standard, sample duplicate, and spiked sample —for
every eight samples received. Success in duplication and spike recovery indicated that
the results obtained for the corresponding eight-sample set were correct with a high degree
of confidence. Failure presented two alternatives: rerun the set of eight samples, or dis-
regard the results entirely.
The EPA's quality control results are summarized in Table 3 for the more important
parameters. In all but three instances our Laboratory met the frequency requirement of
^ 1.0 per eight-sample set. With the exception of sulfide, we achieved s 95 percent
accuracy (or spike recovery).
Results and Interpretation. In anticipation of a monumental task of interpreting
the field and analytical data generated during the studies, it was decided to utilize RSKERL's
Computer and Statistical Section for data accumulation. Verified analytical data were
keypunched and machine verified by Computer Section personnel. Uniform rules for
-------
186
rounding off decimal points and not reporting zero values were established for the data.
Negative values were reported as negative values. Toxicities encountered in BOD,- analy-
tical data were treated uniformly by selection of the highest consistent values. Statistical
techniques for data evaluation, interlaboratory and intralaboratory deviation, frequency
distribution of parameters, quality control regression analysis, student "t" testing, and
plotting were paramount to the studies.
Distribution analyses were performed on the raw waste loads per 1,000 barrels of
crude throughout for each of the 22 parameters and each of the five refining classes. Median
values were selected as the best measure of central tendency on all parameters because of the
highly skewed frequency distribution (Figure 1) which is a common occurance in water quality
data. The difference between the average (62) and median (37) in this figure is quite appar-
ent.
Comparison of the 22 parametric results from 17 refineries with the EPA results was
made. A summary of the difference in median values (mg/1) between EPA and API analyses
and the student "t" values is shown in Table 4. The "t" value relates to statistically sig-
nificant differences of analytical results at the 95 percent level. When the calculated
absolute value of "t" at the 95 percent confidence level is ^2.0, the medians are signifi-
cantly different. The parameters for which this occurred are oil and grease, phenolics,
sulfide, total chromium, ammonia nitrogen, cyanide, iron, copper, lead, and zinc. Note
that BOD was a border line case. This observation of value differences between the same
sample points is further justification for a strict quality control program.
ACTIVATED SLUDGE TREATMENT EFFICIENCY SURVEY
The activated sludge treatment process had previously been recognized by the joint
API/EPA committee to be representative of the best level of treatment currently used by
the refining industry. Consequently, this treatment system concept was selected for the
efficiency studies.
Unlike the net raw waste load survey where the sampling was done by refinery per-
sonnel, the activated sludge efficiency survey required all sampling to be accomplished by
EPA staff, with scheduled analysis at RSKERL in Ada. Field sampling work was planned to
extend over fourteen 24-hour periods as compared to a one-day period (24 hours) for the net
raw waste load survey; therefore, the EPA manpower impact would be somewhat extensive
from both the field and the laboratory standpoint. The efficiency study was managed and
coordinated by EPA's RSKERL, with aid in actual field sampling provided by EPA Region VI
representatives from Surveillance and Analysis and Enforcement staffs.
Refinery Selection. In the selection of the participating refineries, significant
consideration was given to the choosing of those refineries with treatment systems of ade-
quate design. Through discussions with the respective refinery personnel concerning their
operating and control procedures, a selection of treatment systems was made to meet this
criteria. In order to study the maximum number of refineries with the EPA personnel avail-
able and still provide data in a quantifiable manner, five activated sludge systems were
chosen from a list of 14 existing facilities provided by the API.
-------
187
Refinery location selection was made so that distribution would approximate the
geographic distribution of the refining industry as shown in Figure 2. Hence, two were
located in EPA Region VI (southwest); two in Region V (Great Lakes); and one in Region X
(west coast). An important aspect of a national study of this nature was to receive the
samples at the laboratory as rapidly as possible; therefore, consideration was given to the
relation between refinery location and airline schedules. Since not all classes of refineries
use activated sludge as their biological treatment system, it was decided to study three Class
B and two Class C systems. The initial plans included a specific Class D refinery; however,
limited time and funds forced the EPA to eliminate this plant from the five-refinery list.
The parent refining company voluntarily submitted 10-day survey data collected and analyzed
in a manner identical to the five-plant investigation for use as an addendum to the final
report.
Efficiency values were calculated from data obtained at two points in the treatment
process: primary settling effluent and final clarifier effluent. No accounting was made for
once through cooling, intake, and bypass waters, for domestic wastes, nor for the finishing
systems following final clarification of the activated sludge treatment process. This was
because the study objective was activated sludge treatment data; not discharge data.
The systems sampled for activated sludge efficiency employed separation or treatment
ranging from API separators through equalization and intermediate treatment steps to final
detention ponds. Schematic flow diagrams of the treatment systems are shown in Figure 3.
In addition to primary separation, the systems included various combinations of the following
individual processes: activated sludge, trickling filters, air flotation, chemical coagulation,
aeration basins, and clarification. Further, most plant systems included sludge ponds,
detention ponds, and lagoons. This diversity of treatment allowed the evaluation of treat-
ment steps other than the activated sludge process.
Daily samples were composited on a two-hour basis; sample preservation, fixing,
analyses, etc., were conducted as in the RWL procedures. Sample splitting between the
EPA and the refiner was done on an irregular basis, at the discretion of the refiner.
Results The data base for the efficiency survey was obtained by analyzing 70 total samples
from the five refineries. A summary of the range of influent and effluent concentrations
and the median treatment efficiencies is shown in Table 5.
Median effluent loads from the five systems studied and the standard deviation of
loading expressed as percent of the median are shown in Table 6. Although the refinery
daily crude throughput and gallons of water discharged remained nearly constant it was
observed that the pollutant load varied with daily samples, reflecting apparent operational
differences within each refinery from one day to the next.
A summary of the median final clarifier loads for 13 parameters is found in Table 7.
CONCLUSION
This study represents the first industry-wide investigation of its type attempted by
-------
188
the EPA or its predecessor agencies. It resulted in a valid data base of raw waste loading
and contaminant removal efficiencies, and it was planned and conducted with the cooperation
of the petroleum refining industry. There were problems encountered: interlaboratory
analytical variability and interferences in analytical methods, particularly cyanide, sulfide,
and oil and grease.
The survey supported our opinion that several needs deserve further consideration.
Variability and interferences in methods require assessment and resolution through extensive
round-robin testing. Also needed are the identification of major organics treatable by and
refractory to the activated sludge process; the measurement of the fate and effects of these
refractories; the study of metals adsorption on sludges; and additional characterization,
especially of unit process streams such as catalytic cracking and sour waters.
We in EPA at the RSKERL would like to recognize the technical cooperation of the
refining industry in our research endeavors and the many courtesies extended to us. We
look forward to a continued coordinated approach to solving the many problems remaining
in the management of refinery wastewaters.
REFERENCES
(1) "1967 Domestic Refinery Effluent Profile," Committee for Air and Water Conservation,
API, 1968.
(2) "Federal Water Pollution Control Act of 1972," Public Law 92-500, 92nd Congress
S-2770, October 18, 1972.
(3) "Annual Refining Report," The Oil & Gas Journal 70:13, 136, March 27, 1972.
(4) "Petroleum Industry Raw Waste Load Survey," Committee on Environmental Affairs, API,
December, 1972.
(5) "National Petroleum Refining Wastewater Characterization Studies," U.S. EPA Project
No. 12050 CNI (Report in preparation).
(6) "Methods for Chemical Analysis of Water and Wastes, 1971," U.S. EPA Report 16020"-
07/71.
(7) "Laboratory Quality Control Manual," 2nd Ed., 1972, U.S. EPA, Ada, Oklahoma.
-------
189
DISCUSSION
J. J. Chavez: I understand that cationic polyelectrolytes are widely used flotation
schemes in the United States. To your knowledge, have there been studies or surveys
of the toxic effects of using polyelectrolytes?
Fred M. Pfeffer: I am not aware of any such studies.
Robert Farnham: I assume that the data you collected would allow you to calculate vari-
abilities in the activated sludge process.
Fred M. Pfeffer: Our data at each of the five plants which were studied for treatment
efficiency were collected for 14 continuous days and included measurements of flow and
27 parameters. The raw data depicted variations in flow and concentrations as it was
collected. Mathematically, we have expressed the effluent variability (Table 6) in terms
of the median loadings from the final clarifier and the standard deviation of these loadings
expressed as a percentage of the median.
Robert Farnham: Did you take one sample from each plant, or more than one?
Fred M. Pfeffer: Each sample was a two-hour composite. We collected 14 samples from
four of the plants studies, and 13 samples from the remaining plant.
BIOGRAPHIES
Leon H. Myers holds a BS in chemistry/biology from Southwestern
Oklahoma State University and a MS in sanitary science from
Oklahoma University. He is currently Chief, Industrial Section
of the Source Management Branch at the Robert S. Kerr Environ-
mental Research Laboratory, Ada, Oklahoma.
Fred M. Pfeffer holds the BA and MS degrees in chemistry
from the University of Cincinnati. He is currently a Research
Chemist at the EPA's Robert S. Kerr Environmental Research
Laboratory at Ada, Oklahoma.
• |
-------
190
Marvin L. Wood holds a BS in chemical engineering from the
University of Arkansas and an MS in instrumental science from
the Graduate Institute of Technology of the University of
Arkansas at Little Rock. He is currently Deputy Director
and Chief of the Source Management Branch of the EPA's
Robert S. Kerr Environmental Research Laboratory at Ada,
Oklahoma.
-------
191
TABLE 1 "REFINERIES MONITORED BY EPA"
REFINERIES MONITORED
REFINERY
CLASS
A
B
C
D
E
QUESTIONNAIRES
RECEIVED
14
79
31
12
14
NO.
2
8
3
2
2
CRUDE CAPACITY
BPSD
13,200
424,400
388,500
397,300
702,000
150
17
1,925,400
TABLE 2 "ANALYTICAL PARAMETERS"
BOD
COD
TOC
PHENOL
SUSPENDED SOLIDS
DISSOLVED SOLIDS
PH
ALKALINITY
ACIDITY
OIL & GREASE
T. PHOSPHORUS
AMMONIA N
KJELDAHL N
NITRATE N
CHLORIDE
CYANIDE
SULFIDE
T. CHROMIUM
IRON
COPPER
LEAD
ZINC
TABLE 3 "QUALITY CONTROL RESULTS"
FREQUENCY PER 8 SAMPLES
BOD,-
COD3
OIL & GREASE
PHENOL
TSS
SULFIDE
AMMONIA N
KJELDAHL N
CYANIDE
PH
BLANK
1.8
2.9
0.7
1.9
-
0.8
3.3
1.6
1.7
-
STANDARD
0.7
1.7
-
3.0
-
2.3
9.8
1.9
5.2
-
DUPLICATE
4.9
2.4
-
2.8
3.0
2.2
1.6
1.9
1.7
0.7
SPIKE
1.8
-
1.9
3.3
1.2
1.7
2.7
AVG. SPIKE
RECOVERY %
98
98
85
98
100
98
-------
192
TABLE 4 "COMPARISON OF EPA & API DATA1
BOD
COD
OIL & GREASE
PHENOL
TSS
SULFIDE
AMMONIA N
CYANIDE
T. CHROMIUM
IRON
COPPER
LEAD
ZINC
pH (Units)
FLOW (GPM)
MEDIAN DIFFERENCE
(EPA MINUS API)
1.0
0.5
8.4
0.02
-10.0
0.5
0.2
0.02
0.03
0.28
0.02
0.04
0.04
0.0
0.0
"t" STATISTIC
1.9
0.4
3.9
4.0
0.4
5.6
3.8
3.8
4.8
5.1
4.7
7.4
3.4
0.4
1.3
SIGNIFICANT
DIFFERENCE
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
TABLES "TREATMENT EFFICIENCY1
INFLUENT (m
BODC
S
COD
OIL & GREASE
PHENOL
TSS
SULFIDE
pH (units)
AMMONIA N
KJELDAHL N
CYANIDE
T. CHROMIUM
IRON
COPPER
LEAD
ZINC
MAX.
288.0
583.0
89.0
16.0
261.0
43.0
11.5
160.0
194.0
2.5
4.25
44.0
0.88
1.73
1.23
MIN.
27.0
93.0
8.0
0.26
15.0
0.10
6.2
6.0
8.0
0.02
0.03
0.50
0.02
0.00
0.07
a/i)
MEDIAN
85.0
213.0
29.0
3.4
36.0
2.9
8.8
12.0
17.0
0.2
1.43
1.02
0.14
0.10
0.22
EFFLUENT (mg/1)
MAX.
53.0
201.0
37.0
4.2
83.0
1.7
8.7
124.0
134.0
0.03
1.45
0.50
0.11
1.04
1.84
MIN.
5.0
46.0
4.0
0.01
5.0
0.0
5.0
0.05
3.0
0.02
0.02
0.18
0.02
0.10
0.04
MEDIAN
8.0
70oO
11.0
0.01
25.0
0.3
7.3
11.0
15.0
0.1
0.26
0.78
0.05
0.11
0.16
MEDIAN
REMOVAL (%)
89.2
54.6
60.0
99.6
23.8
90.0
—
16.2
17.2
30.9
58.2
37.1
57.7
0.0
22.0
-------
193
TABLE 6 "STANDARD DEVIATIONS OF EFFLUENT LOADINGS'
BOD,
COD°
OIL & GREASE
PHENOL
TSS
SULFIDE
AMMONIA N
KJELDAHL N
CYANIDE
T. CHROME
IRON
COPPER
LEAD
ZINC
MEDIAN LOADINGS
(lb/1000 bbl)
O
23.0
2.9
0.006
7.9
0.10
5.7
7.6
0.04
0.083
0.330
0.015
0.046
0.059
STANDARD DEVIATION
OF LOADINGS
(% of median)
55
21
41
1100
44
97
31
25
24
29
29
26
20
55
TABLE 7 "EFFLUENT LOADING FROM FINAL CLARIFIER"
BOD,
COD
OIL & GREASE
PHENOL
TSS
SULFIDE
AMMONIA N
CYANIDE
T. CHROME
IRON
COPPER
LEAD
ZINC
MEDIAN LOADING (lb/1000 bbl)
9973
5.5
42.0
9.2
0.006
17.0
0.10
4.1
0.06
0.025
0.480
0.032
0.397
0.086
2115
3.1
22.0
2.8
0.006
4.0
0.05
17.0
0.01
0.078
0.148
0.005
0.018
0.027
0288
3.9
49.0
7.1
0.006
14.0
0.32
7.5
_
0.136
0.330
0.015
0.064
0.036
6512
4.0
16.0
2.9
0.003
7.9
0.13
0.1
-
0.085
0.083
0.014
0.030
0.059
6693
1.5
23.0
2.4
0.002
6.4
0.04
5.8
-
0.210
0.650
0.030
0.046
0.089
-------
194
50
40
UJ
O 30
ui
cc
u.
UJ
p 20
<
_i
UJ
tr
10
0
fO
I
1
Q
UJ
-60 -30 0 30 60 90 120 150 180 210 240 270 300 330 360 390 420
BOD5 RAW WASTE LOADING
Fig. I. RELATIVE FREQUENCY DISTRIBUTION OF BOD5 RAW
WASTE LOADING FOR CLASS B REFINERY
10
8
.• ••
F.g. 2. DISTRIBUTION OF THE PETROLEUM REFINING INDUSTRY
-------
REFINERY NO. 9973
Once Through Cooling Water-
Process Water
2400 gpm
REFINERY NO. 2115
•>• API Separator
4hr.
•*• Equal ization •
7.5hr.
•Aeration Basin-
I2hr.
-*-Final Clarifier
2.9 hr.
(S)
•1200 gpm
I
(S)
Once Through Cooling Water
Oily Water
3000 gpm
REFINERY NO. 0288
Clean Water
Oily Water
6600 gpm
REFINERY NO. 6512
Separator
.57 hr.
-^-Primary Clarifier
.65hr.
(S)
Trickling Rlter—r^- Aeration Basin—»-Final Clarifier
.23hr. I 2.7hr. 3.5hr
(S) ^—900 gpm '
(S)
il
^ A? i ocpuruTor
i 2. 5 hr.
0.5hr.
fc
"• MCI UIIUM LJUSIII •
I3,hr- (s)
^ C PRnnnnm.
*• Final Clarifier • ^
2. 5 hr.
All Waste Water
2270 gpm
REFINERY NO. 6693
API Separator
Equalization
48 hr.
Chemical Coagulation
6hr.
(S)
•»- Activated Sludge-
12 hr.
(S)
All Waste Water
620gpm
Separator
2.5 hr.
(S)
*- Aeration Basin
48 hr
100 gpm-
•*• Final Clarifier
2hr.
I
(S)
(S)
( S) Denotes Sample
Point
Fig. 3. REFINERY WASTE WATER SYSTEMS CHOSEN FOR THE ACTIVATED
SLUDGE EFFICIENCY STUDY
•o
Ul
-------
196
-------
SESSION IV
"SLUDGE MANAGEMENT"
Chairman
Milton Beychok
Consulting Engineer, Irvine, California
Speakers
Carl E. Adams, Jr.
"Sludge Handling Methodology For Refinery Sludges"
Jacoby A. Scher
"Processing of Waste Oily Sludges"
C.B. Kincannon
"Oily Waste Disposal by Soil Cultivation"
R.L. Huddleston
"The Disposal of Oily Wastes by Land Farming"
197
-------
198
-------
"SLUDGE HANDLING METHODOLOGY FOR REFINERY SLUDGES"
Carl E. Adams, Jr., President
and
Robert M. Stein,
Director of Operational Services
Associated Water and Air Resources Engineers, Inc.
Nashville, Tennessee
SOURCES OF REFINERY SLUDGES
There are several major sources of sludges generated from normal refinery
operations. These sludges can be classified as storage tank bottoms, bottoms from
API separators, crude desalting sludges, catalytic solids, spent clays, coking fines,
and solids from the utilities and the biological wastewater treatment system. The
basic source of many of the solids is the crude oils which contain materials that
originate either in the well or result during transportation and storage of the crude
oil. The solids generally associated with the crude oil include iron rust, rust
sulfides, clay, sand, salt crystals, wax, and paraffin. These solids will settle out
either in the tank bottoms or gravity API separators. A delineation of types and
characteristics of refinery solids are given in Tables 1 and 2. Estimates of quan-
tities of some of these sludges are given in Table 3. The waste sludge from API
separators plus usual secondary treatment facilities will contain approximately 1,500
to 2,000 Ib. of dry solids per million gallons of effluent treated.
Tank Bottoms. A summary of the tank bottom sludge characteristics is pre-
sented in Table 4. The tank bottoms can be divided into two basic types of solids:
leaded and non-leaded. The leaded solids are generated from those processes which
have been in contact with tetra-ethyl lead, e.g., gasoline and related products or
the slop oil. These solids must be handled carefully because of their contact with
the tetra-ethyl lead (TEL). TEL sludge from storage tanks will contain about 0.12
percent by weight TEL or 120 mg/1. A procedure recommended for handling TEL
sludges is as follow (9):
Manually remove the sludge and transport it to a drying
bed area where it should be spread in a layer of 3-in. thick-
ness or less. The sludge layer should be "weatered" for at
least 4 weeks at ambient temperatures of above 32°F (if the
temperature falls below 32°F for part of the time, the drying
period should be correspondingly increased). Protective cloth-
ing should be used during the transporting and spreading of the
sludge but face masks should not be necessary. At the end of
4 weeks, the sludge should have weathered down to a TEL con-
tent of 20 ppm wt. which Ball states to be a non-hazardous level.
199
-------
200
The non-leaded solids originate from those processes which have not been in
contact with tetra ethyl lead. These sludges result from the crude oil storage, lube
oil, specialty oils and oil intermediates, gas oils, furnace oils, and injection gas oil
or from asphalt. These solids range in concentration from about 10 to 30 percent
solids by weight, averaging about 20 percent. The principal method of removing this
sludge from the tanks is to add water to transfer by pump or vacuum truck. There are
some problems experienced in the removal of the residual waxy crude tank bottoms
because this material is extremely viscous and difficult to handle. It has been found
that the heat addition minimizes pumping transfer problems.
Crude Desalting Sludges. The crude desalting process is another source of solids.
Crude desalting is employed to separate the salts from the oil by the use of an
emulsifier and settling tank. Sludges produced from this process include free and
emulsified oils and suspended solids. These solids are usually discharged directly to
the treatment system in the main waste flow.
Catalytic Solids- The catalytic processes may or may not provide a mojor sludge
problem^ The noble metal catalysts used in reforming and hydroprocessing are usually
extremely valuable and should be completely recovered and reprocessed. The non-
noble catalysts can be recovered and sold to a chemical processor for metal salts
recovery. The remainder of the catalysts are not usually .reprocessed, since they are
inert, but they can be disposed of in a landfill operation. Many times the catalytic
fines which are discharged from fluid beds catalytic crackers can be collected in the
electrostatic precipitator system. These fines might have value as a sludge conditioner.
Spent Clays. Clay or acid treatment can be used to remove color framing
and other undesirable materials from solvent refining and dewaxed lube oil stock.
The acid treatment produces high concentrations of suspended solids, sludges and stable
oil emulsions. The clay treatment results in a sludge which contains clay, fuel oil,
and emulsified oils.
Coking Fines. The coking operation produces a sludge which consists of
finely divided coke particles and waxy materials. These sludges can be land-filled
or may possible have some reclamation value.
Utilities and Wastewater Treatment Sludges* The other major sources of
sludge are from the utilities and wastewater treatment system. The utilities solids are
usually sludges generated from the addition of lime or alum to treat raw water for use
in the plant. Since these solids are inert and have no heat content, they should be
dewatered and landfilled. Sludges generated by the wastewater treatment fcailities
include oily sludges from API separators and air flotation systems and biological
sludges from trickling filter and activated sludge processes. These solids can be
dewatered and incinerated with other combustible solids, or disposed of in combina-
tion with the utilities and other inorganic solids.
The materials from a refinery which are readily combustible (See Table 1) in-
clude the waxy bottoms, oil sludges, coke fines, and waxy tailings. The non-com-
-------
201
bustible sludges are sand, rust, silt, TEL-sludge, salt, spent catalyst, and lime
sludge. The excess biological sludge, although not readily combustible, can be
dewatered to a combustible state.
Both in-plant recovery and reclamation of amenable solids combined with final
sludge handling and ultimate disposal are required in order to implement a successful
sludge management program in a refinery. In-plant management consists of tight
operational control, preventive maintenance for leak control and the location of
separators at critical points to capture materials before they become contaminated and,
thus uneconomical to recover. Sludge handling methods consist of gravity or air
flotation thickening, dewatering by vacuum, pressure or centrifugation, and final
disposal by incineration, landfill, landfarming, and barging to sea.
IN-PLANT CONTROL, RECOVERY, AND RECLAMATION
An important element in a refinery sludge handling program should be a good
in-house control program. This program should include the reduction of oil leakage
by preventive maintenance of pipe lines and equipment and the repairing of leaks as
soon as possible after detection. The refinery should avoid discharges of oils by the
necessary elimination of the water from process units, proper dispostion of solids and
spent chemicals, careful handling of emulsions and awareness by all personnel that
waste treatment begins at the process unit. It has been reported that a refinery can
expect total oil losses of 0.1 to 2.0 percent by volume of throughput; however,
losses in excess of 0.4 percent indicate the need for improved housekeeping.
The control system should be designed to allow for good housekeeping and
should be equipped with adequate sampling and monitoring facilities to indicate
problems immediately. The system design should provide facilities for cleaning up
after charging new catalysts so that materials do not have to be flushed into the
sewers. Adequate installation of both oil-water separators and coke-water separators
at crucial locations is important to a good housekeeping program.
Proper attention should be given to the prevention of the formation of oil
emulsions or, when these emulsions do exist, the system should be capable of isolation
and separate treatment of these materials. API separators for removal of floating oils
should be located as close to the source as possible.
Acidification of caustic wastes with sulfuric acid breaks some emulsions for oil
recovery. Acid wastes may be reused as a source of fuel or to produce by-products,
such as oils, tars, asphalts, resins, fatty acids and chemicals. Many refineries sell
or trade these spent acids and acid sludges to acid manufacturers. Sometimes the
spent acids are sold to other industries, where they may be used directly or fortified
with acid of full strength. Hydrochloric acid alkalation and spent aluminum chloride
sludges are sometimes burned in special incinerators. Additionally, the burning of
oil sludges and salt-bearing sludges in power station furnaces has been practiced.
However, air pollution considerations must be considered in these cases.
A well-planned slop oil and oil solids program is also important to the solids
management program. An emulsion-breaking system can be used to recover oils which
-------
202
can be reused within the refinery- The water from the emulsion breaking steps can be
sent to the wastewater treatment plant, and the oily solids sent to the sludge recovery
system. A summary of the slop-oil emulsion composition is presented in Table 5.
SLUDGE HANDLING AND DISPOSAL
There are basically three types of sludges which must be handled by a refinery
wastewater treatment and sludge handling system. These are: recoverable oils, oily
sludges, and excess biological sludges. Figure 1 indicates the sludges from a typical
refinery waste treatment system. Although many refineries handle the oily sludges in
combinatior. ,
-------
203
tion is highly variable from day to day. Sometimes tank bottoms contain heavy
emulsions which require special treatment, and often the sludge and skimmings from
air flotation systems can present special difficulties in handling. In many cases, the
skimmings from air flotation are sent to a skimmed oil tank from which they can be
taken to landfill or pumped to an oil recovery unit for subsequent reuse of the oils in
the refinery. Chemical treatment is usually required for oil recovery
Thickening. Gravity thickening methods are successfully utilized for oily
sludges. The performance of gravity-type thickeners varies considerably depending on
the nature of the oily sludges being treated. Solids loading rates in the range of 5
to 30 Ib/sq ft-day are reported. The range of thickened solids concentrations is from
3 to 10 percent. Normally, 3 to 7 percent solids can be obtained from a gravity
thickener.
Studies which have been conducted on gravity thickening of oil sludges indicate
that there is no significant correlation of surface solids loading and solids recovery
(10). Surface hydraulic loadings above 400 gal/sq ft-day appeared to cause signifi-
cantly decreased solids recovery; however, there is a wide scatter of data. The
effluent from the gravity thickener would be unsatisfactory for direct release due to
the high oil, solids and COD concentration. Basically, solids and oil recovery are
low from gravity thickeners.
Another alternative for thickening of oily sludges is by use of dissolved air
flotation. A number of studies have been conducted on dissolved air flotation and
the results of some bench-scale evaluations with oily sludges are presented in Table 8.
Flotation thickening of some types of oily sludges was unsuccessful. Tests with heavy
particles indicated some dispersion after setting for 10 minutes. Thus, removal of
heavy particles by partial gravity settling should improve air flotation thickening.
Facilities for chemical addition should be included in any flotation thickener
for use when necessary. The results of tests indicated that dissolved air flotation
would be successful for thickening of froth flotation sludges, but flotation thickening
of other refinery oily sludges is not practical.
Dewatering. Dewatering alternatives for oil sludges include centrifugation,
vacuum filtration, and pressure filtration.
Centrifugation
There have been a number of reports on the use of centrifugal dewatering for
refinery sludges. A schematic of a system that is being utilized for a number of oily
sludges is presented in Figure 3. This consists of a two-stage system whereby the first
stage is used to dewater the thickened sludge and the second stage is utilized to
separate oil from the centrate. In the first stage, a horizontal solid bowl centrifudge
is employed and in the second stage a disc-type cetrifuge is utilized. In this process
the sludge is heated to 180 to 200°F prior to centrifugation.
The results of tests on an oily sludge with a Sharpie's P600 centrifuge are pre-
sented in Figure 4. Note that the total suspended solids recovery is relatively flat
-------
204
as compared to the volatile suspended solids recovery. The volatile suspended solids
are indicative of waxes and heavy tars. Consequently, most of the heavy organic
materials go into the oil phase at high feed rates. Since it is desired to separate
most of the solids from the oil, centrifuge selection should be based on the relatively
flat portions of the recovery curve to maintain a high solids recovery in the cake.
The results of another test for dewatering oily sludges on a solid bowl centrifuge is
presented in Figure 5.
After using the solid bowl centrifuge as a first-stage, a second-stage disc-type
centrifuge should be utilized„ The function of a second-stage nozzle-type centrifuge
is to remove the fine solids from the solids-stabilized emulsions and thus separate the
oil phase from the solids. The results of a test using the disc-type centrifuge are
shown in Figure 6 and indicate that successful operation was obtained for all feed
rates studied.
The performance of a full-scale centrifuge system treating an oily sludge is
presented in Table 9. An internal inspection of the scroll centrifuge following 9
months of operation revealed that the hard surface coating on the internal flights had
worn away exposing the stainless steel. Layered deposits 0.5 in. thick of carbon
fines and grit were found inside the scroll (8).
The maximum period that the disc centrifuge has been on-stream without plug-
ging has been 25 days. The downtime required for removal of the disc stack and
cleaning can extend to three days.
The results of centrifuge testing and experience can be summarized as follows:
1 . A vertical solid bowl centrifuge is not recommended for dewatering most
oily wastes.
2. A horizontal solid bowl centrifuge followed by a high-speed nozzle or disc
centrifuge is best suited for dewatering mixtures of contaminated API
bottoms, sludge decant pit material and truck bottoms.
3. A super decanter centrifuge dewatering oily sludges is anticipated to recover
75 to 90 percent of the solids in the cake when charged with heated oily
sludges. The oily phase will consist of 1 to 5 percent solids and the cake
will contain approximately 50 to 60 percent solids.
4. A nozzle ejector centrifuge separates 95 to 98 percent of the feed oil in
the oily phase and 2 to 5 percent in the ringdam nozzle and nozzle water.
Thirty to 50 percent of the nozzle ejector feed solids will be removed with
the oily phase with the remainder being in the two water phases.
Vacuum Filtration
A significant number of refineries use vacuum filtration for dewatering oily
sludges. If properly implemented, vacuum filtration of oil sludges renders the solids
suitable for direct landfill or subsequent incineration. In order to accommodate oily
solids, a precoat vacuum filter should be used and the incoming solids should be
heated to temperatures greater than 170 F. Many variables affect the performance of
filtration processes; however, the relative filterability of various sludges and estimates
-------
205
of vacuum filter loading rates can be made using standard Buchner funnel and filter
leaf tests.
The results of some filter resistance tests (7) using oily sludges are presented in
Table 10. These tests indicate that oily sludges quickly blind the filter cloth when a
precoat was not used. These data indicate that heating of oily sludges to 165°F
decreased specific resistance and improved the characteristics of the cake. The effect
of chemical addition on filterability is also indicated in Table 10. The addition of
lime and ferric chloride did not improve the filterability of oil sludges.
A vacuum filtration filter leaf test was conducted on the oily sludges in the same
study. The design filter loading rate for precoat vacuum filters for various types of
oil sludges is presented in Table 11. Using an operating temperature of 160 to 180°F
improves the vacuum filter performance in terms of oil recovery, filtration loading,
and cake solids concentration. Several noteworthy observations can be made con-
cerning the effect of clay addition on vacuum filtration performance. First, the oil
recovery decreases with increased clay addition. Secondly, the effective filtration
rate decreases with clay addition. However, the effective filtration rate is based on
solids, excluding the clay, therefore, the total solids filtration rate may not be
decreasing as rapidly as it might seem. The major conclusions regarding vacuum fil-
tration of oil sludges are:
1. Increased feed temperature greatly improves vacuum solids performance.
2. Addition of spent clay decreases oil recovery and solids filtration rate
on a clay-free basis.
3. The measured filtration loadings of 0.8 to 3.0 Ib/sq ft-hr were required
for oil recovery.
Pressure Filtration
A fixed plate, high pressure filter press may be used to dewater many types of
oily sludges. Sludge is applied to a cloth covered filter plate at high pressure and
water and oil are passed through the cloth media while solids are retained. Para-
meters which are important in evaluating the process are feed temperature, lime
addition, spent clay addition, cycle time, cake characteristics, and filtrate character-
istics.
The results of filter press tests on an oily sludge are presented in Table 12.
The feed to the pressure filter consisted of a base mixture of refinery oily sludges
which contained API separator bottoms, sludge decant pit material, and tank bottoms.
These tests indicated that heating the feed was required to obtain satisfactory filtra-
tion. Effective breaking of solids-stablilized emulsions is obtained. Cake solids in
excess of 50 percent can be obtained on pressure filtration with oil concentrations in
the range of 5 to 20 percent.
The filter press cake solids concentration correlated with average filtrate temp-
erature as shown in Figure 7. It is apparent that filtration temperature has a dramatic
effect on cake solids and cake oil concentrations at constant cycle times. The
addition of lime improved filtration somewhat, although the effect is not conclusive;
-------
206
however, experience at other refineries indicates that a definite increase in filtration
efficiency can be obtained by adding lime. A significant decrease in cake oil content
can be obtained by washing the cake with hot water. The filtrate is usually readily
separable into an oil and water phase. Test results indicate that the oil will probably
require processing through the high chloride slop system. The water phase has not been
found to create any problems in the wastewater system.
Disposal. The ultimate disposal of oily sludges can be by barging to sea, land-
fill, land farming, and incineration with landfill of the ash. However, sea disposal of
oily materials is viewed as a short-term alternative and will eventually be eliminated
as an option. The disposal of the oily sludges on soil is acceptable if it can be
shown that such disposal will not contaminate groundwaters or contribute to contami-
nated storm runoff, and will not create a potential seepage problem. A proper land
farming operation using soil bacteria for degradation of oils would satisfy the above
requirements. The utilization of a lined landfill with leachate treatment would also
meet these requirements.
Land farming of oily sludges has been successfully practiced by refineries where
sufficient land area is available for proper decomposition of the oil-containing solids.
Land farming involves spreading the sludge in 4 to 6-in. layers, allowing the sludge to
dry about one week, adding nutrients, and then discing the sludge into the soil. A
Texas refinery land farms crude oil tank bottoms containing high molecular weight, oil-
containing olefinic or aromatic components and waxy raffinate containing high para-
finic components. Decomposition rates were found to be approximately 0.5 Ib/mo-cu
ft without nitrogen and phosphorus addition and 1 .0 Ib/mo-cu ft with nutrient
addition. Disposal costs were estimated at $7 per barrel of oil including the cost of
nutriel addition. The species of microorganisms which were found to remove the oils
were pseudomonas, flavobacterium, nocardia, corynebacterium, and artherobacterium.
Incineration of oily sludges and landfilling of the ash should provide an
acceptable means of final disposal. Important factors in disposal by incineration
sould include the following:
1. Selection and cost of the incineration equipment.
2. Operating costs.
3. Potential air problems by particulates, sulfur dioxide and other pollutants.
Table 13 presents the results of an economic evaluation of flow and optimum
solids concentration to incineration. The fuel costs associated with evaporation of
excess water in an incineration system are presented in Figure 8. Oil and volatile
matter in the contaminated sludges can be destroyed by incineration. The type of
incinerator which is most applicable to a particular waste depends on the nature of
the waste. Sludges and solids of high solids contents are conventionally handled by
multiple-hearth incinerators.
A fluid bed incinerator operates with a bed of ash and sand particles fluidized
by forced air feed. The fluidized bed incinerator is best suited for feeds that are
partially liquid so that the incinerator can be fed by pumps and screw conveyers.
-------
207
While feeding dry solids to a fluid bed incinerator is possible, the multi-hearth
incinerator will be more economical if most of the feed is in the form of cake of non-
pumpable solids.
Rabb indicates "suitable" incinerators as follows:
2-10wt.% solids feed - fluid sand bed
5-70 wt. % solids feed - rotary kilm
40-70 wt. % solids feed - multiple-hearth
Biological Sludges. Basically, the sludge handling methodology employed
for processing biological sludges is similar to that previously described for oily
sludges. However, due to the putrescible potential of biological sludge, stabilization
is required prior to disposal by sea or land. The basic process alternatives are
depicted in Figure 9 with the most feasible alternatives having been summarized in
Figure 1 . Consequently, the process sequency for biological sludge handling consists
of stabilization, thickening, dewatering, and final disposal. Oftentimes, the
thickening step will precede stabilization in order to reduce the stabilization costs
which are highly dependent on hydraulic flow. Detailed design procedures and
examples for these processes are given elsewhere (1) and will not be described herein.
Biological Sludge Stabilization. Stabilization of excess biological solids is
required because of the degradable nature of the microbial sludge. There are three
basic methods of sludge stabilization which are technically available: aerobic
digestion, anaerobic digestion, and wet air oxidation. Because of the complexity of
anaerobic digestion and the problems experienced in operating the process, this
method is not being utilized in new refinery sludge stabilization systems. However,
there are several older treatment systems which have used anaerobic digestion. There
have been considerable operational and materials problems experienced with wet air
oxidation and, to date, this method has not received acceptance in the refinery
industry.
The most common method of stabilizing water biological sludge in the refining
industry is by utilization of aerobic digestion. Aerobic digestion is used chiefly to
stabilize the sludge and render it suitable for land disposal. Additionally, the
overall sludge quantity may be somewhat reduced and the dewatering characteristics
improved slightly. Aerobic digestion is an extension of aerobic biological oxidation
in the absence of an external food supply. Thus, the microbes digest each other due
to endogenous respiration and autooxidation. A simplified summary of the aerobic
conversion of organic materials into cellular materials and the subsequent breakdown
into by-products and residue is shown in Figure 10.
The results of aerobic digestion on various biological sludges are shown in
Figure ll. Approximately 10 to 15 days detention are required on an annual average
to achieve 50 to 60 percent reduction of the volatile content of the sludge. This
reduction represents about 80 to 90 percent reduction of the degradable content and
is generally suitable for land application. Tests on refinery sludges (3) have
indicated that about 0.5 to 1 .2 Ib of oxygen are required per Ib of VSS destroyed
per day.
-------
208
Thickening of Biological Sludges. There appear to be three alternatives by
which refineries can concentrate biological sludges. These are flotation thickening,
gravity thickening, and centrifugal thickening. The primary objective of a thickner
is to reduce sludge volume by removing the bulk water and concentrating the sludge,
thereby increasing the system effectiveness and lowering the cost of subsequent dewater-
ing processes
Gravity Thickening
Gravity thickeners utilize gravity sedimentation and compression for sludge con-
centration. The key design parameters are hydraulic and solids loading rate. If the
surface loading rate is excessive, poor solids recovery will result. If the solids load-
ing (Ib solids applied/sq ft-day) is too great, then there will also be poor solids
recovery and low underflow solids concentration. Solids loading rates for gravity
thickening of waste refinery sludges range from 2 to 15 Ib/sq ft-day. Biological
refinery sludges can be concentrated by gravity to a concentration of 2 to 4 percent
solids by weight.
Gravity thickeners do not require much operator attention and will perform fairly
consistently provided the influent hydraulic flow and solids loading do not vary sub-
stantially. In extremely warm climates, gravity thickeners may generate obnoxious
biological odors if they precede the stabilization process.
Centrifugation
Centrifugation has been utilized sparingly for thickening waste biological sludges
in the refinery industry. There is little information available on the application of
disc centrifuges for thickening of refinery biological sludges. Basket centrifuges are
capable of thickening waste activated sludge to levels of 5 to 6 percent concentration'
with 80 to 95 percent solids capture. One refinery has reported using basket centri-
fuge for thickening to a concentration of 8 percent. However, excessive maintenance
was experienced because of vibrational problems.
Air Flotation
A third method of thickening is by air flotation whereby air bubbles are released
from solution and attach themselves to the sludge floes. The air-solids mixture than
rises to the surface of the basin where it concentrates and is removed. The primary
variables with regard to flotation thickening are recycle rate, feed solids concentration,
air-to-solids ratio, and solids and hydraulic rates. Pressures between 50 and 60 psi
are commonly employed for flotation thickening. Recycle rates on the order of 100 to
500 percent, solids loadings of 2 Ib/sq ft-hr, and hydraulic loading rates (including
recycle) in the range of 1 to 4 gpm/sq ft are common of flotation thickeners.
Results of tests on refinery sludges indicate that these sludges are quite amenable
to flotation thickening (3). These results are summarized in Table 14 and Figure 11A.
These data indicate that the optimum pressure was about 50 to 55 psig and that polymer
polymer addition was not justified. The correlation of these data in Figure 12 shows
the optimum air-to-sol ids ratio to be about 0.01 Ib/lb to obtain a float solids of 3.5
percent.
-------
209
Dewatering of Biological Sludges.
Vacuum Filtration
Vacuum filtration is one of the most common methods used to dewater waste-
water sludges. Water is removed from the sludge under an applied vacuum through a
porous media which retains the solids. Chemicals are usually required in order to
decrease the specific resistance of the sludge, thereby allowing higher sludge loading
rates (Ib/sq ft-hr) and, thus, smaller, more economical filters. The results of several
tests on refinery wastewaters indicate that ferric chloride or a combination of lime and
ferric chloride usually provides the optimum coagulant from an economical and per-
formance standpoint. The optimum ferric chloride dosage usually ranges from 200 to
400 Ib/ton of dry solids. Tests conducted on waste activated sludge indicated that
there is not significant advantage of a precoat filter based on overall economics.
Tests indicate that vacuum filtration of refinery biological sludges is usually able to
obtain solids concentrations in the range of 10 to 16 percent at filter loading rates
ranging from 1 to 5 Ib/sq ft-hr.
The results of filter leaf tests on both a raw and digested refinery waste
activated sludge are presented in Table 15. This table indicates that the digested
sludge presented better dewatering characteristics than the undigested sludge.
Cenrrifugation
Basket centrifuges have been found to be the best centrifugal method of con-
centfatu^j waste biological solids. The tendency for solid particles to separate from
the water portion of the slurry due to a slight difference in density is accelerated by
the application of centrifugal force. In fact, by spinning the slurry at appoximately
1,500 rpm, a centrifugal force of over 1,000 times that of natural gravity is
extended and the particulate matter is concentrated on the external wall of the
revolving basket or bowl. A schematic of the internal components of the basket
centrifuge is presented in Figure 13.
There are a number of applications of utilizing basket centrifuges in the refining
industry. Normally, the centrifuges provide a cake solids ranging from 8 to 18 per-
cent concentration with biological sludges. A correlation of the cake concentration
and solids recovery and cake solids and distance from the wall is presented in Figure
14. It should be indicated that there have been some problems in the utilization of
the basket centrifuge due to mechanical vibrations of the centrifuge.
Pressure Filtration
Pressure filtration is another alternative for the dewatering of the biological
solids. This process usually can obtain a solids concentration approaching 50 percent
with biological sludges, thus producing the driest cake for disposal. The pressure
levels for the pressure filtration systems range from 50 to 225 psig. Normally, it has
been found that the pressure filtration system requires higher chemical dosages than
for the vacuum filter and centrifugation system; however, these produce a higher
solids concentration, In order to obtain an acceptable cake, the manufacturers
recommend a pressure specific resistance of less than 3 x 101
cm
-------
210
Heaf Content of Sludge. The heat content of various samples of dewatered
secondary refinery sludge has been determined in a bomb calorimeter (3). The average
fuel value for the undigested secondary sludge was approximately 8,000 BTU's per Ib
of dry solids. This value is within the expected range of secondary waste activated
sludge with 70 percent voatile content. Approximately 6,000 BTU's per Ib of dry
solids were found after aerobic digestion which indicates that part of the vlatile con-
tent matter is destroyed by aerobic digestion. It is obvious that stabilization would
not be desirable if incineration were the means of final disposal.
Ultimate Disposal. The alternatives for ultimate disposal of the biological
sludge are similar to those for the waste oily sludge with the expection of land
farming. In many cases, a thick slurry of activated sludge can be placed in 4 to 8
in. layers and utilized as a limited nutrient and soil conditioner. In this case, additional
water is removed by natural evaproation and infiltration into the underlying soil.
Following initial periods of drying, the sludge layer can be disced to encourage the
activity of aerobic soil bacteria. The soil bacteria working in conjunction with
plant growth will further stabilize the waste solids and transform them into a hemocyte
material suitable for agricultural use as a soil conditioner.
SUMMARY
A design summary table for the thickening and dewatering alternatives of oily sludges
is presented in Table 16. A summary on biological sludges is presented in Table 17.
REFERENCES
(1) Adams, Carl E., Jr. and W. Wesley Eckenfelder, Jr. Process Design Techniques
for Industrial Waste Treatment. Nashville, Tenn.:Enviro Press, 1974.
(2) Adams, Carl E., Jr., W. Wesley Eckenfelder, Jr., and Gary M. Davis.
'Characterization and Biological Treatability Investigations of a Refinery Water
Discharge." Unpublished report for Ashland Oil, Incorporated, March 1972.
(3) Adams, Carl E.,Jr., W. Wesley Eckenfelder, Jr., and Robert C. Lasater.
"Fina I Process Design for Treatment of a Combined Industrial Wastewater."
Unpublished report for Lower Neches Valley Authority, December 1972.
(4) Beychok, M.R. Aqueous Wastes from Petroleum and Petrochemical Plants.
London: John Wiley & Sons, 1967.
(')) Cross, F.L. and J.R.Lawson. "A New Petroleum Refinery." American
Institute of Chemical Engineering Symposium Series, Vol. 70, No. 136, p.812.
(6) Engineering-Science, Inc. "Wastewater Treatment Facil ities and Ballst Handling
Systems." Unpublished report for the Suntide Refining Company, December 1972.
-------
211
(7) Ford, Davis, L. "A Preliminary Engineering Study for Wasrewafer Treatment and
Pollution Abatement." Unpublished report- for The Sun Oil Company, Sept. 1971.
(8) McCrodden, B.A. "Treatment of Refinery Wastewater Using FilfraHon and
Carbon Adsorption." Proceedings of the 29th Industrial Waste Conference,
Purdue Univeristy, 1974.
(9) Peoples, R.F., P. Krishnan,, and R.N. Simonsen. "Nonbiological Treatment
of Refinery Wastewater. " Water Pollution Control Federation Journal, Vol.44,
No. 11 (November 1972), pp. 2120-2128.
(10) Sebesta, Ed C. and David L. Ford. "A Preliminary Engineering Study for Solid
Waste Disposal and Pollution Abatement." Unpublished report for The Sun Oil
Company, February 1972.
(11) "Session on Sludges and Spent Clays." 18th Mid-Year Meeting, American
Petroleum Institute, 1964.
(12) "Treatment of Recovered Oil Emulsions." Disposal of Refinery Wastes. New
York: American Petroleum Institute, 1969.
DISCUSSION
John Smith : You spoke of the rules-of-thumb reference to air oxygen input rates
and sludge concentrations on bio-sludges. Does this also refer to those sludges
generating on the rotating bio-disk units or have you found a difference?
Carl Adams: We have not conducted much sludge handling on bio-disk sludge,
although we have performed about six pilot studies with the bio-disk. One was
on a very large scale and we are presently evaluating the bio-disk at Cranston, Rhode
Island on a fairly large scale. We are not far enough along into the study to
comment on sludge handling. I would assume that the sludges dewater as wel I or
better than conventional activated sludge. I say this because the material sluffs
off in chunks. Most of the wastes we have evaluated have been high-strength,
difficult-to-settle wastes using the activated sludge process. Consequently, we
examined the bio-disk in order to establish a fixed medium. Thus, these
wastes are bad examples for me to quote as representative of bio-disk performance.
However, with the pulp and paper industry the sludge from a bio-disk settled
very well. We did pilot a large 3-meter unit, with a pulp and paper waste,
and observed very good settling.
John Smith: The work we have done to-date indicates the settling is quite
good but the oxygen input required for digestion is considerably higher?
W.C. McCarthy: From a leaching point of view, what criteria are used to decide
-------
212
if a treated oily sludge would be suitable for landfill?
Carl Adams: I can gef ouf of this one easily by referring this question to the next
apeaker. Genrally, you are required to contruct a monitoring well if you can get
permission from the authorities. The easy way for the regulatory authorities is to
require you to line the area and treat the runoff. If you do not want to do that, you
generally have to convince the authorities that you do not have heavy metals which will
leach out. You can genrally do this; however, ash from incinerators, boilers and so on
contain quite a bit of heavy metals from the coal that was burned. We usually argue that
the rain water won't leach these metals because pH is sufficiently high that you will not
get leaching . The only way you can resolve it if there is an argument is generally
to construct monitoring wells. If a leaching problem then is observed to develop, you
may have to line the pond and treat the storm water runoff from the landfill. We argue
against it generally and hope we are not filling the pond with anything that would cause
leaching and problems.
-------
213
BIOGRAPHY
Carl E. Adams, Jr. is President of
Associated Water and Air Resources
Engineers, Inc. (AWARE) located in
Nashville, Tennessee, and President
of AWARE Engineering in Houston,
Texas, and has consulted to over
200 United States and foreign indus-
tries. He holds the following degrees:
B.E., Civil Engineering, Vanderbilt
University; M.S. , Environmental
Engineering, University of Texas.
He is a professional engineer, re-
gistered in 18 states. Dr. Adams
serves as an Adjunct Professor in
the Department of Environmental
Water and Resources Engineering
of Vanderbilt University. He holds
membership in more than six pro -
fessional societies, has authored and
presented more than 50 technical
papers, and edited and co-authored
three books dealing with industrial
wastewater treatment technology.
Robert M. Stein is Director of
Operational Services for Associated
Water and Air Resources Engineers,
Inc. (AWARE) of Nashville, Ten-
nessee. He holds the following
degrees: B.A., Applied Science,
Memphis State University, Memphis,
Tennessee; and M.S. , Environmental
Engineering, Vanderbilt University
Nashville, Tennessee. He is a pro-
fessional engineer, registered in the
State of Tennessee. Mr. Stein was
associated for several years with the
State of New York in the Department
of Transportation.
-------
TABLE 1 TYPE OF REFINERY SOLID WASTE
214
Source
Crude Oil Storage
Product Storage
Crude Processing
Catalytic Cracking
Catalytic Reforming
Alkylation
Coking
Feedwater Treatment
Wastewater Pollution Control
API Separators
Flotation
Separators
Secondary
Clarifiers
Sludge Thickeners
Combustible Non-Combustible
Waxy Bottoms Sand, Rust, Silt
TEL Sludge
Sand, Rust, Silt,
Salt
Catalyst
Catalyst
Sludge Corrosion Products
Coke, Waxy
Ta i 1 i ngs
Lime Sludge
Slop Oils Sand, Silt
Float Solids
Lime Sludge
TABLE 2
Biodegradable
Bio Sludge
CHARACTERISTICS OF REFINERY SOLIDS WASTES (REF.5)
Waste Type
API Separator Sludge
Tank Bottoms
Chemica1 Treatment Sludge
Air Flotation Froth
Precpat Vacuum Filter
Sludges
Biological Treatment Sludges
Raw
Mechanical 'y Thickened
Centrifuged
Vacuum Fi'tered
Screw Pressed
Waf<=r Treatment Sludge
Typical Composition, Percent
Oil or Volatile Inert S
Hydrocarbon Water So'ids So'ids
15 66 6 13
48 40 4 8
5 90 5
22 75 3
22 29 49
0 98 1.5 0.5
0 94 4 2
0 85 10 5
0 75 15 10
0 40 40
00 95 5
Characteristics
F'uid slurry of oil,
water and sand
Oil -water mixture
Slightly viscous Puid.
Thick,oily fluid
Temperatures
Water Consistency
Thick, but pumpab'e
Viscous-peanut butter
consistency
Wet crumb'y solid.
Intact, solid cake.
Pumpable Puid, some-
times ge'atinous
-------
Waste Types
TABLE 3
ESTIMATES OF REFINERY SOLID WASTE QUANITIES
Unit Loads
API Separator Sludge
Chemical Treatment
(API Separator Effluent)
Biological Sludges
Water Treatment Sludge
A. Lime Soda Ash
B. Ion Exchange
Office Wastes
Cafeteria
200 mg/1 Suspended Solids
50 mg/1 Suspended Solids
Removed Only
0.7 Ib Dry Solids per Ib
BOD Removed
2 parts Dry Sludge per
1 part Hardness Removed
0.4 Ib Salt per 1,000 Grains
Hardness
1 .0 cu yd per Employee
per month
0.6 ''3 per Meal
Water
Oil
Volatile Solids
Ash
Total
TABLE 4
SUMMARY OF TANK BOTTOMS SLUDGE CHARACTERISTICS (REF.ll)
ANALYSIS (PERCENT BY WEIGHT)
Average Maximum
Minimum
39.4
47.9
4.4
8.3
100.0
98.0
95.0
13.0
55.3
5.0
1.3
0.0
0.0
Oil
Water
Sediment
TABLE 5
COMPOSITION OF SLOP-OIL EMULSION (REF- 11)
Maximum
Percent by Weight
Minimum
80.9
81.0
8.0
13.2
18.0
11.1
Average
40.2
54.8
5.0
-------
216
TABLE 6
OPERATING CONDITIONS FOR PRECOAT FILTERS
(Treating Slop-Oil Emulsions, Separator Bottom Sediment,
and Acid Oil )
Precoat Medium = Medium-grade diatomaceous earth
Temperature, F 180
Vacuum, in. of mercury 18 to 20
Knife-cut depth, in. per revolution 0.0040
Drum speed, min per revolution 3 to 4
TABLE 7
COMPOSITION OF API SEPARATOR BOTTOMS (REF.ll)
Analysis as Taken from the Separator (Percent by Weight)
Average Maximum
Water 82.4 98.3
Oil 10.9 20.7
Volatile Solids 1.4 2.3
Ash 5.3 15.8
Total 100.0
TABLE 8
BENCH SCALE FLOTATION THICKENING (REF. 10)
Untreated Water Float Solids Sludge
Sludge TSS Phase TSS Volume Solids Volume
Sludge Mixture (mg/1 ) (mg/1) (ml) (%) (ml)
WEMCO Float Solids 730 73 15 4.4 0
WEMCO Float Solids 1,136 96 30 3.5 0
WEMCO Float Solids 420 66 10 5.5 Trace
WEMCO Float Solids 2.,260 830 40 5.2 Trace
Contaminated API Btms = 10% — — 100 — 100
Tank Bottoms = 20%
Decant Pit Sludge = 10%
Water = 60%
Contaminated API Btms = 20% — — 100 -- Trace
Water = 80%
Tank Bottoms =50%
Water =50% ~~ "" "" '7U
Minimum
61.2
1.1
0.5
0.4
Reco-
very
(%) Comments
90 Water Phase
COD = 270 mg/1
BOD, =140 mg/1
o
92
84 Some of the float
was dispersed
after 10 min.
93 Some of the float
was dispersed
after: 10 min.
Solids dispersed
rapidly.
Solids dispersed
during 10 min.
Most of solids sank.
Float dispersed
rapidly
= " -r,r,H;t-;0ns: 1 ) 100 percent Recycled Based on Sludge Volume. 2.) 50 psig Saturation Pressure
-------
TABLE 9
FULL-SCALE CENTRIFUGE PERFORMANCE
217
Percent
Solids
Scroll Centrifuge Feed 1 .6
Scroll Centrifuge Centrote 0.7
Scroll Centrifuge Sludge 59
Disc Centrifuge Water Discharge 0.01
Disc Centrifuge Solids Discharge 0.1
Scroll Centrifuge = 30 gpm
Disc Centrifuge = 30 gpm
TABLE 10
FILTERABILITY OF OILY SLUDG ES (REF.7)
Special Test Specific Resistance
Test Sample Conditions sec /gm x 10
Composite of Plant Oily Sludges- D , .... - „
T. . . . ' Precoated Filter 0.35
Thickened
Composite of Plant Oily Sludges- Precoated Filter- _
Percent
Oil
6.9
5.9
2.5
0.5
0.1
Comments
Recovered
mg/1 of oi
Recovered
36,900
1 from sludg
40,000
Thickened
Thickened Oily Sludge
Composite of Plant Oily
Thickened
Composite of Plant Oily
Thickened
Sludges-
Sludges-
Composite of Plant Oily Sludges
Thickened
Cloth Media
Precoated Filter-
Heated to 165° F
Precoated Filter-0.67 mg
Spent Water Treating Lime/
1 .0 mg of oily solids heated
to 165 F
Precoated Filter-2,600 mg/1
of FeCI, Heated to 165°F
•J
0.19
0.48
mg/1 of oil from sludges
Blinded Filter Badly
Recovered 74,000 mg/1
of oil from sludge
Recovered 85,000 mg/1
of oil from sludge
Recovered 31,000 mg/1
of oil from sludge
TABLE 11
AVERAGES OF VACUUM FILTER RATES AND PERCENT OF
FILTER TIME REQUIRED
Service
Slop-oil emulsion
Separator sediment
Flocculation sludge
Acid Oil
Pre coating
Downtime
Average Charge
to Filters (Barrels/day)
427
377
627
129
—
—
Average Filter Rate
(gal/hr-sq ft)
1.7
2.8
8.6
2.4
--
—
Filter Time
Required (%)
37.8
20.1
10.7
7.8
8.5
15.1
-------
TABLE 12
SUMMARY OF RESULTS FROM FILTER PRESS PILOT UNIT (REF. 10)
a
Run3
No. Feed Mixture
6
7
8
9
10
11
12
13
14
15
16
Base
Base
Base
Base
Base
Base
Base
Base
Mixture0
Mixture5
Mixture
Mixtureb
Mixture5
Mixture5
Mixture5
Mixture5
Non-Contaminated
API
Separator Sludge
Thickened WEMCO - 78%
Base
Base
Mixture5 » 22%
Mixture0
Admix
Added
Clay &
0.6 Ib/lb D.S.
Clay @
0.6 Ib/lb D.S.
Clay @
0.6 Ib/lb D.S.
Clay @
0.6 Ib/lb D.S.
Clay @
0.6 Ib/lb D.S.
Clay @
0.56 Ib/lb D.S.
Clay @
0.56 Ib/lb D.S.
Clay
-------
219
TABLE 13
INCINERATOR CONSIDERATIONS OF FLOW AND SOLIDS CONCENTRATION (REF. 4)
Effluent flow (gpm)
2 wt. % lean sludge (gpm)
Optimum Sludge Concentration
to incinerator (wt. %)
50
100
200
500
1,500
1
2
4
10
30
3
7
17
50
b
7
-------
220
TABLE 15
FILTER LEAF TESTS RESULTS (REF.3)
Time
Cake
Sludge Run No.
Raw° lb
2C
3C
4C
5C
6C
7C
8C
9°
ioc
Digesfed 1
b
2
3b
b
4
b
5
b
6
b
7
b
8
9b
Feed Solids
(%)
1.04
1.04
1.48
1.48
1.48
1.48
1.48
1.48
2.29
2.44
1.73
1.73
1.73
1.73
1.73
1.73
2.74
3.19
2.18
Vacuum
(in. Hg)
15
15
15
10
18
15
10
15
15
15
15
18
10
15
15
15
15
15
15
Form
(min)
2.00
2.00
2.00
2.00
2.00
1.00
1.50
3.00
2.00
2.00
2.00
2.00
2.00
1.00
3.00
1.50
2.00
2.00
2.00
Dry
(min)
3.00
3.00
3.00
3.00
3.00
1.50
2.25
4.50
3.00
3,00
3.00
3.00
3,00
1.50
4.50
2.25
3.00
3.00
3.00
Solids Thickness
(%) ( in.)
8,9
9.3
9.2
8.6
7.9
7.2
7.5
7.9
7.1
7.0
13.1
13.6
12.1
11.5
12.3
10.8
11.0
11.0
15.9
1/64
1/8
1/8
1/8
3/16
3/16
3/16
1/4
5/16
3/8
3/16
1/4
5/16
1/4
7/16
3/8
7/16
7/16
3/16
Loading
(Ib/sq fr-hr)
0.16
0.73
0.96
0.92
0.99
2.31
1.54
0.88
2.05
1.98
2.24
2.57
3.13
5.08
3.00
4.62
4.63
4.65
1.95
Chemical dosage = 20% FeCl • 6H,O +5% CaO.
b
Media No. POPR-873, polypropylene from Eimco Envirorech
°Media No. POPR-929, polypropylene from Eimco Envirorech
d.
Chemical Dosage - 15% FeCl,
6H2O +5% CaO
-------
221
TABLE 16
SUMMARY FOR THE THICKENING AND DEWATERING ALTERNATIVE
OF OILY SLUDGES
Type of
Sludge
Oily
Sludge
Process
Thickening
Average
Performance
Design
Parameters
Comments
Gravity
Flotation
Dewatering
Centrifugal-ion
3-7 percent solids
50% recovery
84-93% solids recovery
3.5-5.5% float solids
5-30 Ib/sq ft-day
Untreated sludge 420-2,200
mg/l TSS, 100% recycle
50 psig saturation pressure
75-90% solids recovery when Flow rate 50-350 gal/hr
charged with heated oil sludges
Oily phase wil I consist of
1-5% solids and cake will
contain 50-60% solids.
Nozzle ejector centrifuge
separates 95-98% of oil
in oily phase and 2-5% in the
nozzle and nozzle water.
30-50% of nozzle ejector
feed solids will be removed
with oily phase.
Low solids and oil
recovery
Sucessfu! for froth
flotation sludges
but impractical for
other oily sludges
A vertical solid
bowl type not re-
commended for de-
watering most oily
sludges.
Oily
Sludge
(Cont'd)
Vacuum
Filtration
Pressure
Avg . Filter Rates gal/hr-
ft slop-oil emulsion-1 .7
separator sediment-2.8
flocculation sludge-8.6
acid oil-2 .4
Filter Time Required,per
cent,slop-oil emulsion-
37.8 separator sediment-
20.1, flocculation sludge-
10.7, acidoil-7.8
precoating-8.5, downtime-
15.1
5-20% oil, 40-70% solids 2-hr cycle time, feed con-
cake fents 12-38.5% TSS,6-23%
oil. Temperature of feed
58-180° F.
To dewateroily sludges,
a precoat vacuum filter
should be used and solids
should be heated above
170°F. Sol ids should be
suitable for landfill
following vacuum filtra-
tion. Increased tempera-
ture for feed improves
psrformance. Addition of
Spent clays decreases oil
recovery and sol ids F. H .
rate
Heating of feed required
for satisfactory filtration.
Lime or spent caustic added.
-------
222
TABLE 17
SUMMARY OF BIOLOGICAL SLUDGES
Type of
Sludge Process
Biological
Sludges
Average
Performance
Design
Parameter
Comments
Similar to oily sludges
but should be stabilized
due to putrescible nature
Biological
Sludges
Thickening
Gravity
Centrifugation
Flotation
Dewatering
Vacuum
Fi I (ration
Centrifugation
Pressure
Filtration
2-4% solids con-
centration by
weight
5-6% with basket cen-
trifuges, 80-95% solids
capture
Cake solids 2.6-4.0%
Solids concentration
10-16%
8-18% cake solids
Solids recovery 20-90%
50% solids
Loading rate 2-15
Ib/ft day
Feed rate 6-200 gpm
50-60psi, 100-500%
recycle. Solids loading
216 sq ft-ht hydraulic
1 -4 gpm/ft
Loading rate 1-5 Ib/sq
ft-hr
Feed rate 4-90 gpm
Pressure levels 50-255
If surface loading rate
is excessive poor
solids recovery results
There is little informa-
tion avaiable
Sludges quite amenable to
flotation thickening
Chemical usually re-
quired to decrease
specific resistance.
FeCLor lime and FeCL
usually. Optimum FeCL
200-4001 b/ton of solids
Basket type found to be
best. There have been
problems due to mechanical
vibrations in the centrifuge
Normal ly requires
higher chemical dos-
ages than for vacuum
filters or centrifuges.
-------
Refinery
Waste wafer
EXCESS
BIOLOGICAL SLUDGES
FIG. I. SLUDGE HANDLING AND DISPOSAL OPTIONS FOR REFINERY SLUDGES
CO
-------
224
INFLUENT Jv
SLUDGE
TOP LIQUID
ILEVEL
EMULSION
WATER
INTERFACE
EXCESS
WATER
TO API SEPARATOR
-WATER DRAW PIPE
SLUDGE AND EMULSION
LEVELS DETERMINED BY
PERIODIC SAMPLING
V
.STEAM ON OUTSIDE OF TANK
SLUDGE
TO SLUDGE STORAGE
TO SLUDGE STORAGE
FIG. 2. PRIMARY OILY SLUDGE DECANT TANK
FIRST STAGE
CENTRIFUGE
(Horizontal Solid Bowl)
SOLIDS
FINE SCREEN
RECOVERED OIL
WATER
SOLIDS
SECOND STAGE
CENTRIFUGE
(Disk-Type)
FIG 3. TYPICAL OIL RECOVERY AND SOLIDS
REMOVEL SYSTEM FOR OILY SLUDGE
-------
CHARGE MIXTURE
CONTAMINATED API SEPARATOR BOTTOMS
SLUDGE DECANT PIT SLUDGE
398 TANK BOTTOMS
209 TANK BOTTOMS
253 TANK BOTTOMS
28
36
12
225
ID
o
tr
<
a:
o
LU
>Ł
<
O
a:
LU
>
o
o
LU
a:
co
o
_i
o
CO
100
80
60
40
20
Volatile Suspended
(oil free)
I
_Total Suspended
Solids (oil free)
Recommended
Design
\
I 2 3
CHARGE RATE TO P600 (gpm)
FIG. 4. SOLIDS RECOVERY FROM
OILY SLUDGE ( REF 10 )
100
o^
>-
cc
CO
LU
O
o:
LU
Q.
90
9 80
o
CO
70
6Q.
0
O.A 925 G S
• A 2100 G'S
70
50 100 150 200 250
SLUDGE FEED RATE, (gals./hour)
300
FIG. 5. PERFORMANCE OF PILOT SOLID-BOWL
CENTRIFUGE (REF. 9 )
60
50
40
30
350
CO
a
d
CO
LU
3
O
LU
O
-------
CHARGE MIXTURE !S CENTRATE FROM
SHARPLES P600 CENTRIFUGE
i
3E SOLIDS RECOVERED SOLIDS RECOVERED IN
N OIL PHASE (%) NOZZLE WATER (%)
-f» CD OD
0 0 00 - M C
CC
X
o
20
(
CENTRIFUGE: T-
SH
•
- TOTAL
~ oru in
oUL 1 U
AV U L A
SOLID
1. TUBULAR BOWL
<\RPLES NOZLJECTC
. SUSPENDED
S (OIL FREE
FILE SUSPEN
S (OIL FREE
A - —
A
•
•
,
J
^
DED
'-) A
___
>
'
0
O P
• P
° Ł
• T
P
A P
A N
0 N
^ s
CO
Q
_J
O
OT 55
_J
<
1-
0
1-
UJ
^ 50
O
CO
CO
UJ
a:
°- 45
tr
UJ
i-
D 500 1,000 1,500 U.
DISK CENTRIFUGE FEED RATE (cc/min) 4O
H ADJUSTED WITH LIME, ONE HOUR RUN
H ADJUSTED WITH LIME, TWO HOUR RUN
H ADJUSTED WITH LIME, ONE HOUR FILTRATION,
NE HOUR HOT WATER WASH
HICKENED WEMCO FLOAT SOLIDS ADDED,
H ADJUSTED WITH LIME, TWO HOUR RUN
H ADJUSTED WITH SPENT CAUSTIC, TWO HOUR RUN
0 pH ADJUSTMENT, TWO HOUR RUN
0 pH ADJUSTMENT, TWO HOUR RUN, FEED MIXTURE
IMULATED PROCESSING ALL OF THE SPENT CLAY
A
0A
•
A
•
<
\
cj
)
•
o
CN
CM
FIG. 6. SOLIDS RECOVERY FROM
A SECOND - STAGE DISK
CENTRIFUGE (REF. 10 )
60 70 80 90 100 110 120
AVERAGE FILTRATE TEMPERATURE (°F)
FIG. 7. EFFECT OF FILTER PRESS
OPERATING CONDITIONS ON
CAKE SOLIDS (REF. 10 )
-------
BASIS
UJ
H-
2 OC
UJ <
r™ til
i J:
o
•w-
o:
UJ u.
I;
o o
UJ (0
13 ID
O O
X
(O h-
o
o
UJ
D
U.
350
300
250
200
150
Z 100
FUEL OIL COST= $ 3.60 / F.O.E. BBL
GROSS HEATING VALUE OF OIL = 18,500 BTU / Ib 227
DENSITY OF FUEL OIL = 7.35 Ibs /gal.
STACK TEMPERATURE = 800°F AND EXCESS AIR = 100%
GROSS HEAT REQUIRED FOR WATER EVAPORATION = 2400BTU/lb
2 4 6 8 10
SOLIDS PRODUCTION
12 14 16 18
Ibs DRY SOLIDS
YEAR
20
0 10,000 20,000 30,000 40.00O 50,000
SOLIDS PRODUCTION ( Ibs DRY SOLIDS / DAY )
FIG. 8. INCINERATOR FUEL COSTS DUE
TO SLUDGE WATER CONTENT
ALONE ( REF. 10 )
-------
228
STABILIZATION
THICKENING
DEWATERING
DISPOSAL
FIG. 9. BIOLOGICAL SLUDGE HANDLING
ALTERNATIVES
-STANDARD ACT SLUDGE
REMOVED
WASTE I
ACTIVATED I
SLUDGE (ACTIVATED
SLUDGE SOLIDS
t I '
SOLIDS DESTROYEE
IN AEROBIC
DIGESTER
RESIDUAL1 SOLIDSj
--—
FOR DISPOSAL-
OXIDATION
ENDOGENOUS
RESPIRATION
TIME OF AERATION •
FIG. 10. SCHEMATIC DEPICTION OF AEROBIC BIOLOGICAL TREATMENT
-------
229
Q
LJ
cr
CO
to
Bio-chemical
Mixed Pulp 8 Paper Waste
Refinery
A Mixed Domestic Sewage
T Textile 8 Domestic Waste
2 4
6 8 10 12
AERATION TIME (days)
FIG. 11. TYPICAL VOLATILE SOLIDS
REDUCTIONS BY AEROBIC DIGESTION
1000
Recycle = 230%
SS = 0.9%
0 0.20 0.40 0.60 0.80 1.00 1.20 1.40
TIME (min)
FIG. 12. RISE OF SLUDGE INTERFACE WITH
TIME
-------
NO
CO
O
0.040
0.032
0.024
CO
co 0.016
0.008
0
0
.0 2.0 3.0
PERCENT FLOAT SOLIDS
4.0
FIG. 13. DETERMINATION OF
OPTIMUM AIR/SOLIDS
RATIO (REF. 3 )
-------
231
FEED RECYCLE
VALVES
SKIM DIVERSION
VALVE
FEED FROM
AEROBIC DIGESTER
RECYCLE TO
AEROBIC DIGESTER
UNLOADER
ASSEMBLY -v
T _Ł._f__f ^* J J f^^f f^,f Ł J f J J f f
CURB
BOWL
OR BASKET
FIG. 14. INTERNAL COMPONENTS OF BASKET
CENTRIFUGE
-------
232
FIGURE A.
MEAN CAKE SOLIDS
CONCENTRATION
(WT. %)
20"
15"
10
5 +
0 20 40 60 80 100
MEAN SOLIDS RECOVERY (%)
FIGURE B-
KNIFE
CAKE SOFT SKIM CAKE
CAKE SOLIDS 15
(WT. %)
12345 6
DISTANCE FROM BOWL WALL (IN.)
FIG. 15. PERFORMANCE OF BASKET
CENTRIFUGE ON REFINERY
BIOLOGICAL SOLIDS (REF 6 )
-------
PROCESSING OF WASTE OILY SLUDGES
Jacoby A. Scber, Principal Environmental Engineer
Fluor Engineers and Constructors, Inc., Houston, Texas
INTRODUCTION
Petroleum crude oil, fuels, and lubricants become waste components when emul-
sified with water, solids, and/or debris and are potential contaminants of surface and
subsurface waters. The volume of oily waste is potentially large due to large quan-
tities of petroleum oils handled and used. Sources of the waste are from spillage of
crude oil and refined products and from petroleum refinery operations. This paper
deals with the refinery operation, but the findings are expected to be applicable for
most petroleum sludges. The title of this paper needs some clarification because the
use of the word "sludges" is all-inclusive-oils, emulsions and sludges will all be
treated as an umbrella topic.
SOURCES
Almost all operations of the petroleum industry, including exploring, producing
(extracting), storing, transporting, and refining of crude oil and the storing, distribut-
ing, and handling of products are potential sources of oily sludges.
Accidental spills of crude oil and petroleum products during the handling,
storing, and transporting operations are the principal cause for the formation of oily
sludges in large quantities. At refineries, accidents seldom cause oily sludge pro-
duction. Common sources are incoming crude oil, ship ballast water, tank and
vessel cleanings, oil-water separators, and numerous miscellaneous sources such as
sewer boxes and emulsion breaking facilities (demulsifier). The quantity of oil sludge
for disposal at a refinery does not depend only upon the nature of the crude and
processing units. Oily sludge formation can be minimized by prudent operating
practices, sensitive attitudes and suitable control methods. Generally, most of the
oily sludges accumulate in oil-water separators and in tank bottoms.
Crude oil shipped to the refinery contains emulsified material which is commonly
referred to as bulk, sediment, and water (BS&W). The BS&W concentration in the
crude oil ranges from about 0.01 to 0.1 percent by volume (%v). This BS&W
fraction contains about 30%v oil, 50%v water, 10%v carbonaceous matter, and in-
organic salts equivalent to about 10%v. The carbonaceous material which is the part
of BS&W that could eventually become oily sludge for disposal, amounts to about
0.005 percent of the incoming crude oil to the refinery. However, in the refining
steps, much of this carbonaceous material is utilized into fuels. The quantity of
waste sludge from crude oil is small for the following reasons: 1) Crude oil
receipts at a refinery are often stored in tanks for a few days before being fed to
processing units. During quiescent periods, most of the BS&W material settles out as
tank bottoms and is removed during tank cleanings at three to five-year intervals.
Fortunately, tank bottoms have been in demand for oil and wax recoveries by
independent operators and for road bed additives. When crude tank bottoms must be
handled at the refinery, a concentration step is possible with conventional slop oil
233
-------
234
demulsifying facilities. The BS&W emulsion is broken and some of the carbonaceous
material is included with the recovered oil phase and some with the water phase. The
recovered oil phase is fed to a refining unit and the water phase flows into an oil-
water separator. 2) In the event a settling time is not possible, the crude oil con-
taining BS&W material is fed to a crude distilling unit. The BS&W material is
removed in the pretreatment (heater and desalter) section of the distilling unit. The
oil fraction of the BS&W remains in the crude. Water fraction in the desalter
effluent flows into the feedwater stream of an oil-water separator. Emulsified
materials in the oil-water separator will either enter the recovered oil stream, settle
with the bottom sludge inside the separator box, or remain in the water phase flowing
to secondary treating facilities.
Ship ballast water-handling facilities vary at different refineries as does the
quantity and quality of both the water and oil phases of the incoming ballast water.
Ballast water may be fresh water, salt water, or a mixture of the two. Salt water is
usually clean with respect to suspended solids, whereas fresh water of the surface
category, such as that obtained from a large river, may contain large amounts of silt
that will become deposited in the tanks of vessels. In addition to silt, ballast water
may contain finely divided rust and precipitated chemicals used in tank cleaning.
The oil contained in ballast water will range from gasoline to heavy cracked fuel oils.
Ballast water that is discharged from gasoline or light-distillate oil tanks will
contain little oil until almost all the water has been pumped out. On the other hand,
ballast water that is discharged from a tank that contained heavy fuel oil may have
oil dispersed throughout the entire volume. This is also true of the emulsion from
tank cleaning.
Tank bottoms obtained during cleaning of tanks vary in composition and the
residual oil content can usually be recovered or at least reduced with emulsion
breaking facilities. However, the real bottoms and final washings from a tank are
often oily sludges not suitable for feed to a demulsified unit and are potential feed
for the soil cultivation process.
Oily sludge from oil-water separator and holding pond cleanings is usually low
in oil content. The sludge cleanings from the water box or pond usually contains
about two percent carbonaceous material. Oil box cleanings which are mostly oil
can be fed to the demulsifying unit, but sometimes mixtures of oil, straw, grass, and
dirt make this material suitable only for disposal.
Solids and oily waste at the demulsifying unit usually accumulate in tanks which
must be cleaned occasionally, and these cleanings cannot be further improved with
additional treatment in demulsification facilities.
Process unit shutdowns include complete cleaning ofa piece of equipment and/or
vessel before maintenance work begins. In the shutting down of a process unit, oil
is returned to storage. Immediately after shutdown and where possible, the equipment
is washed with water which removes residual traces of oil and carbonaceous material.
This material is sent to an oil sewer which connects to an oil-water separator.
-------
235
Many refineries will require additional oil and suspended solids removal by
pretreatment processes in order to meet process requirements of subsequent treatment
stages. The quantity of additional oily sludge produced can be estimated based on
material balances for the suspended solids removal process such as air flotation. The
additional oil solids production as shown in Table I was estimated based upon median
TSS removals by the air flotation process at median flow rates. Median suspended
solids removal for all classes of refineries was indicated to be 15.3 pounds per thousand
barrels of crude. The total oily sludge production was estimated based on the sum of
the oily sludge production without additional oil and TSS removal equipment and the
oily sludge production anticipated for dissolved air flotation. Total oily sludge pro-
duction as a function or refinery size is indicated in Figure I.
Stating the sludge production in another fashion, it can be expected that the
waste sludge from API separators plus the usual secondary treatment facilities will con-
tain 1,500-6,000 pounds of dry solids per million gallons of effluent treated. The wet
sludge composition will be approximately 1-3 wt.% dry solids, 10-30 wt. % oil, and
65-90 wt. % water.
On an average sludge composition basis of 2 wt. % dry solids, the flow or
accumulation of wet sludge will therefore by 75,000-300,000 pounds per million
gallons of effluent, or about 0.9-3.6 vol. % of the effluent (Rabb states the range to
be 1.5-4.0 vol. % of the effluent).
Weston, Merman, and De Mann report 1,500 pounds of dry solids per million
gallons of effluent from API separation plus chemical coagulation. They also report a
range of 120-710 for API separation along.
For an existing refinery, the quantities and compositions of oily sludge are
attainable by records and measurement. For the design of a new grass roots facility,
these items must be estimated. As long as reasonable allowances are made for con-
servative design, the exact compositions are not very critical. It is highly unlikely
that any such estimate would be duplicated precisely in real life.
The compositions shown in Tables II-V are from the API Manual on Dispoal of
Refinery Wastes, Volume VI, Solid Wastes, 1963.
There are other oily sludges generated in refineries which may be classified as
oily but usually require special consideration in handling and disposal. Typical among
these are sludges generated in Tetraethyl Lead Gasoline Blending Operations and
Hydrofluoric Acid Alkylation Processing. Oily sludges generated from these and other
exotic operations must of necessity be segregated and handled separately because of
their toxicity and other environmental hazards.
TRANSPORT AND HANDLING
Precaution should be taken in any wastewater collection and treatment plant
design to minimize pumping where both oil and water are present together. One of
the most difficult processes to run successfully is the breaking of oil-water emulsions.
-------
236
Care should be taken during the design phase of a project to minimize pumping before
the one constituent (oil or water) is removed by gravity or other separation.
If at all feasible, process wastewaters containing oil should drain by gravity
into the API or other type oil-water separators. This will preclude the possibility of
creating excessive amounts of emulsion. Oily water, if it must be pumped at all,
should be pumped utilizing screw type lift pumps or low RPM open-impeller centri-
fugal pumps. Any mixing of chemicals with oily water should be accomplished using
the lowest RPM mixing speed which will perform the task. An example of this con-
cept is the chemical addition of alum and/or polymer before dissolved air flotation.
Oily sludges from separators, tanks and other sources are frequently quite
viscous and can contain sand, gravel and other large particles. Screw pumps or
positive displacement pumps are to be preferred for this type of service.
PROCESSING METHODOLOGY
Oily sludge processing techniques include emulsion breaking, gravity, separa-
tion, centrifugation, filtration, and incineration as shown in Figure 2.
Emulsion Breaking
Refinery emulsions may be defined as intimate, two-phase mixtures of oil and
water with one phase dispersed as minute globules in the other phase. The minute
globules are stabilized by an interfacial film or stabilizing agent such that the
globules do not coalesce and do not respond to gravity settling.
Emulsions usually are complex, and emulsion technology is, at best, only
quasi-scientific. Generalizations are risky, and trial-and-error experimentation is
essential. For a systematic approach to emulsion problems, emulsions may be
characterized by:
a. Dispersed phase, as oil-in water (O-W) or water-in-oil (W-O).
b. Dual O-W and W-O emulsions.
c. Stability, as measured by change with time of the dispersed phase
particle size analysis and/or a time-separation study.
d. Type of interfacial film:
Anionic - acid stable or acid sensitive.
Cationic - alkali stable or alkali sensitive.
Nonionic - hot-alkali stable or hot-alkali sensitive.
e. Degree of success in emulsion breaking according to process, e.g.,
physical methods - heating, distillation, centrifuging, or filtration.
Improper treatment of either O-W or W-O emulsions may invert the emulsion
to the opposite type.
The most satisfactory theory of emulsions, which can also account for dual
emulsions, inversion of phases, and effects of a wide range of emulsifiying agents,
-------
237
is one which emphasizes the nature of the interface between oil and water. An
interfacial film, which is a complex mixture of dissolved and colloidal matter and
suspended solids, generally is understood to exist. Such an interfacial film may be
highly viscous, gelatinous, and/or possess an electrical charge. Some experimenters
consider the interfacial film as a third phase. Generally, the stability of an emulsion
is not a function of the degree of dispersion; rather, it is a function of the stabilizing
agents and properties of the interfacial film, such as film viscosity and electrical
charges. This is illustrated by the following major factors which affect emulsion
stability - four out of five of these factors being directly related to the interfacial
film:
a. A coherent, mechanically strong interfacial film.
b. An oriented unimolecular layer of emulsifying agent or tightly
packed solid particles at the interface.
c. Viscosity of the continuous phase and of the emulsion.
d. Nature and amount of emulsifying agents.
e. Physical storage or historical conditions, such as temperature,
agitation, and dilution.
Properties of emulsions of refinery waste control are best approached by
understanding what generally contributes to the stability of an emulsion and, therefore,
what can be done to cause an emulsion to become unstable and separate or "break".
Treating from this point of view requires an understanding of causative emulsifying and
stabilizing agents.
Substances which cause or stabilize emulsions may be classified as ionic or
nonionic and, generally, are of a colloidal nature or consist of solid particles which
are surface active. Hydrated gels or gums, soaps, sulfonated oils, asphaltic residues,
waxes, silt, finely divided coke, and sanitary sewage solids typify the range of
emulsifying agents encountered. Colloidal matter may be further classified as hydro-
philic (waterside active) or hydrophobic (oilside active).
The important ionic emulsifiers are chemically active because of the active
carboxyl, sulfonic, amine, of hydroxyl groups. Such compounds, singly or in com-
bination with electrolytes, may readily result in a stable emulsion. When there is no
chemical reaction between the electrolyte and emulsifying agent, the electrical charge
that may exist on the dispersed globules in an emulsion has little significance in
regard to stability or to type of emulsion.
The following rules generally apply for emulsion formation:
a. O-W emulsions are formed when soaps are colloidal in the water
phase (hydrophilic).
b. W-O emulsions are formed when soaps are salted or precipitated
from the aqueous phase.
c. Soaps partially salted out may form and stabilize either type emulsion.
d. Electrolytes favoring O-W emulsions are salts of univalent cations and
sdlts of di- and trivalent onions.
-------
238
e. Electrolytes favoring W-O emulsions are salts of di- and trivalent cations.
One approach to emulsion breaking is to counter-balance emulsifying agents of
one type of emulsion (W-O or O-W) with one of the opposite type.
Methods for emulsion breaking can be grouped into physical, electrical and
chemical. Each is briefly described below. Frequently, the system used in any
particular refinery will include more than one combination of these methods.
Physical Methods
Heating
The process of heating emulsions to obtain a break is the simplest and
most economical approach available; the use of heat is nearly always advantageous.
Heat markedly reduces the viscosity of the oil phase, melts interfacial waxy films,
and, in some instances, weakens the interfacial films of various emulsifying agents.
Coalescence and separation of the oil and water phases can then take place.
For W-O emulsions, heat increases the vapor pressure of the water which
tends to rupture the films around the globules. The formation of steam within the
globule generally will break a W-O emulsion, but the rate of steam generation must
be controlled to avoid excessive foaming.
A heat-and-pressure system may be used to break many types of emulsions.
The emulsion is heated in a suitable pressure vessel wherein water and sediment are
precipitated. Adjustment of the pH by addition of acustic or acid may help facilitate
the breaking of an emulsion with heat. There appears to be no general rule to serve
as a guide in determining optimum pH. This is a situation where trial by experiment
is the best solution.
Distillation
Distillation methods for breaking emulsions are effective in that they involve
the use of heat as described previously. In addition, water and light ends of the oil
are removed as overhead products, and the emulsifying agent remains with the residue.
Centrifuging
Comparatively stable mixtures sometimes may be separated by centrifuging.
If there is a difference in specific gravity between the oil and water, centrifugal
force may be used to accelerate separation. Loosely occluded water is readily
separated from oil or emulsion by centrifuging.
Precoat Filtration
Stable W-O emulsions, particularly those stabilized by finally divided solids,
can be resolved by filtration. Rotary vacuum precoat filters are commonly used for
this purpose. In this operation, the emulsion is filtered through a layer of diatomaceous
earth deposited on a continuously rotating drum filter. The suspended solid material
and as a result of removing the stabilizing of the precoat layer during that portion of
the cycle in which the drum is submerged in the material being filtered. During the
-------
239
remainder of the cycle the filter cake and precoat are dried. A knife edge advances
automatically into the precoat layer to a predetermined depth for each revolution of
the filter drum, thereby removing the sediment deposited on the precoat layer along
with a very thin layer of the precoat material.
The breaking of eumlsions occurs as a result of rupturing the globules in the
dispersed phase when passing through the interstices of the filter cake and precoat
material and as a result of removing the stabilizing solids. The oil and water phases
of the filtrate readily separate by gravity difference. These phases may still contain
dissolved emulsifiers. If so, care should be taken to prevent the re-formation of the
emulsions by excessive turbulence or mixing before separation of the oil and water
layers.
Electrical Methods
Some emulsions can be broken under the influence of a strong electrical field
by application of the principle of the Cottrell precipitator. The emulsion is passed
between two electrodes and subjected to a high-potential pulsating current. The water
globules are attracted electrically and coalesce until the masses are of sufficient size
to settle by gravity. This method is used in the dehydration and desalting of crude
oil.
Where electrical desalting is used for treating crude oil feeds, some emulsions
may be blended along with the regular crude oil charge. Emulsions can sometimes be
blended at rates ranging from 1 to 10 percent of the crude oil charge if the water and
sediment contents of the combined mixture are not too high. Before attempting to
dispose of emulsions through a crude oil desalter, consideration should be given to the
possibility of cracked material in the slop oil throwing virgin products from the crude
oil unit off specifications. Equally important for consideration are emulsifiers in the
slop oil that might throw the oil or water from the desalter off specification.
A combination of electrical and chemical methods is frequently successful
where neither method alone is effective.
Chemical Methods
Emulsions can be broken by chemicals which will balance or reverse the inter-
facial surface tension on each side of the interfacial film, neutralize stabilizing
electrical charges, or precipitate emulsifying agents.
Anionic and cationic surf ace-active agents are not compatible and will tend
to neutralize each other. Rules for breaking emulsions stablizied with electrolytes are:
-2
a. Reactive onions, such as OH and PO, , will break W-O emulsions.
b. Reactive cations, such as H , Al , and Fe , will break W-O
emulsions.
The use of a heavy metal salt to form a flocculent precipitate of a heavy
metal hydroxide to break dilute O-W emulsions is typified by the flocculation
process. The resulting clarification or emulstion breaking is believed to result from
-------
240
destruction of the emulsifying agents by neutralizing or reversing interfacial-film
electrical charges and, in turn, by agglomeration of the precipitated floe and ad-
sorption of suspended oil matter on the floe surface.
Any wetting agent will break a W-O emulsion provided the agent is
effectively surface-active on the waterside of the interfacial film and a certain
critical amount is not exceeded. An example of the use of a surface-active chemical
to break an emulsion is the use of a detergent to break an asphaltic, solids-stabilized
W-O emulsion by causing water to wet the solid preferentially - thereby displacing
the loosely adsorbed asphaltic emulsifying agent.
Effective organic chemicals for breaking W-O emulsions generally have either
an inorganic group (phosphate) or a reactive group (such as OH or -NhL) and have
at least three focuses of polarity occurring in different terminal groups.
Organic chemicals used in breaking W-O emulsions should:
a. Be soluble in both phases, i.e., the treating chemical contains both
hydrophilic and hydrophobic groups.
b. Form an adsorption complex with the predominantly hydrophobic
emulsifying agent present thereby creating a hydrophilic complex that
can be wetted by water.
c. Contain one or more radicals able to react with calcium (Ca) or
magnesium (Mg) salts in the aqueous phase and result in insoluble salts
that are wetted by water.
In general, the dewatering of recovered oil or breaking of recovered oil
emulsions in refinery practice is a series of batch operations involving several stages
of water and solids removal. The simplest system consists of holding tanks operated
on a batch basis, usually with the addition of heat and use of recirculation. How-
ever, both cold and hot settling are used, with cold settling time varying from one
day to four weeks, and hot settling time usually varying from one to three days at
temperatures in the range of 150 F to 200 F.
At the end of each stage of treatment, substantially oil-free water is drawn
off and returned to the refinery separators and low-water-content oil is decanted and
returned to the refinery for reprocessing* In most cases, a layer of emulsion remains
after these separations. This emulsion is generally given further treatment. The
methods used vary widely and are usually adaptations of the mthods described earlier
in this chapter.
The dewatered oil may be charged to crude oil units, with or without
desalters; to thermal or catalytic cracking units; to coking units; to visebreakers; or
to separate distillation units. Water content of recovered oils fed to crude oil units
or separate distillation units may run as high as 2 percent, but the water content of
recovered oil fed to the other types of units is generally limited to 0.3 to 005 per-
cent.
-------
241
The following figures 3 and 4 illustrate a rather simple continuous process for
breaking emulsions emanating from storage tanks and a rather sophisticated process for
de-emulsifying all refinery^wide sources.
Gravity Sedimentation and Phase Separation
The processes of gravity sedimentation and phase separation combine the
effects of specific gravity, viscosity, and time to provide for the separation of sub-
stances of different densities. Conical bottom vessels are utilized for most applica-
tions, especially where a continuous drawoff or a concentrating effect is desired.
Centrifugation
The use of Centrifugation permits the mechanical dewatering of sludges through
centrifugal force. Dewatering by this method may be defined as sedimentation under
the influence of forces greater than gravity. Within a centrifuge, centrifugal force
acts on a suspended particle in the sludge, causing it to settle through the liquid.
It would seem logical to design machines for maximum centrifugal force, but this is
not the only consideration. While additional centrifugal force helps settle solids, it
also increases the difficulty of discharging the settled solids. Thus, for maximum
efficiency, a compromise must be made.
Several variables which affect gravity sedimentation also effect sedimentation
within centrifuges; these variables are:
1 . Particle size and shape,
2. Density differential between solid particles and the liquid,
3. Concentration of the particles, and,
4. Liquid viscosity.
Sludge solids in suspension are a combination of granular, fibrous and
flocculent particles. Granular particles in dilute solution are generally considered to
settle at terminal velocity, independent of one another, and without change in size
or shape. When the particles become concentrated however, volumetric considerations
predominate and the observed settling velocity appraoches zero. Flocculent and
fibrous particles do not settle at a constant velocity since they tend to cluster during
sedimentation, bringing about a continuous change in size, shape, and relative
density. Extremely fine particles which may not settle under normal gravity separa-
tion can be made to settle by increased gravitational forces. In general those sludges
which spearate most readily and concentrate to the greatest degree by sedimentation
are also those which are dewatered most efficiently by centrifuges.
Two of the major types of centrifuges in industrial use today are the solid
bowl and the basket type. These two types may be used singly or in combination to
remove solids from solution. When used together, flow is first directed to the solid
bowl centrifuge where the large solids are removed and then to the basket type
centrifuge where finer solids are removed and emulsions broken due to the higher
centrifugal forces in this unit. The solid bowl sludge dewatering centrifuge assembly
consists of an outer horizontal cone (bowl) and an inner screw conveyer, each
rotating at slightly different speeds. Sludge is fed to the centrifuge through the
central screw and enters the bowl about halfway along the side of the cone.
-------
242
Centrifugal action forces both liquid and solid phases to the walls of the cone and
down the cone to the large end, where water in the top layer of the liquid pool is
drawn off. The solids, which accumulate closer to the wall, are conveyed by the
screw to the small end of the cone where they are discharged. Turbulence due to
the screw, limits the solids concentration obtained, and maximum solids concentrations
from this type centrifuge are normally in the range of 5-10 percent.
As opposed to the continuous operation of solid bowl centrifuges, basket type
centrifuges operate on a timed cycle. The units resemble the rotating tub of a spin
dryer used in home laundering with the exception that the tub has a hopper at the
bottom. Slurry is fed to the tub (basket) during the first portion of the cycle, and
the spin of the basket creates centrifugal action up to 13,000 times the force of
gravity which forces liquid and solids against the outer walls. Part of the liquid
travels up the wall and solids layer and out over the top lip of the basket and into
a discharge trough. According to a timed cycle, remaining liquid is then drawn off
by a pipe which slowly moves toward the wall from the center of the basket. AS
the pipe approaches the wall, it begins picking up a concentrated solids slurry, and
a valving change routes these to the bottom hopper. A new cycle then begins.
After several such cycles, a hard solids layer accumulates near the wall and this
layer is then removed with a knife edge at a slower rpm and these solids fall to the
bottom of the hopper where they combine with the other concentrated slurry and are
discharged. Using this type of centrifuge, overall sludge concentrations of 8-15
percent are possible.
The introduction of flocculation aids, such as polymers, has increased the
range of materials that can be dewatered satisfactorily by centrifuges. The degree of
solids removed can be regulated over a wide range depending on the amount of
chemical coagulant used. Wetter sludges result from chemical aids because of the
increased capture of fines.
Rotary Drum Vacuum Filtration
Vacuum filters are the most widely used type of mechanical sludge dewatering
device. The filtration process is used to separate the solids from the liquid by means
of a porous media which retains the solids but allows the liquid to pass through.
Filter media employed include Nylon, Dacron, and Polyethylene cloth; steel mesh; or
tightly wound steel coils. The types of rotary drum vacuum filters currently in use
differ primarily in the way cake is removed from the drum. The filtration process is
accomplished by means of a horizontal drum covered with the filter media. The drum
is rotated in a tank with approximately one quarter or more of the drum submerged in
wet sludge. Valves and piping are arranged so that as the drum rotates a vacuum
applied on the inner side of the filter media draws water from the sludge and holds
the layer of sludge to the drum. The application of a vacuum is continued as the
drum rotates out of the slurry. This pulls additional water from the sludge, leaving a
moist cake of sludge on the outer surface which is removed prior to re-entering the
tank. Cake is removed by a knife that moves on the filter drum as it rotates,
Precoats and filter aids are often used to speed filtration rates or collect more
of the fine particles in the slurry. The precoat, usually diatomaceous earth, is a
-------
243
finely divided, hard-structured solid that forms an open, noncompressible cake.
When applied on the filter cloth, the precoat acts as the primary filter medium and
permits high removal of fine particles from the slurry. A filter aid may also be mixed
with the slurry where it becomes distributed throughout the cake, thereby keeping the
cake relatively open for flow and continuously supplying a large surface for adhesion
of finely divided solids. This method is particularly good when filtering colloidal
solids that are not to be recovered from the cake.
Filter media may consist of cloth, paper, or woven or porous metal as men-
tioned before. Criteria on which a filter medium is selected must include the ability
to remove the solid phase, high liquid throughput for a given pressure drop,
mechanical strength, and chemical inertness to the slurry or any wash fluids. In
operation, some filter cake solids usually penetrate the filter medium and fill some of
the pores. In normal cases, five to 25 percent of the pore volume of the filter
medium if filled with solids. As a result, resistance to flow through the medium
increases greatly - in some cases to such an extent that the filtration rate is
seriously reduced.
Properties which affect the vacuum filtration dewatering process include:
1 . Solids feed concentration,
2. Sludge and filtrate viscosity,
3. Sludge compressibility,
4. Chemical and physical characteristics of the sludge,
5. Operating vacuum,
6. Drum rotational speed,
7. Drum submergence,
8. Type and porosity of the filter media, and
9- Effect of chemical additions.
Pressure Filtration
The filter press is the simplest of all pressure filters with the most widely
used type being the plate-and-frame press. As the name implies, the plate-and-
frame is an assembly of alternate solid plates, the faces of which are grooved to
permit drainage. Between these plates are hollow spaces in which the cake collects
during filtration. A filter medium, usually some sort of a fabric, covers both faces
of each plate. As filtration proceeds, cake builds up on the filter cloths until the
cakes being formed on each face of the frame meet in the center. When this
happens, the flow of filtrate which has been decreasing continuously as the cake
thickness increases, drops off rapidly. Usually, filtration is stopped well before this
occurs. After the filtration sequence, the press is disassembled and the solids are
collected either manually or mechanically and discarded.
Filter presses can be built for high pressures and can handle the filtration of
either heavy or light slurries. The filter press has many advantages, the greatest of
which are its simplicity, flexibility, and ability to operate at high pressures. A
filter press, with proper operation, will produce a denser, drier cake (up to 60 per-
cent solids) than any other type of filter. The sludge and filtrate properties which
-------
244
affect the filter press dewatering process are the same as those for vacuum dewatering.
The operation of separating waste oil from oily emulsion by means of a filter press
opearated at elevated temperatures is being practiced. A process in which oily
sludges are subjected to high pressure filtration enabled the oil to be removed in the
form of a separable phase. This process produced a cake with a high concentration of
solids and low concentration of oil which would be suitable for land farming. The
cake produced by high pressure filtration contained very little volatile material which
could be removed by incineration. Significant oil recovery in the process reduces
the overall cost of the disposal. The economic incentive of recovering oil from oily
sludges will increase as crude oil supplies become less available.
Incineration
A properly designed and controlled incinerator is satisfactory for burning com-
bustible waste provided air pollution standards can be met. A high control tempera-
ture is needed for efficiency, prevention of excess air pollution, and odor control.
Incinerators are readily adaptable to the disposal of virtually all oxidizable wastes,
including streams which contain inorganic or other non-combustible materials. Com-
bustible wastes which contain 10,000 BTU per pound or more can usually be incin-
erated without the requirement of an auxiliary combustion chamber. In some
instances, where the heat value is less than 10,000 BTU per pound, it is possible to
enrich the liquid waste with another fuel. Incineration is being used at many large
plants where final disposal of the solid waste is a problem. It has the advantage of
freedom from most odors, independence of weather conditions, and high reduction of
the volume and weight of the end product to be disposed of. There is a minimum
size of treatment plant below which incineration is not considered economical since
there must be enough waste solids to warrant reasonable use of the expensive equip-
ment involved.
CASE HISTORY
A major Gulf Coast Refinery is presently in the midst of a major process ex-
pansion program to essentially double their throughput and increase their ability and
flexibility in processing a wider slate of crude oil, including high sulfur mid-eastern
crudes. New facilities to be installed include Crude and Vacuum unit with Electro-
static Desalting, FCC, Hydrotreater, Gas Plant, Product Treating Plants and Sulfur
Plant. Inherent in the design of new, additional processing facilities was an
examination and evaluation of existing offsite facilities such as raw water and waste
treating, steam generation and tankage. In conjunction with a consultant, a major
program for modification of and addition to the waste treatment facilities was under-
taken in order to assure compliance with both State and Federal discharge require-
ments. Concurrent with and as a part of this operation, an evaluation of the "oily
sludge" sector of the plant operation was also performed. The following describes
the history of this evaluation.
Old Facilities (Figure 5)
All oily wastes plus cooling tower blowdown water, boiler blowdown water,
lime softener blowdown, and zeolite regenerant water were discharged to the oily
sewer. The oily waste sewer discharged to an API Separator where gravity separation
-------
245
of free oil and water was accomplished, followed by an air flotation unit, employing
a cationic polymer as the sole chemical feed. The API effluent was sprayed to the
atmosphere to accomplish cooling and was collected in a basin and discharged to two
aeration basins connected in series. Aeration was supplied by a 15 hp aerator in each
basin. The waste was discharged to a settling basis and thence to a lagoon. The
lagoon discharged through a measurement weir to the ship channel.
All storm waters were collected and discharged to the second aeration basin
for treatment prior to discharge to the lagoon.
Ballast water was pumped to a ballast water tank where oil was separated by
gravity and discharged to ballast water holding basins and thence to the second
aeration basin for further treatment prior to discharge to the lagoon.
Recovered oil from the API separator was sent to an emulsion breaking process
and the resultant sludge from recovery was discharged to a sludge pond. The float
from the air flotation unit consisting of sludge and some oil was also discharged to
the sludge pond. Efforts have been made to recover this oil in the emulsion breaking
process but the inclusion of the air float sludge rendered the process inoperable, pro-
bably due to the presence of calcium and magnesium salts in the air float sludge.
The sludge ponds did not dewater readily and removal and subsequent land
disposal of the sludge had become a problem.
Due to the inclusion of highly alkaline boiler blowdown, filter backwash
water, lime softener blowdown and softener regenerant waste, the pH of the API
influent averaged 9.0. The average total steam load was 304,000 Ibs/hr and at
10% blowdown, the discharge, at an average pH of 11.5, to the oily water sewer
was 60 gpm, or 17% of the average volume input to the API Separator. In addition,
the discharges from the lime softener, filter backwash, zeolite regenerant wastes as
well as some spent caustics from the refinery process contributed to the high pH.
Since it is well known that high pH conditions tend to cause oil/water
emulsions, the efficiency of gravity separation of free oil at the API Separator was
impaired, reducing the quantity of recoverable oil at that point.
The inclusion of boiler water blowdown in the oily waste stream produced a
softening effect within the oily waste system, resulting in a rather heavy precipitate.
The average blowdown flow of 60 gpm or 715,000 Ibs. per day was the equivalent
to 114 Ibs. of soda ash, 175 Ibs. of caustic, and 44 Ibs. of tri-sodium phosphate
per day. The precipitate formed was removed as float in the air flotation unit con-
tributing to the excessive sludge problem at this point, and prohibiting further
recovery of oil. In addition, the oil adsorbed on the floe particles carrying over
from the air flotation unit contributed to the oil content of the final effluent as did
the emulsified oil due to the pH.
The emulsion breaking process operated very satisfactorily upon recovered oil
from the API separator but could not handle the float from the air flotation unit due
to precipitated calcium and magnesium salts.
-------
246
Sludge from the emulsion breaking process and the float was discharged to the
sludge ponds where after a period of dewatering it was mechanically removed and
sent to land disposal. The sludge was very difficult to dewater due primarily to the
air float sludge and the amount of rainfall and presented a difficult problem.
INITIAL RECOMMENDATIONS AND SYSTEM UPGRADING
The most economical and feasible point to accomplish more efficient oil
removal and recovery was at the API separator. This could be accomplished by
diverting high alkaline wastes from the oily waste stream to lower the pH of the
stream entering the API separator to prevent emulsification of oil and achieve more
efficient gravity separation of oil.
The diversion of boiler blowdown alone, which represented 17% of the flow
volume to the API separator, would increase the retention time considerably to ac-
complish better gravity separation of oil.
A second beneficial effect of diverting alkaline waste streams was the elimina-
tion of precipitated calcium and magnesium salts which constitute the greatest part
of the sludge volume discharged to the sludge ponds by way of float from the air
flotation unit. Thirdly, the float, with the absence of precipitated matter, could
be discharged to the emulsion breaking process for additional oil recovery. Of
course, a prime benefit of diversion of the alkaline wastes would be the reduction
of emulsified oil content of the discharge to final waste treatment facilities and
final outfall to the ship channel.
The reduction in sludge volume discharged to the sludge ponds would
facilitate dewatering and ultimate final disposal.
The water treatment plant blowdown was diverted from the main process
sewer and pumped into a holding pond. It was apparent immediately that oil
recovery was more efficient and sludge and emulsion formation less severe. However,
due to the project timing and financial considerations, it was decided not to try to
recover oil from either the API sludge or DAF float. This would be attempted when
the expansion was complete and more permanent facilities installed.
NEW REFINERY SYSTEM
Waste Oil
The refinery has been treating waste slop oils and emulsions in the following
manner - that is, place the slop oils into 500 BBL tanks, heat, add a chemical,
and cool. Emulsions were easily broken and oil and water were separated. The oil
was returned to the crude units and the water was recycled to the waste treating
unit.
This method was very successful excep that during the hot summer months the
two existing 500 BBL tanks were sometimes overloaded due to insufficient cooling.
The operating cycle was as follows:
-------
o 247
1. Fill tank and heat to 130-160 F - this required 4 to 7 days.
2. Cool tank - 4 to 5 days (Atmospheric Cooling.)
With only two tanks and periods of high oil loadings, a tank that has not
sufficiently cooled may need to be drained so that it can be refilled. The warmed
drainage was usually not completely demulsified.
For the expanded refinery, the following treatment of the waste slop oil was
proposed.
1 . Results of the tests by local waste water treatment chemical vendors
indicated that cold settling of oily water emulsions did not result in
demulsification.
2. Add four 500 BBL tanks with mixers, heating/cooling coils and sloped
bottom for treating slop oil. Separated oil from these tanks could be
pumped to existing slop oil tanks for surging prior to pumping the oil
back to the crude unit feed tanks.
3. During periods in which excessive amounts of slop oils were collected,
the existing slop oil tanks could be converted back into demulsification
treating service.
Oil Sludges
The refinery was treating their sludges in the following manner. Both API
bottom sludges and air flotation skimmed material were sent to sludge pits where the
oily solids settled. The oil and water overflowed to second and third pits. The
water was recycled to the air flotation unit and oil routed to slop oil treating. The
thickened sludge was pumped to another pit for drying. This method worked satis-
factorily. However, the new waste water treating facilities would be constructed in
the same area that the existing sludge pits are now located, therefore an intemnin
and permanent solution to the disposal of waste solids was required.
Two disposal alternatives for oily solids were compared in order to determine
the most cost effective method of waste solids disposal. The projected waste oily
sludge inventory is shown in Table 6 and a mass balance diagram for oily solids is
given in Figure 6.
Alternate No. 1 - Contract Disposal
The bottoms from both gravity oil separators would be combined in a cone
bottom 500 barrel tank with the air flotation skimmings being stored in another
identical tank. The separator bottoms will concentrate to approximately 15% solids
and the skimmings will concentrate to approximately 3%. After emulsion breaking
decant oil from the tanks would be routed to the slop oil system, oily water to the
old area API, and sludge to contract disposal,,
Contract Disposal Site: A 60 barrel vacuum truck would be needed at
approximately two week intervals, to haul sludge to a disposal site. Since tank
bottoms and oily sediment will be removed intermittently, it was assumed that a HO
barrel vacuum truck would be used for hauling direct to a disposal site. Current
-------
248
vacuum truck rates are $17.50 to $23.00 per hour, and 4 hours per load was used to
estimate hauling costs. The estimated capital cost and yearly operating cost is in-
dicated in Figure 7 based on the project waste solids volumes given earlier.
Alternate No. 2 - Incineration and Landfill
The separator bottoms, air flotation skimmings, oily sediment, and tank
bottoms would be stored in a 1000 barrel tank before being charged to the incin-
erator. Although variation in the oil content is expected, an oil content of 10%
was used for estimating purposes. An average solids concentration of 4% was used
for the estimate. The cost considerations of incineration can significantly influence
the processing steps required. The major factors are fuel costs, loss of oil recovery
potential, and the capital cost of the incinerator. The principal parameters which
determine the above costs are solids production rate, water concentration, and oil
concentration. Due to the complexity of incinerator design and the absence of data,
the cost for incineration was based on data from similar systems at other refineries.
The capital and operating costs are shown in Figure 7.
On the basis of the preceding discussion and preliminary cost estimates the
most cost effective temporary solution to solid waste disposal at the refinery is to
continue the practice of contract disposal of oily solids following thickening. It
must be recognized that the proposed system is not the ultimate solution to the
refinery's waste solids problems. However, the proposed system is a prerequisite for
most future ultimate solids dispoal systems.
The future of contract disposal sites is presently in doubt as is future regula-
tions governing landfill and land farming operations. In order to determine the
magnitude of the waste solids disposal problem at the refinery, the suggestion is
made that the refinery initiate an investigation to determine the feasibility of com-
bining the oily and biological solids and either incinerating the wastes or dewatering
them prior to land farming or contract disposal. Also to be considered and, in
fact, in progress is a program to reclaim oil from the sludge prior to dewatering
and ultimate disposal. The study should commence after the expanded refinery is in
operation and should encompass all solid wastes generated within the refinery. This
recommendation is made on the basis of limited available land area. The fact that
contract disposal costs will continue to increase must be recognized. It is suggested
that it would be in the refinery's best interest to minimize dependence on off-site
disposal facilities.
BIBLIOGRAPHY
(1) Manual on Disposal of Refinery Wastes, Volume On Liquid Wastes, American
Petroleum Institute, 1969
(2) Manual on Disposal of Refinery Wastes, Volume on Solid Wastes, American
Petroleum Institute, 1963
(3) Economics of Refinery Wastewater Treatment, American Petroleum Institute, 1973
(4) Aqueous Wastes, Milton R. Beychok, John Wiley & Sons, 1967
-------
249
(5) "Sludge Disposal: A Growing Problem", Petroleum Refiner, April, 1965
(6) "Waste Disposal Problems Of The Petroleum Industry", Industrial Wastes,
American Chemical Society Monograph 118, Reinhold Publishing Corp., 1953
(7) Residuals Management in Industry; A Case Study of Petroleum Refining,
Clifford S. Russell, Johns Hopkins University Press, 1973
(8) Oily Waste Disposal By Soil Cultivation Process, C. Buford Kincannon, EPA
Report *R2-72-110, 1972
(9) "Land Disposal Of Oily And Biological Sludge - Some Site Considerations",
Ro C. Chatham and J. A. Scher, AlChE 68th Annual Meeting, 1975
(10) "Industrial Wastewater Pumps", J. A. Scher, Chemical Engineering Deskbook
Issue, October 6, 1975
(11) Private Communications to J. A. Scher From Client and Consultant, 1974-5.
DISCUSSION
Morris Wiley: You mentioned the problem of sludges from ballast reception. The
1973 IMCO convention is ultimately going to require shore reception facilities for
all kinds of wastes from ships and from tankers in particular. There are really about
four major categories of wastes. One of these is clean ballast from tanks that have
been cleaned at sea with the Load-on-Top procedure, one is dirty ballast from tanks
which haven't been cleaned, and one is the slop oil residue which is carried on
board from Load-on-Top, and another is that some refineries accept the total tank
washings without any overboard discharge from the vessel while at sea» But then the
most troublesome kind of waste arises when the ship has to go into ship repair port.
They have to get in and muck out the bottom of the tank and for a very large tanker
this can run at a huge volume of unpumpable tank cleanings. Do you have a
recommendation as to what to do with that type of sludge?
J. A. Scher: That is a difficult question for me to answer. I recognize the
problem, in fact, even a lot of refineries don't have the proper facilities for handl-
ing the quantitites of sludge that are involved. I think you will find in the long
run, in conjunction with the whole toxic material system sludges are going to be
handled more or less on a regional basis. I think, and I am really speculating here,
that the government is interested in handling these in a more uniform system. Of
course, you are familiar with the toxic and hazardous materials legislation that is
being generated and I understand there is a very strong possibility of regional waste
management for sludges, etc. In the long run I think you will see a regional
approach to taking these sludges to a central area for processing rather than trying
to do them on individual basis, because it is very expensive and it is very difficult
to do. As far as an individual facility like you describe it would be a very
expensive proposition for them to put in a complete dewatering, oil handling, oil
removal, sludge disposal system at a shipyard.
Morris Wiley: I think I would agree with that point but then there is a subsidiary
question. P you are going to consider Baltimore harbor, for example, or New York
or Philadelphia, there probably should be a regional sludge disposal facility, but
the question would be should that facility be operated in conjunction with some other
petroleum industry facility such as a refinery which would have a way of getting rid
of any recovered oils from processing of this sludge or should it be a separate
-------
250
authority which has no connection with the petroleum industry? The marine people
would like to give it to the refineries and the refineries don't really want it. It
is really an oil industry problem because they are carrying our materials.
J. A. Scher: This gets into the political sphere. I am not so sure the oil industry
is going to operate these kinds of facilities, I don't think they want to operate
these kind of facilities. I really don't know what the answer to that is. I don't
know whether I can answer it and give you a very good answer but it is something I
think is going to be approached on that type of a scale and who the operators and
who the contractors are I don't know, I could speculate but I am just guessing.
Morris Wiley: I guess it is an open question, I am asking the question, but not
really expecting an answer at this time.
BIOGRAPHY
Jacoby A. Scher is a Principal Environ-
mental Engineer with Fluor Engineers and
Constructors, Inc. in Houston, Texas. He
holds a B.A. in Chemistry from Rice
University and an M.S. in Environmental
Engineering from the University of Houston.
He is a Registered Professional Engineer in
the State of Texas. Mr. Scher has been
involved in the Industrial Waste Treatment
Field for ten years and has been with Fluor
since 1973.
-------
251
TABLE 1 ESTIMATED OILY SLUDGE PRODUCTION FROM AIR FLOTATION TREATMENT
Refinery Class
Class
A
B
C
D
E
Air Flotation
Influent TSS
(mg/1)
29.5
35.2
52.2
99.3
29.9
Air Flotation
Effluent TSS
(mg/1)
8.8
10.6
15.6
29.8
8.9
TSS
Removal
(mg/1)
20.7
29.6
36.6
69.5
21.0
Flow
(gal/bbl)
18.0
38.0
42.7
47.3
87.0
TSS
Removal
(Ibs/mbbl)
3.1
9.4
13.0
27.5
15.3
MEDIAN
Economics of Refinery Wastewater Treatment, American Petroleum Institute, 1973.
15.3
TABLE 2 COMPOSITION OF SLUDGE WITHDRAWN FROM OIL-WATER SEPARATORS
API Units with Removal Equipment
Average Range
(Percent by Weight) (Percent by Weight)
Water 82.4
Oil 10.9
Volatile Solids 1.4
Ash 5.3
61.2 to 98.3
I.I to 20.7
0.5 to 2.3
0.4 to 15.8
All Survey Samples
Average
(Percent by Weight)
66.5
15.0
5.5
13.0
Range
(Percent by Weight)
8.9 to 99.5
O.I to 59.0
0.0 to 49.4
0.3 to 56.2
TABLES COMPOSITION OF OIL-WATER SEPARATOR SLUDGES
(Water-Free Basis)
Average
(Percent by Weight)
Oil 46.0
Volatile Solids 14.7
Ash 39.3
Range
(Percent by Weight)
9.4 to 81.5
0.0 to 54.3
2.3 to 85.3
-------
252
TABLE 4 COMPOSITION OF AIR FLOTATION SLUDGE
(Alum Used as a Coagulant)
Wafer removed by settling I hr, percent by weight
Froth analysis, percent by weight:
Oil-free solids
Oil (by difference)
Water (after settling I hr.)
Ash, percent by weight of solids
Average
15
3
22
75
27
Range
2.5 to 42.0
1.0 to 5.2
60.8 to 90.1
3.1 to 43.8
Water
Oil
Volatile Solids
Ash
TABLES COMPOSITION OF TANK BOTTOM SLUDGES
Average Range
(Percent by Weight) (Percent by Weight)
39.4
47.9
4.4
8.3
5.0 to 98.0
1.3 to 95.0
0.0 to 13.0
0.0 to 55.3
Solid Waste
TABLE 6 PROJECTED WASTE SOLIDS INVENTORY*
Characteristics
Dry Solids % Solids Volume
Ibs/day wt/wt gal/day
1. Gravity Oil Separator Sludge
a. Oil Area 95
b. New Area 79
2. DAF Skimmings
a. Old Area 68
3. Oily Sediment 1,600
4. Tank Bottoms 2,500
1.2 1,200 Oily Sludge Containing
up to 10% Oil &I6%
1.2 1,000 Solids After Settling
0.5 1,700 Oil, Solids & Emulsion
Breaking Chemicals
Remarks
Potential Oil
Recovery
Potential Oil
Recovery by
Chemical &
Heat Treatment
up to 20 Approx.
1,000
Oily Sediment Contain-
ing Clay, Sand, &
Carbonate Precipitates
and up to 10% Oil
up to 20 1,500 Up to 50% Oil
* Excluding Tar and CBM from the HFAIkylation Unit
Potential Oil
Recovery if Leaded
& No n-Leaded
Sludges are
Segregated
-------
253
FIGURE 1. OILY SLUDGE PRODUCTION VERSUS REFINERY CAPACITY
500,000
. 200,000
^
u
a
a
a
u
-------
254
FIGURE 3
TYPICAL AUTOMATIC WATER DRAW-OFF SYSTEM
CRUDE OIL
STORAGE TANK
ixj-
SLOPE
OIL RETURN
SLOPE
OILY-WATER SEPARATOR
?M STEAM TO
HEATING COILS
t
STEAM TRAP
BALL FLOAT TYPE
EMERGENCY SHUT-OFF VALVE
FIGURE 4
RECOVERED OIJ, TREATMENT
HOT-OIL
DRY OIL
OIL
EMULSION
WATEF
-------
FIGURE 5
REFINERY WASTEWATER TREATMENT SYSTEM
3AI.TAST WATR3
RAINFALL
RUNOFF
BREAKER
EFFLUENT
POLYMER INJECTION
pi
Cn
OIL TO CRUDE TANK
-------
256
FIGURE 6
MASS BALANCE DIAGRAM
OILY WASTE SOLIDS DISPOSAL SYSTEMS
TANK
BOTTOMS
OILY
SEDIMENT
1700 ihfi./Day 13 20% = 1850 GPP
4115 Ibs./Day
li\ 5 Ihg/Day
GPD
3 15% =
SEPARATOR
BOTTOMS
DAF
SKIMMINGS
615 Ib/D
THICKENING
15% = 2890
GPD
CONTRACT
DISPOSAL
ALTERNATE
NO. I
IB/I
3 15% =
2890 GPD
INCINERATION
3500 Ib/E
DRY
SOLIDS
LANDFILL
ALTERNATE
NO. 2
FIGURE 7
OILY HASTE SOLIDS DISPOSAL SYSTEMS
ESTIMATED COSTS
THICKENING
CAP. = 349,500
0 & M = $9,000
HAUL COST AT
S92/T10 BBT..=
521,000
CONTRACT
DISPOSAL
590,150 Disposal
Costs ($ 5.09/Gal.
INCINERATION
CAP. = $200,000
C 5, H = 590,000
-------
"OILY WASTE DISPOSAL BY SOIL CULTIVATION"
C. B. Kincannon
Texas Water Quality Board, Austin, Texas
INTRODUCTION
Numerous companies dispose of oily waste materials by adding the sludge to land
areas. This method of disposal may be referred to as land disposal, land treatment, land
spreading or land farming.
Sources of oily sludges include petroleum refinery wastes, oil spills and pipe line
ruptures. Also, numerous waste materials from industrial chemical and petrochemical
plants contain oil. The quantity, quality and type or types of oil in oily sludges may
vary widely and often the oil exists as mixtures with solids. Determination of the optimum
disposal method for each batch of sludge based upon chemical and physical composition of
the sludge is not practical. Generally, the disposal of oily waste sludges on land areas
has been done on a trial and error basis. The control of the disposal site may have been
based solely upon visual observations, with a minimal of attention toward the quality of
surface water runoff or the effects on the soil or on the subsurface groundwater.
The purpose of this paper is to review a study using a soil cultivation process for
disposal of known oil materials made in 1971 under the partial sponsorship of the Environ-
mental Protection Agency (1). In this study oily materials were spread over the soil, then
mixed with the soil using a roto-tiller type cultivator. Soil microorganisms were utilized
to decompose the oil at prevailing soil and climatic conditions. Experimentally determined
oil decomposition rates and other findings during the study suggest that the process can be
controlled and is one method for disposal of oily sludges at the plant site.
OILY MATERIALS USED
Since the oil present in oily sludges from a plant may vary widely in composition and
properties, the selection of representative oils to use in a study may be somewhat arbitrary,
but should include oils similar to those which may be present in the plant waste sludges.
For the cultivation process study, oil feed materials were selected to represent different
combinations of hydrocarbon types. The properties and composition of three oils used in the
experiment are given in Table 1. The oils include the following:
crude oil tank bottoms containing a natural balance of hydrocarbon types.
bunker C (No. 6 fuel oil), a residual fuel oil containing olefinic and aromatic
components.
waxy oil, an intermediate refined product containing highly paraffinic components.
Differences in pour points, viscosity and hydrocarbon types are shown in Table 1 for
the three oil materials.
257
-------
258
TEST PLOTS
The study included three parallel experiments with each oil material. A total of nine
test plots were used. A sketch of the experimental plots is shown in Figure 1 . Test plots
were treated identically except for soil nutrient additions. Plot No. 1 was fertilized
heavily, Plot No. 2 was fertilized moderately and Plot No. 3 was not fertilized.
Each of the nine plots was 12 x 125 feet; i.e., 1,500 square feet. Plots were
separated by levees designed so that rainfall runoff from each plot drained through a pipe
to a ditch. A plug in the pipe was used to control the runoff. The levees prevented
crossflow of water from one plot to another and kept oil within the plot area.
Space between the plots permitted vacuum truck passage for convenience in spreading
oil evenly over the soil. The roto-tiller was cleaned after working each plot to prevent
mixing of oil, soil and fertilizer from one plot to another.
The experimental plots were located in an area which had been used previously for
oily waste disposal. Residual oil in the soil at the start of the experiment was approximately
ten percent by weight. Rainfall runoff was accumulated in a holding pond. The depth of
soil mixed with the roto-tiller was approximately six inches. The six inch depth was used
in calculating the volume of the soil in the test plot.
FERTILIZERS
In agricultural crop production, nitrogen and phosphorus are removed from the soil
with the harvested grain or forage. Replenishing of the nutrients is required. For a waste
disposal system such as the soil cultivation process, there should be an equilibrium condition
established which would minimize the need for supplementing the nutrients. Endogenous
metabolism, autoxidation of cellular protoplasm, results in the release of nitrogen and
phosphorus previously used for synthesis. The released nutrients are made available for
reuse in biological waste treatment systems (2). No reference information was found to
use as a guide in selection of appropriate quantities of nitrogen and phosphorus which
should be added to the biological system. Instead, the selection of fertilizer quantities
was based upon agricultural-oriented experience. Warnings that 1) excess nitrogen
fertilizer elements hinder (poison) bacterial action and 2) excessive total soluble salts may
cause unfavorable osmotic conditions for bacterial growth, were considered in the selection
of the quantities of nutrient additive. Excess phosphate was not considered toxic except
for its contribution to the total soluble salt content.
Urea and calcium hydrogen phosphate were added initially as fertilizers. Plots
numbered 1, heavily fertilized, received 1 ,000 pounds of nitrogen (N) and 200 pounds of
Phosphate (P~0,-) per acre compared to 500 pounds of nitrogen and 100 pounds of (P00,-) f°r
the moderately fertilized plots numbered 2. No supplemental nutrient was added to plots
numbered 3.
Fertilizer materials added during the course of the experiment to plots number 1 and
number 2 were based upon the results of periodic analysis of nitrogen and phosphorus in the
-------
259
soil. After about 6 months, loss of ammonia to the atmosphere was noted and a change in
the type of nitrogen fertilizer material appeared warranted. About midway of the project,
ammonia nitrate (NH ,NOo) was used instead of urea and relatively small doses applied
frequently.
EXPERIMENTAL PROCEDURE
Each plot was cultivated with a roto-tiller type plow. Fertilizer materials were added
by manually spreading solid granular fertilizer over the surface of plots No. 1 and 2.
After spreading fertilizer and before adding oily sludge materials, the plots were again
roto-tilled to mix fertilizer and soil.
Oily feed materials were transported to the test plot area with a vacuum tank truck,
and distributed over the soil surface by manual direction of a discharge hose. After
cultivating with the roto-tiller, the oil-soil mixture appeared uniform and presented no
problem at the ambient May temperature of 80 degrees F. The initial schedule for roto-
tilling was once each two weeks, weather permitting. After six months of cultivation, the
soil was friable, had lost its oily appearance, and based upon test results was suitable for
a second addition of the oily sludge. Differences in the plots were apparent. The color of
the unfertilized No. 3 plots was darker than for fertilized Plots No. 1 and 2.
The date for the second addition of oils was in early February when the temperature
was in the low 40 degrees F range. Congealing and solidification of the oils was apparent.
Mixing of the viscous oily matter into the soil was not successful until the ambient tempera-
ture had increased to about 80 degrees F.
At the end of the 18-month experiment, the soil had again returned to the friable
condition with the appearance of normal agricultural soil.
Although no attempt was made to grow vegetation on soils previously used for oil
decomposition during this experiment, native grass and plants did sprout and grow on top of
the levee. The soil in the levees was the same as the starting test plots, which contained
about ten percent by weight (10%w) of oily material.
SAMPLING AND TESTING
Oil and nutrient contents were used as controls for the experimental work. Generally,
oil determinations were made once every two weeks and nutrients were determined once
per month. Microbial analyses were made monthly, hydrocarbon types of oil added and
extracted from the soil were made for each new addition of sludge to the soil, and oil
content of core soil samples were made at the start, midway and at the end of the 18-month
experiment.
Obtaining representative samples was a recognized problem from the start of the project.
The sampling procedure included combining portions of soil from three sampling points to
form subsamples a, b, and c, as shown in Figure 2. Subsamples a, b, and c were compo-
sited to represent the whole test plot.
-------
260
EXPERIMENTAL RESULTS
The experimental results when using the three oily materials as feed were similar. The
relationships of the percent weight (%w) oil in the soil with time are shown in Figure 3 for
the waxy oil test plots and these results are representative of the other two oily material
test results. From the first addition of oil in May to about November 1970, the oil content
of the soil decreased markedly. During the period November 1970 to February 1971, a
reduction of oil was not apparent. From February, when a second dosage of oil was added
to October 1971, the oil content again decreased. The reason or reasons for the period of
inactivity, November 1970 to February 1971, may have been due to one, two or all of the
following reasons: During this period, the concentration of oil in the soil was near the
starting content and this residual oil may be nonreactive or slowly reactive with soil
microorganisms. Another reason may have been the lack of available nitrogen and a third
reason may have been the relatively low temperatures during this period.
The relationship of the nitrogen concentration and oil removal rate is shown in Figure
4 and the relationship of temperature and oil removal rate is shown in Figure 5. Jhe oil
removal rate was expressed as pounds of oil removed per cubic foot of soil (Ibs/ft ). During
the period November 1970 to February 1971, the ammonia concentration was 20 mg/l or
less for waxy oil experiments and the nitrate content was nil indicating a possible lack of
the nitrogen nutrient. Also, during this period, the soil temperature, shown in Figure 5,
was less than 40 degrees F and may have been too low for favorable bacterial growth. The
reason or reasons for the period of inactivity was not further determined.
Oil decomposition rates expressed in pounds of oil per month per cubic foot of soil
(Ibs/mo/ft ) are given in Table 2. For the periods May to November 1970, and February
to October 1971, the oil decomposition rate averaged Q.9, 0.9 and 0.7 Ibs/ft /mo for
Plots 1, 2 and 3 for crude oil; L.8, 1.7 and 0.8 Ibs/ft /mo for Plots 1, 2 and 3 for Bunker
C; and 1 .7, 1 .3 and 0.7 Ibs/ft /mo for Plots 1, 2, and 3 for waxy oil feed. For a three
month period of November 1970 to February 1971, the oil decomposition rate was minimum.
The average decomposition rate per year at these test conditions was 1.1, 1.0 and 0.6
Ibs/ft /mo for heavily fertilized No. 1, medium fertilized No. 2 and unfertilized No. 3
plots, respectively.
Tjne general conclusion was that the oil decomposition) rates were approximately 0.5
Ibs/ft of soil per month without fertilizer and 1 .0 Ibs/ft /month when fertilized. Items
for consideration in developing the cost'of the process using fertilizer are given in Table 3.
The fertilizer requirement may be assumed to be 1,000 pounds of nitrogen and 500 pounds
of phosphorus per acre per year. The cost for spreading oil and fertilizer and cultivating
the soil likely would vary with location. An oil decomposition rate of 1 .0 Ib/ft /mo is
equal to 21,780 Ibs/acre/mo or about 70 barrels (bbls)/acre/mo. The cost for delivering
the oil to the area and distributing the oil over the surface of the soil was about $l/bbl or
$70/acre/mo in 1971 . The fertilizer material and labor for spreading costs prorated on a
monthly basis was estimated to be about $23/mo. Cultivation (plowing) would require
about 8 hours (hrs)/acre at $25/hr and two plowings/mo totals $400/mo. The total cost/
acre/mo was estimated to be $493 or about $7 per barrel of oil at conditions prevailing in
1971 . The cost for disposal of oily sludge containing 33 percent oil was estimated to be
-------
261
about $3 per barrel in 1971. The cost of disposal of oily sludge would depend upon the
oil content of the sludge. Current costs for a specific entity or industrial plant would of
course require an independent evaluation.
MICROBIAL ACTION
Microorganisms in the soil decompose nitrogenous and carbonaceous materials into
microbial cellular matter. When oil or hydrocarbon is the only source of carbon, oil
degrading microorganisms survive and become the predominant species. Results of monthly
determinations of the total aerobic count indicate that the addition of fertilizer affected
the total aerobic count. The relationship of total aerobic and oil removal rate and the
effect of fertilizer are shown in Figure 6 for the waxy oil experiments. The No. 1 plot
(heavily fertilized) generally was higher in total aerobic count than No. 2 (medium
fertilized) and No. 3 (not fertilized) plots. Also, the microbial count was found to be
sensitive to change in the oil content of the soil as shown in Figure 7. The addition of
oil in May 1970 and February 1971 appeared to upset the soil microbial equilibrium and
caused the total aerobic count to be low.
The predominant species given in Table 4 included Arthrobacter, Corynebacterium,
Flavobacterium, Nocardia, and Pseudomonas. Since the area where the experimental
plots were located had been used previously for oily waste disposal, the presence of a
variety of oil consuming microorganisms might be expected in the soil. Flavobacterium
and Pseudomonas species were present but the microbial population was different on all
plots. Distribution of the microorganism species were not greatly different for the three
oils used in the experiment.
OIL EXTRACTED FROM THE SOIL
Oils added to and extracted from soil plots were quantitatively analyzed for hydro-
carbon types, qualitatively analyzed by infrared for "fingerprint" differences and
qualitatively analyzed by gas chromatography for boiling point profile. Detailed
description and results of the analysis have been reported previously (1). Some of the more
pertinent procedures and results follow.
Adsorbents of clay and silica gel were used to separate the oils into total saturates,
resins (asphaltenes, organic acid, etc.) and aromatics. The hydrocarbon content of the
fertilized plots related with time are shown on Figure 8. Generally, the saturate and
aromatic contents are shown to decrease and the resin content remained unchanged or
increased. This may indicate the saturate and aromatic materials to be decomposed and
the resinous material to be relatively nonreactive with soil microorganisms. Also shown is
the hydrocarbon content of the oil originally in the soil.
Generally from the three parallel experiments, more changes in composition for the
No. 1 and No. 2 plots were noted than for the No. 3, indicating an effect of fertilizer.
Further, differences were noted from infrared analyses.
The oil originally in the soil contained saturate groups, small amount of organic acids
-------
262
and a substantial amount of condensed multi-ring aromatic structures.
Infrared spectra for crude added to the soil and after eight months in the soil are shown
on Figure 9. Some observations follow:
1 . The crude oil feed contained saturates and simple aromatics.
2. The oil extracted from the unfertilized plot was similar to the original oil in the
soil. The aromatic were more representative of the polyaromatic groups, and
saturates were present.
3. The oil extracted from the fertilized plot still contained polyaromatic types and
some saturates but contained organic acids, also.
4. Organic acid contents are significantly different in the fertilized and unfertilized
samples. Saturate peak intensities are related to quantity and are higher in the
unfertilized sample.
In general, oils from plots that were fertilized showed higher concentrations of organic
acids and the concentration appeared to be greatest in the heavily fertilized plots.
INFILTRATION
Oil and nutrient concentrations at depths of 2, 4, and 6 feet, determined at the start,
middle, and end of the project, indicated that these constituents did not infiltrate the soil
under the conditions of the project.
RAINFALL RUNOFF
Analysis of accumulated rainwater for oil, ammonia, nitrate, and phosphate contents
indicated that little, if any, of these constituents were present when drained immediately
after a rain. Long standing appeared to cause higher oil contents.
The oil content ranged from 30 to TOO mg/1 and infrared examination of these extracted
oils indicated the soluble oil to be essentially organic acids with characteristics quite
similar with spectra for naphthenic acids.
Each plot was drained about 36 times during the project and the quantity of oil
discharged with the water was equal to about 0.03 Ibs/ft of oil in the soil.
The ammonia content of the water varied with the soil nutrient level. During periods
when the ammonia was excessive on the No. 1 plots, discharge water contained up to 150
mg/1 ammonia as N. Nitrates and phosphates were generally absent in the discharge.
APPLICATION OF THE PROCESS
This study demonstrated that soil microorganisms decompose petroleum oily waste and
that oil and fertilizer chemicals did not penetrate the soil at the location and conditions
of the test. However, numerous factors are left unresolved and future experiments should
gain knowledge concerning the following uncertainties:
-------
263
1 . The residual oil in the soil after the 18-month project contained organic acids,
and those soluble in rainfall runoff water were primarily naphthenic acids. Since
most organic acids are water soluble, this suggests that naphthenic acids do not
decompose in the soil as rapidly as aliphatic acids or that naphthenic acids are
formed from aliphatic or aromatic materials at the soil conditions. Speculatively,
large concentrations of naphthenic acids may inhibit microbial growth in the soil,
and water washing of these organic acids from the soil to a separate biotreatment
system might improve the oil decomposition rate in the soil.
2. Residual oil extracted from the soil was characterized by infrared to be poly-
aromatic oils, suggesting this hydrocarbon group to have a slow reaction or to be
nonreactive for microbial decomposition at the prevailing conditions. Certain,
and yet unknown, environmental conditions might improve the decomposition rate
of the polyaromatics.
3. Accumulation and buildup of metals or salts contents with long-range usage may
affect the microorganism activity. The acidity or alkalinity range, optimum for
microbial action, may vary with the soil contents.
4. The time required for native soil bacteria to become acclimated oil decomposers
likely depends upon the soil composition and temperature.
For application of the process, one should follow the suggested action items given in
Table 5.
CONCLUSIONS
The decomposition rate oLoily sludges (hydrocarbons) by microbial action in cultivated
soil averaged about 0.5 Ibs/ft /mo without fertilizers and about 1 .0 Ib/ft /mo when
fertilizers were added.
Major bacterial species included Pseudomonas, Nocardia, Flavobacterium, Coryne-
bacterium, and Arthrobacter.
Organic acids are produced and water solubles appear to be naphthenic acids.
Oil and fertilizer did not infiltrate.
Rainfall runoff water contained organic acids (and possibly fertilizer).
The soil cultivation process is one method which can be used to dispose of oily waste
materials.
REFERENCES
(1) Oily Waste Disposal by Soil Cultivation Process. EPA-R2-72-11D, Office of Research
and Monitoring, Environmental Protection Agency, December (1972)
(2) Manual on Disposal of Refinery Wastes, Volume on Liquid Waste, Chapter 13,
Biological Treatment, First Edition, 1969, American Petroleum Institute, Division of
Refining, New York, N. Y.
-------
264
DISCUSSION
Banks: Has this soil been returned to cultivation and if so, what were the crop yields?
C. B. Kincannon: This experiment did not include agricultural crop production. In the
study, the soil was utilized as a bacterial bed for the decomposition of oils. I cannot answer
your question concerning the current use of the experimental area because I am no longer
associated with Shell Oil Company where the work was done. I refer you to Shell Oil
Company, Deerpark, Texas for further information along these lines.
Morris Wiley: We have used an alternate method which does not involve cultivation whereby
you put 0.5 cm of oil sludge over, say 1 square mile and find that not much happens during
the cold season in say New York. During the summer and certainly by September, the oil is
weathed completely into little "pebbles" which the earth worms cover up. You can put
5,000 cubic meters per square kilometer which is a much lower application than you are
discussing. The difference is to eliminate the cost of having to cultivate, till, fertilize,
etc. Would you care to comment on the potential savings of such a method?
C. B. Kincannon: I believe we should know more about the material put on the soil.
Morris Wiley: I should have described that. It was a refinery disposal pit sludge stabilized
for some years leaving principally the heavy ends - 20% oil, 20% water, the rest primarily
solids.
C. B. Kincannon: With my limited knowledge, I would say if that works - power to you -
~ ~
Morris Wiley: Do you think that would be cheaper if the land area were available?
C. B. Kincannon: If you can place the waste on the land area, not fertilize, till, etc. and
the waste will disappear without damage to the receiving water or create an air pollution
problem, the disposal cost should be a minimal cost.
Morris Wiley: Would you think that you would have to collect the land run-off in an
impoundment?
C. B. Kincannon: I would recommend that.
R. A. Farnham: Could you explain the meaning of that No. 9 you had up there - the 25
milligrams per liter, I thought we were about in the 30% oil range?
C. B. Kincannon: As I pointed out, we found from exploratory experiments in the
laboratory that concentrations in excess of 30 percent weight oil caused the oil decomposition
reaction rate to become slow or nil. When the soil particle becomes completely enclosed
with oil globules, water is prevented from getting into the soil and water is a necessary
element for the bacteria.
-------
265
BIOGRAPHY
C. B. Kincannon is an engineer in the Central
Operations Division of the Texas Water Quality
Board (TWQB). He obtained a B.S. degree
from Panhandle A and M College (now Panhandle
State University) and was an employee of Shell
Oil Company for 30 years before joining the
TWQB staff. He is a certified Grade A sewage^
works operator and a professional engineer
registered in the State of Texas.
TABLE 1 "OIL PROPERTIES"
Crude Oil Bunker C Waxy Oil
Lb/Gal 7.12 8.52 7.08
Pour Point (°F) -5 40 95
Viscosity
SU 60°F 60 19000
SF 122°F 120 59
HC Type (Wt%)
Saturates 36 18 90
Resins 8 26 0
Aromatics 56 56 10
-------
266
TABLE 2 "OIL DECOMPOSITION RATE LB/MONTH/CU FT OF SOIL1
Nov 70 to
Feb 71
Min
Min
Min
Min
Min
Min
Min
Min
Min
Feed
Crude
Oil
Bunker
C
Waxy
Oil
Average
Plot
1
2
3
1
2
3
1
2
3
1
2
3
May to
Nov 70
0.7
0.5
0.2
1.8
1.2
1.8
1.8
1.5
0.8
Feb to
Oct 71
1.1
1.4
1.3
1.8
1.5
0.5
1.6
1.0
0.5
Avg
0.9
0.9
0.7
1.8
1.7
0.8
1.7
1.3
0.7
TABLE 3 "ITEMS FOR ESTIMATING COST"
Oil Spreading/Month
Lb/Cu Ft
Fertil izer
Nitrogen Lb/Year
Phosphorus
Cultivation
Quantity
1
70
1000
500
Yearly
Avg
0.7
1.3
1.3
0.6
1.3
0.9
0.5
1.1
1.0
0.6
TABLE 4 "MICROBIAL ANALYSES'
Predominant Species
Arthrobacter
Corynebacterium
Flavobacterium
Nocardia
Pseudomonas
TABLE 5 "FOR APPLICATION OF THE PROCESS1
1 . Prepare Area
2. Provide Treatment Facilities For Rainfall Runoff
3. Obtain Samples To Determine Background Level
4. Add Fertilizer Material
5. Cultivate
6. Distribute Oil Over Surface
7. Cultivate, Analyze For Oil
8. At Regular Intervals, Cultivate and Analyze
9. Oil Content Of Soil Should Not Exceed 25% wt.
10. Determine Oil Penetration Rate Into Soil
-------
267
FIGURE I
EXPERIMENTAL PLOTS
EACH PLOT
12 FT x 125 FT
1500 FT2
750 FT3
TO HOLDING POND
1
^-^
1
J 1
2
3
i
— -
t
1
2
3
t
-—
t
1
2
3
f N
L
CRUDE OIL
BUNKER C WAXY OIL
( NO 6 FUEL OIL)
FIGURE 2
TYPICAL SAMPLING POINTS
SUB SAMPLE c
SUB SAMPLE b
SUB SAMPLE a
-------
268
FIGURE 3
OIL IN SOIL
V)
LJ Od
O Q
en
LJ
a.
30
20
10
30
20
10
30
20
10
HEAVILY
FERTILIZED
MODERATELY
FERTILIZED
UNFERTILIZED
APR
JULY OCT
JAN
APR JULY
OCT
FIGURE 4
EFFECT OF FERTILIZER
800
E
a.
CL
~
LU
O
O
400
200
Q
>d
oo
APR
3 — =
OCT
APR
OCT
APR
-------
269
OIL REMOVED
(Ibs/Ft3 SOIL)
SOIL TEMPERATURE (°F)
->l CO
o o
m
-n
-n
m
o
0
TJOl
m
-I
m
OIL REMOVED
(Ibs/Ft3 SOIL)
— ro
o o
o
o
TOTAL AEROBES (xlO7)
IN)
O
00
o
o
o
CD <*)
m
o
o
-------
270
FIGURE 7
EFFECT OF OIL ADDITION
ON MICROBIAL COUNT
o 120
X
~ 100
m 80
o
LJ 60
^ 40
Ł 20
/ i
/ !
I
JL \
r^-^
30
20
APR
OCT
APR
OCT
-------
271
FIGURE 8
HYDROCARBON TYPES
30
20
10
0
h- 30
X
o
g 20
I-
5 10
o
cc
w ^
CL 0
30
20
10
n
—
—
—
•»
\
S
R
A
I
R
A
c
f
A
S
R
A
S
R
•—•™
A
S
B
A
I-SATUR,
*-RESIN<
k-AROMAl
X.
V.
x»
•^
•x.
^
S,
\
\
^:\
•m. '
**-.
X
X
X
•x
X
X
Al
%
rn
s
R
tMita-Bi
A
§
R
A
S
R
A
TES
CS
\
X
\
••»
x»
"* *•»•.
X
X
X
\
X
^s
1''
MJ
\
\
X
N
t^t^^
••K-^
-«-
\
^
•s
*"***»
"* ^
X
X
N
\
\
\
X
\
X.
—
CRUDE OIL
BUNKER C
WAXY OIL
APR OCT APR OCT
-------
FIGURE 9
INFRA-RED SPECTRA
CN
r\
CN
UJ
o
(O
<
tr
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
OH
00
C-0
=0 CH2 u"u OH
C=C CH3 AROMATIC AROMATIC
CRUDE OIL
UNFERTILIZED
FERTILIZED
6 8 10 12
WAVE LENGTH (MICRONS)
14
-------
THE DISPOSAL OF OILY WASTES BY LAND FARMING
R. L. Huddleston
and
L. W. Cresswell
Continental Oil Company, Ponca City, Oklahoma
At CONOCO we have been studying the disposal of oily wastes by land farming for
about three years. By "land farming" we mean the use of the soil environment for the
planned, orderly treatment of waste. The work has been strictly experimental in nature,
but has been conducted in field plots out-of-doors using actual waste materials. In this
presentation we will describe the need for land farming as an oily waste disposal method,
the results of our studies to date, and our view of factors that must yet be addressed before
land farming can be used routinely on an environmentally safe basis for oily wastes disposal.
SOURCES AND CHARACTERISTICS OF OILY WASTES
Oily wastes are generated by numerous routine refinery operations. Examples of such
wastes and their compositions are shown in Table 1. Amounts of these wastes generated are
not well defined and vary widely from refinery to refinery and from time to time within a
given refinery. However, it is probable that a 100,000 BPD refinery generates from 5,000
to 100,000 bbl/year of such wastes. As increasingly better jobs are done in waste water
treatment and air pollution control, the amount of oily wastes that are generated will
inevitably increase.
Some oil can be recovered from these wastes by chemical and physical means, and it is
very important that we learn how to more effectively and economically deoil such wastes.
It is unlikely, however, that we will ever be able to economically deoil all wastes of these
types.
The characteristics of waste oil prevent its decomposition in conventional waste treat-
ment systems. Some disposal methods used in the past are shown in Table 2. For various
reasons these methods are rarely used today, and we obviously need suitable alternatives.
At present, it appears that incineration and land farming may be the only safe and
possibly economic ways to dispose of oily waste. Although incineration is expensive, a
fire hazard, poses air pollution problems, and is difficult to adapt to wastes containing
non-combustible solids and high water content, it may have to be used at locations where
land is economically unavailable. An alternative and less expensive disposal method may
be land farming.
THE CONCEPT OF LAND FARMING
It is well known that many microorganisms can decompose numerous petroleum components.
We must also recognize, however, that petroleum is an extremely complex blend of tens
of thousands of chemical structures. Our knowledge of how, and to what extent, this
complex blend is biodegradable, is quite small indeed. Nevertheless, many workers have
273
-------
274
shown that the bulk of at least some crude oils is biodegradable (1,2,3).
Soil contains an enormous microbial population as shown in Table 3. Many of these
organisms can decompose hydrocarbons (4). Although the water insolubility of petroleum
constituents (see Table 4) adversely affects biodegradation rate, this property helps to "fix"
oily wastes in place when they are added to the soil. The extremely large surface area
offered by the soil environment (see Table 5) helps spread added oil into thin films making
the oil more vulnerable to microbial attack.
A simplistic scheme of oily wastes disposal by land farming is shown in Table 6. We must
quickly point out that the mechanism shown represents what we "hope" will occur and has
not been proven as yet. Exceptions may be waste oil constituents that are not completely
mineralized and others that may prove to be recalcitrant to degradation.
At present we need the answers to a number of questions to define the suitability of land
farming for oily wastes disposal. Some of these questions are posed in Table 7. At present,
we have partial answers to several, and there is almost certainly justification for using land
farming prior to having complete answers to all of these questions. If we are diligent in
providing sufficient data to show that oily wastes land farming is reasonably safe from an
environmental standpoint, we can logically expect little valid opposition to its use in the
near future. More complete answers and answers to some of the less urgent questions can be
ascertained as land farming is practiced.
CONOCO'S EXPERIMENTAL LAND FARMING STUDIES
Billings, Montana. At a site near Billings, Montana, we have studied the land farming
disposal of oily pond bottoms. A view of the area prior to preparation is shown in Figure 1.
The test site and surrounding area was examined to assess the plant and animal life, soil
microbial populations and soil composition. Numerous 4^ foot cores were taken to document
subsurface soil characteristics. The surface soil was found to be a sandy clay containing a
low humus level and very small amounts of nitrogen and phosphate. The Billings climate is
arid, warm in the summer and very cold in the winter.
The test area was diked with subsoil, plowed and disked. The oily waste, described in
Table 8, was spread as evenly as possible on the soil surface using a vacuum truck. We later
found that by using a large diameter hose instead of a truck to spread the waste, we could
prevent the heavy trucks from compacting and generally disturbing the test plot soil and get
more even waste distribution.
After application, the waste was allowed to dry and weather for three weeks. This step
changed the waste to a crumbly consistency, making it much easier to work into the soil.
Even during this period no significant odor problem was evident. The waste was then
blended into the upper 6-inches of soil with a tractor-mounted rototiller. The appearance
of the area at this point is shown in Figures 2 and 3. A description of the cultivation,
fertilization and sampling pattern used is shown in Table 9.
The test plots were fertilized with nitrogen and phosphate. This was done because the
soil and waste contained very small amounts of these nutrients, and we did not want to limit
-------
275
rate of oil degradation by these controllable factors. An increased soil microbial population
must be preceded not only by available carbon, but also by available nitrogen and phosphate.
These nutrients were added as a mixture of urea and 15-15-15, the only fertilizers available
at the time.
Thus, the soil at the start of this land farm study contained the waste constituents as
shown in Table 10.
During the course of the study, oil contents were measured using a benzene-methanol
soxhlet gravimetric technique. A Leco combustion furnace method was used to determine
total organic carbon content. Paraffin, aromatic and resin-asphaltene contents of oil
extracts were estimated by taking the extracts up in tetrahydrofuran and subjecting them to
thin layer chromatography.
Over an 18-month exposure period, we observed a 70% oil loss from the soil as shown
in Figure 4. As also illustrated, the loss was apparently created by an 88% loss of the
paraffins initially present in the oil, an 85% loss of the aromatics and only a 2.5% loss
of the resin-asphaltenes fraction. The original oil contained 50% paraffins, 30% aromatics
and 20% resin-asphaltenes. After 18-months of exposure in the soil, the residue contained
only 20% paraffins and 15% aromatics, while the resin-asphaltenes fraction had increased
to 65%.
Unfortunately, measurements of the organic carbon content of the soil samples produced
very scattered results as shown in Figure 5. While the scatter is more pronounced than
usual in this case, similar scatter is not uncommon when examining field soil analytical
data. Sample differences are magnified by the difficulty in producing thorough soil-
waste blending using only practical field cultivation techniques. Scatter can be minimized
by compositing large numbers of samples, very careful composite blending and sub-sample
taking. We have found it necessary to accept considerable data scatter to economically
conduct our studies. In spite of the scatter, the TOC data does seem to confirm that the
oil was not only degraded but that much of it was mineralized completely to CO,,.
Accompanying data over the 18-month period showed that the pH of the soil did not
change, there was little nitrogen and phosphate loss, and the soil bacterial population
increased about 10-fold.
As expected, climatic conditions over the l:8-month period of soil exposure were not
entirely conducive to good microbial activity. As can be seen from Figure 4, the air
temperature during 10 of the 18-months of exposure was about 10 C or less. In addition,
the soil moisture content ranged between only 9 and 18% over the 18-month interval.
For optimum soil microbial activity, the soil moisture content should be at least 30% (5).
It should also be emphasized that the oily waste studied came from a waste water treatment
pond (bottoms). As a result, the waste had been exposed to microbial attack for an
extended period prior to land farming. Undoubtedly, the most readily biodegradable
substances had been removed from the waste prior to land farm exposure.
It should also be mentioned that crested wheat was grown on part of the test area.
-------
276
Visual observations judged the plant growth to be normal and identical to the control (no
waste oil) plot crop.
The average rate of waste oil loss in the Billings study is shown in Table 1 1 . Losses are
shown for the 18-month period and on an annualized basis.
Ponca City, Oklahoma. At Ponca City, we have been studying the land farming of an
oily (crude petroleum) sand waste and, to a limited extent, the land farming of a refinery
oily waste. Unfortunately, the refinery waste was inadvertently added to a soil that was
later found to have been previously contaminated with oil and salts. This problem has
complicated evaluation of certain portions of our studies.
Methods used for test site preparation, waste application and analytical work were similar
or identical to those used in the Billings study. The oily sand was extremely viscous and
could not be flowed onto the plots. Due to the small size of the oily sand test plots, we
were able to manually distribute the waste over them.
Views of fhe oily sand and refinery waste plots are shown in Figures 6 through 9. Test
plots were fertilized after waste weathering and soil waste blending.
The composition of the refinery oily waste is shown in Table 12. Addition of this waste
to the soil produced the waste-in-soil levels shown in Table 13. The oil in the refinery
waste was found to contain 21% paraffin, 49% aromatics and 30% resins-asphaltenes
(Table 14). Amounts of nitrogen (as NHL) in the fertilized test plots ranged from 50 mg/kg
soil to 320 mg/kg soil and phosphorus (as PO.) ranged from 10 to 50 mg/kg soil. To date
we have noted no detectable effect of fertilizer level, within these concentration limits,
on the rate of oil degradation. Even in the most heavy fertilized plots, we have not found
over 20 ppm NO,-,.
Over a 24-month exposure period, we have measured at 35% loss of oil from the test plots.
Composition of the oily sand that we have been investigating is shown in Table 15. In
oily sand test plots ammonium nitrogen and phosphate phosphorus levels have been kept above
50 and 30 mg/kg, respectively. Soil water contents have stayed around 15% (about 30%
of moisture holding capacity) during exposure and we have observed no difference in pH
between test plots and control plots.
Soil core samples, 30 inches deep, were taken directly beneath test plots and in central
areas one year after oily waste was applied to the test areas. Results of extractable oil and
organic carbon measurements on these cores are shown in Table 16. As can be seen, these
data indicate that the waste has been strictly held within the shallow soil zone in which it
was originally placed. For these tests we selected plots to which we had added 8 and 10
weight % oil to the soil in order to examine most severe test conditions.
As may be seen in Figure 10, we have observed about 40% degradation of oil in these
oily sand test plots over a 22-month period. Actually this expression is an average of what
we observed over several plots. A plot that initially had 8% oil in the soil was found to
-------
277
e experienced a 41% oil loss and several 5% oil plots showed an average of 45% oil
. This small difference, of course, has a sizable effect on mass oil lost. This is illu-
have
loss
strated in Table 17.
Oil composition measurements (see Figure 8) have shown that about 82% of the paraffins
were lost along with 60% of the aromatics. Only 1% of the resin-asphaltenes fraction was
lost. This oil initially contained 22% paraffins, 28% aromatics and 50% resin-asphaltenes.
The 22-month residue consists of over 80% resin-asphaltenes.
OVERALL PERSPECTIVE OF LAND FARMING DISPOSAL OF OILY WASTES
Several companies have been practicing various forms of land farming as a means of
oily waste disposal for many years. While most have relied on visual oil loss as evidence
of success, several investigators have examined the technical aspects of land farming in
recent years. Among those that have carried out such investigative studies are Shell Oil
Company, Union Carbide and Sun Oil Company. A summary of oil and oily wastes studied
and reported rates of oil degradation are shown in Table 18. We have taken the liberty of
converting all reported rates into annualized figures and identical units of measure so they
can be more easily compared.
All of these data support the contention reached long ago with little or no analytical
data; crude petroleum and many of its components are biodegraded in the soil. Of course,
we still don't know to what extent such materials are degraded in the soil. This we must
learn. We have also some data now that oil in the soil can be degraded at reasonable
rates in climates as cold and as arid as Montana, as well as in warmer and more humid
climates such as Houston and Corpus Christi. We have also learned that if we can
mechanically handle the presence of an initial oil addition of 10% in the soil, the rate
of oil degradation is greater. Here at CONOCO, we found that 5% additions are much
more manageable, especially during the initial stages of waste oil and soil blending. We
now have data that indicates that as little as 50 ppm nitrogen and 20 ppm phosphorus
(109 Ib. and 44 Ib. per acre, respectively) are sufficient to support the degradation of
some oils at their maximum rates in some climates. We also have data that indicates oil
in soil can be decomposed at reasonable rates even when the water content of the soil
stays below 20%. We have evidence that waste oil tends to stay tightly bound in soil to
which it is added while degradation is occurring. There is some data that indicates normal
crops can be grown in soil where oily waste is being decomposed. All of these data are
encouraging for those of us who hope we can prove the acceptability of land farming as an
oily waste disposal technique.
There remains, however, a number of very important factors that must be more clearly
established before we can safely recommend the full scale use of land farming for the
disposal of oily wastes. These factors are listed in Table 19. We believe that until these
factors are reasonably well established, the disposal of oily wastes by land farming must
be very carefully controlled and safeguarded against any unforeseen environmental hazards.
-------
278
ECONOMICS OF OILY WASTE DISPOSAL BY LAND FARMING
At present it appears that land farming represents a relatively inexpensive disposal
method. An example of what costs might be are shown in Table 20. We are aware that
costs will vary widely with land cost, labor costs, amount of analytical work required, type
of oily waste applied and other factors. In this example we have used a land cost of
$l,000/acre, costed borrowed capital at 10% for 10-years and assumed very small fertilizer
requirements. It is obvious that we need more clearly defined bases for establishing oily
wastes land farming costs.
In summary, we are encouraged by data that indicates land farming to be an effective,
safe and economic disposal method for at least some oily wastes. We believe, however,
that considerably more investigative effort must be expended before the concept can be
safely used on a wide scale basis. Finally, it is imperative that sound data be utilized to
construct reasonable but precise ground rules for the effective, safe and economic use of
land farming for oily wastes disposal.
REFERENCES
(1) Atlas, R. M. and R. Bartha, "Simulated Biodegradation of Oil Slicks Using Oleophilic
Fertilizers," Environmental Science & Technology, Vol. 7, No. 6, pp. 538-541 June
(1973).
(2) Bartha, R., "Biodegradation of Oil on Water Surfaces," NTIS, AD-DOOO 722, March 13
(1975).
(3) Kator, H., et al, "Microbial Degradation of a Louisiana Crude Oil In Closed Flasks and
Under Simulsated Field Conditions," Proceedings of the 1971 Conference on Prevention
and Control of Oil Spills, Washington, D. C.
(4) Crow, S. A., et al, "Microbiological Aspects of Petroleum Degradation in the Aquatic
Environment," La Mer, Bulletin de la Soc. Franco-Japonaise d'Oceanographic, Vol.
12, No. 2, pp. 95-112, (NTIS COM-75-10298, May 1974) (1974).
(5) Alexander M., Introduction to Soil Microbiology, John Wiley & Sons, Inc., New York
(1961).
(6) Lai, M.G., etal, "Determination of the Molecular Solubility of Navy Oils in Water,
NTIS, AD-784414, June 6 (1974).
(7) Ministry of Defense Working Party, "The Fate of Oil Spills at Sea," NTIS, AD-763042
(1973).
(8) Kincannon, C. B., "Oily Waste Disposal by Soil Cultivation Process," Environmental
Protection Technology Series, EPA-R2-72-110, December (1972).
(9) Raymond, R. L., et al, "Assimilation of Oil by Soil Bacteria," Preprint No. 24-75,
API Division of Refining Meeting, Hyatt Regency O'Hare, Chicago, III., May (1975).
(10) Francks, H. C., et al, "Disposal of Oil Wastes by Microbial Assimilation," NTIS, Y-
1934, May (1974).
-------
279
DISCUSSION
John Smith: This is not a particular question, but just a general comment concerning the
sol ids wastes, the problem facing the petroleum industry. As most of you know the EPA
has contracted a study for the potentially hazardous waste and that document is being
prepared to reference the guidelines. They have now advertised for contractual bids
another study in the same line reference to the investigation of various methods of elimina-
tion of these hazardous wastes. The first study was predicated on the incineration, land
farming, etc. and now they are coming out asking contractors to look into other methods
of waste disposal from petroleum operations. The point being here that EPA is very aware
of and is going to be spending money reference to solids waste disposal so it is a very timely
topic.
Robert Huddleston: I will speak for all of us - I hope they are successful in finding more
economical and acceptable disposal methods.
R. A. Farnham: We have been doing some land farming experiments and you mention soil
water contents. Now you are talking weight percent water rather than percent field
capacity, are you?
Robert Huddleston: Yes, for the soils we have examined, the water holding capacities tend
to average about 60 weight %. Thus, when these soils contain 30% moisture, the moisture
content is condusive to good microbial activity.
R. A. Farnham: That's for clay?
Robert Huddleston: Yes, clay type soils.
Anonymous: For those of us who have trouble with numbers, your number describing Shell's
rate of oil degradation experience was 165 g oil lost per kg of soil per year. Was that
fertilized or unfertilized?
Robert Huddleston: I used the fertilized rate in the table.
-------
280
BIOGRAPHIES
Lewis W. Cresswell is a Research Scientist in
the Environmental Group of Continental Oil
Company's Research and Development Department.
He has a B.S. degree in Chemical Engineering
from Mississippi State University. Prior employ-
ment includes Continental's Process Engineering
Environmental Group and the U.S. Environmental
Protection Agency, Denver Region.
Robert L. Huddleston is supervisor of the
Environmental Research Group in the Research and
Development Department of Continental Oil
Company, Ponca City, Oklahoma. He holds B.S.
and M.S. degrees in Microbiology from the
University of Oklahoma and has taken additional
studies from Oklahoma University and Oklahoma
State University in Microbiology and in Environ-
mental Science. He is currently Vice President
of the Society For Industrial Microbiology. Bob
has been a member of Conoco R&D for almost 20
years and has authored more than 25 papers
dealing with various aspects of industrial
microbiology and waste treatment.
-------
281
TABLE 1 "SOURCES OF OILY REFINERY WASTES'
Crude Oil Tank Bottoms
Slop Tank Bottoms
API, CPI Separator Bottoms
Biotreatment Solids
Coker Slowdown Solids
Water Impoundment Pond Bottoms
Typical Weight % Composition
Water
15
64
36
96
20
40
Solids
45
4
47
4
40
40
Oil
40
32
17
0.4
40
20
TABLE 2 "HISTORICAL DISPOSAL OF OILY SOLIDS'
Burning
Landfill
Surface Pitting
Road Surfacing
Ocean Dumping
Burying
TABLE 3 "MICROBIAL DISTRIBUTION IN SOIL1
(TAKEN FROM ALEXANDER, 1961) (5)
Organisms Per Gram of Soil
Depth
(cm)
3-8
20-25
35-40
67-75
135-145
Aerobic
Bacteria
7,800,000
1,800,000
472,000
10,000
1,000
Anaerobic
Bacteria
1,950,000
379,000
98,000
1,000
400
Actinomycetes
2,080,000
245,000
49,000
5,000
-
Fungi
119,000
50,000
14,000
6,000
3,000
Algae
25,000
5,000
500
100
-
TABLE 4 "HYDROCARBON SOLUBILITY IN WATER"
(TAKEN FROM MING, 1974, AND "FATE OF OIL SPILT AT SEA," 1973) (6,7)
PPM in Distilled Water
Benzene
Xylene
Octane
Anthracene
C Paraffin
Cnrysene
Gas Oil
Kerosene
Bitumen
1800
175
1
0.075
0.002
0.002
0.0003 - 0.00000008
0.0001 ,-0.2
-------
282
TABLE 5 "SOIL PARTICLE SIZE AND SURFACE AREA1
(TAKEN FROM ALEXANDER, 1961) (5)
Diameter No. Particles Surface Area
mm Per g Soil sq cm/g
Fine Gravel 2.00-1.00 90 11.3
Sand 1.00-0.05 722-722,000 22.7-227
Silt 0.05 - 0.002 5,780,000 454
Clay 0.002 90,300,000,000 11,300
*Based on spherical shapes and maximum diameters
TABLE 6 "LAND FARMING OILY WASTE"
Waste Analysis
Land Selection and Preparation
Waste Application
Fertilizer Application
Cultivation and Other Care
Oil + 00 + N + PO. Microbial Microbial Cells + COO + H0O
2 4 Activity 2 2
TABLE 7 "QUESTIONS ABOUT LAND FARMING OILY WASTE SOLIDS'
1 . What type of waste can be applied?
2. How long does it take for the oil to be decomposed?
3. Is there any oil residual? How much? What is it?
4. Are there any environmentally harmful materials formed?
5. Will the oil or other waste components migrate through the soil?
6. Mechanically, how is the waste best applied and blended with the soil?
7. How much fertilizer must be added?
8. Is subsequent cultivation required?
9. Should plant growth be established on the site?
10. How do different climatic conditions affect the process?
-------
283
Analysis
TABLE 8 "BILLINGS REFINERY WASTE POND BOTTOMS"
WT % Metal Content mg/kg Metal Content
m
g/kg
•-
Extractable Oil 15
Water 72
Solids 13
Total Organic Carbon 10
Silicon
Calcium
Chromium
Sodium
Barium
Lead
Vanadium
TABLE 9 "TREATMENT OF BILLINGS
0 1 2
Cultivation R(l) R R
Fertilization 1
Sampling (3) 115
(1) R= Rototiller
(2) D = Disc
(3) Numbers refer to samplings per month
TABLE 10 "INITIAL BILLINGS LAND
Analysis WT %
Extractable Oil 0.93
TOC 1.2
Months
3 4
4 2
12,900
10,900
1,900
810
370
50
10
Iron 12,000
Aluminum 4,300
Magnesium 890
Zinc 690
Copper 360
Nickel 10
LAND FARM TEST PLOTS"
of Exposure
7 12
1 1
FARM SOIL-WASTE COMPOSITION
Analysis
Nitrogen
Iron
Chromium
Zinc
Lead
mg/kg
130
720
115
41
3
13 14 15 17 18
D (2) D D
4 413
DUE TO WASTE ADDITION"
Analysis mg/kg
Phosphate 20
Calcium 650
Sodium 50
Copper 22
-------
284
TABLE 11 "OIL LOSS - BILLINGS LAND FARM"
18 Months Annual ized
15,840 Ib/Acre Soil
7.2gAg Soil
10,560 Ib/Acre Soil
4.8gAgSoil
TABLE 12 "PONCA CITY REFINERY TEST OILY WASTE1
Analysis
Extractable Oil
Water
Solids
Total Organic Carbon
Analysis
Extractable Oil
Water
Solids
Total Organic Carbon
WT%
Metal Content
fig/kg
Metal Content mg/kg
39
24
37
53
TABLE
WT% of
4.8
3.0
4.6
6.5
Silicon
Aluminum
Lead
Sodium
Zinc
Barium
Copper
Manganese
Vanadium
13 "PONCA CITY REFINERY
Soil Metal Content
Silicon
Aluminum
Lead
Sodium
Zinc
Barium
Copper
Manganese
Vanadium
21,000
3,200
1,400
870
400
240
150
130
21
WASTE IN
mg/kg of
2,600
390
170
107
50
30
18
16
3
Calcium 13
Iron 3
Magnesium
Potassium
Chromium
Strontium
Titanium
Nickel
Cadmium
SOIL"
Soil Metal Content
Calcium
Iron
Magnesium
Potassium
Chromium
Strontium
Titanium
Nickel
Cadmium
,000
,000
880
480
370
180
150
21
3
mg
mg/kg of Soil
1,600
370
108
60
46
22
18
3
0.4
-------
285
TABLE 14 "OIL COMPOSITION IN PONCA CITY REFINERY WASTE SUBJECTED TO LAND FARM STUDY'
Analysis WT %
Paraffins 21
Aromatics 49
Resins-asphaltenes 30
TABLE 15 "PONCA CITY OILY SAND WASTE"
Analysis VVT %
Extractable Oil 27
Water 4
Solids (sand) 69
Total Organic Carbon 10
Metal Content mg/kg Metal Content mg/kg
Sodium 2,000 Calcium 1,000
Magnesium 260 Chromium 90
Copper 90 Lead 90
Vanadium 90 Nickel 130
TABLE 16 "CORE ANALYSES - OILY SAND LAND FARM TEST PLOT AND CONTROL AREAS"
Sampling A Sampling B
Core Depth Control Test Area Control Test Area
Extractable Oil 0-6" 0.15 4.79 0.11 6.35
(Wt. %) 6-9" 0.20 0.19 0.05 0.10
9-12" 0.09 0.17 0.02 0.05
12 - 15" 0.09 0.08 0.03 0.05
15 - 18" 0.11 0.18 0.03 0.05
18 - 24" 0.14 0.05 0.05
24 - 30" 0.06
Organic Carbon 0-6" 1.90 5.74 1.48 6.15
(Wt. %) 6-9" 0.79 0.79 0.87 1.53
9- 12" 0.26 0.31 0.93 1.04
12 - 15" 0.29 0.37 0.81 0.91
15 - 18" 0.25 0.60 0.62 0.79
18 - 24" 0.78 0.52 0.60
24 - 30" 0.61 0.56
-------
286
TABLE 17 "OIL LOSS % AND RATE - PONCA CITY OILY SAND LAND FARM STUDY"
Initial % % Oil Loss Mass of Oil Lost
Oil In Soil Per Year Per Year
25 27,500 Ib/acre soil (12 g/kg soil)
22 38,720 Ib/acre soil (18 gAg soil)
TABLE 18 "REPORTED RATES OF OIL LOSS DURING LAND FARM STUDIES"
(TAKEN FROM SUN, SHELL, UNION CARBIDE, NEW ZEALAND DATA) (8,9,10)
Gm Oil Lost Approximate Initial Oil in Soil
Per kg/soil/year Concentration (wt. %)
Refinery Oily Waste
CONOCO - Billings 7 1
Ponca City 9 5
Shell - Houston 165 10
Petroleum Crude Oil
CONOCO - Ponca City 17 5
Sun - Marcus Hook, Tulsa, Corpus Christi 14+2 1
Lube Oil Waste
Sun - Marcus Hook, Tulsa, Corpus Christi 18+5 5
New Zealand 27~ 2
38 5
69 7
454 9
Waste Vacuum Pump Oil
Union Carbide - Oak Ridge 171 7
#6 Fuel Oil
Shell - Houston 311 10
Sun - Marcus Hook, Tulsa, Corpus Christi 20+7 2
-------
287
TABLE 19 "FACTORS TO ESTABLISH PROPER USE OF LAND FARMING"
1. Firm establishment that oily waste is not mobile with respect to waste composition, soil composition
and climate.
2. Final extent of waste oil decomposition in the soil as related to oil composition.
3. Demonstration that no dangerous substances are present or produced when oily wastes are land
farmed.
4. What types of vegetation can be grown on oily waste land farms and what effects does such growth have
on rate and extent of oil degradation.
5. Need precise nitrogen and phosphate requirements for oily wastes land farming.
6. Establishment of ground rules for safe and economic application of land farming for oily wastes
disposal.
TABLE 20 "ESTIMATED COST OF LAND FARMING OiLY WASTES"
100 tons of waste
25% oil
Application limited by oil degradation
Oil degradation rate - 15 g/kg soil/yr.
Farm life - 10 years
Annual Cost ($) % of Total
Land 355 16
Site Preparation 540 24
Diking, Fencing, Cultivation
Waste Hauling (5 miles) 275 13
Waste Spreading 160 7
Fertilizer, Spreading 115 5
Cultivation of Soil-waste (4/yr) 160 7
Samples - Analytical Support 600 27
TOTAL COST $2,205
Cost per ton of waste: $22.05
-------
288
FIGURE 1
Photograph of Billings' Land Farm Area Prior to Waste
Spreading
FIGURE 2
Photograph of Billings' Land Farm Area After Waste Tilled In
.-vf
" - .
..
• :. ~-;r-"-«»-
• " .-» ^* Tit^ •
^Ł*&&,
. . ,
,
"
^ .-;,
*
-------
o
FIGURE 3
Photograph of Billings' Land Farm Area After Waste
Tilled In - Close up
289
H
Z
UJ
(A
UJ
or
Q.
o
100
80
60
40
20
FIGURE 4
BILLINGS WASTE OIL LAND-FARM STUDY
ASPHALTICS
4
100
80 =1
O
60 Z
UJ
40 0=
u.
o
20 ^
10
12 14
16 18
MONTHS OF EXPOSURE
u
o
^**
a.
UJ
ir
<
30
20
10
-IQL
-------
290
o
UJ
a:
o
o
100
80
60
40
20
FIGURE 5
TOO - BILLINGS LAND FARM
I I
0
8 10 12
14
16 18
MONTHS OF EXPOSURE
FIGURE 6
Photograph of Ponca City Refinery Waste Land Farm Area
Prior to Preparation
- •
%* -,** ... .
3
•
-------
FIGURE 7
Photograph of Ponca City Refinery Waste Land Farm Area
After Waste Addition
291
- -*w
-*-w~ .^.
'iS
•
-.
'
FIGURE 8
Photograph of Ponca City Refinery Waste Land Farm Test
Area - Wheat Growth
-------
292
us
CO
UJ
cr
Q.
o
00
FIGURE 9
Photograph of Ponca City Oily Sand Land Farm Test
Sys tern
W:x?^jr-jag ^ _ ^
FIGURE 10
PONCA CITY CRUDE OIL LAND-FARM STUDY
100
-ilOO
80
J
0
60
40
U.
0
20 -
20
18 20 22
O
o
30
20
MONTHS OF EXPOSURE
x
o
(D
- 10
-------
SESSION V
"ACTIVATED CARBON TREATMENT"
Chairman:
Nicholas D. Sylvester
Professor of Chemical Engineering
University of Tulsa, Tulsa, Oklahoma
Speakers
Davis L. Ford
"Current State of the Art of Activated Carbon Treatment"
C.T. Lawson
"Cautions and Limitations on the Applications of Activated
Carbon Adsorption to Organic Chemical Wastewaters"
Joyce A. Rizzo
"Case History: Use of Powdered Activated Carbon in an
Activated Sludge System"
Leon H. Myers
"Pilot Plant Activated Carbon Treatment of Petroleum
Refinery Wastewater"
M.A. Prosche
"Activated Carbon Treatment of Combined Storm and Process Waters'
293
-------
294
BIOGRAPHY
Nicholas D. Sylvester
Nicholas D. Sylvester is Professor of Chemical
Engineering and Director of the Environmental Protection
Projects program at the University of Tulsa. Dr.
Sylvester received his B.S. degree from Ohio University,
and his Ph.D. from Carnegie-Mellon University both in
Chemical Engineering. Before coming to the University
of Tulsa he taught at the University of Notre Dame.
Nick is a member of the American Institute of Chemical
Engineers, Society of Petroleum Engineers, American
Chemical Society, Society of Plastics Engineers, Society
of Rheology and the American Society of Engineering
Education.
Professor Sylvester is currently conducting research
in the following areas: two-phase flow, drag reduction,
environmental protection, chemical reaction engineering
and improved oil recovery. Dr. Sylvester has more than
50 technical publications and has been principal investi-
gator of funded research totalling nearly $600,000.
In addition to his professional interests, Nick is
active in youth athletic programs, having coached junior
high basketball and elementary school baseball and
soccer. He is also an avid, though, inept, golfer.
-------
"CURRENT STATE OF THE ART OF ACTIVATED CARBON TREATMENT"
Davis L. Ford, Ph.D., P.E.
Senior Vice President
Engineering-Science, Inc., Austin, Texas
The treatment of wastewaters using activated carbon has received wide attention for
several years - more recently catalyzed by the development of effluent quality guidelines
pursuant to Public Law 92-500. These effluent quality criteria and the accompanying
development documents prominently mention carbon as an applicable and attractive
treatment concept, particularly in the Best Available Treatment Economically Achievable
(BATEA) process mode currently stated as necessary to produce the 1983 quality level
objective. Until recently, most of the literature has dealt with exploring theoretical
concepts and documenting experimental results. As more information on full-scale opera-
tions is becoming available, however, it is considered appropriate to review the current
state of the art of activated carbon treatment - both in municipal and industrial sectors.
It is the purpose of this treatise, therefore, to present pertinent and current information
relative to the activated carbon treatment of municipal and industrial wastewaters. A
brief discussion of adsorption concepts and carbon characteristics also is included.
ADSORPTION CONCEPTS AND THEORY
Molecules are held together by cohesive forces ranging from strong valence bonds to
the weaker van Der Waals forces of attraction. These attractive forces are satisfied in the
solid phase interior molecules, having the ability to capture certain fluid molecules as
they contact the surface, van Der Waals forces are the bases for the adsorption of waste-
water constituents onto carbon which has been activated to maximize this interphase
accumulation of liquid constituents at the surface or interphase of the solid phase.
The rate at which substances are removed from the liquid phase (adsorbate) to the
solid phase (adsorbent) is of paramount importance when evaluating the efficacy of acti-
vated carbon as a wastewater treatment process. Unfortunately, the task of quantifying
the many forces acting at the solid-liquid interface is a formidable one. Developing a
mathematical expression which describes the dynamic phenomenon occurring in a continuous-
flow/fixed-bed reactor has been difficult because of multi-variable influences. The overall
adsorption rate represents the combined effects of diffusion through a laminar layer of fluid
surrounding the constituent, surface diffusion, and adsorption on the internal pore surfaces.
Most mathematical solutions for equations which describe concentration/time profiles are
limited to the special case in which only one of these phenomena controls the overall rate
of adsorption (1).
One expression for the continuous-flow regime assumes the diffusion of the constituent
through the liquid phase and through the pores of the carbon (which are rate-limiting),
then combining these resistances in an overall mass coefficient term. Using this rationale:
= k2r(Cs'C)
295
-------
296
where:
q = flow rate
C = concentration of the adsorbate
s
D = adsorbent bed depth
C = equilibrium adsorbate concentration
k r = overall mass transfer coefficient
A more convenient expression of Equation 1 is in terms of the adsorbate rate with
respect to the weight of the carbon in the columns:
qdC/dM = ly/X (Cs-C) (Eq. 2)
where:
M = weight of the carbon in the column
X = packed density of the carbon in the column
Another proposed model predicts four successively decreasing adsorption rates would be
observed as the adsorption proceeds to equilibrium. The initial rate would be limited by
the rate of adsorbate transfer across the film layer, film diffusion, or, if sufficient turbulence
existed, control would be exerted by the combined rate of external surface adsorption and
macropore filling. After the external surface adsorption capacity was exhausted, there would
exist three secondary adsorption rates controlled, respectively, by the filling of the macro-
pore (an effective radius of 5,000 to 20,000 A°), the transitional pore (20 to 100 A°), and
the micropore (10 to 20 A effective radius). This model is illustrated in Figure 1 (2). It
is inherent in this model that the intraparticle transport occurs as a series of adsorption/
desorption steps, each linear with respect to time, and their summation resulting in a time/
linear function.
The development of adsorbate removal kinetics on a batch basis can be used to appro-
ximate carbon effectiveness and predict organic residual levels. The adsorption isotherm
is used for this purpose and is defined as a functional expression for the variation of adsorption
with concentration of adsorbate in bulk solution at a constant temperature. The isotherm is
expressed in terms of removal of an impurity - such as BOD, COD, and color - per unit
weight of carbon as a function of the equilibrium impurity remaining in solution. Linear
plots as shown in Figure 2 can be expressed in terms of the empirical Freundlich equation.
This expression relates the amount of impurity in the adsorbed phase to that in solution:
-fi = KC1/n (Eq. 3)
where:
X = amount of impurity adsorbed
M - weight of carbon
C = equilibrium concentration of impurity in solution
K,n = constants
The Freundlich isotherm is valid within the context of a batch test for pure substances
and some dilute wastewaters. As shown in Figure 2, its application is limited in certain
-------
297
cases when a significant portion of the organic impurities are not amenable to sorption,
resulting in a constant residual, regardless of the carbon dosage.
The constants "n" and "K" can be used to define both the nature of the carbon and
the adsorbate. A high "K" and "n" value, for example, indicate good adsorption through-
out the concentration range studied. A low "K" and "n" value would infer low adsorption
at dilute concentrations with high adsorption at the more concentrated levels. Variations
of the constants for selected wastewaters are shown in Table 1 (3).
FACTORS WHICH INFLUENCE ADSORPTION
There are many factors which influence both the rate and magnitude of adsorption -
underscoring the difficulty in developing predictive models which would apply to all
complex wastewaters. A brief discussion of the more important factors is presented herein.
Molecular Structure. The molecular structure, or nature of the adsorbate, is partic-
ularly important in dictating the degree of adsorption that can actually occur. As a rule,
branched-chain compounds are more sorbable than straight-chain compounds, the type and
location of the substituent (functional) group affects adsorbability, and molecules which
are low in polarity and solubility tend to be preferentially adsorbed. Unless the screening
action of the carbon pores actually impedes, large molecules are more sorbable than small
molecules of similar chemical nature. This is attributable to more solute chemical bonds
being formed, making desorption more difficult.
Inorganic compounds demonstrate a wide range of adsorbability. Disassociated salts -
such as potassium chloride and sodium sulfate - are essentially nonsorbable. Mercuric
chloride and ferric chloride are relatively sorbable, and iodine is one of the most adsorbable
substances known. Generally, however, a significant reduction in inorganic materials is
not expected in carbon systems.
Organic compound sorbability can be classified to some extent. Primary alcohols and
sugars, for example, are resistant to adsorption, while ethers and certain organic acids
are highly sorbable. Recently published experimental data presented in Table 2 are
indicative of the sorbability of many organic compounds (4). Additional sorbability data
conducted independently are presented in Table 3.
Solubility. An increase in solubility acts to oppose the attraction of the adsorbate to
carbon. Thus, polar groups which have a high affinity for water usually diminish adsorption
from aqueous solutions. Conversely, the greater adsorption of the higher aliphatic acids
and alcohols is attributed in part of their relatively lower solubility in an aqueous solution.
There are exceptions to this, as in the case of the highly soluble chloracetic acid (5).
lonization. lonization is generally adverse to adsorption by carbon as strongly-
ionized materials are poorly adsorbed. Hydrogen ions, which are significantly adsorbed
under some conditions, would be an exception to this. Some negative ions, therefore, are
more sorbable when associated with hydrogen ions. For this reason, mineral acids - such
as sulfuric acid - are sorbable at higher concentrations.
-------
298
A change in ionization can drastically affect adsorption. A low pH, for example,
promotes the adsorption of organic acids whereas a high pH would favor the adsorption of
organic bases. Phenol adsorbs strongly at neutral or low pH while the adsorption of the
phenolate salt at a high pH is poor. The optimum pH is therefore solute-specific and must
be determined for each wastewater.
Temperature. As adsorption reactions are generally exothermic and high temperatures
usually slow or retard the adsorption process, lower temperatures have been reported to favor
adsorption (1,5). Very little information has been presented, however, which documents
significant shifts in adsorbability within the temperature range of 65 F to 90 F (typical of
most wastewaters). Lower temperatures should increase adsorption, but the effect in aqueous
solutions is very small.
Adsorption of Mixed Solutes. Most wastewaters contain a myriad of compounds which
may mutually enhance, interfere, or act independently in the adsorption process. Factors
which affect overall adsorption of multiple adsorbates include the relative molecular size
and configuration, the relative adsorptive affinities, and the relative concentrations of the
solutes (1). Predictive models obviously require validation for complex wastewaters, as
extrapolation from investigations using synthesized wastes containing controlled concen-
trations of selected adsorbates may not reflect all of the interactions occurring in the waste.
A summary of the factors which potentially influence adsorbability is presented in
Table 4.
PROPERTIES OF ACTIVATED CARBON
Activated carbons are made from a variety of materials including weed, peat, lignin,
bituminous coal, lignite, and petroleum residues. Granular carbons produced from medium
volatile bituminous coal or lignite have been most widely applied in the treatment of
wastewater as they are relatively inexpensive and readily available. The activation of
carbon is essentially a two-phase process which includes burning off the amorphous decom-
position products and enlarging the pores in the carbonized material (6). The burn-off, or
carbonization, phase involves drying the carbon at approximately 170 C, heating the
material to 270 C to 280 C with the evolution of carbon monoxide, carbon dioxide, and
acetic acid, and, finally, completing the carbonization process at a temperature of 400 C
to 600 C. The yield following carbonization is approximately 80 percent. The intermediate
product is then activated by using carbon dioxide or steam at a temperature of 750 C to
950 C, burning off decomposition products, exposing and widening the pores in the development
of macroporous structure. In the activation process, the kind of adsorptive powers developed
are determined by:
1 . the chemical nature and concentration of the oxidizing gas;
2. the extent to which the activation is conducted;
3. the temperature of the reaction; and
4. the amount and kind of mineral ingredients in the char (5).
The proper activation conditions provide an oxidation which selectively erodes the
surface so as to increase the surface area, develop greater porosity, and leave the remaining
-------
299
atoms arranged in configurations that have specific affinities.
Activated carbons from coke, coal, or lignite have specific properties, depending on
the material source and the mode of activation. Property standards are therefore helpful
in specifying carbons to perform a specific task.
As a rule, granular carbons made from calcined petroleum coke have the smallest
pore size, the largest surface area, and the highest bulk density. Lignite carbon has the
largest pore size, least surface area, and the lowest bulk density. Bituminous coal has a
bulk density equal to that of petroleum coke and an average pore size and surface area
somewhere between those of petroleum coke and lignite-based carbons (7). A brief
description of carbon properties follows:
Total Surface Area; This is the surface area of carbon expressed in square meters per
gram, normally measured by the adsorption of nitrogen gas by the BET method (8).
Carbon Density: Apparent density is the weight in grams of one ml of carbon in air.
Bulk density, backwashed and drained, is often used and is usually expressed in
pounds per cubic foot.
Particle Size Distribution: The particle size distribution is critical in terms of
hydraulic loading and backwash rates. Commonly manufactured particle size ranges
for granular activated carbons expressed in limiting U.S. Standard Sieve Sizes
include 8 x 16, 8 x 30, 10 x 30, 12 x 40, and 20 x 40. Effective sizes (sieve
opening at which 10 percent of the material passes) range from 0.55 mm to 1 .30 mm.
In general, the uniformity coefficient (the millimeter opening at which 60 percent
of the material passes divided by the millimeter opening at which 10 percent of
the material passes) for granular activated carbon should not exceed 2.1.
Adsorptive Capacity: The best measure of adsorptive capacity is the effectiveness of
the carbon in removing the critical constituents (BOD, COD, color, etc.) from the
wastewater in question. Various tests, however, have been developed to give
relative removal capacities of activated carbon under specific conditions. Phenol
number is used as an index of a carbon's ability to remove taste and odor compounds;
tannin is representative of organic compounds added to water by decayed vegetation;
and iodine and molasses numbers are used to show if a carbon is activated. The
iodine number, defined as the milligrams of iodine adsorbed by one gram of carbon
(with an iodine concentration in the residual filtrate is 0.03) is probably the most
widely used method of expressing carbon capacities. It generally can be correlated
to the ability of an activated carbon to adsorb low molecular weight substances
while the molasses number correlates the carbon's ability to adsorb higher molecular
weight substances. The iodine number measures the micropores having an effective
radius of less than 20 angstroms and the molasses number measures the transitional
pores ranging from 20 to 500 angstroms.
These are the principal general parameters used in specifying carbons, the objective
being the selected carbon
-------
300
1 . has the adsorption capability to meet the effluent requirements;
2. incurs minimal losses occurring during carbon transport and regeneration;
3. has good hydraulic characteristics with respect to head loss or pressure drop; and
4. represents the most cost-effective media to accomplish the prescribed task.
These, of course, represent general parameters of specification and should be augmented by
test data and process requirements developed from bench- or pilot-scale evaluation using
representative wastewater samples and selected carbons.
Comparative properties of the most widely used carbons in wastewater treatment -
those from lignite and bituminous coals - are shown in Table 5 (6); a broader presentation
of properties from commercial carbon manufacturers also is tabulated in this Table (9).
Typical properties of a petroleum-base powdered carbon developed but not yet marketed by
AMOCO are shown in Table 6.
REGENERATION
Because of economic and solid waste disposal considerations, it is generally more
feasible to regenerate spent carbon for subsequent reuse than to dispose of it. In the
regeneration process, the objective is to remove from the carbon porous structure the
previously-adsorbed materials, thus reinstituting its ability to adsorb impurities. There are
several modes of regeneration which can be applied including thermal, steam treatment,
solvent extraction, acid or base treatment, and chemical oxidation. Of these, only
thermal regeneration using a multiple-hearth or rotary-tube furnace is widely applied in
wastewater treatment. The discussion on regeneration therefore centers around thermal
treatment.
Thermal regeneration refers to the process of drying, thermal desorption, and high
temperature (1,200 F to 1,800 F) heat treatment in the presence of limited quantities of
oxidizing gases such as water vapor, flue gas, and oxygen (10). Multiple-hearth furnaces
are the most commonly used, although rotary kilns or fluidized-bed furnaces are occasionally
applied.
The sequential stages in thermal regeneration are shown in Table 7. Of these steps,
the gasification stage is the most critical. The system should be controlled in order to
selectively gasify the sorbed organic material while minimizing the gasification of the
carbon structure. Basically, there are three major variables involved in thermal regenera-
tion. These include furnace temperature, residence time, and the carbon loading. Of
these three, furnace temperature is controllable, although it may take several hours to
adjust, the residence time can be changed by varying the rabble-arm speed in a multiple-
hearth furnace, or the rotation rate and slope of the tube in a rotary kiln. Very little can
be done to change the carbon loading, which affects the severity of regeneration required.
The multiple-hearth furnace is the most commonly used system for granular carbon
regeneration. A schematic diagram of a typically-designed multiple-hearth furnace is
shown in Figure 3. The wet spent carbon is added at the top of the furnace and drops from
hearth to hearth, being raked along by the rabble arms. The temperatures shown for the
various hearths are typical gas temperatures for granular carbon regeneration. It was
-------
301
found that the addition of steam on hearths four and six gave a more uniform distribution
of temperatures throughout the furnace. The effect of steam is to reduce the apparent
density and increase the iodine number of the regenerated carbon (6). Normally, about
one pound of steam per pound of carbon is used. Natural gas or fuel oil is usually added
to supply the auxiliary heat. Although the fuel requirement varies, generally about
3,000 BTU/lb of carbon and 1,300 BTU/lb of steam generated is required.
Rotary kiln furnaces are not being installed in new facilities but many are still in use.
The kilns are usually direct-fired, counter-current units with steam injection with up to
10 percent excess air. Again, natural gas or fuel oil is commonly used as auxiliary fuel.
The heat efficiency of rotary kilns is less than that of multiple-hearths.
Fluidized-bed systems for powdered carbon regeneration have been evaluated on a
pilot scale and appear to have promise (11, 12, 13). In this process a bed of inert
granular material such as sand is fluidized by the upward flow of hot gases and the wet
spent carbon is injected directly into the bed. The inert bed particles provide a reservoir
of heat which is rapidly transferred to the spent carbon particles. The heat economy of
the fluidized-bed furnace is less favorable than the multiple-hearth furnace or the rotary
kiln - providing afterburners are not required.
There is considerable debate as to the effect of regeneration on carbon capacity,
carbon losses during the regeneration cycle, and hydraulic characteristics of the regen-
erated carbon. Much of this is focused on the regeneration effects on lignite as compared
to bituminous carbon. The manufacturers of lignite claim that, after a number of regen-
eration cycles, both carbons tend to become more like each other and perform the same (9).
Specifically, lignite - while softer and lighter than bituminous - is less susceptible to
change during regeneration and can be regenerated under less severe conditions (lower
temperature and shorter residence time). The bituminous manufacturers note that lignite
carbon losses per cycle are higher than bituminous, indicating a higher operating cost,
and bituminous adsorption capacities based on molecular weights ranging from phenol at
94 to dextran at 10,000 are significantly higher than lignite (14). It is not the intent in
this writing to favor one over the other, but simply to present information from manufacturers
of both types. Based on discussions with independent evaluators of bituminous and lignite,
several trends in thinking evolve. First, few full-scale plants using lignite are currently
in operation, necessitating an evaluation based on pilot-plant studies. There appears to
be little firm evidence that bituminous carbon capacities are higher than those of lignite,
although screening tests should be conducted in any event to determine the optimum carbon
to perform a specific task. Somewhat higher losses have been noted during regeneration
of lignite as compared to bituminous (six to nine percent for lignite, and four to seven
percent for bituminous over three cycles). Lignite requires lower temperature for its
regeneration and more skilled regeneration operation and control. There has been little
evidence to show significant hydraulic differences as demonstrated by head loss charac-
teristics in the two carbons following regeneration. Carbon selection is therefore predi-
cated on process test results, capital cost requirements, and an annualized cost evaluation.
-------
302
EXPERIENCE IN CARBON TREATMENT OF MUNICIPAL WASTEWATERS
Treatment of municipal wastewaters using activated carbon has been practiced on a
full scale for over 10 years. For this reason, ample performance records are presently
available for a realistic evaluation of this application, both in the physical/chemical and
biological effluent polishing modes of treatment. A summary of these systems has been
tabulated elsewhere, but it is the intent in this section to document selected municipal
facilities and discuss salient features of these systems which are pertinent to this overall
discussion on carbon. The two effluent polishing systems selected for discussion are the
South Lake Tahoe and Colorado Springs facilities. The physical/chemical discussion
centers around the new plant being constructed for the Cleveland Regional Sewer District,
with the Garland, Texas; Rosemount, Minnesota; and Pomona, California plants also
included. A syllabus of these selected case histories is presented as follows.
South Tahoe Tertiary Treatment System. The first major and most widely publicized
application of activated carbon in treating domestic wastewater was the system constructed
for the South Tahoe Public Utility District in 1965. The plant was conceived to polish the
effluent from an existing activated sludge facility to a quality level which would have
little or no impact on the receiving waters in an ecologically-sensitive area.
The facility consists of a chemical mix-coagulation, precipitation, and clarification
unit, an ammonia stripping tower (which has been used infrequently), a recarbonation and
settling basin, mixed-media filters, carbon adsorption, and final chlorination. A simplified
flow diagram is shown in Figure 4 (15). The water quality at various points in the process
following 18 months of operation is presented in Table 8 (16). The more recently reported
water quality at the various stages of treatment is presented in Table 9 (9). The carbon
(bituminous) efficiency per regeneration period at the South Tahoe Plant is cited in
Table 10 (6).
It should be recognized that this facility was a demonstration project, involving expensive
capital and annual expenditures. With the possible exception of nitrogen, the effluent
quality should represent the best level obtainable in treating a domestic wastewater effec-
tively using maximum biological treatment polished by chemical treatment, filtration, and
carbon adsorption.
Colorado Springs Tertiary Treatment System. As at Tahoe, the Colorado Springs
treatment system involves chemical treatment, filtration, and carbon adsorption for polishing
a slipstream of biologically treated effluent. There are basically two regimes of historical
data from this facility. The first was when the slipstream was taken from an overloaded
trickling filter effluent (the flow diagram schematically illustrated in Figure 5). The second
regime follows the addition of a contact stabilization facility to the biological treating
component of the system. This indicates that the influent to the polishing portion of the
system, the design data for which are presented in Table 11, has a lower concentration of
biodegradable organics in the second regime than in the first (17). This is detected in a
general comparison of the effluent quality observed during the two periods. Effluent quality
data from a three-month period in 1972 indicate an approximate level obtained during the
first regime. The average quality values for the reactor-clarifier influent and effluent,
-------
303
filter effluent, lead carbon column effluent, and polishing carbon effluent are tabulated
in Table 12. The data averaged during the second regime of operation are shown in
Table 13 (17). Considerable improvement is noted in the comparison, underscoring the
need for adequate biological pretreatment if the tertiary system is to realize maximum
performance. It is noted that the Tahoe and Colorado Springs (second regime) effluents
are relatively similar - both plants have adequate biological pretreatment. The deteriora-
tion of effluent quality when this is not the case is evidenced in Table 12.
The Colorado Springs facility is using a bituminous coal activated carbon, using a
multiple-hearth furnace regeneration system. As in the other carbon systems, there has
been sulfide generation problems in the carbon vessel. This problem has been ameliorated
by adding a copper sulfate solution (100 mg/l as CuSOJ to the top of the carbon columns
and altering the backwashing frequence to remove biological solids.
Cleveland Regional Sewer District Physical/Chemical System. A physical/chemical
wastewater treatment plant of which activated carbon is an integral part is presently being
constructed for the Cleveland Regional Sewer District (CRSD) at the Cleveland Westerly
plant site. This system is designed to receive an average flow of 50 MGD with peak flows
of 100 MGD. A flow diagram of this system is shown in Figure 6. Extensive pilot-plant
treatability studies were conducted prior to finalizing the design as there is a significant
industrial contribution in the raw waste load (18). Moreover, assurance of meeting the
NPDES requirements of 20 mg/l BOD (30-day average) and 30 mg/l (seven-day average)
was required. These studies, combined with those more recently conducted by CRSD,
provided some interesting results with respect to carbon adsorption.
The raw waste load of the Westerly collection system documented during the two
pilot studies (1970-71 and 1974-75) is shown in Table 14. The strength of this wastewater,
combined with the organic nature of the industrial component, created some difficulty in
meeting the required permit levels during the pilot-plant evaluation. Although these
results could be overly-pessimistic for various reasons, CRSD evaluated several process
alterations with the objective of improving process efficiency and effluent quality. Most
of this investigation centered around the use of ozone injection - both as a post-carbon
and as a precarbon mode of operation. The postozonation step was originally conceived
as strictly a disinfection step, although some further reduction of BOD through ozone
oxidation was anticipated. However, studies indicated the ozone requirement for disin-
fection to the required level of 200 fecal coliforms per 100 ml ranged from 2.6 to 10.9
mg/l (19). The higher doses were attributed to the ozone demand from the sulfides
generated in the carbon columns and ferrous bicarbonate resulting from iron slugs in the
influent. This ferrous species was solubilized in the recarbonation step and was not
altered while passing through the carbon columns. As the higher demands of ozone are
economically prohibitive, reducing the ozone-demanding constituents or looking closer
at chlorine was merited. Moreover, postozonation did not improve organic effluent
quality - as indicated in Table 15. Ozonation prior to the carbon columns was then
applied by CRSD with encouraging results. This quality improvement in terms of a
frequency analysis of carbon column effluent BOD is shown in Figure 7. The same trend
in terms of COD is shown in Figure 8. These results infer that some nonsorbable compounds
are converted to more sorbable intermediates through ozone transformations. This possibility
-------
304
is substantiated in the literature, documenting ozone oxidation of aldehydes, ketones, and
alcohols to carboxylic acids, the product being more sorbable than the reactants (20, 21).
Based on these results, the possibility of injecting ozone prior to the carbon columns is
being seriously considered because of the potential advantages of:
1. possibly transforming nonsorbable compounds to sorbable compounds, enhancing
carbon adsorption efficiency;
2. reducing the bacterial load to the carbon columns through particle sterilization; and
3. increasing the dissolved oxygen level in the carbon column influent, reducing the
possibility of sulfide production and anaerobic bacterial activity.
There may be problems concerning this approach, such as the technical problems involved
with injecting ozone in the pressurized carbon influent line, the high decomposition rate of
ozone, and enhancing corrosion potential in the carbon reactors. Moreover, disinfection
following carbon adsorption still needs to be resolved. However, the advantages presently
appear to outweigh the disadvantages and this approach is being seriously considered as a
process modification.
Garland Physical/Chemical System. A physical/chemical treatment plant is presently
being constructed at Garland, Texas. This 30 MGD plant is being designed to produce an
effluent having less than 10 mg/l BOD and TSS. The raw wastewater quality used for design
is presented in Table 16 (22). The flow diagram for this facility is shown in Figure 9. The
carbon basins are common-wall concrete facilities, each having a surface area of 950 square
feet and a depth of 10 feet. Nine basins can treat the 30 MGD, while the remaining basin
is off-line for the backwashing or regeneration. The wastewater is in contact with the carbon
for a minimum of 30 minutes and the upflow rate is 2.5 gpm/ft at design flow. A 14.5 foot
multiple-hearth furnace is designed to regenerate 80,000 pounds of carbon daily. Based on
the treatability studies, an effluent quality of 15 mg/l COD and 10 mg/l BOD is predicted
(22).
Rosemount Physical/Chemical System. The Rosemount, Minnesota physical/chemical
facility has been in operation since 1974. The system receives between 0.3 and 0.6 MGD
and consists of screening, chemical clarification using either lime or ferric chloride,
prefiltration, upflow activated carbon adsorption, postfiltration, clinoptilolite ammonia
exchangers, and disinfection. During the first year of operation, the average BOD removal
was 90 percent, the average effluent BOD concentration being 23 mg/l. More recently,
however, operating problems have been reduced and the reported effluent BOD has been in
the range of five to 10 mg/l (23).
Pomona Physical/Chemical Pilot Plant. A pilot-scale physical/chemical system has
been operating for 27 months in Pomona, California, under the auspices of the Los Angeles
County Sanitation District and the Environmental Protection Agency. The objectives of
this study included an evaluation of the long-term effectiveness of granular activated
carbon in the removal of soluble organic matter from chemically-clarified municipal waste-
water, controlling hydrogen sulfide generation in the carbon columns, and determining the
effects of repeated thermal generations on carbon characteristics and performance.
A flow diagram of the pilot system is shown in Figure 10, the design data for which are
presented in Table 17 (24). The primary organic control parameters used in these studies
-------
305
were total chemical oxygen demand (TCOD) and the dissolved chemical oxygen demand
(DCOD). A summary of the overall performance of the physical/chemical system is
presented in Table 18. These quality levels represent average values observed during the
27 months of study. For comparative purposes, the effluent quality from an 8.0 MOD
activated sludge system treating the same wastewater is also included. As noted in Table
18, the major portion of pollutants, with the exception of DCOD, was removed in the
clarification phase.
The problems of excessive biological growths in the columns, particularly the
biochemical reduction of su I fates, has persisted in many of the case histories cited. A
review of this problem was therefore an integral part of the Pomona study. As micro-
organisms have the hydrogen acceptor preference of molecular oxygen, nitrate, sulfate,
and oxidized organics, it follows that an environment having the presence of sulfates and
organics and the absence of molecular oxygen and nitrates would favor the anaerobic
sulfate reducers. The sulfide levels in the effluent approached as high as six mg/l. This
was a major concern as sulfide production was correlated directly to net head loss in the
column (head loss before the daily backwash minus the head loss after backwash). The
fact that the column head loss was caused primarily by abundant biological growths in the
column is underscored by the graphical relationship depicted in Figure 11 (24). Several
methods of ameliorating the biological proliferation and sulfide production were under-
taken. This concluded oxygen addition, intensive air/water backwash, chlorination, and
sodium nitrate addition to the carbon column at an average dosage of 5.4 mg/l (as N) was
determined to be the most effective in terms of retarding sulfide generation in the carbon
column.
The spent carbon used in the regeneration studies was backwashed, dewatered to
approximately 50 percent moisture, conveyed to a six-hearth furnace, and regenerated at
temperatures ranging from 1,650 F to 1,790 F. Steam in the amount of 0.6 Ib/lb of
carbon was added to the lower two hearths to enhance the regeneration. The regeneration
cycle normally took from 53 to 66 hours to complete. The effects of regeneration on the
physical properties of the carbon are shown in Table 19 (24). A reduction in the iodine
number (a measure of the extent to which the micropores have been cleared) is noted,
although the molasses number (a measure of the extent of macropore clearing) is essentially
unchanged. The ash content of the carbon, which measures the amount of calcium and
other inorganic residues picked up by the carbon during service increased over 60 percent
from the virgin level during the first regeneration. The ash increase during the second
regeneration was less, however, with a decrease during the third regeneration. The over-
all carbon regeneration loss (bituminous coal) ranged from 2.5 to 6 percent with an average
loss of 4.3 percent over three regenerations.
Summary. A summary of the municipal wastewater treatment using the tertiary appli-
cations of activated carbon has been published recently and is shown in Table 20 (6). A
similar summary for a straight physical/chemical application of carbon is shown in Table
21 (6) and in Table 22 (25).
EXPERIENCE IN CARBON TREATMENT OF INDUSTRIAL WASTEWATERS
Although ample carbon treatment data from full-scale facilities are becoming available
-------
306
from the municipal sector, much of the performance data in industrial applications evolves
from pilot-plant studies. There are several full-scale carbon systems treating industrial
wastewaters, however, and this section includes results from pilot studies and selected
operational systems. Specifically, results from pilot-plant studies - primarily in the refinery
and petrochemical sector - will be cited, as well as the full-scale systems at the ARCO
Refinery near Wilmington, California, the Reichhold Chemicals system in Tuscaloosa,
Alabama, and the British Petroleum Refinery in Marcus Hook, Pennsylvania.
Pilot Carbon Studies. There have been pilot studies conducted in the industrial sector
evaluating activated carbon as both a total (physical/chemical) process and as an effluent
polishing (tertiary) unit. Nine of these studies have been conducted by the author for
various petrochemical and petroleum refining facilities (26). The efficiency ranged from
50 to 86 percent COD removal, as noted in Table 23. It should be recognized that these
results were obtained using virgin carbon with controlled hydraulic and feed rate regimes.
The attendant problems of full-scale operation and the effect of using regenerated carbon
on process efficiency should be considered when translating these results to what might
occur in an operating carbon treatment system.
A comprehensive pilot-plant study treating petroleum refinery effluents by activated
carbon was conducted recently by the Environmental Protection Agency (27). Both API
separator effluent and biologically-treated effluent were charged to the columns in order
to obtain a comparative evaluation. The results of this study in terms of BOD and COD
removal are plotted in Figures 12 and 13, respectively. It is apparent that, when operated
in parallel, the biological system was more effective in removing BOD and COD -
particularly the former. This is consistent with the results observed in pilot studies
conducted by the author. It is also noted that carbon adsorption following biological
treatment was particularly effective in reducing both the BOD and COD to low levels. The
residual COD is in the same range as that cited in Table 23 when the carbon application
mode and influent COD levels were similar. The complete results of this study are shown
in Table 24. It is noted that there is no removal of cyanides or ammonia although there
was a reduction of the cited organic constituents, particularly phenols. There was surpris-
ingly good removal of chromium, copper, iron, and zinc although the exact mechanisms
of removal were not determined. The carbon capacity observed in the columns was 0.31
Ib TOC removed/lb of carbon, while the isotherm determined capacity was 0.12 Ib/lb.
This difference was attributed to biological activity observed in the column. The carbon
regeneration activity analysis is reported in Table 25, the change in iodine and molasses
numbers showing trends similar to those observed in previous studies. The investigators
emphasized that these data were generated using virgin carbon in the columns, and cautioned
that the presence of iron or aluminum salts present in the effluent could have a deleterious
effect on the carbon through the regeneration cycle. They hypothesized that aluminum
salts can remain on the surface of the carbon during regeneration, reducing the effective
surface area of the carbon and reducing its adsorption capacity. Moreover, iron salts can
catalyze oxidation reactions of the carbon and the gases during regeneration, thus perma-
nently damaging the carbon structure (27).
Extensive pilot-plant studies have been conducted recently by Union Carbide evaluating
activated carbon as a tertiary process treating effluent from an activated sludge plant
-------
307
receiving petrochemical wastewaters (28). The objectives of this study were both to
establish a basis of justification for applying activated carbon and to optimize its design.
Controlled organic and hydraulic application rates were applied to pilot-scale carbon
columns. The results of the various throughput rates are listed in Table 26. The first
verity which can be deducted from this Table is the fact that, as the activated sludge
improves in quality, the efficiency of the adsorber in terms of COD or soluble organic
carbon (SOC) improves. The improved efficiency with decreasing hydraulic throughput
rates is also apparent. There was still a 30 to 50 percent COD residual in the adsorber
influent attributable to organic compounds which are neither biologically degradable nor
sorbable on activated carbon. There was a high BOD residual at the same throughput
rate, although this residual was reduced to 35 to 50 percent of the adsorber influent at the
uneconomical throughput rate of 0.15 bed volumes per hour. This BOD residual includes
low molecular weight-oxygenated organics such as aldehydes, ketones, alcohols, glycols,
and other polar organics which are adsorbed to a very limited extent. Other observations
from this study are cited as follows:
1. breakthrough curves for multi-component wastewaters tend to be poorly defined
and sporadic rather than sharp sigmoidal wavefronts;
2. in multi-bed series adsorbers, the lead bed removes the more readily-adsorbable
organics and has the highest adsorptive capacity - typical capacities through both
beds ranged from 0.2 to 0.4 Ib COD/lb of carbon and 0.1 to 0.3 Ib BOD/lb of
carbon (the observed capacities for the first bed generally were twice as high as
those observed in the second); and
3. the maximum hydraulic application rate that appeared to be technically justifiable
was 0.5 bed volumes per hour (through both columns of the series).
Full-scale Carbon Studies. There are currently several full-scale operational carbon
treatment systems in the United States. Three systems from which data are available are
cited here. These include the Atlantic Richfield (ARCO) system in California, the
Reichhold system in Alabama, and the British Petroleum (BP) system in Pennsylvania.
ARCO Carbon Treatment System (Watson Refinery)
The first full-scale carbon treatment system in the petroleum refining industry was
installed by ARCO at their Watson Refinery near Wilmington, California. This facility
was necessitated by a resolution from the Los Angeles Regional Water Quality Control
Board which limited the amount of COD that could be discharged by industry into the
Dominguez Channel in Los Angeles County. At the time, process wastewaters could be
treated in the County's primary treatment unit, but they could not accommodate storm
runoff from the Refinery. This presented a problem as the storm water collection system
and the process sewers were interconnected. Therefore, during periods of rainfall, the
combined storm runoff and process flow could not be sent to the County nor could it be
discharged to Dominguez Channel because of its COD concentration. A holding basin
and a carbon treatment plant were therefore deemed to be the most efficacious way of
treating this combined flow on an intermittent basis (29).
The treatment system, completed in 1971, is illustrated schematically in Figure 14.
The process wastewater flows through an API separator and then through pH adjustment and
chemical flocculation tanks. Following the addition of both coagulants and polyelectro-
lytes, the flow is routed to one of two circular dissolved air flotation units. At this
-------
308
point, the oil and grease (O & G) levels are normally below 15 to 20 mg/l. Under dry
weather conditions, the flow goes to the County for further treatment. During periods of
rainfall, it is diverted to a 50 million gallon holding basin where it is surged and conveyed
"as required" to the carbon treatment plant. The plant consists of 12 adsorber cells, the
flow to which is controlled by a handwheel-operated slide gate. Each bed can be back-
washed with treated water from a backwash sump, the backwash water being returned to the
holding basin. The carbon handling and regeneration system includes storage hoppers for
spent and regenerated carbon, a pump, eductor, piping, and controls to convey the carbon
from any cell to the regeneration furnace and back, and the multiple-hearth furnace with
gas scrubber. It is gas-fired and supplemental steam and air are added on two hearths. An
afterburner sect-ion is separately gas-fired to raise the off-gas temperatures to approximately
1,450 F (necessary to combust the organic vapors in the exit gas).
The actual design criteria for the system are presented in Table 27. Based on a
probability distribution of two typical runs presented in Figure 15, the effluent COD was in
the range of the predicted level if the influent concentration did not exceed the design
basis. Chlorination was attempted once the effluent COD exceeded the design level on
the premise that the halogenated compounds would be more effectively adsorbed and the
overall performance would be enhanced. There was no noticeable improvement in effluent
COD when this step was concluded, however (29). Although oil and grease was not cited
as a design parameter, the observed effluent concentration and its dependence on the influent
level is shown in Figure 16. The performance of this full-scale system in terms of COD
removal approximates the design basis and the observed reduction of COD and O & G is
consistent with pilot studies using similar wastewaters. It was noted, however, that algal
proliferation in the holding basin adversely affected the carbon plant performance. When
excessive algal growths developed, a deterioration of effluent quality and more frequent
backwashing was characteristic of the system. At one time, the algae became so concen-
trated that the carbon plant had to be shut down. Copper sulfate was added to the holding
basin to minimize algal growth and ease this problem.
The most significant variance of observed performance from the basis of design was the
carbon capacity. Although a precise determination was not possible, the loading based on
several runs ranged from 0.30 to 0.35 Ib COD removed/lb of carbon, rather than the 1.75
Ib COD/lb carbon prediction cited in Table 27.
With the exception of carbon capacity, the carbon plant generally performed as expected.
It is no longer in operation, however, primarily because the treatment requirements imposed
by Los Angeles County have been changed.
Reichhold Carbon Treatment System (Tuscaloosa Plant)
The Reichhold Chemical Plant produces sulfuric acid, formaldehyde, pentaerythritol,
sodium sulfate, sodium sulfite, orthophenylphenol, and a number of synthetic resins. A
carbon adsorption system was designed to treat an effluent flow of 500,000 gpd having an
average BOD concentration of 390 mg/l, a COD level of 650 mg/l, and a pH ranging from
5.4 to 12.3 (30). An agreement between the Company and the Alabama Water Improvement
Commission established an organic removal objective of 90 percent.
-------
309
A physical/chemical treatment plant using activated carbon was designed and
constructed and Figure 17 presents a simplified flow diagram. The process streams are
routed to a 2.5-day retention equalization basin where the water is pumped to an acid
mixing chamber. Concentrated sulfuric acid is added to maintain a pH range of 6.5 to
8.5. A nonionic polymer is added prior to the flocculation chamber. Following gravity
sedimentation, the water is pumped to one of two moving-bed adsorbers. Each adsorber
contains 124,000 Ib of granular activated carbon. At the design flow rate of 175 gpm/
adsorber, the empty bed contact time is approximately three hours. Treated water is
collected in a trough at the top of each adsorber where it flows to a final retention tank
and then to the river. A conventional regeneration furnace with an afterburner and wet
scrubber system is used for reactivation of the carbon.
In the first few months of operation, the results have been reported as meeting the
effluent quality requirements. The effluent BOD has been in the 35 to 40 mg/l range,
while a 90 percent removal of COD infers an effluent at the 65 to 70 mg/l level (30).
British Petroleum Carbon Treatment System (Marcus Hook, Pennsylvania)
The British Petroleum (BP) Refinery is a 105,000 barrel/day Class "B" refinery located
in Marcus Hook, Pennsylvania. In order to comply with discharge standards prescribed by
the Delaware River Basin Commission (DRBC), a preliminary engineering program and
treatability study was undertaken by the Company. Based on these results, the decision to
treat the effluent from the existing API separator to the required level using a filtration/
carbon adsorption system was made.
The treatment system, placed into operation in March 1973, is shown in Figure 18
(31). The API separator effluent flows to an intermediate surge basin where it is pumped to
three downflow filters, operated in parallel. The flow rate ranges from seven to 12 gpm/
ft , depending on the operation, and the media consists of 2.5 ft of anthracite and 4.5 ft
of sand. The fJlters are backwashed using filtered water, with provisions for air scouring
at 7.1 scfm/ft . The design filter pressure is 47.5 psi and the maximum allowable pressure
drop through the media is 6.5 psi. The backwash interval is 12 hours.
Three carbon adsorbers (10 ft diameter and 65 ft high) are operated in parallel. Each
adsorber contains 92,000 pounds of granular activated carbon in a bed depth of 45 feet.
An additional 8,000 pounds occupy the lower and upper cone areas. Flow to the three
adsorbers is controlled by the level in the filtered water holding tank. The upward flow
rate is 8.5 gpm/ft , which gives an empty bed contact time of 40 minutes. Spent carbon
at the bottom of the adsorber is pulsed out at the rate of approximately 1,000 Ib/day.
Fresh carbon is then added to the top of the column from a feed hopper. The design
criteria and activated carbon properties for the system are shown in Table 28 (32).
The regeneration facility is a five-foot diameter multiple-hearth furnace. The design
feed rate to the six-hearth furnace is 125 Ib/hr. In an atmosphere controlled by the addition
of steam, the adsorbed organics are volatized and oxidized. In order to insure complete
oxidation, all flue gases pass through an afterburner fired by refinery fuel gas and maintained
at a temperature of 1,350°F. A wet scrubber is included for gas cooling and the removal
of particulate matter. The design data for the thermal regeneration system is shown in
-------
310
Table 29.
As the full-scale system has been in operation for over two years, performance data are
available (31, 32). During the first months of operation while the adsorber still contained
virgin carbon, the COD removals were determined to be independent of influent concentration
but dependent on contact time. This relationship is shown in Figure 19. The oil and grease
removal during this period, shown in Figure 20, followed two different regimes. Good removal
was observed initially prior to the start of carbon pulsing. However, once the pulsing began,
taking the pulsed bed out of operation, the adverse effect on the remaining two columns is
reflected in terms of deteriorating effluent quality. The reduction of oil removal with increasing
increasing influent concentrations is attributed to both the oil removal mechanism and pulse
bed "off-line" mode of operation.
There have been four distinct phases of operation since the carbon system came on line.
The first period of data reported herein occurred when virgin carbon was in the adsorbers and
the foul water condensate stream from the Fluidized Catalytic Cracking Unit (FCCU) was not
yet included in the raw wastewater stream. The second period included the FCCU stream,
but virgin carbon was still present in the adsorbers as the wave front had not yet reached the
top of the carbon bed. The third period encompassed a time following a complete turnover of
the carbon bed and still included the FCCU stream. The fourth period excluded this stream
and incorporated a modification to the column septum design, still following a complete carbon
turnover. The average and maximum adsorber effluent concentrations for several parameters
observed during each period is tabulated in Table 30. A significant deterioration in quality
following the inclusion of the FCCU foul water condensate and complete carbon turnover in
the reactor is noted in Figure 21 . This has been attributed to inadequate pretreatment of the
API separator effluent in terms of O & G and soluble organic removal as well as significant
buildups of anaerobic biological growths and oily materials in the carbon media. A 40
percent decrease in adsorptive capacity of the regenerated carbon also has been observed.
The iodine number of the regenerated carbon is 560 to 680, while the virgin carbon was in
the range of 950 to 1,000 and the molasses number of the regenerated carbon was 280, or
approximately 50 higher than the virgin carbon. This indicates u decrease in micropores and
an increase in macropores during the regeneration cycle.
The Company is currently reviewing methods of improving process performance over that
observed in the last three periods of operation, or possibly changing the system concept
altogether. Particular emphasis is being placed on more adequate pretreatment, including
biological oxidation, as steps necessary to effectively use the existing carbon system.
SUMMARY
Activated carbon treatment of industrial wastes, while promising, must be carefully
evaluated before process decisions are made and capital funds are committed. As noted in
the pilot- and full-scale case histories presented herein, breakthrough geometry and adsorption
kinetics of multi-component wastewaters are difficult to define, many organic compounds
are not amenable to carbon adsorption, and the effects of regeneration on carbon capacities
are variable and unpredictable. For these and other reasons, comprehensive testing and
technical reviews are a necessary prerequisite to process commitment.
-------
311
Unfortunately, most of the literature regarding carbon treatment of industrial waste-
waters centers around pilot-plant results. There are, however, sufficient data from pilot
studies and the two full-scale systems in the petroleum refinery industry to draw general
conclusions, at least within this industrial category. For example, a distribution of long-
term average COD concentrations in effluents from carbon adsorbers treating refinery
wastewaters is presented in Figure 22 (33). These levels are relatively consistent with
those residuals reported in the petrochemical industry (28). A similar presentation of oil
and grease effluent levels from a full-scale refinery carbon system is shown in Figure 23.
Carbon adsorption capacities in terms of Ibs of COD removed/lb of carbon exhausted have
ranges from 0.2 to 0.4 in the petrochemical industry and from less than 0.1 to 0.55 in the
petroleum refining industry. These are lower than reported carbon capacities for municipal
wastewaters, as shown in Figure 24, emphasizing the inaccuracies which can occur by
extrapolating results from the treatment of one wastewater and using them as the basis for
predicting another.
SYNOPSIS
A review of the current state of the art of activated carbon treatment has been pre-
sented. Basic concepts of activated carbon treatment have been included, as well as
pertinent municipal and industrial case histories with which the author is familiar. It is
recognized that new truths pertaining to this subject become known on a continuing basis.
However, in evaluating process concepts, developing design bases, predicting effluent
quality, and finalizing management decisions in terms of constructing control systems with
attendant capital commitments, one must base these judgments on the current state of the
art. It is toward the objective of defining the art of activated carbon treatment that this
information is presented.
In the pursuit of this definition, certain apparent verities emerge. Some of the more
significant include the following:
1 . Adsorption theory is rigorous for single solutes, but becomes less definitive when
applied to wastewaters containing multiple components with varying molecular
weights and chemical characteristics. A good example is the poorly-defined
and erratic breakthrough geometry observed in the carbon treatment of complex
industrial wastewaters.
2. Many classes of organic compounds are not amenable to carbon adsorption -
particularly oxygenated organics - and show up as residual BOD, COD, or TOC
in carbon column effluents. This limits the overall process efficiency of pure
physical/chemical treatment systems. As many of these residual compounds are
biodegradable, activated carbon as a polishing process is generally capable of
producing a better quality of effluent than is the strict physical/chemical appli-
cation.
3. Anaerobic conditions which prevail in carbon reactors have caused difficulty
through the proliferation of biological growths on the carbon media. Sulfide
generation has been particularly troublesome. Although biological activity in
the column has the potential of increasing carbon capacities and is hypothesized
to offer regeneration possibilities, full-scale experience has indicated that
uncontrolled biological growth in carbon reactors has more negative features than
-------
312
positive.
4. Ozonation of wastewaters prior to carbon adsorption has been demonstrated on a
pilot scale to enhance adsorption in the columns. This is attributed to the ozone
products being more sorbable than the reactants. Other possible attractions of
this application are controlling biological densities on the carbon media through
partial disinfection and maintaining an aerobic environment in the reactor column.
5. Carbon capacities used as design criteria generally have been overstated and have
not been realized in full-scale experience. Breakthrough concentrations which
force premature regeneration cycles in single-operated columns, and desorption
phenomena which cause low capacities in series-operated polishing columns, are
partially responsible for this.
6. Design criteria for carbon adsorption systems should be sensitive to:
(a) wastewater constituents and their classification in terms of adsorbability, and
the effluent residual potential; and
(b) effect of the selected carbon media with respect to capacity, resistance to
abrasion, regeneration impacts, and hydraulic characteristics; and
(c) the biological growth potential with the associated effects on carbon capacity,
backwash requirements, and hydraulic characteristics of the flow through
the column; and
(d) the necessary pretreatment requirements for control of suspended solids
(organic and inorganic), oil and grease, dissolved oxygen, biological
population, and other constituents which affect carbon adsorber performance.
REFERENCES
(1) Weber, Walter J., Physicochemical Processes for Water Quality Control, John Wiley
and Sons, Inc., New York, (1972).
(2) Bell, Bruce A. and Molof, Alan H., "A New Model of Granular Activated Carbon
Adsorption Kinetics," Water Research, Vol. 9, Pergamon Press, London, (1975).
(3) Ford, D. L., "The Applicability of Carbon Adsorption in the Treatment of Petro-
chemical Wastewaters," Proceedings, The Application of New Concepts of Physical-
Chemical Wastewater Treatment, Sponsored by the International Association of Water
Pollution Research and the American Institute of Chemical Engineers, Vanderbilt
University, Nashville, (September, 1972).
(4) Giusti, D. M., Conway, R. A., and Lawson, C. T., "Activated Carbon Adsorption
of Petrochemicals," Journal, Water Pollution Control Federation, (May, 1974).
(5) Hassler, John W., Purification With Activated Carbon, Chemical Publishing Co.,
Inc., New York, (1974).
(6) Environmental Protection Agency, Process Design Manual for Carbon Adsorption,
Technology Transfer Manual, (October, 1971).
(7) American Water Works Association, "AWWA Standard for Granular Activated Carbon,"
Journal, AWWA, (November, 1974).
(8) Smisek, M., and Cerny, S., Active Carbon, Elsevier Publishing Co., New York, (1970).
(9) De John, P. B., "Carbon from Lignite or Coal, Which is Better?", Chemical Engineering,
(April, 1975).
(10) Loven, A. W., "Perspectives on Carbon Regeneration," Chemical Engineering Progress,
(November, 1973).
-------
313
(11) Federal Water Quality Administration, "Development of a Fluidized Bed Technique
of the Regeneration of Powdered Activated Carbon," Water Pollution Control Res.
Series ORD-17020 FBD, (March, 1970).
(12) Environmental Protection Agency, "Powdered Activated Carbon Treatment of
Combined and Municipal Sewage," Water Pollution Control Res. Series 11020 DSO,
(November, 1972).
(13) Berg, E. L., etal, Chemical Engineering Proceedings 67, No. 107, 154, (1970).
(14) Fuchs, JohnL., Private Communication to Engineering-Science, Inc., (June, 1975).
(15) Culp, R. L., and Roderick, R. E., "The Lake Tahoe Water Reclamation Plant,"
Journal, Water Pollution Control Federation, (February, 1966).
(16) Slechta, A. F., and Culp, G. L., "Water Reclamation Studies at the South Tahoe
Public Utility District," Journal, Water Pollution Control Federation, (May, 1967).
(17) Ford, D. L., personal communication with the City of Colorado Springs, Colorado,
(July, 1975). --
(18) Batelle - Northwest and Zurn Environmental Engineers, "Westerly Advanced Waste-
water Treatment Facility, Process Design Report and Appendices," Cleveland, Ohio,
(1971).
(19) Guirguis, W. A., et al, "Ozonation Studies at the Westerly Wastewater Treatment
Center," Second International Ozone Symposium, Montreal (May, 1975).
(20) Evans, F. L., Ozone in Water and Wastewater Treatment, Ann Arbor Science, Inc.,
(1972).
(21) Snoeyink, V. L., Weber, W. J. and Mark, H. B., "Sorption ofPhenol and Nitro-
phenol by Active Carbon," Environmental Science and Technology, (October, 1969).
(22) McDuff, D. P., and Chiang, W. J., "Physical Chemical Design for Garland, Texas,"
The Applications of New Concepts of Physical - Chemical Wastewater Treatment,
Sponsored by the International Association of Water Pollution Research and the
American Institute of Chemical Engineers, Vanderbilt University, Nashville,
(September, 1972).
(23) Engineering-Science, Inc., verbal communication with the Metropolitan Sewer
Board, Rosemount, Minnesota, (1975).
(24) Directo, L. S., and Chen, C. L., "Pilot Plant Study of Physical Chemical Treat-
ment," 47th Annual Water Pollution Control Federation Conference, Denver,
Colorado, (October, 1974).
(25) Engineering-Science, Inc. and Cleveland Regional Sewer District, Report on
Evaluation of Continuing Westerly Pilot-Plant Studies, prepared for CRSD,(June,
1975).
(26) Ford, D. L., and Buercklin, M. A., "The Interrelationship of Biological-Carbon
Adsorption Systems for the Treatment of Refinery and Petrochemical Wastewaters,"
6th International Association of Water Pollution Research Conference, Jerusalem,
(June, 1972).
(27) Short, T. E., and Myers, L. A., "Pilot Plant Activated Carbon Treatment of
Petroleum Refinery Wastewaters," Robert S. Kerr Environmental Research Laboratory,
Ada, Oklahoma, (1975).
(28) Lawson, C. T., "Activated Carbon Adsorption for Tertiary Treatment of Activated
Sludge Effluents from Organic Chemicals and Plastics Manufacturing Plants - Appli-
cation Studies and Concepts," Research and Development Department, Union Carbide
Corporation, South Charleston, West Virginia, (August, 1975).
-------
314
(29) Environmental Protection Agency, Refinery Effluent Water Treatment Plant Using
Activated Carbon, Environmental Protection Technology Series, EPA 660/2-75-020,
(June, 1975).
(30) Shumaker, T. P., "Granular Carbon Process Removes 99 - 99.2% Phenols," Chemical
Processing, (May, 1973).
(31) McCrodden, B. A., "Treatment of Refinery Wastewater Using Filtration and Carbon
Adsorption," Advanced Petroleum Refinery Short Course, Principles and Practice in
Refinery Wastewater Treatment, University of Tulsa, (June, 1973).
(32) McCrodden, B. A., "Treatment of Refinery Wastewater Using Filtration and Carbon
Adsorption," presented at Technology Transfer Conference, Activated Carbon in Water
Pollution Control, Sponsored by the Pollution Control Association of Ontario, and the
Canadian Society for Chemical Engineering, (October, 1974).
(33) Engineering-Science, Inc., Report to the National Commission on Water Quality,
Petroleum Refinery Industry - Technology and Cost of Wastewater Control, (June, 1975).
DISCUSSION
Ed Sebesta: Were these pilot plant data taken with regenerated carbon and equilibrated
carbon?
Davis Ford: No. It was virgin carbon. That's a good point. Many of the pilot plants
studies that we have run have been with virgin carbon. It is not that we don't recognize
that it would be more applicable with regenerated carbon, it is just sometimes that it is
hard to get that much regenerated carbon from the vendor and that has been our problem.
One would think it would be more practical to run all these pilot studies with regenerated
carbon.
Bob Huddleston: Do you have any data that indicates whether or not the effects of ozone
were anything other than simply killing the microbes?
Davis Ford: Yes. They ran some GC work to identify the nature of the compounds after
organization and I don't have the results in this paper. Just in talking with these people,
it's what you really have done is gone from the ketones, aldehydes to carboxylic acids,
there has been a shift and the carboxylic acids being more ameniable to carbon absorption,
so its a combination of the biological growth but also the transformation of these oxygenated
organics to carboxylic acids.
M. J<_. Mutton: Discuss if you will the cost of ozone. Do you have any figures on this?
I expect it is quite expensive.
Davis Ford: I left cost completely out of the paper because that is almost another paper
itself; but considering the concentration of ozone that we think we have to have, it is
probably about 4 or 5 milligrams per liter is within the cost effective parameters that we
have set to shis point. So we think it is cost effective.
Anonymous: Do you have some comments on phenol removal ?
-------
315
Davis Ford: No. Phenol is highly sorbable, most of the data we have seen on phenol
removal has been excellent and the only exceptions to this was when we just had so many
physical problems, the biological growths and oil and grease, that the phenol removal
has gone down. I think that is shown in one table. If you don't have this constraint,
phenol removal should be good, carbon is a good phenol remover.
Morris Wiley: The little village where I live has a new sewer system and biological
treatment plant. It is an extended aeration without any primary treatment, then a second
stage nitrification and finally a sand filter. They are able to do less than one part per
million BOD and less than one part per million ammonia in the effluent. Now I should
caution that it is only operating at one quarter of rated capacity. Thus, it appears that
if one has to go to these very low organic and nitrogen contents in the effluent, that one
can do it by a conventional biological system simply by spending a lot of money. I
estimate that this is about $1,000 per family per year, total annualized cost for this
system including the sewer connections which were put in. Have you done any economic
studies to try to figure out which would really be preferable, an advanced biological
reactor system versus an activated carbon where you need a very low concentration of
pollutants in the effluent?
Davis Ford: Let me speak in general terms here. First, you have a good analytical
chemist at that plant.
Morris Wiley: Well, I have some reservations about the precision of the tests. The plant
operator is a Texaco employee running the plant in his spare time with his wife and his
son, but it is a good quality effluent and the county is accepting their test reports.
Davis Ford: Let me say, from the cost studies that we have run, we have seen biological
treatment is more cost effective than carbon, I don't think there is any question about
that. When you get down to the fact that a biological system can't make a permit and
you have to go to carbon, you know some people are not sympathetic with cost effectives,
but if you can get to those levels biologically, it just intuitively has to be more cost
effective than carbon. Carbon has gone up, it is 50-60<: now for bituminous. When we
started running these studies three or four years ago, it was 32$ or 30<: and just because
of these lower carbon capacities I have stated, biological treatment has to be more cost
effective. If we can make it with biological treatment, that is certainly the theology that
I would like to see.
Morris Wiley: Then it looks as though our research should be concentrated on enhancing
the effectiveness this biological treatment as Jim Grutsch pointed out earlier?
Davis Ford: I do not disagree with that. I also think we have got to be realistic, there
are certain limitations to biological treatment. When you look at BAT, if and when it
comes, it is going to take some real engineering to make those numbers biologically.
Neale Fugate: Do you have any data on the possible adsorption of heavy metals,
conversely desorption of heavy metals on carbon?
or
-------
316
Davis Ford: Yes, sir. I have it, but it is not presented in the paper. I have just seen some
data on that. There is slight reductions of some of the metallic ions probably as metallic
hydroxides. I think in some literature absorption of some metallic ions has been reported.
It has not been significant, something like 10-15% reduction of zinc, mercury, cadmium,
some of these substances but percent removal is quite low, but some removal has been noted.
I am not sure of the mechanism as a physical removal of zinc or sorption of zinc iron, but
one of those two mechanisms.
Robert L. Wortman: It is my understanding that ozone is a very effective phenol remover,
it removes the phenol quite effectively, it is probably even better than activated carbon
(through oxidation). It would seem to me that may be secondary conventional treatment
followed by ozonation may be as effective as secondary treatment followed by carbonation.
My question is, have you had any experience on this in combination with carbon filtration
and I am interested in treating each one individually, that is considering ozonization as a
treatment method?
Davis Ford: Let me answer that two ways. First of all, I think I mentioned in the talk
that we tried preozonation after carbon with bad results. However, there is a treatment
plant, financed in part by EPA which goes on line in about two months and that is at Estes
Park, Colorado, which is exactly the system you described. Other than that plant, I am
not personally familiar with a real good biological treatment system including nitrification
followed by ozone but that system is designed that way. That is the upper sanitation district
in Estes Park, Colorado, the plant is completed and is probably going through shake down
right now and we will get some good full scale information from that facility in the next
few months.
Anonymous: I think there is also some work being done in Pomona, I don't know exactly
what it was, but on a different combination of ozonation and various schemes; but to further
explore the Estes Park, if you say you had a multimedia filter and then go in ozonation, this
may be quite effective to work with.
Fred Gowdy: Davis, I wonder if you would give us some of the reasons you believe why the
carbon capacities that we have realized in some of our full scale plants, haven't been any-
where close to those predicted by various pilot studies?
Davis Ford: I am not really sure that I know the answer to that, I can just hypothesize.
First of all, I think Union Carbide has presented carbon capacities in lead columns as far
as polishing columns were concerned, so that would indicate the operation has a lot to do
with carbon capacity, in other words, how well you can operate the carbon utilization in
a series mode of operation to get the operable capacity of the overall system. You can
pretty well do that in a pilot scale. I am not sure you can do it in a full scale quite so
easily. It is easier to operate that pilot scale, I assure you, than a full-scale plant. So
I think part of the answer lies in the operation. Secondly, it is possible that you had more
physical problems in terms of biological growth, influent constituents in the full scale that
you don't perceive in a pilot scale. You might control the pilot scale a little better and
for this reason are able to realize a better carbon capacity. Those are the two main reasons
I think, I just can't prove it.
-------
317
Ed Soger: I wondered if you would like to elaborate on the back washing features of these
columns. Why is that required if you have adequate pretreatment, are you removing
suspended solids build-up?
Davis Ford: I think the best case history to point to there is the BP, they have got some
data on their filters, good filters which work quite well; but the effluent suspended solids
are probably in the 10-15-20 milligram per liter range and these influent suspended solids
to the carbon column are going to accumulate after a certain period of time, combined
with the biological proliferation in the columns. I think you always have to have the flexi-
bility operational flexibility of back washing columns as well as the filters. I think that
has been the experience to date. You may not have to do it, but it is an operational
flexibility that you should have. It seems like even if you go in with zero suspended
solids, some point in time you are going to have to back wash the filters or the carbon
columns.
Pat DeJohn: One of the things that one of the gentlemen asked about the loadings, most
of the loadings that are reported are reported with virgin carbon and the properties of
activated carbon change quite substantially when you regenerate carbon and I think that is
part of the BP problem, they are not able to maintain their phenol standards because of
the loss in capacity, adsorptive capacity and consequently, their system has been sub-
stantially under design. The other thing that you mentioned about why you should be back
washing, that BP plant is a pulse bed unit and there are fines that are created due to the
regeneration of the carbon and they have no way to remove those fines; consequently they
recycle them back into the absorber and they pulse that unit a lot of times because of
pressure drop problems and that is another thing about that particular plant.
Davis Ford: I agree with that. In the past. Pat, when we have had to use virgin carbon,
we have just applied a factor on the capacity, reduced capacity, just an empirical factor
and that is a less favorable way of doing it than using regenerated carbon; but we just
haven't had that luxury a lot of times.
J. J. Chavez: I was just wondering if you could elaborate on the regeneration phase
from the economic side by which a refinery could put in one regulation facility and the
options open to him if he can't afford one?
Davis Ford: There again I didn't really get into cost. I believe the reported regeneration
cost now is something like 15-20<: per pound. Some manufacturers, I believe, are offering
a regeneration service. If you have a small system, it doesn't make sense because of
economy in scale to put in your own regeneration facility, to have a contractor regenerate
it, if that's possible, but I think if you go into regeneration which in most areas you would
have to for most size of plants, you have to count on about a 15-20$ per pound regeneration
cost.
-------
318
BIOGRAPHY
Davis L. Ford holds a B.S. in Civil Engineering
from Texas A & M University and M.S. and Ph.D.
degrees in Environmental Health Engineering from the
University of Texas at Austin. Dr. Ford is currently
Senior Vice President and Member of the Board of
Directors of Engineering Science, Inc., in Austin,
Texas. Dr. Ford has written 4 books, 20 reports, 60
publications in the field of environmental engineering
and has consulted for over 50 industries, the United
Nations (WHO and PAHO), the EPA and various state
and municipal agencies.
TABLE 1 "COMPARATIVE ANALYSIS OF BATCH ISOTHERM DATA
REFINERY & PETROCHEMICAL WASTEWATERS"
K
OIL SEPARATOR (PRIMARY) EFFLUENT
Refinery - Petrochemical Complex No. 1
Refinery - Petrochemical Complex No. 2
Refinery No. 3
SECONDARY (ACTIVATED SLUDGE) EFFLUENT
Refinery - Petrochemical Complex No. 1
Refinery No. 3
Refinery Secondary Effluent
Refinery Secondary Effluent
Refinery Secondary Effluent
SINGLE ADSORBATE
Phenol
Dichlorethane (pH 4)
(pH6)
(pH 10)
0.0290
0.0036
0.0140
0.0062
0.0043
0.0051
0.0038
0.0020
0.1110
0.0045
0.0038
0.0041
0.77
0.80
0.36
0.60
1.00
0.96
1.08
0.69
5.80
1.82
1.67
1.49
-------
TABLE 2
AMENABILITY OF TYPICAL ORGANIC COMPOUNDS
TO ACTIVATED CARBON ADSORPTION [4]
Compound
Alcohols
Methanol
Ethanol
Propanol
Butanol
n-Amyl alcohol
n Hexanol
Isopropanol
Ally! alcohol
Isobutanol
t Butanol
2 Ethyl butanol
2Ethylhexanol
Aldehydes
Formaldehyde
Acetaldehyde
Propionaldehyde
Butyraldehyde
Acrolcm
Crolonaldehyde
Benzaldehyde
Paralo>riyde
Amines
Di N Propylamme
Butylamtne
Di N Butylamme
Allylamme
Elhylenediamme
Dielhylenelriamme
Mcnelhanolamine
Dicthanotamme
Tntihanclamme
Monoisopropanolamme
Dusopiupanolamme
Pyridines & Morpholines
Pyndine
2 Methyl 5-Ethyl pyridine
N Methyl morpholme
N Ethyl morpholme
Aromatics
Benzene
Toluene
Elhyl benzene
Phenol
Hydfoqumone
Aniline
Styrene
Nitrobenzene
Esters
Methyl acetate
Ethyl acetate
Piopyl acetate
Butyl acetate
Primary a my I acetate
Concentration (mg/l)
Molecular
Weight
320
46.1
60.1
741
88.2
1022
60 1
58.1
74 1
74.1
1022
1302
30.0
44 1
581
72 1
56.1
701
1061
1322
101 2
731
1293
571
601
1032
61 1
1051
1491
751
1332
791
1212
1012
1152
78.1
92 1
1062
94
1101
931
1042
123 1
74 1
88 1
1021
1162
1302
Aqueous
Solubility
(*)
00
00
oo
7.7
17
0.58
OO
oo
85
00
0.43
007
oo
00
22
71
206
155
0.33
10.5
oo
oo
oo
oo
oo
oo
oo
954
OO
oo
87
OO
si. sol
OO
oo
007
0047
002
67
6.0
34
003
019
319
87
2
068
02
Initial (Co)
1,000
1.000
1,000
1.000
1,000
1,000
1.000
1,010
1,000
1.000
1.000
700
1,000
1,000
1,000
1.000
1.000
1,000
1.000
1,000
1.000
1.000
1,000
1.000
1,000
1,000
1,012
996
1.000
1,000
1.000
1,000
1.000
1,000
1,000
416
317
115
1,000
1,000
1.000
180
1,023
1.030
1,000
1.000
1.000
985
Final (C|)
964
901
811
466
282
45
874
789
581
705
145
10
908
881
723
472
694
544
60
261
198
480
130
686
893
706
939
722
670
800
543
527
107
575
467
21
66
18
194
167
251
18
44
760
495
248
151
119
Adsorbability'
g compound/
g carbon
0007
0020
0038
0107
0155
0191
0025
0.024
0084
0.059
0170
0138
0.018
0022
0057
0106
0061
0092
0.188
0148
0,174
0103
0174
0063
0021
0062
0015
0057
0067
0040
0091
0095
0179
0085
0 107
0080
0050
0019
0161
0167
0150
0028
0196
0054
0100
0149
0169
0175
Percent
Reduction
36
100
189
534
71.8
955
12.6
219
419
295
855
985
9.2
11.9
277
52.8
306
456
940
739
802
520
870
31 4
107
294
72
275
33.0
200
457
473
893
425
53.3
950
792
843
806
833
749
88.8
95.6
262
505
752
846
880
Compound
Esters
Isopropyl acetate
Isobutyl acetate
Vinyl acetate
Ethylene glycol monoethyl ether acetate
Ethyl acrylate
Butyl acrylate
Ethers
Isopropyl ether
Butyl ether
Dichloroisopropyl ether
Glycols& Glycol Ethers
Ethylene glycol
Diethylene glycol
Triethylene glycol
Tetraethylene glycol
Propylene glycol
Dipropylene glycol
Hexylene glycol
Ethylene glycol mono methyl ether
Ethylene glycol monoethyl ether
Ethylene glycol monobutyl ether
Ethylene glycol monohexyl ether
Diethylene glycol monoethyl ether
Diethylene glycol monobutyl ether
Ethoxytriglycol
Halogenated
Ethylene dichlonde
Prupylene dichlonde
Ketones
Acetone
Methylethyl kelone
Methyl propyl Ketone
Methyl butyl ketone
Methyl isobulyl ketone
Methyl isoamyl ketone
Diisobutyl ketone
Cydohexanone
Acelophenone
Isophorone
Organic Acids
Formic acid
Acetic acid
Propiomc acid
Butyric acid
Valeric acid
Caproic acid
Acrylic acid
Benzoic acid
Oxides
Ptopylene oxide
Styrene oxide
Dosage 5 g Carbon C/l solution
Concentration (mg/l)
Molecular
Weight
102.1
116.2
86.1
132.2
100.1
128.2
1022
130.2
171.1
62.1
1061
150.2
194.2
761
134.2
1182
76.1
901
1182
1462
1342
1622
1782
990
113.0
581
72 1
861
1002
100.2
1142
1422
982
1201
1382
460
601
741
881
1021
1162
721
1221
581
1202
Aqueous
Solubility
(*)
2.9
0.63
2.8
22.9
20
02
1.2
003
0.17
OO
oo
oo
oo
00
00
oo
oo
oo
00
099
oo
00
oo
0.81
0.30
00
268
43
v si sol.
19
0.54
0.05
2.5
0.55
1 2
OO
OO
oo
00
24
1.1
oo
029
405
0.3
Initial (C,)
1.000
1.000
1.000
1.000
1.015
1,000
1,023
197
1.008
1.000
1,000
1.000
1.000
1.000
1.000
1.000
1,024
1.022
1.000
975
1.010
1,000
1,000
1,000
1.000
1.000
1.000
1.000
988
1.000
986
300
1,000
1,000
1,000
1.000
1.000
1.000
,000
,000
,000
,000
,000
1,000
1,000
Finil (C|)
319
180
357
342
226
43
203
nil
nil
932
738
477
419
884
835
386
886
705
441
126
570
173
303
189
71
782
532
305
191
152
146
nil
332
28
34
765
760
674
405
203
30
355
89
739
47
Adsorbability
I compound/
t carbon
0137
0164
0.129
0.132
0 157
0193
0162
0039
0200
00136
0053
0.105
0116
0024
0033
0122
0028
0063
0112
0.170
0087
0166
0 139
0.163
0.183
0043
0.094
0139
0.159
0169
0.169
0.060
0.134
0194
0.193
0047
0.048
0065
0119
0.159
0.194
0129
0183
0.052
0.190
Percent
Reduction
68.1
82.0
643
65.8
77.7
95.9
800
100.0
100.0
68
262
523
581
116
165
614
135
310
559
87 1
436
827
697
81 1
929
218
468
69.5
807
848
85.2
100.0
66.6
972
966
235
240
326
595
797
970
645
91 1
261
953
CO
-------
320
TABLE 3 "RELATIVE AMENABILITY TO CARBON ADSORPTION OF TYPICAL
PETROCHEMICAL WASTEWATER CONSTITUENTS"
Compound % Adsorption Compound % Adsorption
Ethanol 10 Vinyl acetate 64
Isopropanol 13 Ethyl acrylate 78
Acetaldehyde 12 Ethylene glycol 7
Butyraldehyde 53 Propylene glycol 12
Di-N-propylamine 80 Propylene oxide 26
Monoethanolamine 7 Acetone 22
Pyridine 47 Methyl ethyl ketone 47
2-Methyl 5-ethyl pyridine 89 Methyl isobutyl ketone 85
Benzene 95 Acetic acid 24
Phenol 81 Proprionic acid 33
Nitrobenzene 96 Benzoic acid 91
Ethyl acetate 50
Initial Compound Concentration = 1,000 mg/l Powdered Carbon Dosage =5,000 mg/l
TABLE 4 "INFLUENCE OF MOLECULAR STRUCTURE AND OTHER FACTORS OF ADSORBABILITY"
1 . An increasing solubility of the solute in the liquid carrier decreases its adsorbability.
2. Branched chains are usually more adsorbable than straight chains. An increasing length of the chain decreases solubility.
3. Substituent groups affect adsorbability:
Substituent Group Nature of Influence
Hydroxyl Generally reduces adsorbability; extent of decrease depends on structure of
host molecule.
Amino Effect similar to that of hydroxyl but somewhat greater. Many amino acids
are not adsorbed to any appreciable extent.
Carbonyl Effect varies according to host molecule; glyoxylic are more adsorbable
than acetic but similar increase does not occur when introduced into
higher fatty acids.
Double Bonds Variable effect as with carbonyl.
Halogens Variable effect.
Sulfonic Usually decreases adsorbability.
Nitro Often increases adsorbability.
Aromatic Rings Greatly increases adsorbability.
4. Generally, strong ionized solutions are not as adsorbable as weakly ionized ones; i.e., undissociated molecules
are in general preferentially adsorbed.
5. The amount of hydrolytic adsorption depends on the ability of the hydrolysis to form an adsorbable acid or base.
6. Unless the screening action of the carbon pores intervene, large molecules are more sorbable than small molecules
of similar chemical nature. This is attributed to more solute carbon chemical bonds being formed, making
desorption more difficult.
7. Molecules with low polarity are more sorbable than highly polar ones.
-------
321
TABLE 5
PROPERTIES OF SEVERAL COMMERCIALLY AVAILABLE CARBONS*(6]
PHYSICAL PROPERTIES
Surface area, nWgm (BET)
Apparent density, gm/cc
Density, backwashed and drained, Ib/cu. ft.
Real density, gm/cc
Particle density, gm/cc
Effective size, mm
Uniformity coefficient
Pore volume, cc/gm
Mean particle diameter, mm
SPECIFICATIONS
Sieve size (U.S. std. series)
Larger than No. 8 (max.
Larger than No. 12 (max.
Smaller than No. 30 (max.
Smaller than No. 40 (max.
Iodine No.
Abrasion No., minimum
Ash (%)
Moisture as packed (max. %)
%)
%)
%)
%)
ICI
AMERICA
HYDRODARCO
3000
(LIGNITE)
600-650
0.43
22
2.0
1.4
0.8
1.7
0.95
1.6
8
5
650
1.5
0.9
CALGON
FILTRASORB
300
(8x30)
(BITUMINOUS)
950-1050
0.48
26
2.1
1.3 1.4
0.8 0.9
1.9 or less
0.85
1.5 1.7
5
900
70
8
2
* Other sizes of carbon are available on request from the manufacturers.
'* No available data from the manufacturer.
— Not applicable to this size carbon.
WESTVACO
NUCHAR
WV-L
(8x30)
(BITUMINOUS)
1000
0.48
26
2.1
1.4
0.85 1.05
1.8 or less
0.85
1.5 1.7
5
350
70
7.5
2
WITCO
517
(12x30)
(BITUMINOUS)
1050
0.48
30
2.1
0.92
0.89
1.44
O.bO
1.2
5
5
1000
85
0.5
1
TYPICAL PROPERTIES OF 8 X 30-MESH CARBONS
[9]
Total surface area, m2/g
Iodine number, min
Bulk density, Ib/ft3 backwashed and drained
Particle density wetted in water, g/cm3
Pore volume, cm3/g
Effective size, mm
Uniformity coefficient
Mean-particle dia., mm
Pittsburgh abrasion number
Moisture as packed, max.
Molasses RE (relative efficiency)
Ash
Mean-pore radius
LIGNITE
CARBON
600-650
600
22
1.3-1.4
1.0
0.75-0.90
1.9 or less
1.5
50-60
9%
100-120
12-18%
33 A
BITUMINOUS
COAL CARBON
950-1,050
950
26
1.3-1.4
0.85
0.8-0.9
1.9 or less
1.6
70-80
2%
40-60
5-8%
14 A
-------
322
TABLE 6 "TYPICAL PROPERTIES OF POWDERED ACTIVATED CARBON (PETROLEUM BASE)1
Surface Area (BET m /gm)
Iodine No.
Methylene Blue Adsorption (mg/gm)
Phenol No.
Total Organic Carbon Index (TOCI)
Pore Distribution (Radius Angstrom)
Average Pore Size (Radius Angstrom)
Cumulative Pore Volume (cc/gm)
Bulk Density (gm/cc)
Particle Size Passes: 100 mesh (wt%)
200 mesh (wt%)
325 mesh (wt%)
Ash (wt%)
Water Solubles (wt%)
pH of Carbon
!,300
!,700
400
10
400
15
20
0.1-
0.27
97-
93-
85-
•2,600
•3,300
•600
12
•800
60
30
0.4
0.32
100
98
95
.5
.0
-9
Stage
Drying
Thermal Desorption
Pyrolysis and Carbonization
Gasification
TABLE 7 "STAGES OF THERMAL REGENERATION"
Approximate Temperature ( F) Processes
Ambient to 212
212 to 500
400 to 1, 200
1,200 to 1,900
Water Evaporation
Physical desorption of volatile adsorbed organics
Pyrolysis of nonvolatile organics and
carbonization of the pyrolysis residue
Gasification of pyrolytic residue through
controlled chemical reaction with water vapor,
vapor, carbon dioxide, or oxygen
TABLE 8 "WATER QUALITY AT VARIOUS POINTS IN PROCESS" (16)
Quality Parameter
BOD (mg/l)
COD (mg/l)
Total organic carbon (mg/l)
ABS (mg/l)
PO4 (mg/l as PO4)
Color (units)
Turbidity (units)
Nitrogen organic N (mg/l as N)
Ammonia N (mg/l as N)
NO- and NO- (mg/l as N)
UncRlorinated: Coliforms (MPN/100 ml)
Fecal coliforms (MPN/100 ml)
Virus
Chlorinated: Coliforms (MPN/100 ml)
Fecal coliforms (MPN/100 ml)
Raw
Wastewater
200-400
400-600
2.0-4.0
10-15
25-35
0
Secondary
Effluent
20-100
Separation-
Bed Effluent
Carbon-
Column Effluent
80-160
-
0.4-2.9
25-30
-
30-70
4-6
25-32
0
2,400,000
150,000
20-60
8-18
0.4-2.9
0.1-1
10-30
<0.5-3
2-4
25-32
0
9,300
930
Negative
8.6
1-25
1-6
<0.01-0.5
0.1-1
<5
<0.5-1
1-2
25-32
0
1 1 ,000
930
Negati\
<2.1
-------
323
TABLE 9 "WATER QUALITY AT VARIOUS STAGES OF TREATMENT AT SOUTH LAKE TAHOE1
Effluent
Quality
Parameter
BOD (mg/1)
COD (mg/1)
SS (mg/1)
Turbidity (JTU)
MBAS (mg/1)
Phosphorus (mg/1)
Col i form
(MPN/lOOml)
Raw
Wastewater
140
280
230
250
7
12 6
50 x 10
Primary
100
220
100
150
6
9 6
15 x 10
Secondary
30
70
26
15
2.0
6 6
2.5x 10
Chemical
Clarifier
10
10
0.7
Filter
3
25
0
0.3
0.5
0.1
50
Carbon
1
10
0
0.3
0.1
0.1
50
Chlorinated
Final
0.7
10
0
0.3
0.1
0.1
<2
TABLE 10 "CARBON EFFICIENCY PER REGENERATION PERIOD AT SOUTH LAKE TAHOE1
NOVEMBER 1968 THROUGH JANUARY 1971
Parameter
Carbon Dosage
(Ib regenerated/million gallons treated)
Iodine Number
Apparent Density (gm/ml)
2
Percent Ash
Chemical Oxygen Demand
Spent Carbon
Regenerated Carbon
Spent Carbon
Regenerated Carbon
Spent Carbon
Regenerated Carbon
Percent Removal
Ib COD applied
Ib COD applied/MG
Ib COD removed/MG
Ib COD applied/lb ]
carbon regenerated
Ib COD removed/lb
carbon regenerated
Methylene Blue Active Substances (Methylene Blue Active
Substances (MBAS)) Percent removal
Ib MBAS applied
Ib MBAS applied/MG
I b MBAS removed/MG
Ib MBAS applied/lb
carbon regenerated
Ib MBAS removed/lb
carbon regenerated
Average
207
Maximum
418
Minimum
111
583
802
0.571
0.487
6.4
6.8
49.9
28,250
162
81
0.78
0.39
77
995
5.7
4.4
0.027
0.021
633
852
0.618
0.491
7.0
7.2
63.3
54,970
254
149
1.56
0.71
93
1,675
10.7
8.2
0.045
0.039
497
743
0.544
0.478
5.8
5.8
30.1
15,680
105
32
0.52
0.16
58
457
2.6
1.6
0.012
0.007
1 3 3
-Based on ft of carbon fed to furnace at 30 Ib/ft
November 1968 through November 1970
-------
324
Solid Contact Clarifier
TABLE 11 "COLORADO SPRINGS TERTIARY PLANT DESIGN DATA"
48 ft diameter, 11 ft 9 in sidewall, 2 ft cone depth
1,809 sq ft surface area 168,465 gallon capacity
2 hr detention time and 0.76 gpm/sq ft rise rate @ 2 MGD flow
Spent Lime Tank
New Lime Holding Tank
Recarbonization Tank
Activated Carbon Adsorbers
Dual Media Sand Filters
14 ft diameter, 10 ft sidewall depth, 8 ft cone
1,948 cu ft 14,610 gallon capacity
14 ft diameter, 10 ft sidewall depth, 8 ft cone
l,948cuft 14,610 gallon capacity
14 ft diameter, 14 ft sidewall depth, plus 2 ft freeboard - 154 sq ft
2, 156 cu ft 16, 160 gallon capacity - 12 minutes detention time @ 2 MGD flow
20 ft diameter, 14 ft sidewall depth, 10 ft of carbon media
314 sq ft surface area, 4,396 cu ft 32,970 gallon capacity
total tower detention time is 24 minutes - 4.5 gpm/sq ft (@ 2 MGD flow)
carbon bed detention time is 17 minutes 4.5 gpm/sq ft (@ 2 MGD flow)
3,140 cu ft of carbon Ł 30 Ib/cu ft 94,200 Ibs carbon
1 2 ft diameter, 1 1 ft sidewall depth, 1 ft 6 in top cone
113 sq ft surface area
1,243 cu ft capacity 9,323 gallon capacity
3 ft of e.s. 1.5 mm sand 5 ft of e.s. 2.8 mm anthrafilt
PARAMETER
BOD
COD
TOC
TSS
Turbidity
O.PO (total)
O.PO^ (soluble)
MB AS
PH
Alkalinity (total)
Hardness
C (as C )
C C (as C COJ
r- i a a •*
Color
SuI fates
Sulfides
Flow (MGD)
Lime Dose (ppm Ca 0)
TABLE 12 "TERTIARY TREATMENT PLANT DATA SUMMARY (FIRST REGIME)"
R-C INF. R-C EFF. FILTER EFF. LEAD CARBON EFF. POLISH CARBON EFF.
106
305
76
56
53
32
31
5.32
7.35
197
163
52
130
162
88
0.15
1.93
46.9
120
37.9
31.0
10.1
2.2
0.16
3.22
11.44
282
210
79
197
44
80
0.14
1.93
48.0
114
36.8
3.7
4.2
1.7
1 .6
7.93
69
250
95
237
38
439
1.93
44.0
83
27.9
2.1
3.6
2.2
2.2
1.36
6.91
63
240
84
209
25
439
0.26
1.70
33.0
64
23.8
3.7
3.2
1.9
1.8
0.51
7.12
69
253
92
231
13
401
0.41
1.70
349
NOTE: During this period, the reactor was operated with a deep sludge blanket of approximately six feet.
-------
325
TABLE 13 "TERTIARY TREATMENT PLANT DATA SUMMARY (SECOND REGIME)'
PARAMETER
BOD
COD
TSS
Turbidity
0. Phosphate (total)
0. Phosphate (soluble)
MBAS
Fecal Coliform
Fecal Strep
Average Flow (MGD)
Average Lime Dose
Lead Tower
Polish Tower
INFLUENT
(mg/1)*
102
258
62
54
30
26
4.2
5 x 105/100 ml
6.5x 10 /100ml
EFFLUENT
(mg/1)*
8.0
15.7
2.2
2.2
1.0
1.0
0.15
225/100 ml
1150/100 ml
1.5
345 mg/1
5 ft of twice-regenerated + 5 ft of virgin carbon
8 ft of twice-regenerated + 2 ft of virgin carbon
*All units in mg/1 unless indicated.
The "INFLUENT" is the secondary effluent going to the solids contact clarifier.
The "EFFLUENT" is the final polish carbon tower effluent.
Influent to tertiary plant is from a trickling filter plant with a flow of
23 to 25 MGD; trickling filter plant capacity is 13 MGD.
REMOVAL
(percent)
92,
93.
96,
96,
96,
96,
96.4
99.96
99.82
TABLE 14 "COMPARISON OF WASTEWATER STRENGTH BATTELLE AND CRSD TEST PROGRAMS"
Probability (% of Occurrences)
10 50 90
Battelle Series (1970-71) BOD
CRSD Series (1974-75) BOD
Battelle Series (1970-71) COD
CRSD Series (1974-75) COD
Battelle Series (1970-71) BOD/COD
CRSD Series (1974-75) BOD/COD
180 mg/1
115 mg/1
320 mg/1
215 mg/1
0.56
0.53
240 ma/I
170 mg/1
500 mg/1
350 mg/1
0.48
0.48
320 mg/1
250 mg/1
720 mg/1
580 mg/1
0.44
0.43
Parameter
TABLE 15 "EFFECT OF POST-OZONATION ON EFFLUENT ORGANIC QUALITY"
CRSD PILOT-PLANT TEST RESULTS*
Before Ozonation After Ozonation
BOD (unfiltered)
BOD (filtered)
COD (unfiltered)
COD (filtered)
TOC
32 mg/1
25 mg/1
72 mg/1
59 mg/1
31 mg/1
31 mg/1
24 mg/1
66 mg/1
52 mg/1
29 mg/1
*AII data represent average values over an eleven day period. Ozone dosage ranges from 4 9 mg/1.
-------
326
TABLE 16 "QUALITY OF RAW WASTEWATER"
GARLAND PHYSICAL/CHEMICAL TREATMENT FACILITY
Parameter Results
Total BOD (mg/l) 266
Filtered BOD (mg/l) 236
Total COD (mg/l) 542
Filtered COD (mg/l) 240
Suspended Solids (mg/l) 233
Alkalinity (mg/l as CaCOJ 200
PH 7.2-7.7
P04 (mg/l) 15
TABLE 17 "PCT SYSTEM DESIGN DATA" (24)
CHEMICAL TREATMENT SYSTEM
1 . Flocculation: Detention Time, minutes 45
Chemical Dosage Alum, mg/l Al 22
polymer, mg/l 0.25
2. Sedimentation: Detention Time, hrs. 1.5
Overflow Rate, gpd/sq ft 900
Underflow, % of plant flow 1.25
Underflow solids, % by weight 2
3. Gravity Thickening: Solids Loading, Ib/day-sq ft 12
Underflow solids, % by weight 4
4. Vacuum Filtration: Yield, Ib/hr-sq ft 2
Cake solids 18
5. Sludge Incineration: Solids Loading, Ib/hr-sq ft 2
CARBON TREATMENT SYSTEM
1. Carbon Contacting (8 x 30 mesh carbon) Empty-bed contact time 25
Hydraulic surface Loading, gpm/sq ft 4
Backwash volume, % of plant flow 5
Sodium Nitrate dosage, mg/l N 5.5
Carbon Dosage, Ib/MG 250
Carbon Regeneration loss, % 5
-------
327
TABLE 18 "SUMMARY OF PHYSICAL/CHEMICAL TREATMENT SYSTEM PERFORMANCE" (24)
Average Percent Removal
Parameters
Suspended Solids (mg/l)
Turbidity (JTU)
TCOD (mg/l)
DCOD (mg/l)
BOD (mg/l)
Total Phosphate (mg/l P)
Nitrate (mg/l N)
Color
PH
Notes: 1. Average alum dosage = 25 mg/l Al (275 mg/l alum)
2. Average polymer dosage = 0.3 mg/l Calgon WT-3,000 _
3. AS Effluent refers to Activated Sludge Plant Effluent (8 MGD, existing facility)
Raw
Sewage
199
321
49.4
11.1
7.7
Clarified
Effluent
28.3
22.9
95.8
48.6
36.2
1.3
0.9
20
6.8
Carbon
Effluent
6.7
6.3
19.3
13.5
7.8
0.9
1.3
7.8
6.8
Chemical
Treatment
85.8
70.2
1.6
88.3
Carbon
Treatment
76.3
72.5
79.9
72.2
78.5
30.8
61.0
Overall
96.6
94.0
72.7
91.9
AS Effluent*
11.6
7.7
39.5
25.7
8.0
33.1
TABLE 19 "EFFECT OF REGENERATION ON THE PCT CARBON CHARACTERISTICS"
Spent Carbon Regenerated Carbon (Composite Sample)
(Composite Sample)
Carbon
Characteristics
Virgin
Carbon
1st
Reg.
2nd
Reg.
3rd
Reg.
Before Quenching
1st
Reg.
2nd
Reg.
3rd
Reg.
After Quenching
1st
Reg.
2nd
Reg.
3rd
Reg.
Iodine No. (mg/l)
Apparent Density
(g/cm cu ft)
Molasses No.
Methylene Blue No.
(mg/g)
Ash (%)
Mean Particle Dia.
(mm)
1,040 402 572 570 805 722 773
0.484 0.629 0.585 0.594 0.528 0.537 0.526
751 727 721
0.565 0.548 0.535
222
259
6.4
1.44
120
147
10.3
1.46
168
153
8.22
1.58
154
153
213
223
8.67 10.7
1.48 1.57
233
243
11.6
1.50
230
246
7.81
1.54
189
227
12.0
1.55
221
239
12.2
1.50
204
245
9.0
1.43
-------
TABLE 20
TERTIARY TREATMENT PLANTS
Site
Arlington, Virginia
Colorado Springe, Colo.
Dallas, Texas
Fairtax County, Va.
Los Angeles, Calif.
Montgomery County, Md.
Occoquan, Va.
Orange County, Calif.
Piscataway, Md.
St. Charles, Missouri
South Lake Taho, Calif.
Windhoek, South Africa
Status
1973
Design
Operating Dec. '70
to Present
Design
Design
Design
Design
Design
Construction
Operating Mar. '73
to Present
Construction
Operating Mar. '68
to Present
Operating Oct. '68
to Present
Average
Plant
Design Capacity
Engineer (MDG)
Alexander Potter/ 30
Engineering-Science
Arthur B. Chafet 3
& Assoc.
URS Forest & 100
Cotton
Alexander Potter/ 36
Engineering-Science
City of Los Angeles 53
CH2M/Hill 60
CH2M/Hill 18
Orange County Water 15
District
Roy F . Weston 5
Moran and Cooke 5.5
CH2M/Hill 7.5
National Institute 1.3
for Water Research
Pretoria, So. Africa
CO
KJ
OO
Total
No. of Contact Hydraulic Carbon Effluent
Contactor Contactors Time1 Loading Depth Carbon Requirements2
Type In Series (Min.) (gpm/ftz') (ft) Size (Oxygen Demand)
Down flow 1
Gravity
Down flow 2
Upflow 1
Packed
Down flow 1
Gravity
Downflow 2
Gravity
Upflow 1
Packed
Upflow 1
Packed
Upflow 1
Packed
Downflow 2
Pressure
Downflow 1
Gravity
Upflow 1
Packed
Downflow 2
Pressure
38 2.9 15 8 x 30 BOD
30 5 20 8 x 30 BOD
10 8 10 8 x 30 BOD
BOD
(by
36 3 15 8 x 30 BOD
50 4 26 8 x 30 COD
30 6.5 26 8 x 30 BOD
COD
30 5.8 24 8 x 30 BOD
COD
30 5.8 24 8 x 30 COD
37 6.5 32 8 x 30 BOD
30 3.7 15 8 x 30
17 6.2 14 8 x 30 BOD
COD
30 3.8 15 12 x 40 COD
< 3 mg/1
< 2 mg/1
< 10 mg/1
< 5 mg/1
1980)
< 3 mg/1
< 12 mg/1
< 1 mg/1
< 10 mg/1
< 1 mg/1
< 10 mg/1
< 30 mg/1
< 5 mg/1
< 5 mg/1
< 30 mg/1
< 10 mg/1
1 Empty bed (superficial) contact time
for average plant flow.
2 BOD: Biochemical oxygen demand
COD: Chemical oxygen demand
50 MGD ultimate capacity
-------
TABLE 21
PHYSICAL/CHEMICAL TREATMENT PLANTS
Site
Cortland, New York
Cleveland Westerly,
Ohio
Fitchburg,
Massachusetts
Garland, Texas
LeRoy, New York
Niagara Falls,
New York
Owosso, Michigan
Rosemount,
Minnesota
Rocky River,
Ohio
Vallejo,
California
Status 1973
Design
Construction
Construction
95% Complete
Construction
Design
Construction
50% Complete
Design
Operational
Operational
Design
Average
Plant Contact
CaŁacity Time1
Design Engineer (MGD) (Min.)
Stearns & Wheeler 10 30
Engineering-Science 50 30
Camp Dresser & 15 35
McKee
URS Forest & 30 30
Cotton
Lozier Engineers 1 27
Camp Dresser & 48 20
McKee
Ayres, Lewis, 6 36
Norris & May
Banister, Short, 0.6 66
Elliot, Hendrickson (max)
and Associates
Willard Schade & 10 26
Assoc.
Kaiser Engineers 13 26
Total
Hydraulic Carbon Effluent
Loading Depth Carbon Requirements2
(gpm/TtZ) (ft) Size (oxygen Demand)
4.3 17 8 x 30 TOD 30 mg/1
3.7 17 8 x 30 BOD 20 mg/1
3.3 15.5 8 x 30 BOD 10 mg/1
2.5 10 8 x 30 BOD 10 mg/1
7.3 26.8 12 x 40 BOD 10 mg/1
3.3 9 8 x 30 COD 112 mg/1
6.2 30 12 x 40 BOD 10 mg/1
4.2 36 12 x 40 BOD 10 mg/1
(max)
4.3 15 8 x 30 BOD 15 mg/1
4.6 16 12 x 40 BOD 45 mg/1
(90% of time)
1 Empty bed (superficial) contact time for average plant flow.
2 BOD: Biochemical oxygen demand
COD: Chemical oxygen demand
CO
-------
330
TABLE 22 "SUMMARY OF PCT PILOT-PLANT AND FULL-SCALE PLANT PERFORMANCES"
Effluent Effluent Effluent
Blue Plains
Pilot Plant
Owosso,
Michigan
Pomona,
California
Rosemount,
Minnesota (1st year
Rosemount, Minnesota
(last 3 to 4 months)
Battelle Pilot Plant
at Westerly
CRSD Pilot Plant
at Westerly
Estimated based on BOD similar to COD removals across clarifier.
2
Just around carbon columns.
Raw COD
(mg/l)
320
250-350
321
;ota
«)
t 527
437
COD
16
24-30
19
42
56
% Raw TOC
Removal (mg/l)
95 100
-91
94
-
92
87 90
TOC % Raw BOD
(mg/l) Removal (mg/l)
8 92 150
140
1201
230
240
21 77 206
BOD
(mg/l)
6
8
7.8
23
26
32
RemovaJ
96
84
78.
90
89
84
52
TABLE 23 "CARBON PILOT-PLANT RESULTS FOR PETROCHEMICAL AND REFINING WASTEWATERS"
Type of Wastewater
Refinery
Refinery
Refinery
Petrochemical
Refinery
Refinery
Refinery
Petrochemical
Design Q
(MGD)
28
1.9
22
3
26
28
8
29
Process
Application
Physical/Chemical
Physical/Chemical
Physical/Chemical
Tertiary
Tertiary
Tertiary
Tertiary
Tertiary
Influent
COD (mg/l)
600
800
670
150
100
300
100
150
Effluent
COD (mg/l)
103
201
143
49
41
50
40
48
Percent
Removal
83
75
79
67
59
83
60
68
-------
331
TABLE 24 "REFINERY WASTEWATER TREATMENT RESULTS"
Parameter
BOD (mg/l)
COD (mg/l)
TOC (mg/l)
Oil & Grease (mg/l)
Phenols (mg/l)
Chromium (mg/l)
Copper (mg/l)
Iron (mg/l)
Lead (mg/l)
Zinc (mg/l)
Sulfide (mg/l)
Ammonia (mg/l)
Cyanides (mg/l)
Turbidity, (TTU)
Color (Std. Color Units)
API Separator
Effluent
97
234
56
29
3.4
2.2
0.5
2.2
0.2
0.7
33
28
0.25
26
30
Carbon Treated
Effluent
48
103
14
10
0.004
0.2
0.03
0.3
0.2
0.08
39
28
0.2
11
15
Biologically Treated
Effluent
7
98
30
10
0.01
0.9
0.1
3
0.2
0.4
0.2
'27
0.2
17
15
Biological-Carbon
Treated Effluent
3
26
7
7
0.001
0.02
0.05
0.9
0.2
0.15
0.2
27
0.2
5
1
TABLE 25 "CARBON REGENERATION ACTIVITY ANALYSIS"
Iodine No.
Virgin Carbon
API Separator Effluent to Carbon
Biologically Treated Effluent to Carbon
1,010
906
991
Molasses No.
216
405
304
COD (mg/l)
SOC (mg/l)
BOD (mg/l)
TABLE 26 "PILOT-PLANT RESULTS-TERTIARY CARBON APPLICATION"
Influent
Concentration
600
500
400
300
200
300
200
100
50
250
200
150
100
50
20
^Throughput Rate = 0.6 - 1.2 Bed Volumes per hour
^Throughput Rate = 2.4 Bed Volumes per hour
Throughput Rate = 1.2 Bed Volumes per hour
Effluent
Concentration
280 ]
230
175
120
65
1402 2503
95 155
30 55
10 25
884 21 85
70 186
54 135
36 68
20 30
10 12
2
.Throughput Rate
Throughput Rate
Average
Percentage Removal
531
54
56
60
68
532 173
53 23
70 45
80 50
654 135
65 7
64 10
64 32
60 40
50 40
= 0.5 Bed Volumes per hour
= 0.15 Bed Volumes per hour
-------
332
TABLE 27 "DESIGN CRITERIA FOR THE ARCO CARBON PLANT"
Number of Rainfall Days Per Year (max.)
Maximum Rainfall Runoff Rate (per day)
Maximum Rainfall Runoff Rate (per year)
Average Influent COD Concentration
Average Effluent COD Concentration
Carbon Capacity
30 days
12.6 Million Gallons
378 Million Gallons
250 mg/l
37 mg/l
1 Ib carbon exhausted per
1000 gal. water treated (1.75
Ib COD removed/lb carbon)
TABLE 28 "ACTIVATED CARBON ADSORPTION DESIGN DATA AND ACTIVATED CARBON PROPERTIES"
Rated Flow (Each of Three Adsorbers)
Adsorber Diameter
Adsorber Bed Depth
Contact Time (Empty Bed)
Hydraulic Loading
Design Inlet Pressure
Pressure Drop Through Carbon
Carbon Inventory Carbon Bed
Adsorber Total
Theoretical Carbon Capacity
Carbon Dosage
667
gpm
10
45
40
8.5
60
35
92,000
100,000
0.3
ft
ft
min
gpm
psi
psi
Ib
Ib
IbT
Ib TOC/lb carbon
0.86 Ib carbon/1,000 gal Throughput
Activated Carbon Properties Filtrasorb 300
(8 x 30 Bituminous Coal)
Total Surface Area (N, BET Method)
Bulk Density
Particle Density Wetted in Water
Mean Particle Diameter
Iodine Number, minimum
Ash
Moisture
950-1050
26
1.3-1.4
1.5-1.7
950
Max 8%
Max 2%
m2/
Ib/ft"
g/cc
mm
TABLE 29 "THERMAL REGENERATION DESIGN DATA"
Furnace
Regeneration Rate
Steam Addition Rate
Fuel
Fuel Rate
Combustion Air Rate
Design Temperatures
Hearth 4
Hearth 6
Afterburner
Hearth 4
Hearth 6
Afterburner
Hearth 4
Hearth 6
Afterburner
60" x 6 Hearth with Integral Afterburner
125 Ib/hr
125lb/hr
Refinery Fuel Gas
188CFH
68 CFH
310 CFH
5,000 CFH
1,800 CFH
8,120 CFH
1,725° F
1,750° F
1,250° F
-------
FIGURE I
ADSORPTION OF DBS TO EQUILIBRIUM
IN CONTINUOUSLY STIRRED
FLOW SYSTEMS [2]
160
140
120
100
80
60
40
20
T 1 1 T
T T
k, = 1.25 mg/g-hr
= 1.78 mg/g-hr
k, = 2.80 mg/g-hr
-k0 = 4.28 mg/g-hr
\ I
1 I I I
15 21 27 33 39
TIME (hours)
45 51
ORGANIC CONSTITUENT IS DODECYL-BENZENESULFONATE (DBS)
57
333
2
O
00
rr
<
o
o
Ł
O
o
o
Figure 2
FREUDLICH ISOTHERM APPLICATION
' DATA CONFORM TO
-FREUNDLICH
- ISOTHERM
DECREASING
WASTEWATER
COMPLEXITY
NON-SORBABLE
RESIDUAL
FREUNDLICH ISOTHERM
APPLICABILITY
RESTRICTED TO
DEFINED LIMITS
EQUILIBRIUM CONCENTRATION, C (mg/l)
-------
334
FIGURE 3
CROSS-SECTIONAL VIEW
OF MULTIPLE-HEARTH FURNACE
CARBON IN
GAS OUT
Rabble Arm
Rabble Teeth
Steam
HEARTH
I (200°-300°F)
2 (300°-450°F)
3 (400°- IOOO°F)
4 (IOOO°-I600°F)
5 (I600°-I800°F)
6 (1600°-1800° F)
CARBON OUT
FIGURE 4
SIMPLIFIED FLOW DIAGRAM
SOUTH TAHOE
0=]
MIXED
MEDIA
FILTRATION
MIXED
MEDIA
FILTRATION
RECARBONATION
AND
SETTLING
AMMONIA
STRIPPING
TOWER
CHEMICAL
CLARIFICATION
FINAL EFFLUENT
PUMP STATION
-------
FIGURE 5
SIMPLIFIED FLOW DIAGRAM
COLORADO SPRINGS PLANT
Recycle
Effluent from
Primary Clarifier
Clarifier
Waste
Solids
Return
High Rate Filter
Dist. Box
,
J
ary
er
Dist. Box
Chlorination
Chamber
Irrigation
Canal
Reactor
Clarifier
Recarbonation
Reactor C02
H2S04
Carbon Columns
LAS Hi
MIX! "\
*- FLOCCULATOR
1
Raw
Waslewater
FIGURE 6
SCHEMATIC FLOW DIAGRAM
WESTERLY ADVANCED WASTE TREATMENT PLANT
PILOT PLANT
HELECTROLYTIC UNIT
FOR HYPOCHLORITE
GENERATION
CO
CO
Oi
-------
FIGURE 7
FREQUENCY ANALYSIS FOR EFFLUENT BODL25j
CRSD PRECARBON OZONATION STUDY
WESTERLY PILOT PLANT
FIGURE 8 r
FREQUENCY ANALYSIS FOR EFFLUENT COD1
CRSD PRECARBON OZONATION STUDY
WESTERLY PILOT PLANT
o
o
00
LJ
_
Ld
100
90
80
70
60
50
40
30
20
2 5 10 20 30 40 50 60 70 80 90 95
% OF THE VALUES LESS THAN STATED VALUE
• NO PRECARBON OZONATION
A OZONE <6.2 mg/l
O OZONE =6.2 mg/l
Q OZONE =3.2 —6.2 mg/l
98 99
Q
O
O
LJ
UJ
2 5 10 20 30 40 50 60 70 80 90 95 98
% OF THE VALUES LESS THAN STATED VALUE
9 NO PRECARBON OZONATION
A OZONE <6 2 mg/l
O OZONE =6.2 mg/ I
a OZONE =3.2 —6.2 mg/l
CO
CO
-------
337
FIGURE 9 r 1
FLOW DIAGRAM FOR THE GARLAND PLANTLJ
INFLUENT
t
t
IEQUALIZATION!
<
IPRETREATMENTI
1
* PRIMARY
9 CLARIFIER
1
4-
SLUDGE
CONDITIONING
"
TRICKLING
"" FILTERS
FILTER
PRESS
1
llNCINER/i
1
CHEMICAL
9 CLARIFICAT^
ILANDFI
TION|
LQ
1
1
FINAL
CLARIFIER
j -^RECARBONATIONl-^FILTRATIONlJ
L CARBON
ADSORPTION
< '
(DISINFECTION |
EFFLUENT
FIGURE 10
PHYSICAL/CHEMICAL PILOT PLANT SCHEMATIC
POMONA, CALIFORNIA
FLOCCULATION
ATMOSPHERE
""~r
1 MAKE-UI-
T r—WATER
buENCH
I JTANK
— ^t-*—v
EDUCTOR
MOTIVE WATER
COMPRESSED
AIR FOR
• PULSED A/ff
CLEANING OF BAGS
-------
338
g>
en
Q.
CO
V)
o
LU
X
LJ
FIGURE II
SULFIDES AND HEADLOSS vs TIME
CT
E
CO
LU
O
to
0
0
345 678
STUDY PERIOD (weeks)
FIGURE 12
BOD REMOVAL TREATMENT RESULTS
Q
O
CD
API EFFLUENT
A BIOLOGICALLY
TREATED
EFFLUENT
O CARBON TREATED
EFFLUENT
D CARBON TREATED
BIOLOGICAL
EFFLUENT
567 8 9 10
STUDY PERIOD (days)
12
13
14
-------
FIGURE 13
COD REMOVAL TREATMENT RESULTS
339
5 67 8 9 10 II
STUDY PERIOD (days)
12 13 14
API EFFLUENT
& BIOLOGICALLY
TREATED
EFFLUENT
O CARBON TREATED
EFFLUENT
D CARBON TREATED
BIOLOGICAL
EFFLUENT
FIGURE 14
ARCO FLOW DIAGRAM
WATSON REFINERY
PROCESS
WASTEWATER
API SEPARATOR
I
NEUTRALIZATION
I
CHEMICAL
FLOCCULATION
I
DISSOLVED AIR
FLOTATION
T
To LACSD
HOLDING BASIN FLOW TO
LACSD OR TREATMENT
SYSTEM DEPENDING
ON WATER QUALITY
CARBON
HOPPERS
FLOW DIVERTED TO HOLDING BASIN
WHEN RAINFALL EXCEEDS
O.I INCHES
REGENERATION
FURNACE
~l
L 1
*• FINAL EFFLUENT
CARBON
-------
FIGURE 15
PERFORMANCE OF ARCO CARBON PLANT
COD REMOVAL
400
O
-3-
CO
10 20 30 40 50 60 70 80 90
% OF THE VALUES LESS THAN STATED VALUE
99
FIGURE 16
PERFORMANCE OF ARCO CARBON PLANT
08G REMOVAL
I 10 20 30 40 50 60 70 80 90
% OF THE VALUES LESS THAN STATED VALUE
INFLUENT • EFFLUENT
D INFLUENT!
O EFFLUENT
-------
FIGURE 17
CARBON TREATMENT SYSTEM
REICHHOLD CHEMICALS CO.
341
Influent
1
EQUALIZATION
Nonionic
Poly
Acid.
ACID MIXING
TANK
FLOCCULATION
TANK
CARBON
ADSORBER
FEED SUMP
MULTIPLE-
HEARTH
FURNACE
CARBON
ADSORBERS
Sludge To
Disposal
V
I
I
I
I
CARBON
FIGURE 18
BP TREATMENT SYSTEM FLOW DIAGRAM
MARCUS HOOK REFINERY
Backwash
API Bottoms
itrifuge
API SEPARATOR SURGE BASIN
r--| T 1
1000
Treated
Process
Effluent
Once- through
Cooling Water
^
s<
\
Si
^
MULTIPLE-
HEARTH
FURNACE
Fuel
-— Steam
QUENCH
TANK
Air
CARBON ADSORBERS
-------
FIGURE 19
COD REMOVAL AS A FUNCTION r i
OF ADSORBER CONTACT TIMEl32J
PERIOD I
FIGURE 20
ADSORBER OIL REMOVAL AS A r ,
FUNCTION OF INFLUENT CONCENTRATION L32J
PERIOD I
O
•^
LJ
(T
Q
O
CJ>
100
90
80
70
60
50
40
30
20
10
0
100
90
80
70
< 60
O
LU 50
DC
O 40
30
20
10
0
40 42 44 46 48 5O 52 54 56 58 60
CONTACT TIME (minutes)
0 10 20 30 40 50 60 70 80 90 100
INFLUENT OIL (mg/l)
CN
^t
n
NITIAL REMOVALS
REMOVALS AFTER START OF BED PULSING
-------
CO
-*
n
FIGURE 21
FULL-SCALE PROCESS PERFORMANCE
BP MARCUS HOOK
FIGURE 22
EFFLUENT COD ATTAINABLE FROM
ACTIVATED CARBON SYSTEMS
100
90
80
70
60
^ 50
O 40
ŁE
PERIOD
30
20
10
PERIOD
PHENOLS
COD
a
o
o
z
Ul
3
u.
LL.
UJ
<
UJ
a:
UJ
o
z
o
I VIRGIN CARBON, NO FCCU FOUL CONDENSATE
2 CARBON BEDS NOT TURNED OVER, FCCU
CONDENSATE INCLUDED
3 CARBON BEDS TURNED OVER, FCCU
CONDENSATE INCLUDED
4 EFFLUENT SEPTUMS BENT, NO FCCU
CONDENSATE, ONE YEAR OPERATING
450
400
350
300
250
200
150
100
50
Systems treating
biological treatment
system effluent
A Only full-scale refinery
system treating total
waste
915
10
20 30
50
70 80 90
99
PERCENT OF PILOT-PLANT OR FULL-SCALE SYSTEMS
TREATING REFINERY OR RELATED WASTEWATERS
WITH LONG TERM MEAN COD LESS THAN
STATED VALUE
-------
FIGURE 23
INFLUENT AND EFFLUENT OIL AND
GREASE DISTRIBUTIONS FOR A
FULL-SCALE ACTIVATED
CARBON SYSTEM
o>
E
UJ
cr
(C
^
O
ISO -
I60
I40
I20
UJ IOO
CO
80
60
40
20
0
Average Influent = 30.2 mg/l
S D. = 36,3 mg/l
Coef. Var = 1.20
No. of Data Points - 55 over 5 months
Average Effluent - 7.4 mg/l
S.D. = ll.l mg/l
Coef. Var = I. 5
No. of Data Points = 50 over 5 months
FIGURE 24
CARBON ADSORPTION CAPACITY
FOR VARIOUS PLANTS
u.y
— 0.8
c
o
_a
o 0.7
X)
^ 0.6
T3
-
OQ.2
Q_
<
0 O.I
°c
A A
/
rl-
A
— Blue
S ^
Ay-Ł
\-Tahoe, Ci
C
/-
/0
-Owosso, Mich.
Plains (DC.)
A / '
fc-Clevt
'osemount, Minn.
I/if.
) ^^
^
'-Colorado Sprin
land Westerly
^'
^
gs, Colo.
/-^1
0
^-^
[
E
o
^-^"
3
3
) 100 200 300 400 500
INFLUENT COD TO CARBON COLUMN (mg/l)
PERCENT OF THE VALUES LESS THAN
STATED VALUE
A Municipal Wastewater
O Refinery Wastewater
D Petrochemicals Wastewater
CO
-------
CAUTIONS AND LIMITATIONS ON THE APPLICATION OF ACTIVATED
CARBON ADSORPTION TO ORGANIC CHEMICAL WASTEWATERS
C. T. Lawson
and
J. C. Hovious
Union Carbide Corporation, South Charleston, West Virginia
OBSERVATIONS ON ACTUAL WASTEWATERS
Normal practice in assessing the feasibility of activated carbon adsorption is to first
conduct batch adsorption isotherms with powdered activated carbon. These tests provide
a relative indication of the amount of organic removal achievable by adsorption and the
ultimate adsorptive capacity of the carbon. Unfortunately, these data are useful only in
a relative sense - for comparing the relative merits of two different carbons, or for
comparing the relative amenability of different wastewaters to adsorption. Isotherm data
are not suitable for designing continuous granular caibon adsorption columns, since the
dynamic effects and interactions in a continuous bed differ too greatly from the batch
equilibrium situation in an isotherm.
Batch adsorption isotherms have been performed on many organic chemical bio-
treated effluents and raw wastewaters of widely varying composition and overall organic
concentration (COD or TOC basis). Figure 1 shows curves of percent organic removed vs.
carbon dosage. It is obvious the amenability of these wastes to adsorption is rather wide -
ultimate percentage organic removals vary from 45% to 90%. Even for those cases where
90% organic removal can be achieved by massive carbon dosing in batch tests, the
organic removals achievable in continuous adsorbers are lower, often significantly so.
Comparing batch and continuous data for Plant A bio-treated effluent:
Ultimate % removal of TOC in isotherm test 90%
%TOC removal in continuous bed adsorber prior to breakthrough
(1 bed vol/hr) 52%
%TOC removal before breakthrough (0.5 bed vol/hr) 64%
Not only was the organic removal predicted by isotherms not achieved, even in early
stages of the column run with fresh carbon, but doubling the bed depth (halving the
throughput rate) only raised the amount of TOC removal from 52 to 64%.
When batch adsorption data are plotted as log (wt. organic adsorbed/wt. carbon) vs.
Jog (organic concentration remaining in solution), the Freundlich isotherm results - X/M
vs. C . When the logarithmic plot is extrapolated to C = C (the initial solution concen-
tration), the resulting (X/M)_ is called the "adsorptive capacity" of the carbon. These
N_-
isotherm capacities are usually°greater, often very much greater, than the ultimate
adsorptive capacity exhibited by a granular adsorption bed at exhaustion. Typical examples
from prior publications (1):
345
-------
346
AdsorpHve Capacity, gm COD/gm carbon
Wastewater Isotherm Column at Exhaustion
Plant A bio-effluent 1.3 0.26
Plant D bio-effluent 0.65 0.24
Plant G raw wastewater 0.075 0.05
Note that carbon had a very low adsorptive capacity for the constituents in Plant G
wastewater.
It was further noted that increasing the carbon bed depth yielded not only the afore-
mentioned limited increase in percentage organic removal, but also a diminished adsorptive
capacity. Operating two beds in series until each bed was exhausted at an overall through-
put of 0.5 bed vol/hr (1 BV/hr/column) showed:
gm COD adsorbed/gm carbon
CoTTlCol. 2
Plant A 0.26 0.17
Plant D 0.23 0.13
One may postulate that selective adsorption of the "more adsorbable" organics occurs in the
lead bed. Not only that, when the first bed is exhausted, the second bed has little
remaining capacity - implying that the "less readily adsorbed" species, which are initially
removed in the second bed, block its further usefulness in removing "readily adsorbed"
species (available when the first bed breaks through). This phenomenon is shown by the
breakthrough curves of Figure 2. Just how "easily adsorbed organics" may be differentiated
from "less readily adsorbed" species is a difficult question. The rate of adsorption at the
carbon surface, the rate of diffusion within the carbon pores, and the ultimate capacity of
the carbon for the adsorbates all play a part in affecting "adsorbability." In isotherm tests
these factors are reflected in the measured adsorptive capacity and limiting percentage
organic removal. In continuous columns they are reflected in the organic removal achieved
and the rapidity with which organic breakthrough and exhaustion occur. The absence, or at
least minimization, of diffusion limitations in powdered carbon isotherms can give a markedly
different picture of "adsorbability," in terms of both extent and capacity, compared to the
complex inter and intra particle diffusion phenomena which occur in a granular carbon bed.
PURE COMPOUND ADSORPTION STUDIES
Because of these observations of low percentage removals of TOC and COD, limited
benefits of increasing carbon bed depth, and rather limited adsorptive capacities (in
continuous columns), an extensive investigation was conducted (2) to quantify some aspects
of "adsorbability" of specific organics. The effects of functionality, molecular weight and
structure, aqueous phase pH, solubility, polarity, carbon surface structure, and adsorbate
interactions were investigated. Column vs. isotherm comparisons were also made for several
pure chemicals and simple mixtures.
In the first phase, 93 organic chemicals from 11 functional group families were
subjected to single dosage adsorption tests.
Functional group families:
Alcohols
-------
347
Aldehydes
Amines and alkanolamines
Aromatics
Pyridines and morpholines
Esters
Glycols and glycol ethers
Ketones
Organic acids
Ethers and oxides
Halogenated aliphatics
Test conditions:
Initial compound concentration = 1000 mg/l (or solubility, if 1000 mg/l)
Carbon: 5 gm/l of powdered Westvaco Nuchar WV-G (surface area,= 1100 m /gm)
Contact time = 2 hours
Typical graphs of adsorbate loading vs. molecular weight are shown in Figures 3, 4,
and 5 for the alcohols, esters, and organic acids tested. The unfavorable effect of branching
and the favorable effect of unsaturation can also be seen. Figure 6 is a "map" of percent
of compound adsorbed (by 5 gm/l carbon) vs. molecular weight showing the regions which
encompass all the experimental data.
Region A Aromatics (including ring-substituted compounds)
Region B General trend for aliphatic mono-functional oxygen and nitrogen-
containing compounds
Line C Saturated organic acid line, shown for comparison
Region D Glycols, glycol ethers
Note that the aromatics, even those containing polar groups (-OH, -NO~, -Cl) are quite
amenable to adsorption, while the poly-functional oxygenated compounds (glycols, glycol
ethers) are especially difficult to adsorb.
Qualitatively, it was concluded that adsorbability, as reflected in constant dosage
tests, is favored by increasing molecular weight, aromaticity, and degree of unsaturation.
Increasing polarity, solubility, branching, and degree of dissociation (for amines and
organic acids) tended to severely limit the extent of adsorption.
The relative ease of adsorption for simple oxygenated organic compounds may be
summarized as follows:
Number of carbon atoms 4
acids aldehydes esters ketones alcohols glycols
Number of carbon atoms 4 (straight-chain compounds)
acids aldehydes alcohols esters ketones glycols
The relatively low adsorption, in general, of compounds containing 4 carbon atoms in
these tests is particularly noteworthy. The carbon dosage of 5 gm/l was a rather massive
dosage - in a continuous column adsorber it would be equivalent to treating only about
24 gallons of wastewater per pound of carbon. This poor organic removal at high carbon
dosage implies that carbon adsorption may have limited application to many organic
chemical manufacturing wastewaters, from a cost-effectiveness viewpoint. Low molecular
-------
348
weight oxygenated organic compounds represent a very large fraction of the production
volume in the industry.
In the next phase of the study, Freundlich adsorption isotherms were obtained on 1000
mg/l solutions of five different organic compounds (all containing four carbon atoms) using
four different carbons at different aqueous-phase pH levels. Figure 7 shows that pH had a
marked effect on the isotherm adsorptive capacity, (X/M)- , for butyraldehyde and ethyl
acetate. Data for Carbon A, Witco 517, are shown; similar trends were observed for other
carbons. Gas chromatographic analyses indicated that, at acidic or basic pH levels, ethyl
acetate was hydrolyzed into its less adsorbable (lower molecular weight) components. At
elevated pH, butyraldehyde underwent an aldol condensation into a higher molecular weight
(more adsorbable) component. More detailed discussion of differences between the carbons
tested and the effects of carbon surface properties are contained in Ref. (2).
A potential problem in successfully applying activated carbon to multi-component
wastewaters is indicated. It is not possible to select a pH level that assures maximal
adsorption of all wastewater constituents. An acidic pH is required to facilitate adsorption
of organic acids (to avoid ionization), near neutral pH favors adsorption of esters, while a
highly alkaline pH is necessary to maximize adsorption of aldehydes.
MULTI-COMPONENT ADSORPTION STUDIES
Interactions in multi-component systems were examined in the third phase of the study
by adsorption isotherm tests on binary and four-component mixtures. The results are shown
in Table 1.
In a test (at pH = 7) of a butanol/ethyl acetate binary mixture (500 mg/l of each
compound), the total adsorptive capacity was observed to be 0.237 gm cpds/gm carbon,
while the sum of the two adsorptive capacities (at 500 mg/l initial concentration) from
pure-component isotherms was 0.27 gm/gm. By gas chromatographic analysis of the residual
solutions, the butanol adsorbate loading was 1.05 times the loading in a pure component test
while the ethyl acetate loading was 0.76 times the pure component loading. Even though
more ethyl acetate was adsorbed from the mixture, on a weight basis, than butanol (as
expected from Figure 7), the interaction had a greater deleterious effect on ethyl acetate, in
terms of using pure component data to predict mixture adsorbate loadings.
By contrast, in the butyraldehyde/MEK binary test, the total adsorptive capacity from
the mixture (0.275 gm cpds/gm) was greater than the sum of the pure component values
(0.270 gm/gm), more butyraldehyde (the "more adsorbable component") was adsorbed on a
weight basis, and the ratio of mixture loading to pure component loading was greater for
butyraldehyde, 1.28 vs. 0.60.
In both four-component mixture isotherms (Table 1), the total adsorptive capacity (0.29
and 0.225 gm/gm for Carbons A and C, respectively) was significantly less than the sum
of the pure component capacities (0.575 and 0.312 gm/gm for Carbons A and C, respectively).
Also, in the Carbon C test, less butyric acid was adsorbed than butyraldehyde or ethyl
acetate, despite the previous pure component conclusion that acids are more adsorbable than
-------
349
other organic compounds.
Thus, while pure component adsorption data are useful in determining which waste
streams are potential candidates for activated carbon treatment, the data cannot be used
to quantitatively predict adsorption from multi-component systems.
CONTINUOUS vs. BATCH ADSORPTION STUDIES
In the fourth and final phase of the study, continuous column, granular carbon
adsorbers were operated to exhaustion for comparison of ultimate capacities with isotherm
capacities. Each column was operated at a throughput rate of 1 bed volume/hour. The
feed concentration was 1000 mg/l in the pure component tests, 2000 mg/l (500 mg/l each
compound) in the four component test (feed TOC = 1200 mg/l). The four component
mixture was butyraldehyde, ethyl acetate, butanol, and butyric acid. In all cases, the
ultimate capacity (at exhaustion) was less than that predicted by isotherms (83 to 89%).
Adsorptive Capacity, gm/gm carbon
Column Column
Organic Isotherm at Exhaustion at Breakthrough
MEK 0.160 0.132 OVK)3
Butyraldehyde 0.220 0.192 0.141
4-Component Mixture 0.170 (TOC) 0.151 (TOC) 0.124 (TOC)
Note that the usable capacity in a single-column system - the capacity achieved at
"breakthrough," where significant organic concentrations appear in the treated effluent -
is significantly less than indicated by the isotherms (64 to 72%). Figure 8 shows the experi-
mental breakthrough curve for the four-component mixture.
A second important criterion when applying activated carbon in continuous beds is the
amount of wastewater that can be treated before breakthrough (often called the "carbon
dosage"). Breakthrough in full-scale practice (where the carbon is removed from service)
would be set at the maximum organic concentration allowed by effluent standards. For
purposes of discussion, breakthrough was set at the 95% organic removal level in these
lab tests. Observed dosages were:
Test Breakthrough "dosage" gal WW treated/lb carbon
MEK 12.7
Butyraldehyde 19.4
4-Component Mixture 12.4
Treating large volumes of wastewater at these inlet concentrations will, thus, lead to
rather short service life (rapid breakthrough), even with the fairly ideal adsorption wave-
front shape in Figure 8.
Hydraulic effects (channeling and axial dispersion) are a partial explanation for the
failure of column adsorbers to achieve adsorbate loadings predicted by isotherms. The
inherent differences between the static equilibrium attained in batch isotherms and the
dynamic situation in a continuous column are also limiting factors, particularly when
treating complex multi-component ( 4) wastewaters.
A recent paper (3) by Keinath has shown that competitive adsorption, even in binary
-------
350
solutions, can lead to chromatographic displacement effects in column adsorbers treating
wastewaters of varying composition. In this situation, the weakly-adsorbed solute is retained
by the carbon for a time and then is eluted as a rather concentrated "peak" upon prolonged
contact with the feed wastewater. Keinath further showed that fluidized bed adsorbers are
much less susceptible to this effect than more conventional packed beds - a finding that could
have significant impact on adsorber design. Chromatographic displacement and related
interaction effects are almost certainly key contributors to the diffused, spread-out adsorption
wavefronts and fluctuating organic removals observed when treating multi-component organic
chemical wastewaters of high component and concentration variability. The question of
dynamic interactions and effects in multi-sorbate systems is an area where further research and
development is sorely needed, not only to shed further light on which wastes can be efficiently
treated by activated carbon adsorption, but also to ascertain how best to design and operate
continuous adsorption systems to minimize the undesirable effects of interactions.
CONCLUSIONS
1. In pure component studies, specific organic chemicals have been shown to differ widely
in their amenability to adsorption, depending on molecular weight, structure, polarity
and solubility. Low molecular weight oxygenated organics are particularly difficult to
adsorb efficiently.
2. The relative ease of adsorption of different functional group compounds can vary strongly
with pH, depending on the chemical nature of the adsorbates. An optimum pH cannot
be predicted for a multi-component wastewater of unknown or varying composition.
3. While pure component data could be used to predict binary adsorption capacity in
isotherms fairly closely, a four-component mixture isotherm showed only about 60% of
the adsorptive capacity predicted. Mutual solubility effects competition for adsorption
sites, and inability to maintain a pH level optimum for all components contributed to
this inability to extend pure component data to more complex mixtures.
4. While isotherm capacities were somewhat extrapolatible to continuous column behavior
in pure component adsorption and simple mixture studies, such extrapolations have not
proved to be possible with real wastewaters (complex mixtures), particularly with bio-
treated effluents.
5. More importantly, the key parameters of interest in real wastewater treatment situations -
percentage organic removal achievable and water volume treated per pound of carbon
before breakthrough - cannot be predicted from isotherm tests.
6. The physical differences between an equilibrium adsorption situation in a powdered
carbon isotherm and the dynamic, multi-component interactions in a continuous granular
carbon bed are too great to permit prediction of column performance from isotherms or
pure component data. The dynamic interplay of adsorption rates, pore diffusion rates,
hydraulic effects, and pure component chemical properties in continuous columns is an
area where further research efforts must definitely be applied.
REFERENCES
(1) Lawson, C. T., and J. A. Fisher, "Limitations of Activated Carbon Adsorption for
Upgrading Petrochemical Effluents," Water - 1973, AlChE Symposium Series No. 136
(1974)
-------
351
(2) Giusti, D. M., R. A. Conway, and C. T. Lawson, "Activated Carbon Adsorption
of Petrochemicals/1 JWPCF, 46, No. 5, May (1974)
(3) Keinath, T. M., "Design and Operation of Activated Carbon Adsorbers Used for
Industrial Wastewater Decontamination," paper presented at 68th Annual Meeting -
AlChE, Los Angeles, Nov. (1974)
DISCUSSION
Frank Manning: Would you care to comment on how you would personally combine
activated carbon with biological treatment together, which one is put first?
C. T. Lawson: Our normal thinking is that we would put carbon before biological treatment
on specific streams where it has an application. In other words, on a stream say that's
noxious for some reason but not biodegradable or perhaps biologically inhibitory, then we
consider that a very good application for consideration of carbon. As a post-biological
polishing step, it depends on how close we are to meeting a permit. I guess, if a little
more treatment would get us to the permit then I would say a tertiary carbon system might
be justified. I don't think, and we are convinced of this fact, that you can simply add on
a carbon system and make up for deficiencies in your biological system. If you are far
away from meeting a permit, the first think to do is make the bio-system work. Then if you
need carbon to close the gap and it is effective in doing so, put it on; but I don't think that
carbon can ever be considered as a prescription for a poorly operated treatment system.
Sterling Burks: Have you performed any studies on absorption of trace heavy metals? I
know this question was asked of Davis Ford.
C. T. Lawson: No, we haven't.
Milton Beychok: As you know, I have congratulated you once before because I think it is
long overdue that someone brought some science to this field. If we look at the data you
have given us this morning, the isotherms in thermodynamic terms, the isotherms are really
equilibrium. Your column data telling us you can't approach that equilibrium the way you
think you should and that involves kinetics, and if we look again and make an analogy
between that and other refining processes and what has been done in improving kinetics and
the approach to equilibrium through catalysis. Do you intend or do you know of any research
going on this pretreatment of carbon, trying to do something to improve the kinetics as well
as perhaps making kinetic models and seeing what you can learn?
C. T. Lawson: Well the question of pretreating carbon, I really don't know of any work
that is going on. It seems to be the kind of thing that the carbon vendors would be
interested in and they are pretty competitive and closed-mouth fellows; they don't tell you
much about what they are doing. There is some potential here, I would say, for tailoring
carbons through pretreatment or through activation processes; but it is not being really
studied much to my knowledge. On the question of developing a diffusion model or a
kinetic model, there is quite a bit of academic interest I think in doing this. The only
papers that I have seen are really for binary systems and that is not of too much interest in
a practical application sense. I know Dr. Keinath at Clemson University and Dr. Weber
at the University of Michigan have done some work in this modelling area.
-------
352
Anonymous: It just seems to me from what we all know that the regenerated carbon behaves
differently than the fresh carbon and you showed us today percentage wise some remarkable
differences between two different carbons, that it would be a fruitful area for research work
whether it be on the part of Union Carbide or the carbon manufacturer.
C. T. Lawson: Yes, I don't really want to give a commercial for anybody, but I will say
that Witco's carbon in the studies we looked at performed quite well, very good carbon. I
think there has been a lot of reluctance over the years to use anything but bituminous coal
carbons because its "sturdier, more rugged, easier to regenerate, not as friable as the other
carbons." Looking at waste treatment with some of the speciality carbons, coconut-shell
carbons, has shown some pretty good results, - a little bit better than bituminous coal carbons.
But when you talk about 75-85$ per pound for carbon, it is really out of the question for
large volume waste water treatment.
Nick Sylvester: You mentioned at the end of your talk that you have done some more research
that you have not reported on and you also mentioned this apparent chromatographic type
phenomena in the columns. Have you done more work on that and are you going to report
that?
C. T. Lawson: That is what we have tried to do, we have tried to take a column and take
samples down through the column at different times and see just where specific compounds end
up in the column on a dynamic basis, to see if there is some way that you can predict when
something is going to be spit out as a big eluted peak, and I hope that data will be in shape
for the Cincinnati paper.
Dave Skamanca: Davis Ford mentioned that if you are going to put in an activated carbon
system, you are really better off with biological system in front of it to decrease the loading
and apparently improve the performance of the carbon. But as you decrease the loading,
the pounds in the concentration to the activated carbon system, you also decrease the isotherm
data, you don't get as good a loading on the activated carbon. Are you getting ahead of
the game in most of these systems by trying to remove more of these pounds up stream with
a carbon, are you gaining on the problem, are you just staying even so to speak?
C. T. Lawson: My own opinion is that the kind of capacities that you see in isotherm tests
are really an interesting qualitative observation, but they have no practical significance.
Anything you can take out before the carbon column by reasonable treatment means, I think
is worthwhile. From what we have seen, running continuous column systems, this 25% or
so by weight absorptive capacity (on a COD basis) to petrochemicals seems to be a pretty
real number and relatively independent of the feed concentration in continuous systems for
the wastes we have examined thus far. The fact remains that isotherm capacities obviously
do vary strongly with concentration, you can go up and down. We really see no way to
relate that to a continuous system. We use our isotherm data just for relative comparison
of one waste water to another and one carbon to another; we don't use it in a quantitative
sense at all.
-------
353
BIOGRAPHY
Cyron T. Lawson is a Project Scientist in the
Water Quality Development Group of Union
Carbide's Chemicals and Plastics Division,
Research and Development Department. He holds
the B.Ch.E. and M.S.Ch.E. degrees from Georgia
Institute of Technology. He has taught graduate
courses in water and wastewater treatment as an
Adjunct Instructor at the West Virginia College of
Graduate Studies. Cyron is a Registered
Professional Engineer (Chemical) in West Virginia
and is a member of the AlChE and its Environmental
Division.
Joseph C. Hovious is a Group Leader/
Technology Manager in the Research and
Development Department of Union Carbide's
Chemicals and Plastics Division. Mr. Hovious
is an M.S. Environmental Engineer from the
University of Illinois. He teaches graduate
courses at the West Virginia University
College of Graduate Studies. He is a member
of the WPCF, AWWA, and a Registered
Professional Engineer in West Virginia.
-------
GO
TABLE 1 "COMPARISON OF MULTI-COMPONENT ISOTHERM DATA WITH SINGLE COMPONENT DATA'
Mixture 1 - Carbon A
Butanol
Ethyl Acetate
Mixture 2 - Carbon A
Methyl Ethyl Ketone
Butyraldehyde
Mixture 3 - Carbon A
Butyraldehyde
Ethyl Acetate
Butanol
Butyric Acid
Mixture 3 - Carbon C
Butyraldehyde
Ethyl Acetate
Butanol
Butyric Acid
Equilibrium Carbon Loading
in Mixture Test
g cpd/g carbon (a)
0.116
0.121
0.237
0.063
0.212
0.275
0.072
0.075
0.031
0.112
0.290
0.072
0.066
0.023
0.064
0.225
Equilibrium Carbon Loading Fraction of Single Component
in Single Component Tests, Equilibrium Loading
g cpd/g carbon (b) from Mixture (c)
0.110
0.160
0.270
0.105
0.165
0.270
165
160
110
140
0.575
0.080
0.080
0.068
0.084
0.312
1.05
0.76
0.60
1.28
0.44
0.47
0.28
0.80
0.90
0.83
0.34
0.76
(a) Compounds each initially present at 500 mg/l. Mixtures 1 and 3 dosed at 1 gm carbon/liter and Mixture 2 at
2 gm carbon/ liter. Equilibrium loadings (X/M)_ are calculated from total TOC loadings and chromatographically
determined fractional compositions of removed material.
(b) Equilibrium loadings measured at C of 500 mg/l.
(c) Calculated by dividing the equilibrium loading of the individual components in the mixture by the single-component
equilibrium loading.
-------
FIGURE 1
ACTIVATED CARBON TREATMENT OF BIOLOGICAL EFFLUENTS
BATCH ADSORPTION ISOTHERM TESTS
355
100 ._
CQ - initial COD (or TOO concentration before carbon treatment
Plant E, C0 = 288 mg COD/1
Plant A, C0 = 360 mg TOC/1*
~
D Plant F, CQ = 2300 mg TOC/1
Plant C
C0 = 1363 mg COD/1
Plant A, C0 = 1030 mg COD/1*
*Plant A tests run at different times,
several months apart
l i i
i i i i
20
Carbon Dosage, gm/1
30
FIGURE 2
BREAKTHROUGH CURVE FOR PLANT A BIO-TREATED WASTEWATER
40
hfl
e
Q
o
u
1000
800
600
400
200
Feed = C
0.5 BV/hr
0.88 liters/BV
(2 col.)
I
50 100 150 200
Wastewater Throughput, bed volumes (BV)
-------
FIGURE 3
EFFECT OF MOLECULAR WEIGHT ON AMENABILITY OF ADSORPTION OF ALCOHOLS
FIGURE 4
EFFECT OF MOLECULAR WEIGHT ON AMENABILITY OF ADSORPTION OF ESTERS
0.200 .
»
z~
-o
m
CO
0.150 _
0.100
0.050
C0* 1000mg/ I at Alcohol
5gm/l Carbon Dosage
*C0- 700 mg/l
Butanol
n-Hexanol
O 2 Ethyl -Butanol
n-Amyl Alcohol
OIsobutonol
O t- Butanol
Propanol
Q Olsopropanol
"yl Alcohol
Ethanol
Methonol
40 60
Molecular Weight
2-Ethyl Hexanol
120
0.200
0.150
9
z"
•V
•x.
0.100
0.050
C,,= 1000 mg cpdY I
5 gm carbon/ I
j L
Butyl
Acrylat*
Butyl
Acetate
Ethyl Acrylcte j/
Propyl Acetate
Isopropyl Acetate
n-Amyl
Acetate
Ethyl Acetate
1 Methyl Acetate
4-
40 80
Molecular Weight
120
-------
FIGURE 5 357
EFFECT OF MOLECULAR WEIGHT ON AMENABILITY OF ADSORPTION OF ORGANIC ACIDS
0.200
0.130 _
1
o
t»
»
a
X
0.100
O.050
C0»IOOO mg cpd/ I
Sgm carbon/1
O Btnzoic
Acrylic
O
'Propionlc
r Acetic
Formic
40 60
MolMular WilgM
120
FIGURE 6
PERCENT OF COMPOUND ADSORBED VS. MOLECULAR WEIGHT FUNCTIONALITY EFFECTS
100
80
Seo
m
,Aromatic
s
in
o
20
Q) Glycols,
Gly-Ethers
i ' ' I 1 1 1 1 1 1 1
20 40 60 80 100 120
MOL. WT.
140 160 180 200
-------
358
FIGURE 7
pH EFFECTS ON ADSORPTION REMOVALS OF SELECTED ORGANIC COMPOUNDS
WITH CARBON A
0.5
0.4
0.3
I
o
u
Q.
U
I
x 0.2
O.I
Butyroldehyde •
Ethyl Acetate
wuiuin/i x.
Methyl Ethyl Ketone'
-L
6
PH
10
12
FIGURE 8
u
o
1200
1000
800
600
400
200 L
BREAKTHROUGH CURVE FOR ACTIVATED CARBON ADSORPTION OF A
FOUR-COMPONENT MIXTURE IN A CONTINUOUS SYSTEM
Feed
TOC adsorbed' 0.151 gmTOC/gm carbon
40 60
Throughput , bed volumes (at I BV/Hj/*
80
-------
CASE HISTORY: USE OF POWDERED ACTIVATED CARBON
IN AN ACTIVATED SLUDGE SYSTEM
Joyce A. Rizzo
Sun Oil Company, Marcus Hook, Pennsylvania
INTRODUCTION
In an effort to expand the performance of existing biological treatment facilities, full
scale trials utilizing powdered activated carbon were conducted at Sun Oil Company's
Corpus Christ!, Texas Refinery. The main objective of the trials was to reduce the effluent
suspended solids loading for compliance with the 1977 NPDES and State permit limitations.
Three separate trials were initiated over a one-year period comprising five months of
operation with powdered activated carbon addition to the aeration system of the 2.16
MGD biological treatment plant. Although the powdered carbon addition could not
guarantee compliance with the 1977 suspended solids' criteria, improvement in the existing
system's performance was significant for COD and BOD removal as well as suspended solids.
The treatment costs ranged from 1.7 to 4.3^/10 gallons depending on the influent flow
and quality. Data have been compiled showing reductions in the refinery's final effluent
loadings of up to 56% for suspended solids, 36% for COD and 76% for BOD.
FACILITIES
A simplified flow diagram of the treatment scheme at the Corpus Christi Refinery is
shown in Figure 1. The system's operating conditions are outlined on Table 1 . The daily
average charge to the plant is only half the design capacity; although the peak design
rate is often operated during periods of rain and high ballast loading. The refinery waste-
water flow (averaging about 600 GPM) is equalized with ballast and contaminated storm
water in a 10 MM gallon pond. The combined flow from the pond (averaging about 750
GPM) is pumped through a dissolved air flotation unit for removal of any excess oil and
suspended solids. The effluent from the DAF is split between two 760,000 gallon
rectangular aeration basins. Air is introduced into each of the basins by two 75 hp
floating mechanical aerators. When operating both basins in parallel at average flow
conditions, the retention time is about 22 hours; at peak flow, this is reduced to 11 hours.
The mixed liquor volatile suspended solids is maintained at about 2200 ppm. With both
systems operating, the food to mass ratio averages only about 0.13 Ib COD/lb MLVSS or
0.07 Ib BOD/lb MLVSS in each basin; this, of course, can be doubled by operating only
one aeration system.
Each aeration system is coupled with a 55' diameter x 12' deep clarifier. The over-
flow from the clarifiers combines for discharge. The settled sludge is collected from each
clarifier, combined and returned to the aeration basins. The return rate usually averages
about 50% of the charge rate.
359
-------
360
OBJECTIVE
Basically, the performance of the biological treatment system is satisfactory and the
effluent readily complies with the refinery's NPDES and State permits. With the exception
of suspended solids, the limits established for compliance in 1977 are also currently being
met. The consistently effective removal of suspended solids in the final clarifier seems to
be the only area of concern. Even with substantial organic removal in the biological
system, there appear to be two major factors causing the extra suspended solids carryover
from the final clarifiers:
Variability of influent organic and hydraulic loadings. The biological population
is thus unstable with erratic growth and variable efficiency.
Excessive aerator foaming in the aeration basins. This foam, laden with biological
solids, tends to carryover into the clarifier and out to the effluent. The problem
becomes most severe at high flow periods.
Interim limitations had been set by the State and NPDES permits allowing time to
investigate in-plant improvements which could possibly permit compliance with the 1977
TSS criteria without construction of additional facilities. Full scale trials appeared
necessary due to the nature of the problems noted — bench scale studies tended to show
excellent settling characteristics which do not reflect the hydraulics of the operating system.
Initial testing was performed with the addition of polymers to the aeration basin effluent.
The polymers did improve the solids settling somewhat but their benefit was overshadowed by
the problems still being caused by aeration foaming and shock loads. Conversely, defoamers
were effective in controlling the foam, nothing more.
In the refinery, every precaution is taken to reduce the possibility of shock loads and
spills. However, no matter how efficient, there is no way to completely eliminate the
variability: turnarounds, emergency shutdowns and the like. This, coupled with the influx
of highly variable loads of ballast, dictated that the treatment plant operation would have
to compensate for unstable influent conditions to the plant.
APPLICATION OF POWDERED ACTIVATED CARBON
The addition of powdered activated carbon was initiated in an effort to resolve all the
operating problems causing the suspended solids carryover — the objective being compliance
with the 1977 suspended solids' limitations. Although this objective was not achieved, some
interesting data resulted. A summary of the three separate carbon trials conducted is shown
on Table 2. Unfortunately, it was not possible to operate a "blank" system simultaneously
with the carbon system. Therefore, the data obtained 30 days prior to the start of each
trial has been utilized as the base period for that trial. As can be noted on Tables 3, 4 and
5, for the most part, the influent loadings of each base period and respective trial are
comparable; this is especially true of Trial 3. During each trial, operation was maintined
as closely as possible to the conditions observed during the base period. No special problems
were encountered that would lead one to believe that any of the trials and base periods'
operation were different, especially on the average. Although, the comparison is not ideal,
it is relevant, especially in respect to the "real world" effect on the plant's effluent. The
-------
361
percent reductions shown between the base period and each trial are meant as a guide to
point up the improvement in effluent quality; the actual values, of course, cannot be
taken as absolute. From an operation's standpoint, there appears to be no question that
the addition of the powdered activated carbon to the aeration system realizes a significant
improvement in the effluent quality and reduces its variability.
The first trial was considerably longer than the last two — mainly, because the
carbon dose was gradually built up in the system over a four-week period; whereas in the
other trials, the carbon was batch loaded in a couple of days up to the operating level.
The gradual addition of the carbon allowed the daily observation of its effect in order to
determine the appropriate loading to be maintained. In order to minimize the amount of
carbon needed to load and maintain the system, only one aeration basin was operated
during the trials; the same was true of the respective base period. The carbon was
manually batch added daily to the dissolved air flotation effluent as it entered the
aeration basin. During the first trial it was found that the optimum operating carbon level
needed in the aeration system was 450 ppm with approximately 1000 ppm resulting in the
recycle sludge. All three trials were conducted at these levels. This requires about
6100 pounds of carbon to charge one aeration system up to operating levels. An easy
indication of the carbon level in the system was found to be the foam level in the basin —
the appropriate amount of carbon would eliminate the foam completely; as the carbon
level dropped, the foam would build up. Of course, laboratory analysis utilizing a
standardized comparison procedure confirmed the operating level of carbon on a concen-
tration basis.
While gradually adding the carbon daily, it was noted that a distinct improvement
in the plant performance did not come about until the 450 ppm level was attained —
however, by letting the level drop off gradually, the peak performance could be main-
tained down to 300 ppm. It was also noted that more frequent additions of smaller
quantities appeared more effective than larger batch additions less frequently.
Referring to Table 2, you will note that the carbon addition rate was much higher
during the first two trials; this came about for a couple of reasons:
' More frequent batch additions of larger quantities ( 1000 pounds at a time) were
necessary to compensate for shock loadings. Although the presence of the carbon
in the system greatly reduced the chances of a shock that could inhibit the
biological activity, it could not guarantee that a reduction in activity would not
occur. However, in general, it was found that if caught in time with the addition
of a large batch of carbon, a lapse in system efficiency could be effectively
brought back up to normal almost immediately without adverse effects on the
effluent quality.
' At the time of the third and most recent trial, the biological system was more
stable and it took less carbon to obtain the same results. Better methods of
equalization had been put into operation and ballast was being successfully
treated at a rateable basis.
The flow rate during the third trial averaged 800 GPM versus 625 - 660 GPM for the
first two trials (this was due primarily to increased ballast water treatment). The higher
-------
362
3
flow with the reduced carbon addition rate accounts for the lower 1.7^10 gallon treatment
cost compared to Trial Ts 4.3$. It is important to note that these costs only reflect the
carbon usage, the manpower utilized in adding the carbon is not included. More efficient
means of carbon addition could, of course, be employed with minimal capital investment
but was not deemed necessary for a trial basis.
On Table 6, the treatment costs obtained are summarized along with the reduction in
effluent loading observed during each trial.
From this data, one is erroneously led to believe that the efficiency seemed to improve
because of a lower carbon dose (and subsequently lower cost). As explained, the higher
carbon dosages observed during Trials 1 and 2 are directly attributable to a higher frequency
of shock loadings to the system. The only valid conclusions to be drawn from the costs
vepsus efficiency data presented is that the carbon treatment costs range from 1 .7$ to 4.3$/
10 gallons depending on the variability of the influent and the frequency of batch additions
to solve overload problems. Again, it is important to remember that the percent reductions
are determined from the trial versus its pre-trial base case and should only be thought of as
a guide for comparison purposes.
Total Suspended Solids Reduction. The average effluent suspended solids data obtained
are outlined on Table 3. The reduction in the effluent suspended solids' loading was
dramatic, between 49.3% and 55.7%. The existing permit limit is easily met; however,
the effluent, at the very low level of 405 Ibs/day daily average for Trial 3, still would not
comply with the 327 Ibs/day limit of the 1977 permit. In addition, although the daily
maximum values while using carbon were significantly lower and under better control,
several peaks still would occur. Referring to Table 7, in the case of Trial 3, four data
points out of 20 were above the 561 Ib/day 1977 allowable limit, for 80% compliance;
100% compliance is easily obtainable with the existing permit. Of course, the odds of
being able to comply with the 1977 limits are better, but still not acceptable.
Referring to the frequency distributions for Trial 3 plotted on Figures 2 and 3, the
reduction in effluent suspended solids with carbon addition becomes very obvious. The
chance of complying with the 561 Ibs/day 1977 maximum was increased from 40% during
the base period to 82% with carbon; of course, these percentages are only relative not
absolute. Although the permit is written on a mass basis, the effluent concentration of
suspended solids is a basis more easily compared between various facilities and is there-
fore, shown in Figure 3. On each figure, the flat slope of the carbon trial versus the
base period also indicates the more consistent effluent obtainable with carbon treatment,
the variability having been significantly reduced.
Organic Removal Efficiency - as mentioned, the addition of powdered carbon also
improved the organic removal efficiency of the system. A summary of the influent and
effluent COD loadings of the system are shown on Table 4. This table also points up the
uniformity of influents between the base period and trial period for Trials 1 and 3. As
mentioned, the operation was maintained as consistent as possible for each case. During
Trial 2 the influent COD loading was about 25% higher than during the base period —
yet, the effluent loading was 30% lower showing a significant reduction on a relative
base. The removal efficiency during Trials 1 and 3 was increased significantly up to 70%
-------
363
from 62.9% and 54.8% respectively for each trial. The ultimate reduction in COD
loading to the effluent being 35.7% for Trial 3 as compared to base period 3.
The daily COD removal efficiencies for Trial 3 are plotted in Figure 4. It can be
seen that not only was the removal efficiency significantly increased, the variability
of the efficiency was reduced. Again, resulting in an effluent of consistently better
quality.
Figure 5 is a frequency distribution plot of the effect of the carbon addition on the
effluent COD concentration for Trial 3. Not only does the slope again show more
consistent results but, for example, the 95% point has been reduced from 280 ppm down
to 150 ppm, a 46% reduction assuming a comparable base.
The effect of powdered carbon on BOD is summarized in Table 5. It is obvious from
the influent loadings that the data obtained during Trial 3 are most relevant since the
base period and trial had the most comparable influent conditions. Again, as in the case
of COD, during Trial 2, the influent loading was 38% higher than the base period; yet,
the effluent was 51% lower. The reduction in BOD was most sharply noted during Trial
3 when the effluent BOD never exceeded 6 ppm — the reduction from the base period
was 76%.
The frequency distribution comparing the base period and Trial 3 is shown in Figure
6. Again the results are significant. The curve for the effluent BOD during the carbon
addition period is virtually flat, showing definite uniformity and consistency of effluent.
Figure 7 denotes the distribution of the percent BOD removals obtained — again, the
curve is relatively flat, the variability being virtually eliminated.
Oil Removal. No data were compiled on oil and grease removal since the system is
not required to handle any large amounts of oil. The effluent oil and grease usually
averages less than 5 ppm with an occasional peak around 10 ppm; the influent is normally
20-25 ppm. Any improvement would be difficult to observe at these low levels. On
occasion, however, the influent oil level can be higher, as did happen during the post
period of Trial 2 when there was an upset in the Refinery. The carbon still in the system
readily adsorbed the excess oil, keeping it in the sludge. Of course, the sludge had to
be removed from the system, thus spending the carbon. However, a possible effluent oil
and grease violation was avoided at only the expense of a carbon charge.
SPECIAL NOTE
Sludge wasting had always been erratic and unpredictable in this system due to the
variable influent and unstable bug growth. When as stable as possible, about 15,000
gallons per day of recycle sludge would be wasted from one operating aeration system.
During the carbon trials the wasting was reduced to 10,000 gallons per day under similar
circumstances. The basis for this reduction being that the sludge was considerably
thicker so less volume needed to be handled to dispose of the same mass.
-------
364
SUMMARY
As evidenced by the data presented, the addition of powdered activated carbon to the
aeration system of the refinery's existing wastewater treatment facilities did not reduce the
effluent suspended solids enough to comply with the 1977 permit limitations. However,
significant improvements in the system's performance were observed as follows:
Improved organic and suspended solids removals and reduced effluent loadings
% REMOVAL
Trial 3 Without* Carbon With Carbon % Reduction in Effluent Loading
COD 5478 7075 3577
BOd 93.5 98.5 75.8
TSS - - 49.3
*Pretrial base period
More uniform effluent quality
Clearer effluent
Elimination of foam in the aeration system
More consistent sludge wasting at 2/3 the volume
Reduced chances of biological upsets -- the use of powdered carbon does not
eliminate biological upsets but it does appear to reduce the opportunity of their
occurance; and when they do occur, it appears to maintain the effluent quality
under better control. When caught in time, it was found that a reduction in
biological activity can effectively be brought back up to normal by massive batch
addition of carbon to the existing operating level.
These improvements were obtained by building up the aeration system to a 450 ppm
operating level of carbon with about 1000 ppm in the recycle sludge. The system was
maintained at this level by the. batch addition of about 100 Ibs/day of carbon (or 10 ppm)
at an average cost of 1 .7^/10 gallons of water treated. The amount of carbon necessary
to operate the system for peak performance could be readily determined by observing:
The aeration basin foam level - the proper operating level of carbon eliminates
foam in the system.
Clarity of the effluent - the presence of carbon in the system removes the tint
usually characteristic of biological effluent.
It is important to note that the powdered activated carbon utilized in these trials has
a high bulk density of 44 Ibs/cu. ft. This becomes an important factor when improved
settleability is the primary objective of the carbon addition to the system. In addition,
the higher bulk density reduces the opportunity for carbon loss to the effluent.
DISCUSSION
James F. Dehnert: We have done considerable work with powdered carbon on a pilot basis
and we have done some of the work that you did. A couple of problems we ran into and I
wondered if you had solved them, that is the analysis of the mixed liquor solids. Did you
determine actual carbon percentage or was this a calculated value?
Joyce Rizzo: We did it by a standard comparison. What we did was make up standard
-------
365
filters with carbon which we compared. We took samples of our aeration system before
we added carbon to it. We added known dosages of carbon and set up a set of standards.
So all the numbers we have obtained are 50 ppm, but we could keep track of it pretty
easily.
James F. Dehnert: We had not devised a system to determine how much carbon was
actually in our system at any one time. The second question: you said you didn't test
very many parameters. We have a problem in California with meeting fish toxicity, did
you ever run any tests on that?
Joyce Rizzo; Everyone knows that refinery effluent is not toxic. Maybe California
is different - they may have a higher breed of fish down there, I don't know. Obviously,
we didn't evaluate toxicity.
Les Lash; You mentioned holding about 450 milligrams per liter of activated carbon,
Joyce, in your aeration basin, I guess. What is the retention time in that aeration basin?
Joyce Rizzo: About 12 hours at average operating conditions.
Les Lash; When we ran the pilot plant on municipal waste for the EPA on powdered
activated carbon in Salt Lake, we used reactive clarifiers and if you didn't allow enough
settling, of course the effluent went over black. Now that is a little different than the
clear you were talking about.
Joyce Rizzo: The only time we ever had a carryover problem with carbon was during a
time when the bug population was upset and they would tend to take the carbon with
them. If they were going out, they weren't going to leave the carbon behind. On a
normal basis we had no problem with carbon carryover at all. It stayed right with the
sludge. In fact, it improved the settleability tremendously. If you look at the clarifier,
it has the appearance of being black; of course, the aeration system would turn black
versus the dark rich brown color and it appeared like it was all carbon. The carbon
color would take over completely, and if you looked at the clarifier surface, it appeared
black but the effluent water was virtually clear.
Milton Beychok: I am just a bit curious, I don't know if there is anyone here from
duPont or not, but something like four years ago, duPont first used this and called it
PACT, powdered activated carbon treatment, and they were licensing it. Does the use
of this involve any patent problems or royalty on the part of duPont?
Joyce Rizzo: No, this is quite a different system than what duPont is utilizing in their
PACT project, they are also utilizing a regeneration system which I believe is also a
part of their patent on that. We used ICI carbon, and the system that we used was out-
lined to us from ICI as to how to operate. I don't think we did anything really different.
I have talked to many people who are also using powdered carbon in their aeration
system in a similar manner.
Ed Sebesta: I note you have a centrifuge for your waste activated sludge there. Did
-------
366
you continue to use a centrifuge both before and after carbon addition and what was the
effect?
Joyce Rizzo: We don't use a centrifuge all the time by the way, only when we have to.
We have plenty of land out there for land farming and a lot of times put the sludge out wet
from the aerobic digester. We did not see any real difference in our sludge handling
characteristics on the solid waste handling system - none at all. Actually, the amount of
carbon that we used - our aerobic digester is three-quarter million gallons like our aeration
system - and the amount of carbon that we used and the amount that was wasted was
virtually lost in the system.
Ed Sebesta: I have another question and you indicated that "the effluent was very clear. I
think the data indicated about an average of 40 parts per million suspended solids. This
is pretty high for a very clear effluent, do you have any explanation for the concentration
numbers?
Joyce Rizzo: Very fine solids. Color really is what I am talking about. 40 ppm does not
represent optimal operation, 10-20 ppm would be more representative of the best effluent
obtainable with the carbon. The solids that would settle out by the way would be ones
that would settle out after hours. You could take a sample of the effluent and you wouldn't
think there were suspended solids in it, but if you let it sit for a while, the solids that
would settle would be very, very fine. I would say about half the solids that went out in
the effluent were carbon solids. Very, very fine solids.
Pat DeJohn: I'd like to make a couple of comments. One that Mr. Lash made, the carbon
that you were using has a density of about 45 pounds per cubic foot, and consequently it
settles very rapidly. I believe what they were using out at Salt Lake was a carbon that
had a density of about 15 pounds per cubic foot. The higher density is the reason the
carbon doesn't go out over the weir. The other thing about the patent situation, duPont
just got a patent issued about two months ago on the PACT process and right now there are
some negotiations going on with respect to that.
Janis Butler: You mentioned sometime just spreading the wet waste activated sludge. Do
you experience any odor problems?
Joyce Rizzo: No - none at all. If we put it out wet, the solids settle rapidly, and there
is no odor problem at all. We are using (I have been gone now from Corpus Christ! for
about six months), but at the time I left we were using the centrifuge all the time for the
waste.
Anonymous: Do you add carbon continuously?
Joyce Rizzo: Corpus Christ! has just purchased another load of carbon for usage. As I
understand it, their main goal is to use it mostly for batch addition for shock loadings and
things like that.
-------
367
Fred Goudy; Does the use of powdered carbon have any long range implications with
regard to the operation of the activated sludge plant at Corpus Christ!?
Joyce Rizzo: They are using it.
Fred Goudy: Continuously?
Joyce Rizzo; I am not sure, I can't answer that for fact right now. I know they have
just purchased some more carbon, and I know they have it in stock and I know that their
intentions are to use it. Now whether they will use it all the time, I don't know at this
time. I might mention what I think about carbon application in the aeration system.
Like I said, although we really were oriented to suspended solids removal, it is obvious
that some of the other things that I have mentioned about the system are very true. That
is, if you have a plant that is sitting there hydraulically loaded and you need more
capacity, it seems to help and gives you that extra capacity that you need. As I
mentioned none of the reductions that I presented, of course, should be taken as absolute
numbers because it was pretrial versus trial period. But there is no question that the
improvement in the plant performance is there and I think the improvement is really
dependent on the variability of your influent.
Dave Story: You mentioned adding polymers; could you elaborate a little bit more on
that?
Joyce Rizzo: Yes, we didn't do too much work with polymers because it was the very
first thing we tried to reduce suspended solids. We were adding about 5 or 6 parts per
million of some polyelectrolytes to our aeration basin overflow and we saw a reduction
in effluent suspended solids and a little bit better settling characteristics, but nothing
very significant. I think this is mostly due to our foaming problem because we still had
a lot of foam carrying over to the clarifier which is a significant problem.
Anonymous: How do you account for the reduction in sludge wasting?
Joyce Rizzo: We didn't have to maintain as many bugs in the system and they didn't have
to work as hard. But the real reduction was in volume not mass. The recycle sludge
was thicker and more uniform with carbon addition.
BIOGRAPHY
Joyce A. Rizzo is a staff engineer in the
Advanced Management and Methods Department of
Suntech, Inc., a subsidiary of Sun Oil Company.
She is based in Marcus Hook, Pennsylvania. Ms.
Rizzo joined Sun Oil in 1971 as a process engineer at
their Marcus Hook Refinery. Prior to her move into
Suntech six months ago, she spent three years in
Process Engineering at Sun's Corpus Christ! Refinery.
She holds a B.S. in Chemical Engineering from
Northeastern University in Boston.
-------
368
TABLE 1
"OPERATING CONDITIONS OF WASTEWATER TREATMENT PLANT1
FLOW
Daily average 1.08MGD
Daily maximum 2.16 MGD
Design 2.16 MGD
BIOLOGICAL SYSTEM 2 - 760,000 Gal. Rectangular Aeration Basins
2-75 HP. Floating Aerators/Basin
Recycle 50% of Charge Rate
MLSS 3000 PPM
MLVSS 2200 PPM
Retention 11 Hours at Maximum Design Rate
(Per Basin) 22 Hours at Average Operating Rate
Sludge Age 12 Days
SETTLING 2-55 Ft. Diameter Circular Clarifiers 12 Ft. Deep
Average Rise Rate, GPD/FT2 227
Peak Rise Rate, GPD/Ft 455
SOLIDS HANDLING
1 760,000 Gallon Aerobic Digester
1 Solid Bowl Centrifuge
4 Acres of Land Farm
Wasting Rate 30,000 GPD
VSS 7,500 PPM
TABLE 2
"SUMMARY OF POWDERED ACTIVATED CARBON ADDITION TRIALS"
Aeration System: 450 PPM PAC
Recycle System: 1000 PPM PAC
Average Treatment
Carbon Dosage ,, Cost*
Period Duration Lbs/D Lbs/IO Gal PPM $/10 Gal
0.20 24 4.3
0.16 19 3.5
0.076 9 1.7
July, 1975 31 Days
"At 22 <:/Lb
Trial
B>-:se Case
Trial
Base Case
Trial
Base Case
1
1
2
2
3
3
Sept. -Nov., 1974
August, 1974
May, 1975
April, 1975
August, 1975
July, 1975
82
30
28
30
26
31
Days
Days
Days
Days
Days
Days
179
152
87
-------
369
TABLE 3
"EFFECT OF POWDERED CARBON ON DAILY AVERAGE EFFLUENT SUSPENDED SOLIDS'
1977 Permit: 327 Lbs/Day Daily Average
Base Case 1
Trial 1
Base Case 2
Trial 2
Base Case 3
Trial 3
Flow
GPM
626
629
645
657
839
799
Effluent
PPM
115
50
163
72
79
42
Lbs/Day
861
381
1262
565
799
405
Reduction
55.7
55.2
49.3
Base
Trial
Base
Trial
Base
Trial
1
1
2
2
3
3
TABLE 4
'EFFECT OF POWDERED CARBON ON COD"
Base Case 1
Trial 1
Base Case 2
Trial 2
Base Case 3
Trial 3
Flow
GPM
626
629
645
657
839
799
Influent
PPM
459
457
343
444
367
379
Lbs/D
3445
3446
2658
3500
3698
3632
Effluent % %
PPM
170
135
266
183
166
112
Lbs/D
1277
1020
2059
1445
1670
1073
Removal
62.9
70.4
22.5
58.7
54.8
70.5
Reduction
20.1
29.8
35.7
TABLE 5
'EFFECT OF POWDERED ACTIVATED CARBON ON BOD"
Flow
GPM
626
629
645
657
839
799
Influent
PPM
152
213
152
227
188
207
Lbs/D
1144
1607
1173
1898
1895
1981
PPM
15
15
30
13.
12
3
Effluent
Lbs/D
116
114
232
5 113
124
30
Removal
%
89.9
92.9
80.2
94.0
93.5
98.5
Reduction
%
1.7
51.3
75.8
-------
370
TABLE 6
'COMPARISON OF CARBON TREATMENT COSTS
AND SYSTEM PERFORMANCE"
% Reduction*
Trial
Trial
Trial
1
2
3
Dose
PPM
24
19
9
3
-------
FIGURE I
TREATMENT PLANT FLOW DIAGRAM
BALLAST
a
TANK FARM
REFINERY
API
SEPARATOR
EFFLUENT
EQUALIZATION
a
IMPOUNDMENT
POWDERED
| CARBON
| ADDITION
AERATION
FINAL
EFFLUENT
AEROBIC
DIGESTER
^ • v*L.I1 1 l\lr UOC r
^
SLUDGE
TO
LAND a
FARM
-------
372
FIGURE 2
EFFLUENT
TSS
LBS/DAY
2800
2400
2000
1600
1200
800
400
0
I
EFFECT OF POWDERED CARBON
ON EFFLUENT TSS LOADING
CARBON
i i
I I
I
I
10 30 50 70 90 95
% OF VALUES LESS THAN
99
TRIAL 3
240
200
160
EFFLUENT
TSS 120
PPM
80
40
0
I
FIGURE 3
EFFECT OF POWDERED CARBON
ON EFFLUENT TSS
BASE CASE
TRIAL 3
WITH CARBON
j i
i
10 30 50 70 90 95
% OF VALUES LESS THAN
99
-------
COD
REMOVAL
FIGURE 4
COMPARISON OF COD REMOVAL EFFICIENCY
373
TRIAL 3
CONTROL »-H TRIAL
PERIOD AUGUST
JULY, 1975 1975
EFFLUENT
COD
PPM
300
280
240
200
160
120
80
40
0
I
FIGURE 5
EFFECT OF POWDERED CARBON
ON EFFLUENT COD
BASE CASE
CARBON
TRUL 3
J_
10 30 50 70 90 95
% OF VALUES LESS THAN
99
-------
FIGURE 6
EFFLUENT
BOD
PPM
60
50
40
30
20
10
EFFECT OF POWDERED CARBON
ON EFFLUENT BOD
TRIAL 3
I
„- WITH
CARBON
_L
5 10 30 50 70 90 95
% OF VALUES LESS THAN
99
BOD
REMOVAL
100
98
94
90
86
82
78
74
-WITH _..
- CARBON
FIGURE 7
EFFECT OF POWDERED CARBON
ON BOD REMOVAL
i i i i i
TRIAL 3
10
30 50
70
90 95
99
o/
•'O
OF VALUES LESS THAN
-------
"PILOT PLANT ACTIVATED CARBON TREATMENT OF
PETROLEUM REFINERY WASTEWATER"
Jack H. Hale, Leon H. Myers, and Thomas E. Short, Jr.
Robert S. Kerr Environmental Research Laboratory
INTRODUCTION
One of the first documented uses of carbon for environmental control was in 1793
when a physician used charcoal to remove the odor associated with gangrene. About
sixty years later a scientist by the name of Stenhouse recommended the use of charcoal to
remove the odors from sewers. Potable water was "purified" by carbon in 1862. ^ Moon-
shiners in the hills of Tennessee were using charcoal, not to mellow their product, but to
remove "hog track" odors.
Since the mid-1960's, there has been an increasing effort to utilize activated carbon
as a secondary or tertiary treatment system to treat wastewater. Two of the most success-
ful uses of activated carbon were demonstrated at Lake Tahoe and in Pomona, California,
to treat domestic municipal wastewaters. It is, of course, a natural progression to treat
industrial wastewaters with systems that appear successful in treating municipal waste-
waters.
The first domestic petroleum refinery to use activated carbon treatment was the
Atlantic Richfield Refinery in Carson, California. The system was designed for inter-
mittent use to treat rainfall runoff and process wastewater during storm periods. The
second application was designed for the BP Refinery, Marcus Hook, Pennsylvania, to
treat process wastewaters. Unlike the Arco system, the BP system was designed to operate
in a continuous mode. Neither system relied on biological treatment preceding the
activated carbon system.
During the late 1960's, activated carbon treatment of industrial wastewater was
gaining momentum as the treatment system of the future. Statements such as "organic
removal" or "removal of dissolved organics from wastewater has been demonstrated" and
"activated carbon treatment systems will remove dissolved organic contaminants" were
prevalent. These innocent gross statements were translated rather rapidly to imply that if
you had an organic waste treatment problem, it could be solved with granular activated
carbon.
With the complete cooperation of Kerr-McGee Petroleum Refinery, Wynnewood,
Oklahoma, a study program was devised to investigate the adsorptive capacity of
activated carbon.
ISOTHERM STUDIES
Adsorption isotherms are used to indicate the effectiveness of an activated carbon
for a specific wastewater under controlled conditions. When the isotherm is to be used as
375
-------
376
a predictive tool, the wastewater should be as representative as possible, pH and
temperature adjustments should not be made, and sample collections need to be made in
glass containers.
CARBON EVALUATION
Eleven commercially available carbons were evaluated for this petroleum refinery's
wastewater. It is necessary to point out that adsorption varies with the carbon being
evaluated and the water sample. The data presented is based on specific activated carbon
for one petroleum refinery. Comparable results may or may not be achieved using the
same carbon at another refinery.
The controlled conditions used for comparison purposes of the eleven carbons are
shown in Table 1 . Both pulverized and granular modes were evaluated.
The waste sample to be evaluated was collected from the API separator discharge,
settled for four hours, and the candidate water was drawn from the middle of the vessel.
The adsorption capacity of one gram sample of the candidate carbons, both pulverized
and granular, is shown in Table 2.
Graphically expressed, the comparison of pulverized and granular adsorption
capacities appears in Figure 1 .
CONTACT TIME
A major factor relating to adsorption is the amount of time the water to be treated is
in contact with the activated carbon. A sample of wastewater from another refinery was
obtained for these studies. Pulverized activated carbon (Filtrasorb 400) was weighed to
obtain 0.1, 0.5, 1.0, 2.0, and 5.0 grams of sample, and the isotherm procedure was
followed for three hours. Samples were obtained at 20, 40, 60, and 180 minutes,
filtered, and analyzed for total organic carbon. Table 3 represents the typical results.
ISOTHERM TESTING
Figure 2 shows the results of an isotherm carried out on wastewater from the refinery
used in this study and a candidate activated carbon. Obviously, this curve does not
follow the typical "Freundlich" isotherm. Instead of the usual straight line, a curve
resulted. At the lower end of the curve the amount of carbon added to the wastewater is
increased but the concentration of TOC does not decrease as much as would be expected.
In fact, the concentration of TOC changes only very slightly at high carbon dosage.
Refinery wastewater is composed of a rather complex mixture of materials. Some of these
materials are readily adsorbable while others are not. The initial amount of carbon picks
up the easily adsorbed materials and leaves behind the others. Therefore, increasing the
carbon dosage does not reduce the concentration of TOC to the same degree. In fact,
there is a possibility that this wastewater contains non-adsorbable components that cannot
be removed by carbon treatment.
-------
377
From the "isotherm," the adsorptive capacity of the carbon, tested with this
specific wastewater, projects to approximately 0.1 Ib of TOC per Ib of carbon. This was
considered to be a reasonable value and indicated that further investigation of activated
carbon treatment was justified.
Results from isotherm tests are considered to be valuable only from an initial
screening standpoint. Their results are generally not sufficient for predicting results of
full-scale granular activated carbon treatment. For a better evaluation of activated
carbon treatment, continuous column pilot plant investigation at the waste stream site
yields much more reliable results.
ANTHRAFILT FILTRATION
When treating oily wastes by activated carbon, it appears that a pretreatment system
is needed to assist in the removal of insoluble oils and suspended matter. An anthrafilt
downflow filter system was evaluated for effectiveness, again using filtrable TOC as the
prime parameter. The glass column used for this exercise was five feet in length and 1 .5
inches in diameter. A diagram of the system is shown in Figure 3. Total organic carbon
results obtained at these time intervals are shown in Table 4.
This particular study lasted 2.5 hours before a pressure drop was noted. The charge
water contained an appreciable quantity of oil, and the oil percolated into the anthrafilt
layer during the study period. When the column was backwashed with distilled water,
the percolated oil was easily removed from the anthrafilt, leaving an apparently clean
bed.
MINI-COLUMN STUDY
A downflow mini-column system was designed using stock one inch ID glass tubing
that was six feet in length. The system was designed with an electronic sampler to
composite hourly samples. A diagram of this system is shown in Figure 4.
The primary study involved dividing the six columns into two sets of three columns
each to evaluate comparison of treatment effectiveness. Each column was packed with
1,000 cc of Filtrasorb 400 using potable water was the wetting liquid. The flow rate was
set at 400 ml/minute and a column pressure of 6-7 psig was used. Total organic carbon
was the primary parameter used for treatment effectiveness. Analytical results of the
primary study are recorded in Table 5.
Comparing each column for removal efficiency, using the anthrafilt effluent TOC as
the base, the percentages obtained are shown in Table 6. Columns 1 and 2 compare very
well for the five-day averages, although they do exhibit considerable daily variances.
Columns A-3 and B-3 are not agreeable, and no reason for this deviation is projected.
A second study was conducted by connecting five columns in an upflow mode. Each
column contained 1,000 cc of Filtrasorb 400 packed in the same manner as the previous
study. TOC results of this study are shown in Table 7.
-------
378
Percentage reductions of each column based on the anthrafilt effluent are shown in
Table 8. The average column trend indicates there is an increase of treatment effective-
ness with each succeeding column, as one would expect.
PETROLEUM REFINERY DATA
Listed in Table 9 is the pertinent information needed to characterize the refinery
where the study was conducted. This refinery was a 30,000 barrel, Class "B" refinery
(using the API classification system); processes included fluid cat cracking, HF alkylation,
catalytic reforming, and asphalt production.
The existing refinery eastewater treatment system is shown in Figure 5. Wastewater
from the refinery includes cooling tower blowdown, boiler blowdown, oily process water,
stripper effluent, and contaminated runoff. These wastes flow to an API separator for oil
removal. The effluent from the separator is treated in a "Pasveer Ditch" activated sludge
treatment system. The biologically treated effluent then flows through a series of holding
ponds and the effluent from the final pond is discharged to a small stream.
Two complete pilot plants were installed and operated simultaneously, one on the
refinery's API separator effluent (secondary) and the other on the clarifier effluent from
the biological treatment system (tertiary).
ACTIVATED CARBON TREATMENT PILOT PLANT
Figure 6 contains a flow diagram of the pilot activated carbon treatment systems.
The wastewater to be treated first flows through a dual-media filter constructed of 4-inch
PVC pipe. This filter consisted of an 18-inch layer of sand over pea gravel, topped with
a 6-inch layer of anthrafilt.
Dual-media filtration pretreatment was chosen because of its reliability and effective-
ness, and because it does not require the use of iron or aluminum salts as coagulants.
This latter point is particularly important insofar as regeneration of activated carbon is
concerned, since aluminum salts during regeneration can remain on the surface of the
carbon at high temperatures. These salts can become permanently attached to the surface.
Thus, the effective surface area of the carbon is reduced and its adsorption capacity is
seriously reduced. Iron salts present a similar problem . In addition, these salts can
catalyze oxidation reactions of the carbon and the gases in the regenerator. Thus, the
structure of the carbon becomes permanently damaged.
After pretreatment, the wastewater entered a "Calgon" activated carbon pilot plant.
This plant was set up so that the wastewater flowed down three of the 5-inch ID columns.
The first column contained an 18-inch layer of granular activated carbon while the
remaining columns had a 36-inch layer of carbon.
The flow rate through each pilot plant was adjusted to 1/4 gpm. During the
operation of the pilot plant, samples of the API separator effluent, biological treatment
effluent, and both pilot plants' effluents were taken every two hours. These samples were
-------
379
composited and preserved according to recommended EPA methods. Twenty-four hour
composite samples were analyzed daily for a spectrum of water pollution control
parameters using EPA analytical methodology and analytical quality control.techniques.
The dual-media filter and carbon columns were backwashed whenever the pressure in
the first column exceeded 20 psi. The pilot plants were operated over a 10-day period,
at the end of which time the first columns in both plants were near exhaustion.
Figure 7 shows the BOD5 daily composite analysis for the API separator effluent
before and after treatment by the various schemes studied. Activated carbon treatment
was not able to achieve the same level of BOD5 reduction as bio-treatment. In fact,
bio-treatment did a considerably better job. However, carbon treatment following bio-
treatment did show further reductions in BOD.
Figure 8 shows the COD daily composite results for the same wastewater streams.
Apparently, both bio-treatment and carbon treatment effect about the same COD
reductions. Carbon treatment following bio-treatment yielded the best reduction of COD.
The TOC results for the daily composites are shown in Figure 9. Unlike the BOD5
and COD results, the TOC indicates increased removal can be obtained by the API-
carbon combination over the bio-treatment system alone. There is an erratic behavior
pattern which indicates unadsorbable organics passing through the carbon columns.
From the BOD5 and COD plots, it may be concluded that for secondary treatment
alone, on the wastewater studied, the biological treatment system is preferable because
it gives the greatest BOD^ reduction and gives COD reduction equivalent to carbon
treatment. The best levels of reduction were obtained with biological treatment followed
by carbon adsorption.
Table 10 gives the results of other parameters evaluated in this study. As would be
expected, both biological and activated carbon treatments are able to produce signifi-
cant reduction in the organic parameters, such as BOD, COD, TOC, oil and grease, and
phenols. It should be noted that carbon has a particular affinity for the removal of
phenols. The color and turbidity are also improved. Cyanides and ammonia, for all
practical purposes, were not removed by either of these treatments.
Sulfides are a peculiar problem for activated carbon treatment. Apparently, the
sulfide content increases as it goes through the carbon s/stem. This is probably due to
biological activity occurring at the surface of the carbon. The sulfur containing
compounds which are adsorbed upon the carbon are reduced anaerobically to h^S giving
rise to the increase in sulfide content. Bio-treatment, on the other hand, decreases the
sulfide content very significantly.
One of the least expected removals by activated carbon was observed for the metals
—chromium, copper, iron, and zinc. These metals are significantly reduced by the
carbon. Whether these removals are due to adsorption, filtration, or other phenomena
has not been evaluated. However, it must be kept in mind that these removals were
-------
380
accomplished with virgin carbon. The effect of/some of these metals upon carbon
regeneration has not been determined. Also, whether regenerated carbon can remove
equivalent amounts of metals has not been established.
As mentioned previously, the isotherm study indicated that the carbon could adsorb
0.1 Ib TOC/lb of carbon. The columns on the API separator e/fluent indicated a capacity
of 0.31 Ib TOC/lb of carbon, as shown in Table 11 . The most probable explanation for
the difference in the capacity as determined in the isotherm test and the pilot plant study
is biological activity in the columns.
In the case of the API separator effluent carbon column, this biological activity
manifested itself by the anaerobic production of H^S. While the increased capacity is a
desirable situation, the production of sulfides is not. Although anaerobic activity was
not apparent in the clarifier effluent columns, it was observed that algae started growing
in the carbon beds. In fact, it was necessary to cover the columns to prevent sunlight
from making this growth possible.
Because of the considerable cost of granular activated carbon, regeneration of the
spent carbon is essential if the cost of treatment is to be kept at a moderate level.
Samples of spent carbon were taken from the first column of each pilot plant and sent to
Calgon Corporation for regeneration studies. Calgon regenerated these two carbons using
their standard evaluation techniques in a muffle furnace at 1750° F and with a normal air,
flue gas, and steam atmosphere.
As shown in Table 12, both carbon samples were regenerated to a good activity as
indicated by the iodine numbers obtained. The regenerated clarifier carbon appeared to
have a little better activity than the API separator carbon. In general, the iodine number
can be related to the surface area of pores larger than 10 A in diameter. The molasses
number, likewise, is related to the surface of pores larger than about 30 A. As can be
observed in Table 15, there is an increase in the molasses number with a slight decrease
in iodine number. This indicates a shift in pore size distribution. Apparently, the
smaller pores are being destroyed while larger pores are being created. Final clarifier
carbon appeared to have less destruction of the pore matrix than did the API separator
carbon.
GAS CHROMATOGRAPHY STUDIES
Studies by Dr. T. C. Dorris and Dr. S. L. Burks at Oklahoma State University have
shown that some organics present in refinery wastewaters are toxic to fish, and some
organics are refractory to biological degradation. These organic chemicals persist in
lakes and streams for long periods of time. It is also noted that a refractory organic can-
not be detected by the TOD test that has been used in the past to evaluate the efficiency
of treatment systems. Tests such as TOD, TOC, and COD are much more realistic for the
indication of organics in a water sample. Of these tests, TOC is the one that measures
only organics.
The laboratory researchers must go further and use sophisticated methods for actually
identifying and characterizing the refractory organics so that engineers will know the
-------
381
type of organics that must be removed, thus designing treatment systems to improve the
quality of water discharged to the receiving streams.
Carbon treatment is a relatively efficient method for removing organics from refinery
wastewater and may be used as a secondary or tertiary treatment system to reduce the BOD
of effluent water to a very low level. In studies to learn more about carbon treatment,
carbon columns were installed at a petroleum refinery wastewater treatment plant at a
point immediately after the API separator and at a point after the final clarifier for the
activated sludge treatment unit. The carbon removed from these columns was extracted
with several organic solvents to determine which solvents were most efficient and to
provide samples of the organics for silica gel, gas chromatographic (GC), infrared, and
GC-mass spectrometric analyses.
About 100 ml of an air dried carbon sample were weighed and placed directly into
the soxhlet extraction apparatus with a glass wool plug at the bottom to prevent the
carbon from getting into the siphon tube. Three hundred ml of solvent were used in the
flask. The rate of solvent reflux was adjusted to give four cycles per hour, and the
extraction was continued for 22 hours. After 22 hours, the solvent-oil mixture was
cooled to room temperature and filtered through a 10 cm depth of anhydrous sodium
sulfate crystals contained in a 20 mm by 25 cm glass tube to dry the mixture and remove
any carbon particles that might be present. The dried solvent-oil mixture was then
distilled through a 20 mm diameter column containing 12 cm of glass beads until the
volume remaining in the flask was slightly less than 50 ml. The concentrated solvent-oil
mixture was measured into a sample vial, and the volume adjusted to exactly 50 ml with
rinsings from the flask.
A gas chromatograph, with a flame detector and an electronic integrator, was used
to determine the amount of oil extracted with each solvent. A 1/8-inch by 20-foot
stainless steel column, packed with 70/80 mesh Chromosorb G, AW-DMCS, with 5%
OV-101 stationary phase, was programmed to hold at room temperature for four minutes
with the oven lid open, then the lid was closed and the program proceeded from 35° to
65° C at 8 degrees per minute and from 65° to 350° C at 6 degrees per minute. The
temperature was held at 350° C for up to 20 minutes, if it appeared necessary to elute all
of the sample. Injector and detector temperatures were about 370° C.
A solution of 3 gm of diesel fuel diluted to 50 ml with solvent was used as a standard
since it was available, although an oil with a higher aromatic content would have been
more desirable. The total area of the chromatogram of the standard, excluding the
solvent peak and correcting for baseline drift near the end of the chromatogram, was
compared with the total areas obtained from the 50 ml samples of extract. Data obtained
from the carbon extractions and GC analyses are shown in Table 13.
Chloroform, benzene, and methylene chloride showed about the same efficiency for
extracting oil from the carbon, while hexane and methanol were less efficient. Figures
10 through 14 show the chromatograms obtained on the extracts from API carbon using the
five different solvents. Comparison of the benzene, methylene chloride, and chloroform
extracts (Figures 10, 11, and 14) shows that the oils extracted with these solvents were
-------
382
almost identical in composition. However, chloroform showed an advantage in the
recovery of one low boiling compound which was an appreciable amount of the sample as
indicated from the size of the peak. This peak had a retention time of 11.3 minutes
which corresponds to the retention time of benzene. This peak could not be seen when
benzene was used as the solvent, and shows at a much lower concentration in the
methylene chloride extract.
Chloroform was used as the solvent for extracting a sample of carbon that had been
used to treat effluent water from the final clarifier of the activated sludge treatment
system. This carbon is referred to as FC carbon, and the resultant chromatogram is shown
in Figure 15. The FC carbon contained about 1/3 as much oil as the API carbon. A
sample of new carbon also was extracted with chloroform to determine whether the new
carbon contained any extractable material. This extract contained no measurable oil, as
illustrated by Figure 16. This figure also serves as a good illustration of the purity of the
chloroform used as the solvent. Figure 17 is a chromatogram of the standard mixture used
to quantitate area data in terms of grams of oil extracted.
The API carbon extract and FC carbon extract were each separated into three
fractions by selective elution and desorption from silica gel. The columns were pre-wet
with hexane and as the last drop of hexane disappeared into the surface of the gel, the
sample-gel mixtures in the 10 ml beakers were emptied into their respective columns.
The samples were then fractionated by eluting the saturates with 10 ml of hexane, the
aromatics with 10 ml of benzene, and finally desorbing the polar fraction with methanol.
Each of the fractions was then reduced in volume to 0.5 ml by evaporating the solvents
with a gentle stream of air.
Gas chromatography was used to obtain an estimate of the amount of saturates,
aromatics, and polar material in each of the samples. The oil extracted from the FC
carbon contained such small amounts of saturates and aromatics that it was necessary to
increase the sensitivity of the GC by a factor of 10 to obtain measurable areas. These
areas were then divided by 10 to get back to the same basis as the area measurements on
the polar fraction. The results are tabulated in Table 14.
These data indicate that the activated sludge treatment system has reduced the
saturates and aromatics to very low levels in the final clarifier effluent but has been
relatively ineffective in removing some of the polar type organic compounds.
Figures 18, 20, and 22 are chromatograms of the saturate, aromatic, and polar
fractions obtained by the silica gel separation of the oil from API carbon. Figures 19, 21,
and 23 are chromatograms of the saturate, aromatic, and polar fractions obtained by the
silica gel separation of the oil from FC carbon.
There was insufficient time for complete GC-MS identifications, but Table 15 lists
some of the compounds and types that have been found to date in the saturate, aromatic,
and polar fractions. The organic structures shown in the table are those normally found in
petroleum and petroleum products, but it is noted that three sulfur compounds were found
in the aromatic fraction of the FC carbon extract. It may be that these compounds are not
-------
383
biologically degraded as much as the aromatics and are a more significant part of the
aromatic fraction after biological treatment.
The polar fractions could not be analyzed by GC-MS because of the inability of the
gas chromatograph to separate the complex mixture. The complexity of the mixture is
illustrated in Figures 22 and 23 by the lack of baseline separation. Differences in the
two chromatograms indicate that some of the components have been degraded in the
activated sludge treatment system. It is probable that the reduction in phenolic com-
pounds, that js known to occur, would account for some of the differences noted in the
ch rom a tog ram s.
Infrared scans on the polar fractions were almost identical. Functional groups noted
were OH and C=O, and CH2 and CH3 adsorption bands indicate aliphatic type structures.
Further work must be done on the polar fractions to identify a number of the individual
compounds in the mixture so that an assessment of their toxicity can be made and treat-
ment methods developed to remove these compounds.
REFERENCES
1. Hansler, John W. Activated Carbon. Chemical Publishing Company, Inc.,
New York, New York, 1963.
2. Burks, S. L., T. C. Dorris, and G. L. Walker. Identification of Toxic Organic
Chemical Compounds in Oil Refinery Wastewaters. Technical Completion Report to
U.S.D.I. Office of Water Resources Research, Project BO-17, Oklahoma, 1970.
Unpublished.
3. Hale, J. H. and L. H. Myers. The Organics Removed by Carbon Treatment of
Refinery Wastewater. Presented at the Oklahoma Industrial Waste and Advanced
Water Conference, Stillwater, Oklahoma, April 1973. Robert S. Kerr Environmental
Research Laboratory, Ada, Oklahoma. Unpublished.
BIOGRAPHIES
Jack H. Hale holds a BS in Chemical Engineering from
Oklahoma State University. He is currently a research
chemist in the Industrial Section of the Source Management
Branch at the Robert S. Kerr Environmental Research
Laboratory, Ada, Oklahoma.
Leon H. Myers holds a BS in chemistry/biology from
Southwestern Oklahoma State University and a MS in
sanitary science from Oklahoma University. He is
currently Chief, Industrial Section of the Source Manage-
ment Branch at the Robert S. Kerr Environmental Research
Laboratory, Ada, Oklahoma.
-------
384
Thomas E. Short holds a BS in Chemical Engineering from
Lamar University and a MS and Ph.D. in Chemical
Engineering from Oklahoma State University. He is currently
a Chemical Engineer at the Robert S. Kerr Environmental
Research Laboratory at Ada, Oklahoma.
DISCUSSION
Bill McCarthy: When you used solvent extraction for recovering your adsorbent components,
could you give us an idea of what efficiency or recovery you got?
Leon H. Myers: No I can't, because we did not attempt a mass balance; however, the quan-
titative value of three solvents produced similar results. The quantity recovered agrees with
the projected isotherm loading.
Bill McCarthy: Do you think solvent regeneration might be a viable process?
Leon H. Myers: Yes, I do believe it is both feasible and viable, and this is determinant on
the mixtures of hydrocarbons to be removed. In some cases, particularly with light hydro-
carbons, steam regeneration might be the most viable regeneration process. Neither of these
alternate regeneration modes have been thoroughly proven.
MYERS HALE
-------
385
Table 1. CONTROLLED CONDITIONS FOR ISOTHERM STUDY
1 . 325 mesh (pulverized) 5. 100 rpm agitation
2. 200 ml sample 6. 23° C temperature
3. 1 hour contact 7. 75 mg/l TOC
4. 0.05, 0.20, 0.50, 1.0 gm 8. pH 7.4
carbon
Table 2. ADSORPTION DATA FOR ELEVEN ACTIVATED CARBONS
Carbon Sample mg/l TOC Adsorbed on 1 gm Carbon
Pulverized Granular
1
2
3
4
5
6
7
8
9
10
11
50
60
56
57
56
52
47
51
49
35
46
38
37
37
48
29
31
41
47
38
13
34
TableS. CONTACT TIME ADSORPTION
Time in Minutes
Gm Carbon
(Blank)
0.1
0.5
1.0
2.0
5.0
20
(206 mg/l)*
124
63
57
55
56
40
(260 mg/l)*
96
45
47
52
54
60
(260 mg/l)*
82
51
106
77
50
180
(260 mg/l)*
101
73
45
34
24
* Results are expressed as mg/l filtrable total organic carbon.
-------
386
Table 4. ANTHRAFILT FILTRATION ORGANIC CARBON REMOVAL
TOC
Turbidity
Charge
After 5 minutes (0.3 gal.)
After 1 hour (3.2 gal.)
After 2.5 hours (7.9 gal.)
255 mg/l
220 mg/l
210 mg/l
206 mg/l
11.0
2.75
Table 5. MINI-COLUMN TOC RESULTS
Date
3/17
3/18
3/19
3/20
3/21
Final Clarifier
Effluent*
35
33
48
37
35
Anthrafilt
Effluent*
31
34
40
48
38
Columns
A-l*
13
11
13
17
18
B-l*
9
11
16
23
19
A-2*
9
10
14
22
17
B-2*
10
12
14
16
25
A -3*
7
10
8
11
16
B-3*
12
18
12
22
32
All concentrations are mg/l TOC.
Table 6. COMPARISON OF MINI-COLUMN TOC REMOVAL EFFICIENCY
Date
3/17
3/18
3/19
3/20
3/21
A-l
58
68
68
65
53
B-l
71
68
60
52
50
Percentage Removals
A-2
71
71
65
54
55
68
65
65
67
34
77
71
80
77
58
B-3
62
47
70
54
16
Average
62
60
63
60
73
50
-------
387
Table 7. TOC RESULTS OF MINI-COLUMN UPFLOW MODE
Date
3/24
3/25
3/26
3/27
Final Clarifier Anthrafilt
Effluent*
50
35
41
38
Effluent*
38
33
33
35
1*
19
17
20
17
Columns
2*
12
13
24
19
3*
11
10
18
17
4*
13
10
13
13
5*
13
9
12
10
All concentrations are mg/l TOC.
Table 8. PERCENTAGE REMOVAL OF TOC FOR UPFLOW MOJDE
Date
3/24
3/25
3/26
3/27
]
50
48
39
51
2
68
61
27
46
Columns
3
71
70
45
51
4
66
70
61
62
5
66
73
64
71
Average
47
51
59
65
69
Table 9. REFINERY PROCESS DATA
Capacity
API Class
Wastewater Volume
Refinery Processes:
30,000 BPSD
B
18 gal. per minute
1 . Crude Distillation
2. Crude Desalting
3. Vacuum Crude Distillation
4. Fluid Cat Cracking
5. HFAIkylation
6. Hydro Cracking
7. Catalytic Reforming
8. Asphalt Production
-------
388
Table 10. REFINERY WASTEWATER TREATMENT RESULTS
Median Values
Parameter
BOD5
COD
TOC
Oil and Grease
Phenols
Chromium
Copper
Iron
Lead
Zinc
Sulfide
Ammonia
Cyanides
Turbidity*
Color**
API Separator
(mg/l)
97
234
56
29
3.4
2.2
0.5
2.2
0.2
0.7
33
28
0.25
26
30
Bio-Treated
(mg/l)
7
98
30
10
0.01
0.9
0.1
3.0
0.2
0.4
0.2
27
0.2
17
15
Carbon-Treated
(mg/l)
48
103
14
10
0.004
0.2
0.03
0.3
0.2
0.08
39
28
0.2
11
15
Bio-Carbon
Treated
(mg/l)
3
26
7
7
0.001
0.2
0.05
0.9
0.2
0.15
0.2
27
0.2
5
1
* Turbidity given in Jackson Turbidity Units.
** Color given in color units.
Table 11. CARBON ADSORPTION CAPACITY
Capacity Lbs TOC Adsorbed Per Lbs Carbon
Isotherm Study
API Separator Effluent
Carbon Column
0.12
0.31
Table 12. CARBON REGENERATION ACTIVITY ANALYSIS
Iodine No.
Molasses No.
Virgin Carbon
API Separator Carbon
Final Clarifier Carbon
1010
906
991
216
405
304
-------
389
Table 13. SOLVENT EXTRACTION EFFICIENCY DATA
Carbon
Identification
API
API
API
API
API
FC
NEW
Solvent
Chloroform
Benzene
Methylene Chloride
Methanol
Hexane
Chloroform
Chloroform
Grams
Carbon
49.3008
50.6917
54.5660
56.3179
55.3762
56.6447
44.6169
Grams Oil
Extracted
5.04
5.01
5.44
3.71
4.03
1.95
0.00
Gram Oil/
Gram Carbon
0.102
0.099
0.100
0.066
0.073
0.034
0.000
Table 14. COMPOSITION BY TYPES OF ORGANICS
Organic Type
Saturates
Aroma tics
Polar Material
Oil from
API Carbon
11.1%
24.7
64.2
TOO
Oil from
FC Carbon
0.2%
1.3
98.5
100.0
Table 15. COMPOUND TYPES INDICATED BY GC-MS
API Carbon Extract
FC Carbon Extract
Saturate Fraction
Aromatic Fraction
Polar
C9H18
C10H20
C12H24
Cg through Cy) n-paraffins
Ethylbenzene
Xylenes
Co Benzenes
C ]0 Benzenes
Naphthalene
Cii Benzenes
C-\-\ Naphthalene
C^ Naphthalene
C13 Naphthalene
Phenol
Hydroxy toluene
Ethylbenzene
Xylenes
€9 Benzenes
C10 Benzenes
C7H8 S
C7H10S
C9H8 S
-------
6O
PULVERIZED
GRANULAR
II
o
o-
FIGURE - I
CARBON SOURCES
-------
ON
500
III
CO
CE
Q 0
< CD
DC
0 <
Ł ° 100
U—
u. o
o
Ł < 50
< 0
oc
° LJ
_l 0.
i
i
10
K.
-
200ml /min.
/+'
P
/
/
1
Q
/ WASTE - API SEPARATOR, EFFLUENT
6 CARBON-FILTRASORB 400
I
1
o
i
l
1
4.4g/m/ft.2
r
k
.PUMP
=A
API SEPARATOR
-h
^
^
//
%•
a'4,
0 «
•"*•'
*V
".'*
1'*-
.sl5"ANTHRAFILT
^30" SAND (20 -30 Mesh)
Br
;>i
/«
•^
i
i^
WATER SAMPLE ^
HLl2" LIMESTONE (0.5"Dia.)
^
^
restful K"r:i Aocvi/nrn
10 50 100 500
TOC- MILLIGRAMS PER LITER
TO RESERVOIR
FIGURE 2 - ADSORPTION ISOTHERM
FIG. 3-MULTI-MEDIA FILTER
-------
392
COLUMN
234
t
ELECTRONIC
SAMPLER
FINAL
CLARIFIER
FINAL
CLARIFIER
SAMPLE
DDDDD
12345
COLUMN
EFFLUENT
IVIULTI- MEDIA
FILTER
SAMPLE
FIG.4-MINI COLUMN SYSTEM
SAMPLE
POINT-!
SAMPLE
POINT-2
COOLING TOWER
SLOWDOWN
BOII FR Bl OWDOWN „
OILY PROCESS WATER
STRIPPER EFFLUENT
CONTAMINATED RUNOFF
TO HOLDING
POND
API SEPARATOR
PASVEER DITCH
CLARIFIER
FIGURE 5 - REFINERY WASTE WATER TREATMENT SYSTEM
-------
FEED
SAMPLE
393
Pea
^Gravel
DUAL-MEDIA FEED
FILTER PUMP
"CALGON" ACTIVATED
CARBON PILOT PLANT
FIGURE 6 - ACTIVATED CARBON PILOT PLANT
FLOW DIAGRAM
a:
UJ
a:
UJ
a.
a:
o
o
o
200-
160-
120 -
API SEPARATOR
EFFLUENT
40 -
CARBON TREATED
EFFLUENT
BIO-TREATED EFFLUENT
8 10
DAYS INTO STUDY
14
BIO-CARBON EFFLUENT
16
FIGURE 7 TREATMENT RESULTS, BOD
-------
394
280
240
ct
UJ
3 200
IT
UJ
Q.
V)
5
I
Q
O
160
120
80
40
API SEPARATOR
EFFLUENT
BIO-TREATED
EFFLUENT
CARBON TREATED
EFFLUENT
BIO-CARBON
TREATED
EFFLUENT
I
6 8 10
DAYS INTO STUDY
12
14
16
FIGURE 8- TREATMENT RESULTS, COD
70
60
ce
UJ
t 50
or
Ul
Q.
c/> 40
cc
H 30
i
o
2 20
10-
0
I
I
API SEPARATOR
EFFLUENT
BIO-TREATED
EFFLUENT
CARBON TREATED'
EFFLUENT
BIO-CARBON
EFFLUENT
6 8 10 12 14
DAYS INTO STUDY
FIGURE 9-TREATMENT RESULTS, TOC
-------
395
TEMPERATURE, °F
I35°F 260 380 490 610 720 825 895 I35°F 260 380 490 610 720 825 895
111
UJ
to
z
o
Q.
V)
Ul
cc
FIGURE 10- BENZENE EXTRACT OF
API CARBON
8
I I T
FIGURE II - METHYLENE CHLORIDE EXTRACT
OF API CARBON
\JV
16 24 32 40 48 56 64 0 8 16 24 32 40 48 56 64
TIME, min.
FIGURE 12- METHANOL EXTRACT OF FIGURE 13 - HEXANE EXTRACT OF
API CARBON API CARBON
TEMPEATURE, °F
I35°F 260 380 490 610 720 825 895 I35°F 260 380 490 610 720 825 895
~g fe 24 32 40^48 56 64 08 16 24 32 40 48 56 64
TIME, min.
FIGURE 14- CHLOROFORM EXTRACT OF FIGURE 15 - CHLOROFORM EXTRACT OF
API CARBON FC CARBON
-------
396
Temperature , °F
tn
z
o
a.
to
UJ
tr
I35°F200 320 440 550 665 780 865
1
1 1
1 1 1 1 1 1 1 I 1 I 1 1 1 1
1 1 1 1 I t 1 1 I 1 1 1 I I
) 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64
TIME, min.
FIGURE 16 - CHLOROFORM EXTRACT OF NEW ACTIVATED CARBON
I35°F 200
320
Temperature °F
440 550
665
780
865
to
z
o
a.
CO
UJ
ae
8 12 16 20 24 28 32 36 40 44 48 52 56 60 64
TIME, min.
FIGURE 17 - STANDARD OIL MIXTURE
-------
397
Ul
CO
z
o
a.
V)
UJ
tr
TEMPERATURE, °F
I35°F26O 380 490 610 720 825 895 I35°F 260 380 490 610 720 825 895
SENSITIVITY x 10
TIME, min
FIGURE 18 -APIC EXTRACT SATURATE FIGURE 19-FCC EXTRACT SATURATE
FRACTION FRACTION
TEMPERATURE, °F
SENSITIVITY x 10
I I I I I I I I I
8
16 24 32 40 48 56 64
16 24 32 40 48 56 64 0 8
TIME, min.
FIGURE 20-APIC EXTRACT AROMATIC FIGURE 21 -FCC EXTRACT AROMATIC
FRACTION FRACTION
V)
z
o
0.
CO
TEMPERATURE, °F
I35°F 260 380 490 610 720 825 895 I35°F 260 380 490 610 720 825 895
J L
_L
1 1—L
i I I I
0 8 16 24 32 40 48 56 64 08 16 24 32 40 48 56 64
TIME, min.
FIGURE 22- APIC EXTRACT POLAR FIGURE 23- FCC EXTRACT POLAR
FRACTION FRACTION
-------
398
-------
ACTIVATED CARBON TREATMENT OF COMBINED STORM AND PROCESS WATERS
M. A. Prosche
Atlantic Richfield Company, Watson Refinery, Carson, California
The Watson Refinery of the Atlantic Richfield Oil Company is located adjacent to the
Dominguez Channel in Los Angeles County. This Channel is a non-navigable dredged
tidal estuary which is lined with rip-rap and used specifically for the discharge of refinery
and chemical plant waste waters and for rain water runoff.
In 1968, the Los Angeles Regional Water Quality Control Board made a study of the
Dominguez Channel and determined that petroleum and chemical plant discharges were
causing a problem due to the oxygen demand of their waste waters entering the Channel.
The Control Board, in accordance with these findings, issued a resolution in February,
1968, which limited the total chemical oxygen demand (COD) from all industrial discharges
into the Channel. These discharges also included any contaminated rain water runoff.
The resolution was to be complied with by February, 1971 .
As defined by the resolution, the Watson Refinery was limited to 1330 pounds per day
of COD in its discharge water to the Channel. Meeting this requirement meant reducing
the COD in its discharge waters by 95 per cent.
Fortunately, the Watson Refinery as a taxpayer of Los Angeles County was able to make
arrangements with the Los Angeles County Sewer District to have its process waste water
handled in the County's primary treatment unit. However, due to limitations in the County
unit, the County was unable to handle rain water runoff. This presented a problem for
the Watson Refinery due to the fact that the rain water collection facilities were inter-
connected with the process waste water collection system. Therefore, during periods of
rainfall, the process waste water and rain water mixture could not be sent to the sewer
district facilities, nor could it be sent to the Dominguez Channel due to high COD content
of the process waste water.
To solve this problem a system was needed which would treat all the process water
plus rain water during the rainy season for the removal of COD and allow its discharge to
the Channel. A system was needed which could be started up easily when rain fell and
then shut down when no longer required. The system selected was impounding of rain plus
process water during the storm followed by activated carbon treatment to adsorb the COD
material.
ADSORPTION SECTION DESIGN AND DESCRIPTION
Since it was planned to operate the plant only during the rainy season, it was felt
there was no need to design for continuous carbon regeneration and changeout of the beds.
Therefore, the design was based on use of only the carbon beds during the rainy season,
and then regeneration of all the beds during the dry summer months in preparation for the
next rainy season. The significant design criteria for the plant and impounding basin were
as follows:
399
-------
400
Design Criteria
Flow Rate 100,000 BPD
3,000 GPM
Inlet COD 250 ppm
Outlet COD 37 ppm
Impounding Basin 1 .2 million barrels capacity
Following the accumulation of 0.1 inches of rainfall, regulations call for diverting
process water from the sewer system. This is accomplished automatically by means of a
solenoid operated valve actuated by a rain gauge. Process water plus commingled rain
water is then switched to the 1 .2 million barrel impounding basin. Depending upon the
specific situation, the carbon plant can be started up immediately for discharge to the
Channel or it can be started up later following the rain. A flow diagram of the adsorption
section of the plant is shown in Figure 1 .
Impounded water is delivered to the plant through a 14" line from the basin. The water
is then delivered to a distribution trough where it is distributed to 12 adsorber cells by
adjustable slide gates. Each of the twelve cells is 12' by 12' square and 26" feet deep.
Each cell originally contained 13' of carbon having a dry weight of about 50,000 pounds.
The water passes down through the carbon bed where it collects in the underdrain system.
Supporting the carbon is a one-foot layer of gravel on top of a Leopold tile underdrain
system. The treated water then flows through 6" lines from each cell to a 24" collection
header leading to the effluent retention sump. Each 6" discharge line has a sample point
and an air-operated pinch valve which can be shut during backwashing.
If necessary to further control COD, a chlorine-water solution may be injected into the
incoming treated water stream. Approximately 15 minutes retention time is allowed for
chlorine contact in this sump. From the sump the water flows by gravity to the Channel.
Each carbon bed must be backwashed whenever it will not pass its share of water flow
due to buildup of solids on top of the carbon. This is indicated by the rise in the level of
the water in each carbon cell. When the level rises to the height of the backwash troughs,
the flow to and from the bed is stopped, and treated water from the backwash sump is pumped
up through the bed to expand it and flush out accumulated solids. The turbid water overflows
into the backwash troughs to the backwash effluent sump where it is pumped back to the
reservoir for settling and retreating.
EPA DEMONSTRATION GRANT
Following completion of the plant, a demonstration grant was received sponsored by the
Water Quality Research Division of Applied Science and Technology of the Environmental
Protection Agency. The specific objectives of the demonstration project were as follows:
1 . Determine feasibility of activated carbon as a treatment system for storm runoff and
refinery process waters.
2. Evaluate performance of the system.
3. Determine operating costs.
4. Assess reliability of the system.
-------
Length of
Run, Hours
44
48
38
18
95
22
" Table 1
Feed
Rate, GPM
3000
3000-2000
2000
2000
1000
2000
Average COD
Feed
326
360
374
310
237
147
Effluent
43
48
86
67
100
93
401
The report relative to this project has now been published by the EPA.
ADSORPTION SECTION OPERATION AND PERFORMANCE
The carbon plant was first placed in operation in May, 1971 for some preliminary test
work prior to the rainy season. At that time the test water was synthesized using process
water diluted with service water. Operation with rain water and process water was not
required until later that year in December. At that time sufficient rain fell to require
placing the unit in full operation at the design rate of 3,000 gpm total to all twelve cells.
Subsequent to that a total of 6 runs were made processing impounded water from the inter-
mittent rain storms that followed. The performance data for all of these runs are shown in
Table 1.
Typical Performance Data First Rainy Season
Run
Number
~~1
2
3
4
5
6
During the initial runs the COD of the feed was higher than design and the plant was
unable to produce the desired removal. Therefore, the feed rate was reduced to 2,000
gpm. Even at this lower rate the effluent COD content was greater than the regulations
allowed and feed rate was again reduced to 1,000 gpm. By the end of the season the
reservoir COD had dropped to 147 ppm due to rain dilution and the rate was increased to
2000 GPM. However, performance continued poor even at this low concentration.
A good evaluation of the performance of the plant during this first season of operation
is difficult due to the significant variations in feed COD and the necessity of having to
reduce the feed rate from 3000 to 1000 gpm. However, COD data for each cell was
recorded during the course of the run and some generalizations are possible. At the end
of the season the COD loading in each cell varied from 0.2 pounds of COD per pound of
carbon to a high of 0.3 pounds of COD per pound of carbon. During the season a total
of 52,000,000 gallons of water were processed. The average feed COD concentration
was 377 ppm while the effluent averaged 67 ppm. The carbon loading averaged 0.23
pounds COD per pound of carbon. It also appeared that the potential maximum carbon
loading at a constant effluent COD concentration was quite sensitive to the COD of the
feed.
Other data collected during the first season's operation included measurement of the
carbon's adsorptive ability at various depths in the bed after a period of operation. One of
the cells was taken out of service after 208 hours of operation and the carbon was sampled
at various depths to test its relative adsorptive efficiency as compared to virgin carbon.
The data are shown in Figure 2. The data show that the top two feet of carbon were
completely exhausted while the last five feet were still 65 per cent of virgin adsorptive
-------
402
ability.
Prior to the second rainy season, we made several significant changes in operation. In
order to increase the adsorptive capacity of the unit, 15,000 pounds of carbon were added
to the existing 50,000 pounds in each of the cells. This increased the bed depth from 13 feet
to 17 feet. It was also decided to recycle effluent water back to the feed in order to
control the feed COD to a concentration of not greater than about 250 ppm. This change in
operation was made after noting during the first rainy season of operation that the effluent
COD level varied directly with the feed COD concentration and the specification COD
quantity to the channel would be exceeded during periods of high feed COD. It was also
felt that controlled feed COD would allow a higher ultimate COD loading of the carbon.
One other modification made was the aeration of the effluent recycled back to the feed to
prevent growth of anaerobic bacteria in the beds.
For the second rainy season of operation, we elected to operate part of the plant on a
continuous basis, regenerating each bed as it became spent, and returning the regenerated
carbon for use again. To accomplish this five of the cells were placed in continuous
staggered operation along with the regeneration furnace. This was the largest number of
cells operating in a progression mode that could be accommodated by the capacity of the
regeneration furnace. The remaining seven cells were used for once-through processing as
was done during the first rainy season. In all cases the flow to each cell was held constant
at 250 gpm which included the waste water feed from the reservoir plus the recycle dilution.
Typical performance data for one of the five cells operated in staggered mode are shown
in Figure 3. This figure shows the relationship between feed and effluent COD and the COD
loading on the carbon. During initial operation of this cell, it was elected to discontinue
service and regenerate after the effluent from the cell reached a COD concentration of
about 50 ppm. This corresponded to a carbon loading of 0.20 pounds of COD per pound of
carbon and a run length of 22 days.
The carbon was regenerated and returned to the same cell. In the second run the cell
was allowed to operate until the effluent reached a COD concentration of about 100 ppm.
This corresponded to a carbon loading of 0.40 pounds COD per pound of carbon and a run
length of 38 days.
The performance of the seven cells that were operated on a once through basis with no
regeneration is shown in Figure 4. Effluent COD concentration was allowed to reach 50
ppm before all the cells were shut down for later regeneration. As shown in the figure, the
corresponding carbon loading attained in all of these cells was 0.30 pounds COD per pound
of carbon and the run length was approximately 30 days.
The evaluation of performance of the carbon plant for this second rainy season was again
based on the combined operation of all the cells. During the season a total of 102,000,000
gallons of water were processed. The average diluted feed COD concentration was 233 ppm
while the effluent averaged 48 ppm. The carbon loading averaged 0.26 pounds COD per
pound of carbon.
-------
403
A summary of the performance of the plant during the two rainy seasons of operation is
shown in Table 2.
Adsorption Data
Table 2
Commercial Feed COD, PPM Effluent COD, PPM Carbon Loading Lb. COD/
Operation Average Average Lb. Carbon Average
First Rains 377 67 OS
Second Rains 233 48 0.26
As noted previously, problems with varying feed COD and algae growth were experienced
during some of the runs. These problems undoubtedly limited the performance to poorer
levels than could have been attained under ideal steady state conditions. However, these
are the real problems that exist when treating refinery waste waters and we therefore feel
the carbon loadings attained are indicative of the true performance of the commercial plant.
CARBON HANDLING AND REGENERATION
This system is shown in Figure 5. The spent carbon is removed as a water slurry through
a carbon removal trough and nozzle even with the top of the gravel layer. The slurry flows
by gravity and with the aid of water jets in the carbon removal trough to the spent carbon
transfer pump sump. From here a rubber lined pump transports the carbon slurry into the
spent carbon tank. The tank has capacity to contain the contents of one cell. From this
tank the carbon is educted with high pressure water to the dewatering screw above the
regeneration furnace. A timer valve on the spent carbon line feeding the eductor opens
and closes at timed intervals to allow close control of the carbon flow to the regeneration
furnace. The dewatering screw separates most of the water from the spent carbon slurry
and the drained water is returned to the reservoir. The carbon discharging into the furnace
contains approximately 50 per cent by weight of water.
The regeneration furnace is a 56" I.D. multiple hearth type with a total of six hearths.
It is gas fired and internal temperatures are controlled to about 1600 F. A center shaft with
rabble arms moves the carbon across each hearth and downward through the furnace. An
elaborate flue gas quench and scrubber system is installed to meet local Air Pollution
Control District requirements. The furnace is designed to regenerate 8,000 pounds of carbon
per day or one bed in about 6 to 7 days.
The regenerated carbon drops into a quench tank just below the furnace from where it is
educted into the regenerated carbon tank. From this tank the carbon is then educted and
transported through a hose back to the same cell where it originated.
The carbon handling system has worked very well. The original 1" cast iron eductors
were too small and soon lost capacity due to erosion. Larger stainless steel eductors and
a reduction of motive water pressure from 110 psig to 50 psig have remedied this. Also
originally we had some problem with displaced support gravel from the bottom of the cells
plugging up the eductors. This was remedied by constructing a gravel screen in the top
of the spent carbon tank to remove the gravel. The time required to transfer the carbon
out of the cell to the spent carbon tank is about 14 hours while transfer back to the cell
from the regenerated carbon tank takes about 7 hours.
-------
404
Operation of the regeneration furnace has been excellent. The major effort relative to
this operation has been to develop a reliable method for determination of the quality of the
regenerated carbon. After operational guidelines were developed, the furnace consistently
regenerated the spent carbon to near virgin activity.
Our experience indicates that about a 5 per cent loss is incurred during carbon handling
and the regeneration. We feel the majority of this is due to attrition of the carbon and
resulting production of fines. These fines are backwashed out of the cells prior to being
placed in operation and end up back in the reservoir where they settle out.
OPERATING COSTS
Actual operating costs for the second year of operation are shown in Table 3.
Cost Data
December, 1972 - March, 1973
Table 3
Cents Per Cents Per
Thousand Gallons Pound of COD
Cost Areas of Water Treated Removed
Utilities 9 4
Repair Labor 3 1
Operating Labor 15 7
Carbon 1 1 5
Miscellaneous 2 1
Total 40~ f8"
Updating these costs to 1976 levels would give about 56 cents per thousand gallons of water
treated, or about 25 cents per pound of COD removed.
CONCLUSIONS
The plant has proved to be easy to maintain, easy and quick to start up, simple to shut
down and leave in a standby stage, and very reliable in its performance.
For the unique intermittent type operation required at the Watson Refinery, we feel the
Carbon Adsorption Process has been successful. However, we do not feel that this conclusion
would necessarily be the same for a continuous operation requirement or some other unique
situation.
DISCUSSION
Gantz: Did you experience any solids handling problems?
M. A. Prosche: We had no solid handling problems. We had some concern about the silt,
sand, and carbon fines building up in the bottom of the reservoir. But during the summer months
we just moved in a front-end loader and removed all solids.
-------
405
Ben Buchanan: I was wondering what control methods you used on your reactivation furnace
to control the reactivation.
M_. A. Prosche: What we finally did was measure the ABD, apparent bulk density, of the
carbon being fed to the furnace and the regenerated carbon from the furnace. We found
that a reduction of gm/cc in ABD was sufficient to bring the carbon back up to 100%
adsorption efficiency. In actual practice the operators had a pre-weighed cube and a
balance at the unit. The cube was then used to weigh a fixed volume of feed and product
carbon and the difference used as the guide to furnace operation. As a generality, spent
carbon had an ABD of 0.56 gm/cc and the regenerated 0.50 gm/cc.
Ben Buchanan: How did you determine how much of your carbon was lost by attrition, and
how much was lost through burning in your regeneration furnace?
M. A_. Prosche: The only insight we had was to roughly know how much carbon we put
back into the cells after regeneration, and the amount of carbon that had accumulated in
the bottom of the reservoir after a season of operation. It appeared to us that the accumulated
fines in the reservoir about equaled the loss in volume in the beds.
Ben Buchanan: There is a little problem with measuring volumes because I think the losses
that you have from burning the carbon do not necessarily reduce the size but many of those
losses are inside the particle and do not show up. You will get a weight loss but you do
not necessarily get the same volume loss.
M. A. Prosche: That could very well be. This was not a research project but strictly a
demonstration project to determine feasibility and costs.
Larry Echelberger: Has there been any attempt to reuse this water in your refinery after you
clean it up? fsTt high in TDS? Do you attempt to reuse the water at all?
M. A. Prosche: The discharge is primarily refinery waste water diluted with rain water.
The waste water consists of untreated service water, desalter waters, and water from tank
bottoms. So this water would require extensive pretreatment prior to any reuse. Our primary
thrust now is to reuse stripped sour water which is essentially contaminated consensate.
Following the successful utilization of all of this water, we can then turn our attention to
the more difficult problems of reusing the waste water.
-------
406
BIOGRAPHY
Marvin A. Prosche is Manager of Refinery
Technology at Atlantic Richfield's Watson
Refinery. He holds a B.S. in ChE from the
University of Notre Dame and an M.S. in
ChE from the University of Illinois. His
experience includes 25 years of petroleum
refinery process design and process engineering.
-------
FIGURE I
407
WASTE WATER ACTIVATED CARBON TREATMENT PLANT
TO RESERVOIR
FROM RESERVOIR
U * U
CARBON
GRAVEL
LEOPOLD TILE
BACKWASH
TROUGH
:HLOR:NATOR
PUMpf~~l
BACKWASH
EFFLUENT
SUMP
PUMP[~~1
CARBON ADSORPTION CELL
12 CELLS
EACH 12' * 12' x 26' DEEP
Relative Efficiency Profile of Cell
After 208 Hours of Operation @ 250 GPM and
650 to 600 PPM COD in Feed
EFFLUENT BACKWASH
RETENTION SUMP
SUMP
TO DOMINGUEZ
CHANNEL
Figure 2
10
11
Depth Down into Carbon Bed, Feet
-------
Typical Performance Data—Staggered Operation
Second Rainy Season
Figure 3
Q_
Q_
c"
g
•*—'
CO
-•—I
c
CD
o
c
o
O
O
O
O
400
300
200
100
0
r*.
6
80246
Millions of Gallons of Water Treated
c
o
_Q
03
cn Q
1 3
c
o O
JD
is w
<5 3
oo
o
0
80246
Millions of Gallons of Water Treated
8
10
12
14
-------
Performance Data—Bulk Processing Cells
Second Rainy Season
Figure 4
CL
DL
cf
-»^*
CO
"c
o
o
O
Q
O
O
c
O
.a
aj
a) o
1 3
c
o O
JD
J= CO
400
300
200
100
0.4
0.3
0.2
0.1
0
3456789
Millions of Gallons of Water Treated
45678
Millions of Gallons of Water Treated
10
10
11 12
11
12
-------
FIGURE 5
CARBON TRANSFER AND REGENERATION SYSTEM
CARBON
BED
CARBON REMOVAL
TROUGH
SPENT
CARBON
TANK
REGENERATED
CARBON
TANK
1
WATER
EDUCTOR
DEWATERING
SCREW
REGENERATION
FURNACE
SCRUBBER
SEALING
POT
EDUCTOR
EDUCTOR
-Jiw
CD
-------
SESSION VI
"MISCELLANEOUS TOPICS"
Chairman
M.K. "Don" Mutton
Manager, Mechanical and Environmental Engineering
Kerr-McGee Refining Corporation
Oklahoma City, Oklahoma
Speakers
Lial F. Tischler
"Inherent Variability in Wastewater Treatment"
R.T. Milligan
"Reuse of Refinery Wastewater"
Sterling L. Burks
"Biological Moniforing of Petroleum Refinery Effluents"
Ronald G. Ganfz
"API - Sour Water Stripper Studies"
411
-------
412
BIOGRAPHY M. K. "Don" Hutton
Graduate of Oklahoma State University. B.S. in Mechanical
Engineering, 1956. Upon graduation employed as Engineer at Shell
Oil Company, Deer Park Refinery until March of 1964 at which
time he was transferred to Shell Oil Company's Odessa Refinery as
Plant Engineer until July 31, 1966. August 1, 1966 employed as
Engineering Assistant to the General Manager of Petroleum
Refining, Kerr-McGee Corporation, Oklahoma City, Oklahoma.
Present position is Manager, Mechanical & Environmental
Engineering, Kerr-McGee Refining Corporation, Oklahoma City,
Oklahoma.
-------
INHERENT VARIABILITY IN WASTEWATER TREATMENT
Lial F. Tischler
Engineering Science, Inc., Austin, Texas
INTRODUCTION
The term "inherent variability" as applied herein can be defined as that vari-
ability in effluent quality from a properly designed and operated waste treatment
system which is attributable to the basic nature of the treatment processes, the charac-
teristics of the wastewaters, and climatological conditions - none of which can be
significantly altered by externally applied changes. Simply stated, it is the minimum
variability which can be practically obtained assuming proper system design, manage-
ment, and operational control (including in-plant changes, equalization, and ef-
fluent holding capacity).
Variability in effluent quality has assumed more importance in recent years since
the effluent limitations guidelines promulgated by the U.S. Environmental Protection
Agency (EPA) for the petroleum refining industry and for other industrial point source
categories are based on selected long-term attainable effluent qualities and allowable
daily and monthly variability for each controlled constituent. The published EPA
effluent limitations are based on maximum allowable daily and monthly average dis-
charges; since traditionally treatment system design has been on more of an "average"
or steady-state basis, it becomes mandatory to consider the potential variability in a
treatment system's effluent quality if it is to comply with the limits in a cost-effective
manner.
To accomplish this it is necessary to consider the effect of each unit process in
a complete system on the overall effluent variability. Each unit process can either
attenuate or increase variability, depending on process characteristics, wastewater
quality, and external stimuli. This paper examines the inherent effluent variability
of the major unit processes which are found or will probably be found in petroleum
refinery treatment systems designed to comply with the 1977 (BPCTCA) or 1983
(BATEA) EPA effluent limitations. This is done by analyzing data collected from a
number of full-scale refinery treatment units. Bench-scale and pilot-scale data are
inadequate for an analysis of effluent variability since the phenomena which are
responsible for variability are not amenable to scaling down. After this evaluation
of effluent variability for various treatment systems, a basis for using the results of
this analysis to estimate potential effluent variability in petroleum refinery treatment
units is presented.
DEFINITION OF VARIABILITY PARAMETERS
Before proceeding it is desirable to define variability in a probabilistic sense in
order to be consistent in all analyses so that data can be compared directly. The
easiest way to define variability is in the form of a ratio of the value of a consti-
tuent at a specified maximum probability to the mean value of that constituent (or
413
-------
414
median for log-normal distributions). One way of doing this is to use a multiple
number of standard deviations from the mean. This is convenient for a normal
(Gaussian) population but is not as useful for log-normal distributions. In this investi-
gation, variability is defined as the ratio of the 99 percent probability of occurrence
value to the 50 percent value. Although the 50 percentile value for a log-normal
distribution is the median, the arithmetic mean was used to conform to the basis used
by EPA in the guidelines (Ref. 1). It is important to note when the variability is for
monthly average data as opposed to daily data. Monthly average and daily vari-
ability ratios cannot be compared since the averaging technique masks variability to a
considerable extent. The same fact holds true, of course, for grab samples versus
daily composites, however, in most cases this paper has considered the latter two
sample types as interchangeable.
The ideal approach in formulating variation allowances is to measure inherent
variability in exemplary treatment units which reflect a wide spectrum of raw waste
load, temperature, and dissolved solids. These data can be resolved statistically and
the long-term (one year or more) distribution can be determined. The shorter-term
limits can be calculated statistically from this distribution. The EPA has used this
concept in developing variability factors for 24-hour maximum and maximum 30-day
average variations based on annual observations. (It should be recognized that states
have the prerogative under the law (Ref. 2) to add grab sample provisions to the
NPDES permit, and the proper establishment of these limits requires a comprehensive
review of performance data and a detailed statistical analysis (Ref. 3, 4).) A
significant difference in the variability of effluent quality between flow-weighted
samples and grab samples during the same period was recently demonstrated and it
emphasized the potential inaccuracy of using an insufficient number of grab samples
as a basis for determination of compliance or as an input for operational decision
(Ref. 5).
Almost all of the data distributions analyzed in the investigation were found to
approximate log-normal distributions; most other distributions, being of a higher order,
could be approximated by assuming a log-normal distribution. The variability factors
were calculated in the following manner for each constituent of the refinery waste-
water data base used in this study:
(1) the effluent loading data were grouped, rank-ordered, and
plotted on log-normal probability paper;
(2) the line of best fit was determined;
(3) a daily variability factor was defined as the ratio of that effluent
loading encompassing 99 percent of the expected daily
variation to the long-term arithmetic average effluent loading;
and
(4) a monthly variability was defined as the ratio of that effluent loading
encompassing 99 percent of the expected monthly variation to the
longH'erm arithmetic average effluent loading.
The 99 percentile values were taken directly from the best fit curves. The
-------
415
99:50 variability ratio can be interpreted as a multiplier which, when multiplied by
the long-term average concentration or mass discharge rate, will give a value which
will only be exceeded once in every 100 samples collected. If it is assumed that the
effluent data follow a log-normal probability distribution, the 99:50 ratio can be used
to define the median value of the distribution, the geometric standard deviation, and
thus the concentration at any other desired probability of occurrence.
CAUSES OF EFFLUENT VARIABILITY
There are many factors which cause "inherent variability." The most signi-
ficant "cause and effect" aspects are discussed below.
Variations in Raw Waste Load (RWL)
Although the impact of raw waste load variation can be minimized by in-plant
changes and by installing equalization, surge, and off-specification storage basins
prior to treatment, transitory loads from most refinery processes cannot be completely
sequestered. Changes in feedstock, such as crude oils of varying specific gravities,
sulfur, nitrogen, and trace metal content, result in raw waste loads which vary both
in quantity and quality, with subsequent effects on treatability. Changes in product
mix and product specifications have a similar effect. The reduction of wastewater
flow from production units through water reuse and good housekeeping practices
normally results in higher influent concentrations of organic constituents to the treat-
ment plant. This, according to most biochemical kinetic models, will correspondingly
result in an effluent of higher concentration. Thus, cyclic changes in raw waste
loads not fully sequestered will result in effluent quality variation in each type of
treatment used. Other events, such as maintenance turnarounds, production of
specialty items, changes in calendar production patterns, and contrasts in dry weather
and wet weather operations simply increase the variability range. It should be
emphasized that the effluent variation due to RWL as discussed here includes only
those changes necessary for manufacturing the product and excludes effects due to
accidental dumps or spills and poor housekeeping.
The inclusion of storm water in waste treament facilities can also increase vari-
ability, depending on the handling procedures used. If separate storm sewers, im-
poundment facilities, and controlled discharge to treatment are used, the effect on
variability will be low. Conversely, if a combined storm/process sewer system is
used and impoundment facilities are not available, variability in treatment performance
will be high.
Table 1 presents variability ratios for the raw waste loads at four "case history"
refineries. It can be seen that the ratios between the mean and peak values are not
too extreme for these refineries. This illustrates an important concept in comparing
variability ratios for various unit processes and treatment systems: variability expres-
sed as a ratio to the mean value for a given parameter is often a function of the
magnitude of the mean value. For example, when the mean is a relatively large
number as in the raw waste load figures, the variability expressed as a ratio will
tend to be low. Conversely, at extremely low effluent concentration values for some
-------
416
constituents, the chance of getting one sample out of 100 which is three or four times
the mean value is much greater. This concept will be illustrated graphically later in
this paper. Note that the raw waste loads given in Table 1 are not truly untreated
wastes; these samples are taken after primary oil/solids separation. These ratios can
be considered to be representative of the variability attainable for each parameter
shown if storm water is carefully controlled and the separators are adequately designed.
Variations in Temperature
The effect of temperature on biodegradation rates has been extensively investi-
gated and is generally well documented (Ref. 6, 7), This component of inherent
variability is therefore accepted, although the significance of the temperature effect
on variability is open to debate. Although this influence is primarily evident on a
long-term or seasonal basis, short-term variations are also possible in some geographi-
cal areas. The degree of variation is not only attributable to water temperature, but
also to the nature and complexity of the wastewater. When most of the organic con-
stituents are in the colloidal or suspended form, for example, the performance of an
activated sludge system is less subject to temperature fluctuations than when treating
soluble organic materials. This is attributed to the fact that biochemical reactions
removing soluble organics are more temperature-sensitive than are physical or "bio-
sorption" mechanisms which remove colloidal or suspended organic matter. Observed
temperature effects on biological systems treating complex soluble wastewaters (such as
petrochemical discharges) are more pronounced than when more readily degradable
soluble wastewaters are treated (such as food processing effluents).
Physical and chemical treatment processes can also be affected by temperature
particularly rapid temperature changes. Density qurrents can form in units which rely
on gravitational process for solids/liquid or liquid/liquid separation. These density
currents can cause short circuting and loss of treatment efficiency. An example of
this would be the sudden discharge of cold storm water into a gravity separator filled
with warmer water. Thus, careful attention must be given to the design of treatment
units to minimize these effects. Most chemical reactions are also affected by tem-
perature, including chromium reduction/precipitation and similar processes. In
general, reactions proceed more slowly at lower temperatures and a chemical treat-
ment unit process designed for optimum operation at high temperatures may perform
poorly if the wastewater cools substantially.
Data showing the effect of temperature on treatment processes have been pre-
sented in the literature. However, the effect of temperature variation on a full-
scale refinery waste treatment plant has not been rigorously treated. Figure 1 shows
the effect of temperature on effluent variability of BOD for a refinery using an
aerated lagoon system (Ref. 8). An aerated lagoon, because of its large surface
area and long hydraulic retention time, can be expected to show more temperature
impact on efficiency than any other type of biological treatment unit with the ex-
ception of facultative waste stabilization ponds. As an illustration, if all of the 13
monthly averages are averaged to determine the annual average mass discharge of
BOD, and this average is compared to the maximum of the monthly averages (Feb-
ruary), a ratio of 2.56 is obtained. However, if the data are arbitrarily divided
-------
417
into two groups, warm weather (May through November) and cold weather (December
through April) the corresponding seasonal variabilities are 1.15 and 1.48, respectively.
This example illustrates the potential effect of temperature on effluent variability from
a biological treatment system and this effect must be considered in the design of these
units.
Variations in Dissolved Solids
Fluctuations of inorganic dissolved solids concentrations are inherent in refinery
wastewater discharges. This can be attributed to changing volumes of cooling tower
or boiler blowdown (high TDS), cyclic treatment of contaminated ballast waters (high
TDS), or varying discharges of process wastewater (high or low TDS). Biological
treatment systems normally function more efficiently when treating low TDS waters,
although effective activated sludge treatment of wastewaters with a TDS approaching
that of sea water has been demonstrated (Ref. 9, 10). Abrupt changes in dissolved
solids, however, have a pronounced effect on biological system efficiency. The^
change of osmotic pressure, for example, disrupts biochemical mechanisms and reduces
the organic removal capacity of the biological population until the system can readjust.
This reduction in efficiency is often magnified by a noticeable increase in effluent
suspended solids discharged from the final clarifier when the wastewater TDS level
increases. This can be attributed to several factors; namely, a biological shift from
flocculation to nonflocculation microorganisms, a partial biological die-off, and an
increase in water density which adversely affects gravity sedimentation of biological
particles. Although sea water has a specific gravity of only 1.025, the effect is
significant when attempting to settle most biological floes.
Analytical Accuracy
An important factor in defining effluent variability is the precision and accuracy
associated with the analytical tests used to measure the concentration of each of the
controlled contaminants. Chemical analysis of wastewaters isadifficult task; these
wastes contain a great array of potentially interfering materials which can drastically
alter the results of an analysis with respect to the true value. Precision is a measure
of the repeatability of a certain analysis; accuracy is a measure of how close the
analytical results are to the "true" value. Precision data are usually easy to obtain,
but the accuracy is often difficult or impossible to ascertain for an actual sample.
Thus, accuracy measures for a given analysis are usually based on a "spiked" sample
and may have no relation to the actual recovery obtained in a wastewater sample.
In order to provide a basis for estimating the potential impact of analytical
variability on effluent variability, Table 2 is presented. The data presented in this
Table are taken from the recently published EPA manual on wastewater analyses (Ref.
11). It should be noted that many of these analyses were conducted on synthetic
samples and thus represent the best attainable analytical performance, which probably
cannot be duplicated on a refinery wastewater. The analytical variability of the BOD,
COD, TSS, and oil and grease methods are particularly important to the refinery
guidelines.' This will be particularly true for the 1983 limitations as they are
currently promulgated. The case-history effluent data given in this paper reflect this
analytical variability and thus the results obtained from the analyses presented herein
-------
418
should be tempered by this variability. Another detailed study of variability conducted
for API using data from the 1972 Raw Waste Load Survey (Ref. 12) showed that a sub-
stantial variation in calculated mass loadings could be attributed to analytical accuracy
(Ref. 13). Neither of these evaluations include potential sampling errors, which also
increase the variability of the effluent quality data.
Miscellaneous Factors
There are miscellaneous factors which cause effluent variability from biological
processes, such as changed in pH, nutrient deficiency, loss of dissolved oxygen, or the
presence of substances in the aeration basin at toxic or inhibitory concentrations.
These can be controlled, however, in the design and operation of a biological treat-
ment plant and are therefore excluded as determinants responsible for inherent vari-
ability. Various factors which can be included in this category and not previously
discussed are: abrupt changes in wet weather/dry weather flow patterns or climato-
logical disturbances; extended changes in organic loading attributable to emergency
situations in production units; and diurnal effects on photosynthetic activity in polish-
ing lagoons and facultative waste stabilization ponds.
INHERENT VARIABILITY IN TREATMENT UNITS
Primary and Intermediate Treatment
In petroleum refineries, primary treatment is invariably oil/solids separation,
often in a separator of conventional API design. The raw waste load data shown in
Table 1 reflect the effluent variability from typical, well-operated and well-designed,
gravity oil separators (without chemical aids).
Secondary oil/solids separation is used extensively in the petroleum refining
industry to reduce the levels of these contaminants prior to subsequent treatment for
organics removal. Chemically assisted air flotation is the most frequently used unit
process, although granular-media filtration has been successfully used in this applica-
tion at some refineries. Table 3 presents case-history data on effluent variability for
four air flotation units and one granular-media filter. Oil is the primary constituent
of interest for these unit processes. Figure 2 presents the data from one of these case
histories. Note that the variability ratio for the unit effluent is increased relative to
that for the influent; this is due to the extremely low average effluent concentration
from this unit and the magnitude of the variation between the average and 99 per-
centile values is much less than the corresponding influent values. Another air flo-
tation unit shown in Table 3 showed a decrease in variability of the effluent relative
to the influent; this is probably due to highly effective unit operation and chemical
dosing plus the fact that the effective unit operation and chemical dosing plus the
fact that the influent had an extraordinarily high variability considering the magnitude
of the mean. The one granular-media filter for which data were available for this
application showed considerably more afluent variability than the flotation units.
-------
419
Biological Treatment Processes
Case-history information from 17 different petroleum refinery wastewater treatment
systems was analyzed. The data base consisted of:
(1) 11 sets of effluent data, each of which covered greater than
12 consecutive months of system operation;
(2) three sets of effluent data, each consisting of 12 consecutive
months of system operation;
(3) two sets of data, each covering 11 months of operation; and
(4) one set of data covering four months of operation.
The important characteristics of the refinery treatment systems are summarized in Table
4.
As noted in the Table, some of the refineries used in this analysis were taken
from the EPA variability analysis presented in Supplement B to the Development Docu-
ment (Ref. 14).
All available effluent quality data for the case-history refinery treatment plants
were analyzed and the long-term average concentrations for the contaminants con-
trolled by the effluent guidelines are presented in Table 5. These concentrations can
be considered as annual averages although, in most cases, the data base used was
greater than 12 months in length. These data illustrate that the refineries used in
this analysis represent the "state of the art" in terms of biological treatment effective-
ness and water use and the variability of the effluent from these plants should be
indicative of the minimum attainable effluent variability to be expected in a well-
designed and well-operated facility.
The daily and monthly variability factors calculated for the case-history
refineries are shown in Tables 6 and 7, respectively; including both the range and
the average factors for each pollutant parameter. Comparison of the variability
factors for the two major types of end-of-pipe treatment systems shown reveals that
the effluent from an activated sludge system is more variable than that from an
aerated lagoon and polishing pond system.
The lower variability of the aerated lagoon and polishing pond effluent is pro-
bably attributable to the long hydraulic detention time required in this type of system
to obtain the long-term effluent qualities consistent with those attainable with acti-
vated sludge systems. For example, the aerated lagoon/polishing pond system used in
this study with the lowest short-term effluent quality variability has better than six
days of hydraulic detention time. Moreover, this refinery is in the southwestern
United States and temperature effects on effluent variability should be negligible. In
general, the aerated lagoon/polishing pond system provides low effluent variability
although the control of suspended solids in these effluents may be a problem. An
activated sludge plant followed by a large holding pond also produces low effluent
variability. An activated sludge plant with large holding pond (six days hydraulic
-------
420
detention time or greater) could be expected to have variabilities similar to the
aerated lagoon/polishing pond system
For some wastewater constituents, an aerated lagoon followed by a dissolved air
flotation unit with chemical addition decreased the effluent variability. However,
this system did not produce minimum effluent concentrations for most wastewater con-
stituents (see Table 5).
No full-scale data are available for rotating biological surface (RBS) systems.
However, for comparison, variability data from an RBS pilot-plant operated at a large
refinery is included. The effluent variability from this unit is quite similar to that of
the activated sludge units.
Effluent Filtration
Long-term operating data on the use of granular-media filtration for polishing
of biological effluents were available from two refineries. The variability ratios for
these two plants are shown in Tables 6 and 7, while the long-term average effluent
quality is shown in Table 5. For comparison, data from the pilot study of polishing
filtration on a pilot RBS effluent is also presented in Table 6. The addition of the
filtration unit reduced both the monthly and daily variability for the activated sludge
plant, however, the aerated lagoon with polishing filtration showed higher variability
than the aerated lagoons followed by large polishing ponds. This does not necessarily
indicate that the latter had better effluent quality, as is evidenced by the operating
data in Table 5.
Carbon Adsorption
There are two existing full-scale carbon plants in the industry: one which treats
refinery process wastewaters after secondary deoiling and another which intermittently
treats contaminated storm waters. Neither has the biological treatment plant preceding
the activated carbon which EPA has used as its 1983 BATEA model. Thus, all
performance estimates must be made using data from these two full-scale systems and
pilot studies operated on biologically treated refinery wastewaters.
Case History 1 data are from a refinery process wastewater treatment system con-
sisting of gravity oil/solids separation, sand filtration, and activated carbon adsorp-
tion. There have been operational and process problems both within the refinery and
in the treatment plant since start-up, so the variation is higher than may be expected
in an "exemplary" system. Daily effluent concentration data were analyzed by the
method outlined previously and daily variability factors are presented in Table 8.
The data did not cover a sufficient period of time to calculate monthly variability
data. Figures 3, 4, and 5 show the effluent quality probability plots for this plant.
Case History 2 is a refinery treatment system consisting of an equalization/
settling basin followed by activated carbon adsorption. A mixture of retaind con-
taminated storm water runoff and process wastewater is routed through this system at a
constant rate. This wastewater is considerably more dilute than normal refinery
-------
421
process wastewater. Again, monthly variability factors could not be calculated
because of the short duration of the data base. The calculated daily variability
factor for COD for this case is also listed in Table 8. An important consideration in
the evaluation of the variability for this plant is that the wastewater is fed to the
system at a constant volumetric rate because it is retained storm water. Thus, these
data are not influenced by the flow variations normally encountered in refinery process
wastewater systems.
Case History 3 is an activated carbon pilot plant following a pilot RBS system.
The variabilities are low, probably for two reasons. The first is that the loading on
the biological system was relatively closely controlled. In addition, the preceding
biological treatment process probably reduces the carbon effluent variabiality due to
lowered feed variability compared to the Case History 1 system.
The variability data shown above for activated carbon treatment are consistent
with that shown in Table 6 for the biological systems. There seems to be little im-
provement in effluent variability from a carbon system compared with a well-designed
and well-operated biological system. This is not to indicate, however, that carbon
adsorption following biological treatment does not improve the overall effluent quality.
DESIGN CONSIDERATIONS
Variability data for major refinery wastewater treatment processes have been pre-
sented. The inherent variability in refinery wastewater treatment processes should be
carefully considered in designing a treatment plant to comply with effluent limitations.
Based on a rather limited data base, some variability ratios for the maximum 24-hour
and 30-day average concentrations can be postulated using the data in Tables 3, 6,
7, and 8 in this paper. If adequate control of storm water is practiced, spills are
minimized, and sufficient equalization capacity is available in the system, the
average values shown in these Tables can be used as a design basis for variability.
If there is a large amount of equalization provided ahead of a biological unit and/or
carbon system, or a large polishing pond following complete treatement, the lowest
variabilities presented herein should provide a conservative estimate of effluent vari-
ability. Conversely, if equalization is nonexistent, the units are underdesigned,
and/or storm surges can be expected, variabilities greater than the higher recorded
values for the exemplary plants used in this investigation can be expected.
In general, as the effluent concentration or loading from a treatment system is
decreased, the effluent variability expresses as a ratio increases. This phenomenon is
primarily due to the fact that the distribution of effluent concentrations or 1 adings is
bounded at the lower end by the nonremovable portion of a particular constituent and
by the sensitivity of the analytical tests involved. There is no similar boundary on
the upper end of effluent concentrations. For example, a 15 mg/1 variation in BOD
when the long-term effluent concentration is 15 mg/1 represents a variability factor of
2.0 whereas the same 15 mg/1 variation represents a variability factor of 3.0 if the
long-term average is 5.0 mg/1 .
-------
422
Another important consideration is the fact that some of the variability factors
are derived from concentrations rather than mass discharges. To illustrate the effects
of deriving variability factors from concentration data rather than from mass loading
data, monthly variability factors for several constitutents have been calculated by both
methods. From the results in Table 9 it can be seen that mass-basis factors may ex-
ceed concentration-basis factors by as much as 42 percent due to wastewater flow
variations.
Examination of the variability factors for each individual system studied revealed
that the variability may be high for one waste constituent and low for another in any
given system. This fact is undoubtedly related to the differences in constituents in
different refinery wastewaters which can be related to product mix and refiner com-
plexity and the inherent variability in biological systems. These are often manifested
in the form of different effluent qualities from two similar treatment systems receiving
the same type of wastewater. It should be noted that the results presented herein
are in general agreement with an earlier study of variability of refinery wastewater
effluents conducted for API (Ref. 8).
One final observation concerning the variability factors developed as a result
of this study is the fact that the choice of a variability factor encompassing 99 per-
cent of the expected variation implies that the maximum daily effluent loading pre-
dicted on this basis may be expected to be exceeded four days of the year (one
percent of 365 days). Likewise, the maximum monthly average may be exceeded one
month out of every 50 months of operation if the 98 percent!le is used. Such
occurrences are inherent in a probabilistic approach to effluent variability and present
no major problems with respect to design, provided they are properly considered.
It is informative to establish the effect that low variability ratios would have on
noncompliance of a refinery effluent. Basically, the question must be asked: if we
choose a ratio of the 99 percent concentration to the annual mean and we are in
error by 25 percent in the low direction, how many days can be expected to be non-
compliant with the effluent limitations in a year? This could be answered by a
rigorous statistical analysis; however, a graphical analysis can be used to serve the
same purpose. Referring to Figure 6, let us choose an arbitrary log-normal distri-
bution with a median of 10 units and a 99 percent value of 35 units. The log stan-
dard deviation for this distribution is 0.536 and the mean is 11.6. The variability
ratio is thus 3.0 for this distribution. Now, assume that the variability ratio should
have actually been 4.4 and the distribution median remains at 10 units. This implies
that the previous 99 percent concentration of 35 now occurs five percent of the time.
This results in the effluent being noncompliant 18 days of the year, or 14 days over
and above the four days per year inherent in the selection of the 99 percent pro-
bability value. The reasoning can be extended to the variability ratios show in this
paper and illustrates the impact of selecting variability ratios when the resulting
effluent limits can result in large economic and criminal penalties.
-------
423
CONCLUSIONS
Several important conclusions may be drawn from this analysis:
(1) Inherent variability, which is defined as the effluent quality
variability attributable to the basic nature of the unit process
and the pattern of the raw waste loads, is an unalterable
characteristic of any industrial wastewater system. The raw
waste load variability, the effects of water temperature and
dissolved solids, and other treatment process idiosyncracies
which cannot be altered to enhance overall stability are the
primary factors responsible for inherent vraiability. The use
of flow equalization will lessen this variability, but will not
eliminate it.
(2) The most accurate approach in developing variability baseline
information from which effluent quality ranges can be pre-
dicted and effluent limitations formulated is to analyze case-
history information from selected operating plants. The plants
should be selected on the basis of good design and operation
so that the demonstrated variation is essentially that which is
inherent to the treatment process and wastewater being con-
sidered. Observation of pilot-plant data is much less realistic,
as inherent variability of full-scale plants is considerably more
pronounced than that observed in pilot plants in most cases.
Plants selected for this study were deemed to satisfy the afore-
mentioned criteria.
(3) The trend toward higher variability at lower effluent concen-
trations is supported by the fact that aerated lagoon systems
produce a less variable effluent than the more efficient
activated sludge systems.
(4) Failure to consider effluent variability in the design and
operation of refinery wastewater treatment plants will pro-
bably result in frequent noncompliance with effluent
limitations.
REFERENCES
(1) U.S. Environmental Protection Agency, "Development Document for Effluent
Limitations Guidelines and New Source Performance Standards for the Petroleum
Refining Point Source Category," Washington, D.C., (April 1974).
(2) Public Law 92-500.
(3) Texas Water Quality Board, "Hearing Commission Report - Public Hearing on
Format and Content of Permits Granted by the Texas Water Quality Board"
(March 14, 1974.)
(4) Sparr, T.S., Malouf, J.B., Cooper, H.B.H., and Hann, R.W., Interrelation-
ship of Water Quality Standards and Effluent Permits," Texas A&M University
Environmental Engineering Division (September 1972).
-------
424
REFERENCES (continued)
(5) Tarazi, D.S./ Hiser, L.L., Childers, R.E.7 and Boldt, C.A., "Comparison of
Wastewater Sampling Techniques," Report for the Texas Water Quality Board,
prepared by Southwest Research Institute (February 1969).
(6) Eckenfelder, W.W., and Ford, D.L., Water Pollution Control - Experimental
Procedures for Process Design, Pemberton Press, Austin, Texas (1970).
(7) Ford, D.L., Chapter on Treatment of Refinery and Petrochemical Wastes,
Industrial Wastewater Management, Ed. by H. Azad, McGraw-Hill, New York
(in press).
(8) Brown & Root, Inc., "Variability of Refinery Wastewater Effluent," Interim
Report No. CEA-7, Committee on Environmental Affairs, American Petroleum
Institute, Washington, D.C. (August 1974).
(9) Engineering-Science, Inc./Texas, Confidential Report (1972).
(10) Associated Water and Air Resources Engineers, Inc., Confidential Report (1973).
(11) Environmental Protection Agency, "Manual of Methods for Chemical Analysis of
Water and Wastes," Office of Technology Transfer, Washington, D.C. (1974).
(12) API/EPA Raw Waste Load Survey (1972).
(13) Brown & Root, Inc., "Analysis of 1972 API/EPA Raw Waste Load Survey," API
Publication No. 4200, Washington, D.C. (1974).
(14) U.S. Environmental Protection Agency, "Supplement B to the Development Docu
ment for Effluent Limitations Guidelines and New Source Performance Standards
for the Petroleum Refining Point Source Category," Washington, D.C., (April
1974).
DISCUSSION
Frank Manning, "Would you please compare your variability ratios with those
implied by the petroleum refining effluent guidelines?"
Lial Tischler, At BPCTCA level technology (1977), the average daily and
monthly variability factors for all constituents controlled by the petroleum refining
industry guidelines exeed the respective variability ratios used by EPA. In a few
cases, those systems with very long hydraulic detention times, the EPA variability
ratios or better are obtained for some, but never for all, constituents. Although
realistic long-term data for BATEA technology are not available to adequately define
variability ratios, it seems likely that the problem is exacerbated since EPA lowered
the 1983 ratios with respect to the 1977 values.
-------
425
BIBLIOGRAPHY
Lial F. Tischler is manager of the
Austin, Texas office of Engineering-
Science, Inc. He holds the degrees of
B.S. in Civil Engineering from the
University of Texas at El Paso and M.S.
and Ph.D. in Civil Engineering from the
University of Texas at Austin. He is a
registered professional engineer in the
State of Texas. Prior to joining
Engineering-Science, Inc., he was
Director of Systems Engineering at the
Texas Water Development Board.
TABLE
VARIABILITY FN RAW WASTE LOAD
RATIO OF 90%: 50%
MONTHLY DAILY
PARAMETER
Note;
Flow
BOD
COD
Oil &
Case 1
Case 2
Case 3
Case 4
AVERAGE
Case 1
1.4
1.7
1.5
Grease
includes stormwater
median oil and grease =
median oil and grease -
median oil and grease -
DATA
Case 2 Case 3 Case 4
1 .3
2.9
—
3.9 3.33 1.54
220 mg/l
68 mg/l
580 mg/l
-------
426
Constituent
BOD
COD
TOC
Phenolics
Oil and Grease
Sulfides
TSS
Ammonia
Chromium, Total
Cyanide, Total
TABLE 2 VARIABILITY ASSOCIATED WITH ANALYTICAL TESTS
FOR TYPICAL REFINERY WASTE CONSTITUENTS
20°C
Regular
Low Level
Chloride > 1000 mg/l
Infrared, Combustion
Extraction
Direct Photometric
Infrared - Freon
Separatory - Freon
Soxhlet - Freon
Titrameteric Iodine
Gravimetric
Distillation
Selective Ion Electrode
Automated
Atomic Absorption
Distillation
Test
Concentration
(mg/l)
175
2.1
270
12.3
0.0096
0.0483
0.0935
4.7
48.2
97.0
13.86](17.5)2
13.0 (12.6)
12.3 (14.8)
STATED
0.21
0.26
1.71
1.92
1.00
0.77
0.19
0.13
0.43-1 .41
0.353 (0.370)4
0.380 (0.407)
0.072 (0.074)
0.084 (0.093)
0.0102(0.0074)
0-.016 (0.015)
0.06
0.13
0.28
0.62
Standard
Deviation
(mg/l)
+ 26
+ 0.7
+ 17.8
+ 4.15
NOT
NOT
99%
Probabilty
(mg/l)
+ 67
+ 1.8
+ 45.9
+ 10.7
AVAILABLE
AVAILABLE
0.00099 0.0025
0.0031
0.0042
0.18
0.48
1.58
1.4
0.9
1.1
NOT
0.008
0.011
0.46
1.24
4.08
3.6
2.3
2.8
AVAILABLE
PRACTICAL RANGE 10-1000
0.122
0.07
0.24
0.28
0.04
0.17
0.007
0.003
0.005
0.105
0.128
0.029
0.035
0.0078
0.009
0.005
0.007
0.031
0.094
0.315
0.18
0.62
0.72
0.10
0.044
0.018
0.008
0.013
0.27
0.33
0.075
0.09
0.02
0.023
0.013
0.018
0.08
0.22
Variability
99/50
1.38
1.86
1.17
1.87
1.26
1.16
1.12
1.1
1.02
1.04
1.26
1.18
1.23
mg/l
2.5
1.69
1.36
1.37
1.1
1.06
1.09
1.06
1.03
1.76
1.87
2.04
2.07
2.96
2.44
1.22
1.14
1.28
1.35
Represents recovery of 14 mg/l of a known oil mixture from sewage
Represents recovery by each method from a sewage sample
Using two-tailed normal distribution
4
True concentration in "spiked" synthetic sample
-------
427
TABLE 3: SECONDARY OIL/SOLIDS SEPARATION
INFLUENT OIL
EFFLUENT OIL
Median
Concentration
UNIT TYPE mg/|
Induced Air Flotation 55
Dissolved Air Flotation 580
Dissolved Air Flotation
Dissolved Air Flotation
Granular
Average
Adjusted
Case
History
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
68
Media Filtration
Variability
Variability
Median
Variability Concentration
90/50 mg/l
1.75 9
3.33 AS
1.54
90/50 - Flotation Units Only
99/50 - Assuming log normal probability -
- Assuming Gaussian
TABLE 4: REFINERY CASE
probability
HISTORIES
Average
Refinery
End-of-pipe Wastewater
Subcategory Treatment System Flow (gal/bl)
C
C
B
B
B
D
B
D
D
B
C
D
B
C
C
E
D
Activated Sludge
Activated Sludge
Activated Sludge
Activated Sludge
Activated Sludge
Roughing Filter
Activated Sludge
Activated Sludge
Polishing Pond
Aerated Lagoon
Polishing Pond
Aerated Lagoon
Aerated Lagoon
Polishing Pond
Aerated Lagoon
Polishing Pond
Aerated Lagoon
Dissolved Air Flo-
tation with Chemical
Addition
Activated Sludge
Sand Filtration
Aerated Lagoon
with Alum Addition
Mixed-media Filtration
Activated Sludge
Polishing Pond
Aerated Lagoon
Polishing Pond
Activated Sludge
17.8
17.9
9.6
19.6
14.8
i
12. 71
31 .2
28.6
82
8.4
33.4
82
14.8
31
21.5
44.2
33.3
12
15
19
Flotation Units
- Flotation
Units
Variability
90/50
2.4
1 9
1 • 7
2.3
1 7
1 • /
3 8
\J *\J
2.1
3.8
3.0
- BIOLOGICAL TREATMENT
EPA
line
Guide-
Flow
(gal/bl)
22
19
25
22
31
53
31
4]
58
16
21
58
31
29
.7
.9
.0
.8
.0
.5
.0
.6
.4
.4
.8
.4
.0
.1
29.4
50.0
36.4
Sampling Data Base
Frequency (months)
Daily Composite
Daily Composite
Daily Composite
3 days/week
Daily Composite
3 days/week
Three -day
Composite
Daily Composite
Daily Composite
Daily Composite
2 days/week
Daily Composite
Daily Composite
2 3 days/week
Daily Composite
Daily Composite
Three -day
Composite
Daily Compositie
Daily Composite
Daily Composite
Daily Composite
22
26
31
36
22
35
12
11
19
11
35
19
17
12
34
12
4
Location
Within
U. S.
South Central
South Central
South Central
South Central
Midwest
South Central
Midwest
Midwest
Midwest
South Central
West Central
Midwest
Midwest
East Coast
West Central
South Central
South Central
Vhis does not constitute the total process effluent discharge at this refinery.
-------
428
TABLE 5: CASE-HISTORY LONG-TERM EFFLUENT CONCENTRATIONS
BIOLOGICAL TREATMENT
Annual Average Concentrations in mg/l
Throughout
Case Capacity Treatment Oil and
Number Subcategory (1,000 b/sd) System BOD COD TOC TSS Grease Phenol NHgH Sulfide Cr Cr
0.41 -
1.35 --
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
C
C
B
B
B
D
B
D
D
B
C
D
C
C
E
D
B
C
60
63
135
63
205
293
160
175
331.6
63
85
331.6
53
95
420
90
205
53
AS 13
AS 16
AS 18
AS 28
AS 12
TF+AS 1 8
AS+PP —
AL+PP —
AL 36
AL+PP 21
AL+PP 1 0
AL+DAF 15
AL 25
AS+PP 24
AL+PP 7
AS+PP 8
AS+GMF 1 1
AL+GMF16
174
130
132
183
112
121
87
68
—
187
148
—
—
142
98
174
104
—
78
30
23
41
15
19
7
28
47
24
20
15
58
41
21
30
13
27
9.4
9.0
—
—
9.9
—
—
5.1
35
--
—
5
13
—
—
7.0
8.6
7.8
0.053
0.15
—
—
0.04
—
—
0.04
0.43
—
—
0.35
0.80
—
—
0.08
0.05
0.47
1.10
6.7
—
—
8.2
—
13.7
28
3.9
—
—
3.8
54
~
—
—
3.4
53
O.Z
0
—
—
0
—
—
—
—
—
—
—
1.3
—
—
—
0
0.9
0.12
AS - Activated Sludge AL - Aerated Lagoon GMF - Granular Media Filtation DAF - Dissolved Air Flotation
PP = Polishing Pond
TABLE 6: DAILY CASE-HISTORY VARIABILITY FACTORS
Parameter Hexa-
Ammonia Sul- Total valent
BOD COD TOC TSS O&G Phenols Nitrogen fide Chrom. Chrom.
BIOLOGICAL TREATMENT
Activated Sludge
(8 plants)
Range 3.7-7.6 1.8-5.0 -1
Average 5.2 3.4 3.5"
Aerated Lagoon and Polishing Pond
(7 plants)
Range 1.6-3.6 1.4-2.4 NA
Average 2.5 2.0 NA
Aerated Lagoon & Dissolved Air Flotation with
Chemical Addition (1 plant)
Average 5.8 NA NA
Rotating Biological Surface (Pilot Plant)
Average 6.7 2.6 NA
EFFLUENT FILTRATION
Activated Sludge, Sand Filtration (1 Plant)
Average 4.7 4.3 NA
Aerated Lagoon, Mixed-Media Filtration
with Alum Addition (1 plant)
Average 4.9 NA NA
Rotating Biological Surface,
Sand filtration (Pilot)
Average 6.1 3.7 NA
I 2 3
Dash indicates one set of values NA indicates information not available Three-day composite used to
determine daily variability Factor.
2-5.2
3.6
8-2.7
2.5
2.3
NA
4.2
2.5
NA
3.9-5.4
4.7
2.9-6.1
4.5
3.6
NA
3.5
3.1
NA
4.5-7.5
5.2
2.4-6.7
4.5
2.2
9.0
8.6
2.1
NA
9.5-10.9
10.2
2.8-3.4
3.1
3.2
NA
7,2
6.0
NA
T
5.6
NA
NA
NA
NA
NA
NA
NA
4.6
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
11.4
NA
NA
-------
429
TABLE 7: MONTHLY CASE-HISTORY VARIABILITY FACTORS
BOD COD TOC TSC
O&G
Hex-
Ammonia Sulf-Total avalent
Phenols Nitrogem fide Chrom. Chrom.
3.4
1.7-2.6
2.2
BIOLOGICAL TREATMENT
Activated Sludge (7 plants
Range 2.2-4.6 1.5-1.9 -1 1.5-3.3 1.8-2.9 1.4-3.4 2.1-4.1
Average 2.7 1.8 2.9 2.5 2.2 2.5
Aerated Lagoon and Polishing Pond (6 plants)
Range 1.7-2.2 1.4-2.1 NA 1.4-2.3 1.9r5.7 2.2-4.1
Average 2.0 1.8 NA 1.9 3.8 3.1
Aerated Lagoon and Dissolved Flotation
with Chemical Addition (1 plant
Average 3.2 NA
EFFLUENT FILTRATION
Activated Sludge, Sand Filtration (1 plant)
Average 4.1 2.2
Aerated Lagoon, Mixed Media Filtration
with Alum Addition (1 plant)
Average 2.6 NA
-- 1.8-2.3 NA
1.6 2.0 NA
NA
NA
NA
1.6
3.2
2.0
1.9
1.9
5.0
1.7
3.2
NA
6.2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA NA
NA 2.9
NA NA
1 2
Dash indicates one set of values NA indicates information not available.
TABLE 8: ACTIVATED CARBON EFFLUENT VARIABILITY
Case History
1
1
1
2
3
Parameter
Oil & Grease
Phenols
COD
COD
COD
Phenols
Daily Variability*
7.8
8.6
4.0
2.6
2.9
5.7
Remarks
After Secondary Oil Removal
After Secondary Oil Removal
After Secondary Oil Removal
Stormwatsr
After Biological Treatment
(Pilot Scale)
After Biological Treatment
(Pilot Scale)
Ratio of 99% concentration to long-term average.
TABLE 9: COMPARISON OF CALCULATED VARIABILITY FACTORS FOR
MASS-BASIS AND CONCENTRATION-BASIS
Parameter
BOD
COD
TSS
Oil and Grease
Phenols
CrT
Case No.
5
9
12
1
5
1
5
9
12
1
5
9
12
5
Monthly Factor
Mass Basis
2.8
2.2
3.9
1.8
2.7
1.8
3.2
1.4
1.6
2.9
1.8
2.0
2.1
3.6
2.6
Monthly Factor
Concentration Basis
2.2
3.2
1.6
1.9
1.5
2.9
1.4
1.5
2.4
1.7
1.8
1.9
3.4
2.3
Ratio
1.07
1.00
1.21
1.12
1.42
1.20
1.10
1.00
1.06
.20
.05
.11
.10
.05
1.13
-------
430
MONTHLY AVERAGE VALUES
NOTE : Approximate average high and
low temperatures are 55"and
36°F ( Ref. 8)
ao
00
in
a
o
CD
J J A
S 0 N D J
MONTH
A M J
Figure I . EFFECT OF TEMPERATURE ON EFFLUENT
VARIABILITY OF BOD
0>
E
O
Figure 2. INDUCED AIR FLOTATION PERFORMANCE
OIL REMOVAL
120 -
INFLUENT
90% Volue=98 mg/l
50% Value = 55 mg/l
90 % Value= 22 mg/l (77.5% eff.
50%Value=9mg/l
10 50 90 99 99.9 99.99
PERCENT EQUAL TO OR LESS THAN
-------
UJ
(S)
<
UJ
rr
o
c
o
.J
O
Figure 3. INFLUENT AND EFFLUENT OIL AND GREASE DISTRIBUTIONS
FOR FULL-SCALE ACTIVATED CARBON SYSTEM
431
180
160
140
120
100
80
60
40
20
0
~ Average lnfluent= 30.2 mg/l
S.D. = 36.3 mg/l
Coef.Var. = 1.20
No. of Data Points = 55 over 5 months
— Average Effluent: 7.4 mg/l
S.D. = II.I mg/l
Coef. Var. = 1.5
No. of Data Points = 50 over 5 months
/
INFLUENT
EFFLUENT
10 20 30 50 70 90 99
PERCENT OF THE VALUES LESS THAN STATED VALUE
V)
_j
o
z
UJ
i
a.
Figure 4. INFLUENT AND EFFLUENT PHENOLS DISTRIBUTIONS
FOR FULL-SCALE ACTIVATED CARBON SYSTEM
90
80
70
60
50
40
30
Average Influent = 13.06 mg/l
S.D. = 14.15 mg/l
Coef.Var. = 1.08
No. of Data Points = 77 over 5 months
— Average Effluent = 9.22 mg/l
S.D. = 11.14 mg/l
Coef. Var. = 1.21
No. of Data Points = 77over 5 months
/
/ INFLUENT
EFFLUENT
10 20 30 50 70 90 99
PERCENT OF THE VALUES LESS THAN STATED VALUE
-------
432
Figure 5. INFLUENT AND EFFLUENT COD DISTRIBUTIONS
FOR FULL SCALE ACTIVATED CARBON SYSTEM
Q
o
o
f-
z
UJ
o
z
o
a
<
z
o
a
900
800
700
600
500
400
300
200
100
0
Average influent: 368.2 mg/l
S.D. = 189.6 mg/l
Coef. Var. = 0.515 /
No. of Data Points = 62 over 5 months /
Average Effluent= 251.8 mg/l
S.D.= 180.2 mg/l /
Coef. Var. = 0.716 /
No. of Data Points = 63 over 5months/
//
/
INFLUENT /
EFFLUENT
I
I 10 20 30 50 70 90 99
PERCENT OF THE VALUES LESS THAN STATED VALUE
Figure 6. IMPACT OF VARIABILITY RATIO SELECTION ON
EFFLUENT COMPLIANCE
100
80
60
40
20
10
8
6
5
C99 = 35
C50 = IO
MEDIAN
10 20 30 50 70 90 99
°ERCENT OF THE VALUES LESS THAN STATED VALUE
-------
REUSE OF REFINERY WASTEWATER
R. T. Milligan
Becht-el Corporation, San Francisco, California
The enactment of Public Law 92-500, the Federal Water Pollution Control Act Amend-
ments of 1972, established goals for the clean-up and elimination of pollutants discharged
into navigable waters. These goals are to achieve Best Practicable Control Technology
Currently Available (BPT) by July 1977, Best Available Treatment Economically Achievable
(BAT) by July 1983, and Zero Discharge of Pollutants by July 1985. Maximum reuse of
wastewater must become a normal operating practice if we are to achieve these goals at
any practical cost.
(I 2 3)
Numerous papers ' ' directed toward water reuse have been written in the past
few years. The purpose of this paper is to summarize ideas for the reduction of water use
and for reuse of wastewaters in the petroleum refining industry, and to suggest a step-by-
step progression from BPT to BAT and to Zero Discharge for existing refineries. A "broad
brush" approach has necessarily been used in this paper and some of the concepts proposed
may not be applicable in specific cases.
EXISTING REFINERY PRACTICES
Most petroleum refiners are now far along with their plans to achieve the 1977 goal of
Best Practicable Treatment (BPT). Allowable limits on the quantities of pollutants permitted
within this goal have been set in the individual NPDES (National Pollutant Discharge
Elimination System) permits issued to each refinery. Many refineries will achieve these
discharge limits by the addition of secondary treatment such as biological oxidation to such
existing refinery treatment as sour water stripping, API separators, air flotation units and
water retention basins. Conservation measures which have been used in refineries to reduce
both flows and contamination of wastewaters include:
o Elimination of once through cooling water
o Use of air cooling instead of water cooling
o Better sewer segregation to exclude uncontaminated runoff and to permit
separate treatment of highly contaminated streams
o Hydrotreating instead of caustic washing for sulfur removal
o Eliminating direct contact of water with oil such as in barometric condensers,
jet ejectors, open steam stripping, etc.
o Eliminating bleeds of clean condensate to waste
o Use of sour water stripper bottoms as feed water to the desalter unit. (This
practice not only reduces demand for treated make-up water and the dis-
charge of wastewater but also helps to recover some of the phenols present
in the sour water by their adsorption in the crude oil.)
433
-------
434
o Cascading blowdowns from boilers and cooling towers - i.e., using blow-
down from one system as make-up to a lower pressure or more concentrated
system
o Use of treated wastewater for fire water
o Better refinery housekeeping and maintenance practices
To better visualize the wastewater treatment steps for a representative refinery after
implementation of BPT, a simplified flow sketch, Figure I gives approximate flows for a
modern 100,000 bbl/day integrated refinery in which most of the cooling is provided by
air coolers. With this schematic representation of treatment steps for BPT, some prorposed
alterations required to meet BAT and Zero Discharge are discussed in the following para-
graphs.
COOLING TOWER OPERATION AND TREATMENT OF MAKE-UP WATER
The cooling towers are key factors in the pollution control strategy for refineries.
They contribute a large part of the wastewater by blowdown and bleed losses and their
make-up water is a large part of a refinery's water requirement. However, they can
tolerate a certain amount of organic pol lutants and dissolved salts in their make-up water
and they serve to oxidize and concentrate these pollutants. Thus cooling tower make-up
offers the greatest potential for wastewater reuse in a refinery.
The bleed losses referred to above include the cooling water used on a once through
basis in bearing cooling, pump gland seals, and flare seals. It is important to minimize
these bleeds and to control them so as to operate the tower near the maximum level of
dissolved solids that can be tolerated without excessive fouling and corrosion. Minimiz-
ing the cooling tower bleed of course means lowering demand for feedwater and for
chemicals for controlling corrosion and scaling.
The quality of cooling tower make-up water is a prime factor in setting the permis-
sible cycles of concentration which is usually dictated by a calcium sulfate or silica
scaling problem. Methods for cooling tower and boiler make-up water treatment to
prevent such scaling problems include:
o Filtration with polymer addition
o Cold lime softening
o Weak acid zeolite
o Sodium zeol ite
o Hot lime soda
o Hot lime zeol ite
o Demineral ization
o Reverse osmosis
-------
435
Each of these treatment methods has a different function and/or ion removal efficiency
and a combination of methods is usually required. Selecting the optimum overall economic
treatment scheme (i.e., softening plus organic removal) is most important in the water
management planning. It is indicated that with the increasingly stringent discharge
limits dictated by pollution control regulations, the optimum economic balance will
involve a much greater make-up water treatment than is now being applied.
(3)
Sun Oil's Toledo Refinery has demonstrated, that a cooling tower can operate
satisfactorily with make-up water containing significant amounts or organics. In fact the
cooling tower can actually serve as a biological treatment stage in wastewater reuse as
long as the level of organics is low enough to avoid excessive fouling or corrosion. Thus
stormwater runoff and other low salt containing refinery wastewaters, except heavily
contaminated foul waters, could be used as make-up water to the cooling towers with less
pretreatment than is needed for discharge from the refinery. Oil removal and/or filtration
would most probably be required of these streams before reuse but a highTdegree of bio-
logical treatment would not be necessary. However, wastewaters which are highly
contaminated with organics would certainly require further treatment before reuse because
of the possibility of biological fouling problems. Stripped sour water high in phenol would
normal y be used for desalter make-up and would, therefore, not be a candidate for make-
up to the cooling tower.
BOILER FEEDWATER TREATMENT AND REUSE OF CONTAMINATED CONDENSATE
As is cooling tower make-up, boiler feedwater is a big water use. Furthermore,
boiler feed normally receives a high degree of treatment, and it would seem that treating
wastewater to make boiler feed would combine two necessary treatments and save on both
cost of treatment and cost of fresh water. This strategy has considerable potential but not
enough is known about the kinds and concentrations of organic contaminants which boilers
can tolerate. Oily materials which would foam or carbonize on hot surfaces certainly
require almost total removal, whereas low molecular weight water-soluble organics which
activated carbon will not remove, may be far less damaging. Surface waters now treated
for use as boiler feed water often contain significant amounts of organic matter but uni-
formly recognized specifications on permissible kinds and concentrations are lacking.
Stripped sour condensates, because they are low in dissolved salts, are also potential
sources of boiler feed water, but besides the organics, residual sulfides would have to be
removed as well.
A CONCEPTUAL DESIGN FOR BAT
In planning ahead for the 1983 goal of BAT, consideration must be given to stepwise
or evolutionary changes in the BPT design which can achieve the new goal at minimum
cost. In general, large steps rather than piecemeal measures will be required. A
proposed scheme and water balance is presented in Figure 2 for the BAT case using the
same refinery basis as was used in Figure I. The principal difference is that this case
segregates the high salt content wastewaters and reuses the low salt wastewaters as cooling
-------
436
tower make-up. Since some of the low salt wastewaters will have high oil and organics
contents, an oil separation step and a biological oxidation step are included ahead of the
reuse of this water as cooling tower make-up. However, since the cooling tower can
tolerate some organics in the make-up water, a high degree of organics removal in the
biological oxidation step is not necessary.
Reuse of the stripped low phenol condensate as low pressure boiler feed water was
not considered for this design, as the cooling tower can use all the low salt wastewater.
The wastewater streams passing to the low salt treatment system and on to the cooling
tower include:
o Contaminated stormwater
o Sanitary wastewater (if not sent to a municipal system)
o Oily wastewaters from the process and process area washdowns
o Sour water stripper bottoms in excess of those fed to the desalter
(low phenol condensate)
Reuse of these streams in the case studied reduces the raw water make-up by 750 gpm.
By improved treatment of the raw water used as make-up to the cooling towers to permit
operation with more cycles of concentration, it is estimated that the blowdown from the
cooling tower could be controlled at 100 gpm ( 9 cycles of concentration with the drift
loss) with essentially no additional fouling or corrosion problems.
The wastewaters passing to the high salt treatment system would include:
o Desalter water
o Cooling tower blowdown
o Ballast water (if any, and if saltwater)
o Any neutralized spent caustic and acid streams not specially handled
After oil removal, these waters would be equalized, treated by biological oxidation and
filtration before finally passing to an activated carbon unit.
Regeneration waters from the ion exchange treatment of feed waters and boiler
blowdown could flow directly to a retention basin for mixing with the other treated
wastewater before discharge from the plant.
A CONCEPTUAL DESIGN FOR ZERO DISCHARGE
The stepwise changes required to convert the wastewater treatment system from BAT
to Zero Discharge primarily involve removal of salt from the final effluent of the BAT
scheme and reuse of the clean, salt-free water either as boiler feed (if it is as clean as
condensate) or as cooling tower make-up water.
-------
437
Since brine concentration and salt removal are quite expensive, part of the Zero
Discharge strategy should involve minimizing the amount of salt brought into the refinery
water. Several techniques for this are:
o Minimize the input of raw water by reusing as much as possible and
reducing evaporative losses by use of air cooling and control of steam
losses
o Use of reverse osmosis rather than ion exchange for demineralizing
boiler feed water
o Hydrotreat to sweeten and desulfurize instead of caustic treat
o If possible, reduce the amount of salt ballast water received
There are numerous possible alternatives for brine concentration and salt removal
and further study and plant demonstrations are required to achieve the optimum design.
A proposed conceptual design shown in Figure 3 includes reverse osmosis followed by an
evaporation system and a flash dryer. A wet coil concentrating cooling tower is a good
alternative to the reverse osmosis unit but may prove to be more expensive than reverse
osmosis when one considers modification of an existing refinery which already has sufficient
cooling capacity.
This study is based on concentrating the waste salts to dryness and transporting these
solids to an arid disposal site. Although the waste solids are concentrated to dryness,
their disposal will not be easy and an alternative using direct disposal of the evaporator
blowdown slurry should be considered.
CAPITAL COSTS
Order-of-magnitude capital costs of $6 million and $10 million respectively, have
been estimated in 1975 dollars for the modifications and additions required in changing
the 100,000 bbl/day refinery treatment system from BPT to BAT and to Zero Discharge
treatment technology. These costs indicate the large capital investment required to meet
the 1983 and 1985 goals.
CONCLUSIONS
Some of the principal conclusions from this study regarding the reuse of refinery
wastewater and achievement of the Public Law 92-500 goals are:
o The large water requirements of refinery cooling towers and their ability
to oxidize organic impurities and concentrate salts, make them logical
candidates for water reuse.
o
The correct degree of treatment for make-up water to refinery cooling
towers and boilers is important in obtaining an economic overall water
and wastewater treatment system.
-------
438
o More informaHon is needed about acceptable levels of organic impurities
in boiler feed water if refinery wastewaters are to be reused in this service.
o Petroleum refineries can probably reach the goals set in PL 92-500 using
existing technology but further verification of this technology needs to be
made by demonstration tests.
o Because BAT and Zero Discharge treatment will be complex, adequate
surge storage and equalization will be needed to get the necessary degree
of reliability.
o Large capital investments will be required by petroleum refiners to meet
the 1983 and 1985 goals.
REFERENCES
(I) Porter, J. W., Blake, J. H. and Milligan, R. T. "Complete Industrial Wastewater
Reuse Goal of Refining Study" 71 No. 40 70-74 (1973).
(2) Willenbrink, R. "Waste Water Reuse and Inplant Treatment" presented at AlChE
Meeting, Houston March 1973.
(3) Mohler, E. F. Jr., and Clere, L. T. "Development of Extensive Water Reuse and
Bio-Oxidation In A Large Oil Refinery" presented at National Conference on
Complete Water Reuse April 23-27, 1973.
DISCUSSION
Milt Beychok: Before I make this comment, Bob, I would like everyone to know that you
and I have been good friends for quite some while. I know that you weren't here the first
day when I discussed the impracticality of zero discharge. I submit that the process
designs you have shown us are not zero discharge, but rather the transformation of a saline
water problem into a solid waste problem. Obviously, we engineers can evaporate water
to remove dissolved inorganic salts (such as chlorides, carbonates and bicarbonates).
But unless we store the salts indefinitely in closed silos or lined ponds, they will return
to Mother Nature during the next rainfall whether by surface runoff or by leaching into
underground aquifers. I therefore repeat what I said in my keynote address, that we
cannot justify the large expenditure of capital and energy required simply to remove
water-solluble inorganic salts from our industrial and municipal effluents. To think
otherwise is to delude ourselves.
Bob Milligan: I agree with what you say Milton regarding the lack of justification for
removing non-toxic salts and I am sorry to have missed your talk the first day. However,
I believe that what I have presented is what at least the law, Public Law 92-500 implies
as zero discharges or elimination of discharge and it is zero discharge of pollutants to
navigable waters if the solid salts are transported far enough away from navigable waters
such as to a desert area.
Bill Ruggles: I would like to echo Mr. Beychok's comments. One of the difficulties that
those of us in Industry have had over the years with '"his environmental situation is ihat
-------
439
vendors of equipment and chemicals often get over-enthused with the efficiency of their
products. They make public utterances about them from time to time. They "sell" the
enforcement people on them. Then, those of us who are in industry are stuck with trying
to explain why we often can't use them.
This has been a real serious problem. It is therefore urged that your written paper
indicate these methods you are advocating are methods that may be possible under some
circumstances and that the situation not be oversimplified. As an example, you
mentioned the use of cooling towers for handling organic-bearing waste. It is well known
that in some cases this works very nicely. We very blithely started out trying it ourselves.
Talk about clogged heat exchangers — you never saw the like of that biogrowth. One
might say "why don't you use a biocide?" Well, there seems to be a reason or two these
days why we can't use biocides as we might like, including the fact they themselves can
sometimes be objectional in plant effluents.
There is'also, as Mr. Beychok said, the question about disposal of solid materials
that result from some of these processes you (Milligan) describe. This is also true in
using an organic-bearing water for such things as better make-up. A\\ you will probably
do in most cases, is just concentrate the organic in the blowdown. Then what do you do
with it? You are back to about the same situation as where you started. So, you really
have to think these things out carefully. People such as yourself in the consulting
business, in making positive statements on these methods instead of putting them in true
perspective as procedures that possibly may work in certain instances, may be selling us
in industry (your clients) down the river.
Bob Milligan: You will note that in the design presented that a treatment system to remove
the bulk of the organics was incorporated ahead of reuse of the wastewater as feed to the
cooling tower. The small amount of organics that are remaining after this pretreatment
does not have to be removed in most cases before feeding to the cooling tower if we
believe Sun Oil's experience and the experience of a number of other companies. I am
not suggesting that refineries incorporate this water reuse scheme blindly but I am saying
that this is a good possibility and that it should be considered not only from the standpoint
of further removal of organics but also from the standpoint of conserving the amount of
raw water taken into the refinery.
-------
440
BIOGRAPHY
Robert T. Milligan is Manager of Pollution
Control Technology for the Bechrel Corporation
in San Francisco, California. He has the respon-
sibility for the operation of three sections which
provide consulting services on air, water and
so I ids and noise pollution control.
Dr. Milligan holds the following degrees
in Chemical Engineering: B.S. from University
of Illinois and Ph.D. from Ohio State
University. He is a registered Professional
Engineer in the State of California. Dr.
Milligan has had 28 years of experience in
process engineering and technical management
with Shell Development and Shell Chemical
Companies before joining Bechtel.
-------
50
PROCESS
STRIPPED
SOUR
WATER
+ 300
STORM WATER
BASIN
OILY
WATER
300
CONTAMINATED WATER SEWER
CLEAN WATER SEWER
NUMBERS SHOWN REPRESENT
APPROXIMATE FLOWS IN GPM.
IF 150
200
RAW 3200
WATER
' BACKWASH
AND
REGENERANTS
WATER
TREATMENT
200
1800
DISCHARGE
SLOWDOWN
STEAM
1600
SLOWDOWN
600
PLANT
SERVICE
WATER
CONDENSATE 800
RETURN
EVAPORATION
AND DRIFT
1400
2000
Figure 1. BEST PRACTICABLE TREATMENT (BPT) CASE
-------
BALLAST _C_A
WATER °ESA
1
r 50 i
Trr, ^ PROCESS STORM WATER
LTIIn " BASIN
STRIPPED OILY
^Dl I R wATrn
WATpp
300 ,r ir 300 ^150 750 C
LOW-SALT CONTAMINATED " SEPA
WATER SEWER
r 200 350 ^ C
HIGH-SALT CONTAMINATED WATER SEWER M " SEPA
HIGH-SALT CLEAN WATER SEWER
NUMBERS SHOWN REPRESENT
APPROXIMATE FLOWS IN GPM.
)IL BIO-TREAT
RATOR * FILTER
)IL BIO-TREAT ACTIVATED
RATOR FILTER CARBON
200
RAW 1950^
WATER *
' BACKWASH
AND
REGENERANTS
SLOWDOWN
STEAM
1600
WATER
TREATMENT
200
PLANT
SERVICE
WATER
CONDENSATE 800
RETURN
SLOWDOWN
EVAPORATION
AND DRIFT
1400
750 750
550
DISCHARGE
Figure 2. BEST AVAILABLE TREATMENT (BAT) CASE
-------
NUMBERS SHOWN REPRESENT
BALLAST „„
WATER DESA
'
r 50
^n ^ PROCESS STORM WATER
LTCn * BASIN
STRIPPED OILY
1 SOUR WATER
WATFR
300 , r ir300 U150 750
LOW-SALT CONTAMINATED "" SE
WATER SEWER
r 200 350
HIGH-SALT CONTAMINATED WATER SEWER jl SE
HIGH-
SALT CLEAN WATER SEWER 2
APPROXIMATE FLOWS IN GPM.
OIL ^ BIO-TREAT
PARATOR * FILTER
OIL ^ BIO-TREAT ^ ACTIVATED
PARATOR FILTER CARBON
i
00
RAW
WATER'
1400
BACKWASH
AND
REGENERANTS
SLOWDOWN
STEAM
1600
WATER
TREATMENT
PLANT
SERVICE
WATER
200
SLOWDOWN
100
CONDENSATE 800
RETURN
EVAPORATION
AND DRIFT
1400
550
PRE-TREAT
REVERSE
OSMOSIS
400_I_MOO_
200
EVAPORATOR
350
SOLID
WASTE
200
1
Figure 3. ZERO DISCHARGE CASE
fe
CJ
-------
444
-------
BIOLOGICAL MONITORING OF PETROLEUM REFINERY EFFLUENTS
Sterling L. Burks
Oklahoma State University, Stillwater, Oklahoma
Increasing awareness of the problem in predicting the environmental effects of waste
water discharges based upon chemical parameters has caused a renewed interest in biologi-
cal toxicity tests. The 1972 amendments to the Federal Water Pollution Control Act were
enacted by Congress to give some support to states efforts to control pollution and to insure
uniformity in regulation of water quality. The goal of the act was to achieve "zero pollu-
tant" discharge by 1985. Where a pollutant was defined as any substance which, directly
or indirectly through the food chain, causes a deleterious effect upon any organism within
the aquatic environment. Thus, the problem facing water pollution scientists and engineers
is the definition of a "zero pollutant" discharge. Effluent criteria attempt to achieve this
goal by regulating the maximum amount of chemical substances which may be discharged to
the receiving stream. This approach has achieved a beneficial effect in preventing the
discharge of gross quantities of known toxic chemical compounds, however, this approach
to establishing effluent criteris has some limiting features which may cause inequities by
restricting maximum allowable concentrations at an unnecessarily low level or the criteria
may not adequately protect the organisms when the combined action of the chemical
pollutants is more than additive. In addition, many chemical toxins have not been identi-
fied yet and therefore can not be regulated based upon chemical criteria. As a consequence,
effluent guidelines based upon biological toxicity tests alone, or in combination with
chemical criteria, may be more equitable to both the regulated point source dischargers and
the organisms in the receiving body of water. The results of several different projects using
both biological toxicity tests and chemical analyses to evaluate the quality of petroleum
refinery effluents are presented in this paper. The objective is to illustrate how biological
toxicity tests can be utilized, either alone or in conjunction with chemical analyses, to
more effectively manage petroleum refinery waste waters.
The most commonly used biological test for determining water quality is the fish bioassay.
Basically, the fish bioassay is performed by exposing a number of fish, usually six to twenty,
to several different concentrations of a prepared chemical pollutant solution or an effluent
and the effect is determined by recording the percent mortality (quantal response) at
preselected time intervals. The percent response is then plotted versus the log of the
concentration to graphically interpolate (may be calculated by statistical methods) the
concentration which causes 50% mortality, usually expressed as the median tolerance limit
(TLM) or median lethal concentration LC50 (APHA Standard Methods, 1971).
The mechanisms of performing fish bioassay tests generally fall into two categories as
far as the method of exposing the fish to the test solution; i.e., a static method and a
continuous-flow method. In static tests, the test solution is prepared and the fish exposed
to the original test solution without any replacement or renewal of the test solution for the
duration of the test interval. In continuous-flow tests, the test solution is continually
passed through the test containers so that renewal or replacement occurs many times during
the test interval.
445
-------
446
The static test requires much less effort by trained personnel, less complicated mechanical
equipment and may be performed at laboratories some distance from the source of sample.
Continuous-flow tests require more complicated mechanical equipment to pump the test sample
through the bioassay containers. It is difficult to maintain a uniform constant rate of flow and
thus must frequently be checked by trained personnel. The volumes of water required to
perform continuous-flow bioassays make it very expensive and inconvenient to perform the test
at locations other than the source of the sample.
The Oklahoma State University Reservoir Research Center, in cooperation with the Oil
Refiners Waste Control Council, has been performing static acute fish bioassay tests on effluent
samples since 1959. The oil refiners collected samples of final effluent and shipped them to
OSU for analyses. The samples were testes for acute toxicity by a static 96 hour Fathead
minnow bioassay in accordance with APHA Standard Methods (1971). The chemical parameters
ammonia, pH, sulfide, phenol, chemical oxygen demand, and alkalinity were analyzed by
personnel of the oil refineries.
The combined chemical and fish bioassay data have been used as indicators of water quality
to detect upsets in water waste treatment systems and to determine the relative efficiency of
the systems for producing good quality final effluents. A comparison of the mean annual
LC50, expressed as percent effluent, of the influent and the effluent of a lagoon waste water
treatment system illustrates the value of fish bioassays (Fig. 1). The long range trend from
1960 to 1964 was a definite improvement in the influent to the treatment system. Since the
influent had received only oil separation through an API separator, the improved quality
must have been due to changes either in crude oil or processes within the refinery. From
1964 to 1975, the long range trend shows the influent has become more toxic. The increase
in toxicity of the influent to the lagoon system places a greater load on the biodegradation
capacity of the system until eventually the capacity is exceeded. This is reflected by the
LC50 of the effluent from the lagoon. Prior to 1970, the mean annual LC50 was always 95%
or greater. Since 1970, the mean annual LC50 of the effluent has decreased to below 95%.
This indicates that the load of chemical poisons is exceeding the capacity of the lagoon
system to degrade them. Examination of the mean annual COD data for this system shows
that the COD in the influent was increasing up to a maximum in 1971 but has decreased
since 1971 (Fig. 2). The mean annual COD of the effluent, although showing a slight
increase, has not changed much since 1961 . Thus the capacity of the system to degrade
COD has not been exceeded and the use of COD data alone would not permit the problem
to be identified.
A significant change in water quality occurred when a refinery installed an activated
sludge waste water treatment system (Fig. 3). The mean annual LC50 had gradually improved
from 1960 to 1964 as the refinery had made improvements in their treatment system, however,
installation of new processes and increase in crude capacity caused a decline in effluent
water quality from 1965 to 1968. The company decided to install an activated sludge system
and a dramatic improvement in water quality occurred from 1970 to 1972.
These two examples illustrate the value of fish toxicity tests as indicators of water
quality. Fish bioassay data have often indicated problems were occurring in a treatment
system where chemical parameters did not detect a problem. Used in conjunction with
-------
447
chemical analyses, fish bioassay tests can help identify problem process effluent streams
which can be segregated and subjected to more intensive treatment than possible with the
total combined waste water flow from a refinery. This approach has been successfully used
by several members of the Oil Refiners Waste Control Council.
The trend today in fish acute toxicity testing is towards a continuous-flow method of
exposure. It has been suggested that static acute toxicity tests might not be obtaining a
reliable measure of the toxicity of an effluent because the effluent may contain biode-
gradable poisons, volatile toxins and was not homogeneous. A project was initiated to
compare the LC50's measured by static and continuous-flow methods of exposing fish to
effluents. The large volumes of water required to perform continuous-flow tests necessi-
tated that the tests be performed on-site at a refinery. A mobile trailer (96 ft of floor
space) equipped with heating and air conditioning was modified for performing fish bioassay
tests. The trailer was equipped with aquaria racks, plumbing, air compressor, and serial
dilution devices for delivering selected percentages of test effluent. In order to perform
toxicity tests with different percentages of effluent, it was necessary to obtain dilution
water. For purposes of simulating the conditions which occur in the receiving stream, the
most logical source of dilution water would be the stream just above the outfall of the test
sample. However, the depths of the stream banks and its inaccessability prevented the
use of the stream. The dilution water was obtained from a reservoir approximately 15
miles upstream from the outfall. The dilution water was transported to the refinery test
site with a 5,000 gallon tank truck. The test refinery effluent, treated with activated
sludge, was taken from the final clarifier prior to the final discharge to the receiving
stream.
The test fish, Fathead minnows (Pimephales promelas), were raised in the laboratory
from a breed stock which has been maintained since 1970. The fish were exposed to
duplicate concentrations of 0, 18, 42, 55, 74, and 100 percent effluent. The percent
dilutions for the continuous-flow tanks were prepared by continuously passing the effluent
through a glass serial diluter (modified from Brungs and Mount, 1971) where it was diluted
with water from the reservoir on the stream. The percent mortality was recorded at 0, 2,
4, 8, 16, 24, 48, 72, and 96 hours when possible. The exact time sequence could not
be maintained for all samples. The data were interpreted by the methods of Lichtfield
and Wilcoxon (1949) when possible. Most of the mortality occurred between 55 and 74
percent effluent and could not be analyzed by Lichtfield and Wilcoxon's method, so the
LC50's were determined by graphical interpolation. An LT50, median lethal time, was
also calculated by graphical interpolation (Sprague, 1969).
A comparison of the LT50 values from eight tests (Table 1, Fig. 4) indicates that the
median lethal time was of shorter duration in the static tests than in the continuous-flow
tests. Thus, the fish in the static test were under more stress and succumbed to the toxic
effects earlier than fish in the continuous-flow tests. However, in only two of the tests
does the difference appear to be significantly different. A comparison of the LCSO's
shows no significant differences between the static and continuous-flow bioassay tests
(Table 2, Fig. 5). Based upon these results, it would appear that there is no significant
difference in estimating an LC50 by the static 96 hour acute toxicity bioassay as contrasted
with a continuous-flow 96 hour acute toxicity bioassay.
-------
448
The most significant difference between the two tests was the difficulty in maintaining
uniform flow of the test sample in the continuous-flow tests. The flow of the effluent was
dependent upon a pump. If the pump failed, then the flow was stopped. If the flow of
effluent stopped, the proportional diluter became inoperative and there was no flow of water
to the test containers.
In my opinion, the added expense of obtaining dilution water, building diluters and of
maintaining pumps does not justify the slight improvements in LT50's or LCSO's. The major
advantages of continuous-flow tests accrue when the sample of analyses contains volatile
toxins which might escape in a static test or biodegradable toxins which could be degraded
in a static test. Apparently, these conditions did not cause a significant effect in the oil
refinery effluent tested during this study.
By mid-1977, all industries must attain a level of waste water treatment equivalent to
Best Practicable Treatment Control Technology (BPTCT) and Best Available Treatment Control
Technology (BATCT) by mid-1983. For the petroleum refining industry, BPTCT has been
defined as equivalent to sequential treatment with activated sludge followed by dual media
filtration. The BATCT level of treatment was defined as equivalent to sequential treatment
with activated sludge, dual media filtration, followed by adsorption on activated carbon.
Based upon reduction of specific chemical contaminants, these levels of treatment will
meet effluent criteria recommended by the regulatory agencies responsible for attaining goals
established by the 1972 amendments to the Federal Water Pollution Control Act. However,
effluent criteria based upon maximum allowable concentrations of chemical contaminants
were developed by attempting to relate acute toxicity concentrations measured in the labora-
tory with anticipated responses of organisms within the aquatic environment. The ultimate
test of the acceptibility of a treatment method must be the effect of the waste water discharge
upon the biological organisms within the receiving environment.
The fiscal year 1976 and 1977 project of the Oil Refiners Waste Control Council-USDI,
Office of Water Resources Technology-OSU Reservoir Research Center was designed to use
biological monitoring to evaluate the effectiveness of BPTCT and BATCT for producing a
non-deleterious effluent. This concept of using biological organisms to determine the
effectiveness of a treatment method is in accord with the intent of the 1972 amendments,
since "zero pollution" was defined by biological criteria.
Although this research project has only begun, the approach used for evaluation of the
advanced waste water treatment methods is unique and will be presented as an example of
the potential of biological monitoring.
Effluents from a control stream, an activated sludge unit, a combined activated sludge-
dual media filter unit (BPTCT), and sequential activated sludge-dualmedia filter-activated
carbon (BATCT) units will be continuously pumped through fish bioassay tanks to determine
fish toxicity, through artificial streams to determine effects upon diversity of benthic
organisms and periphyton.
The purposes of the fish bioassay tests are to relate the research results to commonly
-------
449
accepted short-term bioassay procedures and to explore the effects of long term exposure of
the fish to the test effluents. The fish will be considered to have completed the life cycle
if they produce viable eggs.
In addition to the fish bioassays, effects of the test effluents on benthic organisms will
be tested in simulated streams. Changes in diversity of the benthic aquatic invertebrates
have been shown to be useful parameters for determining the quality of an aquatic
environment (Wilhm and Dorris, 1968). The benthic invertebrate assemblage is composed
on many species of organisms, some of which are sensitive to changes in environmental
quality and others which are relatively tolerant to changes. Adverse changes in environ-
mental quality cause reduction in the numbers of individuals and species of sensitive organisms
and may cause an increase in numbers of individuals among species of tolerant organisms.
The changes can be quantified by a diversity index (Shannon and Weaver, 1963):
_ n. n.
d= - —— Iog0 -i-
Where d = species diversity
n. - numbers of individuals in the ith species group
n = total number of individuals
Nondisturbed benthic communities usually exhibit a diversity index of 3 or greater,
whereas the diversity index of benthic communities in polluted environments is less than 1.
Benthic organisms for the artificial streams were obtained by installing artificial substrate
samplers in a nearby unpolluted stream and allowing the natural organisms to colonize the
samplers for a period of six to eight weeks. The colonized samplers were then transported
to the refinery test site and installed in the effluent streams from the test treatment units.
Mean diversity of the colonized samplers was determined by analysis of four samplers before
exposure to the test effluents. Two artificial substrate samplers were removed from each
duplicate stream at 2, 8, 16, and 30 days after start of exposures to the test effluents.
The diatom community was also used as a biological test monitor. New microscope
glass slides were placed in each artificial stream in specially constructed racks. The slides
were supported vertically and oriented parallel to the current of the stream. Slides were
removed at 8, 16, and 30 days after start of exposure. Two kinds conformation were
obtained from the diatom exposures. The composition and diversity (d) of the assemblages
were determined after full colonization and compared among control and test streams to
yield the same kind of information as in the benthic communities. In addition, a physio-
logial measure of growth was obtained by using ash free biomass and numbers of individuals
(Kevern, Wilhm, and Van Dyne, 1966).
A thirty day test has just been completed at a refinery, from Nov. 4 to Dec. 4, 1975,
and the data which have been analyzed will be presented as an example of the potential
biological monitoring has for determining waste water quality. The results of the continuous-
flow fish bioassay tests show that no mortality occurred in the first six days of the exposure
(Fig. 6). These results indicate that in the normal 96 hour bioassay test, the response of
-------
450
the fish would not be sensitive enough to perform an evaluation of the "zero pollutant" level
of the treatment methods. After thirty days exposure to the test effluents, there was 70%
mortality in the activated sludge effluent, 20% in the combination activated sludge-dual
media filter (BPTCT) effluent, and 15% in the combined activated sludge-dual media-acti-
vated carbon (BATCT) effluent. Thus, on the basis of fish mortality, the additional treatment
by dual media filters or activated carbon improved the quality of the effluent. However, the
activated carbon did not improve the quality as much as would be expected, since it is advo-
cated as the Best Available Treatment Control Technology.
Analyses of species diversity values have been completed on 26 samples (40% of the total
exposed) of benthic invertebrate organisms. After 9 days of exposure, no significant changes
in species diversity of the benthics had occurred (Table 3). Thus the benthic organisms, like
the fish, showed almost no response the test effluents for short-term exposure.
The data presented in this report are too preliminary to permit any conclusive evaluations
of Best Practicable or Best Available Treatment Control Technologies, but the approach illu-
strates how biological monitoring may be used to evaluate advanced waste water treatment
technologies. Such an approach could be very valuable to personnel in the petroleum industry
who must make a decision on the type of treatment facility to install to meet the "zero pollutant"
discharge by 1985.
REFERENCES
(1) American Public Health Association, "Standard Methods for the Examination of Water and
Wastewater," APHA, Washington, D. C. 13th edition (1971.
(2) Kevern, N. R., J. R. Wilhm, and G. M. VanDyne, "Use of Artificial Substrata to
Estimate Productivity of Periphyton," Limnol. and Oceanogr. 11:499-502 (1966).
(3) Litchfield, J. T., Jr., and F. Wilcoxon, "A Simplified Method of Evaluating Dose-Effect
Experiments," J. Pharm. Exp. Ther. 96:99-113(1949).
(4) Mount, D. A., and W. A. Brungs, "A Simplified Dosing Apparatus for Fish Toxicological
Studies," Water Res. 1:21-29 (1967).
(5) Sprague, J. B., "Measurement of Pollutant Toxicity to Fish I. Biocssay Methods for Acute
Toxicity," Water Res. 3:793-821 (1969).
(6) Wilhm, J. L., and T. C. Dorris, "Biological Parameters for Water Quality Criteria,
Bioscience 18:477-481 (1968).
DISCUSSION
Neale Fugate: In the next to the last slide, I believe you showed the mortality at the end of
the 30-day period, so there is a 15% mortality, is that a crack of doom?
Sterling Burks: No, I don't think so. I think we really don't understand going beyond acute
toxicity, we are out there in a brand new world. We don't know other than the laboratory
studies that have been conducted on chemical poisons by the Duluth National Toxicology
Laboratory, we know very little about the long term effects upon aquatic organisms. When
we get past 30 days, what is a significant effect? Let's use the organisms in the receiving
stream, let them tell us what is a significant effect and that is why we proposed the species
-------
451
diversity index (d). Now I am hoping that within a year we will have a lot more data and
will be able to tell you what the biological evaluation of BPT and BAT is. I think they will
give us an answer.
Frank Martin: I would like for you to comment on the Gambasia test that the Texas Water
Quality Board is talking about as a quick test to determine toxicity?
Sterling Burks: Well, I think this is a good test that will give you some reliable indices of
toxicity. I think any time we talk about an acute test, really what we are looking at is an
index of toxicity. We find that there is a lot of variability in biological organisms. As
you all know when you run a BOD test. You can't always predict what those critters are
going to do - the environment effects them tremendously. The dilution water, temperature,
pH, everything, so use a biological test as an index. I hate to see hard and fast numbers put
on them, but I think they are a good index of toxicity.
Lial Tischler: I want to speak to that particular point and get your comment on the use of a
number three or better for species diversity as being polluted or unstressed if you like. I
think in certain types of natural water, stressed eco-systems, I am thinking particularly of
some of the tidal estuaries in the Texas Gulf Coast, you do see species diversities in "natural
or unpolluted waters" which are quite a bit lower. I just think that you ought to emphasize
that species diversity to 1 .5 may be caused by natural conditions; also, I think that is the
case.
Sterling Burks: Maybe I didn't make that point clear and I appreciate your bringing that
point up. The species diversity value of three or greater is generally found in an unpolluted,
unstressed situation. The sampler trays that I showed you were installed in a natural stream.
In Oklahoma we are subject to variabilities of the weather. We installed these samplers in
a stream that had a nice flow of about one foot and a half deep, six or eight feet wide and
a nice clear looking stream. Six weeks later we went back and the stream had stopped flow,
was down in stagnant pools, lot of leaves, decomposing in the water, etc., so we had a
natural stress on that system. But the d on the organisms that we took out of this stream
was 2.75, so I think that natural communities have an ability to respond to stress. Some
organisms may disappear, but other organisms take their place. What I am trying to say is
let's strive for diversified aquatic communities. I am not saying that we should protect every
rare very sensitive species out there, I am saying let's strive for a balanced community in the
receiving stream. If we have a stream that should support a large mouth bass, then I think
we should have large mouth bass there; but if he is not very sensitive, not as sensitive as
some of the invertebrate organisms that might disappear.
Bob Huddleston: How big a job is it to carry out the assay on the benthic organisms?
Sterling Burks: I can tell you more before too long. We hove people who have to take
these rockTaTid wash off the debris, sort the organisms, and then classify the organisms. It
is not an easy job and it is a fairly time consuming operation. I would guess about three
hours to wash and sort, and somewhere between five and six hours to identify the organisms,
depending upon the taxomony of the invertebrates present. It is not an easy task.
-------
452
Ben Buchanan: Do you control the dissolved oxygen in the flow systems that you run the bioassays?
Sterling Burks: We had to, because we found the water coming out of the final clarifier was
coming out of the final clarifier was running about two parts per million so we had to aerate in
order to get the water up to adequate desired oxygen level so we would not stress the fish due
to low dissolved oxygen. The reason I did this is because we feel like we want to measure
dissolved poisons or pollutants that might be toxic in the effluent not low dissolved oxygen. I
don't want to measure that effect. I want to measure the effect of the dissolved pollutants.
Anonymous: Of course, the high dissolved oxygen, the higher it is well that usually provides
a better environment for the fish and so they will survive. Better if you had ten parts per
million than you had five or six.
Sterling Burks: True. In this particular refinery they have a good lagoon system following
their activated sludge system and their dissolved oxygen going to the stream is very good.
Out of this clarifier, it just was not very high at this time.
Anonymous: What has been your experience with dissolved salts on the osmotic effect on your
minnows?
Sterling Burks: We have not set up any tests, to test for this specifically, but I would think
that fish can tolerate quite a range in dissolved solids as long as they are not instantly changed
from maybe a 100 parts per million total dissolved solids to 5,000. Sudden changes really
affect aquatic organisms, gradual changes they can acclimate to, quite a range in chemical
environment.
BIOGRAPHY
Sterling L. Burks is an Assistant Professor of
Zoology at Oklahoma State University, Stillwater,
Oklahoma. He is the Assistant Director of the
Reservoir Research Center. Sterling received his
B.S. degree from Southwestern State College,
Weatherford, Oklahoma, in Biology and Chemistry.
M.S. and Ph.D. degrees in Zoology (Aquatic
Biology) were earned at Oklahoma State
University. Sterling spent two years on a Federal
Water Pollution Control Administration Post
Doctoral Fellowship before joining the faculty
at Oklahoma State University in 1972. Research
interests are the analyses and determination of
the effects of chemical contaminants upon
aquatic organisms.
-------
453
TABLE 1 "COMPARISON OF MEDIAN LETHAL TIMES OF FATHEAD MINNOWS
EXPOSED TO OIL REFINERY EFFLUENTS IN CONTINUOUS-FLOW AND
STATIC TESTS"
LT50(min.) LT50(min.)
Date Continuous-Flow static
5/5-5/10 690 360
5/12-5/17 380 380
5/21-5/25 550 90
5/28-6/3 1,600 500
6/4-6/9 2,100 2,100
6/9-6/15 1,460 1,075
6/16-6/20 800 800
6/23-6/27 1,150 890
TABLE 2 "COMPARISON OF MEDIAN LETHAL CONCENTRATIONS OF OIL
REFINERY EFFLUENTS TO FATHEAD MINNOWS IN CONTINUOUS-
FLOW AND STATIC TESTS"
LC 50 LC 50
Date Continuous-Flow static
375^5/10 60% 60%
5/12-5/17 55% 53%
5/21-5/25 42% 32%
5/28-6/3 77% 75%
6/4-6/9 81% 81%
6/9-6/15 78% 79%
6/16-6/20 74% 64%
6/23-6/27 70% 61%
-------
454
TABLE 3 "SPECIES DIVERSITY OF BENTHIC MACROINVERTEBRATES EXPOSED
TO OIL REFINERY EFFLUENTS TREATED BY ADVANCED
WASTE WATER TREATMENT"
Effluent Stream
Peavine Crk.
Peavine Crk.
Peavine Crk.
Peavine Crk.
Control
Control
Control
Control
Act. Sludge
Act. Sludge
Act. Sludge
Act. Sludge
Dual Media
Dual Media
Dual Media
Dual Media
Act. Carbon
Act. Carbon
Act. Corbon
Act. Carbon
Control
Control
Control
Control
Act. Sludge
Act. Sludge
Act. Sludge
Act. Sludge
Dual Media
Dual Media
Dual Media
Dual Media
Act. Carbon
Act. Carbon
Act. Carbon
Act. Carbon
Blanks in species
time (12/30/75)
Date
1975
10727"
10/27
10/27
10/27
11/6
11/6
11/6
11/6
11.6
11/6
11/6
11/6
11/6
11/6
11/6
11/6
11/6
11/6
11/6
11/6
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
diversity
Stream
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
1
1
2
2
column
# Rep. #
1
2
3
4
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
represent samples
Species
Diversity
d
1.84
2.76
3.69
2.71
2.59
2.62
2.59
2.34
2.32
2.29
2.41
2.86
2.82
2.46
2.50
3.32
3.37
1.54
2.52
2.91
2.61
2.83
2.69
2.16
2.31
2.43
which have not been
Mean of
Reps.
X
d
2.75
2.53
analyzed at this
-------
455
100J
90l
80J
70J
c
-------
4OCLT
35O.
3OO.
25O_
O)
O
u
15O_
1OCL
O_
\
influent
effluent
T—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r~
196O 61 62 63 64 65 66 67 6869 70 71 72 73 74 75
YEAR
1OO_
90L
80j
70.
60.
-------
1x1O3J
en
c.
E
o"
in
2
1x10 J
continuous
flow
*r _ O
100-f
c
-------
CO
1 —
Zi
<-L
1
r*^
O
-
"\
^^_
h-
LU
u
o:
Ld
DL
o.
10.
20.
30.
4O_
50.
60.
7O
/ ^>^f —
80_
90_
1OO
C
nt _ _ ~~ " " ~~ — ~ — ~ TT A r- r\ n t PO I
~~~~--— ^^^^ ""--^^ dual-media
Nx A^^^— ""^~%y
\ \ ' K-,x
\ V . u v
\ ^act. carbon ^
\
\
\
\
\
\
\
\
\
\
\
act sludge.
s\ \
i
\
\
*^ ** v
i i i i i i i i i i i i i i
) 2 4 6 8 10 12 14 16 18 20 22 24 26 28
TIME, DAYS
A
i
30
Response of Fathead minnows in continuous-flow exposure to control, activated
sludge, acitvated sludge-dual media, and acitvated sludge-dual media-acitvated
carbon test effluents.
-------
API - SOUR WATER STRIPPER STUDIES
Ronald G. Gantz
Supervising Process Engineer
Continental Oil Company, Ponca City, Oklahoma
ABSTRACT
Promulgation of national effluent guidelines and stringent new air pollution regulations
make the sour water stripper a critical unit in most refineries. Yet, many plants report
difficulty in achieving low ammonia levels . Others experience severe corrosion and
fouling in the overhead reflux system. In response to these problems, the API - Division
of Refining initiated in 1972 a program to improve stripping technology. First step in the
study was to circulate two comprehensive questionnaires throughout the industry.
Bench-scale studies conducted in 1974 by the Committee on Refinery Environmental
Control showed widely different stripping characteristics for sour waters from three
refineries. Despite excessive steam usage and a large number of theoretical trays,
ammonia concentrations could not be reduced below 520, 100, and 38 mg/l, respectively.
Analyses of the acidic bottoms revealed the presence of organic acids and amides. Adjust-
ment of the pH to 9.0 allowed further ammonia reduction, without fixing HLS in solution.
Similar stripping tests on synthetically prepared sour waters achieved 10-20 ppm NhL with
even modest steam rates.
Tray-by-tray calculations using Van Krevelen NhL-r-LS-r-LO equilibrium data showed
good correlation with actual performance on a full-scale sfripper down to the point that
only fixed ammonia remained in solution.
Also during 1974, the Technical Data Subcommittee conducted fundamental vapor-
liquid equilibrium studies to check and expand the range of the Van Krevelen data. In
addition, vapor pressure data were obtained for synthetic sour water containing HCN,
ethyl mercaptan, and cresol.
The Subcommittee on Refinery Corrosion conducted research using a dynamic corrosion
loop system to study metallurgy problems related to refluxed strippers.
INTRODUCTION AND PURPOSE
In 1972, the API Division of Refining initiated an extensive program to develop sour
water stripping technology. This effort was in response to concern expressed by member
companies regarding tough new environmental regulations. Initially, the study was con-
ducted by the Committee on Refinery Environmental Control (CREC) and the Subcommittee
on Corrosion of the Committee on Refinery Equipment (CRE). In 1974, the program was
expanded to include fundamental vapor-liquid equilibrium studies by the Committee on
Research Data, and Information Services (RDS). During the course of the studies, CREC
served as overall coordinator to ensure that the technology developed was sufficient to
meet environmental requirements.
459
-------
460
The purpose of this paper is primarily to discuss results of the CREC program to-date.
Simultaneously, the roles of CRE and RDS will be described and their programs briefly
mentioned. A detailed discussion of the CRE and RDS studies is impossible in this write-
up, since each easily constitutes an individual paper. All three committees are preparing
detailed reports concerning their work.
The CREC program first included a detailed survey of the industry to define the state
of the art. Then in 1974, numerous bench-scale tests and analytical studies were performed
to compare the stripping characteristics of actual and synthetically prepared sour waters.
Available vapor-liquid equilibrium data are based on such synthetic solutions.!') Results
of tests on a full-scale tower were then used for comparison with the bench-scale data and
tray-by-tray calculations using the procedure outlined by Beychok.^ '
BACKGROUND
Until recent years, sour water strippers were built to remove mainly hLS. Many units
were installed with a minimum of engineering attention. The sour off-gases were usually
burned in a flare or furnace.
New air and water pollution regulations have now made the stripper a critical unit
in most refineries. Careful attention must be given both to design and operation. Best
Practicable Control Technology Currently Available (BPCTCA) and Best Available
Technology Economically Achievable (BATEA) wastewater effluent guidelines^) promul-
gated by the Federal Environmental Protection Agency (EPA) severely limit NHL and hLS
discharges. Some states and localities have requirements that are even more stringent.
When expressed on a concentration basis, most refineries will have to meet an
effluent NH value by 1977 ranging between 2 and 20 mg/1. In 1981, this range must be
reduced to about 1-8 mg/1, if the BATEA guidelines are implemented and water usage is
not significantly reduced per barrel of crude processed.
To compound these problems, EPA is now considering cyanide for possible inclusion
on the list of toxic pollutants that must be controlled. The principal source of cyanides
in refinery waste waters is sour water from cracking operations. Very little is known about
the types of cyanides that occur nor to what degree they are removed during stripping.
Efforts in this regard have been hampered by lack of a reliable analytical method due to
sulfide interferences.
Today, air pollution regulations prohibit the direct burning and release of stripper
off-gases to the atmosphere in some localities. Recovery in a sulfur plant is usually
necessary in these cases. This requires refluxing of the stripping tower to minimize over-
head water and complicates control of the Claus unit due to the NH_ present. In addition,
contaminants in the sour water concentrate in the overhead system and cause numerous
corrosion problems.
-------
461
Ammonia is much more difficult to strip than HLS due to its strong affinity for water.
In years past, a 90 percent removal efficiency for rfls was satisfactory to meet most
state and federal effluent requirements. Consequently, only three or four equilibrium
stages were necessary in the stripper. However, current limitations sometimes require a
bottoms ammonia value below 50 mg/1 . This usually corresponds to 98 percent
removal of NH^. Consequently, the number of theoretical stages required is drastically
increased.
It should be emphasized that the calculation procedure outlined by Beychok, using
Van Krevelen data, will take the bottoms ammonia concentration down to essentially zero,
if enough stripping stages and steam are supplied. To the contrary, field experience
indicates that low ammonia values are difficult to achieve in many cases.
PROJECT CONCEPTION
Verbal reports at the semiannual meetings of the CREC in the early 1970's indicated
that some refineries, particularly those on the West Coast, have difficulty in achieving
bottoms NH~ values below 100-200 mg/1. The bottoms pH simultaneously drops to the
acidic side, although it should be alkaline if the system contains only NhL-hLS-hLO.
Other individuals reported that their plants could achieve 50 ppm NH- under similar
stripping conditions. This suggested the possibility that acidic components may be
present in some sour waters.
Stripping is dependent on the hydrolysis reaction(2):
NH.SH NHQ + H0S
4 J 2.
Small amounts of strong acid would shift the equilibrium to the left and prevent complete
stripping of NH_ from solution.
There were conflicting reports in both CREC and CRE regarding corrosion problems
with refluxed towers. These problems ranged from complete loss of the overhead
exchanger tube bundle in a few months to no noticeable corrosion after several years.
In order to clarify the conflicting stories and establish a better overall understanding
concerning stripper problems, both the CREC and CRE circulated detailed questionnaires
throughout the industry in mid-1972.
RESULTS OF QUESTIONNAIRES
A. CREC Questionnaire
The questionnaire circulated by CREC was intended to obtain information con-
cerning all aspects of sour water stripping. Sixty-three (63) plants participated
and supplied information on 73 units. Williams Brothers Waste Control Inc. collated
and evaluated the voluminous amount of data received. The information is available
as API Publication 927.
-------
462
2.5-3.5 feel". If was of interest to note that about" 75 percent of the plants
use trayed columns. All columns in the survey built since 1970 indicated
use of trays. Greater turndown capability is the most logical explanation,
especially where valve trays are utilized.
Data reported on phenol stripping showed that, for towers with 90 percent
NhL removal, refluxing reduced the average phenol removal from 55 percent
to 39 percent. Fourteen companies reported information on cyanide removal.
The data were very erratic but suggested an average removal of 37 percent.
This information should be viewed with caution since the available anolytical
procedures are subject to sulfide interference. Also, there was no breakdown
in regard to cyanide type (i.e., simple, complexed, etc.).
Questionnaire No. 34 supplied data fora more in-depth review of stripping
performance. The data are plotted in Figure I along with tray-by-tray
calculations using the procedure outlined by Beychok. If appears that
slightly less than five theoretical frays (i.e., 4.6-4.8) give a good fit
with measured data down to about 50 mg/1 . Below this point the measured
NHL concentration drops almost straight down and appears to bend a little
bacR to the left. If is very unlikely this could happen in a system with con-
stant feed conditions. It is suspected the feed characteristics changed or an
analytical error occurred. Additional data are needed for confirmation;
however, if is speculated that the curve normally bends to the right in a
manner illustrated by the dashed line. If 15 feet of 3-inch Raschig rings
in the tower equals 4.7 theoretical stages, the Height Equivalent to a
Theoretical Plate (HETP) equals 3.2 feet. Calculations using Fractionafion
Research Procedures W gave a predicted HETP of 4.5 feet. A more detailed
discussion of fray-by-fray calculations will be given later in this paper.
2. Corrosion
Accumulative data in the questionnaires indicated only minor corrosion
problems with the tower, frays, and feed-to-bottoms exchangers. Most of
the towers are fabricated out of carbon steel. Some cases of localized
corrosion were indicated due to lack of proper insulation around nozzles
and manhole covers. This encourages condensation at the metal surface and
promotes corrosion. It appears tower infernal linings and metals other than
carbon steel are seldom necessary. Corrosion on frays was also minor
regardless of whether carbon steel or stainless steel was utilized. Perhaps
stainless steel is warranted in the tope of the tower if the reflux is fed
directly back to the stripper. Very little corrosion was reported in the
shell or tubes of the bottoms exchanger although both are almost always
made of carbon steel.
As suspected, the major area of corrosion was the overhead system on
refluxed columns. However, the 18 replies to this questionnaire were
-------
463
The questionnaires were first classified by stripper hype as follows:
Refluxed Strippers 31
Nonrefluxed Strippers 24
Flue Gas Strippers 5
Fuel Gas Strippers 2
Oxidizing Strippers 4
Reject Questionnaires (insufficient
information) 7
73
Only the 55 replies concerning refluxed and nonrefluxed steam strippers need
be discussed. The others are limited in regard to meeting BPCTCA and BATEA
guidelines.
I. Stripping Efficiency
As expected, ammonia stripping efficiency varied widely between plants.
It was difficult to find common grounds for comparison due to the various
types and sizes of towers and contacting media utilized. Also, many towers
were operating at nonoptimum vapor and liquid loadings.
Perhaps the most important point observed was that seven refiners indicated
bottoms NHL values averaging 50 mg/1 and twenty plants reported 100
mg/1. As shown in Table I, an even larger number of plants indicated
achievement of these levels periodically. A review of the data indicated
that perhaps a few more plants could average below 50 and 100 mg/1,
respectively, if higher steam rates were used. Several questionnaires
reported only average bottoms data and did not indicate the minimum
attainable NH0 concentration. One should not conclude from the data
that refluxed units strip better than nonrefluxed units. The refluxed units
look better in the table simply because they are generally newer and
probably better designed. A detailed listing of the individual plants is
given in Table 2.
One surprising result was that none of the reported bottoms pH values
dropped below 7.0. However, as shown in Table 3, there were a few
cases where bottoms ammonia fixation could possibly have been a
problem. However, it must be emphasized that the data reported are
insufficient to reach a firm conclusion. The observation is merely made
that from the author's experience and a few calculations using Van
Krevelen data, the bottoms NH,, data reported appear high considering
the stripping ratios and tower internals reported.
The report concluded that the tray efficiency ranged between 30 and 50
percent in most cases and the HETP for 3-inch Raschig rings averaged
-------
464
inconclusive regarding the besf metallurgy. Experience with aluminum was
particularly erratic. In some cases poor performance appeared to be caused
by errosion due ro excessive velocities.
3. Miscellaneous
Table 4 gives the final disposition of the reported stripper overheads and
bottoms. As shown, the majority of plants still burn the sour off-gas and
release it to the atmosphere as SOO, NL, and H_0. This is usually done in
a flare, CO boiler, or furnace. However, the trend throughout the industry
is toward recovery in Claus sulfur units.
There were reports of hydrocarbon carry-over to the stripper causing problems
in the combustion furnace and sulfur units. This was particularly true for
hydrocarbon slugs entering the system. Very little information was obtained
about sulfur plant operations. Air cooling is reportedly utilized in about 50
percent of the overhead reflux systems.
Over one-half of the plants practice water conservation by using the
stripper bottoms for desalter make-up. Most towers have a feed-to-bottoms
exchanger. A typical heat transfer coefficient appears to be 100-150 Btu/
Hr/Sq Ft/°F-
Only three plants reported problems with excessive foaming; however, some
form of solids deposition was indicated in about one-half of the questionnaires.
Carbonate scale was experienced in several plants due to hard water
inadvertently entering the tower. Others reported phenolic and amine
based polymeric sludges. One plant reported a ferrocyanide deposit.
B. CRE Questionnaire
In May 1972, the CRE sent its questionnaire throughout the refining industry.
Eighty-seven (87) replies were received from 67 plants. This questionnaire con-
centrated solely on corrosion and is consequently much more detailed on this
subject than the CREC questionnaire. Results of the survey are available in API
Publication No. 944 prepared by Exxon Research. In addition to conventional
single-stage refluxed and nonrefluxed strippers, information is also tabulated on
two-stage units with pH adjustment before each stage.
Similar to the CREC study, the report concluded that the only area of severe
corrosion in refluxed and nonrefluxed columns is the overhead on refluxed systems.
In general, carbon steel and stainless steel tubes in the reflux condenser were
unsatisfactory. Experience with aluminum was mixed, primarily due to erosion
from excessive velocity.
-------
465
CREC 1974 STUDIES
A. Background and Objectives
During 1974, CREC conducted an extensive bench-scale testing program using
actual and synthetically prepared sour water solutions. This study was initiated
to help answer some of the questions unresolved from the CREC questionnaire.
The program had the following general objectives;
I. Determine if stripping characteristics of actual refinery sour waters are the
same as synthetically prepared solutions for which Van Krevelen vapor-
liquid equilibrium data are available.
2. Compare theoretical tray-by-tray stripping calculations to actual field
performance.
3. Outline methods of handling stripper off-gases to effectively achieve sulfur
recovery and ammonia destruction.
4. Prepare a written discussion of sour water stripper operating problems and
special design considerations.
A detailed report was completed by the contractor, Bechtel Corporation, in April
1975.
B. Laboratory Experimental Procedures
Laboratory stripping studies were made using the Oldershaw column apparatus
shown in Figure 2. The column is insulated by a vacuum jacket and the number of
plates (sieve trays) can be varied by combining sections with various numbers of
plates. One section each of 5, 10, and 30 plates was available for this work.
To start a run, the equipment was brought up to temperature and a level
attained in the reflux receiver before initiating feed to the column. After the
column had been operating a few minutes,the NH,, content of the bottoms was
measured periodically to determine when it attained a constant value, indicating
that the column had reached equilibrium.
All of the condensed overhead was returned to the column feed as reflux
ahead of the preheater. The reflux rate was controlled by holding a constant
level in the reflux receiver with a variable speed condensate return pump.
This pump, like those for the fresh feed and bottoms, was a positive displacement
peristaltic pump, the speed of which could be manually set.
The feed preheater was closely coupled to the top of the column and the
short connection insulated with aluminum foil and heated with a radiant lamp
to minimize temperature changes. Likewise, the vapor connection between
-------
466
the column and the condenser was similarly insulated and heated.
The overhead condenser was changed from time to time during the laboratory
investigation. Cooling water was used for the first seven runs. This overly cooled
the overhead and had to be discontinued. For runs 8-18, the reflux receiver was
kept warm and held at a constant temperature with a heating tape. To be sure
that the reflux was in equilibrium with vapor at the measured condenser temperature,
the heated, water jacketed, condenser shown in Figure 2 was utilized for the
remainder of the program.
Part of the steam generated in the reboiler of the stripper passed overhead
into the condenser as stripping steam, and part condensed in the column (mostly
on the top tray) to heat the downflowing liquid. The latter is heating steam and
its quantity depends on the temperature of the feed and reflux. Since most of the
heating steam travels up through nearly all of the column, it also does considerable
stripping. Flow of total steam was measured, and the flow of stripping steam was
calculated by subtracting the quantity of heating steam required.
C. Laboratory Test Program
For the bench-scale tests, actual sour water samples were collected from two
West Coast and one Gulf Coast refineries. These waters were labeled A, B, and
C, respectively, and stripping tests performed as quickly as possible in the
contractor's laboratory. Synthetic samples of NH^-H-S-H-O were then prepared
to simulate the actual sour waters. The synthetic samples were stripped in a
similar manner using the same bench-scale equipment and operating conditions.
I . Sa mp I e Ag i ng
Initial synthetic samples were prepared by dissolving a commercially available
solution of (NH J^S in distilled water (designated as Synthetic I). It was soon
discovered that rhe synthetic samples prepared in this manner could not be
stripped below 125 mg/1 NH,,. It was suspected that partial oxidation of the
stock solutions produced acid impurities which fixed NH~ in solution since the
bottoms pH dropped below 7.0.
Further experiments using commercial (NH .)~S solutions were abandoned,
and throughout the rest of the program syntnetic sour waters were prepared
by dissolving H~S gas in dilute NH ,OH (Synthetic II samples) immediately
before stripping. This is similar to van Krevelen's experiments.
Additional tests were run to determine if aging affected Synthetic II and actual
sour waters. A series of samples were taken from a large flask containing
Refinery B sour water. An air space was allowed above the liquid level in the
conta iner during the tests. Each individual sample was batch boiled and then
checked for residual ammonia. Distilled water was periodically added to
-------
467
replace any wafer lost during boiling. A similar series of tests were run on
Synthetic II sour water.
Figure 3 shows a plot of the residual NHL values after batch boiling. These
data show that sample aging is detrimental to high ammonia removal
efficiencies for both Synthetic II and actual sour waters if oxygen is allowed
into the system.
2. Stripping of Synthetic II Sour Waters
Synthetic II solutions stripped very easily in all tests. As shown in Figure 4,
residual ammonia levels of 10-15 ppm were attained using very low steam
rates. As further illustrated, the addition of several hundred ppm phenols
and 30-93 mg/1 cyanide failed to fix significant ammonia in solution.
3. Stripping of Actual Sour Waters
Bench-scale stripping of actual sour waters gave results considerably different
than when using Synthetic II sour waters. Figure 5 shows stripping data for
Refineries A and B. Figure 6 shows Refinery C. As illustrated, the bottom
of each curve appears to have flattened and very little additional NH,,
removal is possible despite a large steam usage. The fixed ammonia in the
bottoms appears to average about 520, 100, and 38 mg/1, for A, B, and C,
respectively. A comparison of these curves with Figure 4 indicates con-
siderably poorer stripping.
4. Characterization of Stripped Sour Waters
Since Refineries A and B had considerable fixed NH , a brief analytical
program was conducted in an attempt to identify contaminants present in the
stripped bottoms. A similar characterization was conducted on another
Mid-Continent sour water (Refinery D) which had a fixed bottoms NH,, of
approximately 40 mg/1 . The original samples were first analyzed for
total organic carbon (TOC) and metals content. Next they were boiled
to dryness and the residues subjected to infrared analyses.
As seen in Table 5, Refinery A had a TOC of 3,800 mg/1 but very little
inorganic carbon. The sample contained 1,580 mg/1 nonvolatile residue,
most of which would burn as evidenced by an ash content of only 80 mg/1 .
Refineries B and D were considerably different than Refinery A. Each had
TOC values of only 360 and 330 mg/1, respectively. Although Refinery
B had a high nonvolatile content similar to Refinery A, the residue was
largely inorganic as evidenced by the high ash content. This is verified
by high Na, Ca, Mg, and Si concentrations in the residue. Refinery D
had only 460 mg/1 nonvolatiles, which was much lower than either of the
other plants. An ash content was not run on the residue.
-------
468
Concentrates of samples from Refineries A and D were analyzed by both infra-
red and mass specfroscopy. The organic content of Refinery A was identified
as a complex mixture of low molecular weight amides and inorganic salts.
High concentrations of thiocyanate were identified in the infrared spectra.
Analysis of Refinery D disclosed no discernible organic functional groups.
Thiocyanates were evident. Attempts to identify specific compound types
were unsuccessful. The amides in the sample from Refinery A were
probably formed during evaporation, from ammonium salts of carboxylic
acids. Infrared spectra did not show the presence of sulfonic or naphthenic
acids.
5. Effects of Caustic Addition
Brief pH adjustment studies were conducted using continuous and batch boiling
apparatus. The continuous study results were erratic due to poor pH control.
However, the batch boiling results given in Table 6 clearly show that addition
of caustic allowed ammonia to be further stripped from solution. Since pH
adjustment studies were not in the original project scope and limited funds
remained, this program could not be pursued. Future studies are needed to
determine the optimum operating conditions and caustic injection point(s).
D. Full-Scale Stripping Test Program
Initially, full-scale studies were to follow the bench-scale program for
Refineries A, B, and C. Since Refinery A had such high fixed ammonia, it was
decided to abandon field studies in lieu of the sample characterization program
previously discussed . Due to similar reasons and limited funds, plans were
abandoned for Refinery B as well. Therefore, full-scale tests were conducted
only on Mid-Continent Refinery C. The objective was to provide a comparison
of field and bench-scale data plus theoretical tray-by-tray calculations.
I. Experimental Results
Figure 7 shows how NH~ in the bottoms varied for:
o
o the full-scale stripper at Refinery C.
o the bench-scale stripper using Refinery C sour water.
o the bench-scale stripper using Synthetic II sour water.
As shown, the full scale stripper using ten valve trays performed slightly
better than the small-scale unit using five sieve trays. However, the
stripping curve for Synthetic II sour water is quite different. A lower
I imit of residual NH- (i.e., 15 mg/1 versus 38 mg/1) is reached at high
steam rates. The synthetic sour water also obviously strips easier at low
steam rates.
-------
469
2. Tray-by-Tray Calculations
To provide a further comparison, the Van Krevelen equilibrium data were used
to perform stage-to-stage calculations to determine bottoms NH^asa function
of stripping steam rate. It was found that six theoretical stages (i.e., five
trays plus bottoms reboiler) gave results corresponding most nearly to the large
scale test data. Both the actual and calculated data are for refluxed systems.
The actual system contained ten valve trays, but was not reboiled. Results
are shown in Figure 8.
As might be expected, the calculated NH» concentration in the bottoms
continues to fall as the steam flow increases. When these results are compared
with the actual sour water, it is obvious that the calculations do not predict
the residual level of NHL actually measured once the calculated value drops
below 50 mg/1 . The correlation seems to be good above this point.
A quick review of Refineries A and B bench-scale tests indicated that cal-
culated bottoms ammonia values for these cases would also be inaccurate
below the fixed ammonia levels measured. Questionnaire No. 34 previously
discussed is a little harder to predict.
The data given indicate a fixed ammonia level probably exists much of the
time. If this is true, the same calculational error is predictable.
3. Tray Efficiency
To provide a further comparison, overall stage efficiencies were estimated
for large-scale Refinery C, small-scale Refinery C, and small-scale
Synthetic II. The efficiencies were obtained by running tray-by-tray
calculations for each column using a computer program based on the
Beychok procedure. Various numbers of theoretical trays were assumed
for each case until the assumed value gave a good curve fit with the
measured data. The efficiency was then obtained by dividing the number
of theoretical trays by the number of actual stages.
All three columns had an overhead reflux system. It was assumed that
equilibrium existed in each reflux receiver. Therefore, the term "tray
efficiency" applies only to the column internals and bottoms reboiler.
The full-scale unit contained ten actual stages and the bench-scale
system six (i.e., five trays plus bottoms reboiler). The calculated number
of theoretical stages and tray efficiencies are as follows:
-------
470
STAGE EFFICIENCY
Experimental
Series
Large Scale,
Refinery C
Small Scale,
Refinery C
Small Scale,
Synthetic II
* Excluding overhead
~^ ^ PI \/Q f rn \ /c r\ 1 i ic r\r\ f f ,
Estimated
Number of
Equilibrium
Stages*
6.3
5.4
9.0
condenser system.
t~i m c rarN<~» i 1 o r
Actual
Stages*
10
6**
6**
Tray
Efficiency
63%
90%
150%
Obviously the 63 percent tray efficiency for the full-scale tower indicates
the system is running wel! . The CREC questionnaire indicated most towers
are in the 30-50 percent range. A check of the tray liquid and vapor
loadings showed the tower to be operating in a range where efficiency
should be opl imum .
The efficiency of the smal' -scale unit on Refinery C water was even higher
(i.e., 90 percent versus 63 percent) than the efficiency of the large-scale
tower. This is not surprising since the small unit has tiny tray openings
and gives better contact between liquid and vapor phases. In addition,
it may have had significant wall effects due to the small column diameter.
The tray efficiency required to force a curve fit for the Synthetic II data
is 150 percent. The difference in performance between the two waters in
the same apparatus could not be explained. The vapor-liquid equilibrium
data used in the calculations were originally obtained by Van Krevelen
using similar synthetically prepared sour waters. This point definitely
needs more study.
4. Equilibrium Data at High Concentrations
At the high concentrations of HLS and NHL found in the reflux streams,
the equilibrium data, as extrapolated by Beychok, predict liquid concen-
trations much higher than those observed (see Table 7). This discrepancy
can be explained by two facts:
o recent equilibrium measurements by RDS at high partial pressures of
NhL and H_S indicate that ionic strength effects decrease solubilities
to about half of what the extrapolated Van Krevelen data predict.
-------
471
o reflux condensate temperatures measured in both the small and full-scale
strippers may not indicate the temperature at which liquid and vapor are
in equilibrium, if they actually do approach equilibrium.
In particular, the measured temperatures could be too low due to subcooling
of condensate.
E. Stripping Calculation Procedures
I. Beychok Procedure
Ultimately a calculational model is needed which takes into account all com-
ponents present in refinery sour waters. Such a model is not generally avail-
able throughout the industry today. Further research is needed before this is
possible. In the meantime, Beychok's calculation procedure can be successfully
applied in most instances if allowance is made fo.r the fixed ammonia .
Before initiating calculations, one should determine the fixed ammonia value
using bench-scale apparatus. Several samples should be taken over a period
of time to evaluate the effect of feed fluctuations. The fixed ammonia value
chosen should be subtracted from the design feed concentration and tray-by-
tray calculations conducted on the remainder. Then the fixed ammonia value
must be added back to the calculated bottoms value. If this level is tolerable,
a solution exists. If the value is too high, the steam rate and/or number of
stages can be increased. Once these parameters are optimized and the
bottoms NHL is still too high, pH adjustment must be considered.
O
2. Selection of Proper Bottoms NH3 Concentration
A word of caution should be given at this point concerning selection of the
proper stripper bottoms ammonia concentration to be used for design purposes.
Start with the pounds per day NhL allowed in the plant effluent and back
calculate the maximum concentration permissible in the stripper bottoms.
Remember that the average for the worst (i.e., highest) consecutive 30 days
must meet the permit conditions. Otherwise the permit may be met on a
long-term average basis but contain some 30-day periods above the prescribed
limits. Also, remember that although most effluent NH in refineries comes
from the sour water stripper, a significant amount may come from other
dilute sources. An accurate effluent ammonia balance is needed for the
entire plant. In addition, refineries with biological oxidation will need
some NH for nutrients in the biosystem. This generally averages about
4.4 pounds (5) of nitrogen per 100 pounds BOD<- removed. Ammonia utilized
in this manner will end up in the bacterial eel? mass or be converted to
nitrate. Very little ammonia removal is to be expected with conventional
surface and submerged aeration unless the feed temperature is well above
IOO°F. Biocultures should not be exposed to basin water temperatures much
above 105-110° F. (6)
-------
472
F. Stripper Design Considerations
The last two sections of the Bechtel report cover the important topics of off-
gas disposal and stripper design considerations. These sections prepared by
Milton R. Beychok summarize the state of the art and offer several interesting
observations. However, due to space limitations, it is not the author's intent to
discuss the numerous items. Instead, only three will be mentioned.
First, requirements for low bottoms ammonia concentrations will increase
steam stripping ratios unless more trays are utilized. Today's fuel shortage and
high prices present special incentives to minimize steam usage. Boosting the
stripping ratio one pound per gallon will require 288 M pounds per day
additional steam for a 200-gpm unit. With an 80 percent boiler efficiency and
a fuel with 5.9 MM Btu per barrel this amounts to 22,300 barrels of fuel per
year. At $8 per barrel the added operating cost, considering only the steam,
is $178,200. In addition, higher stripping ratios boost utility requirements on the
overhead reflux condenser and increase the flow rate of water in the tower
bottoms. From this illustration, it is obvious that additional trays are easily
justified when building a new tower. In some instances, refiners may even find
it advantageous to replace existing columns. On an industry-wide basis, proper
tower design can save considerable fuel each year.
Secondly, turndown capability should be given special consideration during
tower design . Sour waters come from several areas of the plant. At any given
moment, all or part of these streams will be feeding the stripper. When some
refinery units are down, such as during turnaround, it is necessary to continue
efficient stripping of the remaining sour waters. Valve trays probably offer the
best solution to this problem since turndown ratios of 3:1 are possible. Feed
storage can also be considered if trace amounts of oxygen can be kept out of
the system.
Finally, verbal reports from refiners confirm satisfactory recovery of
stripper off-gases in Claus sulfur plants. Several plans indicated special burner
designs are needed to ensure complete combustion of the ammonia. However,
most plants would not discuss details of burner design due to proprietary reasons.
Likewise, very little information could be obtained regarding system instrumen-
tation and controls.
RDS 1974 STUDIES
During 1974 the Committee on Research Data and Information Services (RDS)
sponsored an active program at Brigham Young University to obtain new vapor-liquid
equilibrium data for refinery sour waters. The objective was to check some of the Van
Krevelen NH^-H«S-H«O data plus obtain new data points throughout the concentration
and temperature ranges anticipated in well-designed refluxed units. In addition, some
data were obtained to determine the effect of other components such as ethyl mercaptan,
-------
473
hydrogen cyanide, and cresol. Results of this study are to be presented in a written
report to API about midyear 1975.
CRE 1974 STUDIES
During 1974, Battelle Memorial Institute conducted corrosion studies on refluxed
sour water strippers under the sponsorship of the Corrosion Subcommittee of the Committee
on Refinery Equipment. Laboratory studies were to explor the effects of the following
variables:
I. Velocity
2. Temperature and pressure
3. Phase (liquid, vapor, condensing, impingement)
4. Composition (H2S, NH3, Cl~ in HO)
5. Contaminants (cyanides, phenol, acids, caustics, etc.)
6. Materials (metals and coatings)
7. Inhibitors
A detailed report of the results will be submitted by Battelfe to CRE in the fall of
1975.
FUTURE API STUDIES
API will continue a program of research on sour water strippers throughout 1975.
Each of the three committees in 1974 experienced unexpected difficulties with portions
of their programs. Attempts will be made to resolve these problems in 1975 and also
explore new areas. CREC will concentrate on determination of proper methods of caustic
injection, cyanide removal, plus further identification of components causing ammonia
fixation. RDSwill continue its program of expanding available equilibrium data. In
addition, data will be gathered on systems containing CO~ to assist in technology
related to stripping sour water from synthetic fuel processes. CRE will continue its dif-
ficult task of defining corrosion mechanisms and potential corrective measures for
refluxed strippers.
SUMMARY AND CO NCLUSIO NS
To date, the following conclusions have been reached as a result of the API Sour
Water Stripper Studies.
I. Refinery sour waters frequently contain "fixed" ammonia which cannot be
stripped out of solution unless the pH is adjusted. The amount of "fixed"
ammonia varies considerably between plants and can be quite high.
2. Additional work is necessary to determine the best point(s) for caustic
addition and the optimum bottoms pH.
-------
474
3. ConHnuous flow bench-scale or batch boiling techniques can be successfully
used to determine the amount of "fixed" ammonia in solution.
4. "Fixed" ammonia is due to the presence of acidic impurities. The Beychok
calculation model does not take these additional components into consideration.
5. Phenols and cyanide do not appear to explain observed ammonia "fixation."
6. Exposure of sour waters to trace quantities of oxygen can slowly create
oxidation products and cause some ammonia "fixation. "
7. The Beychok calculation model appears to give satisfactory correlation with
measured bench-scale and full-scale results down to the point the "fixed"
ammonia value is approached.
8. Measured H^S and NH concentrations in the reflux drum are considerably
less than predicted by extrapolation of the Van Krevelen data.
9. Development of a new multicomponent stripper calculation procedure is
desirable with capability of handling external pH adjustments. This would
take considerable new research effort.
10. The overhead system on refluxed strippers is subject to severe corrosion in
many cases. Continued research is needed to fully understand the mechanisms
and determine the best corrective measures.
REFERENCES
(I) Van Krevelen, D. W=, Hoftijzer, P.J., and Huntjens, F. J., "Composition and
Vapor Pressures of Aqueous Solutions of Ammonia, Carbon Dioxide, and Hydrogen
Sulfide," Rec. Trav. Chim. Pays-Bas 68, 191-216, (1949).
(2) Beychok, M. R., "Sour Water Strippers," Aqueous Wastes from Petroleum and
Petrochemical Plants 158-198, John Wiley and Sons, London, (1967).
(3) "Petroleum Refining Effluent Guidelines and Standards," Federal Register, Vol. 39,
[91], 16560-16575, (May 9, 1974).
"Petroleum Refining Effluent Guidelines and Standards," Federal Register, Vol. 39,
[202], 37069-37071, (October 17, 1974).
(4) "Packing," Fractionation Tray Design, Vol. 2, Section 8.3, Page I; Section 8.4,
Page 3; Fractionation Research Inc., Alhambra, California, (1968).
(5) Eckenfelder, W. W., Jr., "Principles of Biological Oxidation," Industrial Water
Pollution Control, 146-147, McGraw-Hill, New York (1966).
-------
475
(6) "Biological Treatment," Manual on Disposal of Refinery Wastes - Volume on Liquid
Wastes, 1st edn., 13-9, American Petroleum Institute, New York, 1969.
DISCUSSION
Milt Beychok: I have two comments to add concerning unknown acids found in the
Bechtel work. First, I would like to emphasize that these acids affected only the ammonia.
For those of you who are designing strippers and worrying about HLS removal, there is no
problem with the H«S stripping. The acids fix only the basic ammonia. The second point
was on Grant Wilson's work at Brigham Young University. His equations are theoretically
more exact and he has much more data and better data than Van Krevelen did some 30
years ago. But in the region of most sour water strippers as compared to fractionators
(that is where the molar ratio of ammonia to hLS is 1-1/2 or higher) there doesn't appear
to be very much difference at the usual concentrations levels of the 2-3,000 parts per
million. When you get up to the 60,000 parts per million range, in the reflux accumu-
lators Grant's data is much much better than Van Krevelen's. So I guess what I am really
trying to say is that in the ordinary sour water stripper application other than very low
ammonia, Van Krevelen's data is still usable because it is so much more simple than
Grant's data which will require a computer to use.
Bob Farnham: Did you look at chlorides in the stripper bottoms as a source of acidic
material ?
Ronald Gantz: As I recall we took very little data on chlorides.
Bob Farnham: I think people tend to ignore the fact that chlorides do get into refinery
sour waters via such things as leaded gasoline that finds its way into processing units,
chloride addition for catalyst activity, and materials such as neutralizing amines .
I think chlorides could be one of the problems with the fixed ammonia. Another
problem may be fixation of the ammonia due to exposure to air. Does that have anything
to do with CO2?
Ronald Gantz: Yes that could be a problem if air is allowed into the system. We have
very little CO data but obviously if ammonium carbonate were present you would have
some ammonia Tixation. However, most refinery sour water collection systems are
blanketed with inert gas to prevent the entry of air.
-------
476
BIOGRAPHY
R. G. Gantz is a Supervising Process Engineer
in Continental Oil Company's corporate Process
Engineering Department in Ponca City, Oklahoma.
He currently directs activities of the Environmental
Conservation Group which is involved with all
phases of environmental engineering. Experience
includes nine (9) years of assignments on a very
wide variety of projects. Current responsibilities
include preliminary and detailed process design
work, pilot and bench-scale testing, plant
technical services, stream surveys, and in-company
consulting. Work areas include waste water treat-
ment, air pollution abatement, and solid waste
disposal for Conoco's Refining, Chemicals, and
Natural Gas Departments. He is a member of
API's Committee on Refinery Environmental
Control and author of several technical public-
ations. Mr. Gantz holds a B.S. degree in
Chemical Engineering from Oklahoma State
University. Prior experience includes product
development work in DACRON textile fibers for
the DuPont Company.
-------
TABLE 1 - CREC - QEUSTIONNAIRE DATA ON NH, STRIPPING
\3
BOTTOMS NH3 DATA
477
Number Refluxed
Towers Reporting
Number Nonrefluxed
Towers Reporting
Totals
Total
31
24
55
Averaging
50 mg/1
3
7
Periodically
Achieving
50 mg/1*
5
14
Averaging
100 mg/1
13
_7
20
Periodically
Achieving
100 mg/1*
15
23
*Several questionnaires reported only average data. Numbers shown would probably be larger if minimum
NHQ had been reported in all cases.
o
TABLE 2 - SOUR WATER STRIPPERS WITH HIGH NH, REMOVAL
o
Questionnaire No.
3 -Refluxed
I3B -Refluxed
22A- Refluxed
22B-Refluxed
22C-Refluxed
23 -Refluxed
25 -Refluxed
26A-Refluxed
26B-Refluxed
27 -Refluxed
34 ^Refluxed
36 -Refluxed
43 -Refluxed
44 -Refluxed
55 -Refluxed
7 -Nonrefluxed
10 -Nonrefluxed
32 -Nonrefluxed
33 -Nonrefluxed
52 -Nonrefluxed
53 -Nonrefluxed
54A-Nonre fluxed
58 -Nonrefluxed
Tower Media
10 Valve Trays
8 Valve Trays
30 Sieve Trays
30 Sieve Trays
24 Sieve Trays
23 Sieve Trays
52 Valve Trays
5 Glitch Trays
20'-3" Raschig Rings
10 Flex Trays
15'-3" Raschig Rings
18 Trays
20 Bubble Cap Trays
12 Socony Trays
20 Sieve Trays
8 Bubble Cap Trays
6 Shower Trays
!5'-3" Raschig Rings
15'-3" Saddles
8 Valve Trays
28 Bubble Cap Trays
5 Valve Trays
8 Flex Trays
Average
Stripping
Steam
Pound/Gal.
Fresh Feed
1.4
1.0
I.I
2.5
0.6
1.6
-
7.8
1.3
1.8
1.5
-
-
1.2
-
1.5
0.3
0.6
0.2
1.9
0.8
0.4
2.7
Average
Feed
2,500
1,200
1,720
430
74
4,000
1,600
5,410
3,550
2,000
1,400
19,000
2,000
32,200
1,600
960
1,850
1,200
2,600
5,450
2,625
215
4,400
Average
Bottoms
78
25
68
64
63
100
65
45
-
200
80
80
15
56
25
50
96
65
200
56
10
76
I!
Minimum
Bottoms
25
-
-
-
_
40
-
19
37
25
7
-
10
-
7
30
-
36
34
-
-
-
10
Bottoms
PH
9.4
-
-
_
_
-
9.2
8.4
9.3
9.0
-
-
7.3
8.5
9.0
9.4
-
-
8.6
8.6
-
9.3
8.3
-------
478
TABLE 3 - SOUR WATER STRIPPERS WITH POSSIBLE BOTTOMS NH., FIXATION
Questionnaire No.
22A-Re fluxed
22B-Refluxed
37A-Refluxed
41 -Refluxed
18 -Nonrefluxed
(a)2l -Nonrefluxed
47 -Nonrefluxed
(a)48 -Nonrefluxed
(a) 2-Stage Unit
Tower Media
30 Sieve Trays
30 Sieve Trays
35'-2" Raschig Rings
16 Valve Trays
l6'-3" Saddles
9 Bubble Cap Trays
15 Ballast Trays
19 Bubble Cap Trays
20'-3" Raschig Rings
12 V-Grid Trays
Average
Stripping
Steam
Pound/Gal.
Fresh Feed
I.I
2.5
0.8
0.9
3.9
0.07
0.5
1.9
0.4
0.9
Average
Feed
1,720
430
1,400
1,400
4.460
2,800
-
1,000
2,500
-
Average
Bottoms
68
64
600
400
265
300
115
-
115
Minimum
Bottoms
_
-
-
-
150
-
200
-
-
-
Bottoms
PH
_
_
-
9.5
_
-
-
8.0
-
_
TABLE 4 - DISPOSITION OF OVERHEAD GASES AND STRIPPER BOTTOMS
Overhead Disposition
Incinerated in Some Manner
Sulfur Plant
Vented to Atmosphere
Scrubbers
Gas Plants
Bottoms Disposition
Desalter
Sewer
Biosystem
Cooling Towers
FCC Units
Oxidizers
NUMBER OF REPLIES
Refluxed Nonrefluxed
19
10
0
2
_0_
31
14
10
4
1
1
J_
31
19
2
6
0
J_
28*
24
*Four units were 2-stage.
-------
479
Sample Source
Refinery A
Refinery B
TABLE 5 - CHARACTERIZATION OF STRIPPER BOTTOMS SAMPLES
CONCENTRATION (mg/1) BASED ON ORIGINAL SAMPLE
Sample Source
Refinery A
Refinery B
Refinery D
Total
Organic
Carbon
3,800
360
330
Inorganic
Carbon
14
2
10
Thiocyanate
560
114
NonvolaKles*
1,580
1,230
460
Ash**
80
850
Na
10
100
EMISSION SPEC SEMI-QUANT (mg/1)
Fe Ca MŁ
O.I
10
1
0.8
10
4
5
Al
0.1
0.8
* Nonvolatiles are defined as the total materials (residue) remaining after evaporating samples to dryness.
** Material left after burning the residue.
TABLE 6 - EFFECT OF CAUSTIC ADDITION ON FIXED AMMONIA
Run
B-l
B-2
B-3
Sample
A-Raw Refinery A
B-Raw Refinery B
Refinery B
No Caustic
Product of B-l
Product of B-2
PH
Initial
PH
Final
BEFORE BOILING**
8.6
AFTER BOILING
8.3
8.3"
9.0*
9.0*
AFTER BOILING
6795
6.70
8.25
**
***
Duration
of Test***
A-l
A -2
A -3
A -4
A -5
A -6
Refinery A
No Caustic
Product of A-l
Product of A-2
Product of A-3
Product of A-4
Product of A-5
8.6 6.2
9.0* 7.35
9.0* 7.55
9.0* 8.1
9.0* 7.1
9.0* 7.5
BEFORE BOILING
4 Hour:
1 Hour
1 Hour
1 Hour
1 Hour
1 Hour
4 Hours
1 Hour
1 Hour
pH adjusted to 9.0 by using caustic.
VAII experiments conducted by batch boiling.
*Time considered to be in excess of that required to reach NH3 residual level,
Ammonia
4,960
560
280
120
62
42
28
3,370
115
70
5
-------
480
TABLE 7 - COMPARISON OF CALCULATED AND OBSERVED REFLUX
COMPOSITIONS REFINERY C
Total Steam, REFLUX NH3 , mg/1 REFLUX _NH3, mg/1
Pound/Gallon
Fresh Feed Actual Calculated* Actual Calculated*
1.20 32,600 65,800 16,300 47,400
1.41 36,300 90,800 30,600 73,300
1.58 42,600 102,000 28,600 83,100
1.76 26,300 61,900 17,700 42,100
1.90 30,300 73,300 8,600 54,900
2.23 10,000 78,200 9,200 58,500
2.87 52,300 103,200 12,500 86,500
*Actual operating conditions were used as a basis for all calculations.
-------
481
TRAY-BY-TRAY CALCULATION
4 THEORETICAL TRAYS
TRAY-BY-TRAY CALCULATION
5 THEORETICAL TRAYS
OLDERSHAW
PERFORATED PLATE
COLUMN I'l.D.
CAUSTIC INJECTION
OH VAPOR
CONSTANT
TEMPERATURE
JACKETED REFLUX
CONDENSER
REFLUX
RECEIVER
FRESH FEED
PUMP
HEATING
MANTLE
(CONTROLLED BY
VARIABLE
TRANSFORMER)
REFLUX
PUMP
1.0 2.0 3.0 4.0
STRIPPING STEAM, LB/GAL FRESH FEED
5.0
FEED
RESERVOIR
FIGURE I. COMPARISON OF TRAY'BY" TRAY CALCULATIONS
WITH QUESTIONNAIRE NO 3'. STRIPPING CURVE
FIGURE 2.
CONTINUOUS BENCH SCALE
SOUR WATER STRIPPING APPARATUS
20
SYNTHETIC H
SOUR WATER
2600 MG/L NHj
IN FEED
4 •>
AGING TIME, DAYS
FIGURE 3. AGING EFFECTS ON SOUR WATER STRIPPING
220
200
i eo
! 60
_j
(3
a i 10
ro
CO
2
V-
5
" nn
o -
f
)
\
\
T v
V
V
o
O 1
n i
Q 1
i TRAYS
3 TRAYS
3 TRAYS
3 TRAYS
O 10 TRAYS
\
\
~~D
1 — o
00
n
NO CYANIDE OR PHENOLS
NO CYANIDE OR PHENOLS
125 MG/L PHENOLS
30 MG/I. CYANIDE
550 MG/L PHENOLS
93 MG/L CYANIDE
93 MG/L CYANIDE
04 08 1.2 16 20
STRIPPING STEAM, LB/GAL FRESH FEED
NOTES -COLUMN HAD BOTTOMS REBOILER AND OVERHEAD CONDENSER
-THE FEED CONCENTRATION OF NHj VARIED BETWEEN
2600 AND 5600 MG/L
FIGURE 4 BENCH SCALE STRIPPING OF SYNTHETIC I SOUR WATER
-------
482
2000
O REFINERY A— 50-35
TRAYS , > 3200 MG/L NH,
IN FEED *
REFINERY B— 40 TRAYS;
14600 MG/L NH3 IN FEED
O
0 1.0 2.0 3.0 4.0 50
STRIPPING STEAM, LB/SAL. FRESH FEED
NOTE' COLUMN HAD BOTTOMS REBOILER AND OVERHEAD REFLUX
CONDENSER.
FIGURE 5. BENCH SCALE STRIPPING RESULTS REFINERIES A AND B
2000
400
1 00
3
\
\
\
\
0 \
3
A
\0
N
COLUMN CONTAINED 5
FEED CONTAINED 260C
"-
TRAYS
MG/L NH3-
I .0 2.0 3.0 4.0 5.0
STRIPPING STEAM, LB/GAL FRESH FEED
NOTE i COLUMN HAD BOTTOMS REBOILER AND OVERHEAD REFLUX
CONDENSER
FIGURE 6. BENCH SCALE STRIPPING RESULTS REFINERY C
800
400
60
|
\
\
\
\ °
H\
\
\
v
\
\
i
0 1
o
\
"\-
\N
^^s
^- *-
D FULL
REFINE
REFLU)
REFINE
A BENCH
5 TRAY
CONDE
n^.
— eP=o— = =
•^
SCALE TEST
RY C, 10 TF
C CONDENSE
SCALE TES
RY C, 5 TRA
CONDENSER
SCALE TES1
:TIC n SOUF
S PLUS REf
NSER t RE
h-O
RUN,
(AYS PLUS
^
T RUN,
rs PLUS
(. REBOILER
RUN ,
LUX
30ILER
0 2.0 3.0 4.0 5.
I
STRIPPING STEAM, LB/GAL FRESH FEED
NOTE' APPROXIMATELY 2600 MG/L NHj IN FEED
FIGURE 7 SOUR WATER STRIPPING CHARACTERISTICS - REFINERY C
FULL SCALE TEST RUN , REFINERY
C,IO ACTUAL TRAYS PLUS REFLUX
CONDENSER
BENCH SCALE TEST RUN, REFINERY
C, 5 ACTUAL TRAYS PLUS REFLUX
CONDENSER AND REBOILER
TRAY-BY-TRAY STRIPPING CALCULA-
TIONS, 6 THEORETICAL TRAYS AND
REFLUX CONDENSER
1.0 Z.O 3.0 4.0
STRIPPING STEAM, LB/GAL. FRESH FEED
NOTE' APPROXIMATELY E600 MG/L NH3 IN FEED
FIGURE 6. COMPARISON OF TRAY-BY-TRAY CALCULATIONS
WITH REFINERY C STRIPPING CURVES
-------
SESSION VII
"FUTURE RESEARCH"
Chairman
Peter B. Lederman
Director, Industrial and Extractive Processes
U.S. EPA, Washington, D. C.
Speakers
Wilson K. Talley
"The EPA's Role in Future Research"
Arne E. Gubrud
"The Industry's Role"
Joseph F. Molina
"The University's Role in Future Research"
483
-------
484
BIOGRAPHY
Peter B. Lederman
Peter B. Lederman has just joined Research-Cottrell as
Manager of Technical Development, where he is responsible
for the Product Improvement, Warranty Service, and Gas
Dynamics Departments. Prior to his current position, he
spent four years with the United States Environmental
Protection Agency as Director of Industrial and Extractive
Processes. In that position, he was responsible at various
times for planning, coordinating and implementing EPA's
R&D program to develop improved technology for minimizing
pollution from industrial and mining sources. Dr. Lederman
holds B.S., M.S., and Ph.D. degrees in Chemical
Engineering from the University of Michigan and is a
Professional Engineer. He has published widelyjand has
served on the faculties of the University of Michigan and the
Polytechnic Institute of Brooklyn, as well as having
industrial experience with Exxon, Shell and General Foods.
-------
"THE ERA'S ROLE IN FUTURE RESEARCH"
Wilson K. Talley
Assistant Administrator, Research and Development
U.S. EPA, Washington, D.C.
I would like to talk about future research, more specifically, EPA's role. Let me
preface my remarks by saying that it is my personal belief that the Nation's development
of adequate, and even abundant, energy resources can be done in consonance with the
protection as well as the enhancement of environmental quality. I think that that is a
particularly appropriate thought to carry to this meeting because we are going to need
more petroleum refineries in the future, not fewer, and they must be clean.
Before talking about any thrusts in research priorities, let me explain how the
research arm of EPA functions. The first thing to recall is that EPA, when it was put to-
gether, was structured to be a regulatory agency. Its primary mission was not to be
research and development. I would like to stress "primary," for we do need research and
development, that is, we need the scientific and technological fruits of such labors. We
do do research, but otrr rhrust in the research arm is to support the rest of the Agency.
Ignoring for the moment whether standards and regulations not based on science and
technology could stand legal tests, I don't believe that such arbitrary restrictions would
serve the national interest. If they are too loose, they wouldn't protect the environment
and such laxity is just no longer a publicly acceptable mode of operation. If they are too
restrictive, they could cripple our struggle for new energy resources and this, by the way,
might resound ultimately to the detriment of the environment itself: energy supplies,
economic stability and growth, and environmental quality are now recognized as being
inextricably intertwined. On the other hand, regulations and standards that are based on
firm information as to health and ecological effects and on the existence of proven,
adequate and effective technologies will prove, I think, acceptable to industry and to
public alike. The question is how to develop that scientific and technological base.
I have been up on the Hill quite frequently in the last few months, engaged in dis-
cussions with Congress and with my fellow agencies on the proper structuring of research
and development for regulatory agencies. One suggestion has been to "free" R&D from
the regulatory mode. The examples that are given for this are our good friends in OSHA
and NIOSH. The Occupational Safety and Health Administration is in the Department of
Labor. It is supposed to be served by the research performed by the National Institute for
Occupational Safety and Health in the Department of Health, Education and Welfare.
There are some advantages to this separation, the most notable being the freedom to allow
a research program to follow its own internal logic. There are also disadvantages, and I
would like to mention just two, timing and feedback. If an R&D program yields its results
months after the regulatory arm has had to make a decision, the late information may be
worse than useless. The best you can hope for is confusion and loss of public confidence.
And by feedback, I mean that the researchers, if they are separate, don't live with the
same tempo as the regulators. The feeling of urgency in the researchers to solve actual
problems, rather than interesting ones, isn't developed. If you have a more intimate
contact, the problems that become interesting to the researchers are those that are vital
485
-------
486
to the regulators. If they are solved, they will win for the researcher the gratitude and
approbation of the regulators. The regulators become the first constituency of trte-
researchers, and their fellow scientists and engineers a secondary constituency - an
important one for peer review, but still secondary. Another extreme, which I haven't
heard raised outside the Agency, is that all research should be directed, controlled,
conducted - use any term you like - by the regulatory arm. Leaving aside the possibility
that the rest of the world might view such research results as biased, there is a more
dangerous and subtle disadvantage. Granted that we would have'the tempo of the
research geared absolutely to the temp of the regulatory program, the draw back is then
that the researchers will adopt the same sense of time as held by the regulators. This is a
disadvantage because regulators live in the present; they move from court-set dates to
congressionally mandated time frames. The researchers, if they adopt this attitude, are
likely to bury their noses in the present, with all of their time occupied by solving the
immediate, the obvious problems. That will occupy all their time — Parkinson's law
codifies this -- the work will expand to take all their time and energy. Who then will
look ahead to eliminate future problems, perhaps before they are even generally recog-
nized ?
EPA operates in a middle course. The Office of Research and Development is an
independent Office within the Agency. As Assistant Administrator for Research and
Development, I am co-equal with the Assistant Administrator for Air, the AA for Water,
the AA for Enforcement and the like. In the Office of Research and Development, we set
our own priorities, but we do it cheek to jowl with the rest of the Agency. If we more too
far from their perception of their problems, they let us know it in no uncertain terms. We
do back-stop them. Our research results may be the subject of litigation, and we have to
be prepared to back them up as expert witnesses. This is in addition to facing the peer
review as any other researcher does. Yet if all we did was to perform according to their
needs, we would be failing the opportunity that we are given because we are an indepen-
dent office. Edmund Burke said something to the effect that a representative owes his
constituency more than the diligence of his labors on their behalf; he also owes them his
judgment. I interpret this to mean that we owe the program offices the satisfaction of
their perceived needs, but, if we see something that we feel is more important to them,
we have the obligation to refuse to respond absolutely to their requests. We must reserve
the right to act on our best judgment. That means we have to respond to them, but in a
way to husband our resources so we can allow some work to anticipate what is going to
happen. Another term is that we must perform "problem definition" research. If we can't
spot a problem in the future and eliminate it so it never arises, at least we can flag it so
that it doesn't arrive as an unheralded disaster. We know that it is coming and we can
get ready for it. Now let me admit that all is not skittles and cream. If ORD does its job,
there are but two options — either we do our job well or we don't do it well. If we do it
well, and in the main we do, we will produce solutions before the problems become
critical or, better, we will eliminate problems in advance. In the main, that is what we
do do. We are then hard pressed to justify our share of the Agency resources, because
what have we done that is important to the Agency? On the other hand, if we fail to
supply the technology, or cannot guarantee criteria based on unassailable dose-response
functions, or the like, then we haven't helped the Agency, we're vilified and the result
is the same: we are hard pressed to justify our resources. As I keep telling my people,
-------
487
the best we can hope for is that we will be treated with, perhaps, benign neglect. Des-
pite this, there is one advantage, a very important advantage, to our place in the
organizational structure. We cannot be ignored when a decision is being made,
especially one that has to be based upon science and technology. Were we totally
removed, as NIOSH is from OSHA, then the regulatory side of tbe house could pick and
choose selectively that information that supports their preconceived notions of what their
decision should be. If we were completely captive to the regulatory side of the house,
then we would have no voice at all.
Given the present organizational structure, how do we get the research done?
Philosophically, I can see absolutely no reason to replicate with federal funds and people,
resources that already exist in the private sector, in industry, and universities. It is a
good thing that I have this attitude, because we don't have anywhere near the resources
to even attempt that replication. We conduct a largely extramural program, out of
necessity. Even if it were not, I would do it out of choice. We need and welcome
participation from industry and other elements of the private sector. Roughly speaking,
our budget is about $250 million. Of this, about $70 million is spent inhouse to support
our own research staff. Another roughly $70 million is passed in interagency transfers to
other federal agencies. The rest is let by contract and grant to industry, nonprofits,
universities, and State and local governments.
Let me pause to mention our involvement with industry. There are some people who
do not think it is a good thing for a regulatory agency to have part of its research con-
ducted by people who may turn out to be part of the elements that are regulated. I don't
subscribe to that. In the eleven years that I have been speaking and working in the
environmental arena, I have yet to meet anyone whose primary purpose in life was to be
a polluter. There aren't any all-black hats in this business. And there aren't any all-
white hats, either. No one is pure; we all wear varying shades of grey. Given the
opportunity to conduct his business or to live his life in a way that will achieve his basic
goals, but will also protect the environment, I believe that anyone will opt for that
choice. The remarkable thing is that industry and the public in general are willing to
pay a penalty, to give up something in order to protect the environment.
I only require one thing when we go cooperative with an industry in a research pro-
gram. There must be no suppression of data by either side. When research is done, it's
got to be widely disseminated so that there is the possibility of examination by the rest of
the research community. It is as simple as that.
Being debated by the National Academy of Sciences, the Office of Technology
Assessment, the Congress, and the Agency is the subject of setting research priorities.
Sometimes neglected is the problem of communicating those priorities to all interested
parties, including not only the people who are going to receive the results of the research,
but those who will take part in conducting the research, as well. If one can perform that
communication and can allow feedback on the ordering of these priorities, the result may
produce a better integrated and a more responsive R&D program. Success is not
guaranteed, but if you set your priorities in splendid isolation, there is little chance that
you are going to wind up with a research program that is acceptable in its performance or
-------
488
acceptable in terms of delivery of the results. Let's begin my admitting that the budget
is the ultimate arbiter of priorities. If you say that project X is your No. 1 priority but
give it zero dollars, it is not your No. 1 priority. But the budget is not the sole measure
of priorities. If all of our projects took less than a year to complete and if the develop-
ment of the budget took long enough and was conducted in a way to allow enough inter-
action to take place, then the budget would serve as the appropriate vehicle to order our
priorities. Unfortunately, our projects average longer than a year, and the "windows"
available for observation while the budget is being prepared close very rapidly. Simply,
I do not consider the budget an appropriate priority setting mechanism.
In the past the budget system has been the defacto system for setting priorities. How-
ever, this year - in fact in about 2 weeks - we are going to publish our first attempt at a
five year plan: The Agency Research Statement or the ARS. While by no means perfect,
it is the first and we hope it is perfectable. Each year we are going to produce a new
ARS. It is a rolling plan, what is now the fifth year will become the fourth year, in the
next version it will be the third year, the next the second, finally it will become the
current year and guide the preparation of the budget. One sees that the "current year"
will have been examined four times as it comes forward. While we will never reach the
fifth year unchanged, and I suspect that you can't plan five years in the future, the
decisions we make today will be based on our looking as far forward as we can. At least
our decisions will have input other than the grim necessities and the urgencies of today.
All comments on the ARS are going to be welcome. Our object is to let the ARS
become as detailed as possible, specifying as milestones the outputs of the research pro-
gram: the years they will be available; how much it is going to cost to produce them.
This has to be done with as much external and internal participation in the process as
possible. I should add also that the ARS is our annual report to Congress. Another feature
of the planning system is our "report card" to the Agency. Each year we have solicited
satements of needs from the rest of the Agency. We have asked them what they would like
in the way of research output, what problems they have. In the past that has been a one
way street; the needs come in and sometimes they hear what has happened to them, more
frequently not. From now on, when we send out the request for needs we are also going to
say what we accomplished as contrasted to what we intended to do. In the subsequent
years, I am going to try to get us to explain why we didn't solve or even attack some
problems.
It is part of our obligation to try to explain in the ARS the rational for the research
and development thrusts. For instance, I would like a debate on end-of-pipe pollution
abatement technology versus process modification. My inclination is toward the latter.
Simply put, as an engineer, I view end-of-pipe technology as a poor engineering solution.
Since most of the waste stream is either the raw material — that the manufacturer paid
for -- or it is the finished product — that he would like to sell on the market -- process
modification that reduces that waste stream has got to be more profitable. (It's a return to
the old conservation ethic — use less, use it more efficiently — and you will waste less.
In these times of energy crises and material shortfalls, I think it is a good ethic to return
to.) But I may be wrong, end-of-pipe has some advantages, and in some instances it may
be the method of choice. Also, in some industries, the search for process modification
-------
489
may be fruitless. Rather than commit budget dollars to a major shift to process modifi-
cation, I would like to announce our intentions in advance and solicit advice, using the
Agency Research Statement as the forum. That "straw man" characteristic is the main
feature of the ARS. It is to serve as a guide to the budget only if the ideas and proposals
contained in it withstand criticism. Without something specific down on paper, that you
can react to, that includes resource estimates, it is just too easy to engage in generalized
philosophical discussions about all the good things we might do. Frustratingly, we end
up with everything as the No. 1 priority and budget totals that are beyond the realm of
reason.
Let me illustrate by using the industrial waste program. It has been an effective
government/industry cooperative effort. Public funds have served as seed money to pro-
duce new, economical abatement technology. The present pace places in serious doubt
our ability to meet the 1983 standards, as most recently noted by the National
Commission on Water Quality. Begin with that, and also note that R&D funds that are
expended in 1981 and 1982 are going to do little toward achieving in-place technologies
in 1983. Finally, note that if very large increases for funds for the program were to
occur in fiscal year 1977, there is a good chance that there would be insufficient prepa-
ration for handling them correctly. One result might be the loss of the multiplier factor,
the leverage that these monies have enjoyed. (By the way, I should point out to you that
it is that leverage, the fact that Federal dollars aren't the only money in the game, that
has persuaded OMB to allow us to continue this program. If that leverage is in jeopardy,
the whole program becomes vulnerable.) In the ARS, we have suggested, and so noted to
OMB, that the answer to all these issues may be a "spike" budget increase in 1978 and
1979. The amounts would be $45 million and $50 million respectively, better than a
tripling of the projection we have for 1977. To get the money we will need input: will
that amount be effective? Is it enough? Are there better ways to meet our national
goals? By putting the issue into the ARS, we will have the time to debate it. We would
welcome your help.
Let me turn toward a nearer future — 1977. Our budget for research and develop-
ment in EPA in the Office of Research and Development for 1977 comes in two portions.
One is called the "base program" and the other is related to the environmental impact of
new energy resources. The industrial program is within our base. Whether we're rising
or shrinking depends on whether you look at the base or the energy portion and with what
it is compared. In relation to the President's budget, delivered to Congress in January
1975, we are up $300,000 in the base program and down $16 million in energy. On the
other hand, compared to what Congress appropriated for fiscal year 1976, we are down
$7.3 million in base and down $3.6 million in energy. We believe at this point that we
will be able to hold the 1977 industrial waste program at the $14 million level. This is
our total industrial program, all industries, all media—air and water. We need assistance
in developing the case for our budget. There is very little that outside people can do on
a year to year basis, but with the Agency Research Statement, we are moving far enough
into the future that external assistance can be useful. For example, OMB has not
approved a tripling of the dollars for the industrial program, but they are willing to dis-
cuss the merits of the issue. I have put in three scenarios: a spike increase; the total
program staying level; and a spike increase in the industrial program at the expense of
-------
490
other elements of the research program. If industry, if the academic community, if know-
ledgeable people, feel that one or another of these modes of operation are reasonable,
then that can be communicated to us, to Congress, to the Office of Management and
Budget. That will be of great help when the budget issues are to be resolved.
My title is "Future Research: The Role of EPA," so let me pull the threads together.
Our research program, as one of its functions, serves the intent of Public Law 92-500,
the Federal Water Pollution Control Act of 1972. We cannot perform all the R&D alone.
We need the active assistance of industry and other elements of the private sector, in
particular academic institutions. We have conducted and we will continue to conduct a
large extramural contract and grant R&D program. What is new is that we are attempting
to develop a mechanism for an expansion of your assistance. This is to be an expansion
of outside involvement into the planning, the setting of the priorities of the R&D effort.
The potential benefits of success are high enough to make the attempt worthwhile.
In conclusion, meetings such as this, meetings that are reports of research in progress,
are probably the most effective way to get research results into the hands of users. They
are also a very good way for potential users to present their problems to researchers. The
innovations we have made in planning are not intended to replace meetings such as this;
they should rather supplement or complement such face-to-face dialogues. One of the
drawbacks to my present position is that I seldom get to attend entire sessions and have to
be content with reading the proceedings or the summaries. You miss a lot that way. How-
ever, I am grateful for the invitation and the chance to explain the role of EPA, the tacks
we may be taking in the future, and how you can assist us in setting our research priorities.
DISCUSSION
Milt Beychok: Dr. Talley, you said I believe that the research office spends $100 million
on extramural work outside contracted work. How much money is also spent on outside
contracted work by the Office of Air Programs and in the Office of Water Programs -
roughly ?
W. K. Talley: I cannot say exactly.
Milt Beychok: Would it be the same order of magnitude or half of it?
W. K. Talley: No, I doubt it. The Agency has about 10,000 employees. We have about
2,000. The allocatable budget for the entire Agency is about $750 million. Against those
numbers our budget is about $300 million. Hence, we represent about 40% of the allo-
catable funds and about 20% of the employees. 10% of the total budget of the Agency
goes to pay the salaries of and the research conducted by that 20%, so I imagine that most
of the money spent by the other program offices, goes to support salaries and fringe
benefits of employees. I don't know how much they have left for outside studies and con-
tracts, and they are not required to supply me with that sort of information.
Milt Beychok: We can assume they do spend some significant sums.
-------
49]
W- K- Talley; TheX spend some sums, I don't know how significant they are.
Milt Beychok; And that ERDA also does?
W- K. Talley: ERDA spends, in its environmental program, something on the order of
$160 million. Recognize that most of that money was inherited from the Atomic Energy
Commission so that that is primarily nuclear and Dr. Liverman, my counterpart in ERDA,
is trying to reprogram to shift from nuclear to non-nuclear.
Milt Beychok: Even if a third of their $150 million is spent on non-nuclear it is half of
what you spent in EPA.
W. K . Talley: I doubt that they spend as much as a third.
^^——— ~~~~^^~ I'*-*; ,..:'
Milt Beychok; Well my real question is this, that with those three sources of funding,
yours, ERDA and the Office of Air Programs and the Office of Water Programs, funding
of contract research to support their development of regulations, has anyone ever considered
a centralized liaison group to avoid redundancy which does exist and before any one of
the three award outside contracts the other two must agree that it hasn't already been done
or will be done shortly?
W. K. Talley: Yes, that has been considered - If you had asked that as the question at
the beginning, I would have been able to give you the figure that the total amount
estimated by the National Science Foundation spent by the Federal Government for
research and development in the environmental area, is about $1 .3 billion; of that,
against the same base, I spent $300 million and in those figures you can forget about the
Office of Air and the Office of Water, because they are in the noise level. While the
federal government spends $1 .3 billion, the total amount of money spent by industry
dwarfs that. The question of coordination is a very important one and we are trying
various means of doing it. I would like to point out that coordinating the existing pro-
grams is a very difficult task. It takes us, at best, six months to let a contract and it is
quite easy for it to take nine months to a year. To then stick into that process the time to
coordinate with 18 or 20 or 25 other agencies would mean that we would be taking two
and three years to let contracts.
Up until now, we have never had a central planning document. No agency has made
its projections other than for the budget year. For example, I have been reluctant to go
see Dave Rauscher at the National Cancer Institute, with his $700 million budget, of which
$100 million is in environmental carcinogenesis, and say, "Really, Dave, we ought to get
our acts together," because I haven't been able to tell him what I would like him to do
beyond the budget year. We hope that the ARS will serve as a focus for a constituency
document. You raise a very important issue. I hold this job temporarily, but all my life I
am going to be a taxpayer. I am very concerned that the government spend the money
wisely. All my life I am going to be a consumer, I want the regulations that come forth
from the government, that affect how industry operates and ultimately the price the con-
sumer pays, to be adequate but not overly restrictive. I don't want to keep taking from
one pocket of the taxpayer and the other pocket of the consumer.
-------
492
Milt Beychok: Dr. Talley, I wasn't raising that question from the viewpoint of a taxpayer
that is worried about efficiency in government, but you did make the point that the
research arm's program should support the program's branches, and even try and think
ahead o." them to find the problems before they run into them, but so far I doubt very
seriously if any of the regulations; both, air, water or noise that have been adopted have
been based on contract reports other than those funded by the Office of Air and Water
Programs themselves. Outside contract reports to support the guidelines and the new
source performance standards have all been done by contract work funded by the Office of
Air and Water Programs, or at least in large part.
W. K. Talley: That is not true with respect to effects research—human and environmental.
It might be a true statement about the socio-economic research that we conducted, for we
had a small minority of the economic analysis capability of the Agency, so that the program
offices relied upon their own in-house staff.
Milt Beychok: The development documents for all of the effluent guidelines were written
by outside contractors funded by the Office of Water Programs.
Peter Lederman: That is true, however, in writing the guidelines, the Office of Effluent
Guidelines, Al Lynch's group, did use in addition to those development documents, any
R&D developments that were available to them at that time or looked like they were
successful and were viable. But they did use independent development contractors for a
number of reasons. They felt that this was the way to go. However, the research results
that were available over the last three years and were available on a timely manner, were
utilized by them in setting criteria. Now how they used them may be another question,
and I don't think it is one that we can address from the research arm.
W. K. Talley: I would like to mention that I know of three decisions that the Agency has
made that were almost entirely the result of OR&D's function. One was the Administrator's
decision that depended upon the possible production of sulfuric acid from oxidation
catalysts on the 1975, 1976 and later model vehicles, that was based on the non-regulated
emission program which we have been conducting for three years. Two others, fuels and
fuel additives and equivalency regulations, those too, I know depend entirely upon OR&D
because they are the only two regs that OR&D has.
Milt Beychok: I agree with you on the mobile sources, the automobiles and so forth, but
in the industrial sector, the petroleum industry for example; I am sure that of the 634 law
suits we spoke of the other day, that the defense of those in court will be done by the
contractors who worked on the guidelines and air emissions and not the research department.
Doug Lofgren: I am very interested in the remarks you make, Dr. Talley, concerning
looking ahead and the future research in areas that should be looked into. I assume for the
moment that the Clean Air Amendments are passed as proposed with respect to the non-
degradation regulations and that the requirements of zero discharge for 1985 are realized,
could you comment a little bit on the proposed, how we are going to meet the national
goal of energy independence?
-------
493
W. K. Talley; Facetiously, the correct way to achieve it would be to have Congress
repeal the first and second laws of thermodynamics, as was once done by a State legis-
lature setting pi equal to 3.000.
Seriously, the last committee print on the Senate side of the Clean Air Act has more
areas of uncertainty than certainty. They have been working for over a year, it's an
election year, the coalitions seem to have little to do with political party, and so on.
So I really don't think it is worthwhile to comment on the pending legislation because I
don't know what is finally going to come out from that process.
Frank Martin: Dr. Talley, I would like to ask you to get down to something closer if
your future program includes any work that would help the refiner reduce the amount of
testing in order to stay within the permit? As has been pointed out many times it seems
like the number of requirements are great and that the testing could be reduced for many
people and I wonder if there was any work along this line.
Peter Lederman; We have nothing planned in detail in terms of reduction; we are trying
to establish standard methods. I am sure that as these guidelines are utilized there will
be efforts in this area and I think this is the type of information that is important for us in
the planning area and then to work with API and the refineries or industry in general to
determine what is realistic, what can be utilized, what is good enough and what is not
onerous, but this is the type of interchange I think that is important and that we try to
establish with the various industries and industrial groups.
Frank Martin: I have one more question. Recently, in fact last week in Houston, the
President's White House Consumer Representation Plan had a meeting. There were a
number of agencies, including the EPA there, that indicated that they were going to go
out and go to the grass roots and get the consumers' ideas on what they needed and what
was going to happen and so forth and so on. I realize, or at least one of the fears I had
here is, that we are going to be listening to housewives and not industry in this because
industry, although it is a consumer, is not really considered a consumer. I wondered if,
how this information is going to reach your organization or the EPA or your department of
the EPA on what the consumer wants?
W. K. Talley: What the consumer wants in the way of research?
Frank Martin: Yes, what the consumer wants in maintaining the environment and how
effective it is going to be, how is this information going to get to you. For instance, you
mentioned a budget and the budget is a big part of it and there were objections raised
that many of the agencies there and not only the EPA but Agriculture and State and every-
body else, figure up their budgets ahead of time and then they are going out and ask the
consumer and it isn't going to do them any good to have somebody say we need something
when the budget is already set.
W. K. Talley: I have to admit that the rest of the Agency has not generally adopted the
point of view that we ought to look any further than the next budget year. But there are
hopeful signs. One is that Dr. Andrew Breidenbach, the Assistant Administrator for Water
and Hazardou's Materials, in looking at our five year plan draft, told his people, "We
-------
494
can't respond to a research plan five years into the future, unless we know what we want
to do in water programs beyong next year or so. " He is now exploring five year pro-
jections of the programs under his charge.
Industry, as you well know, has very effective and capable lobbyists in Washington.
They serve a very useful function, they communicate with us directly, they are helpful,
they are helpful, they are good contacts. My principal contacts are generally the vice
presidents of research in companies. I get good feedback, early warnings of problems and
so on.
We are all consumers, as you have mentioned. The initiative you described is an
attempt on the part of the Administration to reach out after recognizing that there isn't any
formal mechanism for contacting the average citizen. If you rely on those who write in and
the like, then you are dealing with only a small activist corps. What Mr. Train has done,
is to acquire a special assistant, a woman whose job it is to try to communicate with all
organized constituency groups having to do with environmental affairs or consumer affairs.
This initiative on the part of Mr. Train may or may not work, but at least he has made it.
I can't predict how effective it's going to be, the decision to give her this responsibility
was made very recently.
Umesh Mathur: EPA's approach to controlling industrial effluents into municipal treatment
systems is to first designate an industry for control, then the pollutants to be controlled.
Many of the larger industries are designated for control, but the contribution of a specific
pollutant within an industrial category being controlled may be small compared to the
pollution contribution of an industry not designated to be controlled. As an example, a
toxic material discharged from the relatively unregulated electroplating and metal
finishing industries may exceed that from a refinery which has been designated for control.
When will we get guidelines from R&D documenting the state-of-the-art for effluent
control? We would eventually like to know how the small industries who are discharging
into sanitary sewers will be regulated?
W. K . Talley: Do you understand the division of responsibilities in the Agency? That
the Office of Research and Development issues criteria (dose response functions, ecological
effects or availability of technology), but that the issuance of standards and regulations
comes from the Office of Water Planning and Standards.
The Office of Water Planning and Standards, through the Effluent Guidelines
Division, is currently developing the technical information necessary to issue pretreatment
standards.
Peter Ledeman: We are doing some work on the compatibility or incompatibility of heavy
metals, and, I want to be very careful of the terminology I use, in joint treatment
facilities. That work is going on right now. There is work going on and there will
continue to be work. Of course one of the things about the effluent guidelines is in
addition to what Joe Moore said the other day about a mid course correction.
W. K . Talley: Two things occur to me; one is that Cincinnati has always been the major
-------
495
home of our municipal sewage research. We have now concentrated the responsibility
for the industrial research program there also. This will effect a better communication
between the municipal side, the ultimate recipients of what finally comes out of
industrial discharges, and the people responsible for developing the industrial technology.
The other item serves to illustrate that we do re-examine problems. We are now re-
examining the draft criteria document which ORD produced on land disposal of municipal
sludge. If we can work out with USDA the issue of restricted land usage, rather than
"any crop for all time," we can permit more land applications. As you know, it is the
heavy metals, cadmium in the municipal waste, that are the major problem.
Umesh Mathur: The only reason why I brought this up is because I really believe that
this has become a very critical question, particularly for the guys who are doing long
term planning under things like Section 303 and 208. Unless we have some definitive
answers to these problems we are going to be stuck with as many different local ordinances
as there are cities in the country and each of them is unlikely to be enforced or admin-
istered at all well because the numbers they are putting in there are pulled out of the air,
since there is no definitive guidance from EPA. Once the bad ordinance goes in and
somebody goes in and builds a facility to comply with that ordinance and then you go
back a few years later and say now that EPA has made up its mind, I want you to go back
and now shoot for this target and of course two or three years later when you have your
annual reviews or whatever you might have another ordinance and so obviously people
will be shooting for moving targets and these are big bucks what you are talking about so
I think the question should be answered definitively and quickly. I think the research has bee
been done by Battelle because I have looked at that study and I am a chemical engineer
and I feel that it is defensible; they have done an economic analysis which takes care of
the size question and they did determine that some smaller industries may have to merge
or in some ways consolidate themselves, otherwise they will have to go out of business.
However, that is the principal behind the laws, If you pollute very badly and you can't
afford to clean up, then you go out of business.
-------
496
BIOGRAPHY
Dr. Wilson K. Talley was named
Assistant Administrator for Research and
Development at the U. S. Environmental
Protection Agency in December 1971.
In this position, Dr. Talley is responsible
for planning, directing, and coordin-
ating all EPA research activities covering
air, water, pesticides, radiation, noise,
and solid waste which are necessary in
guiding national policy for environ-
mental protection, developing pollution
control strategies, and establishing
adequate standards and regulations.
Prior to joining EPA, Dr. Talley was
Study Director of the Commission on
Critical Choices for Americans. He has
served in a variety of academic and
administrative positions with the Uni-
versity of California. In 1969 Dr.
Talley was named a White House Fellow
and served as Special Assistant to the
Secretary of Health, Education, and
Welfare until 1970. While at HEW he
was detailed to the "Ash Council" to
work on the study that led to the establish-
ment of EPA. He has been a member of
EPA's Hazardous Materials Advisory
Committee since 1971 .
A graduate of the University of
California with an A.B. in Physics, Dr.
Talley earned his S.M. in Physics from
the University of Chicago and his Ph.D.
in Nuclear Engineering at the University
of California in 1963.
-------
"THE INDUSTRY'S ROLE"
Arne E. Gubrud
Director, Environmental Affairs
American Petroleum Institute, Washington, D.C.
It is always an advantage to be second on a program like this, for one has the
opportunity to respond to the earlier speaker. I would like to do that briefly before dis-
cussing API's programs. I was particularly heartened by the attitudes expressed by Dr.
Talley regarding the need for open disclosure of research data, for cooperation, and for
basing regulations on sound scientific information and demonstrated technology. All of
these are principles that we in the oil industry have long espoused.
It is unfortunate that Dr. Talley can't speak for the EPA program offices, and that
the attitude of objectivity that he put forward is not always reflected in the actions of
those offices in the development of their regulations. To a large extent, the problems
we in industry have with the Agency's regulatory policies are not all of EPA's making,
but rather a result of rigid legislative deadlines and a regulatory framework which does
not give the agency the flexibility it needs to regulate wisely in accordance with
scientific principles and good economic principles. Unfortunately, this is a fact of life,
and 1 am afraid the place where that problem will have to be solved is on Capitol Hill
and not at the Waterside Mall, in Cincinnati, or down in Research Triangle Park.
Turning now to the field of water pollution control in refining, I think it is pretty
clear that we have gone about as far as we can with research on that "add-on"
technology Dr. Talley talked about. We know what the more sophisticated systems are
and what they can do. We have the technological capability to discharge only distilled
water, if we could stand the costs. So there is no lack of end-of-pipe technology,
although the cost-effective technologies are, of course, limited.
End-of-pipe treatment is not the only answer, of course. There is a lot more that
can be done by individual plants in better water management and in waste reduction to
reduce the load imposed on the final treatment plant. Moreover, there may be refining
process changes around the corner, revolutionary process changes, which will make
possible new innovations.
While at the moment there does not seem to be much need for research on end-of-
pipe treatment technology, this may change, because the rules of the water pollution
game are changing and the concerns of the public and the Congress and those of us in the
environmental field are changing.
We used to worry about overloading the assimilative capacity of receiving waters
and depleting the oxygen needed to support sport fish and other biota. No longer. In
fact, now-a-days we refuse to concede that receiving waters can assimilate wastes at all.
The 'reason for the change in attitudes is the shift in focus in the 1972 amendments to the
Federal Water Pollution Control Act. Today we are increasingly concerned about the
possibility of more subtle, perhaps very long-term effects of very small quantities of
substances about which we know very little.
497
-------
498
I am pleased that the Office of Research and Development in EPA is going to research
some of those problems, particularly in view of Dr. Ta I ley's comments about the need to
establish valid dose-response relationships, because we are surely going to confront regu-
lation of some of these materials in the years ahead. With the passage of the Safe
Drinking Water Act, the EPA Administrator has been given sweeping powers to take what-
ever action he deems appropriate whenever he discovers a threat to health through
drinking water. Thus, industries have to be concerned about the contribution they may be
making to the presence of harmful materials in sources of drinking"water supplies.
The petroleum industry is concerned about this problem, and the API last year
launched a very ambitious program which will probably continue for the next several years.
In this program we are looking at the nature and quantities of these more exotic materials
in refinery effluent following application of "best practicable treatment" and "best avail-
able treatment" as currently defined by EPA. At the same time, we are looking at the
effluent from a well-managed and designed municipal waste treatment plant employing
secondary treatment essentially to see what materials are present in such an effluent and
compare them to those present in refinery effluent. We are also subjecting both of these
effluents to chlorination to see what changes take place and whether chlorinated hydro-
carbons are formed. A final report on Phase I of this work, which was a simple cracking
refinery with reforming capacity, will be available in mid-1976. API plans to expand
this study to look at other refineries with perhaps lube oil operations and petrochemicals.
To do the job right, we feel we should eventually cover the whole broad spectrum of
refinery categories, but this may take several years.
The API is also quite aware of the fact that so-called "toxic" pollutants are going to
be receiving regulatory attention for the next several years and that toxicity of refinery
effluent is a parameter which may be included in future permits. The biological-
monitoring requirements in the Federal Water Pollution Control Act have yet to be
implemented, so API has been working to develop a quick, simple bioassay method of
determining gross toxicity in refinery effluent. We have several contracts with the
Virginia Polytechnic Institute and State University to develop a series of tests which
would be applicable to different situations: a coastal refinery discharging into an
estuary, and inland refinery discharging into fresh water, and a refinery discharging into
the ocean. We have already held several workshops to acquaint refinery waste-treatment
personnel with one technique, which is now being used by a number of refineries in the
country. We think that by gathering data from the users on the effectiveness and
reliability of the techniques we are building the kind of data base which will support the
adoption of one or more of these methods by government at some later date. So we are,
I think, going to continue actively supporting research of this kind.
The focus of regulatory attention on small quantities of exotic materials and their
potential toxicity both to aquatic forms of life and to human beings makes it all the more
urgent it seems to me that the kind of objectivity that Dr. Talley spoke about be
encouraged. As physicists — and I was once a "would-be" physicist — I think we both
recognize the difference between an hypothesis which is as yet without data to support it
but seems plausible and a theory, which is an hypothesis that has found support in data
and has never been contradicted by data. A theory is a sound basis for predictive action
-------
499
and for planning and policy, but an unproved hypothesis, however plausible, is not.
Unfortunately much environmental legislation ignores the distinction between
hypothesis, as the basis for research, and theory, as the basis for policy. Thus, much
environmental regulation is not based on logic, scientific principles, or anything but
the fact that the law requires the Agency to act this way or that. The requirements may
ignore the laws of thermodynamics or require government agencies to violate good
scientific method, but the law must be obeyed. That is the bind that EPA finds itself in
again and again. Those in the public sector who have to devise regulations which seem
both to meet the requirements of the law and to be in the public interest and those in
industry who have to comply should not be too critical of one another. I think we ought
to work together to convince the Congress that we really do want to clean up the
nation's air and waters, that we do want to have wholesome drinking water, that we do
want to be responsible, and that we can do and are doing the job to the best of our
ability. I think that this is true and that there is no question that the industry I represent
and most major industries in this country are dedicated to doing the best job they can
and still stay in business.
Yesterday, as I was on the plane coming to Tulsa, I sat next to a young man who is
in shoe sales with a firm based in Maryland. He was writing a memo to his boss.
Although I shouldn't have, I peeked at what he was writing and the first line read some-
thing like: "The figures below indicate that we have a considerable problem." Then he
had three columns headed "promised", "actually ordered for production", and
"deliveries". The figures were in hundreds of thousands, though I couldn't make out
whether they were units of production, dollars, or what. Nevertheless, it was clear
that deliveries were 50 per cent behind what had been promised to clients. He finished
his memo with the statement, "I would be happy to discuss this with you" and then stuck
it in his briefcase. After apologizing for reading the memo, I told him that he hadn't
really finished the job. He had pinpointed a very real problem for his management, but
hadn't offered any suggestions for a solution. I told him that the memo would upset his
management and that since he was the source of the upset, they weren't going to be too
happy with him. It is a first principle that one never bucks a problem upstairs without a
recommendation. I asked whether he had any ideas about what was causing the problem
and how it could be straightened out. He said the problem was in the manufacturing
operation, so I asked whether he had talked to the person in charge of manufacturing. I
suggested that if he could make a recommendation in his memo and say that he had dis-
cussed it with the manufacturing department, then he would really be performing a
service for his organization rather than simply calling its attention to the fact that it was
not performing well.
The point I want to make with this story is that the same advice I gave this young
salesman applies in the current political situation surrounding the study of the National
Water Quality Commission. The Commission will shortly submit its report to the Congress.
It is to be hoped that this report will contain recommendations for action rather than
simply an analysis of the problems with the present law. I would just say to my industry
friends who are in the audience that if any of you or your companies are planning on sub-
mitting comments to the Commission's staff before the February 6 deadline, one of the
-------
500
things you can and should urge is that the Commission not simply sidestep the issue of
what legislative changes are needed to correct the problems it perceives in the present
Act, but that it make affirmative recommendations as to what kinds of legislative changes
are needed. If the Commission is unable or unwilling to make such recommendations, it
will have failed to fulfill its mission.
BIOGRAPHY
Arne E. Gubrud is director of the American Petroleum
Institute's Environmental Affairs Department. In this post,
he supervises a staff of nine professionals. The function of
the Department, whose annual budget for the past several
years has ranged from two to three million dollars, is to
conduct research and disseminate information about environ-
mental conservation. Gubrud was deputy director of the
Committee on Environmental Affairs from 1970 to 1973, when
he assumed his present position.
Prior to joining API as an environmental writer in 1966,
Gubrud had spent eight years with the American Water
Works Association, starting as an assistant editor of the
Association's Journal and rising to the post of assistant
director of publications and public information.
Previously, he had been an editor of foreign language
books and English translations of several journals of the
Soviet Academy of Sciences.
Born in Redwood Falls, Minnesota, in 1934, Gubrud
holds a Bachelor of Arts degree from the University of
Chicago. He has also done graduate work in linguistics and
physics at the University of Wisconsin.
Gubrud is a member of the Water Pollution Control
Federation and the Air Pollution Control Association. He
lives in Allview Estates, Maryland, with his wife and
three children.
-------
"THE UNIVERSITY'S ROLE IN FUTURE RESEARCH"
Joseph F. Molina, Jr., Gerard A. Rohlich, and Earnest F. Gloyna
Professors of Civil Engineering, College of Engineering,
The University of Texas at Austin
After hearing the proposed activities of the Federal government and industry in the
area of future research related to management and pollution control of petroleum refinery
wastewaters, there is some question as to what remains for the universities. However, if
history repeats, a major role of the university will be to develop fundamental research
programs leading to unique solutions to the environmental pollution abatement problems
of the petroleum refinery industry. The universities have been involved in research
related to pollution abatement in the refining industry for more than two decades. How-
ever, university efforts in the past were supported by industry and governmental agencies
and in the future must not be carried out in a vacuum. The continued cooperation of the
Federal government, industry, and the universities is essential to effective management
of wastewaters and residuals generated at petroleum refineries.
The capabilities of universities to conduct research related to industry-wide problems
have been recognized by the American Petroleum Institute, the Federal government
through its past and present research components including the U.S. Public Health Service,
the Federal Water Quality Administration, and currently, the Environmental Protection
Agency, as well as by individual firms and installations through the support of a variety of
research projects. The initial studies on the performance of oil-water separators and
design of inlet devices for this unit process were performed by university personnel in
laboratory studies and field evaluations under the sponsorship of API. Additional field
studies on API separators were performed in cooperation with industry by the same
university personnel at individual refineries. The effects of discharges of refinery waste-
waters on receiving streams and some of the toxic effects of refinery effluents also were
conducted by universities and funded by the API. More recent studies sponsored by the
API at universities include for example:
"Bioassay of Refinery Effluents"
"Biological Effects of Pelagic Oil"
"Field Study on Effects of Oil on Marine Animals"
"West Falmouth Follow-up Studies"
"Fate of Oils in a Water Environment"
"Oyster Field Studies"
"Natural Biodegradation of Oil in Aqueous Environments"
"Clinical Studies of Toxicity of Oil in Water"
"Survey of Sub-Lethal Effects of Biota of Natural Chronic Exposure to Oil"
The EPA also has sponsored university research dealing with the treatment of
petroleum refining wastewaters and the effects of effluents on ecosystems. The trend in
EPA is away from extramural basic research in favor of in-house efforts, contracts, and
demonstration grants. However, not all demonstration projects have been preceded by
adequate laboratory studies, with the result that many pilot-scale or demonstration plant
operations have met with difficulties and problems which often could have been identified
and solved in bench-scale studies.
501
-------
502
The design of some existing treatment facilities at refineries was based on research
conducted by universities under grants funded by a particular company for a specific
refinery. Therefore, it is likely that in the future the universities will be involved in
research related to developing new solutions to another generation of environmental
problems which face the petroleum refining industry.
One of the major roles of the university is in the education and training of people to
carry out research at the universities, but more importantly, in industry and in govern-
mental institutions. In the past, the support of this effort primarily was by the Environ-
mental Protection Agency or its predecessor agencies. Students, mainly at the Master's
level, received support and specialized in either air pollution control, water pollution
control, or in solid waste management. At the completion of the university program,
many of the trainees found employment in industry and became involved in design, oper-
ation, or research. However, the current trend in EPA is to limit training grants to
individuals who are presently employed by regulatory agencies or who will seek employ-
ment in this type of agency after receipt of their advanced degree. If this trend continues,
the number of qualified engineers and scientists trained specifically to direct their talents
to environmental problems and available to industry will be markedly reduced. Industry
then may be required to provide for education and training of individual employees in the
area of environmental pollution abatement.
One of the difficulties facing universities at the present time is the curtailment of
federal funds and to a certain extent, industrial support for advanced studies. This is
particularly true for Ph.D. students who generally require a stipend for a period of three
years. It is at the Ph.D. level that training in research is accomplished. Therefore, the
university efforts require support through research grants or other means to assist in
training Ph.D. students who will lead the way in developing unique solutions to existing
and future problems.
Universities are uniquely suited to conduct research on problems which are of interest
to the entire petroleum refining industry as well as to address specific industrial problems
which may face a small facility. A wide range of talents are frequently available among
the faculties and graduate students at the university. In Engineering for example, people
in Chemical, Environmental, and Petroleum Engineering can be called upon in addition to
individuals in Chemistry, Biology, and Microbiology. Universities also frequently are
well-equipped with analytical instrumentation and supporting facilities to conduct funda-
mental research and bench-scale and pilot plant studies.
For example, at The University of Texas at Austin, the Environmental Health
Engineering Labs at the Center for Research in Water Resources have been involved in
bench scale and pilot scale studies related to the treatability of a variety of industrial
wastewaters including petroleum refinery wastewaters. Efforts at the University of Tulsa
through their Environmental Protection Projects program provide support for the education
of undergraduate and graduate students in the College of Engineering and Physical
Sciences.
The implications of PL 92-500, Water Pollution Control Act of 1972, and the Safe
-------
503
Drinking Water Act, PL 93-523 lead to many areas where universities can make an
impact in the immediate future. As long as the approach to the solution of the problem
involves end of pipe treatment, treatability studies will have to be performed to develop
more effective designs of facilities resulting in a more economical operation of the treat-
ment plants.
Identification and quantification of organic and inorganic toxicants as well as
techniques for ameliorating the effects of these materials is another area of industry-wide
concern. The accumulation of toxic materials in sediments and in the food chain in the
aquatic environment has not been established in sufficient detail. The entire question of
quantifying toxicity is an area requiring considerable basic research. The toxicity bio-
assay procedure must be carefully evaluated especially in terms of selection of test
organisms. The chemical species of materials discharged into the natural environment and
interaction of the particular species with other components in the environment must be
identified for many of the heavy metals discharged into natural systems. Some of these
metals may react similar to mercury which when discharged into natural water can under-
go conversion from the inorganic form to an organic compound more toxic than the
inorganic form.
The Committee on Water Quality Criteria of the National Academy of Sciences and
National Academy of Engineering (1) has defined some research needs. Universities will
play an important role in these research efforts directed at establishing and mitigating the
impact of industrial effluents on freshwater and marine aquatic life and wildlife. Concern
for biological monitoring toxicity bioassays and assimilative capacity of receiving waters
is a logical sequel to approaching the solution to the problem by end of pipe treatment.
Some of the recommendations of the Committee report are pertinent to the topic under
discussion and are quoted at this time.
". . .to undertake integrated studies of fresh water ecosystems in order to
understand and ultimately predict the response of an ecosystem to various
pollutants and to physical manipulation of the environment."
". . .biological scale models that can be manipulated in the laboratory
provide a research tool that is convenient as well as productive. To
deliberately control addition of pollutants to small unpolluted bodies of
water has been used to great advantage in the past in determining the
broad ecological impact of pollutants."
The report also directs comments to biological monitoring and bioassays.
"The first concern of future efforts in biological monitoring should be
development of standardized procedures having application to a diversity
of aquatic systems. Such studies could most properly be conducted along
with chemical analyses of water with the ultimate objective of identifying
useful predictors of potentially toxic conditions."
-------
504
"Research should be initiated to determine those species best suited for
monitoring various kinds of pollutants."
"With such extensive use of bioassays there has developed a need for
standardized procedures and methods for reporting particularly with regard
to determining chronic and thermal effects in interactions of mixed toxic
compounds. Also, yet to be resolved are questions of the comparability of
results of bioassays involving different organisms and the applicability of
laboratory data to the natural environment. Despite the progress that has
been made bioassays have not yet been conducted for a number of
important compounds."
"Research is needed to develop more effective bioassay methods for
potential toxicants. The most sensitive marine organism should be identified
and cultured on a generation-to-generation basis under laboratory con-
ditions to provide genetic variability information on the test species and to
form a basis for intercomparison of experiments and results. "
"The effects on organisms of accumulation of sub-lethal levels of
pollutants and the effects on man using these organisms in food should be
investigated. "
"Research is needed on the extent of bioaccumulation of various pollutants
and the effects of transfer from organism to organism."
"It is essential to understand the interrelationships among organisms and
between organisms including the highest trophic levels and their environ-
ment. This understanding is required for evaluation of the subtle and
secondary effects which may be more important than the immediate and
direct effects of an organism observed in a bioassay for acute toxicity."
"To make specific water quality recommendations for ecosystems it will be
necessary to acquire information on the physical and chemical character-
istics of the system, on distribution and abundance of species at all trophic
levels, and on the normal variations in these characteristics over at least a
representative annual cycle."
"Studies on the species diversities in a wide variety of marine ecosystems
are needed as a basis for interpreting changes as indications of trends in
overall water quality."
"The natural recovery potential of man stressed ecosystems as well as
techniques for stimulating recovery require investigation in order to assess
the permanence of damage and urgency in designating water quality for
specific use. "
The assimilative capacity of the ecosystem also must be defined.
-------
505
"Fundamental studies of the physical, chemical, and biological processes
determining assimilative capacity are needed to develop methods of dis-
posal that can be applied to a variety of locations. Each location how-
ever, will have some unique characteristic. Local conditions and the
process operating there which determine the fate, concentration, and
distribution of a pollutant must be studied in order to assure that a proposed
rate of disposal not exceed levels harmful to the ecosystem."
"The understanding of the rates of biological, chemical, and geological
modifications is needed to evaluate the long-term impact of various
pollutants on the marine ecosystem. Research is needed on the rates of
biological degradation and decomposition, on the chemical reactions
various compounds and elements undergo when introduced into seawater,
and on the rates of absorption and precipitation to the sediments."
The non-living components of the marine ecosystem also must be considered.
"Research is needed on the synergistic and antagonistic effects of seawater
on organic and inorganic chemicals in the marine environment as well as
on synergism or antagonism among the chemicals. A series of combinations
and permutations should be tested under different conditions in both fresh
water and seawater to understand more fully the mechanisms of synergism
and antagonism."
On the topic of oils, the report mentions,
"The presence of flowing oil in visible amounts is generally considered
unsightly as well as harmful to aquatic life and wildlife. Nonetheless,
invisible surface films of oil should be examined closely for adverse bio-
logical effects attributable to adsorption onto the body surface of
organisms or to interference with the exchange that normally occurs at the
air-water interface. The toxicity of oil dispersants of various crude and
manufactured oils also require further investigation."
"Sedimented oils pose a problem that demands more immediate attention in
both fresh water and the marine environments because of their persistence
and widespread occurrence. Safe levels of sedimented oils have yet to be
determined and the impact's availability to concentrate other toxic
substances is unknown."
The development of effective biological monitoring systems, more universally applic-
able bioassay procedures and a more complete understanding of the biological, chemical,
and physical reactions and interrelationships in fresh water and marine ecosystems will
require the cooperative efforts of industry, universities, and government agencies.
In areas other than end of pipe considerations future research activities at universities
must be oriented to the industrial activities, specifically:
-------
506
a) secondary recovery and reuse of raw materials;
b) process improvement to eliminate waste;
c) waste treatment with emphasis on liquid, solid, and gaseous waste; and
d) disposal of residuals and sludges.
A wide variety of organic solvents and detergents are used in secondary and tertiary
recovery. The basic research on techniques to contain these materials and to minimize
the effects on the enviroment still must be initiated. A great deal of effort has been
expended on tertiary oil recovery with but little attention to the processes which would be
required to make these operations environmentally acceptable.
Many of the companies involved in petroleum refining also are active in developing
alternative energy sources. Many of these activities will require closer attention to their
impact on the environment. Nuclear energy will require basic research directed at the
disposal of spent fuels. Programs for developing effective methods of coal gasification are
ongoing. But almost no information is available to the industry regarding the treatment or
the disposal systems necessary to handle the liquid, solid, and gaseous wastes generated
along with the usable product. It is in the development of unique solutions to these types
of problems that the universities can make the most significant contribution.
A considerable amount of basic research also is required to modify unit processes
involved in petroleum refining in order to eliminate, or at least minimize, the amount of
waste generated. This type of problem can best be solved as a joint venture with the
university and industrial personnel cooperating to develop a satisfactory solution using
modern process control technology. Spills and leaks in process areas result in the loss of
materials which find their way into waste ditches and eventually into the wastewater treat-
ment system. Much of these materials are reusable in the refining process and in some
cases are recovered and returned to the process areas.
The requirements for a higher quality effluent has led to the application of physical
and chemical processes to treat effluents of biological treatment facilities. These pro-
cesses result in increased sludge volumes which require disposal as well as in a higher total
dissolved solids concentration in the effluent. A vast amount of basic research is needed
to answer the myriad of questions that have been raised about the effects of toxicants on
the biota in treatment plants and in receiving streams as well as on the methods of removal
of these toxic substances. In many of the processes these toxicants are not destroyed but
merely removed from the wastewater stream and concentrated as either a brine or in the
sludge. The ultimate disposal of the brine or sludge requires basic research to define suit-
able methods of treatment or disposal of this material. This information is essential before
large scale and industrial pilot plants are designed and constructed.
The area of solid waste management offers opportunities for much basic research.
Solid wastes at a refinery range from plant trash to slop oils, tank bottoms, sludges,
catalysts and other materials which are not carried away in the wastewater stream. Much
of these materials are placed on the land, and in the case of slop oils, tank bottoms, and
biological sludges, the soil bacteria are effective in decomposing the materials. However,
a number of questions come to mind which require answers. This information can be
-------
507
developed through a basic research program. Some of these questions include: a) At
what levels are various toxicants present in the sludges? b) If toxicants are in the
sludges, what is the rate of uptake of these toxicants by the crops grown on these lands?
c) What is the mechanism by which the toxicants are transported from the sludge to the
soil and possibly through the soil into groundwaters? d) What is the tolerance level of
the soils for salts which may be present in the sludges? If the organic and inorganic
toxicants can be transported from these sludges, and slop oils through the soil system, it
is essential to define those characteristics of the soil which inhibit the transport of these
materials so that proper site selection criteria can be developed.
There is no unified policy in the petroleum refining industry concerning solid waste
management. In many cases, the problems of residuals handling and disposal are not
clearly identified or understood. The characteristics of these residuals and their respective
half-lives must be defined before any long-term solutions to this phase of waste disposal
can be developed. Most universities have considerable capabilities for analyses and
instrumentation which can be used to characterize various components of industrial waste.
Another problem which requires definition and quantification is that of non-point
sources of pollutants. The run-off resulting from rainfall in process areas and product
storage areas generally cannot be discharged directly into receiving streams. The
characteristics of these materials must be quantified.
The problems of land use planning also require an objective review which may be
considered as fundamental research. The approach to land use planning must be reason-
able otherwise legislation may be forthcoming which will make it impossible to dispose of
wastes generated by industry within the close proximity of industry. Land use legislation
which does not recognize waste disposal as a reasonable use for certain lands will
markedly affect the overall industrial efforts related to liquid, solid, and gaseous
pollution abatement.
However, it should be emphasized once again that a primary function of the
university will be to produce research engineers and scientists recognizing that these pro-
grams must continue as a cooperative effort with industry and agencies at all levels of
government. In order to meet this responsibility universities will continue to be actively
engaged in research related to industrial pollution abatement problems.
REFERENCE
(1) "Research Needs in Water Quality Criteria 1972", a report of the Committee on
Water Quality Criteria to the Environmental Studies Board of the National Academy
of Sciences, National Academy of Engineering, Washington, D.C., 1973.
DISCUSSION
Bob Huddleston: Just a comment - one of the problems that I have seen in looking at our
Company's waste disposal problems, or waste treatment problems has been that we would
like very much to not see variability in the material that we have to treat but we have
-------
508
got to recognize the fact that it is there. One of the problems that we encounter is how to
handle this variability and this is an area of research I think that needs to be looked into
much more than it has in the past.
Joe Molina; I agree that variability will exist. However it is a matter of what degree of
variability can be expected and to reduce the variability to something that is more or less
predictable. Some work on equalization systems has been done at The University of Texas
at Austin involving the types of models available to predict the size of basin needed
based on the incoming variations, as well as the design of a facility for equalization
which also provides some treatment in addition to equalizing the flow.
BIOGRAPHY
Joseph F. Molina, Jr. is Professor
of Civil Engineering and Director of
Environmental Health Engineering
Laboratories at The University of Texas
at Austin. He holds the Bachelor in
Civil Engineering Degree from
Manhattan College and M.S. and
Ph.D. in Sanitary Engineering from
the University of Wisconsin. He is
registered as a professional engineer
in the State of Texas, and has been .^W^
with The University of Texas at Austin
since 1961.
-------
SUMMARY
"REVIEW OF OPEN FORUM"
Francis S. Manning, P.E.
Professor of Chemical Engineering,
University of Tulsa, Tulsa, Oklahoma
How does one review such a successful meeting? Obviously it is impossible to
recapitulate the highlights of 32 presentations by experts far more learned and experienced
than I. Let me start by thanking all those people who contributed so much to this open
forum: -
1. You the participants of this open forum. Without your attendance this symposium
would have been a disaster. Your excellent attendance, not only guaranteed the
success, but also demonstrates the petroleum-refining industry's genuine concern for
our environment.
2. Our speakers - all of them. I am sure the other 29 will forgive me if I make special
mention of my old friends - Milt Beychok; Wes Eckenfelder; and Davis Ford - who
have also contributed so much to my previous short courses.
3. The EPA for sponsoring this open forum. May I make special mention of my friends at
the R. S. Kerr Laboratory in Ada, Oklahoma. Without Fred Pfeffer, Leon Myers,
Bill Galegar and their colleagues there would have been no open forum.
4. The API and NPRA. The excellent attendance - over 300 - is due to Herb Bruck,
NPRA and Arnie Gubund, API. They were the publicity.
5. Last but not least, my colleagues at T.U.: - Nick Sylvester, Katie Whisenhunt,
Shirley Clymer and Paschal Twyman.
Please make no mistake about it. Many of you were kind enough to offer congratu-
lations to me for "putting on such a success". With all due respect, my conscience
insists that the whole truth be known. A successful open forum is not a one-man show.
Of necessity it is a team effort. I merely had the privilege of coordinating their efforts
with a lot of help from Fred Pfeffer.
To return to my original charge. How does one summarize this open forum? Let us
content ourselves by discussing some of the major recurring themes: -
1. The large attendance and lively interchange of ideas constitutes "proof positive" that
the petroleum-refining industry is a leader in preserving the environment. May I
digress to offer two more "exhibits" in "evidence" of this concern for the environ-
ment: -
(i) Since 1974, 14 petroleum and petroleum-related corporations have sponsored
509
-------
510
the University of Tulsa's Enviromental Protection Projects program (UTEPP).
Current UTEPP research includes: chromate removal from cooling tower blowdown;
phenol removal from wastewater by carbon adsorption; phenol removal from waste-
water; flocculation of oil-water emulsions; oil removal from wastewaters by
induced air flotation; and oil removal from wastewaters by ultrafiltration. UTEPP
was started after a two-year EPA Training Grant in oil-related water pollution
control was terminated. UTEPP not only supports graduate and undergraduate
education in oil-related pollution control; but also constitutes convincing
evidence of the petroleum industry's concern for the environment. It is hoped
that this cooperative research effort will not only grow but will also serve as a
model for future research commitments by other industries at other universities.
The Federal government should actively seek methods of participating in these
joint efforts as this is a very efficient way of minimizing duplication of research
efforts.
(ii) Since 1970 the petroleum industry has also supported 10 one-week short courses
at T.U. on industrial pollution control as applied specifically to petroleum
refining.
2. There has been a remarkable agreement among the views expressed - not only con-
cerning the technologies but also the guidelines.
3. Sure there were differences in opinions and some keen arguments. These differences
reflect honest attempts to state one aspect of our most complex task - optimizing the
management of petroleum refinery wastewaters.
4. As Milt Beychok so eloquently noted, such dialogue between all the concerned parties
is absolutely necessary. Progress will be made only if all parties are not only willing
to state their viewpoints but also equally willing to allow their viewpoints to be
cross-examined in depth. This adversary procedure is not the easiest on everybody's
temper or nerves, and is not the least time-consuming; but is it not the "best avail-
able technology" for discovering "the truth, the whole truth, and nothing but the
truth"?
Next let us sum up the "evidence" of the last 32 "witnesses". Some simple truths
emerge: -
1 . The petroleum refining industry should be justifiably proud of the technology it has
developed. The technology described by Fred Weiss and Jim Grutch (to state just
two examples) is second to none. Sure we can do even better and honest self-
ctriticism is constructive. But there is no need for despair, no benefit in submitting
quietly to unreasonable critics.
2. As Nick Gammelgard stated so forcefully it is senseless to strive blindly for "zero
discharge" with no regard to any other variable. The "Energy Crisis" has forced us to
step back and re-evaluate "the impossible dream" of zero discharge.
3. All our "expert witnesses" clearly demonstrated how complex it is to try to optimize
-------
511
petroleum refinery wastewater management. We examined the subtleties in great
detail - e.g.
Allen Cywin, Martha Sager, Joe Moore revealed the legal and political
complexities.
Carl Schafer, Bob Silvus and Harless Benthul demonstrated the pitfalls of applying
the legalities and politics!
Then after this introduction we reviewed at length the available technologies and
speculated on the future advances.
4. This mass of evidence convinces me that rational optimization of wastewater manage-
ment is achieved most readily if we use a model to help us. At the very least such
a model should identify our problems. At best we get some useful results.
5. Please do not misunderstand me. I am NOT stating that all progress demands the
largest computer. On the contrary, I am a most fervent believer in Occam's Razor.
"Entia non sunt multiplicanda sine necessitate". In the current vernacular it is
called, I believe, the KISS system; "Keep it simple, Stupid"!
6. Applying the KISS principle we must first determine what we are going to optimize.
Very little thought is required to agree that we would like to minimize:
(i) the valuable natural resources consumed
(ii) the pollution load on the environment
(iii) the financial burden on the economy
(iv) the technical manpower required, etc.
7. Because we cannot simultaneously minimize all these objectives; trade offs are
required. In other words we must convert or express "energy loss" or "pollution load"
to $. This is a huge problem; and, in the long run, the "dollar value" or "con-
version factors" must reflect public opinion. In many cases more research will be
needed before we can say how bad a particular chemical is. The current activity in
toxicity studies is addressing just such a facet of our objective function.
8. Next we must model our refinery. At a minimum we should know the energy con-
sumed and pollution load generated first for a typical (say class A) refinery and
second for each of the major operations, crude distillation, cracking, etc. This
second-level detail would permit comparison and optimization of various modes of
operation.
9. Please do not construe the previous ideas as a neglect of wastewater treatment
technology. Such technology can perform wonders; but, it is only one alternative -
another alternative may very well be - don't generate that particular pollutant, as
Cyron Lawson observed. And, of course, the demand for refinery gas, heating oil,
kerosene, etc., will also control plant operation.
10. Technical manpower can be minimized by concentrating on the bottlenecks and so
avoiding spinning our wheels. Poor data or lack of data do not deserve a very
detailed model and, of course, vice versa.
-------
512
Please pardon this digression on modeling. But, I hope, that you experts will agree
that this open forum has identified some key bottlenecks. At least we have discussed all
the key ingredients for the model: -
o should the guidelines reflect a sensible trade-off between energy-consumption,
economics, and technology?
o should the technology discussions help us improve our treatment processes?
o the recommendations on the direction for future research efforts should have
helped us accelerate our treatment technology.
If this open forum has resulted in any of these goals it deserves to be labelled a
success. At the very least it has encouraged and fostered 3-1/2 days of friendly and
constructive dialogue - which is the ideal catalyst for progress.
To those impatient majority of you who restlessly seek to speed up in this optimization
process, may I venture to calm you with the Latin proverb "festina lente" (hasten slowly).
Rest assured the refinery industry can and will serve as leader for municipalities and other
industries. Your interest in this open forum proves that you well realize the key for suc-
cess. It is perhaps best stated in the State of Oklahoma's official motto, "Labor omnia
vincit" (Work overcomes all obstacles).
BIOGRAPHY
Francis S. Manning is Professor and Chairman of
Chemical Engineering at The University of Tulsa.
He holds the following degrees in Chemical
Engineering: - B. Eng. (Hons.) from McGill
University and M.S.E., A.M., and Ph.D. from
Princeton University. He is a professional
engineer, registered in Oklahoma, Pennsylvania
and Texas. Frank taught at Carnegie-Mellon for
9 years before joining the University of Tulsa in
1968 . The author of 1 book and over 60 papers,
Frank's current research interests lie in thermo-
dynamics, reaction kinetics, and industrial
pollution control. In 1969 he received the
R. W. Hunt Silver Medal from the AIME.
------- |